0% found this document useful (0 votes)

2K views

Operation and Control in Power Systems by PSR Murty Wwweeeunive

Uploaded by

Sowjanya Sowji

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

2K views

Operation and Control in Power Systems by PSR Murty Wwweeeunive

Uploaded by

Sowjanya Sowji

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 423

Downloaded From : www.EasyEngineering.

net

ww
w.E
asy
En
gin
eer
ing
.ne
t

**Note : Other Websites/Blogs Owners we requested you, Please do not Copy

(or) Republish this Material.
This copy is NOT FOR SALE.
**Disclimers : EasyEngineering does not own this book/materials, neither created nor
scanned. we provide the links which is already available on the internet. For any quarries,
Disclaimer are requested to kindly contact us. We assured you we will do our best. We DO
NOT SUPPORT PIRACY, this copy was provided for students who are financially troubled but deserving
to learn. Downloaded From Thank
: www.EasyEngineering.net
You and God Bless!
Downloaded From : www.EasyEngineering.net

ww
w.E
a syE
ngi
nee
rin
g .ne
t

Downloaded From : www.EasyEngineering.net

Operation and Control in

Power Systems

ww
w.E
Prof. P. S. R. MURTY
asy B.Sc. (Engg.) (Hans.) ME.,
Dr. - lng (Berlin), F.I.E. (India)

En
Life Member - ISTE

gi
(Formerly Principal O.U. College of Engineering &

nee
Dean, Faculty of Engineering, O.U. Hyderabad)

r ing
. net

BSP BS Publications
=== 4-4-309, Giriraj Lane, Sultan Bazar,
Hyderabad - 500 095 A.P.
Phone: 040 - 23445688

Downloaded From : www.EasyEngineering.net

Copyright © 2008, by Publisher

No part of this book or parts thereof may be reproduced, stored in a retrieval system
or transmitted in any language or by any means, electronic, mechanical,
i photocopying, recording or otherwise without the prior written permission of the
~ublishers.

ww
Published by :

w.E
SSP BS Publications
asy
4-4-309, Giriraj Lane, Sultan Bazar,
Hyderabad - 500 095 A. P.

E
Phone: 040 - 23445688, Fax: 91 +40-23445611

ngi
e-mail: [email protected]

nee
rin
Printed at
g. net
Adithya Art Printers
Hyderabad.

ISBN: 978-81-7800-181-0

Downloaded From : www.EasyEngineering.net

ww
w.E Contents
asy
En
gin
1 Introduction
ee
2 Load Flow Analysis rin
2.1 g.n
Bus Classification ............................................................................................... 9
2.2
2.3
e
Modelling for Load Flow Studies ...................................................................... 10
Gauses - Seidel Iterative Method ...................................................................... 13 t
2.4 Newton - Raphson Method .............................................................................. 16
2.4.1 Rectangular Coordinates Method ..................................................... 17
2.4.2 The Polar Coordinates Method ........................................................ 19
2.5 Sparsity of Network Admittance Matrices ........................................................ 22
2.6 Triangular Decompostion .................................................................................. 23
2.7 Optimal Ordering ............................................................................................... 25
2.8 Decoupled Methods ........................................................................................... 27

Downloaded From : www.EasyEngineering.net

(xiI) Contents

2.9 Fast Decoupled Methods ............................................................................•...... 27

2.10 Load Flow Solution Using Z Bus ...................................................................... 29
2.10.1 Bus Impedance Formation ............................~ .................................. 29
2.10.2 Addition of a Line to the Reference Bus .......................................... 29
2.10.3 Addition ofaRadial Line and New Bus ........................................... 30
2.10.4 Addition of a Loop Closing Two Existing Buses in the System ...... 30
2.10.5 Gauss - Seidel Method Using Z-bus for Load Flow Solution ......... 31
2.11 Load Flow Solution with Static Load Model .................................................... 32

ww 2.12 Comparision of Various Methods for Power Flow Solution ............................. 33

Questions ............................................................. ................ .......... .......... ........ 71

w.E
Problems ......................... ................... .... ..... ................................... ................. 72

3.1 asy
3 Economic Operation of Power Systems
Characteristics of Steam Plants ........................................................................ 86
3.2
En
Input Output Curves ......................................................................................... 87
3.3
3.4
3.5
gi
The Incremental Heat Rate Characteristic ........................................................ 88

nee
The Incremental Fuel Cost Characteristic ........................................................ 88
Heat Rate Characteristic .................................................................................... 89
3.6
3.7
r ing
Incremental Production Cost Characteristics ................................................... 89
Characteristics of Hydro Plants ........................................................................ 90
3.8
3.9
.
Incremental Water Rate Characteristics ............................................................ 91
Incremental Production Cost Characteristic ..................................................... 92 net
3.10 Generating Costs at Thermal Plants .................................................................. 93
3.11 Analytical Form for Input-Output Characteristics of Thermal Units ................ 93
3.12 Constraints in Operation .................................................................................... 94
3.13 Plant Scheduling Methods ................................................................................. 96
3.14 Merit Order Method .......................................................................................... 97
3.15 Equal Incremental Cost Method: Transmission Losses Neglected .................. 97
3.16 Transmission Loss Formula - B. Coefficients .................................................. 99
3.17 Active Power Scheduling ................................................................................. 103
3.18 Penalty Factor .................................................................................................. 106
3.19 Evaluation ofl for Computation ....................................................................... 107

Downloaded From : www.EasyEngineering.net

(xfu) Contents

3.20 Hydro Electric Plant Models ............................................................................ 119

3.21 Pumped Storage Plant ...................................................................................... 120
3.22 Hydro Thermal Scheduling .............................................................................. 120
3.23 Energy Scheduling Method .............................................................................. 121
3.24 Short Term Hydro Thermal Scheduling ........................................................... 125
3.24.1 Method of Lagrange Multipliers (losses neglected) ........................ 125
3.24.2 Lagrange Multipliers Method Transmission Losses Considered .... 126
3.24.3 Short Term Hydro Thermal Scheduling using

ww B-Coefficients for Transmission losses .......................................... 127

Questions ........................................................................................................ 151

w.E
Problems ........................................................................................................ 152

4 Optimal Load Flow

4.1 asy
Reactive Power Control for Loss Minimization ............................................... 155
4.2
En
Gradient Method for Optimal Load Flow ......................................................... 156
4.3
4.4
4.5
gi
Non - Linear Programming .............................................................................. 157

nee
Lagrange Function for Optimal Load Flow ..................................................... 158
Computational Procedures ................................................................ _............. 159
4.6
rin
Conditions for Optimal Load Flow ................................................................... 159
4.7
g.n
Implementation of optimal conditions .............................................................. 161
Questions ........................................................................................................ 168

et
Problems ........................................................................................................ 169

5 Unit Commitment
5.1 Cost Function Formulation .............................................................................. 171
5.2 Constraints for Plant Commitment Schedules ................................................. 173
5.3 Priority - List Method ......................................................................................·174
5.4 Dynamic Programming .................................................................................... t:z5
5.5 Unit Commitment by Dynamic Programming ................................................. 177
Questions ........................................................................................................ 180
Problems ........................................................................................................ 180

Downloaded From : www.EasyEngineering.net

Contents

6 Load Frequency Control

6.1 Speed Governing Mechanism .......................................................................... 183
6.2 Speed Governor ................................................................................................ 183
6.3 Steady State Speed Regulation ......................................................................... 185
6.4 Adjustment of Governor Characteristics ......................................................... 185
6.5 Transfer Function of Speed Control Mechanism ............................................ 186
6.6 Transfer Function of a Power System ............................................................ 188

ww 6.7 Transfer Function of the Speed Governor ....................................................... 190

6.8
6.9 w.E
Governing of Hydro Units ................................................................................ 191
Penstock Turbine Model .................................................................................. 193
6.10
6.11
asy
Modal for a Steam Vessel ................................................................................ 196
Steam Turbine Model ...................................................................................... 197
6.12 En
Reheat Type Steam Turbine Model .................................................................. 198
6.13
6.14
gi
Single Control Area ........................................................................................... 199
nee
The basics of Load Frequency Control ........................................................... 200
6.15
rin
Flat Frequency Control .................................................................................... 201
6.16
6.17 g.n
Real Power Balance for Load Changes ............................................................ 202
Transfer Function of a Single Area System ..................................................... 203
6.18
6.19
et
Analysis of Single Area System ........................................................................ 205
Dynamic Response of Load Frequency Control Loop .................................... 208
6.20 Control Strategy ............................................................................................... 209
6.21 PID Controllers ................................................................................................ 212
6.22 The optimal Control Problem ........................................................................... 222
6.23 The Linear Regulator Problem ......................................................................... 222
6.24 Matrix Riccati Equation .................................................................................... 224
6.25 Application of Modern Control Theory ............................................................ 224
6.26 Optimal Load Frequency Control - Single Area System .................................. 225

Downloaded From : www.EasyEngineering.net

Contents

6.27 Optimal Control for Tandem Compound Single Reheat Turbine -

Generator System ............................................................................................. 229
6.28 Optimal Control of Hydro Speed Governing System ....................................... 232
6.29 A Review of Optimal Control ........................................................................... 235
6.30 Load Frequency Control with Restrictions on the Rate of
Power Generation '" .................... :..................................................................... 236
6.31 Load Frequency Control using Output Feedback ............................................ 237
6.32 Load frequency Control and Economic Dispatch ............................................ 238

ww Questions ....................................................................................................... "39

Problems ........................................................................................................ 240

w.E
7 Control of Interconnected Systems
7. 1
7.2
a syE
Interconnected Operation ................................................................................. 241
Flat Frequency Control of Interconnected Stations ......................................... 241
7.3
ngi
Flat Tie-Line and Flat Frequency Control ........................................................ 244
7.4
7.5 nee
Tie-Line Bias Control ........................................................................................ 247
Complete Tie-Line Bias Control ....................................................................... 250
7.6
rin
Two Area System - Tie-Line Power Model ..................................................... 253
7.7
g.n
Block Diagram for Two Area System .............................................................. 254
7.8
7.9
7.10
Analysis of Two Area System .......................................................................... 255
et
Dynamic Response ........................................................................................... 257
Tie-Line Bias Control- Implementation ........................................................... 266
7.11 The Effect of Bias Factor on System Regulation ............................................ 267
7.12 Scope for Supplementary Control .................................................................... 269
7.13 State Variable Model for a Three Area System ................................................. 269
7.14 State Variable Model for a Two Area System ................................................... 274
7.15 State Variable Model for a Single Area System ................................................ 275
7.16 Model Reduction and Decentralised Control .................................................. ,284
Questions ........................................................................................................ 287
Problems ........................................................................................................ 288

Downloaded From : www.EasyEngineering.net

(XVl) Contents

8 Voltage and Reactive Power Control

8.1 Impedance and Reactive Power ....................................................................... 289
8.2 System Voltage and Reactive Power ................................................................ 293
8.3 Reactive Power Generation by Synchronous Machines .................................. 294
8.4 Effect of Excitation Control ............................................................................. 295
8.5 Voltage Regulation and Power Transfer ........................................................... 296
8.6 Exciter and Voltage Regulator .......................................................................... 297

ww 8.7 Block Schematic of Excitation Control ............................................................ 299

w.E
8.8
8.9
Static Excitation System .................................................................................. 300
Brushless Excitation Scheme ........................................................................... 301
8.10
8.11 aAutomatic Voltage Regulators for Alternators .................................................. 302

syE
Analysis of Generator Voltage Control ............................................................. 303
8.12
8.13 ngi
Steady State Performance Evaluation .............................................................. 306
Dynamic Response of Voltage Regulation Control ........................................... 306
8.14 nee
Stability Compensation for Voltage Control ..................................................... 307
8.15
rin
Stabilizing Transformer .................................................................................... 307
8.16
8.17 g.n
Voltage Regulators ............................................................................................ 309
ieee Type 1 Excitation System ......................................................................... 310
8.18
8.19
et
Power System Stabilizer .................................................................................. 313
Reactive Power Generation by Turbo Generator ............................................. 314
8.20 Synchronous Compensators ............................................................................ 314
8.21 Reactors 315
8.22 Capacitors315
8.23 Tap---Changing Transformers ........................................................................... 316
8.24 Tap-Staggering Method ................................................................................... 317
8.25 Voltage Regulation and Short Circuit Capacity ................................................. 318
8.26 Loading Capability of a Line ............................................................................. 320
8.27 Compensation in Power Systems ..................................................................... 320

Downloaded From : www.EasyEngineering.net

(xviI) Contents

8.28 Load Compensation .......................................................................................... 321

8.29 Static Compensators ........................................................................................ 328
8.30 Steady State Perfonnance of Static var compensators ................................... 331
8.31 Overvoltages on Sudden Loss of Load ............................................................ 334
8.32 Voltage Dips ...................................................................................................... 335
8.33 Subsynchronous Resonance ............................................................................ 337
Questions ........................................................................................................ 343

ww Problems ........................................................................................................ 344

