A CRITICAL EVALUATION AND REFINEMENT OF THE PERFORMANCE PREDICTION OF WET-COOLING TOWERS by Johannes Christiaan Kloppers Dissertation presented for

the degree Doctor of Philosophy (Mechanical Engineering) at the University of Stellenbosch Promoter: Prof. D.G. Kroger Department of Mechanical Engineering University of Stellenbosch December 2003 UNIV.STELLENBOSCH III' II II "' "" "'" II II300 783 1557 . IDECLARAnON I, the undersigned, declare that the work contained in this dissertation is rriy own original work, and has not previously in its entirety or in part been submitted at any uiliversity for a degree. sl#r(~???????????????? ...!:c.P!??.:~. !/:-:::.C? 'l. .Date ii ABSTRACT The thermal performance prediction ofwet-cooling towers is critically analyzed a nd refined. Natural draft counterflow towers and mechanical draft counterflow and crossflow towers are con sidered. The Merkel, Poppe and e-NTU heat and mass transfer methods of analysis are derived from firs t principles, as these methods form the cornerstone of wet-cooling tower performance evaluation. The cr itical differences between these methods, when applied to fill performance analyses and cooling tow er performance evaluations, are highlighted. The reasons for these differences are discussed wi th the aid of psychrometric charts. A new extended empirical relation for the loss coefficient of fills is p roposed where the viscous and form drag effects are accounted for as well as the buoyancy, momentum and fi ll height effects. The empirical equation for the transfer characteristic of fills is extended to inclu de the effects of fill height and the inlet water temperature. Empirical equations to predict the temperature inve rsion profile, height of the temperature inversion and the height from which air is drawn into the cooling to wer are developed. The influence of temperature and humidity inversions on the performance of wet-cooli ng towers is subsequently investigated. A comprehensive analytical computer program is develo ped to predict and optimize the performance of wet-cooling towers. Computer programs are also devel oped to generate cooling tower performance curves, analyze fill performance test data and plot ps ychrometric charts. Keywords; Wet-cooling tower, Merke~ Poppe, e-NTU, natural draft, mechanical draft, fill, t emperature inversion. -', 111 SAMEVATTING Die tenniese werksverrigtingvoorspelling van nat kooltorings word krities geanal iseer en verfyn.

Natuurlike trek teenvlooi kooltorings en meganiese trek teen- en dwarsvlooi kool torings word ondersook. Die Merkel, Poppe and e-NTU warmte- en massaoordrag berekeningsmetodes word afge lei vanuit eerste beginsels omdat hierdie metodes die hooksteen is van die berekening van die werk verrigting van kooltorings. Die kritiese verskille tussen hierdie metodes, wanneer dit toegepas word op die berekening van die werksverrigting van pakkings en kooltorings, word beklemtoon. Die redes vir hierdie verskille word verduidelik aan die hand van psichrometriese kaarte. 'n Nuwe uitgebreide em piriese vergelyking vir die verlieskoeffisil!nt van pakkings word voorgestel waar daar voorsiening gemaa k word vir die viskeuse? en vormsleur effekte, asook die vlotkrag, momentum en pakkingshoogte effekte. Di e empiriese vergelyking vir die oordragskol!ffisil!nt van pakkings word uitgebrei om die eff ekte van die pakkingshoogte en die inlaat watertemperatuur in te sluit. Empiriese vergelyking s om die profiel van temperatuuromkerings, die hoogte van temperatuuromkerings en die hoogte waaruit lug in die kooltoring ingesuig word, word ontwikkel. Die invloed van die temperatuur- en humiditeitomk erings op die werkverrigting van kooltorings word vervolgens ondersoek. 'n Omvangryke analitie se rekenaarprogram word ontwikkel om die werkverrigting van nat kooltorings te voorspel en te optim eer. Rekenaarprogramme word ook ontwikkel om koeltoring werksverrigtingskurwes te gen ereer, pakkingstoetsdata te analiseer en psychrometriese kaarte te genereer. Sleutelwoorde: Nat kooltoring, Merkel, Poppe, e-NTU, natuurlike trek, meganiese trek, pakking, temperatuuromkering. / IV Dedicated to Oupa Chris van der Walt (1914 - 2000) vACKNOWLEDGEMENTS Thank you Lord Jesus Christ, my heavenly Father, for giving me the strength and the will to do everything in Your name. c? The author would also like to thank the following persons and institutions for t heir assistance and contributions during the course of this study: My promoter Prof. D.G. KrOger for his world class guidance and support and for t he countless discussions concerning my research project. Cobus Zietsman for his friendship, help with the experimental work and knowledge and experience he has shared with me. Calvin Hamerse for his help with the experimental work. SASOL for their financial support and Hein Botes, my mentor at Saso\. Dr. P.J. Erens of Industrial Water Cooling for the financial support with the fi ll tests and for supplying the fill materials. My wife, Lynette, for her love and support and for being the inspiration in my l ife and her parents for their support.

My best friend for his calming influence and for always lying at my feet when I am working into the early hours of the morning - my dog Boetie. My mother for her interest in my work and for being a pillar ofstrength in my li fe. My family and friends for their interest and support. VI TABLE OF CONTENTS Declaration Abstract 11 Samevatting iii Dedication iv Acknowledgements v Table ofContents vi List of Symbols viii CHAPTER 1 INTRODUCTION -............................ 1.1 1.1 Overview ofWet-Cooling Towers 1.1 1.2 Outline ofThesis 1.5 1.3 Computer Software Development 1.7 CHAPTER 2 WET-COOLING TOWER HEAT AND MASS TRANSFER ANALySIS............ 2.1 2.1 Introduction 2.1 2.2 Psychrometric Chart 2.1 2.3 Merkel Analysis 2.3 2.4 Poppe Analysis 2.8 2.5 e-NTU Analysis 2.9 2.6 Other Analyses for Evaluating Cooling Tower Performance 2.10 2.7 Conclusion 2.11 CHAPTER 3 FILL PERFORMANCE EVALUATION........................................... ............................ 3.1 3.1 Introduction 3.1 3.2 Loss Coefficient 3.1 3.3 Transfer Characteristic 3.6 3.4 Conclusion 3.10 CIIAPTER 4 WET-COOLING TOWER PERFORMANCE EVALUATION............................. 4.1 4.1 Introduction 4.1 4.2 Natural Draft Cooling Tower 4.1 4.3 Mechanical Draft Cooling Tower 4.4 4.4 Investigation into the Draft Equation of Natural Draft Cooling Towers 4.9 4.5 Consistent Application of Cooling Tower Performance Evaluation Analyses 4.10 4.6 Lewis Factor 4.11 4.7 Atmospheric Pressure 4.13 4.8 Improved Merkel Energy Equation 4.13 4.9 e-NTU Approach 4.14 4.10 4.11 Constant Heat Rejection Conclusion VII 4.15 4.15 CHAPTER 5 THE INFLUENCE OF TEMPERATURE AND HUMIDITY INVERSIONS ON COOLING TOWER PERFORMANCE 5.1 Introduction 5.2 Temperature and Humidity Inversion Profiles 5.3 Height from which Air is Drawn into a Cooling Tower 5.4 Effect ofTemperature and Humidity Inversions on Tower Draft and Inlet Condit ions 5.5 Conclusion 5.1

5.1 5.1 5.2 5.2 5.3 CHAPTER 6 CONCLUSION 6.1 REFERENCES...................................................................... .................................................................... 7.1 APPENDICES Appendix A Appendix B Appendix C Appendix D AppendixE Appendix F Appendix G Appendix H Appendix I Appendix J Appendix K Appendix L Appendix M Appendix N Appendix 0 Appendix P Appendix Q Appendix R Appendix S Appendix T Appendix V Appendix V Properties ofFluids Heat and Mass Transfer in Counterflow Wet-Cooling Towers Heat and Mass Transfer in Crossflow Wet-Cooling Towers Loss Coefficients and Transfer Characteristics Effect of Atmospheric Conditions on the Operation of Cooling Towers Lewis Factor Counterflow Fill Analysis According to the Poppe Approach Crossflow Fill Analysis According to the e-NTU, Merkel and Poppe Approaches Analysis of a Natural Draft Wet-Cooling Tower Employing the Poppe Approach Analysis ofan Induced Draft Wet-Cooling Tower Employing the Merkel Approach Fill Test Facility and Processing ofFill Test Data Temperature Distribution during Nocturnal Inversions Atmospheric Humidity Modelling of a Cooling Tower as a Circular Jet and a Point Sink A Critical Cooling Tower Performance Evaluation Wet-Cooling Tower Performance Evaluation Software Cooling Tower Performance Curves Trickle Fill Performance Test Results Splash Fill Performance Test Results Film Fill Performance Test Results Cooling System Optimization Sample Calculation of the Influence of Temperature and Humidity Profiles on Wet? Cooling Tower Performance A.I B.I C.I D.I E.I F.I

G.I H.I I.I J.I K.I L.l M.I N.I 0.1 P.I Q.I R.I S.I T.! V.I V.I Aa B b C c D DALR d e F f G g H h Ilg J K k L I M m NTU n p P, p Q q R Ry r viii LIST OF SYMBOLS Area, m2 Surface area per unit volume, mol, or coefficient Breadth, m Exponent Coefficient, or heat capacity rate mcp, WIK, or fluid capacity rate ,Cm,JCm,,, o r cost, $ Concentration, kg/m', or constant Specific heat at constant pressure, J/kgK Diffusion coefficient, m2/s

Dry adiabatic lapse rate, Kim Diameter, m Effectiveness Force, N Optimization objective function Mass velocity, kg/m's Gravitational acceleration, mis', or inequality constraint function Height, m Heat transfer coefficient, WIm'K, or equality constraint function Mass transrer coefficient, m/s Mass transfer coefficient, kg/m's Enthalpy, Jlkg Latent heat, J/kg Momentum flux, kg mis' Loss coefficient, or thermal eddy diffusivity, m'/s, or kinematic momentum flux, m4/s' Thermal conductivity, WImK, or turbulent kinetic energy Length, m Characteristic length Molecular weight, kg/mole Mass flow rate, kg/s Number oftransfer units Number Power, W Perimeter, m Pressure, N/m', or Pa Heat transfer rate, W Heat flux, W/m' Gas conslant, JlkgK Characteristic flow parameter, mol Radius, m, or correlation coefficient ix S Sum ofthe least squares T Temperature, ?C or K t Time, s, or thickness, m u Internal energy, J/kg, or Cartesian velocity component V Volume flow rate, m'/s, or molecular volume, or characteristic velocity, or vo lume, m' v Velocity, mis, or Cartesian velocity component W Width,m w Humidity ratio, kg water vaporl kg dry air, or Cartesian velocity component x Coordinate Y Approach velocity factor y Coordinate z Coordinate, or elevation, m, or exponent Greek Symbols a Thermal diffusivity, leiIX'p, or m'ls, or relaxation factor a. Kinetic energy coefficient T/ (} J1 v p Boundary layer thickness Differential Virtual kinematic viscosity, m'ls, or dissipation rate of turbulent kinetic ener gy Non-dimensional coordinate Angle, ? Dynamic viscosity, kg/ms

Kinematic viscosity, rn2/s Non-dimensional coordinate Density, kg/m' Area ratio, or surface tension, N/m Stream function Shear stress, N/m2 Relative humidity, or expansion factor Dimensionless Groups FrD Desimetric Froude number, {JI"/(!J.pdg) Le Lewis number, leI(p::pD), or Sc/Pr Lej Lewis factor, h/(cphd) Me Merkel uumber, hdaflLfllGw Nu Nusselt number, hLik Pg Vapor pressure gradient Pr Prandtl number, cpp/k Re Reynolds numher, {JI'L/J1 x8c Schmidt number, #(PD) 8h Sherwood number, hDLID 81 Stanton number, h/(pvcp ), or Nul(RePr) 81m Mass transfer Stanton number, ""I(PV), or 8hl(ReSc) Subscripts a Air, or above b Below c Convection heat transfer, or critical cl Cooling tower clc Cooling tower contraction cle Cooling tower expansion D Drag d Drop, or discharge, or day de Drift eliminator dij Diffuser e Evaporative, or expansion, or e-NTU theory em Empirical ex Experimental F Fan Fldij Fan/diffuser fi Fill Ir Frontal g Gas Inlet il Inlet louver il Inversion top M Merkel theory m Mean, or mass transfer max Maximum min Minimum n Nozzle o Outlet op Orifice plate P Poppe theory pi Plenum chamber r Reference, or radial coordinate rz Rain zone s Saturation, or shell sp Spray Xl ss Supersaturated I Total, or top, or turbulent Ir Trough Is Tower support

up Upstream v Vapor w Water wb Wetbulb wd Water distribution system x Coordinate y Coordinate z Coordinate 1.1 CHAPTERl INTRODUCTION 1.1 OVERVIEW OF WET-COOLING TOWERS Heat is discharged in power generation, refrigeration, petrochemical, steel, pro cessing and many other industrial plants. In many cases, this heat is discharged into the atmosphere wi th the aid of a cooling tower. Figure 1.1 shows an example of the application of a cooling tower in a si mple steam power plant. Heat is discharged into the atmosphere by the cooling tower via a secondary cycl e with water as the process fluid. Condenser Pump Transformer Generator Cooling tower Figure 1.1: Simple steam power plant with cooling tower. Wet-cooling towers are considered in this study. Wet-cooling takes place when th e water is in direct contact with the air. Cooling is the result of sensible and latent heat transfer where the latent heat transfer component generally dominates. Cooling towers can be classified according to the type of draft through the towe r. Figure 1.2 shows an example of a natural draft wet-cooling tower. The draft in natural draft towers is established by the buoyancy of the hotter air inside the tower shell compared to the cooler ambient air on the outside of the tower shell. Although the art of evaporative cooling is quite ancient, the first natural draft cooling tower was only constructed in 1916 at the Emma Pit in the Netherlands by the Dutch Sta te Mines [97BOI]. The worlds tallest cooling tower is 200 m high and is situated at the Niederaussern power plant in Germany [02BUl,02HAI]. 1.2 Tower shell Tower support Figw-e 1.2: Natural draft counterflow wet-cooling tower. Drift eliminator Sprays Fill Water basin Figw-e 1.3 shows an example ofa mechanical draft wet-cooling tower. Draft in mec hanical draft towers is established by fans that force or draw air through the towers, usually referred to as forced draft and induced draft respectively. A further distinction between cooling towers is whether they are counterflow or erossflow towers. Figw-es

1.2 and 1.3 show examples of counterflow cooling towers, while figw-e 1.4 shows an example of a erossflow cooling tower. In a erossflow tower the fill is usually installed at s ome angle to the vertical to make prOVision for the inward motion of the droplets due to drag forces caused b y the entering cooling air [98KR1]. Less pumping power is needed for modern counterflow towers, as the towe rs are generally not as high as erossflow cooling towers. Icing and wind effects are more prevalent i n erossflow towers than in counterflow towers [95BLl]. When a single cooling tower incorporates a wet and a dry section, this is also s ometimes referred to as a hybrid system [98KR1]. Figw-e 1.5 shows an example of a hybrid cooling tower [98 KR1, 98STl]. Hybrid cooling towers are generally used for plume abatement and in regions where water is relatively scarce. The operation of a wet-cooling tower relies on relatively simple principles. Hen sley [92HE I] and Krllger [98KR1] discuss the operating principles ofwet-cooling towers while Willa [92WIl ] presents a history of the development of wet-cooling towers during the last cenlW"y. The rest of this section deseribes the development, operating principles and limitations of the basic components ofwetcooling towers. 1.3 DiJfuser Fan Plenmn chamber Drift eliminator Sprays Fill SSSSSSSSSS~SSSSSSSSSSSSSSSSSSSSSSSSSSSSS$SlSSSSSSSSSS Figure 1.3: Induced draft counterflow wet-cooling tower. Fill Figure 1.4: Induced draft crossflow wet-cooling tower. Diffuser I-V Fan 1.4 Dilfuser Fan Mixers Heat exchanger Noise attenuator Drift eliminator Fill Rain zone Noise attenuator Water basin Figure 1.5: Hybrid cooling tower. Hot water is sprayed over 1he fill material. The spray zone can account for as m uch as 25% of 1he total heat transfer in a tower [01TUIl It is very important 1hat 1he water is distribu ted uniformly over 1he fill. Maldistribution of liquid flow is often cited as a cause of reduced performance in packed towers [93KR1, 95L1I). Mohiuddin and Kant [96M02] present different spray system designs. A poor water tlow distribution over 1he fill is commonly experienced at water fl owrates in excess of

around 4.2 kg/m's [96M02]. If 1he tlowrate is increased beyond this value, 1he w ater cascades in thick streams instead of falling as a spray, so 1hat 1he effective area is reduced. Th is condition is called flooding. On 1he o1her hand, if1he water flowrate drops to about 0.8 kg/m's or l ess, surface tension causes the waterflow to charmel. This gives a poor water distribution, and hence a mark ed drop in performance. The fill increases 1he transfer area by breaking 1he water up into smaller dropl ets or by forming a 1hin film depending. on 1he type of fill. The fill also increases 1he contact time between 1he water and 1he air [83MA I]. The factors influencing 1he choice of fill are its heat transfer perfo rmance, quality of water, pressure drop, cost and durability [96M02]. Over 1he last 30 years, 1here has be en a gradual change in 1he types of fill used in process cooliog towers [99WAI]. The most dramatic change h as been 1he introduction of film fills !bat provide significantly higher tbennal performance through tbe increase of 1.5 water-to-air contact area and a reduction in pressure drop. This results in a re duction in capital expenditures, lower operating costs and smaller tower footprint. However, in man y applications, due to poor water quality or potential process contamination, these benefits are forfei ted and the older splash fill technology is still used. The fihn fill designs can be grouped in three broad ca tegories: cross corrugated, vertical offset and vertical flow as can be seen in figure 1.6. Mirsky and Bauth ier [93MB] present a history of the development of wet-cooling tower fills. Aull and Krell [OOAUI] in vestigated the performance of various fihu fiUs. Air Water Vertical Fill T I t I Air Water Vertical Offset Fill Water Air (for counterflow) Cross-Corrugated Fill Figure 1.6: Fihufill designs [99WAI]. Air (for crossflow) From the fill the water falls unobstructed through the rain zone into the water basin. A significant amount of heat and mass transfer takes place in the rain zone. The drift eliminator is situated on the air downstream side of the fiU as shown in figures 1.2 to 1.3. Drift refers to the s mall droplets of circulating water that are carried out of the cooling tower by the exhaust air. Inertial imp action separators, known as drift eliminators, are used to strip the water droplets from the warm exhaust ai

r. In this type of separator, the two-phase exhaust flow is forced to abruptly change direction. This causes t he dense drift droplets to hit the eliminator walls and become trapped inside the cooling tower. Drift elim inators have evolved from early single-pass wood lath to multiple-pass wood and then to sinusoidal-wave sh apes. These were roUowed by combinations of sinusoidal and honeycomb shapes. Currently, various s tyles of ceUuiar drift eliminator packs are constructed from thermoformed sheets of polyvinylchloride ( pVC). The performance of these drift eliminator packs is measured by two criteria: droplet coUection e fficiency and system pressure loss caused by the eliminator pack To achieve peak operating efficiency of the overall cooling tower system, it is desirable that the system pressure loss from the eliminators be minimized [93BE1]. 1.2 OUTLINE OF THESIS 1.2.1 CHAPTER I Chapter 1 presents a broad overview of wet-cooling towers. The basic terminology and operation of natural draft and mechanical draft wet-cooling towers are explained The outline of the thesis is also presented in chapter I. 1.2.2 CHAPTER 2 1.6 Chapter 2 presents the theoretical heat and mass transfer analyses employed in w et-cooling tower performance evaluation. These analytical models or approaches are the foundation of any theoretical and experimental investigation into cooling tower performance. Three different analy tical models, the Merkel [25MEl], Poppe [91POl] and e-NTU [89JAI], are considered. These models are later employed to determine the transfer characteristics of the cooling tower fill materials and s ubsequently employed to determine cooling tower performance. 1.2.3 CHAPTER 3 The Merkel [25ME1], Poppe [9lPOl] and e-NTU [89JAl] approaches, discussed in cha pter 2, are employed to determine fill performance characteristics including the loss coeffi cients and the Merkel numbers, or transfer characteristics, according to each method of analysis. A ne w extended empirical relation for the loss coefficient is proposed where the viscous and form drag ef fects are accounted for as well as the buoyancy, momentum and fill height effects. It will be shown that th e proposed empirical relation gives very accurate correlations for splash, trickle and fihn fill type s, over a wide range of air and water mass flow rates when compared to other forms of empirical relations common ly found in the literature. The dependence of the transfer characteristic on the height of the f ill, inlet air drybulb temperature and inlet water temperature is investigated. It is shown that the tr ansfer characteristic per unit height is a function of the water and air flow rates as well as the fill height and the inlet water temperature but not of the air inlet temperature. 1.2.4 CHAPTER 4

The performance of natural and mechanical draft counterflow cooling towers is cr itically evaluated by respectively employing the Merkel, Poppe and e-NTU methods of analysis at differ ent operating and ambient conditions. The WCTPE software program, presented and developed in the a ppendices, is employed in the investigation. The importance of using a particular method of an alysis when evaluating the performance characteristics of a certain fill material and subsequently empl oying the same analytical approach to predict cooling tower performance, is investigated. Procedures to ev aluate and improve the accuracy of the Merkel and e-NTU methods, when compared to the more rigorous Pop pe method, are discussed. 1.2.5 CHAPTER 5 The effect of temperature and humidity inversions on cooling tower performance i s investigated. A very simple empirical relation of the nocturnal temperature inversion profile is deve loped in the appendices and presented in chapter 5. This empirical relation correlates measured data mor e accurately than models found in the literature which require more input data. An equation to determine the height of the temperature inversion is also developed. An analytical model to determine the he ight from which air is drawn into a cooling tower is developed. 1.2.6 CHAPTER 6 1.7 Chapter 6 gives a summary of all the main recommendations made and the conclusio ns drawn during the thesis. Most of the conclusions are repeated from the conclusions drawn at the e nd of each chapter. It serves as a complete overview of the main results and recommendations. The compu ter software programs developed to aid in the performance analysis of cooling towers and cool ing tower fills are summarized. 1.2.7 APPENDICES Most of the research, development and presentation of theoretical and analytical models, equations and computer programs are presented in the appendices. Most appendices are self-cont ained chapters with results and conclusions. The most important results of the appendices are summar ized and presented in the main chapters of the thesis while the details of calculations and the method s followed are presented in the appendices. ,,', 1.3 COMPUTER SOFIWARE DEVELOPMENT A program is developed to process and analyze fill performance test data. This p rogram is presented in appendix K and processes the pressure transducer and thermocouple data, determin es the transfer and loss coefficients and fits relatively complex curves through the test data with mathe matical optimization algorithms. A comprehensive program is developed to predict wet-cooling tower pe rformance. This program is presented in appendix P. Natural draft counterflow and mechanical dra ft counterflow and crossflow cooling towers can be analyzed by the program. The latest empirical an

d heat and mass transfer models found in the literature are included in the solution algorithms of the so ftware. The analytical and empirical models, developed in this thesis from theoretical and experimental inv estigations, are also included in the software. The geometrical dimensions of a natural draft cooling tower can be optimized by the program to obtain the minimum combined capital and operational cost compound ed over the economic life of the cooling tower. Furthermore, programs are developed to plot psychrometric charts (chapter 2) and generate cooling tower performance curves (appendix Q). Comment In the numerical examples, given in the appendices, values are often given to a large number of decimal places. These numbers are usually as given directly by the computer program outp ut and do not necessarily imply a corresponding degree of accuracy. 2.1 CHAPTER 2 WET-COOLING TOWER HEAT AND MASS TRANSFER ANALYSIS 2.1 INTRODUCTION The analytical models or approaches that predict heat and mass transfer in cooli ng towers are the foundation of any theoretical and experimental investigation into cooling tower performance. It is thus imperative to understand their limitations and applications. Three different ana lytical models, referred to as the Merkel, Poppe and e-NTU approaches respectively, are employed in this stu dy to evaluate the heat and mass translior processes in wet-cooling towers. These models are later emplo yed to determine the transfer characteristics of cooling tower fill materials and subsequently to det ermine cooling tower performance. The heat and mass transfer processes are presented graphically with the aid of psychrometric charts. 2.2 PSYCHROMETRIC CHART Psychrometric charts are useful and widely accepted tools for the design and ana lysis of heat and mass exchange involving moist air [82STl]. Properties of air-water vapor mixtures are presented in graphical form on psychrometric charts. The state of air at a specified pressure is comple tely specified by two independent intensive properties. The basic features of the psychrometric chart are illustrated in figure 2.1. The drybulb temperatures are shown on the abscissa and the specific humidit y or humidity ratio is shown on the vertical axis. On the left end of the chart there is a curve called the saturation curve where the relative humidity is 100%. All the saturated air states are located on this curve. Other constant relative humidity curves have the same general shape. The enthalpy has an inclined coordi nate. Although it is not shown in figure 2.1, wetbulb temperature and specific volume can also be present ed on a psychrometric chart. Supersaturated region curve Drybulb temperature, K

Figure 2.1: Psychrometric chart coordinates. 2.2 Figure 2.2 illustrates the various processes of air-vapor mixtures on a psychrom etric chart. If only heat transfer is present and no mass transfur, it can be seen that the humidity ratio remains constant, since the moisture content of the air remains constant, and pure heating or cooling of the air occurs. Heatinglllld HumidifICation Heating and Dehumidification Cooling lIlld Dc::hum idiflCation Cooling +--~E---.Heating Dehumidification Drybulb temperature, K Figure 2.2: Various processes depicted on a psychrometric chart. Psychrometric charts only describe what happens to the air in wet-cooling proces ses and the charts can generally not be used in the supersaturated region, illustrated in figure 2.1. T he outlet air in practical wet? cooling towers, however, is generally in this latter region. However, the psychr ometric chart is an excellent tool to analyze and describe the direction of enthalpy transfer in coo ling towers. Drybulb temperature, K (a) Figure 2.3: Psychrometric charts w (b) , , , T' Drybulb temperature, K Enthalpy, J/kg imosw l?w. " , , , , , ' , ' , \ : \ w Enthalpy, J/kg The enthalpy potential provides a qualitative indication of the direction of net t heat flow in the fill region of cooling towers. Air at condition x (refer to figure 2.3) is in contact with w ater at temperature Tw? Figure 2.3 represent two different cases that can occur inside a cooling tower fill. Co nsider the case in figure 2.3 (a) where w,w > w, thus, the latent heat transfer is from the water to the air a nd Tw> Ta, where the sensible heat transfer is from the water to the air. The total enthalpy transfer is from the water to the air since im~ > ima and since both the latent and sensible heat transfer are from the wate r to the air. The air is heated and the water is cooled. 2.3

However, both the air ood the water coo he cooled, while the nett enthalpy troos fer is still in the direction of the air. Consider the case presented in figure 2.3 (b), where w,. > w, thus, the latent heat troosfer is from the water to the air ood Ta > T., where the sensible heat troosfer is from the air to the water. The nett enthalpy traosfer is from the water to the air since ima,. > ima? Furthermore, G oyal [00G02] states that it is a common misconception that cooling towers coonot operate when the inlet air is saturated. Even though the inlet air is saturated there still exists a potential fur sensible ood laten t heat troosfer. The excess water vapor troosferred to the free stream air will condense as a mist [98KRI]. Figure 2.4: Dialog window ofthe psychrometric chart generator computer program. Figure 2.4 shows the dialog window of a compute~program that is developed to gen erate psychrometric charts. The psychrometric charts are used in conjunction with the heat ood mass troosfer models, presented next, to graphically represent the troosfer process. 2.3 MERKEL ANALYSIS Merkel [25MEI] developed the theory for the perfurmooce evaluation of cooling to wers in 1925. This work was largely neglected until 1941 when the paper was trooslated into English . Since then, the model has been widely applied [9IOSI]. The Merkel model is a very popular model ood it s employment is recommended by international stoodards [88BRI, 90COI, 97COI]. The Merkel theory relies on several critical assumptions to reduce the solution of heat ood mass traosfer in wet-coo ling towers to a simple hood calculation. Because of these assumptions, however, the Merkel method does not accurately represent the physics of heat ood mass troosfer process in the cooling tower fil l. The critical simplifying 2.4 assumptions of the Merkel theory are that the Lewis factor, Lef> is equal to uni ty, the exiting air is saturated and the reduction of the water flow rate, due to evaporation, is negle cted in the energy balance. Appendix B gives a detailed derivation, from first principles, of what is common ly referred to as Merkel's equation for a counterflow configuration. Merkel's equation is given by equation (B.21) and is repeated here, (2.1) where MeM is the Merkel number according to the Merkel theory. The Merkel number is a non-dimensional coefficient of performance. The right-han d side of equation (2.1) can be solved if the water inlet temperature, water outlet temperature, ai r inlet drybulb temperature, air inlet wetbulb temperature, water mass flow rate and airflow rate are known. The mass transfer coefficient, hd, and the surface area per unit volume, aft, of a particular fill are practically impossible to determine [OIROI]. However, hd and aft exist as a product inside the Merkel numb er, as seen in equation (2.1), and it is therefore not necessary to specify them explicitly. The heat tr ansfer coefficient, h, also does not have to be specified explicitly as it is coupled to the mass transfer c oefficient, as can be seen

from equation (F.lO), through the assumption that the Lewis factor is equal to I . In the literature the notation frequently used for the Merkel number is KaV/L where K = ~, a = aft and L = mw? Refer to appendix C for the derivation ofthe governing equations for a crossflow configur ation. It can be seen from equation (2.1) that the Merkel number, or transfer character istic, can be obtained from the evaluation ofa simple integral. Equation (2.1), however, is not self-suffici ent so it does not lend itself to direct mathematical solution [6IBAI, 82MIl]. The usual procedure is to integr ate it in conjunction with an energy balance expressed by (2.2) Figure 2.5 shows the enthalpy curves of the air in a counterflow wet-cooling tow er. The fill test results, .from which figure 2.5 is generated, are given in the beginning of appendix G. I . The ima curve, i.e. the enthalpy of the air as it moves through the fill, shown in figure 2.5, is linear due to the linear nature of equation (2.2). The ima<w curve is the saturation curve of the air at the water interface temperature. The potential for heat and mass transfer at a particular water temperature is the di fference between im~w and ima? The Merkel number, MeM, of equation (2.1), is a function of the area under the lI(ima<w - ima) curve as shown in figure 2.5. The integral in equation (2.1) needs to be evaluated by numerical integration te chniques. The British Standard [88BRI] and the Cooling Tower Institute [90CAI, 97CAI] recommends that the four-point Chebyshev integration technique be employed. A discussion of the Chebyshev integ ration technique can also be found in Oosthuizen [95001] and Mohiuddin and Kant [96MO1]. 2.5 312310308306304302 L..__..J....__-l. L..__..J....__....I..__---1I.-_--J 1.1x1 0-5 314 20000 300 170000 1.6x10-5 1/(i rnasw - i rna ) 140000 1.5x10-5 i ~ ,.....:;110000 1.4x10?5 ~....., .~ ;>: f '"80000 1.3x10-5 ~.~ ~ ..... 50000 1.2x10-5 Water temperature, K Figure 2.5: Enthalpy diagram ofthe Merkel approach. The four-point Chebyshev integration technique essentially uses four intervals f or the determination of the integral. Li and Priddy [85LIl] and Mills [95MIl] use thirteen and seven int ervals respectively for numerical integration to determine the change of water and air enthalpy through the fill for a cooling range of approximately 14?C. Li and Priddy [85LIl] effectively employ a Riemann

sum [90ELl] to determine the integral while Mills [95MIl] employs the composite trapezoidal rul e [97BUl]. The composite Simpson rule is employed in this investigation to determine the ac curacy of the Chebyshev procedure. The Simpson rule is chosen because of its superior accuracy compared to the trapezoidal rule. The error of the trapezoidal rule is ofthe second order while that of the Simpso n rule is ofthe fourth order [92MAl]. Any number of intervals can be chosen for the Simpson rule while that o f the Chebyshev procedure is fixed. (2.3) The integral of the 1I0m",w - ima) curve in figure 2.5 multiplied by Cpw is 0.68 468 while employing the Chebyshev procedure. If the Simpson rule is employed with 150 intervals the inte gral is 0.684876, which is only a 0.03% change from the value obtained by employing the Chebyshev proced ure. If only one interval is used in conjunction with the trapezoidal rule, the integral, or Merk el number is given by cpwm(Twl -Two )[ 1 1]~M~ .. +. . 2 'masw, -'mal 'maswo -'mao For two intervals the Merkel equation is (2.4) 2.6 cp.m(T., -T?.)[ 1 2 1] MeM '" 4 . -' + + lmaswi lmai imaswm - imaM imaswo - imao The Merkel numbers are 0.667587 and 0.685014 when equations (2.3) and (2.4) are respectively employed. This is 2.5% and 0.05% change respectively from the value determined b y the Chebyshev procedure. For the case investigated above is it found that the Chebyshev procedure is very accurate if compared to results ofthe Simpson procedure with many intervals. Kelly [76KE2] states that t he Chebyshev procedure lacks accuracy when the approach (i.e. the difference between the water outlet t emperature and the air inlet wetbulb temperature) is small (down to 0.56 0c). The accuracy of equations (2.3) and (2.4) depend on the degree of curvature of the lI(im". - ima) curve as shown in figure 2.5. A ny integration technique can be employed to solve equation (2.1) but it is strongly recommended that the same integration technique be employed in the fill performance analysis and the subsequent coolin g tower performance analysis. This point will be substantiated in further investigations later in th is study. As already mentioned, the driving potential in wet-cooling towers is the differe nce between the enthalpies im... and ima as shown in figure 2.5. The ima curve is obtained from equation (2 .2) that ignores the change in water flow rate due to evaporation. The effect of evaporation on the energy b alance is thus ignored for a second time. It was first ignored when equation (2.1) was derived as seen in a ppendix B. Baker and Shryock [6IBAI] investigated the effect of this second time the effect of evapor ation is ignored in the

energy balance. They've considered three different cases and found that the Merk el number increases with the more accurate representations of the energy balance. However, the Merke l number increases not as much for the most accurate case investigated as for the second most accurate case. The maximum increase in the Merkel number is 4.4%. Again, it is stressed that the same energ y balance be employed in the fill performance analysis and the subsequent cooling tower performance analy sis. Curves are published in the literature to determine the Merkel number in equatio n (2.1) by graphical means from known air and water temperatures and air and water mass flow rates. C urves to determine the tower characteristic for counterflow towers are given by the cn [67CTI] and for crossflow towers by Kelly [76KEI]. Figure 2.6 is an example of such a curve for a counterflow tower for a particular cooling range and wetbulb temperature. Since the advent of high speed digital computers, these curves are less frequently used. The cooling process shown in the enthalpy diagram of figure 2.5 can also be indi cated on a psychrometric chart as shown in figure 2.7. The Merkel approach is shown as a broken straight line in figure 2.7. The line for the Merkel approach is presented as a broken line because straight line s can only be used on psychrometric charts if the temperature of the water surface is constant. The li ne for the Merkel approach is presented as a straight line because no other information is given by the Mer kel theory about the humidity of the air, except that it is saturated at the air outlet side. That is why the air at the outlet of the cooling tower is assumed to be on the saturation line as shown in figure 2.7. 2.7 1. m"Im. Figure 2.6: Counterflow tower characteristic curves [67CTl). Enthalpy, kJlkg dry air 35302520 50 60 70 80 90 100 15 40 10 Poppe Merkel 5 ~A~t~m~o~.p~h~en~?c:.:pres~~.~ure~:1~0~1~71~2:::.2~7~p~at---jIH-~~~7~~~~~:f 0.02 5 "i; ~ .., 0.020 JP i' 0.015 i ci ~jjlll!~~~0'010i0.005 ?0.000 40 20 Drybulb temperature, Ta., ee Figure 2.7: Psychrometric chart

2.8 2.4 POPPE ANALYSIS The poppe model was developed by Poppe and ROgener [84POl, 91POl] in the early s eventies. The method of Poppe does not make the simplifying assumptions made by Merkel. The de rivation of the governing equations of heat and mass transfer, for a counterflow configuration, according to Poppe, is given in appendix B. Equations (B.24), (B.25) and (B.30) are the governing equations for the counterf low Poppe method when the air is not saturated with water vapor while equations (BA2), (BA3) and (BA7) are the governing equations when the air is supersaturated. For a crossflow configuration, the two -dimensional partial differential equations are given in appendix C. Again, as in the case with the Merkel method, hd and aft appear in the Merkel nu mber. The heat transfer coefficient, h, is obtained from the Lewis factor, Left but this time the Lewis factor is not assumed to be unity. It must be stressed, however, that the heat and mass transfer coefficient s are never specified explicitly when the governing equations are solved. The Lewis factor, Lef, and its application to evaporative air-water systems is d iscussed in detail in appendix F. Poppe employs equation (F.16) to express the Lewis factor in his mod el. Equation (F.l6), developed by Bosnjakovic [65BOl], is also the preferred equation to express the Lewis factor in this study. Other equations, given in appendix F, can be employed to express the Lewi s factor. It will be shown later in this study that it is very important to employ the same equation or definition for the Lewis factor in the fill performance analysis and in the subsequent cooling tower perf ormance analysis. This consistent usage of definitions, in all aspects of the governing equations, in t he fill performance analysis and in the subsequent cooling tower performance analysis, for all the transfer m odels, is very important to obtain accurate and reliable results. It is expected that the Poppe approach will lead to more accurate results than t hat obtained by employing the Merkel approach, as it is the more rigorous approach. The comparison between the Poppe and Merkel approaches is shown on the psychrometric chart in figure 2.7. The humidity of th e air through the entire cooling process is predicted by the Poppe approach, unlike the Merkel approach w here only the outlet condition of the air is known, i.e. it is saturated. Figure 2.8 shows the differences in the enthalpy diagrams between the Merkel and Poppe approaches. The ima,. curves of the two approaches fall on top of each other. There is a sma ll discrepancy in the ima curves of the two different approaches, especially at the hot water side. It can be seen that the Poppe approach predicts an appoximately linear variation of the air enthalpy for this specific case, but the gradient is different from that predicted by the Merkel approach. The l/lima,. ima) curve of the Poppe approach lies above the l/(im~. - ima) curve of the Merkel approach. As the tran

sfer characteristic, or Merkel number, is a function of the area under the lI(im~ - ima) curve, the Merk el number according to the Poppe approach will be greater than the Merkel number predicted by the Merke l approach. 2.9 1.1x10-5 314312310308306304302 20000 300 170000 1.6x10-5 lI(imasw - ima) 140000 1.5x10-5 ~ OIl -Merkel ~~ 110000 1.4x10?5 r:. __ Poppe J>; I ~80000 1.3x10-5 00J::::::' 50000 1.2x10?5 Water temperature, K Figure 2.8: Enthalpy diagram of the Merkel and Poppe approaches. Figure 2.9 shows an example of a psychrometric chart for a water cooling process solved by the Poppe method. The inlet air to the cooling process is very hot and dry. It can be seen that the temperature of the outlet air is cooler than the inlet air. This scenario is explained in section 2 .2 with the aid of figure 2.3(b). -i 140/' /7-Atmospheric pressure: 84100 Pa I 13O!: 90 1/ 7 '>0/ /' 80 I 119:::: "/ /17O:~~;Enthalpy, kJ/kg dry air ../5./ 01 .-/./ 6O-t.<E_ 7n Y-V ,..,..- ~ ...50~;:; BQ...;' :;.....--........ .................. ,.-/ ...... 40 i-30~~ .......;:...-.........- ...-~ ~ 20""2 " -~ 10 " 10 15 20 25 30 35 0.045 0.040 ~ 0.035 'ii ~ '"0.030 If 0.025 } i 0.020 f 0.015 ~ :ll 0.010 5 :I: 0.005 0.000 40 Drybulb temperature, Ta. ?C Figure 2.9: Psychrometric chart with a process determined by the Poppe method.

2.5 e-NTU ANALYSIS The e-NTUmethod, developed by Jaber and Webb [89JAl], is based on the assumption s made by Merkel. The results obtained by employing this approach is therefore not very different from that obtained by employing the Merkel approach. The difference between these two models is essent ially that a different '\" .~'~' ':, r''''i, ii: 1" integration procedure is employed to obtain the area under the l/(ima"" - im,) c urve, as shown in figure 25. A briefdescription of the e-NTU model is presented in appendix B5, The Merkel number according to the e-NTU approach, Me" is given by = Cpw NTU Me, d' jdT 'masw w (2.5) if ma is greater than mwCp,j(dim~,jdTw). If m, is less than mwCp,j(dim~dTw) the Merkel number according to the e-NTU approach is given by Me = m, NTU, mw (2.6) where NTU is given by equation (B.73) for counterflow cooling towers. The great advantage of the effectiveness-NTUapproach is its simplicity in the application of crossflow conf igurations. For crossflow, however, it must be specified whether the air and water streams are mixed or unm ixed or a combination of mixed and unmixed. Thus, there exists a choice of four possible flow geometri es for crossflow. The question now is which geometry will yield the most accurate results for a partic ular fill material. It is not important which flow geometry is chosen, as long as the same geometry is assumed for both the fill performance analysis and the subsequent cooling tower performance analysis. Ifused consistently, the four different geometries together with the Merkel approach and Poppe approach w ill predict practically identical water outlet temperatures if all other variables are assumed to be con stant and if the fill test analysis and the subsequent cooling tower performance analysis is conducted at t he same operating conditions. 2.6 OTHER ANALYSES FOR EVALUATING COOLING TOWER PERFORMANCE Papers are regularly published in the literature that present heat and mass tran sfer models in cooling tower applications. These models differ in complexity and are essentially based on the Merkel or Poppe models. Some of these models are according to Nahavandi et al. [75NAl], Montakhab [78MOl ], Bourrilot [83BOl, 83B02], Sutherland [83SUI], Feltzin and Benton [9IFEl], Bernier [95BEl], Ibrahim et al. [95IBl], Sadasivam and Balakrishnan [95SAl], EI-Dessouky et al. [97ELl], AI-Nimr [98ALl], Soylemez [99S01], Makkinejad [OIMAl], Milosavljevic and Heikkila [OIMIl] and Fis

enko et al. [02FIl]. There are two-dimensional models that calculate the flow field in the cooling to wers by finite difference equations. These models can therefore accommodate non-uniform air and water flow distributions. These models are according to Majumdar [83MAl, 83MA2, 83MA3] and Hawlader and Lui [02H A2]. Johnson [89JOl] presents a comprehensive list of assumptions used for some of th e models mentioned above. Mohiuddin and Kant [96MOl] present a summary and overview of the some of the models found in the literature. 2.11 2.7 CONCLUSION The consistent employment of the heat and mass transfer model in the fill perfur mance evaluation and then using the same model in the subsequent cooling tower performance analysis i s stressed. If used consistently the different models ought to give the same cooling ranges for the water in a particular cooling tower if all the operating conditions are exactly the same for each mode l. Because the Poppe approach is the more rigorous approach, it will predict the water evaporation ra te, the total heat transfer rate and thus the air outlet temperature more accurately than the other approach es. This may lead to situations where the predicted cooling tower operating conditions will not be th e same as those predicted by the other approaches, and it may therefore predict cooling ranges different f rom those predicted by the Merkel or e-NTU approaches. For example, the draft through natural draft cooling towers is a function of the air outlet temperature and the Poppe method will thus predict more accurate tower draft and tower performance. The Poppe method also has distinct advantages in the analysis of hy brid cooling towers since the humidity of the outlet air is calculated, even if the air is supersatu rated [OIROI]. This information is important to ensure that the correct amount of heated dry air is mixed with the wet plume to ensure no visible plume after mixing of the two streams. 3.1 CHAPTER 3 FILL PERFORMANCE EVALUATION 3.1 INTRODUCTION The Merkel [25MB1), Poppe [91P01) and e-NTU [89JAI) methods to evaluate cooling tower performance were discussed in the previous chapter. In this chapter the differen t methods of analysis are employed to determine fill performance transfer characteristics. Fill transfer a nd loss coefficient correlations given in the literature for wet-cooling tower fills are relatively simple and are generally not accurate over a wide range of operational conditions. A new extended empirical r elation for the loss coefficient is proposed where the viscous and furm drag effects are accounted fo r as well as the buoyancy, momentum and fill height effects. It is shown that the proposed empiri cal relation gives very

accurate correlations for splash, trickle and film fill types, over a wide range of air and water mass flow rates when compared to other forms of empirical relations commonly found in the literature. The dependence of the transfer characteristic on the height of the fill, inlet air d rybulb temperature and inlet water temperature is investigated. It is shown that the transfer characteristic per unit height is a function of the fill height and the inlet water temperature but not of the air inlet temp erature. The empirical relations fur the loss and transfer coefficients do not include effects of diffe rent spray types or ageing effects. 3.2 LOSS COEFFICIENT The loss coefficient of a cooling tower fill is determined by measuring the pres sure drop over the fill during the testing phase. Empirical relations are then obtained for the loss coe fficient of the fill as a function of the air and water mass flow rates. These empirical relations are sub sequently employed in the design of cooling towers to determine the draft through the cooling towers. Suit able fans for mechanical draft cooling towers are selected based among others, on the loss coefficient of the fill. The draft in a natural draft cooling tower is a function of the fill loss coefficient. It is th us important to represent the fill loss coefficient accurately, as inaccurate representation of the loss coefficien ts in the form of empirical relations can have financial implications if the cooling tower does not meet des ign specifications. The fill loss coefficient is defined as (3.1) where /',pji is the measured stalic pressure drop across the fill. The static pressure drop across the fill (lipji) is due to viscous drag (frictio nal drag) and form drag resistance in addition to the acceleration of the air due to heating and mass tr ansfer, while the buoyancy 3.2 due to the difference in density of the air in the fill and that in the manomete r tube external to the test section will tend to counteract these effects in cases ofcounterflow [98KRl], i. e. /':.p fi = lipid +(P~ v;'" - P~jV~j)- (P~ - p~.)gLfi (3.2) where the subscriptfd refers to frictional and drag effects and Po", is the dens ity of the ambient air which is essentially equal to the density of the air entering the fill i.e. Po,j. The density of the air leaving the fill is Po'" and the mean harmonic density Po,m ~ 2/(l/Pow + l/Po",). The second term on the right-hand side of equation (3.2) represents the momentum change experienced by the air stream whil e the third term considers buoyancy effects. This equation assumes that the porosity of the parti cular fill, which is defined as the ratio of the free flow area at a cross-section to the corresponding cross -sectional area of the fill, is unity. In the absence of momentum changes a loss coefficient which is determined by frictional and drag effect can be defined, i.e. Kid = 2/':.pId /(pv') = 2[tipft - (Powv;w - Powv;.)+ (P~, - p".)gLfi ]/(pv') (3. 3)

In practice the reference conditions chosen for the denominator in equation (3.3 ) differ. For example, the loss coefficient for a particular fill can be defined in terms of the mean air-v apor flow rate and its density through the fill i.e. Kid. = 2[/':.pfi - (Pow v;" - Powv;,,)+ (P"", - P". )gLft lo".A~ /m;'. (3.4) where rna"" ~ Po,m V.,mAf Per unit height of the fill it follows from equation ( 3.4) that Kfdmi = KjdmlLfi. The following measurements are generally made during fill tests where the transf er coeffcients and loss coefficients are determined: the air inlet drybulb temperature (To,), and the ai r wetbulb temperature (Twb), the water inlet temperature (Tw'), the water outlet temperature (Two), the water mass flow rate (mw) and the air mass flow rate (rna). The atmospheric pressure (Po) is also measured to dete rmine the humidity ratio of the inlet air (Wi). The air outlet drybulb temperature (Tao) is generally not me asured since it is relatively difficult to measure accurately because of condensation, drift and supersaturati on of the outlet air. The outlet air temperature is not employed in the Merkel [25MEl] or Poppe [91POl] th eories to determine the transfer coefficient. However, the outlet temperature can be predicted by these theories. Merkel assumed that the outlet air is saturated which enabled him to determine the outlet air t emperature from a simple energy balance. In the case of the Poppe theory the outlet air temperature is ev aluated as Poppe did not make the simplifYing assumptions ofthe Merkel approach. The loss coefficient as given by equation (3.4) is dependent on the air outlet t emperature. Since the Poppe approach generally predicts higher air outlet temperatures than the Merkel metho d, the loss coefficients will differ. This difference, however, is generally small. 3.2.1 EMPIRICAL EQUATIONS Lowe and Christie [6ILOl] used the following form of equation to represent the l oss coefficients of counterflow splash and film type fills. Kfl=C{~:)+C' (3.5) 3.3 where c, and c, are empirical constants that depend on the fill design. The empi rical relations of Lowe and Christie [6ILOl] are widely applied and cited by other researchers [83MA3, 9 6M02, 98KRI]. Majumdar et al. [83MA3] correlated the data in Kelly [76KEI] by employing equati on (3.5). Johnson [89JO I] gives fill loss coefficient test results for counterflow cellul ar type fills with variable heights as (3.6) where Lfl is the height of the fill. If the fill height is constant then equatio n (3.6) becomes (3.7) Baard [98BA I] conducted extensive tests on expanded metal type fills in various configurations and employed equation (3.7) to correlate his pressure drop data. The correlation coe fficients obtained by Baard [98BA1] indicates that equation (3.7) does not necessarily correlate the m easured data accurately for some fill configurations. He obtained correlation coefficients ranging from

0.61 to 0.98. Milosavljevic and Heikkilli [OIMII] tested seven types of counterflow film type fills and correlated their pressure drop data with !'J.pfilLfi =c,(I+G;<~J; (3.8) Goshayshi and Missenden [OOGOI] tested seven types of counterflow film type fill s in various arrangements. Their tests were conducted in a 0.15 m x 0.15 m counterflow test s ection where Ga is varied between 0.2 and 1.5 kg/m's, and G. is varied between 0.45 to 2.22 kg/m's. These mass velocities are very low and are not typical for industrial applications [85Lll]. Their fill test data is correlated by (3.9) where c, and c, are constant for all the fills tested. Goshayshi and Missenden [ OOGOI] reported a maximum error of?3% for equation (3.9) when applied to their tests. 3.2.2 NEW EMPIRICAL EQUAnON The loss coefficient is essentially a drag coefficient. Figure 3.1 shows the dra g coefficients of two simple shapes as a function of the Reynolds number. The total drag on a body placed in a stream of fluid consists ofskin friction and of form or pressure drag. The sum ofthe two is called the to tal drag [60SCI]. It can be seen in figure 3.1 that the drag coefficient at low Reynolds numbers d ecreases for increasing Reynolds numbers. This is due to the filet that viscous or friction effects pred ominate. The curve flattens out and remains essentially constant at high Reynolds numbers. Form drag is pred ominant in this region. The reason for the existence of form drag lies in the filet that the boundary la yer displaces the external, potential flow [60SCI]. 3.4 "~, '" " ,~ ~"- ---- -, '---. I L::= -- Sphere - - lnfinite Cylinder (3.10) 102 103 Reynolds number, Re Figure 3.1: Drag coefficient for bodies of revolution (adapted from Daugherty et al. [89DA ID. The Ergun [52ERI] equation for the pressure drop through packed beds is given by dp 150jlV 1.75pV' - =--+---'--dx [' [ where V and [ are the characteristic velocity and characteristic length respecti vely. The first terms accounts for the viscous drag, and the second term accounts for form drag. The c haracteristic length is constant fur a specific packed bed while the characteristic velocity is a functi on of the air velocity. If a

cooling tower fill is approximated by a packed bed, V and [ will also be a funct ion of water mass flow rate. Water droplets may be retained in the fill area or be entrained by the air when the drag force acting on the droplets is greater or equal to the weight of the water droplets. This ph enomenon is a function of the air velocity and the water droplet size and ultimately on the type and confi guration of the fill. Wet? cooling tower fills differ from packed beds as the pebbles (or water droplets in this case) are not static, and of variable shape, quantity and size. The fill, of course, is stationary. Ho wever, equation (3.10) gives a basis of what form a generalized correlation for pressure drop in fills must t ake. The pressure drop is a sum of two terms where each term is a function of the air and water mass flow ra tes. Thus, a new general empirical relation is proposed which accounts for the form drag and viscous drag effects as well as the effects that are dependent on the water mass flow rate and the configuration of the fill, Le. K - G"G" G"G"fl - C1 w 8 +C" W Q (3.11) 3.2.3 ACCURACY OF EMPIRICAL EQUATIONS Splash, trickle and film type fills are tested to show the accuracy and generali ty of equation (3.11) compared to that of equation (3.7), that is commonly found in the literature, an d the equation, (3.12) The form of equation (3.12) is commonly encountered in the literature to represe nt the Merkel number, but it is applied here to represent the loss coefficient. 3.5 The experimental results fur the trickle, splash and film type fills are respect ively presented in appendix R, appendix 8 and appendix T. The results in these appendices are obtained by em ploying the methods and computer program presented in appendix K. Table R.14 shows the empirical equations of the loss coefficients of three diffe rent trickle fill heights obtained from experimental tests presented in appendix R. It can be seen that th e correlation coefficient, while employing equation (3.11), is very accurate when compared to that of equat ions (3.7) and (3.12). Figures R.3, R.5 and R.7 show the comparative curve fits of the three different fill heights. The superiority of eIi~ation (3.11) to accurately represent the measured data is evi dent from these figures. Table 8.18 shows the loss coefficient empirical equations of splash fills for fo ur different splash fill spacings. It is again evident from the correlation coefficients that equation (3 .11) is superior to equations (3.7) and (3.12) to represent the measured data accurately. Figures 8.3, 8.5, 8. 7 and 8.9 show the comparative curve fits for fill spacings of 0.1, 0.2, 0.3 and 0.4 m respectively . It can be seen that equation (3.11) represents the measured data very accurately when compared to equations ( 3.7) and (3.12). Table T.14 shows the loss coefficient empirical equations of cross-corrugated fi lm fills for three different fill heights. It is again evident from the correlation coefficients that equatio n (3.11) is superior to

equations (3.7) and (3.12) in representing the measured data accurately. Figures T.3, T.5 and T.7 show the comparative curve fits for fill heights of 0.6, 0.9 and 1.2 m respectively. It c an be seen that equation (3.11) represents the measured data accurately when compared to equations (3.7) and (3.12). Majumdar et al. [83MA3] correlated the data in Kelly for employment in their VER A2D program for the heat and mass transfer analysis of wet-cooling towers. As already mentioned, the y employed equation (3.5). Figure 3.2 shows Kelly's [76KEI] data for a type F fill correlated by Maj umdar et aI. [83MA3] by employing equation (3.5). The air flow range employed in the experiments of Kell y is relatively narrow compared to the experiments conducted in this investigation. Correlations of Kel ly's data by employing equation (3.7) and (3.11) are also shown in figure 3.2. In this instance, equati on (3.7) and equation (3.11) give virtually identical results with correlation coefficients for both equal to 0.9991. 3.2.4 EFFECT OF FILL HEIGHT AND AIR AND WATER TEMPERATURES ON THE L088 COEFFICIENT Equation (3.11) correlates the measured data presented in appendices R, 8 and T for trickle, splash and film type fills respectively, and the data in the literature, relatively accurat ely. It is further investigated in appendix R if the loss coefficient is a function of the fill height, water inlet temperature and air inlet temperature. Equation (R.4) and figure R.IO show that the loss coefficient per u nit height of fill is a function of the height ofthe fill. Equation (3.11) must therefore be extended to include the effect of the height ofthe fill on the loss coefficient per unit height ofthe fill, K _I G" G" G" GC6 \ 7C7fil -\C1 W Q +c4 W Q JLft (3.13) 3.6 It can be seen from equations (R.5) and (R.14) that the loss coefficient is not a strong function of the water inlet temperature, T." and air inlet temperature, Ta;, respectively as the exponents of T.; and Ta; in these equations are very small. Equation (3.13) is thus adequate for correlating loss coefficient data. <> Gw -1.36 ? Gw = 4.07 t;. Gw = 6.78 ? Gw = 9.49 I0 Gw = 12.03 ? Gw = 16.27 . . ? ? Kelly --Equations (3.7) and (3.11) . . . . . ? . . o _ 0 . ?. ? 0 . . n' 0 0 . . . ? . . . .

? . . ? ? 0 . ? ? . . . o ? ? ? . ? ? 0 . 0 ? o ,.. . . -. ? . 0 ? ? . ? ? . --? 0 . ~ ? 0 0 . ? . A A 2 5 6 3.33.23.132.92.82.72.62.5 o 2.4 \, '"s 4 E :'3 7 Ga. kg/rri's Figure 3.2: Data by Kelly [76KEI] correlated by Majumdar et al. [83MA3] (equatio n (3.5)) and equations (3.7) and (3.11). 3.3 TRANSFER CHARACTERISTIC In the fill perfonnance test phase, the water outlet temperature, together with the other variables mentioned in appendix K, are measured under controlled operating conditions. An empirical equation for the transfer characteristic or Merkel number is then detennined from these measu rements. In the subsequent cooling tower performance analysis, the water outlet temperature is d etermined from the known transfer characteristic or Merkel number. 3.3.1 EMPIRICAL EQUATIONS

Lowe and Christie [6ILOI] used the fullowing form of equation to represent the M erkel numbers of counterflow splash and film type fills. (3.14) where c\ and c, are constants. KrOger [98KRI] and Baard [98BA I] use the following form ofequation to represent the Merkel number (3.15) where Lft is the length of the fill and Cb C, and c, are constants. Johnson [89101] expresses the Merkel number fur counterflow cellular type fills with the relation ~''""'''' ;1:"":,0,,' fi1 I:' c' ':' F I'.'., i: I ;" 3"7 O "0 "T"Me/LjI ==c1 w a wi where the Merkel number is a function ofthe water inlet temperature. (3.16) The Merkel numbers in equations (3.14) and (3.15) are only functions of the air and water mass velocities. These empirical equations, to represent the Merkel number or transfe r characteristic, are gross simplifications of a very complex heat and mass transfer process" Equation (3.14 ) assumes that the absolute values of c, and C3 in equation (3"15) are equal. Equation (3.16) makes provision for changes in the inlet water temperature. 3.3.2 ACCURACY OF EMPIRICAL EQUATIONS A more general equation is proposed for expressing the Merkel number, i.e", H /L - G"G" + G"G"JV.J.e fi - c 1 w a C4 w a (3.17) where equation (3" 17) is the same form as equation (3.11) that is proposed for the loss coefficient Splash, trickle and film type fill tests are tested to evaluate the accuracy and general ity of equations (3.14), (3.15) and (3"17)" ( Table R" 13 summarizes the empirical equations of the Merkel numbers for the t rickle fill for three different fill heights" It can be seen from the correlation coefficients in tabl e R.13 and figures R2, RA and R.6 that equation (3"14) is the least accurate. The accuracy of equations (3.15) and (3.17) is of the same order. Table S.17 summarizes the empirical equations and correlation coefficients of th e Merkel numbers for the splash fill for fill spacings of 0.1, 0.2, 0.3 and OA m. Equation (3"14) is agai n the least accurate" Equations (3" 15) and (3.17) have the same order of accuracy" Table T.13 summarizes the empirical equations of the Merkel numbers for the cros s-corrugated film fill for three different fill heights" It can be seen from the correlation coefficien ts in table T.13 and figures T.2, T.4 and T.6 that equation (3.14) is the least accurate. The accuracy of equ

ations (3.15) and (3.17) is of the same order. It is therefore clear that equation (3"14) employed by Lowe and Christie [6ILOI] does not always represent the test data accurately. It is only accurate in limited conditions wh ere the exponents c, and C3 in equation (3.15) is close to each other. Equations (3.15) and (3.17) correlates f ill performance test data with approximately the same degree of accuracy for all the types of fills tested . Equation (3.15) can be used instead ofequation (3.17) as it is the simpler of the two equations. 3.3.3 EFFECT OF FILL HEIGHT ON THE MERKEL NUMBER Equation (R I) shows that the Merkel number per unit height of fill is a functio n of the height of the fill. Figure R.8 graphically compares equation (R" I) and the measured values of the M erkel number for the different fill heights. Equation (3"15) must therefore be extended to include th e effect of the height of the fill on the Merkel number per unit height of the fill, 3.8 (3.18) 3.3.4 EFFECT OF INLET WATER TEMPERATURE ON TIlE MERKEL NUMBER Fill tests, for a 1.53 m high trickle fill, where the water and air flow rates a re varied, are conducted at different inlet water temperatures. These tests are presented in sections R.S an d R.6. Equation (R.6) express the Merkel number as a function of the inlet water temperature fur the c ombined data in sections R.S and R.6. It is therefore clear that equation (3.1 5) must be extended to inc lude the effect of the inlet water temperature as 10hnson [89101] did in equation (3.16). Including the effec t of the inlet water temperature, equation (3.18) can be extended to the general form (3.19) Sections R.7 and R.8 show the experimental results of 1.08 and 1.98 m high trick le fills respectively where only the water inlet temperature is varied during the testing periods. Fig ures R. I3 and R. I 7 shows the variation of the Merkel numbers, according to the Merkel, e-NTU and Poppe ap proaches, as the inlet water temperature varies, for fill heights of 1.08 and 1.98 m respectively. Figu re R. I7 is repeated here as figure 3.3. It can be seen from figure 3.3 that the Merkel numbers according to the different approaches are relatively strong functions of the water inlet temperature. The exponents of Twl in equations (R.7) and (R.8) for the 1.08 and 1.98 m fill respectively are ...{l.2471 and ...{l.2774. 0.85 0.80 0.75 "E j 0.70 ;; :;; 0.65 0.80 "'l~ ~ 00 ?? ?oQo?A,

~~ o~""'~o .~ -.,OO~~<>. xX> Oil !:be IJ~ ,d'""'___ 'O'oo'Q$' x~Yxx~ ",,0 .~~ ol:OOCb o "0"0 "'OS'~ ""~l\o<c Merkel 'Xx o Poppe ,,.. xo-NTU 0.55 25 30 40 50 55 Figure 3.3: Transfer coefficients according the e-NTU, Merkel and Poppe approach es where only TWi is varied. It is interesting to note from figure R.B and figure R.I 7 (figure 3.3) that the Merkel numbers, according to the e-NTU, Merkel and Poppe theories, decrease for increasing water inlet tem peratures. Thus as design loads of cooling towers increase above design specifications, the cooling tower will be less effective. r r ,.," , ~~.1\, :fi: :illIi '1!: k ji\i 3.9 3.3.5 EFFECT OF INLET AIR TEMPERATURE ON THE MERKEL NUMBER According to the Merkel theory, the Merkel number is not a function of the inlet air drybulb temperature. This is because the assumed linear increase in the air enthalpy is indeed linear as indicated by equation (2.2). The cooling tower perfonnance curves compiled by the Cooling Tower Instit ute [67CTl] and Kelly [76KE I], for counterflow and crossflow fills respectively, do not present the M erkel number as a function ofair temperature (refer to fignre 2.6). Roth [OIROI], however, found from experiment that the Merkel number appears to b e a function of the air inlet drybulb temperature decreasing for increasing air temperatures. Figure 3.4 shows the results of Roth [OIROI] where the Merkel number is a function of the inlet air drybulb temp erature. The water inlet temperatures in figure 3.4 are varied until a constant cooling range is obtained . Roth [OIROI] does not give values for the low medium and high air temperatures or states whether the a ir temperatures are approximately constant or not. It will be shown that the apparent dependence of the Merkel number on the inlet air drybulb temperature, according to Roth [OIROI], is actually the de pendence of the Merkel number on the inlet water temperature.

Constant cooling range: 1OK 21---+--+--+----.,;.'f-,.><....j,.L-''"--------------i 3 1L-~---L~____L~---L~..l-----.l. ~ __l 0.5 0.6 0.7 0.8 0.9 1 2 GalGw Fignre 3.4: Merkel number for a fill according to the Poppe theory (Adapted from Roth [OIROI] ). _ r The effects of the inlet drybulb and wetbulb temperatures, on the Merkel number or transfer coefficient, are investigated experimentally for a trickle fill. The summary of the results o f this investigation is presented in section R.ll. Fignre R.23 is repeated as fignre 3.5 and shows the M erkel numbers for three different inlet air drybulb temperatures versus the right hand side of equation (R.IO) where the air temperature is omitted. It is evident from fignre 3.4 that there is no significa nt temperature effect on the Merkel number. Refer to section R.II for a detailed discussion of the fact that the inlet air drybulb and wetbulb temperatures do not influence the empirical equation for the Merkel numb er significantly. The cooling range for all the data presented by Roth [OIROl] in figure 3.4 is co nstant. In order to achieve a constant cooling range, with variable air inlet temperatures, it is necessary to vary the water inlet 1.21.1 3.10 0.4 0.5 0.6 0.7 0.8 0.9 1.952087 G;;O.611513G~?664630T;;t210834 0.3 ? Tal =16.0?C I ?f--EI Tal =25.8 'C A ? iii f--'" Tal =33.3 'C "'4 ?Ll '~ I~" "'* ~~ I.~ --jJ'r' 'wr .., 1.2 1.1 0.2 0.2 0.9 "IS 0.8 . ... 0,7 ~ ~ 0.6 0.5 0.4 0.3 r"'wI ~ Lir~l perature The effect of the inlet water temperature on the Merkel n umber is therefore hidden in the ',J, tern . I't.li!~ analysis ofRoth [OlROI].

iJ! '1m iji 'I:i Figure 3.5: The measured Merkel number per unit length offill versus right hand side of equation (R.lO) where the temperature of the air is omitted. 3.4 CONCLUSION A new empirical relation is developed that correlates measured pressure loss coe fficients accurately for all types of fills under all types of operational conditions as it is based On f undamental principles that make provision for forces due to shear and drag. Other types of equations may co rrelate observed trends accurately, but they generally lack generality and are only applicable for limit ed ranges of water and air flow rates. Both the empirical relations for the loss and transfer coefficients per unit hei ght of fill of wet-cooling tower fills are extended to include the effect of the height of the fill. In add ition, the empirical relation for the Merkel number is extended to include the effect of the water inlet temperatu re. The inlet water temperature has no significant effect on the loss coefficient. It is also found that the inlet air drybulb and wetbulb temperatures have no significant effect on the loss or transfer coeffici ents. It is recommende# that as much information as possible be supplied with the empi rical relations of the loss coefficients, such as the ranges of applicability of Ga and Gw ? The goodne ss of fit must also be supplied in the form of a correlation coefficient. This will enable the designer of wet-cooling systems to take the necessary precautions to compensate for any uncertainties. If possible, the same water spray system must be 'employed in the fill test and the subsequent cooling tower appli cation of the fill. This will eliminate the effects of droplet size and distribution on the loss coefficient. Ageing effects of the fill are not investigated in this study. 4.1 CHAPTER 4 WET-COOLING TOWER PERFORMANCE EVALUATION 4.1 INTRODUCTION The performance of natural and mechanical draft counterflow cooling towers is cr itically evaluated by respectively employing the Merkel [25MB I], Poppe [9IPOl] and e-NTU [89JAI) meth ods of analysis at different operating and ambient conditions. The Wet-Cooling Tower Performance Ev aluation (WCTPE) software program, presented in appendix P, is employed in this investigation. Th e importance of using a particular method of analysis when evaluating the performance characteristics of a certain fill material and subsequently employing the same analytical approach to predict cooling tower performance, is investigated. By employing the different approaches at different ambient conditions for natura l and mechanical draft cooling towers, the resultant predicted performances are compared. The differenc

es in performance of natural draft and mechanical draft towers, for the same ambient conditions, are evaluated. The performance of the natural draft cooling tower, specified in appendix I, is eval uated in the WCTPE program in this investigation. Furthermore, the performance of the mechanical dr aft tower, presented in appendix J, is employed in this investigation, with the exception that the fill height, Lft, and the water mass velocoty, G., are the same as those of the natural draft tower. The fan spe ed of the mechanical draft tower is also adjusted so that the air mass velocity and cooling range is the sa me as that of the natural draft cooling tower, at the ambient conditions as specified in appendix I. Ambient air drybulb temperatures of 280, 290, 300 and 310 K are considered in th e analysis. At each of these temperatures, the ambient humidity is varied from dry to saturated conditi ons. The results of the cooling tower analyses are given in graphical form in appendix O. Most of the gr aphs in ligures 0.1 to 0.19 are presented in the same general form. Subfigures (al) to (a,) in each fig ure respectively illustrate the heat rejected, Q, the water outlet temperature, Two, the air outlet temperat ure, Tas . the mean air-water vapor mass flow rate, ""',11, and the water evaporation rate, mw("",p), at an am bient temperature of280 K at ground level, where the humidity of the inlet air is varied from dry to saturate d conditions. The subligures (bl) to (b,), (CI) to (c,) and (dl) to (d,) show the same variables as subfigllr es (al) to (a,), except that they are for ambient temperatures at ground level of 290, 300 and 310 K respectively. 4.2 NATURAL DRAFT COOLING TOWER The natural draft cooling tower specified in appendix I is taken as the referenc e tower. Figures O.I(at>? O.I(a,) respectively illustrate the heat rejected, Q, the water outlet temperatu re, Two. the air outlet temperature, Tas, the mean air-water vapor mass flow rate, maviS, and the mass f low rate of the water evaporated from the water stream, m.(,,,,,p~ as the inlet air is varied from dry to saturated conditions where the ambient temperature is equal to 280 K. The solid line in each of the figures represents the results according to the Merkel approach while the broken lines represent the results ac cording to the more rigorous poppe approach. The same method of analysis is?used for both the fill p erformance evaluation and the subsequent cooling tower analysis, i.e., for example, the fill performan ce characteristics, determined by the Poppe approach, are used in the cooling tower performance calc ulations, while employing the Poppe approach. r """ '",.:.l'} 'ft?, 1 , R:In, milI'"~ 4.2

4.2.1 HEATRElliCTED The heat rejected by the cooling tower at ambient temperatures of 280, 290, 300 and 310 K in dry to saturated conditions, can be seen in figures O.I(a,), O.I(bd, O.I(c,) and O.1(d, ). It can be seen that the heat rejection predicted by the Poppe approach is higher than that predicted by the Merkel approach at all the ambient conditions considered in this investigation. The Poppe approach pred icts heat rejection rates that are approximately 3% higher at 280 K and 4% higher at 290 K than the values predicted by the Merkel approach at all the inlet humidity conditions. At 300 K, in very dry cond itions, the difference between the Poppe and Merkel approaches is approximately 7% and 4% in the case o f saturated inlet conditions. The difference is 13% in very dry inlet conditions and 4% in the cas e of saturated inlet conditions, at an ambient temperature of 310 K. Thus, it is evident that the dif ference in heat rejection rates between the Merkel and Poppe approaches increases as the inlet air becomes dryer and hotter. 4.2.2 WATER OUTLET TEMPERATURE The heat rejected by the cooling tower at ambient temperatures of 280,290,300 an d 310 K in dry to saturated conditions, can be seen in figures O.1(a,), O.I(b,), O.I(c,) and O.I(d ,).It can be seen that the water outlet temperatures predicted by both the Merkel and Poppe approaches are practically identical. The water outlet temperatures determined by the two different approaches are pra ctically identical, because the same approach (i.e., Merkel or Poppe) is used in the fill performanc e analysis and the subsequent cooling tower performance analysis. Thus, if the same method is used in both the fill and cooling tower analysis, it will result in the same cooling range, if all other v ariables remain unchanged. In hot, dry conditions, there is however a discrepancy between the water outlet tem peratures predicted by the Merkel and Poppe approaches, i.e. the ~ater outlet temperature, predicted by the Poppe approach, is less than that predicted by the Merkel approach. This is because there is a discrepan cy between the air outlet temperatures predicted by the Merkel and Poppe approaches. 4.2.3 AIR OUTLET TEMPERATURE It can be seen in figures 0.I(a3), 0.1(b3), O.I(c,) and 0.I(d3) that the air out let temperatures predicted by the Poppe approach are higher than those predicted by the Merkel approach in all the ambient conditions considered. When the ambient temperature is low, the discrepancy between the pre dicted air outlet temperatures is the smallest. When the temperature of the ambient air increases, the discrepancy between the predicted air outlet temperatures increases in very dry conditions. When the humidity increases at a given temperature, the discrepancy decreases. 4.3 The air outlet temperature, according to the Merkel approach, can only be determ ined after the assumption that the air after the spray zone is saturated. The Poppe approach do es not make this

simplifying assumption and calculates the outlet humidity directly from the gove rning equations for heat and mass transfer presented in appendix B. The condition of the outlet air, dete rmined according to the Poppe approach, can therefore be unsaturated, saturated or supersaturated. For low ambient temperatures (280 K) at any ambient humidity, the outlet air is, according to the Poppe approach, always supersaturated. There is no discrepancy in the air outlet tempe rature trend, according to the Poppe approach, as the humidity increases at 280 K, compared to the trend of the Merkel approach, as seen in figure O.I(a,). However, for a very low ambient humidity at higher tempe ratures, there is a discrepancy between the values predicted by the Merkel and Poppe approaches. Thi s occurs because the outlet air, as predicted by the Poppe approach, is unsaturated. As the outlet ai r becomes saturated and supersaturated, the trend is the same as that predicted by the Merkel approach f or a given ambient humidity ratio. This point of saturation is approximately at W, ~ 0.003 kglkg in figure 0.1(1),) and WI ~ 0.022 kg/kg in figure O.1(c,). The reasons for the difference in the air outlet temperatures predicted by the Merkel and Poppe approaches are discussed in section 4.3.3 for the mechanica l draft cooling tower. Because the operating processes in mechanical draft towers are not as strongly c oupled as in natural draft towers, the differences in the air outlet temperatures between the two approache s can be explained without secondary influences ofother variables. Another interesting phenomena evident in figure O.I(d,) for dry conditions is th at the outlet air is colder than the inlet air. The nett enthalpy transfer is still from the water to the ai r as explained in section 2.2 with the aid of a psychrometric chart. Notwithstanding the fact that the air out let temperature is colder than the ambient temperature, there is still a draft through the tower. Draft th rough the natural draft tower is still possible, because the molar mass of vapor is less than that of air at t he same temperature. Thus, a potential for draft still exists because the density of the air-vapor mIxture in side the tower is less than that ofthe hotter less humid air on the outside of the tower. 4.2.4 MEAN AIR-WATER VAPOR MASS FLOW RATE The mean air-water vapor mass flow rates, determined by the Poppe approach, are higher than those predicted by the Merker approach at all the ambient conditions considered, as se en in figures 0.1(...), 0.1 (b4), O.I(C4) and O.I(d4). The mean air-water vapor mass flow rate is strong ly coupled to the air outlet temperature. This is because the density of the air inside the cooling tower is a function of the air temperature. The mass flow rate of air through the tower is, in turn, a function of the density differential of the air internal and external to the cooling tower. Thus, the draft through t he natural draft cooling tower is strongly coupled to the air outlet temperature. The draft, in turn, will infl uence the heat rejection rate in the cooling tower. It is clear that the processes in anatural draft cooling towe r are strongly coupled.

!fthe outlet air is unsaturated, according to the Poppe approach, the mass flow rates are much higher than the mass flow rates predicted by the Merkel approach, than when the air is super saturated, according to 4.4 the poppe approach. The mass flow rate according to the Poppe approach is higher than that predicted by the Merkel approach, because the air outlet temperature, predicted by the Poppe approach, is higher than that predicted by the Merkel approach. At temperatures of 280, 290 and 300 K (see figures 0.1(14), O.I(b,) and O.I(c,)) the air-vapor mass flow rates increase as the inlet ambient humidity ratio is increased. It is very inte resting to note that this is not the case at an amhient temperature of 310 K. Both the Merkel and Poppe approache s predict this interesting phenomenon. At amhient temperatures of 290 K (figure O.I(h,)) and 30 0 K (figure O.I(c,)) the predicted mass flow rates are decreasing for increasing ambient humidity rat ios, according to the Poppe approach, when the air is unsaturated. Figures O.I(e,), 0.1(14) and 0.1(84 ) illustrate the mass air? vapor mass flow rate at air inlet temperatures 0005, 307.5 and 308.75 K respecti vely. It can be seen that there is a gradual decrease of the slope of the mass flow rate, predicted by Mer kel, as the ambient temperature is increased. 4.2.5 WATER EVAPORATION RATE The predicted water evaporation rates in natural draft cooling towers are always higher according to the Poppe approach than according to the Merkel approach. This is the case even if t he outlet air is unsaturated, according to the Poppe approach. The air can be unsaturated, accord ing to the Poppe approach, but the predicted evaporation rate is still higher than that predicted by the Merkel approach where the outlet air is saturated, because of the strongly coupled draft and ene rgy equations. The outlet air temperatures predicted by the Poppe approach are higher than (hose predicted by the Merkel approach. The hotter the air, the higher the draft. The higher the draft, the more heat an d mass transfer and thus higher evaporation rates. 4.3 MECHANICAL DRAFT COOLING TOWER The mechanical draft tower employed in this section' has the same fill depth as the natural draft tower employed in the previous section. It also has exactly the same water and air mas s flow rates per unit area as the natural draft tower at the rererence conditions. At the reference point, the heat rejected and evaporation rates per unit area will exactly be the same in the reference mechan ical and reference natural draft cooling towers. The ambient conditions ofthe natural draft cooling tower i n the previous section are repeated here with the analysis of the mechanical draft tower. Figure 0.2 shows the variation of the heat rejected, water and air outlet temperatures, air-water vapor mass flow rates and evaporation rates for different ambient temperatures and humidities. 4.3.1 HEAT REJECTED

The trends of the heat rejected at the different ambient conditions are the same as those of the natural draft tower discussed in the previous section. The average difference between th e heat rejection rates, predicted by the Merkel and Poppe approaches, is approximately 2.8% in the case where Tal = 280 K (figure 0.2(al)). The average difference is approximately 3.2%, 3.7% and 4.6% at ambient temperatures 4.5 of 290, 300 and 310 K respectively. Thus, at higher ambient temperatures, the di fferences between the Merkel and Poppe approaches are the greatest. The percentages given for the mechanical draft tower, correspond approximately t o those obtained for the natural draft cooling tower, where T,] = 280 K and T,t ~ 290 K. At higher temper atures, in dry conditions, the discrepancy between the Poppe and Merkel approaches is higher fo r the natural draft tower than that for the mechanical draft tower, because the governing equations are strongly coupled for the natural draft towers, which is not the case for mechanical draft towers. 4.3.2 WATER OUTLET TEMPERATURE The water outlet temperatures, predicted by the Merkel and Poppe approaches, are practically identical at all the different ambient temperatures and humidities considered. This is not th e case with natural draft cooling towers in very dry, relatively warm ambient conditions. This is again be cause of the strongly coupled energy and draft equations for natural draft towers. The energy and draf t equations for mechanical draft towers are not as strongly coupled, and therefore are there ess entially no discrepancies between the Merkel and Poppe approaches. 4.3.3 AIR OUTLET TEMPERATURE It can be seen from figures 0.2(a,), 0.2(b,), 0.2(c,) and 0.2(d,) that the air o utlet temperatures, for the mechanical draft tower, follow the same trend as those ofthe natural draft tower , as discussed in section 4.2.3, for both the Merkel and Poppe approaches. The differences between these t wo models are explained by the discussion that follows. Merkel assumes that the air above the transfer areas is saturated with water vap or. The enthalpy of the air at this point, according to the Merkel approach, is known from a simple energy b alance with the cooling water stream, where the water loss, due to. evaporation, is neglected. By assumi ng that the air is saturated, with the known air enthalpy, the temperature of the air can be determined. The a ssumption made by Merkel that the air is saturated leads to greater errors when the air is unsatur ated, according to the Poppe approach, than when it is supersaturated. The assumption by made by Merkel that the loss in the water mass flow rate, due to evaporation, could be neglected in the energy balance pla ys a secondary role, especially during hot and dry ambient conditions. A procedure to minimize the er ror introduced by neglecting the loss of water due to evaporation in the energy balance is discuss ed in section 4.8. \

Refer to figure 4.1 which is the same as figure O.2(a,). As already mentioned in section 4.2.3, the Merkel and Poppe approaches predict air outlet temperatures that follow the same trends for variable ambient humidities at low ambient temperatures. At higher ambient temperatures, as shown in figure 4.2, which is the same as fig ure O.2(d,), the respective trends predicted by the Poppe and Merkel approaches are not the same. The outlet air, according to the Poppe approach, in figure 4.1 is supersaturated across the whol e range of inlet ambient 'r'''' _..r:? .' t'::' , 4.6 humidities. The outlet air, according to the Poppe approach, in figure 4.2 is, h owever, unsaturated across the whole range of inlet amhient humidities. It is therefore clear that the disc repancy between the Merkel and poppe approaches is the greatest when the outlet air according to Poppe is u nsaturated. If the outlet air is supersaturated, according to the Poppe approach, the discrepency is consi derably smaller. Mechanical draft -, ./ T a l=280K " ",- V , / -Merkel - / ,,"~--Poppe " ", / , ,,/," V V,'--, , V ",X 298.5 298.0 2'J7.5 ~ 297.0 Vi ? ..... 296.5 296.0 295.5 295.0 0.000 0.001 0.002 0.003 0.004 w., kglkg 0.005 0.008 0.007 0.008 Figure 4.1: Air outlet temperature of a mechanical draft cooling tower for a low air inlet temperature. 0.0500.0400.0300.0200.010 -/-----

----- V ---- /-------,,- _.... V -,,/-/oj"' Mechanical draft ./ I--V T a l=31OK-Merkel V -- Poppe 312 310 302 300 0.000 308 304 WI' kg/kg Figure 4.2: Air outlet temperature ofa mechanical draft cooling tower for a high air inlet temperature. 4.7 Figure 4.3 shows the heating path of the air in the cooling tower for cold inlet air, which is saturated with water vapor. These inlet conditions are indicated with an 'X' in figure 4.1. Sin ce the inlet air is saturated with water vapor, indicated by point I in figure 4.3, it immediately becomes sup ersaturated, according to the Poppe approach, as it enters the fill. As the air is heated and the humidity ratio increases, due to the latent heat transfer from the water, it follows the saturation curve very closel y. This is because as the air is heated, it can contain more water vapor before it reaches the point of saturatio n. Point 2b in figure 4.3 shows the state of the air at the outlet of the fill, according to the Poppe app roach. Point 2a in figure 4.3 shows the outlet air state according to the Merkel approach. It can be seen that the air is saturated at the outlet according to Merkel. The outlet air temperatures according to the Merkel and Poppe approaches are relatively close to each other in figure 4.3. The same trends are therefore pred icted by the two approaches as shown in figure 4.1 when the outlet air is supersaturated according to the Po ppe approach. The assumption of Merkel that the outlet air is saturated, is therefore a very good assumption, if the actual outlet air temperature is supersaturated. Enthalpy, kJ/kg dry air 30 40 50 60 70 80 90 100 r;::=:r:::::::r::::::r:=:rr:::=:::c:;n-TlY5IT77""""";itj:::-::::;...---,- 0.030 Atmospheric pressure 84100 Pa 5 1 10 15 20 25 30 35 0.015 ~ IS,.. E 0.010 a-

0.000 40 Drybulb temperature, Ta, ?C ,-' Figure 4.3: Psychrometric chart of cooling process for cold saturated ambient ai r. The degree of supersaturation does not have a great influence on the relative di fference between the outlet air temperatures predicted by the Merkel and Poppe approaches. This is because t he lines of constant air enthalpy, in the supersaturated region, are very close to vertical as seen in fi gure 4.3. It therefore does not matter how much water vapor and mist are present in the supersaturated air, for a specific air enthalpy, the air temperature will be approximately constant. The difference in the air te mperatures at point 2a and 2b in figure 4.3, for the Merkel and Poppe methods respectively, can be reduced by improving the energy balance employed by the Merkel approach where the approximate loss of water, due to evaporation, in the 4.8 energy balance is neglected. Refer to section 4.8 where the loss of water, due t o evaporation, is accounted for in the energy halance. Figure 4.4 shows the heating path ofthe air in the cooling tower for hot inlet a ir, which is virtually void of water vapor. These inlet conditions are indicated with a 'X' in figure 4.2. Poin t I in figure 4.4 shows the state of the inlet air on a psychrometric chart. Point 2b in figure 4.4 shows th e state of the air at the outlet of the fiJI, according to the Poppe approach. It can be seen that the outlet air is colder than the inlet air. This scenario is described in chapter 2 with the aid of figure 2.3. Point 2a sho ws the outlet air state according to the Merkel approach. 0.000 40353025201510 Atmospheric pressure 84100 Pa _ poppel-+__f-I-__I---+-.." - -Merkel L=:===1I=l=+==r+==l~-+--l-l,J(.A-A~~~-=~~0.025 ~ .. ?:' ""-7'q~~~~L-h"L:J=:~~d 0.020 ~ ~ ~:;:~=1;~~j;~::::.?-10.015 ~cj., ~:b::SF:::;;~~f~7Rl 0.005 :i 5 Enthalpy, kJ/kg dry air 30 40 50 60 70 80 90 100 Drybulb temperature, Ta, DC Figure 4.4: Psychrometric chart ofcooling process for hot and very dry ambient a ir. It can be seen in figure 4.4 that the outlet air is saturated according to the M erkel approach. The outlet air temperatures according to the Merkel and Poppe approaches are not very close to each other. The outlet air temperatures predicted by the Merkel and Poppe approaches lie approximately on the same constant enthalpy line in figure 4.4, as was the case in figure 4.3 when the outlet air w

as supersaturated according to the Poppe approach. In the unsaturated region, however, the lines of constant enthalpy are far for vertical and therefore the large discrepancy in the temperatures. The assumption of Merkel that the outlet air is saturated with water vapor, is not as accurate if the actual outlet air i s unsaturated as when it is supersaturated. 4.3.4 MEAN AIR-WATER VAPOR MASS FLOW RATE I I III r 4.9 ~ Figures 0.2('4), 0.2(b.), 0.2(c.) and 0.2(d.) show the air-vapor mass flow rat es predicted, by the Merkel I and Poppe methods, at ambient temperatures of 280, 290, 300 and 310 K respecti vely. At each ambient fI temperature, the differences between the Poppe and Merkel approaches are prac tically negligible. I) I';i However, it is still informative to discuss the respective trends. Ifthe trends in figure 0.2, ofthe air-water vapor mass flow rate, are compared to the natural draft cooling tower in figure 0.1, it can be seen that the trends are the inverse ofeach other. The Poppe approach predicts a smaller air-w ater vapor mass flow rate through the mechanical draft tower than the Merkel approach. For the natural dra ft cooling tower it is the opposite. If the inlet air humidity is increased for a given ambient temperature , then the mass flow rate through the mechanical drafl tower decreases. Again, the opposite is true in nat ural drafl cooling towers. The reason for this is that a lower air density at the outlet of the fill will i ncrease the draft in natural draft towers, due to the increased pressure differential between the inside and outsid e of the tower. A lower density of the air, at the outlet of the fill of the mechanical draft tower, mea ns that less air passes through the fan and hence the lower mass flow rate at higher air temperatures. 4.3.5 WATER EVAPORATION RATE The Poppe approach always predicts higher mass flow rates than the Merkel approa ch in natural draft towers. For mechanical draft towers it is not always the case. In hot, dry condi tions (see figures O.2(c,) and 0.2(d,? the outlet air can be unsaturated, according to the Poppe approach a nd hence the lower evaporation rate than in the Merkel approach where the outlet air is always satu rated. 4.4 INVESTIGATION INTO THE DRAFT EQUATION OF NATURAL DRAFT COOLING TOWERS A detailed draft equation is employed in the performance evaluation of the natur al draft cooling tower presented in appendix 1. The detailed draft equation is also employed in the ana lysis of the natural drafl tower discussed in the section 4.2. The detailed draft equation accounts for the moist air that is raised in a gravitational field, adiabatic cooling and condensation. In this section the inf luence that the detailed drafl equation has on the perfurmance evaluation of a natural draft cooling tower is i

nvestigated and compared to a simplified equation that is commonly employed by other researchers [83B01, 96M02]. In its simplest form, the draft eqnation of a cooling tower can be expressed by !!.po -!!.p, =LJ( pv12' (4.1) , I Iii where Apo is the pressure differential outside the tower and !!.p, is the pressu re differential inside the tower while the ftow resistances are represented by LK?0.5pv'. The right-hand side of equation (4.1) is the same for both the detailed and simplified drafl equations. The difference between the detailed and simplified draft equations is thus on the left-hand side of equation (4.1). The simplified draft equation is given by <A'I-A,,) g (H, - H, -Lfil2) = LK pv'/2 (4.2) Figure 0.3 illustrates the heat rejected, water outlet temperature, air outlet t emperature, air-water vapor mass flowlate and the evaporation rate for temperatures of 280, 290, 300 and 310 K respectively. !fall il i 4.10 the graphs in figure 0.3 are compared to the corresponding graphs in figure 0.1 it can be seen that the graphs are practically identical. Thus, for this specific cooling tower with the specified ambient conditions the simplified draft equation gives accurate results. Notwithstanding this fact, the detailed draft equation is still employed in all the investigations that follow. Figures O.4(a) to 0.4(b) show the pressure differentials of the detailed and sim plified draft equations in the Merkel and Poppe approaches at ambient temperatures of 280 and 290 K respect ively. It can be seen from figure O.4(a) that the difference in the pressure differential between the detailed and simplified draft equations, in both the Merkel and Poppe approaches, is approximately only 3 Pa a cross the entire range of ambient humidities considered. The difference in the pressure differential betwe en the detailed and simplified draft equations decreases to approximately 2 Pa at an ambient tempera ture of 290 K, as can be seen in figure 0.4(b). At 310 K the difference is less than 1 Pa as can be seen in figure 0.4(d). Thus, the higher the ambient temperature, the smaller the difference in the pressure diffe rential between the detailed and simplified draft equations for these particular ambient conditions. 4.5 CONSISTENT APPLICATION OF COOLING TOWER PERFORMANCE EVALUATION ANALYSES It can be seen in figures 0.1 and 0.2 for the natural draft and mechanical draft towers respectively that the water outlet temperatures, predicted by the Merkel and Poppe approaches, are practically identical. It is concluded that the water temperatures are practically identical, because the same method of analysis is used for the evaluation of the performance characteristics of the fill and then subsequently used for the prediction of the cooling tower performance. This is denoted as the consistent a pplication of a particular

approach. It will be investigated what the influence on the accuracy will be if the approaches are applied inconsistently. A common error by cooling tower designers is to apply the approa ches inconsistently. Empirical relations for fill performance characteristics, derived by employing t he Merkel approach, are commonly available. These fill performance characteristics are then employed inc orrectly in cooling tower performance calculations, while employing, for example, the more rigorous Poppe approach. Figures 0.5 and 0.6 show the performance graphs of the natural draft and mechani cal draft cooling towers respectively for the range of ambient conditions similar to those in figu res 0.1 and 0.2. The fill performance characteristics obtained by the Merkel approach are applied inconsis tently to the cooling tower performance evaluation while employing the Poppe approach. The results for the Merkel approach are not included in the graphs. The Poppe approach with consistent application o f the fill performance characteristics is compared to the Poppe approach with inconsistent application of the fill performance analysis. The legend denoted 'Merkel' in each of the graphs in figures 0.5 and 6 refers to the inconsistent application of the fill performance characteristics in the cooling tower perform ance evaluation. In this inconsistent application, the filk performance characteristics are determined by employing the Merkel , approach while the performance of the cooling tower is evaluated by the Poppe ap proach. The legend, denoted 'Poppe', refers to the consistent application and is identical to the co rresponding plots in figure 0.1. 4.11 4.5.1 HEAT REJECTED Although it seems from figures 0.5(a,), 0.5(b,), 0.5(c,) and 0.5(d,) for natural draft towers and from figures 0.6(a,), 0.6(b,), 0.6(c,) and 0.6(d,) for mechanical draft towers, that the discrepancy between the consistent and inconsistent analysis, decreases with increasing ambient temp eratures, it is actually not always the case. The discrepancy between the consistent and inconsistent analyse s of the heat rejection rate is approximately 2-2.5% at all the ambient temperatures and humidities. The heat rejection rates fur the consistent analysis of the Poppe approach are approximately 2-2.5% higher th an the inconsistent analysis, where the fill characteristics are obtained by employing the Merkel ap proach. 4.5.2 WATER OUTLET TEMPERATURE Figures 0.5(a,), O.5(b,), 0.5(c,) and O.5(d,) for natural draft towers and figur es 0.6(a,), 0.6(b,), 0.6(c,) and 0.6(d,) for mechanical draft towers present the water outlet temperatures of the consistent and inconsistent application of the Poppe approach to cooling tower performance. The discrepancy between the consistent and inconsistent analyses is approximately 0.4 K for both natural draft and mechanical draft cooling towers. For the mechanical draft tower, at an ambient temperature of 280

