University of Cape Town: Power Station Thermal Efficiency Performance Method Evaluation

Power Station Thermal Efficiency
Performance Method Evaluation
n
w
To
e
ap
C
Prepared by:
Heeran Heerlall
of
HRLHEE001
ty
Department of Mechanical Engineering

si
University of Cape Town

r
ve
Supervisor:
Dr. Ryno Laubscher
ni
U
February 2021
Submitted to the Department of Mechanical Engineering at the University of Cape Town in partial
fulfilment of the academic requirements for a Master’s of Science degree in Mechanical
Engineering
Key Words: STEP, thermal performance, coal-fired, power plant, sequential perturbation,
uncertainty propagation.
n
w
To
The copyright of this thesis vests in the author. No
quotation from it or information derived from it is to be
e
published without full acknowledgement of the source.
ap
The thesis is to be used for private study or non-
commercial research purposes only.
C
Published by the University of Cape Town (UCT) in terms
of
of the non-exclusive license granted to UCT by the author.

ty
r si
ve
ni
U
Abstract
Due to global warming, there is an escalated need to move towards cleaner energy solutions.
Almost 85% of South Africa’s electric energy is provided via Eskom’s conventional coal-fired power
plants. Globally, coal-fired power plants have a significant share in the power generation energy
mix and this will be the case over the next 20 years. A study, aligned with the aspiration of
improving the thermal efficiency of the coal-fired power plants, was initiated, with a focus on the
accuracy of energy accounting. The goal is that: if we can accurately quantify efficiency losses, the
effort can be prioritized to resolve the inefficiencies.
Eskom’s thermal accounting tool, the STEP model, was reviewed against relevant industry
standards (BS 2885, BS EN 12952-15, IEC 60953-0/Ed1) to evaluate the model uncertainty for
losses computed via standard correlations. Relatively large deviations were noted for the boiler
radiation, turbine deterioration and make-up water losses. A specific review of OEM (Original
Equipment Manufacturer) heat rate correction curves was carried out for the determination of
turbine plant losses, where these curves were suspected to have high uncertainty, especially when
extrapolated to points of significant deviation from design values. For an evaluated case study, the
final feed water correction curves were adjusted based on an analysis done with the use of power
plant thermodynamic modelling tools namely: EtaPro Virtual Plant® and Steam Pro®.
A Python® based computer model was developed to separately propagate systematic (instrument)
and combined uncertainties (including temporal) through the STEP model using a numerical
technique called sequential perturbation. The study revealed that the uncertainties associated
with thermal efficiency, heat rate and individual thermal losses are very specific to the state of
operations, as demonstrated by individual unit performance and the power plant’s specific design
baseline performance curves. Whilst the uncertainties cannot be generalized, a methodology has
been developed to evaluate any case. A 3600 MWe wet-cooled power plant (6 x 600 MWe units)
situated in Mpumalanga was selected to study the impact of uncertainties on the STEP model
outputs.
The results from the case study yielded that the thermal efficiency computed by the “direct
method”, had an instrument uncertainty of 0.756% absolute (abs) versus the indirect method of
0.201% abs when computed at the station level for a 95% confidence interval. For an individual
unit, the indirect efficiency uncertainty was as high as 0.581% abs.
A study was conducted to find an optimal resolution (segment size) for the thermal performance
metrics to be computed, by discretizing the monthly data into smaller segment sizes and studying
the movement of the mean STEP model outputs and the temporal uncertainty. It was found that
i
the 3-hour segment size is optimal as it gives the maximum movement of the mean of
performance metrics without resulting in large temporal uncertainties. When considering the
combined uncertainty (temporal and instrument uncertainty) at a data resolution of 1 minute and
segment size of 3 hours, the “direct method”, had a combined thermal efficiency uncertainty of
0.768% abs versus the indirect method of 0.218% abs when computed at the station level for a
95% confidence interval. This would mean that the temporal uncertainty contribution to the
combined uncertainty is 2.915% for the “direct method” and 14.919% for the “indirect method” of
the above-stated uncertainties.
The term “STEP Factor” can be used synonymously with effectiveness (percentage of the actual
efficiency relative to the target efficiency). For the case evaluated, the mean “indirect method”
STEP Factor at the station level moved from 86.698% (using monthly aggregated process data) to
86.135% (when discretized to 3-hour segments) which is roughly a 0.189% abs change in the
station’s thermal efficiency. This would appear fairly small on the station’s overall efficiency but
had a significant impact on the evaluation of the STEP Factor losses and the cost impact by the
change in the plant efficiency, e.g. the final feed water STEP Factor loss at a unit level moved from
2.6% abs to 3.5% abs which is significant for diagnostic and business case motivations.
Later the discrepancy between the direct STEP Factor and indirect STEP Factor were investigated
as the uncertainty bands did not overlap as expected. The re-evaluation of the baseline
component performance data resulted in the final feed water and the condenser back-pressure
heat rate correction curves being adjusted. The exercise revealed that there could be potentially
be significant baseline performance data uncertainty. The corrected indirect STEP Factor
instrument uncertainty was now found to be 0.468% abs which translates to 0.164% abs overall
efficiency. The combined uncertainty was corrected to 0.485% abs at a 3-hour segment size which
translates to 0.171% abs overall efficiency.
It has been deduced that the figures stated above are case-specific. However, the models have
been developed to analyse any coal-fired power plant at various operating conditions.
Furthermore, the uncertainty propagation module can be used to propagate uncertainty through
any other discontinuous function or computer model.
Various recommendations have been made to improve: the model uncertainty of STEP, data
acquisition, systematic uncertainty, temporal uncertainty and baseline data uncertainty.
ii
Declaration
I, Heeran Heerlall, hereby declare the work contained in this dissertation to be my own. All
information that has been gained from various journal articles, textbooks or other sources has
been referenced accordingly. I have not allowed, and will not allow, anyone to copy my work with
the intention of passing it off as their own work or part thereof.
_________________________
H Heerlall
iii
Acknowledgements
The Almighty is acknowledged in all my achievements and endeavours.
I would like to thank my wife, Karen Heerlall and children for the support given during the period
of this study and always.
Thanks to my academic supervisor for providing guidance, insight on the subject matter of this
study and steering me to acquire new valuable skills in the process of carrying out the study.
Thanks to Eskom’s: EPPEI, Energy Efficiency Care Group, Steering Committee of Technology,
Hendrina Power Station management and Production Integration Coal management for the
support.
Thanks to my team (Hendrina Power Station - Process Engineering Department) for affording me
time through their flexibility and diligence in their duties.
Thanks to Matimba Power Station’s –Performance and Test Department for the data provided.
Thanks to Kusile Power Station’s –Performance and Test Department for including me in their
performance tool development.
Thanks to Duvha Power Station’s –Performance and Test Department for access to plant data, and
signal processing tools.
Thanks to Nick Moolman for sharing his wisdom and information on the subject matter.
Thanks to the academic team at the Department of Mechanical Engineering at UCT for
contributions and insight and advice during course work and colloquiums.
iv
Table of Contents
List of Figures .................................................................................................................................... viii
List of Tables......................................................................................................................................... x
1. Nomenclature ................................................................................................................................. xi
1.1. General symbols ....................................................................................................................... xi
1.2. Acronyms and abbreviations .................................................................................................... xi
1.3. The naming convention for equations and program variables ...............................................xiv
1.4. Definitions ................................................................................................................................xv
2. Introduction ..................................................................................................................................... 1
2.1. Purpose ...................................................................................................................................... 2
2.2. Goals and objectives.................................................................................................................. 4
3. Literature review .............................................................................................................................. 6
Thermal power plants ...................................................................................................................... 6
World energy outlook ...................................................................................................................... 6
Thermal efficiency performance methodology ............................................................................... 9
Boiler efficiency principles ............................................................................................................. 10
Turbine plant efficiency .................................................................................................................. 11
Cycle losses ..................................................................................................................................... 12
Fundamental principles of STEP ..................................................................................................... 13
Uncertainty analysis ....................................................................................................................... 14
Law of Error Propagation ............................................................................................................... 14
Sequential Perturbation (SP) .......................................................................................................... 15
Monte Carlo.................................................................................................................................... 15
4. Methodology .................................................................................................................................. 18
4.1. Understanding STEP ................................................................................................................ 18
4.2. Data validation ........................................................................................................................ 25
4.3. Modelling for uncertainty analysis .......................................................................................... 28
5. Results and discussion.................................................................................................................... 42
5.1. Model uncertainty ................................................................................................................... 42
5.2. Instrument uncertainty ........................................................................................................... 47
v
5.4. Temporal uncertainty .............................................................................................................. 56
5.5. Baseline data uncertainty ........................................................................................................ 61
6. Conclusions and summary of recommendations .......................................................................... 63
6.1. Conclusion ............................................................................................................................... 63
6.2. Recommendations................................................................................................................... 63
7. References ...................................................................................................................................... 68
Appendix A. Equations and correlations from literature ................................................................... 73
A.1. Boiler performance principles [14] ............................................................................................. 73
A.1.1. Boiler efficiency ....................................................................................................................... 73
A.1.2. Flue gas / chimney loss ............................................................................................................ 74
A.1.3. Unburned CO loss .................................................................................................................... 74
A.1.4. Ash sensible and unburned combustible heat loss ................................................................. 74
A.1.5. Radiation and convection losses.............................................................................................. 75
A.2. Turbine plant performance principles [15]................................................................................. 75
A.2.1. Turbine deterioration Loss [15] .............................................................................................. 75
A.2.2. Turbine back pressure Loss...................................................................................................... 76
A.2.3. Other turbine plant losses ....................................................................................................... 76
A.3. Cycle losses[33] ........................................................................................................................... 78
A.3.1. Make-Up water loss ................................................................................................................. 78
A.4. Error propagation techniques ..................................................................................................... 78
A.4.1. Gaussian Quadrature (GQ) ...................................................................................................... 78
A.4.2. Method of Moments (MM) ..................................................................................................... 79
A.4.3. Sigma Point (SP) ....................................................................................................................... 80
A.4.4. Polynomial Chaos Expansion (PCE) and Probabilistic Collocation Method (PCM) ................ 81
Appendix B. STEP calculations ........................................................................................................... 82
Appendix C. Model uncertainty analysis .......................................................................................... 125
C.1. Carbon in refuse as per STEP [33] ............................................................................................. 125
C.2. Carbon in refuse as per BS EN standard [14] ............................................................................ 126
C.3. Carbon monoxide loss as per BS EN 12952-15 [14] .................................................................. 126
C.4. Dry flue gas (DFG) loss as per STEP [33] ................................................................................... 127
C.5. Dry flue gas (DFG) loss as per BS 2885 [5] ................................................................................ 128
vi
C.6. Dry flue gas (DFG) loss as per BS EN 12952-15 [14] ................................................................. 128
C.7. Hydrogen and moisture loss as per BS EN 12952-15 [14] ........................................................ 131
C.8. Hydrogen and moisture loss as per STEP [33] ......................................................................... 132
C.9. Make-up water [33] .................................................................................................................. 133
C.10. Radiation and convection loss as per BS EN 12952-15 [14] ................................................... 135
C.11. Radiation loss as per STEP [14] ............................................................................................... 135
C.12. Turbine deterioration loss as per STEP ................................................................................... 136
C.13. Turbine deterioration loss as per the IEC [72] ........................................................................ 136
C.14. Turbine deterioration loss as per ASME PTC 6 Report-1985 [46].......................................... 136
Appendix D. STEP program inputs details ........................................................................................ 137
Appendix E. STEP program output tag list ....................................................................................... 143
Appendix F. Results .......................................................................................................................... 149
Appendix G. Model verification ....................................................................................................... 155
vii
List of Figures
Figure 2-1: Diagram of a typical coal-fired thermal power station [9] ................................................ 3
Figure 2-2: Illustration of the impact of averaging inputs rather than the result ............................... 4
Figure 3-1: World projected energy mix [7] ........................................................................................ 7
Figure 3-2: World projected electricity generation mix [12] ............................................................... 8
Figure 3-3: South African overall efficiency vs availability (courtesy: Y. Maharaj-Eskom) .................. 8
Figure 3-4: Efficiency of public conventional thermal power production in Europe [11] ................... 9
Figure 3-5: Illustration of Monte Carlo randomized iterations for Gaussian distributed inputs and
outputs - adapted from [47] .............................................................................................................. 16
Figure 3-6: Demonstration of Monte Carlo convergence as a function of the number of iterations
[54] ..................................................................................................................................................... 17
Figure 4-1: Illustration of STEP Factors .............................................................................................. 20
Figure 4-2: STEP calculation architecture .......................................................................................... 20
Figure 4-3: STEP boiler calculation sequence .................................................................................... 21
Figure 4-4: STEP turbine calculation sequence .................................................................................. 22
Figure 4-5: STEP cycle calculation sequence...................................................................................... 23
Figure 4-6: Development of the Mathcad STEP model ..................................................................... 24
Figure 4-7: Power plant data flow ..................................................................................................... 26
Figure 4-8: Example of processed signals/tags .................................................................................. 28
Figure 4-9: Distribution of error on repeated measurement [47] ..................................................... 30
Figure 4-10: Distribution of error as a function of time [16] ............................................................. 31
Figure 4-11: Gaussian distribution probability density function [47] ................................................ 33
Figure 4-12: Uncertainty propagation model architecture ............................................................... 36
Figure 4-13: Step 2 - Example of monthly composite input data ...................................................... 38
Figure 4-14: Step 3 - Example of base results .................................................................................... 39
Figure 4-15: Step 4 -Uncertainty propagation to calculated variable per segment run for each input
variable ............................................................................................................................................... 40
viii
Figure 4-16: Step 6 - Illustration of segmented operational data ..................................................... 40
Figure 4-17: Step 8 - Example of combined uncertainty of calculated variables per segment run .. 41
Figure 5-1: Turbine deterioration loss comparison ........................................................................... 46
Figure 5-2: Instrument uncertainty on SF losses ............................................................................... 51
Figure 5-3: Variable significance on station cycle indirect STEP Factor uncertainty ......................... 52
Figure 5-4: Final feed water - original heat rate correction factor curves ........................................ 53
Figure 5-5: Variable significance on station cycle direct STEP Factor loss uncertainty ..................... 53
Figure 5-6: Variable significance on uncertainty on the station cycle and turbine plant SF losses .. 54
Figure 5-7: Variable significance on uncertainty on the station boiler SF losses .............................. 55
Figure 5-8: Generated power profile by resolution ........................................................................... 56
Figure 5-9: Station STEP Factor results with uncertainty bounds ..................................................... 58
Figure 5-10: Station STEP Factor uncertainty by resolution .............................................................. 60
Figure 5-11: Unit 4 segmented SF loss evaluation ............................................................................ 60
Figure 5-12: Unit 5 segmented SF loss evaluation ............................................................................. 61
Figure 5-13: Station STEP Factor results with uncertainty bounds – corrected ................................ 62
Figure 5-14: Station STEP Factor systematic vs combined uncertainty - corrected .......................... 62
ix
List of Tables
Table 2-1: Legend to Figure 2-1 ........................................................................................................... 3
Table 3-1: Summary of the evolution of coal-fired power plants (adapted from[9][10]) ................... 6
Table 3-2: STEP loss categories .......................................................................................................... 13
Table 4-1: Case study - power plant description ............................................................................... 18
Table 5-1: Summary of loss review .................................................................................................... 42
Table 5-2: Sensitivity analysis on the fraction of make-up water lost at main steam conditions ..... 44
Table 5-3: Truncated results greater 5% – instrument uncertainty propagation ............................. 48
Table 5-4: Truncated results drill-down approach – instrument uncertainty propagation .............. 50
Table D- 1: Input variable instrument uncertainty .......................................................................... 137
Table E- 1: STEP program output list ............................................................................................... 143
Table F- 1: Systematic uncertainty propagation full results ............................................................ 149
Table G- 1: STEP model verification ................................................................................................. 155
x
1. Nomenclature
1.1. General symbols
Cp Specific heat capacity at constant pressure (J/kg.K)
Eff or η Efficiency
h Specific enthalpy (J/kg)
h Hour
i Range variable – each power cycle unit (i.e. 1..6 for a 6 unit power plant)
j Range variable - generic
J Energy
k Empirical constant
L Loss (%)
m Mass
M Molar mass (kg/kMol)
ṁ or `m or MFR Mass flow rate
N ,n Number
P Pressure
P or Pow Power
PF Power factor
Q Heat
S Sample standard deviation
T Temperature
t Time
u Uncertainty
V Volume
w Absolute humidity (kg/kg)
x Mass fraction (kg/kg)
y Volume fraction (V/V)
Greek symbols
η or Eff Efficiency
µ Mass ratio relative to the mass of fuel
σ Population standard deviation
1.2. Acronyms and abbreviations

Abs/abs Absorbed / absolute
ACC Air Cooled Condenser
Acc /Acct Accounted
Act Actual
Adj Adjustment
AH Air heater
AI Abrasive Index
Amb Ambient
AP Available plant (running plant)
AR As-received / as-fired
xi
AS Atomizing steam(for fuel oil firing)
ASME American Society of Mechanical Engineers
AT Acceptance test
Aux Auxiliary
AVL Average load generated
BA Bottom ash
BD Blow down
BFPT Boiler feed pump – turbine driven
BMCR Boiler maximum continuous rating
Boil Boiler
Bool Boolean
BP Backpressure (Condenser Pressure)
BR Burn rate
BS British Standard
BSI British Standards Institute
Ce Carbon equivalent
BCEGB British Central Electricity Generating Board
CEP Condensate extraction pump
CF Correction factor
CIR /CIG Carbon In refuse / Carbon in grit
CLF Capability loss factor (unavailability)
CO Carbon Monoxide
Comb Prod Combustion products
Comm Commissioning
Cond Condensate
Cons Consumption
Corr Correction or corrected
CSV Comma-separated values
Cu Unburned coal (fraction of coal-fired)
CV Calorific value
DA Dry Air
Depres Depression (difference between LP turbine exhaust temperature and steam
saturation temperature at low-pressure exhaust pressure
Det Deterioration
DFC Mass ratio of Dry Flue Gas to 1kg Carbon burned
DFG Mass ratio of Dry Flue Gas to 1kg fuel burned
DIN Deutsches Institut für Normung (German Institute for Standardization)
DMAIC Define, Measure, Analyze, Improve and Control
EFP Electric (driven) feed water pump
EM Excess moisture
EPPEI Eskom Power Plant Engineering Institute
Equiv Equivalent
Extr Extraneous
FA Fly ash
FC Fixed Carbon
FD Forced draught fan
FO Fuel oil
FW Feedwater
Gen Generator
GQ Gaussian Quadrature
GT Generator transformer
xii
HGI Hard Grove Index
HPT High-pressure turbine
HR Heat rate
HV High voltage
IEC International Electrotechnical Commission
Imp Import
Incr Incremental
IPT Intermediate pressure turbine
ISO International Organization for Standardization
L Loss (%)
Leak Leakage (%)
LF Load factor
LH Left hand
LPT Low-pressure turbine
LS Live Steam (steam at high-pressure turbine inlet)
LV Low voltage
MCR Maximum continuous rating
MCS Monte Carlo simulations
MF Moisture-free Basis
moist Moisture
MP Measurement point
MR Mill rejects
MS Main steam
MU Make-up (demineralized water)
MV Medium voltage
NER Normal economic rating
NL No load
OL Off -load
Opt Optimum
PCLF Planned capability loss factor (%) -the measure of planned unavailability
Pract Practical
R&E Revenue and expenditure
Rad Radiation
RC Radiation and convection
ref Reference
ReH Re-heater
Rej Rejection (To Atmosphere)
RH Right hand
RSS Root sum square
RT Run time
RW Raw water
SA Steam air heater
SAH Sulphur
Sat Saturation
SB Station board
Sby Standby
SD Shutdown
SF STEP factor (actual – target)
SFP Steam (driven) feed water pump
SM Surface moisture
STEP Station Thermal Efficiency Performance
xiii
Chapter 2. Introduction
2. Introduction
South Africa and Africa as a whole is currently faced with a significant challenge of meeting its
energy demand whilst transforming to cleaner electric energy production. Eskom is a state-owned
power utility that manages the distribution, transmission and dispatching of electricity on the
South African power grid. Almost 85% of South Africa’s electric energy is provided via Eskom’s
conventional coal-fired power plants [1]. One of the challenges identified is the ageing of Eskom’s
power generation fleet with limited capital to meet the maintenance requirements for optimal
performance and reliability. Over the last decade or two, there has been a great focus on
minimizing load shedding which might have meant that some of the assets would have been
operated at higher load factors and in some cases exceeding recommended maintenance
intervals, to compensate for diminished capacity. A significant portion of the power production
fleet is in the final 10 years of its designed operating life. Six out of the thirteen fully commercial
power plants are to be decommissioned by 2030 as per the Integrated Resource Plan 2018 Update
[2]. Naturally, significant investments in these power plants would be curbed as it would be
difficult to justify a return on investment for a short remaining operating life.
A consequence of the cited challenges and constraints is the decline in the fleet’s thermal
efficiency. Thermal efficiency is an indication of the state of the plant's performance and
correlates with the plant’s availability. A lower thermal efficiency will result in: plant components
having to work near design limits for extended periods which reduces the remnant life of the
components, or the power plant not being able to achieve the rated generating capacity. The
improvement of a power plant’s thermal efficiency should contribute positively to the power
generation capacity and is a significant aspect in reducing the cost of operations.
The path to recovery would require some engineered methods of getting the maximum
performance gains with minimum capital injection. Whilst there are various tools employed in
industry, Eskom being the largest power producer in Africa has adopted a program called Station
Thermal Efficiency Performance (STEP) to evaluate the thermal performance of their coal-fired
power plants.
The goal of this research is to improve the means of efficiency and thermal loss evaluation for
coal-fired power plants so that improvement initiatives can be confidently justified. A key focus of
the study is the evaluation of the accuracy of the stated method that is used.
1
2.1. Purpose
South Africa is the largest power producer in Africa [3]. Eskom claims to produce 95% of this
power [1], with 90% of its generating capacity in the form of conventional coal-fired power plants
[4]. These plants employ the typical regenerative-reheat Rankine cycle where: the turbines are
driven by superheated steam (see Figure 2-1), the LP turbine exhaust steam is condensed via
condensers supplied by wet or dry cooling water systems, the condensate is preheated by
regenerative feedwater heaters and steam is generated by pulverized fuel (PF) boilers. Eskom uses
a thermal performance model called STEP to evaluate the Rankine cycle performance and identify
losses in the power generation process. Although there are differences in technology, such as
drum-boilers vs once-through, air-cooled condenser (ACC) vs wet natural draught cooling towers,
subcritical vs supercritical boilers, reheat and non-reheat cycles, the model uses a standardized
approach with a few differences in the calculation methodology applied. The differences in the
calculation methodologies are augmented with customized inputs for each power station. There is
also functionality to compensate for differences in the configuration (positions of measured
inputs) and technology (mainly the cooling water system employed).
The STEP model intends to provide an assessment to quantify the state of operations and
monetary impact for the deviations from ideal or expected performance (see section 4.1 for more
details). The calculation principles of STEP are based on industry standards for acceptance testing
of boilers [5] and turbines [6], which typically employ energy balances based on the Rankine cycle
thermodynamics (see Figure 2-1: Diagram of a typical coal-fired thermal power station [9], the
principles are further elaborated in section 4.1 and Appendix B).
The current topic for research and development was initiated from the need for continuous
improvement on the thermal efficiency performance/accounting tools. In the current global
context, there is a strong focus on cleaner electricity production coupled with affordability. Even
though coal-fired power plants are viewed as a “dirty” production process, it is envisaged that
coal-fired generation will have a significant contribution to the energy mix in the next fifty years to
come [7].
In the South African context, the power plants are ranked in merit order for the dispatching of
electricity to the national grid. The ranking is determined predominantly by the cost of primary
energy. The power plant industry is transitioning to competitive dispatching of energy that
considers the cost of productions, thermal efficiency and emission offsets such as carbon tax [8].
STEP is an instrument that can be used to drive the medium-term objectives of environmental
sustainability and cost reduction through energy efficiency maximization. In the current work, the
STEP model is extended to evaluate instrument and temporal (random) uncertainties on the
2
model output values. The quantification of these uncertainties, and insights gained from the STEP
modelling approach, could lead to an increase in accuracy. To implement the uncertainty
estimation, a computer model of the STEP calculation methodology is developed using the Python
3.8 programming language. During the development of the computer model, the calculation
approach will also be reviewed.
Figure 2-1: Diagram of a typical coal-fired thermal power station [9]
Table 2-1: Legend to Figure 2-1
No. Description No. Description No. Description

1 Cooling tower 10 Steam control valve 19 Superheater
2 Cooling water pump 11 High-pressure steam turbine 20 Forced draft fan
3 Transmission line (3-phase) 12 Deaerator 21 Reheater
4 Step-up transformer (3- 13 Feedwater heater 22 Combustion air intake
phase)
5 Electrical generator (3- 14 Coal conveyor 23 Economiser
phase)
6 Low-pressure steam turbine 15 Coal hopper 24 Air preheater
7 Condensate pump 16 Coal pulverizer 25 Precipitator
8 Surface condenser 17 Boiler steam drum 26 Induced draft fan
9 Intermediate 18 Bottom ash hopper 27 Flue-gas stack
pressure steam turbine
3
The function evaluated at the average input, x = 2 (4) does not equate to the average of the
results for x = 1 and x = 3 (8). The above principle demonstrated would be much more critical
in the case of flexible operations i.e. load swings for non-baseload power plants. With the
increasing penetration of renewable power generation, it is expected that the power plants
would operate with an increase in load transients due to the nature of renewable power
production.
5
Chapter 3. Literature review
3. Literature review
This chapter will focus on the evolution of thermal power plant technology and the development
of performance standards, the adoption of these standards in the development of the STEP
modelling approach and uncertainty propagation methods employed in power plant efficiency
evaluation applications.
Thermal power plants
Thermal power plants were developed during the 18 th century with the first commercial plant
being commissioned in 1882 based on the reciprocating steam engine technology [9]. The first
steam turbine was developed in 1884 with a thermal efficiency of only 1.6% [10]. By 1905 steam
turbines replaced all reciprocating engines in large central power stations. During the 1900s
boilers were constructed with evaporators, economizers and superheaters. By the 1910s
condensing turbines with extractions for feed heating operated at a thermal efficiency of 15%. In
the 1930s, pulverised fuel boilers with reheat steam had become prominent. Supercritical steam
plants were produced by the 1960s. A summary of the technology evolution for coal-fired power
plants can be found in Table 3-1.
Table 3-1: Summary of the evolution of coal-fired power plants (adapted from[9][10])
Period Technology changes Thermal