9.1
w.E
9 Introduction to Advanced Topics
Facts Controllers .............................................................................................. 346

asy
9.1.1 Series Controllers ............................................................................ 346

En
9.1.2 Shunt Controller .............................................................................. 347
9.1.3 Series - Series Controllers .............................................................. 347

9.1.5
gi
9.1.4 Series - Shunt Controllers .............................................................. 348

nee
Power Flow Control ...................................................................... 348

r
9.1.6 Static Var Compensator(SVC) ........................................................ 349

ing
9.1.7 Unified Power Flow Controller ....................................................... 349

9.2 .
9.1.8 Advantages due to FACTS devices ................................................. 349

net
Voltage Stability ................................................................................................ 350
9.3 Power Quality ................................................................................................... 352
9.3.1 Power Quality Index ....................................................................... 353
9.3.2 Voltage Sags .................................................................................... 353
9.3.3 Rectifier Loads ................................................................................ 355
9.3.4 Flicker ............................................................................................. 355
9.3.5 Power Acceptability or Voltage Tolerance ....................................... 356
9.3.6 Solutions to Power Quality.problem ............................................... 356
9.4 Data Base for Control ....................................................................................... 357
9.5 State Estimation ................................................................................................ 358
9.6 Power System Security .................................................................................... 360

Downloaded From : www.EasyEngineering.net

(xviii) Contents

9.7 Steady State Security Assessment ................................................................... 361

9.8 Application to Outage Studies .......................................................................... 362
9.9 Pattern Recognition Methods ........................................................................... 363
9.10 Power System Control Centres ........................................................................ 365
9.11 Level Decomposition in Power Systems ......................................................... 367
9.12 NetworkAutomation ........................................................................................ 368
9.13 LoadPrediction ................................................................................................. 369

ww 9.14 Load Prediction using Matereological Data ...................................................... 371

9.15
9.16 w.E
Spetral Expansion Method ................................................................................ 376
Prediction by Scaling a Standard Load .......................................................... 377
9.17
9.18
a
Short - Term Load Forecasting Using Exponential Smoothing ....................... 378

syE
Peak Power Demand Prediction ....................................................................... 378
9.19
9.20 ngi
State Estimation in Load Forecasting ............................................................... 379
Generating Capacity Reliability and Outage Probabilities ................................. 380

nee
Questions ........................................................................................................ 386

rin
Objective Questions .......................................................................... 387

Answers to Objective Questions ...................................................... 400 g.n

References ......................................................................................... 401 et
Index ................................................................................................. 407

Downloaded From : www.EasyEngineering.net

ww
w.E
1 INTRODUCTION
asy
En
gin
Elecrical energy is the most popular form of energy, because it can be transported easily at
high efficiency and reasonable costs. Thomas Edison, established the first power station in

eer
1882 at New York city, United States of America. The lower Manhattan area was supplied DC
power from this station. Underground cables were used for distribution. At Appelton, Wisconcin

ing
the first water wheel generator was installed. Under Edison's patents several companies started
functioning in USA. However, these companies could supply energy to small distances due to
I2R power loss being excessive at low voltage distribution.
.
In 1885, Wililiam Stanley invented the transformer. which revolutionized the AC net
transmission. The invention of induction motor in 1888 by Nnikola Tesla caused dramatic
change in electrical power consumption through AC replacing many DC motor loads.
It is now an acknowledged fact that HV and EHV transmission alone can reduce
substantially the losses and bulk power transmission is feasible at these voltages. Nevertheless,
it is also well established that HVDC is convenient and more economical from operation and
control point of view under certain circumstances such as at distances of more than 500 kms.
A detailed discussion of this aspect is not within the purview of this book.
In India, two third of the electrical power generated is from coal based power stations.
Of the rest, about 24% comes from hydroelectric, 8.7% from Gas fired plants, 2.4% from
nuclear power plants. At the time of independence, the per capita consumption of electric

Downloaded From : www.EasyEngineering.net

2 Operation and Control in Power Systems

energy was 1.3 units. It is now about 3 units while China's per capita consumption is about 6
units. Developed countries have per capita consumptions of as high as 8,500 units. This
shows the great disparity that exists between the rich and the poor countries. However, Indian
Power Sector has undergone revolutionary changes. While in 1947 the installed capacity was
at 1300 MW, today it has surpassed 1,00,000 MW. In India, Nuclear Power has a target of
350 GW, and Hydro Power is estimated at 84 GW by CEA.
In India regional and national power grids are established to facilitate transfer of power
within and across the regions with reliability, security and economy on sound commercial
principles. The Power Grid Corporation was established in August 1991 and it started its
commercial operations from 1992-93. It is one of the largest transmission utilities in the

ww
world. The power grid is an ISO 9001 company with complete capability in AC transmission
upto 765 kV level and HVDC transmission upto ± 500 kV. Challenging jobs in operation and

w.E
maintenance of the national demand which is expected to reach a peak value of 114000 MW by
2006 are undertaken by Power Grid Corporation. Power Grid is also engaged in activities such

asy
as unified load dispatch which facilitates close monitoring of grid with real time data for
economic dispatch of power between the five regional grids and states.

En
Planning,design, operation, control and protection of power systems requires continuous
and comprehensive anatysisJo_e~aluate the current states and remedial control, if any, needed.

gi
Manual computation of power flows isexrremely time consuming even for very simple networks.

nee
In 1929, AC network analyzer, an analog computer was devised. Most of the early system
studies were performed on the network analyzer.

r ing
The Indian Power Grid System is divided into five regional grids. The southern region
comprises of Andhra Pradesh, Tamilnadu, Karnataka, Kerala and Pondichery. All these State

.
Electricity Boards are integrated for operation into Southern Regional Electricity Board.

net
Likewise, other boards are formed. Each state has inter State and Inter Regional links. For
example Andhra Pradesh and Maharastra have a tie-line at Ramagundam - Chandrapur 400 kV
link at Chandrapur.
In a similar way Andhra Pradesh has tie-line connections with Orissa, Madhaya Pradesh,
Tamilnadu and Karnataka.
The stipulated system frequency in India is 50 Hz. Since, there is deficit of generation
in the southern region, the operating frequency goes to 48.5 Hz. States like, Maharastra,
Madhyapradesh, and Orissa operate at 50 to 52 Hz. Andhra Pradesh imports power through
HVDC back to back system at Chandrapur through the 400 kV AC double circuit line from
Chandrapur-Ramagundam. In a similar way, through the HVDC back to back system at Gajuwaka
power is transmitted via 400..kV Jaypore-Gajuwaka double circuit AC line from Orissa.
While, we have dealt with the frequency scenario, it is worthwhile, to look at the other
performance index of electric power, the voltage profile. 400 kV lines have their voltage falling
to 340.3 kV at Cuddapah, 220 kV lines reaching 160.8 kV at Sullurpet and 132 kV lines
operating at 95.8 kV at Nagarkurnool.

Downloaded From : www.EasyEngineering.net

Introduction 3

From the above, it can be seen that there is very heavy demand for electric power in this
areas seriously compromising the power quality. While more generation is needed to be added
continuously a thorough knowledge of various aspects involved in the study of power system
operation and control is very essential for electric power engineers. This book is dedicated to
this task in a manner that students of power engineering grasp the essential concepts involved
in operation and control.
The system variables are continuously changing in both magnitude and number. The
system never reaches a steady state so as to permit any tests to be carried out on it, so that its
dynamic behaviour can be ascertained. In practice, it requires both continuous and discrete
controls. The spread of the power systems over vast geographical areas contributes to its

ww
vulnerability to environmental changes. The system's dynamics extends over a broad band
width ranging from micro seconds to several minutes.

w.E
Planning operation and control of isolated or inter connected power systems present a
large variety of challenging problems, the solution of which requires application of several

a
mathematical techniques from various branches of it. Knowledge of optimization techniques

syE
and optimal control methods is very essential to understand the multi level approach that has
been very successfully utilized. Various mathematical techniques that needed to be applied are

ngi
explained at the appropriate places while dealing with the subject.

Models for analysis and control

nee
Power system engineering is a branch where practically all the results of modern control
theory can be applied. Such an application will result in economy, better quality of service and

rin
the least inconvenience under abnormal situations, both anticipated and unforeseen.

g.n
Control system design, in general, for its analytical treatment, requires the determination
of a mathematical model from which the control strategy can be derived. While much of the

suitable technique to determine the models for the process to be controlled. Thus, it is
required to model and identify power system components using both physical relationships
et
control theory postulates that a model of the system is available. It is also necessary to have a

and experimental or normal operating data. The objective of system identification is the
determination of a mathematical model characterizing the operation of a system in some form.
The available information is either system outputs or some functions of outputs which may
contain measurement noise. The inputs may be known functions applied for the purpose of
identification, or unknown functions which it may be possible to monitor somehow, or a
combination of both.
The identified model may be in the form of differential equations, difference equations,
transfer functions, etc.
Even though all systems are nonlinear to some extent, the assumption of a linear model
leads to simpler models which can yield meaningful results with fairly good accuracy. A
system may be classified as stationary or non stationary. During the period of operation, when

Downloaded From : www.EasyEngineering.net

4 Operation and Control in Power Systems

controls are implemented, the system is normally assumed to be stationary. The system equations
may be formulated either in the continuous mode or in the discrete mode. While measurements
and predicted values are available at discrete intervals, continuous representation is the most
familiar mode. Transformation from continuous to discrete formulation is a straight forward
process. A power system invariably, is stochastic in nature since the load demand is the most
uncertain aspect in the operation of it. In addition, measurement uncertainty, errors, non
availability of readings, etc. all contribute to its stochastic nature in the model. In most of the
power system studies only a deterministic model is assumed, but when the situation demands
the probabilistic model is also used. A modern power system required identification of the
model and optimization of the same with reference to a performance criterion with computer

ww
predictive, adaptive, non interacting, sampled-data control for efficient operation.
Various models that are needed in analysis and for control are discussed and presented

w.E
through out the book.
Chapter 2 deals with load flow studies. They are performed to determine voltages,

asy
active and reactive powers, etc. at various points in the network for different operating conditions
subject to the constraints on generator capacities and specified net interchange between operating

En
systems and several other restraints. Load flow solution is essential for continuous evaluation
of the performance of the power systems so that suitable control measures can be taken in

variety of conditions. gi
case of necessity. In practice it will be required to carry out numerous load flows under a

nee
Economic system operation can be defined in a more general sense as making the best

rin
use of the resources available, subject to a variety of requirements over any desired period of

g.n
time. Economic power system operation deals with the means and techniques for achieving
minimum operating cost to supply a given predicted load demand. It may be pointed out that

et
extensive research has been carried out in this field covering topics such as economy of fuel,
maintenance and overhaul schedules of equipment, starting and shut-down of generating plants,
scheduling of generation to different units, exchange of power between neighbouring utilities
and a wide range of problems related to hydrogeneration, like water usage, policies for different
types of hydro plants (reservoir, pondage, run-off river and pumped storage) and their integration
with the system, both hydraulically and electrically. Different combinations of thermal and
hydrogenerating plants give rise to different cost structures. Also, for a given combination of
plants, the operational requirement of scheduling generating plants to supply the predicted load
demand and subsequent formulation of loading pattern to be imposed on individual units
committed to service to minimize the cost of supplying a given load is another important aspect
of the problem. In general, the ordering or committing plant to operate on one hand and
loading of plant in operation on the other are the two facets of the economic operation, both
considered separately in Chapter 3 and 4.
The division is mainly from the period_of time over which cost minimization is affected.
The loading of plant in operation'is related to'Cost minimization over short periods of time.