K, the discrepancy is less at approximately 0.15 K. 4.5.3 WATER EVAPORATION RATE The discrepancy in the water evaporation rate between the consistent and inconsi stent analyses of both natural and mechanical draft cooling towers is approximately 2.5% in all the amb ient conditions considered. 4.6 LEWIS FACTOR A detailed history and description of the Lewis factor are presented in appendix F. It is assumed in the Merkel approach that the Lewis factor is equal to unity. The Lewis factor, howev er, is specified explicitly in the Poppe approach. The equation of Bosnjakovic [65BOl) is employed in the Po ppe approach to specify the Lewis factor in the investigation in sections 4.2 to 4.5. The value of the Lewis factor, calculated by the equation of Bosnjakovic [65BOI), is approximately 0.92. HAszle r [99HAI) fuund that the Lewis factor could vary from 0.5 to 1.3, depending on the state of the air i n the boundary layer of the interface between the air and the water. Three different specifications of the L ewis factor are employed in this section to determine the effect of the Lewis factor on the results of the P oppe approach. The equation ofBosnjakovic [65BOI) is employed as well as arbitrarily chosen Lewis factors of 0.5 and 1.3. The various Lewis factors are applied consistently to the evaluation ofthe fill perfurmance characteristic and the subsequent cooling tower performance evaluation, i.e. the same definitio n of the Lewis factor is employed in the fill performance analysis and the subsequent cooling tower perfo rmance analysis. The results achieved by the consistent application of the Lewis factor in the natura l draft tower are shown in figure 0.7 and those achieved in the mechanical draft tower are shown in figure 0.8. The trends of the variables in the graphs in figure 0.7 and figure 0.8, of the natural draft and m echanical draft towers 4.12 respectively, are approximately the same, except for the air-vapor mass flow rat e as discussed in section 4.3.4. 4.6.1 HEAT REJECnON RATE The heat rejection rates for the different specifications of the Lewis factor ca n be seen in figures 0.7(a,), O.7(bJl, 0.7(c,) and 0.7(d]) for natural draft towers and in figures 0.8(aJl, 0. 8(b]), 0.8(c,) and 0.8(d]) of mechanical draft towers. The higber the Lewis factor, the more heat is reject ed. In the natural draft cooling tower at an ambient temperature of 280 K the differences in heat rejecti on rates, between the analyses of Lewis factors of 0.5 and 1.3, are approximately 2.4%. The difference is 0.8% at 290 K and approximately zero at 300 K. At 310 K in very dry conditions, the difference is almost 5% where the heat rejected, due to the smaller Lewis factor, is more than that predicted by the hi gber Lewis factor. The mechanical draft tower follows approximately the same trends. 4.6.2 WATER OUTLET TEMPERATURE

Because more heat is rejected at higber Lewis factors, the corresponding water o utlet temperature is lower. In the natural draft cooling tower, the discrepancy between the water out let temperatures, by applying Lewis factors of 0.5 and 1.3 respectively, is approximately 0.65 K at a n ambient temperature of 280 K. This discrepancy is practically zero at 300 K. At 310 K, however, in very dry conditions, the discrepancy is 0.6 K. The mechanical draft tower follows approximately the same trend. 4.6.3 WATER EVAPORAnON RATE The water evaporation rate is higher when applying smaller Lewis factors than wi th higher ones. Thus, the air becomes saturated more quickly with lower Lewis factors. The discrepancy between the water evaporation rates in natural draft cooling towers with Lewis factors of 0.5 and 1.3, is approximately 15% at 280 K and reduces to 6% at 310 K. The mechanical draft tower follows approxim ately the same trend. 4.6.4 DISCUSSION ON THE CONSISTENT APPLICATION OF THE LEWIS FACTOR The Lewis factor has little influence on the water outlet temperature and the he at rejected from the cooling tower in very humid ambient air. In dry conditions, at all ambient tempe ratures considered, the differences between the results of the different Lewis factors can be quite sign ificant. The rate of water evaporation is strongly dependent on the Lewis factor for both the natural draft and mechanical draft towers. This is because the Lewis factor is an indication ofthe relative rates o fheat and mass transfer in an evaporative process. The Lewis factor can therefore be tuned to represent the physically measured evaporation rates and outlet air temperatures more closely in fill performance a nalyses. It is therefore important to perform the fill performance tests in conditions that closely repre sent actual operational conditions, especially if the cooling tower is operated at a very low ambient hu midity. If the fill performance test data is insufficient to accurately predicts the Lew is factor of a particular fill, it is recommended that the equation of Bosnjakovic be used as the numerical value i s approximately 0.92, which is approximately the mean between the limiting values of0.5 and 1.5 given by HAszler [99HA I]. ! I 1 4.13 4.6.5 THE INCONSISTENT APPLICATION OF THE LEWIS FACTOR The analyses of the natural and mechanical draft cooling towers are repeated wit h an inconsistent application of the Lewis factor specification. The equation of Bosnjakovic is us ed in the fill performance evaluation, while Lewis factors of 0.5 and 1.3 are used in the cooling tower per formance evaluation. Figure 0.9 and figure 0.10 show the results of the natural draft and mechanical draft towers respectively for the inconsistent application ofthe Lewis number. If figures 0.9 and 0.10 are compared to figures 0.7 and 0.8, it can be seen that the inconsistent

application of the Lewis factor results in larger discrepancies than is the case with the consistent application of the Lewis factor. The discrepancy between the heat rejection rate of both natural and mechanical draft cooling towers, is approximately 8% at an ambient temperature o f 280 K. The discrepancy is only 2.4% where the Lewis factors are applied consistently. The d iscrepancy reduces at higher ambient temperatures to approximately 2% at 310 K. This is consistent wit h the conclusion reached previously, that the influence of the Lewis factor diminishes at higher ambient temperatures. The discrepancy in the water outlet temperature for the natural draft cooling tower, for the inconsistent analysis of the Lewis factor, is larger than the consistent application. Ironica lly, the discrepancy between predicted water evaporation rates is smaller during the inconsistent application of the Lewis factor than during the consistent application. 4.7 ATMOSPHERIC PRESSURE An atmospheric pressure of 84200 Pa has been used in all the cooling tower analy ses discussed so far. The atmospheric pressure is increased to 101325 Pa to see what the effect on coo ling tower performance is in both the natural and mechanical cooling towers. Figures 0.11 and 0.12 show the cooling tower performance curves in the natural draft and mechanical draft towers respectively . The Merkel and Poppe approaches are employed in the cooling tower performance evaluation. If figures 0.11 and 0.12 are compared to figures 0.1 and 0.2 respectively for na tural and mechanical draft towers, where an atmospheric pressure of 84200 Pa is specified, it can be seen that the trends are practically identical for all the plotted variables. It is difficult to give exact quantification of the differences between the resu lts at low and high atmospheric pressures. The reason is that the ambient humidity and the draft thr ough the respective towers are functions of the atmospheric pressure. The processes in a cooling tow er are strongly coupled and it is therefore difficult to make accurate predictions. However, if all othe r variables remain unchanged, the cooling towers at higher altitude, and thus lower atmospheric pre ssure, will perform better than those towers at lower altitudes. 4.8 IMPROVED MERKEL ENERGY EQUATION 'r"'"/"'", , ~ t,; 4.14 The poppe approach predicts higher heat rejection rates than the Merkel approach . This is hecause the Merkel approach ignores the loss in the water mass flow rate, due to evaporation , in the energy equation. In a cooling tower analysis, employing the Merkel approach, the heat transfer ra te is generally given by Q~ m.Cpwm (T.,- Two) (4.3) The effect of the change in water mass flow rate is not included in the energy b alance in equation (4.3). If

it is assumed that the air is saturated at the outlet of the fill, then the mass flow rate of the evaporated water can be approximated by the equation, (4.4) A new improved equation ofthe heat rejection rate, according to the Merkel appro ach, is proposed where the water loss, due to evaporation, is included in the energy equation, i.e., (4.5) I~. Figures 0.13 and 0.14 show the results of the cooling tower performance llIlalys es for the natural draft and mechanical draft towers respectively, where the improved Merkel equation is employed in the analyses. It can be seen that the results of the improved Merkel approach and th e Poppe approach are practically identical for both the natural and mechanical draft cooling towers, At hot, dry conditions there is a discrepancy between the improved Merkel and Poppe approaches. The reason fo r this discrepancy is discussed in detail in section 4.3.3. In a nutshell it is because of the Merkel assumption that the air is saturated at the outlet when the Poppe approach predicts that the air is unsatur ated. At low ambient temperatures, the outlet air is generally supersaturated and the assumption of M erkel is relatively accurate. Thus, the improved energy equation of the Merkel approach can be employed in the relatively simple Merkel approach to predict heat rejection rates and water outlet temperatures th at are within very close tolerance with the predictions by the more rigorous Poppe approach. This is espe cially the case, as mentioned earlier, when the ambient air is cold or relatively humid. 4.9 e-NTU APPROACH The e-NTU approach is employed in both the natural draft and mechanical draft co oling towers. The results of the e-NTU approach are compared to those of the Merkel approach. They are compared to the Merkel approach because the governing equations of both approaches are derived w hile making the same simplifying assumptions, i.e. the Lewis factor is equal to unity and the water e vaporation rate is omitted from the energy balance. The fill performance characteristics are applied consistently in the cooling tow er performance evaluation for the comparison between the Merkel and e-NTU approaches. It is found that the transfer characteristic, obtained by the e-NTU approach, is approximately I% lower than that derived by e mploying the Merkel approach, I, I'; 1"; I" I'i' I f I "r i

I I 4.15 Figures 0.15 and 0.16, of natural draft and mechanical draft towers respectively , show the comparison between the Merkel and e-NTU approaches. The heat rejected, water and air outlet temperatures and water evaporation rates, predicted by both approaches, are practically identical across the entire range of ambient temperatures and humidities. Figure 0.17 shows the results of the natural draft tower where the fill performa nce characteristic, obtained by the Merkel approach, is applied inconsistently to the cooling tower performance analysis while employing the e-NTU approach. It can be seen that the results are practica lly identical to those in figure 0.15, where the fill performance characteristic is applied consistently. The fill perfurmance characteristics, obtained by the Merkel and e-NTU approaches, can therefore be a pplied inconsistently in cooling tower perfurmance evaluations employing either approach. This is because the transfer characteristic of the e-NTU approach is, as mentioned above, only I% less than t he transfer characteristic obtained while employing the Merkel approach. 4.10 CONSTANT HEAT REJECTION The water inlet temperature is 313.15 K in the investigation in sections 4.2 to 4.9. However, at power stations, the heat to be rejected is known and the inlet water temperature is no t known beforehand. This is because the condenser will have to absorb the heat load it receives from the tur bine exit stream. Figures 0.18 and 0.19 show the results of the natural draft and mechanical draft towers respectively where the heat rejection rate is known and the water outlet temperature is unkno wn. The plotted trends are approximately the same for both natural draft and mechanical draft cooling t owers. The water inlet temperatures, for the natural draft cooling tower, calculated by both approaches , are practically identical' at low ambient temperatures. At higher temperalures, in dry conditions, the disc repancies are greater. For mechanical draft towers, the water inlet and outlet temperatures are practically identical for both approaches at all ambient conditions considered. 4.11 CONCLUSION It is very important that the same model, definitions and assumptions be employe d in the fill perfurmance analysis to determine the transfer coefficient and in the subsequent analysis to determine cooling tower performance. This will ensure that the water outlet temperature in cooling tower perfurmance analyses, predicted by the different models, are practically the same when all other varia bles are assumed constant. The predicted water evaporation rates in natural draft cooling towers are always higher according to the Poppe approach than according to the Merkel approach. This is not the case for m echanical draft cooling towers at very hot and dry ambient conditions. The performance prediction of natural draft cooling towers is not as strongly in

fluenced by the accuracy of the left-hand side of the draft equation given by equation (4.1). If the left hand side of the draft 4.16 equation accounts for the moist air that is raised in a gravitational field, adi abatic cooling and heating by condensation, the same order of results are obtained when the simplified equatio n, given by equation (4.2) is employed. When the Poppe approach is employed during hot ambient conditions, the value of the Lewis factor has little influence on the prediction of the water outlet temperature, if it is emp loyed consistently or inconsistently. The heat transfer rate, water outlet temperature, draft, air outlet temperature and evaporation rate of the Merkel approach can be brought within closer tolerances of the more rigorous Pop pe approach, when the reduction of the water mass flow rate, due to evaporation, is included in the en ergy balance. The assumption ofMerkel that the outlet air is saturated with water vapor, leads to tower performance that are within close tolerance of the tower performance predicted by the Poppe approach, for cold or humid ambient conditions. The e-NTU and Merkel approaches predict virtually the same tower perfurmances wh en the models are applied consistently or inconsistently. (S.I) 5.1 CHAPTERS THE INFLUENCE OF TEMPERATURE AND HUMIDITY INVERSIONS ON WET-COOLING TOWER PERFORMANCE 5.1 INTRODUCTION The influence of temperature and humidity inversions on the performance of wet-c ooling towers is investigated. Hofman [97H0I] investigated the effect of temperature stratificati on in the atmospheric boundary layer on the performance of natural draft dry-cooling towers. Hoffinan [97HOI] followed a predominantly numerical approach in addressing the problem. A semi-empirical app roach is followed in the current analysis with the emphasis on simplicity. Pure theoretical approache s that predict temperature profiles during nocturnal inversions are generally impractical due to the vagari es of nature and the complexity ofthe models. Relatively simple equations are developed to predict th e vertical temperature profile or distribution and the height of the inversion throughout the course of the year. The diurnal and annual variations of atmospheric humidity are investigated. The buoyancy force that drives the air through a natural draft tower is negative ly affected when temperature inversions occur. Cooling tower designs are generally based on the a mbient air drybulb temperature, measured at, or near, the ground. The average temperature of the ai r at the inlet of the tower may deviate significantly from the measured air temperature near the ground due to temperature inversions. The effective inlet humidity ratio of the air may also deviate due t

o the presence of humidity inversions that occur during the night. The deviation of the effective inlet air temperature and humidity ratios occur because the tower draws in air from high above the ground. A simple equation is recommended to determine this height for a particular tower. 5.2 TEMPERATURE AND HUMIDITY INVERSION PROFILES Appendix L presents the development of an empirical approach to extrapolate temp erature profiles during nocturnal inversions from ground based measurements. Equation (L.S), repeated he re as equation (S.1), gives the temperature inversion profile after the first few hours after the ince ption of an inversion. T=(T, +273.1S(:')' where T, and z, are the reference temperature and reference height respectively. T, (0C) is measured at z, which is about 1 m above ground level. The value of the exponent, b, varies thro ughout the course of the year and is given by equation (L.9), repeated here as equation (S.2). b =0.003Ssin(0.0177nb - 2.32392) +0.006S (S.2) where nb is the number of the day of the year (I January is the first day of the year). Equation (S.2) is only valid for the specific geographical location, as mentioned in appendix L. 5.2 The height of the inversion is of importance in a cooling tower performance anal ysis, especially when the cooling tower draws in air from above this height. The height of an inversion, d uring a specific inversion period, is given by equation (L.26), repeated here as equation (5.3). 1 [ 0.00975 ]b-l (5.3) zn = b{T, + 273.15) I Equation (5.3) will predict, in conjunction with equation (5.2), inversion hei ghts of approximately 300 m ! in the winter months and approximately 90 m in high summer at the specific geo graphical location, as mentioned in appendix L. Equation (5.3) is only valid after the first few hours of the inception of an inversion. Equation (5.1) requires less data than that required by equations (L.1), (L.2) a nd (LA), which are found in the literature. Appendix M describes the complexity of determining the vertical vapor profiles i n the atmosphere, especially in the atmospheric boundary layer. Empirical relations are given in t he literature that predict atmospheric humidity profiles, but these equations are not accurate in the atmos pheric boundary layer during humidity inversions. Huniidity profiles are very unpredictable and depend on the vegetation and meteorological conditions. Examples of nocturnal humidity inversion profiles are given in figures M.2 andM.3. 5.3 HEIGHT FROM WHICH AIR IS DRAWN INTO A COOLING TOWER The height from which air is drawn into a natural draft cooling tower, H" is inv estigated analytically in appendix N for windless conditions. Refer to figure N.2 for a graphical descript ion of H,. It is shown in appendix N that H, is only a function of the diameter of the cooling tower and i

s constant some distance away from the cooling tower in the radial direction. For the cooling tower prese nted in appendix I, H, is approximately 127 m, according to the results in appendix N. Wilber et al. [85WI l) state that H, is generally between 50 and 100 m while Lauraine et al. [88LAI) estimate H, to be b etween 50 and 150 m. The effect of H, on cooling tower performance is investigated in appendix V.6. I t is shown that tower performance is relatively insensitive to the choice ofH,. It is therefore recomm ended that H, be arbitrarily taken as half the height of the cooling tower shell, (SA) where H6 is the height of the tower shell, as shown in figure I.l. 5.4 EFFECT OF TEMPERATURE AND HUMIDITY INVERSIONS ON TOWER DRAFT AND INLET CONDITIONS Appendix V contains a sample calculation that calculates the effective inlet air temperature and humidity ratio during prescribed ambient temperature and humidity inversions. Since the h eight of the temperature inverSion, Zn, throughout the course of the year, as determined by equation (5.3 ) or (L.26), is generally higher than H" it has no effect on the effective inlet air temperature and humid ity ratio. 5.3 Figure V.5 shows the heat rejected hy a particular cooling tower as a function o f the exponent, b, of equation (5.1) or (L.5). The particular tower heat rejection is reduced by appro ximately 20 % when the magnitude of temperature inversions is strongest during winter. Tower performanc e is 8 % down in summer when temperature inversions are generally not as strong. The reduction in performance due to the temperature inversion is very high when compared to the reduction in performance due to the humidity inversion. The effect ofthe humidity inversion on tower perfurmance is generally very small (1.5 %). A sample calculation of the pressure differential between ground level and the t op of the tower shell is also presented in appendix V. Approximately 20 % of the reduction in heat reject ed is due to the reduction in draft and approximately 80 % is due to the increased effective inle t air temperature and humidity. Ifz" is less than the height of the cooling tower shell (Zit = 90 m fr om equation (5.3) when b = 0.003), it does not influence the reduction in performance significantly if z" i s increased to the same height as the tower shell. 5.5 CONCLUSION A very simple empirical relation of the nocturnal temperature inversion profile is developed in appendix L that correlates measured data more accurately than more complex equations foun d in the literature which require more input data. The height of the temperature inversion and the h eight from which air is drawn into the cooling tower are obtained by relatively simple equations. It is found that the choice ofpractical values of the height from which air is d rawn into the cooling tower, H" does not influence tower performance significantly. The effect of the inversi on height, if it is less than

the tower shell and higher than H" also does not effect tower performance signif icantly. The influence of the humidity inversion on the reduction of tower performance is relatively small when compared to the effect of the temperature inversion on the reduction of cooling tower performanc e. Temperature inversions reduce a particular tower heat rejection by approximately 20 % in win ter and by approximately 8 % in summer. The reduction in tower perfurmance due to the adversely affected pressure differential (draft equation) on the outside of the tower, during temperature inversions, acc ounts for approximately 20 % of the total loss and the increased effective inlet temperature (transfer p rocess) for approximately 80 % of the reduction in tower performance. 6.1 CHAPTER 6 CONCLUSION 6.1 INTRODUCTION This chapter presents a summary of all the main recommendations made and the con clusions drawn during the thesis. Most of the conclusions are repeated from the conclusions dra wn at the end of each chapter. It serves as a complete overview of the main results and recommendation s. The computer software programs developed to aid in the performance analyses ofcooling towers are summarized. 6.2 WET-COOLING MEmODS OF ANALYSIS The consistent employment of the heat and mass transfer methods of analysis in t he fill performance evaluation and then using the same model in the subsequent cooling tower perform ance analysis is stressed. If used consistently the different models ought to give the same cooli ng ranges for the water in a particular cooling tower if all the operating conditions are exactly the same fo r each model. Because the Poppe approach is the more rigorous approach, it will predict the water evaporat ion rate, the total heat transfer rate and thus the air outlet temperature more accurately than the other approaches. This may lead to situations where the predicted cooling tower operating conditions will not be the same as those predicted by the other approaches, and it may therefore predict cooling ranges d ifferent from those predicted by the Merkel or e-NTU approaches. For example, the draft through natu ral draft cooling towers is a function of the air outlet temperature and the Poppe method will thus predi ct more accurately tower draft and tower performance. The Poppe method also has distinct advantages in th e analysis of hybrid cooling towers since the state of the outlet air is calculated [OIROI). This inf ormation is important to ensure that the correct amount of heated dry air is mixed with the wet plume to ensure no visible plume after mixing of the two streams. 6.3 FiLL PERFORMANCE A new empirical relation is developed that correlates measured pressure loss coe fficients accurately for all types of fills under all types of operational conditions as it is based on f undamental principles that

make provision for forces due to shear and drag. Other types of empirical equati ons, found in the literature, may correlate observed trends accurately, but they generally lack ge nerality and are only applicable for limited ranges ofwater and air flow rates. Both the empirical relations for the loss and transfer coefficients per unit hei ght of fill of wet-cooling tower fills are extended to include the effect of the height of the fill. In add ition, the empirical relation for the transfer coefficient is extended to include the effect of the water inlet te mperature. The inlet water temperature has no significant effect on the loss coefficient. It is also found that the inlet air drybulb and wetbulb temperatures have no significant effect on the loss and transfer coeffic ients. 6.2 It is recommended that as much information as possible be supplied with the empi rical relations of the loss coefficients, such as the ranges of applicability of Go and Gw? The goodnes s of fit must also be supplied in the form of a correlation coefficient This will enable the desigoer of wet-cooling systems to take the necessary precautions to compensate for any uncertainties. If possible, the same water spray system must be employed in the fill test and the subsequent cooling tower applic ation of the fill. This will eliminate the effects of droplet size and distribution on the loss coefficient. Ageing effects of the fill are not investigated in this study. 6.4 WET-COOLING TOWER PERFORMANCE EVALUATION As already mentioned, it is very important that the same model, definitions and assumptions be employed in the fill performance analysis to determine the transfer coefficient and in th e subsequent analysis to determine cooling tower performance. This will ensure that the water outlet temp erature in cooling tower performance analyses, predicted by the different models, are practically the sam e when all other variables are assumed constant. The predicted water evaporation rates in natural draft cooling towers are always higher according to the Poppe approach than according to the Merkel approach. This is not the case for m echanical draft cooling towers at very hot and dry ambient conditions. The perfurmance prediction of natural draft cooling towers is not as strongly in fluenced by the accuracy of the left-hand side of the draft equation given by equation (4.1). If the left hand side of the draft equation accounts for the moist air that is raised in a gravitational field, adi abatic cooling and heating by condensation, the same order of results are obtained when the simplified equatio n, given by equation (4.2) is employed. When the Poppe approach is employed in hot ambient conditions, the value of the Lewis factor has little influence on the prediction ofthe water outlet temperature, if it is employed co nsistently or inconsistently. The results of the Merkel approach can be brought within close tolerance of the more rigorous Poppe approach, when the reduction of the water mass flow rate, due to evaporation, is

included in the energy balance. The assumption of Merkel that the outlet air is saturated with water va por, leads to tower performances that are within close tolerance of the tower performance predicted by the Poppe approach, for cold or humid ambient conditions. The outlet air is generally supersaturated when the inlet ambient air is cold or relatively humid. The air drybulb temperature at a specific air entha lpy is practically constant for air in the saturated state or air in any degree of supersaturation. This is the reason why the results of the Merkel and Poppe methods are within close tolerance with each other. The Mer kel method predicts saturated outlet air while the Poppe method generally predicts supersaturated ou tlet air when the inlet ambient air is cold or relatively humid. The discrepancy between the performance evaluations according to the Merkel and Poppe methods are greater when the outlet air according to the Poppe method is 6.3 unsaturated. This is because the temperature of unsaturated air is a strong func tion of the humidity of the air at a specific air enthalpy. The e-NTU and Merkel approaches predict virtually the same tower performances wh en the models are applied consistently or inconsistently. 6.5 TIIE INFLUENCE OF TEMPERATURE AND HUMIDITY INVERSIONS ON COOLING TOWER PERFORMANCE A very simple empirical relation of the nocturnal temperature inversion profile is developed in appendix L that correlates measured data more accurately than models found in the literat ure which require more input data. Drybulb temperature measurements at two heights are sufficient to de termine the temperature inversion profile. The one measurement is typically taken at I m above ground el evation while the second measurement must be taken as high above ground elevation as possible (typically 10 m). The height of the temperature inversion and the height from which air is drawn into the coolin g tower are obtained by relatively simple empirical equations. It is found that the choice ofpractical values of the height from which air is d rawn into the cooling tower, H" does not influence tower performance significantly. The effect of the inversi on height, if it is less than the tower shell and higher than H" also does not effect tower performance signif icantly. The influence of the humidity inversion on the reduction of tower performance is relatively small when compared to the effect of the temperature inversion on the reduction of cooling tower performanc e. Temperature inversions reduce a particular tower heat r?iection by approximately 20 % in win ter and by approximately 8 % in summer. The reduction in tower performance due to the adversely affected pressure differential (draft equation) on the outside of the tower, during temperature inversions, acc ounts for approximately 20 % of the total loss and the increased effective inlet temperature (transfer p rocess) for apprOximately 80 % of the reduction in tower performance.

6.6 SOFTWARE DEVEWPMENT A program is developed to process and analyze fill performance test data. lhis p rogram is presented in appendix K and processes the pressure transducer and thermocouple data, determin es the transfer and loss coefficients and fits relatively complex curves through the test data with mathe matical optimization algorithms. A comprehensive program is developed to predict wet-cooling tower pe rformance. This program is presented in appendix P. Natural draft counterflow and mechanical dra ft counterflow and crossflow cooling towers can be analyzed by the program. The latest empirical an d heat and mass transfer models found in the literature are included in the solution algorithms of the so ftware. The analytical and empirical models, developed in this thesis from theoretical and experimental inv estigations, are also included in the software. As discussed in appendix U, the geometrical dimensions of a natural draft cooling tower can be optimized by the program to obtain the minimum combined cap ital and operational cost compounded over the economic life of the cooling tower. It is shown in appe ndix U that the inlet height of the cooling tower is generally the critical dimension influencing the combined operational and I f 6.4 capital cost of a cooling tower. A comprehensive cooling tower performance evalu ation tool, that is very user friendly, is therefore developed that predicts and analyzes the thennal per formance of wet-cooling towers. Furthennore, programs are developed to plot psychrometric charts (chapte r 2) and generate cooling tower performance curves (appendix Q). 6.7 SCOPE FOR FUTURE WORK All of the results obtained in this thesis are obtained by essentially employing one-dimensional analytical models. However, some problems can only be solved satisfactorily by three-dimens ional numerical modelling when the air or water mass flow rates are non-uniform, or a combinatio n of counterflow and erossflow exists in the fill. 7.1 REFERENCES Each reference is identified by a code in square brackets which consists of two digits for the year of publication, the first two letters of the first authors surname and a sequential ly assigned digit to make the reference unique. [08HAI] [22LEI] [25MBI] [33LEI] [42HUI] [49JOI] [52ERI] [54MOl] [56Z11] [58MBI]

[59MCI] [60SCI] [61BAI] [61BEI] [61LOI] Hann, J., Handbuch der Klimatologie, 1. Band: Allgemeine Klimalebre, Verlag von J. Engelhom, Stuttgart, 1908. Lewis, W.K., The Evaporation of a Liquid into a Gas, Transactions of ASME, Vol. 44, pp. 325-340, May, 1922. Merkel, F., VerdunstungskOhlung, VDI-Zeitchrift, Vol. 70, pp. 123-128, January 1 925. Lewis, W.K., The Evaporation of a Liquid into a Gas - A Correction, Mechanical Engineering, Vol. 55, pp. 567-573, 1933. Hutchison, W.K. and Spivey, E., Design and Performance of Cooling Towers, Transa ctions of the InstiMe ofChemical Engineers, Vol. 20, pp. 14-29, 1942. Jordan, RC. and Priester, G.B., Refrigeration and Air Conditioning, Constable an d Company, Ltd., London, 1949. Ergun., S., Fluid Flow through Packed Columns, Chemical Engineering Progress, Vo l. 48, pp. 89-94, 1952. Monin, A.S. and Obukhov, A.M., Basic Regularity in Turbulent Mixing in the Surfa ce Layer of the Atmosphere, Frud. Geofig. Inst. Akkad. Nauk. SSSR, Vol. 24, pp. 151 -163, 1954. Zivi, S.M. and Brand, B.B., An Analysis of the Crossflow Cooling Tower, Refriger ating Engineering, Vol. 64, pp. 31-34 & 90-92,1956. Meteorological Office, Tables of Temperature, Relative Humidity and Precipitatio n for the World, Part IV, Africa, The Atlantic Ocean South of 35?N and the Indian Ocean, H er Majesty's Stationary Office, London, 1958. McKelvey, K.K. and Brooke, M., The Industrial Cooling Tower, Elsevier Publishing Company, Amsterdam, 1959. Schlichting, H., Boundary Layer Theory, Fourth Edition, McGraw-Hill, New York, 1 960. Baker, DR and Shryock, B.A., A Comprehensive Approach to the Analysis of Cooling Tower Performance, Transactions of the ASMB, Journal of Heat Transfer, pp. 339-3 50, 1961. Berman, L.D., Evaporative Cooling of Circulating Water, 2nd Edition, Chapter 2, pp. 94-99, ed. Sawistowski, H., Translated from Russian by R Hardbottle, Pergamon Press, Ne w York, 1961. Lowe, HI. and Christie, D.G., Heat Transfer and Pressure Drop Data on Cooling To wer Packings and Model Studies of the Resistance of Natural Draft Towers to Airflow, Proceedings of the International Heat Transfer Conference, Colorado, Part V, pp. 933-950, [65BOI] [65GEI] [67CTl] [72MCl] [74LAI]

[75NAI] [75YOI] [76ANI] [76KEI] [76KE2] [77M11] [78MOI] [780KI] [80PAI] [8IBRI] [8IGOI] [82CAI] [82M11] [82SNI] 7.2 1961. Bosnjacovic, F., Technische Thennodinmik, Theodor Steinkopl; Dresden, 1965. Geiger, R, The Climate Near the Ground, Harvard University Press, Cambridge, 196 5. cn, Cooling Tower Performance Curves, The Cooling Tower Institute, Houston, 1967 . McGee, O.S., The Content of Water Vapor in the Atmosphere Over Southern Africa, SA Geographer, Vol. 4, No. I, pp. 25-32,1972. Launder, RE. and Spalding, D.R, The Numerical Computation of Turbulent Flows, Computer Methods in Applied Mechanics and Engineering, Vol. 3, pp. 269-289, 1974 . Nahavandi, AN., Kershah, RM. and Serico, RJ., The Effect of Evaporation Losses i n the Analysis of Counterflow Cooling Towers, Journal of Nuclear Engineering and Desig n, Vol. 32, pp. 29-36, 1975. Yoshino, M.M., Climate in a Small Area, An Introduction to Local Meteorology, Un iversity of Tokyo Press, Tokyo, 1975. Anfossi, D., Bacci, P. and Longhetlo, A, Forcasting of Vertical Temperature Prof iles in the Atmosphere during Nocturnal Radiation Inversion from Air Temperature Trend at Sc reen Height, Quarterly Journal of the Royal Meteorological Society, Vol. 102, pp. 173 -180, 1976. Kelly, N.W., Kelly's Handbook of Crossflow Cooling Tower Performance, Kansas Cit y, Missouri, Neil W. Kelly and Associates, 1976. Kelly, N.W., A Blueprint for the Preparation of Crossflow Cooling Tower Characte ristic Curves, Paper Presented before the Cooling Tower Institute Annual Meeting, Janua ry, 1976. Miller, D.H., Water at the Surface of the Earth, An Introduction to Ecosystem Hydrodynamics, Academic Press, New York, 1977. Montakhab, A., Waste Heat Disposal to Air with Mechanical and Draft - Some Analy tical Considerations, Heat Transfer Division of the ASME, Winter Annual Meeting, San Francisco, 1978. Oke, TR, Boundary Layer Climates, Methuen & Co Ltd, London, 1978. Patankar, S.V., Numerical Heat Transfer and Fluid Flow, Hemisphere Publishing Co ., New York, 1980. British Standard 1042, Measurement of Fluid Flow in Closed Conduits, Part I, Sec tion 1.1,

1981. Gorchakov, G.!., Kostka, O.K. and Krikunov, G.A, Statistical Properties of Humid ity Profiles and the Backscattering Coefficient in the Lower Troposphere, Atmospheri c and Oceanic Physics, Vol. 17, No. 10, pp. 777-781,1981. Cale, SA, Development of Evaporative Cooling Packing, Commission of European Communities, Report EUR 7709 EN, Luxembourg, 1982. Missimer, J. and Wilber, K., Examination and Comparison of Cooling Tower Compone nt Heat Transfer Characteristics, IAHR Cooling Tower Workshop, Hungary, October 1215, 1982. Snyman, J. A, A New and Dynamic Method for Unconstrained Minimization, Appl. Mat h. Modelling, Vol. 6, pp. 449-462, 1982. [82STl] [83BOI] [83B02] [83MAI] [83MA2] [83MA3] [83SNI] [83SUl] [84BRI] [84POl] [85Lll] [85SNI] [85Wll] [86SUl] [87HOl] [87SUl] [88BRI] [88DRI] [88LAI] 7.3 Stoecker, W.F. and Jones, J.W., Refrigeration and Air Conditioning, McGraw-Hill Book Co., Singapore, 1982. Bourillot, C., TEFERl, Numerical Model for Calculating the Performance of an Eva porative Cooling Tower, EPRI Report CS-3212-SR, August 1983. Bouri1lot, C., On the Hypothesis of Calculating the Water Flowrate Evaporated in a Wet Cooling Tower, EPRI Report CS-3144-SR, August 1983. Majumdar, AK, Singhal, AK and Spalding, D.B., Numerical Modeling of Wet Cooling Towers - Part I: Mathematical and Physical Models, Transactions of the ASME, Jou rnal of Heat Transfer, Vol. 105, pp. 728-735, November 1983. Majumdar, AX., Singhal, A.K., Reilly, H.E. and Bartz, J.A, Numerical Modeling of Wet Cooling Towers - Part 2: Application to Natural and Mechanical Draft Towers, Tra nsactions of the ASME, Journal of HeatTransfer, Vol. 105, pp. 736-743, November 1983. Majumdar, A.K, Singhal, AK and Spalding., DR, VERA2D: Program for 2-D Analysis of Flow, Heat, and Mass Transfer in Evaporative Cooling Towers, EPRI Report CS 2 923, Volume I and 2, March 1983. Snyman, J. A., An Improved Version of the Original Leap-Frog Dynamic Method for Unconstrained Minimization LFOPI(b), Appl. Math. Modelling, Vol. 7, pp. 216-218,

1983. Sutherland, J.W., Analysis of Mechanical-Draught Counterflow AirlWater Cooling T owers, Transactions of the ASME, Journal ofHeat Transfer, Vol. 105, pp. 576-583, August 1983. British Standard 1042, Measurement of Fluid Flow in Closed Conduits, Part I, Sec tion 1.2 and Section 1.4, 1984. Poppe, M. and R5gener, H., Berechnung von Rtickkiih1werken, VDI-Wl!rmeatlas, pp. Mhl? MhI5,1984. Li, KW. and Priddy, A.P., Power Plant System Design, John Wiley & Sons, 1985. Snyman, J. A., Unconstrained Minimization by Combining the Dynamic and Conjugate Gradient Methods, Quaestiones Mathematicae, Vol. 8, pp. 33-42, 1985. Wilber, K.R., Yost, J.G. and Wheeler, D.E, An Examination of the Uncertainties i n the Determination of Natural Draft Cooling Tower Performances, Joint AMSElIEEE Power Generation Conference, Milwaukee, Wisconsin, October 20-24, 1985. Surridge, AD., Extrapolation of the Nocturnal Temperature Inversion from GroundBased Measurements, Atmospheric EnVironment, Vol. 20, No.4, pp. 803-806, 1986. Hoffmann, J.E., Bedryfspunt Voorspelling vir Nat Koeltorings, M.Eng Thesis, Univ ersity of Stellenbosch, Stellenbosch, South Africa, 1987. Surridge, AD., On the Evolution of the Height and Temperature Difference Across the Nocturnal Stable Boundary Layer, Boundary-Layer Meteorology, Vol. 40, pp. 87-98, 1987. British Standard 4485, Water Cooling Towers, Part 2: Methods for Performance Tes ting, 1988. Dreyer, AA., Analysis of Evaporative Coolers and Condensers, M.Eng Thesis, Unive rsity of Stellenbosch, Stellenbosch, South Africa, 1988. Lauraine, H., Lemmens, P. and Monoie, M., Experimental Data Coupling Atmospheric [88PRI] [89DAI] [89JAI] [89JOl] [90COI] [90ELl] [90SUl] [9IFEI] [91OSI] [9IPOl] [9IWHI] [92HEI] [92MAI] [92W11] [93BEI] [93BE2] [93KRI] [93MB] [94DUl] 7.4 Temperature Inversions and Cooling Tower Perfonnances, Proceedings of the 6th lA HR Cooling Tower Workshop, Pisa, Italy, 1988. Preston-Whyte, RA. and Tyson, P.D., The Atmosphere and Weather of Southern Afric a,

Oxford University Press, Cape Town, 1988. Duagher!y, RL., Franzini, J.B. and Finnemore, E.J., Fluid Mechanics with Enginee ring Applications, SI Metric Edition, McGraw-Hill, Singapore, 1989. Jaber, H. and Webb, RL., Design of Cooling Towers by the Effectiveness-NTU Metho d, Journal of Heat Transfer, Vol. III, pp. 837-843, November 1989. Johnson, B.M. (ed.), Cooling Tower Performance Prediction and Improvement, Volum e I, Applications Guide, EPRI Report GS-6370, Volume 2, Knowledge Base, EPRI Report G S? 6370, EPRl, Palo Alto, 1989. Cooling Tower Institute, cn Code Tower, Standard Specifications, Acceptance Test Code for Water-Cooling Towers, Part I, Part II and Part III, cn Code ATC-105, Revised , February 1990. Ellis, R and Gulick, D., Calculus with Analytic Geometry, Fourth Edition, Harcou rt Brace Jovanovic College Publishers, Fort Worth, 1990. Surridge, A.D., Swanepoel, D.J.deV., Held, G., Research on Thermal Feedback Caus ed by Dry-Cooling Power Generating Stations, Confidential Report, EMA-C 9086, CSIR, Pr etoria, 1990. Feltzin, A.E. and Benton D., A More Exact Representation of Cooling Tower Theory , Cooling Tower Institute Journal, Vol. 12, No.2, pp. 8-26, 1991. Osterle, F., On the Analysis of Counter-Flow Cooling Towers, International Journ al of Heat and Mass Transfer, Vol. 34, No. 4/5, pp. 1313-1316, 1991. Poppe, M. and ROgener, H., Berechnung von RllckkUhlwerken, VDI-Wlirmeatlas, pp. Mi 1? Mi 15, 1991. White, F.M., Viscous Flnid Flow, Second Edition, McGraw-Hill, New York, 1991. Hensley, J., Maximize Tower Power, Chemical Engineering, pp. 74-82, February, 19 92. Mathews, J.H., Numerical Methods fur Mathematics, Science, and Engineering, Seco nd Edition, Prentice-Hall International, Inc, 1992, Willa, J.L., Evolution of the Cooling Tower, cn Journal, Vol. 13, No. I, pp. 404 9, 1992. Becker, B.R and Burdick, L.F., Drift Eliminators and Cooling Tower Performance, ASHRAE Journal, pp. 28-36, June 1993. Beckwith, T.G., Marangoni, RD. and Lienhard, J.H., Mechanical Mearurements, Fift h Edition, Addison-Wesley Publishing Company, 1993. Krane, SC, Performance of Counterflow Cooling Towers with Structured Packings an d Maldistributed Water Flow, Numerical Heat Transfer, Part A, Vol. 23, pp. 115-127 , 1993. Mirsky, G.R and Bauthier, J., Evolution of Cooling Tower Fill, cn Journal, Vol. 14, No. I, pp. 12-19,1993. Du Preez, A.F. and Krllger, D.G., The Influence of a Buoyant Plume on the Perfor mance of a Natural Draft Cooling Tower, 9th lAHR Cooling Tower and Spraying Pond Symposiu m, [94GRI] [94SNI]

[94WHI] [95ANI] [95BEI] [95BLI] [95COI] [95ffil] [95KII] [95MB] [95LI1] [95001] [95SAI] [96HAI] [96MOI] [96M02] [97B0I] [97BUl] 7.5 Brussels, 1994. Grange, J.L., Calculating the Evaporated Water Flow in a Wet Cooling Tower, Pape r presented at the 9th IAHR Cooling Tower and Spraying Pond Symposium, von Karman Institute, Brussels, Belgium, September 1994. Snyman, J. A, Stander, N. and RoW<, W. J., A Dynamic Penalty Function Method for the Solution of Structural Optimization Problems, Appl Math Modelling, Vol. 18, pp. 453-460, 1994. White, F.M., Fluid Mechanics, Third Edition, McGraw-Hill, New York, 1998. Anderson, Jr., J.D., Computational Fluid Dynamics, The Basics with Applications, McGraw -Hill, New York, 1995. Bernier, MA., Thermal Performance of Cooling Towers, ASHRAE Journal, pp. 56-61, April 1995. Bland, C., A Cool Solution to a Hot Problem, Process Engineering, pp. 33, June, 1995. Conradie, AE., Performance Optimization of Engineering Systems with Particular Reference to Dry-Cooled Power Plants, Ph.D. Thesis, University of Stellenbosch, South Africa, 1995. Ibrahim, G.A., Nabhan, M.B.W. and Anabtawi M.Z., An Investigation into a Falling Film Type Cooling Tower, International Journal of Refrigeration, Vol. 18, No.8, pp. 5 57-564, 1995. Kintner-Meyer, M. and Emery, AF., Cost-Optimal Design for Cooling Towers, ASHRAE Journal, pp. 46-55, April 1995. Mills, A.F., Basic Heat and Mass Transfer, Irwin, Chicago, 1995. Liffick, G.W. and Cooper, Jr, J.W., Thermal Performance Upgrade of the Arkansas Nuclear One Cooling Tower: A "Root Cause" Analysis Approach, Prooeedings of the American Power Conference, Vol. 57, No.2, pp. 1357-1362, 1995. Oosthuizen, P.C., Performance Characteristics of Hybrid Cooling Towers, M.Eng. T hesis, University ofStellenbosch, Stellenbosch, South Africa, 1995. Sadasivam, M. and Balakrishnan, AR, On the Effective Driving Force for Transport in Cooling Towers, Transactions of the ASME, Journal of Heat Transfer, Vol. 117, pp . 512? 515, May 1995. Hardy, R, Weather, Teach Yourself Books, Hodder & Stoughton, London, 1996.

Mohiuddin, AK.M. and Kant, K., Knowledge Base for the Systematic Design of Wet Cooling Towers. Part I: Selection and Tower Characteristics, International Journ al of Refrigeration, Vol. 19, No. I, pp. 43-51, 1996. Mohiuddin, AK.M. and Kant, K., Knowledge Base for the Systematic Design of Wet Cooling Towers. Part II: Fill and other Design Parameters, International Journal of Refrigeration, Vol. 19, No. I, pp. 52-60, 1996. Bowman, C.F. and Benton, D.J., Oriented Spray-Assisted Cooling Tower, cn Journal , Vol. 18, No. 1,1997. Burden, R.L. and Faires, J.D., Nwnerical Analysis, Sixth Edition, Brooks/Cole Pu blishing [97COI) [97DEI) [97ELI) [97HOI) [97HUI) [98ALI) [98BAI) [98COI) [98KRI) [98SEI] [98SNI) [98STl) [99DEI) [99HAI) [99S01) [99WAI) [99WlI) [OOADI] 7.6 Company, 1997. Cooling Tower Institute, cn Code Tower, Standard Specifications, Acceptance Test Code for Water-Cooling Towers, Vol. I, cn Code ATC-I05(97), Revised, February 1997. De Villiers, E. and KrOger, D.G., Analysis of Heat, Mass and Momentum Transfer i n the Rain Zone of Counterflow Cooling Towers, Proceedings of the 1997 IJPGC, Vol.2, P WR? Vol. 32, pp. 141-149, Denver, November 1997. EI-Dessouky, R TA, AI-Haddad, A and AI-Juwayhel, F., A Modified Analysis of Coun ter Flow Wet Cooling Towers, Journal of Heat Transfer, Vol. 119, No.3, pp. 617-626, 1997. Hoffmann, J.E., The Influence of Temperature Stratification in the Lower Atmosph eric Boundary Layer on the Operating Point ofa Natural Draft Dry-Cooling Tower, Ph.D Thesis, University of Stellenbosch, Stellenbosch, South Africa, 1997. Huser, A., Nilsen, P.l and Skatun, H., Application of k-& Model to the Stable AB L: Pollution in Complex Terrain, Journal ofWind Engineering and Industrial Aerodyna mics, Vol. 67 and 68, pp. 425-436, 1997. AI-Nimr, M.A., Dynamic Thermal Behaviour of Cooling Towers, Energy Conversion Management, Vol. 39. No.7, pp. 631-636, 1998. Baard, T.W., Performance Characteristics of Expanded Metal Cooling Tower Fill, M .Eng Thesis, University ofStellenbosch, Stellenbosch, South Africa, 1998.

Conradie, A.E., Buys, J.D. and Kroger, D.G., Perfonnance Optimization of Dry-Coo ling Systems for Power Plants through SQP Methods, Applied Thermal Engineering, Vol. 18, Nos. 1-2, pp. 25-40, 1998. KrOger, D.G., Air-Cooled Heat Exchangers and Cooling Towers Thermal-Flow Performance, Evaluation and Design, Begell House, Inc., New York, 1998. Seinfeld, lR and Pandis, S.N., Atmospheric Chemistry and Physics, From Air Pollu tion to Climate Change, John Wiley & Sons, Inc., New York, 1998. Snyman, J. A., ETOPC: A Fortran Program for Solving General Constrained Minimiza tion Problems by the Conjugate Gradient Method without Explicit Line Searches, Resear ch Report, Department of Mechanical Engineering, University of Pretoria, 1998. Streng, A., Combined WetIDry Cooling Towers of Cell-Type Construction, Journal o f Energy Engineering, Vol. 124, No.3, pp. 104-121, December 1998. De Villiers, E. and Kroger, D.G., Inlet Losses in Counterflow Wet-Cooling Towers , Joint Power Generation Conference, Vol2, PWR-Vol. 34, ASME, 1999. Hliszler, R, Einflusz von Kondensation in der Grenzschicht auf die WlIrme- und Stoffilbertragung an einem Rieselfihn, Fortschritt-Berichte VDI, Reihe 3, Nr. 61 5,1999. S1lylemez, M.S., Theoretical and Experimental Analysis of Cooling Towers, ASHRAE Transactions: Research, Vol. 105, No. I, pp. 330-337, 1999. Wallis, lS. and Aull, Rl, hnproving. Cooling Tower Performance, Hydrocarbon Engineering, pp. 92-95, May, 1999. Williams, T. and Kelly, C., Gnuplot, MS-Windows 32 bit Version 3.7, Patchlevel 0 , 1999. Aull, R.J., and Krell, T., Desigo Features of Cross-Fluted Film Fill and Their E ffect on 7.7 Thermal Performance, cn Journal, Vol. 21, No.2, pp. 12-33,2000. [OOCAl] Castro, M.M., Song, T. W. and Pinto, J.M, Minimization of Operational Co sts in Cooling Water Systems, Transactions of the Institution of Chemical Engineers, Vol. 78, P art A, pp. 192-201, March, 2000. [OOGOI) Goshayshi, RR and Missenden, J.F., The Investigation of Cooling Tower Pa cking in Various Arrangements, Applied Thermal Engineering, Vol. 20, pp. 69-80, 2000. [OOG02) Goya~ O.P., Maintenance and Retrofitting, Guidelines and Troubleshooting , Hydrocarbon Processing, Vol. 79, No. I, p. 69, 2000. [OOSNI) Snyman, J. A., The LFOPC Leap-Frog Method fur Constrained Optimization, Computers Math. Applic., Vol. 40, No. 8/9, pp. 1085-1096,2000. [00SN2) Snyman, J. A. and Hay, A. M., The Dynamic-Q Optimization Method: An Alte rnative to SQP?, Proceedings of the International Workshop on Multidisciplinary Desigo Optimization, University ofPretoria, Pretoria, South Africa, pp. 163-172, August , 2000. [OIMAI) Makkinejad, N., Temperature Profile in Countercurrent/Cocurrent Spray To wers, International Journal of Heat and Mass Transfer, Vol. 44, pp. 429-442, 2001. [OIMII) Milosavljevic, N. and Heikkila, P., A Comprehensive Approach to Cooling Tower Desigo, Applied Thermal Engineering, Vol. 21, pp. 899-915,2001. [OIROI) Roth, M., Fundamentals of Heat and Mass Transfer in Wet Cooling Towers.

All Well Known or are Fwther Developments Necessary? 12th IAHR Symposium in Cooling Tower s and Heat Exchangers, UTS, Sydney, Australia, pp. 100-107, November, 2001. [OITUI) Turpin, J.R (ed), Wantto Save Energy? Look at your Cooling Tower, Engine ered Systems, Vol. 18, No. 10, p. 48, 2001. [02BUl) Busch, D., Harte, R., Krlltzig, W.B. and Montag, U., New Natural Draft C ooling Tower of 200 m ofHeight, Engineering Structures, Vol. 24, pp. 1509-1521,2002. [02FIl) Fisenko, S.P., Petruchik, A.I. and Solodukhin, A.D., Evaporative Cooling of Water in a Natural Draft Cooling Tower, International Journal of Heat and Mass Transfer, Vo l. 45, pp. 4683-4694, 2002. [02HAI) Harte, R. and Krlltzig, W.B., Large-Scale Cooling Towers as Part of an E fficient and Cleaner Energy Generating Technology, Thin-Walled Structures, Vol. 40, pp. 651-6 64, 2002. [02HA2) Hawlader, M.NA and Lui, B.M., Numerical Study of the Thermal-Hydraulic P erformance of Evaporative Natural Draft Cooling Towers, Applied Thermal Engineering, Vol. 2 2, pp. 41-59,2002. [02THI) Thiart, G.D., Preliminary CFD Analysis ofa Solar Chimney, HEFAT2002, I" International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Kruger Park, So uth Africa, pp. 449-452, 2002. r A.I APPENDIX A PROPERTIES OF FLUIDS AS SUMMURISED BY KROGER [98KRl) A.l THE THERMOPHYSICAL PROPERTIES OF DRY AIR FROM 220K TO 380K AT STANDARD ATMOSPHERIC PRESSURE (101325N/m'). Density: Po = p.t(287.08 1), kglm' Specific heat: c"" ~ 1.045356 X 10' - 3.161783 X 10-1 T+ 7.083814 X 10-4 T' - 2.705209 X 10-7 T', JlkgK Dynamic viscosity: ~ = 2.287973 X 10'" + 6.259793 X 10'" T - 3.131956 x 10-11 T' + 8.15038 X 10-15 T', kglsm Thennal conductivity: ka = -4.937787 x 10-4 + 1.018087 xlO-4 T - 4.627937 x 10'" T 2 + 1.250603 X 10-11 T', W/mK (A. 1.1) (A.1.2) (A.1.3) (A. 1.4) A.2 A.2 THE THERMOPHYSICAL PROPERTIES OF SATURATED WATER VAPOR FROM 273.15K TO 380K. Vapor pressure: p,~ 10', N/m' z = 10.79586(1- 273.16/1) + 5.028081oglO(273.16/1) + 1.50474 x lo-'[I_10--8?29692{(TI273.l6)-1}] + 4.2873 x 10-4[10 '.76Oll(l-273.16!!) - 1] + 2.786118312 Specific heat:

cp, ~ 1.3605 X 10' + 2.31334 r - 2.46784 x 10-10 r '+ 5.91332 x 10-13 r 6, J/kgK Dynamic viscosity: J4 ~ 2.562435 X 10-6 + 1.816683 X 10--8 r+ 2.579066 x 10-11 r' - 1.067299 x 10-14 r', kg/sm Thermal conductivity: k, ~ 1.3046 X 10-' - 3.756191 xl0-l r+ 2.217964 x 10-7 r' -1.111562 x 10-10 r', W/mK Vapor density: A ~ -4.062329056 + 0.10277044r - 9.76300388 x 10-4 r' + 4.475240795 x 10-6r ' -1.004596894 x 1O-8r 4 + 8.9154895 x IO-I'r ' , kg/m' Temperature: r~ 164.630366 + 1.832295 x IO-'p, + 4.27215 x 1O-IOp/ + 3.738954 x 10'p,-' -7.01204 X lO ' p,-' + 16.161488 In p, - 1.437169 X 10-4 p, In p", K (A.2.1) (A.2.2) (A.2.3) (A.2.4) (A.2.5) (A.2.6) A.3 A.3 THE THERMOPHYSICAL PROPERTIES OF MIXTURES OF AIR AND WATER VAPOR. Density: p.,= (I + w) [1- w/(w + O.62198)]p",1(287.08n, kg air-vapor/m3 Specific heat: cPO' = (cpa +wcp,)/(1 +w), 11K kg air-vapor or the specific heat of the air-vapor mixture per unit mass of dry air: cpm, = (cpo + wcp,), 11K kg dry air Dvnamic viscosity: f./a' = (x"J.loM,o., +X,p.M,Oj)! (X"M,05 +X,M,o.,), kg/ms where M, ~ 28.97 kg/mole, M, = 18.016 kg/mole, Xa = 11(1 + 1.608 w) and X, =w/(w + 0.622) Thermal conductivity: ka, ~ (X,kaM,O.33 +X,k,M,O.33)! (X"M,OJ3 +X,M,O.33), W/mK Humidity ratio: ( 2501.6-2.3263(Tw' -273.15) J( 0.62509p~, J w= 2501.6+1.8577(T-273.15)-4.184(Tw, -273.15) Pob' -I.005p_, ( . I.00416(T-Twb ) ) 2501.6 +1.8577(T - 273.15) - 4.184(Tw' - 273.1 5) Enthalpy: i" = [cpo(T - 273.15) +w{ifgwo +cp,(T-273.15)}]/(I +w), Ilkg air vapor or the enthalpy oftbe air-vapor mixture per unit mass of dry air: im, =cpo(T - 273.15) +w[ifgwo + cp,(T-273.1 5)], I/kg air vapor where the specific heats are evaluated at (T + 273.15)/2 and the latent heat ig.o' is evaluated at 273.15K according to equation (AA.5). (A.3.1) (A.3.2a) (A.3.2b) (A.3.3) (A.3A) (A.3.5) (A.3.6a) (A.3.6b) r A.4 A.4 THE THERMOPHYSICAL PROPERTIES OF SATURATED WATER LIQUID FROM 273.15K TO 380K.

Density: Pw = (1.49343 X 10-' -3.7164 x 10"'T+ 7.09782 x IO-'T' -1.90321 x 10-lOT'rl , kg /m' Specific heat: Cpw = 8.15599 X 10' - 2.80627 x 10 T+ 5.11283 X 10-' T' - 2.17582 X 10-13 T', Jl kgK Dynamic viscosity: Pw = 2.414 x 10-' x 10 '47.81(T-140\ kg/sm Thermal conductivity: kw = -6.14255 X 10-1 + 6.9962 xl0-' T-1.01075 x 10-' T'+ 4.74737 X 10-12 T 4, W/ mK Latent heat ofvaporation: ifgw ~ 3.4831814 X 10' - 5.8627703 xl0' T + 12.139568 T' - 1.40290431 X 10-' T', JIK Critical pressure: Pwo = 22.09 X 10', N/m' Surface tension: OW = 5.148103 X 10-' + 3.998714 X 10-4 T-1.4721869 x 10'" T'+ 1.21405335 x 10-' T' (A.4.1) (A.4.2) (A.4.3) (A.4.4) (A.4.5) (A.4.6) (A.4.7) B.1 APPENDIXB HEAT AND MASS TRANSFER IN COUNTERFLOW WET-COOLING TOWERS B.IINTRODUCTION The governing equations for heat and mass transfer in the fill of a counterflow cooling tower are derived in this appendix. The governing equations for Merkel's and Poppe's models are pr esented. The Merkel theory relies on several critical assumptions to reduce the solution to a simple hand calculation. Because of these assumptions, however, the Merkel method does not accurately represent t he physics of heat and mass transfer process in the cooling tower fill. The critical simplifying assumptions of the Merkel theory are [83B02]: o The Lewis factor relating heat and mass transfer is equal to 1. This assumptio n has a small influence but affects results at low ambient temperatures. o The air exiting the tower is saturated with water vapor and it is characterize d only by its enthalpy. This assumption regarding saturation has a negligible influence above an ambient temperature of 20?C but is of importance at lower temperatures. o The reduction of water flow rate by evaporation is neglected in the energy bal ance. This energy balance simplification has a greater influence at elevated ambient temperatures. Bourillot [83B02] stated that the Merkel theory is simple to use and can correct ly predict cold water temperature when an appropriate value of the coefficient of evaporation is used. In contrast, it is insufficient for the estimation of the characteristics of the warm air leaving t he fill and for the calculation of changes in the water flow rate due to evaporation. These quantities are impor tant to estimate water consumption and to predict the behavior of plumes exiting the cooling tower.

The method of Poppe does not make the simplifying assumptions ofMerke\. Predicti ons from the Poppe formulation result in values of evaporated water flow rate that are in good agre ement with full scale cooling tower test results. In addition, the Poppe method predicts the water con tent of the exit air accurately [83BOI, 83B02]. Sections B.I and B.2 are adapted from Bourillot [83B01], Poppe and Rllgener [9IP OI], Krllger [98KRI] and Baard [98BAI]. B.2 GOVERNING EQUATIONS FOR HEAT AND MASS TRANSFER IN FILL FOR UNSATURATED AIR Figure B.I shows a control volume in the fill of a counterflow wet-cooling tower . Figure B.2 shows an airside control volume of the fill illustrated in figure RI. B.2 dz ,I I , , I I I I I ma(l + w + dw) ima dima ---------- -I I I, , I I I I , m.(l + w) ima Figure B.I: Control volume of counterflow fill m.(l+w+dw) ima + ima dz r---------------I I m.(l +w) ima -, , I, , I, , I , Figure B.2: Air side control volume ofthe fill A mass balance for the control volume in figure B.t yields, dm. =madw

The energy balance for the control volume of the fill in figure B.t is as fullow s: madintQ - mwdiw - iwdm"", = 0 where ima is the enthalpy ofthe air-vapor mixture, expressed by equation (A.3.6b ). Substitute equation (B.1) into equation (B.2) to find upon rearrangement, dT = rna (_t_ di - T dw)w ma w mw Cpw (B.t) (B.2) (B.3) (B.4) Consider the interface between the water and the air in figure B.2. An energy ba lance at the interface yields, dQ =dQm + dQ, where dQm is the enthalpy transfer due to difference in vapor concentration betw een the saturated air at the interface and the mean stream air and dQ, is the sensible heat transfer due to the difference in temperature. The mass transfer at the interface is expressed by, dm. = h,,(W,.- w)dA (B.5) B.3 The corresponding enthalpy transfer for the mass transfer in equation (B.S) is dQm = i, dmw~ i,h,,(w,w- w)dA The enthalpy ofthe water vapor, i" at the bulk water temperature, Tw, is given b y (B.6) ill = iJkwtl + cpy T.III The convective heat transfer from figure B.2 is given by (B.7) dQ, = h(I'w - T,JdA (B.8) The temperature differential in equation (B.8) can be substituted by an enthalpy differential. The enthalpy of saturated air evaluated at the local bulk water temperature is given by imasw = cpa Tw + wsw(i/gwo + Cpv rw) Substitute equation (B.7) into equation (B.9) and find upon rearrangement (B.9) (B.10) The enthalpy of the air-water vapor mixture per unit mass of dry air, which acco rding to equation (A.3.6b) is expressed by ima = cpa Ta+ w(if.., + cp, T,J The specific heat of the air-water vapor mixture fur unsaturated air is defined by (B.1I) epma = cpa + w Cpy (B.12) (B.13) Subtract equation (B.II) from (B.IO). The resultant equation can be simplified i f the small differences in specific heats, which are evaluated at different tempe~tures, are ignored. T -T =(i....... -ima}-(w,. -w), w a Cpma where cpma is given by equation (B.12). Substitute equation (B.13) into equation (B.8). Substitute the resultant equatio n and equation (B.6) into equation (B.4) to find upon rearrangement, (B.14) is known as the Lewis factor, Lef, and is an indication of the relative rates of

heat and mass cpmahd dQ=hd[ h (ima,. -im.)+(1 h )i,(W,. -w}JdA cpmahd cp.."hd h (B. IS) transfer in an evaporative process. Bosnjakovic [6SBOI] developed an empirical r elation for the Lewis factor, Lef, for air-water vapor systems. The Lewis factor for unsaturated air, according to Bosnjakovic [6SBOI] is given by ( W,. +0.622 ) Lef = 0.86So."" w + 0.622 I tn( w,. +0.622) w+0.622 Refer to appendix F for a discussion on the derivation of equation (B.1S). Alter native approaches for the determination of the Lewis factor are also given in appendix F. The enthalpy transfer to the air stream from equation (B.14) is B.4 (B.l6) (B.l8) (B.l9) (B.20) For a one-dimensional model of the cooling tower fill, where the available area for heat and mass transfer is the same at any horizontal section through the fill, the transfer area for a section dz is usually expressed as dA = aft AI' dz (B.17) where aft is the area density of the fill, i.e. the wetted area divided by the c orresponding volume ofthe fill and Afr is the corresponding frontal area or face area. Substitute equation (B.l7) into equation (RI6) and find dima hdaftAI, [(. .) ( \. ( )] --= Lef 'mallw -lrna + 1- Lef Jlv W$\II - Wdz rna To simplifY the analysis of an evaporative process Merkel [25ME 1) assumed that the evaporative loss is negligible, i.e. dw = 0 from equation (B.3), and that the Lewis factor is equal to unity. The governing equations (B.18) and (B.3) of the counterflow evaporative process simplifY respe ctively to dima _ hdafiAfr (. .) - 'masw -lmo dz rna and by dividing equation (B.3) by dz on both sides of equation (B.3) to dTw rna 1 dima --=----dz rnw Cpw dz Equations (B.19) and (B.20) describe respectively the change in the enthalpy of the air-water vapor mixture and the change in water temperature as the air travel distance changes. Equations (B.l9) and (B.20) can be combined to yield upon integration the Merkel equation, (B.21) where MeM is the Merkel number according to the Merkel approach. It is not possi ble to calculate the state of the air leaving the fill according to equation (R21). Merkel assumed th at the air leaving the fill is

saturated with water vapor. This assumption enables the air temperature leaving the fill to be calculated. Poppe and Rogener [9IPOI) did not make the simplifYing assumptions Merkel made. They derived the governing equation through the fill by following a different strategy than Merke l [25MEI). Whereas the governing equations (RI9) and (B.20) according to the Merkel theory describe the changes of the enthalpy of the air-water vapor mixture and ofwater temperature to the change of air travel distance (i.e. dim/dz and dT,/dz), Poppe and Rilgener [9IPOl) describe the change of the humidi ty ratio and the enthalpy of the air-water vapor mixture to the change of water temperature (i.e. dw/dTw and dim/dTw). Bourillot [83BOI) presented the governing Poppe equations as three equations des cribing the change of water temperature (dTw), air enthalpy (dlma) and humidity ratio (dw) to the chan ge of air travel distance (dz). The method of Poppe and Rilgener [9IPOI) is employed in the derivation of the governing equations in this study. B.5 Substitute equations (B.5) and (B.16) into equation (B.2) to find upon rearrange men~ mwdiw= hddA&masw -ima + (Lef -l)imasw -ima -(w", -w);,]-(w", -w}cpwTJ Find upon rearrangement of equation (B.3), (B.22) dw 1 dima 1 mw --=------dTw cpwTw dTw Tw ma or dw dima 1 mw --=----dTw Twdiw Tw ma (B.23) Substitute equations (B.16) and (B.22) into equation (B.23) and find upon rearra ngemen~ Substitute equation (B.24) into equation (B.23) and find upon rearrangemen~ From equations (B. I ) and (B.5) find hddA= madw w'" -w (B.24) (B.25) (B.26) Divide both sides by mwand introduce dTwldTwto the right h""d side of equation ( B.26) and integrate to find From equation (B.2?) find hdA = t a dwldTw dT w mw mw Wsw -w Equation (B.28) is defined as the Merkel number according to the Poppe approach i.e., Me = [ma dwldTw dTp w mw Wsw -w (B.2?) (B.28) (B.29) Upon substitution of equation (B.24) into equation (B.29) and differentiation of the latter with respect to

the water temperature, find dMep Cpw dTw == imasw -irna + (Le f -l}imasw -ima -(wsw -w~v]-{W.IW -w)cpwTw (B.30) The ratio of the mass flow rates, m w Im a , changes as the air moves towards the top of the fill. The change in the mass flow rate is determined by considering the control volume of a portion of the fill illustrated in figure B.3. B.6 ------------, , , , , , I , -, Figure B.3: Control volume ofthe fill. The varying water mass flow rate can be determined from the known inlet water ma ss flow rate, mw/. From the control volume in figure B.3 a mass balance will yield, (B.3I) Upon rearrangement of equation (B.3I) find, (B.32) From equations (B.15), (B.24), (B.25) and (B.32) the air outlet conditions in te rms of enthalpy and humidity ratio can be calculated. The preceding system of equations is only applicable for unsaturated air. In som e cases, the air can become saturated before it leaves the fill [98KRI]. Because the water temperatur e is still higher than the temperature of the air, the potential for heat and mass transfer still exists. U nder these conditions, the excess water vapor will condense as a mist. B.3 GOVERNING EQUATIONS FOR HEAT AND MASS TRANSFER IN FILL FOR SUPERSATURATED AIR The control volumes in figures B.l and B.2 are also applicable for supersaturate d air. Since the excess water vapor will condense as a mist, the enthalpy of supersaturated air is expre ssed by i" ~ cpa Ta + w,,(ijgwa + cp, T.) + (w - w,.)cl""Ta (B.33) where w" is the humidity ratio of saturated air at temperature Ta? Assume that the heat and mass transfer coefficients for supersaturated and unsat urated air are the same, as proposed by Bourillot [83BOI] and Poppe and RlIgener [9IPOl]. The driving potent ial for mass transfer is the humidity ratio difference between the saturated air at the air-water inte rface and the saturated free stream air, thus dmw= hd (w,w - w,.) <fA (B.34) (B.35) B.7 The enthalpy driving potential for supersaturated air can be obtained by subtrac ting equation (B.33) from equation (B.IO). By introducing, (w - wsa ) Cpw T." - (w - wsa) Cpw Tw + Wsa cpv Tw - Wsa Cpv Tw which add up to zero, into the resultant enthalpy differential, the temperature differential can be obtained

by manipulation. imasw -iss -{wsw -wso)iv +{w-wsa)cpwTwT w - T a = ..==--=-~....:;:.'----'----='----~--=-'-'----'cpmas where cpma, is the specific heat of supersaturated air per unit mass and defined as cpmas =Cpo +wsacpv +(w-wsa)cpw (B.36) Proceeding along the same lines as in the case of unsaturated air, using equatio ns (B.34) and (B.35) instead of equations (B.5) and (B.l3), find for supersaturated air, (B.37) (B.38) where the Lewis factor, Lef' is equal to h!hdc pma' ? The empirical relation of Bosnjakovic [65BOI] can be used to calculate the Lewis factor, which for supersaturated air is given by ( W~+0.622 I)0.667 WM +0.622 Lef = 0.865 ( )In w~ +0.622 WM +0.622 Substitute equations (B.34) and (B.37) into equation (B.2) to find upon rearrang ement, . _ _ [Lef~masw -i" -(wsw -w,a), +(W-W,a)cpwTw~ mwdlw- mwcpwdTw - hddA ( \; ( )c + Wsw -WsaPv - wsw -wsa pwTw (B.39) By introducing, limasw -i" -(wsw -w,a)i, +(w-w,a)cpwTw]-~masw -i" -(wsw -wsa)i, + (w-wsa)cpwTJ into the main parenthesis on right hand side of equation (B.39) the following eq uation yields after rearrangement. Substitute equation (B.34) into equation (B.l) and find upon rearrangemen~ hddA = madw (wsw -W,a) Substitute equation (BAI) into equation (BAO) to find upon rearrangement, (BAO) (B.41) B.8 (BA2) Substitute equation (BA2) into equation (B.23) and fmd upon rearrangement, dima mw --=c dT f'Ww rna cpwTw(wsw -Wsa)1+ -------;=-----!:::.....::-'----'7-~~~=__----. . (L l{imasw - i" - (w,w - w,a ~,] ( )c T lmasw -las + ef - +( _ )c T + W-Waw pw w W Wsa pw w (BA3) From equations (B,l) and (B.34) find rnadw (B.44) Divide both sides of equation (B.44) by mw, introduce dT,JdTw to the right hand side of equation (B.44) and integrate to find Equation (BA5) is defined as the Merkel number according to the Poppe approach i .e., Me p = hdA = frna dw/dTw dT w

mw mw wsw - waa (BA5) (BA6) Upon substitution of equation (BA2) into equation (BA6) and differentiation of t he latter with respect to water temperature, fmd ( { i -i -(w -w "] . _ . L -1 masw s,s .l'W sa lv (- )c T lmasw las + ef + (W _W )c T + W Wsw pw w sa pw w (BA7) From equations (B.32), (B.38), (B,42) and (BA3) the air outlet conditions in ter ms of enthalpy and humidity ratio can be calculated. BA. SOLVING THE SYSTEM OF DIFFERENTIAL EQUATIONS The fourth order Runge-Kutta method [83BOI, 92MAI, 97BUI] is used to solve the s ystem of differential equations for unsaturated and supersaturated air. The system of equ ations for unsaturated air (including saturated air) is represented by equations (B.24), (B.25) and (B.30). The system of equations for supersaturated air is represented by equations (BA2), (BA3) and (B.47). In t he equations that follow, ima must be replaced by i" for supersaturated air. Refer to the example problems in appendices G and I for a description of the conventions that is used, i.e. the conventions of the varia ble subscripts, fill intervals and fill levels. B.9 Equations (B.24), (B.25) and (B.30) for unsaturated and saturated air or equatio ns (B.42), (B.43) and (B.47) for supersaturated air can be respectively written as. (B.4S) (B.49) (B.50) (B.52) (B.51) (B.53) The fill is divided into one or more intervals with the same water temperature d ifference across each interval. In addition to the intervals, levels are specified (a level is an imag inary horizontal plane through the fill at the top and bottom of the fill and between two fill intervals). Init ial values of the variables, w, ima and Tw ' are required on a particular level, say level (n). The values of the variables can then be determined at level (n + I) with the aid ofequations (B.51) to (B.53). w(n+l} =w(n) + (/(n+I,I) + 2j(n+l,2) + 2j(n+I,3) + j(n+l,4J/6 ima(n+l) = ima(n) + (k(n+I,I) + 2k(n+l,2) + 2k(n+l,3) + k(n+I,4?);6 Mep(n+l) = Mep(n) + V(n+l,!) + 21(n+I,2) + 21(n+l, 3) +l(n+I,4) )/6 where J(n+l,I) = /';.Tw ? f(Tw(n) ,ima(n)' W(nJ k(n+l,I) = /';.Tw ? g(Tw(n) ,ima(nl' W(nJ l(n+l,l) = /';.Tw ' h(Tw(n) ,ima(n)' W(n?) . (/';.Tw . k(n+l,I) J(n+I,I) ))(n+l,2) = /';.Tw ' f Tw(n) +-2-' lma(n) +-2--' w en) +-2-( /';.Tw ? k(n+l,l) l(n+l,l) ) k(n+I,2) = /';.Tw . g Tw(n) + -2-,lma(n) +--2-' wen) +--2-

. (/';.Tw . k(n+I,2) 1(n+l,2) ) )(n+l.3) = /';.Tw ? f Tw(n) +Z,lma(n) + 2 ' wen) + 2 ( /';.Tw ? k(n+l,2) J(n+l,2) ) l(n+l,3) = /';.Tw ' h Tw(n) +-2-' lma(n) + 2 ' wen) + 2 (B.54) (B.56) (B.57) (B.5S) (B.59) (B.60) (B.61) (B.62) B.IO l(n+l,4) = !1Tw ? I(Tw(n) + !1Tw' ima(n) + k(n+!,3!' Wen) '+ 1(n+!.3)) k(n+!,4) =!1Tw ? g(Tw(n) + !1Tw' ima(n) + k(n+l,3)' wen) +1(n+l.3?) [(n+I,4) = fj,Tw . h(Tw(n) + fj,Tw, ima(n) + k(n+!,3)' wen) + 1(n+!,3?) where !iT. ; (T.; - T..,)/(Nurnber ojlntervals) (B.63) (B.64) (B.65) (B.66) The four variables in the Rwtge-Kutta method are T., W, lma or I" and Mep from t he left-hand side of equations (B.24), (B.25) and (B.30) for Illlsaturated air and equations (B.42), (B.43) and (B.47) for supersaturated air. For this reason equations (B.48) to (B.50) are functions of only w, I"", or I" and Tw? Most of the other variables are functions of these variables. Equations (B.48) t o (B.50) are not functions of Mep because dMep/dTw is a function of dw/dTw as can be seen from equation (B. 46). Thus, equations (B.24) and (B.25) for Illlsaturated air, or equations (B.42) and (B.43) for supe rsaturated air can be solved without equation (B.30) or equation (B.47) respectively. B.5 e-NTU MEmOD Jaber and Webb [89JAl] developed the equations necessary to apply the e-NTU meth od directly to counterflow or crossflow cooling towers. The approach is particularly useful in the latter case and simplifies the method of solution when compared to a more conventional numerical procedure as discussed in appendix C. KrOger [98KRI] gives a detailed derivation and implemen tation of the e-NTU method applied to evaporative air-water systems. It can be shown according to Jaber and Webb [89JAI] that _1JdA rna (B.67) Equation (B.67) corresponds to the heat exchanger e-NTU equation (B.68) Two possible cases of equation (B.67) can be considered where rna is greater or less than rn,.cpwl(dlm",JdTw)' The maximum of rna and rn,.cpwl(dl_JdT.) is denoted by C""" , and the minimum by Cm'n' The gradient of the saturated air enthalpy-temperature curve is _d_ima_ sw_ =",im",a",sw!.!.i_-_l.!!.?m""aswo=dTw TWi -Twa The fluid capacity rate ratio is defined as

The effectiveness is given by (B.69) (B.70) B.11 (B.71) where). is a correction factor, according to Berman [6IBEI], to improve the appr oximation of the imosw versus T. curve as a straight line. The correction factor, A, is given by (B.72) (B.73) (B.74) (B.75) where imoswm donates the enthalpy of saturated air at the mean water temperature . The number of transfer units for counterflow cooling towers is given by NTU O'_I_ln l-eC I-C I-e If m, is greater than m.cp.J(dim".ldT.) the Merkel number according to the e-NTU approach is given by c Me = Pw NTU , di""". /dT. Ifm, is less than m.cpwl(dimddT.) the Merkel number according to the e-NTU appro ach is given by mMe, O'-"NTU mw C.l APPENDIXC HEAT AND MASS TRANSFER IN CROSSFLOW WET-COOLING TOWERS C.IINTRODUCTlON In 1956, Zivi and Brand [56ZIl] extended the analysis of Merkel to the fill of c rossflow cooling towers. In 1976, Kelly [76KEI] used the model of Zivi and Brand [56ZIl] along with labor atory data to produce a volume of crossflow cooling tower characteristic curves to be used in graphica l solutions of cooling tower performance. The present analysis does not make the simplifying assumptions of Merkel and is also known, as in the case with counterflow towers, as the Poppe approach. A different approach is fol lowed in the derivation of the governing equations for crossflow cooling towers than was the case in app endix B for counterflow cooling towers. A more fundamental approach is followed to prevent confusion wit h sign conventions and partial derivatives because ofthe two dimensional nature of the problem. C.2 GOVERNING EQUATIONS FOR HEAT AND MASS TRANSFER IN FILL FOR UNSATURATED AIR Figure C.l shows a control volume in the fill of a crossflow wet-cooling tower. " z X x+At/J _ . imalx+nx --+---. /,/~" z+6.z t::./_'.? -------------" 1 z Figure C.1: Control volume of crossflow fill A mass balance ror the control volume in figure C.l yields.

(C.2) (C.I) C.2 G.k AxAy - G.!r+", AxAy + G. AyAz wi, - G. AyAz wk.", = 0 Divide equation (C. I) by AxAyAz and let Ax, Az ~ 0 oG.=_G Ow az ? ax The energy balance for the control volume ofthe fill in figure C.I is as follows : cp.(T. G.)k AxAy - cp.(T. G.)!r+", AxAy + G. AyAz im.k - G. AyAz im.!r+", = 0 (C .3) Divide equation (C.3) by AxAytiZ and let Ax, Az~ 0 and find after using the chai n rule ofdifferentiation, T oG. GoT. G aim. - 0Cpw .--+cpw w--+ .--az az ax Substitute equation (C.2) into equation (C.4) to find upon rearrangement, aT. = G. (T Ow __I_aim.) az G. ? ax Cpw ax The mass balance for the water stream in the control volume is expressed by G.k AxAy - G.k.", AxAy - hd aft (w,.-w)AxAytiZ = 0 where hd aft (w,.-w)AxAyAz is the amount ofwater evaporated in the control volum e in figure c.l. Divide equation (C.6) by AxAyAz and let Ax, Az~ 0 aG?=_G hdaft(w .-w) az ?G '" a Substitute equation (C.7) into equation (C.2) rearrange and find, Ow hdafl ax = G(w", -W) a The sensible heat transfer to the air stream in the control volume is ~xpressed by q,k AyAz - q,!r+", AyAz + h aft (T.-Ta)AxAyAz = 0 (CA) (C.S) (C.6) (C.7) (C.S) (C.9) where h aft (T. -Ta)AxAyAz is the amount of sensible heat transferred to the air stream in the control volume io figure C.l. Divide equation (C.9) by AxAyAz and let Ax, Az~ 0 (C.IO) The latent heat transfer to the air stream in the control volume is expressed by qmk AyAz - qm!r+", AyAz + i,G.k AxAy - i,G.k.", AxAy = 0 (C.II) Divide equation (C.lI) by AxAytiZ, let Ax, tiZ ~ 0 and substitute equation (C.S) into the resultant equation Oqm . oG. . h ( ) --=-1'1'--='''' daft W,rw-Wax oz An energy balance at the air/water interface ioside the control volume yields, aq oq, aqm -=-+-ox ax ax (C.12) (C.I3) (C.14) C.3 Substitute equation (B.13) into equation (C.10). Substitute the resultant equati on and equation (C.12) into

equation (C.B) to find upon rearrangement, : =hdafi [ hh (ima", -imJ +(I- hh )i,(W", -W)] cpma d cpma d h --- is the Lewis factor, Lef. The Lewis factor for unsaturated air, according to Bosnjakovic [6SBOI] cpmahd is given by equation (B. 15). The enthalpy transfer to the air stream from equation (C.14) is 8ima 1 8q hdafi [. . (L 1\(' . . ( )}] --=--=--fmo.sw -'rna + ef - fl'masw -lrna -Zv Wsw-wax Ga ax Ga Substitute equations (C.8) and (C.IS) into equation (C.S) to find upon rearrange ment, 8Tw 1 Ga hdafi [(W'" -w)cpwTw -(ima", -ima )- J 8z = cpw Gw ----c;: (Le f -1:!ima,w -ima -(w", -w},] (C.IS) (C.16) Thus, the system of equations to be solved for unsaturated air fur the crossflow fill are equations (C.7), (C.8), (C.IS) and (C.16). C.3 GOVERNING EQUATIONS FOR HEAT AND MASS TRANSFER IN FILL FOR SUPERSATURATED AIR The governing equations fur supersaturated air can be manipulated as was done fo r the unsaturated case using the same arguments as in the counterflow case to obtain the following equa tions for supersaturated air in crossflow, 8Gw = -hdafi(W'" - W,a) 8z aw hdafi ax =G(W'" -W,a) a 8i" =_1_8q = hdafi [ima,.. -i" + (Le f -1)(ima", -i" -i,(w", -W,a)}+] ax Ga ax Ga LefcpwTw(w-w,J 8Tw=_I_Ga hdafi [(W'" -W,~)cpwT~ -(ima", -i,,)-.LefCpwTw(W-w,J] az Cpw Gw Ga (Le f -1)lma", -I" -(W", -W,a),] (C.17) (C.18) (C.19) (C.20) C.4 C.4 SOLVING THE SYSTEM OF PARTIAL DIFFERENTIAL EQUATIONS Figure C.2 illustrates an example of a grid of a crossflow fill that is divided into four intervals in both the vertical and horizontal directions. 2,5) 3,5) 4,5) + -.;,.i 1,5)(1,1) (1,2) (1,3) (1,4) ( (2,1) (2,2) (2,3) (2,4) ( (3,1) (3,2) (3,3) (3,4) ( (4,1) (4,2) (4,3) (4,4) ( Air inlet side --.ttl (5,2) (5,3) (5,4) (5,5) Figure C.2: Example ofa crossflow fill that is divided into four intervals in ea ch direction. To simplify the solution process of the governing' equations the fill dimensions can be non?

dimensionalized. Poppe and Rogener [91POI] presen,ted the governing equations fo r crossflow fills in non-dimensional form. Thus, in non-dimensional form the fill can be analyzed wit hout any reference to fill dimensions. All the governing equations are of the first order. These first derivatives can be approximated by first? order rearward finite difference expressions. An example of the application of t his finite difference technique to first derivatives can be seen in figure C.3. (. First-order rearward difference with respect tox (:;L = './ U;,; ? ?1/.14 I!ix ) t:.x (+) j.l,) i,) Figure C.3: An example of a first derivative approximated as a first-order rearw ard finite difference with respect to x for an arbitrary variable u [95ANI]. C.S Where the fill dimensions are non-dimensional, equations (C.7), (C.8), (C.l5) an d (C.16) respectively become OG hdufl -a-=-Gaa(w", -w) '7 a Ow hdufl 8.g =o:-(W", -w) 8i h U-..!!!?..=~&ma", -ima +(Lef -IXima", -ima -i,(w,w -w)}]8'; Ga 8Tw 1 Ga hdufl [(W'" -w)cpwTw-(ima", -ima )- J 8'7 = cpw Gw 0:- (Lef -1)ima", -ima -(w", -w)i,] where .g =x/Lx and '7 =z/L, with L, and L, the fill lengths in the xand z direct ions respectively. (C.21) (C.22) (C.23) (C.24) Figure C.4 illustrates an excerpt of four grid points from the conputational gri d in figure C.2 for generalized non-dimentional coordinates. It is essential that the fill is divide d into equal intervals in both the horizontal and vertical directions fur the non-dimensional fill analysis and thus is 1i.'7 = Ii..g . '7 (i-l ,j), Point b Figure C.4: Four generalized grid points ofone cell ofa crossflow fill. (i ,j-l) Point a (i ,J) By applying first-order rearward differences and letting Me~ =hduflli.I;/Ga = hd uflli.'7/Ga, equations (C.7), (C.8), (C.l5) and (C.16) respectively become GWO.j) = GW(;.}_l)-GaMe~(w", -W)!a (C.25)

w(;,}) =WO-1,}) +Me~(w", -wt (C.26) 1 Ga [(W'" -w)cpwTw-(ima", -ima )- J Two.}) = TW(',J-') +-;;--GMe~ (Le -1'Ii _ i _ (w _w)i ] (C.28) pw IN f,{ musw rna sw 11 a C.6 The la and I. symbols in the last terms in equations (C.25) to (C.28) refer to p oint a and b respectively in figure CA. Point a refers to the average value ofthe last term of equation (C.25 ) or (C.28) between points (i,j) and (i,j-l) while point b refers to the average value of the last term of equation (C.26) or (C.27) between points (i,j) and (i-I, j). Take for example the average value of the las t term of equation (C.28) between points (i,j) and (i,j-I), i.e., _1_ Ga Me" Kw", - w)cpwTw - (ima", - ima ) - (Lef -l!;ma", - ima - (w", - w)i,~ =Cpw Gw a (C.29) where Ga and Me!; are constant throughout the solution domain. Equation (C.29) c an be substituted into equation (C.28) to obtain the value of TW(/J). Equations (C.25) to (C.27) are tr eated in a similar manner to obtain average values for the last terms ofthese equations. The governing partial differential equations are solved by an iterative techniqu e. Gw and Tw are known at the water inlet side. i.a and ware known at the air inlet side. Ga is constant t hroughout the solution domain. Equations (C.25) and (C.28) are used to solve respectively for Gw and Tw at the air inlet side while equations (C.26) and (C.27) are used to solve for w and i.a at the water i nlet side. Equations (C.25) , to (C.28) can be solved simultaneously throughout the rest ofthe domain. All of the other variables in equations (C.25) to (C.28) are functions ofTw, Gw, i.a and w. If the air is supersaturated at a point in the fill, the governing equations for supersaturated air must be solved instead of the equations for unsaturated air. The mean water outlet temperature can be obtained by integrating the water tempe rature values at the water outlet side of the fill, i.e., 1 f'Twom =- Two de; n" where n!; is the number of fill intervals in the ~or x direction (C.30) The mean outlet air enthalpy and humidity can be obtained by integrating these v alues at the air outlet side of the fill, i.e., . 1 f" dlmaom = - lmoo 1] n" (C.3I) C.? w =_1 !'w dTfam a n" where n. is the number offill intervals in the TJ or z directioo (C.32) Me~ in equations (C.25) to (C.28) can be referred to as the local M...kel number according to the air stream in the horizontal direction where (C.33)

(C.34) At every point in the solution domain the local Merkel number according to the w ater stream is determined by Me (" ") = Go Me, "'.J G 'w(i,j) The Merkel number for the fill, Me, is obtained by integrating Me"JJJ across the entire fill. Firstly, determine the average ofthe Me'l(iJJ quantities at the cent... ofeach cell of th e entire fill. The mean Merkel number, Me .,m(iJ), at the cell center is calculated from figure C.5 as follows, MeomU,j) = (Me,u,j) + Me"(,+"j) + Me"(i,j+1) + Me"U+l,j+l))j4 (C.35) Me'lli-l,j-I) Me"",(, ,J) ?Tf Figure C.5: Average value ofMe'l(tJ) at the cell center. The mean quantity of all the Me.,m(;';) values is donated by Memwh...e Me = L,Me"m(i,j) m n~n" The Merkel number of the fill is given by (C.36) (C.37) e.8 The Merkel number for a crossflow fill is determined from experimental data by t he following approach. A value for Me!; is guessed. This value is constant throughout the computational domain. The water outlet temperature is determined by equation (C.30) after the governing equations have converged. Me!; is varied until the water outlet temperature from equation (C.30) matches the known water outlet temperature. The Merkel number is then determined by equation (C.37). (0.2) f".'.::., D.I APPENDIXD LOSS COEFFICIENTS AND TRANSFER CHARACTERISTICS 0.1 INTRODUCTION The loss coefficients and transfer characteristics that are employed in the perf ormance evaluation of wet? cooling towers are presented; Most of this section is abridged from KrOger [98KR l) where a detailed derivation and presentation of the information presented in this section can be found. 0.2 LOSS COEFFICIENTS Frictional resistance, abrupt changes in cross section, inlets and outlets, amon gst others, reduce the "mechanical energy" between any two sections of a duct. The mechanical energy is converted to thermal energy. The "mechanical energy" refers to the pip + aev'/2 terms in the equation of the first law of thermodynamics, i.e., p+Q=m[(u, +;: +a.;v; +gz,)-(u1+;: +a.;vl' +gZI)] (0.1) The subscripts I and 2 refer to a control volume between sections I and 2 respec tively. P and Q respectively represent the power and the rate ofheat input into the fluid. A dimensionless loss coefficient, also referred to as the total pressure loss co efficient, can in general be

defined between two cross-sections in a horizontal duct as (12+ a'lv~)_(p, + a.,v;)PI 2 P, 2K = 'O":""'-----"7--~----"? v'/2 where v is usually based on conditions at either section I or 2. For incompressi ble and uniforms flow with ae"' I, equation (0.2) can be written as K = Pll - P" pv'/2 Pli - P" (:)' 1(2p) (0.3) where Ptt and pa are the total pressures at sections I and 2 respectively. To simpli/)' the solution process the loss coefficients are usually referred to the mean fill conditions by using the principle of the conservation of mass. Referred to the same conditions , the values of the loss coefficients can then be added. The relative magnitudes of the loss coefficients can then be observed. For example, refer the loss coefficient, at one set of conditions, to another set of conditions. These conditions are denoted by subscripts I and 2 in the following discussion. If the pressure d rop is equal at both sets of conditions, find from equation (0.3), D.2 (D.4) It follows from equation (D.4) that the loss coefficient K, referred to the cond itions at I, is equal to (0.5) Loss coefficients are usually expressed by, empirical relations that are obtaine d from numerical or experimental work, or, directly as a value. 0.2.1 SPRAY REGION Data presented by Cale [82CAI] suggests that the loss coefficient in the spray z one may be expressed approximately as (0.6) 0.2.2 DRIFT ELIMINATOR, INLET LOUVERS The loss coefficient of a drift eliminator is determined experimentally for each eliminator. A typical empirical relation for a drift eliminator loss coefficient is K - Rybd?de - ade de 0.2.3 WATER DISTRIBUTION - ! (0.7) Krilger [98KRI] gives an approximate for a water distribution system loss coeffi cient as Kwd = 0.5. 0.2.4 RAIN ZONE According to De Villiers and Krilger [99DEI], the loss coefficient for a circula r rain zone is given by 0.2246 - 0.31467 app. + 5263 .04a.,u. + 0.775526 {1.4824163 exp(71.52aL d d )- 0.91} K~ =3a,vw (H,ld d x{0.39064exp(0.010912aL d,)-O.17} x ~.0892(a, v",. t'3944 + 0.14} [ {0.84491n(aA 12)- 2.312} l xexp x {0.37241n(a,v",.)+0.7263} x 1n ~06.757(aLH;)-2.83" + 0.43} where 6{ _4 9 )0.25 a" =3.061xlO- IF g law

(0.8) D.3 ap = 9981 Pw ( 5 3 3 )0.25a, =73.298 g CTw1Pw The equation is valid under the following conditions: O?C "To " 40?C; 10?C" Tw " 40?C; 0.927 kg/m3 "Po " 1.289 kg/m3 992.3 kg/m3 "p"" 1000 kg/m3; 1.717xIO?' kg/ms"/l." 1.92xl0?' kg/ms 0.0696 N/m " 0;, " 0.0742 N/m; 0.002 m " d.J S 0.008 m; 9.7m/s' S g S 10 mis' 30 m " d,/2 S 70 m; 4 m " II, S 12 m; 0.00075 m/s S Vw S 0.003 m/s; 1 mis " V"'O S 3 mis The loss coefficient for a rectangular rain zone is given by 0.219164- 0.30487ap Pa +8278.7ap #a + 0.954153{0.328467 exp(135.7638aLdd)+ 0.47} Kn = 1.Sa, VW (H, 1dd)x x ~6.28482(aLH, )"",.95729 +0.56) { 1n(0.204814exp(0.066518aLw,)+ 0.21)} x exp x (3.9186exp(- O.3aL H,)) x (0.310951n(aLdd)+ 2.63745) x ~.177546(a,v",0)""1.46"1 + 0.21) where the range of applicability is the same as for the circular tower except th at 1 m/s :s; v"o :s; 5 mis; 2 m S H, " 8 m; 4 m S W, " 40 m (0.9) (0.10) D.2.5 TOWER INLET For a round counterflow cooling tower with an isotropic fill (e.g. splash or tri ckle type fill) operating in the absence of a rain zone, the loss coefficient is according to De Villiers and KrOger [99DEl]. K,,(.on) =0.011266exp(0.093d, 1H, )K~ -0.3105exp(0.1085d, 1H,)K fi -1.7522 + 4.5614exp(0.131d, /HJ [{ (I0970.2 exp(- 0.2442Kfi )+1391.3)/(d, 1H, -15.7258)1] + sin-1 +1205.54exp(--0.23Kfi)+ 109.314 J x {2r, 1d, -0.01942/(d, I H, - 27.929)-O.OI6866} which is valid for 7.5" d,IH,,, 15,5 S Kfi S 25 and 0" r;lli,,, 0.02. Kfi in thi s case is the sum of the loss coefficients in the vicinity of the fill. This value must be multiplied by the correction factor en as given by equation ( 0.11) to obtain the correct inlet loss coefficient in the presence of a rain zone. D.4 c = [O.2394+80.l{O.0954/(d, I H, )+dd }exp(O.395G. IGJ ] " -0.3195(G.IG.}-966{dd I(d, / H,)}exp(O.686Gw/GJ x{1- O.06825Gw)K~09667exp{s.7434(1/ d, - O.Ol)} This correction factor is valid in the range 7.5 ~ d,lH, ~ 20, 5 ~ Kfi ~ 25, 3 ~ d.i ~ 6 mm, 1 ~ G. ~ 3 kg/m2s, 1.2 ~ Ga ~ 3.6 kg/m2s and 80 ~ d, ~ 120 m. The tower inlet loss coefficient in the presence ofa rain zone is given by K" =CnK,,(nan) (D.l1 ) (D.l2) (0.13) (D.l4) (D.l5) The inlet loss coefficient in isotropically packed induced draft rectangular tow ers is according to De Villiers and Kroger [99DE1], K"(n<"') =O.2339+~.919xlO?3 K~ -6.840xlO?2 Kft +2.5267) xexp{ ;: (0.5143-0.l803xexp~.0163Kft })}