Efficiencies
Achieved (%)
1882 Reciprocating Steam Engines 6
1884 First Steam Turbine 1.6
1900s Condensing turbine, the addition of economizer,
superheater,
1910s Addition of feed heating and air preheaters 15
1920s Once thought boilers
1930s Inclusion of re-heaters
1940 Subcritical main steam temperature increases 37
1960 Supercritical steam generation (typically above 250 47
bar and 565 oC)
2016 Ultra-Super critical steam generation (>600oC) 50
World energy outlook
The permanence of coal-fired power generations technologies is threatened by the penetration of

renewable energy technologies into the energy and power generation sectors. However, from
Figure 3-1 and Figure 3-2, it can be seen that the projections by the International Energy Agency
6
(IEA) and Statista suggest that coal and other fossil fuels will maintain a significant share in the
world’s energy mix in the medium-term. Coal-fired technologies are still maintained by 2050 even
on the “New Policy Scenario”, hence significant effort must be made to improve the thermal
efficiencies of conventional fossil plants to reduce the carbon footprint as well as NOx and SOx
emissions. According to The World Bank, about 40 countries have already implemented carbon
pricing globally with significant commitments from others to move in this direction [8]. These
countries particularly in Europe have demonstrated this improvement [11].
The South African average power generation efficiency has shown a steady decline over recent
years as observed from Figure 3-3 (x and y labels removed for confidentiality reasons, indicative
only) The planned and unplanned unavailability has increased from the financial year 2012. A
direct relationship between thermal efficiency and plant availability can be observed. In the
current state, there are significant opportunities to improve thermal efficiency through enhanced
deficiency identification and resolution. There are significant factors that may have a detrimental
effect on thermal efficiency which include: inadequate funding, execution of maintenance
strategies, remnant design lifespan, skills, coal quality, etc.
Other parts of the world have shown an upward trend in thermal efficiency where carbon pricing
has been implemented, particularly in Europe. Thermal efficiency had been on a decline in this
region from 2010 but has recovered post-2014 as demonstrated in Figure 3-4 by the European
Environment Agency (EEA).
Figure 3-1: World projected energy mix [7]
7
Figure 3-2: World projected electricity generation mix [12]
Figure 3-3: South African overall efficiency vs availability [13]
8
Figure 3-4: Efficiency of public conventional thermal power production in Europe [11]
Thermal efficiency performance methodology
Standard energy balance methods have been employed in the power plant industry since the
1800s. The first ASME boiler test code was established in 1884 (ASME established in 1880) [14].
The DIN 1942 and BS 2885 standards can be traced back to 1975 and 1974 respectively.
Of late, the number of performance standards has reduced due to merging and collaborations of
standards by the standards institutes, especially in Europe. The major standards applied to
Eskom’s generating fleet thermal performance analysis are the: BS-EN standard [15], BS–EN(IEC)
standards [6][16] and ASME standards [17][18][19].
From the mid-’90s to current, there is a growing interest in exergy analysis which has yielded
various studies on fossil-fired power plants. The exergy efficiency method accounts for non-useful
energy by using the 2nd law of thermodynamics i.e. entropy change of a stream to be brought to
equilibrium with its surroundings (destruction of exergy) and hence usually yields higher
efficiencies when compared to standard thermal efficiency calculation techniques. Many authors
deem exergy analysis to be the preferred method of analysing an energy conversion process as it
would differentiate between losses to the environment vs internal irreversibilities. Rashad and
Elaihy [20] presented both an energy and exergy analysis on a 1260 MWe natural gas power plant
in Egypt in 2009 which demonstrated the above point. Aljundhi presented an exergy study done
for 396 MWe heavy fuel oil power plant in Jordan also in 2009 [21], and Mahamad et al. presented
9
across and within the boiler tubes coupled with a CFD simulation to acquire fluid properties
external to the wall tubes. The CFD simulation incorporated combustion dynamics. The method
could further be expanded to explore the impact of coal properties on the dry flue gas loss. It
would be interesting to explore the heat transfer impact as a function of the coal elemental
composition, that influences the thermodynamic losses.
Turbine plant efficiency
The turbine plant efficiency is computed conventionally by the method below [16] in generalized
form:
Pnett
Turb.Plant  (3.3)
  m j hj 
where Pnett is the power measured at the generator terminal less the power consumed for
auxiliaries and j is the flows to which heat is added from external sources. The above method was
demonstrated by Geete and Khandwawala [28] which had been applied to a 120 MWe thermal
power generating unit. They used the method to study the impact on heat rate by varying the
main steam (throttle) temperature. Similar work was carried out by Rout et al. [29] where the
output power was also studied. Karakurt and Gunes [30] used this methodology to investigate the
turbine performance relative to partial loads.
The specific efficiency of a turbine cylinder is computed by the “Enthalpy Drop” method:
hin  hOut
Turb.Cylinder  (3.4)
hin  hIsentropic @ POut
where hIsentropic @POut is isentropic enthalpy (or enthalpy of steam at the turbine outlet pressure and
inlet entropy). The word cylinder implies all turbine stages including the inlet valves.
Turbine plant losses or the indirect method are detailed in Appendix A.2. The quantification of
losses are derived from the heat rate correction factors associated with parameter deviations
from design. These curves are typically supplied by the plant designers. GE Power Generation [31]
presented examples of typical heat rate correction curves which are aligned with the requirements
of the ASMEs performance test code [32]. There are various turbine process modelling tools that
can be used to generate the heat rate correction curves; Akpan and Fuls [33] recently (2020)
demonstrated a methodology for developing and validating such models and went on to develop a
generalized means of predicting the turbine cycle heat rate as a function of load. The validation
11
included a comparison to actual performance data from 11 South African coal-fired power plants
against models developed with the ETAPro® -Virtual Plant software.
Cycle losses
Works power loss is typically expressed as a percentage of electrical energy used for internal
operation in the power plant to the electrical energy that is supplied to the national grid [34],[35].
Naturally, electricity will be consumed by components in the power generation cycle i.e. motors,
heating elements, controllers, etc. Therefore, this loss cannot be avoided but it is important to
quantify and compare to a baseline, to be able to detect plant performance abnormalities. The
measurement is done at strategic locations i.e. unit transformers, station transformers, station
electrical boards and unitized electrical boards. Some online performance tools such as EtaPro®
can calculate the power consumption by using the current transducer readings for each motor
installed in the power plant. Whilst this method is very useful for diagnostics, the uncertainty is
considered to be too large for energy accounting purposes and not all components are measured.
Cycle Make-up water losses are usually quantified by measuring the top-up that is required to
maintain the deaerator or condenser levels and is expressed as a percentage of the total steam
raised. Ideally, the steam-condensate-feed-water system is a closed loop, however practically it is
not the case, as losses can be experienced from passing drain/blowdown valves across the cycle,
boiler tube and pipe leaks, venting, pump gland leakage, etc. Generally, the boiler and turbine
plant manufacturers estimate an acceptable loss relative to the MCR live/throttle steam flow.
Eskom targets these losses to be between 0.8 to 1%. The heat efflux through these losses is
attempted to be quantified through simple thermodynamics which employs various assumptions
(see Appendix A.3.1 for details on a methodology used in industry). A study conducted by Sewlall
[36] in 2019 indicated that: for purpose of water accounting in Eskom, it is assumed that 50% of
the total make-up is evaporated and the balance is recovered as a liquid into the station drains. It
is stated that actual plant data is required to improve this estimate of a typical split between
losses at steam and feedwater conditions. There are various texts [37],[38],[39],[40] involving
power plant water balances, energy conservation and cycle isolation, that mention feedwater
make-up losses, but omit detail in localized quantification and costing. However, these texts do
suggest that make-up loss targets could be as low as 0.5%. A study by Kaushik and Khanduja [41]
was carried out on a combined cycle thermal plant in 2007. A great effort was made to quantify
the contributions to demineralized water make-up using the Six Sigma DMAIC methodology. It was
assumed that all of the make-up water supplied is due to losses at main steam conditions.
12
Fundamental principles of STEP
STEP was developed by the Central Electricity Generating Board (CEGB) in the U.K in 1966. In 1970
STEP was adopted by Eskom and over the years been adapted based on operational experience
gained from plants within their fleet [42]. The STEP program baselined the power plant
performance as per the acceptance test data acquired during the initial commissioning of each
type of unit design. The acceptance tests were based on the boiler performance standards [5],[43]
and the turbine performance standard [6]. The current version of STEP is prescribed in an internal
standard [34] and has been implemented to run on a centralized application server in 2015 (the
STEP model is computed using the application software - ETAPro®). A superseded manual [44],
also contains a detailed calculation methodology on how the calculations were previously
performed which give some history in the development.
In general, heat rate targets for acceptance tests are produced by applying correction factors
deduced by design sensitivity analysis to account for externally imposed conditions that differ
from design parameters namely: load factors, ambient temperature, ash in coal, H2 and moisture
in fuel feedstock, power factor, turbine backpressure associated with ambient temperature, etc.
The target is then compared to the actual performance test results. The STEP model differs by
applying these correction factors when the actual condition deviates from the parameters
established during acceptance testing and compare to the actual performance. In South Africa,
most of the coal-fired plants employed the British and IEC standards for acceptance testing.
The resolution of the STEP program input is one month (monthly average operating parameters
are inputted to the model). The typical outputs of STEP include targets, actuals, SF (STEP Factor)
and cost impact of losses refected in Table 3-2.
Table 3-2: STEP loss categories generated from [33]
Cycle Boiler Turbine

Cycle efficiency DFG loss Condenser BP loss
Heat rate CIR loss FW temperature loss
Aux power loss H2 & moisture loss Turbine Deterioration loss
MU loss Radiation loss Re-heat steam temperature loss
Mill reject loss Main steam temperature loss
Exergy analysis considers the conditions of surroundings to determine the quality of energy
available, thereby giving a more realistic measure of effectiveness. It only considers the useful
portion of energy that is to be transformed as a maximum base for effectiveness analysis. The SF
approach achieves this by calculating the deviation in thermal performance from the ideal case
(maximum base), which essentially focuses on what is within the utility’s control. The SF method
goes further to consider the impact of other externalities e.g. the grid-induced power factors and
13
load factor. Refer to Appendix B for a detailed calculation methodology in the form of sample
calculations or section 4.1 for a summarized calculation flow in the form of a diagram.
Uncertainty analysis
There are two main approaches to uncertainty propagation, namely the derivative approach which
is also known as the Law of error propagation [44] and the probabilistic approach [52]. Both these
approaches have various numerical methods for their implementation. Whilst there is a multitude
of approaches and variants of these methods, this study will be limited to some of the more
popular methods. General statistics and probability principles are explained in more detail in
section 4.3.2.
Law of Error Propagation
Uncertainty indication in power plant thermal performance evaluation has been included in the
performance standards in recent revisions and has also been included as a quality measure for
acceptance testing contracting. Typical turbine efficiency uncertainty for turbine acceptance
should be within 0.6 -0.9% [16], the typical uncertainty for steam generators that is coal-fired is
between 0.4% and 0.8% [17] and the maximum acceptable for the cycle is 3.0% using the direct
method [18]. The performance standards as per previous references prescribe a method for
uncertainty propagation as per the Laws of Error Propagation. Most texts on error propagation
make some reference to the GUM- Guide to the expression of uncertainty in measurement by ISO
[45], which demonstrate these laws by the equations below (derived from the 1st order Taylor
expansion of a function. The laws are synonymous with other standard publishers such as VDI[46]
and ASME[47]. BSI has also adopted the GUM which is described by the equations below:
2
  R 
L 
uR      .uxi  (3.5)
i 1   xi  x  x 
 i 
The equation (3.5) is used for uncorrelated inputs:

2
  R 
L  L 1 L
R R
uR      .u   2  . .u( xi , x j ) (3.6)
i 1   xi  x  x  i 1 j  i 1 xi x j
xi
 i 
The equation (3.6) is used for correlated inputs:
K
uxi  u
k 1
k
2
(3.7)
The uncertainties are combined by the Root Sum Square (RSS) method as per the equation (3.7),
14
where:
R - result
uR - propagated uncertainty to the result of a function at the mean of each input variable
xi - each variable in a function
x j = xi 1 - sequential pair of variables
uxi - combined uncertainty of variable for a single measurement point
u( xi , x j ) - estimated covariance of associated with xi and x j
uk -each type of uncertainty (systematic, random-temporal, random-resolution…)
Sequential Perturbation (SP)
Sequential Perturbation is a controlled numerical method that gives the same result as the
equation (3.5) as demonstrated by Figliola and Beasley [48]. The method circumvents the need for
carrying out the first-order differentiation of complex functions. SP is ideal for smaller amounts of
input variables and is relatively computationally inexpensive. This method had already been
employed by Moffat [49] in 1988 and demonstrated by Mantheufel [50] in 2013 for thermal fluid
applications. Sequential Perturbation is further detailed in section 4.3.1. There are variations in
sequential perturbation such as the Sigma Point methods, where the perturbations for each
variable are weighted as a function of its probability density function [51].
Monte Carlo
Monte Carlo simulation (MCS) is a probabilistic numerical method where the uncertainty of each
variable in a function is randomly varied as a function of its probability density function. MCS was
experimented with during the 1930s and developed in the late 1940s for use in nuclear weapons
projects [52]. Monte Carlo simulation for uncertainty analysis is currently used in a broad
spectrum of applications i.e. finance, engineering, physical science, supply chain, risk
management, applied statistics, computational biology, etc.
The MCS method is used in complex functions with a significant multitude of the input variables
but can be computationally expensive as the iterations could run into the order of 10 4 to 106
before the distribution of the result converges to the specified standard deviation [48]. Padulo et
al. [51] have deemed this method as impractical for robust engineering design due to the high
number of iterations required for numerical convergence to be reached (see also Figure 3-6:
Demonstration of Monte Carlo convergence as a function of the number of iterations [55]).
15
MCS is the most commonly used method due to its flexibility, minimal limitations and ease of
scripting. There are variations in the employment of the method (i.e. Sequential Monte Carlo,
Quasi-Monte Carlo, Markov Chain Monte Carlo) to reduce the computational expensive by the
improved sampling of the input variables [53]. The method has a reasonable comparison with the
GUM [45] as proven by Papadopoulos and Yeung [54] with an average deviation of 0.03% and
maximum sample deviation of 0.8% shown in their study.
Figure 3-5 demonstrates the application of Monte Carlo simulation where x' is the true value of
any input variable which is estimated to be x and R' is the true value of any output variable.
Monte Carlo simulations were recently applied (2017) by Keshavarzian in reliability analysis for
cyclic fatigue [55]. Figure 3-6 is an example of a convergence graph for a bolt hole damage
parameter demonstrated in this study which further demonstrates the computational expense of
the method.
The MCS is commonly used as a benchmark in various comparative uncertainty studies. There are
multitudes of numerical and probabilistic methods with their variations that are lessor used in the
research work involving uncertainty propagation. These include Gaussian quadrature [56], method
of moments [51], sigma point [57], polynomial chaos expansion and probabilistic collocation
method [51],[58],[59],[60],[61],[62],[60],[63],[64]. Refer to Appendix A.4 for more detail on these
techniques.
Input variables are randomly varied iteratively within this range to produce a results data
set. The iteration continues until  estimated   result or when the result converges within a
specified tolerance,  Result  niterations   tolerance . In many cases the variance,  2 is used
instead.
Figure 3-5: Illustration of Monte Carlo randomized iterations for Gaussian distributed inputs and outputs - adapted
from [48]
16
Figure 3-6: Demonstration of Monte Carlo convergence as a function of the number of iterations [55]
17
Chapter 4. Methodology
4. Methodology
A 6 x 600MWe wet-cooled power plant situated in Mpumalanga-South Africa, was modelled using
the STEP methodology. The power plant was selected based on its location, availability of process
data, and support of site personnel. See Table 4-1 for technical description.
Table 4-1: Case study - power plant description
Power rating 6 x 600MWe

Twin tower, once through, single re-
Boiler type heat
Main steam pressure 16.6 MPa fixed
Main steam temperature 540 oC
Fuel Sub-bituminous coal
Cooling Wet natural draught
4.1. Understanding STEP

Eskom’s thermal accounting models (server-based EtaPro® application and Excel versions) were
reviewed against Eskom’s standard [34] and its referenced industry standards. The STEP
calculations are based on general power plant thermodynamics, heat and mass balances. In a few
calculations, organizational intelligence such as turbine deterioration factors and start-up heat had
been included in the calculations. Certain power cycle component efficiency calculations are
technology-specific and require certain modifications; for example, the calculation of wet-cooled
vs. dry-cooled condenser performance and the effect on the overall power cycle efficiency it has.
It is important to reiterate some of the concepts and terminology that is used in STEP to
understand the findings and results of the current study. Thermal performance can be
measured/calculated using two methods. The first being: the “Direct Method” also known as the
“Input-Output Method”. Here the thermal efficiency of a system is determined by only quantifying
the energy flowing into a system and the desired energy output. So the conventional rule applies:
EOutput
Direct  (4.1)
EInput
In the case of a power plant, the desired output is the electrical energy sent out to the grid, the
input energy would be the chemical energy of the fuel. The second approach is the “Indirect
Method” also known as the “Energy Balance Method”. Here the thermal efficiency of a system is
determined by quantifying energy losses as a percentage of the total energy input. The principle
employed is that the energy output must be equal to the input energy minus all of the losses.
18
Indirect  100   Losses (4.2)
Ideally, both methods should give the same value; however, a deviation can be expected due to
the uncertainty of the measurements used in both approaches.
From Figure 4-1 it can be observed that STEP intends to normalize thermal performance against
the target efficiency, so the STEP Factor (SF) is the percentage of the target efficiency achieved.
The STEP Factor can be viewed as a thermal effectiveness metric. The STEP Factor losses are also
relative to the target, so the ∑STEP Factor losses and the indirect STEP Factor must add up to
100%. The unaccounted STEP Factor losses will then be the difference between the direct STEP
Factor and the indirect STEP Factor. The mathematical expressions can be found below:
Direct
SFDirect   100
Target (4.3)
Loss.SFj  Loss j.Actual  Loss j.Target (4.4)
N
SFIndirect  100 -  Loss.SFj (4.5)
j 1
where j is the individual losses i.e. Carbon in Refuse, Dry Flue Gas, Hydrogen & Moisture,
Backpressure, etc.
Figure 4-2: STEP calculation architecture, illustrates the basic architecture of STEP, where the
calculations are categorized into four sections namely the boiler, turbine, cycle and station
sections. The outcomes of each section are listed in Figure 4-2. Figure 4-3, Figure 4-4 and Figure
4-5 show an overview of the inner workings of the STEP calculation methodology. The detailed
STEP calculations are provided in Appendix B. The station section of the calculation as depicted in
the architecture diagram (Figure 4-2), are merely the aggregating of the calculations performed at
the unit level to acquire station results.
19
Figure 4-3: STEP boiler calculation sequence
21
Figure 4-4: STEP turbine calculation sequence
22
`
Figure 4-5: STEP cycle calculation sequence
23
The model uncertainty is appreciated by reviewing the thermal losses computed in STEP against its
latest version of its referenced standards [6],[15]. In some cases, other standards are explored
where there are significant conflicts. Sample calculations are performed for losses that are
omitted in the STEP program, to appreciate the magnitude of such model omissions. The model
uncertainty would then be relative to the adopted set of standards as used in the performance
baseline testing [65]. The losses that are reviewed include Carbon in Refuse, Dry Flue Gas, Carbon
Monoxide, Make-up Water, Radiation, Turbine Deterioration, and Hydrogen and Fuel Moisture.
The turbine plant losses are mostly a function of parameter sensitivity studies performed on
thermodynamic modelling packages. (Refer to Appendix C for calculation methodology and results
of the sample calculations). The results are discussed in Section 5.1.
4.2. Data validation

The section describes how the data used in this research, otherwise known as the specific case
data, is acquired and processed before being applied to the STEP and uncertainty propagation
models for observation.
4.2.1. Data acquisition

Figure 4-7: Power plant data flow, illustrates how the data used in this research is acquired. Online
plant data was sourced from the power plant’s data server. The VA View ® (database software)
web interface had been utilized to query raw data on a minute interval for one month (this was
the set limitation for the server queries). Electrical metering data had been made available
through integrating software namely EMDAS® which publishes to the GPSS database at a
resolution of one hour. Laboratory data for coal parameters and carbon in refuse was acquired via
site personnel, which is typically stored on another database (LIMS®) with a minimum reporting
resolution of one day.
25
Figure 4-7: Power plant data flow
4.2.2. Data screening

It was intended to process raw data and use typical instrument uncertainties found from literature
without exploring the detailed elements of each variable’s systematic uncertainty (this scope has
been included in a different study that was to run concurrently with this study). It would be
nonsensical to include instrumentation drift when the unit is not producing power, hence data
during off-load conditions were disregarded from the analysis.
4.2.3. Data pre-processing

Plant process data is converted to measurement units that are required in STEP, e.g.:
kg 60s 1Gg
See point 2 in Figure 4-8, mFW .1 min  mFW .avg  
s 1 106 kg
A significant amount of the STEP input parameters are not directly measured and had to be
derived. This is particularly in the case of component running hours, where time is accumulated by
the summation of the time step where the motor current signals are filtered to be above the
26
assigned off-load drift tolerance. In the very same manner power (MW) is converted to energy
totals (GWh) for electric power, e.g.:
kg 1 min
See point 1 in Figure 4-8, RTBoil . for 1 min  If ( mMain _ Steam  20 , ,0 )
s 60 min/ hour
STEP has two types of input data, namely average input parameters and cumulative input
parameters. The STEP computer model internally segments the data into specified chunk sizes. For
each segment analysed, the average input parameters are weighted by the electrical units
generated, e.g.:
NSegment _ size
 ( PCondenser .Acti  PowGeneratedi )

see range 4 in Figure 4-8 : PCondenser .Act _ Segment  i 1
NSegment _ size
i 1
PowGeneratedi
For cumulative input parameters, the entries are summed and extrapolated for a month as the
STEP program is designed for monthly inputs, e.g.:
NSegment _ size
44640
See range 3 on Figure 4-8 : RTBoil .segment  i 1
RTBoil .1 mini 
NSegment _ size
The average input parameters include temperature, pressure, the carbon in refuse (CIR), and
specific mass flow rates. The data processing is further demonstrated in Section 4.3.5.
In certain cases, data had to be up-scaled to a minute resolution i.e. coal flow rates were
discretized to a minute resolution by calibrating the mill feeder mass flow signals to the summed
monthly conveyor belt mass meter indications. Although some unquantifiable uncertainties would
be introduced as feeders degrade over time and mill configurations change within the month, this
was thought to be relatively insignificant and necessary to compare the direct and indirect
efficiency computation.
Figure 4-8 is an example of the CSV file produced from the processed signals. This file is read in as
input data to the STEP model. The data file had consisted of 112 tags per unit x 44640 temporal
entries.
27
4.3.1. Principles of sequential perturbation

The sequential perturbation method is illustrated by equations 4.6 to 4.12:
R0  f  x1 x2 ,, xL  (4.6)
R1  f  x1  ux1,x2 ,, xL  (4.7)
R1  f  x1  ux1,x2 ,, xL  (4.8)
 Ri   Ri   R0 (4.9)
 Ri   Ri   R0 (4.10)
 Ri    Ri   R 
 Ri    ux i (4.11)
2  xi  xi  xi
2
L   R  
uR     i  uxi  or uR  RSS(  Ri ) (4.12)
i 1   xi  x  x 
 i 
The above method is detailed by Fiogiola and Beasly [48]. The significance or impact of each
variable relative to the total uncertainty can be derived as follows:
2
  R  
  i  uxi 
  xi  x  x 
SignificanceUi   i   100%
2
(4.13)
L  
 Ri 
   uxi 
i 1   xi  x  x
 i 
or
uRi 2
SignificanceUi   100% (4.14)
uR2
where:
R0 -evaluation of the function with the inputs being the mean of each variable
Ri  -is the result of the function where the absolute uncertainty for each variable is added to the
mean sequentially
Ri  - is the result of the function where the absolute uncertainty for each variable is subtracted
from its mean sequentially
29
uR - Propagated combined uncertainty to the result of a function at the mean of each input
variable
ui - Propagated individual uncertainty to the result of a function at the mean of each input variable
4.3.2. Statistic & probability theory

The term error is used in various texts to describe the deviation from an individual measurement
to that of the true underlying mean of the collective sample. The error or deviations from the true
mean can be attributed to various elements. The elements are typically classified as random
(temporal) or systematic errors.
Figure 4-9 and Figure 4-10 demonstrate the difference between random and systematic errors
where the systematic error is fixed in time which is a result of some static influence such as
measurement equipment calibration. On the other hand random errors change due to a temporal
influence on repeatability. An example of where the random error could occur is human
subjectivity when a measure resolution is too low, variations in operating and environmental
conditions.
Figure 4-9: Distribution of error on repeated measurement [48]
30
Figure 4-10: Distribution of error as a function of time [17]
For repeated measurements, the calculated statistical parameters are given below [66]:
Mean of the measured value:
1 N
x  xi
N i 1
(4.15)
where:
N = number of measurements
xi = individual measurement in a set of measurements
The standard deviation for a sample:
1 N
Sx  ( xi  x )2
N  1 i 1
(4.16)
The standard deviation of means or standard error for a sample is given by:
Sx
Sx  (4.17)
N
For N >30, Sx is approximated to  x [67]. An expected or true value based on a set of

measurements is generally represented by its mean and uncertainty as a function of the
confidence level or percentage probability ( P% ). The form as per probability theory is given
below:
x'  x  ux ( P%) (4.18)
31
where:
P% = probability percentage
ux = uncertainty associated with a set of measurements from a single point of measurement.
In the case of multiple, independent and similar measurements points to acquire the same
measurand such as in the case of two thermocouples measuring the same parameter, the
combined uncertainty is given by [47]:
ux
uc  (4.19)
NMP
where:
uc - combined uncertainty for a variable
NMP - number of measurement points
The above equation is derived from the equation (4.17) where the numerator remains unchanged
for additional measurements but the denominator becomes N  NMP .
In probability theory, repeated observations or repeated measurements will generate a typical

distribution profile when the histograms are plotted. When the data is normalized in terms of Sx
and expressed in percentage of frequency, it would likely fit a standard probability density
function. The typical distributions for engineering data are normal distribution, log-normal,
rectangular, triangular, Poisson and binomial. Refer to [48] for more detailed descriptions of the
standard distributions used in engineering sciences. Figure 4-11 is the most common distribution
profile in engineering data. It is also known as Gaussian distribution. For this type of distribution,
the normal and t-probability tables indicate the probability of an observation based on the
deviation from the mean and the standard deviation of the data set.
32
assumed that all STEP inputs are non-correlated therefore the covariance had not been analysed
in this study.
The uncertainty model has been limited to the first layer or primary uncertainty propagation,
meaning that a derived variable that may have additional uncertainties from the method of
derivation is omitted in the propagation to the result. A typical example to demonstrate this point
is that, in thermodynamics, the enthalpy is looked up from steam tables using steam pressure and
temperature. The uncertainty in the steam temperature and pressure measurement may be
considered as primary uncertainties. The uncertainty associated with the development of the
steam tables can be considered as secondary uncertainty and is a function of the temperature and
pressure. From literature [68], uncertainty estimates are provided for steam tables, however, it
would require significant effort in coding and introduce additional computational expense. The
processing of the 2nd tier of calculations is of limited value to the objectives of this study. The
second tier would be a consideration for the analysis using STEP models specific to dry cooled
sliding pressure power plants as Eskom employs the direct method in the turbine performance
analysis due to the thermal performance being more sensitive to the load factor and ambient
conditions. The model used in the case study was that of a wet-cooled static pressure power plant
hence did not require any look-ups from steam tables.
The STEP program utilizes coefficients for polynomial performance curves derived from regression
fitting of the acceptance test data. Seeing that these tests were only performed once (not
repeated), there are no estimates in the uncertainty of this data. Similarly, all other parameters
derived from single plant tests have no uncertainty estimates and therefore could not be modelled
at this stage of the study.
For the current study, all parallel or series measurements used in the variable averaging are
assumed to be measuring a homogeneous stream i.e. the uncertainty pertaining to spatial non-
uniformity of the measurement is ignored. Even though in some cases stratification in flue gas
ducting could cause large deviations between points of measurement and consequentially
introduce significant uncertainty of the measured temperature, more detailed practical studies are
required to estimate the impact of these cases which may be very specific to the physical layout of
the plant. The spatial uncertainty can be easily incorporated into the uncertainty model for future
work.
4.3.4. The computer program of the STEP calculation procedure

To evaluate the temporal uncertainty at a minimum of 30-minute intervals (1488 intervals per
month) and perturb 538 of the input variables by the positive and negative uncertainties per
34
interval, would require significant computational resources, seeing as the STEP model would have
to be run ≈ 1.6 M times.
Due to the significant number of calculated variables in STEP (approx. 1700) and the operation
sequence required in the processing of the uncertainties, it had been anticipated that significant
computational processing would be required, especially in the handling of large data sets. This had
necessitated that the modelling be moved to a platform more capable for large data analytics.
Python® with the Pandas library was chosen due to availability (open source), online
support/training, and the experience and recommendation of fellow scholars, academics and
peers in the industry.
4.3.5. Model architecture

