1) TAB stands for Testing, Adjusting, and Balancing, which describes the sequence of work performed. The TAB firm works with other construction teams to deliver a project that meets design intent.
2) Finding a suitable duct traverse location with fully developed airflow is often a challenge for TAB work. Standard guidelines recommend locations downstream of fans or bends based on duct diameter.
3) When direct airflow measurements are not possible, TAB technicians can use static pressure or temperature measurements to calculate outdoor air quantities. However, these indirect methods introduce additional errors.
1) TAB stands for Testing, Adjusting, and Balancing, which describes the sequence of work performed. The TAB firm works with other construction teams to deliver a project that meets design intent.
2) Finding a suitable duct traverse location with fully developed airflow is often a challenge for TAB work. Standard guidelines recommend locations downstream of fans or bends based on duct diameter.
3) When direct airflow measurements are not possible, TAB technicians can use static pressure or temperature measurements to calculate outdoor air quantities. However, these indirect methods introduce additional errors.
1) TAB stands for Testing, Adjusting, and Balancing, which describes the sequence of work performed. The TAB firm works with other construction teams to deliver a project that meets design intent.
2) Finding a suitable duct traverse location with fully developed airflow is often a challenge for TAB work. Standard guidelines recommend locations downstream of fans or bends based on duct diameter.
3) When direct airflow measurements are not possible, TAB technicians can use static pressure or temperature measurements to calculate outdoor air quantities. However, these indirect methods introduce additional errors.
1) TAB stands for Testing, Adjusting, and Balancing, which describes the sequence of work performed. The TAB firm works with other construction teams to deliver a project that meets design intent.
2) Finding a suitable duct traverse location with fully developed airflow is often a challenge for TAB work. Standard guidelines recommend locations downstream of fans or bends based on duct diameter.
3) When direct airflow measurements are not possible, TAB technicians can use static pressure or temperature measurements to calculate outdoor air quantities. However, these indirect methods introduce additional errors.
The article discusses common issues faced by TAB firms such as suitable traverse locations and design/installation errors. It also discusses how performance curves can be used to analyze field measurements and help solve issues.
Issues with suitable traverse locations, design/installation errors, and lack of straight ductwork upstream of measurement points.
Performance curves from manufacturer testing can be used to analyze field measurements and help predict how changes will impact system performance.
May 2001 ASHRAE Jour nal 19
ASHRAE Journal TAB
W A Primer on Testing, Adjusting and Balancing About the Author By Andrew P. Nolfo, P.E. Member ASHRAE hat is TAB? Most of our industry thinks it stands for Test and Balance. The correct wording is Testing, Adjusting, and Bal- ancing. The sequence of the words also describes the sequence of the associated work. A system or individual piece of equipment must first be tested to determine its operating state, then adjusted, and finally balanced to produce the desired results in accordance with the design docu- ments. The TAB firm is part of the construction delivery team along with the design engi- neer, mechanical contractor, and controls contractor. They all have the same goal: deliver a project that satisfies the design in- tent. If an adversarial relationship develops among team members, the TAB firm often is perceived as the watchdog. Sometimes this watchdog role identifies design and/or installation errors. However, this watchdog role is ancillary to the TAB firms main func- tion: helping the system to work properly by balancing the fluid flows to their correct proportion. This article discusses some common problems that can complicate the TAB firms work. The article addresses air and hydronic systems. It also addresses how these mistakes can be avoided at the de- sign stage or fixed in the field. Addition- ally, the