-sinh .1[2.77 xexp{0.958 ;:}xexp{K ft(2.457 -1.015 ;: }W2 }x(~, -0.013028}] De Villiers and KrOger [99DEI] states that it becomes acceptable to ignore the i nlet loss correction factor for the rain zone in cases where W, / H, :;; 3. In this case, W; / H, = 3, which means that K" = K"(nan)' Where the correction factor for a rectangular tower is needed, the following empirical ! correlation provides the required value, Cn = I-Gw[ O.l23-12.ldd - 272.26d; + 5.04x 10. 4 x eXP{0.466~ }] x ~ -1.16 X 10-3 x exp{GJ) and is valid for 3 ,.:; WilH, ~ 7.5 m, 3 ,.:;d~ 6 mm, I ~ Gw~ 3 kg/m's and 2 ~ G a ~ 6 kg/m's. This equation can only be used with any degree of confidence at high W,IHi values and since th is is not normally the case, it becomes prudent to take the conservative approach by ignoring the influ ence of the rain zone loss on the inlet loss. D.2.6 TOWER SUPPORTS The loss coefficient due to the tower supports, based on the drag coefficient of the particular support geometry is given approximately as Cd"Llsdlsn"A~K - -""~--=-:=C-"-Isfi - (nd,H,r D.2.7 EXPANSION LOSSES The expansion loss coefficient after the fil~ referred to the mean conditions th rough the fill, is given by K,,, = (1- 0",)2 where G. =A,JA, is the expansion area ratio. D.5 (D.16) D.2.8 FILL LOSSES The losses through the fill are usually expressed as an empirical relation ofone ofthe following forms. K - L G b'G "fdm-apjiwa where up, bd and Cd are coefficients specified for each fill. The actual fill loss coefficient applicable to the cooling tower is then given b y K-K +(G;,o G;,,)/G;,mfi - fdm ---- -PaVD Pavi Pavm where IIp.,m ~ O.5(I/p."+1Ip.,,,) and G,,,,, = (G,.+G,,,,)12 D.2.9 OTHER LOSS COEFFICIENTS (D.17) (D.l8) (D.l9) In addition to the above mentioned losses are there also other losses, for examp le, contraction losses and losses due to the fill supports. In mechanical draft cooling towers there are al so fan upstream and downstream losses, plenum losses and diffuser losses. The loss coefficients for these cases can either be specified or obtained from empirical relations in the literature. D.3 TRANSFER CHARACTERISTICS The total transfer characteristic of a wet-cooling tower consists of the transfe r characteristics for the fill, spray zone and rain zone. D.3.l RAIN ZONE The transfer characteristic or Merkel number in the rain zone of a circular cool

ing tower is given by [97DEl], l2(~XH,X P. )Sc?.33 [ln(W' +O.622)/(W, +O.622)JX v.rod, d, PwR,T. w+O.622 0.90757a pPa - 30341.04ap.u. - 0.37564 b.55 + 41.7215(aL dd )080043 lb.713 + 3.741(aL H ,tl.23416 } +4.0401 x {3.llexp(0.15a,v~J-3.13} [ {5.375gexp(- 0.2092a LH,)} ] x exp x In{0.371gexp(0.0019055aL d,)+0.55} (D.20) where the range of applicability is the same as for K" in equation (D.8). D is g iven by equation (D.21) where M, ~ 28.97 and V, ~ 29.9 fur air, while for water vapor V, ~ 18.8 andM, ~ 18.016. 0,6 D=O,04357TL5 (11M. +1/MJ" p[V.O'''3 + V.O. 333r The equation for the rain zone Merkel number in a rectangular tower is, hdnanH, 3.6(---.!2-YH, Y P. JScO'''[ln(W. +O.622)/(W, -W)] Gw v~.dd )ldd JlPwR.Ta w+O.622 (0.21) 4.68851ap Pa -187128.7a"Jla - 2.29322 + 22.4121~.350396(avvMOY'''046 +O.09~.60934(aLHJI.l2083 + O.66} (0.22) x x ~4.6765(aLdd)'732448 +0.45} { { O.o87498exp(o.o26619aLW:)}} xexp 7.738gexp(-O.399827aLH.)1n I , +0.85 The range of applicability for equation (0.22) is the same as that for equation (0.9). 0.3.3 SPRAY ZONE The data of Lowe and Christie [61LOI] can be correlated to give ( ) 0.5 O.2L,p ~: (0.23) D.3.4FlLL The transfer characteristic of the fill is usnally expressed as an empirical rel ation of one of the following forms. (O.24) haL dft fi ft = a L {G I G \b,G d fi~ w aJ w where ad, bd and Cd are coefficients specified for each fill. (0.25) E.l APPENDIXE EFFECT OF ATMOSPHERIC CONDITIONS ON THE OPERATION OF COOLING TOWERS E.! INTRODUCTION The atmospheric conditions prevailing in the region of a cooling tower affect th e operation and perfurmance of a cooling tower. An atmospheric temperature inversion, for exampl e, reduces the performance of cooling towers. This is because the effective temperature of the air entering the cooling tower is higher than during conditions where the adiabatic lapse rate prevails, and the potential driving force or pressure differential is less. The formulas for calculating the pressur

e differential and the approximate effective air inlet temperature are derived in this appendix for var ious atmospheric temperature and humidity profiles. E.2 EFFECT ON TOWER DRAFT In it's simplest form the draft equation ofa cooling tower can be expressed as tip, - tip; = r.K fA"12 (E.l) where tip, is the pressure differential outside the tower and tip; is the corres ponding pressure differential inside the tower. The effect of the atmosphere on the draft equation will be evi dent from the pressure difference external to the tower, tip" where tip, is equal to tip, =PI - p, (E.2) where P, and p, refer to the atmospheric pressure at points! and 7 respectively shown in figure E.!. Point 6 is on the inside ofthe tower shell at the same elevation as point 7. 6 Figure E.!: Natural draft cooling tower with external points! and 7. The pressure gradient in a gravity field is given by : =-p~g The density ofmixtures of air and water vapor is given by equation (A.3.1), (E.3) P =(I+w{1 w )1., '\ w+0.62198 RT E.2 ,.ij.622 (w+ I) P (w+0.622)RT (E.4) Substitute equation (E.4) into equation (E.3) and find after rearrangement dp _ 0.622(w+l) g dz p-- (w+0.622) R T (E.5) (E.6) The pressure difference between ground level and an arbitrary elevation z can be obtained by integrating equation (E.5) between these two points. After rearrangement find = ex ( 0.622? g r (w+ I) dzJ p p, P R (w+0.622) T Equation (E.6) can be solved if the humidity and temperature profiles as a funct ion of the altitude, z, are known. Refer to appendix M for a detailed discussion on atmospheric humidity. Re fer to appendix L for a detailed discussion on the temperature profile during nocturnal inversions. Kroger [98KRI] shows that temperature distribution for moist air of constant hum idity is, T = T, _ 0.00975(1+ w) z (1 +1.9w) (E.7) (E.8) Assume that the humidity ratio is constant at w, and substitute equation (E.7) i nto equation (E.6) to find upon integration between ground level and H, with g ~ 9.8 mis' and R ~ 287.08 Il kg K that 2.1778(I+l.9w\} ( 0.00975(1 + w)H ) ',+0.62'98P, = p, I ' ,(I + 1.9w, )T, The pressure difference between points 1 and 7 from equation (E.8) is, _ _ [1-(1- 0.00975(I+W')H,)2~1+~';~;:;')] p, P, - PI (I +1.9w, )T, (E.9) If it is assumed that the air is dry, with respect to both the temperature and t he humidity profiles, i.e. w, = 0, equation (E.8) can be simplified to give ( )"P, =p,\1-0.00975 H, IT,

The corresponding pressure difference between points I and 7 from equation (E.l0 ) is, P,-P, =P,~-(I-O.00975 H, IT,)" j (E.IO) (E.11) The pressure difference between points 1 and 7 in figure E.I can also be derived if it is assumed that the air temperature profile corresponds to a dry adiabatic lapse rate and that the a tmospheric humidity is constant at WI. The temperature profile for a dry adiabatic lapse rate is given by T = 1; - 0.00975z (E. 12) Substitute w, and equation (E.12) into equation (E.6). After integration between ground level and H" and after rearrangement find E.3 2.177S(1+wl) P, = PI (1- 0.00975 H, 11;) ,,+0.62198 The corresponding pressure difference external to the tower is, [ '1778(1+"1] PI - P, = PI 1-(1-0.00975 H,ITJ,,+0.'2198 The temperature profile in a temperature inversion, discussed in appendix L, can be expressed as T = (T, +273.15{:.J where T, is in ?C and the exponent, b, is given by b = 0.0035sin(0.0177 ?n. -2.32392)+0.0065 (E. 13) (E.l4) (E.l5) (E.l6) where nd is the number of the day of the year (nd = I on the first of January). Equation (E.16) is developed in appendix L from experimental measurements. If it is assumed that the humidity ratio is constant, substitute WI and equation (E.l5) into equation (E.6). After integration between ground level and H, find, [ ( + I) b H I- b ] = ex -0021232 WI z, 6 P, PI p. (WI +0.622XT, + 273.15Xl-b) (E.l7) If the height ofthe inversion, Zit, is higher than the tower height, H6, then th e pressure difference between points I and 7 in figure E.I is ( [ (w + l)zbH I- b ]J _ _ I-ex -0021232 " 6 PI P, - P, p. (WI +0.622XT, +273.15Xl-b) (?.18) Ifthe top ofthe inversion is lower than the tower height then the pressure at th e inversion top is given by [ ( l)zb I-b ] = ex -0021232 WI + ,zit Pit PI p. (w, +0.622XT,+273.15Xl-b) where PI/ is the pressure at the inversion top at elevation ZI/. (E.l9) Assuming a constant humidity ratio WI for both the temperature and the humidity profiles from the top of

the inversion to an elevation corresponding to the top ofthe cooling tower, find the pressure at this latter elevation 2.1778(1+1.9wl ) = (1 0.00975(1 + WI )(H6 - Z,,)) .,+0.'2198 P, Pit (l+1.9wl )T" where Pit is given by equation (E.19) and T;, is given from equation (E.15) by ( ) b Z., T" = (T, + 273.15) -:: where T, is in ?C. (E.20) (E.21) (E.22) E.4 Zit in the equations above is still unknown and can be determined by referring t o appendix L. If the humidity is not assumed to be constant or zero, but expressed as a function of t he height above ground level, then equation (E.6) has generally to be solved by numerical integration t echniques. E.3 EFFECT ON THE EFFECTIVE AIR INLET TEMPERATURE The height of the air drawn into a cooling tower, 8,., is constant at radial dis tances not close to the cooling tower, as shown in figure E.2. Refer to appendix N for a discussion on this stat ement. The temperature of the air flowing from below the inversion top, Zit, into the c ooling tower taking into consideration adiabatic compression is given by T", = (T,. + 273 .IS{:,J+ O.00975{I - ~:) where H, and H, are shown in figure E.2. ~ ~ H, rnav L ~ H3 ~ (E.23) (E.24) Figure E.2: Illustration of constant approach height to a cooling tower. The air originally from a region above the inversion top enters the cooling towe r at Tail =T" -O.00975(Z- Zil)+O.0097SZ(I- ~:) The last term in equation (E.22) and (E.23) accounts for the heating of the air due to adiabatic compression. The mean effective inlet temperature to the cooling tower is thus the integral o f equation (E.22) and equation (E.23) up to a height of H, ifZit < H,. Ta'm =7<r.";,ra.,)dz =lr, + 273.1S{:,J +O.0097SZ(I- ~:)]~ +~![T" -O.00975(z - zJ+O.0097SZ(I- ~:J];' Substitute equation (E.21) into equation (E.24) and find after integration and r earrangement, E.S T ;(T +273.1Sfz")'[-'_(!.L)+H,-zItJ+o.oo97Iz _Z;' _H,] aim r '\ Zr b+ 1 HI' HI' 1 1t 2H r 2 IfZit> H, then equation (E.24) reduces to Taim ; lr."dz; HJ[(T, + 273.15)(3-)' + O.00975Z(I- H,)] dz

o H, 0 z,. H r HI' After integration of equation (E.26), find T.'m; (1; +273.1/H,)'(_I_)+O.00975(H'X1- H3 ) \ z, b+1 2 H, (E.25) (E.26) (E.27) E.4 CONCLUSION Equations are derived that predict the effects of atmospheric temperature and hu midity on the draft through natural draft cooling towers. Equations are also derived that determine the temperature at the inlet ofcooling towers during nocturnal temperature inversions. F.I APPENDIXF LEWIS FACTOR F,llNTRODUCTION It can be seen in appendix B that the Lewis factor, Left appears in the governin g equations of the heat and mass transfer processes in a wet-cooling tower. Merkel assumed that the Lewis fa ctor is equal to I to simplify the governing equations while Poppe used the equation of Bosnjakovic [6 5B01] to express the Lewis fuctor in his more rigorous approach. The Lewis fuctor and its relation to the Lewis number are investigated in this appendix. F.2 LEWIS NUMBER The rate equation for momentum transfer is given by Newton's law ofviSCOSity, i. e., F Ov B(pv) -=-p-=-v-A 0' 0' The rate equation for heat or energy transfer is given by Fourier's law ofheat c onduction, Q = _k_BT =-a_B(,-pc~pT_) A 0' 0' (F.I) (F.2) (F.3) The rate equation for mass transfer is given bY Fick's law of diffusion, ie., m iJc -=-DA 0' The coefficients v, a and D in equations(F.I), (F.2) and (F.3) respectively have dimensions of [L'/T]. Any ratio of two of these coefficients will result in a dimensionless number. In systems undergoing simultaneous convective heat and momentum transfer, the ratio of v to a would be of importance and is defined as the Prandtl number, i.e., v C J1Pr =-=-pa k (FA) In processes involving simuhaneous momentwn and mass transfer the Schmidt number is defined as the ratio of v to D, i.e., v cpSc=-=-P? a k (F.5) In processes involving simultaneous convective heat and mass transfer, the ratio

of a to D is defined as the Lewis number, i.e., (F.6) From equation (F.6) can it be seen that the Lewis number is equal to the ratio o f the Schmidt to the Prandtl number and is relevant to simultaneous convective heat and mass transfer . The relative rate of (F.7) F.2 growth of the thermal and concentration boundary layers are determined by the Le wis number. The temperature and concentration profiles will coincide when Le ~ 1. Mills [95MII] states that other definitions of the Lewis number are found in the literature, for example, the ra tio of the Prandtl to the Schmidt number and the ratio of the heat to mass transfer conductance. The values of k, p and cp in equation (F.6) can be determined by the equations i n appendix A. According to Mills [95MII] the diffusion coefficient, D, for air-water vapor mixtures can be given by D ~ 1.97 XIO-S( ~ )(;,)"" where Po = 101325 Pa and To ~ 256 K. Equation (F.7) is valid under the following condition, 273 K < T < 373 K. F.3 LEWIS FACTOR In addition to the Lewis number the Lewis factor can be defined. In some referen ces the Lewis factor is referred to as the Lewis relation [9IFEI, 95MII, 99HAI]. The Lewis factor is an indication of the relative rates of heat and mass transfer in an evaporative process. In some of the litera ture encountered there seems to be confusion about the definitions of these dimensionless numbers and t he Lewis factor is often incorrectly referred to as the Lewis number. The Lewis factor, Left is equal to the ratio of the heat transfer Stanton number , St, to the mass transfer Stanton number, Stm where Nu h St ~--~-RePr pvc p St ~~~~ m ReSc pv The Lewis factor can be obtained by dividing equation (F .8) by equation (F.9), i.e., St h pv hLei ~-~-_._=-Stm pvcp hd cphd (F.8) (F.9) (F. 10) Lewis [22LEI] tried to prove analytically that Lei = I for gas/liquid systems. I n a later article Lewis [33LEl] stated that the relation, Lei = I, holds approximately for air/water mix tures but not for all mixtures of liquid and gas. Although the proof given by Lewis was incorrect [88D R1] the ratio h/cphd is today known as the Lewis factor. In chemical engineering practice, the analogy between convective heat and mass t ransfer is widely used in a form recommended by Chilton and Colburn in 1934, namely, F.3

~: =(~~r~ The Chilton-Colburn relation is of adequate accuracy for most external forced fl ows [95MII]. (F.II) Equation (F.11) is obtained for laminar forced flow from the Chilton-Colburn ana logy power law relations, i.e., J 2 St=C.Re"Pr" =_h_ pvcp .~ .~ h St =C?Re'Sc'=..L. m pv Thus, (F.12) (F.B) (F.14)Lef =(~:J~ =Le~ Bourillot [83B02] states that the Lewis number is not constant and is tied to th e nature of the vapor-gas mixture. It also depends on the nature of the boundary layer near the exchange s urfaces and the thermodynamic state of the mixture [83B02, 94GRI]. Bosnjakovic pointed out that the mass transfer is not proportional to the difference (wsw - w). A corrector term, F(~, is applied to equation (F.l4) and the expression for'Lef in the Bosnjakovic form is obtained. ~ I Le = Le' -- (F.15) f F(~) where F(~) = In~ and ~ = w", +0.622 ~-1 w+0.622 Poppe and Rllgener cited that the Lewis factor, Lef, is according to the Bosnjak ivic form, 2 w'" + 0.622 1 Lef = 0.865' w + 0.622 m(w", +0.622) w+0.622 (F.16) where the Lewis number, Le, is taken constant at 0.865. Bourillot [83BOI] and Gr ange [94GR1] state that the Lewis factor for a wet-cooling tower, using equation (F.16), is approximatel y 0.92. Merkel [25MEI] assumed in his classical work on evaporation that Lef = 1. HlIszl er [99HA1] cited that other researchers showed that the assumption of Merkel is not correct and that a ll ofthe researchers find Lewis factors in the range from 0.6 to 1.3. An analysis of both splash and film packings by Feltzin and Benton [91FEI] indicates that for counterflow towers a Lewis factor of 1.25 is m ore appropriate. According to Feltzin and Benton [9IFE1] the Lewis number does not appear to be d ependent on whether the packing is splash type or film type, but only on the configuration (I.e. cou nterflow or crossflow). Sutherland [83SUI] used a Lewis factor of 0.9 in his "accurate" tower analysis. Osterle [910SI] FA developed a wet-cooling tower model that corrects the Merkel [25ME 1] assumption so that the mass of

water lost by evaporation is accounted for. However, he still assumes that the L ewis factor is equal to unity. (p.17) Hliszler [99HAI] states that when the humidity potential (w,w - w) is large, equ ation (F.J4) is not valid any more. Figure F.J illustrates the Lewis factor as a function of the dimension less vapor pressure gradient, Pg, defined by equation (F.J 1). The dimensionless vapor pressure grad ient is a measure of the degree of supersaturation in the boundary layer. If Pg < I then there is no mist in the boundary layer. If Pg> I then mist is present in the boundary layer. H~szler [99HAI] gives a detail ed account for the derivation of the dimensionless vapor pressure gradient, i.e., ( dP,) (I + O.622wo) T,o - T,Pg= T, 0 O.622p, wo-w where the subscript 0 refers to the water film surface. 1,81,61,41,20,80,6 0 Re",SOOO A Re=8500 V Ae=9000 ~~r 0 ? Rea9500?- Re=100004 ~ Re=10800 <> ..... 'A .,. ' 7 .0 ~b~( '/J~( 0.50 0,4 1,50 1,0 Pg Figure F.I: The Lewis factor, Lef> as a function of the dimensionless vapor pres sure gradient, Pg. Le = ... 1,25 'll ~ ;; Q -u 1,00 .:! ~... ~ ..;l 0,75 0.82, Twm = 40?C [99HAI]. The average water temperature for the data in figure F.I is 40?C with Le = 0.82. With Le = 0.82 in equation (F.J4) Lef = 0.88. The discrepancy between equation (F.14) and the data in figure F.J is an indication that equation (F.14) is not valid for all conditions. G.l APPENDIXG COUNTERFLOW FILL ANALYSIS ACCORDING TO THE POPPE APPROACH Oosthuizen [9500I] and Krllger [98KRI] present a sample calculation where the tr ansfer coefficient of an expanded metal fill is evaluated while employing the Merkel method with Cheby shev numerical

integration. Baard [98BA1] presents a sample calculation ror the same experiment al data while employing the Poppe approach. The method Baard [98BAIJ employed to calculate the Merkel number is improved in the sample calculation presented here. During a test of an expanded metal fill of height, Lft = 1.878 m, the following measurements are made [95001,98KRIJ, Atmospheric pressure Pa = 101712.27 Pa Air inlet temperature Taj = 9.7 DC (282.85 K) Air inlet temperature (wetbulb) Twb = 8.23 DC (281.38 K) Dry air mass flow rate ma = 4.134 kg/s Static pressure drop across fill !!'pft = 4.5 Pa Water inlet temperature TWi = 39.67 DC (312.82 K) Water outlet temperature Twa = 27.77"C (300.92 K) Inlet water mass flow rate mw = 3.999 kg/s Refer to section BAfur a discussion on the Runge-Kutta method applied to the gov erning equations. The Runge-Kutta method requires four intermediate calculation steps per fill int erval. Variables for the four intermediate calculation steps for the different intervals are denoted with the subscript (n,m), where n is the fill interval number and the second subscript, m, refers to the intermedi ate calculation step. It must be stressed that the single value subscripts between brackets refer to the level numbers as shown in figure G.1. For this evaluation consider a fill that is divided into two intervals as s hown in figure G.1. According to Kroger [98KR1] the initial values at level (0) are w(O) = Wi = 0.00 616336 kg/kg dry air and ima(O) =imai = 25291.87496 Jlkg dry air By following an iterative procedure find that a humidity ratio Wa = W(2) at the outlet of the fill is 0.02226 kg/kg dry air. The outlet humidity ratio is required in equation (B.32), which i s used in all the intermediate calculation steps of the Runge-Kutta method for all the fill interv als. 0.2 ,''j Level (2) Level (1) Level (0) Interval 2 w Figure G.I: Counterflow fill divided into two intervals Since the fill is divided into two intervals find from equation (B.66) I!>.T. = (T., - Two)/(number ofintervals) ~ (312.82 - 300.92)/2 ~ 5.95 K. From equations (B.5I) to (B.53) find for level (I) at the top of the first inter val of the Runge-Kulla method, W(I) = W(O) + (jO,I) + 'lj(I)) + 2j(I,,) + j(I,4))16 imall) = imalo) + (k(l,I) + 21<{",) + 2kO,') + k(I,4)/6 Mep(l) = Mep(o) + (/0,1) + 21(1,,) + 21o,,) + 10,4))/6 where Mep(o) = 0 is equal to zero at the air inlet side of the fill or at level (0). Commence with the first intermediate calculation step of the Runge-Kutta method for the first fill interval. It can be seen from equations (B.54) to (B.56) thatj(l,I), 1<{1,1) and 1(1,1) are functions of T.eo). ima(o) and weO). Define that

T.(I,I) = T.(o) = Two = 300,92K W(I,I) = w(O) = W, = 0,00616336 kg/kg dry air i,aO,I) = lma(O) = ima, =25291.87496 Jlkg dry air. To calculate jO,IJ, k(1,1) and 10,1) in equations (B.51) to (B.53) respectively, the specific heats have to be evaluated at ([w(I,I) + 273,15)/2 = (300.92 + 273.15)/2 = 287,035K. Specific heat of dry air from equation (A.1.2) Cpa(I,I) ~ 1.045356 X 10' - 3,161783xl0-1x 287.035 + 7.083814 x 10-4 (287.035)' - 2.705209 x 10-7 (287.035)' = 1006,567 J/kgK Specific heat of water vapor from equation (A,2,2) Cp?(I,I) = 1.3605 X 10' + 2.31334 x 287.035 - 2.46784 X 10-10 (287.035)' + 5.91332 x 10-13 (287.035)' = 1874.385 JlkgK Specific heat of water from equation (A.4.2) G.3 Cpw(I,I) ~ 8.15599 x 10' - 2.80627 x 10 x 287.035 + 5.11283 x 10-2(287.035)2 - 2.17582 x 10-13 (287.035)6 ~ 4191.744 JlkgK Pressure ofwater vapor from equation (A.2.1) evaluated at T(1,I) ~ 300.92K. Z(I,I) ~ 10.79586(1 - 273.16/300.92) + 5.02808 loglO(273.16/300.92) + 1.50474 x 1O-4[I_IO-s.29692 ?300.92I273.16)-1}]+ 4.2873xI0-4[1O 4.76955(I-273.16/'OO.92L I] + 2.786118312 = 3.57157 P>(I,I) ~ 10,?m57 = 3729 Pa Humidity ratio for saturated air at T(1,I) ~ 300.92K from equation (A.3.5) w = ( 0.62509(3729) ) = 0.02379 kgikg dry air w,(I,I) 101712.3 -1.005(3729) Latent heat at 273.15K follows from equation (AA.5) itgw(I,I) = 2.5016 X 10' Jlkg The enthalpy ofwater vapor at the local bulk water temperature, TW(I,I? relative to water at O?C, 1>(1,1) = Ijgw(l.l} + Cp,(I,I) TW(I.I) = 2.5016 x 106+ 1874.385(300.92-273.15) ~ 2553650 Jlkg The enthalpy of saturated air at the local bulk water temperature from equation (A.3.6b) im",W(1,I) = 1006.567(300.92 - 273.15) + (0.02379)(2553650) = 88711 Jlkg The Lewis factor from equation (B.l5) ( 0.662 + 0.02379 I) Le - 0 8650.667 0.662 + 0.00616336 ~ 0.9205 /(1.1) - ? In( 0.662 +0.02379 ) 0.662 + 0.00616336 The mass balance from equation (B.32), ( m w ) = 3.999 [1- 4.134 (0.02226 -0.00616336)J~ 0.9512 rna (1,1) 4.134 3.999 From equation (B.54) find j(1,I) = ATw ? f(;rw(o) , lma(O) , w(oJ= ATw ' f(Tw(I,I)' ima(1,I)' W(I.,?) From equation (BA8) find but from equation (B.24), 0.4 j(I.I) = . . 'masw(l,l) -lma(l,l) + pw(I,llw(I,I) = (5.95)(4191.744)(0.9512)(0.02379 - 0.00616336) [ 88711- 25291.87496 ] + (0.9205 -1)[88711- 25291.87496 - (0.02379 - 0.00616336)(2553650)] - (0.02379 - 0.00616336)(4191.744)(300.92 - 273.15) = 0.006982 Combine equations (B.25), (B.49) and (B.55) to find. pW(I.llw(I.I) ) = (5.95)(4191.744)(0.9512) x 1+ -=-__----'('--0.-'-02_3_79_-_0_.0_0_61_6_33_6~)(4_1_9_1._74_4'-'.)(3_0_0._9

2_-_2_7_3._15-,-) ~ [ 88711- 25291.87496+(0.9205 -1) x [88711- 25291.87496 - (0.02379- ] 0.00616336)(2553650)]- (0.02379 - 0.00616336)(4191.744)(300.92 - 273.15) ~ 24537.63 Combine equations (B.30), (B.50) and (B.56) to find (5.95)(4191.744) = ;=-[8::-:8:-::7.,.-11:---:2::-:5:::-29::-:1-:.8::-:7,-;-49::-:6:---'O"--'-''-----'---------~] +(0.9205 -1)[88711- 25291.87496 -(0.02379 - 0.00616336)(2553650)] - (0.02379 - 0.00616336)(4191.744)(300.92 - 273.15) = 0.41635 By proceeding along the same lines j(I.2h 1<{1.2) and 1(1,2) are determined for the second intermediate calculation step of the Runge-Kutta method for the first fill interval: From equations (B.57) to (B.59) can be seen thatj(I~). k(1,2) and 1(1.2) are fun ctions of G.5 I!>.T k .T + _w_ i +~ and w + J(I.l) thus define w(O) 2' ",a(O) 2 (0) 2 Tw(I.2) = Tw(o) + I!>.Tw/2 = 300.92 + 5.95/2 = 303.895K W(I.2) =W(I) +j(l.ly2 = 0.00616336 + 0.006982/2 = 0.0096544 kg/kg dry air ima(l.2) = ima(l) + k(l.d2 = 25291.88 + 24537.63/2 = 37560.67 Jlkg The specific heats have to be evaluated at (Tw(1.2) + 273.15)/2 = 288.5225K Specific heat ofdry air from equation (A.1.2) C,.(1.2) = 1.045356xl03- 3.161783xI0-1x 288.5225 + 7.083814xI0-4 (288.5225)2 - 2.705209xI0-7(288.5225)3 = 1006.603 JlkgK Specific heat of water vapor from equation (A.2.2) Cp"I.2) = 1.3605 x 103+ 2JI334 x 288.5225 - 2.46784 x 10-1?(287.035)' + 5.91332 x 10-13 (288.5225)6 = 1875.654 JlkgK Specific heat of water from equation (A.4.2) C,w(I.2) = 8.15599 x 103- 2.80627 x 10 x 288.5225 + 5.11283 x 10-2(288.5225)2 - 2.17582 X 10-13 (288.5225)6 = 4189.941 JlkgK The vapor pressure and humidity ratio of saturated air are calculated at the loc al water temperature Tw(I.2) = 303.895K Vapor pressure from equation (A.2.1): Po;(I.2) = 4427.4 Pa Humidity ratio for saturated air from equation (AJ.5): W,w(I.2) = 0.028454 kglkg dry air The enthalpy ofwater vapor at the local bulk water temperature, Tw(1.2) , relati ve to water at Oae, io;(I) = iJilw(I.2) + Cpo;(I.2) TW(I.2)= 2.5016 x 106+ 1875.654(303.895-273.15) = 2553650 Jlkg The entalpy ofsaturated air at the local bulk water temperature from equation (A J.6b) imasw(l.2) = 1006.603(303.8950 - 273.15) + (0.028454)(2553650) = 103770 Jlkg The Lewis factor from equation (B.15) ( 0.662 + 0.028454 1) Le - 0 8650.667 0.662 + 0.0096582 = 0.9212 fO.2) - ? In( 0.662 + 0.028454 ) 0.662 +0.0096582 The mass balance from equation (B.32) r ?? Ii" (rn. J = 3.999 (1- 4.134 (0.02226 - 0.0096544)) = 0.9547 'I' rna (I,Z) 4.134 3.999 ,I, From equation (8.57) find j(I,Z) = t1T? ? f(Tw(l,z),ima(I,z)' W(I.ZJ From equation (B.48) find

:; = f(Tw(I,z)' ima(I,z)' W(i.zJ W From equation (B.24) find Combine equations (B.24), (B.48) and (B.57) to find j(I,Z) = , . lmasw(I,2) - lma(I,2) + pw(I,zlw(I,Z) = ;=--__--'(_5.9_5-'-')('-41_8_9._94_1"-')(~0._95_4_'7)co_(0_.0_2_84_5_4_-_0._00 _9_65_4---,4)~_~ [ 103770-37560.67 ] + (0.9212 -1)[103770- 37560.67 - (0.028454 - 0.0096544)(2559265)] -(0.028454 - 0.0096544)(4189.941X303.895 - 273.15) = 0.00717527 Arelation for ,,<,.Z) is obtained by com bining (B.25), (B.49) and (B.58) i.e. k(I.Zl = t1TwCpw(I.Zl(rn w ) rna (1,2) =(5.95)(4189.941)(0.9547) pw(Lzlw(I.2) ) x 1+ -;=c--:--:c:---,----'(0.:cc'0.:cc2=:.:::8_45:-4:--""c'0,-::.0--,-0:..,,96_5 -:44:-,:)(-,::-4=:18:-9-C..9:-4-::!:1):,,:(3:-03:-.=:98_5:---:2-::-73:-c.,-::15,-)-=-=-:-::-::-:-:-'i' [ 103770 - 37560.67 + (0.9212 -1) X [103770 - 37560.67 - (0.028454- 0.0096544)] X (2559265)]- (0.028454 - 0.0096544)(4189.941)(303.895 - 273.15) = 24726 10.z) is obtained by combining equation (B.30), (B.50) and (8.59) G.7 = (5.95X4189.941) =0.3997 [ 103770 - 37560.67 +(0.9212 -1)[103770 -37560.67 j - (0.028454 - 0.0096544X2559265)] - (0.028454 - 0.0096544)(4189.941)(303.895 - 273.15) Proceeding along the same lines, the following values are calculated to complete the Runge-Kutta numerical integration for the first interval ofthe fill. j(1.l) = 0.007150; k(1.3) = 24725; 1(1.3) = 0.4004;j(I.4) = 0.0073899; k(1.4) = 24927; 1(1.4) = 0.3738 The humidity ratio at level (I) follows from equation (B.5I), W(I) = w(O) + (j(1,1) + 2j(,)) + 2j(1,3) + j(1.4)Y6 = 0.00616336 + [0.006982 +(2)0.00717527 + (2)0.007150+ 0.0073899] 16= 0.0133338 kg/kg dry air The enthalpy of the air at level (I) follows from equation (B.52), [ma(l) = [malO) + (~I,I) + 2k(1,2) + 2k(1,3) + ~1,4)/6 = 25291.89 + [24537.63 + (2)24726 + (2)24725.12 + 24927] 16 = 50019.67 J/kg The transfer characteristic or Merkel number at level (I) follows from equation (B.S3), Mep(l) = Mep(o) + (/(1,1) + 21(1)) + 21(1,3) + 1(1,4?/6 = [0.41635 + (2)0.3997 +(2)0.4004 + 0.3738] = 0.3984 The dry bulb temperature Ta(l) and wet bulb temperature TWb(l) at level (I) are determined by assuming that the air is unsaturated, If TWb(lyTa(l) the air is supersaturated and the as sumption of unsaturated air must be corrected. The assumption is then corrected by assuming supersaturated a ir with TWb(l) = Ta(I)' Find the dry bulb temperature at level (ll, Tq(!}. The enthalpy of the air at level (I) is, ima(1) = 50019.67 J/kg. Assume that the air is unsaturated and that the drybulb temperature, Ta(l) = 289 .307 K.

The specific heats are evaluated at (Ta(1)+273.15)/2 = (289.307+273.ISY2 = 28I.2 2S8 K Specific heat of dry air from equation (A.I.2) Cpa(l) = 1006.446 J/kgK Specific heat of water vapor from equation (A.2.2) Cp>(l) = 1869.49S JlkgK Equation (A.3.6b) gives an expression for the enthalpy of an air vapor mixture p er unit mass ofdry air. ima(l) = 1006.446(289.307 - 273.1 S) + 0.0133338x[2501598 + 1869.495(289.307 - 2 73.IS)] = 50020 Jlkg G.8 The value of Im,(I) determined by equation (A.3.6b) is within close tolerance of the value determined by equation (B.52). The assumption of the value of the dry bulb temperature is ther efore correct if the air is unsaturated at level (I). Find the wetbulb temperature at level (l), Tw' flI : The humidity ratio at level (1), W(I) = 0.0133338 kWkg dry air, Assume that the wetbulb temperature at level (I) is Tw'(I) = 291.617 K and find from equations (A.2.1) and (A.3.5) respectively the corresponding vapor pressure and the humidity ratio . Vapor pressure from equation (A.2.1): P'CI) = 2124.092Pa Humidity ratio from equation (A.3.5): w(I) = 0.0133338 kWkg dry air The value of WCI), determined according to (A.3.5) is the same as the value dete rmined earlier. The assumed value of the wetbulb temperature is therefore correct if the air is unsa turated at level (I). Test if air is unsaturated or supersaturated: Since TW'(I) > T,(I) the air is actually supersaturated at level (I). The assump tion that the air is unsaturated for the determination of Ta(l) and Tw'(I) is therefore incorrect. For supersatur ated air at level (1) T,(I) = Tw'(I). Assume a value for T,O) = TW'(I) = 290.8448 K and find at this temperatu re Vapor pressure from equation (A.2.1): pm,(I) = 2023.427 Pa Humidity ratio from equation (A.3.5): W"CI) = 0.012689 kWkg dry air The following specific heats are determined at (Ta(l) + 273.15)/2 = (290.8448 + 273.15)/2 = 281.99 K Specific heat of dry air from equation (A. 1.2): Cpa(l) = 1006.460 J/kgK Specific heat of water vapor from equation (A.2.2): cp,(I) = 1870.138 J/kgK Specific heat of water from equation (A.4.2): Cpw(l) = 4198.815 JlkgK From equation (B.33) it follows that 1"(1) = cp'CI)(Ta(l) - 273.15) + w"(I)[ifgw, + cp"I)(T,(I) - 273.15)] + (W(I) - W"(I))CpwCI)(T,(I) - 273.15) = 1006.460(290.8448 - 273.15) + 0.012689[2501598 + 1870.138(290.8448-273.15)] + (0.0133338 - 0.012689)(4198.815)(290.8448 - 273.15) = 50020 J/kg 1.,,(1), determined by equation (B.33), is within close tolerance of Ima(I), det ermined by equation (B.52), thus, the assumption of the value of the dry bulb temperature at level (1) is th erefore correct. The air temperature at level (1), Ta(l), is therefore equal to 290.8448 K. Apply Runge.Kutta numerical integration to the second fill.interval. Find from e quations (B.51) to (B.53) at level (2), W(2) = W(I) + VC2.1) + 2jC2,2) + 2h,3) + j(2,4))/6 I"(2) = i.'.,(I) + (!c<2,1) + 2!c<2,2) + 2kC2,3) + !c<2,4))/6 G.9 Because the air is supersaturated i"CI) ~ imaCI )' The water temperature at leve l (1) is Tw(l) ~ Twa + AT. ~ 300.92 + 5.95 = 306.87K. For the first intermediate calculation step of the seco nd fill interval, i"C2.1) =

(M(l), TW(2,1) = Tw(l) and WsrJ(2,1) = Waa(l)To calculatejc2,,? 1cc2,1) and /c2,1) for the first intermediate calculation ste p for the second fill interval certain thermophysical properties have to be evaluated at (TW(2,1) + 273.15}/2 = (306.87 + 273.15}/2 ~ 290.01 K Specific heat of dry air from equation (A. 1.2) CpaC2.1) = 1.045356 X 10' - 3.161783xl0-' x 290.01 + 7.083814 x 10-4 (290.01)2 - 2.705209 X 10-7 (290,01)' = 1006.642 J/kgK Specific heat of water vapor from equation (A.2.2) Cp,,2,1) = 1.3605 x 10' + 2.31334x 290.01 - 2.46784 x 10-10 (290.01)5 + 5.91332 X 10-13 (290.01)6 ~ 1876.933 J/kgK Specific heat of water from equation (A.4.2) CpwC2.1) = 8.15599 x 10'_ 2.80627 x 10 x 290.01 + 5.11283 x 10-2(290.01)2 - 2.17582 X 10-13 (290,01)6 ~ 4188..264 J/kgK Vapor pressure from equation (A.2.1) evaluated at T'C~I) = 306.87K ZC2.1) = 10.79586(1- 273.16/306.87) + 5.0280810810(273.16/306.87) + 1.50474 x 10"'[1 _ 10?8.2%92\C306.87I27'.16)-1}] + 4.2873 x 10-4[10 4.76955(1-273.161'06.87) - I] + 2.786118312 =3.71909 P"2.1) = 10,?7100' = 5237 Pa Humidity ratio for saturated air evaluated at TwC2.1)= 306.87K from equation (A. 3.5), w =( 0.62509(5237) J =0.0339417 kg/kg dry air ..1'.1) 101712.3 -1.005(5237) The enthalpy ofwater vapor at the local bulk water temperature, T'C2 ,1), relati ve to water at O?C, ;,,2,1) = iftwC2 ,I) + Cp>(2,I) Twe2,1) ~ 2.5016 x 106+ 1876.933(306.87-273.15) = 2564889 J/kg The entalpy of saturated air at the local bulk water temperature from equation ( A.3.6b) ;mMwC2,1) ~ 1006.642(306.87 - 273.15) + (0,0339417)(2564889) ~ 121001 J/kg The Lewis factor from equation (B.38) G.10 ( 0.662+0.0339417 1) Le - ?865? 66' 0.662 + 0.012689 = 0.9229 /(2.1) - ? In(0.662 + 0.0339417) 0.662+0.012689 The mass balance from equation (B.32) ( m w J = 3.999 (1- 4.134 (0.02226 -0.0133338))= 0.9584 m. (2,1) 4.134 3.999 From equation (B.54) find j(2,1) = !!.Tw. f(Tw(l),i"'(l) ,w(lJ= !!.Tw. f(Tw(2.1),i,,(2.1)' W(2,1?) From equation (B,48) find :; = f(Tw(,), i"(I)' W(l?)= f(TW (2,,), i"(2,1), W(2"J w but from equation (B,42) Cpw(2.,)(mW) (W'W(2.1) -W'.(2.1?) _dw_ = --,-__--,,--__-'-m_.=('72,1"-.)----,-----,----.-------r dTw imasw(2,I) -i.U (2,1) + e f(2,l) -1 'mant'(2,l) -iSS(2,1) - wsw(2,1) - WSQ(2 .1) 'Y(2,l) + w(l.) - wsa(2,1) + (W(l,l) - Wsw(2,})~PW(2,llw{2,1) Combine equations (B.42), (BA8) and (B.54) to find t.TwCPW(2,,)(mwJ (W~(2,,) - W'.(2,1?) . _ ma (2,1) ~(2.1) - ., 1.. . 'masw(2,1) -13$(2,1) + e f(2,1) - lmuaw(2,l) -la8(2,1) - W.rw(2,1) - Wsa(2,1) v( 2,1) + W(l,l) - Wsa(2,1) + (W(l,l) - WSW(2.1)~PW(2,llw(2,1) (5.95)(4188.264XO.9584XO.0339417 - 0.012689)=.".---,-_---:-'-_-"-'-----"--'---~-

------'-------.;121001-50019.67 + (0.9229 -1)[121001- 50019.67 - (0.0339417 - 0.012689X2564889) + (0.0133338 - 0.012689)(4188.264X306.87 - 273.15)] + (0.0133338 - 0.0339417X4188.264)(306.87 - 273.15) = 0.007599 Combine equations (BA3), (BA9) and (B.55) to find, k(2.1} = ATwCpw[ll(mw) rna (2,1) G.ll x [+ lW~(2,1) - W,Q(2,1)) i.wasw(2,l) - is.I'(2,1) + f(2,1) -1 ~(2,1) - is.7(2.1) - wsw(2,I) - wsa(2.1) . (2,1) + w(2,l) - wsa(2,I) + (W(2,1) - WSlf'(2.1)~pw(2,l)Tw(2,I) = (5.95)(4188.264)(0.9548) x 1+ (0.0339417 - 0.012689)(4188.264)(306.87 - 273.15) 121001- 50019.67 + (0.9229-1) x [121001- 50019,67 - (0.0339417 - 0.012689)(2564889) + (0.0133338 - 0.012689)(4188.264)(306,87 - 273,15)] + (0,0133338-0.0339417)(4188.264)(306,87 - 273.15) Combine equations (B.47), (B.50) and (B.56) to find, = 24957 (5,95)(4188,264) =~--------'---'-'----------'--------..". 121001- 50019.67 + (0.9229 -1) x [121001- 50019.67 - (0.0339417 - 0.012689)(2564889) + (0.0133338 - 0.012689)(4188.264)(306.87 - 273.15)] + (0.0133338 - 0.0339417)(4188.264)(306.87 - 273.15) ~ 0.3731 ) By proceeding along the same lines, the following values are calculated for the second to fourth intermediate calculation steps to finish the Runge-Kutta numerical integration f or the second interval of the fill. i(2,2) = 0.008957 ;l?~2) ~ 25346; 1(2,2) ~ 0.33598 i(2,]) ~ 0.008973; k(2,]) ~ 25365; 1(2,]) ~ 0,03364 i(2,4) = 0.010124; k(2,4) = 25771; 1(2,4) ~ 0.2965 The humidity ratio at level (2) follows from equation (B.51), W(2) ~ W(I) + (j(2,I) + 2i(2)) + 2i(2,]) +i(2,4)y6 = 0.0133338 + [0.007599 +(2)0.008957 + (2)0.008973+ 0.010124] /6 ~ 0.02226 kg/kg dry air Since W(2) is equal to WOo which is assumed to be 0.02226 kg/kg dry air in the b eginning of this example, the system ofequations has converged. The air enthalpy at level (2) follows from equation (B.52), i"(2) ~ ;..(1) + (k(2,I) + 2k(2,2) + 2k(2,]) + k(2,4)/6 = 50019.67 + [24957 + (2)25346 + (2)25635 + 25771] /6 = 75378 Jlkg The transfer characteristic or Merkel number at level (2) follows from equation (B.53), G.12 Mep(2) =Mep(l) + (1(2,1) +2le2) + 2le2,,) + 1(2,40/6 = 0,3984 + [0.3731 + (2)0.33598 + (2)0.3364 + 0,2965] = 0.7341 The air was already supersaturated at level (1). Therefore assume that the tempe rature of tbe supersaturated air at level (2) is Ta(2) = TWb(2)= 297.8508 K Tbe partial pressure and bumidity ratio of saturated air, from equations (A.2.1) and (A.3.5), evaluated at Ta(2) are respectively P"a(2) = 311 0.68 Pa W"(2) = 0.01972 kglkg dry air

The specific beat of dry air, water liquid and vapor are evaluated at (Ta(2) + 2 73.15)/2 = (297.8508 + 273.15)/2 = 285.5 K Specific beat of dry air from equation (A.I.2): Specific beat of water vapor equation (A.2.2): Specific beat of water equation (A.4.2)\ C,.(2) = 1006.532 JlkgK Cp'(2) = 1873.084 JlkgK Cpw(2) = 4193.739 J/kgK It follows from equation (B.33) that i"(2) = cpa(2)(Ta(2) - 273.15) + w"(2)[ijgw[l] + cp,(zlTa(2) - 273.15)] + (W(2) - w"(2?)cpw(2)(Ta(2) -273.15) = 1006.532(297.8508 - 273.15) + 0.01972[2501598 + 1873.084(297.8508 - 273.15)] + (0.02226 - 0.01972)(4193.739)(297.8508 -273.15) = 75378.4 K i"(2), determined by equation (B,33), is within close tolerance of iss(2), deter mined by equation (B.52), thus the value of the air temperature assumed at level (2) is therefore correct. Therefore the conditions at the outlet oftbe fill are: imaa = i"(2) =75378 Jlkg; Wa = W(2) = 0.02226 kglkg dry air; Taa = Ta(2) =297.85 08K wbere the Merkel number for the fill is Mep = Mep(2) = 0.7341. This is 7.2% grea ter than the Merkel number obtained by the Merkel approach in KrOger [98KRI]. H.l APPENDIXH CROSSFLOW FILL ANALYSIS ACCORDING TO THE e-NTU, MERKEL AND POPPE APPROACHES H.I INTRODUCTION Experimental test measurements of a counterflow expanded metal fill are presente d in appendix G. Kroger [98KRI] obtained the Merkel number, for the experimental values given in appendix G, according to the Merkel approach. The Merkel number, according to the counterflow Poppe ap proach, is presented in appendix G. The same values of the experimental values are used in this crossflow fill perfo rmance analysis as in the counterflow case presented in appendix G. This is done to evaluate the differenc es between the Merkel numbers obtained for the counterflow and crossflow fill configurations. A sample calculation for the crossflow configuration for the Merkel and Poppe ap proaches can not be presented in the same form as for the counterflow configuration presented in app endix G. The governing partial differential equations for the crossflow configuration are solved by a p oint-by-point Gauss-Seidel [80PAl, 92MAI] iterative procedure across a two-dimensional domain using the pri nciple of finite / differences. It is therefore very cumbersome to present a sample calculation and only the results are therefore presented. The results can be presented graphically for the Merkel and Poppe approaches due to the two-dimensional nature of the crossflow configuration. H.2 POPPE APPROACH The governing equations of the Poppe approach for crossflow fills are solved by an iterative technique as discussed in appendix C. The governing equations must be satisfied on each verte x in the computational domain before convergence can be obtained. Figure H.I shows the solution domain

of a counterflow fill for non-dimensional fill dimensions. The solution domain is divided in 50 interv als in both directions. It can be seen from figure H.I in which parts of the fill the air is unsaturated an d supersaturated, for the experimental measurements specified in appendix G. The dividing line between the unsaturated and supersaturated regions will be smooth if the solution domain is divided in much more intervals. It can be seen that the air becomes saturated soon after entering the fill, especially in the top parts of the fill. The governing equations for unsaturated and supersaturated air are thus solved in th e respective regions shown in figure H.l. Figures H.2(a) to H.2(t) show the distribution of the water temperature, water m ass velocity, Lewis factor, air enthalpy, air temperature and the humidity ratio of the air respecti vely across the non? dimensional solution domain of the crossflow fill. Refer to figure H.! for the c oordinate system 1p , H.2 convention used in figure H2. The water and air inlet sides of the various plots in figure H2 are the same as those illustrated in figure HI. Water inlet side o ~ ~ ~ ~ ~0.-,-''------''------''------''------''---, -. Supersaturated air Unsaturated air ..... '------'------------' 17 ~ Figure HI: State ofair in fill for non dimensional fiU dimensions. The distribution of the water temperature across a vertical section of the fill is illustrated in figure H.2(a). The mean water outlet temperature is determined by equation (C.30) and is equal to 300.92K. It can be seen that water cooling is more effective near the air inlet side. This is becau se the water near the air inlet side is in contact with the cool inlet air aU the time it fans through the fin. The mass velocity of the water, as it passes through the fill, can be seen in figure H2(b). Approximately the sa me trends can be observed as the water temperature in figure H2(a). The water mass velocity is reduced as it passes through the fill because of evaporation. The evaporation loss is larger near the air inlet side b ecause the inlet air is relatively dry compared to the air deeper into the fill. Thus, a greater potenti al for evaporation loss exists where the air is the driest. Figure H,2(c) shows how the value of the Lewis fact or, according to the equation of Bosnjakovic [65BOI], is distributed across the fiU. Figures H2(d) to H.2(f) show the enthalpy, temperature and humidity ratio of the air as it passes through the fil l. It can be seen that the plotted contours of these three variables follow approximately the same trends. The air enthalpy increases more rapidly in the top of the fill because the air is in contact with the hot i nlet water stream the entire

time as it moves through the fill. H.3 MERKEL APPROACH Figures H3(a) and H.3(b) show the distribution of the water temperature and air enthalpy according to the Merkel approach. The results of the Merkel approach can be compared to the r esults of the more rigorous Poppe approach presented in figure H.2. The mean water outlet temperatu re of both approaches is equal to 300.92K. The mean outlet air enthalpy and temperature of the Merkel approach is less than that predicted by the Poppe approach. H.3 (d) (c) (f) Figure H.2: Distribution of water temperature, water mass velocity, Lewis factor , air enthalpy, air temperature and humidity across a crossflow fill, determined according to the Po ppe approach. H.4 (a) Figure H.3: Water temperature and air enthalpy distribution in a crossflow fill according to the Merkel approach. D.4 e-NTU APPROACH The crossflow Merkel number according to the e-NTU approach is not solved by tw( )-dimensional finite differences. Four variants of the e-NTUapproach are employed. e-NTU, and e-NTU, refer respectively to the crossflow cases where both the water and air streams are unmixed and both th e air and water streams are mixed. e-NTU, refers to the crossflow case where Cmax, which is generally th e water stream, is mixed, and C,nim which is generally the air stream, is unmixed. C_ is unmixed and C"n i s mixed for the e-NTU. case. The comparison of the four different e-NfU approaches and the comparison t o the Merkel and Poppe approaches are presented in the next section. D.S COMPARATIVE RESULTS OF THE e-NTU, MERKEL AND POPPE APPROACHES The heat rejected, air outlet temperature, water evaporation rate and the Merkel number for the fill test conditions given above, obtained by employing the Merkel, Poppe and the four e-N TU approaches, are given in table H.t. The 2-d computational domain is divided in 50 intervals in b oth the horizontal and vertical directions for the Merkel and Poppe analyses. Q, Too and m",,~p) determined by all four variants of the e-NTU approach are ide ntical. 'These variables are also identical to the values obtained by the Merkel approach. This is becaus e it is assumed, for both approaches, that the outlet air is saturated with water vapor. The heat rejectio n rate is calculated by exactly the same manner for all the variants of the e-NTU approach and the Merl< el approach, i.e., Q = m.,.cpwm(Twi - T~). The Merkel numbers obtained by the Merkel approach and the e-NTU, approach are p ractically identical Where both streams are mixed, i.e., the e-NTU, approach, the Merkel number is ap proximately 5% higher than that predicted by the Merkel approach. The Merkel numbers for the other two cases, e-NTU, and e?

NTU. respectively, are in between the limiting, e-NTU, and e-NTU" cases. H.5 Table H.I: Fill performance charact...istics of a erossflow fill according to th e M...kel, e-NTU and Poppe approaches. Merkel Poppe e-NTU. e-NTU2 e-NTUJ e-NTU. Q,MW 0.1987946 0.2064673 0.1987946 0.1987946 0.1987946 0.1987946 Tao,K 297.4277 297.8390 297.4277 297.4277 297.4277 297.4277 mi:l:~ 0.05395610 0.06277423 0.05395610 0.05395610 0.05395610 0.05395610 Me 0.7395232 0.7976296 0.7404729 0.7750973 0.7588670 0.7486152 Thus, only the Merkel numbers differ for the respective e-NTU approach variants and the Merkel approach. The fill outlet conditions, predicted by all the variants of the e-NTU approach and the Merkel approach are identical. Thus, any variant of the e-NTU approach or the Merkel ap proach can be used consistently in the fill performance analysis and in the subsequent cooling towe r p...formance analysis. Cooling tower perfurmance, predicted by the all the variant of the e-NTU approac h and Merkel approach, will therefore be practically identical. It is recommended that the fill perform ance evaluation be carried out at approximately the same conditions where the cooling tower will operate. The Merkel numbers according to the M...kel approach and the e-NTU1 approach are practically identical for the fill performance analysis. However, this is only true for the operationa l conditions specified above. Therefore, the empirical relations obtained from fill performance analyses, by e mploying the one approach, cannot be used interchangeably in cooling tower performance calculatio ns while employing the other approach. For other water temp...atures and practical water to air mass fl ow ratios the differences between the Merkel numbers ofthe two approaches can be quite significant. It can be seen from table H.I that the more rigorous Poppe approach predicts hig her heat rejection rates, water evaporation rates and Merkel numbers than the Merkel approach. The M...kel number according to the Poppe approach is approximately 8 % higher than that predicted by the Merkel approach. The predicted heat rejection rate according to the Poppe approach is approximately 4 % high... than that predicted by the Merkel approach. It is important to realize that the comparisons between the different approaches are only for the ambient and operational conditions specified above. The differences between the approach es can vary quite sigoificantly at extreme ambient conditions. The mean outlet air temperature and hwnidity ratio, according to the Merkel appr oach, are obtained by integrating the outlet air enthalpy, at the air outlet side of the fill, and by asswning that the air is saturated at this mean enthalpy. The air outlet temperature and hwnidity ratio can also be obtained, at each grid point at the air outlet side of the fill, by assuming that the air is saturated at each grid point. The mean air temperature and humidity ratio can then be obtained by integration. Therefore, i t does not matt... if the H.6 assumption of saturated air is applied to each air outlet grid point, or to the

mean outlet air enthalpy, the same results are obtained H.6 COMPARISON BETWEEN THE PERFORMANCE OF COUNTERFLOW AND CROSSFLOW FILLS The results of the crossflow fill performance analysis are compared to the resul ts of a counterflow fill performance analysis. The same operational and ambient conditions are used in bo th the erossflow and counterflow fill performance analyses. Thus, the cooling range is identical for both the counterflow and crossflow fill tests. The heat rejection rates, air outlet temperatures, evapora tion rates and the Merkel numbers, according to the Merkel, Poppe and e-NTU approaches for the counterflow fill analysis, are shown in table H.2. [fthe values in table H.2 are compared to the corresponding values in table R.I, it can be seen that the heat rejection rate, Q, the air outlet temperature, Ta" and the water evaporatio n rate, m.(.""p), for the Merkel and e-NTU approaches are identical for both the crossflow and counterflow fills. The heat rejection rates are the same because of the equal cooling ranges for the crossfl ow and counterflow fills. The outlet air enthalpy can be calculated from a simple energy balance. Subseque ntly, the outlet air temperature can be calculated after the assumption that the outlet air is satura ted with water vapor. The Merkel number for the counterflow case, however, is equal to 0.68468, which is a pproximately 7% smaller than that predicted for the erossflow fill. Table H.2; Fill performance characteristics of a counterflow fill according to t he Merkel, e-NTU and Poppe approaches. Merkel Poppe e-NTU Q,MW 0.1987946 0.211380 0.1987946 T..,K 297.428 298.1192 297.428 m"'.....l> 0.05395610 0.0649723 0.05395610k2!s Me 0.68468 0.741356 0.6770926 The heat rejection rates, air outlet temperatures and water evaporation rates fo r the fill performance analyses, for the crossflow and counterflow fills, according to the Poppe approa ch, are not equal as was the case with the Merkel and e-NTU approaches. The heat rejection rate according to the Poppe approach for the counterflow fill is approximately 2.5% higher than that predicted for th e crossflow fill. The air outlet temperature is also approximately 0.3K higher for the counterflow fill co mpared to the crossflow fill. Approximately 3.5% more water is evaporated in the counterflow fill compar ed to the crossflow fill. It is evident that the Merkel numbers for the crossflow fill, for the Merkel and Poppe approaches, are higher than those for the counterflow fill. A larger wetted area is needed in th e crossflow fill to obtain the same cooling load as in the counterflow fill. Thus, a larger volume of fill is n eeded in crossflow towers than in counterflow towers to obtain the same cooling load l.l APPENDIX I ANALYSIS OF A NATURAL DRAFT WET-COOLING TOWER EMPLOYING THE

POPPE APPROACH 1.1 INTRODUCTION The heat rejection rate and the loss in cooling water, due to evaporation, in a hyperbolic natural draft counterflow wet-cooling tower, as shown in figure I.1, are determined while empl oying the Poppe approach for heat and mass transfer in the fill. KrOger [98KRl] employed the Mer kel approach to calculate the heat rejection rate and evaporation rate while employing the same cooling tower and operational specifications. It is assumed that the water and the airflow through the fill are uniform and that the inside diameter of the upper section ofthe tower is constant. IS d6 ~I I I (j) , I , I , I , I , I , H6I , I , I I , I , (l) ~UUUUUUU CD Figure I.1: Natural draft cooling tower with horizontal fill. Ambient conditions: Air temperature at ground level Wetbulb temperature at ground level Atmospheric pressure Ambient temperature gradient Cooling tower and operational specifications: Water mass flow rate Water inlet temperature Rounded tower shell inlet Tower height Tower inlet height Tower inlet diameter Tower outlet diameter Number of tower supports Length oftower supports Diameter of support Drag coefficient of round tower supports Shell thickness at air inlet 1.2 Tal = 15.45aC (288.6K)

T.. ~ 11.05?C (284.2K) pal = 84100 N/m' dTa/dz = -0.00975 Kim from ground level m. = 12500 kg/s TWi ~ T., = 40?C rtld, =0.02 Hb = 147m H, = 10m d, = 104.5 m db = 60.85 m n" ~72 L" = 11.6 m d" ~0.8m CD" = 1.0 t, ~ 1.0 m Fill specificatiDns: The cooling tower is fitted with an expanded metal fill (Lji = 2.504 m) for whic h the performance characteristics are respectively: Transfer coefficient Loss coefficient Frontal area of the fill Other specifications: Depth of spray zone above fill Mean drop diameter in rain zone Loss coefficient for contraction and fill supports based on Afr Loss coefficient for distribution system Kinetic energy coefficient at tower outlet h aMe = dji ji = 0 27928G-o,o"GO.6023 ~ G . 'Ill a ? K =I 85IGI.27"G-I.0316tdm)' 'Ill a ~ 8300 m' L,p = 0.5 m da = 0.0035 m =0.5 = 1.01 The transfer area, i.e., the fill, rain zone and spray zone, is divided into fiv e intervals, with an equal temperature difference across each interval, for the numerical integration of th e governing equations of the Poppe approach. 1.3 1.2 AIR INLET CONDITIONS The enthalpy ofthe inlet air, i..,,, is found according to equation (A.3.6b). At the specified air inlet drybulb temperature of T,t = 288.6K and wetbulb temperature of TWb = 284.2K find the following: Pressure ofwater vapor from equation (A.2.1) evaluated at Twb , where Twb = 284. 2K. Zt = 10.79586(1 - 273.16/284.2) + 5.02808 loglO(273.16/284.2) + 1.50474 x 10-'[1_ 1O-8.29692 {(284.2n73.1b)-I}]+ 4.2873 x 10-4[10 ,.,."'(t-273.tbn84.2) - I] + 2.786118312 = 3.119284 P,t = 103.119284= 1316.086 Pa Humidity ratio for saturated air from equation (A.3.5) ( 2501.6-2.3263(Twb -273.15) )( 0.62509p~b ) WI = 2501.6+1.8577(T-273.15)-4.184(Twb -273.15) p,,, -1.005p~b

( 1.00416(T-Twb) ) 2501.6+ 1.8577(T -273.15) - 4.1 84(Twb - 273.15) ( 2501.6 - 2.3263(284.2 - 273.15) )( 0.62509 ?1316.086 ) - 2501.6 +1.8577(288.6 - 273.15) - 4.184(284.2 - 273.15) 84100 -1.005 ?1316.086 ( - 1.00416(288.6 - 284.2) ) 2501.6 +1.8577(288.6 - 273.15) - 4.184(284.2 - 273.15) = 0.008127 kglkg dry air The enthalpy of the inlet air, i..,I, is found according to equation (A.3.6b) wi th Cpol = 1006.44 JlkgK and Cp,1 = 1869.2 J/kgK being evaluated at (T'I + 273.15)12 = (288.6 + 273.15)/2 = 2 80.875K according to equations (A.1.2) and (A.2.2) respectively. The latent heat is found to be ijg"o = 2.5016 X lOb JlkgK according to equation (A.4.5) at 273.15K. With these values find ;..,t = 36114.7 1 Jlkg dry air. 1.3 INITIAL APPROXIMATION OF VARIABLES Six cooling tower design variables are chosen and solved through an iterative pr ocedure. These variables are m"t', Po" Pob, T", Two and w,. Initial approximations for the variables must be supplied for the first iteration of the cooling tower analysis. A preliminary estimate can be made on t he evidence of empirical results and simple physical models. The initial approximations of the water outl et temperature, Two> and the outlet air temperature, To" are determined from empirical relations found in the literature. The initial approximations of the pressures Po' and Pob are found from a pressure distributi on derived for a constant atmospheric humidity and a dry adiabatic lapse rate. The initial approximation f or the mass flow rate, m"I', is found from a simple heat balance ofthe cooling tower. The air outlet hu midity, w" is determined by assuming that the air is saturated at the air outlet ofthe fill. An empirical formula to determine the approximate water outlet temperature is ac cording to Johnson and Priester [49101], 1.4 = 313.15 + (2)(284.2) + 288.6 = 292.5375 K 4 Where the temperature of the saturated air leaving the tower, T,,, is not known to the designer can it be approximated by the average of the inlet and outlet temperatures as can be seen in figure 1.2. The data in figure 1.2 are obtained from Mohiuddin and Kant [96MOl], Hutchison and Spivey [4 2HUI] and McKelvey and Brook [59MCI]. T = Tw; +T_ = 313.15+292.5375 302.8437K a' 2 2 ? Hutchison and Spivey [42HUI] o McKelvey and Brooke [59MCI] 40 D Mohiuddin and Kant [96MOI] 40 Averal!e of inlet and outlet water temnerature. ?C Figure 1.2: Variation of outlet air temperature with average water temperature. The air pressures inside the cooling tower at elevation 6 and 7 can be approxima ted by using the are pressure distribution relation derived for a constant atmospheric humidity, WI,

and a dry adiabatic lapse rate, i.e., [ ( )] 3.'{1+W f 1- w, J _ _0.00975H3 +Lfl +L,p '\ ,,\+0.62198 P" - P,I 1 T al ) 35(1+0 008127{1 0.008127 J [ 0.00975(10+ 2.504 + 0.5 ] . . 'l 0.008127+0.?2198 =841001-----'--------'-? = 83971.38 Pa 288.6 _ [_ 0.00975H. ]3.'(I+W'{1 "\+0~~2198J Pa. -Pal I T ,I 1.5 35(1+0 812l1- 0,008127 J 000975(147)]' .00 \ 0.008127+0.62198 . ~ 82654.27 Pa 288.6 The initial approximated air mass flow rate can be obtained by a simple heat bal ance for the cooling tower, ma(imoss -ima1 )= mwcpwm(Twl -Two) After rearrangement and aS8uming that Ina,,, = rna find, mwcpwm (TWi - Two) m =--'-+--'---'--,_a'lS (. .) Imas5 -'mal where cpwm is evaluated at (Tw; + Two)/2 ~ (313.15 + 292.5375)/2 = 302.8437 K. F rom equation (A.4.2) find cpwm = 8.15599 X 10' - 2.80627 x 10 (302.8437) + 5.11283 x 10'2 (302.8437)' ? 2. 17582 x \0.13 (302.8437)6 = 4178.721 JlkgK The enthalpy of the saturated outlet air, im~', is found according to equation ( A.3.6b) at the approximated air outlet temperature, Ta, = 288.6 K, Pressure of water vapor from equation (A.2.1) evaluated at Ta" where Ta, = 302.8 437K. z, ~ 10.79586(1 - 273.16/302.8437) + 5.02808 logI0(273.16/302.8437) + 1.50474 x 10-4[1 - 1O,?.29692{(302.84371273.16j-1 1)+ 4.2873 x 10-4[10 4.76"'(1 - 273.16 1302.8437) - 1) + 2.786118312 = 3.61998 p" ~ \0,.6199. ~ 4168.581 Pa Humidity ratio for saturated air from equation (A.3.5) 0.62509p""b W, = Pa" -1.005P""b = 0.62509? 4168.581 ~ 0.0326607 kglkg dr air 83971.38-1.005?4168.581 y The enthalpy of the inlet air, ima" is found according to equation (A.3.6b) with cpa' = 1006.59 JlkgK and Cp,' = 1875.204 JlkgK being evaluated at (Ta, + 273.15)/2 = (302.8437 + 273.15)/ 2 = 287.9969K according to equations (A. 1.2) and (A.2.2) respectively. The latent heat is fou nd to be ifgwa = 2.5016 X 10' JlkgK according to equation (A.4.5) at 273.15K. With these values find I;'a,' = 113412.1 Jlkg dry air.

The air vapor mass flow rate is m = mwcpwm(Twi -Two) = (12500)(4178.721)(313.15-292.5375L 13928.97k /s a,l, Vma,' -i mal ) 113412.1-36114.76 g 1.4 THE ENERGY EQUATION The values for Two, Tas, mavlS, PaS, Pa6 and Ws, determined in the previous sect ion, are only initial approximations. This problem can only be solved by following an iterative proced ure to obtain a solution 1.6 that will satisfy both the energy and draft equations. The choice of an air-vapo r mass flow rate of rna," = 16966.47 kg/s through the fill will satisfy these equations, giving correspondin g pressures of PaS ~ 83937.04 Pa and pa' ~ 82650.57 Pa, and an air temperature of Ta, ~ 299.85626K. T he mean temperature of the recooled water in the basin is T.a = 294.5572K. The humidity of the super saturated air above the spray zone is w, ~ 0.027888 kglkg dry air with w,.s ~ 0.027176 kg/kg dry air. At the specified air inlet drybulb temperature of Ta, ~ 288.6K, wetbulb temperat ure of T.b ~ 284.2K, find the following thermophysical properties employing the equations given in appendi x A. Density ofair-vapor Viscosity of the air vapor mixture Pa,' = 1.0101 kg/m3 /la" = 1.7857 x lO's kg/ms (A.3.1) (A.3.3) (A.2.1) (A.3.1) (A.1.3) (A.2.3) (A.3.3) If the air is assumed to be supersaturated immediately after the drift eliminato r, the wetbulb temperature at 5 will be equal to the given drybulb temperature Tas = 299.8563K at this elev ation. The corresponding thermophysical properties at 5 can be determined accor~ing to the equations give n in appendix A. Saturated vapor pressure P", = 3Sil3.482 N/m' Density of air-vapor .A", =0.95964 kg/m' Dynamic viscosity of air /las ~ 1.8462xl0's kg/ms Dynamic viscosity of vapor j.l,s = 1.0041xl0's kg/ms Dynamic viscosity of air-vapor /la,s ~1.8l82 xl0" kg/ms The enthalpy of the supersaturated outlet air, i"s, is found according to equati on (B.33) with cpas ~ 1006.5548 JlkgK and Cp,S = 1873.9329 J/kgK being evaluated at (Tas +273.15)/2 =( 299.8563 +273.15)/2 ~ 286.50315K according to equations (A.l.2) and (A.2.2) respectively. The latent heat is found to be if.., = 2.5016 x 10' J/kgK according to equation (A.4.5) at 273.15K. With these values find i"s = 96303.4766 Jlkg dry air. The approximate harmonic mean density of the air-vapor in the fill is given by 2 Pa," = 1 1 --+-? Povl Pav5 _--,-__2_-;-_- 0.98424 kg/m

' 1 1 ---+ --,-----:-c 1.01012 0.95964 The dry air mass flow rate can be determined from the fullowing relation: rna," =[ma(l+w,)+ma(l+w,a,)] or rna =2m",,, /(2 + w, + W,as) ~ 2(16966.47)/(2 + 0.008127 + 0.027176) ~ 16672.19 kg/s The respective air-vapor mass flow rates upstream and downstream of the fill are thus m", =rna (1 + w,) ~ 16672.19(1 + 0.008127) ~ 16807.68 kg/s and 1.7 m." = m. (1+ w,.,) ~ (I + 0.027176) ~ 17125.27 kg/s The corresponding mass velocities are G"" =ma,t, I Afr = 16966047/8300 = 2.04415 kg/m's G a = m.lAfr = 16672.1918300 = 2.00870 kg/m's G"t =m"t I Afr = 16807.68/8300 = 2.02502 kg/m's G", = m", I A/, = 17125.2718300 = 2.06329 kg/m's According to equation (A.4.2) the specific heat of water cpwm = 4178.32 J/kg at the mean water temperature of (To; + To.)/2 ~ (313.15 + 294.5573)12 = 303.8536K. At the mean outlet temperature ofthe water Too = 294.5573K find Density of water A. = 997.8629 kg/m3 Surface tension 0;.. = 0.07256 N/m The mass velocity for the water based on the frontal area of the fill is G. = m. I Afr = 12500/8300 = 1.50602 kg/m's (Ao4.I) (Ao4.7) The transfer coefficients can be determined with the above values. To find the t ransfer coefficient in the rain zone, use equation (D.20). The "a" coefficients appearing in the equation f or the rain zone transfer and pressure drop coefficients are as follows: ap =3.061x 10-6(p:" g 9icY"" )0." =3.061x 10-6(997.8629'9.8' 10.07256)??" =1.000 04 ap = 998 I Po. = 998/997.8629 ~ 1.00014 a, = 73.298(g'a':" Ip,:,,)02S = 73.298(9.8 5 X 0.07256 3 1997.8629y-'5 = 1.0008 aL=6. 122(gaoo IPo.)02S = 6.122(9.8x 0.072561 997.8629)?'" = 1.00025 Other quantities required to evaluate the rain zone transfer coefficient are: The humidity ratio of saturated air at To. W,I = 0.019539 kg/kg Diffusion coefficient at inlet conditions D, = 2.29972xI0?' m'/s Furthermore, the Schmidt number is SCi = Pa,' 1(P.,Pt) = 1.7857xI0?'/(1.01012 x 2.29972xlO"') = 0.76865 and the air-vapor velocity before the fill V"3 =m", 1(P..,Afr) = 16807.68/(1.01012x8300) = 2.00473 m/s With these values find (A.3.5) (D.21) 1.8 12( D, J(!!.J-)( Pa' )Sc~'''[ln(W" +0.622)I(W" -w,)] V"",dd Add PwoR.Tal WI +0.622 ~.55 + 41.7215(aLdd )"-8004'}t>.713 + 3.741(aLH3rl."456} x + 4.04016 x {:1.1Iexp(O.l5avv""3)- 3.l3} { ( ) { 0.371gexp(0.0019055aA)}}

xexp 5.375gexp -0.2092aLH 3 In +0.55 [ IJ0.019539+0.622)] =12 2.29972x10' ~ 84100 076865,.33 ll\0.008127+0.622 (2.00473XO.0035X0.0035X997.8629X 461.52x 288.6) . (0.019539-0.008127) 0.90757 x1.00014 x 1.01012 - 30341.04x 1.00004 x1.7857 x10-5 - 0.37564 ~.55 + 41.7215(1.00025 x 0.0035)"-'?043 }t>.713 + 3.741(1.00025 x IOtI.23456 } X + 4.04016 x {3.11 exp(O.l5 x1.0008 x 2.00473<)- 3.l3} j5.375gexp(- 0.2092 x1.00025 x 10) )x exp x 1n{0.371gexp(0.0019055 x 1.00025 x 1 04.5}l +0.55 f = 0.4150354 The Merkel number applicable to the fill is specified in the form ofequation (D. 24) h a Mepft = dft ft = 0.27928G~'094G~6023G. = 0.27928 x 1.878 x (1.50602)"??094 x (2.00870)??6023 = 1.024240 The transfer coefficient in the spray zone is given by (D.23) ( ) 0.5 ( )"0.2L Ga =0.2xO.5 2.00870 . =0.11549 'p G. 1.50602 The total transfer characteristic of the cooling tower is h~anH3 h~aftLft hilipa~L~ Me = Men + MePfl + Me,p = + .+ --=---"'--"'-Gw Gw Gw = 0.4150354 + 1.024240 + 0.11549 = 1.55476 1.9 The empirical relations for the transfer characteristics of the rain zone and sp ray zone were derived by employing the Merkel approach. However, the Poppe approach is used to evaluate c ooling tower performance. The transfer characteristic of this particular fill, determined by the Poppe approach, is approximately 7% larger than that determined by the Merkel approach. Thus, the t ransfer characteristics of the spray zone and rain zone were increased by 7% in another investigation by the author. This was done to determine the influence of the inconsistent application of the fill perf ormance characteristics to the cooling tower performance evaluation. The fill performance characteristics a ccording to the Poppe approach were employed in that investigation. It is found that cooling tower performance, determined by the Poppe approach, wh ere the transfer characteristics of the spray zone and rain zone are increased by 7 %, is practic ally identical to the cooling tower performance where the transfer characteristic ofthe spray and rain zones a re not increased by 7 %. The results are practically identical across a broad range of ambient temperatur es and humidities. Thus, the transfer characteristics of the rain zone and spray zone, obtained by the Me rkel approach, can be employed in cooling tower performance evaluations while employing the Poppe appr oach. The transfer characteristics of the fill, however, must be obtained from the Poppe approach, if the Poppe approach is employed in the cooling tower performance evaluation. The inconsistent applicati on of the transfer characteristics of the rain and spray zones is possible because these characteri

stics are respectively approximately only 27 % and 7 % ofthe total transfer characteristic. Thus, due to the uncertainties associated in converting the transfer characteris tics of the rain zone and spray zone, obtained from the Merkel approach, to that of the Poppe approach, by increasing it by 7 % and the fact that the Merkel approach gives conservative results, can the empiri cal relations of the transfer characteristics of the rain zone and spray, zone be used inconsistently without compromising solution accuracy. 1.5 TRANSFER AREA ANALYSIS ACCORDINGTO THE POPPE APPROACH The fourth order Runge-Kutta method is used to solve the governing equations. As already mentioned, the fill is divided into five intervals with an equal water temperature difference a cross each interval. Refer to figure 1.3 for a layout of the transfer area, i.e., the fill rain zone and spray zone, that is divided into five intervals. Refer to appendix B.4 for a detailed discussion on the application of the RungeKUlla method to the governing equations of the Poppe approach. Refer to appendix G for the conventio n used for the subscripts. The Runge-Kutta method is an initial value problem; therefore, the initial value s at level (0) are as obtained previously w(O) = WI =0.008127 kg/kg dry air ima(Oj =imal = 36114.71 J/kg dry air LlO Since the transfer area is divided into two intervals find from equation (B.66) l;.Tw~ (T.; - Two)/(Number of intervals) ~ (313.15 - 294.5572)15 =3.719K Level (5) Interval 5 W(4), lma(4p ma Level (4) ---------------1"----------Interval 4 :- :- :- : w(3), ima(3), ma Level (3) ~~:~:---------------1"-~t~~~13Level(2) ~~~~:---~~~~~~~2~-:':-1"----------:- :- :- : Interval 2 Level(l) ~~~~:---~~~~~:~~-:':-1"----------:: ::::- Interval I Level (0) y..:..;c-----t------' Figure 1.3: Total transfer area of a counterflow tower, divided into five interv als. From equations (B.51) to (B.53) find for the first interval of the Runge-Kutta m ethod, w(I) = w(O) + UCI,I) + 2j(l,2) + 2jCI,3) +jCI,4?)l6 Imocl )= ImoCo) + (k(l.I) + 2kCI ,2) + 2k(l.3) + A{1,4?)/6 Mep(I) ~ Mep(o) + (/(1,1) + 2/(1)) + 2/CI ,3) + /(1,4))16 where Mep(o) is equal to zero at the air inlet side or at level (0). Commence with the first intermediate calculation step of the Runge-Kutta method for the first interval. It can be seen from equations (B.54) to (B.56) thatj(I.I), k(l,l) and /(1,1) are fu nctions of TwCo), im.(o) and w(O). Define that TwCI,I) = Twco)~ Two = 294.5572K W(I,I) ~ w(O) = WI ~ 0.008127 kg/kg dry air Im'CI,I) ~ Im.(o) = lmal ~ 36114..71 Jlkg dry air.

To calculate jCI.I), k(l,I) and /(1,1) in equations (B.51) to (B.53) respectivel y, the specific heats have to be evaluated at (Tw(I,I) + 273.15)/2 =(294.5572 + 273.15)12 =283.8536K Specific heat of dry air from equation (A.1.2) Cpa(I.I) ~ 1.045356 x 10' - 3.l61783xl 0.1(283.8536) +7,083814 x 10'4 (283.8536) ' I.11 - 2.705209 X 10-7(283.8536)' = 1006.497 JlkgK Specific heat ofwater vapor from equation (A.2.2) Cp,(l.I) = 1.3605 x 10' + 2.31334 (283.8536) - 2.46784 x 10-10 (283.8536)' +5.91332 x 10.13 (283.8536)" = 1871.695 JlkgK Specific heat ofwater from equation (A.4.2) cpw(l.1) = 8.15599 x 10' - 2.80627 x 10 (283.8536) + 5.11283 x 10"2(283.8536)2 - 2.17582 X 10-13 (283.8536)' = 4196.035 JlkgK Pressure of water vapor from equation (A.2.1) evaluated at T(I.I) = 294.5572K. Z(I,I) = 10.79586(1 - 273.16/294.5572) + 5.0280810gI0(273.161294.5572) + 1.50474 x 10-4[1 - 10"?29692{(294.5572I27,.16)-II] + 4.2873 x 10"[10 4.76955(1-273.16129 4.l572) - I] + 2.786118312 = 3.40626 P>(I,I) = 10,?57157 = 2548.33 Pa Humidity ratio for saturated air at T(I,I) = 294.5572K from equation (A.3.5) w =( 0.62509(2548.33) ) = 0.019536 kglkg dry air w'(I,I) 84100-1.005(2548.33) Latent heat at 273.15K follows from equation (A.4.5) ifgw(I,I) = 2.5016 x 10' Jlkg The enthalpy ofwater vapor at the local bulk water temperature, TW(I,I), relativ e to water at O?C, i>(I,I) = ir...(I,I) + Cp>(I.I) TW(I,I) = 2.5016 X 10' + 1871.695 (294.5572-273.15) =2541662 J/kg The enthalpy of saturated air at the local bulk water temperature from equation (A.3.6b) ima.w(I,I) = 1006.497(294.5572 - 273.15) + (0.019536)(2541662) = 71198 Jlkg The Lewis factor from equation (B.15) ( 0,662+0.019536 1) Ie - 0865?.'67 0.662+0.008127 = 0.9160 [(1,1) - ? In(0.662+0.019536) 0.662 + 0.008127 The mass balance from equation (B.32), ( rn W J 12500 [1 rn o (1,1) =16672.19 16672.19 (0.027889 -0.008127)1= 0.730 12500 J 1.12 From equation (B.54) find j(l.I) = tiT? . f{T.(o),i~(o)' w(oJ= tiTw. f{T.(I.I),ima(l.')' W(I.IJ From equation (B.48) find :; = f(Tw(o) ,ima(o)' w(oJ= f(Tw(I,ll' ima(l,I)' W(I,I?) W but from equation (B.24), CPW(I,,)(m W J (W,W(',,) - W(I,IJ dw ma (1,1) -=-------r---;r-'-'-'------r--------.,.---, dTw immW(I,I) - ima(l,l) + LefO,I) -1 imasw(I,I) - ima(l,l) - WSW(I;J) - WO,I) ' 10'(1,1) - (w~(I,I) - W(I,I) ~pw(I,llw(I,I) Combine equations (B.24), (B.48) and (B.54) to find tiTwcpw(I,I)(mwJ (~~(I,') - W(I,I?) . ma (1,1) J(1,1) = -ima-~-(I,-,)---i-ma-('-,'-)+-r(L-'--e-

f -(I,-,)--....'-..1~~.~.:...o.:("",,_) ---ima-(I-,I)---c(,W-",-",-)--W-('-,'l-':} -'v(-',,-)J - (w~(I,I) - W(I,I)~pw(l.Ilw(l,I) = (3.719)(4196.035)(0.730)(0.019536 - 0.008127) =0.003874 [ 71198-36114.71 ] + (0.9160-1)[71198- 36114.71-(0,019536-0.008127)(2541662)] - (0.019536- 0.008127)(4196.035)(294.5572- 273.15) Combine equations (B.25), (B.49) and (B.55) to find. k(I,I) =tiTwCpw(l,,)(mw) mu (1,1) 'W -w L Tx 1+ \ sw(l.l) (1,1)~pw(I,I) w(I,I) ima~(I,I) - ima(I.I) + Lef(l,I) -1 imasw(I.ll - ima(I.I) - W ~(l,l) - W(I,I) - (w~(I,I) - W(",)~pw(I,llTw(I,I) = (3.719X4196.035)(O.730) x 1+ (0.019536-0.008127)(4196.035X294.5572-273.l5) [ 71198-36114.71+ (0.9160-1)x ] [71198-36114.71-(0.019536-0.008127)(2541662)] - (0.019536 - 0.008127)(4196.035)(294.5572 - 273.15) C01Dbine equations (B.30). (B.50) and (B.56) to find = 11739.34 I.13 I = aTwcpw(l,t) (1,1).. 1 . . 'maww(I,I) -'ma(I,I) + e1(1,1) - lmasw(l,l) -'ma(l,l) - W $W(I,I) - W(I,I) V(l,! ) - (WSW(i.l) - W(I,I)~PW(I,I)TW(I,I} = (3.719)(4196.035) = 0.4651566 [ 71198-36114.71 ] +(0.9160 -1)[71198 - 36114.71-(0.019536 -0.008127)(2541662)] - (0.Q19536 - 0.008127)(4196.035X294.5572 - 273.15) By proceeding along the same lines j(I.2), k(l,2) and [(l.2) are determined for the second intermediate calculation step of the Runge-Kutta method for the first interval: From equations (B.57) to (B.59) can be seen thatj(I)), k<1.2) and [(l,2) are fun ctions of T aTw ? k(I,I) j(I.I)w(O) +2,lma(0) +2 and w(O) +-2- thus define TW(I.2) = Tw(o) + IiTw/2 = 294.5572 + 3.719/2 = 296.4147K W(I,2) ~ w(I) +j(l.ly2 ~ 0.008127 + 0.003874/2 = 0.010064 kglkg dry air , ima(I,2) ~ ima(l) + k(I,I/2 = 36114.71 + 11739.34/2 = 41984.43 J/kg The specific beats have to be evaluated at (Tw(I,2) + 273.15)/2 ~ 284.7823K Specific beat of dry air from equation (A.1.2) Cp.(1.2) = 1.045356xl0' - 3.l61783xl0?1(284.7823) +7.083814xl0?4 (284.7823)2 - 2.705209xl0?7 (284.7823)3 = 1006.516 J/kgK Specific heat of water vapor from equation (A.2.2) C,>(1.2) = 1.3605 x 103+ 2.31334 (284.7823) - 2.46784 x 10.10 (284.7823)' + 5.91332 x 10.13 (284.7823)' = 1872.478 J/kgK Specific heat of water from equation (A.4.2) C,w(I.2) = 8.15599 x 10' - 2.80627 x 10 (284.7823) + 5.11283 x 10.2(284.7823)' - 2.17582 x 10.1' (284.7823)' = 4194.72 J/kgK The vapor pressure and humidity ratio of saturated air are calculated at the loc al water temperature TW(I,2) = 296.4147K Vapor pressure from equation (A.2.1): P,(l)) ~ 2853.55 Pa Humidity ratio for saturated air from equation (A.3.5): W",(1,2) ~ 0.021958 kglk g dry air

The enthalpy of water vapor at the local bulk water temperature, Tw(I.2) ' relat ive to water at ODe, i'(1,2) = irgw(I,2) + Cpv(l,2) Tw(1,2) = 2.5016 x 10' + 1872.478(296.4147-273.15 ) ~ 2545161 Jlkg I.14 The entalpy of saturated air at the local bulk water temperature from equation ( A.3.6b) ima",(I.2) = 1006.516(296.4147 - 273.15) + (0.021958)(2545161) = 79304 J/kg The Lewis factor from equation (B.l5) ( 0.662+0.021958 1) Le = 0.865?"., 0.662+0.010064 ~ 0.9164 [(1,2) In(0.662+0.021958) 0.662+0.010064 The mass balance from equation (B.32) (rn.) = 12500 (1 16672.19 (0.027889-0.010064))= 0.7319rna (1,2) 16672.19 12500 From equation (8.57) find 1'(1.2) = !'1T? ? f(T.(I,2)' ima(I,2) ' WO?2?) From equation (B.48) fmd :; = f(TW (I.2)'i",a(I,2) , W(I.2J ? From equation (B.24) find v([,2) dw C pw(I.2)(:.J (W'.(I.2) - W(!,2?) = -,.-__........,ar-"(,"'2)'--- --r ----._---, tiT. i"''''.(1.2) - ima(I.2) + Le[(1,2) -1 i""".([,2) - i"",(I,2) - W",(1,2) - W ([,2) - (W",([,2) - W(I,2) ~pw([,2l.o,2) Combine equations (B.24), (B.48) and (B.57) to find !'1TwcPW(1,2/ m. J (W'.(1.2) - W(I,2?) . lma (1,2) J (1,2) =-i""""--(I,-2)---I-''''''-(-I,2-)-+'(L-e-j-([-,,-)---1~-?,""",-'-('-I,, '-)---i-=-(l-.,-) --'(w-",-([.-,)---w-([.-,)')-?,(-[.,-)1 - (w",([.,) - W(I.') ~pw(I"lw(J,,) = (3.719)(4194.72)(0,7319)(0.021958-0.010064) = 0.00381818 [ 79304-41984.43 ] + (0.9164-1)[79304- 41984.43 -(0.021958 -0.010064)(2545161)] - (0.021958 - 0.010064)(4194.72)(296.4147 -273. I 5) A relation for ~,.') is obtained by combining (8.25), (B.49) and (B.58) i.e. U5 k(l,2) =/iTw Cpw(I,2) (rnw ) rna (1,2) X~+ (W~(I,2) -W(I,2)fpw(I,2lw(I,2) ] i~(I,2) - i...,(I,2) + Le j(I,2) -1 ?~(I.2) - i""(I.2) - W~(1.2) - W(I.2) "(1.2) - (W'W(I,2) - W(I,2)~pw(I,2lw(I,2) = (3.719)(4194.72)(0.7319) X 1 + (0.021958-0.010064)(4194.72)(296.4147 - 273.15) = 11790.64 [ 79304-41984.43+(0.9164-I)X j [79304 - 41984.43 - (0.021958- 0.01 0064)(2545161)] - (0.021958 - 0.01 0064)(4194.72)(296.4147 - 273.15) hi) is obtained by combining equation (B.30), (B.50) and (B.59) 1 _ /iTwcpw(I,2) (1,2)- .. L I 'masw(I,2) -}ma(l,2) + e[(1,2) - 'nulSW(I,2) - jma(I,2) - W sw(l,2) - W(I,l) '\1 (1,2) - (W~(1.2) - W(I,2)~PW(I,2)Tw(I,2) = (3.719)(4194.72) = 0,438578

[ 79304-41984.43 ] + (0.9164-1)[79304- 41984.43 - (0.021958-0.0 I0064)(2545161)] - (0.021958 - 0.01 0064)(4194.72)(296.4147 - 273.15) Proceeding along the same lines, the following values are calculated to finish t he Runge-Kutta numerical integration fur the first interval. }(I) = 0.00382916; 1?1.3) = 11791.28; 1(1,3) = 0.4388265 }(1,4) = 0.00380129; k(I,4) = 11844.59; 1(1,4) = 0.4080234 The humidity ratio at level (I) fullows from equation (B.51), W(I) = w(O) + V(I,I) + 2}(1) + 2}(I,3) +}(1,4))/6 = 0.008127 + [0.003874 +(2)0.00381818 + (2)0.00382916+ 0,00380129] 16 =0.011955 kglkg dry air The enthalpy of the air at level (1) fullows from equation (B.52), ima(l) = ima(o) + (k(1.I) + 21?1.2) + 2k(1,3) + 1?1,4?/6 = 36114.71 + [11739.34 + (2)11790.64 + (2)11791.28 + 11844.59] /6 = 47906.31 Jlk g The transfer characteristic or Merkel number at level (I) follows from equation (B.53), Mep(l) = Mep(o) + (/(1.1) + 21(1) + 21(1,3) + 1(1,4?/6 = 0 + [0.4651566 + (2)0.438578 + (2)0.4388265 + 0.4080234] = 0.4380259 p,(!) = 1578.146 Pa w(t) = 0.011955 kglkg dry air I.I6 The dry bulb temperature, Ta(l), and wet bulb temperature, T.h(!), at level(l) a re determined by assuming that the air is unsaturated. If T.h(!?Ta(!) the air is supersaturated and the as sumption of unsaturated air must be corrected. The assumption is then corrected by assuming supersaturated a ir with T'h(l) = Ta(I)' Find the dry bulb temperature at level ell, T.O): The enthalpy of the air at level (I) is, ima(!) ~ 47906.3 I Jlkg. Assume that the air is unsaturated and that the drybulb temperature, Ta(!) ~ 290 .6444K. The specific heats are evaluated at (Ta(I)+273.15)12 = (290.6444+273.15)12 = 28I.8972K Specific heat of dry air from equation (A.U) Cpa(l) = 1006.458 J/kgK Specific heat of water vapor from equation (A.2.2) cp>'(!) = 1870.054 JlkgK Equation (A.3.6b) gives an expression for the enthalpy ofan air vapor mixture pe r unit mass of dry air. ima(l) = 1006.458(290.6444 - 273.15) + 0.011955 x[2501598 + 1870.054(290.6444 - 273.15)] = 47906.31 Jlkg dry air The value of ima(t) determined by equation (AJ.6b) is equal to the value determi ned by equation (B.52). The assumption ofthe value of the dry bulb temperature is therefore correct. Find the wet bulb temperature at level (1). T'h(lh The humidity ratio at level (1), w(1) ~ 0.011955 kglkg dry air. Assume that the wet bulb temperature at level (1) is T.h(1) =286.9630K. From equations (A.2.l) and (A.3.5) find at T.h(!) = 286.9630 K respectively the vapor pressure and the humidity ratio. Vapor pressure from equation (A.2.1): Humidity ratio from equation (A.3.5): The value ofw(l), determined by equation (A. 3. 5) is equal to the value determi ned earlier. The assumption of the value of the wet bulb temperature is therefore correct. T.h(!) < Ta(l) and hence the air is still unsaturated. By following the same procedure as above find the following the values for the r elative humidity, air enthalpy, Merkel number and air temperatures on the next three levels in figure 1.3,