Figure 4-12 illustrates the computer program process flow sequence covering the high-level steps
in the modelling architecture. The steps illustrated in the mentioned figure are further elaborated
with illustrations. Based on the acquired data as per 4.2.1, it would be statistically inaccurate to
consider a sample size of less than or equal to 30 observations (30 x 1minute intervals) as the
sample standard deviated is no longer representative of the population standard deviation, which
is a significant parameter in the uncertainty quantification.
35
Figure 4-12: Uncertainty propagation model architecture
36
Herewith follows a brief discussion of the computer program functions.
Step 1: All variables are declared and set to a value of zero.
Step 2: The base input variables are read in from a CSV file to the input data frame in the STEP
computer program. The CSV file is generated from an MS Excel workbook where data is manually
copied in from the EtaPro database client. The data is thereafter automatically processed for input
to the developed STEP computer program so that various monthly composite datasets can be
easily imported to the model as per Figure 4-13.
Step 3: The STEP module is run and the results are exported for verification (see also Appendix G).
Figure 4-14 is a truncated view of the calculated variables. These values had been compared with
the MathCAD calculation sheet mentioned in section 4.1 to verify the model and have proven to
be extremely useful in eradicating errors such as unit conversions and code debugging e.g. divide
by zeros or empty entries and syntax errors for the STEP module of the program. Refer to
Appendix E for the description of the calculated variables.
Step 4: The uncertainty model is run with only the systematic uncertainties read in. The results are
exported to an MS Excel file (refer to Figure 4-15).
Step 5: The plant operational data on a 1-minute resolution is read into a data frame from a CSV
file.
Step 6: Each segment/chunk is processed to acquire the average or sum of the variables (refer to
Figure 4-16). The standard deviation for each segment and each variable is computed to
determine the temporal uncertainty at a 95% confidence interval. The associated uncertainty of
each input variable (rows) to the calculated variable (columns) is reported per segment run. The
table as shown in Figure 4-15 is produced for each segment run. The overall impact of each input
variable on a calculated variable is the weighted average value of all segments by the USO.
Step 7: The temporal uncertainty for each segment is combined with the systematic uncertainty as
shown in Figure 4-12.
Step 8: Recall equations (4.6) to (4.12). The δRi for each calculated variable is exported for each
segment run. The RSS of the δRi is the combined impact of all input uncertainties per calculated
variable. Each row entry in Figure 4-17 represents the results of a segment run. The overall
uncertainty of each calculated variable is the weighted average of the RSS values by the USO for all
segment runs.
Step 9: The results are fed into post-processing tools to generate visuals, see section 5.
37
Unit(Cycle) Number
Input Variable Station 1 2 3 4 5 6 Unit Step
of Code
measure
AH_LH_Gas_Out_T 0.000 113.824 113.824 129.384 129.384 125.661 125.661 °C IDB001
AH_RH_Gas_Out_T 0.000 115.843 115.843 129.468 129.468 126.529 126.529 °C IDB002
FD_LH_AIR_In_T 0.000 29.303 29.303 32.741 32.741 29.455 29.455 °C IDB004
FD_RH_AIR_In_T 0.000 31.447 31.447 30.268 30.268 31.544 31.544 °C IDB005
AH_LH_Gas_In_O2_x 0.000 3.677 3.677 3.712 3.712 3.892 3.892 % IDB007
AH_RH_Gas_In_O2_x 0.000 3.623 3.623 3.350 3.350 4.169 4.169 % IDB008
Boil_Run_hrs 0.000 744.000 744.000 627.800 627.800 727.970 727.970 h IDB010
FW_Tot_Mass 0.000 1 144.900 1 144.900 1 003.250 1 003.250 864.360 864.360 Gg IDB011
BD_Tot_Mass 0.000 0.000 0.000 0.000 0.000 0.000 0.000 t IDB012
AH_LH_Gas_Out_O2_x 0.000 5.677 5.677 5.712 5.712 5.892 5.892 % IDB013
AH_RH_Gas_Out_O2_x 0.000 5.623 5.623 5.350 5.350 6.169 6.169 % IDB014
Boil_Starts_0to7hrs 0.000 0.000 0.000 0.000 0.000 0.000 0.000 IDB022
Boil_DownTime_0to7hrs 0.000 0.000 0.000 0.000 0.000 0.000 0.000 h IDB026
HotSby_Time 0.000 0.000 0.000 0.000 0.000 0.000 0.000 h IDB030
Boil_ColdStarts 0.000 0.000 0.000 0.000 0.000 0.000 0.000 IDB031
Mill_Reject_Mass 0.000 0.000 0.000 0.000 0.000 0.000 0.000 ton IDB070
Mill_Reject_CV 0.000 15.000 15.000 15.000 15.000 15.000 15.000 MJ/kg IDB071
AH_Gas_In_CIR 0.000 3.043 3.043 3.105 3.105 3.793 3.793 % IDB076
CoarseAsh_CIR 0.000 0.761 0.761 0.777 0.777 0.947 0.947 % IDB077
MILL_A_Run_Hrs 0.000 641.830 641.830 573.330 573.330 556.070 556.070 Hrs IDB091
MILL_B_Run_Hrs 0.000 676.270 676.270 617.390 617.390 665.890 665.890 Hrs IDB092
MILL_C_Run_Hrs 0.000 289.170 289.170 442.030 442.030 357.510 357.510 Hrs IDB093
MILL_D_Run_Hrs 0.000 706.260 706.260 460.380 460.380 118.770 118.770 Hrs IDB094
MILL_E_Run_Hrs 0.000 597.400 597.400 445.860 445.860 657.570 657.570 Hrs IDB095
MILL_F_Run_Hrs 0.000 742.280 742.280 529.430 529.430 677.540 677.540 Hrs IDB096
Gen_LV_Main_J 0.000 375.157 375.157 340.861 340.861 316.510 316.510 GWh IDT012
Gen_LV_Check_J 0.000 375.157 375.157 340.861 340.861 316.510 316.510 GWh IDT013
AuxPower_Main_J 0.000 19.295 19.295 23.511 23.511 26.116 26.116 GWh IDT014
AuxPower_Check_J 0.000 19.295 19.295 23.511 23.511 26.116 26.116 GWh IDT015
Gen_AboveNER_J 0.000 0.000 0.000 0.000 0.000 0.000 0.000 GWh IDT019
Gen_AboveNER_RT 0.000 0.000 0.000 0.000 0.000 0.000 0.000 h IDT025
Gen_Reactive_J 0.000 -42.369 -42.369 -17.357 -17.357 4.281 4.281 GVARh IDT020
Gen_RT 0.000 744.000 744.000 622.060 622.060 716.040 716.040 h IDT024
AuxPower_Import_Trip_J 0.000 0.000 0.000 1 256.840 1 256.840 1 239.740 1 239.740 MWH IDT100
AuxPower_SB_J 0.000 20.470 20.470 25.820 25.820 28.680 28.680 GWh IDT089
A_EFP_RT 0.000 0.000 0.000 632.790 632.790 742.990 742.990 h IDT085
B_EFP_RT 0.000 0.000 0.000 628.420 628.420 619.870 619.870 h IDT100
C_EFP_RT 0.000 0.000 0.000 0.000 0.000 0.000 0.000 h IDT110
… 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Figure 4-13: Step 2 - Example of monthly composite input data
38
Calculated Variable Unit No. Base Result Unit

of
Measure
AH_GAS_OUT_T [0] 0.000 °C
AH_GAS_OUT_T [1] 125.610 °C
AH_GAS_OUT_T [2] 125.610 °C
AH_GAS_OUT_T [3] 129.385 °C
AH_GAS_OUT_T [4] 129.385 °C
AH_GAS_OUT_T [5] 124.055 °C
AH_GAS_OUT_T [6] 124.055 °C
FD In T [0] 0.000 °C
FD_In_T [1] 30.330 °C
FD In T [2] 30.330 °C
FD_In_T [3] 31.575 °C
FD_In_T [4] 31.575 °C
FD_In_T [5] 32.065 °C
FD_In_T [6] 32.065 °C
AH_Gas_In_O2_x [0] 3.898 %
AH_Gas_In_O2_x [1] 3.705 %
AH_Gas_In_O2_x [2] 3.705 %
AH_Gas_In_O2_x [3] 3.845 %
AH_Gas_In_O2_x [4] 3.845 %
AH_Gas_In_O2_x [5] 4.215 %
AH_Gas_In_O2_x [6] 4.215 %
Stm_Mass [0] 6025.020 Gg
Stm Mass [1] 1144.900 Gg
Stm_Mass [2] 1144.900 Gg
Stm_Mass [3] 1003.250 Gg
Stm_Mass [4] 1003.250 Gg
Stm_Mass [5] 864.360 Gg
Stm_Mass [6] 864.360 Gg
Boil_LF [0] 0.000 %
Boil_LF [1] 84.311 %
Boil_LF [2] 84.311 %
Boil_LF [3] 87.554 %
Boil_LF [4] 87.554 %
Boil_LF [5] 65.054 %
Boil_LF [6] 65.054 %
Boil_LF_CF [0] 1.003
Boil LF CF [2] 1.003
CIR_Targ_CF [0] 0.000 %
…
Figure 4-14: Step 3 - Example of base results
39
AH_GAS_OUT_T[0] AH_GAS_OUT_T[1] AH_GAS_OUT_T[2] AH_GAS_OUT_T[3] AH_GAS_OUT_T[4] AH_GAS_OUT_T[5] AH_GAS_OUT_T[6] …

Systematic 0 5.024937811 0.707106781 0.707106781 0.707106781 0.707106781 0.707106781
RSS 0 5.032899573 2.084726672 0.707106781 3.560308546 3 223841703 0.707106781
RSS 0 5.025765818 1.472842196 0.707106781 1.392738594 2.151293904 0.707106781
RSS 0 5.026524906 2.83563277 0.707106781 4.703723519 2.37324494 0.707106781
RSS 0 5.026084669 1.06587837 0.707106781 1.156305313 1 037540841 0.707106781
RSS 0 5.049999961 0.959689442 0.707106781 1.15288757 1.16643581 0.707106781
RSS 0 5.049373475 1.602211387 0.707106781 1.041553388 1 029528412 0.707106781
RSS 0 5.025789483 1.956804872 0.707106781 1.263126888 0 982108146 0.707106781
RSS 0 5.028715952 1.670217565 0.707106781 1.208252706 1.665883187 0.707106781
RSS 0 5.026666392 1.027063943 0.707106781 1.400642445 0 859778209 0.707106781
RSS 0 5.027752136 0.970189789 0.707106781 1.689729181 1.173250192 0.707106781
RSS 0 5.027576158 0.8129707 0.707106781 1.212563435 0 938796495 0.707106781
RSS 0 5.025562183 0.882086764 0.707106781 0.903267328 1.124062103 0.707106781
RSS 0 5.025635933 0.858340378 0.707106781 2.694344468 2 897349003 0.707106781
RSS 0 5.027408529 0.945148627 0.707106781 1.524432005 0 924474577 0.707106781
RSS 0 5.025796259 0.866217942 0.707106781 2.010073418 0 983406335 0.707106781
RSS 0 5.025768914 0.821017242 0.707106781 11.61855408 0.730042311 0.707106781
RSS 0 5.026349504 0.828206233 0.707106781 2.73457951 0.782894435 0.707106781
RSS 0 5.025436179 0.836239513 0.707106781 17.77619234 0 881203098 0.707106781
RSS 0 5.042984907 0.82350228 0.707106781 6.38205624 2.115300198 0.707106781
RSS 0 5.049843416 2.132214134 0.707106781 6.898385759 0 914492033 0.707106781
RSS 0 5.035788783 1.887947292 0.707106781 15.49715185 0 953875703 0.707106781
RSS 0 5.037867608 1.979184128 0.707106781 4.381363057 0 948808566 0.707106781
RSS 0 5.026077206 1.572599673 0.707106781 1.080280393 2.148516144 0.707106781
RSS 0 5.025937698 1.200461683 0.707106781 2.93480708 1.453197476 0.707106781
RSS 0 5.026834389 0.848329487 0.707106781 1.264197507 1 350386547 0.707106781
RSS 0 5.025949977 1.021542622 0.707106781 1.358757236 0.772267363 0.707106781
RSS 0 5.026712682 2.097274151 0.707106781 1.471749487 1 345109652 0.707106781
RSS 0 5.025910009 1.267933714 0.707106781 0.779716012 0 902858395 0.707106781
RSS 0 5.025864606 0.99091987 0.707106781 0.71152324 1.620624455 0.707106781
…
Figure 4-17: Step 8 - Example of combined uncertainty of calculated variables per segment run
41
Chapter 5. Results and discussion
5. Results and discussion

5.1. Model uncertainty
In the section the model uncertainty had been studied by comparing the STEP loss computations
to that of the standard performance codes listed in Table 5-1. The table below is a summary of the
results further explained in this section.
Table 5-1: Summary of loss review
Losses STEP Standard Deviation Code

% % %
Carbon in refuse 0.940 1.097 0.157 BS EN 12952-15: 2003

Carbon monoxide 0.170 0.170 BS EN 12952-15:2003
Dry flue gas 4.797 4.660 -0.182 BS EN 12952-15: 2003
Dry flue gas 4.797 4.843 0.046 BS 2885: 1974
Hydrogen & moisture 4.897 5.007 0.110 BS EN 12952-15: 2003
Make-up 2.657 2.261 -0.396 66.7% vs. 50% steam fraction
Radiation 0.568 0.289 -0.279 BS EN 12952-15: 2003
Turbine deterioration 0.850 3.264 2.414 IEC 60953-0/Ed1: 1996
Turbine deterioration 0.850 1.402 0.552 ASME PTC 6 Report -1985
5.1.1. Carbon in Refuse

From equations (A.8) and (B.42), it can be deduced that STEP calculation does not account for the
sensible heat of the ash leaving the boiler. The impact on the loss due to the difference in
methodology was demonstrated to be +0.157% (or +0.029 MJ/kg coal-fired) for the case evaluated
as per calculations (C.4) and (C.10) in Appendix C. A sensitivity study was further conducted by
changing the fly ash in flue gas efficiency to 90% and 85% which yielded a difference in thermal
loss of 0.246% and 0.335% respectively which is significant in the indirect determination of the
boiler efficiency.
5.1.2. Dry Flue Gas loss

Both the BS 2885 [5] and BS EN 12952-15 [15] do not give any means to calculate the combustion
products in flue gas as it would be measured during an acceptance test or if the flue stacks are
fitted with gas analysers. The STEP methodology attempts to approximate the gas composition
from an oxygen balance calculation and assuming that the gas products will comprise only of N2,
42
CO2 and O2. Typical gas analysers measure these gases on a dry volumetric basis which suits the
calculation of the Dry Flue Gas loss. It is observed that the heat capacity of the flue gas ( cP .FG ) is
static as per STEP calculations and the BS 2885 [5]. The approach is different as per the BS EN
12952 -15 as the cP .FG is calculated based on the CO2 mass fraction and temperature of the flue
gas.
For the case demonstrated, the STEP calculation and the BS 2885 [5] yielded very similar results as
the same gas composition was used in the calculation. The loss as per STEP was calculated to be
4.797% or 0.901 MJ/kg coal (refer to Appendix C.4) versus 4.843% or 0.909 MJ/kg coal using the BS
2885 approach (refer to Appendix C.5). The difference between the two methods is that STEP
calculates the mass ratio of dry flue gas to a kg of coal-fired and BS 2885 [5] leaves it on a molar
basis. They both calculate the relative gas flows based on the CO2 concentration in the flue gas.
The principle is based on 1 mol C → 1 mol CO2. The method employed in STEP to calculate the CO2
concentration in the flue gas is not referenced but appears to be based on some empirical
relationships of the coal composition (some parameters being on a proximate analysis base).
The BS EN 12952-15 [15] calculates the mass ratios of gas products through stoichiometric
reactions of the elements within the coal, hence the ultimate analysis of the constituents is
required. The excess air is determined by the oxygen depletion method (see Appendix C.6 –
calculation (C.31)). In this case, the result is marginally different yielding a loss of 4.660% or 0.875
MJ/kg Coal (see Appendix C.6 - calculations (C.42) and (C.43)).
5.1.3. Carbon Monoxide (CO) loss

Carbon monoxide loss is the loss of energy release due to incomplete combustion due to
insufficient mixing or residence time in the combustion chamber i.e. the formation of CO instead
of the full reaction of carbon with oxygen to produce CO2. This loss is omitted from the STEP
calculation possibly due to low levels normally generated in the fleet. The Fossil Fuel Firing
Regulations (FFFR) [69] ensures a sufficiently adequate oxidising environment to prevent the
significant formation of CO. The threshold for CO at the economiser outlet is 200ppm (V/V). In the
case evaluated, with a CO concentration of 50ppm (V/V) the loss yielded is only 0.043% and
reaches 0.426 % (or 0.08 MJ/kg coal-fired) at 500ppm (see Appendix C.3 for the methodology [15]
and assumptions).
5.1.4. Make-up Water loss

The make-up water loss is used to quantify the energy lost due to the efflux of water from the
Rankine cycle through losses such as leakage from passing valves, boiler blowdowns, soot blowing
and venting/draining. There is limited information on these losses in literature as the performance
43
test codes assume that these defects are resolved before acceptance testing. These flows are
typically not measured at a power plant. The make-up water loss is one of the highest losses
experienced in South African power plants as per the STEP program and hence had been identified
for critical review.
The calculation methodology is demonstrated by calculations in Appendix C.9. The methodology
was deemed to be fundamentally satisfactory. However, STEP makes approximations of these
flows by assuming that a 3rd of the make-up water is lost at the boiler feed water conditions and
2/3rd is lost at main steam conditions (see Appendix C.9 calculation (C.54)). The approximation is
not referenced in the STEP manuals [34][42] nor could be found in the literature. For the case
evaluated the make-up loss was determined to be 2.657%. See Table 5-2 for the results of the
sensitivity study.
Table 5-2: Sensitivity analysis on the fraction of make-up water lost at main steam conditions
The fraction of make- Thermal Loss

The fraction of make-
up lost at FW (%)
up lost at MS
conditions
conditions
0.90 1.317
0.10
0.80 1.552
0.20
0.67 1.866
0.33
0.50 2.261
0.50
0.33 2.657
0.67
0.20 2.976
0.80
From the above table is it quite obvious that the fraction of make-up water lost at main steam
(MS) conditions has a significant impact on the thermal loss determination. The spread or range
would become even wider when conditions of high make-up are experienced.
From calculation (C.59) in Appendix C.9, It can be seen that demineralized water flows that are
external to the Rankine cycle are included in the hot make-up volume. It becomes difficult to
standardize these calculations done in STEP as the various plant metering configurations are
different. It is incorrect to include the water losses in the thermal loss calculations e.g. the water
supplied to the hydrogen plant for electrolysis or cooling water to the compressed air systems
cannot be considered to be lost at main steam condition neither at feed water conditions.
Another major risk to the accuracy of the make-up determination is the method to distinguish
between hot make-up and cold make-up flow especially if this is done manually. Manual methods
44
of recording the integrated volume of make-up may not be synchronized to the unit
synchronization and uncoupling from the grid. Hot make-up is the demineralized water top-up to
the Rankine cycle when the unit is synchronized to the electricity grid. The make-up used during
commissioning should also be discounted as the input energy to the cycle (coal & oil-fired) during
commissioning is discounted in the calculation of the running plant heat rate. Make-up flow during
cold conditions (unit offload and cooled) has negligible thermal energy value. In many cases where
power plants are equipped with modern data systems, the quantification of the make-up volume
can be determined through the integration of the instantaneous flow meter signals and can be
filtered directly by the plant data systems. In this case calculation (C.59) in Appendix C.9 may be
circumvented to a direct input.
5.1.5. Boiler Radiation loss

Generally, during acceptance testing of a boiler, a full-scale radiation test is omitted due to the
vast number of measurement points required across the boiler skin casing and piping, it is
preferred to agree on a figure based on a similar design implemented elsewhere. Both ASME [17]
and BSI [5] recommend sampling surface temperatures and ambient temperatures measurement
at various points and applying related heat transfer coefficient to calculate heat loss per area. The
STEP calculation methodology uses the former principles for its loss calculations, where previous
test results for similar designs are available in the form of manufacturer’s curves. The boiler
radiation loss is estimated by BSI [15] as a function of the boiler output in MWs. It is assumed that
the ash hopper losses are inclusive as the revision does not specifically mention. The losses
calculated in STEP yielded a value of 0.568 % versus a value of 0.289% calculated as per the
method prescribed by BSI (see Appendix C.10 and C.11 respectively).
5.1.6. Turbine Deterioration loss

The turbine deterioration in STEP is 0.1% abs/year which is applied monthly by multiplying by an
availability factor i.e. the percentage of time on-load. The IEC’s correlation [16] yields a much
higher deterioration as demonstrated in Appendix C.13 calculations (C.68) - (C.70). Whilst ASME’s
PTC 6-1996 [32] does not recommend corrections for ageing when carrying out the acceptance
test, however, the related report [47] does provide a method. The deterioration had been
calculated over ten years. The calculations as per STEP yielded a deterioration of 1.0 % compared
to the IEC’s method yielding a deterioration of 1.76%. It should also be noted that the quoted loss
constants of 0.1% abs/month and 0.06% abs/month are applicable for the first and second year of
operation relatively. The BCEGB formerly stated 0.117% abs/month for the first year, 0.0417%
abs/month for the 2nd and 0.0083% abs/month for the balance of years [34]. In Figure 5-1: Turbine
deterioration loss comparison below, the dashed lines are linearized extrapolations for the ASME
45
5.1.8. Unitized cycle efficiency - direct method

The STEP model apportions the coal flow to each unit based on its fraction of total electrical units
generated applied to the station’s total coal consumption. The methodology adopted is due to
very high uncertainty associated with mill feeder calibration repeatability, drift as a function of
time (wear on spiral feeders and ploughs on table feeders) and inconsistency of the coal density,
hence the lack of cycle efficiency calculation at a unit level by the “direct method”.
5.2. Instrument uncertainty

The instrument uncertainty of each of 545 input variables had been propagated to each of the
1785 calculated variables via the Python computer program of the STEP calculation method. Refer
to Appendix D for the details of each input variable uncertainty assumption. See Table F- 1 in
Appendix F for the full results of the instrument uncertainty propagation. It is recommended that
this section be read with the review of terminologies explained in sections 4.1 and 1.4.
47
Table 5-3: Truncated results greater 5% – instrument uncertainty propagation
Calculated Variable Unit Level Station Level

Base Result u (abs) u (%) Base Result u (abs) u (%)
MS_L_Cost 14 767.224 30 984.097 209.817 150 304.440 68 109.383 45.314
MS_SF_T_L 0.038 0.079 209.716 0.070 0.032 45.312
HPT_Diff_In_T -1.405 2.800 199.322
IPT_Diff_In_T -4.102 2.800 68.253
ReH_Stm_L_Cost 46 573.048 31 505.089 67.647 286 757.185 69 619.781 24.278
ReH_SF_T_L 0.119 0.080 67.504 0.134 0.033 24.275
BP_L_Cost 674 718.199 419 453.943 62.167 5 593 366.533 927 751.450 16.587
BP_SF_L 1.720 1.068 62.113 2.614 0.433 16.582
CW_Rise_T 14.170 7.920 55.891
MU_SF_L -0.339 0.087 25.755 0.640 0.063 9.782
MU_L_Cost -132 936.388 34 233.172 25.752 1 369 881.210 134 120.223 9.791
Radiation_L 0.005 0.001 25.397 0.006 0.000 7.584
AuxPower_L_Cost -325 313.576 78 290.221 24.066 -670 246.404 133 849.399 19.970
BP_Targ 5.416 1.270 23.441 5.322 0.514 9.651
AuxPower_SF_L -0.829 0.192 23.142 -0.313 0.063 19.966
DFG_L_Cost 224 778.035 49 812.081 22.161 2 169 978.309 116 838.251 5.384
DFG_SF_L 0.573 0.125 21.825 1.014 0.054 5.369
MU_Act_L 0.516 0.107 20.685 1.780 0.076 4.283
MU_Hot_F 0.516 0.107 20.684 0.000 0.000 0.000
FW_L_Cost 1 490 785.753 302 380.180 20.283 14 302 061.830 743 058.851 5.195
Radiation_L_Coeff 0.408 0.082 20.100
AH_LH_Leak 13.043 2.609 20.006 13.070 1.072 8.200
AH_RH_Leak 12.999 2.600 20.000 13.018 1.067 8.193
Boil_L_Cost 322 772.480 63 325.683 19.619 3 068 035.665 196 518.510 6.405
FW_SF_L 3.800 0.732 19.257 6.685 0.346 5.179
Boil_SF_L 0.823 0.154 18.728 1.434 0.092 6.388
Cycle_L_Cost 2 910 659.322 539 693.164 18.542 28 459 290.381 1 227 700.769 4.314
Cycle_SF_L 7.420 1.326 17.877 13.302 0.571 4.294
Turb_L_Cost 3 046 136.806 530 535.547 17.417 24 691 619.910 1 191 696.155 4.826
Turb_SF_HR 718.825 124.034 17.255
Turb_SF_L 7.765 1.297 16.708 11.541 0.555 4.809
L_Corr_H2andMoist 1.384 0.229 16.575 0.000 0.000 0.000
CIR_L_Cost 199 713.018 28 531.468 14.286 1 381 066.295 142 917.428 10.348
L_Corr_Ash 0.361 0.049 13.510
FW_Diff_T -44.326 5.661 12.772
CIR_SF_L 0.509 0.065 12.724 0.645 0.067 10.334
Cond_DepresT 38.075 3.960 10.400
MR_SF_L 0.000 0.000 7.370
MR_Act_L 0.000 0.000 7.370 0.000 0.000 3.180
48
AH_Out_CO2_Coeff 1.942 0.132 6.779

H2_Moist_L_Cost -101 718.572 5 807.997 5.710 -483 008.940 12 108.292 2.507
AuxPower_Act_Frac 3.806 0.214 5.635 8.310 0.072 0.871
FD_In_T 30.374 1.617 5.322
Boil_LF 81.869 4.175 5.099
H2_Moist_SF_L -0.259 0.013 5.055 -0.226 0.006 2.473
Stm_Mass 1 113.935 55.697 5.000 6 019.247 123.377 2.050
The Table 5-3 is a truncated form of the stated table where the variables with uncertainties
greater than 5% are shown at a unit level. The results above are valuable to understand the
uncertainty of each calculated variable but it must be emphasized that even though the
uncertainty of the calculated variable may be extremely high, it may be relatively insignificant to
the final result of STEP i.e. cycle efficiency, turbine efficiency, boiler efficiency and associated STEP
Factors (% of targets achieved). The above information, however, may be vitally important for
specific business case sensitivity analysis, where an individual parameter may be of high
importance e.g. the cost of feed water temperature loss may be a significant input parameter in a
feasibility study to replace the feedwater heaters.
It is observed that the parameter uncertainty at a station level is lower than that of a unit level by
the typical factor, Nunits , where the station parameters are a mere composite of unit parameters.
This is not the case where the unit parameters are influenced by some common input parameters
e.g. coal quality parameters, common plant auxiliary power consumption, common plant
demineralized water consumption.
From the results above is already noticeable that some significant contributors to the uncertainty
of the listed variables are temperature measurement (thermocouples), make-up water flow
(orifice plates), condenser pressure transducer, coal quality parameters (total moisture) and
electrical metering. The method of flagging the items in the table by the percentage uncertainty is
very sensitive to the base run conditions, hence a more holistic and in-depth analysis was
performed on the parameters of specific interest with consideration for the absolute uncertainty
with a drill-down approach.
It is interesting to observe from Table 5-4 (highlighted in blue), there is quite a large difference
between the direct and indirect STEP Factor uncertainty i.e. 0.571% vs 2.210%. It is vastly accepted
that the indirect method should yield lower uncertainties due to the fact it is not directly
dependent on the coal flow rate measurement, which is prone to significant measurement noise.
49
Table 5-4: Truncated results drill-down approach – instrument uncertainty propagation

Cycle_Act_Eff 0.000 0.000 0.000 31.911 0.756 2.370
Cycle_Indirect_SF 73.479 1.683 2.291 86.698 0.571 0.659
Cycle_Direct_SF 0.000 0.000 0.000 90.487 2.210 2.443
Cycle_Unacc_L 0.000 0.000 0.000 -3.788 2.230 58.878
Cycle_Acc_L 0.000 0.000 0.000 13.302 0.571 4.294
Cycle_Act_HR 0.000 0.000 0.000 3.134 0.074 2.369
Cycle_HR_Targ 2.872 0.034 1.179 2.836 0.015 0.516
Cycle_SF_L 26.521 1.683 6.346 13.302 0.571 4.294
Cycle_Unacc_L_Cost 0.000 0.000 0.000 -8 105 289.618 4 767 717.845 58.822
Incr_NLtoNER_Act_HR 3.313 0.051 1.532 2.970 0.015 0.520
AuxPower_CF 1.008 0.001 0.067 1.004 0.000 0.018
AuxPower_SF_L 1.675 0.332 19.801 -0.313 0.063 19.966
AuxPower_Targ_Frac 10.001 0.086 0.863 7.879 0.027 0.340
Boil_Act_Indirect_Eff 88.201 0.281 0.318 88.735 0.246 0.277
Est_MU_F 1.486 0.032 2.187 0.000 0.000 0.000
MU_Targ_L 1.192 0.026 2.208 0.996 0.008 0.811
MU_Targ_CF 1.003 0.000 0.021 0.000 0.000 0.000
MU_Hot_F 3.694 0.284 7.693 0.000 0.000 0.000
MU_Act_L 3.695 0.284 7.695 1.780 0.076 4.283
MU_SF_L 2.045 0.233 11.371 0.640 0.063 9.782
Turb_Est_Det_L 2.046 0.000 0.004 2.037 0.000 0.014
Boil_Act_L 11.799 0.281 2.380 11.265 0.246 2.183
Boil_Eff 0.000 0.000 0.000 88.735 0.246 0.277
Boil_SF_L 2.117 0.175 8.252 1.434 0.092 6.388
MS_SF_T_L 0.047 0.079 167.343 0.070 0.032 45.312
ReH_SF_T_L 0.138 0.079 57.333 0.134 0.033 24.275
BP_Targ 4.875 1.170 23.995 5.322 0.514 9.651
BP_Act_CF 103.643 0.148 0.143 0.000 0.000 0.000
BP_Targ_CF 0.996 0.011 1.122 0.000 0.000 0.000
BP_SF_L 4.048 1.127 27.836 2.614 0.433 16.582
Turb_SF_L 20.684 1.615 7.809 11.541 0.555 4.809
FW_SF_L 14.405 1.164 8.078 6.685 0.346 5.179
FW_Targ_T 233.103 0.833 0.357 238.775 0.338 0.142
FW_Diff_T -97.852 5.662 5.786 0.000 0.000 0.000
CIR_Act_L 1.880 0.079 4.190 1.615 0.066 4.086
CIR_SF_L 0.980 0.089 9.125 0.645 0.067 10.334
DFG_Act_L 5.223 0.131 2.506 4.971 0.056 1.132
DFG_SF_L 1.342 0.130 9.721 1.014 0.054 5.369
Gen_HV_LF 73.910 1.280 1.732 81.166 0.616 0.759
50
H2_Moist_Act_L 4.686 0.230 4.909 4.672 0.229 4.902

