See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/305452638

Automatic Generation Control of Thermal Power System under Varying Steam

Turbine Dynamic Model Parameters Based on Generation Schedules of the
Plants

Article  in  The Journal of Engineering · July 2016

DOI: 10.1049/joe.2016.0178

CITATIONS READS

22 5,321

3 authors:

Nikhil Pathak T. S. Bhatti

The University of Queensland Indian Institute of Technology Delhi
27 PUBLICATIONS   230 CITATIONS    173 PUBLICATIONS   3,782 CITATIONS   

SEE PROFILE SEE PROFILE

Ashu Verma
Indian Institute of Technology Delhi
116 PUBLICATIONS   1,348 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Stability analysis and optimisation of AC-DC power systems View project

Transmission Expansion Planning View project

All content following this page was uploaded by Nikhil Pathak on 09 April 2019.

The user has requested enhancement of the downloaded file.

Automatic generation control of thermal power system under varying steam turbine
dynamic model parameters based on generation schedules of the plants
Nikhil Pathak, Ashu Verma, Terlochan Singh Bhatti
Centre for Energy Studies, Indian Institute of Technology Delhi, New Delhi 110016, India
E-mail: [email protected] 24 6 2016

Published in The Journal of Engineering; Received on 13th June 2016; Revised on 24th June 2016; Accepted on 29th June 2016

Abstract: This paper investigates the automatic generation control (AGC) of thermal power system operating under different power generation
levels/schedules of the plants. The power system load varies considerably over the period of 24 hours and accordingly thermal power plants are
planned to operate at different generation schedules. It has been discovered that the steam turbine dynamic model parameters also changes
along with generation schedules of the plant. Typical steam turbine model parameters are the turbine time constants and power fractions.
The literature survey reveals that steam turbine dynamic model parameters are assumed to be constant in all preceding studies pertaining
to AGC systems, irrespective of generation schedules of the plant. In this paper, steam turbine dynamic model parameters have been computed
for an actual 500 MW thermal unit at different generation schedules. These steam turbine dynamic model parameters have been used to study
the dynamic performance of AGC system. The paper also presents the general mathematical procedure to calculate steam turbine dynamic
model parameters from an actual generating unit using the heat balance data of a thermal power plant. Results for two area thermal power
system demonstrate the potential of the proposed approach.

Nomenclature and system frequency to their nominal values. The AGC system is
divided into two control loops named as primary and secondary
List of symbols
control loops. Primary control loop chases the power system load
continuously using speed-governing action. However, the system
i and j index number referring to the ith and the jth frequency settles down at a new equilibrium point. The secondary
areas (where i and j = 1, 2, …) control loop helps to restore the system frequency to its initial
Δfi deviation in frequency of the ith area (Hz) value. The preliminary studies on AGC system were conducted
ΔPtieij deviation in tie-line power (pu) on tie-line power and frequency control of electric power system.
Di load damping constant of the ith area (MW The dynamic performance of AGC system was investigated under
pu/Hz) the variation of different system parameters: namely, tie-line
Kpi 1/D power system rotating mass gain of the power controller gains, frequency controller gains, speed regulation
ith area (Hz/MW pu) parameter, changing tie-line capacity [1, 2] etc. The detailed math-
Tpi power system time constant in the ith area (s) ematical modelling of interconnected AGC system was presented
ACEi area control error signal in the ith area and classical control theory had been used first time to optimise
Pri rated capacity of the ith area (in MW) the controller gains [3]. Later, state space model of AGC system
Rieqv equivalent speed droop characteristic of was also developed and optimal control theory had been utilised
generating unit in the ith area (Hz/MW pu) to obtain the optimum gain settings [4]. The studies were also con-
i
Heqv equivalent inertia constant of all coherent ducted on multi-areas interconnected thermal power systems in the
parallel operating unit of the ith area (s) presence of governor dead-band effects, generation rate constraints
FHPi, FIPi and power fractions of high-, intermediate- and [5, 6]. Furthermore, the study was carried out in discrete mode of
FLPi low-pressure turbines of the ith area AGC system. The secondary control loop consists of supplementary
TCHi, TRHi and time constants of steam chest, reheater and controllers which are updated with the power system data informa-
TCOi cross-over piping of the ith area tion in every 2 s time interval [7–9]. Supplementary controller
aij area’s capacity ratio (−Pri /Prj ) chases the area control error signal continuously and tries to dimin-
Bi frequency bias constant of the ith area (in pu ish it. However, this leads to continuous movement of governor
MW/Hz) control valve which eventually results to excessive wear and tear
ΔPdi step load perturbation of the ith area (in pu) of steam turbines. To avoid this difficulty, AGC controller is oper-
Gnon-reheat (s) equivalent transfer function of non-reheat ated in discrete mode. The control signals are fed to the controllers
turbine after a fixed interval of time known as sampling period of the con-
Greheat (s) equivalent transfer function of reheat turbine troller. Various intelligent control techniques have been developed
to optimise the supplementary controller gains [10–16]. The studies
are further extended to the thermal power systems operating at dif-
1 Introduction
ferent generation schedules of the plants. The power system load
To operate the system economically and reliably, the neighbouring changes considerably over the period of 24 hrs and accordingly
power system areas are interconnected through tie-lines. The power the thermal power plants are scheduled to operate at different gen-
exchanges between these areas are frequently scheduled on contract eration levels. The literature survey shows that the AGC power
basis via tie-lines named as ‘prescheduled interchanges’. The load system model parameters which are considered to be changed
disturbance in these areas causes the deviations in prescheduled along with the generation schedules of the plants are the power
interchanges and system frequency. The automatic generation system rotating mass time constant Tp, power system gain constant
control (AGC) is used to maintain these prescheduled interchanges Kp and bias factor Bi [17–21]. However, it has been found that these

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
Fig. 1 Heat balance diagram of 500 MW unit of thermal power plant for rated power output condition