W(2) ~ 0.015760627 kglkg dry air; ima(2) = 59806.97 J/kg Me(2) = 0.8126387; Ta(2) ~ 292.8226K; T'h(2) = 291.1911K T.h(2) < Ta(2) and hence the air is still unsaturated. W(3) = 0.019614896 kglkg dry air; ima(l) ~ 71826.70 Jlkg ~: 1.17 Me(,) ~ 1.119698; Ta(,) = 294.9653K; Twb(,) = 294.6191K TWb(') < Ta(,) and hence the air is still unsaturated. W(.) ~ 0.0235572 kglkg dry air; ima'(4) = 83973.91 J/kg Me(4) = 1.364207; T a(4) = 296.986IK; TWb(4) '" 297.5420K TWb(') > T a(4) and hence the air is supersaturated. The assumption that the air is unsaturated for the determination of Ta (4) and TW b(4) is therefore incorrect. This can he correcte d by assuming that the air is supersaturated where Ta(4) ~ TWb(4)= 297.4155K From equation (B.52) the enthalpy at level (4) is equal to ima(4) = 83973.91 J/k g The partial pressure and humidity ratio of saturated air evaluated at Ta(4), Partial pressure from equation (A.2.1): p"a(') ~ 3030.684 Pa Humidity ratio from equation (A.3.5): W,a(4) '" 0.0133173 kglkg dry air The following specific heats are determined at (Ta(4) + 273.15)/2 '" (297.4155 + 273.15)/2 ~ 285.28275K Specific heat of dry air from equation (A.I.2): Specific heat of water vapor from equation (A.2.2): Specific heat of water from equation (A.4.2): Cpa(') '" 1006.527 J/kgK cp,(') '" 1872.9 J/kgK cpw(') '" 4194.033 J/kgK From equation (B.33) it follows that ;"(4) ~ Cpa(4Ta(') - 273.15) + W,a(4)[i,,,wa + Cp,(.)(Ta(4) - 273.15)] + (W(.) W,a(4)Cpw(4Ta(4) - 273.15) '" 1006.527(297.4155 - 273.15) +0.0133173[2501598 + 1872.9(297.4155-273.15)] + (0.0235572 - 0.0133173)(4194.033)(297.4155 - 273.15) '" 83974 J/kg i,,(.), determined by equation (B.33), is within close tolerance of i.a(4)' dete rmined by equation (B.52), thus, the assumption of the value of the dry bulb te",perature at level (I) is t herefore correct. The air temperature at level (4), Ta(4? is therefure equal to 297.4155K. Apply Runge-Kutta numerical integration to the fifth interval. Find from equatio ns (B.51) to (B.53) for the fifth interval, we') '" W(4) + V('.I) + 2j('.2) + 2j(,.,) + j('.4?)/6 i,,(,) '" i"(4) + (k('.l) + ~'.2) + 2k(,.,) + kt.s.4)/6 Mep(s) '" Mep(s) + (/(S.I) + 21('.2) + 21(s.,) + l(s.4)/6 Because the air is supersaturated 1~'(4) '" i.a(.). The water temperature at lev el (I) is T w(4) ~ T w(') + I'>T. '" 305.712 + 3.719'" 309.4311K. For the first intermediate calculation step of the fifth interval, i,,(,.I) '" i"(4? Tw(5,1) = TW(4) and Wsa(5,1) =Wsa(4). U8 To calculate j(S.Ij, k(s.l) and h5.1) for the first intermediate calculation ste p for the fifth interval certain thermophysical properties have to be evaluated at (TwC5?1) + 273.15)/2 = (309.43 11 + 273.15)/2 = 291.2905K Specific heat of dry air from equation (A.1.2) Cpa(5.1) = 1.045356 x 10' -3.161783xIO'I(291.2905)+ 7.083814 x 10-4(291.2905)' - 2.705209 x 10'7 (291.2905)' = 1006.676 JlkgK Specific heat of water vapor from equation (A.2.2) Cp>(5.1) = 1.3605 X 10' + 2.31334 (291.2905) - 2.46784 x 10,10 (291.2905)5

+ 5.91332 X 10.13 (291.2905)' = 1878.043 J/kgK Specific heat ofwaler from equation (A.4.2) Cpw(5.1) = 8.15599 X 10' - 2.80627 x 10 (291.2905) + 5.11283 x 10'2 (291.2905)' - 2.17582 x 10.13 (291.2905)' = 4186.92 JlkgK Vapor pressure from equation (A.2.1) evaluated at T.e5,1) = 309.43 11K Ze5.1) = 10,79586(1 - 273.16/309.4311) + 5.02808 10810(273.16/309.4311) + 1.50474 x 10-4[1 _ 10?,?29692!('09.43I1I273.l6).!)] + 4.2873 x 10-4[10 4.769"(1-273.161'09.4311) - I] + 2.786118312 = 3.780603 P>(5.1) = 10,?78060' =6033.964 Pa Humidity ratio for saturated air evaluated at T.e5.1)= 309.43tJK from equation ( A.3.5), w =( 0.62509(6033.964) ) = 0.0483338 kg/k dry air "(5,1) 84100-1.005(6033.964) 8 The enthalpy of water vapor at the local bulk water temperature, T.es?I ), relat ive to water at O?C, i>(5.!) = iJKw('.I) + Cp>(5.1) T.e5?1) = 2.5016 x 10' + 1878.043(309.4311-273.15 ) = 2569736 Jlkg The entalpy of saturated air at the local bulk water temperature from equation ( A.3 .6b) i,,_c'.I) = 1006.676(309.4311 - 273.15) + (0.0483338)(2569736) = 160728.4 J/kg The Lewis factor from equation (B.38) ( 0.662 + 0.0483338 I) Le = 0865,?'67 0.662 +0.0133173 = 0.92529 f('.I)? In(0.662 + 0.0483338) 0.662+0.0133173 The mass balance from equation (B.32) Fl."f'''' ."" ( m w ) _ 12500 (I m, ('.I) 16672.19 I.19 16672.19 (0.027889-0.0235572)1= 0.74542 12500 ) From equation (B.54) find i(,.]) = !'J.Tw ? f(Tw(4) , i"(4)' W(4?)= !'J.Tw. f(Tw('.') ' i"(,,,), W('.I?) From equation (B.48) find :; = f(Tw(4) , i"(4) ,W(4J= f(Tw(",)'i"(,,,), W('.IJ w but from equation (B.42) Combine equations (B.42), (B.48) and (B,54) to fmd !'J.Twcpw(',I) ( mw ) (W'W(5,]) - W,a(',I?) i(5,,) = 1m., (',I). 1 r-----l"] . _. (L -I Imasw(S,l) -1&s(5,1) -\WSW(5.l) -wsa(S,l)}v(S,I) I 1""""(',1) 1"(',1) + ef(5.1) ( \. + W(S,I) -Wso(S,l)PPW(S,I)Tw(S,I) + (w(,,]) - W"'(5,1)~pw(5,]lw(5,1) =-=--__-'-(3_,7_1-'9)'-'.(4_1_86_.9_2.:..:)(_0,_74_5_42.:..;)(:..,0,_04_8_33_3_8 -_0_.0_1_33_1_73-,-)__---=: 160728.4 -83973.91 +(0,92529-1)[160728.4 - 83973.91- (0.0483338-0.0133173)(2569736) +(0,0235572-0.0133173)(4186.92)(309.4311- 273.15)] + (0.0235572- 0.0483338)(4186.92)(309.4311- 273,15) = 0.0040213\ Combine equations (B.43), (B.49) and (B.55) to find, =!'J.TwCPW(5,])(mw) ma (',1) 1.20 IW -W )x 1+ \ sw(5,1) .1'0(5,1)

. _. (L -lll""""(,,,) -I"(S,I) -(W~("I)-W'Q(S'l)}~Jlmarw(~,l) 'SS(S,l) + ef (5,1 ) ( \. + W(5,l) -WSQ(5.1)~PW(S.I)TW(S,1) + (W("l) - W~e',l) ~pw(s,I)Twe',!) ~ (3.719)(4186.92XO.74542) X (0.0483338 - 0.0133173X4186.92)(309.4311- 273.15)1+ ."--,-,-.,...,...~--,-,---,--------'--'---'-'---------"----------.,,. 160728.4 - 83973.91 + (0.92529 -1)[160728.4 - 83973.91- (0.0483338 - 0.0133173X2569736) + (0.0235572 - 0.0133173)(4186.92)(309.4311- 273.15)] + (0.00.0235572 - 0.0483338)(4186.92)(309.4311- 273.15) = 12217.76 Combine equations (B.47), (B.50) and (B.56) to find, . _. (L _1{1""",,(S'!) - 1,,(s.1) - (w~CS.I) - W,QC,.[SV(S.I)] 'mosw(5,1) 'ss(S,l) + e1(5,1) ( \. T + W(S,I) - W'QCS.I) Fpw(S,!) w(S,I) + (WCS ,[) - W~(S,l) ~pwcs,I)Tw(S,l) (3.719)(4186.92)=-=- -2._-'--'--_--''-- ...". 160728.4 -83973.91 + (0.92529-1)[160728.4- 83973.91-(0.0483338- 0.0133173)(2569736) + (0.0235572 - 0.0133173)(4186.92)(309.4311- 273.15)] + (0.0235572-0.0483338)(4186.92)(309.4311- 273.15) = 0.2161235 By proceeding along the same lines, the following values, are calculated for the second to fourth intermediate calculation steps to finish the Runge-Kutta numerical integration f or the fifth interval. jC5,2) = 0.00433930; k(5,2) = 12328,45; /(5,2) ~ 0.01900152 jCS~) ~ 0.00434170; ~"') = 12331.31; /c"') ~ 0.1900797 jes,4) ~ 0.00460331; ~S,4) = 12440.13; l(s.4) = 0.1670177 The humidity ratio at level (5) follows from equation (B.51), W(5) ~ w(') + U(',I) + 2j(5,2) + 2j(s,,) + je5,4))/6 = 0.023557229 + [0,004020131 +(2)0.00433930 + (2)0,00434170+ 0.00460331] 16 , ~ 0.027888 kglkg dry air = w,. The value ofWs is equal to the value given initially in the, example. The air enthalpy at level (5) follows from equation (B.52), iV'c') ~ i"(4) + (kc"I) + 2kC,,2) + 2~s.,) + k(s.4)/6 = 83973,91 + [12217.76 + (2)12328.45 + (2)12331.31 + 12440.13] 16 ~ 96303.48 Jlk g Cpa(l) ~ 1006.555 JlkgK Cp>(l) ~ 1873.932 JlkgK Cpw(l) ~ 4192,421 J/kgK 1.21 The transfer characteristic or Merkel number at level (5) follows from equation (B.53), Mep(l) ~ Mep(4) + (1(1.1) + 21e",) + 21(1.3) + 1(1,4?/6 ~ 1.364207 + [0,2161235 + (2)0.01900152 + (2)0.01900797 + 0.1670177] ~ 1.554762 This value is almost identical to the value obtained by adding the transfer coef ficients in the three wet zones which means that the water outlet temperature, Two ~ 294.5572K is correct. The air was already supersaturated at level (4), Therefore assume that the air i s still supersaturated and that Ta(l) ~ TWb(I)~ 299.855K The partial pressure and humidity ratio of saturated air, from equations (A,2.1) and (A.3.5), evaluated at Ta(l) are respectively P"a(l) ~ 3503.218 Pa W..(I) ~ 0,027176 kglkg dry air The value of W..(I) is equal to the value ofw.., given initially in the example. The specific heat of dry air, water liquid and vapor are evaluated at (Ta(l) + 2

73.15)/2 ~ (299.855 + 273.15)/2 ~ 286.5025K Specific heat of dry air from equation (A. 1.2): Specific heat ofwater vapor from equation (A.2.2): Specific heat of water from equation (A.4,2): It follows from equation (B.33) that i,,(I) ~ cpa(I)(Ta(l) - 273.15) + w..(I)[ifgw(l) + cp>(I)(Ta(I)-273.15)] + (W(I) - w..(I?cpw(I)(Ta(l) -273.15) ~ 1006.555(299,855 - 273.15) + 0.027176[2501598 + 1873.932(299.855-273.15)] + (0,027888 - 0.027176)(4192.421)(299,855 - 273.15) ~ 96303 J/kg i,,(I), determined by equation (B.33), is within close tolerance ofi,,(I), deter mined by equation (B. 52), thus the value of the air temperature assumed at level (5) is therefore correct. Ta(l ) ~ Tol ~ 299.85K is also within close tolerance of the temperature given initially in the example. Therefore the conditions at the outlet of the fill according to the Poppe approa ch are: im05 ~ i,,(I) ~ 96303.48 Jlkg W, ~ W(I) ~ 0.027888 kglkg dry air W..I = W..(I) ~ 0.027176 kg/kg dry air Tal ~ Ta(l) ~ 299.8563K where the Merkel number is Mep ~ Mep(l) ~ 1.554762, The heat rejected by the cooling tower is given by Q=ma(imal - imaJ~ 16672.19(96303.48 - 36114,71) ~ 1003.4775 MW r.? If "'". '.:: ,,;:'~\'" \ ,\' 1.22 The path of the air through the cooling tower, predicted by the Poppe approach, is shown in figure LA. Atmospheric pressure: 84100 Pa I 10L ~90 .... 90 70710/ / .-I I I /' / 60 ! 83/V J' , /"Enthalpy, kJ/kg dry air V .....V 7 -"" ..,..., / -so J?' '",,-10""...../v .............~~ V ~~--..,60 / ........... -'" :?,,40 -"~ .....-:z~~,.,. ..-" -'~?-.c~ Il""'"__ .--I--' .~ ... 30 --~ 'Ii " II! 20 10 0.030 0.025 ~ iii ~ "0.020 '" :!!: ~ 0.015 ~ o ;::

l! 0.010 ~ "E ::l 0.005 J: 15 17 19 21 23 25 27 29 31 33 0.000 35 Drybulb temperature, Ta, ?C Figure 104: Airside heating process indicated on a psychrometric chart. 1.6 THE DRAFT EQUATION The specified loss coefficient due to the support structure of the fill is refer red to the mean conditions through the fill i.e. K K =0 5(Pa'15)( ma.1)2 =0 5( 0.98424)( 16807.68)2 = 0.4781096fiji + "cfi ? P~I m~15 . 1.01012 16966.47 According to the specified fill loss coefficient find Kldm =1.851LjiO~27520~Lo356 =1.851 x 2.504 x (1.50602)12752 (2.00870rI.0356 ~ 3. 79414 It follows that the actual fill loss coefficient applicable to the cooling tower is given by Kji = K tdm + (G;'5 _0;', )/(0;'" )';;3.79414+(2.063292 P~5 P~, P~" 0.95964 2.02502 2 )/(2.044152 ) = 3.88284 1.01012 0.98424 The expansion loss coefficient after the fill, referred to the mean conditions t hrough the fill, follows from equation (D. 16). K,,,ji =(1- At, J2(p.."J(~J2 =(1 A3 Pavs mavIS 8300 x 4 )2(0.98424)(17125.27)2 =0.001088 Il' x 104.52 0.95964 0.95964 ...-.,.'1i':'1@I'l'l!':"I' 1.23 The loss through the spray zone above the fill referred to the mean conditions t hrough the fill is given by equation (D.6) K,pji = L,p[O.4(Gw )+1](P=I')( m=, )' ~ 0.5[0.4(1.50602) + lrO.98424XI7125.27)' Go P=5 m=l, 2.00870 0.95964 16966.47 ~ 0.679143 The specified loss coefficient due to the water distribution system is referred to the mean conditions through the fill i.e. K wdji = K Wd (p =15 )( m=, )2 = 0.5(0.98424XI7125.27)' = 0.522458P"" m=15 0.95964 16966.47 The loss coefficient for the specified type c drift eliminator [98KRI) based on fill conditions is K =27 4892Ry-O.l4247(PO'I' XmO., )'deft . PavS mavIS ( )-O.l4247()()'=27.4892 17125.27 0.98424 17125.27 = 5.471011.81816xl0-' x8300 0 .95964 16966.47 The sum of these loss coefficients in the vicinity ofthe fill is K HE = K fiji + Ke" + K ji + Kele + K,pji + KWdji + Kdeji ~ 0.4781096 + 3.88284 + 0.001088 + 0.6791:43 + 0.522458 + 5.47101 = 11.03465 " The inlet loss coefficient for a circular cooling tower with isentropic fill ope

rating in the absence of a rain zone Can he determined according to equation (D.IO) which is applicable if the i nlet is roun'ded with r,ld, = 0.02 and 10 ~ d,lH, = ~ 15. The sum ofthe loss coefficients in the vicinity of the fill (effective fill loss coefficient =KHE = 11.03465) also falls within the range of the applicability of the equation, i.e. 5 ~ Kji ~ 25. K,,(twn) =0.011266exp(0.093d, 1H, )K~E - 0.3105exp(O.l085d, 1H, )KHE -1.7522 + 4.5614exp(O.l31d, 1HJ [ { (10970.2exp(-0.2442KHE)+ 1391.3)/(d, 1H, -15.7258)lj + sin -I + 1205.54exp(- 0.23KHE )+ 109.314 f x {2r, 1d, - 0.01942/(d, / H3 - 27.929)- 0.016866} K"(oo,,j =0.011266exp(0.093 x 104.5110) 1.03465' - 0.3105 exp(0.1085 x 104.5 /10 ) xll.03465-1.7522+4.5614exp(0.131xl04.5/10) [{ (10970.2exp(- 0.2442x 11.03465)+ 1391.3)/(104.5/10 -15.7258)}] + sin -I + 1205.54exp(- 0.23x 11.03465)+109.314 ~ 6.15873 x {2 x 0.Q2 -0.01942/(1 04.5110- 27.929)-0.016866} 1.24 This value must be multiplied by the correction factor C" as given by equation ( D.15) to obtain the correct inlet loss coefficient in the presence of a rain zone. Present values fa ll in the range of applicability ofequation (D.15). c" =[0.2394+80.1{0.0954/(d, 1H,)+ dd }exp(0.395G. IGJ ] -0.3195(G. IG')-966{dd I(d, 1H,)}exp(0.686G. IGJ x(1- 0.06825G. )K~~9667exp{8.7434(1 1d, - 0.01)} [ 0.2394 +80.1{0.0954/(1 04.5 110) +0.0035}exp(0.395 x 1.50602 12.00870) J = -0.3195(1.50602/2.00870)- 966{0.0035 1(1 04.5/10)}exp(0.686 x 1.50602/2.00870) x(l- 0.06825 x 1.50602) 1.03465"9667 exp {8.7434(1 1104.5 - 0.01)} ~ 0.9233078 Referred to the mean conditions through the fill, the inlet loss coefficient fro m equation (D.l2) becomes, K =C K (P""5 X~)'(Afr)'ctfi rzct A Pavl maviS 3 = 0.9233078x 6.15873 (0.98424)(16807.68)'( 4x8300 )' = 5.686403 1.01012 16966.47 Jrxl04.5' With equation (D.8) find the loss coefficient fur the rain zone. 0.2246- 0.31467ap P", + 5263.04a;'u." + 0.775526{1.4824163exp(71.52aLdd)- 0.91} x {0.39064exp(0.010912a/d,)- 0.17}~.0892(a,V a"t1.3944 + 0.14} j(0.84491o(a Ld, 12)- 2.312) Ix exp x (0.3724In(a,va,,)+ 0.7263) x 10(0.206.757 exp(0.06651 SaLH, t'8344 + 0.43) where the values of the "a" coefficients are identical to those employed in the mass transfer coefficient equation with v w3 =Gw / Pwo = 1.50602 I 997.8629 = 1.5092 xl0' mis, the value of the coefficien t is found to be K nfi =3x 1.0008 x1.5092 x10-3 (10/0.0035) 1.25 0.2246- 0.31467x 1.0oo14x 1.01012 + 5263.04 x1.00014 xl.7857x 10-' + 0.775526~.4824163exp(71.52 x1.00025 x 0.0035)- 0.91}x {0.39064exp(0.010912 x1.00025 x 0.0035)-0.17}x ~.0892(1.0008x2.00473)-1.3944 + 0 .14~ {

(0.8449In(1.00025 x 104.5/2)- 2.312)x (0.37241n(1.0008x 2.00473)+ 0.7263)l exp x In(0.206.757exp(O.066518 x1.00025 X10)-,?'344 +0.43) f X(0.98424XI6807.68)'( 4x8300 )' ~ 6.39225 1.01012 16966.47 I<x104.5' The loss coefficient due to the tower supports referred to the fill follows from equation (D.15) K"ft = [C.",L"d"n"A~ ](PavIS )( mavl )'(m13H3 ) P",I m",I' =[I x 11.6 x 0.8 x 72 x 8300'10.98424)(16807.68)' ~ 1.24393 (7fxI04.5xIOf 1.01012 16966.47 At this stage it is possible to confirm the value ofpa' where _ [ 0.00975(H3+ Lft 12)]3.'(I+W,{I-W,+~622J Pa' - Pall T al " 0.008127 1 = 101325[1 0.00975(10 + 2.504/2)]3,'(1+0.008127 \1 0,0081'7+0.622 288.6 ( 16966.47)' ( 1.24393 + 5.686403 +6.39225 + 0.4781096 +3.88284) 8300 - = 83937.04 Pa + 0.001088 +0.679143 + 0.522458 + 5.47101 2x 0.98424 -.I This value is in agreement with that used previously in calculations in this exa mple. To find the temperature lapse rate inside the tower, the specific heat of water is evaluated at (299.8563 + 273.15)/2 = 286.5032 K. According to equation (A.4.2) find cpw = 4192.42 JlkgK. Using the previously obtained values for the specific heat of dry air and water vapor at this tempera ture and cpma ~ cpa' + W,a'XCp,' = 1006.5548 + 0.027176xI873.9329 = 1057.481, find =-(I+w ) [1+0.42216XlO- IlW;a,Pa,eXP(5406.l915ITa,) ] C;T" 'a' g X ~kwa -(Cpw -cp'XTa, -273.15)}/{(w,a, +0.622)RTa,} Ilcpma +3.6696xlO-?w;a,Pa, exp(5406.l9151Ta, x,>gwa -(Cpw -cp' XTa, - 273.15)}IT a~ J 1.26 ,..-?I' C f r 1 [ I +0.42216xI0-1I xO.027176' x eXP(5406.1915/299.8563)] = -(I +0.027176)9.8 x ~.5016xlO' -(4192.43 -1873.932X299.8563 - 273.15)} 1{(0.0271 76+ 0.622}287.08 x 299.8563} 1[1057.481 +3.6696 x 10-' x 0.027176' x eXP(5406.1915/299.8563)] ~ _ 0.0034113K/ m ~;5016x10' -(4192.42 -1873.9329X299.8563 - 273.15)}/' According to du Preez and KrOger [94DUlj the difference in the mean pressure at the tower outlet and the ambient pressure at the same elevation is given by Pa' - Pa' =[0.02Fr;"'-0.141 FrD Kma'5 I A,Y I P<N' To find the pressure difference (P" - p,,) given by equation above the air prope rties and corresponding desimetric Froude number must be determined at the outlet of the tower. Using th e lapse rate obtained above, and assuming it essentially constant over the height of the cooling tower , the air temperature at 6 may be determined. Ta, =T'5 + qT" (H, - H, - Lft - L,p) = 299.8563 - 0.0034113 (147 -10 -2.504 - 0.

5) = 299.399K The corresponding density of the air-vapor mixture at this temperature is accord ing to equation (A.3.1) (I II w..5 ] P a6Pav6 = +WSQ$ -w,a' +0.62198 RTa6 = (1+0.02717l1- 0.027176 ] 82650.57. -0.9464 kg/mJt 0.027176+0.62198 287.08x299. 399 The ambient temperature avelevation 7 follows from equation (E.12) with H, ~ H,. Ta, = T'I - 0.00975H, ~ 288.6 - 0.00975x147 = 287.167K The pressure at 7 may be determined from equation (E.13). ( 2.1778(l+~) 2.1778(1+0,008127) P =P 1-000975 HIT )W,....6219. =84100(1-0.00975 xI47/288.6)o.oosl27+0.'219'07 al ' 601 ~ 82654.27 Pa The corresponding density of the ambient air at elevation 7 assuming a uniform a mbient humidity ratio WI, is according to equation (A.3.1) P ,= (I + W/I- WI ] p" = (I + 0.008127{1 a' t WI +0.62198 RT,,1.. =0.9977 kg/mJ With no cold inflow these values yield 0.008127 ] 82654.27 0.008127 +0.62198 287.08x287.167 1.27 Frn = (m"" IA,,) 1[P",,(P"'7 - p",,)gd,] =(17125.27/~.25 x If X 60.85 2}j 1[0.9464(0.9977 -0.9464)l.8x 60.85]= 1.19697 Substitute this expression for FrD into the equation of Du Preez and KrBger [94D Ul] presented above and findpa6' Pa' =Pa7 + [0.02Fr;I.S - 0.141 FrD?m"" IAoy IP",6 = 82654.27 + [0.02 x 1.1969T" - 0.14/1.19697117125.27/(0.25 x If x 60.852 )j 10. 9464 = 82650.57 N/m2 which is in good agreement with the value given initially. The draft equation from equation (E.!) may now be solved using the above values (Pal - Pa7)- (Pal - Pa34 )-(Pa34 - Pa,)- (Pa6 - Pa7)= where (Pal - Pa7) and (Pal - Pa'4) is given by equation (E.l4) \ [ 2.1778\I+W,)](P -p )=p 1-(1-000975 H IT )",+0.6219.01 a7 al ? 6 at [ 2.1778(1+0.008127)] =841001-V-0.00975xI47/288.6)0.oo8127+0.6219. = !445.732Pa [ f, ( ) 2.1778(1+11'1 lJ(Pal - Pa'4)= Pal 1-1'-0.00975 H6+Lji121Tal w,+0.62198 [ 2.1778(1+0.008127)] =84100 1- {I- 0.00975(147 + 2.504/2)/288.6} 0.008127+0.6219' = 111.2947 Pa [ -(l+OO27176J{I- 0,027176} 9.' ] { 147 -10 - 2 504/2}' 0.027176+0.6219' 287.08+-0.0034113) = 1- 1+ 0.0034113 . =1267.954 Pa 299.8563 (Pa6 - Pa')= [0.02Fr;LS -0.141 FrDlm", 1A6)' IP"6 =[0.02xI.1969T" -0.14/1.19697KI7125.27/~.25xIf X 60.862}j 10.9464= -3.726184 Pa Upon substitution the left-hand side of the draft equation yields, If?:. ,., , 1.28

(Pal - Pa7 )-(Pal - Pa34 )-(Pa34 - Pa6)- (Pa6 - Pa7) ~ 1445.732 -111.2947 - 1267.954 + 3.726184 ~ 70.20883 Pa The value on the right-hand side of the draft-equation is (~;5J (mZ)'(K,Sf/ + Kctfl + Krzft + K jsji + Kctcji + K Ii + KC'tifi + Kspfi + K Wdfi + Kdef/) + a e6 -"----'-----'-2Pav15 2Pav6 ( 16966.47)2 = (1.24393 +5.686403+6.39225 +0.4781096+ 3.88284) 8300 +0.001088+0.67914+ 0.522458 + 5.47101 2 xO.98424 ( 17125.27 )' +1.01 O.25x1TX60.86' ~ 70.20879 Pa 2xO.9464 Since the value ofthe right-hand side ofthe draft equation is essentially the sa me as the left-hand side, the chosen air-vapor mass flow rate is correct. The amount ofwater lost due to evaporation is given by m.(m,p) = ma(w, -w,) = (0.027888-0.008127) = 329.464 kg/s - . ..1 J.I APPENDIXJ ANALYSIS OF AN INDUCED DRAFT WET-COOLING TOWER EMPLOYING THE MERKEL APPROACH J.1 INTRODUCTION A sample calculation is presented for the performance evaluation of an induced d raft wet-cooling tower while the Merkel method of analysis is employed. The main dimensions of the towe r are shown in figure J.1. ? ? H9 H9 ~ (]) ~~ UH3 ~CV ~CD ~ V; Figure J.I: Induced draft wet-cooling tower Ambient conditions: Atmospheric pressure at ground level Air temperature at ground level Wetbulb temperature at ground level P,I = 101325 Pa T'l ~ 306.65 K (33.5 0c) Tw? = 298.15 K (25?C) Cooling tower and operational specifications: Tower height Fan height Tower inlet height Tower inlet width Tower breadth Fill height Height of spray zone Inlet rounding Plenum chamber height Inlet water temperatnre Water mass flow rate Fill specifications: Transfer coefficient hdfiafi =0.2692G. -o,O"G a O,6023 G.

Loss coefficient K =1 9277G 1.275'G -1.0356fdml" wa The frontal area of the fill Afr = B, x W, = 144 m' Other specifications: Mean drop diameter in rain zone Loss coefficient for inlet louvers Loss coefficient for fill support system Loss coefficient for water distribution system Fan upstream losses J.2 H, ~ 12.5 m H6~9.5m H,~4m W,= 12m B,= 12m Lfi ~ 1.878 m L" = 0.5m r, ~ 0.025Wi Hp/=2.4m Tw, ~ 314.65K (41.5 0c) m. = 412 kg/s ti<J = 0.0035 m Kit = 2.5 K/,= 0.5 K.d =0.5 K,p= 0.52 The loss coefficient for the drift eliminator (type c) is given by equation (D. 7). Fan/diffuser with rounded inlet dimensions andperformance characteristics: Fan diameter dF~ 8 m Fan rotational speed NF ~ 120 rpm Fan model diameter dF, = 1.536 m Reference rotational speed NF, = 750 rpm Reference air density p, = 1.2 kg/m' Fan/diffuser static pressure: D.pFld/f =320.85 - 6.9604VF1d/f + 0.31373V;,d/f - 0.021393V';!d'[ Pa Fan shaft power: J.3 PFld/f = 4245.1-64.134VF1d/f +17.586V;,d/f -O.71079V';,d/f W J.2 SOLUTION It is assumed in this problem that the atmospheric pressure and humidity fields have no influence on the draft equation. This problem is solved by following an iterative procedure to fi nd a solution that will satisfy both the energy and draft equations. The choice of an air-vapor mass flo w rate ofma,I' ~ 442.1426 kg/s through the fill will satisfy these equations, giving a corresponding press ure ofpa' ~ 101170.6 Pa, an air temperature Ta, ~ 306.7645 K and an exit water temperature Two ~ 303.4677 K. At the specified air inlet drybulb temperature of Tal ~ 306.65 K, wetbulb temper ature of Twb ~ 298.15 K and atmospheric pressure at ground level Pal ~ 101325 Pa, find the following the rmophysical properties employing the equations given in appendix A. Density of air-vapor Humidity ratio of inlet air Viscosity of the air vapor mixture P." ~ 1.1397 kg/m3 WI ~ 0.016569kg/kg f./avl ~ 1.85928 x 10" kg/ms

(A.3.I) (A.3.5) (A.3.3) The enthalpy of the inlet air, imaI. is found according to equation (A.3.6b) wit h Cpal ~ 1006.64 J/kgK and Cp,1 ~ 1876.84 J/kgK being evaluated at (Tal + 273.15y2 ~ (306.65 + 273.l5y2 ~ 2 89.9 K according to equations (A.1.2) and (A.2.2) respectively. The latent heat is found to be ifgoo ~ 2.5016 X 10' J/kgK according to equation (A.4.5) at 273.15K. With these values find imal ~ 76213.73 J/kg dry air. If the air is assumed to be saturated immediately after the drift eliminator, th e wetbulb temperature at 5 will be equal to the given drybulb temperature Ta, ~ 306.7645 K at this elevatio n. The corresponding thermophysical properties at 5 can be determined according to the equations give n in appendix A. Saturated vapor pressure Humidity ratio Density of air-vapor I Dynamic viscosity ofair Dynamic viscosity of vapor Dynamic viscosity of air-vapor p" ~ 52060305 N/m' W, ~ 0.033922 kg/kg p." ~ 1.12634 kg/m' )./a, ~ 1.8779xW-' kg/ms p" ~ 1.0254xlO" kg/ms )./a" ~ 1.84273x 10" kg/ms (A.2.l) (Ao3.5) (A.3.1) (A.1.3) (A.2.3) (A.303) The enthalpy of the inlet air, im~', is fuund according to equation (A.3.6b) wit h cpa, ~ 1006.640 J/kgK and cp,' ~ 1876.887 J/kgK being evaluated at (To, + 273.15Y2 ~ (306.7645 + 273.15)/2 ~ 289.9573 K according to equations (A. 1.2) and (A.2.2) respectively. The latent heat is fou nd to be ifgwa ~ 2.5016 X 10' J/kgK according to equation (A.4.5) at 273.15K. With these values find ima" ~ 12 0836.9 J/kg dry air. The approximate harmonic mean density of the air-vapor in the fill is given by 2 p"" = 1 1 --+-Pavl Pa\l~ J.4 _::--_2_::--_- 1.132982 kg/m3 1 1 ---+--1.1397 1.12634 The dry air mass flow rate cao be determined from the following relation: m"" = [ma (1 + w,)+ ma(1 + w, )], or m. = 2m"" 1(2 + w, + w,) =2(442.1426)1(2 + 0.016569 +0.033922) = 431.2545 kg/s The respective air-vapor mass flow rates upstream aod downstream of the fill are thus ma,' =ma (1 + W,) ~ 431.2545(1 + 0.016569) = 438.4000 kg/s, aod m", =ma (1 + w,) = 431.2545(1 + 0.033922) = 445.8834 kg/s

The corresponding mass velocities are Go," =m"" 1AI' = 442.1426/144 ~ 3.070429 kg/m's Ga =malAfr ~431.2545/144~2.994823kg/m's G", = m",1 AI' = 438.40001144 = 3.044444 kg/m's G." =m", 1AI' = 445.8834/144 = 3.096413 kg/m's According to equation (A.4.2) the specific heat of water cpwm ~ 4176.992 Jlkg at the meao water temperature of(Tw; + Two)/2 = (314.65 + 303.4677)/2 ~ 309.0589 K. At the meao outlet temperature ofthe water Two ~ 303.4677 K find Density ofwater {J"o = 995.6046 kg/m3 Surface tension OWa = 0.0711808 N/m The mass velocity for the water based on the frontal area of the fill is G w =m w 1 At, =412J144~2.86111kg/m's (A.4.1) (A.4.7) The transfer coefficients cao be determit)ed.with the above values. To find the traosfer coefficient in the rain zone, use equation (D.22). The "a" coefficients appearing in the equation f or the rain zone traosfer aod pressure drop coefficients are as follows: ap =3.061 xl 0-6 (,0:0 g' 1a wo)""" =3.061 x 10-6 (995.6046'9.8' 10.0711808)''' =1.0026 a p =9981 P"o = 998/995.6046 ~ 1.0024 a, =73.298(g'a';" 1p,;" )025 = 73.298~.8' X 0.07118083 1995.6046)??" ~ 0.9882 aL =6. 122(gawo 1Pw.)''' ~ 6.122(g.8x 0.0711808/995.6046)?" ~ 0.9960 0.2692LfiG .0.094G 0.6023=0.2692x 1.878 x (2.86111).00094 x (2.994823)??602' ~ 0 .8866959 ? ? .', 00j:.,', l.5 Other quantities required to evaluate the rain zone transfer coefficient are: The humidity ratio ofsaturated air at Two w" ~ 0.027848 kglkg Diffusion coefficient at inlet conditions D, = 2.09061 xl 0" m'/s Furthermore, the Schmidt number is Sc, = P.'l 1(P.,!DJ = 1.85928xI0?'/(1.l397 x 2.0906IxI0?') = 0.7803 and the air-vapor velocity before the fill v"' = mo'! l(po"A/,) = 438.4000/(1.l397xI44) = 2.67127 m/s With these values find from equation (D.22) hdnQnH, 3.1~XH, X P'l )SC\oo"[ln(w,\ +0.622)/(W n -wJl G. ,\V",dd dd p.,R,T,\ w\ +0.622 J 4.68851ap p.,! -187128.7a.pov! - 2.29322 + 22.4121~.350396(a.v." 10'8046 + 0.09~ .60934(aLH,tI208' + 0.66} X x ?4.6765(aLddt732448 +0.45} { ) { 0.087498exp(0.026619aLw,)}} x exp 7.738gexp(- 0.399827aLH, In +0.85 =36( 2.09061xl0?s y 12 X 101325 ) . 2.67127 x 0.0035 )l0.0035 995.6046 x 461.52 x 306.65 x 0.7803300" [In(0.027848 + 0.622)/(0.027848 - 0.016569)1 0.016569 + 0.622 J 4.68851x 1.0024 x 1.1397 -187128.7 x 1.0026 x 1.85928xl0?' - 2.29322 + 22.4121~.350396(0.9882 x 2.67127)138046 + 0.09} x x t.60934(0.9960x4)"1.12083 +0.66} ~ 0.2851664 x ~4.6765(0.9960 x 0.0035)??732448 + 0.45} j7.738geXP(-0,399827 x'O.9960x 4) )x exp x In{0.087498exp(0.026619X 0.9960 x12)} +0.85 The Merkel number applicable to the fill is specified i.e.

hdflQflLfl G. The transfer coefficient in the spray zone is given by Lowe and Christie [6ILOI] (A.3.5) (D.21) J.6 0.2L (Go )0" =0.2x 0.5(2.994823)??, =0.102310 '" G w 2.86111 (A. 1.2) (A.2.2) The total transfer characteristic ofthe cooling tower is According to equation (B.2I) the Merkel number, MeM, is rI" cpwdTwMeM = (. .)T..o 'masw -Imo It the four point Chebyshev integral is applied to this relation, the integral o n the can be expressed as rI" CpwdTw Cpwm (TWi - Tw.)( 1 1 1 1)Me = ,., --+--+--+--M r~ (i....,., - i... ) 4 M(l) !'J.i(2) M(3) M(4) A detailed discussion of the Chebyshev integration method applied to equation (B .2I) is given in the literature [88BRI, 90COI, 95001, 96MOI, 97COI). The enthalpy differentials are dependent on the following intermediate temperatu res: TW(I) = Two +0.1(314.65 - Two) = 303.4677 +0.1(314.65 - 303.4677) = 304.5859 K TW(2) =Two + 0.4(314.65 - Two) = 303.4677 + 0.4(314.65 - 303.4677) = 307.9406 K TW (3) =Two + 0.6(314.65 - Two) = 303.4677 + 0.6(314.65 - 303.4677) = 310.1711 K TW(4) =Two + 0.9(314.65 - Two) =303.4677+ 0.9(314.65 - 303.4677) = 313.5318 K The bracketed subscript numbers refer to the intervals in the Chebyshev integral and should not be confused with the numbers indicating various positions in the cooling tower. To find the correspondi'1!L increments in enthalpy, determine the enthalpy of sa turated air at TW(I) = 304.5859 K. The relevant specific heats of air and water vapor respectively are evaluated at (T.(l) + 273.15)/2 = (304.5859 + 273.15)/2 = 288.868 K. Specific heat of air Cpo(l) = 1006.612 JlKg K Specific heat of water vapor Cp,(I) = 1875.950 Jlkg K The pressure of saturated water at TW(l) follows from equation (A.2.I) and the c orresponding humidity ratio evaluated at PalS follows from equation (A.3.5). Where POlS = (pal +Pa,)/2 = (101325 + 101170.6)/2 = 101247.8 Pa. Pressure of the water vapor Humidity ratio P"(I) = 4605.056 JlKg K W'(l) = 0.02979283 Jlkg K (A.2.I) (A.3.5) J.7 With these values determine the enthalpy of saturated air at Tw(l) according to equation (A.3.6b) im""'(lj ~ cpa(i)(Tw(1) - 273.15) + w,(I)[ijgwa + cp >,(Ij(Tw(l) - 273.15)] ~ 1006.612 (304.5859 - 273.15) + 0.02979283 x [2.5016xI0' + 1875.950 (304.5859 - 273.15)] = 107930.4 J/kg The enthalpy of the air at TW(I) can be determined by applying equation (B.21) L e. ima(l) = mwCpwm(Tw(I) - Two)lma + imal = 412 x 4176.992( 304.5859 - 303.4677)/431.2545 + 76213.73 = 80676.0 J/kg With these values find the difference in enthalpy

!l.i(,) = ima"'(l) - ima(l) = 107930.4 - 80676.0 = 27254.42 Jlkg dry air Repeat the above procedure in the case ofthe other three intermediate temperatur es and find ru(2) = 34190.49 J/kg dry air; !l.i(3) = 40692.23 Jlkg dry air; !l.i(4) = 53759. 63 J/kg dry air Substitute these values into the approximate expression for the integral and fin d = 4176.992(314.65-303.4677)( I + I + I + I ) = 1.274150 4 27254.42 34190.49 40692.23 53759.63 This value is almost identical to the value obtained by adding the transfur coef ficients in the three wet zones which means that the water outlet temperature, Twa = 303.4677 K is correct . The heat rejected by the cooling tower is given by Q ~ mwCpwm(Twi - Twa) = 41'2 x 4176.992 (314.65 - 303.4677)= 19.243874 MW The correctness of the temperature of the saturated air leaving the spray zone, Ta" can thus be confirmed from the relation Q = l71a(jm~' - imal) The enthalpies ima" and imal are already known, thus Q~ 431.2545 (120.836.9 -76213.73) ~ 19.243942 MW The values of Q are in agreement which means that the value for Ta, is correct. The specified loss coefficient due to the louvers is refurred to the mean condit ions through the fill i.e. = (Pa,,,)( W,B, )( m~l )' = 2.5(1.132982)( 144 X438.4000)' = 5.497609 Kilfi Kil P~l 2H,W, m~" 1.1397 2x4xl2 442.1426 J.8 With equation (D.9) find the loss coefficient for the rain zone with v 3 = Gw = 2.86111 =2.873743xl0-3 m/s w Pwo 995.6046 0.219164-0.30487a p Pwl +8278.7ap Pwl + 0.954153{0.328467 exp(135.7638aLd d )+ 0.47} ( In(0.204814 exp(0.066518aLW,)+ 0.21)l x ~6.28482(aLH3t'''129 +0.56}exp x (3.91 86 exp(-0.3aLH 3? x (0.31095ln(aLdd )+ 2.63745) x~.177546(a,vw3t1.46S41 +0.21} = 1.5 x 0.9882x 2.873743xl 0-3 (41 0.0035) 0.219164-0.30487x 1.0024x 1.1397 +8278.7x 1.0026x 1.85928x 10-' + 0.954153{o.328467 exp(135.7638x 0.9960x 0.0035)+ 0.47} ( In(0.204814 exp(0.066518 x0.9960x 12)+ 0.21)l x x ~6.28482(0.9960x4t'9"'9 +0.56}exp x (3.9186exp(-0.3 x 0.9960 x 4? x (0.31 095ln(0.9960x 0.0035)+ 2.63745) x ~.177546(0.9882x 2.67127t1...."1 +0.21} = 2.072855 The rain zone loss coefficient referred to fill conditions is K =K (p",,)( ma,l )' =1072855(1.132982)(438.4000)' = 2.025911 "fi "p", rna,,, . 1.1397 442.1426 The specified loss coefficient due to the support structure of the fill is refer red to the mean conditions through the fill i.e. K = K (p",,)( ma,l)' = 0 5(1.132982)(438.4000)' =0.4886764 I'fi I' Pa,l rna," . 1.1397 442.1426 According to the specified fill loss coefficient K Idm = 1.9277LfiG w 1.27"G

o -1.035. = 1.9277 x 1.878 x (2.8611 1)127" (2.994823f1.03'. = 4.441997 From the note at the end of example 4.3.1, it follows that the actual fill loss coefficient applicable to the cooling tower is given by K =K +(G;" _GL1)/(G;,,,) ~4.441997+ (3.096413' fi fdm PavS Pa"l Pavl5 1.12634 ~ 4.487634 3.044444' )/(3.070429') 1.1397 1.132982 J.9 The loss through the spray zone above the fill referred to the mean conditions t hrough the fill is given by equation (D.6) K = L[O.4(Gw )+ 1](P"'" Xm"" )' = 0.5[0.,,( 2.86111 )+ 1](1.132982X445.8834)' 'pfi 'p G" Po,' m",,, l. 2.994823 1.12634 442.1426 =0.7069580 The specified loss coefficient due to the water distribution system is referred to the mean conditions through the fill Le. K wdji = K wd (pad' )( m"" )2 = 0.5(1.l32982Y445.8834)' =0.5114949P"" m"''' 1.1263 4 A442.1426 From equation (D.7) the loss coefficient for the specified type c drift eliminat or based on fill conditions is Kd,ji =27.4892Ry--<l.l4'47( P"'l' Xrna,' )2 Pav~ maviS ( )--<l.l424'( X )2= 27.4892 445.8834 1.132982 445.8834 = 5.064901.84273xl0-' x1 44 1.12634 442.1426 The inlet loss coefficient in isotropically packed induced draft rectangular tow ers is according to equation (D.13), K"C'"") =0.2339+(3.919xlO-3 K~, -6.840xl0-2 Kfi, +2.5267) x exp{:: (0.5143 -0.1803exp~.0163K jiJ)} -Sinh-'[2.77exp{0.958 :JxexP{Kji,(2.457-1.015 :: )XI0-2 }x(i, -0.013028)] = 0.2339 +(3.919x 10-3 x 11.259662 - 6.840x 10-2 x 11.25966+ 2.5267) x exp{l: (0.5143 - 0.1803exp[0.0163 x 11.25966D} - Sinh-'[2.77 exp{0.958 1:} x exv{t 1.2596{2.457 -1.015 1:)x 1O-2 } x( ~; - 0.01 3028)] = 5.212728 where Kji, is the loss coefficient in the vicinity of the fill Le. Kfie = K/sfi +Kji + K SPfi +KWdfi + K defi =0.4886764 + 4.487634 + 0.7069580 +0.5114949 + 5.06490 = 11.25966 r"" I I llO De Villiers and Krllger [99DEI] state that it becomes acceptable to ignore the i nlet loss correction factor in cases where ~ / H, ~ 3. In this case, Wi / H3 = 3, which means that K" = K"(n on) Referred to the mean conditions through the fill, the inlet loss coefficient bec omes, K = K (p~" Yrn~, )' ~ 5.212728 (1.132982)(438.4000)' ~ 5.094674 'rf/ , "P~l Arn~I' 1.1397 442.1426 The specified fan upstream loss coefficient is referred to the mean conditions t hrough the fill i.e. K = K(p,.,,)( m,., )'(Afr )' = 0.5(1.132982)(445.8834)'(144 X 4)' ~ 4.365760 'Pfi 'P P~, rna," A, 1.12634 442.1426 7l'x8' The actual air volume flow rate through the fan is

VF=rna,' /Pa,' =445.8834/1.12634 = 395.8676 m'ls Since the actual air density and the rotational speed of the fan are not the sam e as the reference conditions for wbich fan performance characteristics were specified, the relevant fan laws as given in Krllger [98KR1] are employed. According to the fan conversion law [98KRI], V ,= V (NFld'i ldF')' = 395.8676( 750 X1.536)' = 17.51192 m'ls F Idil F N d 120 8 F F At this flow rate the fan static reference pressure drop is given by the specifi ed relation, , , I!.pFldift =320.85 - 6.9604VFldij + 0.31373V';ldif - 0.021393Vildif =320.85 - 6.9604(17.51192) + 0.31373(17.51192)2 - 0.021393(17.51192)' ~ 180.2833 Pa The actual change in fan static pressure as expressed by the fan conversion law [98KRI] t.. = t.. (~}'(:!LJ'(Pa" J= 180.2833(120)'(_8_)'(1.12634) = 117.5123 PaPF, PFldi j N d 750 I 536 1 2 FldJj Fr P r .? At the reference condition the fan shaft power is ~'Idil = 4245.1-64.134VFld/1 +17.586V';ldil - 0.71079V}ld'l = 4245.1- 64.134(17.51192) +17.586(17.51922)' - 0.71 079(17.51922)' ~ 4697.863 W The actual fan shaft power follows from the fan conversion law [98KRI] P F =PFldi/(~)'(!!.L)'(P~6) ~ 4697.863(120)'(_8_)'(1.12634) ~ 69222.04 W N Fldil d", P, 750 1.536 1.2 J.II The fan static pressure rise coefficient follows from Krllger [98KRI] 2 x117.5123 x1.12634 = 3.364199 ( 4 X445.8834)' Jl' x 8' At this stage, it is possible to confirm the value ofPa' according to -{ O.016~69) =101325[1- 0.00975(4 + 1.878/2)],?'(1+0,01.".\1-0,01""?+<1.622 306.65 ( 442.1426)' ( 5.497609+ 2.025911 + 0.4886764+ 4.487634+ 0.7069580) 144 - - 101170.6 Pa + 0.5114949+ 5.064900+ 5.212728 2 x1.132982 This value is in agreement with that used previously in calculations in this exa mple. It is assumed that the condition of the air at the inlet of the fan is equal to that at the outlet of the fill, i.e. Ta? = Ta" Pa. =pa, and therefore are Pa,,; =Pa" = 1.12634 kg/m'. Ignoring pressure differences due to the gravity field, the draft equation can b e expressed as The terms on the left-hand side ofthe equation give (~~' )2 (m~:, )' (Kilfi + K,.q; + K ftfi + K fi + K,pfi + K Wdfi + Kd'fi + K't/i + KuPfi ) Kft -' ---=--.?-2Pa,lS 2Pa,' ( 442.1426)2 = (5.497609 + 2.025911 + 0.4886764 + 4.487634 + 0.7069580) 144 + 0.5114949+ 5.064900 + 5.212728+ 4.365760 2 x 1.132982

( 445.8834X 4)' _ 3.364199 Jf x 8' = -0.00482 2x 1.12634 This value is close to zero and the draft equation is thus satisfied. J.12 The amount of water lost due to evaporation is given by mw(evap) =ma(ws - WI) =431.2545(0.033922 - 0.016569) =7.48343 kg/s The path ofthe air through the counterflow cooling tower, according to the Merke l approach, is shown in figure 1.2. Only the inlet and outlet states of the air is known if the Merkel a pproach is employed. The path of the air according to the more rigorous Poppe approach is also shown in f igure J.2. The outlet air is not saturated according to the Poppe approach. r;::=:r::=:c=::r::=:::r:=~--r-r-1OO-=;Z:::::::J0.040 38363428 30 32 9 2624 Enthalpy, kJ/kg dry air 22 ~==+===!===+===1=="-l---t'1"'2"0-LL=~~~~;:J0.035 ~ 'iii ~ 'tljl./r4F:"?~~:;;~t=-;;;-J0.030 ~) ---t 0.025 ~ o .. l! _=:t o.020 ~ 'tl~~~~~!~~~~~~~~~t~~~~~~~~? 'E0.015 :f0.010 4020 Atm~spheric pressure: 101325 Pa \Drybulb temperature, T.,?C Figure J.2: Psychrometric chart J.3 COMPARISON BETWEEN DIFFERENT DRAFT EQUATIONS Table J.l contains the comparative results of the induced draft cooling tower an alyzed in this appendix for two variants of the draft equation. The first variant oflbe draft equation is eq uation (1.1), which is used in the analysis in the previous section. The second variant oflbe draft equation is equation (J.2). Equation (J.2) thus replaces equation (J.1) in the analysis above. The draft equation for the induced draft wet-cooling tower in the previous secti on, where the pressure differential between the inside and the outside of the tower is ignored, is acco rding to equation (J.]), (J.1) ~, IW". J.13 For the case where the pressure differential between the inside and the outside of the tower is taken into account, the draft equation is according to equation (J.2), (1.2) where the pressure differentials are similar to those employed in appendix I. Table J.I: Differences in results fur different draft equations. Draft equation (J.I) (J.2) Air-vapor mass flow rate, m.,,,, kws 442.1426 443.4894

Air outlet temperature, Ta" K 306.7645 306.7506 . Water outlet temperature, Twa' K 303.4677 303.455 Heat rejected, Q, MW 19.243874 19.26578 It can be seen in table J.l that the differences between the results for equatio n (J.I) and equation (J.2) for each model are very small. It can therefore be concluded that the pressure diffe rentials due to the pressure gradient in a gravity field can be ignored for mechanical draft cooling towers. This is because the pressure differential due to the fan is the dominating term in equation (J.2). r ' '1'" . . '~'. '., ,,:'" I I ~I f, I I I I I I I I Kl APPENDIXK FILL TEST FACILITY AND PROCESSING OF FILL TEST DATA K.IINTRODUCTION The fill test facility where 1he performance of fill mllterial is tested is desc ribed io section K.2. A computer program 1hat is developed to process and analyze 1he experimental data of a fill test is presented io section K.3. A sample calculation of the processing of 1he experimental data is presented io section K.4. K.2 FILL TEST FACILTIY A schematic layout of 1he wet/dry cooliog test facility io 1he Department of Mec hanical Engioeering at the University of Stellenbosch is shown io figure K.l. The test facility consist s of crossflow and counterflow test sections to test and analyze 1he performance of cooliog tower p ackiog material and spray-cooled heat exchangers. The crossflow section has a height and wid1h of 2. 5 m and a dep1h of 2 m. The counterflow section has a cross sectional area of 1.5 x 1.5 m. The counterfl ow test section can be extended to any practical height by 750 mm modules which are bolted toge1her. Hot water is pumped from an underground storage tank to the test sections. The s torage tank has a capacity of45 m'. The water is heated by recycling it through a 100 kW diesel-fi red boiler. During a test, 1he heated. water is pwnped from 1he top of 1he storage tank to 1he test section where it is cooled. The cooled water is 1hen fed back to 1he bottom of 1he storage tank. This ensures th at stratification occurs io 1he storage tank and 1hat 1he subsequent supply temperature will remaio almost c onstant for short test

runs. The water flow rate is determioed from the pressure drop measured across an orif ice plate iostalled io 1he supply lioe aceordiog to British Standard 1042 [8lBRl). The water flow rate is v aried by a manually operated gate valve. Air is drawn through 1he tunnel by a 50 kW centrifugal fan wi1h variable speed c ontrol. The mass flow rate of1he air is determioed by measuriog the pressure drop across one or more o f 1he five ASHRAE 51? 75 elliptical nozzles mounted io 1he horizontal section of 1he wiodtunnel as sho wn in figure K.l. The pressure drop is determioed by a calibrated electronic pressure transducer. The temperatures are measured usiog calibrated copper-constantan 1hermocouples. Refer to figure K.2 for 1he location of1he thermocouples on the counterflow section. ~~-~~' - - ,:0:"_-:: =.. c_~~ / Water distribution k-waterinlel system 1'"1'000 175 1000 3750 3500 =l-~1Isection Fan Water exlraction troughs ---? ;;<: Turning vanes I Settling chamber N .\ ,- I ? ~I Crossflow test Inlet .... . , , section Electrical motor Flow measuring nozzles Screen 100 nnn Insulation Figure K.l: Schematic layout ofwet/dry cooling test facility. /Air drybulb temperature (4 thermocouples) Air wetbulb tempemture (4 thermocouples) K.3 Water outlet temperature (4 thermocouples) Water inlet temperature t.....r-Ir-(3 thermocouples)Air outlet temperature(10 thermocouples) Cyclone, drybulb, wetbulb Air inlet drybulb temperature (4 thermocouples) -_.....Air inlet wetbulb tempemture (4 thermocouples) I i I, i Figure K.2 Thermocouples installed on counterflow section. The air temperature is measured before the nozzles to accurately predict the den sity of the air flowing through the nozzles. The drybulb and wetbulb temperatures are the average of fou

r thermocouples each, distributed across a vertical plane. The air drybulb and wetbulb temperatures are again measured below the water extr action troughs. They are the average of four thermocouples each, distributed across a horizontal plan e. The average temperatures of the air below the troughs will be used to determine the inlet pr operties to the test section. These temperatures may differ from the temperatures measured before the nozzles due to the influence of the fan and the leakage of warm water through the troughs. The air outlet temperature can be measured by four drybulb/wetbulb measuring pro bes or in a cyclone in which the water entrained in the air is separated from the air. Ten thermocouple s are available to measure the outlet air temperature. The pressure drop across the fill and troughs is measured by four static pressur e probes. Two are installed below the troughs and two are installed above the fill. The pressure drop across the troughs is subtracted '-, from the total pressure drop to obtain the pressure drop across the fill. Refer to Oosthuizen [95001) and Baard [98BAl) for a detailed description of the pressure probes, the water distr ibution system, the water extraction troughs and psychrometric probes. The data logging system consists of two Isolated Measurement Pods (IMPs). The IM Ps are connected to a Pentium Personal Computer (PC) via an S-Net cable and Schlumberger PC card. The data logger has an internal reference point, which eliminates the use of an ice bath needed for tem perature measurement purposes. The data logger converts all temperature readings from millivolts to d egree Celsius before r i KA transferring them to the PC. The pressure transducers adapt pressure readings to voltage signals, which are transferred to the data logger. K.3 DEVELOPMENf OF A COMPUTER PROGRAM TO PROCESS AND ANALYZE COOLING TOWER TESf FACILITY DATA. K.3.1 INTRODUCTION, A Windows 95 computer program, named Natklos, is developed to analyze the experi mental data from the approximately forty thennocouples and electronic pressure transducers of the cooling tower test facility. The aim is to process and analyze the data filst and efficiently. Diff erent calculation uptions can be selected with the press of a button on the user-friendly graphical user inter fuce (GUI). The GUI is develuped in Visual C++ 6 while the nwnerical algotitbms are develuped in Fortra n 77 and Fortran 90. K.3.2 MAIN PROGRAM DIALOG The main dialog window of the Natklos program is shown in figure K.3. The three main options that can be accessed from the main dialog window are the processing of the experimental d ata, the detennination of the transfer and loss coefficients according to different analytical models a nd the determination of

empirical relations to represent the transfer and loss coefficients. K.3.3 PROCESSING OF THE EXPERIMENTAL DATA As already mentioned, the data logger saves the various temperatures measured by the thennocouples in degrees Celsius while the output from the pressure transducers are saved in volt . The pressures and mean temperatures are calculated with the aid of the dialog window shown in figure K. 4. The calibration curves, to convert the voltage signals of the pressure transducers can be suppli ed in the dialog window shown in figure K.4. There are calibration curves for the pressure transducers t hat measure the pressure drup over the fill, over the nozzles in the upstream windtunnel and over the ori fice plate in the water, supply line. The air and water flow rates are detennined from these calibration curves. If certain selections are made, some of the ftmctions of the dialog window are disabled. By disabling the non? relevant ftmctions, the dialog window is more user-friendly. The pressure drup o ver the water troughs are subtracted fr01I\ the measured pressure drup over both the fill and the troughs. The data logger writes the data in a file that it receives from the Schlwnberger card in forty different channels, twenty channels fur each IMP. Due to maintenance or breakdown of therm ocouples, it is sometimes necessary to change or disable some the channels on the IMPs. Also, if the discrepancy of one thennocouple is great compared to other ones that are measuring the same quantit y, it is sometimes necessary to exclude this temperature from the calculation of the mean temperatu re. An efficient method is therefore implemented to assign the channels for different quantities on the data processing software. K.5 ------.-----------..--4 SAVE )-( Je '(x) CH frace?? e.perimental data I Iranster and 1= coefficients .Empirical equation. ~ave configu,ation I Ready Figure K.3: The main dialog window of the Natklos program. 54 ? nozzles I Z...nllldngll""'...,,.-;.::::. 1 ----------"".,:., 'll'lll.tfilernme: Bromel ;:",,?ulpulfilename:)------- BrowseI r. ~'WiJ!Jf lkJw r5te from British Stand5d 1042-P::O?I'ptel.wr,,-'tranSducer- ?I(pa)- j 10759 mV+ j ?:Bl.33 !High preaa"e Il!!lNducer ~ (pa). I 4830 S mV + I -.s66 2 ------------- ---------- -----':,.. W_fIow rale hom drun cSb'alion-?--?-------------''',"~ i: '(:, II1l'i via electrcnc pleulJlO tumducei lPfeULftl tt8l'\$ducet~::-.::0'.). 10627:. oN + I -62?.92 i <L;' .. .__1 ;> "-:Alnaphericpre~' 100800 Pa ;':;:)',~euue IIMlduc", cl!llbation for pleStUfe drop evel fil1-1

(" M5U9IinpJI: 1 ',':', ,:j II'1lUt frem fil,,:..,, _ Fik Ildn.;: ldpfil~~ LinJ'.',,':"ej I' I I Figure K.4: Dialog window for processing of the experimental results. K.6 Figure K.S: Dialog window from where the channels are set for the different meas urements. Figure K.S shows the dialog window from where the channels are set for the diffe rent experimental quantities. The dialog shown in figure K.S is obtained by pressing the 'Channels ' button shown in figure KA. If each of the variable buttons shown in figure K.S is pressed a dialog wind ow as shown in figure K.6 appears. For example, the air inlet temperature to the test section will be the average of the four selected data logging channels shown in figure K.6. If the 'Process' button in figure KA is pressed, the data of each fill test is w ritten to an output file in following sequence: Pa. Tai, Twh , Two, rna, mw t?Jfi. K.3A CALCULATION OF THE TRANSFER AND LOSS COEFFICIENTS Figure K.7 shows the dialog window where all the transfer and loss coefficients are calculated. The dialog window shown in figure K.7 appears when the 'Transfer and loss coefficients' but ton is pressed in the dialog window shown in figure K.3. The coefficients according to the Merke~ Popp e and e-NTU can be selected. The integration settings for the Merkel and Poppe approaches can be sp ecified. If a rain zone K.7 andlor a spray zone exist, then the effects of these zones can be subtracted fro m the total measured transfer and loss coefficients. Refer to appendix D for the formulas used to sub tract the influences of the rain and spray zones for both the transfer and loss coefficients. Figure K.6: Dialog window where the channels for a certain measurement are set. The output file generated by the dialog window shown in figure K.4 is used as th e input file in the dialog window shown in figure K.7. An output file with the air and water mass velocitie s and the transfer and loss coefficients according to the e-NTU, Merkel and Poppe approaches is generat ed when the 'Calculate coefficients' button is pressed in the dialog window shown in figure K.7. K.3.5 DETERMINATION OF EMPIRICAL EQUATIONS The dialog window in figure K.8 appears when the 'Empirical equations' button is pressed in the dialog window shown in figure K.3. Empirical equations are generated in this dialog win dow with a least squares curve fitting method. Conventional linear regression with straight-line transformations can not be used to minimize the objective function, because the empirical equations have ge nerally more than two unknown variables. Thererore, mathematical optimization algorithms are employed to obtain the best possible fit for the empirical equation. The form of the equation can be selecte d on the dialog window. The coefficients G, b, C and d are obtained to give the best possible fit for th e selected form of the

equation shown in figure K.8. As already mentioned, the method of least squares is employed to obtain empirica l correlation through the experimentally determined data. The sum ofthe least squares is given bY the func tion [93BE2], . 2 8' =flY, -y(X,)] i=l K.8 (K.l) where Y, are the experimentally determined ordinate for measured values ofx, whe re x, is a vector in It'. j is the number of experimental observations used. Figure K.7: Dialog window fur the calculation oftransfer and loss coefficients. To indicate the reliability ofthe fit, the correlation coefficient, r, is define d where, (K.2) whereYm is the mean of the measuredYI [93BE2]. It can be seen from figure K.8 that the sum ofthe least squares can be minimized by the LFOPC [82SN1, 83SNl, 85SNl], ETOPC [98SNl, OOSN1] or DYNAMIC-Q [94SN1, OOSN2] optimization alg orithms. LFOPC is a gradient method that generates a dynamic trajectory path, from any gi ven stating point to a local optimum. ETOPC is a conjugate gradient method. The DYNAMIC-Q algorithm app lies the dynamic K.9 trajectory optimization algorithm LFOPC to successive spherical quadratic approx imations of the actual optimization problem. Figure K.8: Dialog window to determine empirical relations. It can be seen in figure K.8 that the values of the coefficients can be fixed to a certain value. The optimization problem then changes from an unconstrained to a constrained problem . After the first run of the optimization problem is it sometimes useful to round some of the exponents o ff to less significant digits. The coefficients then can be fixed and the other coefficients can then b e optimized to obtain the best fit. The correlation coefficient,?, will be an indication how the accuracy of the fit is affected by this procedure. K.4 SAMPLE CALCULATION OF EXPERIMENTAL DATA PROCESSING KA.! INTRODUCTION K.lO A sample calculation is presented of how the air mass flow rate, the water mass flow rate, the pressure drop over the fill and the energy balance are calculated from experimental measu rements. Temperature measurements: Water inlet temperature Water outlet temperature Air temperature below fill Wetbulb temperature below fill Air temperature (nozzles) Wetbulb temperature (nozzles) Air temperature in cyclone at air exit Other measurements: Atmospheric pressure Pressure drop - orifice plate (voltage)

Pressure drop - nozzles (voltage) Calibration curves: Pressure drop - orifice pressure transducer Pressure drop - nozzle pressure transducer Pressure drop - fill pressure transducer Pressure drop over troughs Pressure drop ( pressure upstream of troughs minus atmospheric pressure) Other specifications: Number of nozzles Nozzle diameter Frontal area of the fill Wind tunnel area Water pipe inside diameter Orifice plate diameter TWi = 46.J60C (319.51 K) Two =31.89?C (305.04 K) Tal =22.35aC (295.50 K) TWb = 20.WC (293.29 K) To. = 22.WC (295.29 K) Twbn = 19.0s0C (292.20 K) Tao =38.31?C (311.46 K) po = 100500 Pa 2.098510 mV 1.377270 mV lipop = 4830.80x(mV) - 4986.20 lip. = 200.89x(mV) - 206.48 lip, = 627.9x(mV) - 626.92 lip, = 7.72930G;"'70 Iip,p. = 9.7219G;,9"9 nn= 3 d. = 0.3 m AI' = 2.25 m' Atus =4m2 dw~O.13m dop = 0.62 m K.4.2 AIR MASS FLOW RATE The formulas used in this section to obtain the air mass flow rate can be found in [98KRI]. The pressure drop over the nozzles is given, at a pressure transducer voltage of 1.37727OrnV, by lip. = 200.89x(mV) - 206.48 = 200.89x1.377270 - 206.48 =70.19977 Pa The discharge area ofone nozzle is (Ao3.l) (Ao303) K.II Pressure of water vapor from equation (A.2.1) evaluated at Twb., where Twb? = 29 2.20K. z. ~ 10.79586(1 - 273.16/292.20) + 5.0280810g lO(273.16/292.20) + 1.50474 x 10.4[1 - 1O.8.29692{(292.20/273.16)-1 1]+ 4.2873 x 10'4[10 4.76955(1- 273.16/2 92.20) - I] + 2.786118312 ~ 30343007 P~b. = 103.343007 = 2202.538 Pa It is assumed that the nozzle upstream pressure is equal to atmospheric pressure . The humidity ratio then follows from equation (Ao3.5) ( 2501.6 - 2.3263(Twb? - 273.15) )( 0.62509p~b ) W. = 2501.6+1.8577(T?? -273.15)-4.1 84(Twb? -273.15) Pa -1.005P~b' (

1.00416(Ta. -Twb.) ) 2501.6+1.8577(Ta? -273.15)-4.184(Twbn -273.15) ( 2501.6 - 2.3263(292.20- 273.15) X 0.62509? 2202.538 ) = 2501.6 +1.8577(295.29- 273.15) - 4.184(292.20 - 273.15) 100500-1.005? 2202.538 _( 1.00416(295.29-292.20) )= 0.0127123 k Ik 2501.6 +1.8577(295.29 - 273.15) - 4.184(292.20 - 273.15) g g At the measured air drybulb temperature of Ta? = 295.29K, wetbulb temperature of Twb. ~ 292.20K, find the following thermophysical properties employing the equations given in appendi x A. Density ofair-vapor Pa", = 1.176541 kg/m3 Viscosity of the air vapor mixture Po" ~ 1.811918 X 10'5 kg/ms The mass flow rate of air containing water vapor through one nozzle can be given by rna",' = C.?gYtrd;(2Pa"f"p.t' /4 where the gas expansion factor is given by ?g =1- 3f"p. /(4P.I.4) = 1-(3)(70.19977)1[(4)(100500)(1.4)] = 0.9996258 The approach velocity factor is given by Y =1+ 0.5(A. 1Atu,)' + 2(A. 1A",,)2 fj.P. 1(1.4Pa) ~ 1 + 0.5(0.0706858/4)' + 2(0.0706858/4)' 70.19977/[(1.4)(100500)] = 1.000156 The coefficient of discharge, C" is a function of the Reynolds number and is obt ained by an iterative procedure. Assume that the Reynolds number, Re" for the air vapor mass flow thro ugh one nozzle is 210935.1. The coefficient is given by I. K.12 c. =0.9758+1.08xl0-7 Re. -1.6xl0-13 Re; = 0.9758 + 1.08 x 10-7 X 210935.1-1.6 X 10-13 X 210935.1 2 = 0.991462 The mass flow rate of air containing water vapor through one nozzle can be deter mined by maml = C,?.YA. (2Pa"I:1p.)0'> = (0.991462)(0.9996258)(1.000156)[(2)(1.176541 )(70.19977)]?1 = 0.9005343 kgls The corresponding nozzle Reynolds number is: Re = m"".ld. = (0.9005343XO.3) ? A.!l"". (0.0706858ls119l79x W') 210936 This value is within good agreement with the Reynolds number assumed above. The mass flow rate through all three nozzles is m"". = 3m"".1 = (3)(0.9005343) = 2.701603 kgls The corresponding dry air mass flow rate through the system is given by rna ~ rna" I (I +w.) =2.7016031 (I + 0.0127123) ~ 2.667697 kg/s K.4.3 WATER MASS FLOW RATE The water mass flow rate is determined with the aid of an orifice plate accordin g to the British Standard 1042 [8IBRI, 84BRI]. The upstream pressure tapping is one pipe diameter upstream of the orifice plate while the downstream pressure tapping is half the distance. The method used to d etermine the water mass flow rate is almost the same as that used to determine the air mass flow rate. The pressure drop over the orifice plate is given, at a pressure transducer volt age of2.098510mV, by !'>.pop = 4830.80x(mV) - 4986.20 = 4830.80x2.0985I0 - 4986.20 ~ 5171.282 Pa The density and viscosity of the water is determined at the water inlet temperat ure, T.; = 319.51 K Density ofthe water Pw ~ 989.6611 kg/m' Viscosity of the water /.Iw = 5.773254 X 10" kg/ms The diameter ratio, 0; ofthe orifice plate is given by, (j'~ do/d. = 0.62/0.13 ~ 0.4769231

The approach velocity factor is given by y= (I - dro" = (I -0.4769231'r0 1= 1.026917 (A.4.1) (A.4.3) Because the water is incompressible the expansion factor is equal to unity and t hus falls away. The coefficient of discharge, Cd, is a function of the Reynolds number and is obtain ed by an iterative K.13 procedure. Assume that the Reynolds number, Reaw, referred to the inside pipe di ameter, is equal to102022. The coefficient of discharge is given by Cd =0.5959 + 0.0312p'l - 0.1 840P' + 0.0029 p,.,(IX 10 6 )0,75 + 0.09L] 0.0390 _ 0.0337 L,p3 Redw =0.5959 + 0.0312(0.4769231)'1 _ 0.1840(0.4769231)' +0.0029(0.4769231)25( 1x 10 6 )0,75 + 102002 0.09(1.0)0.0390 - 0.0337(0.47)(0.4769231)3 =0.6063129 where LI = I ana L, =-0.47 are accordilig to the orifice plale.setup. The mass flow rate of the water can be determined according to m w = CdYmi: P (2p w L'.Pop t' =(O.6063129)(1.026917)(1r)(O.62)'[(2X989.6611)(5171.282?)o, = 6.014 kg /s The corresponding velocity of the water in the pipe is given by Vw~ 4mj{fJw 1r dw) = (4X6.{)14)/[(989.6611)(n)(ll.13)J=1l,4571lO8'5m1s The corresponding nozzle Reynolds number is: Re = Pwvwdw = (989.661lX0.4578085XO.13) =102021.8 dw J.lw (5.773254 xl0-' ) This value is within good agreement with the Reynolds number assumed above. KAA PRESSURE DROP OVER FILL The pressure drop over the fill and troughs is given, at a pressure transducer v oltage of 1.04819OmV, by !'>p, = 627.9x(mV) - 626.92 ~ 627.9xI.048190 - 626.92 = 31.23854 Pa To obtain the pressure drop over the fill, the effect of the troughs must be sub tracted from the value determined above. Go ~ mjAfr ~ 2.667697/2.25 = 1.185640 kg/m's. !'>p" = 7.7293OG~.88670 =7.72930(1.185640)'-88670 =10.65779 Pa The pressure drop over the fill is !'>pfi ~ L'.P, -!'>p" = 11.23854 - 10.65779 ~ 20.58075 Pa KA.5 ENERGY BALANCE The pressure of the air, where the air drybulb and wetbulb temperatures are meas ured, is higher than atmospheric pressure. This pressure increase is due to the fan. A correlation wa s obtained by measuring the pressure difference between atmospheric pressure and the pressure below the troughs, were the air inlet temperatures are measured. The properties of the air are then determined a t this pressure. The atmospheric pressure is,P. = 100500 Pa. K.14 The pressure increase, due to the fan, from atmospheric pressure is given by Iip.p.= 9.7219G~?9459 =9.7219(1.185640)1.9459= 13.5 Pa

Thus, the pressure upstream of the troughs is given by po( = Pa + lip"". = 100500 +13.54116 = 100513.5 Pa The enthalpy of the inlet air, ima" is found according to equation (A.3.6b). At the specified air inlet drybulb temperature of To( = 295.50K and wetbulb temperature of Tab = 293.29K fi nd the fullowing: Pressure of water vapor from equation (A.2.1) evaluated at Twb, where Twb = 293. 29K. Z, = 10.79586(1 - 273.16/293.29) + 5.02808 loglO(273.16/293.29) + 1.50474 x 10-4[1 - 10..?29692{(293.29I273.I6)-\ l] + 4.2873 x 10-4[10 4.76955(1- 273.16/ 293.29) - I] + 2.786118312 = 3.37242 Pvwb = 103.37242= 2357.34 Pa It is assumed that the nozzle upstream pressure is equal to atmospheric pressure . The humidity ratio then follows from equation (A.3.5) ( 2501.6 - 2.3263(293.92- 273.15) J( 0.62509?2357.34 ) w, = 2501.6 +1.8577(295.50 _ 273.15) - 4.184(293.92 - 273.15)}.100513.5-1.005. 2 357.34 _( 1.00416(295.29-292.20) )= 0.01439305 kglk 2501.6 +1.8577(295.50 - 273.15) - 4.184(293.92 - 273.15) g The enthalpy of the inlet air, imal, is found according to equation (A.3.6b) wit b cpa' = 1006.508 JlkgK and cpw = 1872.142 J/kgK being evaluated at (Ta, + 273.15)12 = (295.50 + 273.15)/2 = 284.325K according to equations (A.1.2) and (A.2.2) respectively. The latent heat is found to be ifgwa = 2.5016 x 106 JlkgK according to equation (A.4.5) at 273.15K. With these values find imal = 59225.82 J/kg dry air. The pressure downstream of the fill is given by Pao = Poi -lipfl = 100513.5 - 20.58075 = 100493.0 Pa The outlet air is assumed to be saturated with water vapor. The enthalpy of the outlet air, imao, is found according to equation (A.3.6b). At the measured saturated outlet temperature of Tao = 311.46K find the following: Pressure ofwater vapor from equation (A.2.1) evaluated at Too = 311.46 K. Zo = 10.79586(1 - 273.16/311.46) + 5.02808 loglO(273.16/311.46) + 1.50474 x 10-4[1 - I0..?29692{(311.461273.16)-l l] + 4.2873 x 10-4[10 4.76955(1 - 273.16 /311.46) - 1] + 2.786118312'= 3.82852 P,o = 103.82852 = 6737.78 Pa It is assumed that the nozzle upstream pressure is equal to atmospheric pressure . The humidity ratio then follows from equation (A.3.5) K.15 w =( 0.62509?6737.78 ) = 0.04969278 kg/kg dry air o 100493.0-1.005.6737.78 The enthalpy of the outlet air, imoo, is found according to equation (A.3.6b) wi th cpao = 1006.730 JlkgK and cp"" = 1897.695 JlkgK being evaluated at (Too + 273.15)12 ~ (311.46 + 273.15 )12 = 292.305K according to equations (A.l.2) and (A.2.2) respectively. The latent heat is foun d to be itlf"O ~ 2.5016 x 10' JlkgK according to equation (A.4.5) at 273.15K. With these values find imoo = 16 8374.7 Jlkg dry air. The enthalpy gained by the air is given by Qo =ma(i..,o -i..,,) = 2.66769(168374.7-59225.82)= 291175.5 W The heat lost by the water is given by Qw =mwcpw,(Twl -273.15)-(mw-mw("apJ}pwo(Two -273.15) It is assumed that the outlet air is saturated with water vapor. The amount of w

ater lost due to evaporation is then given by mw(,,"p) =ma(wo -w,) =2.667691(0.04969278 - 0.01439305) = 0.094169 kgls CPWI and cpwo are evaluated by equation (AA.2) at temperatures of TWI = 319.51 K and Two ~ 305.04 K respectively to obtain values of4177 .766 and 4177.094 JlkgK respectively. Thus, Qw = mwcpw' (Tw, -273.l5)-(mw-mW(,,"PJ}pwo(Two -273.15) =(6.0 13992)(4177.766)(319.51-273.15)-(6.013992-0.094169)(4177.094)(305.04-273.1 5) ~ 298850.5 W The energy balance is given by 100 291175.5 - 298850.5 -2.64% 291175.5 A difference of 2.64% can be considered as a very good agreement due to the unce rtainties associated with experimental measurements. 1.1 APPENDIXL TEMPERATURE DISTRIBUTION DURING NOCTURNAL INVERSIONS L.IINTRODUCTION To determine the effect of nocturnal inversions on the performance of cooling to wers it is necessary to have information concerning the vertical temperature and humidity distributions or profiles. Equations are available in the literature that predict the vertical nocturnal temperature prof iles from ground-based measurements. It will be shown that these equations are largely incorrect, relat ively complex and unsuitable for analytical integration. Existing equations are generally inaccura te in predicting both the , height and magnitude of inversions. In this study we rely on a relatively simple empirical relation that predicts inversion temperature profiles throughout the course of a year, with an acceptable degree of accuracy requiring a minimum amount of input data The proposed empirical relatio n can be analytically integrated and employed in the analysis ofthe influence of inversions on cooling tower performance. L.2 BACKGROUND The structure of the stable thermal boundary layer is outlined in figure 1.1. Th e temperature inversion itself is contiguous with the earth's surface and exhibits an increase in temper ature with height. The inversion region is capped by and isothermal region, which sometimes is accompan ied by a "wind jet". Above that the atmosphere exhibits an adiabatic lapse rate. The two most importa nt parameters pertaining to the stability of the inversion are the magnitude of the temperature distribut ion across the inversion and height of the inversion, which are shown in figure 1.1 [90SUl). The magnitude of the temperature differential, I'!.T, is defined as the difference between the maximum temperatur e and the ground level temperature measured 1 to 2 m above the ground. The height of the inversion top, Z'h is defined as the height at which the actual temperature gradient first becomes zero. i Jldiatlaticlapse o

'--__=-__~ ""Gr~o""Wl""d"__' -+ Temperature, T Figure 1.1: Structure of the stable boundary layer. L.2 Many theoretical studies have been made to predict the nocturnal atmospheric tem perature profile from ground based measurements. The classical method is to assume a semi-infinite med inm bounded on one side by the earth's surface, which is considered to radiate a constant heat flux . Under these conditions Anfossi et al. [76ANl] found that the temperature profile 1('.1) takes the form, ~ [ 2] [ JO., [2 ]0.5}-z z 1C zT -1'. - T -1'. - -- - erfc -(,,1) - <p,O) [(0,0) (0,1) exp 4Kt 2 Kt 4Kt) (L.l) where z is the height above ground level, t (s) is the elapsed time since the di urnal maximum temperature and K is the thermal eddy diffusivity (assumed to be constant), It will be shown that the thermal eddy diffusivity is not constant, especially in the surface boundary layer. The time origin is taken when the diurnal t~perature wave reaches a maximum, i.e. 1(0,0) is the maximum daily temp erature at ground level. In order to improve the accuracy of equation (L.l) Anfossi et al. [76ANl] modifi ed the boundary conditions used in the above-mentioned classical solution. Instead of regarding the medium as semi? infinite, an upper boundary Zu was introduced, The atmosphere was then bounded b y a fixed earth's surface and a time varying inversion "top". This modified solution is given by 1(,,1) =1(0,0) - [1(0,0) - T(o,') ~exp[~] - [.:....]1C 0.' erfc[':""] + O.27l .: ....]}~ zit zit zit 1Zit where Zit is defined by Anfossi et al. (76ANl] as zit = (4Kt)0' This new solution showed a marked improvement over the distribution predicted by equation (L.l). (L.2) (L.3) The temperature profile, as predicted by equation (L.2) requires input data of t he diurnal surface temperature maximum, 1(0,0), the surface temperature at the time of profile extr apolation, T(O,I)' and Zu which can be computed from the time lapse since the diurnal temperature maximum and a "suitable value" of the thermal eddy diffusivity K from equation (L.3). By introducing an extra measurement T , being the temperature measured at the he ight z", there would(ZII'/) be four data points available to extrapolate the temperature profile from the gr ound based temperatures 1(0,0), T(o,i? 1(,.,0) and 1(".", This is double the amount of information used by Anfossi et al. [76ANl]. Of th.e four input temperature data available, only three are required. Of all the combinations of the above four temperatures, the difference T(,..I) - T(o,,) and T(o,o) - T(o,l) can be de termined experimentally to the greatest degree of accuracy. With this information Surridge [86SUl] derives the following equation, (LA) This relation describes the ratio of the vertical to temporal temperature differ ence as a function of the normalized height z"/z,,. The value of z,,/Zit can be determined by solving equa

tion (LA) by successive (L.S) L.3 approximations. Once the value of z.lz" is determined, the inversion height z" c an be calculated by substituting the value ofZm. Thus the value ofZit has been obtained without assu ming a value of K or any temporal variation. The value of Zit may hence be substituted into equation (L.2 ), from which the temperature profile may be calculated up to a maximum ofZit. Equation (L.2) does not generally correlate the data well but does give a reason able indication of the inversion height. The method proposed by Surridge [86SUl] does follow the partic ular data closely over a part of the inversion but Wlder-predicts the inversion height. L.3 SIMPLIFIED INVERSION PROFll.E It is obvious that both of the above-mentioned approaches have their limitations . Furthermore, the equations require considerable temperature data, are relatively complex and do n ot readily allow for further analysis of the influence of inversions on cooling tower performance. In view of these complications, an approximate Kelvin temperature distribution of the form T=(T, + 273.lS{:,J is assumed to be applicable in the stable boWldary layer, where T, and z, are th e reference temperature and reference height respectively. T, (OC) is measured at z, which is aboot I m abov e groWld level. The value of the exponent, b, varies throughout the course of the year. An empirical relat ion for b as a fimction of the day of the year is developed in section LA to section L.6. L.4 MEmOD TO DETERMINE THE EXPONENT OF THE SIMPLIFIED INVERSION PROFILE An optimization technique is employed to determine optimum values of the exponen t, b, ofequation (L.S) from experimental data An objective fimction, Le. the fimction to be minimized, is selected to aid in the determination of b so that equation (L.S) closely represents measured inversion profiles. The objective fimction is selected to be the modulus of the difference between the area under the proposed curve given by equation (L.S) and the experimentally determined area. The areas referred to are the areas Wlder the temperature versus height curves. Thus, the objective is to minimize the differe nce in the area between these two curves. The unit of area is mK. As T, and z, are known, the only solut ion variable is the exponent, b. The area under the curve given by equation (L.S) is obtained throug h integration. (L.6) where z, is the height of the measured air temperature closest to the ground (us ually I to 2 m above ground level) and z, is the height ofthe temperature sensor on top of the weathe r mast. Since the area under the temperature versus height curve, expressed by equation (L.6), includes the area from the datum ofthe temperature axis (0 K) to the temperature T, at height z" i t must be subtracted from the area given by equation (L.6). Therefore, the new area is given by LA

A =~+273.l5)(b+l_ b+I}_(7' 273l5X _ )b( ) Z, ZI 1 1 + . Zt Zlz, b+l (L.7) The areas lUlder the experimentally determined temperature dis1ributiOllS at eac h time interval are approximated, as shown in figure L.2, through the usage of rectangles and triang les. The approximated area lUlder the experimentally determined temperature-height curve is denoted by Au. It can be seen that an error is made due to the choice of the triangles, but as the curvature betwee n two adjacent points is not significant, the error is not significant. Ts---------------T3 -------T2 ----Zs Figure L.2: Area under experimentally determined curve The objective function, F, is the modulus of the difference between the area det ermined by equation (L.7) and the area obtained by the method illustrated in figure L.2, i.e., F=IA-A 1- (T,+273.l5)( b+l_ b+I}_T.( - )-A u - b ( ) z/ Zl I Zt Zl ex z, b+l where Au is the is area determined by the graphical method illustrated in figure L.2. L.5 REFERENCE HEIGHT AND TEMPERATURE (L.S) The influence of the reference height, z" on the value of the exponent, b, and t he accuracy of equation (L.5) are determined in this section. The calculations are performed for referen ce heights of 1 m, 2 m and 5 m respectively. Reference heights of 1 to 2 m are preferred as the temperature at this height is usually measured by meteorologists. L.5 100 10 20 re...,el'olure. ?C 10 Figure L.3: Temperature inversion profiles during first six hours after inceptio n of an inversion [98KRI]. Krtiger [98KRI] presents hourly temperature data measured at eight different hei ghts on a 96 m high weather mast. Figure L.3 shows the temperature inversion profiles, presented by Krtiger [98KRI] during the first six hours after the inception of the inversion. The exponent, b, for e ach hourly interval during the period when nocturnal inversions occur, is determined by the method discussed in the previous section. Table L.I gives the optirmnn values of the exponent b of equation (L.5) at hourl y intervals for the different reference heights for the data given in Krtiger [98KRI], determined by the method discussed in the previous section. The last row of table L.I gives the average values of the exponent b during the inversion period for the three different reference heights. It can be seen that the optimal value of the exponent is approximately 0.0 I for reference heights of I m and 2 m. For a refe rence height of 5 m, however, the optimum value of the exponent b is approximately 0.013. Table L.l: The optimum value of the exponent b for reference heights of 1m, 2 m and 5 m respectively.