H2_Moist_SF_L -0.204 0.016 7.659 -0.226 0.006 2.473
MR_SF_L 0.000 0.000 7.495 0.000 0.000 0.000
Radiation_L 0.011 0.002 15.740 0.006 0.000 7.584
A negative unaccounted cycle loss as indicated in the table above (In blue text) would indicate
either that the direct-method cycle efficiency is overstated or that the losses are overstated, or
both. The unaccounted SF loss (3.788%) is the difference between the direct STEP Factor and the
indirect STEP Factor, and is used as an indication of the uncertainty of either of the stated metrics.
If one combines the systematic uncertainty of both parameters (0.571% + 2.210%), it could only
justify a maximum deviation of 2.781%, which would be only 73 % of the unaccounted loss, hence
it can be deduced that other significant uncertainties affect these computations, which will be re-
examined in Section 5.5 of this document.
From Figure 5-2, the backpressure loss and final feed water loss have the highest instrument
uncertainty for the case analysed. The other loss uncertainties are significant but relatively much
lower.
1.400 0.500
Uncertainty (%-abs)
Uncertainty (%-abs)
1.200 0.450
0.400
1.000 0.350
0.800 0.300
0.250
0.600 0.200
0.400 0.150
0.100
0.200 0.050
0.000 0.000
BP_SF_L
MR_SF_L
MU_SF_L
DFG_SF_L
Radiation_L
FW_SF_L
AuxPower_SF_L
CIR_SF_L
H2_Moist_SF_L
Turb_Est_Det_L
MS_SF_T_L
ReH_SF_T_L
AuxPower_SF_L
MR_SF_L
BP_SF_L
DFG_SF_L
Radiation_L
MU_SF_L
H2_Moist_SF_L
FW_SF_L
ReH_SF_T_L
CIR_SF_L
Turb_Est_Det_L
MS_SF_T_L
SF Losses - Unit 5 SF Losses - Station
Figure 5-2: Instrument uncertainty on SF losses
51
The aim was to find the segment size that would yield the highest number of stable load segments
and lowest number of transition segments by a statistical method. Naturally one would expect this
to be achieved by discretizing to the highest possible resolution at which the data is available,
however, the temporal uncertainty within each segment had to be quantified with the reduction
in the sample data size i.e. the segment size. The minimum sample size of 30 minutes or 30
observations was set to conform to the principles of statistics, where the sample standard
deviation can be approximated to the population standard deviation as per equation (4.17). It is
important to note that, it is a requirement that the load remains stable for 4 hours to deem an
acceptance test to be successful, hence the baseline performance data is very specific to the unit
loads rather than the average of transient conditions [14] and therefore should only be applied to
actual stable loads rather than averages.
Each loss has a significant parameter e.g. condenser backpressure loss is a function of the LP
exhaust pressure, final feedwater temperature loss is a function of the economizer inlet
temperature, etc. These significant parameters also have some degree of variability relative to
time or resolution. If one examines Figure A- 1 in Appendix A.2.2, it can be deduced that large
deviations in either the load or the significant parameter (in this case LP-turbine condenser
pressure) from the design point will result in a significant change in the heat rate which is used to
determine the losses, especially when the deviated points are in the exponential zones of the
curves.
Figure 5-9 compares the STEP Factors when run at various resolutions. As expected, the direct
STEP Factor appears to be fairly static due to the simple calculation of the actual direct efficiency
(see Appendix B equations (B.255) & (B.256)) which is not affected by plant process parameters.
The most significant parameter: coal CV, is weighted by the coal mass within and across all
segments. The coal mass and USO are summative parameters hence does not vary with the time
resolution analysis. The target efficiency is derived through the application of correction factors,
but these are normally small adjustments to the heat rate to compensate for external conditions
and not as significant as plant deficiencies.
57
minimal movement of the mean for the direct SF. The indirect SF accrued an uncertainty of
0.571% abs when only considering the systematic uncertainty propagation and a combined
uncertainty of 0.617% abs at a 3-hour resolution. Here the findings suggest that the temporal
uncertainty contribution is significant and the movement of the mean is also significant when
compared to a monthly composite run. The above phenomenon is also expected as previously
explained. When zooming into the uncertainty profile as per Figure 5-10, it would suggest that
there is a limit to the resolution at which STEP should be run, as the temporal uncertainty sharply
increases from a time resolution of 3-hour. The phenomenon is due to the reduction of the sample
Sx
size for each segment which makes the temporal uncertainty higher, recall Sx  equation
N
(4.17).
One of the objectives of STEP is to accurately quantify the losses at a unit level so that
maintenance work or defect resolution can be prioritized especially in a capital-constrained
environment. Figure 5-11 reveals that the final feed water temperature loss could be perceived to
be 2.378% when evaluated at a month resolution, however, increases to 3.188 % when evaluated
by a 3-hour segment run which is significant. The turbine back-pressure loss moved from 3.511%
to 3.164%. The auxiliary power consumption is baselined with the boiler steam feed pumps (BFPTs
– pumps driven by a steam turbine) in service. When the BFPTs are unavailable the energy
consumed for the operations of the electric motor driven boiler feed pumps (EFPs) are estimated
by the power vs load curve as indicated in equations (B.106) and (B.107) in Appendix B. The target
auxiliary power is adjusted by the estimated power consumption, which in principle gives a zero
nett effect on the STEP Factor loss. For the data used in this study, it is not the case, which
indicates that there is a possible deviation with the power vs load curve (i.e. it overestimates the
power consumption of the EFPs).
Unit 5 results indicate that the unit suffers from the same deficiencies as Unit 4 however the
magnitude is much higher. The weighted average final feed water temperature deviation from the
design is 99 oC, which is typical of operating the unit without both banks of HP heaters. The
operating condition translates to a STEP Factor loss of approximately 14.8% as indicated in Figure
5-12. The uncertainty in this value is expected to be extremely high as the operating point will fall
in the extrapolated zone of Figure 5-4 with the largest distance from actual performance data.
59
Chapter 6. Conclusions and recommendations
6. Conclusions and summary of recommendations

6.1. Conclusion
The program used in Eskom for thermal efficiency and loss quantification was reviewed with a
focus on 4 facets pertaining to the accuracy of the system employed:
 model uncertainty (the method used in quantifying the losses),

 instrument uncertainty (systematic uncertainty),
 temporal uncertainty (as a function of input resolution),
 baseline performance uncertainty (heat rate correction curves)
A semi-quantitative/qualitative approach was used to appreciate the potential uncertainty of the

STEP model by studying the difference between performance standards and STEP. A quantitative
assessment was carried out for the case studied, however, it was not intended to establish which
standard is superior and should be used as a generalized ultimate reference for the expression of
model uncertainties. The models produced during the study, provide a means of quantifying the
accuracy of STEP relative to a specific case, and the expected parameter uncertainties. From the
analysis of the case studied, it was deduced that the systematic uncertainty is much more
dominant than the temporal uncertainty, and has highlighted key parameters/measurements that
should be maintained or enhanced to reduce the overall impact on the STEP results. Whist the
temporal uncertainty may be of a lower magnitude, it may have a significant impact on the losses
at a unit level, this could be crucial for investment consideration. The study has yielded that the
optimal period (segment size) that STEP should be run to negate the impact of the temporal
uncertainty is three hours. The input data processing methodology can be considered as a pilot
study, however, can be improved further with the implementation on existing data systems. The
baseline performance data (heat rate correction curves) was compared to that produced via
commercial modelling software. The review had led to the conception that significant
uncertainties are introduced with erroneous baseline performance data. The recommendations
below, speak to the conclusions made above and the findings made in Section 5.
6.2. Recommendations
6.2.1. Model Uncertainty

STEP intends to apportion cycle heat loss as accurately to the lowest possible level. The following
recommendations will drive lower unaccounted losses calculated by STEP:
63
 The calculation of the sensible heat loss in ash that is emitted by the boiler is to be
accounted for in the STEP calculations.
 The hydrogen and fuel moisture losses should be updated as per BS 2885-1974 [5] or the
sample the calculation performed in Appendix C.7.
 The losses associated with partial combustion or CO production are to be expressed in
STEP.
 The demineralized water loss appropriation assumption is to be reviewed through a
practical study (the assumption that a 3rd of make-up water is lost is at feed water
condition and the balance at the main steam condition). A practical study using a
calibrated instrument is to be carried out. The existing water balance model should be
configured in the future to estimate this ratio based on plant data, which then could
become a volatile input to STEP. In the case of input errors then default values being based
on some proven principle from baseline tests be used.
 All empirical correlations derived from losses as a function of load factor to be re-based
lined over time. STEP baseline testing should include sensitivities at extreme operating
condition so that the polynomial regression fitting is done with a data scatter over the
entire operating range, e.g. heat rate test and modelling should be done with both banks
of HP heaters offload or with the LP turbines operated at maximum allowable
backpressure as per the operating philosophy.
 Throttle losses are included in overall turbine losses as a function of load. Deviations from
expected to be identified and quantified in STEP. The mentioned loss could be beneficial in
identifying issues with final steam pressure control such as defects with emergency shut-
off valves, governor valves and strainers.
 The direct unitized cycle efficiency calculation should be performed for power stations
fitted with new technology coal feeders such as the gravimetric belt feeders. The
calculation could be carried through to stations with spiral and table feeders for trending
purposes only (identify feeder calibration issues). The uncertainty in the latter would be
too high in the case of volumetric coal flows as the density of coal can vary significantly
with mineral composition, moisture content and size grading.
 Coal quality is considered for Carbon in Refuse losses and Hydrogen and Moisture losses,
but, the ash impact on Dry Flue Gas loss (radiation impact) and Auxiliary Power loss are to
be further studied and incorporated into STEP.
 The unitized Auxiliary Power STEP Factor losses are erroneous, based on the treatment of
off-load auxiliary power that is equally shared by all units. The unit that supplies the loop
64
supply (interconnected auxiliary power reticulation between units and common plant) is
not credited for its contributions to the common auxiliaries and other units.
 The auxiliary power STEP Factor loss is baselined with the boiler steam feed pumps (BFPTs
– pumps driven by a steam turbine) in service, when the BFPTs are unavailable the energy
consumed for the operations of the electric motor driven boiler feed pumps (EFPs) are
estimated by the power vs load curve as indicated in equations (B.106) and (B.107) in
Appendix B. The target auxiliary power is adjusted by the estimated power consumption,
which in principle should give a zero nett effect on the STEP Factor loss. STEP does not
account for the cycle heat rate gains by running the EFPs instead of the BFPTs. The above
philosophy could have adapted due to the number of tests that would have to be done to
cover all possible configurations. It is therefore recommended that validated
thermodynamic models be used to generate the load sensitivity correlations as done in
software such as Virtual Plant or Steam Pro and be expressed as a separate loss. The target
auxiliary power should be adjusted by the measured actual EFPs electric energy
consumption rather than using a correlation to determine.
 A study to determine the turbine deterioration factors should be initiated to consider the
local history of turbine cylinder efficiency test or enthalpy drop test for specific turbine
designs and sizes.
6.2.2. Input data

In some cases, totalized values such as mass flows were averaged and multiplied by the duration
of a period. This practice should be avoided as there can be a considerable difference when
compared with an integrated value at the lowest possible resolution (dt).
A system should be designed/configured and proved to process data to enable the running of STEP
at a higher resolution. Section 4.2.3 can be considered a pilot exercise in determining the
methodology. Future work could entail the automation and integration with existing plant data
systems.
6.2.3. Baseline performance data

Polynomial correlated data based on an acceptance test to be refitted with credible regression
fitting tools that can be extrapolated to worse case operational data. Once-off repeated testing
should be done to establish the typical uncertainty of the acceptance test data. In the world of
power generation, this would be very difficult to justify based on both resource and plant
availability costs. A similar desktop study as discussed in section 5.5 should be performed for all
STEP applications within Eskom and the industry.
65
6.2.4. Instrument uncertainty

The study intended to determine the typical uncertainty of performance metrics in the South
African power industry. Whilst it has been proven that it is very dependent on a specific case, a
tool has been developed to compute the uncertainty of these performance metrics for any power
plant that employs STEP, for a specific condition. Instrument uncertainty can be reduced by
increasing the number of transmitters, increasing the calibration frequency and choosing
measurement devices of a higher measurement class.
Key parameters for improvement include the:
 cooling water inlet temperature,

 economizer inlet feed water temperature,
 coal CV,
 coal mass,
 units generated,
 coal moisture,
 mass of feed water supplied to the boiler,
 air heater gas outlet temperature,
 air heater gas outlet oxygen content,
 ash content of coal, and
 volatile matter content of coal.
6.2.5. Temporal uncertainty

The optimal resolution is 3 hours as there could be a significant change of the resultant means,
particularly for the losses at a unit level, with minimal increase in the temporal uncertainty.
The STEP input or segment resolution should be kept a minimum of 3-hours, even though 6 hours
would be adequate the segmenting windows would have to be aligned to separate the peaks for
off-peak data. The 3-hour segment size is the size up to overcome alignment issues and also the
next size down from the 1-hour segment size, where an increased temporal uncertainty at this
resolution is observed.
66
6.2.6. Future work

Coal quality parameters were limited to daily composites as per the station laboratory analysis.
During the study, some work had commenced at other sites to correlate online coal scanner data
to real-time burn conditions. Each power plant has a specific coal flow and storage configuration
based on conveyor belt layout, measurement points and storage capacities. A time duration
sensitivity/uncertainty study should be carried out to determine the duration of coal movement
from the measurement point (location of the coal scanner) to the boiler, taking into account belt
loading, silo storage, bunker storage, mill configuration and confirmation of the assumption of
plug flow (no mixing). Should an uncertainty in the time lag be considerably low translating into a
relatively low temporal parameter uncertainty, the data can then be processed through the
existing models developed, however, configured for the specific power plant.
The water balance models developed should be integrated or linked to the STEP program to
ensure the improvement of the method by reasonably appropriating the losses based on the
enthalpy of the water at its point of release.
STEP intends to provide a thermal performance effectiveness assessment. Exergy efficiency

analysis is deemed to provide a more realistic measure of effectiveness; it would be an interesting
study to include exergy computations to the STEP baseline/ acceptance test data and produce
exergy STEP factors for comparison with the current method.
The spatial uncertainties associated with the location of the measurement points and sampling
methods were omitted from this study as practical case-specific studies would be required to
determine these uncertainties. It would be interesting to carry out a study that investigates the
temperature, oxygen content, and unburned carbon in ash by traversing the flue gas ducting at
various points.
67
Chapter 7. List of References
7. References
[1] Eskom, “Company Info Overview.” [Online]. Available:
http://www.eskom.co.za/OurCompany/CompanyInformation/Pages/Company_Information
.aspx. [Accessed: 18-Dec-2019].
[2] Department of Energy South Africa, “Integrated Resources Plan for South Africa (2018),”
2018.
[3] Central Intelligence Agency, “Country Comparision : Electricity Production,” The World Fact
Book. [Online]. Available: https://www.cia.gov/library/publications/the-world-
factbook/fields/252rank.html. [Accessed: 28-Dec-2019].
[4] Eskom, “Electricity technologies.” [Online]. Available:
http://www.eskom.co.za/AboutElectricity/ElectricityTechnologies/Pages/Electricity_Techno
logies.aspx. [Accessed: 18-Dec-2019].
[5] British Standards Institution, “Acceptance tests on stationary steam generators of the
power station type,” BS 2885, London, 1974.
[6] International Electrotechnical Commission, “Rules for steam turbine thermal acceptance
tests,” IEC 953-1, Frankfurt, 1990.
[7] International Energy Agency, “World Energy Outlook,” Paris, 2015.
[8] The World Bank, “Pricing carbon,” 2019. [Online]. Available:
https://www.worldbank.org/en/programs/pricing-carbon. [Accessed: 25-Jun-2019].
[9] Wikipedia Commons, “Typical coal thermal power station,” Wikipedia. [Online]. Available:
https://en.wikipedia.org/wiki/Thermal_power_station. [Accessed: 02-Jun-2018].
[10] A. Harvey, A. Larson, and S. Patel, “History of Power: The Evolution of the Electric
Generation Industry,” in Connected Plant Conference, 2017, no. September 1882, pp. 1–16.
[11] “Efficiency of conventional thermal electricity and heat production in Europe,” European
Environment Agency, 2018. [Online]. Available: https://www.eea.europa.eu/data-and-
maps/indicators/efficiency-of-conventional-thermal-electricity-generation-4/assessment-2.
[Accessed: 24-Jun-2019].
[12] “Global outlook on electricity generation by energy source 2015-2050,” Statista, 2019.
[Online]. Available: https://www.statista.com/statistics/238610/projected-world-electricity-
generation-by-energy-source. [Accessed: 21-Jun-2019].
[13] Y. Maharaj, “South African overall efficiency vs availability,” Eskom. 2019.
[14] The American Society of Mechanical Engineers, “History of ASME Standards.” [Online].
Available: https://www.asme.org/codes-standards/about-standards/history-of-asme-
standards.
[15] British Standards Institution, “Water-tube boilers an auxiliary installation -Acceptance
tests,” BS EN 12952-15, London, 2003.
[16] International Electrotechnical Commission, “Rules for steam turbine thermal acceptance
tests Part 0: Wide range of accuracy for various types,” IEC 60953-0/Ed1, Frankfurt, 2018.
68
[17] The American Society of Mechanical Engineers, “Fired Steam Generators,” ASME PTC 4,
New York, 2008.
[18] The American Society of Mechanical Engineers, Overall Plant Performance. New York: ASME
PTC 46, 2013.
[19] The American Society of Mechanical Engineers, “Procedures for Routine Performance Test
of Steam Turbines,” ASME PTC 6S, New York, 1988.
[20] A. Rashad and A. A. Elmaihy, “Energy and Exergy Analysis of a Steam Power Plant in Egypt,”
in Aerospace Sciences and Aviation Technology, vol. 9, no. 4, p. 399.
[21] I. H. Aljundi, “Energy and exergy analysis of a steam power plant in Jordan,” Appl. Therm.
Eng., vol. 29, no. 2–3, pp. 324–328, 2009.
[22] R. Mahamud, M. M. K. Khan, M. G. Rasul, and M. G. Leinster, “Exergy Analysis and Efficiency
Improvement of a Coal Fired Thermal Power Plant in Queensland,” 2016.
[23] P. Rousseau and W. Fuls, Systems Analysis. Department of Mechanical Engineering,
University of Cape Town, 2018.
[24] L. S. Paraschiv, A. Serban, and S. Paraschiv, “Calculation of combustion air required for
burning solid fuels (coal / biomass / solid waste) and analysis of flue gas composition,”
Energy Reports, vol. 6, no. September, pp. 36–45, 2020.
[25] A. M. Lahijani and F. Kalantari, “a Review of Indirect Method for Measuring Thermal,” vol.
34, no. 1, pp. 1825–1832, 2018.
[26] A. Chirag, P. N. Mehta, and P. J. Dabhi, “Research paper on Analysis of Boiler losses to
improve Unit heat rate of coal fired thermal power plant,” Int. J. Adv. Eng. Res. Dev., vol. 1,
no. 05, pp. 1–6, 2014.
[27] R. Laubscher and P. Rousseau, “Numerical investigation on the impact of variable particle
radiation properties on the heat transfer in high ash pulverized coal boiler through co-
simulation,” Energy, vol. 195, p. 117006, 2020.
[28] A. Geete and A. I. Khandwawala, “Thermodynamic analysis of 120 MW thermal power plant
with combined effect of constant inlet pressure (124.61 bar) and different inlet
temperatures,” Case Stud. Therm. Eng., vol. 1, no. 1, pp. 17–25, 2013.
[29] I. Rout, “Thermal Analysis of Steam Turbine Power Plants,” IOSR J. Mech. Civ. Eng., vol. 7,
no. 2, pp. 28–36, 2013.
[30] A. S. Karakurt and Ü. Güneş, “Performance analysis of a steam turbine power plant at part
load conditions,” J. Therm. Eng., vol. 3, no. 2, pp. 1121–1128, 2017.
[31] GE Power Systems, “Thermal Performance Evaluation and Assessment of Steam Turbine
Units,” GER-3953, New York, 1996.
[32] The American Society of Mechanical Engineers, “Performance Test Code 6 on Steam
Turbines,” ASME PTC 6-1996, New York, 1996.
[33] P. Akpan and W. Fuls, “Methodology for Developing and Validating Representative Process
Models for Coal- Fired Power Plants,” no. October, 2020.
[34] Eskom, “Station Thermal Efficiency Performance,” 240-133560870, Johannesburg, 2018.
69
[35] Eskom, “Title : Metering and Measurement Systems for Power Stations in Generation,” 240-
53459028, Johannesburg, 2018.
[36] P. Sewlall, “A methodology for implementing a power plant water balance on the online
condition monitoring software EtaPRO,” UCT, 2021.
[37] A. M. Carpenter, “Water conservation in coal-fired power plants Water conservation in coal-
fired power plants,” IEA Clean Coal Cent., no. February, pp. 1–72, 2017.
[38] Electric Power Research Institute, “Heat Rate Improvement,” TM-114073, Palo Alto, 2014.
[39] Electric Power Research Institute, “Range and applicability of heat rate improvements,” CA:
2014. 3002003457, Palo Alto, 2014.
[40] United States Department of Energy, “Power Plant Water Usage and Loss Study,” 2007.
[41] P. Kaushik and D. Khanduja, “DM make up water reduction in thermal power plants using
Six Sigma DMAIC methodology,” J. Sci. Ind. Res. (India)., vol. 67, no. 1, pp. 36–42, 2008.
[42] P. Moolman, STEP Training Material. Johannesburg, 2019.
[43] British Standards Institution, “Assessing thermal performance of boilers for steam, hot
water and high temperature heat transfer fluids,” BS 845-1, London, 1987.
[44] P. Moolman and J. Murphy, Station Thermal Efficiency Performance Manual. Johannesburg:
Eskom, 2014.
[45] International Organization for Standardization, “Uncertainty of measurements - Guide to
the expression of uncertainty in measurement,” ISO/IEC 98-3, Geneva, 2008.
[46] The Association of German Engineers, “Uncertainties of measurement during acceptance
tests on energy conversion and power plants - Fundaments,” VDI 2048 Part 1, Dusseldorf,
2000.
[47] The American Society of Mechanical Engineers, “Guidance for Evaluation of Measurement
Uncertainty in Performance Tests of Steam Turbines,” ASME PTC 6 -Report 1985, New York,
1997.
[48] R. S. Figliola and D. E. Beasley, Theory and Design for Mechanical Measurements, 5th ed.
Hoboken: John Wiley and Sons Inc., 2011.
[49] R. J. Moffat, “Describing the uncertainties in experimental results,” Exp. Therm. Fluid Sci.,
vol. 1, no. 1, pp. 3–17, 1988.
[50] R. D. Manteufel and A. Karimi, “Influence of uncertainties and assessment of significant
digits in thermodynamics,” ASEE Annu. Conf. Expo. Conf. Proc., 2013.
[51] M. Padulo, M. S. Campobasso, M. D. Guenov, M. S. Campobasso, and M. D. Guenov,
“Comparitive Analysis Of Uncertainty Propagation Methods For Robust Engineering Design,”
Int. Conf. Eng. Des., no. August, pp. 1–12, 2007.
[52] N. Metropolis, “The beginning of the Monte Carlo Method,” 1987.
[53] K. N. Hasan, R. Preece, and J. V. Milanović, “Existing approaches and trends in uncertainty
modelling and probabilistic stability analysis of power systems with renewable generation,”
Renew. Sustain. Energy Rev., vol. 101, no. December 2017, pp. 168–180, 2019.
[54] C. E. Papadopoulos and H. Yeung, “Uncertainty estimation and Monte Carlo simulation
70
method,” Flow Meas. Instrum., vol. 12, pp. 291–298, 2001.