article discusses the application of fan and pump curves to TAB work. Fi- nally, the article discusses how TAB firms use other diagnostic tools and data to ob- tain unknown data. Traverse Location Measuring airflow in ducts is important. Duct traverses are used for everything from measuring total airflow to determining cor- rections factors for direct reading hoods. The lack of a suitable traverse location is prob- ably the single greatest issue in TAB work. A suitable location is one where there is fully developed airflow, i.e., one where the ve- locity profile is reasonably uniform across the plane of the traverse location. While the industry debates the accuracy of various duct traverse protocols, the larger issue is determining and using a suitable traverse plane to obtain accurate, repeatable read- ings. Many of todays TAB specifications re- fer to ANSI/ASHRAE Standard 111-1988, Practices for Measurement, Testing, Adjust- ing, and Balancing of Building Heating, Ventilation, Air-Conditioning, and Refrig- eration Systems. Standard 111 is being re- viewed with a new edition expected in 2001. The current version lists several character- istics of an ideal traverse plane: a. A uniform velocity distribution means that 80%90% of the velocity pres- sure measurements are greater than 10% of the maximum velocity pressure. b. Airflow should be at right angles to the traverse plane. c. The cross section of the traverse plane should not be an irregular shape, and the shape area should be uniform in the vicin- ity of the traverse plane. d. The traverse plane should be located to minimize the effects of leaks. Appendix D of the standard offers addi- tional guidance in locating a suitable traverse plane. It suggests an effective length of 2 duct diameters downstream of a centrifugal or axial fan outlet. This length is based on a velocity of 2,500 fpm (12.5 m/s) or less. For velocities greater than 2500 fpm (12.5 m/s), add one more duct diameter for each 1000 fpm (5 m/s) in excess of 2500 fpm (12.5 m/s). For rect- angular duct, Equation 1 provides an equivalent diameter: (1) where, a and b are the rectangular duct dimensions. As an example, a 10,000 cfm (5000 L/s) system with a 30 in. 20 in. (76 cm 51 cm) discharge duct (2,400 fpm [12 m/s]) would require about 5.8 ft (1.7 m) of unob- structed, straight duct upstream of an ideal traverse location. Another 1 diameter, or 2.7 ft (0.8 m) of downstream duct should also exist before a fitting, takeoff or other ob- struction is encountered. This would require that the mechanical equipment room be de- signed to accommodate an air-handling unit with a straight, horizontal discharge duct 8.5 ft (2.6 m) long. As an example, Figure 1 shows this air- handling unit (10,000 cfm, 30 in. 20 in. [4719 L/s, 76 cm 51cm] duct) with a traverse plane 3 ft (0.9 m) from the dis- charge elbow. This location is 1.1 diam- eters from the disturbance. A traverse taken there would not be accurate. The elbow would also create a substantial pressure drop due to turbulence associated with sys- tem effect. This discharge is in a broken- back condition. The NEBB Procedural Standard for the Testing, Adjusting, and Balancing of Envi- ronmental Systems says the accuracy of a Andrew P. Nolfo, P.E., is the techni- cal director for the National Environ- mental Balancing Bureau (NEBB). He is a corresponding member of TC 9.7, TAB, and a member of TC 9.9, Build- ing Systems Commissioning. ( ) 2 1 4 ab E L = 20 ASHRAE Jour nal www. ashr aej our nal . or g May 2001 ASHRAE Journal pitot tube traverse is determined by the availability of a location to perform the traverse. The standard con- tinues to identify a location that has six to 10 diam- eters of straight duct upstream of the test location. The standard also has a statement regarding the practicality of finding such an ideal location: this condition will not be found very often in the field, therefore, use the best location available. The procedural standard also discusses ways to correlate total airflow when a less than ideal location exists. If the traverse location cannot avoid elbows, off- sets, transition, branch take-offs or other items that would cause turbulence, one solution is to obtain more readings closer together for a better average reading. If a total supply cannot be directly read, it might be possible to measure several main