Downloaded From : www.EasyEngineering.net

Introduction 5

This problem is called instantaneous, static or optimal point minimization. Committing of

plants to service, by comparison, relates to larger periods of time and gives rise to a variational
form ofthe problem in which the minimization required is that ofthe time integral of operating
costs over the period for which programmes of plant (unit) commitment are formulated.
A division of scheduling studies related to operation and control can be made as foHows:

Scheduling problem Time I Period

(a) Long-range scheduling for plant maintenance and Month /Year
for short term availability or resources
(b) Short term scheduling for unit commitment and I Day / 2 Weeks

ww unit hourly energy schedules

.
w.E
(d) Tie-line interchange, system frequency control Seconds

asy
(e) Plant and unit control Continuous
In Chapter 5 optimal load flow problem and certain guide lines to obtain an optimal

E
solution are presented, but the information is only at an introductory level.

ngi
There are several problems associated with hydrothermal combined operation such as:

nee
1. Fuel ordering, i.e., given the operating pattern over the period of interest (say two
weeks), determine the station fuel requirements and the optimal fuel procurement
policy.

rin
2. Plant ordering or unit commitment, i.e., the scheduling of the start up and shut
down of generating units, a dynamic optimization problem.
g.n
3. Hydrothermal scheduling, i.e., the optimal use of available water to coordinate
with the thermal generation, a dynamic optimization problem, etc.
et
The economic objective of a schedule can be assumed either by the profit from energy
sales or by the cost of energy production. If the loads are fixed, then both the criteria result in
the same schedule. However, under more realistic circumstances, i.e., when load is a function
ofa voltage, the two criteria yield different optimal policies. Minimization of production costs
is taken generaHy as the criterion for economic operation.
Each minimization problem is subject to a number of constraints arising from the
characteristics of plants and their safe operating conditions and from the requirements of
technically favourable operating conditions in the transmission system interconnecting various
power stations. The requirements of security are superimposed on these constraints. Added
to these are the requirements of marginal or reserve generating capacity in excess of the
minimum necessary to supply a predicted load demand to complement the probabilistic nature
of load predicted and to cover unforeseen operational occurrences. A wide number of
formulations and analytical solution techniques have been pursued in this direction. A few
important methods only are discussed in Chapters 3, 4 and S.

Downloaded From : www.EasyEngineering.net

6 Operation and Control in Power Systems

The disturbances to which a power system is subjected can be roughly classified as

small scale and large scale disturbances. Slowly varying small magnitude changes can be
effectively controlled, using governors, exciters, etc. The control of the system frequency
using speed governors and supplementary controls is discussed in detail in Chapter 6 under the
title "Load frequency control"
Power systems are often interconnected to improve reliability and quality of power
supply to the consumer, to reduce the spinning reserve requirements of individual systems and
for similar other advantages. The operating state of a power system can be divided into four
modes: .

ww 1. Nonnal mode
2. Preventive mode

w.E
3. Emergency mode, and
4. Restorative mode.

asy
In the nonnal mode of operation, the system has to maintain scheduled voltages, frequency
and load flow profile maintaining the scheduled tie line power flows. In this mode of operation
control is required to
En
I. Maintain scheduled voltages and frequency

gi
2. Maintain scheduled tie-line flows, and
3. Obtain economic generation
nee
In the emergency mode of operation, i.e., when the contingency has occurred, control
is required to
I. Maintain the specified frequency, and
r ing
2. Maximize the amount of load demand being met.
. net
During the restorative mode of operation, the system is brought from emergency mode
of operation into either nonnal mode or preventive mode.
The most important aspect in any mode of operation is the matching between load
demand and generation. The frequency deviation of the system is a direct measure of the
mismatch between the total generation and combined load demand. It is only when the frequency
is maintained at the rated value that the generation balances the load demand. An accelerating
frequency means that the generation is high while a decelerating freqttency indicates insufficient
generation.
A transmission line may be a connection between a generating station to a system or
may be an inter-tie between two large systems.
Assuming the line losses to be negligible, it can be proved that a more or less natural
way of operating a transmission line would be to seek to maintain the voltage levels through
regulating the reactive power flow and to provide for the variations of the active power demand

Downloaded From : www.EasyEngineering.net

IntrtJduction 7

by allowing the phase angle between the two end voltages 0, to change. This is brought about
by adjusting the throttle ofthe prime movers in the generating stations at one end or both ends
of the line. The pewer transfer over the line is given by

p=VsVRSino
XL
where Vs and V Rare sending end and receiving end voltage magnitudes respectively.
By slowly increasing the load, the maximum power transfer can be obtained when
o ~ 90°. Further, increase in load will not increase the power transmitted, but instead decreases
it. This point is referred to as the static stability limit or static transmission capacity of the line.

ww
This capacity can, of course, be increased by increasing the voltage magnitudes, but there are
limits for this increase.

w.E
The incremental increase in transmitted power ~p caused by a small increment ~o in
the phase angle is a measure of the electrical stiffness of the transmission line. The quantity

(~)
asy
is also called synchronizing coefficient.

E ngi
It can be seen that the transmission capacity can be increased also by reducing the
effective reactance of the line. This can be achieved by paralleling the lines, using bundled
conductors or inserting series capacitors.
nee
The analysis, operation and control of inter connected power systems or simply areas

rin
are discussed comprehensively in Chapter 7. The objective of system voltage control is to
maintain a satisfactory voltage profile in the system during both periods of maximum and

g.n
minimum loadings. A detailed analysis of excitation control and means adopted for reactive
power generation in addition to synchronous machine are presented in Chapter 8.

var compensators etc., are discussed in Chapter 8. The role of a power system stabilizer is
also presented.
et
Various devices such as tap changers, reactors, capacitors, induction regulators static

In Chapter 9, certain advanced topics that are related to operation and control are
introduced. These are, state estimation, FACTS controllers, Voltage stability, Power quality,
load prediction, energy control centers etc. The inclusion of the topics and the presentation of
the information is by no means exhaustive.

Downloaded From : www.EasyEngineering.net

ww
w
2 LOAD
.EaFLOW ANALYSIS
syE
ngi
Load Flow or Power Flow is the solution for the Power System under static conditions of

nee
operation. Load Flow studies are undertaken to determine:

rin
I. The line flows
2. The bus voltages and system voltage profile

g .ne
3. The effect of changes in circuit configuration, and incorporating new circuits on
system loading

loading and accompanied effects

5. The effect of in-phase and quadrature boost voltages on system loading.
t
4. The effect of temporary loss of transmission capacity and (or) generation on system

6. Economic system operation

7. system transmission loss minimization
8. Transformer tap settings for economic operation and
9. Possible improvements to an existing system by change of conductor sizes and
system voltages.
For the purpose of load flow studies, a single phase representation of the power network
is used since the system is generally balanced. When systems had not grown to the present
size, networks were simulated on network analyzers for power flow studies. These analyzers

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 9

are of analogue type, scaled down miniature models of power systems with resistances,
reactances, capacitances, autotransformers, transformers, loads, and generators. The generators
are just supply sources operating at a much higher frequency than 50Hz to limit the size of the
components. The loads are represented by constant impedances. Meters are provided on the
panel board for measuring voltages, currents, and powers. The load flow solution is obtained
directly from measurements for any system simulated on the analyzer.
With the advent of the modern digital computer possessing large storage and high speed,
the mode of load flow studies have changed from analog to digital simulation. A large number
of algorithms are developed for digital power flow solutions. Some of the generally used
methods are described in this chapter. The methods basically distinguish between themselves

ww
in the rate of convergence, storage requirement and time of computation. The loads are generally
represented by constant power.

(i) w.E
In the network at each bus or node there are four variables viz.
Voltage magnitude
(ii)
(iii) asy
Voltage phase angle
Real power and
(iv)
En
Reactive power.

gi
Out of these four quantities two of them are specified at each bus and the remaining two

nee
are determined from the load flow solution. To supply the real and reactive power losses in
lines which will not be known till the end of the power flow solution, a generator bus, called

r
slack or swing bus is selected. At this bus, the gen~rator voltage magnitude and its phase angle

ing
are specified so that the unknown power losses are also assigned to this bus in addition to
balance of generation if any. Generally, at all other buses, voltage magnitude and real power

. net
are specified. At all load buses the real and the reactive load demands are specified. Table 2.1
illustrates the types of buses and the associated known and unknown variables.

2.1 Bus Classification

Table 2.1

Bus Specified variables Computed variables

Slack - bus Voltage magnitude and its phase angle Real and reactive powers
Generator bus Magnitudes of bus voltages and real Voltage phase angle and
(PV - bus or voltage powers (limit on reactive powers) reactive power.
controlled bus)
Load bus Real and reactive powers Magnitude and phase
angle of bus voltages

Downloaded From : www.EasyEngineering.net

10 Operation and Control in Power Systems

2.2 Modelling for Load Flow Studies

Bus admittance formation
Consider the transmission system shown in Fig. 2.1.

ww
w.E Fig. 2.1 Three bus transmission system

asy
The line impedances joining buses 1,2 and 3 are denoted by z 12' Z 22 and z31 respectively.
The corresponding line admittances are Y12' Y22 and Y31

E
The total capacitive susceptances at the buses are represented by YIO' Y20 and Y30'
Applying Kirchoff's current law at each bus ngi
II = VI YIO + (VI
nee
- V 2) YI2 + (VI - V 3) YI3

+ (V 2 - VI) Y 21 + (V2 - V 3) Y23

rin
12 = V 2 Y20

13 = V3 Y30 + (V 3 - VI) Y 31 + (V 3 - V 2) Y32

In matrix from
g.
lll] f' +!I' +y"
12
13
_. Y12
-YI3
Y20
- YI2
+ Y12 + Y23
- Y23 Y30
- Y13
- Y23
net
+ YI3 + Y23
]
x

l
VI]
V
V3
[YII
2 == Y2I
YJI
YI2
Y22
Y32
y"]n
Y23 '
Y33
~2
3

where
YII == YIO + YI2 + YI3
Y22 = Y20 + YI2 + Y23
Y 33 = Y30 + YI3 + Y23

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 11

are the self admittances forming the diagonal terms and

Y 12 =Y 21 =-YI2

Y 13 =Y 31 =-YI3

Y 23 = Y 32 = -Y23
are the mutual admittances forming the off-diagonal elements of the bus admittance
matrix. For an n-bus system, the elements of the bus admittance matrix can be written down
merely by inspection of the network as
diagonal terms

ww
n
YII = YiO + LY,k
k=1
k ..i

w.E
off and diagonal terms

asy Y. k =-Y.k
If the network elements have mutual admittance (impedance), the above formulae wi\l

E
not apply. For a systematic formation of the y-bus, linear graph theory with singular
transformations may be used.
System Model for Load Flow Studies ngi
nee
The variable and parameters associated with bus i and a neighboring bus k are represented in
the usual notation as follows :

rin ..... (2.1 )

Bus admittance,
g.n
e .k = I Y .k I (Cos q.k + j sin e.k)
Complex power,
Y. k = I Y ik I exp j

S. = p. + j Q. = Vi [\
et
..... (2.2)

..... (2.3)
Using the indices G and L for generation and load,
p. = PG• - PLi = Re [Vi [.J ..... (2.4)
Q. = QG. - QL. = 1m [Vi 1. 1] ..... (2.5)
The bus current is given by
IBus = YBUS . VBUS ..... (2.6)
Hence, from eqn. (2.3) and (2.4) from an n-bus system

I~I = P, -V,.jQ I =YII V,

~
+ L Y,k V k ..... (2.7)
I k=1
k .. 1

Downloaded From : www.EasyEngineering.net

12 Operation and Control in Power Systems

and from eqn. (2.7)

..... (2.8)

Further,
n
P, + jQ, = V, LYi~ V: ..... (2.9)
k=1

In the polar form

ww P, + jQ, = L
n
lV, Vk Y,klexpj(o, -Ok -e,k) ..... (2.10)

so thatw.E k=1

asLy P, ==
n

k=1
lV, Vk Yiklcos(o, -Ok -e,k) ..... (2.11 )

and
E n
Q, == L lV, ngi Y,k Isin (0, ..... (2.12)

nee
Vk -Ok -e,k)
k=1

i = 1, 2, ..... n; i -:t:- slack bus

rin
The power flow eqns. (2.11) and (2.12) are nonlinear and it is required to solve 2(n-1)
such equations involving 1V, I, 0" P, and Q, at each bus i for the load flow solution. Finally, the
powers at the slack bus may be computed from which the losses and all other line flows can g.n
be ascertained. V-matrix interactive methods are based on solution to power flow relations
using their current mismatch at a bus given by
n
et
L\ I, = I, - L Y,k V k ..... (2.13)
k=1

or using the voltage from

L\I·
L\ V.==-' ..... (2.14)
I Y II

The convergence of the iterative methods depends on the diagonal dominance of the
bus admittance matrix. The self-admittances of the buses, are usually large, relative to the
mutual admittances and thus, usually convergence is obtained. Junctions of very high and low
series impedances and large capacitances obtained in cable circuits long, EHV lines, series and
shunt compensation are detrimental to convergence as these tend to weaken the diagonal
dominance in the V-matrix. The choice of slack bus can affect convergence considerably. In

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 13

difficult cases, it is possible to obtain convergence by removing the least diagonally dominant
row and column of Y. The salient features of the V-matrix iterative methods are that the
elements in the summation terms in eqn. (2.7) or (2.8) are on the average only three even for
well-developed power systems. The sparsity of the V-matrix and its symmetry reduces both
the storage requirement and the computation time for iteration (sec. 4). For a large, well
conditioned system of n-buses, the number of iterations required are of the order of n and total
computing time varies approximately as n2•
Instead of using eqn (2.6), one can select the impedance matrix and rewrite the
equation as
v = y- I 1= Z.I

ww
..... (2.15)
The Z-matrix method is not usually very sensitive to the choice of the slack bus. It can

w.E
easily be verified that the Z-matrix is not sparse. For problems that can be solved by both
Z-matrix and V-matrix methods, the former are rarely competitive with the V-matrix methods.