[55] H. Keshavarzian and T. Nesari, “Reliability Analysis for Cyclic Fatigue Life Prediction in
Railroad Bolt Hole,” vol. 11, no. 9, pp. 1227–1232, 2017.
[56] G. S. Reber, D. D. Frey, and E. Greitzer, “Gaussian Quadrature for Computer Aided Robust
Design,” pp. 1–88, 2004.
[57] S. Bhonsale, D. Telen, B. Stokbroekx, and J. Van Impe, “An analysis of uncertainty
propagation methods applied to breakage population balance,” Processes, vol. 6, no. 12, pp.
1–14, 2018.
[58] L. Di Persio, G. Pellegrini, and M. Bonollo, “Polynomial chaos expansion approach to interest
rate models,” J. Probab. Stat., vol. 2015, 2015.
[59] S. Yang, F. Xiong, and F. Wang, “Polynomial Chaos Expansion for Probabilistic Uncertainty
Propagation,” Intech, vol. i, no. tourism, p. 13, 2016.
[60] F. Xiong, S. Chen, and Y. Xiong, “Dynamic system uncertainty propagation using polynomial
chaos,” Chinese J. Aeronaut., vol. 27, no. 5, pp. 1156–1170, 2014.
[61] A. Aleti, C. Trubiani, A. van Hoorn, and P. Jamshidi, “An efficient method for uncertainty
propagation in robust software performance estimation,” J. Syst. Softw., vol. 138, pp. 222–
235, 2018.
[62] A. Kaintura, T. Dhaene, and D. Spina, “Review of polynomial chaos-based methods for
uncertainty quantification in modern integrated circuits,” Electron., vol. 7, no. 3, 2018.
[63] S. Tyson, D. Donovan, B. Thompson, S. Lynch, and M. Tas, Uncertainty Modelling with
Polynomial Chaos Expansion Stage 1 Final Report, no. November. 2015.
[64] M. D. . Webster, M. A. . Tatang, and G. J. McRae, “Application of the probabilistic
collocation method for an uncertainty analysis of a simple ocean model,” MIT Jt. Progr. Sci.
Policy Glob. Chang. Rep., vol. 4, 1996.
[65] Eskom, “STEP Testing Standard,” 240-92464418, Johannesburg, 2020.
[66] D. M. Diez, C. D. Barr, and M. Cetinkaya-Rundel, OpenIntro Statistics, 3rd ed. Creative
Commons license, 2017.
[67] D. M. Diez, C. D. Barr, and M. Cetinkaya-Rundel, “OpenIntro Statistics,” 3rd ed., Creative
Commons license, 2017, pp. 429–430.
[68] K. Watanabe and P. Alto, “The International Association for the Properties of Water and
Steam Advisory Note No . 1 1 Introduction 2 Uncertainty of the IAPWS-95 Formulation in
Enthalpy,” no. 1, pp. 1–15, 2006.
[69] Eskom, “Fossil Fuel Firing Regulations Standard,” 240-105453648, Johannesburg, 2016.
[70] W. Luo, “Wiener Chaos Expansion and Numerical Solutions of Stochastic Partial Differential
Equations,” System, vol. 2006, p. 400, 2006.
[71] S. Yang, F. Xiong, and F. Wang, “Polynomial Chaos Expansion for Probabilistic Uncertainty
Propagation,” Intech, vol. i, no. tourism, p. 13, 2016.
[72] H. U. Fuchs, The Dynamics of Heat. New York: Spring Science + Business Media, 1996.
[73] British Standards Institution, “Rules for steam turbine thermal acceptance tests,” IEC 953-2,
71
London, 1990.
[74] Eskom, “Fossil Fuel Firring Regulations Standard,” 240-105453648, Johannesburg, 2016.
[75] British Standards Institution, “Solid mineral fuels — Determination of gross calorific value by
the bomb calorimetric method and calculation of net calorific value,” BS ISO 1928, London,
2009.
[76] Eskom, “Coal Quantity and Quality Accounting Standard for Thermal Efficiency
Determination,” 240-87607698, Johannesburg, 2018.
[77] International Organization for Standardization, “Solid mineral fuels — Determination of
gross calorific value by the bomb calorimetric method and calculation of net calorific value,”
BS ISO 1928, London, 2009.
[78] British Standards Institution, “Tracked Changes Hard coal and coke — Mechanical sampling
Part 2 : Coal — Sampling from moving streams,” BS ISO 13909-2, 2016.
[79] South African National Standards, “Solid mineral fuels — Determination of ash content,”
SANS 131, Pretoria, 2011.
[80] South African National Standards, “Hard coal — Determination of total moisture,” SANS
589, Pretoria, 2007.
[81] South African National Standards, “Hard coal and coke — Determination of volatile matter,”
SANS 50, Pretoria, 2011.
[82] South African National Standards, “Coal — Determination of forms of sulfur,” SANS 402,
Pretoria, 2005.
[83] D. Ratlhagana and F. Lombard, “Abrasive and Hardgrove Index : Realistic Boiler Chemical
Analysis,” Eskom, 2013.
[84] South African National Standards, “Hard coal — Determination of Hardgrove grindability
index,” SANS 5074, Pretoria, 1994.
[85] Eskom, “Title : Coal Quality Specifications for ESKOM Power Stations at Staithe / Silo Inlet
rev 1,” 240-71273834, Johannesburg, 2019.
[86] The American Society for Testing and Materials, “Standard Test Method for Heat of
Combustion of Liquid Hydrocarbon Fuels by Bomb Calorimeter,” ASTM D240 -19, West
Conshohocken, 2019.
72
Appendix A. Formulas from literature
A.1.5. Radiation and convection losses
QLoss .RC  kQBoil

0.7
.Out
(A.12)
Where:
k =0.0315 for brown coal (lignite)

k =0.0220 for hard coal (anthracite)
QBoil .Out = Boiler heat output (MW)
A.2. Turbine plant performance principles [16]
A.2.1. Turbine deterioration Loss [16]
Turbine cylinder efficiency tests are performed on the HP and IP turbine sections by the enthalpy
drop method to assess the actual performance or condition of the turbine cylinders. The stated
method is preferred as per the IEC [16] to determine the turbine loss due to ageing, however, a
theoretical method to estimate the target turbine efficiency degradation as a function of age for
the cylinder efficiencies is given by:
150
DetTurbine  k f (A.13)
P
Where:
DetTurbine -Target turbine efficiency deterioration [% per month]
P - Power rating of the turbine [MW]

k =0.1 for 2-12 months of operation and 0.06 for 12-24months
f =1 for thermal performance. For power output on reheat cycles f =1.45
Cylinder tests are typically conducted during initial commissioning and thereafter after major plant
outages therefore In STEP the actual turbine deterioration is set equal to the target for monthly
performance evaluation and baselined after each cylinder efficiency test.
75
Figure A- 3: Example of turbine throttle temperature HR correction curve [31]
Figure A- 4: Example of reheat temperature HR correction curve [31]
Figure A- 5: Example of reheat pressure drop HR correction curve [31]
77
The generalized expression for computing the mean and variance of a function subjected to input
variability can be found below [51],[56]:

y  E  f ( x )   f ( x )p ( x )dx
x (A.17)


  f ( x )  E  f ( x ) p ( x )dx
2
2  x (A.18)

px is the joint probability function

The above integrals are approximated by applying the GQ integration technique and one-
dimensional integration rule for multivariate functions to achieve the formulas below[51].
N  N  N 
yGQ  Wi1  Wi2  ...Win f ( xi1 ,i1 ,...,in )   (A.19)
 i 1 
i1 1 2  in 1 
N  N  N 
 y2  Wi  Wi  ...Wi f ( xi ,i ,...,i )  yGQ   (A.20)
 
 
GQ 1 2 n 1 1 n
i1 1  i2 1 in 1
A.4.2. Method of Moments (MM)
MM is a derivative base technique with a reduced computational expense, however, relies on

methods to process the derivatives as per the Taylor Series expansion. MM used in uncertainty
propagation is relatively inaccurate for inputs with large variation if the second and higher-order
derivatives are eliminated. MM has been significantly used in Robust Design Optimization in the
aircraft design [51]
MM is a derivative base technique with a reduced computational expense, however, relies on
methods to process the derivatives as per the Taylor Series expansion [51]:
1 n  2 f  2 1 n  3 f  3 1 n  4 f  4
yMM  f ( x )  
2 p1  x p2
 xp    3  p xp    4
6 p1  x p  24 p1  x p
 p xp (A.21)
 
M1 M2 M3 M4
2
n f  n  2
 f  f  n n 
3 f  f  2 2
    p xp    2  x  p xp 
 2

2
  
3
 2 
 x   xp  xq
p 1  x p  p 1  x p p 1 q 1  x p xq
yMM
 p  qp
 q 
2 2
(A.21)
n  2 f
n  2 2 1 n  3 f  f  1 n  2 f 
   p  1 xp
1
     xp  xq    3    p x4p    2 4
2 p1 q 1  x p xq  3 p1  x p  x 

 p  4 p1  x p 
qp
79
The above equations apply to independent input variables and the Taylor series expansion to the
4th order. M = moments. MM used in uncertainty propagation is relatively inaccurate for inputs
with large variation if the second and higher-order derivatives are eliminated. MM has been
significantly used in Robust Design Optimization in the aircraft design [51]
A.4.3. Sigma Point (SP)
SP methods are synonymous with the sequential perturbation of uncertainties described in section
4.3.1, however, a weight is applied to each variable perturbed ( Wp ) as a function of its
distribution. The SP method is further categorized as Unscented Transformation (UT) and Divided
Difference Filter (DDF), where UT only considers the first-order polynomial approximation and DDF
creates a second-order polynomial approximation. A significant advantage of this method is the
ability to propagate uncertainties through discontinuous functions and relatively lower
computational expense (2n +1), n is the number of variables with uncertainties. The DDF has a
higher accuracy when compared with UT, however increases the computational expense. SP has
been recently demonstrated by Bhonsale et al. [57] to be the most favourable method to be
employed to breakage population models in the pharmaceutical design process. The comparison
with MCS yielded a good correlation for significantly fewer function evaluations.
The SP method is described by the equations below [51] :
n
yUT  W0 f ( x0 )  Wp  f ( x p )  f ( x p ) (A.22)
p 1
 
n
 2 2

 y2  W0  f ( x )  yUT   Wp  f ( x )  yUT    f ( x )  yUT  
2
(A.23)
UT  0
 p1  p
  
p
1 n  2 2

 2
  Wp  f ( x p )  f ( x p  2 
)  ( Wp  2Wp f ( x p )  f ( x p )  2 f ( x0 )  (A.24)
2 p1     
yDDF

where:
x p - positive perturbation of each variable by its uncertainty,
f ( x0 ) - evaluation of the function at its mean input variables

Wp - weight associated with the distribution of each input variable
80
A.4.4. Polynomial Chaos Expansion (PCE) and Probabilistic Collocation Method

(PCM)
PCE is a probabilistic numeral method that was first introduced by mathematician Robert Wiener
in 1938. It has since been improved by various authors such as Martin & Cameron, Ghanem &
Spanos , Bhabuska,Schwab [38],[58]. There have been various comparative studies to determine
the suitability of PCE for various applications; Persio et al. tested PCE for interest rate models
[58],Yang et al. [71] and Fenfen et al. [60] for engineering design.
There is very little use of PCE in the application of this method possibly due to its high
computational expense especially with a large number of input variables, mathematical
complexity and the requirement to express a model as a single function. Some examples of
applied uncertainty analysis include software performance estimation by Aleti et al. [61] and
robust design of Integrated Circuits (ICs) by Kaintura et al. [62]. PCE makes use of appropriate
orthogonal polynomials (Hermite, Legendre, Jacobi, Laguerre, and General Laguerre) relative to
probability density functions (Gaussion, Uniform, Beta, Exponential, and Gamma) also known as
the Askey Scheme to estimate a function. The properties of orthogonality are used to simplify the
calculations of the statistical moments(mean, variance, kurtosis and skewness) of the outputs
relative to input distribution properties, see [51],[58],[59],[60],[61],[60],[63] for mathematical
detail.
PCM is one of the non-intrusive methods in solving the coefficients of the orthogonal polynomial
in the Estimate Function. This entails selecting collocation points similar to the GQ integration
technique. These points are generally the roots of the next higher-order polynomials in the Askey
Scheme [64].
81
Appendix B. STEP calculations

Note: The number of significant digits is maintained from plant-specific source information. The
coefficients used in the polynomial equations below are variables specific to a boiler design.
TGas.LH.AH.Out  TGas.RH.AH.Out (B.1)

i i
TGas.AH.Out 
i 2
TAir.RH.FD.In  TAir.LH.FD.In
i i
TAir.FD.In 
i 2 (B.2)
xO2.RH.Eco.Out  xO2.LH.Eco.Out
i i
xO2.Eco.Out 
i 2 (B.3)
xO2.LH.AH.Out xO2.LH.Eco.Out
i i

% % (B.4)
LeakAH.LH 
i xO2.LH.AH.Out
i
21 
%
xO2.RH.AH.Out xO2.RH.Eco.Out
i i

% % (B.5)
LeakAH.RH 
i xO2.RH.AH.Out
i
21 
%
mStm.Boil  mFW  mBD (B.6)

i i i
LFBoil  0 if tRT.boiler 0 (B.7)

i i
mStm.Boil
i
otherwise
tRT.boiler  MCRBoil
i i
82
QStart.0to7hrs  0 if NStarts.0to7hrs 0 GWh (B.19)

i i
  QStart.Cold.1 
i 
  kQ.OffLoad 
  i GWh 
  NStarts.0to7hrs 
QStart.Cold.1 1  e  N
i
 Starts.0to7hrs otherwise
i i

i i
i 
  i GWh 

QStart.Cold.1 1  e  i  N
i i

i i
i 
  i GWh 

QStart.Cold.1 1  e  i  N
i i

i i
i 
  i GWh 
QStart.Cold.1 1  e  N
i
i i
QStart.Hot  QStart.0to7hrs  QStart.7to15hrs  (B.23)

i i i
 QStart.15to24hrs  QStart.24to168hrs
i i
QStart.Cold  QStart.Cold.1 NStarts.Cold (B.24)

i i i
tSby.Hot  kQ.OffLoad  1
QSby.Hot 
i
QStart.Cold.1  1  e
i 
hr

i i (B.25)
85
6
LCIR.Act.Stn   LCIR.Act FracTot.Stm 
 i i (B.59)
i  1
LCIR.SF.Stn  LCIR.Act.Stn  LCIR.Targ.Stn (B.60)
6
LDFG.Targ.Stn   LDFG.Targ FracTot.Stm 
 i i (B.61)
i  1
6
LDFG.Act.Stn   LDFG.Act FracTot.Stm 
 i i (B.62)
i  1
LDFG.SF.Stn  LDFG.Act.Stn  LDFG.Targ.Stn (B.63)
6
LH2_Moist.Targ.Stn   LH2_Moist.Targ FracTot.Stm 
 i i (B.64)
i  1
6
LH2_Moist.Act.Stn   LH2_Moist.Act FracTot.Stm 
 i i (B.65)
i  1
LH2_Moist.SF.Stn  LH2_Moist.Act.Stn  LH2_Moist.Targ.Stn (B.66)
6
LRad.Stn   LRad FracTot.Stm 
 i i (B.67)
i  1
6
LMR.Act.Stn   LMR.Act FracTot.Stm 
 i i (B.68)
i  1
90
LBoil.Targ.Stn  LDFG.Targ.Stn  LH2_Moist.Targ.Stn  LCIR.Targ.Stn  LRad.Stn (B.69)
LBoil.Act.Stn  LDFG.Act.Stn  LH2_Moist.Act.Stn  LCIR.Act.Stn  LRad.Stn (B.70)
LBoil.SF.Stn  LBoil.Act.Stn  LBoil.Targ.Stn (B.71)
6 (B.72)
QOffLoad.Stn   QOffLoad.Boil 
 i 
i  1
 CFLF.Boil mStm.Boil 
 i i (B.73)
i  1
CFLF.Boil.Stn 
mStm.Stn
 CFFuel.Boil mStm.Boil 
 i i (B.74)
i  1
CFFuel.Boil.Stn 
mStm.Stn
6 (B.75)
xCIR.Stn   xCIR FracTot.Stm 
 i i
i  1
6
xO2.AH.Gas.In.Stn   xO2.Eco.Out FracTot.Stm 
 i i (B.76)
i  1
6
LeakAH.RH.Stn   LeakAH.RH FracTot.Stm 
 i i (B.77)
i  1
6
LeakAH.LH.Stn   LeakAH.LH FracTot.Stm 
 i i (B.78)
i  1
JGen.LV.Main  JGen.LV.Check
i i
JGen.LV 
i 2 (B.79)
91
JAuxPower.Main  JAuxPower.Check
i i
JAuxPower.UB 
i 2 (B.80)
i i 
 i

JGen.HV  JGen.LV   1  BinaryMP JAuxPower.UB GT  JAuxPower.UB
 i
(B.81)
BinaryMP  1 For electrical measurement point after the unit transformer
BinaryMP  0 For electrical measurement point at generator terminals

6
JGen.HV.Stn   JGen.HV
i (B.82)
i  1
JStn1.Main  JStn1.Check  JStn2.Main  JStn2.Check  JStn2.Main  JStn2.Check (B.83)

JImp.Stn   JExttr
2
6
JAuxPower.Stn   JAuxPower.UB  JImp.Stn
i (B.84)
i  1
JUSO.Stn  JGen.HV.Stn  JAuxPower.Stn (B.85)
JAuxPower.Common.Stn  JImp.Stn  JAuxPower.Offload.Stn  (B.86)

6
   JUnit.AuxPower.SB
i
i  1
JGen.HV
i
FracGen 
i JGen.HV.Stn (B.87)
JAuxPower  FracGen JAuxPower.Common.Stn  JUnit.AuxPower.SB  JAuxPower.UB (B.88)

i i i i
JUSO  JGen.HV  JAuxPower (B.89)

i i i
92
AVLNLtoNER  0 if tRT.Gen 0 (B.90)

i i
JGen.HV  JGen.>NER
i i
otherwise
tRT.Gen  tGen.>NER
i i
 QIncr.NLtoNER AVLNLtoNER QNL 

i i i 
QCons_hr.NLtoNER   
kW hr 
kWh
i  kW hr kW (B.91)
 tRT.Gen tGen.>NER 
i i
QAT.NLtoNER  QCons_hr.NLtoNER   
i i  hr hr  (B.92)
 QIncr.NLtoNER NER QNL 

i i i 
QNER   
kW hr 
kWh
i  kW hr kW (B.93)
AVLNERtoMCR  0 if tGen.>NER 0 (B.94)

i i
JGen.>NER
i
otherwise
tGen.>NER
i
 QIncr.NERtoMCR AVLNERtoMCR QNER 

i i i
QCons_hr.NERtoMCR   
kWh 
kWh
i  kWh kW (B.95)
 tGen.>NER 
i
QAT.NERtoMCR  QCons_hr.NERtoMCR  
i i  hr  (B.96)
QCons.AT  QAT.NERtoMCR  QAT.NLtoNER (B.97)

i i i
93
6
LFUSO.Stn   LFUSO FracGen 
 i i (B.111)
i  1
6
LFUSO.Stn   LFUSO FracGen 
 i i (B.112)
i  1
  LFUSO 
4
 LFUSO 
3 
 i i 
FAuxPower.Est     %     %    (B.113)
i
     
2
  LFUSO  LFUSO 
i i
   %     
   % 
FracAuxPower.Targ  0 if JUSO 0 (B.114)

i i
 JEFPs.Targ 
F FractAuxPower.min.ACC 
i
otherwise
 AuxPower.Esti i JUSO 
 i 
CFAuxPower  0 if JUSO 0 (B.115)

i i
 FracAuxPower.Targ  JEFPs.Targ 100

i i
100   
 %  JUSO
i
otherwise
FractAuxPower.min.ACC
i
100 
%
96
PBP.Targ  P6deg if TACC.Air.In /°C  6 (B.126)

i i i
P10deg  P6deg 
i i
 TACC.Air.In /°C  6  P6deg if 6  TACC.Air.In /°C  10
 10  6  i  i i
i i
  ACC.Air.In
T /°C  10 10deg
 P if 10  TACC.Air.In /°C  15
 15  10  i  i i
i i
  ACC.Air.In
T /°C  15 
 15deg
P if 15  TACC.Air.In /°C  20
 20  15  i  i i
i i
  ACC.Air.In
T /°C  20 20deg
 P if 20  TACC.Air.In /°C  25
 25  20  i  i i
i i
 30  25  i  i i
i i
 35  30  i  i i
i i
 40  35  i  i i
i i
 42  40  i  i i
P42deg otherwise
i
  PBP.Targ 
4
 PBP.Targ  
3
 i i
CFBP.Targ.40%    kPa     kPa   (B.127)
i
     
2
  PBP.Targ  PBP.Targ 
i i
   kPa     
   kPa 
 
4
3
 i i
CFBP.Targ.60%     kPa     kPa   (B.128)
i
     
2
i i
   kPa     
   kPa 
 
99
  PBP.Targ 
4
3
 7 i i
CFBP.Targ.80%     kPa     kPa   (B.129)
i
     
2
i i
   kPa     
   kPa 
 
4
3
 i i
CFBP.Targ.100%     kPa     kPa   (B.130)
i      
  PBP.Targ 
2
PBP.Targ 
    i
  
i
 
  kPa  kPa 
 
  
CFBP.Targ  CFBP.Targ.100% if LFGen  1 (B.131)

i i i
CFBP.Targ.100%  CFBP.Targ.80%  LFGen  

 i i i
 
 100  80  % 80   if 1  LFGeni  0.8
 
 CFBP.Targ.80% 
 i 
 i i i
 
 80  60  % 60   if 0.8  LFGeni  0.6
 
 i 
 i i i
 40  if 0.6  LFGen  0.4
 
60  40  %   i
 i 
CFBP.Targ.40% otherwise
i
 FracMU.Targ 
i
1  RQ.MU 
CFMU.Targ 
 i %  (B.132)
i  Frac MU.Targi 
1  RQ.MU FMU.Est 
 i % i
100
6
VMU.Other  VMU.MeterA.Stn  VMU.MeterB.Stn   VMU
i (B.159)
i  1
FMU.Hot  0 if mStm.Boil 0 (B.160)

i i
VMU  FracGen VMU.Other  VMU.Cold
i  i i otherwise
mStm.Boil
i
FMU.Hot 1  RQ.MU FracMU.Targ 

LMU.Act 
i  i i
i ML  FMU.Hot  (B.161)
i
1  RQ.MUi
Gg
ML 
 
 Gg 
6
VMU.ExclU  VMU.Stn   VMU
i (B.162)
i  1
LMU.SF  0 if LFGen 0 (B.163)

i i
LMU.Act  LMU.Targ  RQ.MU 100 otherwise

 i i i 
LU.SF  LTurb.SF  LMU.SF  LAuxPower.SF  LTot.Boil.SF (B.164)

i i i i i
LU.SF.derived  1  LU.SF (B.165)

i i
MCRLV  MW (B.166)
i
LFGen.LV  0 if tRT.Gen 0 (B.167)

i i
JGen.LV
i
otherwise
CFAmb tRT.Gen MCRLV
i i i
105
HRIncr.Act  0 if LFGen 0 (B.202)

i i
6
HRIncr.AT CFBP.Targ 10
i i
otherwise
 LTurb.SF   LTot.Boil.Act   FractAuxPower.min.ACC 
i  i  i
100    100    100  
 %  %  % 
HRIncr.MCR.Act  0 if LFGen 0 (B.203)

i i
6
HRIncr.MCR.AT CFBP.Targ 10
i i
otherwise
 LTurb.SF   LTot.Boil.Act   FractAuxPower.min.ACC 
i  i  i
100    100    100  
 %  %  % 
6
TMS.Targ.Stn   THPT.Targ.In FracGen 
 i i (B.204)
i  1
6
TIPT.Targ.In   TIPT.Targ.In FracGen 
 i i (B.205)
i  1
6
TMS.Act.Stn   THPT.Act.In FracGen 
 i i (B.206)
i  1
6
TIPT.Act.Stn.In   TIPT.Act.In FracGen 
 i i (B.207)
i  1
6
LMS.SF.Stn   LMS.SF FracGen 
 i i (B.208)
i  1
6
LReH.SF.Stn   LReH.SF FracGen 
 i i (B.209)
i  1
113
6
PBP.Targ.Stn   PBP.Targ FracGen 
 i i (B.210)
i  1
6
PBP.Act.Stn   PCond FracGen 
 i i (B.211)
i  1
6
LBP.SF.Stn   LBP.SF FracGen 
 i i (B.212)
i  1
6
LTurb.Det.Stn   LDet.Turb FracGen 
 i i (B.213)
i  1
6
TFW.Targ.Stn   TFW.Targ FracGen 
 i i (B.214)
i  1
6
TFW.Act.Stn   TFW.Act FracGen 
 i i (B.215)
i  1
6
LFW.SF.Stn   LFW.SF FracGen 
 i i (B.216)
i  1
LTot.Turb.SF.Stn  LReH.SF.Stn  LMS.SF.Stn  LBP.SF.Stn  LTurb.Det.Stn  LFW.SF.Stn (B.217)
6
LMU.Targ.Stn   LMU.Targ FracGen 
 i i (B.218)
i  1
114
6
LMU.Act.Stn   LMU.Act FracGen 
 i i (B.219)
i  1
6 (B.220)
RQ.MU.Stn   RQ.MU FracGen 
 i i
i  1

LMU.SF.Stn  LMU.Act.Stn  LMU.Targ.Stn RQ.MU.Stn100  (B.221)
6
FractAuxPower.Act.Stn   FractAuxPower.Act FracGen 
 i i (B.222)
i  1
6
FractAuxPower.Targ.Stn   FracAuxPower.Targ FracGen 
 i i (B.223)
i  1
LAuxPower.SF.Stn  FractAuxPower.Act.Stn  FractAuxPower.Targ.Stn (B.224)
6
HRIncr.Act.Stn   HRIncr.Act FracGen 
 i i (B.225)
i  1
6
HRIncr.MCR.Act.Stn   HRIncr.MCR.Act FracGen 
 i i (B.226)
i  1
CostCoal.Incremental
CostQ.Incr 
CVCoal
(B.227)
115
QTot.Targ Cost
i Q.Incr
GWh c
kWh
CostL  1000 (B.228)
i LU.SF.derived
i
%
LMS.SF
i
CostL.MS  CostL FracGen
i % i i (B.229)
LReH.SF
i
CostL.ReH  CostL FracGen
i % i i (B.230)
LBP.SF
i
CostL.BP  CostL FracGen
i % i i (B.231)
LDet.Turb
i
CostL.Det.Turb  CostL FracGen
i % i i (B.232)
LFW.SF
i
CostL.FW  CostL FracGen
i % i i (B.233)
CostL.Turb  CostL.FW  CostL.Det.Turb  CostL.BP  CostL.MS  CostL.ReH (B.234)

i i i i i i
LMU.SF
i
CostL.MU  CostL FracGen
i % i i (B.235)
LAuxPower.SF
i
CostL.AuxPower  CostL FracGen
i % i i (B.236)
116
LCIR.SF
i
CostL.CIR  CostL FracTot.Stm
i % i i (B.237)
LDFG.SF
i
CostL.DFL  CostL FracTot.Stm
i % i i (B.238)
LH2_Moist.SF
i
CostL.H2_Moist  CostL FracTot.Stm
i % i i (B.239)
CostL.Boil  CostL.H2_Moist  CostL.DFL  CostL.CIR (B.240)

i i i i
Indirect.Act.Boil  1  LTot.Boil.Act (B.241)

i i
Indirect.Targ.Boil  1  LTot.Boil.Targ (B.242)

i i
Indirect.Var..Boil  Indirect.Act.Boil  Indirect.Targ.Boil (B.243)

i i i
4 3
 LFBoil   LFBoil 
i i
Load.AT.Boil    %     %   (B.244)
i    
2
 LFBoil 
i
   %  
 
LFBoil
i
   
%
CostL.Tot  CostL.AuxPower  CostL.MU  CostL.Turb  CostL.Boil (B.245)

i i i i i
117
QStn.In
HRAct.Stn 
JUSO.Stn  JAuxPower.Offload.Stn
(B.255)
1
overall 
HRAct.Stn
(B.256)
QTot.Targ.Stn (B.257)
HRTarg.Stn 
HRTarg.Stn
SFDirect.Stn 
HRAct.Stn
(B.258)
LAcc.Stn  LBoil.SF.Stn  LTot.Turb.SF.Stn  LMU.SF.Stn  LAuxPower.SF.Stn (B.259)
SFIndirect.Stn  1  LAcc.Stn (B.260)
LSF.equiv.Stn  1  SFDirect.Stn (B.261)
LUnacc.Stn  LSF.equiv.Stn  LAcc.Stn (B.262)
mCoal.StnCostCoal_Ton
CostCoal_USO 
JUSO.Stn
(B.263)
mFO.Stn CostFO_ton
CostFO_USO 
JUSO.Stn
(B.264)
CostFuel_USO  CostCoal_USO  CostFO_USO (B.265)
119
CostCoal.Incremental
CostQ.Incr 
CVCoal
(B.266)
CostQ.Incr_USO  CostQ.Incr HRIncr.Act.Stn (B.267)
6 (B.268)
P'Cap.Installed.USO.Stn   P'Cap.Installed.USO
i
i  1
mCoal.StnSFDirect.Stn
m'Coal.Stn 

JUSO.Stn 1  LAcc.Stn  LUnacc.prev.month.Stn  (B.269)
LFStd.USO.Stn 
P'Cap.Installed.USO.Stnthr_month
(B.270)
6
FractAuxPower.min.ACC.Stn   FractAuxPower.min.ACC FracGen 
 i i (B.271)
i  1
JEFPs.Targ.Stn
1  FractAuxPower.Targ.Stn  JUSO.Stn
(B.272)
CFAuxPower.Stn 
1  FractAuxPower.min.ACC.Stn
JUSO.Stn (B.273)
LFSby.Stn 
JSby.Stn
JUSO.Stn  JAuxPower.Offload.Stn (B.274)

LFAP.SO.Stn 
P'Cap.Installed.USO.StntGen.Stn
120
VRW.Stn (B.275)
Litre_USORW 
JUSO.Stn
VWW.Stn (B.276)
Litre_USOWW 
JUSO.Stn
QTot.Targ.Stn CostQ.Incr (B.277)