branches and add the resulting airflows together. When all else fails, the TAB technician must use the best available location and correlate the airflow reading to other gathered data, such as brake horsepower calculated from fan pressure drop and electrical performance. Determining Outside Air Quantity If supply and return airflow are measured accurately, outside airflow could easily be determined from Equation 2: outside return supply Q Q Q + = (2) When accurate direct measurements are not available, another method to determine the outdoor air uses static pressure. If accu- rate supply airflow can be determined, the outdoor air can be de- termined by measuring static pressure in the return duct. The out- door air must be completely sealed off for this test. As an example, consider a 10,000 cfm (5000 L/s) supply fan with a minimum out- door air requirement of 10%. With the fan delivering the required 10,000 cfm (5000 L/s) and the outside air intake sealed, measure the static pressure in the return duct just before the mixing box. Since the outdoor air is completely sealed, the return air must equal the supply air. Suppose the measured return static pressure is 1.0 in. w.c. (250 Pa). Since flow varies with the square root of the pressure, the TAB technician can then open the outdoor air damper until the static pressure in return duct equals 0.81 in. w.c. (202 Pa), as illustrated by Equation 3: ( ) 0.5 2 1 2 1 P P Q Q = (3) Where, Q 1 = Return airflow with minimum outdoor air (9,000 cfm [4500 L/s]) Q 2 = Return airflow with zero outdoor air (10,000 cfm [5000 L/s]) P 1 = Return air static pressure at minimum outdoor air P 2 = Return air static pressure at zero outdoor air (1.0 in. w.c. [250 Pa]) Substituting into Equation 3 would give: ( ) ( ) 5 . 0 1 5 . 0 1 250 000 , 5 500 , 4 0 . 1 000 , 10 000 , 9 P P = = Solving for P 1 :
= 0.81 in. w.c. (202 Pa) Many TAB technicians and engineers are familiar with using temperatures to estimate outside air quantity. This technique uses Equation 4 in conjunction with Equation 2:
outside return supply Q OAT Q RAT Q MAT + = (4) Where: MAT = Mixed air temperature RAT = Return air temperature OAT = Outside air temperature The accuracy of the temperature measurements introduces another level of potential error. Unless the air in the return air duct is thoroughly mixed, the measured temperatures could be off as much as 1F to 4F (0.5C to 2C). The outdoor air mea- surement will probably experience an error. The mixed air tem- perature is the most suspect and could be off as much as 5F to 20F (2C to 10C) depending on the temperature. Many air-handling unit mixing chambers do not really mix the airstreams well. Some use opposed blade dampers that provide excellent throttling capabilities but poor mixing. Parallel blade dampers provide better mixing but poor control. As a result, per- fect mixing is rarely achieved. The errors are magnified when the supply, return, and outdoor air are in the range of 45F to 70F (7C to 21C) because the differences are small. In colder climates, errors in this method of reading the outdoor air quantity diminishes when the temperature is in the range of 0F to 35F ( 18C to 2C). As an example, an air-handling unit is to provide 10,000 cfm (5000 L/s), with a minimum of 1500 cfm (750 L/s) of outside air. The actual field conditions have made it impossible to measure the supply and outdoor airflows accurately. The return air, out- door air, and mixed air temperatures are respectively measured as: 74F, 45F, and 70F (23.3C, 7.2C, and 21.1C). The return airflow is accurately measured at 8,900 cfm (4450 L/s). Substi- tuting these values into Equations 2 and 4 and solving would yield: Q supply = 10,324 cfm (5154 L/s) Q outside = 1,424 cfm (704 L/s) The supply and outdoor airflows appear to be within 5% of the design requirements. Suppose measurement errors and incomplete mixing meant the actual temperatures were: 75F, 44F and 72F (23.9C, 6.7C, and 22.2C). Substituting these values into Equations 2 and 4 and solving would yield: Figure 1: Air-handling unit pitot traverse location (10,000 cfm [4719 L/s] SA; 8,500 cfm [4011 L/s] RA; 1,500 cfm [708 L/s] OA [min.]