2.3
asy
Gauses - Seidel Iterative Method
In this method, voltages at all buses except at the slack bus are assumed. The voltage at the

En
slack bus is specified and remains fixed at that value. The (n-I) bus voltage relations.

[pl-jQI-~Ygv]
V=_I
Y ine V· ..... (2.16)

er
I L. Ik k
II I k=1
k",]

i = I, 2, ..... n; i 7= slack bus ing

. net
are solved simultaneously for an improved solution. In order to accelerate'the convergence, all
newly-computed values of bus voltages are substituted in eqn. (2.16). The bus voltage equation
of the (m + I )th iteration may then be written as

..... (2.17)

The method converges ~low1y because of the loose mathematical coupling between the
buses. The rate of convergence of the process can be increased by using acceleration factors
to the solution obtained after each iteration. A fixed acceleration factor a (I ::; a ::; 2) is
normally used for each voltage change,

AV = a AS: ..... (2.18)

I •
VjYII

Downloaded From : www.EasyEngineering.net

14 Operation and Control in Power Systems

The use of the acceleration factor amounts to a linear extrapolation of VI' For a given
system, it is quite often found that a near-optimal choice of a exists as suggested in literature
over -a range of operating conditions. Even though a complex value of a is suggested in
literature, it is more convenient to operate with real values given by

..... (2.19)
Alternatively, different acceleration factors may be used for real and imaginary parts of
the voltage.
Treatment of a PV - bus

ww
The method of handling a PV -bus requires rectangular coordinate representation for the voltages.
Lettering

w.E ..... (2.20)

asy
Where v; and v;' are the real and imaginary components of Vi the relationship.

En V '2
I
+v·"2
I
::; 1V I 12schedules ..... (2.21)

gi
must be satisfied, so that the reactive bus power required to establish the scheduled bus

nee
voltage can be computed. The estimates of voltage components, v;(m) and V;-(m) after

voltage is

1
r
m iterations must be adjusted to satisfy eqn. (2.21). The Phase angle of the estimated bus

ing
oem)
I
=tan-I
[
~"Cm)
'em)
Vi
. net ..... (2.22)

Assuming that the phase angles of the estimated and scheduled voltages are equal; .then
the adjusted estimates of V'(m) and V;'cm) are

ICm)
vi(new) -IVIIScheduiedcosuls:Cm)
- ..... (2.23)

and "(m)_I I .
vi(new) - VlscheduledsmBj
(m)
..... (2.24)

These values are used to calculate the reactive power Q\m) . Using these reactive powers

Q~m) and voltages Vi\~~W) a new estimate V lm+l) is calculated. The flowchart for computing
the solution of load flow using gauss-seidel method is given in Fig. 2.2.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 15

While computing the reactive powers, the limits on the reactive source must be taken
into consideration. If the calculated value of the reactive power is beyond limits. Then its value
is fixed at the limit that is violated and it is no longer possible to hold the desired magnitude of
the bus voltage, the bus is treated as a PQ bus or load bus.

ww
w.E
a syE
Yes

1 Y
IJVcm) ngi
VCm ) =_1 [<PJ-.iQJ) ~Y
k =1 Ik
v(m+l)_
k
i
k=I+1
Y VIm)]
Ik k

nee
1

rin
g.n
et
Yes

Yes
m=m+ 1

Calculate line flows and slack bus

power

Fig. 2.2 Flowchart for Gauss - Seidel iterative method for load flow solution using Y-Bus

Downloaded From : www.EasyEngineering.net

16 Operation and Control in Power Systems

2.4 Newton - Raphson Method

The generated Newton-Raphson method is an interactive algorithm for solving a set of
simultaneous nonlinear equations in an equal number of unknowns. Consider the set of nonlinear
equations:
fj (xl' x2 ' ...... , xn) = YI' i = 1,2, .... , n ..... (2.25)
with initial estimates for XI
(0) (0) (0)
XI ,x 2 , ••••• , xn
which are not far from the actual solution. Then using Taylor's series and neglecting

ww
the higher order terms, the corrected set of equations are

(X\O) + ~XI' x~O) + ~X2 ,...... , <0) + ~xn)= YI ..... (2.26)

w.E
w here ~ XI are the corrections to X. = (i = I, 2, ..... , n)
A set of linear equations, which define a tangent hyperplane to the function ~ (x) at the

a
given iteration point (x~O»)
syE
are obtained as
~Y= J~X .....(2.27)
where ~
ngi
Y is a column vector determined by

~ (x\O) ,..... ,x~O»)

~X
YI _
is the column vector of correction terms ~neeXI' and J is the Jacobian matrix for the

rin
function f given by the first order partial derivatives evaluated at x~O) The corrected solution is
obtained as
X(I) = x(O) + ~X g.n ..... (2.28)

et
I I I
The square Jacobian matrix J is defined by
Ofl
J lk = Ox ..... (2.29)
k
The above method of obtaining a converging solution for a set of nonlinear equations
can be used for solving the load flow problem. It may be mentioned that since the final voltage
solutions are not much different from the nominal values, Newton - Raphson method is
particularly suited to the load flow problem. The matrix J is highly sparse and is particularly
suited to the load flow application and sparsity - programmed ordered triangulation and back
substitution methods result in quick and efficient convergence to the load flow solution. This
method possesses quadratic convergence and thus converges very rapidly when the solution
point is close.
There are two methods of solution for the load flow using Newton - Raphson method.
The first method uses rectangular coordinates for the variables, while the second method uses
the polar coordinate formulation.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 17

2.4.1 Rectangular Coordinates Method

The power entering the bus i is given by
SI = PI + j Q I

" V,~,i=I,2, .... ,n

=Vi I: =VILYI~ ..... (2.29)
k=1

Where v I
+ jVI
= VI
and Y lk = Gik + j Bik

Xv~ - v~)1'j V~ - j y'~

ww (Pi + jQ) = ((v; + jv;·)t (G lk - jB'k
k=1
..... (2.30)

parts. w.E
Expanding the right side of the above equation and separating out the real and imaginary

asy . .... (2.31 )

En v~ -G lk v~)-V;'(Glk v~ +G lk v~)]
gi
QI = t[v;(G lk
k=1

nee
These are the two power relations at each bus and the linearized equations of the form
(2.27) are written as
..... (2.32)

~P1 oP1 ~ OP1 r ~ ing

av';

oPn-1
av~_1

oPn-1 oPn-1
av~ av~-1

oPn-1
. net
~Pn-1
av'l av~_1 av~ av~_1 ..... (2.33)
=
00 1 00"-1 00 1 00"-1
~QI
av'; av " av'; av "
"-I "-I
oon-I 00"-1 oon-I OOn-1
~Qn-I av'l av n-I av'; av "0-1 ~v 0-1

Matrix equation (2.33) can be solved for the unknowns~v; and ~v;' (i = 1,2'00" n -1),
leaving the slack bus at the nth bus where the voltage is specified. Equation (2.33) may be
written compactly as

..... (2.34)

Downloaded From : www.EasyEngineering.net

18 Operation and Control in Power Systems

where H, N, M and L are the sub-matrices of the Jacobian. The elements of the Jacobian are
obtained by differentiating Eqns. (2.31) and (2.32). The off-diagonal and diagonal elements of
H matrix are given by

oP· "
-.-1 = G 1k V k + Blk V k ,i ~ k ..... (2.35)
Ov k

..... (2.36)

ww The off-diagonal and diagonal elements ofN are:

w.E oP·
-.-1 =Gikv
Ov k
"
k
.
-BikV ,k~i ..... (2.37)

asy ..... (2.38)

E ngi
The off-diagonal and diagonal elements of sub-matrix M are obtained as,

nee ..... (2.39)

rin
g.n ..... (2.40)

Finally, the off-diagonal and diagonal elements of L are given by et

..... (2.41 )

..... (2.42)

It can be noticed that

Lik =- Hik

and Nik = Mik

This property of symmetry of the elements reduces computer time and storage.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 19

Treatment of Generator Buses

At all generator buses other than the swing bus, the voltage magnitudes are specified in addition
to the real powers. At the ith generator bus
2 '2 "2 ..... (2.43)
IVj 1 =v, +v,
Then, at all the generator nodes, the variable llQ j will have to be replaced by
2
lllVd

2 _ a~ll,12) , a~t1,12) .
But, 1t1Vj 1 - ,llV, + . t1v.

ww av, av,

w.E
This is the only modification required to be introduced in eqn. (2.40)
..... (2.44)

a
2.4.2 The Polar Coordinates Method.
syE
The equation for the complex power at node i in the polar form is given in eqn. (2.10) and the

ngi
real and reactive powers at bus i are indicated in eqn. (2.11) and (212). Reproducing them here
once again for convenience.

Pj = t nee
lv, Vk Yjklcos(o, - Ok - ejk ) ..... (2.11)

rin
k=1

n
L
and Q, =
k=1
IV j Vk Vjk Isin (OJ -Ok - e,k)
g.n ..... (2.12)

The Jocobian is then formulated in terms of IV I and 0 instead of Vj' and Vj" in this case.
Eqn. (2.27) then takes the form et
[:~l = [: ~l[II~~ Il ... (2.45)

The off-diagonal and diagonal elements of the sub-matrices H, N, M and L are determined
by differentiating eqns. (2.11) and (2.12) with respect to 0 and IVI as before. The off-diagonal
and diagonal elements of H matrix are

OP =IVj Vk V'klsin(o,-ok-e'k1j;~k
j
..... (2.46)
aO k

..... (2.47)

Downloaded From : www.EasyEngineering.net

20 Operation and Control in Power Systems

The Jocobian is [hen formulated in terms of IV I and 8 instead of V,' and V," in this case.
Eqn. (2.27) then takes the form

[~]=[: :][~~~I] ..... (2.45)

The off-diagonal and diagonal elements of the sub-matrices H, N, M and L are determined
by differentiating eqns. (2.11) and (2.12) with respect to 8 and IVI as before. The off-diagonal
and diagonal elements of H matrix are

oP, = lV, Vk Y'klsin(8, -8 k -e'k~i;t:k ..... (2.46)

o8 k

ww aP
-' = I
aD. K=I
n
lv, Vk Y,kI COS(D, -Dk -8,k1i;t: k ..... (2.47)

w.E
The off-diagonal and diagonal elements ofN matrix are
~=IV Yklsin(o
al V i ' "
-Ok -8 k)
,
..... (2.48)

a ap
al V, I
k

syE k=1
n
--'- = 21 Vi YII ICose'i + L IV. V,. ICos(8, - 8 k - O,k) ..... (2.49)

ngi k;t'
The off-diagonal and diagonal elements ofM matrix are
aQ, = -Iv, Vk
aO k
aQ n
nee
Y,k ICos (0, - Ok - e,k) ..... (2.50)

- ' = L IV; Vk V'kICOS(O, -Ok -e,.)

aO, k=1 rin ..... (2.51 )

Finally, the off-diagonal and diagonal elements of L matrix are

g.n
01oQ,
Vk I
= lV, Y.k Isin (8 - 8 k - e'k)
" " et
..... (252)
.

~ =12V, Yi, ICOSe'i

oI V, I
+ f
k=1
IV k Yik ISin (8, - Ok - e,k) ..... (2.53)

It is seen from the elements of the Jacobian in this case that the symmetry that existed
in the rectangular coordinates case is no longer present now. By selecting the variable as ~8
and ~ IVI / IVI instead equation (2.45) will be in the form

[ ~P] [H N] [~~'l
~Q
= M L IVI ..... (2.54 )

In this case it will be seen that

Hik = L,k

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 21

and N,k = - M,k

or, in other words, the symmetry is restored. The number of elements to be calculated for an
n-dimensional Jacobian matrix are only n + n2/2 instead of n2, thus again saving computer
time and storage. The flow chart for computer solution is given in Fig. 2.3.

ww
w.E
asy ~p(m) = P (scheduled) _ P (m)
"

En
l\Q:m) '" Q, (scheduled) _ Q,(m)
,

gi nee
r Yes
ing
.
Calculate line flows
net
and power at slack
bus

Solve the equation:

[~ ~]=[: ~] [~ I~I]
Calculate the changes in variables
IV 1(111 + 1)= IV 1(111) + 6 V 1(111)
8(111+ 1)=8(111)+68 (111)
1

m=m+ I

Fig. 2.3 Flow chart for Newton - Raphson method (Polar coordinates) for load flow solution

Downloaded From : www.EasyEngineering.net

22 Operation and Control in Power Systems

Treatment of Generator Nodes

For a PV-bus, the reactive power equations are replaced at the ith generator bus by

IVI 12 =v '2j +Vj"2

The elements of M are given by

a~vlY .
M
I
k =--
aO k
=0'1'*
'
k

ww
and

Then elements of L are given by

w.E Iv 1= L· =
Ik
afty. 12)
q
alvkl
I O·j '* k

asy a~VjlY
k ,

and
En d L j, = ~Vd IV,I =21 V
2

gin
Newtons method converges in 2 to 5 iterations from a flat start ( {V} = 1.0 p.u and 0 = 0)

eer
independent of system size. Previously stored solution can be used as starting values for rapid
convergence. Iteration time can be saved by using the same triangulated Jacobian matrix for

in
two or more iterations. For typical, large size systems, the computing time for one Newton -
Raphson iteration is roughly equivalent to seven Gauss - Seidel iterations.

g.n
The rectangular formulation is marginally faster than the polar version because there are
no time consuming trigonometric functions. However, it is observed that the rectangular
coordinates method is less reliable than the polar version.