system parameters are barely able to affect the power system inferences drawn in the preceding AGC studies, assuming the steam
dynamic response at different generation schedules of the plant. turbine dynamic model parameters constant are in doubt. Also, the
Typically, the generation level of a plant is controlled using one optimum gain settings of the controllers obtained assuming the
of the following control strategies: constant pressure mode, steam turbine dynamic model parameters constant may not be ac-
sliding pressure mode or hybrid mode [22–25]. Mostly hybrid ceptable at different generation schedules of the plant.
mode, which is the combination of constant pressure and sliding In this paper, steam turbine dynamic model parameters have been
pressure mode is adopted. The steam turbine dynamic model para- computed for an actual thermal unit of 500 MW capacity, at differ-
meters depend on the type of control strategy used to change the ent power generation schedules of power plant. These computed
generation level of the plant. In the subsequent AGC studies, differ- steam turbine dynamic model parameters have been used to study
ent kinds of steam turbine dynamic models are used depending on the dynamic performance of AGC system under varying generating
the type of steam turbines being employed in thermal power plants. schedule condition. This paper also presents the general mathemat-
The IEEE committee taskforce has developed a variety of approxi- ical procedure to extract the steam turbine dynamic model para-
mate linear models for all kinds of steam turbines available in meters from an actual power generating unit. From the literature
thermal power plants [26]. The main parameters of steam turbine review, it is observed that there is no mathematical procedure pre-
dynamic models which describe the dynamics of steam turbines sented yet to compute the steam turbine dynamic model parameters.
are the time constants of steam chest (SC) TSC, reheater (RH) Hence, in this paper the mathematical procedure is developed to
TRH, cross-over (CO) TCO and power fractions FHP, FIP and FLP. compute these parameters using heat balance data of thermal
The literature survey shows that either primary/secondary control power plant.
loops are the main focus areas of the researchers for the AGC dy-
namics study. Surprisingly, no one has paid attention on the chan-
2 Thermal power plant cycle with heat balance data
ging dynamic behaviour of steam turbine dynamic model
parameters along with the generation schedule of thermal power The complete steam cycle of a thermal generating unit of 500 MW
plant. The previous AGC studies reveal that steam turbine model capacity has been presented with its heat balance data for rated
parameters are assumed to be fixed irrespective of generation sche- output condition as shown in Fig. 1. The heat balance data for
dules of the plant. However, it is discovered that steam turbine partial generation levels have been given in Tables 9 and 10 of
dynamic model parameters also change as the generation schedules Appendix 2. The main components of steam units of a thermal
drifts from its nominal value. Therefore, the dynamic responses and power plant are RH, boiler, condensers, turbine sections, feed

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
 Fig. 3 Steam turbine dynamic model of
a Non-reheat steam turbine
b Reheat tandem compound steam turbine

Fig. 2 Steam turbine configurations

a Non-reheat turbine pipe and the power fractions FHP, FIP and FLP of HP, IP and
b Tandem compound single reheat turbine LPTs, respectively. The transfer function models of non-reheat
and single reheat steam turbines derived from Figs. 3a and b are
as follows.
water heaters, feed water pumps (FWPs) etc. The steam cycle con-
Equivalent transfer function model of non-reheat turbine
sists of two high-pressure heaters (HPH) for high pressure turbine
(HPT) denoted by HPH1 and HPH2, five low pressure heaters
1
(LPH) for intermediate and lower-pressure turbines denoted by Gnon-reheat (s) = (1)
LPH1, LPH2, LPH3, LPH4 and LPH5, respectively. The boiler (1 + sTSC )
converts the water into high temperature and pressure steam using
constant pressure heating process. At high pressure (HP) and Equivalent transfer function of reheat tandem compound steam
temperature, steam passes through the main control valves and turbine
enters into HP steam turbine. Furthermore, it passes through inter- 
 
mediate-pressure turbine (IPT) and low pressure turbine (LPT) and FLP 1 1
GReheat (s) = FIP + + FHP
at last reaches to the condenser. The steam may be drained into heat (1 + sTCO ) (1 + sTRH ) (1 + sTSC )
regenerative cycle of feed water heaters when it passes through (2)
HPT, IPT and LPTs. The condensate pump pumps the condensate
water into the heat regenerative cycle. Before heading toward the FHP FIP
GReheat (s) = + 
steam generator, condensed steam passes through the gland steam (1 + sTSC ) 1 + sTSC (1 + sTRH )
condenser, HP and LP heaters, FWP etc. It is important to consider
FLP
the dynamics of control valve system, turbines, SC, RH and CO for +  (3)
the power system dynamic stability analysis. The other components 1 + sTSC (1 + sTRH )(1 + sTCO )
of the power plant are likely being effective for long-term stability     
analysis. These turbines generate the mechanical output power FHP 1 + sTRH 1 + sTCO + FIP 1 + sTCO + FLP
GReheat (s) =  
which is further transformed into electrical output power using 1 + sTSC (1 + sTRH )(1 + sTCO )
synchronous generators. Furthermore, detailed information about (4)
heat balance diagrams is available online on websites of thermal
power plants [27]. The comprehensive study of complete steam (see (5))
cycle and heat balance diagram has been carried out in [24, 28].
As the sum of all power fractions is 1 pu, i.e. FHP + FIP + FLP = 1.
3 Steam turbine dynamic model parameter calculations for Therefore, GReheat (s) can be written as
power system dynamic studies
 
The functional block diagrams of non-reheat and reheat steam tur- 1 + TCO FIP + TRH FHP + TCO FHP s + TRH TCO FHP s2
GReheat (s) =  
bines are shown in Figs. 2a and b, respectively. The model consists 1 + sTSC (1 + sTRH )(1 + sTCO )
of governor controlled valves (CVs), used to regulate the steam flow (6)
at the inlet of HPT. The steam flow cannot change instantly with the
change in the CVs position. The steam flow changes with some The methodology for finding the values of these parameters using
time lag which is caused by the SC, RH and CO pipes. In the mod- heat balance diagram are as follows.
elling of steam turbine system, these time lags (delays) are incorpo-
rated by using time constants for each component causing time lags
3.1 Time constant calculation
in the system. The dynamic models of both turbines are shown in
Figs. 3a and b, respectively [26]. The typical parameters of these The steam turbine dynamic model parameters contain the time con-
models are the time constants TSC, TRH and TCO of SC, RH, CO stants of SC, RH and CO. These time constants can be calculated

 
(FHP + FIP + FLP ) + TCO FIP + TRH FHP + TCO FHP s + TRH TCO FHP s2
GReheat (s) =   (5)
1 + sTSC (1 + sTRH )(1 + sTCO )

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
 3.2 Power fraction calculation
The power fractions determine the proportion of total power pro-
duced by HP, IP and LP steam turbines. Assuming, the total
power shared by these turbines has been represented on unit base.
It can be written as
Fig. 4 Steam vessel containing inlet and outlet ports for steam passage
FHP + FIP + FLP = 1 (9)

These power fractions in terms of power output can be written as

FLP PLP
= = a1 (10)
FHP PHP

and

FIP P
= IP = a2 (11)
FHP PHP

where PHP, PIP and PLP are the power outputs of HP, IP and LP
steam turbines. The turbine power outputs can be obtained by
using the enthalpy and steam flow data available in the heat
Fig. 5 Reheater configuration of the thermal power plant
balance diagram. The power output can be obtained using the
following equation

using the expression derived for the calculation of steam vessel’s

n

time constant as shown in Fig. 4. This expression derived using PX = Qk (hin, k − hout, k ) (12)
k=1
the mass continuity equation is given as [29, 30]
where PX is the output power and X stands for HPT, IPT or LPTs,
P k = indices of turbine stage between draining points, in kilojoules/
TV = o Vvessel × Kvessel (7)
Qo seconds, Qk is the flow rate of the kth stage, hin,k and hout,k are
the inlet and outlet enthalpies of the kth stage. From (10)–(12),
power fractions can be written as
where
1 a2
 FHP = , FIP = and
∂r  1 + a1 + a2 1 + a1 + a2
Kvessel =  (8) (13)
∂P T (◦ C) a1
FLP =
1 + a1 + a2