Time b z -1m zr- 2m Zr= 5m 19hOO 0.008898 0.008722 0.008734 20hOO 0.009632 0.01018 0.010082 21hOO 0.01205 0.011314 0.012426 22hOO 0.009235 0.009528 0.011352 23hOO 0.0093 0.010126 0.012104 OOhOO 0.01 1304 0.013027 0.016928 OlhOO 0.009983 0.01 1525 0.015631 02h00 0.009359 0.010535 0.012583 03hOO 0.008853 0.010051 0.013042 04hOO 0.010067 0.011842 0.01589 05hOO 0.009796 0.011362 0.014982 06h00 0.009949 0.011317 0.014312 07h00 0.009684 0.010472 0.01239 Average 0.009855 0.010769 0.013112 L.6 ? Measured values /.., --zr=1 I..,??? %1"2 , - -zr=5 l' 11/I J/ I' V .' l?' I-.~Y, ~ -.I-~. ~ "r. - -- -o ~w~~~~m~~~~mm~~~m~m Temperature. K 10 20 30 80 70 80 90 e ~ 50 J 40 100 Figure LA: Height versus temperatw:e profiles of equation (L.5) fur three refere nce heights with the optimmn value of the exponent b determined for each reference height. Figure LA i11ustrates a sample case of the temperature versus height profiles at 23hOO for cases where the different reference heights are employed together with the optimmn exponents giv en in table L. J. The measured data is also shown. It can be seen that nearly all the curves approxima ted by equation (L.5), at different reference heights, with optimmn b values, represent the measured data relatively accurately. If the curve, generated with a reference height of 5 m is extrapolated, it may not represent the real inversion profile as accurately as for reference heights of I and 2 m. It is strongly reco mmended that the same

reference height be employed when the value of b is determined as in the subsequ ent employment of equation (L.5), with b known, to obtain the temperature profile. L.6 SEASONAL VARIATION OF INVERSIONS In this section, the seasonal variation of the exponent, b, of equation (L.5) is investigated. The same procedure is followed to obtain the optimum values of the exponent, b, as discus sed in section LA. The reference height, Zn in equation (L.5) is chosen as 2 m in this investigation. Data was collected on a 96 m weather mast from 16 March 1994 to 30 January 1995 in a relatively arid area near Lephalale (Ellisras) (23?40'8, 27?47'E). Wind direction, wind speed an d drybulb temperatures were measured at heights of I, 2, 5, 10,20,40,65 and 96 m approximately every si x minutes. There were many times during the above-mentioned period where no data was collected. When d ata was collected, it was done intermittently with the result that continuous data sets, extending ove r periods where inversions and low wind conditions (less than 2 mls at all measured heights) were present, are relatively scarce. Table L.2 gives the hourly temperatures at different elevations during the occur rence of nocturnal L.7 inversions throughout the course of a year. Low wind conditions were present dur ing these selected periods of investigation. Table L.2 (a): Hourly temperatures, wind speed and wind direction at different h eights. A1t. 21/22 March 1994 m 19hOO 20hOO 21hOO 22h00 23hOO OOhOO OlhOO 1 21.2 19.9 19.2 19.9 18.5 16.5 17.3 2 22.5 21 19.8 20.4 19.3 17.3 18 5 24 22.2 21.1 20.9 19.8 18.4 19.5 ~ 10 25.3 24.7 21.8 21.4 20.9 19.1 20.1 ,.; 20 26.2 25.9 25.5 22.6 22.2 20.3 20.7 40 25.8 25.9 25 23 23.5 20.9 21.6 65 25.7 25.8 25.2 24 25 21.7 21.9 96 26.1 26.1 25.7 24.9 25.1 22.7 21.8 10 0.8 0.6 0.6 1.1 0.5 0.9 0.9 "' 20 0.9 0.4 0.6 1.8 0.4 0.6 1.3 '8 40 1 0.3 0 1.6 0.7 0 1.9 ,: 65 0.6 0.4 0.1 1.8 2.8 0.9 2.2 96 0.4 0.4 0 2.2 2.7 2 1.4 10 235 246 181 41 234 215 98 0 20 189 175 222 29 241 208 101 is 40 184 179 139 44 72 150 9965 181 178 166 53 73 83 93 96 181 159 139 62 57 8 70 Table L.2 (b): Hourly temperatures, wind speed and wind direction at different h eights. Alt. 23/24 April 1994 m 19hOO 20hOO 21hOO 22hOO 23hOO OOhOO 01hOO 02hOO 03hOO 04hOO 05hOO 06hOO 07h00 I 18.4 16.4 16 15.6 14.6 12.5 11.1 10.9 10.4 10.1 11.4 11.1 12.6 2 19.2 17.2 16.6 16.1 15.2 13.6 11.6 11.5 10.8 10.6 11.9 11.3 13 5 20.2 18.4 17.7 16.9 15.8 14.8 12.7 12.6 11.5 11.8 12.5 12.3 14.2 U 10 20.9 19.3 18.3 17.6 16.9 16.7 14.4 13.3 13.2 13.3 13.1 13.1 15 0 ,.; 20 21.9 21.1 19.4 19.3 18.7 18.7 16.1 15.4 17.2 15.9 14 13.9 15.2 40 24.5 22.2 20.9 21.8 20.1 19.5 17.4 17.8 18 16.8 16.2 15.7 15.9

65 24.7 23.4 22.3 22.4 21.5 19.8 17.8 18.1 18.5 17.3 18 17.1 16.4 96 24.8 23.9 23.2 23 23 20.7 19.3 19 20 18.1 18.6 17.1 16.5 10 0.4 0.3 0.3 1.3 0.6 0.7 0.1 0.7 1.1 0.9 1.8 1.9 1.8 ~ 20 0.1 0.3 0.2 1.5 0.7 0.8 0.1 0.8 1.2 0.9 1.6 1.7 1.640 4 4.3 4.2 5 4.3 4.2 3 .7 4.3 4.7 4.2 4.6 4.5 4.3 ,: 65 I I I 2.2 1.3 0.8 0.1 1.1 1.4 0.9 2 1.9 1.8 96 1.2 1.1 1.5 2.5 1.4 0.7 0.1 1.1 1.3 0.9 1.5 1.5 1.4 10 18 66 36 129 134 145 120 134 143 149 159 162 164 0 20 61 111 116 138 141 141 163 139 142 149 152 156 159 .~ 40 182 181 183 174 179 178 183 177 175 179 176 177 179 0 65 209 191 197 161 163 129 172 156 141 127 121 105 103 96 187 178 181 158 161 128 111 151 138 129 117 102 101 L.8 Table L.2 (c): Hourly temperatures, wind speed and wind directioo at different h eights. Alt. 8/9 June 1994 m 19hOO 20hOO 21hOO 22hOO 23hOO OOhOO OlhOO 02h00 03hOO 04hOO OShOO 06h00 07h00 I 10.4 10.3 II 9.2 7.9 6.6 504 4.6 4.3 3.9 4 4.2 2.7 2 11.4 II 11.3 10 8.6 7.5 604 5.6 4.9 404 4.7 4.7 3 5 14.2 12.06 12.1 1104 9.8 9.9 8.1 704 5.7 5.3 5.7 5.2 3.5 U 10 1604 15.6 13.6 12.5 10.6 lOA 9.3 904 7.1 6 6.8 8 4.80 .... 20 17.2 1704 14.5 13.7 12.2 12 11.8 11.8 11.1 8 9.1 10.2 904 40 16.9 17.8 15.5 14.9 14.7 13.6 12.5 12.8 12.6 11.8 10 11.9 11.5 65 17.1 17.8 16.8 16 15.7 15.7 14.6 13.9 14 13.5 11.9 12.7 12.3 96 17.7 17.8 17.2 16.9 16.4 16.7 15.9 14.7 1404 14 13.3 12.9 12.7 10 104 0.7 0.7 1.1 0.6 0.7 0.7 1.1 1 0.6 0.3 0.7 0.9 '" 20 1.6 2.3 0.7 0.3 104 0.5 1.5 1.5 2.2 0.7 0.1 0.3 1.3 8 40 1.7 1.9 0.6 1.1 2 2.1 2.3 1.8 1.9 2.6 0.2 0.9 0.1 ,: 65 2.6 2.2 1.8 1.6 1.6 2.6 3.1 0.8 104 2.2 0.9 1.4 1 96 2.2 2.2 2.7 1.6 2.2 2.8 2.7 I 1.2 2.6 1.5 1.3 1.9 10 152 113 279 223 267 175 194 211 195 145 158 301 226 0 20 151 109 269 202 249 228 230 212 205 187 303 320 189 is 40 139 113 183 184 196 218 204 180 181 186 323 58 15765 114 104 110 157 179 2 09 183 109 138 158 37 87 88 96 109 101 107 134 156 186 179 109 129 159 65 95 83 Table L.2 (d): Hourly temperatures, wind speed and wind direction at different h eights. Alt. 23/24 July 1994 m 19hOO 20hOO 21hOO 22hOO 23hOO OOhOO 01hOO 02h00 03hOO 04hOO 05hOO 06h00 1 11.6 8.9 8.6 7.7 7.1 6.7 5.3 4.6 3.7 2.9 2.8 2.6 2 12.5 9.9 10.1 8.5 7.8 7.6 6.2 504 4.6 3.6 3.3 3.3 5 13.9 12.9 12.6 II 9.9 8.9 8.2 6.6 5.8 5.2 4.2 404 ~ 10 17.2 15.1 13.9 13.7 12.7 11 11.6 8.2 9 7.8 6.2 6.2 .... 20 20.3 16.5 17.3 15.9 16.6 14.3 1304 11.9 13 10.8 lOA 8.4 40 20.9 18.9 19.6 17.9 17.7 1504 15.3 13.4 13.8 12 12.2 10.5 65 21.5 2Q.4 19.8 18.2 17.7 16.9 16.1 13.9 14.9 14.2 12.6 12.6 96 21.9 20.8 20.2 18.8 17.6 17.3 16.6 15 15.2 15 15.8 14.3 10 1.3 0.9 0.1 0.2 0.2 1.5 0.1 1.5 0.9 0.6 104 004 '" 20 1.9 1.3 1.4 1.9 204 1.2 0.8 204 0.9 0.2 2.6 0.3 8 40 2.3 1.7 3.2 2.5 2.9 0.8 204 2.2 1.1 0.9 2 0.8 ,: 65 2 2.5 3.3 204 3.3 2.9 2.6 1.2 3 204 0.6 1.9 96 2 2.9 2.7 1.9 2.9 2.8 2.5 0.9 2 1.7 1.3 2.1 10 0 114 179 101 91 238 221 2.Q4 226 239 208 288 0 20 356 95 89 98 83 219 105 220 176 120 197 331 is 40 359 59 101 99 73 122 98 209 99 97 183 4665 349 32 92 86 71 95 74 194 82 64 180 70 96 341 16 65 52 64 77 58 113 66 27 184 58 L.9

Table L.2 (e): Hourly temperatures, wind speed and wind direction at different h eights. Alt. 5/6 September 1994 m 19h00 20hOO 21hOO 22hOO 23hOO OOhOO OlhOO I 16.4 15 13.3 1\.8 10.2 9.6 9 2 17.8 17.1 14.6 12.8 II 10.4 9.8 5 20 20 16.7 14.1 12.3 I\.9 I \.3 u 10 2\.7 22.7 19.8 15.8 15.7 15.2 14.20 Eo-: 20 24.3 23.1 23.1 19.5 19.5 17.4 16.7 40 25.4 22.7 23.2 21 20.5 18.8 18.3 65 25.4 23.5 23.4 21.3 2\.8 21.5 20.6 96 25 24.8 24.3 22.7 22.2 2\.8 2\.2 10 0.8 0.9 1 0.6 0.4 0.4 0.1 ~ 20 2.4 0.9 0.7 1.3 0.7 0.6 I 13 40 2.8 0.9 0.5 1.1 2.2 2.4 3.1 ,:: 65 2.4 1.4 0.7 0.8 2.7 3.4 3.4 96 2.5 2.1 \.4 1.l 2.7 2.3 2.8 10 I 251 191 235 6 186 235 0 20 348 246 266 176 54 92 78 .!i' 40 357 273 286 168 70 94 86 Cl 65 358 305 315 133 66 70 60 96 358 336 350 66 67 51 44 Table L.2 (f): Hourly temperatures, wind speed and wind direction at different h eights. Alt. 19/20 October 1994 m 02hOO 03hOO 04hOO 05hOO 06h00 1 15.6 15.2 14.9 14.3 13.4 2 16.5 16 15.7 15.1 14.1 5 17.6 17.5 17.1 17.2 16 u 10 18.9 18.9 18.9 19 17.50 ?.; 20 20.7 21.1 20.3 20.5 19.4 40 22.3 22.1 2\.7 2\.6 2\.3 65 22.5 22.2 22 22.2 22 96 22.4 22.3 22.1 22.1 22 10 \.4 \.8 0.5 1.4 \.3 ~ 20 \.5 0.9 0.1 \.5 2 13 40 4.2 1.5 \.7 2.9 4.8 ,:: 65 3.5 1.7 1.4 \.9 3.8 96 2.7 2.4 \.6 1.5 2.5 10 98 91 98 95 83 0 20 107 91 221 95 85 a 40 95 96 96 95 9565 89 39 69 70 78 96 69 7 13 47 59 1il't1i1i L.l0 Table L.2 (g): Hourly temperatures, wind speed and wind direction at different h eights. All. 2/3 December 1994 m OOhOO OlhOO 02h00 03hOO 04hOO 05hOO 1 21.8 20.4 18.5 18 18.1 18.2 2 22.4 21.1 19.2 18.7 19 18.9 5 22.5 21.7 20 20 19.8 19.8 U 10 23.2 23.4 20.4 20.9 20.5 20.4 0 .... 20 24.3 24.1 22.6 23.2 21.9 20.8 40 24.9 24.7 24.2 24.1 24.2 21.7 65 25.4 25.2 24.5 24.3 24.6 22.3 96 25.7 25.7 24.6 24.3 25.2 22.8 10 0.4 0.4 0.7 0.8 0.9 1.4 '"

20 0.9 0.3 1.7 1.6 1.2 2.1 '8 40 0.7 0.2 0.7 0.7 2.3 4.6 ,: 65 1.9 1.1 0.9 1.5 2.9 3.9 96 2.8 2 0.8 1.1 2.7 2.5 10 96 48 276 274 166 95 0 20 179 265 339 341 253 99 a' 40 276 274 282 286 278 10065 218 226 304 337 256 87 96 212 227 270 303 236 81 Table L.2 (h): Hourly temperatures, wind speed and wind direction at different h eights. All. 201Z1 December 1994 m 19hOO 20hOO 21hOO 22hOO 23hOO OOhOO 01hOO 02h00 03hOO 04hOO 05hOO 1 30.8 28.2 26.4 24.7 23.7 24.2 22.8 21.1 19.6 19.9 20.2 2 31.3 28.9 27.2 25.4 24.5 24.8 23.4 21.7 20.1 20.6 20.9 5 31.7 29.6 28.5 26.6 25.6 25.1 24.1 22.6 20.8 22 22.4 U 10 32.3 30.7 29.1 27.8 26.5 25.5 24.9 24 22.9 23.1 23.40 ,.; 20 32.9 31.8 29.7 28.3 27.8 26.2 25.9 24.6 24.8 24.6 24.4 40 32.8 32.2 30.7 28.9 28.3 26.7 26.4 25.8 25.7 25.7 25.4 65 32.7 32.2 31.7 30.3 30 28 27 26.6 26.2 26.4 26 96 32.5 32 31.8 31.4 31.4 28.5 27.4 26.9 26.6 26.5 26 10 1.1 0.4 0.6 0.6 J.l 1.9 1.9 1.1 0.3 1.8 1.9 '" 20 0.9 0.3 0.4 0.5 0.9 1.5 1.5 0.9 0.3 1.4 1.5 '8 40 1 0.3 0.5 . 0.6 J.l 1.8 1.8 1 0.3 1.7 1.8 ,: 65 1.1 0.7 0.8 0.8 1 2.2 2 1.2 0.7 2 1.9 96 0.9 0.4 0.7 0.7 0.8 1.8 1.7 . 1 0.5 1.7 1.6 10 137 136 136 136 136 137 137 137 136 137 137 0 20 145 144 145 145 145 145 145 145 146 145 145g 40 139 139 140 139 139 139 139 139 140 139 139 65 108 61 65 68 100 113 117 108 81 115 114 96 115 52 48 42 88 115 118 110 71 117 116 Table L.3 shows the optimum values of the exponent b for each hourly interval sh own in table L.2 with average values of the exponent for each inversion period. L.ll Table L.3: Values of the expooent, b, at hourly intervals. . I 2 3 4 5 6 7 8 Time Mrt Apr June 8/9 July Oct 19/20 Dec21122 23124 23/24 Sept 516 Dec 2/3 20121 19hOO (0.0037) 0.0051 (0.0065) 0.0093 (0.0077) (0.0015) 20hOO (0.0052) 0.0058 (0.0073) 0.0102 (0.0070) (0.0033) 21hOO (0.0057) 0.0052 0.0051 0.0099 (0.0094) 0.0040 22h00 0.0033 0.0059 0.0059 0.0101 0.0088 0.0045 23hOO 0.0049 0.0059 0.0067 (0.0116) 0.0107 0.0050 OOhOO 0.0043 0.0065 0.0077 0.0091 0.0103 (0.0027) 0.0026 OlhOO 0.0038 0.0065 0.0079 0.0102 0.0102 0.0040 0.0034 02hOO 0.0065 0.0085 0.0088 0.0056 0.0050 0.0045 03hOO 0.0080 0.0088 0.0105 0.0063 0.0056 0.0060 04hOO 0.0069 0.0082 0.0104 0.0063 0.0052 0.0055 05h00 0.0052 0.0069 0.0101 0.0071 0.0032 0.0049 06h00 0.0051 0.0080 0.0089 0.0077 07h00 0.0034 0.0092 Ave. 0.0058 0.0075 0.0098 0.0100 0.0069 0.0050 0.0044 nd 81 113 159 204 248 293 337 354 Tro?Ol 33.1 30 27.7 30.9 35.1 38.8 37.8 Time 14h44 14h46 15h20 15h50 15h04 15h59 16h20 The values in table L.3 that are printed in brackets are omitted in the calculat ioo of average values of b. This is because the vertical inversion profiles were not monotonic functioos, or significant winds were

present at ooe or more of the heights where measurements were made. Generally va lues of b were also ignored for a few hours immediately after sunset. 121 151 181 211 241 271 301 331 361 Number ofthe day ofthe year, n d 916131 I? Observed values --Sinusoidal function [ 5 /4 ~ ".>1 x\. ~ / .!<la...x /?3 1 \ ~0.S " " rIl C" / " ?6\" OiOi " g$ 7z .>1 g \~.!<l0 <:- ~... / II 7~.S~ "'- ./' . I T " ... Ib 0.0035sin(0.0177? nb 2.32392) +0.0065rI I0.002 1 0.003 0.004 0.011 0.008 0.009 0.010 ... ~ 0.007 i!!! 0.006 0.005 Figure L.5: Summary ofannual variatioo ofthe expooent b. L.l2 Figure L.5 shows the average values of b for a particular inversion period, give n in table L.3 as it varies throughout the course ofa calendar year. The tilt in the earth's axis ofrotation is the cause ofthe seasons. The earth's axis of rotation is tilted 23.5? with respect to the plane of its orbit arOlmd the SlID. Hoffinann [97HOI] presen ts a summary of important dates during earth's orbit armmd the SlID. When it is winter in the southern hem isphere, the south end of the axis of rotation is tilted away from the SlID, while in summer, it is tilted towards the sun. The day when the axis is tilted exactly towards the SW1 is called the summer solstice, i .e. 23 December in the southern hemisphere. On 21 JIIDe, the axis of rotation is tilted directly away f rom the SlID and is called the winter solstice. On 22 March and 22 September, the tilt is in a plane tangential to the earth's orbit arolIDd the SlID. These days are called the autumn and spring equinoxes, respectively. T hese dates are shown in figure L.5 and indicate the relation of these dates to the sinusoidal function v ariation of the exponent, b. Due to the eccentricity of the SlID with respect to the earth's orbit, the earth is closer to the SlID when it is summer in the southern hemisphere, and there is a tendency fur seasonal differen ces in temperature to be

greater in the southern hemisphere than in the northern hemisphere. The average values of b presented in table L.3 can be correlated by b = 0.OO35sin(0.0177? nb - 2.32392)+ 0.0065 (L.9) where nb is the number of the day of the year (I January is the first day of the year). It can be seen from figure L.5 that equation (L.9) correlates the observed values ofb relatively wel l. Figure M.5 shows the annual variation of relative humidity measured at 08hOO and 14hOO at Pretoria, Germiston and Pietersburg. Figure M6 shows the average monthly minimum and maxim um temperatures at these locations. It can be seen that the annual relative humidity and tempera ture distributions can also be correlated by sinusoidal functions. It would therefore appear that the expone nt, b, at a particular location, is a function of humidity. The exponent b may also be dependent on oth er variables such as wind speed, heat flux, evapotranspiration and the albedo of vegetation or surfac e cover. At other locations the exponent, b, determined by equation (L.9), is not known a nd can be approximated by equation (L.5). This is done by taking temperature measurements at two differ ent heights and solving for b in equation (L.5). The value of b for the 23/24 July inversion period in t able L.3, determined by the method described in section LA, is 0.0098. Ifthe highest elevation temperature m easurements are made at 2, 5 and 10 m, the average value of b is 0.0057, 0.0067 and 0.0091 respectively. It can therefore be concluded that the one temperature measurement be made at as high elevation as p ossible (typically 10 m) and the second measurement at I or 2 m above grolIDd elevation. L.7 COMPARISON OF SIMPLE INVERSION PROFILE TO OTIIER MODELS Equation (L.l), (L.2), (LA) and (L.5) are plotted to compare the accuracy of the se equations to the measured data presented in table L.2. K is taken constant at 0.3 m'/s for equati ons (L.I) and (L.3). Surridge [86SUI] states that Tm must be measured at the top of a low mast, there fore Zm is taken as 10 m L.13 in equation (LA). Lower heights of Zm generally lead to less accurate results in this investigation. In equation (L.5) Z, is taken as 2 m and b is determined by equation (L.9). Figures L.6 to L.8 show sample cases of inversion profiles extrapolated from gro und based measurements for one inversion period at 22hOO, 02hOO and 04hOO during the night of 20 and 21 December. ....... Eq (L.1): Anfossi et al. [78AN1] c : , ? .'l?? - ?Eq (L.2): Anfossi et al. [76AN1] ? " --Eq (LA): Surridge [86SU1j ? "l- ? " -Eq(L.5) ? ? "c Measured values ? .? . . c' . , ?

. . 20/21 December 22h00 ? . , nd = 354; b = 0.0040 , t???? >- , rrT(o.o) = 37.8 ?c (16h20) , .' . .'>- T(o,/) = 24.7 ?c :1 .. JTm = 27.8 DC; Zm = 10 m ..>-- ,T, = 2504 DC; Z, = 2 m ""j--" 1/ , , rT/." '.. ~ 100 90 80 70 E 60 i 50 .2' .. :J: 40 30 20 10 o 15 17.5 20 22.5 25 27.5 30 32.5 35 37.5 40 Temperature, ?C Figure L.6: Comparison between temperature inversion profiles. I I c ? "? .'? "I- 20/21 December 02hOO ? -nd = 354; b = 0.0040 ? "? .I- 1(0.0) = 37.8 ?C(16h20) ,? .' , ? .T(o,/) = 21.1 ?c , ,.ITm = 24?C; Zm= 10m ,.c "IT, -2L7?C;z, -2m ,I- , "j,, , ., V, ., ? , , . , / . . ./ " , . ..... -- Eq (L.1): Anfossi et al. [76AN1] "J .' V, . ./ - ? - ?Eq (L.2): Anfossi et al. [76AN1], '7- V --Eq (LA): Surridge [86SU1]~/~ r./ -Eq(L.5) I'p- c Measured vaiues~ 100

90 80 70 E 60 .,; &. 50 '"'iii :J: 40 30 20 10 o 15 17,5 20 22.5 25 27.5 30 32.5 35 37.5 40 Temperature, ?C Figure L.7: Comparison between temperature inversion profiles. L.14 I c ? .'? ? .' . - 20/21 December 04hOO ? .n" = 354; b = 0.0040 ? .' .' T(o,o) = 37.8?C (16h20) ? . , . 1(0,1) = 19.9 ?C , Jo'ofTm=23.1 ?C;zm= 10m ? .' . a . -T, -20.6 ?C;z, -2m , 0f- ., . o? j, 0, .? ,1 ? 0 /.0 .. /1.- ;-- //' ....... Eq (L.l): Anfossi et al. [76AN1] 'f / - - - .Eq (L.2): Anfossi et al. [76AN1], .'.' ?? '1/ --Eq (LA): Surridge [86SU1]'.' -Eq(L.5)'.' l'~ 0 Measured values~ 100 90 80 70 E 60 i 50III ! 40 30 20 10 o 15 17.5 20 22.5 25 27.5 30 32.5 35 37.5 40 Temperature, "C

Figure L.8: Comparison between temperature inversion profiles. 100 90 80 70 E 60 i 50III ! 40 30 20 10 I ,.I I .,? .0 - 23/24 July OOhOO .' ? . nd = 204; b= 0.0099 , .?, J , . - T(o,o) = 27.7?C (15h20) ? II .., ,. T(o,l) = 6.7 ?C , . f- .Tm= II ?C;zm= 10m ? oj .... T, =7.6?C;z,=2m ? .f- . ? / ...... /?? ? .I? I? ? . " II V,? ./ ....... Eq (L.1): Anfossi et al. [78AN1] ? . LV? . ? - - oEq (L.2): AnIDssi et al. [76AN1].? ? . /' --Eq (LA): Surridge [86SU1]." ~ -Eq(L.5)'.' ..' 0 Measured values o 5 7.5 10 12.5 15 17.5 20 22.5 25 27.5 30 Temperature, 'C Figure L.9: Comparison between temperature inversion profiles. L.J 5 10 12.5 15 17.5 20 22.5 25 27.5 30 32.5 35 Temperature, "C 7.5 ? 'c . ..... _. Eq (L.1): Anlossi et al. [76AN1] .. ? . . - - ? ?Eq (L.2): Anfossi et al. [76AN1j . ? , --Eq (LA): Surridge [86SU1] ? , ? .' -Eq(L.5) . ? .? c Measured values ? ,J " ? J . .'c "? . , / ..? .'. II? , ? 1"? ??*Il / 5/6 September OlhOO f--?? .' nd = 248; b = 0.0096? .. 1 T(o,O) = 30.9 ?C (15h50) f--? .' /? .'.- T(o,t)=9?C f--. P? , Tm= 14.2 DC; Zm= 10m.-' ..~ ~ T, = 9.8 ?C;z,=2 m f-~1'-'.Jl.ool",.." Io

5 70 30 20 10 100 90 80 E 60 i 50III ~ 40 Figure L.IO: Comparison between temperature inversion profiles. 7.5 10 12.5 15 17.5 20 22.5 25 27.5 30 32.5 35 , " .' 23/24 Apri123hOO 11,} = 113; b = 0.0054 T(o,o) = 30?C (14h46) 1(o.~ = 14.6 ?C Tm= 16.9 ?C; Zm= 10m T, =15.2?C;z,=2m ?, ? ...~ .' /1 ?c ?, ? :' J.:j ? 1/'.' , " .",{" ',' ... -? .. Eq (L.1): Anfossi et al. [76AN1] ? c ,.' - ? - .Eq (L.2): Anfoss et al. [76AN1] 1-+_+-J..1-----++----+~,'-'------j~-+----I --Eq (LA): Surridge [86SU1j : .... -Eq(L.5) c Measured values 100 90 80 70 E 60 i 50III ~ 40 30 20 10 0 5 Temperature, 'e Figure L.II: Comparison between temperature inversion profiles. L.16 Figures L.9 to L.II show sample cases of inversion profiles at other periods. It can be seen that equation

(L.5), proposed in this investigation, correlates the measured data very well du ring a particular inversion period and throughout the year. L.S AVERAGE TEMPERATURE AND PRESSURE DISTRIBUTION For the proposed further analysis of the influence of inversions on cooling towe r performance, it is necessary to validate the accuracy of equation (L.5) and equation (L.9). This is done by comparing the average temperature in the stable boundary layer obtained by measurement to that determined by the integration of equation (L.5), The approximate pressure difference over an altit ude of 96 m is also determined from the measured temperatures where the stable boundary layer is 96 m or thicker. This pressure difference is compared to the pressure difference obtained by manipulat ion of equations (L.5) and (E.3). At every hourly interval during all the inversion periods given in table L.2, th e average difference between the average temperature in the stable boundary layer (between z ~ I m an d 96 m) obtained experimentally and empirically is 0.76 K. The maximum difference is 2.68 K. For the pressure distribution, the difference is only 0.50 % with a corresponding maximum differe nce of 1.35 %. The empirical relation for the exponent, b, as given by equation (L.9) substitut ed into equation (L.5) gives very satisfactory results for both the mean temperature and pressure differences in the stable boundary layer. L.9 THERMAL EDDY DIFFUSMTY In the analyses of Anfossi et al. [76ANI] and Surridge [86SUl] the value of the thermal eddy diffusivity is assumed to be constant. It can be seen from figures L.6 to L.ll that these eq uations with this assumption do not predict the temperature profile accurately. It will be shown t hat the reason 'for this inaccuracy of the equations is in part due to the fact that the thermal eddy dif fusivity is not constant. At the present location, the temperature at ground level during an inversion var ies approximately linearly as a function of time. Figure L.12 shows the temperatures at ground level, given in table L.2, for the different periods ofthe year. It can be seen the gradient of all the lines in fi gure L.12 is approximately the same. The temperature at I m above the ground can be approximated by 1; =c,l+c, where T, is in ?C and c1 and C2 are constants. From equation (L.5), with a reference height of I m, the vertical inversion prof ile is given by T =(1; + 273.lS)Zb Substitute equation (L.lO) into equation (L.II) and find T =(Cl+ c2 + 273.lS)Zb (L.IO) (UI) (L.12) r' , , I

i 35 30 (J ? 'i 25 .!! ... c: ::l e 20 '".. .. ! 15il I! .. ... E 10 S .. < 5 1.17 __21/22 March __23124 April --.-8/9 June ___23124 July ~I--. """"*-5/6 September __ 19120 October Il'--- -fr- 2/3 December -a- 20121 December -..r--...r-- r----- '--t'-....... ........... ............ .... ~~ ~ -~ ~ ~ ....... .....,~ -05hoo03hoo23hoo21hoo o 191100 01hoo Time, h Figure L.IZ: Nocturnal air temperatures 1m above ground level at different perio ds. From Fourier's one-dimensional heat conduction equation it follows that ~(k, ~)=pcp(~) Differentiate equation (L.12) with respect to t and find oT b -=cz 8t ' Differentiate equation (L.12) with respect to z and find (L.13) (L.14) oT =b(c,t + c2 + 273 .15 )Zb-loz Substitute equations (L.14) and (L.15) into equation (L.l3) and find ~ [k,b(c,t+c2 + 273.l5)zb-l] =PC1CpZb ' Integrate (L.16) with respect to z to find after rearrangement (L.l5) (L.l6) (L.17) where c, is a constant introduced due to indefinite integration which willlster

be shown to be qr-'). From Fourier's equation ofheat conduction it follows that BT q=k,& Substitute equations (L.15) and (L.17) into (L.IS) to fmd (L.1S) -,1IlW:1:,:: L.18 (L.19) pc C Z'+I q = p I +c b+ I ' From equation (L.19) with z =1 "" 0 find c, =qz=o "" qz=1 where qRJ is the heat flux conducted from the air to the grOlmd due to radiation from the ground ioto the night sky. Figure L.13 shows the temperature vs. time gradients of the data given in table L.2 of the atmospheric air during nocturnal inversions for various times during the course of the year. It can be seen that the temperature gradients are relatively constant throughout the year at ground leve l and at a height of 96 fi ? The average temperature gradient at grmmd level, from figure L.13 is approximate ly ~.7787 Kih or -2.163xI0-4 Kls, thus from equation (L.lO), c1 '" -2.163x 1O-4K1s (L.20) For the inversion period of 23/24 July in table L.2, the number of the day of th e year is, n" = 204 with a corresponding value of b = 0.0099 from equation (L.9). 90 120 150 180 210 240 270 300 330 360 Day of the year 6030 ? 00 ? ? ? Q 0 ~ ? ? 0 0 096m I-.1m-0.1 o o .c liZ _ -0.8 "E .. -0.7 :!:: c " -06.. ' !. Cl. -0.5 e 'tl!! -0.4 ! -0.3 8E -0.2 ~ -0.9 -1 Figure L.13; Temperature drop per unit time at I m and 96 m.

The value ofc" for the inversion period, from equation (L.1O) is 11.103 ?C, i.e. the temperature above the ground at the beginniog of the ioversion period under consideration. This is the temperature I m above the ground at 19hOO as can be seen from table L.2. If it is assumed that cp = 10 06.5 JlkgK and p = 1.2 kg/m' k can be determined as a function ofqRJ from equation (L.17). Figure L.14 shows the conductivity, k, as a function of the qz=<>, the heat flux at ground level. The thermal eddy d iffusivity is equal to the effective thermal conductivity divided by the product of the density and specifi c heat The thermal eddy diffusivity is thus not constant as Anfossi et al. [76ANI} and Surridge [86SUl} assumed. L.l9 100908070 qFiJ 1~V 91V 80 -' ./ V V ~/ ---7/V ~ 70~ -" V/17/'/ L..--'I-~V/v j....--"'" ~~ /.V/V ~L..-- 50 ~~~ t..-~~t:--,.,- 40 ~'?-~ AA20 --... 200 400 ~ 800 ~ 2000 1800 if E 1600 i b 1400 '; 'fl 1200 ::J ... c8 1000 600! u ~ o o 10 20 30 40 50 60 Height, m Figure L.14: Atmospheric conductivity for different values ofq,.o. I ",- t-- .... _ - Huser et al. [97HU1] ". ... I " -Present model I \ . I \, qFiJ = 13.4 W/m2 rI I -\ I ..

I \ \ -1--------- .._-------------~---._-------._- ----------I "...I I " .... ........ ~ ----0.8 0.7 .!!! 1: 0.6 ]) .~ 0.5 1Il ::I !l: 'g 0.4 ? 'g 'g Gl 0.3g .. CIJ 0.2 .c I0.1 o o 30 60 Height, m 90 120 150 Figure L.15: Comparison between present model and the results ofRuser et al. [97 HUl]. Huser et al. [97HUl] employed the similarity theory of Monin and Obukhov [54MOl] in conjunction with a commercially available computational fluid dynamics software to fmd the v ertical distribution of the thermal eddy viscosity in the atmospheric boundary layer. The inlet profiles of the computational L.20 domain, which extends 5000 m in the downwind direction, is approximated by Huser et al. [97HUI]. These inlet profiles include the velocity, temperature, and the k and e turbulen ce parameters of the k-e turbulence model of Launder and Spalding [74LAI]. Figure L.15 shows the thermal eddy viscosity profiles of the present model and the results of Huser et al. [97HUI] for a smoo th surface where the surface heat flux, q,..o = 13.4 W/m2? The surface roughness, scaling velocity, s caling temperature, Obukhov length, the height at which velocity shear vanishes and the wind speed a t 10 m are 0.01 In, 0.23 mis, 0.029 K, 130 m, 196 m and 4 mls respectively in the calculations ofHuser et al. [97HUI]. The present model is developed from temperature measurements in a relatively ari d area near Lephaiale (Ellisras) (23?40'S, 27?47'E) while the model of Huser et al. [97HUI] is applied to Drarnmen in Norway (59?42'N, IOOI2'E). It is therefore not very meaningful to critically compare th

e results, but it is nevertheless interesting to note the respective trends. It can be seen in figure L.15 that the maximum thermal diffusivities of the present model and the model of Huser et al. [97HUI] occur at approximately the same height. Huser et al. [97HUI] also states that their model over-predicts the thermal eddy diffusivity. Therefore, the results of Huser et al. [97HUI] also concludes that the thermal eddy diffusivity is not constant as Anfossi et al. [76ANI] and Surridge [86SUI] assumed. It shoul d also be noted that the temperature distribution as given by equation (L.12) is a good approximation at lower elevation but becomes less meaningful near the top of the inversion. This means that the effec tive conductivity given by equation (L.I 7) is correspondingly less accurate near the top of the inversi on. L.IO INVERSION HEIGHT CALCULATION As mentioned above, the stable boundary layer is generally accepted to evolve as the square root of time as shown in equation (L.3) [86SUl]. However, faster development of the stable bo undary layer has been observed [87SUl]. Values of the thermal eddy diffusivity, K, in equation (L.3) m ay differ measurably depending on site and ambient conditions. Values of K = 0.3 to 0.5 m2/s have bee n observed. K is also assumed to be constant in the analyses ofAnfossi et al. [76ANI] and Surridge [86 SUI]. Due to the uncertainties associated with equation (L.3) to determine the inversi on height, the following procedure can also be employed to determine the inversion height: The DALR temperature distribution immediately prior to the development ofan inve rsion is given by T = Tm - 0.00975z (L.21) where Tmax is the diurnal surface temperature maximum. Experimental evidence indicates that the mean temporal temperature gradient at e levations above the top of the inversion is approximately 0.43 KIh [90SUI]. Once the stable boundary lay er or inversion begins to evolve, the temperature distribution above the inversion can thus be approxim ated by T = Tm~ - 0.00975z - 0.43t where t is the time measured in hours. (L.22) At the top of the inversion, Zit, the temperature as given by equation (L.21) is equal to the temperature given by equation (L.22) Le. { )'Zt(T, +273.15 -:: =T_ -0.00975z" -0.431 L.21 (L.23) For a given initial temperature, T""", the value of Zit can be determined by an iterative procedure for different groWld temperatures at different times. The procedure presented above is still relatively complex, as it requires the ma ximum daily temperature as well as the time since the inception of the inversion. The aim is to develop a theoretical model with the minimum amount of input. Equation (L.5) correlates measured inversion temperature data well for heights o f about 1 m above

groWld elevation and greater heights. Since the temperature gradient of this equ ation can never be equal to zero it is inadequate at the top of the inversion layer. In this region a mor e realistic empirical equation for the temperature distribution would be T - (T, + 273.1S{:'r-0.00975z (L.24) This equation effectively represents the region of transition from the inversion layer to the adiabatic lapse above it. Although this equation could also have been applied at lower elevation s in the inversion layer, the simpler equation (L.5) is adequate and makes possible the solution ofproblem s that would otherwise be less amenable to analysis. An approximate average value for the inversion height can be obtained by differe ntiating equation (L.24) and equating it to zero, i.e. { }'-'dT =b(T, +273.lS!lL -0.00975-0dz zr Rearrange equation (L.25) to find 1 [ 0.00975 ]'-'zit - b(T, +273.15) (L.25) (L.26) Equation (L.26) will predict, in conjWlction with equation (L.9), inversion heig hts of approximately 300 m in the winter months and approximately 90 m in high summer. These heights are typically observed in nature. The inversion heights predicted by equation (L.26) will remain constant for a particular inversion period. In reality, the inversion will grow throughout the night. The growth rat e of the inversion is relatively slow after the first few hours after the inception of the inversion. Equation (L.26) therefore predicts an average value of the inversion height but is not valid during the fi rst few hours after the inception ofthe inversion. After a certain time, I, the enthalpy ofthe air up to the inversion height ofzl/ ' fur a 1 m2 colwnn of air, is ~~ ZIt fpcpTdz=pcp f(C,,+C, +273.IS)z'dz o 0 The heat flux, q,=o, into this 1m' of air is L.22 (L.27) (L.28) Zit is approximately 300 m from equation (L.26) if b ~ 0.01. From equation (L.28 ), with c\ = -2.163 X 10-4 Kls, p = 1.1 kglm' and cp = 1000 J/kgK find from equation (1..28) that qz=O ~ 74.8 W/m'. This is not unreasonable. By employing equation (L.24) in equation (L.B) a more realistic value for the ef fective thermal conductivity will be obtained near the top of the inversion. Figure L.16 shows t he effective thermal conductivity predicted by employing equation (L.24) in the analysis. Compare fig ure L.16 to figure L.14 and see that the effective thermal conductivity is less in figure L.16 than in f igure L.14. q~ ....

~~ ~~.......-~ ~/ V - 70_ ~::::-:- .~ 60~ --~~% - 50::;...-'" -~ -:- . 4020 30 600 400 2000 1800 ~ 1600 i a. 1400 .;; g1200 i!8 1000 ;; E BOO ~ ~ ~ 200 o o 10 20 30 40 50 60 Height, m 70 BO 90 100 Figure L.16: Atmospheric conductivity for different values ofq,"". L.ll CONCLUSION Further investigation is necessary to validate the results obtained in this appe ndix, as limited experimental data was available for analysis. It is anticipated that the valne of the exponen t b in equation (L.5), given by equation (L.9), will differ depending on the geographical location of the pla ce of interest. Nevertheless, a simple empirical relation is obtained that predicts the vertical temperature profiles during stable conditions relatively accurately and that can be analytically integrated. M.l APPENDIXM ATMOSPHERIC HUMIDITY M.I1NTRODUCTION To accurately predict the performance of wet cooling towers, humidity profiles i n the atmosphere have to be known. The atmospheric humidity influences the heat aod mass traosfer as well as the draft through the towers. It is found that hwnidity inversions occur during the same period th at temperature inversions occur. The formulas that are employed in the aoalysis of cooling tower perforrna oce, to determine the effect ofatmospheric humidity on tower draft, are presented in section E.2. M.2 MEASUREMENT OF HUMIDITY Most of the literature devoted to the measurement of the vertical hwnidity profi les in the atmosphere, focuses on the lower atmosphere. The lower atmosphere extends to the top of the stratosphere, which is approximately 50 kIn above the surface of the earth [98SEI]. Meteorological weat her balloons (radiosondes) or remote sensing techniques are generally employed to obtain the humidity profiles [96HAl]. There are a variety of remote sensing techniques, which employ ground-b

ased instruments aod satellites. Empirical models are developed that predict hwnidity profiles as a f unction of the altitude aod the humidity at ground level. Some of these models are according to Hann [08HAI] , Yoshino [75YOI] aod Gorchakov [8IGOl]. Only the humidity profile in the lower parts of the plaoetary boundary layer is of importaocc in the performaoce aoalysis of cooling towers. The planetary boundary layer is approxim ately the first I kIn to 2 kIn of the atmosphere. The empirical models of Hann [08HAI, Yoshino [75YOl] ao d Gorchakov [8IGOI] are unable to predict the diurnal variations of atmospheric humidity in the plaoetary boundary layer, but are relatively accurate above the plaoetary boundary layer, where diu rnal chaoges are small according to McGee [72MCI]. M.3 DIURNAL VARIATION OF ATMOSPHERIC HUMIDITY After sunrise, water vapor is added to the lower atmosphere by evapotraospiratio n. This causes a sharp increase in the humidity of the air. The resulting lapse condition becomes most pronounced at the time of maximum surface heating due to convective mixing aod subsequent dilution of the vapor concentration. By late afternoon, however, convection waoes as the air near the ground becomes stable. Evaporation continues to supply water vapor to the air above the surface, but the rate of di lution due to mixing slows down aod the lapse mte tends toward isothermal. At night, radiative cooling of t he air below the dew? point tempemture causes dew to form on the ground. The extmction of water vapor from the overlying air causes ao inversion to form in the water vapor profile. The depth aod strength o f this inversion is determined by the downward flux of water vapor in a suitably turbulent environme nt The level of turbulence is critical. If it is too low, dew ceases to form since the ground ca nnot be replenished by water M.2 vapor from above. If it is too high, mixing inhibits surface radiative cooling t o below dew-point temperature. Near the surface in the early afternoon, even with strong evapotran spiration, turbulence transfers moisture away from the surface so rapidly that specific humidity usual ly falls to an early? afternoon minimum. The vertically-transported water vapor then produces a maximu m at the same time in the upper boundary layer [88PRI]. Moistening of the atmosphere by evaporation fr om the underlying surface proceeds quickly and invisibly every day when energy and water are avail able. The moistening process goes forward at daily rates up to 3-4 kg/m2? The surface is the source o f the water vapor that is mixed through the earth's atmosphere [77MB]. Vapor moving upward from the source at the earth's surface would, in time, diffuse evenly through the entire atmosphere but it is i ntermittently removed by condensation in the middle levels ofthe atmosphere. The interplay between the su rface as sole source of vapor and the vapor sinks represented by ascending columns or sheets ofair in th

e atmosphere produces a more or less equilibrium balance that is represented by a vertical decrease in v apor concentration [77MII]. Case study by Oke (780Kl) Refer to figure M.l. By day, the profile of vapor concentration lapses with heig ht away from the surface moisture source in the same manner as the temperature profile. Vapor is transpor ted upwards by eddy diffusion in a process analogous to that fur sensible heal In the morning hours the evapotranspiration of surface water (dew, soil water and plant water) into a moderately unstable atmos phere adds moisture to the lower layers and the humidity increases quite sharply. By early afternoon, a lthough vapor flux is at a peak, the humidity concentration drops slightly. This is a result of convective activity having penetrated to such heights in the boundary layer that the vapor concentration becomes dilut ed by mixture with descending masses of drier air from above. In the late afternoon surface cooling is strong and the lower layers becomes stable. Thus, the ability to transport vapor to higher layers is less than the rate at which it continues to be added from the surface. Moisture converges into the lowest layer s and a second humidity maximum is observed. Thereafter evapotranspiration declines into the night perio d. Under certain conditions the vapor profile may become inverted near the surface so that the va por is transferred downwards as a dewfall. This depletes the moisture in the lowest layers and humi dity decrease, until after sunrise when the cycle recommences. Oke [780KI] also found that the humidity pro file above non? vegetated surfaces, characterized by low absolute moisture content, is far from saturation and has a very weak lapse rate in most occasions. Case studies by Geiger (6SGEl) Figure M.2 shows average water vapor pressure profiles in the lowest 100 m of th e atmosphere for 19 clear summer days measured in Rye, England. Before sunrise there is a flow of wa ter vapor from a height of about 40 m toward the ground, to form dew. Vigorous evaporation is initiated by heating after sunrise, as may be seen from the increase in water vapor pressure in the layer nearest th e ground until 06hOO. Since eddy diffusion is still restricted, this supply ofwater vapor is trapped n ear the ground, and the daily maximum of vapor pressure is reached about 08hOO with powerful humidity gradient s (full line). Without any marked change of the gradient, the profiles (broken) then become displaced t oward the region of 11 M.3 lower vapor pressure, because of increasing eddy diffusion. This transport out o f the layer close to the ground brings ahout a minimum value at 14hOO. At 18hOO the decrease of water vap or pressure with height is still normal, but by 20hOO the decreasing eddy diffusion and increasin g water vapor content of the air have reestablished the humidity inversion at the ground, and as time goe

s on it gradually increases in height, in a similar way to the temperature inversion. U.l c:t :J III III U.l O!. 10IX "'"'0 IX'- 9 :JX0? Q,. -< > 7L.-.---.L_-'--_I..--....L._-L----J o 4 8 12 16 20 24 TIME (h) Figure M.I: Diurnal variation of vapor pressure at 3 heights at Quickborn, Germa ny on cloudless days in May [780KI). 4Or---IE-+:.... -,... . .; " ' ""'" f"? I20?- '-' .-..... "";;;---1~ '~ "~'" ! ........ I I. I~ ~ I '\'" ,../ I....... I II 09l-~~:::::L1 __....,.-1O~~~.:~_...~t::??...L_..)~~~...: ....:::?:::....~L__.... ,.12~--=::::::",... Water-vapor pressure (mm-Hg) Figure M.2: Profiles of water vapor stratification in the lowest 100 m on clear swnmer days measured in Rye, England [65GEI). Figure M.3 shows the data of the water vapor pressure in the first seventy meter s of the atmosphere measured at Quickborn, Germany. Here, too, the flow of water vapor is directed d ownward at night toward the ground surface (covered with dew), and the fall is remarkably large i n the hours before M.4 sunrise. During the forenoon, vapor pressure rises close to the ground, and fall s at the 70 m level. The transition from day to night conditions can be recognized easily in the average values for the period 20hOO to 22hOO. --+-22hOlHl2hOO / c____ 02hOD-06hOO -..... 06hOD-OShOO /~OShOO-14hOO --e-14hOD-20hOO /-h- 20hOD-22hOO I \ 1\ h '"'l .~ ~V "F:--::~:a. 70 60 50 E 40 :i .!l' ! 30 20

10 o 10 10.5 11 11.5 12 12.5 13 13.5 Vapor pressure, hPa Figure M.3: Water vapor pressure profiles of the first 70 m of the atmosphere, m easured at Quickborn, Germany. Db " 12* TIme re* Figure M.4: Diurnal variation ofwater vapor at Quickborn, Germany on clear July days. M.5 The daily pattern is shown in figure M.4 for heights of 2, 13 and 70m. The well known double wave of vapor pressure is easily recognized. This appears at all levels, but the amplitu de of fluctuation increases with approach to the ground. In all these layers, the evening maximum is higher than the morning value. Figure M.4 is, as we might expect, dependant on the season of the year, the weat her situation, and the geographic location ofthe place of observation. If the average of all days is ta ken, the picture obtained of the layered structure of water vapor distribution is similar, but a rather defla ted version of the picture ofa clear day. If conditions in the 100 m layer show a clear cut relation, the same cannot be s aid for the first few meters close to the ground. In the majority of cases, there is a repetition, on a small scale, of what we found for the first 100 m. This is true particularly in regions ofdry climate, rich in rad iation. M.4 ANNUAL VARIATION OF ATMOSPHERIC HUMIDITY Figure M.5 depicts the average annual variation of the relative humidity measure d at three different weather stations on the South African Highveld at 08hOO and 14hOO. The data in f igure M.5 for Germiston were obtained during a period from 1932 to 1950, for Pretoria during a period from 1938 to 1950 and for Pietersburg from 1904 to 1950 [58MBI]. Figure M.6 shows the corresp onding minimum and maximum monthly temperatures. -+-Pretoria 08hOO --,!r-Germiston 14hOO ---+-Pretoria 14hOO ......Germiston 08hOO ___ Pietersburg 08hOO -a-Pietersburg 14hOO 90 80 ';II. 70 i "E 60 ".c!! 50 '"..&! 40 30 20 -.. '"- t--- ~~0 ~~,.. ~ !r--Ii~~ ./~

"""'I lib.. ~V~ 1 2 3 4 5 6 7 8 9 10 11 12 Month of the year Figure M.5: Average annual variation of relative humidity at 8hOO and 14hOO meas ured at Pretoria, Germiston and Pietersburg. -+-Pretoria Low -0- Germiston High M.6 ~Pretoria High ........Germiston Low __Pietersburg Low -s-Pietersburg High 40 35 30 ~ 25 ~ 20 fX. 15 E ~ 10 5 o -5 :;::::. - S----.:.::;"- ~ ---'~V'":I-. ~ ~ ./. =I~~ 1 2 3 4 5 6 7 8 9 10 11 12 Month of the year Figure M.6; Average monthly variation of minimum and maximum temperatures measur ed at Pretoria, Germiston and Pietersburg. M.5 CONCLUSION Humidity inversions occur in a similar way as temperature inversions during the night. Oke [780KI] found that the vertical humidity profile is approximately constant above non-veg etated surfaces during the sunshine hours of the day. Geiger [65GEI] found that the water vapor flows to th e ground, from approximately 40m above the ground, to form dew during the night From these conc lusions and the general trends of the humidity profiles shown in figures M.l to M.4, the effect of humidity profiles on cooling tower performance can be determined. ~'i' N.] APPENDIXN MODELLING OF A COOLING TOWER AS A CIRCULAR JET AND A POINT SINK N.t INTRODUCTION In this study, the height at which a cooling tower draws in air from the atmosph ere is detennined for different mass flow rates through the cooling tower. The cooling tower is modell ed as a turbulent circular jet and a point sink. The inlet of the cooling tower is modelled as point sink w hile the exit is modelled as a turbulent circular jet. A solution of the flow field for the turbulent circula r jet is obtained with an analytical method derived from boundary layer theory. The computational domain i n tw(}-dimensional

cylindrical coordinates, at which a solution is sought, extends 3000 m in the ax ial (vertical) direction and 5000 m in the radial direction. N.2 TURBULENT CIRCULAR JET The investigation of the flow field far away from a jet is detennined in this se ction by an analytical solution. There are presently three techniques used to study external flows Le. numerical (digital? computer) solutions, experimentation, and boundary layer theory [94WHl]. Prandtl first formulated the boundary layer theory by making certain order-of-magnitude assumptions to greatl y simplify the Navier? Stokes equations into the boundary layer equations. White [91WHI) gives a summar y of strong inequalities at large Reynolds numbers: o?x v?u au au -?-Ox iJy Ov Ov -?Ox iJy (N.l) (N.2) (N.3) where 0 is the thickness of the boundary layer. These strong inequalities are us ed to simplify the Navier? Stokes equations to obtain the boundary layer equations. Jet velocity profiles are unstable and undergo transition to turbulence early, a t a Reynolds number of about 30, based on exit slot width and mean slot velocity [9IWHl). This is the r eason why only the turbulent circular jet is investigated in this study and not the laminar turbule nt jet. Problems in free turbulent flow are of a boundary layer nature. Schlichting [60SC1) gives the ana lytical solution for the boundary layer flow of a turbulent jet. The boundary layer equations for a tw(}dimensional incompressible flow are: au au au ] aT -+u-+v-=-of ax By pBy 8u+Bv=O Ox By where ris the turbulent shear stress. Prandtl's mixing length theory states that : (NA) N.2 21E"J auf, : pL a.Y1 By where L is the turbulent mixing length. This equation for shear stress in turbul ent flow is unsatisfactory in that the awarent, kinematic velocity 6, vanishes at points of maximum velocity. A hypothesis by Prandtl circumvents this problem by defining a virtual kinematic viscosity which is form ed by multiPlying the maximwn difference in the time-mean flow velocity with a length which is asswned to be proportional to the width, b, of the mixing zone. Thus (N.5) where X donates a dimensional nwnber to be determined experimentally. The width ofa circular turbulent jet is proportional to x such that the centreline velocity U oc X-I . Thus, the

virtual kinematic viscosity becomes: &, : zbU oc X O =const: &0 (N.6) (N.7) which means that it remains constant over the whole ofthe jet. The differential equations for the velocity distribution are identical to the laminar circular jet, with the only difference in the viscosity, where &0 replaces v in the laminar equations. J is the kinematic momentwn flux, which is a measure of the strength ofthe jet. The velocity distribution ofa turbulent jet is 3 J I u: 8,.. cox ( I ')'1+-1/ 4 where '1=~ 3 ..{,j 2:'. 1M 8 0 x According to measurement, the virtual kinematic viscosity is &0 : O.OI61..{,j The flow rate ofthe jet can be expressed as: V =OAI4..[.fx N.3 POINT SINK The equation of continuity for incompressible flow in polar coordinates is: ~~'V, sinO)+~(rv. sinO): 0 ar ao There exists a stream function such that I arp V = , r' sinO 80 (N.8) (N.9) (N.lO) (N.lI) (N.12) (N.13) I otp v =--? rsinO or N.3 (N.14) (N.15) The tangential velocity, v9' is zero for a point sink. Integrating equation (N.t 4) gives the radial velocity: rdtp = -(v,r' sinlliO rp =v,r' cosO where: v V v =-=-, A 4117" N.4 EXAMPLE CALCULATION AND RESULTS (N.t6) (N.17) The outlet of the cooling tower is modelled as a circular jet. The mass flow rat e through an arbitrarily chosen cooling tower is assumed to be 15213.68 kgls, while the outlet diameter o f the cooling tower is 58 m. The density of the air, p, is 1.204 kglm' and the dynamic viscosity, f.l, is !.8 x 10-' kg/ms. The momentum of the jet is given by

J=mv where m=pAv A= 1r d' = lr(58')=2642.08m' 4 4 From equation (N.19) find v =~ = 15213.68 4.782576 mls pA (1.204)(2642.08) From equation (N.18), the momentum flux is J = mv = (15213.68)(4.782576) = 72760.6 kg mis' (N.18) (N.19) (N.20) (N.2I) The exact solution for the flow field of a turbulent jet, approximated by the bo undary layer equations, is given by equation (N.7) to (N.t 0). The results for the turbulent circular jet c an be seen in the first column of table N.!. The streamlines, axial velocity distribution, radial velocity dist ribution, and the velocity magnitude distribution are illustrated in table N.!. The range of the contour le vels for the axial and velocity magnitude distributions is from 0 to I m/s. The pitch black contour lev el represents velocities greater than or equal to I m/s. The negative sign in the table for the radial ve locity distribution indicates a flow in the negative radial direction. To solve equation (N.16) and equation (N.17) for the point sink, the volume flow rate, V, must first be determined. v = m =15213.68 -12625.5 m'ls 2 p 1.205 To maintain the desired flow rate, the strength of the sink must be doubled. For a sink, the volume flux, V, is negative. The results for the point sink in cylindrical coordinates are illus trated in the second column of NA Axial velocity Stream function II (jill' Radial velocityVelocity magnitude II ill_ ~ :z f------+------r------+-----+-I? ~ ~ an 'r ., ~ ~ ~. ~ 1 - -~ ~ 8? .g ~ ~. f-------f-------i--------t------H g' a?go

~ f "'J;'" ,,,. 111M N.5 table N.!. Once again the stream function, axial velocity distribution, radial v elocity distribution and velocity magnitude distribution are ptesented. To combine the results of the turbulent circular jet and the point sink, the pri nciple of superposition is used. Superposition, however, can only be used in potential flows, i.e. when the flow is irrotational and incompressible. Due to the fact that viscous flow is present in the turbulent je t, superposition can not be applied in the region of high velocity gradients. If it is assumed that the flow is laminar and irrotational in the farfield, which is a very reasonable approximation, superposition can be app lied. As the farfield characteristics are required in this study, the assumption of su perposition is sufficient. The last column in table N.I illustrates the results for the superposition of the tu rbulent jet and the point sink. It can be seen that the influence ofthe point sink is almost negligible. Different radial velocity distributions with height are illustrated in figure N. !. Six different distributions are presented. Radial distances from the cooling tower, modelled as a circular j et, of 3000,4000 and 5000 m are considered. At each distance, the effect of the jet and the combined effec t of the jet and sink are presented. 0.00650.0060.00550.0050.00450.004 ~ ~ ~ ? ? ~ ?S - Sink? ? ? ? , ? ? ? ?, ? ? ? ? ?, ? ? ? ? ? ? ?, , ? ? ? ? , ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?, , ? ? ?, ? ? ? ? ?o 0.0035 200 800 1- -~ -3OOOm J __ 3OOOm J+$ - -.. ? 4boomJ _4OOOm J+S - -. -5000m J __5OOOm J+S

I 1000 Radial velocity magnitude, m/s Figure N.l: Radial velocity distribution at different radial distances from the cooling tower. As expected, the velocity magnitude ofthe combined effect ofthe jet and sink is greater that that ofthe jet alone. This is because the sink extracts flow from the flow field and thus incre ases the velocity of the air, which is drawn towards the centre axis of the jet, due to the effect of the turb ulent jet. It can be seen that the velocity remains constant with altitude for the different cases at the vario us radial distances from the cooling tower. N.6 The determination of the height, at which the tower draws in air from the surrou ndings, is presented next. Refer to figure N.2 fur the illustration of the variables used in the calculatio n. The mass flow rate in the atmosphere towards both the jet and the sink can be expressed as mr ;::: pArvr where A, =P,H, and P, is the perimeter. Thus where P, = 2nr. Find H = 15213.68 , (J.204 )21l7V, (N.22) (N.23) (N.24) H, I-.. -.. Vr, mr r v,J,m Figure N.2: lllustration of variables The value ofH, for the example calculation is contained in table N.2 for differe nt air flow rates at various radial distances from the cooling tower. It can be seen that the height, H" is a pproximately 123 m for the analysis with the jet and sink and 127 m for the analysis with the only the jet. The influence of the mass flow rate on the height, H" is investigated. Table N.2 and table N.3 contain the results where mass flow rates of respectively 10000 kg/s and 5000 kg/s are used in the calculations. Alt hough the velocity of the air drawn towards the cooling tower is different for each mass flow rate, the he ight, H" remains unchanged for each case at the different radial distances from the cooling tower . This is consistent with the numerical results ofThiart [02TH1]. Table N.2: Altitude, H" at different radial distances for two different flow cas es with m = 15213.68 kg/s and d= 58 m. m = 15213.68 kg/s, d- 58 m, v = 4.783 mis, J= 72760.6 k~ mis' r,m Flow Case v, m/s Hr,m 3000m Jet 0.0052331 127.032 3000m Jet + Sink 0.0054528 121.915 4000m Jet 0.0039248 127.031 4000m Jet + Sink 0.0040484 123.154

5000 m Jet 0.0031931 127.031 5000m Jet + Sink 0.0032749 123.858 N.? Table N.3: Altitude, H" at different radial es with m ~ 10000 kg/s and d=58m. m - 10000 kg/s, d- 58 m, v = 3.143603 mis, J =31436.03 kg m/s2 r,m Flow Case v, m/s H,.,m 3000 m Jet 0.00344 127.032 3000m Jet + Sink 0.003584 121.915 4000m Jet 0.00258 127.031 4000m Jet + Sink 0.002661 123.154 5000m Jet 0.002099 127.031 5000m Jet + Sink 0.002153 123.858 Table N.4: Altitude, H" at different radial es with m ~ 5000 kg/s and d = 58m. m - 5000 kg/s, d - 58 m, v = 1.57180 mis, J =7859.oI kJ/: mis' r,m Flow Case v, m/s H"m 3000m Jet 0.00172 127.032 3000 m Jet + Sink 0.001792 121.915 4000m Jet 0.00129 127.031 4000m Jet + Sink 0.001331 123.154 5000m Jet 0.001049 127.031 5000m Jet + Sink 0.001076 123.858 Table N.5: Altitude, H" at different radial es with m = 15213.68 kg/s and d= 100 m. m - 15213.68 kg/s, d - 100 m, v = 1.608859 mis, J= 24476.66 k1!:m/s' r,m Flow Case v,m/s H,.,m 3000m Jet 0.003035 219.0209 3000m Jet + Sink 0.003255 204.2403 4000m Jet 0.002276 219.0195 4000m Jet + Sink 0.0024 207.742 5000 m Jet 0.001852 219.0189 5000m Jet + Sink 0.001934 209.7544 Table N.6: Altitude, H" at different radial es with m ~ 10000 kgls and d= 100m. m -10000 kg/s, d- 100 m, v = 1.057508 mis, J = 10575.08 kg mis' r,m Flow Case v, m/s H"m 3000m Jet 0.001995 219.0209 3000 m Jet + Sink 0.002139 204.2403 4000m Jet 0.001496 219.0195 4000m Jet + Sink 0.001578 207.742 5000 m Jet 0.001217 219.0189 5000m Jet + Sink 0.001271 209.7544 N.8 Table N.7: Altitude, H" at different radial es with m = 5000 kg/s and d = 100 m. m 5000 kg/s, d - 100 m, v = 0.5287539 mis, J= 2643.770 kg mis' r,m Flow Case v, m/s Hr.m 3000m Jet 0.000998 219.0209 3000 m Jet + Sink 0.00107 204.2403

distances for two different flow cas

distances for two different flow cas

distances for two different flow cas

distances fur two different flow cas

distances for two different flow cas

4000m Jet 0.000748 219.0195 4000m Jet + Sink 0.000789 207.742 5000 m Jet 0.000609 219.0189 5000m Jet + Sink 0.000636 209.7544 Table N.5, N.6 and N.7 contain the results where the outlet diameter is changed to 100 m with flow rates of 15213.68, 10000 and 5000 kgls respectively. It can be seen that the values of H, are the same for the different mass flow rates, but is substantially larger than the case where the d iameter was 58 m. 1101009060706050 I IH:-Jet I ... '" - - Jet and sink I ...... .................... ",'/~'" ..... V " ............... ",.:::--- ...... ....-V ..y .yY ".., 75 50 30 100 250 12S E ,[150 175 200 225 Tower outlet diameter, m Figure N.3: Height at which tower draws in air from the surroundings versus the outlet diameter ofthe tower. Figure N.3 shows the height at which air is drawn in from the surroundings as th e outlet diameter ofthe tower changes. It can be seen that there is a linear relationship between the ou tlet diameter of the tower and the height, H~ N.S CONCLUSION It can be seen from the tables N.2 to N.7 that the height, H" at different mass flow rates, remains the same for a given outlet diameter of the cooling tower. It is shown in figure N.3 that as the diameter of the tower increases, the height, H" increases. Thus, the height from which air is drawn in to the cooling tower is only a function of the diameter of the tower and not of the air mass flow rate t hrough the tower. The influence of the height of the tower on the height from which air is drawn into the cooling tower can not be determined by analytical approaches but only by numerical analysis. It is the refore recommended that these results and the effect ofthe height of the tower be validated by numerical analysis. 0.1 APPENDIX 0 A CRITICAL COOLING TOWER PERFORMANCE EVALUATION The performance of natural and mechanical draft cooling towers is critically eva

luated in chapter 4 by employing the Merkel, Poppe and e-NTU methods of analysis at different operating and ambient conditions respectively. The figures pertaining to the discussion in chapter 4 a re presented in this appendix. iii' _",,!; 0.2 , , ......... " -, , " ~~ ...... , 1'----.:.' -'" Natural dnUt. r-----.. ............... r a l=290K r--.......' , -MmIrel - - Poppe ""0.000 0,002 0.004 0.008 O,lXlIl 0.0,0 0,012 IUI'I4 0-016 "'''~ "00 ,oro "00 , , , , ....... '" , , "', ~~~ ... ......... , '" , ,........ , '" N__ ,, r-----.. , , I- T'Ji""280K , - """" "-Poppe "'" "" "" "00 "00 0.000 0.00'\ 0.002 0.0::0 0.0001 0,00; o,DOS (HXJ7 0.009

WI,kJIkI Natura1 draft T a l"'290K V -Merkel -- Poppe / 17 ./ :"'",., 0.000 0.002 IMl04 0,006 1I.1XIlI 0.010 0.012 0.014 0.018 IO'hkWkI '" , Natund draft , Ta l=280K V - """", -- Poppe V -'", ./ ." V, 0/ , , ""Il.OOO O,an 1J.1Xl2 0.003 0.0001 1l,0lXI 0.006 0,007 0.00!I "'" kPI .""i ''"' ""''''~-~---r-~--'--r--.,--.-' J~"""L..m, j--+---1~-+---+---1---;>t'/S..-j'=--j ./ ,:;: m."j--+-+--+--I""<~""=-I---+--~ _-1- /' 'ild /~ V ~ 2Z.51--i--_~_tr;__,r/"~-P..-''l--t-;:?N:'_='_:'d?",,==:;-i ....... ' V ",,'" ,.'" f---ce1,'''''/7-F--i--+----1~-1 Tal'" 280 K :::- V _ """" ""1-'==--1----1--./----/---1--1 - _ Poppe ",,,LLJ.-L-LLJ:~ IHIOO 0,001 o,(Xl:2 o.cm 0.0001 O.lXl5 O,OOIJ 0.007 0.008 wI, kI"k& "" N__ </;, T.l=290K ,-Merkel -,I~V- - Poppe -//-' ----", /'V '"0.000 0.002 0.0001 0.009 O.lXlll 0.010 0.012 0,01. o.01e "'ltkt'q: Natural draft T"1~290K / ,- i-- Morl<.J -

- / - - Poppe ..- 7 ----./ V "--V ./ V "." "..IllXXI 0.002 0.004 o.(XIlI O.OO!I 0,010 0.012 0.014 O,Ole ",.kJ/kI "'''' j ::i 18ZXl , , (14) NlI1ura1 draft " T.I~280K "" -Mo<kol .../, -- Poppe ,/' -- / -- / ..-- /' ..-/ /' ""'" ~'oo '"""0.000 0,001 0.002 0,003 O.1Xlol o.OlXI 0,006 0.007 0,009 "',,1<&*& ""'" ""'" """' ,., O,1lXI 0,002 0.004 o.ooe o.ooe 0.010 0,012 0.01' O.01e "I."""" m J--~f_.~,c_,_-1--I--f-+-+-+---1 -" .~ ....:: " ........,: ~~" ,

--.... " '" " -.... ~..._,N__ 1-..... "-T"I:290K \'-..... - Morl<.J - - Poppe (bs) Figure 0.1: Performance curves of a natural draft cooling tower. OJ , Natural daft " Ta l=300K~ "'"~''' - """" - - Poppe ~<, " ~". , ~ ". , '" " ~ ..' .... , , ~ " .... NlItunI1draft ~, " T a l=310K , -Merkel ....... -- Poppe "--''''' "" """ ."' 0.015 "'I, kI/kll ."" '00 ?."" 0.010 ... " kI/1ll '''''' 00," ."" ./V V 4'-V " Natural dtaft --;:-Ta l-3ooK - ""''''' -- Poppe (dz) ."'".""1l0I,O Nallltal draft /' T.I",310K ./ ----

-Merkel - - Poppe / ,/ ? -?" /' ij-", ---: " , " " '" '" '"'?."" ". ."" (cz) .."'".=?.=0,0150.010 .. .. .'"'i 1f'"ka/kI .-Y~ ,,-;:V -, , -' , ;-,.-/ " "" No__ / Ta l=300KMmk. / - Poppe .. ."" /",-, -,' ;/' ..... '--- /--" , ", , / Natudl draft / T.I=310K-Mmk" - - Popp~ "" """ ?."' 0.010 .., "'1,kW1<& .."" .= ?."'" '" '"

'" ? '".: .. 0.010 ?."" ..." .tp.a .... ."" -~---. ..----~_/ -- / /" V Natuml dnft ------Ta l=300K -Mmk_ .-/ -- Poppe ........ , , Natural draft " --- ...... T.I-3IOK " - Merkel '.. - - Poppe '" ...... , ---I----- " ,t--- '., , ~ i l1roJ ;, I! 1'~ "'''' """ ."" ?."' 0010 ,., "',.~ ."" ..'" ."'" i~ ; , ..." .." .."" 0,010 ."" .."'" .... (dt) ."""

>. _dnft " "~ Ta l-3OOK-Mmk"-,- Poppe ~'" ~ ., , ~ :::: , Natural draft '" ~ ....., T.l"'310K , -Merkel -----..:.:.: Poppe ". '.~~" ~, "'.. j 211110 ii W,O .... ,"00 """ .."" 0,010 ,., 1l'1tJcw'kc 00" .'" .."'" (cs) '" ? .."" 0.010 ."" ?."'" .... (ds) .."" Figure 0.1: Perfo11I1ance curves ofanatural draft cooling tower. .. .. ..." ...... , -'.... r--;' -', I~ ....... -~ '.T.. 1"'307.SK --.:R,-Mflkel - - Poppe ~ '<> .. .. MOO IHl05 0,010 0015 0.020 D.1Ji!5 D,mo 0.035 0.040 0.046 w.,kJ'kI; "" .. 0.4 "

NIIluni draft - T~1"305K f~ ""~ - ....... '. --..... --::'~, - - Poppe, ~"'~" -..::K', ~ .. O,OOIl 0.006 a.llIo 0.015 0.020 0,025 o.mD 0.036 0.040 "" Il'"kWkI (f2) NlltU.fa1draft --T.. l =301.5K V-MmbI - - Poppe ../ ,/ /. 1-7'--::- .I -_, " '"'0,000 O,IDS 0.010 0.015 0.020 o.ms 0.000 O,lI3!i O,IKI 11l14!i "'lIkWk1 (e2) Nalllnd draft I.T.. t=3OSK V- ...... -- Poppe -/V /. ~0- ""~~. -1).000 0,005 nOlO 0.016 0.020 o.Q25 cum a.ass 0.040 DO 0.000 000:1 0,010 0.015 0.020 o.C12S 0,030 0.035 0.0?1 .." ....... .. ""0,000 D.tal 0,010 0,015 o.m 0,025 0,100 0.035 0.040 0.045 IIIhklJ'<l "" "" -;/ -;/;./;--' --~ ---- . -- /' -/' Natund draft ..-/ T,,\=30SK - """"I /' -- Poppe ,," --:/V

.. ~_ .. 1.---;/ --- 1/....-. . -- V./ Natural draft ..-/ T"I'"307,SK -Morl<' 1-- -- Poppe w.,1IWIlI ""'1I---r-r-T1I1r:==:~ ...... Natural draft "'"'1--"'+',-,-1--1--+--+--+1 T.. I =307.SK " "'"1--1--1-0...+-+-1--1-1- ......Poppe j "'"+----\---1--1--':::-"-1--=-+---I-"==f=='i='-1; ---- (f4)i "'" I--f----I---I ---I--I--"-"'ct,-,-+--I--l "'"1--+--+-+-+-+--f-'-~-+-c:-+--1 -'-,=+-_-'-_-'-_-1-_-+_-+_--+_-1_-1_-' o.em O.llDll (l010 0.015 O,lI2ll 0.025 !lOX! 1l.035 0.0?1 0.045 -........ , Natw'II1draft --- T"l "'305 K --......... --- ....... -'--Poppe --. - f....----f-- f--' "'"0.000 0.006 0.010 0.015 0,020 0.025 0.000 0,035 0.0?1 "'hk~C ~ 0,000 O.M 0,010 0015 O,1llO O,1%l!i a03Q 11035 OJMO 0-0415 w,,-.: -..::' "~ '-. " ~-N"""''''' ~ --f- T,,1-305K "'"~f- - ....... '- - Poppe f' ., 0.000 llOCl5 0.010 00015 0.020 O.O2!l 0030 IHXli5 O,lMO w,. k&I'kI ", N.nwI draft - T" 1-307.5K ~ '-. (:~. , -M<ri?I ~ - - Poppe~... ~ ~, ~ '-, ~(f5)

Figure 0_1: Perfonnance curves ofa natural draft cooling tower. r I I 0.5 , " N...",""'" ---' ....~ T~l=30B.75K , -"""'" --..::: r--., --Poppe ..... :'::'" ~ "0 0 ~ ""0.000 0JXJ5 !lOlli o.01~ 0,020 0.0215 O.too O.OSlS 0.0?1 o.O'li!l o.aJO -,N...",,,,,,,, /'VT~ l ...lOS.75K V - """"Pop", V .' (.ij -./V,'.. . ? 0.000 O.OOll 0.010 0.015 O,tr2Il 0.025 0.000 o.m5 0.00 0.045 o.O!lO ,,, "'1,.Ilf'\llI: ,-"" --- l? .-.-- V...... '" -./ .-...NallIl'll! draft -- / T a l=]OB,7SK -M_ - - Poppe ./,. 0,000 IllXIIl O-DID 0.015 0.020 0.025 0.030 O.CI3l5 0.040 llOolll O,Cl&I "'1,1lWkK ,,, ?",. .: ., :r"1:,=C0oo =r=r=~~t---+ ~.. ~~ NllIur81draft T"I=30B.75K "'" +---t-+-f'-",,+--t-+-H _ Morl<eI !=::t=::t=::t=j='~?~t=t=~=2P~O=P~"'=~~: "'-,~ -'-..''''''EfEEEB~''3j"- '. "'" .." l-?-,-l:,-l--I---1-_-I--,.J-----1-_-I----J 0.000 0.005 0.010 o.Ol~ 0,020 o.OZI 0,000 O,Da5 0.0*;1 0.045 0.000 "'''~

.0 '0, N...",""'", T","'lOS,75K -.....: ........ -M""" - Poppe ~ "'" ~~, i'. '" ~ 0.000 0,005 0,010 0,015 O.OZI o.ll2!l 0.030 II.lI!l5 0.0?1 0.045 a,lEO Wh~ Figure 0.1: Perfonnance curves ofa natural draft cooling tower. 0.6 ? "? <' ". ~ , , , ", ? "r-:::.-' '" MIlciJanUl draft -----..: ..-...... , Tal -290K -Merkel r---::::, ? - - Poppe ,, 0,000 a.em 0.004 O.OOCI 0.006 0,010 0,012 0,01. 0.016 wltkA , -"-., , '---, -', "'" t--. '" , -, ~ -',, "Mechanical draft '-- -', , Tal =280K , -_d ~, - - Poppe , WI'~ ". (1000 000\ 0,002 0.003 O.lXM 0.0::6 0.00I:I 0.007 0.008 ". Cbz) Mechanical draft -/ Tal =290K V -M<ri<d - - Poppe ./V

/ ../ /" .. 0.000 0.002 0,004 0.006 OJXII 0,010 0.012 0.01. 0.018 w"kt/kI (az) , / V, Mechanical draft Ta1 - 280K 1/, -M<rl<d - - Poppe V , "/' V ./ 2111.0 0,000 0.00\ 0,002 0.003 0.004 C.OO6 0,008 0.007 O.lXMl w"kclkl ?, ,:: , , Mechanical draft . Ta l c 280K ./ ~ ..-, -M<ri<d '/ -- Poppe ~~~~ ..., ----. V , ~ ; ......... / . V ./ "" "" "'., o.1Xll 0.001 0.002 0.003 0.004 0.0::6 0,006 0.007 0,008 ,~" kfIklJ =. ? =. ,:: '" w"kWkg (bs) ""'-- ~ -"-- .. '" ,~ i'-...Mechanical drilft " Ta l=290K ",~ -M..kd - - Poppe " ,':---" , , "" ~-, 0

~ '" -......: ~""'" '" Meclulnical draft Ta L"'290K ~ -- -Merkel - - Poppe " C-o '" ,~ 0.000 0.002 a.l104 o.DOII 0.008 0,010 0.012 0,014 0.016 whkWkI ?0.000 0,002 0-004 0,008 0.008 0&10 0.012 0014 0.018 w" kglkc o i 6.6 i ?"? (l4) (as) Mechanical draft " Ta l=280K -f'---_d "'-. ~ - - Poppe........... t--. ", ~-. .... _- ~" -', ", -r'-'-. -'-----, --"'" , -"--, --"-----Mel;banical dtEt. "'""--Tal ""280K ---f- """" "'"-- Poppe , I

'" '" "0,000 0.001 0.002 0,00:3 0.004 0.00l 0,006 0.007 O,lll5 w"k&o'kc .., .. .. .. OJJXI 0,001 (lim 0.003 O.ClO4 0.005 0,005 0.007 c.oOll Figure 0.2: ?Perfonnance curves ofa mechanical draft cooling tower. 0.7 .... ."" 0,010 ~'" '"~,~ ~, ~" Mechanical cb"aft ~"Ta l=31OK -""""' "" "'-Poppe "-.. " " " ,,,, ."" .", .m',m. ."" ? , " , ~" ? -', ~ -, , ~'" Mectt.nicll draft r--::'-_, T~I=300K "'.' ", - """" "-- - Poppe -"'-':" ?.~ ".,lc&I"l& ..... I<f'k& I Mecbanical dlWl V Tal "'300K ./ -"""" /' - - Poppe ./' ,/ / ./' Mechanical draft / Ta l-310K ./ - """"

- - Poppe .,V ./ V ,/ .'"'I '" '"..~ b."" ,m. "'., ."" .", '" '"' '"' '",."" 0.010 ,,,, .", ."" ,... ."kW\<I .---':;7 ---- V , -- " ./' ---------- ./ /' / Mechmical draft 1/ Ta l=31OK -""""'/ - - Poppe "1, JcWIq: '"' '"' :/ '" _.~~ ---/ -----~~ ./ "" --? --- ./ "," ./~ ~, ~hanicaldnIft ./ Tal '" 300 K '"' - """" '" -' - - POppe '" -' ."" ""'" .m. .., .", ."" ,,,, w"kJ/kI ". .. 0.010 ,,,,, ?.", .... "'-- Mechanical draft ""-

Tal "'JOOK ~ - """" - - Poppe I--~ ..... ~- ~--, -- ~ ..._--~ ~ " 'McchJnicaidraft ""- Ta l=310K I""- - """"Poppe I~~--" ( -, " ""'" - ,~" --~ '" i'" i.", ,., ''''' 0,1)10 .m. ............. ?.", .."" '" 0.010 ,,,, ..,.J<cIk& .."" ?.", <4) " Mechanical draft f-":-' Ta ] = lOOK -, "~ - """" 'h.. - Poppe 1'" ~'-' ;;:. ..... .""0.010 ,~ -.. I~ " -.................... '>-'. Mechanical dIU'\: -'~ Tal=JIOK " -Morl<d - - Poppe I'-----..

" ."" " ..? ,,,, ''''' O.ll'l~ "'" kII'<a 0.010 ."""."" ,.? ..? .. Figure 0.2: Perfonnance curves ofa mechanical draft cooling tower. 0.8 "'" .. 0.000 o.c02 0.004 0008 o.oos 0.010 0.012 O,O1~ 0.016 .... kl"kl '~r--.---r-,-,--r-T-'----' " "" j-----""'+-,-j--+--I---+--t---f---j ...~~, ,=t-''"'-:J-f'------+~',-''+,,-+--+--f--\--j ",.t--+~""';:--r~<cd--t--+---t---j~ '~~~--------W:------~"":J'="':t, ~~11711 t-{ Naturald~ .... ~~, Tal'" 280K f---+-+-+'~~+--+~-1 -Merkel ~ "" 1:::!=-::i-~Po~P~P~?:!=!==t=1==t=J==1 "00llllOO 11001 0.002 0.003 0.004 O.lXl!I on 0,007 0-008 "1.~ "" "00 "'" , " -.... " " -....-.... " '" " ----- ' '" """'" ,.-.ft r-.....: " "T"I""290K , -Mmkol ------> , - Poppe Natural draft T"I-290K V -_d - - Poppe / ./V V p '" ""'.III*1l

"" '"a,OX! 0.002 0.004 o.oos O,OOll 0,010 0,012 o.Ol~ 0.018 NatJ.nI draft I.T,,/-280K -" V -Motd /'-- Poppe V f ,... ,<"" lid 29J,1l ; " ,." "'. "'., m.' m .? "" 0.00) 0,001 0.002 0,003 0.0001 COOS o.lXlEl 0.007 a.Oll'I w,.k-"& ,. Natural draft v:-T"l==290K , ..", - M?l<d ;/V-- Poppe ..-'~~ ;:/V ----; ./ '"0.000 0,!Xl2 o.QOoI 0.1XlII 0.00II O,oto 0.012 0.01. 0.01& ............ """'" ''''' -;;i-"T"l" 280K - ....... ~,~';V - - Poppe V V"..~V --/' _. O.lXXI 0,001 0.002 0,003 0.004 0.005 QOO6 0.007 0.008 "'It~ "".? "'.0 "" "'.? ""?, " ,... ,~ 0.0:0 0.001 0.002 Ol)l)3 0.004 0.005 o.OOlI 0.007 0,(108 ./ Natural draft / Ta l=290K .-- V-_d --, -- Poppe ~~'~- / --

- / --, ---- / .-/ ./ ".. "'00 ""00 "'000.000 0,002 0.004 0.006 0.0CIl 0.010 0,012 0.014 0.016 Wloklfkll: ,... j :i 15l1O:l , ? " """'" ''''' -1',,1 == 280K , ./' -M""" ---- Poppe , / /../- / -, /' -' V V ".. "." """j :10100, .? ? bs) " --:' , ~ ", , -----.: ....~~~~ -----'" I' '" Natural draft -.... I---- ~~~ ....Ta l=290K , - M<rl<d -----I- - - Poppe r-"" ~ 0,000 0.002 0.lXl4 000II 0.008 0.010 0.0\2 0.014 0.016 '" 11 kl!ka (as) -, , '" -, ~--- , ~ ,

--- '- , --- , ----I- NldW1II draft I'-----T"I=280K --------Motd i"-- -Poppe i'---- I-~j i 35Q ;~ "0,000 o.CKl'l 0,002 0.003 0,00( 0,005 0008 llOO7 o.OJIl ... " kl!kg Figure 0.3: Perfonnance curves of a natural draft cooling tower with a simplifie d draft equation. 0.9 " -----::: " " '-...:K""Natural dmft ~.T"1=300K -Morl<d -- Poppe ~ i::,. , " ", ~"'" " -----::.: ,'. ~~dnft '. "T"I"'310K ~, --'1 - - Poppe ~ ''''' .. "" ."" .."" 0,1110 0.015 ."" ,.. '.0 ."" .."" 0.'" ..., w"q\IJ: N....,dnft /'V T"I=300K /-"""'" -- Poppe d V ~ " Natural draft /' T"l=3IOK ./' V - """" - - PappI:! ./ ,/' # 8'/' /.,-

../' ,-, " ..?i h ?."" 0010 ??? "'lt~ .."'" ... 0.. '" ?i"" h '"' ."" ??? 0."'" ."" ... .rp; .-;;/V ~ ... ~~ -.. /--,--.V N....,dnftT.1 =300K -M.ml /' - - Poppe -'V ---",.. /...-,-' /' / ------- / NaAJraldraft ;/ 1.1"" 310K -Morl<d - - Poppe .""? .: .."" 0.010 0,015 .. ,....... ... '"" 0."" '" '" ?... .: "" ''''' 0,010 ."'" ."" .'" ",'" '- '" -" --". -- ----- /'~ ~ Nawraldrafl: ./'V T"1=3llOK """" ----Poppe """ ",m j

:; l1WJ i ? """ """ ""'" "m .."" 0,010 ,., 1I',,1lPll: ."'" 0"" .... >" "-~", ~.", NllturaI draft ~T"t-300K - M<rlreI -, - - Poppe ~ ..::. , " , ~ '" .....~ ~ '~., Natural draft ~ ".. T,,1"'310K , -Merkel --.... -- Poppe (ds) ,.. .'"."" ? ."" (cs) ..""..""0.0100.""''''."" ....kf'1II Figure 0.3: Perfonnance curves ofa natural draft cooling tower with a simplified draft equation. 0.10 98 96 94 ~ ? 92 ~ 90 88 I I I --Detaiild draft equation~ 1---- ~>---L.------j..----- ,.;::..---L.------ I..-- ---- -- L.------f-------- -- ~ - ~---- t.------------ Natural draft --- --

- f------v T a l=280K -,--------Merkel f---- Poppe (a) 86 0.000 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 WI. kgIkg 70 .,---,...---,...---,...---,...---,...---,...---,...-----, 0.0160.0140.0120.0100.0080.0080.0040.002 68 t----+-----+-----t----+----I---V~------_I_--_1 Detaiild draft equation ---- ----...""-..,"'---- - ---~ - ?66t------'I---f----j---::O+""--+-~~_+_-==-~+---1 _--- /::1--'" _-- -- ...-~ 64 t-----+~~""'-f_---------+--,-~...-...-=----,J,---' -":------+__---,-=-"""--~"----__+-_--1 (b) ~ ~::----- ..----"'-"'62 +--~~<:G----~==-------I-____="..cc---+--__+__-__J Natural draft e.______.7 - Tal = 290 K 60 t-------c!~.------~_+_---/----+--__!-____I- Merkel _______ - - Poppe 58 -1----+---.....---+------.+---1---+-1 --4----1 0.000 WI. kgIkg Figure 0.4: The difference in the pressure differential between the simplified a nd detailed draft equations for both the Merkel and Poppe approaches. -r"lffllrn 0.11 43,---,----,----r---..,-----r------, ----------------- --Detailed draft equation '"=? 39 +------t---+-~=_=,..Li---___::_I_L---+--_1(c) .?37 t---=..L:-+_--~'F_---+_---+___1 Natural draft T a 1= 300 K -Merkel --Poppe 0.0300.0250.0200.015 w.. kg/kg 0.0100.005 35 -1-----'----+----1----........----1-----1 0.000 23 22 21 20 '"

19 =-. .?- 18 17 16 15 14 0.000 0.010 0.020 0.030 0.040 (d) 0.050 Wt, kgIkg Figure 0.4: The difference in the pressure differential between the simplified a nd detailed draft equations for both the Merkel and Poppe approaches. 0.12 ,,", 0.000 ClJl1 0.002 0,003 0,0001 OJX!5 0.005 0.007 OJX18 "'1.ka/kI; '"0.000 0,002 O,1XI4 0.008 C.P 0,010 ll.O12 0,014 0.018 IfII,k-"a "'" "'" """ """ -- , "--, -'------, --t--- , , ~ -',, Natural draft , '---, Tal "'280K , -Mob! " - Pop", """ , -...--: ............... ""::::''Natural draft ~ ............. Ta l=290K "-~>, - """" ~--- Poppe (hz) N"""'dnft fT a l-2OOK - """'I ~ ,/Poppe V, ..-

/ -V- /-V> --, -,WI, k:Wk1 '"' ""0.000 0.1)02 0.004 0.008 O.OOB 0,010 0,012 0,014 0.016 '" ? .'"" , Natutal draft Ta l=280K V, . -MsbJ ~V ~~/Poppe "7 --.' J/' , --/ .// 291.0 ""' ,.., ""., "M 0.000 0.00\ O.(lO;! 0.003 0,004 0.005 O.OOIl 0.007 0.006 "' .. tett, '" ~, ?i :290.5 " NaIi1nIldraft VT a l=290K f~....." -MO\<d :7-- Poppe V --~V-~ --'"0,000 0.002 0.004 0.006 O,1Dl 0.010 0.012 D,D'l4 0.1J16 ,.. ..kPI "" ?,,., .: , NIIbmI1 draft _/ ~, Ta l=280K ,:;;;V -M""" ./ -- Poppe ...-;.V , -""---V, V;.. '"'

'"" "'., 0.000 0.001 OJXl2 11003 O,(l(l( 0,006 0,008 (I,our o.CQ1 ..... -....c '"" ? '''' .: N"""'_ J-Tg l=290K , -Morl?1 ",,"V- - Poppe ...-~1/ "/,,, ..-V-'--- ---' l/ ,.." ""'"QlIOO O,llO2 110004 (I,DOll 111X11 (1.010 11012 ll.O14 0,016 WI.kaJkt "'.'" ,.." j "''"' ; ? ?'''''' NlIlur11draft r T a l-280K -. . -Mob! ,---Poppe , -- /.-- 7 . .-- / ............ / 1/ / """'0,000 0.001 0.002 0.003 o.ac. 0,005 0,008 0,001 c.o08 w,,1Ie'<& ""'" """' -~""", , """ ? "',---,---,-,---,--,--,----,--, " .... , '"'t-:--4'<cC+-t---+-+-+--I--/ j "'~ '---, i '" t--I---P,,",,-?2'-''+:--j--t--I--I (as) i N""",_ ~~--, 360 T"I = 280 K f--f--+-"---'1""",:1-'''''4--1- Mob! _ -- Poppe '"L'::=::::::::L-l-L-l-L---LJ a.coo 0.001 O,QO;! 0,008 0,0001 0,005 0.006 0,007 O,OO!! W"kw'ka

" ---; " , ~ " .~~~ " '" '-,..", , Natural draft ~T g l"'290K -', " -MO\<d " -- Poppe ~"'"11000 0,002 1l.0004 ll.lXlI G.OO8 0,010 G.1J12 0,014 0.1J16 WlokJofll Figure 0.5: The difference between the consistent and inconsistent application o f the fill performance cbaracteristics while employing the Poppe approach to determine the performance of a natural draft cooling tower. r'?J"':'" "':. : 0.13 . '. ~R''. ~~N__ , T"l=)OOK ". -Mm. . -- Poppe ~ '" '" " "'" I'....N__ '-.. T"I'"'31OK "'" - "'""" -- Poppe ~ "'" '"0."" 0,010 om. ."" 0."" 0"" 0.010 ."" ."" 0'" 0."" "hJ<&l1c:ll ....~ 0""0."'"0.010 , ? NI\lnl draft .....T"l=31OK /" - "'""" -- Poppe ./7' '" V V /' " '"0."" ?i"" ? 0"""".."".m,

w" kal'kI 0.010 c ,<'Natural draft /-/T~I"'3QOK ..- ....... 4', -- Poppe _/ /.-/}--..--;;.'"0"" ?;'" ? ....0 0."" NoMol'''''' I.fTal-lOOK .) -Mmkd ..-- Poppe I::?",,~ ./'? /' .-/ 0"" .'""0."'"0.010 0 7", NoMol'''''' 0 T"I-310K /'" -MoIooI ..Poppe /'" .-/ /' /'" 310,5 31'. "".. "'oo "".? ?: 310.0 h 0"" .""0.015omo "".0 '"'' "".0 ""., Il",kWk& ~,~ ~-,", ",NoMol'''''' ,.Ta l=300K -Morl<. " , -- Poppe """"'.~ . , " .