2.5 Sparsity of Network Admittance Matrices

et
For many power networks, the admittance matrix is relatively sparse, where as the impedance
matrix is full. In general, both matrices are nonsingular and symmetric. In the admittance
matrix, each non-zero off diagonal element corresponds to a network branch connecting the
pair of buses indicated by the row and column of the element. Most transmission networks
exhibit irregularity in their connection arrangements, and their admittance matrices are relatively
sparse. Such sparse systems possess the following advantages:
I. Their storage requirements are small, so that larger systems can be solved.
2. Direct solutions using triangularization techniques can be obtained much faster unless
the independent vector is extremely sparse.
3. Round off errors are very much reduced.
The exploitation of network sparsity requires sophisticated programming techniques.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 23

2.6 Triangular Decompostion

Matrix inversion is a very inefficient process for computing direct solutions, especially for
large, sparse systems. Triangular decomposition of the matrix for solution by Gussian
elimination is more suited for load flow solutions. Generally, the decomposition is accomplished
by elements below the main diagonal in successive columns. Elimination by successive rows
is more advantageous from computer programming point of view.
Consider the system of equations
Ax=b ..... (2.55)
where A is a nonsingular matrix, b is a known vector containing at least one non-zero element

ww
and x is a column vector of unknowns.
To solve eqn. (2.55) by the triangular decomposition method, matrix A is augmented by
b as shown
w.E
[~.: . asy ·:':··1
a l2 a ln
a 22 a 2n

anIEn a n2 ann bn

gi nee
The elements of the first row in the augmented matrix are divided by all as indicated by
the following step with superscripts denoting the stage of the computation.

J '-
lj -- ( - 1 alJ'J - 2, .........n
a (I)
all
r ing ..... {2.56)

bll)=(_IJb l
all
. net
..... (2.57)

In the next stage a21 is eliminated from the second row using the relations

a (I) (I) . 2
2 ) = a 2) -a 21 a l), J = ,......... ,n
..... (2.58)

b(l) -
2 -
b
2
-a 21 b(1)
I
..... (2.59)

2) =
a (2) (1m )a (1)
2j , J• = 3, ......... , n
a 22

b(2)
2
= (_1_) (I)
b(l)
2
..... (2.60)
a 22

Downloaded From : www.EasyEngineering.net

24 Operation and Control in Power Systems

The resulting matrix then becomes

I 1(1) a(l) a(l) b(l)

12 13 In I
0 I a (2) a(2) b(2)
23 2n 2

ani a n2 a n3 ann bn

using the relations

. - 2
b (J)
3) -- a 3) - a 3l a (I)
l) J - ,............ , n ..... (2.61 )

ww l
b(3)= b3 - a 31 b(l)
I ..... (2.62)

w.E a (2)
3)

b(2) -- b(l)
_
- a 3(I)) -a (I)
(I)b(2)' - 4
(2) • - 3
32 a 2) ,J - ,............. ,n ..... (2.63)

asy 1 3 3 -a 32 3 ,J- , ....... ,n ..... (2.64 )

En a (3)
3
-
- ( I a (I)
a(2) 3j ,J' -
- 4, ....... ,n ..... (2.65)

gi
33

nee ..... (2.66)

r ing
The elements to the left of the diagonal in the third row are eliminated and further the
diagonal element in the third row is made unity.
After n steps of computation for the nth order system of eqn. (2.55), the augmented
matrix will be obtained as
. net
a(l)
12
o I

o o
By back substitution, the solution is obtained as
x = ben) ..... (2.67)
n n

xn _ l -- b(n-I) (n-I) ,n ........ xn

n-I -an_I ..... (2.68)

..... (2.69)

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 25

For matrix inversion of an nth order matrix, the number of arithmetical operations required

is n3 while for the triangular decomposition it is approximately (n; ) .

2.7 Optimal Orde~g

When the A matrix in eqn. (2.55) is sparse, it is necessary to see that the accumulation of
non - zero elements in the upper triangle is minimized. This can be achieved by suitably
ordering the equations, which is referred to as optimal ordering.

ww
Consider the network system having five nodes as shown in Fig. 2.4

w.E 3

asy
En
gi nee
4

Fig. 2.4 A Five Bus system

rin
The y-bus matrix of the network will have entries as follows
g.n
2
x

x
2 3 4
x

x
x x

0 0 =y
et
..... (2.70)
3 x 0 x 0
4 x 0 0 x

After triangular decomposition the matrix will be reduced to the form

234
x x x

2 0 x x =y ..... (2.71)
300 1 x

4 000

Downloaded From : www.EasyEngineering.net

26 Operation and Control in Power Systems

By ordering the nodes as in Fig. 2.5 the bus admittance matrix will be of the form
2 3 4
x 0 0 x

2 0 x 0 x =Y ..... (2.72)
3 0 0 x x

4 x x x x

ww 3

w.E @

asy 4

En
Fig. 2.5 Renumbered five bus system

gi
As a result of triangular decomposition, the V-matrix will be reduced to
234
nee
o
2 0
0 x
o x =Y rin ..... (2.73)
300 x
g.n
400 0
et
Thus, comparing the matrices in eqn. (2.71) and (2.73) the non-zero off diagonal entries
are reduced from 6 to 3 by suitably numbering the nodes.
Tinney and Walker have suggested three methods for optimal ordering.
1. Number the rows according to the number of non-zero, off-diagonal elements before
elimination. Thus, rows with less number of off diagonal elements are numbered
first and the rows with large number last.
2. Number the rows so that at each step of elimination the next row to be eliminated is
the one having fewest non-zero terms. This method required simulation of the
elimination process to take into account the changes in the non-zero connections
affected at each step.
3. Number the rows so that at each step of elimination, the next row to be eliminated is
the one that will introduce fewest new non-zero elements. This requires simulation
of every feasible alternative at each step.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 27

Scheme I. is simple and fast. However, for power flow solutions, scheme 2. has proved to be
advantageous even with its additional computing time. If the number of iterations is large,
scheme 3. may prove to be advantageous.

2.8 Decoupled Methods

All power systems exhibit in the steady state a strong interdependence between active powers
and bus voltage angles and between reactive power and voltage magnitude. The coupling
between real power and bus voltage magnitude and between reactive power and bus voltage
phase angle are both relatively weak. This weak coupling is utilized in the development of the
so called decoupled methods. Recalling equitation (2.54)

ww ~P] [H
[ ~Q - M
NI ~o ]
L IVI/~IVI
w.E
by neglecting Nand M sub matrices as a first step, decoupling can be obtained so that

and asy I ~p I = I HI· I~ol

I ~Q I = I L I . I ~ I VI ! I V I
..... (2.74)
..... (2.75)

En
The decoupled method converges as reliability as the original Newton method from

gi
which it is derived. However, for very high accuracy the method requires more iterations

nee
because overall quadratic convergence is lost. The decoupled Newton method saves by a
factor of four on the storage for the J - matrix and its tri-angulation. But, the overall saving is

iteration is 10 to 20% less than for the original Newton method.

r
35 to 50% of storage when compared to the original Newton method. The computation per

ing
2.9 Fast Decoupled Methods
. net
For security monitoring and outage-contingency evaluation studies, fast load flow solutions
are required. A method developed for such an application is described in this section.
The elements of the sub-matrices Hand L (eqn. (2.54» are given by
H'k = I (VI Vk Ylk ) I sin (01 - Ok - elk)
= I (VI Vk Ylk ) I sin 0lk Cos elk - cos 0lk Cos elk)
= I VI Vk I [G lk sin 0lk - B Ik cos ~\d
where 01 - Ok = 0lk
Hkk = - 2:IVI Vk YI~ Isin(o, Ok elk)
= + I VI 12 1YII I sin elk -I Vi 121 Vlk I sin elk
- 2:IVI Vk Yiklsin(Oi -Ok -elk)

Downloaded From : www.EasyEngineering.net

28 Operation and Control in Power Systems

Lkk = 2 V, Y" sin ell + 2: VkY,k sin(o, -()k - e,k)

With lllVI / IVI formulation on the right hand side,

LKK = 2 IV,2 YII I sin ell + 2: lv, Vk Y,k Isin{o, - Ok - e,k)

Assuming that
Cos O,k == I
Sin O,k == 0

ww Gik sin 0ik

0, ~ Bi , IV?I
~ B,k
and
w.E Hik = -I V, Vk I B,k

asy Hkk = I V,2 1 B"

Lkk = I V,21 B"

En
L,k = I

Rewriting eqns. (2.74) and (2.75)

V,2 1 B,k
gi nee
pi ~VIB' I V~O
r
III = ..... (2.76)

IllOI =1 V I B" I V III I V I ing ..... (2.77)

or
IVI
. net
pi B'[~o]
III = ..... (2.78)
IVI

IllOI = B"[~ I V I] ..... (2.79)

IVI
Matrices B' and B" represent constant approximations to the slopes of the tangent hyper
planes of the functions ~p / IVI and 110 / IVI respectively. They are very close to the Jacobian
sub matrices Hand L evaluated at system no-load.
Shunt reactances and off-nominal in-phase transformer taps which affect the Mvar
flows are to be omitted from [B'] and for the same reason phase shifting elements are to be

omitted from [B"].

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 29

Both ~.] and [B"] are real and spars e and need be triangularised only once, at the
beginning of the study since they contain network admittances only.
The method converges very reliably in two to five iterations with fairly good accuracy
even for large systems. A good, approximate solution is obtained after the pI or 2nd iteration.
The speed per iteration is roughly five times that of the original Newton method.

2.10 Load Flow Solution Using Z Bus

2.10.1 Bus Impedance Formation
Any power network can be formed using the following possible methods of construction.

wwI. A line may be added to a reference point or bus.

w.E
2. A bus may be added to any existing bus in the system other than the reference bus
through a new line, and
3. A line may be added joining two existing buses in the system forming a loop.

asy
The above three modes are illustrated in Fig. 2.6

S,,,,~_i
E ngi B
k
0-+-<>----0 New
System lime bus
i
lime

~
_ _O New System rk
k bus
nee
(a)
Line added to
(b)

rin
Line added to any bus other
(c)
Line added joining two
reference bus than reference line
g.n existing buses

2.10.2
Fig. 2.6 Building of Z - Bus

Addition of a Line to the Reference Bus

et
If unit current is injected into bus k no voltage will be produced at other buses of the systems.
Zlk = ~I = 0, i '* k ..... (2.80)
The driving point impedance of the new bus is given by
..... (2.81 )
i = 1.0
. l ~I----
...

J
System :-+-----<1 ~ l
Fig. 2.7 Addition of the line to reference line

Downloaded From : www.EasyEngineering.net

30 Operation and Control in Power Systems

2.10.3 Addition of a Radial Line and New Bus

Injection of unit current into the system through the new bus k produces voltages at all other
buses of the system as shown in Fig. 2.8
These voltages would of course, be same as that would be produced if the current were
injected instead at bus i as shown.
i = 1.0
. Z ~I----
....

System :>-+------<I J~ l
ww Fig. 2.8 Addition of a radial line and new bus

w.E
Therefore,

asy
therefore,

E Zmk = Zml' m
ngi
-:t:. k
The dimension of the existing Z - Bus matrix is increased by one. The off diagonal
..... (2.82)

nee
elements of the new row and column are the same as the elements of the row and column of
bus i of the existing system.

2.10.4 Addition of a Loop Closing Two Existing Buses in the System

rin
Since both the buses are existing buses in the system the dimension of the bus impedance
matrix will not increase in this case. However, the addition of the loop introduces a new axis
which can be subsequently eliminated by Kron's reduction method.
g. net
~
~ _~o_
System
:TikZ,tne
o___+--4~
System IV -
~
1=1.0

'-------' r
(a) (b)

Figure 2.9 (a) Addition of a loop (b) Equivalent representation

The systems in Fig. 2.9(a) can be represented alternatively as in Fig. 2.9(b).