The values of a1 and a2 are calculated assuming that the process is

where TV is the vessel steam flow time constant in seconds, Qo is
isentropic and the losses (mechanical) are ignored.
the vessel’s steam flow rate in kilograms/seconds, Po is the steam
pressure in vessel, Vvessel is the volume of vessel and Kvessel is
4 Evaluation of steam turbine model parameters for different
the change in steam density with respect to change in pressure at
generation schedules of thermal plant
a particular temperature. The pressure, flow and temperature can
be obtained using heat balance diagram of thermal power plant. The 500 MW unit of thermal power plant has been considered for
The Kvessel can be calculated from the available thermodynamic the calculation of steam turbine model parameters. The complete
data of steam table [31]. It is important to know that the heat balance diagram of the plant is shown in Fig. 1, which repre-
temperatures at which we want to find the thermodynamic data sents the rated generation regime. The heat balance data for partial
sometimes may not be given in the steam table. The linear generation schedule is not shown in this figure, instead it has been
interpolation technique is generally used to approximate the data tabulated in Tables 9 and 10 of Appendix 2. The heat balance
of steam for the temperatures not available in the steam table. diagram consists of pressure, enthalpy, mass flow rate and tem-
Equations (7) and (8) can be used to find the time constants of perature of various components and pipe sections of plant cycle.
SC, RH and CO. The calculations of parameters of steam turbine dynamic model

Table 1 Calculated values of time constant and heat balance data for steam chest at different generation schedules

0% make up (MU) 0.1047 at back pressure

Generation schedules, % Pressure, kgf/cm2 Enthalpy, kcal/kg Flow, t/h Temperature, oC TSC, s

100 170 810.7 1457.009 537 0.2990

80 170 810.7 1166.642 537 0.3746
60 170 810.7 887.632 537 0.4922
50 170 810.7 755.373 537 0.5786
30 170 783.2 552.646 497 0.8947

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
Table 2 Calculated values of time constant and heat balance data for reheater at different generation schedules

0% MU 0.1047 at back pressure

Generation schedules, % Pressure, kgf/cm2 Enthalpy, kcal/kg Flow, t/h Input temperature, °C Output temperature, °C TRH, s

100 40.5 858.2 1301.229 337 565 5.00

80 32.81 859.8 1051.583 337 565 5.01
60 25.26 861.4 808.262 337 565 5.02
50 20.61 862.4 656.828 337 565 5.04
30 12.98 815.6 437.216 437.21 472 5.37

Table 3 Calculated values of time constant and heat balance data for cross-over at different generation schedules

0% MU 0.1047 at back pressure

∂r 2 2
Generation schedules, % Pressure, kgf/cm2 Enthalpy, kcal/kg Flow, t/h Temperature, oC , s /m TCO, s
∂P
100 7.13 735.8 1086.754 309.9 0.003839 0.4000
80 5.80 737.2 883.566 311.4 0.003790 0.3970
60 4.49 738.6 680.817 312.8 0.003770 0.3966
50 3.91 742.3 589.281 319.5 0.003715 0.3932
30 `2.74 714.6 424.432 261.2 0.004125 0.4248

Table 4 Calculated values of MW outputs for HP, IP and LP turbines at different generation schedules

Generation schedule, % PHP, MW PIP, MW PLP, MW PT, MW Ploss, MW Pnet, MW

100 138.864 178.837 191.439 509.173 9.173 500

80 110.646 144.895 151.382 406.923 6.923 400
60 83.5652 111.712 110.9735 306.25 6.250 300
50 73.4910 96.9823 85.4160 255.89 5.890 250
30 50.9830 59.1584 44.3843 154.5415 4.5415 150

using heat balance diagram of thermal power plant are in Section table [31], using simple curve fitting technique is given as
4.1. The conversion factors used in the calculations are as follows:
1 kgf/cm2 = 98.0665 kPa, 1 t/h = 0.277778 kg/s and 1 kcal/kg =  
4.186 kJ/kg. r p = CSC1 + CSC2 · p + CSC3 · p2
+ CSC4 · p3 kg/m3 , for 13, 000 ≤ P ≤ 17, 000 kPa at 537◦ C
4.1 Calculation of time constants for different generation (14)
schedules
All the coefficient values of CSC1, CSC2, CSC3 and CSC4 are given
4.1.1 Steam chest time constant: The average radius and height of in Appendix 1. Differentiating (14), with respect to pressure ( p)
the SC cylinder are ravg = 67.2 cm and havg = 150 cm. The volume of gives
cylinder is given as VSC = phavg ravg 2
= 2.1298 m3. The thermo-
dynamic data of SC extracted from the heat balance diagram is
shown in Table 1. Here, SC time constant calculations are shown ∂r
= CSC2 + 2.CSC3 . p + CSC4 . p2 s2 /m2 , (15)
only for rated power output condition. The same procedure can be ∂P
used for partial generation levels/schedules also. From the
heat balance diagram rated pressure and flow of SC are as follows:
PSC = 170 × 98.0665 = 16671.305 kPa, QSC = 1457.009 × 0.278 =
405.04 kg/s. The pressure–density curve derived from the steam

Table 5 Power fractions of HP, IP and LP turbines for different

generation schedules

Generation schedule, % FHP FIP FLP

100 0.2727 0.3511 0.3760

80 0.2719 0.3560 0.3720
60 0.2728 0.3647 0.3623
50 0.2872 0.3790 0.3338
30 0.3299 0.3828 0.2872 Fig. 6 Power fractions variation along with the generation schedule of the
plant

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
Fig. 7 Transfer function model of single area AGC system with reheat steam turbine dynamic model

∂r  337 to 565°C. Therefore, approximated time constant can be
537◦ C, = 0.00341 s2 /m2 , (16)
∂P calculated by taking the average temperature given as
16671.305 kPa Tavg = ((337 + 565)/2) = 451◦ C. The pressure–density curve
derived from the steam table [31], using simple curve fitting tech-
The time constant (TSC) of SC is nique is given in (18). The coefficient values of CRH1, CRH2,
CRH3 and CRH4 are given in Appendix 1
PSC ∂r
Tsc = V = 0.2999 s (17)  
QSC SC ∂P r p =CRH1 + CRH2 · p + CRH3 · p2 + CRH4 · p3
(18)
for 3800 ≤ P ≤ 4200 kPa at 451◦ C
Similarly, time constant can be calculated for different generation
schedules by extracting the heat balance data. All calculated ∂r
= CRH2 + 2 · CRH3 · p + CRH4 · p2 s2 /m2 , (19)
values of SC time constants for rated and partial generation sche- ∂P
dules have been tabulated in Table 1. 
∂r 
◦ = 0.0031544 s2 /m2 , (20)
4.1.2 Reheater time constant: Generally, RHs are double pipe set ∂P 451 C,
3971.69 kPa
structures and are located between HPT and LPTs and usually
buildup on the furnace heating walls. Fig. 5 shows a simple design
The time constant (TRH) of RH is
of RH used in the plant. The average height, radius and length are
taken as follows: Havg = 12 m, ravg = 47.86 cm and Lavg = 8 m. The
PRH ∂r
volume of the RH can be approximated as VRH = number of TRH = V = 5.00 s (21)
rows × number of pipes in a single row × volume of a single pipe. QRH RH ∂P
Thus, RH volume is given as VRH = 2× Lavg /2r × p · r2 · Havg =
p · r · Havg · Lavg = 144.36 m3 . Here, two rows of pipes are consid- Similarly, RH time constants can be calculated for partial gener-
ered because of the fact that the plant RH has the double pipe set ation schedules. All calculated values of RH time constants for
arrangement. The spacing between the pipes has been neglected. rated and part generation schedules have been tabulated in Table 2.
From the heat balance diagram, PRH = 40.5 × 98.0665 =
3971.69325 kPa, QRH = 1301.229 × 0.278 = 361.741 kg/s. It is 4.1.3 Cross-over time constant: The CO pipe is located between
clear from Table 1, i.e. temperature of the RH increases from IPT and LPTs. The dimensions are radius r = 130.14 cm and