CostL.Stn 
SFDirect.Stn
CostCIR.Stn  LCIR.SF.Stn CostL.Stn (B.278)
CostCIR.Stn  LCIR.SF.Stn CostL.Stn (B.279)
CostH2_Moist.Stn  LH2_Moist.SF.StnCostL.Stn (B.280)
CostBoil.Stn  CostH2_Moist.Stn  CostDFG.Stn  CostCIR.Stn (B.281)
CostMSt.Stn  LMS.SF.StnCostL.Stn (B.282)
CostTurb.Det.Stn  LTurb.Det.Stn CostL.Stn (B.283)
CostBP.Stn  LBP.SF.Stn CostL.Stn (B.284)
CostReH.Stn  LReH.SF.Stn CostL.Stn (B.285)
CostFW.Stn  LFW.SF.StnCostL.Stn (B.286)
121
CostTurb.Stn  CostFW.Stn  CostReH.Stn  CostBP.Stn  CostTurb.Det.Stn  CostMSt.Stn (B.287)
CostMU.Stn  LMU.SF.Stn CostL.Stn (B.288)
CostAuxPower.Stn  LAuxPower.SF.StnCostL.Stn (B.289)
CostAcc.L.Stn  CostAuxPower.Stn  CostMU.Stn  CostTurb.Stn  CostBoil.Stn (B.290)
CostSF.Equiv.Stn  LSF.equiv.StnCostL.Stn (B.291)
CostUnnacc.L.Stn  CostSF.Equiv.Stn  CostAcc.L.Stn (B.292)
JAuxPower.Offload.Stn (B.293)
FAuxPower.OffLoad.Stn 
CVCoal (B.294)
CVAdj.EM 
1  xEM.Coal
CVCoal (B.295)
CVMF 
1  xTM.Coal
QAdj.Gen.Stn  CVAdj.EM mCoal.Stn  mFO.Stn CVFO.Std (B.296)
QAdj.Gen.Stn
HRStn 
JUSO.Stn
(B.297)
1 (B.298)
Stn 
HR Stn
Boil.Stn  1  LBoil.Act.Stn (B.299)
LBoil.SF.Stn  LBoil.SF.Stn (B.300)
122
mCoal.Stn (B.301)
BRStn 
JUSO.Stn
6
QTot.TurbTarg.Stn   QCons.AT CFPF CFBP.Targ 
 i i i (B.302)
i  1
JGen.HV.Stn
Turb.Stn 
QTot.TurbTarg.Stn
1  LTot.Turb.SF.Stn
(B.303)
VMU.Stn
l
RatioMU_Stm.Stn  (B.304)
mStm.Stn
kg
JAuxPower.Stn (B.305)
FractAuxPower.Stn 
JUSO.Stn
JUSO.Stn
LFSby 
JSby.Stn
(B.306)
JUSO.Stn
LFMax.Demand 
thr_month USOMax.Stn
(B.307)
JGen.HV.Stn
LFRunning.Stn 
JCap.Stn
(B.308)
JUSO.Stn
LFUSO.Stn 
P'Cap.Installed.USO.Stnthr_month
(B.309)
123
LFUSO.Stn
EUFStn 
EAFStn
(B.310)
124
Appendix C. Model uncertainty analysis

Note: For most of this section STEP is compare with the BSEN12952-15:2003 [15]. Even though the
total boiler heat input includes the energy of the combustion air and atomizing steam, it is omitted
when calculating the heat loss percentages so that the comparison is done on the same base as
STEP so that key differences can be appreciated and attributed to the methodology. Hence the
losses in MJ per kg coal-fired are divided only by the CV of coal.
C.1. Carbon in refuse as per STEP [34]

MJ MJ
Case: CVCoal  18.780 CVC  33.820 xAsh.Coal  0.309 FA  0.950 xC.FA  0.016
kg kg
xC.BA  0.028
From equation (B.27)
 
xCIR  FA xC.FA  1  FA xC.BA  0.017 (C.1)
The equation below is also derived from 1st principles mass balance at a higher level.
xCu (C.2)
xCIR 
xCu  xAsh.Coal
Then solving for xCu , we derive equation (B.33):
xCIR xAsh.Coal
xCu   0.005
1  xCIR  (C.3)
Simplifying equation (B.42), we get:
xCu CVC (C.4)

LCIR.Act   0.940 %
CVCoal
or in energy per kg coal-fired:
MJ
LCIR.Act  xCu CVC  0.177
kg Coal_fired
(C.5)
125
C.2. Carbon in refuse as per BS EN standard [15]

kJ
Case: xTM.Coal  0.07 xVM.Ash  0.05 BA  0.050 Tref  25 °C Cp.FA  0.840 TFA  130 °C
kg K
kJ kJ
TBA  790 °C Cp.BA  1.000 Cp.FA  0.840
kg K kg K
 
MJ (C.6)
hBA  Cp.BA TBA  Tref  xC.BA CVC  1.712
kg
 
MJ (C.7)
hFA  Cp.FA TFA  Tref  xC.FA CVC  0.629
kg

xAsh.Coal 1  xVM.Ash   xC.BA xC.FA  (C.8)
xCu    BA  FA  0.008
1  xAsh.Coal  xTM.Coal
 1  xC.BA 1  xC.FA


xAsh.Coal 1  xVM.Ash   BA FA  MJ (C.9)
LCIR.Act   hBA  hFA  0.206
1  xCu
 1  xC.BA 1  xC.FA
 kg Coal_fired
or as the percentage of the energy fired:

xAsh.Coal 1  xVM.Ash   BA FA 
 hBA  hFA
1  xCu
LCIR.Act 
 1  xC.BA 1  xC.FA
  1.097 % (C.10)
CVCoal
C.3. Carbon monoxide loss as per BS EN 12952-15 [15]

3
MJ 200 m
kg kg
Case: CVCO  12.633 yCO.DFG   DFG  0.747 DFG  9.455
m
3 1000000 m3
m
3 kg Coal_fired
3 (C.11)
DFG m
VDFG   12.657
 DFG kg Coal_fired
MJ (C.12)
LCO.Act  VDFG yCO.DFG CVCO  0.032
kg Coal_fired
Or
VDFG yCO.DFG CVCO (C.13)

LCO.Act   0.170 %
CVCoal
126
C.4. Dry Flue Gas (DFG) loss as per STEP [34]

Case: xH.Coal  0.031 xC.Coal  0.4682 xS.Coal  0.009 yO2.AH.Out  0.051 TGas.AH.Out  403.65K
kJ
TAir.FD.In  308.65K Cp.DFG  1.009
kg K
 xCIR 
 CV 
xC.Coal  % Coal
  xS.Coal 
Ce    0.375   46.236 (C.13)
%  kJ   % 
 33820 kg 
 
 xC.Coal   xVM.Coal  (C.14)
k  6.28   0.0897   0.023   2.586
 %   % 
yO2.AH.Out
21 
%
yCO2.AH.Out  %  0.124
 xH.Coal  xVM.Coal
2.37   0.375 k (C.15)
1
 %  %
Ce
yN2.AH.Out  1  yCO2.AH.Out  yO2.AH.Out  0.825 (C.16)
The calculations above aim to estimate the volume fraction of CO2 and N2 on a dry basis, by an
approximated molar balance relative to the O2 depleted from the air, where:
Ce - Carbon Equivalent
k - Empirical constant relating to N2 and O2 content in coal (not referenced in [34]). It is assumed
that this term is synonymous with the xN ,O .Coal as mentioned on page 42 [15].
11 yCO2.AH.Out  8 yO2.AH.Out  7 yN2.AH.Out kg

DFC   20.275
3 yCO2.AH.Out kg C_fired
(C.17)
DFC - Flue gas mass ratio to 1kg of carbon burned
 xS.Coal  kg
DFG  DFC  xC.Coal  xCu    9.399
 2.67  kg Coal_fired
(C.18)
DFG - Flue gas mass ratio to 1kg of coal-fired
127

DFG Cp.DFG TGas.AH.Out  TAir.FD.In  (C.19)
LDFG.Act   4.797 %
CVCoal
 
MJ (C.20)
LDFG.Act  DFG Cp.DFG TGas.AH.Out  TAir.FD.In  0.901
kg Coal_fired
C.5. Dry Flue Gas (DFG) loss as per BS 2885 [5]

The dry flue gas loss calculation in STEP is very similar to the method detailed in [5], however, the
calculation (C.18) is limited to yCO2.AH .Out being measure by an Orsat apparatus, which typically
x S .Coal
includes the SO2 content, otherwise, the term should be dropped.
2.67
Now DFG is calculated differently in molar form, see below:
kJ xCu  0.005
Cp.FG.mol  30.6
kmol K [5]
 xS.Coal  kg
xC.Coal  xCu   100
DFG 
 2.67  kg Coal_fired
 0.313
kmol
(C.21)
kg  yCO2.AH.Out yCO.DFG kg Coal_fired
12   
kmol  % % 
 
MJ
LDFG.Act  DFG Cp.FG.mol TGas.AH.Out  TAir.FD.In  0.909
kg Coal_fired
(C.22)

DFG Cp.FG.mol TGas.AH.Out  TAir.FD.In 
LDFG.Act   4.843 %
CVCoal
(C.23)
C.6. Dry Flue Gas (DFG) loss as per BS EN 12952-15 [15]

kg
Case: xVM.Ash  0.05 yCO2.DA  0.00033 yO2.DA  0.20938  DA.STP  1.293
3
m
xCO2.DA  0.000505 yO2.AH.Out  0.051 xN2.DA  0.756 xO2.DA  0.231 xC.Coal  0.468
kg kg kg
xH.Coal  0.031 xS.Coal  0.009 MNO2  46.010 MSO2  64.065 MCO2  44.0098
kmol kmol kmol
kg kg kg
MO2  32.000 MN  14.010 MS  32.065   0.192 S  1
kmol kmol kmol N
The BS EN 12952-15 [15] and provides an alternative method of calculating the gas products and
heat coefficients i.e. calculating the boiler excess air through an O2 balance. Stoichiometric
128
quantities are derived from elemental balances. Properties of air are acquired from DIN 1871 as
referenced in [15]. Other chemical properties not stated were acquired from [72].
 
xN.Coal  0.015 1  xAsh.Coal  xTM.Coal  0.009 (C.24)
The nitrogen content of coal is assumed to be 1.5 % on a dry-ash-free basis as per [15]. The
calculation above converts to as received/fired basis. The combined nitrogen and oxygen, xN ,O .Coal
is required. In this case, it is set to be equal to k = 0.026 as computed in the STEP dry flue gas loss
calculation above.
xO.Coal  xN.O.Coal  xN.Coal  0.017 (C.25)
kg (C.26)
 DA.Stoich  11.5122 xC.Coal  34.2974 xH.Coal   6.421
kg Coal_fired
 4.3129 xS.Coal  4.3212 xO.Coal
3 3
m m
VDFG.Stoich  8.8930 xC.Coal  20.9724 xH.Coal    4.807
 3.319 x  kg kg Coal_fired (C.27)
S.Coal  2.6424 xO.Coal 
 
 0.7997 xN.Coal 
kg (C.28)
 CO2.Stoich  3.6699 xC.Coal  0.0173 xH.Coal   1.719
kg Coal_fired
 0.0022 xS.Coal  0.0022 xO.Coal
kg (C.29)
 CO2   CO2.Stoich   1.720
kg Coal_fired
yO2.AH.Out
  DA.STP VDFG.Stoich x
yO2.DA  yO2.AH.Out CO2.DA
kg
 DA   DA.Stoich   8.422 (C.30)
kg Coal_fired
yO2.AH.Out
  DA.STP VDFG.Stoich
yO2.DA  yO2.AH.Out
kg (C.31)
 DA   DA.Stoich   8.422
kg Coal_fired
yO2.AH.Out
  DA.STP VDFG.Stoich
yO2.DA  yO2.AH.Out
kg (C.32)
 N2   DA.Stoich xN2.DA   6.367
kg Coal_fired
yO2.AH.Out
  DA.STP VDFG.Stoich x
yO2.DA  yO2.AH.Out N2.DA
129
yO2.AH.Out kg
 O2   DA.STP VDFG.Stoich x  0.462
yO2.DA  yO2.AH.Out O2.DA kg Coal_fired
(C.33)
MNO2 kg
 NO2  xN.Coal N  0.006
MN kg Coal_fired
(C.34)
MSO2 kg
 SO2  xS.Coal S  0.018 (C.35)
MS kg Coal_fired
N , S - Nitrogen and sulphur conversions are assumed for wall-fired boilers. The nitrogen
converted is approximated to 100% nitrogen dioxide and sulphur to 100% sulphur dioxide.
kg
 DFG   SO2   NO2   N2   CO2   O2  8.573
kg Coal_fired
(C.36)
For coal-fired boilers, other combustion products are negligible, then:

 CO2
MCO2
yCO2   0.139 (C.37)
 CO2  SO2  N2  O2  NO2
   
MCO2 MSO2 MN2 MO2 MNO2
 CO2 (C.38)
xCO2   0.201
 DFG
4  06 (C.39)
0.1919210 10 0.5883483 10 2
Cp.DA.0 ( T )  1.004173  T  T 
2 3
9  12
0.7011184 10 3 0.3309525 10 4
 T  T 
4 5
 16
0.5673876 10 5
 T
6
The integral heat capacity of dry air relative to 0oC is given by the function above.
130
3  06
0.7661864 10 0.9259622 10 2
P2m( T )  0.1002311  T  T 
2 3 (C.40)
9  12
0.5293496 10 3 0.1093573 10 4
 T  T
4 5
The above polynomial function is given in [15] to calculate the heat capacity of flue gas. See below:
Cp.DFG.0( T )  Cp.DA.0 ( T )  P2m( T ) xCO2 (C.41)
 TGas.AH.Out TGas.AH.Out
Cp.DFG.0  
 K  K
  TAir.FD.In TAir.FD.In
 Cp.DFG.0  
  K  K  kJ kJ (C.41)
Cp.DFG   1.074
TGas.AH.Out TAir.FD.In kg K kg K

K K
 
MJ (C.42)
LDFG.Act   DFG Cp.DFG TGas.AH.Out  TAir.FD.In  0.875
kg Coal_fired

 DFG Cp.DFG TGas.AH.Out  TAir.FD.In  (C.43)
LDFG.Act   4.660 %
CVCoal
C.7. Hydrogen and moisture loss as per BS EN 12952-15 [15]

xH20.Coal  0.08 xH.Coal  0.031  DA  8.422 w  0.013  AS  0
Case:
TGas.AH.Out  403.650K TAir.FD.In  298K
 H20.Fuel  xH20.Coal  8.937 xH.Coal  0.357

(C.44)
 H20   H20.Fuel   DA w   AS  0.357 (C.45)
 
kJ … from steam tables
hH20.FG  hsteam 1bar  TGas.AH.Out  ""  ""  """  2737.718
kg
 
kJ …from steam tables
hH20.Tref.liquid  hsteam 1.bar  TAir.FD.In  ""  """  """  104.301
kg
 
kJ … from steam tables
hH20.Tref.Gas  hsteam ""  TAir.FD.In  ""  1  """  2546.272
kg
 
MJ (C.46)
LH20   H20 hH20.FG  hH20.Tref.Gas  0.089
kg Coal_fired

 H20 hH20.FG  hH20.Tref.Gas 
LH20   0.476% (C.47)
CVCoal
131
It can be deduced from the term  H20 , the fuel moisture is assumed to be in a gaseous phase
hence combined with the moisture in combustion air and atomizing steam to determine only the
sensible heat losses. This would not apply to solid fuels. The calculations below will individually
compute the moisture in fuel loss and the moisture in combustion air which is congruent with BS
2885-19974 [5]. The reference enthalpy for moisture in fuel loss will be considered in the liquid
phase.
 
MJ (C.48)
LH20.Moist_Air   DA w hH20.FG  hH20.Tref.Gas  0.021
kg Coal_fired

 DA w hH20.FG  hH20.Tref.Gas  (C.49)
LH20.Moist_Air   0.112%
CVCoal
 
MJ (C.50)
LH20.Fuel   H20.Fuel hH20.FG  hH20.Tref.liquid  0.940
kg Coal_fired

 H20.Fuel hH20.FG  hH20.Tref.liquid  (C.51)
LH20.Fuel   5.007%
CVCoal
C.8. Hydrogen and moisture loss as per STEP [34]

xH20.Coal  0.08 xH.Coal  0.031 TGas.AH.Out  403.650K TAir.FD.In  298K
Case:
kL.H_Moisture.1  2.466 kL.H_Moisture.2  238.8
xTM.Coal  xH.Coal 
 9 
%  % 
(C.52)
LH_Moist.Act  kL.H_Moisture.1   %  4.897%
CVCoal  T 
  Gas.AH.Out 
/°C
MJ  TAir.FD.In /°C  
kg   2  
  kL.H_Moisture.2  
   
The origins of the above equation cannot be traced back to the BS 2885 [5], however KL.H _ Moist .1 is
the latent heat of water (MJ/kg) and KL.H _ Moist .2 is the inverse of the heat capacity of liquid water
(K.kg/MJ). So it is simplified that the fuel moisture would absorb sensible heat in the liquid phase
to the midpoint between TGas .AH .Out and TAir .Fd .In , thereafter in the form of latent. The sensible heat
absorbed in the gas phase is ignored.
132
C.9. Make-Up Water [34]

Case: VMU.Stn  108128000L JUSO.Stn  477.966GWH VMU.Other  103.518 ML VMU  4610ML
VMU.Cold  20ML FracMU.Targ  1% mStm.Boil  145000000000k g LFUSO  70% Boil  0.9
kg kg kJ kJ
FracGen  1 m'ReH.in  626 m'MS  630 hRW  104.913 hMS  3381.453
s s kg kg
kJ kJ kJ
hFW  1148.252 hReH.Out  3601.370 hReH.in  3003.550
kg kg kg
m'ReH.in (C.53)
 
kJ
h Boil  hMS  hFW  hReH.Out  hReH.in  2827.225
m'MS kg
In the above calculation, h Boil is the total heat absorbed per kg steam produced by the boiler.
2 1
hMS  hFW  hRW
3 3
RQ.MU   0.008 (C.54)
h Boil
100
Boil
where RQ.MUis the make-up heat ratio which can also be expressed as:
Heat of 1kg water lost 1

RQ.MU  
Energy required for 1kg steam produced 100
Therefore in generalized form:
 Heat of 1kg water lost 1   mMU 100 

LMU      
 Energy required for 1kg steam produced 100   mSteam 1 
FracMU (%)
Then:
FracMU.Targ (C.55)
LMU.design  RQ.MU  0.806 %
%
4 3
8  LFUSO  LFUSO
FMU.Est  3.4643 10    8.231 10 6     1.107
 %   %  (C.56)
2
 4  USO
LF LFUSO
 6.1239 10    8.0326 10 3  1.554
 %  %
133
The above factor calculated is a polynomial function to estimate the make-up water to steam
raised, mass fraction relative to design conditions at MCR, as a function of the sent out load factor
and is specific to a Unit design.
FracMU.MCR
1  RQ.MU
%
CFMU.Targ   1.001 (C.57)
FracMU.MCR
1  RQ.MU FMU.Est
%
The target cycle heat rate correction factor relative to the sent-out load factor is expressed above.
FracMU.MCR (C.58)
LMU.Targ  FMU.Est CFMU.Targ RQ.MU  0.893 %
%
It can be seen that there is now compensation for changes in efficiency specific to the deviation in
the make-up water fraction, CFMU.Targ . Initially, the energy input to the cycle was calculated by
setting the boiler efficiency to 90% as in calculation(C.49). The correction factor is necessary, for
the thermal loss to be normalized to one unit of energy input to the cycle.
 
VMU  FracGen VMU.Other  VMU.Cold ML
FMU.Hot   0.032
mStm.Boil Gg
(C.59)
From the above calculation, it is observed that make-up when the unit is off-load or before
synchronization, VMU.Cold is deducted from the total make-up accounted for. This is sensible as
this heat is accounted for in the heat for commissioning and should be included in the running unit
heat rate calculation. FracGen VMU.Other is the means to apportion all other demineralized water
flow that is not directly fed to the unit Rankine cycle to an individual unit. This would typically
include water used in the auxiliaries cooling loops, hydrogen production and other common plant
areas.
In most cases it is difficult to split this flow further into what is recovered to the Rankine cycle
typically via the condenser cleans drains tank and what would be lost externally to the Rankine
cycle such as the flows to the hydrogen production plant or compressed air supply plants. Each
power plant’s metering configuration should be carefully analysed to establish if this flow should
be considered for thermal losses or a factor should be applied which station-specific is. It is noted
that the density of water in the calculation (C.54) is approximated to 1kg/m3 and is therefore
134
FMU.Hot approximated to a mass fraction. A sensitivity calculation with a density of 0.997 kg/m3
proved to have a negligible impact on subsequent loss calculations.
 FracMU.MCR 
FMU.Hot  1  R Q.MU 
LMU.Act 
 %  R (C.60)
Q.MU  2.657 %
ML
%
 FMU.Hot 
Gg 1  RQ.MU 
ML
 % 
 Gg 

LMU.SF  LMU.Act  LMU.Targ  1.764 %  (C.61)
Note: In STEP, the term RQ.MU is dropped from calculations (C.54) and (C.55), but added to the
calculation (C.56) to give the same effect as the above.
C.10. Radiation and convection loss as per BS EN 12952-15 [15]

MJ kg
C  0.02675 for bituminous coal CVCoal  20.466 m'Coal  70.665
kg s
3
Q'Boil.Out  1.363  10 MW @ 97% MCR design
0.7
 Q'Boil.Out 
LRC  C   MW  4.183MW (C.62)
 MW 
LRC
LRC   0.289% (C.63)
m'Coal CVCoal
C.11. Radiation loss as per STEP [15]

LFBoil  97%
4 3
 LFBoil   LFBoil 
CFLF.Boil          (C.64)
 %   % 
2
 LFBoil  LFBoil
       
 %  %
135
4 3
 LFBoil   LFBoil 
kL.Rad         0.551 (C.65)
 %   % 
2
 LFBoil  LFBoil
     
 %  %
kL.Rad
LRad   0.568% (C.66)
LFBoil
CFLF.Boil
%
C.12. Turbine deterioration loss as per STEP

DetTurb.Previous  0 FAvailability  0.85 Nyr  10yr
%
LDet.Turb  DetTurb.Previous  0.1 N F  0.85% (C.67)
yr yr Availability
C.13. Turbine deterioration loss as per the IEC [73]

k1st_year  0.1 P  600MW f  1 k2nd_year  0.06 FAvailability  0.85
150  %   0.05 %
DetTurb.1st_Year  k1st_year f   (C.68)
P  month month
MW
150  %   0.03 %
DetTurb.2nd_Year  k2nd_year f   (C.69)
P  month month
MW

DetTurb.120_Month  12 DetTurb.1st_Year  108DetTurb.2nd_Year FAvailability  3.264% (C.70)
C.14. Turbine deterioration loss as per ASME PTC 6 Report-1985

[47]
BF10yrs  3.95 Base Factor …linearly extrapolated from Fig 3.3 [47]
Pow  600MW P  16.1MPa f  1.000 for fossil fuels
P
BF10yrs psi
DetTurb.10_yrs  f %  1.402% (C.71)
 Pow 
log
2400

 MW 
136
Appendix D. STEP program inputs details

Table D- 1: Input variable instrument uncertainty
List Variable Python Tag reference Variable Description Unit ui i

uLiterature Function Ntx Sourceii
no.
iii o
0 TGAS .LH .AH .OUTi AH_LH_Gas_Out_T[i] Left hand air heater outlet temperature C 1.604 5.6 Abs 12 [47]
o
1 TGAS .RH .AH .OUTi AH_RH_Gas_Out_T[i] Right hand air heater outlet temperature C 1.604 5.6 Abs 12 [47]
o
2 TAir .LH .FD .Ini FD_LH_AIR_In_T[i] Left hand forced draught fan air inlet C 2.268 5.6 Abs 6 [47]
temperature
o
3 TAir .RH .FD .Ini FD_RH_AIR_In_T[i] Right hand forced draught fan air inlet C 2.268 5.6 Abs 6 [47]
temperature
4 xO2.LH .Eco .Outi AH_LH_Gas_In_O2_x[i] Left hand air heater inlet oxygen content % (V/V) 0.212 0.3 Abs 2 [74]
5 xO2.RH .Eco.Outi AH_RH_Gas_In_O2_x[i] Right hand air heater inlet oxygen content % (V/V) 0.212 0.3 Abs 2 [69]
6 xO2.LH .AH .OUTi AH_LH_Gas_Out_O2_x[i] Left hand air heater outlet oxygen content % (V/V) 0.300 0.3 Abs 1 [69]
7 xO2.LH .AH .OUTi AH_RH_Gas_Out_O2_x[i] Right hand air heater outlet oxygen content % (V/V) 0.300 0.3 Abs 1 [69]
8 tRT .boileri Boil_Run_hrs[i] Boiler running hours hrs 5.000 5.0 % 1 default
9 mFWi FW_Tot_Mass[i] Total feed water delivered to the boiler Gg 3.200 3.2 % 1 [47]
10 mBDi BD_Tot_Mass[i] Total mass of water blown down ton 3.200 3.2 % 1 [47]
11 NStarts.0to7hrsi Boil_Starts_0to7hrs[i] Number of boiler starts after being offload - 5.000 5.0 % 1 default
for 0-7 hours
12 tOffload 0to7hrsi Boil_DownTime_0to7hrs[i] Boiler offload time for periods 0-7 hours hrs 5.000 5.0 % 1 default
13 NStarts 7to15hrsi Boil_Starts_7to15hrs[i] Number of boiler starts after being offload - 5.000 5.0 % 1 default
for 7-15 hours
14 tOffload 7to15hrsi Boil_DownTime_7to15hrs[i] Boiler offload time for periods 7-15 hours hrs 5.000 5.0 % 1 default
for 15-24 hours
16 tOffload 15to24hrsi Boil_DownTime_15to24hrs[i Boiler offload time for periods 15-24 hours hrs 5.000 5.0 % 1 default
]
for 24-168 hours
18 tOffload 24to168hrsi Boil_DownTime_24to168hrs[ Boiler offload time for periods 24-168 hours hrs 5.000 5.0 % 1 default
137
i]
19 t Sby .Hoti HotSby_Time[i] Period where the boiler is kept on hot hrs 5.000 5.0 % 1 default
standby
20 NStarts .Coldi Boil_ColdStarts[i] Boiler cold starts (RTS after being offload for - 5.000 5.0 % 1 default
>168 hours)
21 mMRi Mill_Reject_Mass[i] Mass of mill rejects ton 5.000 5.0 % 1 default
22 CVMRi Mill_Reject_CV[i] Calorific value of mill rejects MJ/kg 0.300 0.3 abs 1 [75]
23 xC .FAi AH_Gas_In_CIR[i] Air heater gas inlet carbon in ash % 1.000 1.0 % 5 [17]
(m/m)
24 xC .BAi CoarseAsh_CIR[i] Boiler bottom carbon in ash % 1.000 1.0 % 200 [17]
(m/m)
25 tRT .Mill .Ai MILL_A_Run_Hrs[i] Mill “A” running hours hrs 5.000 5.0 % 1 default
26 tRT .Mill .Bi MILL_B_Run_Hrs[i] Mill “B” running hours hrs 5.000 5.0 % 1 default
27 tRT .Mill .Ci MILL_C_Run_Hrs[i] Mill “C” running hours hrs 5.000 5.0 % 1 default
28 tRT .Mill .Di MILL_D_Run_Hrs[i] Mill “D” running hours hrs 5.000 5.0 % 1 default
29 tRT .Mill .Ei MILL_E_Run_Hrs[i] Mill “E” running hours hrs 5.000 5.0 % 1 default
30 tRT .Mill .Fi MILL_F_Run_Hrs[i] Mill “F” running hours hrs 5.000 5.0 % 1 default
31 JGen.LV .Maini Gen_LV_Main_J[i] Energy measured at the generator GWH 5.000 5.0 % 1 [47]iv
terminals main meter
32 JGen .LV .Checki Gen_LV_Check_J[i] Energy measured at the generator terminals GWH 5.000 5.0 % 1 [47]
check meter
33 JAuxPower .Maini AuxPower_Main_J[i] Energy measures at auxiliary power tap-off GWH 5.000 5.0 % 1 [47]
point main meter
34 JAuxPower .Checki AuxPower_Check_J[i] Energy measures at auxiliary power tap-off GWH 5.000 5.0 % 1 [47]
point check meter
35 JGen.NERi Gen_AboveNER_J[i] Energy measured at generator terminals GWH 5.000 5.0 % 1 [47]
when generating above NER
36 tGen.NERi Gen_AboveNER_RT[i] Time that generator operates above NER hrs 5.000 5.0 % 1 default
37 JReactive .Geni Gen_Reactive_J[i] Generator reactive energy GVARs 5.000 5.0 % 1 [47]
38 tRT .Geni Gen_RT[i] Generator run time (synchronization to Hrs 5.000 5.0 % 1 [47]
separation from the power grid)
39 JUnit .AuxPower .Tripi AuxPower_Offload_J[i] Auxiliary power measured when generator GWH 5.000 5.0 % 1 [47]
not synchronized
40 JUnit .AuxPower .SBi AuxPower_SB_J[i] Auxiliary power supplied from the station GWH 5.000 5.0 % 1 [47]
138
board
41 tRT .EFP .Ai A_EFP_RT[i] Electric feed water pump “A” run time hrs 5.000 5.0 % 1 default
42 tRT .EFP .Bi B_EFP_RT[i] Electric feed water pump “B” run time hrs 5.000 5.0 % 1 default
43 tRT .EFP .Ci C_EFP_RT[i] Electric feed water pump “C” run time hrs 5.000 5.0 % 1 default
o
44 TFW .Acti Eco_FW_In_T[i] Economizer feedwater inlet temperature C 5.556 5.6 abs 1 [47]
o
45 TLH .HPT .Ini HPT_LH_MS_In_T[i] High-pressure turbine left hand inlet C 3.928 5.6 abs 2
temperature [47]
o
46 TRH .HPT .Ini HPT_RH_MS_In_T[i] High-pressure turbine right hand inlet C 3.928 5.6 abs 2
temperature [47]
o
47 TLH .IPT .Ini IPT_LH_Stm_in_T[i] Intermediate-pressure turbine left hand C 3.928 5.6 abs 2
inlet temperature [47]
o
48 TRH .IPT .Ini IPT_RH_Stm_in_T[i] Intermediate-pressure turbine right hand C 3.928 5.6 abs 2
inlet temperature [47]
49 mHPT .Ini HPT_Stm_In_MFR[i] High-pressure turbine inlet mass flow rate kg/s 1.000 1.0 % 1 [47]
50 PHPT .Ini HPT_Stm_In_P[i] High-pressure turbine inlet pressure bar 0.250 0.5 % 4 [47]
51 mIPT .Ini IPT_Stm_In_MFR[i] Intermediate-pressure turbine inlet mass kg/s 1.000 1.0 % 1 [47]
flow rate
52 PIPT .Ini IPT_Stm_In_P[i] Intermediate-pressure turbine inlet bar 0.250 0.5 % 4
[47]
pressure
53 mReH .Ini ReH_Stm_In_MFR[i] Re-heater inlet mass flow rate kg/s 1.000 1.0 % 1 [47]
54 PHPT .Outi HPT_Stm_Out_P[i] High-pressure turbine outlet pressure bar 0.354 0.5 % 2 [47]
o
55 THPT .Outi HPT_Stm_Out_T[i] High-pressure turbine outlet temperature C 3.928 5.6 abs 2 [47]
56 mReH .SWi ReH_SW_MFR[i] Re-heater spray water mass flow rate kg/s 1.000 1.0 % 1 [47]
o
57 TReH .SW .Ini ReH_SW_T[i] Re-heater spray water supply temperature C 5.556 5.6 abs 1 [47]
58 tRT .CEP .Ai CEP_A_RT[i] Condensate extraction pump “A” run time hrs 5.000 5.0 % 1 default
59 tRT .CEP .Bi CEP_B_RT[i] Condensate extraction pump “B” run time hrs 5.000 5.0 % 1 default
o
60 TACC .Air .Ini ACC_Fan_In_T[i] Air cooled condenser air inlet temperature C 2.778 2.8 abs 1 [47]
o
61 TACC .Air .Outi ACC_Out_T[i] Air cooled condenser air outlet temperature C 2.778 2.8 abs 1 [47]
62 PCondi Cond_Act_P[i] Condenser actual pressure kPa 0.354 0.5 % 2 [47]
63 NStart .Hot .TurbI Turb_HotStart_Num[i] Number of turbine hot starts - 5.000 5.0 % 1 default
64 NStart .Cold .TurbI Turb_ColdStart_Num[i] Number of turbine cold starts - 5.000 5.0 % 1 default
139
65 VMUi MU_V[i] Make up water volume ML 1.000 1.0 % 1 [47]