The link between i and k requires a loop voltage
"Yloop = 1.0 (Zii - Z2k + Zkk - Zlk +Z line) ..... (2.83)
for the circulation of unit current

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 31

The loop impedance is

Zloop = ZII + Zkk - 2Z Ik + Zhne ..... (2.84)
The dimension of Z matrix is increased due to the introduction of a new axis due to the
loop 1

and

Zr_m = Zun -Zkm;m 7: e

The new loop axis can be eliminated now. Consider the matrix

ww
w.E
It can be proved easily that

asy ..... (2.85)

En
using eqn. (2.85) all the additional elements introduced by the loop can be eliminated.
The method is illustrated in example 2.3

gi
2.10.5 Gauss - Seidel Method Using Z-bus for Load Flow Solution
nee
An initial bus voltage vector is assumed as in the case of Y - bus method. Using these
voltages, the bus currents are calculated using eqn. (2.6) or (2.7).
rin
P .Q
1= I-J I_ y v
y' I I
g.n ..... (2.86)

where YI is the total shunt admittance at the bus i and

from bus i to ground.
YII VI
et
is the shunt current flowing

A new bus voltage estimate is obtained for an n-bus system from the relation.
Y bus = Zbus Ibus + YR ..... (2.87)

Where YR is the (n - 1) x 1 dimensional reference voltage vector containing in each

element the slack bus voltage. It may be noted that since the slack bus is the reference bus, the
dimension of the Zbus is (n - I x (n - I).
The voltages are updated from iteration to iteration using the relation

..... (2.88)

Downloaded From : www.EasyEngineering.net

32 Operation and Control in Power Systems

Then

i = 1,2, ........ , n
S = slack bus

2.11 Load Flow Solution with Static Load Model

It can be seen that the system loads vary with both freguency and voltage. The static load
model for small changes in voltages is defined by egn. (2.89). The Newton - Rapson method
in polar coordinates form is ideally suited for incorporating the load model for load flow solution.
At nay bus i, with the ~ubscript N stands for nominal values

ww Iv,l
Pi = P,(N) [ IVi(N) I]
3

w.E Iv,l
b ..... (2.89)

asy Q, = Q'(N) [ IV'(N)I ]

En
Differentiating egn. (2.89) with respect to V

oP,
gi [ Iv,l
0IV, I = P,(N) ·a· IV'(N)I
](8-1)
nee
1 OP,(N) [ Iv,l
IV'(N)I + o!V,! IV'(N)I
]8
..... (2.90)

But from egn. (2.49)

r ing ..... (2.91 )

substituting egn. (2.91) in egn. (2.90)

(8-1)
. net
OPt Pi(N) jVi I
-!-'I=a-- I ' - I- I
av, IV,(N) Vi(N) [ ] +2IV,Y,dcos9jj

..... (2.92)

Similarly,

(b-I) [J(b)
OQ , _ . . b.1& _1_+ OQ,(N) 1&
oV - Q'(N)
I
[ IV I] IV I oV Iv I ..... (2.93)
I(N) ',(N) I ,(N)

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 33

But from eqn. (2.53)

alv I =-2I V V.lsinS. + ~IV y

OQi(N)
1
1 11 11 ,,",11k
k~
I sin(o. -0
1 k
-e.ik ) ..... (2.94)
k.. l

Substituting eqn. (2.94) into eqn. (2.93)

ww
w.E ..... (2.95)

asy
En
Inclusion of static load model into load flow equations required the use of eqn. (2.92)
and (2.95) in place of eqn. (2.49) and (2.53). In other words, the diagonal elements of the sub

of eqn. (2.54) is now not feasible. gi

matrices N and L of eqn. (2.45) have to be modified. The advantage that is gained by the use

nee
2.12 Comparision of Various Methods for Power Flow Solution
r ing
The requirements of a good power flow method are - high speed, low storage, and reliability
for ill-conditioned problems. No single method meets all these requirements. It may be mentioned

. net
that for regular load flow studies NR-method in polar coordinates and for special applications
fast decoupled load flow solution methods have proned to be most useful than other methods.
NR-method is versatile, reliable and accurate. Fast decoupled load flow method is fast and
needs the least storage.
Convergence of iterative methods depends upon the dominance of the diagonal elettlents
of the bus admittance matrix. •

Advantages of Gauss-Seidel method:

1. The method is very simple in calculations and thus programing is easier.
2. The storage needed in the computer memory is relatively less.
3. In general, the method is applicable for smaller systems.

Disadvantages of Gauss-Seidel Method:

1. The number of iterations needed is generally high and is also dependent on the acceleration
factor selected.

Downloaded From : www.EasyEngineering.net

34 Operation and Control in Power Systems

2. For large systems, use of Gauss-Seidel method is practically prohibitive.

3. The time for convergence also increases dramatically with increase of number of buses.

Advantages of Newton-Raphsan Method,'

1. The method is more accurate, faster and reliable.
2. Requires less number of iteration for convergence. Infact, in three to four iterations
good convergence is reached irrespective of the size of the system.
3. The number of iterations required is thus independent of the size of the system or the
number of buses in the system.

ww 4. The method is best suited for load flow solution to large size systems.
5. Oecoupled and fast decoupled power flow solution can be obtained from Newton Raphsan

w.E
Polar Coordinates method. Hence, it also can serve as a base for security and
contingency studies.

asy
Disadvantages of Newton-Raphsan Method,'

E
I. The memory needed is quite large for large size systems.

ngi
2. Calculations per iteration are also much larger than Gauss-Seidel method.
3. Since, it is a gradient method, the method is quite involved and hence, programming is

nee
also comparatively difficult and complicated.
E 2.1 A three bus power system is shown in Fig. E2.1. The system parameters are given in

rin
Table E2.1 and the load and generation data in Table E2.2. The voltage at bus 2 is

g.n
maintained at I.03p.u. The maximum and minimum reactive power limits of the
generation at bus 2 are 35 and 0 Mvar respectively. Taking bus 1 as slack bus obtain the
load flow solution using
(a)
(b)
Gauss - Seidel iterative method using YBus
Newton - Raphson polar coordinates method using YBus
et

Fig. E 2.1 A three bus power system

Downloaded From : www.EasyEngineering.net

LOlld Flow Analysis 35

Table E 2.1 Impedance and Line charging Admittances

Bus Code i-k Impedance (p.u.) Zjk Line charging Admittance (p.u) Yj

1-2 0.08 + jO.24 0

1-3 0.02 + jO.06 0

2-3 0.06+jO.018 0

Table E 2.2 Scheduled Generation, Loads and Voltages

Bus No i Bus voltage Vi Generation Lood

ww I 1.05+jO.0
MW
-
Mvar
-
MW
0
Mvar
0

w.E 2
3
1.03 +jO.O
--
20
0
-
0
50
ro
20
25

Solution:
asy
En
The line admittance are obtained as

gi
Y12 == 1.25 - j3.75
Y23 == 1.667 - j5.00
YI3 == 5.00 - j15.00 nee
rin
The bus admittance matrix is formed using the procedure indicated in section 2.1 as

6.25 - j18.75 -1.25 + j3.75 - 5.0 + j15.0j g.n

(a)
YBus=-1.25
[
- 5.0
+~3.73
+ J15.0
2.9167 - j8.75 - jl.6667 + j5.0
-1.6667 + j5.0

Gauss - Seidel Iterative Method using Y BUS

6.6667 - j20.0 et
The voltage at bus 3 is assumed as I + jO. The initial voltages are therefore

viOl = 1.05 + jO.O

v~O) = 1.03 + jO.O
v~O) = 1.00 + jO.O

Base MVA == 100

Iteration 1 : It is required to calculate the reactive power Q2 at bus 2, which is a P-V or voltage
controlled bus

Downloaded From : www.EasyEngineering.net

36 Operation and Control in Power Systems

e~(new) =IV2 1sCh COS0 2 =(1.03)(1.0) =1.03

e~(new) = Y2sch sin o~O) = (1.03)(0.0) = 0.00

Q~O) =[(e~(new») 8 22 + (e;(OCW») 8 22 ] +

ww Substituting the values

Q~O) = 1(1.03.)2 8.75 + (0 Y8.75 j+ 0(1.05)(-1.25) + 0.(-3.75)
w.E - 1.03[(0)(-1.25) - (1.05)(-3.75)]
+ (0)[(1)(-1.6667) + (0)(-5.0)]
a syE
-1.03[(0)( -1.6667) - (1)(-5)]
-= 0.07725
Mvar generated at bus 2
ngi
Mvar injection into bus 2 + load Mvar
0.07725 + 0.2 = 0.27725 p.u. nee
27.725 Mvar
This is within the limits specified. rin
The voltage at bus i is
g.n
y(m+l)
I
= ~[PI
y
- jQI _ ~y. y(m+l) - ~y y(m)]
y(m)*
il
£... Ik k
I k=1
£... Ik k
k=1+1
et
y(1) _ _
2 - y
1_[P1 - jQ2 - Y
y{O)* 21
y
1
- y
23
y{O)]
3
22 I

1
=-----
(2.9167 - j8.75)

- 0.3 - 0.?7725 _ (-1.25 + j3.75)(1.05 + jO.O) + (-1.6667 + j5.0)(1 + jO.O)]

[ 1.03 - JO.O

Yil) = 1.01915 - jO.03249I

= 1.0 I 96673L -1.826°

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 37

An acceleration factor of 1.4 is used for both real and imaginary parts.
The accelerated voltages is obtained using

v~ = 1.03 + 1.4(1.01915 -1.03) =1.01481

v; =0.0 + 1.4(-0.032491- 0.0) =-0.0454874
V~I)(accelerated) = 1.01481- jO.0454874

= 1.01583L - 2.56648°

ww
The voltage at bus 3 is given by

2 - Y
w.E
V(I) - _1_[P3 - jQ3 - y V - Y V(I)]
33
V(O)·
3
31 I 32 2

a
6.6667 - j20 syE
ngi
nee
=1.02093 - jO.0351381
rin
The accelerated value of VP) obtained using
g.n
v~ = 1.0 + 1.4(1.02093 - 1.0) = 1.029302
v; =0 + 1.4(-0.0351384 - 0) =-0.0491933
et
vjl) = 1.029302 - jO.049933

= 1.03048L - 2.73624 0

The voltages at the end of the first iteration are

VI == 1.05 + jO.O

V~I) = 1.01481- jO.0454874

vjl) = 1.029302 - jO.0491933

Downloaded From : www.EasyEngineering.net

38 Operation and Control in Power Systems

Check for convergence: An accuracy of 0.001 is taken tor convergence

,~O) [ , ~I) [ , ~O)
[ ~v21 = v2J - V2J = 1.01481-1.03 = -0.0152
,,~O) [" ~I) [" ']<0)
[~v2J = v2J - V2J = -0.0454874-0.0 = -0.0454874

[~V3"JO) = [,v3 JI) - [,V3J~O) = 1.029302 -1.0 = 0.029302

[~v;r) =[~v;r -[~v;r) =-0.0491933-0.0=-0.0491933

ww The magnitudes of all the voltage changes are greater than 0.001.

w.E
Iteration 2 : The reactive power Q2 at bus 2 is calculated as before to give

L_" 1<1)
8(1) =tan-I~=tan-I -0.0454874 =-2.566480
[ ]

asy 2 . [v~r 1.01481

En [v~r) =IV2schl·coSO~I) = 1.03cos(-2.566480) = 1.02837

gi[v;r = IV2schl·sino~1)
nee
=1.03sin(-2.566480) = -0.046122
[v2new P = 1.02897 - jO.046 122
rin
Q~I) :::: (1.02897)2 (8.75) + (-0.046122)2 (8.75) g.n
+( -0.046 122)[1.05(-1.25) + (0)(-3.75)]
- (1.02897)[(0)(-1.25) - (1.05)(-3.75)] +
et
- (1.02897)[(-0.0491933)(-1.6667) - (1.029302)(-5)]
=-0.0202933
Mvar to be generated at bus 2
= Net Mvar injection into bus 2 + load Mvar
= - 0.0202933 + 0.2 = 0.1797067 p.u. = 17.97067Mvar
This is within the specified limits. The voltages are, therefore. the same as' before
VI = 1.05 + jO.O
Vii) :::: 1.02897 - jO.0.46122

VJI) :::: 1.029302 - jO.0491933

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 39

The New voltage at bus 2 is obtained as

V(2) _ 1 [ - 0.3 + jO.0202933 ]

2 - 2.9167 - j8.75 1.02827 + jO.046122

- (-1.25 + j3.75)(1 .;.. 05 + jO)