Fig. 8 Transfer function block diagram of thermal–thermal power system model

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
 4.2 Calculation of HPT, IPT and LPT power fractions at different
generation schedules of the plant
The power fractions have been calculated using enthalpy and steam
flow data of the plant. The power fraction calculations have been
shown only for two cases, i.e. for 100 and 80% generation schedule
of the plant as given in Appendix 3. The values of a1 and a2 are cal-
culated using (10) and (11), then the power fractions are calculated
using (12) and (13). The same procedure can be used to calculate
the power fractions at other power generation schedules. All calcu-
lated values of megawatt outputs and power fractions of HPT, IPT
and LPTs for different generation schedules have been tabulated in
Tables 4 and 5, respectively.
The power fractions variation of reheat turbine system along with
generation schedule of the plant is shown in Fig. 6. It is very inter-
esting to know that at rated power output HPT, IPT and LPTs are
contributing 27, 35 and 38% of total power output, respectively.
As the generation schedule decreases, LPT contribution in total
power output decreases while the HPT and LPTs contribution
increases. At 30% generation schedule HPT, IPT and LPTs contri-
butions are 32, 39 and 29% which means that LPT contribution has
been decreased by 9% while HPTs and IPTs contributions have
been increased by 5 and 4%, respectively. Hence, the power frac-
tions of HPT, IPT and LPTs are varying considerably with the gen-
eration schedule of the plant. The power fraction values used at
rated power generation are no longer valid for partial generation
schedules and needs to be recalculated in studying the dynamic per-
formance of the AGC system.

5 Simulation results and discussion

The two areas of thermal power system have been considered for
the investigation. The area-1 consists of four parallel units, each
of 500 MW capacities and having the total generating capacity of
4 × 500 MW = 2000 MW. The area-2 is having the two parallel
units of 500 MW, having the total generating capacity of 2 × 500
MW = 1000 MW. Identical steam turbines have been employed in
both areas; therefore, same heat balance data have been used to
Fig. 9 Dynamic responses of the system at different generation schedules evaluate the turbine model parameters of all steam turbines of
assuming the steam turbine model parameters constant irrespective of gen-
both power system areas. The unit size is taken as 500 MW just
eration schedules of the plant
a Frequency deviation of area-1 for the purpose of demonstration of the concept. However, the
b Frequency deviation of area-2 methodology is general enough and can be applied to any other
c Tie-line power deviation system also.
It is assumed that all generating units in each area swings coher-
ently. Therefore, all parallel operating units in their respective areas
length Lavg = 8.5 m. The volume of the CO is VCO = π · r 2 · Lavg = can be replaced by a single generating unit having equivalent inertia
45.23 m3. From the heat balance diagram, PRH = 7.13 × 98.0665 = constant. For a system of ‘n’ generating units operating in parallel
699.214 kPa and QRH = 1086.754 × 0.278 = 302.117 kg/s. The coherently having inertia constants H1, H2, H3… and ratings S1,
pressure–density curve derived from the steam table [31] using S2, S3…. Then equivalent inertia constant Heqv of all generating
simple curve fitting technique is given below. The coefficient units can be written as [32]
values of CCo1, CCo2, CCo3 and CCo4 are given in Appendix 1
  H1 S1 + H2 S2 + . . . + Hn Sn
Heqv = s, (26)
r p =CCo1 + CCo2 · p + CCo3 · p2 + CCo4 · p3 , Hsystem
(22)
for 400 ≤ P ≤ 1000 kPa at 451◦ C
where
∂r
= CCo2 + 2 · CCo3 · p + CCo4 · p2 , (23)
∂P Hsystem = H1 + H2 + . . . + Hn . (27)

∂r 
◦ = 0.003839 s2 /m2 , (24)
∂P 309.9 C, Using (26) and (27), power system rotating mass time constant TP
699.214 kPa and power system gain Kp can be calculated as follows

The time constant (TCO) of CO is 2Heqv 1

TP = s, Kp = Hz/Mw pu. (28)
D·f D
PCO ∂r
TCO = V = 0.400 s (25)
QCO CO ∂P The power system frequency characteristic depends on the com-
bined effect of speed droops of all generator speed governors. For
Similarly, CO time constants can be calculated for partial gener- a system of ‘n’ generating units having droops R1, R2, R3… and
ation schedules as shown in Table 3. composite load damping constant D, the steady-state frequency

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
Table 6 System parameter values at different generation schedules of the plant assuming steam turbine model parameters constant

Total load, MW Generation schedules, % Area-1 Area-2

MW output TP1, s Kp1 B1 MW output TP2, s Kp2 B2

3000 100 2000 10 60 0.433 1000 10 60 0.433

1500 50 1000 20 120 0.425 500 20 120 0.425
900 30 600 33.3 200 0.422 300 33.3 200 0.422

Table 7 Optimised controller gains at different generation schedules of deviation for a load disturbance of ΔPd is given as [30]
the plant assuming steam turbine model parameters constant
−DPd
Generation schedules, % Controller gains Dfss =   Hz (29)
1/R1 + 1/R2 + . . . 1/Rn + D
Ki1 Ki2 −DPd
Dfss =
 Hz (30)
100 1.723 0.320 1/Reqv + D
50 1.627 0.283
30 1.582 0.269 where

1
Reqv = Hz/MW pu. (31)
(1/R1 + 1/R2 + . . . + 1/Rn )

The parameters Reqv, TP and KP are calculated using above equa-

tions are given in Table 11 of Appendix 4. It is found that the
equivalent inertia constant and speed regulation parameter of all
generating units based on system rating (total area capacity) are
same as the values of inertia constant and speed regulation param-
eter of a single unit based on its own individual rating. On the basis
of (26)–(31), the equivalent block diagram of single area AGC
system consisting ‘n’ generating units has been drawn in Fig. 7.
Block diagram includes the equivalent speed-governing system,
steam turbine dynamic model and rotating mass and load suitable
for load frequency analysis. The steam turbine model used in
diagram is based on (6), derived for reheat steam turbine. To
obtain the transfer function of non-reheat turbine, simply put TRH
and TCO equals to zero. The transfer function block diagram of
two area interconnected power system is shown in Fig. 8. System
parameter values are given in Appendix 4. The dynamic responses
have been obtained for 1% load perturbation in area-1 at different
generation schedules of the thermal power plant. The supplemen-
tary controller gains have been optimised based on integral
squared error (ISE) optimisation criterion given as
T