. May 2001 ASHRAE Jour nal 21 TAB Q supply = 9,853 cfm (4938 L/s) Q outside = 953 cfm (488 L/s) The supply airflow is still within 5% of the design require- ment, but the minimum outdoor air is 36% below the design re- quirement. Although this method of can provide accurate data, the best ap- plication is when it is used by trained technicians to provide an acceptable check to a suspect measurement. Duct Leakage A good duct traverse can also be used to determine duct leak- age. Suppose a VAV system needs to supply 10,000 cfm (5000 L/s). The design diversity is 10%, so the sum of all the VAV terminals is 11,000 cfm (5500 L/s). With terminals equal to 1000 cfm (500 L/s) closed (zero airflow), the TAB technician has tested the remaining terminals. The sum of the outlets is 8,500 cfm (4250 L/s). A supply duct traverse, as shown in Figure 1, measures supply airflow of 9,500 cfm (4750 L/s). The sum of all the air outlets is measured as 8,500 cfm (4250 L/s). Which one, if either, is correct? If the supply air duct traverse was reliable, the technician could be justified in concluding that the problem is duct leakage. Determining Pump Flow Like measuring airflow on a fan, pump total flow measure- ments can sometimes be suspect. Since water is non-compress- ible, the problems are not as severe as for fans. Most TAB techni- cians will determine flow by measuring the differential pressure between the pump discharge and the pump suction. Pump manu- facturers provide pressure taps machined into the body of the pump or on the suction and discharge flanges of larger pumps. By closing the discharge valve, the differential pressure at no- flow conditions can also be measured. The discharge valve is then opened to its original condition. By using the manufacturers pump curve and these two pressure measurements, the techni- cian can estimate total flow. The differential pressure at the no- flow conditions is used to verify which impeller is installed in the pump. The differential pressure at the operating condition will determine the actual flow condition. Using the pump curve to determine pump flow can be inaccu- rate when the pump has a flat curve. As an example, Figure 2 is the manufacturers pump curve for a pump selected to provide 800 gpm (50.5 L/s) at 68 ft (204 kPa) head. This design point is identified as Point 1. It shows that a 9 in. (229 mm) impeller will satisfy these design conditions and that a 20 hp (15 kW) motor should be provided. The technician tested the pump and mea- sured a shut-off differential pressure of 75 ft (225 kPa) and an operating differential pressure of 71 ft (213 kPa). These points are identified as Points 2 and 3 respectively. Based on the technicians findings, it appears that the pump Advertisement for the print edition formerly in this space. 22 ASHRAE Jour nal www. ashr aej our nal . or g May 2001 ASHRAE Journal is flowing the desired 800 gpm (50.5 L/s) but has a 9 8 1 in. (23 cm) impeller. The reality is that the instrumentation accuracy is probably 1% of full scale. If the technician used a 0 to100 psig (0 to 300 kPa) gauge, the readings could be off by 1 psi, or 2.31 ft (7 kPa). Even if the shutoff head was measured at 73 ft (219 kPa) and the operating head was read at 68 ft (204 kPa), both readings could be off by 2.31 ft (7 kPa). The actual operating head could be 70.31 ft (211 kPa), which would equate to a flow of 700 gpm (44.2 L/s). The system could be short of flow by 100 gpm (6.3 L/s), or 12.5%. Figure 3 shows the identical requirements with a different pump with a steeper curve. In this scenario, the 1% tolerance in the pressure measurements becomes only a 25 gpm (1.6 L/s) change in pump flow. However, this selection has lower efficiency than the first pump, 75% instead of 80%. Many considerations go into selecting pumps for each applica- tion, including flat vs. steep curves and pump efficiency. To avoid having the pump selection affect the ability to obtain reliable flow measurements, designers can provide a calibrated flow device at the pump discharge. As for air systems, these flow devices require certain minimum straight pipe upstream and downstream of the device. Flow devices can be installed in the field during testing or if problems are encountered, but it is usually easier and less expen- sive to install them during construction. Another method for determining pump flow would be to use system devices and components as calibrated flow devices, e.g., a chiller bundle, a heat exchanger, or a control valve. Each of these components has an associated C v. The chiller manufacturer will identify the pressure drop across the tube bundle at a certain flow. The differential pressure drop across the entering and leav- ing chilled water piping could be measured and the pump affin- ity laws applied as in Equation 3: Figure 2: Flat pump curve.