- (-1.6067 + j5)· (1.029302 - jO.0491933)]
= 1.02486 - jO.0568268

The accelerated value of vi 2) is obtained from

ww v'2 = 1.02897 + 1.4(1.02486 - 1.02897) = 1.023216

w.Ev; = -0.046122 + 1.4(-0.0568268) - (-0.046122 = -0.0611087)

asy v~2)' = 1.023216-'jO.061 1087

The new voltage at bus 3 is calculated as

E
V(2) -
3 -
1 [
ngi 1
- 6.6 + jO.25
6.6667 - j20 1.029302 + jO.0491933

nee
- (-5 + jI5)(1.05 + jO.O)

rin
- (-1.6667 + j5.0)· (1.023216 - jO.0611)]
= 1.0226 - jO.0368715

The accelerated value of VJ2) obtain{;d from g. net

v~ =1.029302 + 1.4(1.0226 -1.029302) =1.02
v~ = (-0.0491933)+ 1.4(-0.0368715)+

(0.0491933) =-0.03194278

V~2) = 1.02 - jO.03194278

The voltages at the end of the second iteration are

VI = 1.05 + jO.O

Vj2l = 1.023216 - jO.0611 087

Vfl =1.02 - jO.03194278

Downloaded From : www.EasyEngineering.net

40 Operation and Control in Power Systems

The procedure is repeated till convergence is obtained at the end of the sixth iteration.
The results are tabulated in Table E2.1 (a)

Table E2.1 (a) Bus Voltage

Iteration Bus 1 Bus2 Bus 3

0 1.05 + jO 1.03 + jO 1.0 + jO
I 1.05 + jO 1.01481-jO.04548 1.029302 - jO.049193
2 1.05 +jO 1.023216-jO.0611087 1.02 - jO.0319428
3 1.05 + jO 1.033476 -jO.0481383 1.027448 - jO.03508

ww 4 1.05 + jO 1.0227564 - jO.051329 1.0124428 - jO.034 1309

w.E
5 1.05 +jO 1.027726 - jO.0539141 1.0281748 - jO.0363943
6 1.05 + jO 1.029892 - jO.05062 1.020301- jO.0338074

asy
7 1.05 + jO 1.028478 - jO.0510 117 1.02412 - jO.034802

SI2 vt(vtE
Line flow from bus 1 to bus 2

=
ngi
- V;)vt2 = 0.228975 + jO.017396

Line flow from bus 2 to bus 1

S21 = V2(v; - V; )v;1 nee

rin
= --0.22518 - jO.0059178
Similarly, the other line flows can be computed and are tabulated in Table E2.1 (b). the
slack bus power obtained by adding the flows in the lines terminating at the slack bus, is
g.n
PI + JOI = 0.228975 + jO.017396 + 0.684006 + jO.225

= (0.912981 + jO.242396)
et
Table E2.1(b) Line Flows

Line P Power Flow Q

1-2 + 0.228975 0.017396

2-1 -0.225183 0.0059178

1-3 0.68396 0.224

3-1 -0.674565 -0.195845

2-3 -0.074129 0.0554

3-2 0.07461 -0.054

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 41

(b) Newton - Raphson polar coordinates method

The bus admittance matrix is written in polar form as

19.7642L -71.6° 3.95285L -108.4° 15.8114L -108.40]

Y BUS = 3.95285L -108.4° 9.22331L -71.6° 5.27046L -108.4°
[ 15.8114L -108.4° 5.27046L -108.4° 21.0819L -71.6°

Note that

LY II = -71.6°

ww and LY ik =-180° -71.6° =108.4°

w.E
The initial bus voltages are
VI = 1.05 LOo

asy
via) = 1.03LOO

v~O)
E =1.0LOO
ngi
The real and reactive powers at bus 2 are calculated as follows:

P2 = IV 2VIY21 Icos(&~O) -
nee
&1 - 9 21 )+Ivi y 22 lcos(- 9 22 )+

IV 2V3 Y23 Icos(&~O) - &jO) - 9 23 )

rin
g.
= (1.03) (1.05) (3.95285) cos (108°.4) + (1.03Y (9.22331) cos (-108°.4)

+ (1.03f (9.22331) cos (71°.6) +(1.03)(1.0)(5.27046) cos (-108°.4)

net
= 0.02575

Q2 = IV 2VI Y21 Isin(&iO) - &1 - 9 21 )+ Ivi Y22l sin (- 9 22 )+

IV2V3Y23Isin(&~0) -&~O) -9 23 )
= (1.03) (1.05) (3.95285) sin (-108°.4) + (1.03)2 (9.22331) sin (71.6°)
+ (1.03) (1.0) (5.27046) sin (108.4°)
= 0.07725

Generation of p.u Mvar at bus 2

= 0.2 + 0.07725
= 0.27725 = 27.725 Mvar
This is within the limits specified. The real and reactive powers at bus 3 are calculated
in a similar way.

Downloaded From : www.EasyEngineering.net

42 Operation and Control in Power Systems

P3 "'" IVjOlVIY31ICOS(0~O) -0 1-8 31 )+IVjOlV2 y 32 1 COS(O~Ol -0 2 -8 3J+

IVjO)2 Y33ICOS(-833)
"" (1.0) (1.05) (15.8114) cos (-108.4°) +
(1.0) (1.03) (5.27046) cos (-108.4°)
+ (1.0)2 (21.0819) cos (71.6°)
=-0.3

Q3 = IvjO)VIY31Isin(0~0) -0 1 -8;1)+lvjO)V2 Y32lsin(o~O) -0 2 -9 32 )+

ww IVjO)2 Y33Isin(-933)

w.E "'" (1.0) 1.05 (15.8114) sin (-108.4°) + (1.0) (1.03) (5.27046)
sin (-108.4°)
asy
+ (1.0)2 (21.0891) sin (71.6°)
=-0.9
E ngi
The difference between scheduled and calculated powers are

L\P~O) = -OJ - 0.02575 = -OJ2575

L\PjD) =-0.6 - (-0.3) =-0.3 nee
L\Q~O) = -0.25 - (-0.9) = -0.65
rin
g.n
It may be noted that L\Q2 has not been computed since bus 2 is voltage controlled bus.
since IL\P?) I. lL\PjO) 1and IL\Q ~O) I
are greater than the specified limit of 0.0 1. the next iteration is computed. et
Iteration 1 : Elements of the Jacobian are calculated as follows.

8P2 =IV2 vjO) Y23Isin(0~O) -0~0) -9 23 )

80 3
= (1.03) (1.0) (5.27046) sin (-108.4°) = -5.15
8P2 = -IV2 VI Y211sin(0~O) - OlOl - 9 21 )+
80 2

IV2 vjO) Y23Isin(o~0) -O~Ol -8 23 )

= - (1.03) (1.05) (3.95285) sin (108.4°) +
(1.03) (1.0) (5.27046) sin (-108.4°)
= 9.2056266

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 43

OP2 I (0) (0) I

00 = V2 Y23 COS\02 - 03 - 8 23
)

3
= (\.03) (5.27046) cos (108.4°)
= - \.7166724

OP3 = IV(O)V y Isin(8(O) - 8(0) - 8 )

00 3 1 31 3 2 32
3

= (0.0) (1.03) (5.27046) sin (-108.4°)

= - 5.15

ww ::3 =lviO)v1 Y31Isin(0~0)-81-831)+lviO)V2 Y32lsin(0~O)-0~O)-832)

w.E 3

= - (1.0) (1.05) (15.8114) sin (-108.4°) - 5.15

asy = 20.9

E ngi
nee
Iv 2 Y321 cos(&~O) - 8~0) - 8 32 )

=
rin
2(1.0) (21.0819) cos (71.6°) + (1.05) (15.8114) cos (-108.4°) +

=
(1.03) (5.27046) cos (-108.4°)
6.366604 g. net
00
00 -- -IV(O)V
3
3 1
y Icos(8(O) - &(0) - 8 )
32 3 2 32
2

= (1.0) (1.03) (5.27046) cos (-108.4°)

= 1.7166724

- I
00= V3(0) VI Y32 cosu3
3 (s::(O) I -U
s:: 8)
I - 31 +
00 3

IV(O)V
3 y32 Icos(o(O)2 - 0(0)
2 3 - e32 )
= (1.0) (1.05) (15.8114) cos (-108.4°) - 1.7166724
= - 6.9667

Downloaded From : www.EasyEngineering.net

44 Operation and Control in Power Systems

IV2 Y32lsin(<>~O) - <>~O) - 8 32 )

= 2(1.0) (21.0819) sin (71.6°) + (1.05) (15.8114) sin (-108.4°) +
(1.03) (5.27046) sin (-108.4°)
= 19.1
From eqn. (2.4.5)

ww - 0.325751 [9.20563 - 5.15

-0.3 = -5.15 20.9

w.E [
0.65 1.71667 - 6.9967

asy
Following the method of triangulation and back substations

j
[E -0,35386] [
-0.3
- 0.035386
= -5.15
I - 0.55944 - 0.18648][ .1<>2

ngi
20.9
+ ] .71667 - 6.9667
6.36660
] 9.]
.1<>3
.1!V3!

-0.55944 nee j
[
- 0.35386 J [1
- 0.482237 = 0 18.02 5.40623
rin
- 0.18648J[ .1<>2
.1<>3
+0.7]0746
g.
0 -6.006326 ]9.420]2 .1!V3!

- 0.35386] [I - 0.55944 net

Finally, - 0.0267613 = 0 ]
[
0.55 0 0

Thus, .1!V3! = (0.55) /(2] .22202) = 0.025917

.10 3 = - 0.0267613 - (0.3) (0.025917)

= - 0.0345364 rad
= - 1.98°
.10 2 = - 0.035286 - (-0.55944) (-0.034536) - (-0.18648) (0.025917)

= - 0.049874 rad
=- 2.8575°

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 45

At the end of the firs iteration the bus voltages are

V I == 1.05 LOo
V2 == 1.03 L2.85757°
V3 == 1.025917 L-1.9788°
The real and reactive powers at bus 2 are computed :
p?) = (1.03)(l.05)(3.95285)[cos(-2.8575) - 0(-108.4 0 )

+ (1.03)2 (1.025917)(5.27046) cos[( -2.8575) - (-1.9788) - 108.4 °

== -0.30009

ww Q~1) == (1.03)(l.05)(3.95285)[sin( -2.8575) - 0(-108.4 0 )

w.E + (1.03)2 (9.22331)sin[( -2.85757) - (-1.9788) -108.4°)]

== 0.043853

asy
Generation of reactive power at bus 2
== 0.2 + 0.043853 == 0.243856 p.u. Mvar

En
== 24.3856 Mvar

gi
This is within the specified limits.
The real and reactive powers at bus 3 are computed as
nee
pjll = (1.025917)(1.05)(l5.8117)cos[( -1.09788) - 0 -108.4°)]

r ing
+ (l.025917)(l.03)(5.27046)cos[(-1.0988) - (-2.8575) -108.4]
+ (1.025917)2 (2 1.08 I 9)cos(71.6° )
=-0.60407

Q~l) = (1.025917)(1.05)(l5.8114)sin[( -1.977) -108.4 0)]

. net
+ (l.025917)(l.03)(5.27046)sin[( -1.9788) - (-2.8575) -108.4°)]
+ (1.025917)2 (21.0819) sin(71.6°)
== -0.224
The differences between scheduled powers and calculated powers are

L1p~l) == -0.3 - (-0.30009) = 0.00009

L1pjll = -0.6 - (-0.60407) =0.00407

L1Q~I) = -0.25 - (-0.2224) = -0.0276
- (0.3) - (0.30009)
- (0.6) - (0.60407)
- (0.25) (0.2224)

Downloaded From : www.EasyEngineering.net

46 Operation and Control in Power Systems

Even though the first two differences are within the limits the last one, Q~I) is greater
than the specified limit 0.0 I. The next iteration is carried out in a similar manner. At the end of
the second iteration even dQ 3 also is found to be within the specified tolerance. The results are
tabulated in table E2.1.2(a) and E2.1.2(b)

Table E2.1.2(a) Bus voltages

Iteration Bus I Bus 2 Bus 3

0 1.05LOo 1.03LOo I.LOo
I 1.05LOo 1.03L-2.85757 1.025917L-I.9788

ww 2 1.05LOo 1.03L-2.8517 1.02476L-1.947

w.E Line
Table E2.1.2(b) Line Flows

P Power Flow Q

asy 1-2 0.2297 0.016533

En2-1
1-3
- 0.22332
0.68396
-0.0049313
0.224

gi
3-1
2-3
- 0.674565
-0.074126
nee
-0.0195845
0.0554
3-2 0.07461

r ing
E2.2 Obtain the load flow solution to the system given in example E2.1 using Z-8us. Use
- 0.054

Gauss - Seidel method. Take accuracy for convergence as 0.000 I.