   2 
J= Df (s)2 + DP  dt
1 tie 12 (s) (32)
0

where ‘T’ is the total simulation time. The classical control algo-
rithm has been used for the tuning of controller gains. In this ap-
proach sequential optimisation is adopted where one parameter is
optimised at a time, keeping other parameters fixed using ISE

Table 8 Optimised controller gains at different generation schedules of

the plant considering the variation of steam turbine model parameters
along with generation schedules

Generation schedules, % SC time constant Controller gains

(Tt1 and Tt2), s
Fig. 10 Dynamic response of the system at different generation schedules Ki1 Ki2
after considering the variation of steam turbine time constant along with
generation schedules of the plant 100 0.2999 1.732 0.320
a Frequency deviation of area-1 50 0.5786 0.618 0.73
b Frequency deviation of area-2 30 0.8947 0.301 0.086
c Tie-line power deviation

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
criterion. The process is repeated for other parameters until 6 Conclusions
optimum value is reached. This method is suitable for cases
Dynamic performance of AGC system has been studied at different
where the numbers of optimisation variables are less; otherwise,
generation schedules. The AGC performance very much relies on
heuristic algorithms are used. However, in this case there are only
the steam turbine dynamic model parameters. The steam turbine
two variables to optimise; therefore, classical control approach
model parameters are found to be dependent on the generation sche-
will be self-sufficient to optimise the controller gains.
dules of thermal power plants. This paper incorporates the effect of
Here, two cases have been studied. In the first case, the dynamic
steam turbine model parameters variation in the dynamic perform-
responses of the system have been obtained at different generation
ance of AGC system. A realistic 500 MW units have been consid-
schedules assuming the steam turbine model parameters constant ir-
ered for study. Heat balance data is used to extract the steam turbine
respective the generation schedules of the plant. The power system
model parameters at different generation schedules. Some sturdy
parameters Kp, Tp and Bi are only considered to be variable along
points have been highlighted which are the main contributions of
with the power plant generation schedules. In the second case,
the work, are as follows:
steam turbine model parameters are also considered to be varying
with the plant generation schedules.
(i) The IEEE committee has just recommended the values of
5.1 Case-1: Steam turbine dynamic model parameters are steam turbine dynamic model parameters; however, it has
assumed to be constant irrespective to the generation schedules of not been described how to estimate these parameters from a
the plant realistic generating unit of thermal power plant. This paper
presents the mathematical procedure to calculate the
In this case, steam turbine dynamic model parameters are assumed
steam turbine model parameters by simply using the heat
to be fixed. Only power system model parameters Kp, Tp and Bi are
balance data (most likely available data) of the thermal
considered to be variable along with generation schedules.
power plant.
Frequency and tie-line power deviations of power system are
(ii) The IEEE committee has recommended the value of steam
shown in Fig. 9. It is found that the dynamic response of the
turbine (SC) time constant in the range of 0.2–0.5 s and in
system almost remains same at all generation schedules of the
most of AGC studies 0.3 s has been selected without any jus-
plant. Though the power system parameters Kp, Tp and Bi are chan-
tification [1–21]. The committee also did not mention any de-
ging with the generation schedules, still they are barely able to
pendency of the steam turbine time constant on the generation
affect the dynamic responses of the system. The power system
schedule of the plant. However, authors found that the steam
model parameter values and optimum controller gain settings for
turbine time constant depends on the generation schedule of
different generation schedules of the plant have been presented in
the plant and varies in the range of 0.2–1.0 s, particularly in
Tables 6 and 7, respectively. It is also clear for Table 7, i.e. con-
this case at 30, 50 and 100% generation levels, the calculated
troller gain settings are also slightly affected by the generation
values of time constants are 0.8947 s, 0.5786 s and 0.2999 s,
schedules of the plant.
respectively. Hence, the range of time constant defined by
the IEEE committee needs some modifications. Also, the
5.2 Case-2: Steam turbine dynamic model parameters are
value of the turbine time constant should be selected based
considered to be varying along with the generation schedules of the
on the generation level of the plant. As we can observe that
plant
the value of the steam turbine time constant affects the
In this case, steam turbine model parameters are also considered to dynamic performance of the entire thermal power system
be variable along with the generation schedule of the plant. The model; therefore, exact value of the time constant at a particu-
power system dynamic responses are shown in Fig. 10. It is clear lar generation level should be calculated using the mathemat-
from figure that as the generation schedule of the plant drifts ical procedure presented in this paper. The RH and CO time
down from the plant’s rated output; the dynamic responses of the constant almost remains the same for all generation levels of
system becoming more oscillatory in nature, i.e. number of oscilla- the plant. Hence, at partial generation schedules, only SC
tions and settling time are increasing considerably. The first peak of time constant needs to be recalculated.
oscillation has been increased considerably at 30% generation (iii) The IEEE committee also recommended the values of steam
schedule of the plant. It is happening due to the fact that the turbine power fractions only for rated power output
steam turbine time constant is increasing as the generation schedule condition of the plant. However, it has been discovered that
is shifting downwards from the rated value. After comparing Figs. 9 the power fractions of the steam turbines are also varied
and 10 of cases 1 and 2, respectively, it is also clear that the steam considerably along with generation schedules of the plant.
turbine dynamic model parameters are dominating over the power It has been observed that typically, at 100% loading
system parameters in deciding the dynamic performance of the condition HPT, IPT and LPT shares to the total power
system, as the power system parameters Kp, Tp and Bi are hardly outputs are 27, 35 and 38%, respectively. However, at 30%
able to affect the dynamic response of the system, while the generation level 5% increase in HP, 4% increase in IP and
steam turbine model parameters variation with the generation sche- 9% decrease in LPT have been recorded. Hence, in generalised
dules of the plant making the system more oscillatory. The optimum form, it can be stated that as the generation shifts down from
gain settings of the controllers of the thermal power system model its rated value, the power share/fraction of LPT decreases
are presented in Table 8, for different generation schedules of while this share is compensated by the increase of power
the plant. It can be seen from this table that the controller gain share/fractions of IP and HP turbines. These power fraction
settings are also changing considerably with generation schedule changes must be accounted in the dynamic studies of
of the plant. The optimum gain settings used at 100% generation AGC system. The power fractions at partial generation
scheduling are no longer valid at partial scheduling of the plant levels/schedules can be calculated using the method presented
and needs to be recalculated. in this paper.
Hence, it can be concluded that in AGC studies, the variation (iv) At last, it has been observed that the AGC dynamic perform-
steam turbine model parameters along with plant generation sche- ance degrades as the scheduled generation drifts down from
dules should be considered in order to get more realistic response rated output of the plant. The steam turbine time constant is
of the system. The dynamic responses obtained while keeping the the main parameter which deteriorates the dynamic perform-
turbine model parameters constant in such studies are very far ance of the system. The optimum controller gains obtained
from the actual responses of the system. for 100% generation schedules are completely unacceptable