* * (hp 0.746 = kW); (ft of water 2.99 = kPa); (gpm 0.0631 = L/s) Advertisement for the print edition formerly in this space. May 2001 ASHRAE Jour nal 23 ( ) 5 . 0 2 1 2 1 P P Q Q = (3) Where: Q 1 = Actual Chilled Water Flow (gpm or L/s) Q 2 = Design Chilled Water Flow (gpm or L/s) P 1 = Actual Measured Pressure Differential (ft or kPa) P 2 = Manufacturers Design Pressure differential (ft or kPa). As an example, a chiller has a design flow of 1,000 gpm (63.1 L/s) with a stated differential pressure of 20 ft (60 kPa) head. The TAB technician measured differential pressure across the dedi- cated pump and estimated the flow at 850 gpm (53.6 L/s) from a flat pump curve. The technician wants to check the reading so it measures differential pressure across the chiller at 19.2 ft (57.6 kPa) of head. Substituting into Equation 3: ( ) ( ) 5 . 0 5 . 0 0 . 60 6 . 57 1 . 63 0 . 20 2 . 19 000 , 1 = = X X Solving for X:
= 979 gpm (61.8 L/s) The technician concludes that the flow through the chiller is 979 gpm (61.7 L/s). There is a caveat to this scenario and to all other data associated with equipment. The manufacturers rating is based on laboratory conditions. In this case, long, straight runs of pipe are at the chiller. Additionally, the pressure drop is measured im- mediately at the equipment connection. The measured pressure drop as shown in Figure 4 must be corrected for the pressure drop associated with the additional pipe, valves and fittings at the chiller. One method is to calculate the pressure drop associated with the additional pipe, valves, and fittings. This pressure drop must be subtracted from the measured pressure drop and the corrected pres- sure drop be used to verify chiller flow. Sizing Balancing Valves This final common mistake is easy to correct at the design stage but costly to correct in the field. A balancing device is similar to a control device and should be sized accordingly. Figure 3: Steep pump curve.* Advertisement for the print edition formerly in this space. 24 ASHRAE Jour nal www. ashr aej our nal . or g May 2001 ASHRAE Journal Specifying line size balancing valves can cause a balancing night- mare, especially at terminal units. Balancing devices should be sized based on the design pressure difference at that part of the system. Line size balancing valves are often so oversized that they must be throttled to the point of wire-drawing or exces- sive noise before they can effectively provide the resistance re- quired to balance the flow in the circuit. Suppose a balancing valve has to eat up 10 psi (70 kPa) of extra pressure difference. A smaller valve that can achieve that pressure drop at 50% to 80% open will be easier to adjust than a larger valve that must be squeezed down to 10% or 20% open. Fan and Pump Curves Manufacturers performance curves are graphic representations of measured performance under laboratory conditions. The per- formance curves only represent the equipment being tested, not the system. System curves are graphic representations of the per- formance of a particular installed system. Fan laws and pump laws cannot directly be applied to equipment performance, espe- cially variable flow systems, either air or water. The test rigs for these equipment performance tests bear little, if any, resemblance to real world, installed applications. Also remember that manu- facturers test equipment at normal operating conditions of tem- perature, pressure, density, etc. For example, a fan curve for 75F (24C) air at sea level will not represent performance of a fan operating at 250F (120C) in Denver. Most fan manufacturers have their products performance-tested by an independent testing agency. The products may bear a label as proof of the testing, and the catalogued data may state that the products have been tested by the independent agency. The testing procedures are published along with a statement as to the accuracy of the data and the range of performance that is to be expected within normal manufacturing tolerances. Typically, a centrifugal fan is tested with