66 VMU .Coldi MU_Cold_V[i] Make up water volume consumed when ML 1.000 1.0 % 1 [47]
unit not synchronized
67 DetTurb.Pr evious Turb_Previous_Det[i] Turbine deterioration from previous period % 0.000 0.0 % 1 n/a
I
o
68 TCondi Condensate_T[i] Condensate temperature C 3.928 5.6 abs 2 [47]
69 PowCap.Installed .USOI Capacity_USO_Pow[i] Unit sent out capacity MW 0.000 0.0 % 1 n/a
70 PowCap .LVI Capacity_LV_Pow[i] Unit generating capacity MW 0.000 0.0 % 1 n/a

71 JStn1.Main Stn_Tx1_Main_J Station transformer “1” energy imported GWH 5.000 5.0 % 1 [47]
measured by the main meter
72 JStn1.Check Stn_Tx1_CkeckMeter_J Station transformer “1” energy imported GHW 5.000 5.0 % 1 [47]
measured by the check meter
74 JStn2.Check Stn_Tx2_Check_J Station transformer “2” energy imported GHW 5.000 5.0 % 1 [47]
76 JStn3.Check Stn_Tx3_Check_J Station transformer “3” energy imported GHW 5.000 5.0 % 1 [47]
77 JExtr Extr_WP_J Extraneous works power MWH 5.000 5.0 % 1 [47]
78 mCoali Coal_Mass[i] Mass of coal fired ton 0.500 0.5 % 1 [76]
v
79 CVCoal Coal_CV Calorific value of coal MJ/kg 0.464 0.5 abs 1 [77],[78]
80 x Ash .Coal Coal_Ash_x Ash content in coal (as received) % 0.300 0.3 abs 1 [79]
(m/m)
81 xTM .Coal Coal_TM_x Total moisture of coal % 1.500 1.5 abs 1 [80]
(m/m)
82 xSM .Coal Coal_SurfMoist_x Surface moisture of coal % 1.500 1.5 abs 1 [80]
(m/m)
83 xVM .Coal Coal_VM_x Volatile matter content of coal % 0.500 0.5 abs 1 [81]
(m/m)
84 xFC .Coal Coal_FC_x Fixed carbon content of coal % 0.300 0.3 % 1 IM
(m/m)
85 x S .Coal Coal_S_x Sulphur content of coal % 0.150 0.2 abs 1 [82]
(m/m)
140
86 AICoal Coal_Abrassiveness Abrasive index of coal mgFe/k 15.000 15.0 % 1 [83]

g
87 HGICoal Coal_HGI Coal Hardgrove index - 5.000 5.0 abs 1 [84]
88 Cos tCoal .Incremental Coal_Incr_Cost Coal incremental cost R/ton 0.000 0.0 % 1 n/a
89 Cos tCoal _ ton Coal_PerTon_Cost Coal cost per ton R/ton 0.000 0.0 % 1 n/a
90 mFO .Stn FO_Mass Mass of fuel oil burned for all units ton 5.000 5.0 % 1
91 CVFO FO_CV Calorific value of fuel oil MJ/kg 0.300 0.3 abs 1 Typical
certificate
vi
92 QComm.Stn Heat_Comm Heat for commissioning GWH 5.000 5.0 % 1 default

93 VMU .MeterA.Stn MU_MeterA Make-up water through measuring point ML 1.000 1.0 % 1 [47]
“A”
94 VMU .MeterB .Stn MU_MeterB Make-up water through measuring point ML 1.000 1.0 % 1 [47]
“B”
95 VRW .Stn RW_V Raw Water intake volume ML 1.000 1.0 % 1 [47]
96 VWW .Stn WW_V Wastewater recovered volume Ml 1.000 1.0 % 1 [47]
97 LUnacc .prev .month .Stn Unacc_Previous_L Previous unaccounted loss % 0.000 0.0 % 1 n/a
98 USOMax .Stn USO_Max_Stn Maximum electrical units sent out (power) MW 0.000 0.0 % 1 n/a
99 UCFStn UCF_Stn Station unit capability factor % 0.000 0.0 % 1 n/a
100 xEM .Coal Coal_EM_x Coal excess moisture content % 1.500 1.5 abs 1 [80]
101 JSby .Stn Tot_Sby_Load_J Total standby load GWH 5.000 5.0 % 1 [47]
102 tGen .Stn Stn_Run_Time Station running hours hrs 5.000 5.0 % 1 default
103 Cos tFO _ ton FO_Cost Fuel oil cost R/ton 0.000 0.0 % 1 n/a
104 thr .month Month_Hrs Hours for the specific month hrs 5.000 5.0 % 1 default
105 EAFStn Stn_EAF Station availability factor % 0.000 0.0 % 1 n/a
141
uLiterature
i
The uncertainty, ui  , where uLiterature is transformed to a % form by dividing by the typical variable value x 100 if given in units.
Ntransmitters
ii
The word “default” in the reference column indicates that a quantitative assessment had not been determined, a chosen default value to test the sensitivity of the
variable. Used typically for manual human errors however these should be neglected when the estimation of running hours is done through an automated electronic
process.
iii
The subscript “i” or “[i]” = 1-6 (for each unit number). If omitted from the variable name, then the property refers to a composite value for the entire station.
iv
Even though the referenced text indicates that the uncertainty for uncalibrated electrical metering can be in excess of 5% a more conservative figure of 2.5% has been
evaluated based on the authors experience.
v
ISO [78] indicates that mechanical sampling systems should be design to a 1% uncertainty on ash content which was translated to the CV using the South African ash/CV
correlation[85] to achieve an uncertainty of 0.37 MJ/kg on an as received basis. ISO [77] suggests an uncertainty of 0.3 MJ/kg on an air dried basis. The uncertainties were
combined by the RSS method.
vi
ISO [86] indicates a 0.1% uncertainty, however typical supplier test certification indicate 7.5% uncertainty in the measured CV
142
Appendix E.STEP program output tag list
Appendix E. STEP program output tag list

Table E- 1: STEP program output list
Python Tag STEP Description Unit

tag
A_EFP_RT IDT085 (A) Electric Feed Pump On Load Hours h
Cond_CW_In_T IDT003 Mean CW Inlet Temperature °C
AH_Gas_In_CIR IDB076 Carbon in Dust %
AH_Gas_In_O2_x IDB009 Mean Economiser Gas Outlet O2 %
AH_Gas_Out_CO2_x IDB079 Mean Air heater Gas Outlet CO2 %
AH_GAS_OUT_T IDB003 Mean Air Heater Gas Outlet Temperature °C
AH_LH_Gas_In_O2_x IDB007 L/H Economiser Gas Outlet O2 %
AH_LH_Gas_Out_T IDB001 L/H Air Heater Gas Outlet Temperature °C
AH_RH_Gas_In_O2_x IDB008 R/H Economiser Gas Outlet O2 %
AH_RH_Gas_Out_T IDB002 R/H Air Heater Gas Outlet Temperature °C
AT_Heat IDT053 Test Heat Consumption GWh
AT_NLToNER_Heat IDT051 Fixed Heat Consumption GWh
AuxPower_Act_Frac IDT073 Actual Works Unit On-Load %
AuxPower_CF IDT064 Works Power Correction Factor NA
AuxPower_Check_J IDT015 Actual Unit Aux Power (Check Meter) GWh
AuxPower_F IDT062 Factor for Estimating Works Power NA
B_EFP_RT IDT100 (B) Electric Feed Pump On Load Hours h
AuxPower_J IDT022 Total Works Units Used GWh
AuxPower_L_Cost IDT076 Cost of Works Units Loss R
AuxPower_Main_J IDT014 Actual Unit Aux Power (Main Meter) GWh
Auxpower_OffLoad_F IDS109 Works Units Not On-load %
AuxPower_OffLoad_J IDS156 Works Units While Not Generating MWh
AuxPower_SB_J IDT089 Works Power from Station Board GWh
AuxPower_SF_L IDT074 STEP Factor Works Units Loss %
AuxPower_Targ_Frac IDT063 Target Works Units On-Load %
AuxPower_UB_J IDT017 Mean Auxiliary Units From Unit Board GWh
C_EFP_RT IDT110 (C) Electric Feed Pump On Load Hours h
BD_Tot_Mass IDB012 Water Blowdown (SI)t
Boil_Act_L IDB054 Total Actual Boiler Loss %
Boil_ColdStarts IDB031 No. of Cold Starts NA
Boil_DownTime_0to7hrs IDB026 Total Time Off Load Type 1 Start h
Boil_Fuel_CF IDB065 Boiler Fuel Correction Factor NA
Boil_L_Cost IDB057 Total Cost of Boiler Losses R
Boil_LF IDB061 Average % Load of Practical MCR Capacity %
Boil_LF_CF IDB062 Boiler Load Correction Factor NA
Boil_OffLoad_Heat_GWH IDB069 Total Off-Load Heat GWh
143
Boil_Run_hrs IDB010 Steaming Hours h

Boil_SF_L IDB055 Total Boiler STEP Factor Loss %
Boil_Starts_0to7hrs IDB022 No. of Hot Starts Type 1 NA
Boil_Targ_L IDB053 Total Target Boiler Loss %
BP_L_Cost IDT042 Cost of Backpressure Loss R
BP_SF_L IDT040 STEP Factor Backpressure Loss %
BP_Targ IDT031 Target Condenser Backpressure kPa
BP_Targ_CF IDT058 Backpressure Correction Factor NA
Capacity_LV_Pow IPT029 Specified Turbine MCR Capacity MW
Capacity_USO_Pow IDT158 Standard Output Capacity MW
CarbonEquiv CE.U01 Carbon Equivalent of Fuel %
CIR IDB035 Mean Carbon in Ash %
CIR_Act_L IDB038 Actual Carbon in Refuse Loss %
CIR_L_Cost IDB041 Cost of Carbon in Refuse Loss R
CIR_SF_L IDB039 STEP Factor Carbon in Refuse Loss %
CIR_Targ_CF IDS151 Target Carbon in Refuse Correction Factor %
CIR_Targ_L IDB037 Target Carbon in Refuse Loss %
Coal_Abrassiveness IDS188 Abrasiveness NA
Coal_Ash_x IDS140 Ash in Coal %
Coal_BurnRate IDS173 Estimated Firing Rate kg/USO
Coal_C_x IDS148 Carbon in Coal (total) %
Coal_CV IDS146 CV of Coal as Fired (Received) MJ/kg
Coal_EM_x IDS066 Excess Moisture in Coal %
Coal_FC_x IDS142 Fixed Carbon in Coal %
Coal_H_x IDS139 Hydrogen in Coal %
Coal_Heat IDS036 Heat in Coal GWh
Coal_HGI IDS185 Hardgrove Index NA
Coal_Incr_Cost IDS144 Cost of Coal (Incremental) c/ton
Coal_Incr_PerUSO_Cost IDS157 Incremental Cost per Unit Sent Out c/kWh
Coal_Mass IDS014 Total Station Coal Burnt (SI)t
Coal_PerTon_Cost IDS169 Total Cost of Coal c/ton
Coal_PerUSO_Cost IDS160 Station Total Cost per USO c/kWh
Coal_S_x IDS143 Sulphur in Coal %
Coal_TM_x IDS138 Total Moisture in Coal %
Coal_VM_x IDS141 Volatiles in Coal %
CoarseAsh_CIR IDB077 Carbon in Rough Ash %
Cond_Act_P IDT010 Actual Condenser Backpressure kPa
Cond_TSat IDT043 Backpressure Equivalent Saturation °C
Temperature
Condensate_T IDT071 Condenser Hot well Condensate °C
Temperature
Cu_Targ_x TCU.U0 Target Unburnt Carbon in Refuse %
Cu_x CU.U01 Unburnt Carbon in Refuse %
144
Cycle_Act_Eff IDS041 Station Overall Thermal Efficiency %

Cycle_Act_HR IDS047 Actual Station Heat Rate kWh/kWh
Cycle_Direct_SF IDS042 Direct STEP Factor %
Cycle_HR_Targ IDS046 Target Station Heat Rate kWh/kWh
Cycle_Indirect_SF IDT077 Derived Unit STEP Factor %
Cycle_L_Cost IDT084 Cost of Total STEP Factor Loss R
Cycle_SF_L IDT080 Total Unit STEP Factor Loss %
Cycle_Unacc_L IDS119 Unaccounted STEP Losses %
Cycle_Unacc_L_Cost IDS121 Cost of Unaccounted Losses R
DFC DFC.U0 Dry Flue Gas per kg of Carbon Burnt NA
Dfg DFG.U0 Dry Flue Gas per kg of Fuel Burnt NA
DFG_Act_L IDB043 Actual Dry Flue Gas Loss %
DFG_L_Cost IDB046 Cost of Dry Flue Gas Loss R
DFG_SF_L IDB044 STEP Factor Dry Flue Gas Loss %
DFG_Targ_L IDB042 Target Dry Flue Gas Loss %
Eco_FW_In_T IDT011 Actual Final Feedwater Temperature °C
EFPs_Targ_J IDT087 Estimated Target Units Used by EFP, SFP GWh
O/C
Equiv_SF_L IDS116 Station STEP Factor Loss Equivalent %
Equiv_SF_L_Cost IDS118 Cost of STEP Loss Equivalent R
Est_MU_F IDT060 Factor for Estimating Make-up NA
Extr_WP_J IDS155 Total Extraneous Supply MWh
FD_In_T IDB006 Mean FD Fan Air Inlet Temperature °C
FD_LH_AIR_In_T IDB004 L/H FD Fan Air Inlet Temperature °C
FD_RH_AIR_In_T IDB005 R/H FD Fan Air Inlet Temperature °C
FO_Cost IDS176 Cost of Fuel Oil R/ton
FO_Heat IDS037 Heat in Fuel Oil GWh
FO_Mass IDS015 Total Station Fuel Oil Burnt (SI)t
FO_PerUSO_Cost IDS177 Fuel Oil Cost per USO c/kWh
Frac_Tot_Stm IDB036 Fraction of Total Steam Raised NA
Fuel_PerUSO_Cost IDS178 Total Fuel Cost per USO c/kWh
FW_L_Cost IDT047 Cost of Final Feedwater Temperature Loss R
FW_SF_L IDT045 STEP Factor Feedwater Temperature Loss %
FW_Targ_T IDS091 Average Target Final Feedwater °C
Temperature
FW_Tot_Mass IDB011 Total Feed Water To Boiler Gg
Gen_Frac IDT067 Unit Load fraction of Total Units NA
Generated
Gen_HV_J IDT018 Total Units Generated (HV) GWh
Gen_HV_J[0] IDS029 Station Total Units Generated - L.V. GWh
Gen_HV_LF IDT054 Average Load Generated % of Specified %
MCR Capacity
Gen_LV_Check_J IDT013 Actual Units Generated (Check Meter) GWh
Gen_LV_J IDT016 Mean Units Generated - L.V. GWh
Gen_LV_Main_J IDT012 Actual Units Generated (Main Meter) GWh
Gen_Reactive_J IDT020 Reactive Load Generated GVARh
145
Gen_RT IDT024 Time Generating h

Generating_Heat IDS038 Heat for Generation GWh
H2_Moist_Act_L IDB048 Actual H2 and Fuel moisture Loss %
H2_Moist_L_Cost IDB051 Cost of H2 and Fuel Moisture Loss R
H2_Moist_SF_L IDB049 STEP Factor H2 and Fuel Moisture Loss %
H2_Moist_Targ_L IDB047 Target H2 and Fuel Moisture Loss %
H2andMoist_Targ_CF IDS137 Target H2 Fuel and Moisture Correction NA
Factor
Heat_Boil_Cold_Starts IDB067 Heat for Unit Cold Starts GWh
Heat_Boil_Hot_Starts IDB066 Heat for Unit Hot starts GWh
Heat_Boil_HotSby IDB068 Heat for Unit Hot Standby GWh
Heat_Comm IDS020 Heat for I.C. Sets Calculated GWh
HotSby_Time IDB030 Time for Hot Standby h
HPT_LH_MS_In_T IDT004 Main Steam Turbine Inlet L/H Side °C
Temperature
HPT_RH_MS_In_T IDT005 Main Steam Turbine Inlet R/H Side °C
Temperature
HPT_Targ_In_T IDT029 Target Main Steam Temperature °C
HR_Targ IDT065 Running Sent Out Heat Rate kWh/kWh
Incr_Heat_Cost IDS147 Cost of Heat (incremental) c/kWh
Incr_NLtoNER_Act_HR IDT094 Actual Incremental Heat Rate kWh/kWh
IPT_Act_In_T IDT009 Actual Mean Reheat Steam Temperature °C
IPT_LH_Stm_in_T IDT007 Reheat Steam Turbine Inlet L/H Side °C
Temperature
IPT_RH_Stm_in_T IDT008 Reheat Steam Turbine Inlet R/H Side °C
Temperature
IPT_Targ_In_T IDT030 Target Reheat Steam Temperature °C
L_Corr_Ash IDB063 Correction for Fuel Ash Content NA
L_Corr_H2andMoist IDB064 Correction for H2 and Moisture Content NA
L_Corr_VM IDB078 Correction for Volatiles in Coal NA
L_Cost IDB016 Cost of Loss Factor R/%
MILL_A_Run_Hrs IDB091 Total (A) Mill On Load Hours h
MILL_B_Run_Hrs IDB092 Total (B) Mill On Load Hours h
MILL_C_Run_Hrs IDB093 Total (C) Mill On Load Hours h
MILL_D_Run_Hrs IDB094 Total (D) Mill On Load Hours h
MILL_E_Run_Hrs IDB095 Total (E) Mill On Load Hours h
MILL_F_Run_Hrs IDB096 Total (F) Mill On Load Hours h
Mill_Reject_CV IDB071 Mean CV of Mill Rejects MJ/kg
Mill_Reject_Mass IDB070 Weight of Mill Rejects (SI)t
Month_Hrs ISS001 Hours in Period h
MR_Act_L IDB058 Actual Mill Reject Loss %
MR_L_Cost IDB075 Cost of Mill Reject Loss R
MR_SF_L IDB073 STEP Factor Mill Reject Loss %
MS_Act_T IDT006 Actual Main Live Steam Temperature °C
MS_L_Cost IDT036 Cost of Main Steam Temperature Loss R
MS_SF_T_L IDT034 STEP Factor Main Steam Temperature Loss %
146
MS_Targ_T IDS072 Average Target Main Steam Temperature °C

MU_Act_L IDT069 Actual Make-up loss %
MU_CF IDT061 Make-up Correction Factor NA
MU_Cold_V IDT096 Cold Make Up Used ML
MU_Hot_F MU.U01 Unit of Make-up as a Fraction of Total NA
Steam Raised
MU_L_Cost IDT072 Cost of Make-up water Loss R
MU_MeterA IDS001 Make-Up Water Meter (A) Ml
MU_MeterB IDS002 Make-Up Water Meter (B) Ml
MU_Other_V IDS175 Station Make-Up Excluding Unit Make-Up ML
MU_PerUSO_V IDS135 Demin Water Used L/USO
MU_SF_L IDT070 STEP Factor Make-up Water Loss %
MU_Targ_L IDT068 Target Make-up Loss %
MU_V IDT079 Unit Make-Up Water Used ML
MU_V[0] IDS003 Station Meter Make-up Water Used ML
OnLoad_Targ_Heat IDT066 Target Turbine On-Load Heat GWh
Consumption
PF IDT056 Average Power Factor NA
PF_CF IDT057 Power Factor Correction Factor NA
Radiation_L IDB052 Radiation Loss %
Range_Start_Heat IDT028 Heat for Turbine Starting GWh
ReH_SF_T_L IDT037 STEP Factor Reheat Steam Temperature %
Loss
ReH_Stm_L_Cost IDT039 Cost of Reheat Steam Temperature Loss R
Reh_Targ_T IDS077 Station Target Reheat Steam Temperature °C
RW_PerUSO_V IDS134 Raw Water Used litres/USO
RW_V IDS025 Total Raw Water Used ML
Sby_LF IDS133 Standby Load Factor %
StdCoal_Ash_x IPS027 Ash in coal (Standard) %
StdCoal_C_x IPS029 Carbon in coal (Standard) %
StdCoal_CV IPS025 CV of coal (Standard) MJ/kg
StdCoal_H2_x IPS030 Hydrogen in coal (Standard) %
StdCoal_S_x IPS031 Sulphur in coal (Standard) %
StdCoal_TM_x IPS026 Moisture in coal (Standard) %
StdCoal_VM_x IPS028 Volatiles in coal (Standard) %
Stm_Mass IDB017 Total Steam Raised Gg
Stm_Mass[0] IDS044 Station Total Steam Raised Gg
Stn_AuxPower_CommonPlant_J IDS174 Station Common Aux Power Consumption GWh
Stn_Run_Time IDS034 Station Time Generating h
Stn_Tx1_CkeckMeter_J IDS005 Station Tx.1 Check Meter GWh
Stn_Tx1_Main_J IDS004 Station Tx1 Main Meter GWh
Stn_Tx2_Check_J IDS007 Station Tx 2 Check Meter GWh
Stn_Tx2_Main_J IDS006 Station Tx 2 Main Meter GWh
Stn_Tx3_Check_J IDS009 Station Tx 3 Check Meter GWh
Stn_Tx3_Main_J IDS008 Station Tx 3 Main Meter GWh
StnTx_Imp_J IDS010 Station Transformers Works Units GWh
147
Tot_Sby_Load_J IDS021 Total Standby load GWh

Tot_Targ_Heat IDT081 Total Target Heat Consumption GWh
Turb_ColdStart_Above30Hrs IDT027 Number of Turbine Cold Starts (Offload > NA
30 Hours)
Turb_Det_L_Cost IDT044 Cost of Turbine Deterioration Loss R
Turb_Est_Det_L IDT032 STEP Factor Turbine Deterioration Loss %
Turb_HotStart_Below30Hrs IDT026 Number of Turbine Hot Starts (Offload < NA
30 Hours)
Turb_L_Cost IDT050 Cost of Total Turbine Losses R
Turb_LF_CF IDT055 IDT055 Correction Factor NA
Turb_SF_L IDT048 Total Turbine STEP Factor Loss %
UCF_Stn IDS024 Generation Load Factor NA
Unacc_Previous_L IDS172 Previous Months Unaccounted STEP Loss %
USO_AP_LF IDS132 Installed Sent-out Load Factor %
USO_J IDT023 Units Sent Out - L.V. GWh
USO_LF IDT059 Average Running Load % of Max Output %
Capacity
USO_Max_Stn IDS131 Max Sustained Load Sent Out in a Specific MW
Hour of a Day
USO_Std_LF IDS179 Standard Installed Sent Out Load Factor %
WW_PerUSO_V IDS136 Station Total Waste Water L/USO
WW_V IDS027 Waste Water ML
148
Appendix F. Results
Appendix F. Results
Table F- 1: Systematic uncertainty propagation full results

MS_L_Cost 14 767.224 30 984.097 209.817 150 304.440 68 109.383 45.314

MS_SF_T_L 0.038 0.079 209.716 0.070 0.032 45.312
ReH_Stm_L_Cost 46 573.048 31 505.089 67.647 286 757.185 69 619.781 24.278
ReH_SF_T_L 0.119 0.080 67.504 0.134 0.033 24.275
BP_L_Cost 674 718.199 419 453.943 62.167 5 593 366.533 927 751.450 16.587
BP_SF_L 1.720 1.068 62.113 2.614 0.433 16.582
CW_Rise_T 14.170 7.920 55.891
MU_SF_L -0.339 0.087 25.755 0.640 0.063 9.782
MU_L_Cost -132 936.388 34 233.172 25.752 1 369 881.210 134 120.223 9.791
Radiation_L 0.005 0.001 25.397 0.006 0.000 7.584
AuxPower_L_Cost -325 313.576 78 290.221 24.066 -670 246.404 133 849.399 19.970
BP_Targ 5.416 1.270 23.441 5.322 0.514 9.651
AuxPower_SF_L -0.829 0.192 23.142 -0.313 0.063 19.966
DFG_L_Cost 224 778.035 49 812.081 22.161 2 169 978.309 116 838.251 5.384
DFG_SF_L 0.573 0.125 21.825 1.014 0.054 5.369
MU_Act_L 0.516 0.107 20.685 1.780 0.076 4.283
MU_Hot_F 0.516 0.107 20.684 0.000 0.000 0.000
FW_L_Cost 1 490 785.753 302 380.180 20.283 14 302 061.830 743 058.851 5.195
Boil_L_Cost 322 772.480 63 325.683 19.619 3 068 035.665 196 518.510 6.405
FW_SF_L 3.800 0.732 19.257 6.685 0.346 5.179
Boil_SF_L 0.823 0.154 18.728 1.434 0.092 6.388
Cycle_L_Cost 2 910 659.322 539 693.164 18.542 28 459 290.381 1 227 700.769 4.314
Cycle_SF_L 7.420 1.326 17.877 13.302 0.571 4.294
Turb_L_Cost 3 046 136.806 530 535.547 17.417 24 691 619.910 1 191 696.155 4.826
Turb_SF_L 7.765 1.297 16.708 11.541 0.555 4.809
L_Corr_H2andMoist 1.384 0.229 16.575 0.000 0.000 0.000
CIR_L_Cost 199 713.018 28 531.468 14.286 1 381 066.295 142 917.428 10.348
L_Corr_Ash 0.361 0.049 13.510 0.000 0.000 0.000
CIR_SF_L 0.509 0.065 12.724 0.645 0.067 10.334
MR_SF_L 0.000 0.000 7.370
MR_Act_L 0.000 0.000 7.370
H2_Moist_L_Cost -101 718.572 5 807.997 5.710 -483 008.940 12 108.292 2.507
FD_In_T 30.374 1.617 5.322
Boil_LF 81.869 4.175 5.099
H2_Moist_SF_L -0.259 0.013 5.055 -0.226 0.006 2.473
149
Appendix F. Results
Stm_Mass 1 113.935 55.697 5.000 6 019.247 123.377 2.050