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
 at partial generation schedules and needs to be recalculated as [19] Sharma Y., Saikia L.C.: ‘Automatic generation control of a multi-area
the generation schedule changes. ST – thermal power system using Grey Wolf optimizer algorithm
based classical controllers’, Int. J. Electr. Power Energy Syst.,
2015, 73, pp. 853–862
[20] Puja D., Saikia L.C., Nidul S.: ‘Automatic generation control of
multi area thermal system using Bat algorithm optimized PD–PID
7 Acknowledgment cascade controller’, Int. J. Electr. Power Energy Syst., 2015, 68,
Authors want to thank to Mr. Manish Kumar who shared his knowl- pp. 364–372
edge and insight work experience and assisted in research work. He [21] Sahu R.K., Panda S., Padhan S.: ‘A hybrid firefly algorithm and
pattern search technique for automatic generation control of multi
is currently working as a Project Engineer in NTPC Ltd. (National area power systems’, Int. J. Electr. Power Energy Syst., 2014, 64,
Thermal Power Corporation Limited), India. pp. 9–23
[22] Jonshagen K., Genrup M.: ‘Improved load control for a steam cycle
combined heat and power plant’, Energy, Elsevier, 2010, 35, (4), pp.
8 References 1694–1700
[23] Haisheng Y., Shuping C., Ruitao W.: ‘Analysis of the sliding pressure
[1] Concordia C., Kirchmayer L.K.: ‘Tie-line power and frequency operation for throttle controlled steam turbine generation unit’. IEEE
control of electric power systems’, Power Appar. Syst. III, Trans. Conf. on Power and Energy Engineering (APPEEC), Asia-Pacific,
Am. Inst. Electr. Eng., 1953, 72, (2), pp. 562–572 2012, pp. 1–4
[2] Concordia C., Kirchmayer L.K.: ‘Tie-line power and frequency [24] Sairam A., Kaushik S.C.: ‘Energy and exergy analysis of super
control of electric power systems – part II’, Power Appar. Syst. III. critical thermal power plant at various load conditions under constant
Trans. Am. Inst. Electr. Eng., 1954, 73, (1), pp. 133–146 and pure sliding pressure’, Appl. Therm. Eng., Elsevier, 2014, 73, (1),
[3] Elgerd O.I., Fosha C.: ‘Optimum megawatt frequency control of pp. 51–55
multi-area electric energy systems’, IEEE Trans. Power Appar. [25] Eggenberger Markus A., Callan Patrick C.: ‘Turbine control system
Syst., 1970, PAS-89, (4), pp. 556–563 for sliding or constant pressure boilers’, U.S. Patent 4253308 A, 1981
[4] Fosha C., Elgerd O.I.: ‘The megawatt-frequency control problem: a [26] IEEE Committee Report: ‘Dynamic models for steam and hydro tur-
new approach via optimal control theory’, IEEE Trans. Power bines in power system studies’, IEEE Trans. Power Appar. Syst.,
Appar. Syst., 1970, PAS-89, (4), pp. 563–577 1973, 92, (6), pp. 1904–1915
[5] Asadur R., Saikia C.L., Nidul S.: ‘Maiden application of hybrid [27] Online resource: ‘Heat balance diagram of typical coal-fired thermal
pattern search biogeography based optimisation technique in auto- power station’. Available at http://www.indianpowersector.com/
matic generation control of a multi-area system incorporating interline home/power-station/thermal-power-plant/
power flow controller’, IET Gener. Transm. Distrib., 2016, 10, (7), [28] Ameri M., Pouria A., Armita H.: ‘Energy, exergy and exergoeco-
pp. 1654–1662 nomic analysis of a steam power plant: a case study’, Int. J. Energy
[6] Chvez H., Baldick R., Matevosyan J.: ‘The joint adequacy of AGC Res., 2009, 33, pp. 499–512
and primary frequency response in single balancing authority [29] Ion B.: ‘Synchronous generators’ (CRC Press, Taylor & Francis,
systems’, IEEE Trans. Sustain. Energy, 2015, 6, (3), pp. 959–966 New York, 2016, 2nd edn.), pp. 75–77
[7] Kumar A., Malik O.P., Hope G.S.: ‘Discrete variable structure con- [30] Kundur P.: ‘Power system stability and control’ (McGraw-Hill,
troller for load frequency control of multi-area interconnected New York, 2009), pp. 422–424
power systems’, IEE Proc. C, Gener. Transm. Distrib., 1987, 134, [31] Van W.G.J., Sonntag R.E., Borgnakke C.: ‘Fundamentals of thermo-
(2), pp. 116–122 dynamics’ (John Wiley & Sons, Hoboken, NJ, USA, 2009, 7th edn.)
[8] Tripathy S.C., Bhatti T.S., Jha C.S., ET AL.: ‘Sampled data automatic pp. 838–877
generation control analysis with reheat steam turbines and governor [32] Grainger John J., Stevenson W.D.: ‘Power system analysis’
dead-band effects’, IEEE Trans. Power Appar. Syst., 1984, 103, (McGraw-Hill, New York, 1994), pp. 703–706
(5), pp. 1045–1051
[9] Tripathy S.C., Juengst K.P.: ‘Sampled data automatic generation
control with superconducting magnetic energy storage in power
systems’, IEEE Trans. Energy Convers., 1997, 12, (2), pp. 187–192 9 Appendices
[10] Ibraheem N., Bhatti T.S.: ‘AGC of two area power system intercon-
nected by AC/DC links with diverse sources in each area’, 9.1 Appendix 1
Int. J. Electr. Power Energy Syst., Elsevier, 2014, 55, pp. 297–304
[11] Shiva C.K., Mukherjee V.: ‘Automatic generation control of multi- Coefficients of pressure–density curves
unit multi-area deregulated power system using a novel
quasi-oppositional harmony search algorithm’, IET Gener. Transm. (a) For steam chest
Distrib., 2015, 9, (15), pp. 2398–2408
[12] Sahu R.K., Panda S., Rout U.K., ET AL.: ‘Teaching learning based op-
timization algorithm for automatic generation control of power
system using 2-DOF PID controller’, Int. J. Electr. Power Energy CSC1 = −0.030063, CSC2 = −0.0026774,
Syst., Elsevier, 2016, 77, pp. 287–301 ,
[13] Shiva C.K., Mukherjee V.: ‘Design and analysis of multi-source
CSC3 = 1.5478 × 10−8 and CSC4 = 2.464 × 10−13
multi-area deregulated power system for automatic generation
control using quasi-oppositional harmony search algorithm’,
Int. J. Electr. Power Energy Syst., Elsevier, 2016, 80, pp. 382–395
[14] Mohanty B., Hota P.K.: ‘Comparative performance analysis of fruit (b) For reheater
fly optimisation algorithm for multi-area multisource automatic gen-
eration control under deregulated environment’, IET Gener. Transm.
Distrib., 2015, 9, (14), pp. 1845–1855
[15] Bevrani H., Daneshfar F., Hiyama T.: ‘A new intelligent agent-based CRH1 = −0.49492, CRH2 = 0.003372,
AGC design with real-time application’, IEEE Trans. Syst. Man
Cybern., 2012, 42, (6), pp. 994–1002
CRH3 = −6.6696 × 10−8 and CRH4 = 8.7542 × 10−12
[16] Chang-Chien L.R., Wu Y.S., Cheng J.S.: ‘Online estimation of
system parameters for artificial intelligence applications to load fre-
quency control’, IET Gener. Transm. Distrib., 2011, 5, (8), pp.
895–902 (c) For cross-over
[17] Ramakrishna K.S.S., Pawan S., Bhatti T.S.: ‘Automatic generation
control of interconnected power system with diverse sources of
power generation’, Int. J. Eng. Sci. Technol., 2010, 2, (5), pp. 51–65
[18] Nikhil P., Ashu V., Bhatti T.S.: ‘Study the effect of system parameters CCO1 = 0.00076084, CCO2 = 0.0037128,
on controller gains for discrete AGC of hydro-thermal system’. IEEE
Int. Conf. INDICON, 2015, pp. 1–5
CCO3 = 8.8771 × 10−8 and CCO4 = 1.8648 × 10−12 ,