an open inlet and a long, straight discharge duct. This is not the normal situation in the field. The testing agency has exacting procedures for testing and measurements of the fan. A typical fan family may have one size tested at multiple speeds. The results for airflow, efficiency, horsepower, etc., are then interpolated or extrapolated for the entire range of fan condi- tions. AMCA publication 203-90 discusses field performance testing of fans. It, as well as other AMCA publications, is an excellent reference source. Pumps, on the other hand, are normally tested by the manufac- turer. The manufacturers testing requirements may be as stringent as those of the independent fan-testing agency, but the testing is still accomplished under ideal conditions. Pump suction and dis- charge piping may consist of long straight runs of piping. Again, this is a condition that seldom occurs in the field. As with fan test- ing, a pump may be tested with several impellers and the results are then interpolated for the entire pump range. For these reasons, when technicians mea- sure a piece of equipment, such as a fan, they measure total airflow, differential static pressure, temperature, amperage, voltage, etc. The combination of these field measure- ments and associated calculations could be plotted on the fan curve. If accurate, care- ful, thoughtful measurements can be taken, the measured data usually matches the pub- lished data within the normal tolerances of the performance tests. Performance curves are normally used as a diagnostic tool when trying to solve a field issue. When interpreting actual data, the TAB technician will plot the measured system curve on the performance curve. The system curve is the only true graphic representation of how the equipment is in- teracting as only one component in the system. This information can then be used, with the affinity laws, in attempting to pre- dict how changes to the system will im- pact the performance. Discernable Data With many systems, optimizing design to assist the TAB technician to achieve reliable, repeatable measurements is a function of equipment layout and space allocation. This is especially true for air- handling systems. So what other tools are available to the TAB technician when Advertisement for the print edition formerly in this space. May 2001 ASHRAE Jour nal 25 key data cannot be directly measured? As previously discussed, airflow measurement of mixed air- streams can be determined by the temperature measurement. This can also be applied to coils. If the entering and leaving air tempera- tures can be adequately determined, airflow can be estimated by applying the conservation of energy equations. As an example, a hot water heating coil has the following conditions: Entering air temperature 55F (13C) Leaving air temperature 90F (32C) Hot water flow 50 gpm (3.16 L/s) Entering water temperature 180F (82C) Leaving water temperature 150F (66C) Since the energy from the hot water must be transferred from the water to the airstream, the energy balance can be written as Equation 5: ( ) ( ) ( ) ( ) 13 32 1.232 66 82 4200 3.16 L/s(air) 55 90 1.10 150 180 500 50 cfm = = Figure 4: Chiller pressure differential pressure measurement.* (5) Substituting and re-arranging gives: air water air water 7 1.232 L/s 7 4200 L/s 7 1.10 cfm 7 500 gpm = = cfm = 19,480 cfm (9071 L/s) Conclusion Most TAB firms feel their objective is to orchestrate all pieces of the mechanical system to a workable, operable system. The world is not an exact science. Their ability to perform these ser- vices depends on taking accurate, repeatable measurements. The design professionals and installing contractors can greatly enhance the TAB firms work by understanding the difficulties that most of todays projects present to the TAB work. Bibliography National Environmental Balancing Bureau (NEBB) Procedural Stan- dard for Testing, Adjusting, Balancing of Environmental Systems, 1998/ Sixth Edition. ANSI/ASHRAE Standard 111-1988, Practices for Measurement, Test- ing, Adjusting, and Balancing of Building Heating, Ventilation, Air-Con- ditioning, and Refrigeration Systems. Air Movement & Control Association InternationalAMCA 203-90. Rishel, J.B. 2001. Applying affinity laws for centrifugal pumps. HPAC Engineering, February. Advertisement for the print edition formerly in this space.