H2_Moist_Act_L 4.644 0.228 4.909 4.672 0.229 4.902
H2_Moist_Targ_L 4.904 0.229 4.678 4.898 0.229 4.683
AuxPower_J 13.702 0.622 4.541 157.738 0.880 0.558
Coal_Mass 193 529.322 8 680.859 4.486 1 045 752.689 4 294.749 0.411
Frac_Tot_Stm 0.185 0.008 4.467
Cu_Targ_x 0.486 0.021 4.395
CIR_Act_L 1.495 0.063 4.188 1.615 0.066 4.086
AH_Gas_In_O2_x 3.640 0.150 4.121 3.664 0.062 1.681
CIR_Targ_CF 1.429 0.058 4.059
AH_Gas_Out_O2_x 5.640 0.212 3.761
Cu_x 0.908 0.032 3.525
Coal_H2_x 3.293 0.099 3.007 3.293 0.099 3.007
Dfg 10.919 0.323 2.959 0.000 0.000 0.000
Turb_Det_L_Cost 819 292.583 22 857.333 2.790 4 359 129.921 17 867.034 0.410
L_Cost 392 288.889 10 943.851 2.790 2 139 534.043 8 764.325 0.410
DFG_Act_L 4.543 0.124 2.722 4.971 0.056 1.132
CarbonEquiv 52.099 1.402 2.691
Boil_Act_L 10.687 0.272 2.546 11.265 0.246 2.183
Boil_Targ_L 9.864 0.225 2.277 9.831 0.222 2.262
MU_Targ_L 0.931 0.019 2.083 0.996 0.008 0.811
Est_MU_F 1.162 0.024 2.067
USO_LF 84.560 1.572 1.859 72.690 0.859 1.181
OnLoad_Targ_Heat 1 017.657 18.331 1.801
Tot_Targ_Heat 1 017.657 18.331 1.801 5 424.765 41.222 0.760
Gen_LV_LF 84.042 1.456 1.732
Gen_HV_LF 84.062 1.456 1.732 81.166 0.616 0.759
Turb_Targ_Heat 886.429 15.026 1.695 4 877.733 33.646 0.690
CIR_Targ_L 0.986 0.017 1.679 0.970 0.011 1.084
USO_J 360.040 5.641 1.567 1 898.263 11.929 0.628
AH_Gas_Out_CO2_x 12.218 0.176 1.440
Cycle_Indirect_SF 92.580 1.326 1.433 86.698 0.571 0.659
AuxPower_UB_J 21.190 0.300 1.414
Gen_LV_J 375.157 5.306 1.414
Gen_HV_J 373.741 5.284 1.414 2 056.001 11.896 0.579
DFC 20.585 0.283 1.374
AT_Heat 890.866 11.767 1.321
AT_NLToNER_Heat 890.866 11.767 1.321
Gen_Frac 0.182 0.002 1.268
Cycle_HR_Targ 2.827 0.032 1.116 2.836 0.015 0.516
HR_Targ 2.827 0.032 1.116
Targ_Turb_HR 8 538.378 91.766 1.075
Turb_Targ_Eff 42.163 0.452 1.071
BP_Targ_CF 0.997 0.011 1.068
Heat_Boil_HotSby 0.000 0.000 1.000
150
Appendix F. Results
Heat_Boil_Cold_Starts 0.000 0.000 1.000

CIR 3.055 0.031 1.000 3.291 0.015 0.456
AH_GAS_OUT_T 115.029 1.143 0.994
DFG_Targ_L 3.970 0.037 0.942 3.956 0.031 0.779
Incr_NLtoNER_Act_HR 2.816 0.024 0.864 2.970 0.015 0.520
AuxPower_F 1.048 0.009 0.823
AuxPower_Targ_Frac 4.635 0.038 0.823 7.879 0.027 0.340
Boil_OffLoad_Heat_GWH 0.000 0.000 0.736 4.983 0.023 0.455
IPT_Act_In_T 530.898 2.800 0.527 530.359 1.146 0.216
MS_Act_T 533.595 2.800 0.525 532.417 1.146 0.215
Heat_Boil_Hot_Starts 0.000 0.000 0.481
FW_Targ_T 239.365 0.831 0.347 238.775 0.338 0.142
Boil_Act_Indirect_Eff 89.313 0.272 0.305 88.735 0.246 0.277
Boil_Fuel_CF 1.020 0.003 0.271 1.020 0.003 0.271
Boil_Targ_Indirect_Eff 90.136 0.225 0.249
Turb_LF_CF 1.021 0.001 0.119
BP_Act_CF 101.432 0.088 0.087
H2_Moist_LF_Targ_L 3.520 0.001 0.037
AuxPower_CF 1.002 0.000 0.036 1.004 0.000 0.018
PF 0.994 0.000 0.031
MU_Targ_CF 1.001 0.000 0.016
PF_CF 0.998 0.000 0.004
Turb_Est_Det_L 2.088 0.000 0.004 2.037 0.000 0.014
Boil_LF_CF 1.003 0.000 0.001 1.003 0.000 0.011
AuxPower_Min_Frac 4.421 0.000 0.000
Cycle_Unacc_L_Cost -8 105 289.618 4 767 717.845 58.822
Cycle_Adj_HR 11 369.525 269.469 2.370
Auxpower_OffLoad_F 0.776 0.005 0.624
MU_PerUSO_V 0.055 0.002 3.591
Equiv_SF_L 9.513 2.210 23.236
Generating_Heat 5 995.096 137.004 2.285
MS_Targ_T 535.000 0.000 0.000
Equiv_SF_L_Cost 20 354 000.762 4 741 859.072 23.297
EFPs_Targ_J 57.534 0.597 1.037
Fuel_PerUSO_Cost 21.068 0.140 0.665
Coal_BurnRate 0.551 0.004 0.751
USO_AP_LF 0.733 0.009 1.179
Reh_Targ_T 535.000 0.000 0.000
RW_PerUSO_V 1.211 0.061 5.039
Cycle_Adj_Eff 31.664 0.751 2.371
Cycle_Direct_SF 90.487 2.210 2.443
Cycle_Acc_L 13.302 0.571 4.294
FO_Heat 30.781 1.623 5.274
Cycle_Unacc_L -3.788 2.230 58.878
Cycle_Act_Eff 31.911 0.756 2.370
151
Appendix G. Model verification

The table below is a verification of the Python STEP computer model against the Mathcad STEP
model. The key deviations in the results can be attributed to the number of significant digits used
in the baseline performance data polynomial coefficients. These coefficients were hardcoded into
the MathCAD STEP model with up to 5 significant digits. The Python STEP computer model was
coded to read in the data as per the source data. The deviations were scrutinized for each
calculation to ensure that it indeed stemmed from the polynomial coefficients or carried over
from previous calculations. The deviation listed below was considered to be negligible.
Table G- 1: STEP model verification
Calculation
No. Variable [Unit No.] MathCAD Python Difference
1 AH_GAS_OUT_T[1] 130.50000 130.50000 0.00000
2 FD_In_T[1] 35.50000 35.50000 0.00000
3 AH_Gas_In_O2_x[1] 4.35000 4.35000 0.00000
4 AH_LH_Leak[1] 5.76920 5.76923 0.00003
5 AH_RH_Leak[1] 3.70370 3.70370 0.00000
6 Stm_Mass[1] 1 450.00000 1 450.00000 0.00000
7 Boil_LF[1] 92.93206 92.93206 0.00000
8 Boil_LF_CF[1] 1.00050 1.00053 0.00003
9 CIR_Targ_CF[1] 0.96829 0.96829 0.00000
10 L_Corr_Ash[1] -0.03942 -0.03942 0.00000
12 Coal_H2_x[1] 3.09715 3.09715 0.00000
13 L_Corr_H2andMoist[1] 0.17021 0.17021 0.00000
14 L_Corr_H2andMoist[1] 0.17021 0.17021 0.00000
16 L_Corr_VM[1] 0.00000 0.00000 0.00000
17 L_Corr_VM[1] 0.00000 0.00000 0.00000
18 Boil_Fuel_CF[1] 1.00147 1.00147 0.00000
23 Heat_Boil_Hot_Starts[1] 1.09043 1.09043 0.00000
24 Heat_Boil_Cold_Starts[1] 0.46440 0.46440 0.00000
25 Heat_Boil_HotSby[1] 0.36133 0.36133 0.00000
26 Boil_OffLoad_Heat_GWH[1] 1.91616 1.91616 0.00000
27 CIR[1] 1.65964 1.65964 0.00000
31 AH_Gas_Out_CO2_x[1] 12.41016 12.41016 0.00000
32 DFG_AT_L[1] 4.97287 4.97287 0.00000
33 Cu[1] 0.52199 0.52199 0.00000
34 CIR_Targ_L[1] 1.21408 1.21408 0.00000
35 Tcu[1] 0.69662 0.69662 0.00000
36 DFG_Targ_L[1] 5.04333 5.04333 0.00000
38 DFC[1] 20.27211 20.27211 0.00000
39 Dfg[1] 9.45542 9.45542 0.00000
155
40 DFG_Act_L[1] 4.82614 4.82614 0.00000

41 DFG_SF_L[1] -0.21719 -0.21719 0.00000
42 CIR_Act_L[1] 0.94003 0.94003 0.00000
43 CIR_SF_L[1] -0.27405 -0.27405 0.00000
44 H2_Moist_LF_Targ_L[1] 4.42900 4.42900 0.00000
45 H2_Moist_Targ_L[1] 4.59921 4.59921 0.00000
46 H2_Moist_Act_L[1] 5.01700 5.01700 0.00000
47 H2_Moist_SF_L[1] 0.41779 0.41779 0.00000
48 Stm_Mass[0] 8 700.00000 8 700.00000 0.00000
49 Coal_Mass[1] 282 981.00000 282 981.00000 0.00000
51 Radiation_L[1] 0.34663 0.34663 0.00000
52 MR_Act_L[1] 0.00085 0.00085 0.00000
53 MR_SF_L[1] 0.00085 0.00085 0.00000
54 Boil_Targ_L[1] 10.86098 10.86034 -0.00064
55 Boil_Act_L[1] 10.78838 10.78774 -0.00064
56 Boil_SF_L[1] -0.07260 -0.07260 0.00000
57 Frac_Tot_Stm[1] 0.16667 0.16667 0.00000
58 CIR_Targ_L[0] 1.21408 1.21408 0.00000
59 CIR_Act_L[0] 0.94003 0.94003 0.00000
60 CIR_SF_L[0] -0.27405 -0.27405 0.00000
61 DFG_Targ_L[0] 5.04333 5.04333 0.00000
62 DFG_Act_L[0] 4.82614 4.82614 0.00000
63 DFG_SF_L[0] -0.21719 -0.21719 0.00000
64 H2_Moist_Targ_L[0] 4.59921 4.59921 0.00000
65 H2_Moist_Act_L[0] 5.01700 5.01700 0.00000
66 H2_Moist_SF_L[0] 0.41779 0.41779 0.00000
67 Radiation_L[0] 0.00437 0.00373 -0.00064
68 MR_Act_L[0] 0.00085 0.00085 0.00000
69 Boil_Targ_L[0] 10.86098 10.86034 -0.00064
70 Boil_Act_L[0] 10.78754 10.78774 0.00020
71 Boil_SF_L[0] -0.07345 -0.07260 0.00085
72 Boil_OffLoad_Heat_GWH[0] 11.49698 11.49698 0.00000
73 Boil_LF_CF[0] 1.00053 1.00053 0.00000
74 Boil_Fuel_CF[0] 1.00147 1.00147 0.00000
75 CIR[0] 1.65960 1.65964 0.00004
76 AH_Gas_In_O2_x[0] 4.35000 4.35000 0.00000
77 AH_RH_Leak[0] 3.70370 3.70370 0.00000
78 AH_LH_Leak[0] 5.76923 5.76923 0.00000
79 Gen_LV_J[1] 510.00000 510.00000 0.00000
80 AuxPower_UB_J[1] 62.00000 62.00000 0.00000
81 Gen_HV_J[1] 508.20800 508.20800 0.00000
82 Gen_HV_J[0] 3 049.24800 3 049.24800 0.00000
84 AuxPower_J[0] 401.99400 401.99400 0.00000
85 USO_J[0] 2 647.25400 2 647.25400 0.00000
87 Gen_Frac[1] 0.16667 0.16667 0.00000
156
88 AuxPower_J[1] 66.61567 66.61567 0.00000

89 USO_J[1] 441.59233 441.59233 0.00000
92 AT_NLToNER_Heat[1] 1 085.69877 1 085.69877 0.00000
97 AT_Heat[1] 1 091.12323 1 091.12323 0.00000
98 Amb_CF[1] 0.99398 0.99399 0.00001
99 Gen_HV_LF[1] 96.66554 96.66530 -0.00024
100 PF[1] 99.87922 99.87922 0.00000
105 PF_CF[1] 0.99752 0.99752 0.00000
106 EFP_Est_Pow[1] 25 099.40390 25 099.53210 0.12820
107 EFPs_Targ_J[1] 16.66575 16.66584 0.00009
108 Turb_LF_CF[1] 1.00361 1.00361 0.00000
109 AuxPower_ACC_Min_Frac[1] 12.28572 12.28567 -0.00005
110 USO_LF[1] 94.31341 94.31374 0.00033
112 USO_LF[0] 94.31341 94.31374 0.00033
113 AuxPower_F[1] 1.02241 1.02240 -0.00001
114 AuxPower_Targ_Frac[1] 16.33499 16.33495 -0.00004
115 AuxPower_CF[1] 1.00245 1.00245 0.00000
116 Est_MU_F[1] 1.02851 1.02851 0.00000
126 BP_Targ[1] 12.30903 12.30909 0.00006
131 BP_Targ_CF[1] 1.00020 1.00021 0.00001
132 MU_Targ_CF[1] 1.00024 1.00024 0.00000
133 HR_Targ[1] 2.74033 2.74035 0.00002
134 OnLoad_Targ_Heat[1] 1 210.11755 1 210.11773 0.00018
135 Range_Start_Heat[1] 0.42249 0.42249 0.00000
136 Tot_Targ_Heat[1] 1 212.45621 1 212.45638 0.00017
137 MS_Act_T[1] 562.00000 562.00000 0.00000
138 IPT_Act_In_T[1] 572.00000 572.00000 0.00000
139 Turb_Est_Det_L[1] 0.00759 0.00759 0.00000
140 FW_Targ_T[1] 265.53095 265.53116 0.00021
141 AuxPower_Act_Frac[1] 15.08533 15.08533 0.00000
142 AuxPower_SF_L[1] -1.24966 -1.24962 0.00004
144 FW_SF_L[1] 0.00682 0.00682 0.00000
146 MS_SF_T_L[1] -0.05571 -0.05571 0.00000
148 ReH_SF_T_L[1] -0.06003 -0.06003 0.00000
153 BP_Act_CF[1] 99.99388 99.99388 0.00000
154 BP_SF_L[1] -0.02650 -0.02724 -0.00074
155 Turb_SF_L[1] -0.12783 -0.12856 -0.00073
156 MU_Targ_L[1] 1.02876 1.02875 -0.00001
158 MU_PerUSO_V[0] 0.04085 0.04085 0.00000
160 MU_Hot_F[1] 0.89802 0.89802 0.00000
161 MU_Act_L[1] 0.89802 0.89802 0.00000
163 MU_SF_L[1] -0.10982 -0.10982 0.00000
164 Cycle_SF_L[1] -1.55992 -1.56060 -0.00068
165 Cycle_Indirect_SF[1] 101.55992 101.56060 0.00068
167 Gen_LV_LF[1] 96.61774 96.61811 0.00037
157
173 Targ_Turb_HR[1] 7 756.92498 7 757.01250 0.08752

179 MS_BP_LF_CF[1] 318 971.27120 318 972.50933 1.23813
180 MS_Act_h[1] 3 380.74358 3 380.78186 0.03828
181 ReH_Act_Out_h[1] 3 596.80526 3 597.15783 0.35257
182 FW_Act_h[1] 1 162.53340 1 162.43766 -0.09574
183 HPT_Targ_Out_P[1] 53.02903 53.03444 0.00541
184 HPT_Targ_Out_T[1] 335.13283 335.16984 0.03701
185 HPT_Act_Out_h[1] 3 020.55539 3 020.55677 0.00138
186 SW_Targ_T[1] 183.98667 183.95927 -0.02740
187 SW_Act_h[1] 779.87254 779.85319 -0.01935
189 CEPs_Est_J[1] 0.64974 0.64974 0.00000
190 Pumps_J[1] 17.31549 17.31558 0.00009
191 Turb_Act_HR[1] 6 717.79854 6 719.22020 1.42166
192 Turb_SF_HR[1] -1 039.12644 -1 037.79231 1.33413
193 Turb_Targ_Eff[1] 46.41014 46.40962 -0.00052
194 Turb_Act_Eff[1] 53.58898 53.57765 -0.01133
195 Turb_SF_Eff[1] 7.17884 7.16803 -0.01081
201 ACC_HeatRejected[1] 971.58625 971.48644 -0.09981
202 Incr_NLtoNER_Act_HR[1] 2.58083 2.58338 0.00255
203 Incr_NERtoMCR_Act_HR[1] 2.58083 2.58338 0.00255
204 MS_Targ_T[0] 560.00000 560.00000 0.00000
205 Reh_Targ_T[0] 570.00000 570.00000 0.00000
206 MS_Act_T[0] 562.00000 562.00000 0.00000
207 IPT_Act_In_T[0] 572.00000 572.00000 0.00000
208 MS_SF_T_L[0] -0.05571 -0.05571 0.00000
209 ReH_SF_T_L[0] -0.06003 -0.06003 0.00000
210 BP_Targ[0] 12.30903 12.30909 0.00006
211 Cond_Act_P[0] 13.50000 13.50000 0.00000
212 BP_SF_L[0] -0.02724 -0.02724 0.00000
213 Turb_Est_Det_L[0] 0.00759 0.00759 0.00000
214 FW_Targ_T[0] 265.53095 265.53116 0.00021
215 Eco_FW_In_T[0] 263.00000 263.00000 0.00000
216 FW_SF_L[0] 0.00682 0.00682 0.00000
217 Turb_SF_L[0] -0.12783 -0.12856 -0.00073
218 MU_Targ_L[0] 1.02875 1.02875 0.00000
219 MU_Act_L[0] 0.89802 0.89802 0.00000
221 MU_SF_L[0] -0.10982 -0.10982 0.00000
223 AuxPower_Targ_Frac[0] 16.33499 16.33495 -0.00004
224 AuxPower_SF_L[0] -1.24966 -1.24962 0.00004
225 Incr_NLtoNER_Act_HR[0] 2.58083 2.58338 0.00255
226 Incr_NERtoMCR_Act_HR[0] 2.58083 2.58338 0.00255
228 L_Cost[1] 1 052 702.40510 1 052 702.40512 0.00002
229 MS_L_Cost[1] -58 642.05830 -58 642.08862 -0.03032
230 ReH_Stm_L_Cost[1] -63 190.21575 -63 190.25988 -0.04413
158
231 BP_L_Cost[1] -28 675.61352 -28 674.73422 0.87930

232 Turb_Det_L_Cost[1] 7 989.89553 7 989.84301 -0.05252
233 FW_L_Cost[1] 7 178.20440 7 179.06421 0.85981
234 Turb_L_Cost[1] -135 339.88734 -135 338.17549 1.71185
235 MU_L_Cost[1] -115 610.07628 -115 608.12876 1.94752
236 AuxPower_L_Cost[1] -1 315 530.56778 -1 315 474.33942 56.22836
237 CIR_L_Cost[1] -288 497.69595 -288 495.79255 1.90340
238 DFG_L_Cost[1] -228 633.89873 -228 632.39306 1.50567
239 H2_Moist_L_Cost[1] 439 813.90237 439 811.01132 -2.89105
240 Boil_L_Cost[1] -77 317.68249 -77 317.17430 0.50819
241 Boil_Act_Indirect_Eff[1] 89.21162 89.21226 0.00064
242 Boil_Targ_Indirect_Eff[1] 89.13902 89.13966 0.00064
243 Boil_Var_Indirect_Eff[1] 0.07260 0.07260 0.00000
245 Cycle_L_Cost[1] -1 643 737.81798 -1 643 737.81798 0.00000
246 EFPs_Targ_J[0] 99.99452 99.99503 0.00051
247 Tot_Targ_Heat[0] 7 274.73724 7 274.73831 0.00107
248 Cond_TSat[1] 52.02536 52.02536 0.00000
249 Cond_DepresT[1] -1.84536 -1.84536 0.00000
250 ACC_Rise_T[1] 28.00000 28.00000 0.00000
251 Capacity_J[0] 3 167 108.03549 3 167 108.03549 0.00000
252 Coal_Heat[0] 8 992.18812 8 992.18812 0.00000
253 FO_Heat[0] 6.25000 6.25000 0.00000
254 Generating_Heat[0] 8 998.43812 8 998.43812 0.00000
255 Cycle_Act_HR[0] 3.39621 3.39621 0.00000
256 Cycle_Act_Eff[0] 29.44460 29.44460 0.00000
257 Cycle_HR_Targ[0] 2.74565 2.74565 0.00000
258 Cycle_Direct_SF[0] 80.84444 80.84446 0.00002
259 Cycle_SF_L[0] -1.56076 -1.56060 0.00016
260 Cycle_Indirect_SF[0] 101.56076 101.56060 -0.00016
261 Equiv_SF_L[0] 19.15556 19.15554 -0.00002
262 Cycle_Unacc_L[0] 20.71632 20.71614 -0.00018
265 Fuel_PerUSO_Cost[0] 29.74885 29.74885 0.00000
269 Coal_BurnRate[0] 0.51055 0.51055 0.00000
270 USO_Std_LF[0] 86.24015 86.24015 0.00000
272 AuxPower_CF[0] 1.00242 1.00176 -0.00066
273 Sby_LF[0] 88.24180 88.24180 0.00000
274 USO_AP_LF[0] 2.38819 2.38819 0.00000
275 RW_PerUSO_V[0] 0.07114 0.07114 0.00000
276 WW_PerUSO_V[0] 0.00000 0.00000 0.00000
277 L_Cost[0] 7 934 723.82235 7 934 724.98922 1.16687
278 CIR_L_Cost[0] -2 174 531.40765 -2 174 531.72265 -0.31500
279 DFG_L_Cost[0] -1 723 312.19849 -1 723 312.45162 -0.25313
280 H2_Moist_L_Cost[0] 3 315 066.90983 3 315 067.39708 0.48725
281 Boil_L_Cost[0] -582 776.69631 -582 776.77719 -0.08088
282 MS_L_Cost[0] -442 013.36932 -442 013.66284 -0.29352
159
283 Turb_Det_L_Cost[0] 60 223.28571 60 223.29456 0.00885

284 BP_L_Cost[0] -216 141.87692 -216 135.28100 6.59592
285 ReH_Stm_L_Cost[0] -476 295.01732 -476 295.41999 -0.40267
286 FW_L_Cost[0] 54 105.21985 54 112.06424 6.84439
287 Turb_L_Cost[0] -1 020 121.75800 -1 020 109.00503 12.75297
288 MU_L_Cost[0] -871 401.45957 -871 394.14120 7.31837
289 AuxPower_L_Cost[0] -9 915 389.58289 -9 915 363.60410 25.97879
290 Cycle_L_Cost[0] -12 389 689.49677 -12 389 643.52751 45.96926
291 Equiv_SF_L_Cost[0] 151 994 049.23174 151 993 977.24915 -71.98259
292 Cycle_Unacc_L_Cost[0] 164 383 738.72851 164 383 620.77666 -117.95185
293 Auxpower_OffLoad_F[0] 0.08681 0.08681 0.00000
295 Coal_MF_CV[0] 20.52459 20.52459 0.00000
297 Cycle_Adj_HR[0] 12 236.36917 12 236.97357 0.60440
298 Cycle_Adj_Eff[0] 29.42049 29.41904 -0.00145
299 Boil_Act_Indirect_Eff[0] 89.21246 89.21226 -0.00020
300 Boil_SF_L[0] -0.07345 -0.07260 0.00085
301 Coal_BurnRate[0] 0.64138 0.64138 0.00000
302 Turb_Targ_Heat[0] 6 531.85561 6 531.90338 0.04777
303 Turb_Act_Eff[0] 46.74240 46.74240 0.00000
304 MU_Stm_Ratio[0] 1.24285 1.24285 0.00000
306 Sby_LF[0] 88.24180 88.24180 0.00000
307 Max_Demand_LF[0] 122.55806 122.55806 0.00000
308 Running_LF[0] 96.27825 96.27862 0.00037
309 USO_LF[0] 86.16529 86.16529 0.00000
310 EUF[0] 97.91510 97.91510 0.00000
160

University of Cape Town: Power Station Thermal Efficiency Performance Method Evaluation

Uploaded by

Copyright:

Available Formats

University of Cape Town: Power Station Thermal Efficiency Performance Method Evaluation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

University of Cape Town: Power Station Thermal Efficiency Performance Method Evaluation

Uploaded by

Copyright:

Available Formats

Power Station Thermal Efficiency

Performance Method Evaluation

Department of Mechanical Engineering

University of Cape Town

of the non-exclusive license granted to UCT by the author.

Figure 3-1: World projected energy mix [7] ........................................................................................ 7

Figure 3-2: World projected electricity generation mix [12] ............................................................... 8

Figure 4-1: Illustration of STEP Factors .............................................................................................. 20

Figure 4-2: STEP calculation architecture .......................................................................................... 20

Figure 4-3: STEP boiler calculation sequence .................................................................................... 21

Figure 4-4: STEP turbine calculation sequence .................................................................................. 22

Figure 4-5: STEP cycle calculation sequence...................................................................................... 23

Figure 4-6: Development of the Mathcad STEP model ..................................................................... 24

Figure 4-7: Power plant data flow ..................................................................................................... 26

Figure 4-8: Example of processed signals/tags .................................................................................. 28

Figure 4-9: Distribution of error on repeated measurement [47] ..................................................... 30

Figure 4-10: Distribution of error as a function of time [16] ............................................................. 31

Figure 4-11: Gaussian distribution probability density function [47] ................................................ 33

Figure 4-12: Uncertainty propagation model architecture ............................................................... 36

Figure 4-13: Step 2 - Example of monthly composite input data ...................................................... 38

Figure 4-14: Step 3 - Example of base results .................................................................................... 39

Figure 5-1: Turbine deterioration loss comparison ........................................................................... 46

Figure 5-2: Instrument uncertainty on SF losses ............................................................................... 51

Figure 5-8: Generated power profile by resolution ........................................................................... 56

Figure 5-10: Station STEP Factor uncertainty by resolution .............................................................. 60

Figure 5-11: Unit 4 segmented SF loss evaluation ............................................................................ 60

Figure 5-12: Unit 5 segmented SF loss evaluation ............................................................................. 61

Table 3-2: STEP loss categories .......................................................................................................... 13

Table 4-1: Case study - power plant description ............................................................................... 18

Table 5-1: Summary of loss review .................................................................................................... 42

Table 5-3: Truncated results greater 5% – instrument uncertainty propagation ............................. 48

Table D- 1: Input variable instrument uncertainty .......................................................................... 137

Table E- 1: STEP program output list ............................................................................................... 143

Table F- 1: Systematic uncertainty propagation full results ............................................................ 149

Table G- 1: STEP model verification ................................................................................................. 155

1.2. Acronyms and abbreviations

Figure 2-1: Diagram of a typical coal-fired thermal power station [9]

Table 2-1: Legend to Figure 2-1

No. Description No. Description No. Description

Thermal power plants

Period Technology changes Thermal

World energy outlook

The permanence of coal-fired power generations technologies is threatened by the penetration of

Figure 3-1: World projected energy mix [7]

Figure 3-2: World projected electricity generation mix [12]

Figure 3-3: South African overall efficiency vs availability [13]

Thermal efficiency performance methodology

Turbine plant efficiency

Fundamental principles of STEP

Cycle Boiler Turbine

Law of Error Propagation

The equation (3.5) is used for uncorrelated inputs:

The equation (3.6) is used for correlated inputs:

xi - each variable in a function

x j = xi 1 - sequential pair of variables

uxi - combined uncertainty of variable for a single measurement point

u( xi , x j ) - estimated covariance of associated with xi and x j

uk -each type of uncertainty (systematic, random-temporal, random-resolution…)

Sequential Perturbation (SP)