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
Units of coefficients are (b) IP turbine

s2 Q
CSC1 , CRH1 and CCo1 = ,
m2 PIP = (hin − hout )
s4    
CSC2 , CRH2 and CCO2 = , PIP = [Q22 h22 − h6 + Q22 − Q6 (h6 − h8 )]
m kg
kg kJ
s6 × 0.278 × 4.186
CSC3 , CRH2 and CCO3 = , s kg
kg2
PIP = [1301.229(858.2 − 800.1) (1301.229 − 87.418)
s8 m
CSC4 , CRH4 and CCO4 = 3, kg kJ
kg × (800.1 − 735.8)]0.278 × 4.186
s kg
PIP = 178.837 MW
9.2 Appendix 2
see Tables 9 and 10.

9.3 Appendix 3 (c) LP turbine

† Case-1: At 100% loading (500 MW load)
(a) HP turbine
Q Q
PHP = (hin − hout ) PLP = (hin − hout )
    
kg kJ PLP = Q21 h21 − h10 + Q21 − Q10 h10 − h12
PHP = Q1 (h1 − h2 ) × 0.278 × 4.186      kg
s kg + Q21 − Q10 − Q12 h12 − h14 + Q20 h14 − h20 ] × 0.278 ×
kg kJ s
PHP = [1457.009(810.7 − 728.8)]0.278 × 4.186 kJ
s kg 4.186 PLP = [1086.754(735.8 − 689) (1086.754 − 40.720) ×
kg
kJ
PHP = 138753.24 or kW = 138.864 MW (689 − 661.3) (1086.754 − 40.720 − 71.96) ×
s

Table 9 Heat balance data of 500 MW power generation unit at different loads

Location Pressure, kgf/cm2 Enthalpy, kcal/kg

100% 80% 60% 50% 30% 100% 80% 60% 50% 30%
Loading Loading Loading Loading Loading Loading Loading Loading Loading Loading

1 170.0 170.0 170.0 170.0 170.0 810.7 810.7 810.7 810.7 783.2
2 45.00 36.42 28.06 22.87 14.37 728.8 729.2 729.8 726.39 703.925
3 45.00 36.42 28.06 22.87 14.37 728.8 729.2 729.8 726.39 703.925
4 42.75 34.92 27.14 22.20 14.00 728.8 729.2 729.8 726.39 703.925
5 – – – – – 126.9 201.8 191.4 183.6 163.7
6 18.32 14.90 11.52 9.520 6.050 800.1 801.6 803.2 804.3 763.1
7 17.40 14.20 11.04 9.160 5.990 800.1 801.6 803.2 804.3 763.1
8 7.130 5.810 4.490 3.910 2.740 735.8 737.2 738.6 730.915 697.46
9 6.560 5.410 4.230 3.710 3.650 735.6 737.2 738.6 730.915 686.43
10 2.828 2.314 1.794 1.565 1.004 689.0 690.3 691.6 694.3 668.8
11 – – – – – 106.5 101.1 94.90 91.70 81.90
12 1.487 1.198 0.935 0.819 0.544 661.3 661.9 663.2 665.6 644.5
13 – – – – – 68.40 65.50 61.00 58.40 50.80
14 0.351 0.292 0.226 0.196 0.132 608.6 609.9 610.7 611.8 612.6
15 0.333 0.282 0.222 0.194 0.132 608.6 609.9 610.7 611.8 612.6
16 – – – – – 49.70 49.30 48.80 48.60 48.10
17 – – – – – 47.20 47.30 47.50 47.60 47.90
18 22.000 24.390 26.19 26.82 27.63 47.00 47.10 47.20 47.30 47.50
19 0.1047 0.1047 0.1047 0.1047 0.1047 46.30 46.30 46.30 46.30 46.30
20 0.1047 0.1047 0.1047 0.1047 0.1047 573.2 579.5 590.4 599.2 603.9
21 7.1300 5.8100 4.4900 3.9100 2.7400 735.8 737.2 738.6 730.837 682.110
22 40.500 32.810 25.260 20.610 12.980 858.2 859.8 861.4 862.4 815.8
23 – – – – – 99.80 99.80 99.80 99.80 99.90
24 – – – – – 51.80 50.60 49.40 48.90 48.20
25 – – – – – 71.10 67.30 61.90 58.90 50.90
26 – – – – – 73.10 69.10 63.70 60.70 52.00
27 – – – – – 111.2 105.1 98.10 94.40 83.70
28 – – – – – 107.8 161.7 151.2 145.7 145.0
29 – – – – – 214.2 204.0 192.1 183.5 162.5
30 – – – – – 263.4 251.4 237.2 225.9 201.0

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178
Table 10 Heat balance data of 500 MW power generation unit at different loads

Location Flow, t/h Temperature, °C

100% 80% 60% 50% 30% 100% 80% 60% 50% 30%
Loading Loading Loading Loading Loading Loading Loading Loading Loading Loading