.~ ", i ~ " -' I " -Nallll'81draft '" -, - I T"I=300K F:: .-Mukd PPoppe (ds) (<4) 0."" 0.""0.'" 0'" 0,"" ''''' 0."'" 0.010 0.010 "" """..... ......... ""'-Natural draft r"I=3IOK " '--Mukd - - Poppe ,'~~, ' ., ~, . ~~ Natural draft " T"I=310K 1'<,? - ....... ~--Poppe ?0."" "" ""'"0."" (cs) ."" "'" "". ."" ."'"om' 0.010 omo """ 0.. "..0"" ",oo ",., ",oo 121llO ,., ."" .. i 12100

.;; 12lll5O ??"'''' """ "'"' ".. .. " kf'k& '" h kII'cI Figure 0.5: The difference between the consistent and inconsistent application o f the fill performance characteristics while employing the Poppe approach to determine the performance of a natural draft cooling tower. 0.14 "' .....,.. .~.... "',~I ,..,."'" -..;--.........: -....-... -R-Moohanieal draft -.....:~--Tal .. 290K ~-M<ml -- Poppe" " " " ...0.000 0.004 o.llOll 0.006 0.007 n I"I-_.l-_.l-_-"-_-"-_-!-_-+-_-I-_-" 0.000 0.001 1).002 ~ , _-C:'t--=-+---/---/--+-+--!---l r ---t--:-~ "l--+----f">~r__:::::..~---~- f...,h.--:=--+--__ +--+---1 (al) ei 18 Mechanical draft f-- ~~ ~~~~ Tal- 280K ---t-M<ml - - Poppe (~) 0 ,Mechanical draft T al =290K " " -M<ml / , " - - Poppe , ~ ../. V; .... ../ ../" --'" '"0,000 0,00:1 0.004 Q.00I!i 0.006 0.010 0.012 0.014 0.016 .. " ko'kc .'"i

" , Mechanical draft ~ ? Ta I""280 K 0' ", -Morlrel - - Poppe ./ J,J' , , -/ "", L./ ,, 1/./ ./ ./, --, "'? "'-; " m m m. "'.OJXIO 0.001 O.lXX2 1),003 OJXI4 0.005 0,006 0.007 0.008 Mlb,," "'., O,lDl 0.00, 0.002 0.003 0.004 O,1XIi5 tHDl 0.007 a.DOlI w"kf"Ill '"0.1;100 O,lXI2 0.004 0,006 O.lXiB 0.010 0.012 (1,(114 0,016 ..... kllkc "" '"" "'., "" Mechanical draft Ta l=280K ~/ ..... J..-M<ml --- - Poppe "" .... .~ ....-' /....-~ ....... ' .. /---, / Mechanical draft Tal =290K ,- (: -M<ml ,/ -- Poppe '-;/../';V :.::.~~ ----t:=::--- r........., .......~.. 0 --, '"MechBnicai draft ~... ~::~Tal ~ 290K -M<ml -,~ - - Poppe "',r1I',.kWkt

'"O,lIlllI 0.002 0.0001 0.008 0.008 (1,010 0,012 0.01. 0.016 '" j ;"" ? ? Mechanical draft ' ....... Tal =280K ...... ~- '---. - Morlrel~ -- Poppe .......~- h. -- ~~""'''' ~.........~ .... ----r-...-...... j ;'" ? ? '"0,000 O.lXll 0.002 O.lXI3 0,0001 0,005 0.006 IUXl7 o.cUl ", .. kW1<1I: '" (bs) , :::: " , .....~~~~.. Mechanical dtUt -......:~T.I~290K ---M<ml "~-- Poppe '-- r? 11'" kIIkI: ,., ,., ?., 0.00J 0.002 0.004 o.lD6 0008 0.010 0.012 0,014 0,016 i ?I" I (as) -Mechanical draft " Ta !-280K "-. -- """"I"-.. ............... -- Poppe -...... I'-...... -""-- ... --............ -"-- --'.ro

,ro 0,000 0.001 0,002 0.003 MO", QlXl5 0,006 0,007 0.009 WI> kg/k, j ?t 6.00 ; '" Figure 0,6: The difference between the consistent and inconsistent application o f the fill performance characteristics while employing the Poppe approach to determine the performance of a mechanical draft cooling tower, 0.15 0.0.00."0.""'.0 , >-. ? , ---...:: .~~~ .. 0 ~-, ~'-.Mechanical draf\ " K-, T.I-310K -. -Merkel "'"? Poppe -, ""0""''''."""'"..,??0."" , , -"':-~~ ..... """ -"'---Mechanical draft ~-T4 1"300K "0-Mol<d ~Poppe ~ ? 0." " " " " "'"k;I\I:I k:'Meclwlical draft /, T"I"'300K k:"" -Mol<d ./' -' - Poppe --~f --"V" --::/ " (d2) 0."0'"0.010 0 MccharJk:a1 draft /'

T"I-3IOK / - """"1 -- Poppe ..". l'" V /' ? ..:,/'" '"0." .-i """ (C2) 0.'"0"" "'" .., ."k!dkl '.0 '"" '"... Mechanical draft VT"I-300K -Mol<d V-- Poppe "J,,"': ./" !,../ -------0."0."0."0.010 , I VMechanioal dtBft 0 T"I"'310K -Mock<l 4' "....- Poppe /' l.,</ # ~/ , .. 0."0"" "'" 0.00,,"""... (1:4) 0."0.'"0.'"0.0'0 w .. ItWkI 0,010 , "" "" ""Mecbllllical draft t-... T.I-310K ~ -MCfkcl '--""'-"- Poppe "m 0." " '" :i~ ;m ?.~ (ds)

0.'"0.0'"0'"11.010 ,,--~ '-'" "'.Mechanical draftT"I"'310K ~ -Mol<d -- Poppe ~'.0 '0 0." '0 '0 '.0 :i ?! ~.O Ii 3,0 (Cs) 0.'" '"" 0."0.'0,(1\00."" 0 , ~" 0 R, -, '" --, MechBnit:al draft ~T?? -300K ~, 0 -Moe'" ~'" Poppe I ~ "0.. ? , w"JlllkI Wh"alka: Figure 0.6: The difference between the consistent and inconsistent application o f the fill performance characteristics while employing the Poppe approach to determine the performance of a mechanical draft cooling tower. 0.16 .., 0.000 0,002 O.l?! 0.005 1l,0QI!l 0.010 0.012 0-014 0.019 w,,1eA ,~ ,...-,.--r---,...--'---""'-'---'--' "'"1---1---1---1---1--+-+~+- "'""00 ~~ """" ~ Natural draft ~ T a1 =290K ["':. - Lel-a.s ... __ Bosq.kori: .. _ Lel= 1.3 (bz) N"""' .... Tal "'290K ~- Ler=O.S /-. --13osr\iaklMo P_ Lei = 1.3 --# ~ L:-V

.;:; '" '" ~ 0.000 0.002 0.004 1l,1Xl6 0.008 0.010 0.012 0.01.' a.alS "'\o~1 -----::::- ;. -----:::: :;;;;v --:1-::;;;~ Natural draft ;..-- Td l=280K - Lel=0.5 -~ Le/= 1.3 ~, 0.003 0.001 1l,lXl2 lI.OOJo 0004 O.(ICl} ll.!,lOIl 0.007 alDl "'1,kw'k. I--l:/j;;.-+--V I-- V ... -" V Natural draft -9' T"I-290K -- LeI'" 0.5 --~ _ Lef=l.3 '" '"Il.WJ 0.002 0,00( 0,006 D.lXllI 0.010 0.012 0.014 0,016 ... "kPI ? '''' .: , , , k=?;>:. , 1---::::~ ,--;;V , , V-- Natund draft~",,--~~ T d l=280K ';;;-- Lel= 0.5 , --~ Lei = 1.3 m. ~., 0,000 0,001 0.002 0.003 0.004 0.005 Q,1XlI:i 0.007 0,001I .". JlIIk& .'", .:,. "'000,000 0,001 OJX1:2 0.003 OJXl'I 0,00; 0,006 0.007 0,001I "'" ke'kl ............. ,-,---,--,--,--,--,--,--,--, """r=F=F=F=t--t--t--t----j,,,,~,~ t--+--j---+--\--t----p,p'j---j ~ "~t--+--t--r----r:V"'"*":?"-..---j-----1----'1 (b4) ~ 16Zlll---+--1-~b.-~-~--t:::..-1_1_;Nt.mnJj;;;;~d~,,;._ftl ? "'''' t-='i'-7'"'lV~--'----+---'I_-+_~ .T.a I == 290 K _V Ler=O.s ,~t-'-~+_-+_-+_-+_-+-I -_ Ler=U

,~l--l-_L-L....1_.L5==l:::::::..J O.tIIXl aOll2. 0.004 o.tJ:lfi O,QOB 0,010 0,012 O.O1~ 0,016 (34) ---: ---:::~-- '/v --~-' ~ Natural draft ---~'~V l'al=280K--,,---- Lef=O.5 __B~ Ltf=1.3 ,,"" = 0.000 0,002 0.0001 0,1)[1I Q,OOIl 0.010 0.012 00l~ 0.016 -"" ----------------.. '----j,. ------I--!:m ---- ~--- (as)Nohmo1d", ---? Ta l=2SOK~ - Lel=O.5 r"" __ Boor;.J?M:: Lef= 1.3 ""0000 000' ,... 0000 000< '000 '000 0.= 0000 '- """"'d'" ': '--.... T a l"'290K" Lt/= 0.5 "" '" -....-... -" _ LeI"'!.J " ~~, .......::: -....:.' -- -....... "~~~'-~ -....-... (bs) Hlj,kg/kf; Figure 0.7: The difference between the consistent application determined for different Lewis numbers, while employing the performance ofa natural draft cooling tower. "",kJI1<a; of fill performance Poppe approach to characteristics, determine the 0.17

"N111l11'll1 draft T~1"'300K .....~ - UI=O.S__ B~ f"--.. _ Lei'" 1,3 "-".........~ , ~ ....1'..N__ ......... ,..... Tal =310K "- Lel=0.5 --~ ............... _ Let = 1.3 "'oo ''''' ,=O,trlll "" '''''"'" o.ot50.010,....""",. "" 11' .. ,," ...... ko'kI ----....../' l--- I--' / ---------V """""d....----V Ta l=300K_ Ler-a.S __ BasnjalaM;: Lei = 1.3 ./V ./ / "" NII1Ural draft ./" Tdl-JOOK - Lel-a.S .....""" _ Lel=}.3 '''''''''' "',. klIIk& 0,010 N__ ......Ta l=310K - Lel=O.S ./--~ "" _ LeI= 1.3 /'" V V ~ , Natural draft

, Ta l=310K - LeI""O.S ~?-", _ _ Bomjo.koYil --_ Le,-I.3 ~V, ~~ .--;~ ''''' " '" '" " "~ 31 .: (C2) "''' ,...,., II'hl!Wka D.DlD,... '" .. " '"~ "~ .i '" '" ''''' ,... ,., ''''' "'" 0,010 "'oo ,."" ''''' ''''' . .. ,.kt1k1 ..... Iqp'kz """ ''''' -------- ~ ",..--- J.----------NlltWlI1draft Tal-lOOK ?..ia- 0.5 --~ LeI = 1.3 -~ i:"-~~ Natura1dfllf\ ~~ Tal-lIOK -. - Lu=O.s ~~--~ -....::_ Lel= 1.3 """ "'00 "'" j"'" :: 117!'O , E 11~ "'" """ ,."" 0,010 ,=

''''' ,." "'"j , "'"i ? D.01D ,." ,."" (l4) ,."" "l.~ ~'~ ...:x Natural draft 'Tal "'lOOK -, _ LeI""O.5 -'. -~ ~ _ Ut= 1.3~ -....::;~ "'" '" ~ .....~ ...... "Natural draft ........ ......... Ta l-3IOK ....... - LeJ"'O.s ............ __ Booojalcavil _ Lei=> 1.3 (ds) .""''''' M'1.ka/kI ,=0,010 , ,."" (Cs) ,."" ''''' ,....0,015 .. " kw'ta 0,010,..". ''''' Figure 0.7: The difference between the consistent application of fill performanc e characteristics, determined for different Lewis numbers, while employing the Poppe approach to de termine the performance ofa natural draft cooling tower. 0.18 Mechanieal drllfl: ? ~ Tal ""290K~ - Lel-O.S ? BlJlnjakovil '"~ Lef'" 1.3, ........ r-... ~ ---.. to" " "0,000 0.002 0.004 O,!D5 ll.OO8 0.010 0,012 O,O1~ 0.016 "'\. kf'kC

" 1:::--~~ ? '"." Mechanical draft ~~T~l=280K ~ - ur~O.5 _ _ BosojokM: _ Lt/= 1.3 " n O,I))J 0.001 0,002 0.003 o.lJJol 0.005 1I.1XllI 0.007 0-008 " .... 1laI1I1 ,.,. 0,000 0.001 0.002 0.003 0,(1)1 o.lXl5 0.006 1100T 0-008 "'1Ik1"<!1 '"QOOO a.D0:2 D,~ C.ooe 0,00II 0.010 0.012 1l01. 0.018 ....k"" (bz) ./... V ./'I' / ~ Mechanical draft ?" Ta l=290K~ _ !AI=O.S __ B""';'lo:>ri: _ Lel= 1.3 (az) , , Mct:hri::al draft , T",=280K , - Le/~O.S L-;:- .. - - Bosr;l.kovt ~ --~, LeI= 1.3 ---:: ;p-, -, , ; ""r--,r--,---,---,--.,---'--""--'""], , , I., ~~ v-V ,- /-V Mechanical draft -<:V T"l"'280K ;/ Ler- 0.5 -_ BosI\iIlkcM;: ? _ LeI'" 1.3 "". ". "" W',kA '" "., 0.000 0,001 O,oo:l 0.003 0.004 O.OOl5 o.lllIl 0,007 O.lXIl "".~+=+=~~~2!lIl.S f-- Mechanical draft 2!16J) T"I=280K t-'''''~'-F"-odc--+--f----j-f - Lei'"' 0.5a.s ~ __,~_______ ~ ~ ::t:=t==-~+-.::;-;';'~,~,:--j::t!=FL.~f.~I~.3~ (84)

: 294.0 +--+--+--+--f""~'""i~,""k:--+---1 ""'-.- ::::::: "'.1-----i--+--+--I---j-C::::;'""_;;;,;>-'-,--I "'" I--+--+--+--I---j--I---l"---I "".? 1----+--1---1----+--1---1----+--1 j", , , "" ? Mcchenical draft --TaL ""290K ur=O.S :::::::-" -LeI" 1.3 ----,~~ I--I-- -....... -"'-,1---1--+---1---1-----1--1---+---1 0.00) 0.001 0.002 0.003 0.004 0,005 Q,OOB 0007 O.llO!I .....~ '"11000 0,002 D.OO4 OJDl O.OCI8 0,010 0,012 0.0'14 0.016 W" kI/kll " ,. 0,(0) 0002 O_~ 0.006 0,008 0.010 0,012 0.014 0,0l& " r---,----,---,--~-~--r----,.-~ 1-----___. .. t--+--+--I=-kc-+--+--\----1 '.. '" -1---1---_ j .. !-"_b-~f--.::.::j.~-?--+--I--+---Ii 1--__ '''.. (as) i 5.5 Mechanical draft T"I-280K - LeF'" 0.5 -- JJou;ok""'" _ Lel=l.J ?? L'::=~::5::::o-l_--l_-L_--L_--L_-' 0.000 0,001 0.002 0,000 0,004 0.005 0.006 0.007 0.006 " ?? .. i ?t 5.5 , I 5.0 ?., ---, ~ --:..- --, -....... -::.:'-----", Mechanical draft ~-~~~ ... , ------, T a ]=290K r---....., -~~ .....

- Lel=O.5 ...... __ BosnjoktM: , _ Ler= 1.3 (bs) "'11Il1:o'<1 Figure 0.8: The difference between the consistent application determined for different Lewis numbers, while employing the performance of a mechanical draft cooling tower. W" k&Ikll of fill performance Poppe approach to characteristics, determine the 0.19 "".""'''''' ? ~~ ":~ ~ Mechanical drllft "Ta l=310K "-- ?61=0,5 ..... --~ _ Le/= 1.3 " , ."'" " ~ " ci ? 0,0150.010 ?,,-:----,---,----r---,---,--,~ ?.1--''''l2-~+--+--+----I~-_1~k " ~ " OJ 12 w,,1I.J4lI w"Jtw1<t: V LPV ~~..~/ Mechanical dnft k9 Tal :3OOK,-' - Lu=O.5 0/ --~_ Ler-1.3-./ ./ A il" # /;:? Mechanical draft T"l=31OK ~;:/ Ltf-O.s__ BoIJliolo:mo 0/ Lel= 1.3 ."'"i " "","" ... 0.010 .m, "'" ."" "" '" .~

i ""'" .,., .,"" .,"" ,,., .'""'" 0.010 , , ..-..#'" ..;~ ~V Mechanical dlllft ----;;V Ll"310K - ?61=0,5 .. __ ~koY:i: _ ?ef=I.3 '" " ? '''' ..: '''''."" a,OIS0.1)10,... ~ -/ I----" ---/ -.. ' . ./..- Mechanical draft----~ ..' .' / Ta l=300K. - Lef=O.5 ---_B~ _ ?6/=1.3 , Mechanical dTllft ---T.l=300K I-- LeI=O.5 -....:.:: -'-- ~ __ BQbM:. ........... "- "L" 1.3 '" .......... ~"'----........ J-...:.'...: ~ (<4) .,., .'"'''''' 0.010 -:, ----.:~ ?

~ Mechanical dlllf\: ~ T.l=3IOK ", - L6r=0.5 ....... __ B~ ~?6/-1.3 " , '" ,."'"0.0250.0200.Q150,0100.005 '" '" '" ". 0,000 j '" ,;; 284 i .'" "'"kWkll: "- Mechanical draft ":'::'- Ta l=300K " LeJ=O.5 ......~,~ __ BOIllliokoW: " ':::---.. Lel= 1.3 'Z.'K ~""I""-..~ " ..:.:: ~ Mechanical draft "'<: T.I=310K ~ - Ler=O.5_ BomjoIoM> " ~ _ Lef-I.) """''(ds) ,,., ''''''''''' 0.010 , ."'" (cs) "" 110150.010,... " " "" ,., ., w"k&fIqc "'hllW1<i: Figure 0.8: The difference between the consistent application of fill performanc e characteristics, determined for different Lewis numbers, while employing the Poppe approach to de termine the performance of a mechanical draft cooling tower. 0.20

,~ "., r-"00 '-- r--'" "" --- (at)~ "00 r- f-ci --~~--N__ "'" Ta l=2SOK -:-"00 - Lu=O.5 "., - - Ilomjtkovic _ Lei"" 1.3 "'" """ i'-. "'-~~ .... ~I'--... ~ ~~~ .....~ r---:' " , r--... -, Ta l"'290K , .......... t:---... _ Lel",O,5 r--::: '__B~ _ ?el=1.3 '"'0.000 O.llO2 0.004 lI,rolI 0.00ll 0.010 0,012 0.01. 0,016 w,.tH ~, 0,000 0-00\ 0.002 a,om o.~ O.Olli o.OlJl 0.001 0.008 WhkW'kl (~) ....v: ......':.:"" / /'/ '/ ---V v,....... /" Natural draft.;' Tdl =290K / ~,/ -/ i"'" _ Lel-O,S " ./ -" Lei"" 1.3 '" "" '"0.000 0.002 0.001 0.005 0,1UI 0.010 1;1,012 1101. 0.018 w"kWkI '" L-------------

v---V --- V ~~,~, .. -----/--.....'~----NatuIal draft - ,/ Tal =280K v--- _ Ler"O.5 __ B.....;.klMl LeI" 1.3 "i"'" I--I-" ~1---I-- ';/ --~ 1/V -v:V Natural dRIft-_ ........ Td l""290K -- Lel"'O,S ....;' - - Bosnjolm'" _ Lef"'l.3 '"0.0:0 0,002 0.001 0.00ll o.tDB 0.010 0.012 001. 0.0111 Wt,kt'!l '" ? , ; ....--? ......~ L.--1-"/ ..__ J ~, V "", ~ ? - V _'nfl, ....'-' ..TaJ -280K --- V Ler=0.5-~ , Le, = t.J >n >n. '"' " '"'? .: '"'. "'., 0.000 0.001 O.C1'12 0.003 0.004 O.1XI5 0.006 0.001 D.OOll '"" "'I' k&/kI; """'

"'000.000 0,002 0.004 0,1XIlI 0.008 0.010 0.012 0.01. 0.01~ 10'" kl'kl "'00 ,--,--,--,---r----,.--,--..,---, --i "'"t--+-~-+-_J=_+__tL..-----b__1 .- --- ~) : 2I7lOO !-'''-:.--+--_-_-+~+",..4L----''--_j_l Natural dl1lft1----- T. , ? "" K "'"!---1o--4'''-+--+--4-1- LeI = 0.5 , ~-'-+---I--f--i--+-r-=~~ , l---l_..L_L-L_-L':::::i=::i:::::U 0,0l:Il 0,001 0002 0.003 0.004 0.005 O.OOS 0.007 Q,CXlIl w,,"-'k, ,... ,... jll>4OO ; ."'"'E ''''"' ",,,, ---, ~, ... , /' --....---".. -" ~~ ... ~~ ....V Td l=290K --V Lu=O.5-~ ....V _ Lei = 1.3 '"0.0:0 0,(102 0,000l 0.008 O.lUI 0.010 0,012 0.01. 0,018 =t-'~-h~+--+-+-+-+-+---I '" -------~t--..:;:;~cc:t--f='-+=+-+-+--I j "" t--\--:'::':_'.:-:-'I'-'...-'~'F---'"-I=-----\----:1==I---I . '- "--, ( )f :l6O Naturaldraft ~-- ---- as Ta l- 280 K --.... - Le/=Oj __ BoIrlIf,~ [AI- 1.3 '"'L5===:i:=-.L---l----L-L..-l-J 0,000 0.001 0,002 0.003 0004 0.006 0,005 0,001 lum ~ ~ "'" "'-.......:~ Natunli dfllft ""'" '-,--'Td l""290K ----_ Ler=O.S "'-:.:>--~ _ Lel=1.3 " ..... kg/kI ""1~ Figure 0.9: The difference between the inconsistent application of fill performa

nce characteristics, detennined for different Lewis numbers, while employing the Poppe approach to de tennine the performance of a natural draft cooling tower. 'r '.," 0.21 ~ Natural. draftT~I"300K ~~ - Lu=O,S--~Lei'" l.3~"~""'" ~~'~~ ""-1""r--.... Natural draft ....... ""-T"I"'3IOK - Le/=O,S .........--~ "-"'" "" "" ,.. 0,010 O.Ol~ ,... ''''' "" "" 'ro '000 0.011) .,,, "'" A ~ # ;;-'~ /-:p.- Natural draft 0 ? T"l=3ooK" _ ul-O.S '" __ Bo!llljamvE _ Lel= 1.3 Natutal draft ./T a l=310K ..... _ Le/~O.S / --.....- V '"" ..; " ./ ./ .'"';, "" "" ~ '.000 ,.. 0.010 w,. II8/kI "'" ,... ''''' '" ?;"", '"'000 0,010 "'" ''''' '000 "'"t-~=t---t---+--+--+--j~---- --"

(<4) ''''' ... ''''' ."" ...,kWkll ,....0,010 0.Q10 -~ t:--.. ~~ Natural. draft ~-~To l=310K ~,~_ Lef-a.S__ Bo~ ~ , NatunUdrafl: .--, T a l"'31OK _ LIII=O.5 ~V, _ _ Bcmjokovio - ~V-, .--;~ =-... '000 ... j :: 7000 ? ? ''''' .,'",... ,... "'" 0,015 ,., ,., 0,010 ,... ,... , t-... ./V I-- --~- "/,,IY - ~-_...---- '/-~--- Natural draft V T"l "'300 K_ Lel=0.5 -- Boso,Ilk<wi: Lei'" 1.3 ""'.000 j "oooj--f-+=+---IV..-,:'C;::"~---J (C4) .;; 11niC : l\!5lXll-,,,=l::::::::'''i''''''-----=-"i----H N_""" , Ta l=300K ""'J.-~-f--_+--_I_--_1j_ Lel=0.5 --~ ,,000 J.---1---f---I-----u- - Ler" 1.3 ""., l-_L_-l.._J.._J==:C:=!J '000 ? ''''' ..: "'"k8&1

, Natural draft T a l-3OOK I' - Le/=0.5--~ ........ Lef- 1.3 ... " ~ ....~ ,..0,010 "'....... ........... Natural. dreft ....... .......... T. 1 =310K ...... _ Lel=0.5 r---...__ Bo~ '"' ''''' ,... ''''' 0.015 "'"kWka: 0.010 ,ro '000 Figure 0.9: The difference between the inconsistent application of fill performa nce characteristics, determined for different Lewis numbers, wbile employing the Poppe approach to de termine the performance ofa natural draft cooling tower. 0.22 ~ Mecfwmicaf draft ~,~ T.,=290K- Lu=O.5 ~ -~ _ _ Bomjakovi> '- r---.. _ Let"'1.3 ~-~ -..........: -- --. t--...h> r---:::: " " " "O,!XXl 0.002 0.004 0.006 0,006 O.G10 0.012 o.01~ 0.016 .... lIJIkI ........r--~~-~ I'--.... '" ........ "-........ ----r---.. ~"-"" r-... -'- ~ t-Mecharicd draft I'----- ..........T~,=280K ........ --_ lAI"'O.5 ------------- I-_ Le/"'" 1.3 "'.kIIkI "0.000 (l001 0.002 0.003 0.004 0.005 0.006 0,007 o.OO!I

(bz) ......--::~ ~V ~V'"" /.---- --V Mechanical draft'l -< -;/ T.I ""290 K -_ LII=0.5 ?7 __B~ _ Ief=1.3 '"a.1JIXI 11002 O.()()oI o.l108 0.006 lUl10 0.012 0.(114 0.016 w"kWkl v . -----V ...~ ...... " I.--,--...v ---- ~ ... ...-...... /" Mecbanlcal draft T.I-2S0K -' -ur == 0.5 - -_ Le/-1.3 "".D.lXXl O,Cll! 0,002 0.003 0.004 O.1lXI O,OOll 0.007 0,006 w"kWklI ,." ?i :m.5 0m. I--I--f/ e-- /;;V ...............'V ~ddraft ,,- V T.I .. 290K --_...--V _ LeI"'O,S V _B~ ---_ ur-1.3 ,., O,(IllI 0.002 o.QOoI 0.006 Q.l108 0.010 0.012 0.014 0.015 w,.kWka ."", .: '" ? ? V-? ,

, I~ , ,", ? V.....-,? ,,, -----Mechanical draft .............. L------ T.,-280K ? , - lAl-a.s ? ---- B<>sqjok<I\'io ---Lef= 1.3 ? '" 0.000 Cl.OO1 0lXl2 O,DCl3 0,004 O,(Xl!i o.lXl6 0.007 0.00II ""kWlll '" " '''' .: '" '"0,000 O.lXll 0.002 0,003 O.lD4 0.00ll O,lJ[lIl 0,007 Ol;llll (bs) ~ Mechanical draft -~~ T~I-290K...... _ LeI=O.5B~, :::---- _ Ule1.3 --~~ ---......-: ::.:::.::Mechanical draft ....... T.,"'290K r---.. _ Let-QS --- r-----... -~ , ---- LeI'" 1.3 -~~,........ " r-- , -~~~ ::-:---.r--. -I--- - -:::::- tof.. , "(WOO 0002 0.004 QIJtII o.em 0.010 0.012 O.O1~ 0.016 Wlt~ .aOOJ 0.002 aOO4 0.006 0,006 0,010 0.012 O.O\~ 0,016 ..? j 6,0 , ~ 5.5 E 6,0 ,. j" ;

."" ? ---I---- -- I-------: ..::-----~ -----, Mechanical draft r-... -"--. ----T.l=280K ------ .. - Uf=O.S --~ _ Ler-I.] ,.. "j ~ 60 ? w" k&IkI w,. kllkl Figure 0.10: The difference between the inconsistent application of fill perfonn ance characteristics, detennined for different Lewis numbers, while employing the Poppe approach to de tennine the perfonnance of a mechanical draft cooling tower. 0.23 ""0 "''' ?."".""0.010 , "? ~, , 0 ......... Mechanical draft ~T"I=310K _ ?e1"" 0.5 "" "---~ _ Lef= 1.3 """' , ''''' ~ , '" . ''''' ..,.., ? ? .......... " 1000... ? ....., ~Mel::hIInical draft T~I = lOOK -, - 1.eJ'=O.5 ---~ ~ _ L~f=1.3 ? '''''

" "1.kw\<& "'h~ Mechanical dlllft 9'Tal"looK - LII/=0.5 ~ - - BottjokavE A Le/- 1.3 ~V ./ ,/ 0."'"....0""0 "'''' 0.010 ? Mechanicll1 draft -7 T.I=310K /' - ?el=0.5 __ Booliokovi> ~ .... _ Lei 1.3 V ./ .a! """ " " ? ,."" .... j ..... ,,,,, "'''' O,O1~0.010,.. ... ,."" lI'"kar'<g ~~ .7~ -/~ ~~~~.-~-~ ..-' ---~-- /'~.-.- Mechanical draft -_.'-To l=300K L-------' - LU~O.5 --__ a Lel= 1.3 Mechanical draft T.I=JI0K -~- ?er=0.5_ _ BoonjokovW; ,/Lei"" 1.3 ~ ..-'_-~,; <-- ;;7 /"' Z>, ''''' 0."" ??0 0,015 .. "kWl!II: ''''' ''''' ''''' '" '"

'" .. .. 0"" 0.010 0."" ,,,,, ,,,., .... (c4) '''''' 0.'"0.""0.010 , ~ ,-'~ -----.~ ---.:....... ~ Mechanical draft .... I-...Ta j=310K -........ - Ler=O.5 --~ ....... LeI'" 1.3 """'" ~. '" ''''' "" 0.0300.""0.'"0.0150,0100.005 '" I--__l--__l--__>--_-<>--_-I__-I 0.000 ""~ (ds) ,."" ...'"0.""0,010 , ....... ...... I---. "Mechlmical draft , Tal-lIOK , _ ?el=0.5 __ B~ ~_ ?e[.-1.3, ''''' (cs) "'''''''' 0,0150.00."" 0 , Mfcbanical draft "T.l=300K , ~ LeI'" 0.5, -~ ....~ _ I.er"" 1.:3 0 ..... "~ ~ "'"'.00."" ? " ? .,

w" kI/kI Figure 0_10: The difference between the inconsistent application of fill perform ance characteristics, determined for different Lewis numbers, while employing the Poppe approach to de termine the perfonnance ofa mechanical draft cooling tower. 0.24 0,0140.012O.QlO ""' j----j----j----j---J---J---I---j ,,~,-,--,---,--,------r--,---, "", """"""""""""."" -,"'" -, Ta l-280K --- """", , - Poppe r----. ---~ ---------, ~ " ""' """ "" "'" "" "" ""' ,,~ ""~ "'" ""'~ ..... kWlt& -,"'" ./ T.I-290K /-M<rol -- Poppe / ./V /' --~ (az) , Natural draft Tal ==280K ./", -_I -- Poppe V ? V, V --"" V

"''' "'. """ """ "'" ."" """ "", "" """ ?."" """ .'"" "., 0.01'0,010 """""""""""" """"'-T. 1"'290K ---.. -/V- - Poppe -- -;:/V , , -----V / "" "."" '" ,,,,?.""?."".""""" ? N__ , T'l=280K ? -Mmd f~- - Poppe /', V,/,,~ " ,-:;;-, -' --V ", , ~ '" .."" .""; -"" Natural dtaft Tal = 200K , - """" -'/- - Poppe ..-/ /;V ---- ----'V / V - NtIUrll1 draft ... Ta ,-280K --- """" -' ,.." , -- POPPIl ,/' Vi- ..- (14), V;

--, .? ,= ? ,/ ./ " , --/ "'"' / "'"' .."" """ """ """ """ """ """ "", w"JiIlkll: ".'" j llMOO ; ? .""" '''"' "."" """ 0.010 0,012 ".. Na1UnlI draft Ta l:280K ~~~'-Mmd - - Poppe '-----. '"------ ~ -~ " ~ ,NaturIl draft ::.::' T.I-290K -Mok. ~, - - Poppe -~' '--~- , -"" '-~ "":;

i"", '"" "" """ can ."" "."" "."" ."" """ (as) "", """ """ "." (bs) W" ki/kg Wi ........ Figure 0.11: Performance curves ofa natural draft cooling tower at an atmospheri c pressure of 101325 Pa. 0.25 Natunl draft " T~I=300K , , ..,'., - """", - - Poppe, ~'", ~ '" ~ ", , ", , N.....w_ --.....;'." Ta l=310K , -Mm<! K' - - Poppe...., '-.. r::'..... ~ " '"0" 0.010 0.015 """ 0.010 0,,",, """ 0,,,", ,. ~V / V'.'"....-' ~../ . Natural draft -T~I=300K """"-- Pop" (C2) '" /" /' V ~. " Nanni draft .// T~I"'310K -Mm<! " -- Poppe " (dz) 0,010 ... II kIIk& 0.015 0.010 "" 0,,,", .y

,-,--_..V . --' ,,--' V, ",..V """'" d""Ta ,- 300K -Molrel 1/ -- Poppe '''"' O~0,""",.,0,010 , , , --/ . .,' .,./ , . , .' ./", , V N__ 1/ T~,-310K - """" -- Poppe " " '" ? '"..: "'"''"'' ,.,'.00,"'" '"" ... "kWkl """r---,-----r---,.---,----, ,-t----t-----I---+---t--~-___1 ."",- c::.:j 1.m:1 j---+---+===::::r=--r9f'-----I,~ ,_I _---+-_+----4V-fN~;ftl(C 4)~ +-- Natural draft/V T.,-300K -""""1 -- POppe ...r..-:-,.-,-,--,-r::=7:=':"il - Natural draft ... t--'-"-'2.,+-.---+----j--~ T ~ 1= 310 K ~... - Merkel~""..... - - Poppe l---+--~'~'+O---l__-..b=,=;,;d~j'" '.: """ (~) ,,..j----+---j---'''-t---j---___1 """ f'_::::=t=--1~::::?"~~:""".-i.--1 '''"' 0.030 D.o?l """ '.0 0;001-__-'-__-'-__-1-__-1-_--1 '"""" 0.015 0,020'.00.00:1 0.0J!5 ",.~ I " Natuni draft I~ Ta ,""300K-M?kd ", - Poppe ~' .... '" " " ,

~~> 0.010 " -,"" ~ .... , T~l=310K -M?kd ~" - - Poppe ~", '-... , 0"" "" (Cs) ''"'' 0.015 ... "kl!kll '.0 "" Figure 0.11: Performance curves of a natural draft cooling tower at an aunospher ic pressure of 101325 Pa. 0.26 ~"1--''''-.,d---I--_!_-+--I--_!_-~, ,--....,---,--....,---,---,---,---, 0.0140.010 """."" 0."" ''"' , --, "-. --, , "- --? K' '" ---.....:: "Mechanical draft -? '-Ta [-290K ----::::--? -Mai?\ - - Poppe ,, 0.,," """"""""" 18,Q 17.15 \!I.15 j--"''''+----=-:O-''-t~-_!_-_t--_!_-_t-_I l'--- ------_~ ""J-----+--"4'---,.--.--~,_-__J-----+---I-----1 (al) ci l!1.ll tr=O~====:?:=;- -------=1""::::::-f-O-,...,o::::?-~-I--___1 ""_" ''''' I--I--------.p,~-~--I--'':-'''-.,1--1T~I"280K - """" - - Poppe 17.0 J.....!::==+===+=:C..._L__L__L__I-_-I ."" ..... kW1lI 2ll1.0 ."" Mechanical draft TaJ =280K

.--/ - """" -- Poppe / V /' V ./' ./Mechanical draft Ta l-2OOK V-M<rl<d - Poppe ./V / ./ "" ,." "'.0 ."'"/ " "20 m" ,.0 0.001 0."" ... ."" """ ."" (az) ."" '" '"0.,," ."" ."" 0,010 <b2) 0.014 .... Ilf'lq 0.0100.""0,", /:/ V--..----;V --- Mechanical draft - V ./ Ta l=290K - """" ./ - - PoppeI '"0."" '" .'", .: 0.""0.""0""0."0." ."" , I- Mechanical draftTa l=280K --::.--, -Makel -' - - Poppe ---/ ~./~",V ,- V~", ..- V"'.0 ""'

""0 0,," , Mechanical draft t----- Ta l=280K~~~~~ -M<rl<d -POppll -, ~~''''''' -- <~~ --"'" Mechanical draft "-T.l=290K -M""" -- -~ -- Poppe-----, ~"~'. -0 '-::-:-j'" i,,. """ 0.'"' ", .. kc/kI "."" ....~. ."" ''''' ."" .. j ;'", , ."" ."" ."" 0.010 0.0'" Mechanical draft Ta l-280K --'-Moi<_ ---- Poppe ----------------- ----~ '-~ ----(bs) 0,0120,010 .""''''' '-

Mechanical draft T.I"'290K " -M_ -, - - Poppe ~ '-"'--'" -- '-., ~ '0 0.,," ., (as) .""""" 0.""0.""... "'""'" 0.0 " ?., "'" Figure 0.12: Performance curves ofa mechanical draft cooling tower at an atmosph eric pressure of 101325 Pa. 0.27 ? i~, Mechanioal draft ? '~, Td l=310K, - ...... ~ - - Poppe", "'.' "~ ? '" ? ~ ....... , , " Mechanical draft 8 I ....:. ...... T~l"']OOK '-', , - """", ~" - - Poppe ? , '-' " ~", " , ", ,~", ",. , (lOla ."'" ..... VMechanical draft Ta l=300K V-Merkel - POppll V ./ ./V ./ 0,,",

""."'" o,lMa ? ./ / / ./ ./ Mechanical WaftT"l=310K ./ --..Poppe / " '"0 ..O_lM~???0."'" .'"'j '" "'"kA/kI ""'~ 0.... , 0 __---:7 ~_.--~--V -----,----~- - / / Mechanical draft V T"l=JIOK -Moi<d / -- Poppe " '"'... ? '''''~ '"' ,---,---r---,-----,----, ,",l---l---l---l---h~---I 1,,- MecIw1K;a1 draft "-T"1=300K -M<rl<d ~ - - Poppe --- ~'---', - -~--- -~ -,"> w,,~ (<4) 0.010 I~ Mecbanical draftT"I=310K ~ -Moi<d-- Poppe ---- ' .. I'-----'~~-~ ~~-,~--~ ---. 0."'"..,'.0 ."'" ...

... (ds) ..... "~ Mechanical draft , " Td l=310K " i'--. -Moi<d "',0- -- Poppe ,,~ -', "--", " '.0 ,. ... ,.ji 4.11 ; ,. "".'.m."'" I, -,-,"" " "I'-"'C:-.. .~ Mechanical draft T.I-looK '-Mori<d "Poppe , ". '0 '.0 ".... ,. Figure 0,12: Perfonnance curves ofa mechanical draft cooling tower at an atmosph eric pressure of 101325 Pa. 0.28 "'"0.000 0.001 0.002 0.(113 0.004 a.C05 1).006 0.007 0?008 "'"ka:/1lI .. .. 0.000 0002 0.004 0005 0,008 0.010 0.012 0.014 0.11111 "' .......... ". "" ".. ".. "" ~~ " k, ~, I ~-Natural draft ~~T~I"280K -Morl<cl - - Poppe "., "ro

, ~ "'i'-. Natural draft '--..... r"I==290K - Morl<cl '-- Poppe "'.. 0000 0.001 0.002 0.003 0.<:? 0005 0.006 0.007 0.008 "'''Jr." ....... (b2) NaQ.Ifa1 draft T"I=290K /' -Mako1 - - Poppe ./V V ./'"/" '" ""0,000 0,002 0,004 O.iXl8 O.DCI'l 0.010 0012 0.01" 0.016 Natural draft T"I=28OK 1/ -Morl<cl V - - Poppe /' V /' "'-, "'., "'. i:IIII 29:1,5 i " "'" Natural dlU\ V j.-1- T.I~290K -Mo'k" V - - Poppe V ...-/V -----"'0.000 0,002 0,004 0.1308 O.Q08 0.010 0.012 0.014 0.016 "'10""'" , J.--'" ? V -" V I.'"I.>" N,,",,_ ,/ T"I~280K - ......, - - Poppe "'"kI/kI '""

"'" '"" "'" 0.000 0.001 0002 (1(103 0.004 Ooo:l 0.006 0.007 a,OlE ,... ".. ,... a,lUI 0.002 D.OQot 1l.00ll DJXllI 0.010 0,012 0.014 0,016 "'h~ """ """ V ""'" ..,~ """i_ -.~ i2lJollll I;? (1l4) ..? /;;: Natural draft ""'" .:? T"l=;ZSOK ""'" -M?"" "'''' - Poppe ""'" ".. 0,001 ".. ".. "'" ,."" '"" ''"' '""Wltkw'kc ".'" "'00i '! 162al ? ? ,... ....V /'V ----V Nwral drift V V To l=290K - Morl<cl - - Poppe ""0,000 0001 0.002 0.003 0,(J)4 0.005 0.00I'I 0.007 0,008 ",,~I w,.k&*& (bs) r-.... f'--.,., , '-~ ~~-.. Natural dtaft ~ "Tal = 290K -- -Moi<d ------ Poppe I "" ~ 0,000 0.002 D.~ O,OOB o.OOI!l 0.010 0,012 0,0'" 0,016 ""

, " ~ -" ~---''--~ --,-- -- ---'-- r---...NatunJI draft T.<\=280K ~ -....." '-- -Poppe , .. Figure 0.13: Perfonnance curves of a natural draft cooling tower determined with an improved energy equation in the Merkel approach. 0.29 ./.-V --V N81:UBl dmft / k? Ta l=3OOK -M""" - - Poppe I "'" .."kWkI; o.Q2Q 0,1)300.010 ,m, ./Naturlli draft T"l=3IOK ./V -Merkel -- Poppe ./V .... V V."/ ./,' , " ~" ~.'...... "'Natural draft --..... TQI~310K ~ -Merkel -- Poppe ~ '" '" '''' '''' 0.016 ..",,," 0,(110 ,.,,.. '"" '" Natur8l dmft

"" T~l"300K ~ -"""'" ~ - Poppe "-.h.. " " '" ''''' H'" ?i " .. ?." r~':' ';' I . I I I I I I I I I / V ,,7 ,,~;-------[:/J~--V Natural draftT"l=300K -"""'"/' -- Poppe ''''' 0.010 , " ,,--V, ......-"',, .. "/ , -" , ........ V --1/ Natural draft V T a l"31OK, - M""" - - Poppe " "" ''''' " '''''''' 0.015 w,.!lWkI 0.0'10,...'" ''''' --- "" -'.

./,/- -~~-/ / V Natural draft ------T"I=300K -"""'" - Poppe , -'-, Natural draft -, '-T"I-3l0K -, -Mm. --, Poppe r---- "-r---. '-, -~k """'" w,,1l.I/klI: "'''' "'''' """,."" ''''' ,., ,.. ... " ....... ,.'" "'" ''''' 0.010 '''' '''' '''' (<4) ,."" "-.. _d"" ~ T"I=300K-M""" "- - Poppe '" " "" F"~ '" t--.... " '"NII1UJ8i draft ---. '" ,~ T"I=3IOK ... "-Merkel - - Poppe '"

j,., i .'" ? ,.. 0,010 '''' . '":tt 150 i 100 '" , ''''' 0.010 '''' '''' (ds) .,. k&Ika Figure 0.13; Performance curves of a natural draft cooling tower determined with an improved energy equation in the Merkel approach. 0.30 17.!S 0.000 O.lXI1 0,0'2 0,003 o.lXI4 QOOl5 0.006 0.007 0.008 '-.. r--... ?.,. i'-... '""'-. r-.... ? Mechanical draft ~, Ta l-290K '-M?kd l'f- - PopPe " "0.000 0,002 0.0004 o.lXlIl Q.OOll 0.010 0.012 o.01~ 001e Wit kal'tl ""~ " "'~, Mechanical draft ~ " Tal~280K ~ -MorlreI "- - Poppe ,ao '" ", ". 18,5 ~ 19.0 ci Mel;banical draft I-Tdl~290K /- MorlreIPOppe V/ .---;.:V ./ """ I-V ./ V V Mochanical draft

V Tdl~290K ./ -MorlreI -- Poppe .... Ir.D"& '"O.tD:l OCQ;! OJ)Ool 0.006 O.lXIll 0,010 0.012 0.01~ 0,0\6 '" '" '"O.(IXl OJX12 0.004 0,008 0.001 0.010 0.012 0.014 0.016 MIll ...... '" .'"~ '" Mechanical draft T.lI ~280K /" --.? - - Poppe / .//' ./ ,y ./ ,-' Mecbanical draft Tal "'280K / -Merkel -- Poppe V/' V V ,-' "'.? "" 291.5 "'0 ",0 "'. "<0 "'., 291.0 0.0:0 a,WI 0,002 0.00'1 0.(1)1 0.005 D.lXl6 0.007 1I,QCB "'hkW't1 ..' .., 0.000 11001 0.002 ODOS 0,004 0.005 0.006 0,007 D.00iI ... ,,1I.Wk3 '-.... --,~ '''>" "---Mechanical draft ~Ta l=290K -Moi?! r--... - - Poppe ~I T'" '" '"0.000 0,002 O.Q04 0.008 o.OOll 0.0\0 0,012 0,0'" 0,016 (84) Meclwlical draft Ta l"'280K ~ -Moi?! " - - Poppe .." '""-

""-~ '"0.003 0,001 O.lXI2 0.003 0.004 a.1m 0.006 0.007 O.IXIB ", .. ,," '" i, .. i E (bs) ''0, ~ ",,", ~ --,Mechanical draft Tal ~ 200K '----.: ---Mol<d "--- - Poppe , I I ..."k&fkg ?., ?.. 0.000 0.002 0,004 O,1Xli5 O,a:Ill 0.0'0 0.012 0.014 0,016 j 5.5 ?t ? S.D ? ..? '-- -. -'-'-''----J~ ---- ---- ----- '---Mcchmical draft ""--I ------T a \-280K '--MorlreI - - Poppe I " ., 0,000 IHIIl, 0,002 0,003 0.00- 0.001 a.CPo 0,007 a.roe "'II kIA<. '0 Figure 0.14: Performance curves of a mechanical draft cooling tower determined w ith an improved energy equation in the Merkel approach. 0.31 "'''' MOO .."" ... .. kI'kI .."" ? ,

" " " Mechanical draft "--Ta !"'3IOK -""""" " ~Poppe '",?.~ " .""..'"0.015 .." ....... 0,1)10 ? , , Mechanical draft " T~J=300K ? ....... -""""" i'-. - - Poppe " ....... , " "" ?.~ ~ .. ci 13 " ./ ----/ ---~../' ------ /' 1/ 1/ Mel:;hmical draftTQ ]-300K ../' -""""" - - Poppe ."" ."" .."k~ 0.020 O,lnD 0,010 ,.. I ./ Mechanical draft ./T.,l"'310K -Mml:el / V - - Poppe V /' /' ./

, ? -----:?' -- ----:/ :.------~--- --- /' / V Mechanical draft V Tal- 310K -MClkd / - - Poppe ,,, ""..~ (C2) ."" ."" ?.'" .'" ?."" II',.?I. 0.010 0.015 0.020 0.010 ."" Mechanical draft V T~l=300K V-M<rlrel ./Poppe ./V /' ,/' '"?.~ '" I I I I I " Mechanical draft ~ T.,I=300K -""""" Poppe - '" --,,- ~-- ... ,--~----~ " Mechanical draft T.I = 310 K " -M<rl<d - Poppe .

---.. --.. ----'"------- .~ -'-~ , .~ ... ??? ?."" ?."" ."" '" j"" "'" , 'm m 0,010 ,~ (<it) ."" .. I, kW'kI "'-. ", I'---'. " ..~ . '0--'- .. :::--.Ml!Ic1'lanil;a1 draft '~T.l'"300K -""""" , - Poppe "" (ds) .""."".""0.010 , ,,~ , '. ,~ ? '''-.. . " '" ::---.. Mechanical draft "'~ T a l=310K ..~ -Mo<kd - - Poppe ., , .~?.""?."" .""0.015??? ao ?.~ ?? ,.? ,.0" kgr1lg ..... Ilf'k& Figure 0.14: Perfonnance curves of a mechanical draft cooling tower determined w ith an improved energy equation in the Merkel approach. 1\1;'" 0.32

"""'" '''' "" "', "" -:::- "-.:.: ---, --I"""'"'''''' K, T~1:280K N -M?l<d ~-.~NfU ~ ,.., ci", '" N"""'_ "'.- T~l .. 290K ~- - M<ri<oI --- e-NTIJ ~- , " -r-.-.. "00 o.lXXl 0.001 0.002 0,003 0.004 O.DCfi 0.006 0.007 0.008 ""0.000 0.002 0.004 0,00 6 0,008 0.010 0.012 0.014 0.01l!. "'1.kPI W,.kWkI ""'kWkIl ... o,too O.lJJ1 0,002 o,oce 0.00( 0.008 0.006 0.007 0.008 NlItW"8I draft T"," 290 K V -M?l<d --e-NTU 11'-F -/' ~ / .r;--". 0.000 0.002 0,001 0.006 0.008 0.010 0.012 0.014 0.016 "'10 lfIkI NlIbIl'a1draft T~1~28OK l--:-"-M?l<d 1.;::"/ --e-NfU /' ./ ,,-V , ---~,.,. --;~" 191.5 "" "'..

.~. i r:-. 2!IO.0 ""'~ ""O_lXXl 0,001 0.002 0.003 0.004 0,00!:l 0.006 0.007 0,008 Natuml draft /'T~l:290K -Morirel V - - e-NTU V _VV /'V '"0.000 0.002 0.004 0006 0,008 0,010 0,012 11014 0.016 ... 1t ....1 """'"'''''' ,0' TQI~280K V -M.mI ,,7' - - e-NfU V ---V -- -/V--"" ?i 295.5 Nll.tUrllI draft T~I=290K -M?l<d ",~ - - c-NTU J;;;://' /:;/V ./:/ -;-:/ ,.." ,..., O.lXXl 0,002 0.004 0.00!l 0.008 0.010 0.012 0.014 0.018 w\t"1/IlI """ """ """'"''''''TQ I"'280K - Morlrel /:/ -- &oNfU ,/ " , ,-/ --;.---/ / --------""'"0,0lXI 0001 llOO2 0,003 0,004 0.005 0.006 llOO7 0,008 lO'ltkl/kc '"0.000 0.002 0,004 0_008 0,00II 0.010 0.012 0.014 0,018 "'I> k&o"? "1T--r~IT----;r::c::::l Nl!.\unll draft ..1----+--1---1---+--1--1 T ~ 1'" 280 K i ..0--"'-+-+--1---1-----1---1=~ ! .. ~~~~---~--""'- J---l--1-~==l==-1 (as) . ~ ? ~ l------I--+--I---"I"--'=-~---+--I-___.j"""'~?"'1-----I--+-!----f-+='h,-f---1

~" ""I-_I-_-I-_-I-_-I-_-I-_+_---l-_-I o.too 0.001 0,002 0.1XI3 0,004 0,005 0.006 0,007 o,a M',.ka/kI .. ,., "'-I~ ~ '""Natural dnU\ 1"--..T~l=:Z90K "--M""" --&oNTU Figure 0.15: Perfonnance curves of a natural draft cooling tower while employing the Merkel approach and e-NTU approach. 0_33 N__ Ta l-3ooK ,- - Morl?I--e-NTU, k ~ '--..... c-.. .... NomnI_ -......... T a l"'310K -Mok. ~ -- e?NTU ~ ~ ~ ",., ''''" 0.010 ,m '''''.'" w.,1qptcr 0.010 NomnI ,nil /'V Tal "'300K /-Mokd --e-NTU / V / V (d2) ,."'",."'" """ ,., , N__ /' T"I",310K ..-//' -Mokd -- e-NI1J ./ :;:/' V

...V /' " "" ?i" " (C2) ""'""'" ,."'"..,0.010 '" Natural dmft ./V Tal = lOOK V - Morl?I ./,-NfU /" /' ,/ Natural draft V T"I=310K 1/ -M."" - - e-NTU / / / '" """ ''''' ''''' ''''' '" "" ? '"~ .. .. ''''' 0.010 ."'" ""'" ''''" ."'" ",.,kIIka "'" ,,"m i 11m i E 11:lCl1 "'" NlIturaI draft T"I""300K -Mokd -_/~NfU J/ --;/1/ ..-/ Nlllw"a1draft Ta l=310K -Mokd ~NfU ~--...... t-------. f--. ---

(c4) 0.010 ."'" '''''' 0.010 ""'" '''''' ,,,,, "."'" "'''1I.WkI W" kW'k1l "Natural draft Tal = 300K ~. -Mokd~NfU ~ ~ ~ " (ds) ""'" 0.010 ~ ~ NIturaI draft ~ "T"I>o3IOK "-.. "-..-M_ - - e-NTU ~ " """"."'"'''''."'" 0.010 ,m ''''' Figure 0_15: Perfonnance curves of a natural draft cooling tower while employing the Merkel approach and e-NTU approach. 0.34 17.0 0.000 0.001 D.1Xl'! oem Q,1JOoI Q,IXJ!I 0.008 0.007 O.lXlIl w,,1I.w'kl Mcchrilal draft T"I=290K ,/" - ""'" ./- - e-NTIJ , V . /'...Y>:/ .' ./I-Mechanical draft T"I"'290K / -Merkel. --e-NIU V" 4" /' ~--/- ., "---, -'" '""-

Mecbanlcal draft " h T"I-290K -..........-Mclkel --e-NTU " '"'0,000 0.002 0,004 0.006 0,006 0.010 0.012 11014 0.016 ... " k&A<c ""O.DlXl 0,002 OJl.~ 0,006 O.OIJlI 11.010 0.012 0.01<4 0,0115 Wh~ " " "'t, kw'<I "11000 0,002 0,004 0.006 0.008 0,010 0.012 0.0,. 0.016 " " ~ " -.....:: ~"" ~ --Mech8nical. draft ---.::: R,-,_T. I"'280 K - Mo<'" ~ -~NTIJ Mechanical draft ~' T"l"280K // ..' - ""'" --e-NI'U /' ....~~ ........ / .'" ~ ../," " '"" '" '"" ~ 19.0 ci 1&~ 17.5 "" l-_l-_l-_l--I---,-,::---,-I---,-,::--~ 0.00:> 0.001 0.002 now ll.CIQ4 0.005 illXlS 0.007 OJXIIl ""~ '" ~., 0.000 0.001 OIXli! 0.003 o.~ D.IXJ!I 0.008 O~ OlXlll "''' I!WkI ""~~D=D29e.o ~hanical dml: T"I=280K / .......1;:; - ""'"--~NTIJ ,p~ '"" t-L-,-:.::..:;-J-t--l7:;;:r-f-----I----1 (a3) . /.V .:: ,?., t--t--..+-;:/-,CC.-?=t--+-+-+--I "'., j---b""'f'=---+--I----j--I---j--I ._::-Y' "" b""-j--+-+--I----j----jr--i--I 1I',.kI/kI. '"IlIXl:l 0.001 0.002 o,om O.ll04 OJJ(1!l o.lXl6 0.007 0.008 '-",~

'-~ " " ~ Meclwlical draft '-~Ta l",290K '-", -""". --e-NIlJ ,~ "" ""OIXXl 0.002 0.0001 0.006 D.n 0.010 WltkWkt i"" ; ?I'" :---. -- -~~ c--. -'-~Mechanical draft ~T. 1- 280 K - Mo<'" '~" l --e-NTU '" ?? 0,000 0.001 0,002 0,003 0.004 0.1)05 Wl,~ ?., O.(XXI 0.002 0,0(14 ooos a,roe 0,010 "'I,~ " "ji 5.5 ; ""~-,~ -- Mechanical dndt. ~~ Tol =280K ~--Morl?l ~ f' - - e-NTU ., ?., ~5.5 tIi 5.0 ., "" -, " ""-.... ...... Mechanical draft Ta l==290K i'--M""" --e-NTU "'r(bs) Figure 0,16: Performance curves of a ing the Merkel approach and e-NTU approach. 0.35 ",.,0,010 ""-~

0,012 0.01. o.01G

0,006 0.007 0,008 0.012 0.014 0,01&

mechanical draft cooling tower while employ

---........ "'Mechanical draft "-T a l=310K "::::: -MoO<~mu ~ , 0."" " " ''''''''' ,., 'M ? , ........ ? ...... ........ -........: ~hanical drMI: "T"l"'3OOK "- """" "-~mu i'" " " w" k8IkI Mcchri;:al dnft ./ Tq l=300K V-MoO<./~mu / v /' /' '" ''''' '.0 ,., '''' (C2) " 0 / / ./ V IV ./ Mechanical draft V Tal'" 310K -M",," ./ --o-NI'U 0,010 (d2) Whkr/kl "'''~I Mechanical drUt /T., lOOK / - M<rlool ./~mu

V 1/ /' -" ... Mechanical dJUft ./ T"1=31OK V f- -M""" ./ -- e-NTU /' ./ V / ,/' ?.,., .: '" '" ''''' 0,010 ''''' 0.000 '" ?,,,,. .: 0,010 ''''' 0.000 0."" 0."" ..........;,. " '" ""~ Mechanical dmft ~T"I"'looK - M<rlool -- e-NTU (c4) 0.""0."'"0."" w"kPI 0""0.0 "', "'" "' Mechanical draft "'--T"I=]!OK ~ -M""" ~f- - ~mu m = ''''' '" '" i~ ~2fKl ?1m 0."" ''''' 0.""0.015 "'"krJkl ,.,0."" i'", ?I'"

" ''''' "- Mechanical draft " T"l=looK "'-. - M<rlool "-.. -~mu ....... "'"'"-...... i"....... "-~ Mechanical draft "'- T"l"310K ~ -Mmbl -- e-NTU ---........ (ds) 0.""0"'"""00.010 '.0 '.0 " ''''' '0 (cs) 0."" '''' nOlll0,010 '''''' .0 " " ,., i ~.5i ~,C I " "hk&/kl "'" kI/IlI Figure 0.16: Perfonnance curves of a mechanical draft cooling tower while employ ing the Merkel approach and e-NTU approach. 0.36 "" ,,., "" "'" ....:.: -----...; --"- ', NatunlI draft .;;.-TG I=2110K K' --M<d<d ~ '-~NfU

"00 '"'" 000 """'" dmft~, T.d = 290K ~, - """"I"~ ...... --e-NTU ~~ '-, ~, "00 0.000 nall 0,002 0.003 OJlll4 O,tIOS 0.006 0,007 0.006 000 0,000 0.002 0.004 O,(D!I 0.006 0,010 0,012 o.01~ 0.(116 "10 k " '" """'" dmft T Q l"'290K /-Mo"kd ?-- e-NfU ---:: k? , .... -//" --'"0.000 0,002 O,lXW O.OOB (lOOfl 0.010 0.012 0.014 0.016 "'u kW'kc ? ? """"dmftTd L=280K v- , .. -Mo"kd l;:-'-- c-NTU /' ? "/" --? V --, V" , ? -..?".QJIXl OJXl\ 0.002 0,(113 0.004 D.lXI6 O.!Xlll 0.001 O,OD!I w"kWkI ". OJJO(l 0.002 0,004 0.006 Q,(XIl 0.010 0.012 0.014 0.018 "'hll&lkC "". r,::::::r::==C;-IT-r-lTi Natun1 draft f--+--+--+--+="-fl:-'''--IT,,=280K __~t? - ...... 1-+---'1-&'''"'''''-+-+-1--~mu -V~ "...1-'=::r=+~-",,--~~-+--I---J (a3) .: ~ "M j--b-:';--;:;-~-j--j--j--j--j---i;;;::1/ "" /-''-+--+-+--I----I--I---+--I "'. 1c-_1c----,'c-_'c-_h.----,I:::-_'c-_"---1 0.000 0.001 0.002 0.003 O.DOol 0.005 0.006 0.001 0.006 Il'hkW'kl ""

? ''''' .: Nld1mIl draft /" 1TQ l=290K -Mo"kd V --e-NTU 1/ V /, ...... ~;7-N"""" dmft Tdl- 280K -Mo"kd ,~ ..' V --e-NTU .,.. ....../ -;"/ --> .,-" ...'/ --/ ---0.0160,002 0.00<1 0.006 0.r08 0.010 0.012 no\( w\>klP'kc -;/ ,~ ~":;'-I-_-l__+-_-I-_-l__+-_-I-_-l ... ,~ ,... "'''''(14) j :; IG2ClJ , ? ,??oo ,~ ...? an....'.000 ."" .." 0??oo .... ,,"00 (bs) --=s~~- , ""'--Natural draft 1'-l'al=290K '-Merklll "-- e--NTU ,., O.lXlO 0,002 0,00'1 0.006 0,00fI 0010 0,012 0-01~ 0,0\6 (as) """"dmft TQ l=28QK '-, -M<d<d ~ - - e-NI1J -'~ '-~M" ~ " ""OlXXl 0,001 0.002 0.003 Q,OOol 0,006 0.006 0,007 0,006

w"kt!'<& w" ....... Figure 0.17: Performance curves of a natural draft cooling tower while employing the Merkel approach and e-NTU approach. The application of the fill characteristics for the e-NTU ap proach is applied inconsistently. 0.37 N__ , T~1"'300K~"'" - """" -. - ~NTII ~ -""'-. "'--. r...... -... '''''' -....... T a l""310K " -Ms!<O ~ -- c-NTU ~ ~ ~ "'" "" .. .. ''''' '.000 0,010 0.015 .." ........ '''''' "''' .. 0.010 ""kI/IllI ,,., ,."'" v -../ Ta l=300K /-_. .NTU / V ~ .""" ,,.,,..,..,., , Natural drW'l: ../ Tal "310K ../ V -M"""I --e-NTU ../ ,/ ./ V ./V r'

" "",."" '" (ez) '''''."" 0.0150,010'000 ~ ''''' .. ? I " ... l,JlI/kI NBturllidraft ./J-..Tal~300K ./ -Ms!<O ./ -e-NTU ./ l/ ,/ ,,,.,",.,,..,..0,010 , V 'I- -...-Tal ",31QK V ,I- -Mtrkel -- e?NTU / /'" ../' " " ? ''''',; ''''' ,..0.015,.,,."" .. ''''' .""? ,; "" W"kW'klI: IN__ Tal~300K -_.. ./ -&oNTU ---V~/~,/ ---;/V ;/ Natural draft Ta1 ""310K -M"""I ~NTII ~-----I'--;---. ~

jl1~ ? ?l! l1i5t'O """ """ ''''' ,."" 0,010 0.015 '''''' ,."" ,.. ~ ,."" 0.010 ,.. '''''' ,,., (c4) ,,,., "'"Natuml draft '" Tal "'300K " - """" --e-NTIJ ""--......... --.. ........... -........ NaturBi drBft "-.... ..... '- Tal ""310K ''--~ - """" --e-NTU (ds) ,,,., '''''''' 0,010 , ''''' ,"",., "'I, kt/kl 0.c10 ''''' ". ,."" Figure 0.17: Perfonnance curves of a natural draft cooling tower while employing the Merkel approach and e-NTU approach. The application of the fill characteristics for the e-NTU ap proach is applied inconsistently. 0.38 ? Natund draft .../ Tal = 290K V -Mal<' -- Poppe /' V fi

V , '" '" '"0.000 0.002 OJlO'! MOB 0,_ 0;010 0.012 0.014 0.016 ...,......., 0 Nolu>l ,nft ;,.-'b, T.l=280K r;0 - ...... /;:. -- Poppe '/, 1/' 0 ,~ ,7 " .. V0 ..-/' ? "". .,. O.CXXl 0,001 0.002 0.003 0.004 o.OltI 0,006 0.007 0,008 "'hkWkc ,,, "0 0.000 0.001 0.002 nom o.lXl4 O.lXl5 O.CIl/I 0.001 0,0Ill "'\>kf'kg Natllrll1draft Ta l=290K /' -Merkel ./ -- Poppe / V -/ ,/' ;'::' '" '" ", Q.(IXI 0.002 1I.tIl4 o.\Xl!l o.oce 0.010 0.012 0.014 0,016 ... "IlA Natunl draft I-Tal =: 280K V -Mal<' - - Poppe J" V /" .-/ ./' ,/ "M 291.15 291.0 ? ".0i l-. ?.5 I- NaIunlIdraft: /:'T.l=290K ~ /'I- -M""" - - Poppe b? " "-/V

--"/ -? ""0,000 OJJJ2 0,004 O.OOll (l.!XlIl 0,0111 0.012 0.014 0.016 "".kJl'kc ? ''''~ 0 ? Natural draft , Ta l=2.B.OK /;0 -"""'", Pop,. .</ 0 '<..., ,>, ... "/0 '/, /'0 '" ", tLOOl 0.001 0.1;02 0,003 0,Q(l4 0.005 0.006 o.iX17 o.OllIl ",,,kwq. '" "'. ..I- Natunddndt J .... :,..--T a l=290K VI- -Merkel " -- Poppe ;":;V --"",.-'-1.:/ -- V """ """ ",." ",ro ,... ().(Dl 0,002 0.004 0.006 0,008 0.010 0-012 Cl014 0.016 ..... Ir.~ j =11100l, ? N,,,,,nol""" , " Tal =280K ", '/' -""''''' ".' ..- - Poppe .' , " V , /' V .., /' --/ /' ,..., """ ,..., ,..., ",ro

o.cm 0.001 0.002 0-003 0,00<1 0.006 11.006 0.001 0.008 w,,~ ,..., ,... j Ill:!OO ; ?''''''? (bs) -------------~ J-------__ M_ -Nl!ItUtlI1dndl. ----""'Tal ~ 290K '----- ...... -- Poppe '" , ""0.000 0.002 0.004 0,008 0-008 0.0\0 O.OIZ 0,014 QOla "'I!"wk& '" '--'--r'"--'--'-"---'---'--" ., j--c~--j.=c:-1--+--+--J---+---j--4 Figure 0.18: Perfonnance curves ofa natural draft cooling tower where the heat r ejection rate is constant. 0.39 .."'".'" 0,010 w, '"'. 316.01- -1- .... -+- -+- -1 .."" ~17.5 t----+----+---+----+----1 "". rr.;==;:;c.::;----r-----r---r~N"""",oft V-" T"I = 310 K I--+----t,-/:__,,-~,,?jC---< -"""'" /:",-" - - Poppe 1--+-/~-4----I----1 ? /,~--- (d) ,f ".., j----t >5"'f---t---t----J 1 _&/ --' " "'. t-.."...-c/7'I'----I---+----!-----! ,/ ?."".""0,010.... NoIwol ''''' I""T~I"300K -M,<lret V" "" - Poppe , /' k?-' .-L7"""

---::> , .." '" .."" '" '" '" .'"i ? NatunlI draft T"I=300K /'- .......Poppe /' /'./ -;:::: --."'" ,~"',. '''' 0.010 Natural draft ./" T"l-JIOK /' - """" -- Poppe ./ A'r // ", ----'" "","" '" "'" .."'" ,.,0.010,..""?."" "" ,"" '''''..'" 0,010 ? 1 ,/01- Natural draft ,T"I=lIOK ./-M_ -,,'I- -- - Poppe -;? ....."" "..-" / ---/'" "","" '" '"?, ~ 0.016.... ? ,b>'-N""",_ T,,1=300K -:7 - ....... -- Poppe

'/......... _......,....;; ,~,,,~' V , -, .. ' ,/ ./ " "" .. ,. kI/k? ,,"" ,"" " " V.--:;;;:V-~--- , NaP.nId~ ", " / T"I-3OOK - ....... .- - -Poppe I Natural draft T"I=3IOK " - """" --V Poppe , , //V, , ---/ y --, ---/ ",,,, """ j 140Xl i ?",," ",," 0.010 "'" ?."" "'" ,,"" ""'" j12C1OO ; , IE 11000 "'" .."" 0.010 .'" ,,,, (t4) ,."" (ds) ?."'",,,,?.""0,0\0 '~~-- -', " ---.::' ------~ "', ...-"-,

Natural draft ~ T"I-310K ~ -Morl<d - - Poppe '","" (cs) ?."""",..'"0.0150,010.... ~' '" -, ~ " .....,,-~ " '-~ -- -----, ,-Natural draft T,,\- 300K ""-"""" '-.: - Poppe "----I '" ."" j"" i , "" ? WI, kgr1ql Figure 0,18: Performance curves ofa natural draft cooling tower where the heat r ejection rate is constant. 0040 Mechanical dnd't Ta z-290K V -Mo'<d V - Poppe V ...,/ '" /' ""0.000 0.002 0,0l'I4 0,008 0.008 0.010 0.012 0,01-4 0,018 ~,'" '" ,,, '" ,,, ?J 313 '" , , Mcchanicu draft TQ 1=280K -./' -Mal?< ,/, - - Poppe ,/,, , -.:Y , -'</ :::, " " , , 310,'

"" "" 311. "'.O,IXI] 0.001 llOO2 llCUl 0.004 0.005 0.006 llOO1 0.006 "'1.kI/kI 31\. 310. .""~"" Mecblmicu draft T Q I=290K V -Mo'<d - - Poppe V ./ / ./ ./ '"CUlOO 0,002 0.004 0.006 QCIl8 0.010 0.012 001~ 0.018 ~,"" , , MIlCblmical draft , Ta l-280K ..., -Mal?< - - Poppe V, V V , V V -" ?~ 291.:1 :291.0 ~ ~, ,." llooo 0.001 0.002 0,003 OJXI4 0,00fi 0,006 0,001 M08 w" Iqp'kc ,." ~ 0.000 0.001 0.002 0,008 0.004 0.005 0-008 0,007 0,008 .. ""-,,,& Mcch8nical draft , "Ta l=290K ..,I~i:/ -Merkel V -- Poppe -,' " '/ " ----V ~~ .../ -./ ./ '"llCOO 0.002 0.004 0.008 0008 0,010 0,012 0.014 0.016 ?

-""~ "" Mecbanical drBft TQ l"'280K ,- ,-:/v -M"d - - Poppe ~~,,- ;.--, -;/ ..- y " -;/ "" "" ,.,--,----,---,-,--,--,-,--, - "'"'----........ --.:::Mechanical draft -~ TQ !- 290K -~ -Mo'<d --~ - - Poppe .... r:::: '" "" .. 0.000 0,002 0,0lM 0.(Xl6 0.008 0.010 0.012 0.014 0,016 ~,",-,---,----,--,-,--,-,--, , .......... "-.........~ --::::- C-Mechanical draft -"---, Ta l"'280K ---0-Mal?< --- - Poppe "" "" '"O-lIXI 0.001 0,002 0.008 0.004 0,005 0.008 0,001 0,l;1li whkWk. j"" ; ."", ?., ",., " j-~-+--I---+--t--j---+--c::b--~--~-"j--+--1----I--I---+-+--+---I~ ?? j--,--- ---'!''''--.k::-+-+--j---j--j-----jr?.

, ----r--- (as) I; Mechanical draft I---- t-...~ T Q l = 280 K I--+-,I---+--j-""""I-~~ -Mal?< - - Poppe ?.,l!:::::::l="=+:='---L-L-l--l-J-j 0.000 0.001 0,002 0,003 0,0l'I4 0,00fi 0,006 0,001 0,008 ..",," "/---''''-;:1---+---+--+-+-+-+--1 " j---f-"-'-c-'-+o:--f---+~-+----I--+--1 "-- --".. ,~-+--'>k:--:P~""=c-j---+-~I---l , "-- "---j ,.I---+----+-~'--__f-t---+--+--I (bs) f"~~~+=1=~I !I,2 +-rM~lwlnical draft ........ Tal" 290 K I---j--j---'t'--,"-.c-f----j--! -Mo'<d ~ ..t~:~:r-~po~p~PJ'~~j:==t=j==t=~t=j ?? a,coo 0,002 O,lXI4 0.008 0,00'1 0.010 0012 0.0\4 0,016 ~,Figure 0.19: Performance curves of a mechanical draft cooling tower where the he at rejection rate IS constant. 0.41 /' "~hankaldrVtTa l=300K /' , --Poppe /' /' / ...M".."",... Mechanical draft 4Tal ~ lIOK ./-' -, -MoI<ol ~" - - Poppe , ...... J' --'" "". ? /' ... /' '" '" ". ,""..,c,mo.."" '" ,.., '" ". ". ?i " "' .. 1lWk1 Mechanical draft Tal-lOOK .../' --

-- Poppe V V V /' Me<:;hllnkal! draft .-/' Tal =310K ./ - """'" V - Poppe /' / ./ /' /' ...,,., ''''' ...0,010'"'.000 '" ", '" '''''''''' 'W''",."" "" ...,,.,0.010 ? ? ,/ , -",-;?' , --/', , ...... - 1/ ----/' 1/ ~lwUcaldraftTa l=310K / -Mokd / - Poppe '" .. ''''' ? ''''',: ,.""",.?.""0.010."" ? Mechanicl1 draft ,Y' Ta l=300K ~ -"""'. ---- Poppe V , -;-~--~ , " V~"", / / '" .""

"', i'.. - '--~ ~- --, '----------- -- . Mechanical draft Ta l=300K ~ -Mokd -- Poppe " ~ -, ~ , ~-. -- , ",~ Med:lanical draft ..~~~~~ Tal-JOOK ~ -MoI<ol -- Poppe ~ (<4) (ds) ,... .... ,,., '''''''''' 0.010 .., " M?:hanical dnft " Tal~310K ~ -Mai<d -, '" - Poppe , ---,~ "":-,,-'" -----. ', Mechanical draft ~ Ta l""'310K-""",,, , -- Poppe -" ---~ ""'-"~ r::::,.. '-,~~"'-. ~_75 .. '000

j 6.75 i 6.&1 Ii 6,215 i"", ?1 m (es) ,."" ..""",,, "'. ?."" ?."" w"kJIkg 0.010 0,010 ."" ''''' '" " '" ."" ,. ., "'" .. ., j 6,0 fM ? '."kW\<I Figure 0.19: Perfonnance curves of a mechanical draft cooling tower where the he at rejection rate is constant P.l APPENDIXP WET-COOLING TOWER PERFORMANCE EVALUATION SOFTWARE P.I INTRODUCTION A computer software program, Wet-Cooling Tower Performance Evaluation (WCTPE), i s developed to analyze the perfurmance of counterflow and crossflow wet-cooling towers. The gra phical user interface of the software is developed in Visual C++ 6 while the program algorithm is deve loped in the Fortran computer language. All the models and equations cited and derived in this thesis pertaining to wet? cooling towers are included in the software program. For counterflow cooling tow ers the program is essentially a one-dimensional approach, that yields results orders of magnitude faster than full-blown two-, or three dimensional computational models involving the continuity, moment um and energy equations. However, the two- and three dimensional nature of the problem is acco unted for in some ofthe semi-empirical relations, such as those presented in appendix D for the loss and transfer coefficients of the rain zone. The sample calculations presented in appendices I and J for the natural draft an d mechanical draft wet? cooling towers respectively are examples of the solution process of the software program. Due to the iterative processes involved throughout the solution of the program, mathematica

l control measures are applied to prevent numerical instability and hence divergence ofthe solution. Warnings that occur during the solution process are written to an output file. S ome of these warnings occur when empirical relations are employed outside their range of applicability according to one or more variables. Warnings also occur when convergence of iterative processes is not at tained in a specified maximum number of iterations within the specified solution tolerances. There are more than fifty different warnings and a possible remedy or remedies are given for each warning that is written to the warnings output file. Some functions and variable inputs of the program are disabled fur certain choic es made in the program. This is done to make the software user friendly and to prevent confusion, as onl y the active parts of the program requires input from the user. This appendix is not intended to be a deta iled user manual of the program, but it rather gives an overview of the basic architecture, functions an d capabilities of the program. P.2 ASSUMPTIONS AND SIMPLIFICATIONS It is mentioned in the section above that the program is virtually a one-dimensi onal model of cooling tower operation. This can only be achieved by introducing assumptions and simpli fications such as, ? The cooling tower operates under steady-state conditions without wind. P.2 ? Miscellaneous thermal loads such as make-up water additions, pump head gain an d the net heat exchange with the ambient surroundings are negligible. ? Uniform air and water flow rates over the tower cross sectional area. ? For counterflow towers, the thermodynamic properties of the upward airflow and downward water flow vary vertically, but are constant across any cross-section inside the tower . P.3 PROGRAM AND SOFTWARE DEVELOPMENT Figure P.I shows the main dialog window of the computer program after the progra m is executed from the Windows? environment. Figure P.I: Main dialog window of the WCfPE computer software. The toolbar on top of the dialog box consists of nine different buttons. These f unctions on the toolbar can also be accessed from the buttons presented in the bottom part of the dialog win dow. Dialog windows for the specification of the atmospheric conditions, tower specifications, solution control, loss coefficients, transfer characteristics, heat and mass transfer model settings and fan specific ation are accessed from the main dialog window by clicking the appropriate buttons with a computer mouse. Al l the data entered into the program can be saved in files with user specified file names. P.3 P.3.1 AMBIENT CONDITIONS The dialog window where the ambient conditions are specified is shown in figure P.2. The ambient air temperature, pressure and temperature lapse rate are entered in the top left han d side of the dialog window. The atmospheric hwnidity can either be specified by supplying the wetbul

b temperalute, relative hwn idity or hurn idity ratio. Figure P.2: Dialog window to specify ambient conditions. Different options of the vertical atmospheric profiles of temperature and humidi ty can be specified. The characteristics of a stable boundary layer, i.e.. when a temperature inversion i s present, are specified on the right hand side of figure P.2. The characteristics include the height of the ground-based temperature from which the temperature profile is extrapolated. The height of the inversion can either he specified or iteratively determined by the program. The height of the iteratively determined inversion height is a function of the time elapsed since sunset, the thermal eddy diffusivity of the a tmosphere and the maximum daily temperature. The maximwn number of iterations and tolerance for th e iterative process are also specified. Refer to appendix L for a detailed discussion on the stable boundary layer and the variables associated with it PA P.3.1 SOLUTION CONTROL The dialog window for the control of the solution is presented in figure P.3. Fi ve variables, i.e., the water outlet temperature, the air temperature and pressure above the drift eliminator, the internal pressure at the top of the cooling tower and the mean air-vapor mass flow rate, are chosen as th e primary solution variables of the primary iteration loop of the program and are solved by the Jac obi iterative method. All the other so-called secondary solution variables are either explicitly, or itera tively solved from these arbitrarily chosen primary solution variables. The iterative method utilized for the solution of the secondary variables is the Secant iterative method. Refer to numerical analysis textbooks such as Mathews [92MAl] and Burden and Faires [97BUl] for detailed descussions of the ab ove mentioned iterative schemes. Thus, convergence of the program will be reached when all fiv e variables mentioned above change less than the specified tolerance, speCified in figure P.3, from on e program iteration to the other within the specified maximum numher of iterations. Figure P.3: Solution control dialog window. In order for the program to start the iterative process successfully, practical initial estimates must be supplied for the five chosen solution variables. Either these initial values can be supplied by the user, or they can be automatically estimated by the program. Refer to appendix 1.2 for th e procedure followed to initialize the variables. I"? P.5 As already mentioned in the introduction of this appendix, mathematical control measures must be implemented to prevent instability of the iterative process. One way of preventi ng instability is the implementation of relaxation. In the iterative solution of the algebraic equations, it is often desirable to s low down the changes, from

iteration to iteration, in the values of the dependent variables [80PAI). This p rocess is called undeITelaxation. DndeITelaxation is often employed to avoid divergence in the it erative solution of strongly nonlinear equations. In the following discussion A is an arbitrarily ch osen variable and B is an arbitrarily chosen function where A =B . IfA' is added to the right hand side and subtracted find, A=A'+(B-A') A' is the value of A from the previl;lUs iteration. The contents in the parenthe ses represents the change in A produced by the CUITent iteration. This change can be modified by the introduc tion of a relaxation parameter, a, so that A=A'+a(B-A') When the iterations converge, A becomes equal to A'. There are no general rules for choosing the best value of the undeITelaxation factor. The optimum value depends upon a number of factors, such as the nature of the problem and the iterative procedure used. For this program, it was found that relaxation factors of 0.1 for all the selected solution variables prevented divergence for all the sample cases investigated. Another principle to prevent solution divergence is implemented in the program a lgorithm. No control from the user, however, is necessary. For Jacobi-type iterative schemes to conve rge to unique solutions, the arbitrarily chosen matrix C in the linear system C?x = D must be strictly di agonally dominant [92MAI, 97BUI). Patankar [80PAI) refers to this condition applicable to numerica l heat transfer and fluid flow problems as the Scarborough criterion. Thus, the applicable equations in the program algorithm are manipulated to satisfy the Scarborough criterion. P.3.3 TRANSFER CHARACTERISTICS The dialog window to specify the transfer characteristics in the fill, spray zon e and rain zone is shown in figure P.4. The transfer coefficients can be specified either by constant values or by empirical relations. The empirical relation for the spray zone is given by equation (D.23). The empir ical relation for the rain zone is given by equation (D.20) for circular towers and by equation (D.22) for rectangular towers. In addition, the transfer coefficient of a purely counterflow rain zone can also be specified. The empirical relation for the purely counterflow rain zone can be found in Kroger [98KRl). Th e empirical relations for the spray and rain zones are not applicable to the crossflow cooling tower. P.6 Figure P.4: Dialog window to transfer coefficients. The empirical relations for all the transfer characteristics are determined by a ssuming that the cooling water is distilled. In practice, however, the cooling water can be contaminated by various kinds of minerals and impurities. If the evaporation rate and surface tension of a sample of the actual cooling water is determined, the corrected transfer coefficients can be determined. The properties of the actual cooling water, compared to distilled water, can be entered in the bottom left?

hand side of figure PA as percentages. The percentages for the evaporation rate and surface tension are the percentages of the cooling water to that of distilled water. Equation (0.20) for the transfer characteristic of the rain zone in circular towers and equation (0.22) for recta ngular towers are functions of the surface tension, ,,", through the a", a v and aL coefficients specified u nder equation (0.8). The correction of all the transfer coefficients for the surface tension is implement ed as follows. The transfer characteristic for the rain zone is calculated with the surface tension of disti lled water. The calculation of the transter coefficient of the rain zone is repeated with the corrected surface tension of the actual cooling water. The percentage change of these two rain zone transfer coefficients are th en applied to the fill and P.7 spray zones. The total transfer coefficient is multiplied. by the specified evap oration percentage, to account for higher or lower evaporation rates, ofthe actual cooling water compar ed to distilled water. The empirical relation for the transfer characteristic ofthe fill is given on th e right-hand side ofthe dialog window shown in figure PA. The empirical relation can be selected either from a database of49 different counterflow and 7 crossflow fills, or it can be specified by choosing the approp riate form ofthe empirical relation and then specifying the coefficients. The empirical relation for the lo ss coefficient of the fill is also specified in this dialog window. The transfur characteristics in the databa se and the transfer characteristics of fills given in the literature are generally according to the Merkel approach. The transfer characteristic of the fill can be adjusted in the bottom lefl haud side of figur e P.4 to be suitable for employment with the e-NTU and Poppe approaches. P.3.4 COUNTERFLOW TRANSFER MODEL SETTINGS The settings for the counterflow heat and mass transfer models can be selected i n the dialog window shown in figure P.5. The partiCular model ofanalysis is specified at the top oft he dialog window. Figure P.5: Dialog window to specify counterflow heat and mass transfer model se ttings. P.8 For the Merkel approach, the numerical integration algorithm can be selected. Th e four point Chebyshev numerical integration method is the preferred algorithm for cooling tower analys es [88BRl, 90COl, 97COlj, but the Simpson algorithm is also included in the program for comparativ e purposes as the number of intervals can be specified for the Simpson integration algorithm, to o btain very accurate approximations of the integral. The energy equation applied in the? Merkel approach, to calculate the air temper ature above the spray zone, can also be chosen. The common energy equation does not account for the ch ange in the water mass flow rate due to evaporation, while the detailed energy equation does. This cons ideration has far reaching implications for especially natural draft towers, where the draft through the to

wer is a function of the air temperature above the spray zone. The e-NTU approach is employed in a secondary iterative scheme inside the main p rogram algorithm. The variables specified for the e-NTU approach are parameters to control the Sec ant iterative procedure. The Secant differential in figure P.5 is a parameter to determine two initial ap proximations for the Secant iterative scheme. Similar Secant differentials will be required for the counterf low Poppe approach as well as the crossflow Merkel and Poppe approaches. The Poppe approach is also employed in a secondary iterative scheme inside the m ain program algorithm. Iterative control parameters are specified which include the maximum number of i terations and solution tolerances for the water temperature and outlet humidity ratio. The number of in tegration levels can also be specified. For example, 2 levels are chosen for the Poppe approach employed i n the fill analysis in appendix G, and 5 levels are chosen for the Poppe approach in the analysis of th e natural draft cooling tower in appendix L The governing equation of the Poppe approach can be solved by different solution algorithms. The governing equations presented in appendix B can be solved by an iterative Secant algorithm, or the governing equations can be manipulated and solved explicitly. The different appr oaches can be used for comparative studies to evaluate the accuracy of one algorithm compared the other . If the water inlet temperature, T. i, is known and the heat rejected, Q, is unknown, the governing equations are in a different form than when Q is known and T. i is unknown. That is why the explicit or itera tive algorithms can be separately specified, in figure P.5, for each instant where T.i is known or unkn own. The Lewis factor, discussed in appendix F, must be specified when the Poppe appr oach is employed. The Lewis factor can be specified by the equation of Bosnjakovic [65BOlj given by eq uation (F.16). The Lewis factor can also be specified as a constant, or it can be determined by equ ation (F.14), where the exponent given in equation (F.14) as 2/3 ~ 0.667 can also be specified. The Lewi s number, Le, in equation (F.14) is determined by equation (F.6) where the thermophysical propert ies, k, p and cp are solved according to the equations in appendix A and the diffusion coefficient is solved by equation (F.7). P.9 It is highly recommended that the same definition of the Lewis factor be employe d as was the case when the transfer characteristic ofthe fill was determined. P.3.S CROSSFLOW TRANSFER MODEL SETTINGS Figure P.6 shows the dialog window for the crossllow model settings. Due to the two-dimensional nature of the crossllow problem as seen in appendix C, the implementation of the transf er models in the program differs from the counterllo,,: problem. Figure P.6: Dialog window to specify erossflow heat and mass transfer model sett ings.

P.IO The settings for the Merkel and Poppe models in figure P.6 are parameters to con trol the internal Secant and Jacobi iterative schemes associated with these models. The iterative method employed is discussed in the last paragraph of appendix C. The number of intervals specified are equal in the horizontal and vertical directions. Refer to figure C.2 fur an example where the number of inte rvals is chosen as four. P.3.6 COOLING TOWER DIMENSIONS Figures P.7, P.8 and P.9 show the dialog windows where the cooling tower dimensi ons ofmechanical and natural draft towers of counterflow and crossflow configuration are specified. I n addition to the dimensions of the cooling towers, it is also specified in each of these dialog w indows whether the heat rejected or inlet water temperature is known. The water mass flow rate is also s pecified on the dialog windows shown in figures P.7 to P.9. Figure P.7: Dialog window to specify counterflow mechanical drafI tower dimensio ns and operating conditions. P.II Figure P.8: Dialog wiodow to specify natural draft tower dimensions and operatin g conditions. P.3.7 LOSS COEFFICIENTS Figure P.lO shows the dialog window where the loss coefficients are specified. L oss coefficients can be explicitly specified or be determined by an empirical relation if one is availab le. The loss coefficient of the fill is specified in the dialog window shown in figure PA. P.3.8 FAN SPECIFICATION For mechanical draft cooling towers the fan can be specified using the dialog wi ndow shown in figure P.ll. The fan static pressure, fan power and fan efficiency can be specified by sixth order polynomials. P.12 ./ // ///,/, .. [:~T:a~::;-K](" a p.625:Ee. W --------~~~~ IHXIO Wi 111 Figure P.9: Dialog window to specifY crossflow mechanical draft tower dimensions and operating conditions. Figure P.IO: Dialog window to specifY loss coefficients. W~f dislrbuti:::ln SYSlem-'l Kwd 10.5 . . ,,_ ..., ...,--,._-_...._.?._,-_._-Spr~ zone Io?et-----'l r Spe~K'P J' i r. EQUatnn 7.2.20 I_~~~-_---r E........nsion Iosse;---'-'I ~ ? , (" Specily Kcte It_.:_:~?~ r.-Iriet Iol1'lre Io?eS-~-l ! K~ 125 L ....l Fi'"::I~~? .:]

1- F~~;-~~~----I 10 I r. Empiical relation I __, ~ .._ .._._.J r- Rain zone 108888"-- ..----, I r Specify Kr2 i .. E",!n7.2.18 II r Include countefflow section. H $peciied 81: bMisfer UlecificatiJns. [ T~~::~f~::..:::e,-'"_-_'---' 1? r. Ecp.laion 7.3,15 -,._---_ ..~,--------_._If' I I Dr?]rrJ8(er ':'( :;,urrJjQ!( 10 fn I _~.I:~C~'fflC'''' __.~o 1 rT~=,Icm"-~'--~l 10 I r E",,",," 7.1.21 r?ll,mbe.. Q~ SUPp01lc. lD P.B The fan and fan casing dimensions are specified as well as the fan model and ref erence conditions. The fan operating conditions are automatically corrected by the appropriate fan laws . A database of fans can be built by saving the fan specifi~~tions entered in the dialog window shown in figure P.11. P.3.9 COOLING SYSTEM OPTIMIZATION The geometrical dimensions of natural draft cooling towers can be optimized for the minimum combined operational and capital cost over a selected project period. Refer to appendix U for a detailed discussion of this procedure. In addition, the optimum fill height can be determined for al l three types of cooling towers presented above. This simple optimization procedure has only the fill hei ght as a solution variable. At the optimum fill depth the maximum amount of air will flow through the tower while the water is cooled to the minimum temperature for the specific fill height. Figure P.ll: Dialog window to specify fan. P.4 CONCLUSION Due to the simplifications and assumptions made in the development of the softwa re, the program has its limitations. However, it is a very useful tool to predict cooling tower performa nce. It is also a very useful tool to conduct parametric studies of cooling tower performance and behaviour. P arametric studies can be conducted quickly and efficiently with the maximum control on the solution proce ss. P.14 Figure P.12: Dialog window to specify optimization settings. Q.l APPENDIXQ COOLING TOWER PERFORMANCE CURVES Q.l INTRODUCTION Software is developed to generate cooling tower performance curves for different operating and ambient

conditions. A cooling tower performance curve is a graphical tool with which coo ling tower performance can be predicted. The performance, i.e., the water outlet temperature or cooling range, can be graphically determined as a function of the ambient temperature, relative humidity, water ma ss flow rate and water inlet temperature. Cooling tower performance curves can be generated for any coo ling tower that can be specified by the WCTPE software program presented in appendix P. This appendix d oes not serve as a user manual of the software, but illustrates the basic functions and capabilitie s of the software. Q.2 PERFORMANCE CURVES GENERATOR SOFTWARE Cooling Tower Performance Curves Generator (CTPCG) is software that aids in the generation ofcooling tower performance curves for different operating and ambient conditions. in addi tion to the water outlet temperature and cooling range the air outlet temperature, tower draft, heat reje ction rate and water evaporatrion rate can be determined from the cooling tower performance curves ge nerated by the CTPCG software. Figure Q.l shows the graphical user interface of the cooling tower per formance curves generator computer program developed with Visual C++ 6. The program consists of four basic steps. Firstly, the ranges of the ambient air temperature, ambient relative humidity, w ater mass flow rate and inlet water temperature are specified with the number of increments across each variable range. The water outlet temperature. is then automatically calculated for each specified operatin g condition using the WCTPE computer program discussed in appendix P. For the 28 increments of the amb ient air temperature, the 9 increments of the relative humidity, the 6 increments of the water flow rate and the 28 increments of the water inlet temperature, 58870 operating points are calculated by the WCTPE computer program. Step 2 of the computer program defines and calculates the global x-axis and y-ax is coordinates of the 58870 cooling tower operating points. The straight lines of the relative humidit y and water mass flow rates are also defined in this program step. The operating points must be conver ted into a structured grid format to reduce the number of data points and to preprocess it for the contour plot generator. The Gnuplot [99WII] software program is used to generate the contour plots. The dens ity of the structured grid can be defmed as seen under step 3 in figure Q.l. The contour data points of the water outlet temperature and the cooling range ar e generated in step 4 of the cooling tower performance curve generator. Contour curve smoothing can be ob tained by selecting the Bezier smoothing option. The global x-axis and y-axis ranges for the perform ance curves are defined in this program step. The global y-axis coordinates are expressed in terms of th e water inlet temperature. Q.2 The global x-axis coordinates are expressed in terms of lbe ambient air temperat ure. The data points of

lbe performance curves are written to? output files lbat can be imported into Mi crosoft Excel, Gnuplot and Tecplot. The variables to plot, in addition to lbe ambient air temperature, rela tive humidity, water mass flow rate and inlet water temperature, can be selected under step 4 oflbe progra m. The heat rejection rate, Q, lbe tower draft, m~1S and water evaporation rate, mw("ap? must be plotted on a different graph lban lbe water outlet temperature, cooling range and air outlet temperature. The reason f or lbis will be discussed in lbe next section Figure Q.l: Graphical user interface for cooling tower performance curves genera tor program. Q.3 COOLING TOWER PERFORMANCE CURVES Figure Q.2 and figure Q.3 show lbe cooling tower performance curves generated by lbe CTPCG computer program for lbe natural draft cooling tower specified in appendix I. However, lb e Merkel approach is employed in lbe generation of lbe performance curves instead of lbe Poppe approa ch. The sample calculation of the performance evaluation of the natural draft tower according t o lbe Merkel approach is given by KrOger [98KRl]. --~-~~-------~. - """" Water outletlemperature, Two, ?C -9 w 19 20 21 22 23 24 25 26 27 28 291817161514.~ . - . - .. ._I I ~ j"oo!-'l "< l(" .... l,..o7 v i.-' i.o' V ~ II.- ~ '.' , , , , , , , "" " I i-" V k [;0'" Air temperature above fill, Tao, ?C -_31 .... u,.1-- .... .... r, .... k'l it" ~ l..o I", I!( .... " _30 ....~ I-- .... .... fo'" .... "'1 v-C I) /(" Jl ~ i.-' "I- v I>V .... j"oo'" i>' Y "" V29-1\ .... 1\ l..I. I.; L.o 1.-3 lo-: .' 1-1- 28- I-- ~ l>' m( '" ~i-" v J.' Y "'" I).- it" ]);1-1- __ .... 1...1-- Il vI'. J.,l 1)0 li.- I-- Jr l..- I!'. I....... . l,.o I.... V- I 1-1- 26 10- H1~ ~-:: L'I-' 1\. ..... 1""1 vI"" i-" ~ Jr l)( l>- I.- [;0 Y J> I-.....~"'" I-- I-1-1- 25 l-' ~~ ~, ... I\e-1'1\!.<'V ..... I-- ,\ I-- l..- i/ ~ i t"I1-1- 24 23 22 21 20 19 18 1 16 15 14 13 12 11 10 I- _ Cooling range, ?C I I I I I I I Water mass flow rate, mw, kg/s a 0 a a 0 o 0 ot,()ltJOO '" 0 Relative humidity, %ltJt--~g~ "- 0 -t--CO~~~ C") '" 100 1.; ... ~; fo" ... 90... 80 10-'" ... fo"l0- 70 ... t::fo" "'j,oofo" ......... 10-'" 6050 ....... j,oolo- ...... fo" ....... fo"i""' ... "'j"oofo" 4030 ~"'I::: ....v ....... .... ........ .... .... ....fo" ....... 20;.. ~ ...15 ;:::

~",,-= :.. I-'" .... ... .... j"oo ... ....fo" .... ... 10 ;"'r;;,.;'" ....... ............ .... 1-"' .... .... r;:: .... ... .... .... ... . .... ....1-'" '" 1-"' ... .... 1..0 ........ ........ ........ .... ....1-"' ........ .... .... .... 1..0 .... .... ....... 1..0 .... ....:.. ....... .... 1-' .... .... ......... ....... I- I' I" .... 46 ~ 'j 44 Ioj ':; 42 '? <Il C. 40E 2 ]! 38 .S ~ 2 III :s: Figure Q.2: Cooling tower performance curves. 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Ambient air temperature, T," ?C -1?I-+~ I III I I I I I I I, : I I I I I, I I I Relative humidity, % 100 ~~ ... ro' ~ ......~ 1--'1-" 70 1--'''' ~~"'I--~"'I--'''''I-''~ : ~ ... t-I-" ...1..- ....1-"...... ...... I..- ... I-"I-"~ 4030 r:;.j;' I--' I-" ...~ ...1--' ...~ ...1-- I--' I..- ...~ ~~ 20 10~t ...... ~I--''''_~''' ...~ I--~_I--''''I..-~ I-- .... ~I--' ... ~ ~ r.,... ~~ Heat rejection rate, Q, MW U 1800 1700 1600 1600 1400 1300 1200 1100 1000 900 800. 700 600 ? 46 i j-.-I I I' L"" .1.... 10' If" j;. 1 I&!~V v~iI 1.'1.o!1/ V 16 "11 151.;' LI .. .. t.l1f$ !I I I I I I I I I I IH--j r-. 520 ~l7vl1l.; 11.'-" 'f'.I ...~ ~i-"~ ,(I..o1t::: [,J, ~I.;'I" V II',1V ~ 1/ ,/ 500 "'44 f- .a j-.- 500.- 14V ~II:"" I""" VLJ.... I~ ~ ~ JL "'rj V ~ '. !') .I.(~" .~ ... II I'" /.'1 ~ ,.... I~I" v. V I""V I.-! I, 11')1', II' II? Air-vapor mass flow rat e, m", Mgfs '" 42 f- 480 ~rl L.< V -- ""V F- VIr ~ I.. y 1....1'/ V", , '-+-+t+--I-+-~ j-.460 kI IAJ ViJ'~ II.V ~~ ... 1:1' L.. ....,...~r.r ~r;,I" l/~ 1/.:1 1" I~IV 1~ 1 ~40 ' V~ ~ ~ V ~ VV~ ~ ~VI/Y ~~ ~ iJ'V~ ~~ ~ 440 I.' tA l.!l~ iJ' ~ L; f1~ j,>' i'!A~ J.; V 1;1 l' ]I. '" L"II" 1'0' If IfV II

.- j-..& 38 j-.- 420 'II ... kjll'" vI/I'l/1''' Li' V V~ i." 1,,1/ ~ ~~I .... ~ I/. I. .. " I"" V ~ f- 400 380 360 340 320 300 280 260 40 220 200 180 160 140 120tttttttti t-tj-.- Water evaporation rate, mw(e"pl, kgfs I ?lli?lli?lli I I I I 1 j-.I Water mass flow rate, mw, kgfs I II/II! 11 I" 00000 01001C)000 O ..... IC)("I,lOIOQ 10 C") N ".. 0 ..... Ll)....... ..... .,.... ..... ...-- co I''"o I~n.. r.Figure Q.3: Cooling tower perfonnance curves. 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Ambient air temperature, Te" 'C IL Q.S Figure Q.2 shows how the water outlet temperature, cooling range and air outlet temperature are obtained when the ambient air temperature, relative humidity, water mass flow rate and wa ter inlet temperature are known. Figure Q.3 shows the how the water evaporation rate, air-vapor mass flow rate and heat rejection rate are obtained when the ambient air temperature, relative humidity, water mas s flow rate and water inlet temperature are known. Note that the water mass flow rate lines in figure Q.3 are inverted from those presented in figure Q.2. This is the reason why it is necessary to present the performance curves on two different graphs. Table Q.I: Comparison of performance curves and Krllger [98KRlj Variable Performance curve Krllger [98KRlj Two,oC 21.2 21.376 Range,OC 18.8 18.624 TeO,oC 26.0 26.4219 Q,MW 910 972.713 m~", kg/s 16S00 1684S.4 mw(""P), kg/s 28S 308.S173 Table Q.I shows the results, using figure Q.2 and figure Q.3, of the tower speci fied in appendix I while employing the Merkel approach. These graphically obtained results are compared t o the results of a sample calculation of the same tower given by Krllger [98KRI]. It can be seen fr om table I that the outlet water temperature predicted by the performance curves in figure Q.2 is less than 0.2 ?c from the value determined by Krllger [98KRlj. The heat rejection rate and water evaporation rat e is approximately 7% less than the values determined by Krllger. The reason for this discrepancy is t hat the cooling tower performance curves are generated from averaged data. The WCTPE program practical ly gives identical results to those presented by Krllger [98KRlj. The performance curves are nevert heless a useful tool to predict the approximate cooling tower performance. Another application for the CTPCG program is to generate performance curves of c ooling towers for comparison to the original performance curves supplied by the cooling tower manu facturer. If these two

sets of curves compare very well, then the WCTPE program can be employed with co nfidence in other studies of the cooling tower. Q.4 CONCLUSION Cooling tower performance curves are useful tools to predict cooling tower perfo rmance for various operating and ambient conditions. However, it is recommended that the WCTPE prog ram presented in appendix P be employed when greater accuracy.is required, as the curves generate d by the CTPCG program are averaged curves. R.l APPENDIXR TRICKLE FILL PERFORMANCE TEST RESULTS R.t INTRODUCTION The performance characteristics of trickle fills of three different heights are determined experimentally. The results are critically evaluated and presented by extended empirical equatio ns. The trickle fills consist of horizontally stacked cylinders as shown ?in figure R.I. The cylinders have an outside diameter of 90 mm. The height ofthe spray zones above the fiU fur all the tests is 150 mm. 221ayers Lfi =1.98 m 17 layers Lfi =1.53 m 121ayers Lfi =1.08 m (a) (b) Figure R.I: Three heights of trickle fills tested. (c) Each fill in figure R.I is tested at different air and water mass flow rates The results of the tests for the fills shown in figureR.l(a). R.l(b) and R.l(c) are shown in sections R.2, R.3 an d R.4 respectively. The test of the 1.53 m high fill. shown in figure R.I(b), is repeated at colder wate r inlet temperatures to investigate the effect of the inlet water temperature on the transfer coefficien t. The results of this test are shown in section R.6. The 1.08 m and 1.98 m fills are then tested at constant wa ter and air mass flow rates to further investigate the effect of the inlet water temperature on fill perform ance. R.2 "R.2 FILL HEIGHT: 1.08 m Table R.1: Experimental measurements (Po = 100060 Pa). Tai T.b T.",; T.a ma m. dpft Tao ac ?C o,C t ?C kg/s kg/s Pa ?C 1 12.943 11.905 41.150 31.565 2.677 6.219 12.789 35.954 2 12.592 11.397 41.221 28.916 4.101 6.203 28.033 32.398 3 11.837 10.304 41.252 26.834 5.382 6.174 44.656 29.339 4 11.877 9.939 41.255 25.137 6.743 6.156 70.993 27.893 5 12.524 10.155 41.258 23.825 8.149 6.139 112.324 26.580 6 13.063 10.427 41.248 22.650 9.488 6.145 164.570 25.902 7 14.011 12.867 41.535 34.332 2.779 10.250 19.323 36.862 8 13.868 12.564 41.520 32.437 3.884 10.272 32.459 34.974 9 12.847 11.285 41.479 30.094 5.438 10.262 58.188 33.233 10 12.836 10.870 41.420 28.502 6.806 10.194 90.245 31.650 11 13.468 11.025 41.128 27.242 8.120 10.225 136.201 30.194 12 14.510 11.480 40.121 25.728 9.460 10.237 206.566 28.649

13 15.727 14.427 37.939 34.118 2.688 15.259 32.095 34.461 14 15.375 13.989 36.444 31.508 4.025 15.253 49.694 33.636 15 14.640 12.813 34.672 28.908 5.444 15.265 79.608 31.410 16 14.730 12.428 33.867 27.485 6.650 15.264 114.150 29.945 17 15.189 12.251 33.184 26.160 8.012 15.268 174.013 28.335 18 15.985 12.571 32.817 25.248 9.234 15.268 257.127 27.533 Table R.2: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (Lft = 1.08 m). G. Go Me/Lfi Me.,lLfi Mep/Lfi KramlM KramJP TaoP TaoM 1 2.764 1.190 0.612 0.618 0.684 19.107 19.068 35.091 34.456 2 2.757 1.823 0.763 0.775 0.845 17.627 17.588 32.309 31.757 3 2.744 2.392 0.888 0.906 0.981 16.354 16.321 30.197 29.695 4 2.736 2.997 1.012 1.035 1.115 16.625 16.594 28.467 28.012 5 2.729 3.622 1.134 1.161 1.247 18.064 18.034 27.131 26.719 6 2.731 4.217 1.273 1.303 1.396 19.569 19.540 26.229 25.851 7 4.556 1.235 0.494 0.497 0.568 26.113 26.036 38.102 37.381 8 4.565 1.726 0.597 0.603 0.673 22.306 22.239 36.461 35.796 9 4.561 2.417 0.720 0.731 0.803 20.475 20.417 34.313 33.704 10 4.531 3.025 0.810 0.824 0.899 20.369 20.317 32.628 32.065 11 4.545 3.609 0.893 0.911 0.990 21.675 21.625 31.274 30.754 12 4.550 4.204 1.017 1.040 1.126 24.316 24.268 29.936 29.466 13 6.782 1.194 0.331 0.332 0.385 46.237 46.117 35.746 35.098 14 6.779 1.789 0.459 0.461 0.522 31.997 31.921 33.529 32.960 15 6.784 2.419 0.577 0.580 0.645 28.242 28.184 31.070 30.568 16 6.784 2.956 0.656 0.660 0.728 27.246 27.195 29.644 29.181 17 6.786 3.561 0.752 0.757 0.831 28.715 28.667 28.458 28.029 18 6.786 4.104 0.853 0.860 0.941 31.976 31.926 27.844 27.440 R.3 Table R.3: Empirical relations for the Merkel number according to the various me thods (Ljl = 1.08 m). Approacb Eq. Empirical relation Correlation type coefficient I Me, / Lfi =0.834905 G:?465726G~?615841 0.9911 e-NTU 2 Me; / Lfi =1.000951(Gw/GJ-<l?539946 0.9749 Me, / Lfi =0.008168 G~?271388G~?9S03S2 3 + 0.992732 G;0.578730 G~.468523 0.9954 I Me M / Lfi =0.857501 GwO.480158G~?624824 0.9918 Merkel 2 MeM / Lfi =1.021224(Gw /GJ-<l.sSlS02 0.9774 MeM / Lfi = 0.008028 G~?22S3S8G~?9%20S 3 +1.007228 G:?S828S9G~?4820S6 0.9955 I Mep / Lfi = 0.932891 G;O.4S3190G~?S92728 0.9911 Poppe 2 Mep / Lfi = 1.104216(Gw /Gj?.522373 . 0.9760 Mep / Lfi = 0.008383 G~.1S8S42G;,070185 3 +1.073837 G:?S36574 G:.451302 0.9950 ~ ~~....~ .,J;l ~ -;;- ~~ ~ ~.'..... .- .... - - .-' .' ~ ~ n. .- ." --~ ...~--_ . .- ... .--.'.. . -~.-''''' I""" Gw 6.78 kglm's f--~ <>~ . --,;.; Gw 4.55 kglm"s....... 0 0 " Gw - 2.73 kglm's f---Equation type 1 .. _. _.. Equation type 2 f--

-Equation type 3 1.4 1.2 1 'E 0.8 '"..J " 0.6 :::iE 0.4 0.2 o 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Figure R.2: Comparison of experimental data and empirical equations (Ljl = 1.08 m). R.4 Table R.4: Empirical relations for the loss coefficient according to the various methods (Lft = 1.08 m). Approach Eq. Empirical relation Correlationtype coefficient I K = 10 054991 GO.712342G-O.2J7II6 0.8043Idm!' w a Merkel and 2 K = 18 517679(G /G y<19331fdml' w a 0.6199 e-NTU K = 2 330647 G1.551011G-2.065135fdmt '. w a 3 +5.720996 G,:o?432653G~?521191 0.9752 I K = 10 026815 GO.712256G-D236243 0.8039fdml' w a 2 K = 18 484530(G /G )0.418775Poppe fdml' w a 0.6187 K = 2 331334 G1.549666G-2.062807fdml' w a 3 + 5.695113 G,:o?43254SG~.s23194 0.9750 4.54.03.53.02.52.01.5 I Q <> Gw =6.78 kglm's \ 0 Gw =4.55 kglm's!:. Gw =2.73 kglm's ~ --Equation type 1~ ....... Equation type 2 .. -Equation type 3 0. . . . .. . .'. .. - -~""" .......'1Q:"-.: .'. ....... .-.. ' '". -- - -" ", ' .. .... L ?? ~. '. ??1 ..... h .. ?? .. -?? .. "D. - . ~ . .. -'" . ......... .. ... ~ '" ...