1 1457.009 1166.642 887.632 755.373 552.646 537.0 537.0 537.0 537.0 497.0
2 1301.229 1051.563 808.262 656.828 437.216 337.0 329.2 321.5 310.7 255.9
3 149.644 110.122 75.558 95.434 113.636 337.0 329.2 321.5 310.7 255.9
4 149.644 110.122 75.558 58.824 38.129 334.8 327.6 320.5 310.0 255.3
5 1457.009 1166.642 887.632 755.373 552.646 204.8 196.6 186.7 179.2 159.9
6 87.418 68.478 50.422 39.413 13.767 445.0 446.0 447.1 448.0 363.7
7 87.418 68.478 50.422 39.413 13.767 444.5 445.5 446.8 447.8 363.6
8 130.184 102.042 78.980 29.646 00.00 309.9 311.4 312.8 319.5 261.2
9 66.966 50.879 36.028 29.646 33.478 309.3 311.0 312.5 319.3 237.9
10 40.720 31.823 22.994 19.111 12.690 280.7 210.3 211.9 217.1 161.9
11 1152.981 937.163 725.624 627.49 467.272 106.2 100.9 94.8 91.6 81.8
12 71.960 54.349 39.771 33.474 23.830 148.0 148.1 150.0 154.4 108.2
13 1152.981 937.163 725.624 627.490 467.272 68.3 65.4 60.9 58.3 50.7
14 39.658 27.626 15.849 10.905 2.261 0.9644 0.9701 0.9759 0.9801 0.9572
15 39.658 27.626 15.849 10.905 2.261 0.9644 0.9701 0.9759 0.9801 0.9572
16 1152.981 937.163 725.624 627.49 467.272 49.6 49.2 48.7 48.5 48.0
17 1152.981 937.163 725.624 627.49 467.272 47.0 47.2 47.4 47.5 47.6
18 1152.981 937.163 725.624 627.49 467.272 46.6 46.6 46.7 46.7 47.0
19 1152.981 937.163 726.624 627.49 467.272 46.4 46.4 46.4 46.4 46.4
20 934.862 770.214 602.647 526.237 386.371 0.9223 0.9335 0.9524 0.9679 0.976
21 1086.754 883.566 680.817 589.281 424.432 309.9 311.4 312.8 319.5 261.2
22 1301.229 1051.563 808.262 656.828 437.216 565.0 565.0 565.0 565.0 472.0
23 0.356 0.355 0.355 0.353 0.283 99.7 99.7 99.7 99.7 99.9
24 152.338 113.798 78.614 63.490 38.781 51.8 50.6 49.4 48.9 48.2
25 152.338 113.798 78.614 63.490 38.781 71.1 67.3 61.9 59.0 50.9
26 112.68 86.172 62.765 52.585 36.520 73.1 69.1 63.7 60.7 52.0
27 40.720 31.823 22.994 19.111 12.690 111.0 105.0 98.0 94.4 83.6
28 237.062 178.60 125.980 98.237 51.896 169.0 160.2 150.1 144.8 144.2
29 149.644 110.122 75.558 58.824 38.129 209.6 200.2 189.2 181.0 161.0
30 1457.009 1166.642 887.632 755.373 552.646 253.4 242.7 229.8 219.4 195.9

kg kJ (a) HP turbine
(661.3 − 608.6) (934.862)(608.6 − 573.2)] 0.278 × 4.186
s kg
PLP = 191.439 MW Q
PHP = (hin − hout )
kg kJ
PHP = [1166.642(810.7 − 729.2)]0.278 × 4.186
Total power output (PT) = 138.864 MW + 178.837 MW + s kg
191.439 MW = 509.173 MW. kJ
Mechanical losses = 1.444 MW. PHP = 110646.89 or kW = 110.646 MW
s
Generator losses = 7.579 MW.
Power required by turbine-generator (TG) auxiliary = 0.150 MW.
Net turbine output = (total power output) − (mechanical losses +
generator losses + TG auxiliary supply). (b) IP turbine
Pnet = 509.173 MW − (1.444 MW + 7.579 MW + 0.150 MW).
Pnet = 500 MW.
The power fractions of HP, IP and LPTs are as follows
2
PIP = Qk (hin, k − hout,k )
k=1
PLP P
a1 = = 1.3786, a2 = IP = 1.2875 PIP = [1051.563(859.8 − 801.6)
PHP PHP
1 kg kJ
FHP = = 0.2727 × (1051.563 − 68.478)(801.6 − 737.2)]0.278 × 4.186
1 + a1 + a2 s kg
a2 PIP = 144.895MW
FIP = = 0.3511
1 + a1 + a2 (c) LP turbine
a1
FLP = = 0.3760
1 + a1 + a2

4
PLP = Qk (hin, k − hout, k )
† Case-2: At 80% loading (400 MW load) k=1

J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314 This is an open access article published by the IET under the Creative Commons
doi: 10.1049/joe.2016.0178 Attribution License (http://creativecommons.org/licenses/by/3.0/)
 Pnet = 406.923 MW − (1.444 MW + 5.329 MW + 0.150 MW).
Pnet = 400 MW
PLP = [883.566(737.2 − 690.3) (883.566 − 54.349)
× (690.3 − 661.9) (883.566 − 54.349 − 27.626) PLP P
a1 = = 1.3681 , a2 = IP = 1.30956
PHP PHP
× (661.9 − 609.9) (770.214)(609.9 − 579.5)]0.278
1
FHP = = 0.2719
1 + a1 + a2
kg kJ a2
× 4.186 PLP = 151.382 MW FIP = = 0.3560
s kg 1 + a1 + a2
a1
FLP = = 0.3720
1 + a1 + a2
Total power output (PT) = 110.646 MW + 144.895 MW +
151.382 MW = 406.923 MW.
Mechanical losses = 1.444 MW.
Generator losses = 5.329 MW.
9.4 Appendix 4
Power required by TG auxiliary = 0.150 MW.
Net turbine output = (total power output) − (mechanical losses + Thermal–thermal power system model parameter values are calcu-
generator losses + TG auxiliary supply). lated as follows (see Table 11).

Table 11 Calculated parameters of two area AGC thermal power system model

Parameters Area-1 Area-2

area capacities Pr1 = 2000 MW Pr2 = 1000 MW

rating of single generating S = 500 MW S = 500 MW
machine
speed droop of single unit on 2.4 Hz/MW 2.4 Hz/MW
machine base
2.4 2.4
speed droop of single unit on R= = 9.6 Hz/MW pu. R= = 4.8 Hz/MW pu.
(500/2000) (500/1000)
power system area base
number of generating units n=4 n=2
1 1 R 4.8
equivalent speed droop Reqv1 = Reqv2 = Reqv2 = = = 2.4 Hz/MW pu.
(1/R + 1/R + 1/R + 1/R) (1/R + 1/R) 2 2
characteristic for area-1 and area-2
R 9.6
Reqv1 = = = 2.4 Hz/MW pu.
4 4
inertia constant of single H=5s H=5s
generating unit
H1 S1 + H2 S2 + H3 S3 + H4 S4 (5 × 500) + (5 × 500) + (5 × 500) + (5 × 500)
equivalent inertia constant for Heqv1 = Heqv1 =
Ssystem 2000
area-1 and area-2
H1 S1 + H2 S2 H1 S1 + H2 S2
Heqv1 = 5 s Heqv2 = Heqv2 =
Ssystem Ssystem

(5 × 500) + (5 × 500) (5 × 500) + (5 × 500)

Heqv2 = Heqv2 = 5 s Heqv2 = Heqv2 = 5 s
1000 1000
system frequency 60 Hz
synchronising coefficient 2πT12 = 0.545 pu MW
maximum tie-line power transfer Ptie max = 200 MW
limit
maximum allowable angle δ1 − δ2 = 30°
difference
area capacity ratios a12 = − Pr1/Pr2 = − 2
Tp1, Tp2, KP1, KP2, B1, B2 the values are calculated at different generation schedules shown in Table 6
FHPi, FIPi and FLPi the values are calculated at different generation schedules shown in Table 5
TCHi, TRHi and TCOi the values are calculated at different generation schedules shown in Tables 1–3

This is an open access article published by the IET under the Creative Commons J Eng, 2016, Vol. 2016, Iss. 8, pp. 302–314
Attribution License (http://creativecommons.org/licenses/by/3.0/) doi: 10.1049/joe.2016.0178

View publication stats