...... ...... - -..... 35 45 50 20 25 15 10 1.0 40 "E E 30 :l Ga, kg/m's Figure R.3: Comparison of experimental data and empirical relations for the loss coefficient (Lft = 1.08 m). R.5 R.3 FILL HEIGHT: 1.53 m Table R.5: Experimental measurements (Pa ~ 101340 Pa). Tai Tw' TWf Two ma m. dpft Taa ?C ac ?c; ac kg/s kg/s Pa ?C 1 16.351 14.451 48.781 34.326 2.701 6.213 18.695 43.753 2 15.951 13.769 48.451 30.551 4.097 6.166 39.393 40.658 3 15.448 12.893 48.432 27.886 5.398 6.171 63.944 38.397 4 15.457 12.414 48.088 25.820 6.707 6.189 99.235 36.304 5 15.912 12.438 47.651 24.149 8.093 6.204 154.657 34.526 6 16.420 12.627 47.113 22.664 9.450 6.149 227.564 32.952 7 16.963 15.303 45.966 36.997 2.727 10.338 25.337 43.310 8 16.448 14.416 45.487 33.794 4.070 10.370 47.584 41.036 9 15.562 13.119 45.067 31.083 5.415 10.330 78.553 38.787 10 15.371 12.390 44.528 28.987 6.753 10.351 120.953 36.848 11 15.886 12.431 44.254 27.363 8.118 10.340 186.495 35.487 12 16.455 12.640 44.013 25.921 9.372 10.342 274.940 34.491 13 17.260 15.430 43.845 38.091 2.730 15.183 37.761 42.136 14 16.971 15.077 43.765 35.800 4.049 15.188 61.901 40.920 15 15.948 13.783 43.665 33.538 5.425 15.093 100.334 39.515 16 15.355 12.789 43.585 31.780 6.760 15.055 151.818 38.414 17 15.365 12.577 43.364 30.349 8.097 15.122 231.878 37.226 18 15.763 12.606 43.228 28.678 9.445 15.069 375.143 36.357 Table R.6: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (Lft = 1.53 m). Gw Ga Me/Ln Me,jLfi MepiLn Krdm1M Krdm1P TaoP Ta"" 1 2.761 1.201 0.518 0.533 0.597 18.984 18.923 42.240 41.458 2 2.740 1.821 0.660 0.684 0.751 17.228 17.174 38.728 38.060 3 2.743 2.399 0.786 0.816 0.886 16.168 16.122 36.340 35.743 4 2.751 2.981 0.902 0.932 1.007 16.350 16.310 34.205 33.669 5 2.757 3.597 1.028 1.052 1.131 17.601 17.564 32.385 31.904 6 2.733 4.200 1.180 1.189 1.273 19.088 19.053 30.827 30.395 7 4.595 1.212 0.388 0.392 0.458 24.871 24.776 42.894 42.068 8 4.609 1.809 0.502 0.511 0.577 20.842 20.765 40.518 39.776 9 4.591 2.407 0.604 0.620 0.688 19.542 19.475 38.362 37.686 10 4.601 3.001 0.688 0.708 0.778 19.482 19.423 36.378 35.758 11 4.596 3.608 0.775 0.801 0.874 20.881 20.825 34.885 34.316 12 4.597 4.165 0.883 0.914 0.993 23.169 23.113 33.894 33.366 13 6.748 1.213 0.281 0.282 0.342 36.777 36.631 42.012 41.184 14 6.750 1.799 0.382 0.385 0.447 27.272 27.164 40.811 40.040

15 6.708 2.411 0.474 0.481 0.546 24.705 24.612 39.455 38.728 16 6.691 3.005 0.538 0.547 0.612 24.214 24.129 38.040 37.353 17 6.721 3.599 0.596 0.609 0.675 25.908 25.825 36.766 36.121 18 6.697 4.198 0.708 0.727 0.801 30.886 30.796 36.071 35.465 Ii I I I I I R.6 Table R.7: Empirical relations for the Merkel number according to the various me thods (Lft ~ 1.53 m). Approach Eq. Empirical relation Correlationtype coefficient I Me / L = 0 793101 G-o?,S021lGO.674521 0.9955 e 'fi' w a e-NTU 2 Me, / Lft = 0.888688(G w IG a to 62SSS7 0.9905 Me, / Lft = 1.267026 G:),l82591G~?631993 3 _ 0.576009 G~?OS6750G~?60S960 0.9968 I Me / L = 0817071 G ?5810"GO.670746 0.9948M fi' w a Merkel 2 MeM / Lft = 0.910464(Gw IGJ-0 624sss 0.9903 Me / L = 1 299681 G-O?l2lSSSGO.S72612M ft? w a 3 -0.593218 G~?146139G~?SlS2S4 0.9978 1 Mep / Lft = 0.893319 G:'?S42378G~?62S433 0.9947 Poppe 2 Mep / Lft = 0.987536(G. IGa tOS82992 0.9901 Me / L = 1 321142 G-o?1I1S03GO.S710S1 P fi' w a 3 _ 0.557590 G~.166192G~?S3SS37 0.9970 1.4 1.2 1.0 ";E 0.8 j ~ 0.6 0.4 0.2 <> Gw =6.72 kglm's 0 Gw =4.60 kglm's Gw =2.75 kglm's -t> ~--Equation type 1 -... ' .. _Equation type 2 ~- ....P-Equation type 3 p .. ' --'~ ---.. -.. ' ~ - -~ .... ..P-.- ......?? 0 ? .. ' . .... ~ .."",-~.~ 0.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/ro's Figure R.4: Comparison of experimental data and empirical equations (Lft = 1.53

m). R.7 Table R8: Empirical relations for the loss coefficient according to the various methods (Lj. ~ 1.53 m). Approach Eq. Empirical relation Correlation type coefficient I K = 10 855787 GO?SS8271G-o?135404 0.7562ftJml? 11' a Merkel and 2 K = 18, 587998(G IG )0304217fdml ~ wa 0.5234 e-NTU K =8885917 GO,772070G-l.l02970fdml' 11' a 3 +1.716322 G~,330658G~?298676 0.9587 I K = 10829610 GO,557081G-o,134185 0.7548fdml' 11' a 2 K = 18 542800(G IG r03088Poppe fdml' wa 0.5212 K = 8 859530 GO.771492G-IJ04103feintl' w a 3, +1.716283 G~,329653G~?298721 0.9585 I I <> Gw =6.72 kglm"s 0 Gw =4.60 kglm"s ~ /),. Gw =2.75 kglm 2s Equation type 1 2'- ....... Equation type 2r-.. _Equation type 3 0 "', ~~ ...... --'. ~"""-'..". - . .0 .... -. ........ - 1]' .. ' .. .. -"..~ -. '" . - .. - - -.. - --'" . --.. ....... .. -... - .. .. .J-.. A.. -~ .. -.. .. '--t:.-. - ........ - o. ......... '- .. . ' ..... - A '"- L.> -..... "".?::,. 45 40 35 "'E 30 E -1 25 20 15 10 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga, kg/m"s Figure R.5: Comparison of experimental data and empirical relations for the loss coefficient (Lft ~ 1.53 m). R.8 R.4 FILL HEIGHT: 1.98 m Table R.9: Experimental measurements (Pa = 100970 Pal. Taj T wb Twi Two ma m. dpfl Tao

DC DC ?c; ?C kg/s kg/s Pa ?C I 18.001 14.143 52.856 . 34.858 2.665 6.343 22.559 47.974 2 17.891 14.057 52.126 30.604 4.002 6.388 48.068 44.435 3 17.354 13.202 51.514 27.266 5.397 6.332 82.355 41.350 4 17.295 12.712 5\.104 24.928 6.744 6.322 129.508 38.902 5 17.609 12.616 50.562 23.137 8.056 6.319 195.649 36.829 6 18.271 12.898 49.961 21.182 9.395 6.304 286.882 35.047 7 18.283 15.088 49.252 37.985 2.699 10.183 30.269 46.458 8 17.824 14.502 48.325 33.927 4.083 10.124 59.634 43.560 9 16.947 13.244 47.600 30.937 5.362 10.129 96.806 41.015 10 16.868 12.594 47.016 28.598 6.695 10.095 150.050 38.875 11 17.344 12.720 46.296 26.681 8.086 10.081 231.043 37.009 12 18.016 13.014 45.637 25.154 9.343 10.117 341.255 35.723 13 18.525 15.950 44.656 38.312 2.717 15.149 44.267 43.077 14 17.853 15.028 44.307 35.597 4.040 15.080 75.603 41.655 15 16.903 13.540 43.914 33.052 5.397 15.077 122.481 40.044 16 16.631 12.725 43.587 31.128 6.736 15.114 187.688 38.731 17 17.095 12.628 43.669 29.629 8.106 15.073 288.946 37.665 18 17.738 12.915 43.571 28.382 9.250 15.044 426.801 37.244 I I I I I I I Table R.10: Transfer coefficients, loss coefficients and outlet temperatures a ccording to the different methods (Lfl = 1.98 m). Gw Ga Me/?,; Me,.!L. Mep/Lfi K[dmlM KfdmlP TaoP TaoM I 2.819 1.184 0.451 0.473 0.534 17.862 17.789 46.212 45.350 2 2.839 1.779 0.588 0.621 0.684 16.655 16.591 42.524 41.794 3 2.814 2.399 0.722 0.754 0.821 15.805 15.754 39.258 38.627 4 2.810 2.997 0.851 0.870 0.939 16.043 15.999 36.821 36.259 5 2.808 3.580 0.995 0.985 1.058 17.104 17.064 34.847 34.341 6 2.802 4.176 \.169 \.101 \.176 18.541 18.503 33.154 32.697 7 4.526 1.200 0.346 0.352 0.419 23.013 22.904 46.370 45.459 8 4.500 1.815 0.453 0.468 0.532 19.705 19.619 43.324 42.529 9 4.502 2.383 0.541 0.563 0.627 18.679 18.607 40.907 40.189 10 4.487 2.976 0.618 0.645 0.711 18.716 18.654 38.713 38.060 11 4.481 3.594 0.705 0.736 0.804 19.884 19.826 36.803 36.213 12 4.496 4.152 0.802 0.837 0.910 22.098 22.042 35.468 34.928 13 6.733 1.208 0.270 0.270 0.352 33.240 33.098 43.670 42.812 14 6.702 1.796 0.349 0.352 0.419 25.613 25.506 42.036 41.242 15 6.701 2.399 0.427 0.434 0.499 23.366 23.274 40.462 39.719 16 6.718 2.994 0.481 0.492 0.554 23.120 23.037 38.901 38.205 17 6.699 3.602 0.537 0.552 0.615 24.661 24.580 37.800 37.143 18 6.686 4.111 0.603 0.623 0.689 28.028 27.942 37.052 36.432 R.9 Table R.II: Empirical relations for the Merkel number according to the various m ethods (Lft ~ 1.98 m). Approach Eq. Empirical relation Correlationtype coefficient 1 Me / L =0 809060 G-o?6.9841GO.724050 0.9934 eft? w a e-NTU 2 Me, ILft =0.843326(G w /G a )-o706997 0.9928

Me / L =1 534059 G-o?862216GO.630953 e fl' w a 3 -1.024374 G~1.377900G~?393037 0.9956 I Me / L =0 817647 G-o?644301GO.67290. 0.9984M fi' w a Merkel 2 MeM / Lft =0.846533(Gw /Ga yo.6S??26 0.9980 Me / L =1 262879 G-029.0S6GO.6S8S4. M ft? w a 3 _ 0.543287 G~O.064423G~?64'236 0.9990 I Mep / Lft =0.892753 G:?S'9301G~?611772 0.9969 Poppe 2 Me p / Lft =0.917453(Gw /G.)-o600831 0.9662 Me / L =1 808451 G-o??09397GO.558907P fi' . w a 3 -1.270136 G~1.371907G~.4394'2 0.9976 1.2 1.0 0.8 "'E j 0.6 Qj :;: 0.4 0.2 ? GN = 6.71 kgfm"s ~0 Gw =.4.50 kgfm"s --' t>. GN = 2.82 kgfm"s ~~--Equation type 1 ~ .J;I ....... Equation type 2 ~ ...._Equation type 3 .~ .-iii. ,.... .,./;) ~V ~~ :::-~ 0.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m's Figure R.6: Comparison ofexperimental data and empirical equations (Lfl = 1.98 m ). R.JO Table R.12: Empirical relations fur the loss coefficient according to the variou s methods (Lft = 1.98 m). Approach Eq. Empirical relation Correlationtype coefficient 1 K =10539809 GO.525842G-<JI07452 0.7779filml' w a Merkel and 2 K =17 952891(G /G )0.270809fdml' w a 0.4998 e-NTU K '=7 047319 GO.8124S4 G-1.l43846fdml' w a 3 +2.677231 G:.194827G~Ol8498 0.9684 1 K =10501559 GO.524991G-<J?I05825 0.7765fdml' w a 2 K =17 904821(G /G )0.169548Poppe fdml' w a 0.4965 K =6 993131 GO.813936G-1.l47400fdml' wo 3 + 2686735 GO.294068GI.Ol6412 0.9682 ? w a I I <> Gw - 6.71 kg/m's 0 Gw = 4.50 kg/m's ~ fj. Gw = 2.82 kg/m's--Equation type 1 ...... ....... Equation type 2 '-'-. .....

_Equation type 3 0. ' . ., .. l.. n""""':" ."' . ~ .. ' .. ~ v .. -.. -. .' . v . -......~ -r:J. -'. -. '. . .....~ . . -.. -. '" --'. -..... .. -.. ... '. " ' .. . '" . .." . ...........:- . .. - . .. -. -. - ......... ............ '" .... -". r--.... D. ..... -.. . -.. -~ 40 35 30 "'E ? 25E ~ 20 15 10 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m"s Figure R.7: Comparison of experimental data and empirical relations for the loss coefficient (Lft = 1.98 m). I I I I I I I I ! R.lI R.S SUMMARY AND COMBINED RESULTS A summary of the equations for the transfer coefficient, according to the Merkel approach, is shown in table R.I3. It can be seen from figures R.2, R.4 and R.6 that all three types of equations give accurate representations of the measured data. This is also shown in table R.13 where all the correlation coefficients, r'-, are close to unity. Table R.B: Summary ofthe transfer coefficients according to the Merkel approach. L., m Equation type 1 r'

1.08 Me / L =0857501 G--<l?480158GO.624824 0.9918M fi' w a 1.53 Me / L =0 817071 G-O.581055 GO.670746 0.9948M fi' w a 1.98 Me / L =0 817647 G--<l?64430IGO.672908 0.9984M fi' 111 a L .. m Equation type 2 it 1.08 M, / L =1 021224(G /G )--<l551502 0.9774eM fi' 111 a 1.53 Meu / Lfl =0.910464(Gw /Ga )--<l624555 0.9903 1.98 Meu / Lfl =0.846533(Gw /Ga t?658826 0.9980 L.. m Equation type 3 r' 1.08 Me / L =0008028 GO.225358 GI.996205 + 1 007228 G--<l.582859GO.482056 0.9955 M fi' 111 a ? 'It' Q 1.53 Me / L =1 299681 G-O.12I5"GO.572612 - 0 593218 GOI46139GO.515254 0.9979Mfi' 111 Q ? 111 Q 1.98 Me / L =1 262879 G-O.298056GO.658548 - 0543287 G--<l064423GO.64ll236 0.9905 Mft? wa ? 111 a The experimental data for the transfer coefficient for all the different fill he ights can be represented by Me IL =0 932886 G-<l?l4l807GO.64838SL-<l?161078 (R.l) M fi' II' a fi with a correlation coefficient ; ~ 0.986328. Figure R.8 shows the results from equation (R.I) compared to the transfer charac teristics obtained from experimental measurements. Tests 1 to 18 in figure R.8 represent the tests for t he 1.08 m thick fill. Tests 19 to 36 represent the fill test results of the 1.53 m thick fill and tests 37 t o 54 represent the measurements ofthe 1.98 m thick fill. Due to the Iimitatioos of the fill test facility, it is impossible to conduct th e fill tests at a constant water temperature. If the effect of the changing water temperature is included in the correlation, the Merkel number can then be presented by Me M / Lfl =1.930306 G::?"8230G~?641400L;?352733Tw-:.I78670 with a correlation coefficient? = 0.986224. (R.2) A summary of the equatioos for the loss coefficient according to the Merkel appr oach is shown in table R.14. It can be seen from the correlation coefficients, r'-, in table R.14 that equation type 3 gives the most accurate representation of the measured data. Equation type I and equation type 2 do not correlate the measured data well. This is seen in figures R.3, R.5 and R.6. R.12 1.4 1.2 1 "s 0.8 j Gi 0.6~ 0.4 0.2 0 0 9 18 27 Test no. 36 <> Experimental --Empirical 45 54 Figure R.8: Transfer characteristic given by equation (R.1) compared to experiem

tal results. Table R.14: Summary ofthe loss coefficients. L.. m Equation type 1 I' 1.08 K =10054991 00.712342G-o?237116 0.8043fdml' w a 1.53 K =10855787 0?.5582710..... 135404 0.7548fdml' w a 1.98 K =10539809 0?.5258420-0.107452 0.7779fdml' w a L .. m Equation type 2 I1.08 K =18517679(0 /0 )0.419331 0.6199fdml' w a 1.53 K =18 587998(G /0 )0304217 0.5212[dInl' w a 1.98 K =17 952891(0 /G )0270&09 0.4998jd",l' w a L.. m Equation type 3 I' 1.08 K =2330647 Gl.l1l0110-2.061135 + 5720996 O-o?4326l3GO.521191 0.9752fdmt? w a ? w a 1.53 K fdml =8.885917 0~?7720700a1.102970 + 1.716322 0~3306l80;298676 0.9585 1.98 K =7 047319 0??812454 G-1.I43846 + 2 677231 0?.294827 01.018498 0.9684film} " wa ? wa The experimental data for the loss coefficient for all the different fill height s can be represented by K = 11 093641 GO.603212G-lJ?148620rO.203976fdml? waft (R.3) with a correlation coefficient; = 0.769823. Equation (RJ) and the test data of t he three different fill heights are shown in figure R.9. The correlation coefficient suggested that equa tion (RJ) does not correlate the data well and this is evident from figure R.9. R.13 54453627189 v <> Experimental -Empirical , , 0 ~ <~ 0\ < ~ o~ ~ ~ 0~ "o yo , 0 OOV<I><> 0 ~ !'o.. IV I~O IV ~OOY<~-00 10 o 15 40 20 45 35 "E 30 ~ ~ 25 Test no. Figure R.9: Loss coefficients given by equation (R.3) compared to experiemtal re sults. 54453627189 v <> Experimental _Empirical ~ < ~ I ~<l , .1 ~) ~ .I L. b.VO ~IA "0 ~ /?"&0 \.IVV ~O/ rv' ~<>J~ ~v ""

10 o 15 40 20 45 35 "E 30 ~ ~ 25 Test no. Figure R.IO: Loss coefficients given by equation (RA) compared to experiemtal re sults. The data for the tests of all three fill heights can also be correlated by K = (4 547149 01.376770 -1.710200 +5058140 O?.38127400.673116\T-<l.207834jdml' w a . w a JLfi (RA) (R.5) r R.14 with a correlation coefficient? = 0.950862. It can be seen from figure R.IO that equation (R.4) correlates the data accurately. As a function of the inlet water temperature, expressed in ?e, the loss coeffici ent is correlated by K =(3 8163 GI.l49219G-1.773193 + 4 296632 GO.387023GO.673487 h-o.209736To.040794 fdml? w a ? w a JLfi WI with a correlation coefficient? = 0.9508. It can be seen that the loss coefficie nt is a very weak function ofthe inlet water temperature as the exponent ofTWi in equation (R.5) is very cl ose to zero. R.6 FILL HEIGHT: 1.53 m TESTED AT COLDER WATER TEMPERATURES. The fill test presented in section R.3 is repeated in this section at colder wat er temperatures. The fill test measurements are shown in table R.15 with the corresponding transfer and loss co efficients shown in table R.16. The empirical equations are not plotted for this fill test, as they are essentially the SaD3e as those in figures R.4 and R.5. Table R.15: Experimental measurements (Pa = 101340 Pa). Tai T.b TWi Tw, ma m. dpfl Ta, ?e ?e ?e, ?C kg/s kg/s Pa ?e 1 15.686 14.377 42.329 31.718 2.729 6.250 18.812 38.485 2 15.137 13.606 42.027 28.489 4.115 6.220 38.954 36.116 3 14.605 12.763 41.845 26.122 5.403 6.228 63.052 34.109 4 14.414 12.195 41.630 24.255 6.789 6.240 100.579 32.375 5 14.795 12.104 41.387 22.826 8.115 6.218 153.825 30.922 6 15.246 12.312 41.018 21.548 9.534 6.240 230.210 29.691 7 15.763 14.573 40.578 33.771 2.713 10.250 24.190 38.116 8 15.059 13.716 39.926 30.970 4.049 10.270 45.720 36.358 9 14.315 12.813 39.400 28.597 5.450 10.223 77.528 34.691 10 13.956 12.086 38.956 26.819 6.759 10.253 118.283 33.108 11 14.115 11.862 38.574 25.269 8.166 10.120 184.456 31.787 12 14.462 11.950 38.452 24.132 9.365 10.256 266.665 31.027 13 15.414 14.503 38.178 34.002 2.717 15.050 35.886 36.756 14 15.016 14.084 37.595 31.826 4.069 15.020 59.857 35.566 15 14.351 12.973 37.145 29.873 5.433 14.971 97.078 34.237 16 14.029 12.152 36.672 28.221 6.769 15.002 148.028 32.912 17 14.112 11.934 35.909 26.804 8.157 14.981 226.963 31.625 18 14.445 11.981 35.411 25.663 9.304 14.945 332.053 30.901 If is curve is fitted through the data of this fill test and the fill test prese nted in section R.3 for the same

fill, with the hotter inlet water temperatures, then find Me IL = 1907410 O-o?S9""G'?649137T -o?209468Mft" w awl (R.6) with a correlation coefficient, ? = 0.9951 and T. i is in ?C. The exponents for Gw and Ga are within close tolerance of those presented in table R.7 or table R.13. The exponent of T. i in equation (R.6) is within close tolerance of the exponent of TWi in equation (R.2) R.15 The mfluence of the inlet water temperature on the performance of a fill is inve stigated where all other test parameters are held approximately constant. Table R.16: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different methods (Lft = 1.53 m). Gw Ga Me/4 Me.,!L. MepiL. KfdmlM KfdmlP Taop raaM 1 2.778 1.213 0.536 0.545 0.610 19.059 19.019 37.438 36.793 2 2.765 1.829 0.688 0.704 0.772 17.198 17.160 34.667 34.112 3 2.768 2.402 0.817 0.840 0.913 16.186 16.153 32.627 32.131 4 2.773 3.017 0.926 0.954 1.030 16.439 16.409 30.691 30.246 5 2.763 3.607 1.036 1.066 1.146 17.667 17.639 29.180 28.777 6 2.773 4.237 1.173 1.202 1.288 19.218 19.191 27.971 27.609 7 4.555 1.206 0.412 Q.415 0.485 24.554 24.491 38.366 37.676 8 4.564 1.800 0.532 0.537 0.606 20.699 20.645 36.207 35.590 9 4.544 2.422 0.638 0.648 0.718 19.440 19.394 34.204 33.648 10 4.557 3.004 0.721 0.735 0.806 19.381 19.339 32.554 32.043 11 4.498 3.629 0.805 0.823 0.896 20.810 20.769 30.963 30.497 12 4.558 4.162 0.908 0.932 1.011 22.910 22.869 30.340 29.902 13 6.689 1.207 0.293 0.293 0.354 36.258 36.165 36.814 36.153 14 6.676 1.808 0.407 0.409 0.474 26.859 26.790 35.457 34.852 15 6.654 2.415 0.505 0.508 0.575 24.495 24.435 34.133 33.564 16 6.667 3.009 0.584 0.590 0.658 24.160 24.105 32.834 32.300 17 6.658 3.625 0.637 0.644 0.710 25.657 25.605 31.201 30.712 18 6.642 4.135 0.716 0.725 0.796 28.926 28.872 30.355 29.898 R.7 FILL HEIGHT: 1.08 ro, Ga AND Gw CONSTANT The 1.08 m fill is tested at approximately constant air and water mass flow rate s. Only the inlet water mass flow rate is varied during the duration of the fill test. This is done to i nvestigate the dependence of the transfer coefficient on the inlet water temperature. Table R 17 contains the experimental measurements of the fill test and table R 18 contains the corresponding transfer and loss coefficients. Table R.17: Experimental measurements (Pa = 100060 Pa) Tw? Twb TW1 Twa ma mw dpft Taa ?C ?C ?C 1 ?C kg/s kg/s Pa ?C 1 16.656 11.693 56.252 33.671 6.734 10.137 91.434 2 16.673 11.705 56.053 33.585 6.735 9.984 91.366 3 16.631 11.645 55.726 33.499 6.735 10.068 91.131 4 16.626 11.641 55.453 33.399 6.736 10.088 91.178 5 16.583 11.570 54.930 33.269 6.734 9.860 91.053 6 16.536 11.479 54.706 33.118 6.743 10.039 90.700 7 16.510 11.501 53.758 32.900 6.742 10.012 90.711 8 16.466 11.512 53.389 32.778 6.748 10.628 90.745 9 16.414 11.448 53.141 32.657 6.748 9.999 90.437 10 16.398 11.405 52.811 32.568 6.751 9.992 90.297 11 16.395 11.388 52.429 32.435 6.745 9.987 89.879 12 16.400 11.397 52.300 32.387 6.746 9.970 89.883 13 16.393 11.407 51.882 32.278 6.747 9.954 89.944 14 16.359 11.362 51.535 32.107 6.749 9.964 89.655 R.16

Table R.17 (continue): Experimental measurements (Po = 100060 Pal Tai TWb TWI Tw? rna rnw dpfi To. 'C ?0 'C I ?C kg/s kg/s Pa ?C 15 16.345 11.347 51.440 32.084 6.749 9.925 89.709 16 16.336 11.352 51.013 31.968 6.754 9.927 89.766 17 16.293 11.323 50.702 31.842 6.754 9.937 89.540 18 16.299 11.339 50.344 31.721 6.749 9.813 89.536 19 16.336 11.436 49.655 31.486 6.759 9.874 88.673 20 16.329 11.453 49.495 31.429 6.751 9.855 88.987 21 16.339 11.495 49.539 31.420 6.756 9.837 89.064 22 16.297 11.472 49.211 31.340 6.759 9.815 89.077 23 16.226 11.439 48.506 31.077 6.758 9.809 88.911 24 16.253 11.537 48.393 31.005 6.757 9.727 88.617 25 16.270 11.584 48.172 30.975 6.753 9.716 88.521 26 16.219 11.545 47.783 30.838 6.762 9.842 88.812 27 16.179 11.516 47.637 30.804 6.760 9.883 88.884 28 16.128 11.513 47.445 30.718 6.756 9.833 88.716 29 16.073 11.481 47.287 30.650 6.764 9.880 88.839 30 16.029 11.469 46.841 30.523 6.756 9.848 88.393 31 15.869 11.457 46.064 30.199 6.742 9.676 87.983 32 15.853 11.431 46.069 30.159 6.769 9.701 88.225 33 15.823 11.435 46.037 30.120 6.768 9.702 88.161 34 15.806 11.427 45.805 30.053 6.769 9.698 87.868 35 15.763 11.406 45.351 29.874 6.785 9.581 88.247 36 15.704 11.382 45.189 29.761 6.783 9.634 88.136 37 15.661 11.391 45.089 29.719 6.781 9.659 87.919 38 15.633 11.373 44.906 29.648 6.779 9.588 87.806 39 15.631 11.393 44.716 29.556 6.781 9.612 87.672 40 15.570 11.370 44.577 29.483 6.782 9.662 87.747 41 15.519 11.342 44.337 29.391 6.787 9.621 87.553 42 15.494 11.326 44.166 29.312 6.784 9.575 87.531 43 15.485 11.332 43.996 29.247 6.787 9.646 87.561 44 15.435 11.329 43.826 29.165 6.782 9.599 87.601 45 15.300 11.239 43.268 29.299 6.763 10.004 89.561 46 15.267 11.200 43.115 29.254 6.770 10.187 89.534 47 15.245 11.178 42.770 29.122 6.774 10.196 89.394 48 15.182 11.159 42.093 28.755 6.781 10.152 89.094 33.414 49 15.119 11.128 42.024 28.717 6.783 10.168 89.311 33.301 50 15.098 11.088 41.861 28.653 6.782 10.184 89.045 33.255 51 15.095 11.078 41.774 28.603 6.783 10.138 89.071 33.190 52 15.073 11.093 41.541 28.527 6.776 10.179 88.867 33.039 53 15.035 11.087 41.479 28.471 6.779 10.121 88.756 32.976 54 15.015 11.070 41.334 28.428 6.785 10.142 88.899 32.898 55 14.996 11.053 41.097 . 28.321 6.782 10.166 88.796 32.754 56 14.973 11.038 40.902 28.229 6.788 10.147 88.864 32.622 57 14.928 11.032 40.899 28.182 6.790 10.120 88.659 32.564 58 14.906 11.042 40.806 28.172 6.782 10.140 88.488 32.514 59 14.910 11.050 40.565 28.073 6.787 10.125 88.535 32.405 60 14.880 11.047 40.486 27.997 6.788 10.107 88.529 32.343 61 14.851 11.050 40.425 27.976 6.786 10.096 88.515 32.261 62 14.769 11.025 40.388 27.935 6.790 9.965 88.673 32.250 63 14.740 11.021 40.308 27.912 6.790 10.044 88.540 32.188 64 14.698 11.004 40.205 27.863 6.791 10.091 88.350 32.142 R.17 Table R.17 (continue): Experimental measurements (P, = 100060 Pa) Toi T.b Twi T., m, m. dPJi Too0c; ?C ?C l ?C kg/s kg/s Pa ?C 65 14.675 10.999 40.133 27.821 6.790 10.073 88.282 32.119 66 14.664 11.013 40.174 27.818 6.793 9.986 88.427 32.076 67 14.643 10.990 40.156 27.805 6.794 10.021 88.140 32.042 68 14.615 10.955 40.078 27.766 6.794 10.043 88.212 32.019

69 14.603 10.925 40.009 27.727 6.790 10.022 88.033 31.949 70 14.599 10.909 39.994 27.708 6.796 9.995 88.148 31.891 71 14.572 10.907 39.975 27.708 6.788 10.016 88.000 31.853 72 14.505 10.878 39.915 27.677 6.797 10.026 88.196 31.840 73 14.506 10.881 39.877 27.652 6.789 9.982 87.817 31.839 74 14.481 10.854 39.817 27.613 6.791 9.960 87.904 31.762 75 14.405 10.801 39.743 27.585 6.784 10.020 87.731 31.686 76 14.340 10.769 39.711 27.534 6.791 9.986 87.847 31.627 77 14.295 10.746 39.674 27.523 6.786 9.986 87.538 31.635 78 14.236 10.709 39.573 27.645 6.780 10.329 89.057 31.669 79 14.206 10.710 39.525 27.633 6.781 10.205 88.774 31.633 80 14.172 10.710 39.516 27.604 6.790 10.288 88.852 31.618 81 14.119 10.715 39.451 27.603 6.788 10.214 89.148 31.638 82 14.105 10.715 39.289 27.529 6.788 10.298 89.080 31.530 83 14.119 10.741 39.131 27.477 6.791 10.223 89.169 31.396 84 14.101 10.697 38.998 27.380 6.793 10.142 89.185 31.303 85 14.075 10.656 38.893 27.307 6.795 10.278 89.269 31.205 86 14.056 10.637 38.771 27.246 6.802 10.138 89.013 31.109 87 14.025 10.627 38.551 27.183 6.791 10.260 89.118 31.003 88 13.951 10.631 38.389 27.096 6.798 10.240 88.891 30.881 89 13.919 10.686 38.268 27.058 6.800 10.251 89.076 30.798 90 13.900 10.746 38.123 27.015 6.790 10.233 88.571 30.761 91 13.872 10.741 37.973 26.953 6.802 10.257 88.592 30.669 92 13.835 10.765 37.830 26.876 6.797 10.182 88.835 30.527 93 13.809 10.789 31.645 26.821 6.789 10.219 88.367 30.418 94 13.698 10.705 37.224 26.575 6.795 10.267 88.151 30.Q47 95 13.631 10.656 37.069 26.498 6.795 10.256 88.375 29.963 96 13.558 10.615 36.890 26.411 6.792 10.259 87.843 29.836 97 13.523 10.622 36.697 26.315 6.790 10.260 87.851 29.699 98 13.484 10.629 36.526 26.245 6.771 10.214 87.475 29.596 99 13.470 10.642 36.357 26.178 6.771 10.207 87.414 29.457 100 13.421 10.642 36.147 26.091 6.778 10.181 87.270 29.360 101 13.411 10.666 35.850 25.959 6.784 10.178 87.470 29.105 102 13.385 10.676 35.661 25.850 6.790 10.136 87.597 28.939 103 13.340 10.656 35.339 25.693 6.804 10.127 88.006 28.739 104 13.322 10.660 35.155 25.584 6.813 10.180 87.962 28.568 105 14.483 11.601 34.882 25.747 6.740 10.416 88.866 28.570 106 14.404 11.592 34.849 25.740 6.734 10.331 88.680 28.531 107 14.308 11.571 34.745 25.683 6.742 10.422 88.809 28.440 108 14.273 11.576 34.676 25.653 6.747 10.411 88.749 28.351 109 14.303 11.678 34.513 25.606 6.742 10.368 88.775 28.286 110 14.264 11.617 34.326 25.475 6.741 10.347 88.509 28.159 III 14.173 11.481 34.221 25.382 6.748 10.268 88.562 28.060 112 14.171 11.405 34.119 25.318 6.748 10.331 88.387 27.967 113 14.200 11.338 33.933 25.216 6.740 10.194 88.666 27.847 114 14.190 11.286 33.857 25.139 6.741 10.356 88.731 27.742 R.l8 Table R.17 (continue): Experimental measurements (Pa = 100060 Pa) TaJ T.b TW1 Twa rna m. dpfi Tao ac ?c ?C 1 ?c kg/s kg/s Pa ?c 115 14.178 11.286 33.688 25.073 6.745 10.349 88.557 27.619 116 14.182 11.302 33.443 24.949 6.743 10.316 88.633 27.492 117 14.145 11.306 33.278 24.852 6.748 10.305 88.694 27.375 118 14.114 11.275 32.923 24.649 6.746 10.290 88.348 27.075 119 14.125 11.349 32.795 24.629 6.745 10.303 88.298 27.056 120 14.202 11.385 32.570 24.524 6.740 10.283 88.212 26.915 121 14.196 11.467 32.319 24.423 6.738 10.224 88.149 26.771 122 14.163 11.417 32.147 24.286 6.748 10.227 88.235 26.645 123 14.143 11.359 31.853 24.122 6.749 10.247 88.009 26.387 124 14.122 11.351 31.580 23.980 6.746 10.227 88.076 26.185

125 14.117 11.334 31.388 23.841 6.749 10.139 87.812 26.027 126 14.100 11.361 31.226 23.790 6.751 10.075 87.775 25.916 127 14.283 11.260 30.378 23.328 6.743 10.352 88.304 25.340 128 14.299 11.330 30.120 23.234 6.742 10.364 88.155 25.188 129 14.372 11.324 29.895 23.107 6.738 10.357 88.028 25.003 130 14.403 11.266 29.673 22.951 6.737 10.358 87.893 24.827 131 14.435 11.221 29.440 22.805 6.739 10.327 87.934 24.601 132 14.432 11.244 29.214 22.688 6.740 10.329 87.746 24.454 133 14.438 11.241 28.969 22.558 6.738 10.303 87.853 24.276 134 14.384 11.299 28.706 22.426 6.740 10.322 87.552 24.089 135 14.341 11.268 28.505 22.292 6.741 10.311 87.671 23.903 136 14.286 11.258 28.309 22.178 6.746 10.301 87.649 23.773 137 14.246 11.255 28.139 22.083 6.745 10.342 87.540 23.644 138 14.225 11.224 27.981 21.971 6.747 10.310 87.453 23.490 139 14.246 11.170 27.795 21.865 6.744 10.307 87.404 23.357 140 14.234 11.198 27.640 21.766 6.752 10.290 87.293 23.215 141 14.228 11.155 27.538 21.702 6.751 10.294 87.379 23.134 142 14.187 11.039 27.347 21.540 6.752 10.313 87.293 22.940 Table R. 18: Transfer coefficients, loss coefficients and outlet temperatures ac cording to the different methods (Lfi = 1.08 m). Gw Ga Me/Ln Me,)Ln Mep/Ln KfdmlM K fdmlP TaQp TaaM I 4.506 2.993 0.737 0.767 0.840 20.256 20.152 41.798 40.975 2 4.437 2.993 0.736 0.765 0.838 20.274 20.172 41.466 40.649 3 4.475 2.993 0.739 0.768 0.842 20.232 20.131 41.407 40.593 4 4.483 2.994 0.743 0.772 0.847 20.247 20.147 41.306 40.496 5 4.382 2.993 0.736 0.764 0.836 20.311 20.214 40.625 39.826 6 4.462 2.997 0.749 0.778 0.852 20.160 20.063 40.811 40.012 7 4.450 2.997 0.749 0.777 0.851 20.235 20.142 40.215 39.430 8 4.457 2.999 0.752 0.780 0.855 20.228 20.136 40.034 39.255 9 4.444 2.999 0.756 0.784 0.859 20.180 20.089 39.868 39.092 10 4.441 3.000 0.756 0.783 0.858 20.159 20.070 39.646 38.875 11 4.439 2.998 0.760 0.787 0.862 20.122 20.035 39.446 38.680 12 4.431 2.998 0.761 0.788 0.863 20.128 20.041 39.355 38.592 13 4.424 2.999 0.761 0.788 0.863 20.167 20.081 39.079 38.324 14 4.428 2.999 0.768 0.795 0.871 20.107 20.022 38.928 38.178 15 4.411 3.000 0.767 0.793 0.868 20.130 20.046 38.801 38.053 16 4.412 3.002 0.767 0.793 0.869 20.142 20.059 38.538 37.796 R.19 Table R.18: (continue) Transfer coefficients, loss coefficients and outlet tempe ratures according to the different methods (Lfi = 1.08 m). Gw Go Me/L" MeMiL. MepfL. KfdmlM K fdmlP TaoP TooM 17 4.416 3.002 0.772 0.798 0.873 20.109 20.027 38.386 37.650 18 4.361 2.999 0.170 0.795 0.871 20.180 20.100 38.008 37.280 19 4.388 3.004 0.780 0.804 0.881 19.950 19.873 37.723 37.006 20 4.380 3.000 0.781 0.806 0.882 20.079 20.003 37.627 36.913 21 4.372 3.003 0.783 0.808 0.884 20.060 19.984 37.648 36.934 22 4.362 3.004 0.781 0.805 0.881 20.017 20.002 37.384 36.675 23 4.359 3.004 0.787 0.811 0.887 20.087 20.013 36.975 36.278 24 4.323 3.003 0.790 0.814 0.890 20.038 19.966 36.847 36.155 25 4.318 3.001 0.788 0.811 0.887 20.056 19.985 36.686 35.999 26 4.374 3.005 0.795 0.818 0.895 20.017 20.006 36.620 35.936 27 4.393 3.005 0.795 0.818 0.895 20.106 20.036 36.576 35.893 28 4.370 3.003 0.796 0.819 0.896 20.112 20.042 36.409 35.732 29 4.391 3.006 0.800 0.822 0.900 20.100 20.031 36.370 35.694 30 4.377 3.003 0.798 0.820 0.897 20.079 20.011 36.041 35.373 31 4.300 2.996 0.801 0.822 0.899 20.137 20.072 35.317 34.727 32 4.311 3.008 0.805 0.826 0.903 20.032 19.968 35.388 34.738 33 4.312 3.008 0.808 0.829 0.907 20.019 19.955 35.398 34.750

34 4.310 3.008 0.808 0.828 0.905 19.965 19.901 35.232 34.588 35 4.258 3.016 0.805 0.825 0.901 20.001 19.940 34.748 34.114 36 4.282 3.015 0.814 0.835 0.912 19.990 19.929 34.776 34.144 37 4.293 3.014 0.817 0.837 0.915 19.952 19.891 34.764 34.134 38 4.262 3.013 0.815 0.835 0.912 19.959 19.899 34.547 33.921 39 4.272 3.014 0.821 0.840 0.918 19.925 19.865 34.491 33.869 40 4.294 3.014 0.826 0.846 0.924 19.937 19.878 34.490 33.870 41 4.276 3.016 0.825 0.844 0.922 19.885 19.826 34.261 33.646 42 4.256 3.015 0.825 0.844 0.922 19.915 19.857 34.102 33.491 43 4.287 3.016 0.830 0.849 0.927 19.900 19.843 34.100 33.490 44 4.266 3.014 0.831 0.850 0.928 19.954 19.897 33.950 33.345 45 4.446 3.006 0.808 0.825 0.902 20.533 20.475 33.856 33.249 46 4.527 3.009 0.814 0.831 0.909 20.478 20.419 33.976 33.368 47 4.532 3.011 0.815 0.832 0.910 20.440 20.382 33.744 33.142 48 4.512 3.014 0.829 0.846 0.925 20.368 20.313 33.323 32.736 49 4.519 3.015 0.831 0.847 0.926 20.411 20.356 33.296 32.709 50 4.526 3.014 0.831 0.847 0.927 20.370 20.315 33.197 32.612 51 4.506 3.015 0.831 0.847 0.926 20.376 20.322 33.085 32.503 52 4.524 3.011 0.833 0.849 0.928 20.384 20.330 32.992 32.414 53 4.498 3.013 0.834 0.850 0.929 20.347 20.294 32.894 32.318 54 4.508 3.015 0.833 0.848 0.927 20.355 20.302 32.793 32.220 55 4.518 3.014 0.836 0.852 0.931 20.359 20.306 32.679 32.110 56 4.510 3.017 0.837 0.852 0.932 20.353 20.301 32.519 31.954 57 4.498 3.018 0.842 0.858 0.938 20.294 20.243 32.523 31.959 58 4.507 3.014 0.841 0.856 0.936 20.308 20.256 32.479 31.916 59 4.500 3.016 0.841 0.856 0.935 20.308 20.257 32.293 31.735 60 4.492 3.017 0.847 0.862 0.942 20.300 20.250 32.258 31.702 61 4.487 3.016 0.846 0.861 0.941 20.314 20.264 32.205 31.651 62 4.429 3.018 0.842 0.857 0.936 20.347 20.298 32.009 31.459 63 4.464 3.018 0.845 0.860 0.939 20.315 20.266 32.052 31.502 64 4.485 3.018 0.849 0.863 0.943 20.264 20.215 32.046 31.496 65 4.477 3.018 0.850 0.865 0.944 20.261 20.212 31.987 31.439 66 4.438 3.019 0.848 0.863 0.942 20.282 20.233 31.917 31.370 R.20 Table R.18: (continue) Transfer coefficients, loss coefficients and outlet tempe ratures according to the different methods (Lft = 1,08 m). G. Ga Me/Lti Me.,lLfi Mep/Lfi KfdmlM KfdmlP TQop TaoM 67 4.454 3.019 0.850 0.865 0.945 20.212 20.163 31.948 31.402 68 4.464 3.020 0.852 0.866 0.946 20.230 20.181 31.918 31.372 69 4.454 3.018 0.852 0.866 0.946 20.220 20.171 31.849 31.304 70 4.442 3.020 0.851 0.866 0.945 20.218 20.169 31.799 31.255 71 4.452 3.017 0.852 0.866 0.946 20.225 20.177 31.821 31.277 72 4.456 3.021 0.852 0.866 0.946 20.223 20.175 31.770 31.227 73 4.436 3.017 0.852 0.866 0.946 20.190 20.142 31.712 31.171 74 4.427 3.018 0.852 0.866 0.946 20.206 20.159 31.642 31.102 75 4.453 3.015 0.854 0.868 0.948 20.211 20.163 31.664 31.123 76 4.438 3.018 0.856 0.871 0.950 20.199 20.151 31.607 31.069 77 4.438 3.016 0.855 0.870 0.949 20.165 20.117 31.581 31.043 78 4.591 3.013 0.848 0.862 0.942 20.540 20.491 31.789 31.245 79 4.535 3.014 0.840 0.853 0.932 20.485 20.436 31.579 31.039 80 4.573 3.018 0.848 0.862 0.942 20.440 20.392 31.695 31.155 81 4.539 3.017 0.840 0.854 0.932 20.536 20.488 31.527 30.989 82 4.577 3.017 0.846 0.860 0.940 20.520 20.472 31.536 31.000 83 4.544 3.018 0.841 0.854 0.933 20.540 20.493 31.315 30.784 84 4.507 ?3.019 0.842 0.855 0.933 20.546 20.500 31.139 30.611 85 4.568 3.020 0.853 0.866 0.946 20.547 20.500 31.261 30.733 86 4.506 3.023 0.845 0.858 0.937 20.467 20.421 30.976 30.453 87 4.560 3.018 0.849 0.861 0.940 20.560 20.514 30.969 30.448 88 4.551 3.021 0.851 0.863 0.942 20.479 20.434 30.840 30.322

89 4.556 3.022 0.851 0.864 0.943 20.513 20.468 30.773 30.259 90 4.548 3.018 0.850 0.863 0.942 20.465 20.420 30.671 30.161 91 4.559 3.023 0.851 0.863 0.942 20.409 20.365 30.570 30.063 92 4.526 3.021 0.851 0.863 0.941 20.506 20.463 3Q.411 29.909 93 4.542 3.017 0.851 0.862 0.941 20.454 20.412 30.325 29.826 94 4.563 3.020 0.862 0.873 0.953 20.391 20.349 30.116 29.624 95 4.558 3.020 0.861 0.872 0.952 20.454 20.413 29.982 29.493 96 4.560 3.019 0.862 0.873 0.952 20.363 20.322 29.856 29.371 97 4.560 3.018 0.864 0.875 0.955 20.384 20.344 29.738 29.257 98 4.540 3.009 0.863 0.874 0.953 20.429 20.389 29.592 29.115 99 4.536 3.010 0.862 0.872 0.951 2Q.421 20.382 29.457 28.983 100 4.525 3.012 0.859 0.869 0.947 20.365 20.326 29.252 28.784 101 4.524 3.015 0.859 0.869 0.947 20.390 20.352 29.034 28.572 102 4.505 3.018 0.861 0.871 0.949 20.398 20.360 28.869 28.413 103 4.501 3.024 0.861 0.871 0.948 20.436 20.399 28.605 28.156 104 4.524 3.028 0.869 0.878 0.957 20.374 20.338 28.556 28.110 105 4.629 2.996 0.870 0.879 0.960 20.808 20.771 28.797 28.356 106 4.592 2.993 0.863 0.872 0.951 20.815 20.778 28.668 28.230 107 4.632 2.996 0.869 0.878 0.958 20.795 20.759 28.685 28.248 108 4.627 2.999 0.868 0.877 0.956 20.756 20.720 28.614 28.179 109 4.608 2.996 0.864 0.873 0.952 20.804 20.768 28.466 28.037 110 4.599 2.996 0.870 0.878 0.958 20.760 20.725 28.335 27.909 III 4.564 2.999 0.866 0.875 0.954 20.752 20.718 28.144 27.720 112 4.592 2.999 0.870 0.878 0.958 20.710 20.676 28.125 27.702 113 4.531 2.996 0.861 0.869 0.947 20.849 20.815 27.829 27.411 114 4.603 2.996 0.878 0.887 0.967 20.850 20.815 27.991 27.570 115 4.599 2.998 0.874 0.883 0.962 20.796 20.762 27.831 27.413 116 4.585 2.997 0.875 0.883 0.963 20.840 20.806 27.636 27.225 R.21 Table R.18: (continue) Transfer coefficients, loss coefficients and outlet tempe ratures according to the different methods (Lji = 1.08 m). G. G. Me;L; MeJiL. Mep/L. Kfdm1M KfdmJP TaoP T.oM 117 4.580 2.999 0.878 0.887 0.966 20.837 20.804 27.521 27.113 118 4.573 2.998 0.884 0.892 0.972 20.787 20.755 27.274 26.874 119 4.579 2.998 0.879 0.886 0.966 20.785 20.753 27.175 26.779 120 4.570 2.996? 0.879 0.887 0.966 20.804 20.773 27.009 26.617 121 4.544 2.994 0.875 0.882 0.961 20.822 20.792 26.776 26.392 122 4.545 2.999 0.884 0.891 0.971 20.785 20.756 26.681 26.300 12j 4.554 3.000 0.887 0.894 0.973 20.744 20.715 26.481 26.105 124 4.545 2.998 0.887 0.894 0.973 20.798 20.770 26.267 25.897 125 4.506 3.000 0.891 0.897 0.976 20.728 20.700 26.076 25.711 126 4.478 3.000 0.880 0.886 0.964 20.725 20.698 25.856 25.496 127 4.601 2.997 0.902 0.908 0.989 20.924 20.897 25.524 25.174 128 4.606 2.996 0.896 0.902 0.982 20.906 20.880 25.323 24.978 129 4.603 2.995 0.899 0.905 0.985 20.907 20.882 25.166 24.825 130 4.603 2.994 0.907 0.913 0.994 20.896 20.870 25.031 24.694 131 4.590 2.995 0.908 00.914 0.994 20.904 20.879 24.830 24.496 132 4.591 2.995 0.909 0.915 0.996 20.868 20.843 24.671 24.343 133 4.579 2.995 0.907 0.913 0.993 20.922 20.898 24.464 24.141 134 4.587 2.995 0.911 0.916 0.997 20.848 20.825 24.303 23.987 135 4.583 2.996 0.916 0.921 1.002 20.880 20.857 24.163 23.850 136 4.578 2.998 0.916 0.922 1.002 20.860 20.837 24.004 23.696 137 4.596 2.998 0.921 0.926 1.007 20.844 20.822 23.920 23.615 138 4.582 2.999 0.925 0.930 1.011 20.823 20.801 23.790 23.488 139 4.581 2.997 0.922 0.927 1.008 20.844 20.822 23.629 23.330 140 4.573 3.001 0.927 0.932 1.013 20.773 20.751 23.524 23.229 141 4.575 3.001 0.927 0.931 1.012 20.804 20.782 23.439 23.146 142 4.583 3.001 0.938 0.943 1.025 20.793 20.772 23.334 23.041 ""

-Two r-.... -Twi 1". I???? Tair-- TwbI"-..... -!'I-----. t--.t---...f----.-- h r--. r-... -------~ I---I,? , I, ..... J?? " i, ..i?, I I. I? ..... j.. ' " ? .."...?,-,-.?," I?? .. ,? I 50 60 20 10 1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 Testna. (,J ? ? 40 ~ 8E 30~ Figure R.II: Variation of air and water temperatures. R.22 7+----I---II---f--+----\--+--+--r----f--+--+-+-+----I . 6-l--f-........--!---!--f-........--4--......-f--+--4--......-f---l 1 11 21 31 41 51 61 71 Test no. 81 91 101 111 121 131 141 Figure R.12: Air and water mass flow rates during the fill test. 1.05 <> 1.00 0.95 "'e 0.90 .c ...J -.. 0.85::;: 0.80 o Merkel 0.75 <> Poppe xe-NTU 0.70 25 30 35 40 45 TWI.oC 50 55 60 Figure R.13: Transfer coefficients according the e-NTU, Merkel and Poppe approac hes. R.23 The variation of the air inlet drybulb and wetbulb temperatures and water inlet and outlet temperatures are shown in figure Rll. The air and water mass flow during the fill test is shown i n figure R.12. The transfer coefficients according to the e-NTU, Merkel and Poppe approaches are pl otted against the inlet water temperature in figure R.l3. The transfer coefficients according to all the approaches decreases as the water inlet temperature increases. The absolute difference between the trans fer coefficients of the

Merkel and Poppe approaches remains approximately constant as the water inlet te mperature changes. The transfer coefficients of the e-NTU and Merkel approaches is within close tol erance at colder water inlet temperstures, but as the water increases the difference increases. Ifa cur ve is fitted to the data of the Merkel approach in figure R 13 then find MeM / Lft = CT:,O.2471 where c is a constant and the correlation coefficient, ,;' = 0.9500. (R.7) Figure R.14 shows the measured air outlet temperatures compared to those predict ed by the Merkel and Poppe approaches over a wide range of water inlet temperatures. It can be seen t hat the Poppe approach predicts the water outlet temperatures very accurately over the entire range of water inlet temperatures. This is expected, as the Poppe approach is the more rigorous approach of the two approaches. At lower water inlet temperatures, however, the Merkel approach predicts the outlet air t emperatures accurately. This is because supersaturation of the outlet air is not as dominant when the wa ter inlet temperature is colder. The assumption of the Merkel approach that the air is saturated at the a ir outlet is close to reality, althOUgh the air is supersaturated. ~o Experimental ~f- --Merkel and e-NTU -Poppe ~ ~I~.A, ~~ V l':i'.., 17 34 32 30 o ? - 28~ 26 24 22 25 27 29 31 33 35 37 39 41 43 45 Figure R.14: Measured air outlet temperatures compared to air outlet temperature s predicted by tbe e? NTU, Merkel and Poppe approaches. I I I I I I I I I R.24 R.8 FILL HEIGHT: 1.98 m, G. AND G., CONSTANT The 1.98 m fill is tested at approximately constant air and water mass flow rate s. Only the inlet water mass flow rate is varied during the duration of the fill test. Table R.19 contai ns the experimental measurements ofthe fill test and table R.20 contains the corresponding transfer

and loss coefficients. Table R.19: Experimental measurements (p. = 101340 Pa) rai T.b Twi T., m, m. dpfI Too 'C 'C 'C! 'C kg/s kgis Pa 'C I 14.769 12.323 53.548 30.124 6.732 10.406 153.695 42.867 2 14.709 12.004 53.392 30.048 6.727 10.393 153.050 42.756 3 14.729 11.866 53.268 30.026 6.727 10.357 152.906 42.656 4 14.719 11.836 52.666 29.913 6.727 10.397 153.123 42.378 5 14.650 11.709 52.576 29.820 6.728 10.389 152.812 42.226 6 14.639 11.668 52.104 29.731 6.728 10.379 152.577 41.954 7 14.642 11.678 51.537 29.667 6.728 10.410 152.325 41.626 8 14.631 11.673 51.197 29.535 6.731 10.400 152.155 41.405 9 14.604 11.624 50.746 29.452 6.729 10.387 151.791 41.175 10 14.615 11.634 50.223 29.312 6.728 10.400 151.536 40.903 11 14.609 11.615 49.955 29.243 6.732 10.384 151.518 40.742 12 14.603 11.576 49.391 29.136 6.730 10.386 151.341 40.410 13 14.580 11.548 48.844 28.975 6.733 10.383 151.551 40.019 14 14.549 11.501 48.611 28.846 6.740 10.359 151.256 39.821 15 14.560 11.491 48.177 28.769 6.737 10.334 151.072 39.580 16 14.525 11.459 47.787 28.663 6.736 10.309 150.984 39.333 17 14.503 11.443 47.512 28.579 6.739 10.331 151.013 39.117 18 14.478 11.443 47.077 28.493 6.743 10.329 151.094 38.919 19 14.486 11.447 46.805 28.383 6.740 10.347 150.605 38.695 20 14.539 11.486 46.497 28.300 6.737 10.355 150.447 38.563 21 14.525 11.460 46.116 28.201 6.739 10.314 150.324 38.262 22 14.586 11.514 45.727 28.098 6.739 10.323 150.185 37.982 23 14.650 11.590 45.433 28.023 6.742 10.306 150.212 37.847 24 14.648 11.584 45.279 27.979 6.739 10.303 150.180 37.684 25 14.697 11.649 45.054 27.929 6.739 10.298 149.907 37.542 26 14.660 11.621 44.840 27.845 6.745 10.309 150.146 37.440 27 14.610 11.620 44.638 27.772 6.748 10.266 150.123 37.289 28 14.590 11.568 44.439 27.680 6.749 10.293 149.960 37.071 29 14.525 11.527 44.032 27.655 6.746 10.436 150.565 36.894 30 14.529 11.534 43.869 27.618 6.748 10.427 150.597 36.767 31 14.534 11.545 43.703 27.575 6.750 10.356 150.631 36.723 32 14.518 11.600 43.693 27.567 6.752 10.358 150.663 36.683 33 14.476 11.499 43.560 27.497 6.752 10.411 150.254 36.501 34 14.471 11.517 43.479 27.476 6.755 10.429 150.415 36.479 35 14.419 11.459 43.360 27.416 6.755 10.387 150.606 36.307 36 14.424 11.430 43.281 27.370 6.752 10.392 150.452 36.258 37 14.398 11.408 43.160 27.316 6.758 10.387 150.473 36.197 38 14.424 11.493 43.112 27.332 6.752 10.356 150.468 36.196 39 14.406 11.481 43.086 27.310 6.751 10.386 150.366 36.151 40 14.365 11.409 43.029 27.273 6.752 10.392 150.398 36.073 41 14.369 11.382 42.877 27.221 6.759 10.388 150.509 36.009 42 14.334 11.333 42.716 27.122 6.766 10.360 150.540 35.878 R.25 Table R.19 (continue): Experimental measurements (Po = 101340 Pa) Tal T.b T., T.a ma m. dpfi Tao ?C ?C ?C/ ?C kg/s kg/s Pa ?C 43 14.312 11.317 42.533 27.093 6.758 10.348 149.944 35.828 44 14.337 11.287 42.294 27.016 6.757 10.393 150.007 35.603 45 14.368 11.264 42.119 26.916 6.762 10.375 150.194 35.504 46 14.297 11.210 41.823 26.818 6.769 10.256 150.268 35.273 47 14.266 11.197 41.607 26.719 6.766 10.382 150.257 35.079 48 14.259 11.184 41.371 26.654 6.765 10.386 150.050 34.953 49 14.237 11.138 41.095 26.553 6.765 10.369 149.961 34.760 50 14.271 11.149 40.873 26.474 6.766 10.324 149.942 34.576 51 14.220 11.130 40.649 26.345 6.768 10.361 149.531 34.435 52 14.207 11.111 40.436 26.295 6.767 10.348 149.626 34.300

53 14.212 11.088 40.058 26.167 6.764 10.320 149.360 34.009 54 14.202 11.052 39.878 26.046 6.770 10.336 149.420 33.885 55 14.255 11.072 39.566 25.983 6.771 10.341 149.307 33.732 56 14.250 11.035 39.207 25.822 6.774 10.322 149.420 33.427 57 14.210 10.962 38.883 25.682 6.767 10.331 149.181 33.163 58 14.138 10.916 38.449 25.514 6.778 10.310 149.044 32.818 59 14.100 10.849 38.041 25.346 6.780 10.331 148.966 32.556 60 14.101 10.824 37.591 25.171 6.780 10.346 148.914 32.266 61 14.106 10.769 37.137 24.999 6.782 10.345 148.451 31.939 62 14.130 10.775 36.798 24.864 6.779 10.354 148.335 31.690 63 14.182 10.828 36.262 24.693 6.785 10.295 148.349 31.352 64 14.172 10.826 35.830 24.506 6.783 10.344 148.370 30.980 65 14.159 10.737 35.343 24.279 6.786 10.326 148.079 30.579 66 14.074 10.649 34.817 24.008 6.787 10.316 147.977 30.162 67 14.089 10.666 34.404 23.865 6.788 10.313 147.927 29.836 68 14.050 10.623 33.907 23.623 6.791 10.230 147.537 29.485 69 13.988 10.537 33.521 23.405 6.793 10.303 147.307 29.157 70 13.958 10.522 33.174 23.263 6.792 10.289 147.289 28.908 71 14.028 10.568 32.770 23.105 6.793 10.284 147.248 28.626 72 14.025 10.567 32.431 22.954 6.791 10.283 146.831 28.349 73 14.043 10.573 32.112 22.818 6.789 10.303 146.941 28.092 74 14.024 10.559 31.854 22.683 6.798 10.240 147.095 27.869 75 13.997 10.521 31.585 22.533 6.797 10.320 146.893 27.604 76 13.958 10.484 31.228 22.359 6.793 10.295 146.561 27.312 77 13.944 10.468 30.940 22.218 6.794 10.294 146.458 27.078 78 13.930 10.470 30.715 22.116 6.801 10.299 146.741 26.920 79 13.877 10.446 30.502 21.991 6.804 10.272 146.619 26.730 80 13.823 10.403 30.307 21.892 6.805 10.291 146.528 26.584 81 13.845 10.394 30.055 21.769 6.802 10.285 146.310 26.393 82 13.876 10.412 29.944 21.712 6.806 10.221 146.426 26.322 83 13.869 10.398 29.737 21.621 6.806 10.277 146.360 26.169 84 13.876 10.394 29.546 21.533 6.803 10.282 146.105 26.024 85 13.894 10.459 29.375 21.472 6.801 10.262 146.034 25.918 R.26 Table R. 20: Transfer coefficients, loss coefficients and outlet temperatures ac cording to the different methods (LjI ~ 1.98 m). G. G. Me,/Lfi Me..JLfi MepfLfi K fdmlM K/dmJP TaoP TaoM 1 4.625 2.992 0.599 0.634 0.696 18.662 18.579 42.937 42.197 2 4.619 2.990 0.597 0.631 0.693 18.641 18.558 42.788 42.047 3 4.603 2.990 0.593 0.627 0.688 18.640 18.558 42.620 41.880 4 4.621 2.990 0.595 0.629 0.691 18.700 18.620 42.323 41.590 5 4.617 2.990 0.598 0.632 0.695 18.667 18.586 42.274 41.541 6 4.613 2.990 0.598 0.631 0.694 18.672 18.594 41.968 41.241 7 4.627 2.990 0.596 0.629 0.691 18.674 18.597 41.644 40.923 8 4.622 2.992 0.602 0.634 0.697 18.655 18.579 41.459 40.744 9 4.616 2.991 0.600 0.632 0.695 18.653 18.579 41.150 40.440 10 4.622 2.990 0.605 0.637 0.700 18.657 18.584 40.879 40.176 11 4.615 2.992 0.605 0.636 0.700 18.650 18.578 40.681 39.982 12 4.616 2.991 0.604 0.635 0.698 18.681 18.610 40.317 39.625 13 4.615 2.992 0.608 0.638 0.701 18.721 18.652 39.983 39.298 14 4.604 2.996 0.613 0.642 0.707 18.662 18.594 39.829 39.148 15 4.593 2.994 0.610 0.639 0.703 18.689 18.623 39.504 38.829 16 4.582 2.994 0.611 0.639 0.703 18.713 18.648 39.220 38.552 17 4.592 2.995 0.614 0.642 0.706 18.714 18.649 39.080 38.416 18 4.591 2.997 0.612 0.640 0.704 18.728 18.664 38.770 38.112 19 4.599 2.995 0.618 0.646 0.711 18.696 18.633 38.667 38.013 20 4.602 2.994 0.621 0.648 0.714 18.703 18.641 38.502 37.853 21 4.584 2.995 0.620 0.646 0.711 18.707 18.646 38.181 37.538 22 4.588 2.995 0.623 0.649 0.714 18.711 18.651 37.961 37.324

23 4.581 2.997 0.624 0.649 0.715 18.708 18.649 37.757 37.125 24 4.579 2.995 0.625 0.650 0.715 18.730 18.672 37.659 37.030 25 4.577 2.995 0.626 0.651 0.716 18.705 18.648 37.514 36.888 26 4.582 2.998 0.629 0.653 0.719 18.719 18.662 37.391 36.768 27 4.563 2.999 0.628 0.653 0.718 18.713 18.656 37.200 36.582 28 4.575 2.999 0.633 0.657 0.724 18.699 18.643 37.123 36.508 29 4.638 2.998 . 0.631 0.655 0.721 18.803 18.747 36.970 36.357 30 4.634 2.999 0.631 0.654 0.720 18.809 18.754 36.838 36.228 31 4.603 3.000 0.627 0.649 0.715 18.823 18.768 36.618 36.013 32 4.604 3.001 0.629 0.651 0.717 18.811 18.757 36.632 36.027 33 4.627 3.001 0.633 0.655 0.722 18.766 18.712 36.614 36.010 34 4.635 3.002 0.634 0.657 0.724 18.774 18.720 36.584 35.981 35 4.616 3.002 0.633 0.655 0.722 18.808 18.754 36.443 35.843 36 4.619 3.001 0.636 0.658 0.725 18.812 18.758 36.417 35.817 37 4.617 3.004 0.637 0.659 0.726 18.787 18.734 36.322 35.725 38 4.603 3.001 0.635 0.657 0.723 18.823 18.770 36.260 35.664 39 4.616 3.001 0.638 0.660 0.727 18.813 18.760 36.297 35.701 40 4.619 3.001 0.639 0.661 0.728 18.819 18.766 36.258 35.663 41 4.617 3.004 0.638 0.660 0.727 18.807 18.754 36.132 35.539 42 4.604 3.007 0.641 0.663 0.730 18.782 18.731 35.995 35.406 43 4.599 3.004 0.638 0.659 0.726 18.765 18.714 35.839 35.252 44 4.619 3.003 0.641 0.662 0.729 18.792 18.741 35.738 35.154 45 4.611 3.005 0.645 0.666 0.734 18.795 18.745 35.616 35.035 46 4.558 3.009 0.637 0.657 0.723 18.801 18.752 35.212 34.638 47 4.614 3.007 0.649 0.669 0.738 18.812 18.763 35.276 34.703 48 4.616 3.007 0.649 0.669 0.737 18.806 18.758 35.107 34.538 49 4.609 3.007 0.649 0.668 0.736 18.814 18.766 34.887 34.322 50 4.588 3.007 0.648 0.667 0.735 18.824 18.777 34.678 34.118 R.27 Table R. 20 (continue): Transfer coefficients, loss coefficients and outlet temp eratures according to the different methods (Lft = 1.98 m). G. Ga Me/Lfi Me,JLfi Mep/Lfi K fdmlM K(dmlP roop TooM 51 4.605 3.008 0.658 0.677 0.746 18.766 18.719 34.619 34.063 52 4.599 3.008 0.654 0.673 0.742 18.800 18.754 34.426 33.873 53 4.587 3.006 0.654 0.672 0.740 18.807 18.762 34.120 33.574 54 4.594 3.009 0.662 0.680 0.749 18.792 18.747 34.054 33.511 55 4.596 3.009 0.658 0.676 0.744 18.787 18.743 33.801 33.263 56 4.588 3.011 0.662 0.679 0.748 18.804 18.761 33.539 33.007 57 4.592 3.008 0.665 0.682 0.752 18.834 18.791 33.336 32.809 58 4.582 3.012 0.665 0.681 0.750 18.789 18.748 32.973 32.455 59 4.592 3.013 0.668 0.684 0.754 18.790 18.749 32.701 32.189 60 4.598 3.013 0.672 0.687 0.757 18.806 18.767 32.399 31.895 61 4.598 3.014 0.672 0.686 0.756 18.763 18.725 32.045 31.548 62 4.602 3.013 0.676 0.690 0.760 18.783 18.745 31.829 31.338 63 4.576 3.015 0.670 0.683 0.752 18.786 18.750 31.332 30.852 64 4.597 3.015 0.678 0.691 0.762 18.816 18.780 31.103 30.630 65 4.589 3.016 0.681 0.693 0.764 18.789 18.755 30.716 30.252 66 4.585 3.017 0.688 0.701 0.772 18.800 18.767 30.345 29.890 67 4.583 3.017 0.686 0.698 0.769 18.814 18.782 30.008 29.561 68 4.546 3.018 0.686 0.697 0.767 18.780 18.749 29.551 29.115 69 4.579 3.019 0.700 0.711 0.783 18.757 18.726 29.383 28.952 70 4.573 3.019 0.699 0.709 0.781 18.780 18.750 29.093 28.669 71 4.571 3.019 0.700 0.710 0.782 18.788 18.759 28.782 28.366 72 4.570 3.018 0.702 0.712 0.785 18.764 18.736 28.532 28.123 73 4.579 3.017 0.706 0.715 0.788 18.803 18.775 28.317 27.913 74 4.551 3.021 0.704 0.713 0.785 18.790 18.763 28.048 27.651 75 4.587 3.021 0.717 0.726 0.801 18.778 18.751 27.964 27.570 76 4.576 3.019 0.718 0.727 0.801 18.779 18.752 27.668 27.280 77 4.575 3.020 0.721 0.730 0.804 18.771 18.746 27.451 27.069

78 4.577 3.023 0.722 0.730 0.805 18.781 18.756 27.270 26.892 79 4.565 3.024 0.725 0.733 0.808 18.760 18.736 27.094 26.721 80 4.574 3.024 0.727 0.736 0.810 18.758 18.734 26.957 26.588 81 4.571 3.023 0.729 0.738 0.813 18.758 18.734 26.764 26.399 82 4.543 3.025 0.726 0.734 0.808 18.760 18.736 26.615 26.253 83 4.568 3.025 0.730 0.738 0.813 18.760 18.736 26.504 26.145 84 4.570 3.024 0.731 0.739 0.814 18.748 18.725 26.360 26.004 85 4.561 3.023 0.730 0.738 0.812 18.757 18.735 26.213 25.862 It can be seen from figures R.17 and R.18 that the 1.98 m high fill has the same trends as the 1.08 m high fill presented in section R.7. If a curve is fitted to the data of the Merkel approach in figure R.17 then find MeM / Lft = CT;,O.2774 where c is a constant and the correlation coefficient, ? = 0.9831. (R.8) R.28 60,--,--T""----,--,---r--.....,--,....---r---.I --Two _Twi ..??,.,,, Tai 1----1 , Twb ~ --- 40 r---+--+---l---I -~.........:::---l---If---+---l] ---..., l! "'-...~ 30 r-=r===p==t==+=~==r-lr----~t..-l~ ----------h. ~L 20 +----1----!----I----+---+----+----1---I----I 918171615141312111 10 +---+----+----+---+---+----1......--__--+----1 1 Test no. Figure R.15: Variation of air and water temperatures. 11 10 Ul -F 9 S .. ~ ~ 0 0:: 8II .. :E 7 6 - - ..., " ..... , ma -mw . ,. . ,"' " " , .... ' .". ,. " " 11 21 31 41 51 61 71 81 91 Test no. Figure R.16: Air and water mass flow rates during the fill test. R.29 OJ!;~ ~ 00

00A 000A <>~$<? '"-- O'~F!coOO~Q.v"0_~. xX> DC 'bD~~ v OQ0 O,?> fx>f' DXXxx~ ifill!" <m... ..xc -~~ """"'b'b::l'b c [ Dc'b c'i!J &' ""<~'So<'S<D Merkel ,,~~ xo Poppe "AO(~' xe-NTU I 0.85 0.80 0.75 "e j 0.70 ., :;: 0.65 0.60 0.55 25 30 35 40 45 50 55 Figure R.l7: Transfer coefficients according the e-NTU, Merkel and Poppe approac hes. .. D Experimental ~--Merkel and e-NlU -~ - ~_Poppe F V ~r ../ / Jt' 45.0 42.5 40.0 37.5 (,J 0 35.0Ii ..... 32.5 30.0 27.5 25.0 25 30 35 40 45 50 55 Figure R.l8: Measured air outlet temperatures compared to air outlet temperature s predicted by the e? NTU, Merkel and Poppe approaches. II I I I I R.30 R.9 FILL HEIGHT: 1.53 m, TEST REPEATED (24?C < Tal < 27?C) Table R.21: Experimental measurements (Po = 101340 Pal. Tai T.b T./ T.o mo m. dpft Tao ?C ?C ?C j ?C kg/s kg/s Pa ?C I 26.841 23.023 47.799 36.31l 2.634 6.294 19.064 43.201 2 26.154 22.388 47.778 33.573 3.879 6.184 37.21l 40.895

3 25.763 21. 767 47.614 31.349 5.191 6.293 61.082 38.808 4 25.850 21.525 47.267 29.635 6.593 6.299 99.143 37.042 5 26.289 21.647 47.052 28.433 7.91l 6.275 152.017 35.656 6 26.702 21.824 46.874 27.869 8.591 6.320 184.965 34.913 7 26.418 23.135 46.190 38.569 2.690 10.400 26.429 42.987 8 25.785 22.500 45.673 36.038 3.896 10.434 45.773 40.804 9 25.398 21.853 45.369 33.791 5.266 10.308 76.815 38.900 10 25.363 21.590 44.935 32.258 6.485 10.395 115.712 37.406 II 25.733 21.603 44.569 30.930 7.858 10.400 177.934 36.533 12 26.1l1 21.71l 44.201 30.186 8.609 10.351 223.670 35.744 13 26.256 23.495 43.573 38.989 2.646 15.260 37.792 41.623 14 25.670 22.864 43.369 37.073 3.882 15.237 59.495 40.668 15 25.074 21.986 42.926 35.125 5.217 15.152 94.270 39.142 16 24.837 21.468 42.715 33.722 6.534 15.182 142.662 37.943 17 25.236 21.475 42.479 32.564 7.931 15.156 219.553 37.149 18 25.546 21.586 42.212 31.935 8.650 15.202 275.353 36.587 Table R.22: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different methods (Lft = 1.53 m). G. Go Me/L, MeuiLs Mep/Ls KfdmJM KfdmJp Taop T/loM I 2.797 Ll71 0.490 0.501 0.566 19.301 19.244 42.448 41.753 2 2.748 1.724 0.604 0.620 0.686 17.355 17.304 39.756 39.151 3 2.797 2.307 0.728 0.750 0.821 15.981 15.938 37.815 37.274 4 2.800 2.930 0.844 0.866 0.940 16.185 16.147 35.941 35.460 5 2.789 3.516 0.967 0.982 1.061 17.315 17.279 34.649 34.215 6 2.809 3.818 1.054 1.061 Ll44 17.879 17.844 34.211 33.796 7 4.622 1.196 0.368 0.371 0.436 25.391 25.299 43.441 42.692 8 4.637 1.732 0.465 0.473 0.538 20.975 20.902 41.439 40.764 9 4.581 2.341 0.563 0.575 0.642 19.376 19.314 39.523 38.912 10 4.620 2.882 0.640 0.655 0.726 19.351 19.295 38.087 37.526 11 4.622 3.492 0.721 0.740 0.813 20.371 20.318 36.743 36.230 12 4.600 3.826 0.780 0.801 0.877 21.382 21.331 36.048 35.562 13 6.782 1.176 0.275 0.276 0.341 37.556 37.425 42.209 41.495 14 6.772 1.725 0.367 0.369 0.435 27.472 27.377 41.074 40.403 IS 6.734 2.319 0.448 0.453 0.518 24.238 24.159 39.599 38.973 16 6.747 2.904 0.508 0.514 0.580 23.501 23.430 38.375 37.787 17 6.736 3.525 0.565 0.574 0.641 24.654 24.586 37.238 36.691 18 6.756 3.844 0.608 0.619 0.688 26.035 25.967 36.755 36.229 R.31 Table R.23: Empirical relations fur the Merkel number according to the various m ethods (Lji = 1.53 m). Approach Eq. . Empirical relation Correlation tVlle coefficient I Me / L =0 786327 G--<l?590474GO.653933 0.9965eli' w a e-NTU 2 Me, / Lji =0.847612(Gw/GJ--<l620771 0.9941 Me, / Lft =1.831727 G';?231365 G~?709589 3 -1.149633 G';?1l5990 G~?728367 0.9971 1 Me / L =0 805764 G-O.592389 GO.65OS<6 0.9972M Ii? w a Merkel 2 MeM / Lft =O.863446(Gw/Gat620337 0.9952 Me / L =1 836449 G-o?161045 GO.645385M Ii' W Q 3 -1.145760 G';?032776 G~643715 0.9983 I M. / L =0884005 G-0547641GO.599196 0.9966ep fi' w a Poppe 2 Mep / Lft =O.939648(Gw/GJ--<l?572564 0.9947

Mep / Lft =1.863314 G';?147917 G~?657350 3 -1.118364 G;:'?012514 G~?677998 0.9977 1.2 ? GIN =6.75 kgtm"s 0 GIN =4.61 kgtm"s 1.0 l:>. GIN =2.79 kglm's --Equation type 1 ....... Equation type 2 '; 0.8 _Equation type 3 E j -Gl ::E 0.6 0.4 -t--=.,.,.~--:-~,..r..~=-+~~-I-~~--I-~~-j~--I 4.54.03.53.02.52.01.5 0.2 -1----1----l-----4----I----+----I------l 1.0 Ga. kg/m's Figure R.19: Comparison of experimental data and empirical equations (Lji ~ 1.53 m). " R.32 Table R24: Empirical relations for the loss coefficient according to the various methods (Lji ~ 1.53 m). Approach Eq. Empirical relation Correlationtype coefficient I K =11 804507 aO.541354 a-D?244939 0.8282fdml' w a Merkel and 2 K =17 144755(a fa )0362046fdml' "If a 0.7133 e-NTU K =3 699802 a1.l39958a-1.923973fdml' w 3 + 6.599771 a:?320485a~?454993 0.9833 I K =11 774221 aO.540593a-D243942 0.8272fdml' w a 2 K =17 105425(a fa )0361175Poppe fdml' w a 0.71\8 K =3695143 a1.139140 a -1.923604fdml' w a 3 J +6.575536 a:31%65a~456907 0.9832 4.54.03.53.02.52.01.5 \. <> Gw =6.75 kglm"s0 Gw =4.61 kgfm"s ~ A Gw =2.78 kglm"s--Equation type 1 '.~ ..... '. Equation type 2~..? ~ ~Equation type 3 (> '. "~. ' ...-. .. ????v. . -"' .. '. . . .. "'"' . -' .. '" .. ~ '. -. 0 . . - . - . -.... u .. . .-- - ? o. __ '. - - .?? :.:t-'._?1 -'- . ... .. -... '. -.. --. -"'-. 40 30 15 20 35 10 1.0 "'e .; 25 E ~ Ga. kglm's

Figure R20: Comparison of experimental data and empirical relations for the loss coefficient (Lji ~ 1.53 m). R.33 R.10 FILL HEIGHT: 1.53 m, TEST REPEATED (32?C < Tal < 3S 0c) Table R.25: Experimental measurements (P, = 99780 Pa). TaJ T.b TWi T.o ma m. dpji Ta, ?C ?C ?C j ?C kg/s kg/s Pa ?C 1 33.007 25.247 48.274 37.108 2.536 6.260 19.140 44.346 2 32.755 24.599 48.231 34.175 3.903 6.283 39.523 41.939 3 32.744 24.084 48.075 32.139 5.135 6.276 63.230 40.081 4 33.312 23.912 47.844 30.614 6.406 6.279 98.328 38.554 5 34.129 24.172 47.650 29.629 7.534 6.264 143.074 37.549 6 34.687 24.333 47.413 29.031 8.261 6.101 177.376 36.865 7 33.015 25.417 ,,'46.486 39.200 2.555 10.299 25.532 44.121 8 32.807 24.734 46.105 36.614 3.831 10.293 46.271 42.157 9 32.755 24.117 45.869 34.658 5.061 10.311 74.017 40.623 10 33.069 23.768 45.428 32.930 6.401 10.295 117.304 39.091 11 33.946 24.012 45.131 31.651 7.808 10.266 182.734 37.803 12 34.785 24.289 44.706 30.744 8.797 10.231 250.738 36.903 13 32.842 25.418 43.936 39.522 2.625 15.336 38.913 42.569 14 32.724 24.904 43.687 37.671 3.841 15.299 61.385 41.476 15 32.561 24.103 43.327 35.800 5.178 15.295 99.067 40.180 16 32.827 23.682 43.127 34.499 6.430 15.300 146.638 39.133 17 33.649 23.944 42.877 33.416 7.731 15.203 219.963 38.237 18 34.614 24.169 42.674 32.192 9.064 15.116 347.512 37.745 Table R.26: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different methods (Lji = 1.53 m). G. Ga MejLfi MeM/Lfi Me,.JLfi KfdmlM KfdmlP TaoP raoM 1 2.782 1.127 0.496 0.507 0.578 20.289 20.225 43.383 42.675 2 2.793 1.735 0.632 0.651 0.723 17.721 17.665 40.778 40.ll2 3 2.789 2.282 0.753 0.775 0.851 16.494 16.442 39.145 38.242 4 2.791 2.847 0.876 0.S96 0.975 16.574 16.519 37.991 36.709 5 2.784 3.348 1.009 1.017 1.101 ,17.494 17.432 37.348 35.735 6 2.7ll 3.671 1.107 1.100 1.185 18.088 18.021 36.888 34.966 7 4.577 1.136 0.381 0.384 0.463 26.448 26.348 44.270 43.512 8 4.575 1.703 0.483 0.491 0.565 21.371 21.292 42.216 41.530 9 4.583 2.249 0.573 0.585 0.660 19.708 19.639 40.697 39.981 10 4.576 2.845 0.663 0.680 0.757 19.659 19.591 39.339 38.407 II 4.563 3.470 0.763 0.784 0.865 20.664 20.592 38.444 37.236 12 4.547 3.910 0.865 0.888 0.976 22.386 22.308 37.915 36.541 13 6.816 1.167 0.274 0.275 0.346 38.321 38.182 42.699 41.974 14 6.799 1.707 0.364 0.367 0.439 28.280 28.178 41.582 40.899 15 6.798 2.301 0.453 0.458 0.532 25.253 25.164 40.401 39.674 16 6.800 2.858 0.512 0.519 0.591 24.356 24.270 39.466 38.591 17 6.757 3.436 0.581 0.590 0.665 25.342 25.254 38.757 37.689 18 6.718 4.028 0.702 0.717 0.803 29.169 29.069 38.324 37.183 R.34 Table R27: Empirical relations for the Merkel numbllr according to the various m ethods (Lft = 1.53 m). Approach Eq. Empirical relation Correlationtype coefficient 1 Me / L =0 802686 G -0.596023G 0.693986 0.9943 e fi' w a e-NTU 2 Me, / Lft =0;900275(G w /0.)-0641703 0.9889 Me, / Lft =1.271927 G:?263657G~?667591

3 i - 0.554315 G:?036227 G~?650214 0.9950 1 Me M / Lft =0.814045 G:?588025G~?685956 0.9939 Merkel 2 MeM / Lfl =0.913014(Gw /G.)-o?633821 0.9883 Me / L =1292572 G-0.145482GO.59991I M .Ii' w a 3 - 0.589198 G~?1l3166G~?550082 0.9963 1 Me / L =0 894579 G-o?533702GO.624663 0.9924P fi' W Q Poppe 2 Mep / Lfl =0.995347(Gw /G.)-o576918 0.9865 Mep / Lfl =1.325635 G~O.126711G~?585201 3 _ 0.558443 G~.l39244G~?559401 0.9943 1.0 1.21r""'======;----,---I---,--I----, <> GN = 6.78 kg/m's o GN = 4.57 kg/m's A GN = 2.78 kglm's1----+-----+----0-;,;1''''-+----+------1 --Equation type 1 ..... - - Equation type 2 _Equation type 3 4.54.03.53.02.52.01.5 0.2 -1-----1-----1----1----4----1----4------1 1.0 Ga, kg/m's Figure R.21: Comparison ofexperimental data and empirical equations (Lfi = 1.53 m). R.35 Table R,28: Empirical relations for the loss coefficient according to the variou s methods (LjI = 1.53 m). Approach Eq. Empirical relation Correlationtype coefficient I K =11 636035 GO.556388G 0.217645 0.8282fdml' w a Merkel and 2 K =17.781634(G /G )0.351490t;' fdml w a 0.7133 e-NTU K =7 807807 GO. 832OOl G,1.l940l0fdml' w a 3 + 2.698714 G~?254849G~?093894 0.9833 I K =11 598391 GO.556158G-o?217"6 0.8272fdmi' w a 2 K =17 721574(G /G r351331Poppe fdml' w a 0.7118 K =7 781392 GO.83l737 G'1.l94921fdml' w a 3 + 2.694334 G~?255441G;?092209 0.9832 <> 0 Gw - 6.78 kglm"s \ 0 Gw =4.57 kg/m"s ~ l>. Gw =2.78 kglm"s--Equation type 1~ .... ' .. Equation type 2 .'~ ----Equation type 3 0~"." .. ... ' .. r--'. :.::.:.:.::;...... -'"'. ~. '. "~. . '9..... ~'- .""'- ". . --.'. -'. "'. . ' .. .. .....~.. '. . .. . '. -. 1\ ' .. . " .. ~ " '.\-!... u ' .. ..~ ' ........ .... ' ...... - ...A....... A ".6. .. ' ... . " ... 40

35 30 ";E E ~ ll:: 25 20 15 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kglm"s Figure R.22: Comparison of experimental data and empirical relations for the los s coefficient (LjI ~ 1.53 m). R.36 R.ll THE EFFECT OF AIR TEMPERATURE ON FILL PERFORMANCE Fill test results of the 1.53 m high trickle fill are presented in sections R.3, R.9 and RIO respectively for low, intermediate and high tem,peratures. The test results are combined into one set of data and he effect of the air temperature on fill performance can subsequently be evaluated. Table 29 summarizes the empirical equations of the Merkel numbers according to the Merkel approach of th e 1.53 m high trickle grid fill for cool, mils and hot ambient conditions presented in sections R.3, R 9 and R.l 0 respectively. The empirical equations generated from the combined set of data are also present ed in table R.29. Table R29: Summary of the transfer coefficients according to the Merkel approach . Ta.oC Equation type I r 16.0 Me I L =0 817071 G-o.581055GO.670746 0.9948M fi' w a 25.8 MeM / Lft =0.805764 G:?592389G~?650846 0.9972 33.3 MeM / Lft =0.814045 G:?588025G~?685956 0.9939 Comb. Me / L =0811211 G-o?587437GO.671265 0.9896M fi' 'If a T... oC Equation type 2 ,16.0 MeM / Lft =0.910464(Gw/GJ-0624555 0.9903 25.8 MeM / Lft =0.863446(Gw /Ga )-0620337 0.9952 33.3 MeM / Lft =0.913014(Gw/GJ-0633821 0.9883 Comb. MeM / Lft =0.896180(Gw/Gat062758S 0.9855 Tllh ?C Equation type 3 ,16.0 Me / L =1 299681 G-o?l2lS55GO.572612 - 0593218 GO.146139GO.515254 0.9979 Mfi' wa ? wa 25.8 Me M / Lft =1.836449 G:1l68045G:?64l38l -1.145760 G;:?032776G:?64371S 0.9983 33.3 Me M / Lft =1.292572 G:?l4S482G:S99911 - 0.589198 G:?ll3166G:?S50082 0.9963 Comb. Me M / Lft =1.281200 G:?165354 G:?600120 _ 0.576672 G~?08993 G:?lS6793 0.9916 It is evident from table R.29 that the experiments are repeatable as the respect ive coefficients of each type ofempirical equations are numerically within close tolerance of one another. The correlation coefficients for the empirical equations obtained from the combined data set is of the same o rder of accuracy as the other empirical equations obtained separately for the cool, mild and hot air tem peratures. This suggests, as a first approximation, that the air temperature does not influence the Merkel

number or transfer characteristic significantly. The combined data set, as a function of the inlet air drybulb temperature, is co rrelated by the relation M, /L = 0 835550 r;-<J.587470,-,0670643T-<J?OO9106 eM fi' w u a QI where the correlation coefficient,?, is 0.9897 and To, is' expressed in ac. (R.9) When the effects ofthe water inlet temperature and inlet air drybulb are include d then find the empirical relation R.37 Me / L = I 952087 G-o?611Sl3GO.664630r-o?OI011lr-o?210834 Mfi" 11' a at wi where the correlation coefficient, 7', is 0.9894. Toi and Twi are expressed in ? C. (R.IO) As a function of the inlet air wetbulb temperature, the combined data set is cor related by the relation Me / L = 0881611 G-O.586797GO.667188r-O.027211 M fi' 11' a wb where the correlation coefficient, 7', is 0.9901 and Tw? is expressed in ?C. (R.ll) When the effects of the water inlet temperature and inlet air wetbulb are includ ed then find the empirical relation Me / L =1 954254 G-o?60?244GO.6614IOr-o?028767r-o?l.736' Mfi' 11' a wb WI where the correlation coefficient, 7', is 0.9899. Tw? and Twi are expressed in ? C. (R.12) It is evident from equations (R.9) tQ (R.12) that the Merkel number, according t o the Merkel theory, is a relatively weak function of the inlet air drybulb and wetbulb temperatures. This is because the exponents of Toi and Twi in equations (R.9) to (R.12) are relatively small. Similar trends are observed for the Poppe and e-NTU theories. The exponents of Twi in equations (R.IO) and (R.12) are appr oximately equal to the exponent of Twi in equation (R.6), where the effect of the inlet water temperatu re on the Merkel number is determined explicitly. The measured Merkel number or transfer characteristic per unit length offill ver sus the right hand side of equation (R.IO), where the effect of the air temperature is omitted is shown in figure R.23. Figure R.24 shows the measured Merkel number versus equation (R.I 0). Figures R.23 and R.24 are virtually identical and there is no temperature effect visible in figure R.23. Therefore, the inlet air drybulb temperature has no significant effect on the Merkel number. The loss coefficient, for the combined data set, can be correlated by the relati on KI""" = 4.587403 G~?0l2197G:1.7"'26+ 5.513909 G~?322716G~"9124 with a correlation coefficient 7' = 0.9721. (R.13) As a function of the inlet air drybulb temperature, the loss coefficient can be correlated by the relation K =(4 84878 GI.036998G-I.808134 + 5 489596 GO.347548GO.,.,861 \"-0.00066 (R.14)f dml' 11' a . 11' a J1 QI

with a correlation coefficient 7' = 0.9623. It can be seen that the loss coefficient is a very weak function of the inlet ai r drybulb temperature, as the exponent of Toi in equation (R.5) is very close to zero. The loss coefficients p resented in figures R.5, R.20 and R.22 for air inlet drybulb temperatures of 16,25.8 and 33.3 ?C respectively are shown, together with equation (R.13), in figure R.25. It can be seen in figure R.25 that the air inle t temperature has no significant effect on the loss coefficient. R.38 ? Tai =16:0?C .1"flI Tai = 25.8 ?C A ATai = 33.3 ?C lEI .. ?Do I~L...~ ~ ~~ k~~ [,.. ,.,.'" Jt .~ 1.2 1.1 0.9 ~ 0.8 's ~ .... 0.7~ ~ 0.6 0.5 0.4 0.3 0.2 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1 952087 a-D?611513aO.664630T-o.210834 . w a ~ Figure R.23: The measured Merkel number per unit length of fill versus right han d side of equation (R.lO) where the temperature ofthe air is omitted. .Tai 16.0 ?C .1lJEI Tai = 25.8 ?C f!,. Tai = 33.3 ?C liIlI ... ?Do L.; l!lI ~. ", A~ ftJT ,~I .hiJl~ 4lfAil 41 -.lj~ 1.2 1.1 1 0.9 ~ 0.8 's ~ .... 0.7 ~ ~ 0.6

0.5 0.4 0.3 0.2 0.2 0.3 0.4 0.5 1.952087 Q6 07 0.8 Q9 a-0.611513 a 0.664630 T-D?010111 T-D?210834 w a ~ M 1 1.1 1.2 Figure R.24: The measured Merkel number per unit length of fill versus right-han d side of equation (R.IO). ~ ( ? Tai =16.0 ?C III Tai =25.8?C '\. /),. Tai =33.3 ?C Gw '" 6.7 kg/m2s - Equation (R.13) ?, I:'> 'ia""~ A "" Gw '" 4.5 kg/m2s lUI? """li?I '" ... A ar:.o.. f', ? 1m , w_ LJlI ......A ~ ~ A IilL~. A , Gw '" 2.8 kg/m2s I I 40 35 30 "; E ~ 25e J2 :ll:: 20 15 10 1 1.5 2 R.39 2.5 3 3.5 4 4.5 Ga. kglm"s Figure R.25: The loss coefficients at different air inlet temperatures. S.I APPENDIX S SPLASH FILL PERFORMANCE TEST RESULTS S.l INTRODUCTION The performance characteristics of a splash fill are determined experimentally f or four different fill spacings as shown in figure S.l. The results are critically evaluated and presen ted by extended empirical equations. The height ofthe spray zone above the fills is 0.15 m. t.n=0.1 m

25 layers Lft=2.5m (a) t.n = 0.2 m 15 layers Lft=3m _.~~~~~~---~~-'E~~&-~~ ~~~;~~::j'?;""fZ.-:=--72:=.z:=.~% ~ -~~:s-~~~- ? .~~~~~%:?~-: (b) t.n = 0.3 m 10 layers Lft=3 m (c) 6ft = 0.4 m 8 layers Lfi=3.2m (d) Figure S.1: Four different fill spacings for splash pack fill. The test results, for the fills shown in figure S.I(a), (b), (c) and (d), are gi ven in sections S.2, S.3, S.4 and S.5 respectively. The measured data are shown in a table for each test with anot her table showing the corresponding transfer and loss coefficients. The empirical relations for the tr ansfer and loss coefficients for each test are given in separate tables. The empirical relations of the trans fer and loss coefficients for the Merkel approach are compared to the measured data and shown in separate figu res for each test. S.2 S.2 FILL SPACING: 0.1 m Table S.l: Experimental measurements (Pa ~ 100490 Pal. Tai TWb T?? Two rna rn. dpft Tao ?e ?e ?e ?e kg/s kg/s Pa ?e 1 17.547 15.807 43.409 33.007 2.694 6.171 15.103 37.487 2 17.389 15.480 43.126 30.372 4.104 6.208 29.613 34.651 3 17.361 14.719 42.733 28.239 5.492 6.179 45.373 31.971 4 17.635 14.716 42.600 26.934 6.694 6.121 68.038 30.184 5 18.015 14.785 42.312 25.678 8.074 6.100 100.366 28.959 6 18.367 14.889 42.034 24.894 9.223 6.068 122.673 28.323 7 18.607 17.408 40.973 35.663 2.588 11.860 30.137 36.967 8 18.224 16.550 40.705 33.340 3.929 11.719 47.731 36.630 9 17.740 15.700 40.327 31.198 5.392 11.695 73.401 34.939 10 17.715 15.104 39.708 29.451 6.716 11.566 114.345 33.060 11 17.950 14.940 39.449 28.147 8.003 11.435 159.105 32.011 12 18.274 14.943 39..134 27.217 9.242 11.309 193.437 30.990 13 18.587 17.619 38.257 34.511 2.636 14.546 43.420 35.228 14 17.766 16.844 38.265 32.721 4.015 14.389 65.086 35.181 15 17.296 15.907 38.124 31.037 5.376 14.201 93.659 34.607 16 17.224 15.377 37.843 29.525 6.748 13.975 137.129 33.581 17 17.509 14.896 37.475 28.107 8.173 13.574 183.570 31.404 18 17.706 14.622 37.009 26.946 9.502 13.363 233.049 30.828 Table S.2: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (Aft = 0.1 m). G. Ga Me/Lft Me.,lLft MeP/Lf K/<JmIM K/<JmIP TaoP To.OM kg/m's kglm's m?1 m?1 m? l m? l m?1 ?e ?e

1 2.743 1.197 0.284 0.288 0.321 9.805 9.795 37.633 36.979 2 2.759 1.824 0.345 0.351 0.385 8.045 8.031 34.622 34.061 3 2.746 2.441 0.398 0.407 0.442 6.859 6.847 32.131 31.635 4 2.721 2.975 0.442 0.452 0.488 6.917 6.905 30.576 30.127 5 2.711 3.589 0.492 0.503 0.541 7.022 7.011 29.139 28.736 6 2.697 4.099 0.525 0.536 0.576 6.590 6.580 28.062 27.691 7 5.271 1.150 0.182 0.183 0.213 20.361 20.317 38.584 37.901 8 5.208 1.746 0.243 0.245 0.276 13.881 13.848 36.825 36.197 9 5.198 2.396 0.299 0.303 0.336 11.330 11.303 35.067 34.491 10 5.140 2.985 0.342 0.347 0.382 11.414 11.389 33.316 32.789 11 5.082 3.557 0.385 0.391 0.428 11.211 11.188 32.105 31.616 12 5.026 4.107 0.412 0.419 0.456 10.251 10.231 30.898 30.443 13 6.465 1.172 0.158 0.158 0.186 28.177 28.119 36.504 35.892 14 6.395 1.784 0.226 0.227 0.259 18.141 18.100 35.633 35.054 15 6.312 2.389 0.278 0.280 0.314 14.567 14.534 34.434 33.885 16 6.211 2.999 0.323 0.326 0.361 13.571 13.542 33.129 32.617 17 6.033 3.633 0.362 0.366 0.402 12.432 12.408 31.644 31.168 18 5.939 4.223 0.396 0.401 0.438 11.722 11.701 30.384 29.941 8.3 Table 8.3: Empirical relations for the transfer characteristic according to the various methods (.1.jl ~ 0.1 m). Approach Eq. Empirical relation CorrelationtvDe coefficient I Me / L = 0364409 0-0.4228240?571340 0.9900e fl' w a e-NTU 2 Me, / Lft =0.436402(Ow/O.)-o,496508 0.9713 Me / L =1 127969 0-0,071776 0?,255178 e ji' w a 3 _ O.752887 0~?0634690~.l45449 0.9981 1 Me / L =0 374457 0-0,43552?0?,577401 0.9900M fi' 'IV a Merkel 2 MeM / Lft ,,= 0.444827(Ow/0 .)-o,505736 0.9735 MeM / Lft =1.134627 0:0863980~254250 3 _ 0.747394 0~,0540820~,139992 0.9979 1 Mep / Lft = 0.409266 0:,4049030~.540131 0.9913 Poppe 2 Mep / Lft =0.482331(Ow/0 .)-o,472575 0.9741 Me / L = 1 152103 0-0,0778990?,255908 p ft? 'IV Q 3 - O.733838 0~?0656510~,142894 0.9980 4.54.03.53,0252.01,5 <> GN = 6.23 kg/m"s o GN = 5.15 kg/m"s 6. GN =2,73 kglrn"s -I--~~-::.l-'~--+---+---+------1__Equation type 1 -- - .... Equation type 2 -t--'>L----1I-----+-----t-------I------1 -Equation type 3 0.55 0.50 0.45 0.40 'E 0.35 ,;

...I - 0.30 ":E 0.25 0.20 0.15 0.10 1.0 Ga, kg/m"s Figure 8.2: Comparison of experimental data and empirical equations (.1.p = 0.1 m). r 8.4 Table S.4: Empirical relations for the loss coefficient according to the various methods (Aft ~ 0.1 m). Approach Eq. Empirical relation CorrelationtvDe coefficient I K = 4 292617 GO.997913G-o?593524 0.9446fdmlM' w a Merkel and 2 K/dmiM =7.439312{Gw /GJ07l0454 0.8975 e-NTU K = 0 257516 G2.38&3ooG,2.303946fdmlM' w a 3 + 3.729301 G~?646977 G:?041177 0.9947 1 K = 4 289381 GO.997176G-o.5934961 0.9445fdmlP' w a 2 K = 7 425689{G /G )0.710230Poppe fdmIP': w a 0.8974 K = 0 173797 G2.576711 G'2.444I94 fiJmlP' w a 3 +3.881804 G~?668754G:093127 0.9948 \ <> Gw =6.23 kg/m"s0 Gw =5. 15 kg/m"s ~ 6. Gw =2.73 kg/m"s f------Equation type 1;';'-., '-~ - - - - - - - Equation ty pe 2 ~~ Equation type 3 f----~.- f-~~ --.... - ~ ........ . . .. -.. ' .. -... ' .. In.. . 0 . . V' .. . . .. .. '- ". ~ ....t-- "-. ' .. ' .. po, .. ... ' ........ -- . - . .. '. " "" - ... A ...... u-........ --- - --30 25 20 "E E .J 15 10 5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Ga. kglm"s Figure S.3: Comparison ofexperimental data and empirical equations (Aft ~ 0.1 m) . S.5 S.3 FILL SPACING: 0.2 m Table 8.5: Experimental measurements (P. = 100480 Pa). T", T.b T., T.o rna m. dpfi Too DC ?C DC DC kg/s kgls Pa DC 1 14.808 12.959 46.443 34.233 2.866 6.553 16.735 39.922 2 14.476 12.504 46.526 31.783 4.103 6.510 23.930 37.014 3 14.780 11.980 46.610 29.816 5.404 6.525 29.827 34.819 4 14.926 11.724 46.594 28.285 6.727 6.521 43.791 32.964 5 15.439 11.917 )'46.575 26.968 8.171 6.516 71.501 31.778 6 15.977 12.262 46.610 26.080 9.487 6.515 91.641 31.027 7 15.816 14.085 46.847 38.104 2.921 10.636 25.780 41.861 8 15.712 13.715 46.821 35.948 4.024 10.574 34.316 40.383 9 15.299 12.465 46.824 33.603 5.413 10.631 44.740 38.976 10 15.393 12.268 46.812 31.844 6.765 10.620 66.140 37.537 11 15.819 12.282 46.772 30.322 8.165 10.559 102.194 36.438 12 16.254 12.537 46.669 29.082 9.522 10.575 131.165 35.218 13 17.439 17.101 46.623 40.502 2.769 15.177 40.278 43.617 14 16.415 16.103 45.735 37.730 4.130 15.203 51.796 41.755 15 15.515 13.849 43.883 34.693 5.455 15.204 67.582 39.370 16 15.558 13.317 42.794 32.741 6.764 15.211 94.772 38.324 17 15.936 12.662 40.830 30.500 8.128 15.222 134.642 36.361 18 16.314 12.861 38.592 28.323 9.456 15.226 172.635 33.841 Table 8.6: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (~ = 0.2 m). G. Go Me,/Lfl Me,jLfl MeplLj KjdmlM KjdmlP TaoP TaoM kglm's kg/m's m'l m?1 m?1 m-I m-I DC DC 1 2.912 1.274 0.216 0.220 0.244 8.239 8.231 39.840 38.293 2 2.893 1.824 0.255 0.261 0.285 5.596 5.587 36.182 35.518 3 2.900 2.402 0.291 0.298 0.324 3.972 3.965 33.894 33.290 4 2.898 2.990 0.322 0.330 0.357 3.732 3.724 31.921 31.370 5 2.896 3.632 0.356 0.365 0.394 4.115 4.107 30.326 29.825 6 2.896 4.217 0.385 0.394 0.423 3.910 3.902 29.194 28.732 7 4.727 1.298 0.147 0.148 0.168 11.717 11.687 41.499 40.655 8 4.700 1.788 0.181 0.184 0.205 8.101 8.078 39.676 38.895 9 4.725 2.406 0.220 0.224 0.247 5.799 5.781 37.762 37.033 10 4.720 3.007 0.253 0.259 0.283 5.456 5.440 36.167 35.492 11 4.693 3.629 0.285 0.292 0.318 5.782 5.766 34.704 34.078 12 4.700 4.232 0.316 0.324 0.352 5.459 5.444 33.587 33.004 13 6.745 1.231 0.112 0.113 0.131 19.640 19.569 43.270 42.403 14 6.757 1.836 0.150 0.151 0.171 11.345 11.305 40.955 40.167 15 6.757 2.424 0.184 0.186 0.208 8.542 8.514 38.051 37.330 16 6.761 3.006 0.211 0.213 0.236 7.803 7.780 36.006 35.345 17 6.765 3.613 0.240 0.243 0.267 7.725 7.705 33.523 32.930 18 6.767 4.203 0.281 0.285 0.312 7.355 7.339 31.492 30.975 r , 8.6 Table S.7: Empirical relations for the transfer characteristic according to the various methods (~ft ~ 0.2 m). Approach Eq. Empirical relation Correlationtype coefficient 1 Me / L .=0 289737 G-Q4%763 GO.58378S e 14,'? w a 0.9876 e-NTU 2 Me, / Lft =0.322738(Gw/GatS41922 0.9640 Me / L =1 072707 G-Q,08l663GO,243658

eft? w a 3 - O.782625 G~?031I22G2?1S6002 0.9909 I MeM / Lft =0.298810 G;;o,S09221G2.590240 0.9872 Merkel 2 MeM / Lft =0.330365(Gw/GJ-O,S51192 0.9653 MeM / Lft =1.077103 G;;o,078787G~,246703 3 - 0.779099 G~?039S66G2,15708S 0.9920 I Mep / Lft =0.325359 G:?483l28G~.560198 0.9871 Poppe 2 Mep / Lft =0.357981(Gw /Gj'?52326o 0.9640 3 Mep / Lft =1.092143 G:?077I1b~24l133 _ 0.767756 GO 046927 GOlSOl83 0.9905 w a -""'ir"" .J;] .' . .. ' . .. ' . ..J~ ....... -. -...~ . .--V .. ' ."~..... ....-cr 0 Gw 6.76 kglm's -.....,.,. 0 Gw - 4.71 kg/m's t;:::- :;....- !'>. Gw =2,90 kg/m's Equation type 1 ....... Equation type 2 I. Equation type 3 0.45 0.40 0.35 0.30 "'E 0.25 .; ..J 0.20.. :;; 0.15 0.10 0.05 0.00 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m's Figure S.4: Comparison of experimental data and empirical equations (!? = 0.2 m) . S.7 Table 8.8: Empirical relations for the loss coefficient according to the various methods (Aft ~ 0.2 m). Approach Eq. Empirical relation Correlationtype coefficient I K = 430145 GO.921207 G-o?87201 0.9543[dmlM' 141 a Merkel and 2 K fdm1M =4.23192(Gw/GJo900392 0.9518 e-NTU K = 3 179688 Gl.083916G-1.965418fdmlM' W Q 3 +0.639088 G~?684936G:?642767 0.9932 I K = 4 30017 GO.93230 G-o?87192 0.9548fdmlP' 141 a ~ / t 89931

Poppe 2 K fdmlP = 4.22661 Gw Ga 0.9590 K = 3 196078 GI.079447 G-l.964637fdmlP' 141 a 3 + 0.633738 G~685777 G~?645903 0.9932 0 GN =6.76 kg/m"s 0 GN =4.71 kg/m"s I-~ A GN =2.90 kg/m"s--Equation type 1 ------- Equation type 2 I-~ --Equation type 3~ ~' .. ~ -- .~, ~ A .-.~ ,:..;- .-. -.'. . -~ .W A .... ~ 25 20 15 ':"E E ~ 10 5 o 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga, kg/m"s Figure 8.5: Comparison ofexperimental data and empirical equations ('? = 0.2 m). S.8 S.4 FILL SPACING: 0.3 m Table S.9: Experimental measurements (Po = 101560 Pa). To; Twh TWi Twa mo mw dpft Tao ?e ?e ?e ?e kg/s kg/s Pa ?e 1 13.827 11.929 45.383 33.375 2.949 6.027 17.571 38.346 2 13.725 11.536 45.289 31.337 4.074 6.031 23.192 34.870 3 13.901 11.064 44.919 29.214 5.485 6.027 28.215 31.875 4 14.079 10.934 44.929 27.691 6.850 6.035 40.289 30.083 5 14.506 11.153 44.628 26.332 8.159 6.034 68.247 28.881 6 14.994 11.344 44.224 24.988 9.592 6.036 92.966 28.248 7 14.997 13.496 43.782 36.839 2.825 10.657 28.216 39.245 8 14.652 12.835 43.471 34.368 4.110 10.631 35.996 37.757 9 14.186 11.609 43.004 31.979 5.516 10.632 47.188 35.561 10 14.032 11.067 42.770 30.243 6.753 10.637 67.561 34.352 11 14.386 11.054 42.706 28.758 8.126 10.659 99.516 33.176 12 14.851 11.273 42.528 27.597 9.494 10.624 127.475 32.257 13 16.460 16.053 42.371 37.547 2.858 15.228 42.361 38.753 14 15.459 14.793 42.324 35.762 4.114 15.205 53.616 38.233 15 14.653 12.668 42.220 33.941 5.457 15.214 69.178 37.491 16 14.035 11.648 41.950 32.333 6.754 15.220 96.462 36.543 17 14.166 11.268 41.666 30.722 8.237 15.199 136.104 35.288 18 14.478 11.315 41.455 29.307 9.527 15.213 170.933 34.642 Table S.10: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different methods (6.fi = 0.3 m). Gw Go Me/Lft Me;JLft Mep/Lf KjdmlM K/<JmlP TaoP TooM kg/m's kg/m's m,l m'l ?1 m,l m,l ?e ?em. 1 2,679 1.310 0.210 0.213 0.235 8.241 8.235 36.765 36.058 2 2.680 1.811 0.240 0.245 0.267 5.581 5.574 34.043 33.410 3 2.679 2.438 0.277 0.282 0.306 3.706 3.700 31.308 30.747 4 2.682 3.045 0.311 0.318 0.343 3.361 3.356 29.535 29.025 5 2.682 3.626 0.347 0.355 0.382 3.997 3.990 28.174 27.713 6 2.683 4.263 0.389 0.398 0.427 3.939 3.932 26.936 26.519 7 4.737 1.256 0:129 0.130 0.146 13.880 13.852 38.345 37.576 8 4.725 1.827 0.172 0.174 0.193 8.282 8.264 36.487 35.784

9 4.725 2.451 0.214 0.216 0.238 6.000 5.986 34.492 33.843 10 4.727 3.001 0.250 0.254 0.277 5.709 5.695 33.185 32.578 11 4.737 3.611 0.287 0.293 0.318 5.793 5.780 32.094 31.528 12 4.722 4.220 0.317 0.323 0.350 5.437 5.425 30.937 30.412 13 6.768 1.270 0.108 0.109 0.125 19.824 19.773 38.991 38.248 14 6.758 1.828 0.142 0.142 0.160 12.085 12.052 37.662 36.953 15 6.762 2.425 0.171 0.172 0.192 8.869 8.844 36.126 35.435 16 6.764 3.002 0.199 0.201 . 0.222 8.076 8.053 34.796 34.139 17 6.755 3.661 0.233 0.236 0.259 7.663 7.642 33.602 32.985 18 6.761 4.234 0.273 0.278 0.304 7.187 7.169 33.025 32.441 8.9 Table S.II: Empirical relations for lbe transfer characteristic according to lbe different methods (8j, = 0.3 m). Approach Eq. Empirical relation Correlationtype coefficient I Me / L ,;, 0 243744 G-o?454727 GO.645462 0.9799eft,? wa / 6 ( / )-0542662 e-NTU 2 Me, Lfi =0.306 90 Gw Ga 0.9536 Me / L =1 035456 GO.085019GO.36175S e fi' w a 3 _ 0.805507 G~181184 G~?J13734 0.9916 I Me / L =0 249013 G-0 464089 GO.653578 0.9794M fi' w a Merkel 2 MeM / Lfi =0.312837(G.,/Ga )"""S51295 0.9543 Me / L =1 026925 GO 04034 GO.366635 M Ii? w a 3 - 0.792561 G~I84318G~316749 0.9915 I Me / L =0270391 G-o?441671GQ.628913 0.9788P fi w a Poppe 2 Mep /Lfi =0.338802(Gw /GJ-o.528249 0.9521 Me / L =1 051589 GO.088689 GO.3560SO P .fi' w a 3 _ O.797118 G~19360SO G~304773 0.9904 0 Gw = 6.76 kglm2s 0 Gw = 4.73 kg/m's f). Gw = 2.68 kg/m2s ~ __Equation type 1 -.'r...... -LJ_______ Equationtype2 _ :/'- ~ - pO" ??? _Equationt~~,~ '" . _.-~ ..... ~_. :..-?.. -' :T ~,- - ~"--~ -- -..A--;"" ..... ::;..- ...:;.....-~ <P' ~'-'-. ,.. _.~~ <>./ 0.45 0.40 0.35 ~ 'E 0.30 j -Gl 0.25:;: 0.20 0.15

0.10 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m's Figure S.6: Comparison ofexperimental data and empirical equations (<1fi = 0.3 m ). S.lO Table 8.12: Empirical relations for the loss coefficient according to the variou s methods (~= 0.3 m). Approach Eq. Empirical relation Correlationtype coefficient 1 K =4 163824 GO.8972l8 G-o?931478 0.9518fdmlM' w a Merkel and 2 K =3 982891(G /G )0.918727,fdmlM', wa 0.9516 e-NTU K =4 276605 GO 971935 G-1.699623fdmlM' w a 3 +0.319932 G~?642241G~?066190 0.9947 1 K =4166137 GO.895630G-o?931793 0.9517filmlP' w a 2 K =3 975401(G /G YOI8329Poppe fdmlP' w a 0.9515 K =4 283539 GO.969848G-1.699905fdmlP' w a 3 +0.318727 G~?641672G~?067842 0.9947 20 <> Gw = 6.76 kg/m"s 16 0 Gw = 4.73 kg/m"s 6- Gw = 2.68 kg/m"s --Equation type 1 12 ....... Equation type 2 "'E ___Equation type 3 ~ ~ 8 4.54.03.53.02.52.01.5 0+----+---+----+---+----+---+----1 1.0 Ga. kg/m"s Figure 8.7: Comparison of experimental data and empirical equations (~= 0.3 m). S.1l S.S FILL SPACING: 0.4 m Table S.13: Experimental measurements (P, = 100950 Pal. T" Twb TWi Two m, m. dpft Too ?C ?C ?C ?C kg/s kg/s Pa ?C 1 18.799 18.523 49.280 37.710 2.740 6.722 23.267 28.300 2 18.634 18.428 49.755 35.738 3.972 6.745 29.186 30.171 3 18.451 17.944 49.883 33.870 5.311 6.743 34.041 32.958 4 18.718 17.939 49.883 32.427 6.622 6.666 45.814 34.502 5 19.105 17.982 49.818 30.827 8.014 6.695 75.265 34.240 6 19.651 18.214 49.866 29.536 9.379 6.700 99.317 33.388 7 19.873 19.873 49.933 41.803 2.660 10.760 32.021 44.269 8 19.590 19.590 49.956 39.025 4.118 10.783 40.890 42.657 9 18.910 18.687 49.930 36.926 5.380 10.718 50.302 41.343 10 18.799 18.799 49.803 35.090 6.651 10.652 70.643 39.187 11 19.173 18.016 49.552 33.381 8.044 10.687 106.767 37.418 12 19.682 18.184 48.472 31.454 9.411 10.653 135.553 36.772 13 23.036 23.036 43.582 39.557 2.663 15.122 45.347 40.836 14 20.700 20.700 42.113 36.975 3.881 15.128 53.115 38.998 15 19.412 19.412 40.768 34.376 5.410 15.129 70.267 37.126 16 18.418 18.418 39.895 32.619 6.668 15.051 95.215 35.883 17 18.536 18.080 39.364 31.133 8.060 15.129 129.654 35.m 18 18.996 18.144 38.956 29.957 9.364 15.107 167.915 34.302 Table 8.14: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different methods (i? = 0.1 m). G. G, Me/Lft Me,jLft Mep/Lj KjdmlM KjdmlP raop T.oM kg/m's kg/m's m?1 m'l m'l m'l m'l 'C ?C

1 2.987 1.218 0.165 0.168 0.187 11.170 11.149 41.241 40.481 2 2.998 1.765 0.194 0.198 0.217 6.588 6.574 38.769 38.090 3 2.997 2.360 0.223 0.228 0.248 4.276 4.267 36.497 35.885 4 2.963 2.943 0.250 0.256 0.277 3.678 3.670 34.694 34.143 5 2.976 3.562 0.290 0.297 0.321 4.109 4.100 33.516 33.013 6 2.978 4.168 0.332 0.338 0.365 3.954 3.946 32.651 32.190 7 4.782 1.182 0.108 0.109 0.123 15.810 15.760 43.945 43.088 8 4.793 1.830 0.147 0.149 0.166 8.353 8.326 41.835 41.064 9 4.763 2.391 0.177 0.180 0.199 6.012 5.993 40.027 39.307 10 4.734 2.956 0.205 0.210 0.230 5.515 5.498 38.451 37.778 11 4.750 3.575 0.240 0.246 0.269 5.691 5.674 37.222 36.601 12 4.735 4.183 0.285 0.293 0.319 5.287 5.273 35.897 35.338 13 6.721 1.183 0.092 0.093 0.107 21.955 21.902 40.646 39.996 14 6.724 1.725 0.122 0.122 0.138 12.234 12.207 38.210 37.624 15 6.724 2.405 0.160 0.160 0.178 8.379 8.361 35.977 35.440 16 6.689 2.963 0.182 0.183 0.202 7.521 7.505 34.148 33.638 17 6.724 3.582 0.219 0.221 0.243 7.010 6.996 33.258 32.779 18 6.714 4.162 0.256 0.258 0.283 6.725 6.712 32.601 32.152 8.12 Table S.15: Empirical relations for the transfer characteristic according to the various methods (Aft = 0.4 m). Approach Eq. Empirical relation Correlation tYDe coefficient 1 Me, / Lft =0.189421 G~0406131G~?702488 0.9753 ") e-NTU 2 Me, / Lft = 0.274704(G w /GJ-<ll63612 0.9207 Me / L =0292061 G-<l.07l223 G?.460930 e ji' w a 3 - 0.099761 G~.301183G~?244'8l 0.9783 I MeM / Lft = 0.195062 G:.41821lG~?707961 0.9759 Merkel 2 MeM / Lft = 0.280500(Gw /GJ-<l.17l781 0.9251 MeM / Lft = 0.305797 G:? 050807 G~.411411 3 - 0.11 0328 G~?319306G~.2l2678 0.9770 I Mep / Lft = 0.213383 G:.396999G~?680197 0.9742 Poppe 2 Mep / Lft =0.304395(Gw /Ga )-<ll479l3 0.9212 Mep / Lft =0.327906 G:?04lmG~.43'712 3 _ 0.114295 G~?333173G~?238620 0.9774 0.35 -r---....,.---,.---....,.---.,.---.....,.---.,.-----, 0.30 t----T----i---t----j----=-~~f'ZS"::::....--__r_n-__I 0.25 +-------I----j----7"'?b~"---__'_'If_-__=~FF_--,-;c.....--'+_--'-'--_1 ? Gw =2.98 kglm"s o Gw = 4.76 kglm"s /). Gw = 6.71 kglm"s --Equation type 10.10 +---=x:'S.~+-----t-------If_---t--I ....... Equation type 2 __Equation type 3 "E i 0.20 t----+~=J.R-'5:..r--=d'!;;:.o"'-'--=?:~-"'--__:t=-=C-f---__j Gi

:;: 4.54.03.53.02.52.01.5 0.05 -1----I---~----l---_l_-.J;;;;=F"==~==-_l 1.0 Ga, kg/m"s Figure S.8: Comparison of experimental data and empirical equations (Aft = 0.4 m ). S.l3 Table SJ6: Empirical relations for the loss coefficient according to the various methods (Aft = 0.4 m). Approach Eq. Empirical relation CorrelationtVlle coefficient 1 K =3478872 GO.939791G--<l?968447 0.9416fdmlM' w a Merkel and 2 K =3 349466(G /G )0.959075fdmlM" w a 0.9415 e-NTU K =6 010822 GO.800143 G-1.886076fdmlM' w a 3 +0.353521 G~695382G~942744 0.9955 I K =3 472315 G 0939584 G --<l.968594 0.9416fdmlP" w a 2 K =3 341588(G /G y959107Poppe fdmlp? 'It' a 0.9414 K =6 003348 GO.799475G-1.885029fdmlP? w a 3 +0.350371 G~?696669G~945598 0.9955 0 Gw = 6. 71 kg/m's \ 0 Gw = 4.76 kg/m's ~ A Gw = 2.98 kglm's I-__Equation type 1.... - - - Equation type 2 0~ I-_Equation type 3...... ~~ ......... ~U ~ A 25 20 5 o 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m's Figure S.9: Comparison ofexperimental data and empirical equations (Aft = 0.4 m) . 8.14 S.6 SUMMARY OF SPLASH PACK RESULTS Table S.17 presents a summary of the three types of empirical equations of the t ransfer characteristic according to the Merkel approach, for each of the four fill heights tested. Equa tion type 3 gives the best correlation for all the fills tested followed by equation type 1 and equation ty pe 2. Although equation type 3 gives the best accuracy, ?as can be seen from the values of the correlation co efficients in table S.17, equation type 1 gives correlations of the same order of accuracy. This can be se en from figures S.2, SA, S.6 and S.8 for fill spacings of0.1, 0.2, 0.3 and 004 m respectively. Table S.17: Summary ofthe transfer coefficients according to the Merkel approach . A.. m Equation type 1 r' 0.1 MeM / Lfi =0.374457 G-{)?435520 GO.577461 0.9900w a 0.2 MeM / Lfi =0.298810 G-0.509221 GO.S90240 0.9872w a 0.3 MeM / Lfi =0.249013 G-{)?464089 GO.6S3S7ll 0.9794w a 004 MeM / Lfi =0.195062 G-{)?41821S GO.707961 0.9759 " a A.. m Equation type 2 ,.. 0.1 MeM / Lfi =0.444827(G,,!Ga t.50S736 0.9735 0.2 MeM / Lfi =0.330365(Gw!GJ-O.SSll92 0.9653 0.3 MeM / Lfi =0.312837(Gw!GJ-{).5S129S 0.9543 004 MeM / Lfi =0.280500(Gw/Ga t 571781 0.9251 A.. m Equation type 3 r'

0.1 Me M / Lfi =1.134627G:?086398G:?2l42l0 -O.747394G~Ol4082G:?I39992 0.9979 0.2 MeM / Lfi =1.077103G;;"?078787 G:?246703 - O.779099G~.oJ9S66G:?ll708l 0.9920 0.3 Me M / Lfi =1.026925G~?040340 G:,36663l - O.792561G~?184318G:,316749 0.9915 004 Me M / Lfi =0.305797G:Ol0807 G:.4ll4ll - 0.110328G:319306G:?2l2678 0.9770 The data of the coefficients in tables S.2, S.6, S.JO and S.14 are evaluated tog ether to obtain the following general empirical relations, applicable to all the fills tested. Neglecting the effect of the fill height, the experimental data for the transfer coefficient for all the different fill spacin gs can be presented by Me / L =0 164488 G-O.4l9312G0.614346 tl.-O.347798 (S.I) M fi' w a fi with a correlation coefficient? = 0.9748 and where Aft is the fill spacing in me tres. If the effect of the fill height is included in the correlation then find Me M / Lft =1.021213 G:.41l10G:?6IlI14 tl.1..103487 rJ,?369334 with a correlation coefficient? = 0.9845. If the effect of the water inlet temperature, expressed in ac, is included in th e correlation then find M. / L =2011515 G-{).493143GO.604637 tl.-O.ll4237 L-{)?913309T~0.311016 eM fi' w a fi Ii WI with a correlation coefficient? = 0.9858. (S.2) (S.3) S.l5 0.6 r---.,...---,....--,....----,-----,--..,.----;--..., -Empirical 0.5 1~r-__j---I---I--I---I--i,~c----=E~xpe~n~?m~e~nta~1 0.4 .,. E j 0.3 Ii ::I! 0.2 0.1 +----+---+---+---=-+---+---"'---+-----''---t-.;--I 726354453627189 0+---4----1----1-----1--__10_-.......__.....;..__-1 o Test "c. Figure S.IO: Comparison between measured data and equation (S.2). Figure S.l 0 shows the comparison between the measured data and equation (S.2) f or the Merkel approach. Test numbers I to 18, 19 to 36, 37 to 54 and 55 to 72 in figure S.10 r espectively refer 10 the 0.1,0.2,0.3 and 0.4 m spacing fill tests. It is important to note that the equations given above are only accurate for the specific fills tested. It is not intended to extrapolate the equations for other fills heights and fill spacings. For example, if f!I.t; is chosen as 0.4 m then Lft in the equations must be 3.2 m as can be seen in figure S.l(d) .

Table S.l8 presents a summary of the three types of empirical equations of the l oss coefficients according to the Merkel approach, for each of the four fill spacings tested. Equation type I gives a very accurate correlation of the experimental data as the correlation coefficients in table S. 18 is very close to 1. Equation type I and equation type 2 are less accurate and give curve fits of app roximately the same accuracy. This is because the modulus of the exponents of Go and G. are approxim ately equal for equation type 1. If the effect of the fill height is neglected then a correlation through the mea surements of all four of the tested fills is given by K =(1 655585GI.8320"G-1.919182 + 0 805749G".s96214 G,,?J81360 \,--0.310972 (S.4) fdmlM \. w a . w a P fi with a correlation coefficient,:' = 0.9212. Ifthe fill height is included in the correlation then find (8.5) 8.16 K =(10 05368G1.297029G-I.992125 + 7 76125Go.595059G?.251349 \.-0.042869 r-1.5429 47ftJmlM ? }II a ? w a P fi ft with a correlation coefficient'- = 0.9419. Figure 8.11 shows the comparison betw een the measured data and equation (8.5) for the Merkel approach. Table 8.18: 8ummary ofthe transfer coefficients according to the Merkel approach . 11?? m Equation type 1 -,0.1 K filmlM =4.292617 GO.997973 G-O.593524 0.9446w 0 0.2 K =430145 GO.921207 G-o?87201 0.9543fdmlM' w' a 0.3 K fdm1M =4.163824 G~?897218 G;0.931478 0.9518 .'. 0.4 K =3 478872 GO.93979IG-o?968447 0.9416fdmlM' w a 11th m Equation type 2 ,. 0.1 K =7 439312(G /G )0.710454 0.8975[dmIM' w a 0.2 K fdmlM =4.23192(Gw /G.)0900392 0.9518 0.3 K = 3 982891(G /G )0918727 0.9516fdmlM' }II a 0.4 ( / )0959075 0.9415KfdmlM =3.349466 Gw Go 11th m Equation type 3 ,. 0.1 K =0 257516G2.388300G-2.303946 +3 729301Go.646977G-o?041l77 0.9947fdmlM' w a . VI a 0.2 K =3 179688G1.083916G-1.965418 + 0 639088Go.684936GO.642767 0.9932fdmlM' }II a . VI a 0.3 K =4 276605Go.971935G-I.699623 + 0 319932Go.642241GI.066190 0.9947fdmlM' w a ? VI a 0.4 K =6 010822Go.800143G-I.886076 +0 353521Go.695382GO.942744 0.9955[dmIM' w a ? w a I -Empirical c Experimental 0 0 Po ~ ~ \ "on ~ 0\ ~ '- \ c~ \ \0 :~~ \, ~ bo,., "o'"!ljj ri;;tm ~ 30

25 20 "'e E 15 ~ 10 5 o o 9 18 27 36 Test no. 45 54 63 72 Figure S.lI: Comparison between measured data and equation (8.5). T.I APPENDIXT FILM FILL PERFORMANCE TEST RESULTS\.; T.l INTRODUCTION The performance characteristics of cross-corrugated film fills of three differen t heights are determined experimentally. The results are critically evaluated and presented by empirical equations. The film fills are stacked in layers consisting of horizontally stacked rectangular parallelepi peds. A schematic diagram of the film fills is shown in figure T.\. Each layer is stacked 90? relative to the layer below. The parallelepipeds are 0.3 m high, 0.3 m wide and \.2 m long. The height of the spr ay zones above the fill for all the tests is 15 em. 2 layers Lfi=O.6m (a) 3 layers Lfi=O.9m (b) 4 layers Lfi =1.2 m (c) Figure T.I: Three heights ofcross-corrugated film fills tested. Each fill in figure T.I is tested at different air and water mass flow rates The results of the tests for the fills shown in figure T.I(a), T.I(b) and T.I(c) are shown in sections T.2, T.3 a nd T.4 respectively. Di'" " T.2 T.2 FILL IIEIGHT: 0.6 m Table T.l: Experimental measurements (Po = 100620 Pa). Tai Tw? TW1 Twa mo m. dpfI Tao DC DC ?C DC kg/s kg/s Pa DC 1 23.941 19.912 47.597 37.007 2.685 6.297 11.138 40.376 2 23.145 19.006 47.428 34.222 3.989 6.390 22.665 38.435 3 22.347 17.941 47.253 31.801 5.291 6.375 36.567 36.651 4 22.290 17.475 47.301 29.932 6.628 6.300 56.427 35.329 5 22.642 17.491 47.124 28.374 7.960 6.358 83.519 33.802 6 23.312 17.782 46.959 27.036 9.299 6.358 113.950 32.579 7 23.201 19.562 46.662 38.826 2.700 10.449 12.659 42.646 8 22.793 19.122 46.346 36.137 4.028 10.413 25.276 40.424 9 22.007 17.935 46.199 33.905 5.286 10.462 42.241 38.906 10 21.629 17.294 46.045 31.998 6.621 10.384 65.216 37.675 11 22.082 17.316 45.908 30.437 7.990 10.438 96.722 36.153 12 22.726 17.579 45.522 28.966 9.349 10.453 132.885 35.260

13 23.508 20.565 45.175 40.185 2.658 15.343 19.844 42.287 14 22.963 19.943 44.988 38.013 3.953 15.358 32.182 40.841 15 21.995 18.594 44.761 35.808 5.326 15.370 54.915 39.659 16 21.374 17.639 44.200 33.907 6.624 15.354 82.540 38.226 17 21.728 17.520 44:044 32.470 7.980 15.295 122.097 37.193 18 22.252 17.872 43.680 31.110 9.298 15.337 160.147 36.519 Table T.2: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (Lfl = 0.6 m). Gw Go Me/Lfi Me~Lj, Mep/Lf KfdmlM KfdmlP TaoP TaoM kglm's kglm's m?1 m? l m? l m? l m? l DC DC 1 2.799 1.193 0.896 0.908 1.011 27.908 27.809 39.922 39.206 2 2.840 1.773 1.148 1.170 1.286 25.805 25.719 37.439 36.805 3 2.833 2.351 1.397 1.430 1.559 23.864 23.792 35.345 34.771 4 2.800 2.946 1.636 1.676 1.817 23.629 23.566 33.669 33.149 5 2.826 3.538 1.905 1.948 2.104 24.357 24.299 32.555 32.079 6 2.826 4.133 2.217 2.255 2.425 24.411 24.359 31.677 31.243 7 4.644 1.200 0.759 0.765 0.879 30.982 30.844 42.529 41.725 8 4.628 1.790 0.999 1.013 1.137 27.795 27.677 40.397 39.675 9 4.650 2.349 1.217 1.241 1.377 27.169 27.062 38.796 38.121 10 4.615 2.943 1.419 1.453 1.598 26.939 26.844 37.180 36.556 11 4.639 3.551 1.643 1.687 1.847 27.567 27.479 36.053 35.475 12 4.646 4.155 1.907 1.963 2.139 27.762 27.683 35.058 34.526 13 6.819 1.181 0.523 0.525 0.614 50.026 49.808 42.204 41.407 14 6.826 1.757 0.747 0.752 0.862 36.629 36.469 41.102 40.354 15 6.831 2.367 0.974 0.984 1.112 34.625 34.481 39.897 39.184 16 6.824 2.944 1.149 1.166 1.303 33.924 33.795 38.410 37.742 17 6.798 3.547 1.325 1.349 1.497 34.736 34.615 37.372 36.743 18 6.817 4.132 1.549 1.583 1.750 33.628 33.521 36.631 36.045 T.3 Table T.3: Empirical relations fur the transfer characteristic according to the various methods (Lfl = 0.6 m). Approach Eq. Empirical relation Correlationtype coefficient 1 Me / L =1 115805 G -0400945 GO.773647 0.9878e fi' w a e-NTU 2 Me, ILfi =1.757441(Gw /GJ-0582890 0.9047 Me, / Lfi =1.453563 G~?043374 G~655501 3 _ 0.500908. G~?422441 G~.552463 0.9940 I MeM / Lfi = 1.131871 G;;?403632 G~782625 0.9869 Merkel 2 MeM / Lfi =1.796447(Gw /GJ-o?588290 0.9029 MeM / Lfi =1.638988 G~?282648G~?682887 3 _ 0.802755 G~560711G~?644229 0.9970 I Me / L =1 232608 G-o?37092SGO.748191 0.9859P fi' w a Poppe 2 Mep / Lfi =1.952821(Gw/GJ-o?ss6132 0.8931 Me / L =1 497125 GO.276216G0.66573S P fi' w a 3 _ 0.589942 G~?6J4757 G~?622408 0.9962 4.54.0 .' ? .-3.53.02.52.01.5 0.5 -l-..l.:l.--l----I----I------4----I-----I-----l 1.0

2.5 <) GN =6.82 kglm"s 0 GN =4.64 kglm"s /),. GN =2.82 kglm"s 2.0 --Equation type 1 .... '" Equation type 2 ':e -Equation type 3 .'. .j 1.5 .' " ."::;: . . ..' " 1.0 Ga. kg/m"s Figure T.2: Comparison ofexperimental data and empirical equations (Lfl = 0.6 m) . rI, TA Table T.4: Empirical relations for the loss coefficient according to the various methods (4 = 0.6 m). Approach Eq. Empirical relation Correlationtype coefficient I K =17 162976 GO.48S379 G...()?206927 0.8152fdmlM' w a Merkel and 2 K =24 528308 (G IG )0.313113fdmlM' w a 0.6858 e-NTU K =0 00819 GS.46SS33G-3.66631SfdmlM' w a 3 +17.545503 G~?34S860G:?036969 0.9613 I K =17118512 GO.484076G...()?20S697 0.8136fdmlP' W Q 2 K =24 461621(G IG r 1l897Poppe [dmlP' w a 0.6836 K =0 003132 GUSS218 G-3.631669 fdmlP' W Q 3 +17.238242G~.349702G:?030826 0.9591 \ 0 GIl =6.82 kg/m"s0 GIl =4.64 kg/m"s . \ lJ. GIl =2.82 kg/m"s~ \. --Equation type 1 .~ ....... Equation type 2.. ~ __Equation type 3.. ' .... I--.. ------~ '" IV 0... ". - -- .. ..... ,cr-- .. '.~ ' . .. .. '"'' . . '"'' . '. . '. . ... '" ."tJ . .... " .~ .-" ... 0 . '" W.. ' .... 0 ..... ---. .'"":..f----.A. ..... ' .... '. . ... .'. -.... . -. L .. ... - . ' .... .- ....... . ... . ... . -. 50 45 40 ""E E 35 ~ 30

25 20 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kglm"s Figure T.3: Comparison ofexperimental data and empirical equations (Lft = 0.6 m) . T.5 T.3 FILL HEIGHT: 0.9 m Table T.5: Experimental measurements (Po = 100950 Pal. TaJ T.b Twi Two rna m. dpjI Tao ?e ?e ?e ?e kg/s kg/s Pa ?e 1 21.490 17.574 46.252 34.169 2.745 6.201 16.052 40.742 2 21.060 17.028 46.057 31.212 3.985 6.180 31.042 38.582 3 20.546 16.110 45.922 28.443 5.359 6.144 51.440 36.455 4 20.504 15.632 45.674 26.476 6.634 6.123 76.526 34.904 5 21.078 15.810 45.308 24.831 8.021 6.098 114.345 33.192 6 21.719 16.124 45.070 23.618 9.303 6.047 154.239 32.032 7 21.609 18.573 43.850 36.198 2.716 10.043 16.995 40.954 8 21.003 17.749 43.444j, ; 33.401 3.948 9.962 32.715 39.213 9 20.247 16.421 42.934 ; 30.606 5.402 9.869 58.232 37.360 10 20.171 15.825 42.483 28.639 6.669 9.877 86.515 36.065 11 20.566 15.709 42.161 26.965 8.013 9.795 125.962 34.741 12 .21.264 15.932 41.707 25.398 9.379 9.851 175.598 33.409 13 21.724 18.619 41.287 36.431 2.754 14.695 22.724 39.454 14 21.212 17.940 41.165 34.443 3.996 14.595 39.174 38.594 15 20.282 16.572 40.882 32.243 5.368 14.470 67.710 37.363 16 19.965 15.884 40.728 30.610 6.627 14.422 102.540 36.334 17 20.179 15.755 40.734 29.198 8.025 14.333 153.822 35.656 18 20.954 15.891 40.784 27.762 9.338 14.343 204.969 35.070 Table T.6: Transfer coefficients, loss coefficients and outlet temperatures acco rding to the different methods (LjI = 0.9 m). G. Go Me/LjI MeulLjI Mep/Lf KfdmlM KfdmlP Tao? TaoM kg/m's kg/m's m?1 m,l m'l m-I m-I ?C ?C 1 2.756 1.220 0.849 0.866 0.972 26.062 25.977 40.284 39.572 2 2.747 l.771 1.066 1.096 1.207 23.931 23.858 37.679 37.056 3 2.731 2.382 1.317 1.360 . 1.482 22.078 22.018 35.415 34.864 4 2.721 2,949 1.533 1.580 1.711 21.574 21.523 33.606 33.109 5 2.710 3.565 1.802 1.843 1.984 22.165 22.119 32.108 31.664 6 2.687 4.135 2.087 2.104 2.254 22.294 22.253 31.005 30.602 7 4.463 1.207 0.675 0.681 0.797 28.004 27.904 4 l.l86 40.445 8 4.427 1.755 0.887 0.900 1.021 25.474 25.385 39.299 38.628 9 4.386 2.401 1.098 1.122 1.248 24.406 24.330 37.122 36.515 10 4.390 2.964 1.283 1.318 1.452 23.932 23.864 35.589 35.031 11 4.353 3.561 1.482 1.528 1.671 24.260 24.199 34.240 33.728 12 4.378 4.169 1.763 1.822 1.984 24.751 24.696 33.314 32.844 13 6.531 1.224 0.534 0.535 0.660 36.491 36.367 39.992 39.273 14 6.487 1.776 0.714 0.719 0.844 29.805 29.702 38.961 38.284 15 6.431 2.386 0.898 0.908 1.038 28.684 28.590 37.660 37.020 16 6.410 2.945 1.045 1.061 1.195 28.630 28.542 36.597 35.991 17 6.370 3.567 1.194 1.218 1.357 29.394 29.311 35.633 35.064 18 6.375 4.150 1.441 1.481 1.642 28.929 28.851 35.282 34.742 T.6 Table T.7: Empirical relations for the transfer characteristic according to the different methods (Lfl ~ 0.9 m). Approach Eq. Empirical relation Correlationtype coefficient 1 Me / L =1 073109 G--o?444337 GO.770970 e fi' 'W a 0.9921 e-NTU 2

Me. / Lji =1.588654 (GwlG.)--o606334 0.9340 Me / L =1 382687 GO.OOI9S3GO.73922; 3 e fi' 'If' a _ 0.492783 G~344143G:?7149SS 0.9944 'J' 1 Me / L =1 092357 G--o?443876 GO.774531 0.9910Mji', w a Merkel 2 MeM / Lfl =1.625041(Gw /GJ--o?6078os 0.9321 Me / L =1 625618 GO.091940GO.702913 M fl? WI a 3 _ 0.735958 G~?376496G:?66;399 0.9951 1 Mep ILji =1.193519 G;:'.397740G:?724438 0.9900 Poppe 2 Mep / Lji =1.767607(Gw/GJ--oS61449 0.9225 Me / L =1 526182 GO.078237G".69;680 3 P 'fi' 'If' a -;- 0.556982 G~419;84 G:?67S1S1 0.9933 4.54.03.53.02.52.01.5 0.5 -I-....loIl:=:....-I-----1----1---4----t----I-----l 1.0 2.5 <> GN = 6.43 kg/m"s 0 GN = 4.40 kglm"s [). GN = 2.73 kglm"s 2.0 --Equation type 1 ....... Equation type 2 "'e __Equation type 3 :f 1.5 .' .. ::;: -.' , , . .... 1.0 ," .. , Ga. kg/m"s Figure T.4: Comparison ofexperimental data and empirical equations (Lfl = 0.9 m) . T.7 Table T.8: Empirical relations for the loss coefficient according to the various methods (4; = 0.9 m). Approach Eq. Empirical relation Correlationtype coefficient I K =18 216653 GO.33O%8G-{)?140105 0.8506jdmlM - w a Merkel and 2 K =23.089652(G /G )0.215184fdmlM -"',: w a 0.7223 e-NTU K .=1 633204 GL2502E8 G-3.873083fdm1M' w a 3 +16.170094 GO.288861GO.012429 0.9304 w ? I K =18 166799 GO.330291 G-O139161 0.8493fdmlP' w a 2 K =23 033881(G /G )0214362Poppe fdmlP' w a 0.7200 K =1 561219 GL276792G-3.931459fdmlP' w a 3 +16 173258Go.287875 Go.om99 0.9300

. w ? <> 0 GN =6.43 kglm2s \. 0 GN & 4.40 kglm2s~\ 6- GN =2.73 kglm'sEquation type 1 ...,.,..~ ....... Equation type 2. .. ' .... ---Equation type 3 ~ . " ....... ~ <>...,.. <>; ::.:, . ' .. -.... . ~~ . ' .. ../'; . '. .~ .... . - . ' .. .-.. . '" ' .. " . ..... -._-~. ' .. .... - .. '-~ L D...... _" " '. ....... - .. - .. ' .. '. l5. . --' .. . '" .. 37.5 35.0 32.5 ~ 30.0 'E E :i 27.5 25.0 22.5 20.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kg/m"s Figure T.5: Comparison ofexperimental data and empirical equations (Lfi = 0.9 m) . T.8 T.4 FILL HEIGHT: 1.2 m Table T.9: Experimental measurements (P. = 101020 Pal. Tai T.b Twi Two rn. m. dpft Tao 'C ?C ?C ?C kg/s kg/s Pa ?C 1 22.787 17.282 44.500 32.906 2.715 6.372 21.090 40.502 2 22.468 16.739 44.154 29.703 3.988 6.354 40.996 38.385 3 22.374 16.079 43.971 27.221 5.285 6.331 66.197 36.410 4 22.736 15.866 43.568 25.164 6.684 6.275 102.465 34.632 5 23.405 16.100 43.159 23.607 8.006 6.236 149.358 33.225 6 24.076 16.435 42.750 22.319 9.420 6.191 208.263 31.938 7 23.591 18.493 41.830 34.844 2.661 9.997 22.022 39.505 8 23.077 17.833 41.347 31.968 3.935 9.928 42.702 38.096 9 22.553 16.686 40.875 29.321 5.329 9.804 73.625 36.294 10 22.636 16.286 40.504 27.399 "6.615 9.700 110.577 35.052 11 23.465 16.504 40.095 25.746 7.959 9.569 161.016 33.828 12 24.156 16.689 39.754 24.257 9.342 9.528 224.387 32.390 13 24.155 19.208 38.740 34.553 2.673 13.895 27.143 37.470 14 23.690 18.461 38.459 32.348 4.031 13.643 50.272 36.594 15 22.961 17,.227 38.260 30.377 5.298 13.512 81.935 35.818 16 22.914 16.543 38,004 28.700 6.597 13.276 124.537 34.805 17 23.117 16.445 37.632 27.082 ' 7.992 13.100 183.626 33.579 18 23.952 16.772 37.295 25.726 9.341 12.972 241.428 32.655 Table T.IO: Transfer coefficients, loss coefficients and outlet temperatures acc ording to the different

methods (Lfi ~ 1.2 m). G. G. Me/Lfi MeJiLfi Mep/Lj KfdmlM KfdmlP TaoP T.oM kg/m's kg/m's m?1 m'! m,l m,l m'! ?C ?C 1 2.832 1.207 0.753 0.769 0.873 26.321 26.241 40.118 39.420 2 2.824 1.773 0.960 0.990 1.098 23.719 23.652 37.543 36.937 3 2.814 2.349 1.151 1.192 1.304 21.941 21.886 35.407 34.866 4 2.789 2.971 1.358 1.403 1.521 21.369 21.323 33.395 32.916 5 2.771 3.558 1.607 1.647 1.774 21.801 21.760 31.991 31.561 6 2.752 4.186 1.914 1.925 2.058 22.038 22.002 30.757 30.370 7 4.443 1.183 0.584 0,588 0.710 28.509 28.422 40.104 39.395 8 4.412 1.749 0.790 0.801 0.924 25.221 25.144 38,391 37.751 9 4.357 2.368 0.968 0.990 1.112 23.872 23.805 36.454 35.873 10 4.311 2.940 1.131 1.163 1.289 23.386 23.327 34.946 34.415 11 4.253 3.537 1.338 1.382 1.517 23.607 23.554 33.690 33.207 12 4.235 4.152 1.594 1.651 1.802 23.938 23.890 32.725 32.282 13 6.176 1.188 0.454 0.453 0.581 34.961 34.868 37.928 37.282 14 6.063 1.792 0.652 0.655 0.785 28.414 28.335 36.921 36.316 15 6.005 2.355 0.824 0.833 0.967 26.905 26.832 35.986 35.407 16 5.900 2.932 0.946 0.960 1.088 26.502 26.435 34.757 34.212 17 5.822 3.552 1.124 1.146 1.283 26.720 26.658 33.709 33.206 18 5.766 4.151 1.345 1.380 1.534 25.750 25.696 32.980 32.513 I' T.9 Table T.II: Empirical relations for the transfer characteristic according to the different methods (Lft ~ 1.2 m). Approach Eq. Empirical relation Correlationtvoe coefficient 1 Me / L =0 974989 G...?468433GO.788223 0.9909eli? w a e-NTU 2 Me, / Lft =0.1438010{G,,/GJ...?644151 0.9454 Me / L =1 370484 G...?249137 GO.748290 e 'Ii. w a 3 - 0.441297 G~.028141G~?703293 0.9915 I MeM / Lft =0.996604 G:? 469512 G~?790386 0.9905 Merkel 2 MeM / Lft =1.471826{G,,/GJ...?645730 0.9451 Me / L =1 357391 GO. 110577 GO.712196M fl? W Q 3 _ 0.567207 G~443165 G~?669846 0.9942 I Me / L =1 090362 G...?408136GO.725775 0.9891P Ii' w a Poppe 2 Mep / Lft =1.604656{G,,/GJ""85132 0.9349 Me / L =1 380517 GO 112753 GO 698206 P fi? w a 3 -0.517075 G~.461071G~681271 0.9917 2.0 -r----,....-----,.---......----r----r-----,.---..., 1.5 ':E j 1.0 .. ., .. -' . :E 0.5 ~ ~ ..

... o Gw =5.96 kg/m's o Gw =4.34 kg/m's /),. Gw =2.80 kg/m's --Equation type 1 ....... Equation type 2 --Equation type 3 4.54.03.53.02.52.01.5 0.0 -1----+----1-----1----+----+----+----4 1.0 Ga, kg/m"s Figure T.6: Comparison ofexperimental data and empirical equations . T.10 Table T.12: Empirical relations for the loss coefficient according s methods (LJI = 1.2 m). Approach Eq. Empirical relation Correlationtype coefficient 1 K = 19658921 GO.281255G-o?I75117 0.8561fdmlM' w a Merkel and 2 K = 22 448599(G /G )0211132filmlM' Ii' a 0.8154 e-NTU K = 3 897830 GO.7m71G-2.114727fdmlM' w a 3 +15.327472 G~?215975G:?079696 0.9562 1 K = 19 601207 GO.281090G-o.I74400 0.8546fdmlP' w a 2 K = 22 399835(G /G )0210566Poppe fdmlP' w a 0.8133 K = 3 859490 GO.782298G-2.119420fdmlP? Ii' a 3 +15.295976 G~?21531IG~?080546 0.9559 \ I I <> Gm =5.96 kg/rn's 0 Gw =4.34 kg/rn's -~ f). Gm = 2.80 kg/m"s~ " __Equation type 1 ~-"~ __ ..... Equation type 2 -~ __Equation type 3....;. .. ~.__ k?.ii?? ~--.~I-- 0' . ............... . '. ' ... - -. ~ "'" ~. -.~ -.. . .. '.--'" 0~ ' .... ~- ..:.> 0 ' ..... ... '- ' ... --. -. -. -' . .. . .... /\-.. C: ' .. -.. ' .. ' .... ' .. . 35.0 32.5 30.0 "'e ~ 27.5 ~ 25.0 22.5 20.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Ga. kglm's Figure T.?: Comparison ofexperimental data and empirical equations . T.ll T.5 SUMMARY AND COMBINED RESULTS A summary of the equations for the transfer coefficient, according

(Lft ~ 1.2 m) to the variou

(LJI ~ 1.2 m)

to the Merkel

approach, is shown in table T.I3. It can be seen from figures T.2, T.4 and T.6 that equation types I a nd 3 give accurate representations of the measured data. This is also shown in table T.13 where all the correlation coefficients, r, for these equation types are close to unity. Table T.13: Summary ofthe transfer coefficients according to the Merkel approach . L.,m Eouation IYne I ,.. 0.6 MeM / Lfi =d.131871 G-{),403632 GO,782625 0.9869w a 0.9 MeM / Lfi = 1.092357 G~0.443876G:?774531 0.9910 1.2 Me / L = 0 996604 G-{),469512 G 0.790386 0.9905M fi? w a L.. m Eouation tvue 2 ,.. 0.6 MeM / Lfi = 1.796447(Gw /Ga t 588290 0.9029 0.9 MeM / Lfi = 1.625041(Gw/Gat?607805 0.9321 1.2 MeM / Lfi =1.471826(Gw /G.)-{)?64573o 0.9451 Lft> m Equation type 3 ,; 0.6 MeM / Lfi = 1.638988 G~?282648G~,682887 - 0.802755 G~,5607IlG~?644229 0.9970 0.9 Me M / Lfi = 1.625618 G~,091940G:,702913 -0.735958 G~.376496G:?665399 0.9951 1.2 MeM/Lfi =1.357391 G~,110577G:,712196-0.567207 G~,443165G:?669846 0.9942 The experimental data for the transfer coefficient for all the different fill he ights can be represented by Me M / Lft = 1.019766 G~A32896G:,782744L~?292870 with a correlation coefficient r= 0.9880. (T.1) Figure T.8 shows the results from equation (T.!) compared to the transfer charac teristics obtained from experimental measurements. Tests 1 to 18 in figure T.8 represent the tests for t he 0.6 m thick fill. Tests 19 to 36 represent the fill test results ofthe 0.6 m thick fill and tests 37 to 54 represent the measurements of the 1.2 m thick fill. Due to the limitations of the fill test facility, it is impossible to conduct th e fill tests at a constant water temperature. If the effect of the changing water temperature is included in the correlation, the Merkel number can then be presented by M. / L = 1 722176 G-o?448804GO.778434L-o,315917T-o?132799 eM fi' " a fi WI with a correlation coefficient r~ 0.9876. TW1 is expressed in ?e. (T.2) A summary of the equations for the loss coefficient according to the Merkel appr oach is shown in table T.14. It can be seen from the correlation coefficients, r, in table T.14 that eq uation type 3 gives the most T.12 accurate representation of the measured data. Equation type 1 and equation type 2 do not correlate the measured data well. This is seen in figures T.3, T.5 and T.6. 2.5 r--T--i--i---r-;:::::c====:::;-, ? Experimental -Empirical 2 -I-----Jc--+-"...---+--'t-+---+--=+======-J ':" 1.5 E

j ., :::: 1 0.5 -1-----+--~-----jf----_+_-.s:>---l----=-_I_~l_-__I 54453627189 O-l-----+-----!-----+----II----4-------I o Test no. Figure T.8: Transfer characteristic given by equation (T.1) compared to experiem tal results. Table T.14: Summary of the loss coefficients. L fl, m EQuation type 1 i' 0.6 K fdmlM =17.162976 GO.485379 G-O.206927 0.8152w a 0.9 KjiimlM =18.216653 GO.330968 G-O.I40105 0.8506w a 1.2 K =19658921 GO.281255G-0175177 0.8561fdmlM' w a Lift m EQuati()p type 2 i' 0.6 K =24 528308 (G /G )0313113 0.6858fdmlM' w a 0.9 K =23 089652(G /G )0215184 0.7223fdmlM' w a 1.2 K =22 448599(G /G )0211132 0.8154fdmlM' W Q Lift m EQuatiop type 3 i' 0.6 K =0 00819 G5.465533 G-3.666315 +17 545503 G0345860 G-O.036969 0.9613fdmlM' w a . w a 0.9 K =1 633204 G1250268G-3.873083 +16 170094 GO.288861GO.012429 0.9304fdmlM' w a ? w a 1.2 K =3 897830 GO.777271G-2.1I4727 +15 327472 GO.215975GO.079696 0.9562fdmlM' w a . w a The experimental data for the loss coefficient for all the different fill height s can be represented by K -16753566 G'?401127G-o?I704S4r-0.2JS]70ftlml -. w a fi (T.3) with a correlation coefficient? = 0.8276. Equation (T.3) and the test data ofthe three different fill heights are shown in figure T.9. The correlation coefficient suggests that equation (T.3 ) does not correlate the data well and this is evident from figure T.9. T.13 The data for the tests of all three fill heights can also be correlated by -0.236292 K =(5 154914 GO.877646G-1.462034 +10806728 GO 226'78GO.293222 IT (T.4)fdml' w a ? w a jLfl with a correlation coefficient? ~ 0.904013. It can be seen from figure T.IO that equation (T.4) correlates the data relatively accurately. 0 Experimental Empirical \. 0 ~ ~I> \ \ 1\ <>.~ 0\ ~<><>V; \ o~0 6 i'.. 0 ~ ~ ~ ~oq;> ~ ....... 50 45 40 ?s 35 E ~ 30 25 20 15 o 9 18 27

Test no. 38 45 54 Fi gure T.9: Loss coefficients given by equation (T.3) compared to experiemtal resu lts. o Experimental _Empirical I I ~ ~ < ~ \ It <>.~ <\ v~ \ ~0 0 ~. ~ lij:>v 0 ~~ . 50 45 40 .,. 35S ?~ 30 25 20 15 o 9 18 27 Test no. 36 45 54 Figure T.IO: Loss coefficients given by equation (T.4) compared to experiemtal r esults. U.l APPENDIXU COOLING SYSTEM OPTIMIZATION U.I INTRODUCTION The dimensions of a counterflow natural draft wet-cooling tower are optimized to obtain the minimum combined operational and capital cost compounded over a specified economic life of the cooling tower. The performance characteristics of a typical turbo-generator system are shown in figure V.I. It can be seen that the system performs optimally at a certain recooled water temperature, T~. The cooling tower therefore needs to supply the condenser water at a certain temperature at the ba se load to obtain maximum power output. The cooling tower outlet temperature is therefore fixed in a cooli ng tower optimization analysis. 630 620 610 ~ 600 a: ~ 590::J & ::J 5800 .. III ~ 570 a. - 560! 550

540 530 1 I / / ........... 1 '" / " / 1 X 1 . / '\. t\ ./ , ~ '\ " 060 960 5 10 15 20 25 30 35 40 45 50 55 Recooled water ternperature, Two, ?C Figure V.l: Performance characteristics of turbo-generator-condenser system. The analysis that follows is very basic but it can be readily expanded to includ e more detailed cost approximations. However, the present analysis shows the powerful capabilities of the WCTPE computer program, described in appendix P, when it is employed in conjunction with the LF OPC [82SNI, 83SNI, 85SNl] and DYNAMIC-Q [94SNI, OOSN2] mathematical optimization algorithms. More d etailed optimization analyses applied to cooling systems are according to Li and Priddy [85LII], Conradie [95COl], Kintner-Meyer et al. [95KII], Conradie et at. [98COlj, KrOger [98KRI] a nd Castro et at. [OOCAI]. V.2 U.2 THE OPTIMIZATION PROBLEM (D.I)(equality constraints) (inequality constraints)gix),;; OJ = 1,2.....rand The optimization problem can generally be given in the following form: Minimizej{x). x = (XI,x'?... ,x.)T such that h,(x) ~ O. i ~ 1.2.....m ,;; n The solution, which mayor may not be unique, is given by f is the objective function and hi and gJ are equality and inequality constraint functions respectively that defme a feasible region. U.3 OPTIMIZATION VARIABLES There are only three primary geometrical solution variables in the optimization analysis presented here, i.e .? H" H, and d,. Refer to figure 1.1 for a description of the variables. Thr ee more geometrical variables are functions of these primary geometrical variables. i.e .? Afr, LjI and d, whe re d, is a function of d,. Kroger [98KRI] states that to prevent cold inflow at the top of natural draft co oling towers the ratio of the tower outlet to tower base diameters must be approximately 0.6. The area at the tower lip is given by A = 7T: d' , 4 3 (D.2) Afr is less than A3 to make provision for the fill support structure and water d

istribution system. It is assumed in this analysis that the ratio ofApiA, is constant. Afr is therefore a function ofd3? LjI is a function ofAfr. As the diameter, d,. changes during the optimization pr ocess. and Afr is changing. LjI will change to obtain the desired water outlet temperature, for optimum turb o-generator performance as discussed above. for the given fill frontal area, Afr. The transfer effects o f the rain and spray zones are included in the analysis when the height of the fill, LjI, is determined. U.4 THE OBJECTIVE FUNCTION The objective function consists of the sum of the operational costs and capital costs compounded over a specified economic life of the project. The simplified operational and capital c ost components employed in the optimization analysis are presented below. VA.! TOWER OPERATIONAL COST The pumping power can be expressed by Ppump = mwg(H3 + Lft + Lsp ) (U3) The operating cost of the pump for one year is given by ,rl Cpump ;;;; PpumpCe/tfC1: U.3 (U.4) where r is the total hours per year that the pump is working and C"" is the cost of electricity. The effect of the motor-pump efficiency can be included in C,I" , for example, multiply C,I " by 1.1 if the motor? pump efficiency is 90 %. The operating cost of the pump compounded over the selected period, expressed in years, is given by Y'~' Cp"mp = LPpumpC'I" r(l+ f 1l00y-' ~I where/is the inflation rate to account for the increase in the cost ofelectricit y due to inflation. U.4.2 TOWER CAPITAL COST The volume ofthe concrete in the tower shell can be approximated by v, = !!...(d, + d"',H. 2 The capital cost of the tower shell is given by where Ceone is the cost ofconcrete per unit volume. The volume of the fill is given by Vfi = AfrLfi The cost ofthe fill is given by Cfi = VfiCfi where Cfi is the cost of the fill material per unit volume. The total capital cost over the selected period is given by Ceap = (C, +CjlXI + illOOra" where i is the inflation rate. The objective function for optimization is thus given by Minimize (C"", + Cp'mp) (U.5) (U.6) (U.7) (U.8) (U.9) (U.IO) (U.ll) U.S CONSTRAINTS

It is possible to include equality and inequality constraints into the optimizat ion analysis. Only two inequality constraints are included in this analysis, i.e., (U.12) VA H 6 ,;;c2 where Cl and C2 are constants. These inequality constraints are included into th e optimization analysis to ensure that physically realistic results are obtained. U.6 SAMPLE OPTIMIZATION The dimensions of the cooling tower specified in appendix I are optimized in thi s analysis to obtain the minimum combined operating and capital cost over the economic life of the coolin g tower. However, the Merkel method is employed in the analysis instead of the Poppe method. Kroger [9 8KRI] presents the performance analysis of the tower presented in appendix I employing the Merkel m ethod. The cost and optimization variables employed in this example optimization analysis are only f or illustrative purposes and do not necessarily resemble realistic and practical values. However, the val ues are chosen to be as realistic as possible. The water outlet temperature determined by Kroger [98KRI] , Two = 294.526 K (21.376 0C), is fixed in the analysis. It is assumed that this is the water outl et temperature for a given base load (Q = 972.3714 MW) for the given water inlet temperature. The cost and optimization variables: Cost ofelectricity, C"" Operating hours per year, r Inflation rate Economic life ofcooling tower Cost ofconcrete Cost offill Diameter ratio to prevent cold inflow, dJd, Area ratio to make provision for fill supports etc., ApJA, Prime interest rate = 0.03 $/kWh = 8760 hlyear =3% ~ 35 years = 200 $/m' = 25 $/m' = 0.58 (same as tower in appendix I) = 0.9677 (same as tower in appendix I) =7% Table U.I shows the dimensions of the cooling tower specified both in Kroger [98 KRI] and appendix I and the cooling tower dimensions obtained by the cost optimization. The LFOPC [8 2SNI, 83SNI, 85SNI] optimization algorithm is employed in conjunction with the WCTPE program, described in appendix P, to obtain the tower dimensions for the minimum combined capital and operational cost. It can be seen from table 0.1 that the total cost is reduced by 18.7 % for this particu lar optimization analysis. Table U.I: Cooling tower dimensions with corresponding total cost H6,ID d3, m AjhID" H3,m Lji' m d6, m Total cost, m$ Kroger [98KRI] 147 104.5 8300 10 2.504 60.85 330.459 Optimized 139.218 99.897 7584.64 4.925 4.614 57.94 268.715

Figure U.2 shows the magnitude of the normalized objective function as convergen ce commences for the optimization analysis. 10.98 c 0 0.96 '8 c 02 0.94~ '8 .. 0.92:S' 0 "0 .. 0.9 .!!;; E 0.880 z 0.86 1\ If u -U.5 0.84 1 10 20 30 40 50 11lIrations 60 70 80 90 100 Figure U.2: Normalized objective function obtaioed by the LFOPC optimization alg orithm. Figure U.3 shows the solution domaio for the sample optimization problem where T "", ~ 294.526 and where H3 is held constant at 5 m. Figure U.3 is generated manually by employiog the WCTPE program without an optimization algorithm. Figure U.4 shows the contour plot of the solu tion domaio shown io figure U.3. The total cost io figures U.3 and U.4 is shown as a function of d3 a nd H. where H3 is held constant at 5 m. Only d3, H. and H3 are presented io figures U.3 and UA as these are the primary solution variables of the optimization process; as described io section U.3. It can be seen from figure 0.4 that a mioimum value of the total cost does exist for the sample optimization problem defined above. The values for H. and d3 given io table U.I, with H3 '" 5 m, are shown io figure U.4. It can therefore be concluded that the LFOPC optimization a lgorithm accurately determioes the tower dioJensioos to obtaio the mioioJum cost over the economic l ife of the cooliog tower. The absolute values for the minimum cost io table U.I and figure U.4 differ due to the assumption that H, = 5 io figure U.4 and H3 ~ 4.925 io table U.1. As can be seen io figure U.3, the total cost is relatively high when the diamete r of the tower, ~, and the height of the tower, H., decreases at relatively small values of d3 and H6 . The iocrease io cost is due to the fact that the height of the fill iocreases to achieve the same cooliog load to c ompensate for the reduced draft (H6 decreases) and reduced fill frontal area (d3 decreases). Figure U.5 shows the solution domaio where H3 is 5, 10 and 15 m. Figure U.6 show s the contour plots of

the total cost where H3 is equal to 10 and 15 m. U.6 350 340 330 320 Cost, m$ 310 300 290 280 270 200 150 125 100 90 120 Figure V.3: Solution domain where H, =5 m. 305 285 280 Optimum: table UJ---------. Cost, m$ 140 160 170 180 190 200 ,--~----r----q,--------",----"'-------"-'-----' S 150 ~ Figure V.4: Contour plot oftolal cost where H, = 5 m. D.7 500 450 400 S +of '" 3500U 300 250 200 180 160 140 120 100 90 'H.,m 110 100 d3, m 120 Figure U.5: Comparison between solution domains for H3 equal to 5, 10 and 15 m. H,=lOm 170 H,= 15 m '00 ". '90 ... 450-~ .., '80 .,,-----.,43. 170 ---._~ '20115100 1O!i 110 S '80 ~ '50

,<0 '50 120 110 42' 1~ ,. '20'15110,. 150 S 150 ~ ,<0 '50 120 110 '~ 100 105 d),m Figure U.6: Contour plot oftotal cost where H, is equal to 10 and 15 m. U.7 CONCLUSION The WCTPE computer program, discussed in appendix P, is successfully employed, i n conjunction with mathematical optimization algorithms, to obtain the dimensions of a natural draf t cooling wet-tower for the minimum operating and capital costs, compounded over the economic life of th e cooling tower. The total operating and capital cost in this particular optimization analysis is red uced by 18.7 %. The results of the mathematical optimization algorithm are verified by a manual investigation o f the solution domain. V.I APPENDIX V SAMPLE CALCULATION OF THE INFLUENCE OF TEMPERATURE AND HUMIDITY PROFILES ON COOLING TOWER PERFORMANCE V.I INTRODUCTION The reduction in the thermal performance of a natural draft cooling tower is inv estigated during the occurrence of nocturnal temperature and humidity inversions. The natural draft c ooling tower, specified in appendix I, is the reference tower employed in this investigation. The heat r ejection rate, Q, and cooling range, Twi - Two, from appenstil' I are 1003.4775 MW and 18.5928 K respe ctively. It is assumed in appendix I that the vertical temperature distribution follows the dry adiabatic lapse rate (DALR) of -0.00975 Kim and that the vertical humidity ratio profile is constant. The analysis that follows is a comparative investigation where the temperature, specified at ground level during temperature inversions, is not necessarily a practical value. The same te mperature is specified at ground level during inversions as was the case where a DALR was assumed (see app endix I). The temperature at ground level durinll, an inversion period is generally relatively low. The performance, where temperature and humidity inversions are employed in the analysis,is compar ed to the case where a DALR and a constant humidity ratio are employed. The possible error in tower per formance evaluation is determined when no provision is made for the temperature and humidity inversion profiles to determine the effective inlet conditions and the reduction in tower draft. It is assumed in the following sample calculation that the exponent, b, in equat ion (L.5) is 0.01. The maximum value of b in figure L.4 is 0.01 during the winter. The magnitude of the

temperature inversion, as defined in figure L.I, is the strongest when b is at its maximum value. It is assumed in this investigation that T" and Twb are measured I m above ground level. This is gener ally referred to as the ground level measurements. To, in appendix I is 288.6 K. T, in equation (L.5) is therefore (T,t - 273.15) = (288.6 - 273.15) = 15.45 ?C. and z, in equation (L.5) is I m. The effect of b on cooling tower performance is investigated in section V.6. For this cooling tower with d, = 60.85 m, the height from which ambient air is d rawn into the cooling tower, H" is approximately 135 m from figure N.3. It is assumed that no wind is present in the analysis in appendix N. Wilber et al. [85WII] found from experience that H, is between 50 an d 100 m for practical natural draft cooling towers while Lauraine et al. [88LAI] estimate H, to be bet ween 50 and 150 m. It is recommended that the value ofH, be taken as approximately half the height of the cooling tower shell. It . is therefore assumed, in the sample calculation that follows, that H, is 73.5 m. The effect of H, on the results is investigated in section V.5. V.2 V.2 MEAN INLET AIR TEMPERATURE: SAMPLE CALCULATION The height ofthe temperature inversion is given by equation (L.26), 1 1 Z _[ -DALR ]'_1 _[ 0.00975 ]0.01-1=313.51 m ,,- b(r, + 273.15) - 0.01(15A5 + 273.15) Z;, is greater than the height from which air is drawn into the cooling tower, H ,. Equation (E.27) can therefore be employed to calculate the mean temperature at the inlet of the cool ing tower, T,,'m = (1; +273.15{H,)'(_l_) + 0.0097l H'X1- H,) '\.z, b+1 1. 2 H, = (15A5+273.1s173.5)O.Ol(_1_)+0.00975(73.5X1_~) = 298.60 K '\. 1 0.01+1 2 73.5 This is 10K higher than the measured air drybulb temperature at ground level. Th e presence of a temperature inversion can therefore have a pronounced effect on the effective in let temperature compared to the measured value. When b, as shown in, figure LA is at its minimum value of 0.003 during the summer, then z" = 90 m from equation (L.26) and r'im = 291.8018 K from equation (E.27). This is only 3.2 K higher than the measured air drybulb temperature at ground level. V.3 MEAN INLET HUMIDITY RATIO: SAMPLE CALCULATION Figures M.2 and M.3 show examples of inversion vapor pressure profiles that occu r during the night. The humidity ratio profiles follow approximately the same trends as the vapor pressu re profiles. The humidity at 04hOO in figure M.2 and at 02hOO-06hOO in figure M.3 falls by approximately 1 2 % during the first 40 m. The height of the humidity inversion in figure M.2 is approximately 40 m. It is assumed in this " analysis that the height of the humidity inversion profile is 40 m. The humidity is assumed to be approximately constantabove 40 m. The humidity ratio at ground level is found to be 0.008127 kglkg dry

air in appendix I for the specified ambient conditions. The following fourth ord er polynomial gives the humidity ratio to be 0.008127 kglkg I m above ground level and approximately 12 % less at a height of 40 m after which it remains approximately constant, (V.l) where CI to c, are constants where CI = 8.0812xI0-', c, = 4.6139x10-', c, = -7.0 639xI0-', C4 = 4.6179xlO-9 and c, = - 1.0709xlO-II ? Equation (V.I) is plotted in figure V.1. T he humidity profile is expressed by a polynomial since the humidity profile can be specified as constan t, linear and non-linear. The average humidity at the inlet ofthe cooling tower, Wim, is given by the equa tion (V.2) Substitute equation (V. 1) into equation (V.2) and find after integration (V.3) V.3 0.0092 ../'I--'"" / i"" / / I / / 0.0090 . .; ~ ., 0.0088} .. 0.0086g e ~ 0.0084 'll 'E :i! 0.0082 0.0080 o 10 20 30 40 Height, Z, m 50 60 70 80 (VA) (V.5) Figure V.I: Inversion hwnidity profile Substitute constants Ct to c, given below equation (V.1) and H, = 75 m into equa tion (V.3) and find W'm = 0.0089 kglkg dry air. The effective inlet hwnidity ratio, during this humidity i nversioo, is approximately 9.6 % higher than when it is asswned that the humidity ratio profile is constant . V.4 EXTERNAL PRESSURE DISTRIBUTION: SAMPLE CALCULATION The pressure distribution external to the cooling tower is calculated in this an alysis as accurately as possible as both the vertical temperature and hwnidity profiles are considered. The calculation is relatively complex, due to the inclusion of the humidity profile in the analysis , and can generally not be employed in a simple cooling tower performance analysis. This is because it requ ires numerical integration to calculate the pressure distribution. It will be shown, however, t

hat the hwnidity inversion has a negligible effect on tower draft. \ As mentiooed in appendix E, the temperature and humidity profiles also influence the draft through the tower as the pressure distribution on the outside of the outside of the tower is a fimction of both the temperature and humidity profiles. The pressure at the top of the tower shell, e X1ernal to the tower, is given by equation (E.6), - ex ( O.622?g f' (w+l) dzJ p, - p, P R (w +0.622) T where w is given by equation (V.I) and Tby equation (L.5). Substitute equations (V.I) and (L.5) into equation (VA). The integral in equation (VA) is then given by f? (w+I) dz = f' (c, +c,z+c,z' +c4z' +C,Z4 +1) dz(w+O.622) T { J' . (c, +c,z+c,z' +C4Z' +C,Z4 +0.622XT, +273.15\:' V.4 The integral in equation (V.5) can only be evaluated by numerical methods like S impson's composite rule [97BUI]. For H6 = 147 m, z, = I m, T, = 15.45 ?C, b = 0.01 and c, to c, given be low equation (V. I), the integral in equation (V.5) is determined by Simpson's composite rule with 20 int ervals to give 0.7823663. Thus, from equation (VA) and numerical integration result ofequation (V.5) find =84100ex ( 0.622.9.8.0.7823663)=82714A7pa p, p 287.08 The pressure differential on the outside of the cooling tower is PI - P, =84100 - 82714.47 = 1385.53 Pa This pressure difference is 1445.73 Pa in appendix 1. This reduced pressure diff erential will cause a reduction in draft through the tower. V.S REDUCTION IN TOWER PERFORMANCE As already mentioned; the heat rejection rate, Q, from appendix I is 100304775 M W. It is assumed in appendix I that the vertical temperature distribution follows the DALR of -0.009 75 KIm, the vertical humidity ratio profile is constant, w; ='0.008127 kglkg dry air and Tal = 288.6 K. If it assumed that the vertical humidity profile is constant and the increase in the temperature, as calculated in section V.2 (Taim = 298.67), is employed in performance calculatio n then Q = 80204774 MW. This is approximately 20 % less than 100304775 MW. The detrimental effect of the temperature inversion on tower draft is also included in this evaluation where the pressure differential on the outside ofthe cooling tower is calculated by equation (B. 18), i.e., PI - P, ~ 1386.70 P a. If it is not assumed that the vertical humidity profile is constant, but given b y the value calculated in section V.3 (Wlm = 0.0089 kgikg dry air) then Q = 787.9063 MW. This is 21.5 % an d 1.8 % less than 100304775 MW and 802.4774 MW respectively. The reduction in the draft due to the reduced pressure differentia~ as calculated in section VA, is included in the performance evaluat ion. The effect of the humidity inversion on tower draft is negligible. Thus, the relatively strong temperature inversion reduces tower performance by a pproximately 20 % in

this particular case. The effect of the humidity ratio inversion on tower perfor mance is less pronounced than the temperature inversion and causes an additional 1.5 % reduction in perfo rmance. A weak temperature inversion during the summer, where b = 0.003, reduces tower performa nce by 7.8 %. Approximately 20 % of the reduction in performance is due to the reduction in dr aft and approximately 80 % is due to the increased effective inlet air temperature and humidity, where the effect of the increased effective inlet humidity is essentially negligible. V.6 INFLUENCE OF H, ON TOWER PERFORMANCE V.5 The effect of the height from which air is drawn into a cooling tower, H" on the performance of the cooling towers is investigated. H, is varied between 50 m and 150 m. Figure V.2 shows the effect of the variation of H, on the effective inlet air drybulb temperature and humidity rati os. There is only a 2 K increase in the effective inlet air drybulb temperature when H, is varied from 5 0 to 100 m. As already mentioned, Wilber et aI. [85W1I) found that the approach height is between 50 an d 100 ill for practical natural draft cooling towers while Lauraine et aI. [88LAI) estimate H, to be bet ween 50 and 150 m. Figure V.2: Effective air inlet temperature and humidity ratio versus H,. ~'"~ '-., """-... ~~ ~, ............. ............ 14.52 14.47 lo: ?J I; 14.45 .:! 14.50 14.42 .. iii 0.00896 ~ "0 I III 0.00889 .... .Ii ~ 0.00910 0.00903 0.00882 14.40 150 0.00975 150 137.5 137.5 125 125

112.5100 100 112.5 H"m 87.5 87.5 75 75 62.5 62.5 -~ .,. ....V V.,...... ........ V .... V /1,.0'",. ~ .r VV,, i/ 788 787 786 297 50 789 298 785 50 299 302 790 300 301 H"m Figure V.3: Heat r~ected and cooling range versus H,. Figure V.3 shows the heat rejected and cooling range as a function of H, during the occurrence of temperature and humidity inversions. Q is reduced by only 0.3 % and the cooling range by 0.06 K when ,ii',: " V.6 H, is increased from 50 to 100 m. The performance of natural draft cooling tower s is, therefore, relatively insensitive to H~ V.7 INFLUENCE OF b ON TOWER PERFORMANCE Figure VA shows the effect that the exponent, b, in equation (1.5) has on the ef fective inlet air drybulb temperature and the temperature inversion height from equation (1.26) when H, = 73.5 m. The minimum and maximwn values ofb'm figure LA are 0.003 and 0.01 respectively. Zit from equ ation (1.26), with T, = 15045 DC, varies between 90 m and 313.51 m for the minimum and maximum values of b respectively. The corresponding'reduction in the heat rejection rate and cooling range, betwee n the minimum and maximum values ofb, are shown in figure V.5 and are 15% and 2.6 K respectively. Z70 180 330 300 240

E 210 = N 160 120 90 0.010.0090,0080.0070,0080,0050.004 /h -:", ~~/ /. '",'"", V ",4 ./.... ,,-L,.'" ./ ,/.... ", ",'",.. 291 0.003 299 298 297 296 l' J 29S 294 293 292 b Figure VA: Effective air inlet temperature and temperature inversion height vers us b. 18.0 17,0 17.5 14.5 15.0 16.5 II: ii 16.0 ~ .. 15.5 f14.0 0.010.0090.0060.0070.0060.0050.004 """~ -... ,-='..... .... .~ .....~ ..... ~~ ............ """"'" '"~~ '" 840 780 0.003 900 920 820 940 800

880 1880 d b Figure V.5: Heat rejected and cooling range versus b. V.7 V.9 CONCLUSION The presence of a temperature inversion can have a pronoWlced effect on the effe ctive mean inlet temperature compared to the measured temperature value at or near groWld level. It is fOWld in this investigati~? that a relatively strong temp...ature inversion reduces a particul ar tower performance by approximately 20 %. A relatively weak inversion reduces a particular tower perfo rmance by approximately 8 %. The effect of the humidity ratio inversion on the same tower performance is less pronounced than the temperature inversion and causes an additional 1.5 % reducti on in performance. Approximately 80 % of the reduction in p...formance is due to the increased mean inlet temperature and 20 % is due. to the reduction in draft. The performance of natural draft cooling towers is relatively insensitive to H, where H, is practically half the height of the tower shell. Th e predicted heat rejection rates differ by only 0.3 % when H, is increased from 50 m to 100 m. It is recomm ended that H, is taken as approximately half the height of the tower shell.