Solution: . net
The bus impedance matrix is formed as indicated in section 2.10. The slack bus is
taken as the reference bus. In this example, as in example 2.1 bus I is chosen as the
slack bus.
(i) Add element 1-2. This is addition of a new bus to the reference bus

Z = (2)
BUS (2) ! 0.05 + jO.24!

(ii) Add element 1-3. This is also addition of a new bus to the reference bus

(2) (3)
(2) 0.08 + jO.24 0.0 + jO.O
ZBUS = (3) 0.0 + jO.O 0.02 + jO.06

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 47

(iii) Add element 2-3. This is the addition of a link between two existing buses 2
and 3.
Z2_loop =; ZIOop-2 =; Z22 - Z23 =; 0.08+jO.24
Z3-IOOp =; Zloop-3 =; Z32 - Z33 =; -(0.02+jO.06)
Z,oop-Ioop =; Z22 + Z33 - 2 Z23 + Z23, 23
= (0.08+jO.24 )+(0.02+jO.06)(0.06 + jO.18)
=; 0.16 +j0.48

(2) (3)

ww ZBUS =
(2)
(3)
0.08 + jO.024
0.0 + jO.O
0+ jO
0.02 + jO.06
0.08 + jO.24
- (0.02 + jO.06)

w.E e 0.08 + jO.24 - (0.02 + jO.006) 0.16 + jO.48

asy
The loop is now eliminated

En ,
Z 22 = Z 22
Z2-loop Zloop-2
- ----'---'--
Zloop-loop

gi = (0.08 + jO.24) _ (0.8 + jO.24)2

0.16 + j0.48 nee
= 0.04 + jO.12
rin
g.n
= (0.0 + jO.O) _ (0,8 + jO.24)( -0,02 - jO.06)
et
0.16+j0.48
= 0.01 + jO.03

Similarly Z~3 = 0.0175 + jO.0526

The Z - Bus matrix is thus

Z =[0.04+ jO. 12 1 0.01 + jO.03 ]=

Bus 0.01 + jO.03 0.017 + jO.0525

0.1265L71.565° I 0.031623L71.565°
[ 0.031623L71.565° 0.05534L71.565°
1

Downloaded From : www.EasyEngineering.net

48 Operation and Control in Power Systems

The voltages at bus 2 and 3 are assumed to be

V~O) ::::: 1.03 + jO.O

V~O) ::::: 1.0 + jO.O

Assuming that the reactive power injected into bys 2 is zero,
O2 = 0.0
The bus currents I~O) and 1~0) are computed as

1(0) ::::: -0.3 + jO.O ::::: -0.29126 _ '0.0 = 0.291 26LI 80°

ww 2 1.03 _ jO.O J

1(0) ::::: -0.6 + j0.25 ::::: -0.6 _ '0.25::::: 0.65LI57.38 0

w.E 3 1.0 + jO.O

Iteration 1 : The voltage at bus 2 is computed as
J

asy Vi) v) + Z22 I~O) + Z23 1(0)

En 1.05LOO + (0.1265L71.565° (0.29126LI800 +

gi (0.031623L71.565° )(0.65LI57.3So

nee
1.02485 - jO.05045
1.02609L-2.8182
The new bus current I~O) is now calculated. rin
g.n
et
::::: 1.02609L - 2.8182 x ( 1.03 _ 1)::::: 0.0309084L _ 74.38320
0.1265L71.565° 1.02609

0(0) = Im[v~1) L11~0)* ]

= Im[I.02609L - 2.8182° !0.0309084L74.383°
=0.03

o~) = O~O) + L10~0) = 0.0 + 0.03 = 0.03

1(1)= -OJ-jO.3 = 0.29383LI 82.8918°

2 1.02609L _ 2.8182°

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 49

Voltage at bus 3 is now calculated

V(I) = V +Z 1(1) +Z 1(0)
3 I 32 2 33 3

= 1.05LO.Oo + (0.031623L71.5650) (0.29383LI82.8320) +

(0.05534L71.565° )(0.65LI57.38°)

= (1.02389 - jO.036077) = 1.0245L - 2.018°

1(1) = 0.65LI57.38° = 0.634437 LI55.36°

3 1.0245L2.0 180

ww The voltages at the end of the first iteration are:

w.E
VI = 1.05 L 0°
Vii) = 1.02609L: - 2.8182°

a Vjl) = 1.0245L - 2.018°

syE
The differences in voltages are

ngi
11 V~I) = (1.02485 -10.05045) - (1.03 + jO.O)
= -0.00515 - jO.05045
11 vjl) = (1.02389 - jO.036077) - (1.0 + jO.O)
nee
= (0.02389 - jO.036077)

rin
Both the real and imaginary parts are greater than the specified limit 0.00 I.

, = VI + Z 22 1(1) Z 1(1) g.n

Iteration 2 : V_!2)
=
2 + 2J 3

1.02 LOo + (0.1265L71.5650)(0.29383 L 182.892°) +

(0.031623L71.5650) (0.63447 L 155.36°) ....
et
= 1.02634 - jO.050465
= 1.02758 L-2.81495°

= 1.02758L - 2.81495° [ 1.03 _ I]

111~1) 1.1265L-71.5650 1.02758

= 0.01923L -74.38°

I1Q~ll = Im[ Vi 2l (111~1) r]

= Im(1.02758L - 2.81495)(0.0 1913L74.38°)

= 0.0186487

Downloaded From : www.EasyEngineering.net

50 Operation and Control in Power Systems

Q~2) = Q~I) + f1Q~1)

= 0.03 + 0.0186487 = 0.0486487
-0.3 - jO.0486487
1(2) = ----=-------:-0
2 1.02758L2.81495

= 0.295763LI86.393 °
vj2) = 1.05LOO + (0.31623L71.565° (0.295763LI86.4 ° +
0.05534L71.565° )(0.634437 LI55.360)

ww 1(2) = 0.65LI57.38° = 0.6343567 L155.434 °

w.E
3 1.02466LI.9459 0

f1 V~I) = (1.02634 - jO.050465) - (1.02485 - jO.05041) = 0.00149 - jO.OOOO 15

asy
f1 vjl) = (1.024 - jO.034793) - (1.02389 - jO.036077) = 0.00011 + jO.00128

En
As the accuracy is still not enough, another iteration is required.
Iteration 3 :
gi nee
V~3) = 1.05LOo + (0.1265L71.565° )(0.295763LI86.4 0 ) +

r
(0.031623L71.565° )(0.63487 L155.434 0)
= 1.0285187 - jO.051262 ing
= 1.0298L - 2.853° . net
1(2) = 1.0298L-2.853° [~-I]=0.00158IL74.4180
2 0.1265L71.565 0 1.0298

!lQ~2) = 0.00154456
Q~3) = 0.0486487 + 0.001544 = 0.0502

1(3) = -0.3 - jO.0502 = 0.29537 LI86.6470

2 0.0298L2.853 0

v1 3
) = 1.05LOO + (O.031623L71.5650) + (0.29537 L186.6470) +

(0.05534L71.565° )(0.634357 L155.434 0 )

=1.024152 - jO.034817 =1.02474L -1.9471°

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 51

1(3) = - 0.65 - LI57.38° =0.6343LI55.4330

3 1.02474L1.9471 0

!:l V~2) = (1.0285187 - jO.051262) - (1.02634 - jO.050465)

=0.0021787 - 0.000787
!:l V~2) = (1.024152 - jO.034817) - (1.024 - jO.034793)
= 0.000152 - jO.00002
Iteration 4 :

ww vi 4
) = 1.02996L - 2.852°
~1~3) = 0.0003159L - 74.4170

w.E !:lQ~3) = 0.0000867

asy Q~4) = 0.0505

En 1~4) = 0.29537 LI86.7°

vi
gi 4
) = 1.02416 - JO.034816 = 1.02475L -1.947°
~ V~3) = 0.000108 + jO.000016
nee
The final voltages are
!:l Vj3) = 0.00058 + jO.OOOOO 1
r ing
VI = 1.05 + jO.O
V2 = 1.02996L-2.852°
. net
V3 = 1.02475L-1.947°
The line flows may be calculated further if required.
E 2.3 Consider the bus system shown in Fig. E 2.3.

4 5

E2.3 A six bus power system

Downloaded From : www.EasyEngineering.net

52 Operation and Control in Power Systems

The following is the data:

Line impedance (p.u.) Real Imaginary

1.4 0.57000 E·I 0.845 E·I
1·5 1.33000 E·2 3.600 E·2
2·3 3.19999 E·2 1.750 E·I
2·5 1.73000 E·2 0.560 E-I
2-6 3.00000 E·2 1.500 E·I

4·5 1.94000 E·2 0.625 E·I

ww
Scheduled generation and bus voltages :

Bus Code P
w.E Assumed bus voltage
MWp.u.
Generation
Mvarp.u MWp.u.
Load
Mvarp.u

I
asy 1.05 +jO.O
(specified)
... ... _.. ...

2
En ... 1.2 0.05 ... ...
3
4
5
gi ...
...
...
1.2
...
...
0.05

nee
...
...
...
1.4
0.8
...
0.05
0.03
6 ... ...
r
-..
ing
0.7 0.02

Taking bus - I as slack bus and using an accelerating factor of 1.4, perform load flow
by Gauss - Seidel method. Take precision index as 0.000 I.
Solution:
. net
The bus admittance matrix is obtained as :
Bus Code Admittance (p.u.)
P-Q Real Imaginary
I-I 14.516310 -32.57515
1-4 -5.486446 8.13342
1-5 -9.029870 24.44174
2-2 7.329113 -28.24106
2-3 -1.011091 5.529494
2-5 -5.035970 16.301400
2-6 -1.282051 6.410257
Contd. ..••

Downloaded From : www.EasyEngineering.net

Load Flow Analysis S3

Bus Code Admittance (p.o.)

P-Q Real Imaginary
3-2 -1.011091 5.529404
3-3 1.011091 -5.529404
4-1 -5.486446 8.133420
4-4 10016390 -22.727320
4-5 -4.529948 14.593900

ww 5-1
5-2
-9.029870
-5035970
24.441740
16.301400

w.E 5-4
5-5
-4.529948
18.595790
14.593900
-55.337050

asy
6-2 -1.282051 6.410257
6-6

En 1.282051 -6.410254

gi
All the bus voltages, y(O), are assumed to be I + jO except the specified voltage at bus

nee
I which is kept fixed at 1.05 + jO. The voltage equations for the fist Gause-Seidel iteration are:

v 2(I) -
-
_1_ [ P2 -
Y
2
V
2
.iQ 2
(0)'
_ Y
21
v3 (0) _ Y

r2"
V (0) _ Y
26
V

ing
(0) ]
6

y(l) =
3
_1_[P3
y
33
-jQ3 - y
y(O)*
.1
32
y(I)]
2 . net
y(l) =
4
_1_[P4 - jQ4 - Y41 Y- y 45 y(O)]
y y(O)' 5 I
44 4

y(l) -
5 - y
_1_[P5y(O)'
- jQ5 - Y y - Y y(l) - y y(l)]
51 I 51 2 54 4
55 5

y(l) =
6
_1_[P6 - jQ6 - y
y y(O)' 62
y(l)]
2
66 6

Downloaded From : www.EasyEngineering.net

54 Operation and Control in Power Systems

Substituting the values, the equation for solution are

vel) =( 1 _ '28.24100) x [1.2.- jO.05]

2 7.329113 J l-jO
- (-1.011091 + j5.529404) x (I + jO) - (-5.03597 + jI6.3014)(l + jO)
- (1- 282051 + j16.30 14)(1 + jO)
= 1.016786 + jO.0557924

=( 1.011091
ww
V(I) 1 '5.52424)X[I.2- jO.05]
3 J l-jO
- (-1.011091 + j5.529404) x (1.016786 + jO.0557924)

w.E = 1.089511 + j0.3885233

asy
v(1) =( 1 _ '22.72732) x [-1.4 + jO.005]
4

E 10.01639 J

ngi
l-jO
- (-5.486446 + j8.133342) x (1.05 + jO.)
- (-4.529948 + jI4.5939)(1 + jO)
= 0.992808 - jO.0658069
nee
=( 18.59579 rin
V(I)
5
1 _ '55.33705}X[-0.8+ jO.03]
J 1- jO
- (-9.02987 + j24.44174) x (1.05 + jO)
g. net
- (-5.03597 + jI6.3014)(1.016786 + jO.0557929)
- (-4.529948 + jI4.5939)(0.992808 - jO.0658069)
= 1.028669 - jO.O 1879179

V(I)
6
=( 1.282051
1 _ '6.410257) x [-0.7 + jO.02]
J 1- jO
- (-1.282051 - j6.41 0257) x (1.016786 + jO.0557924)
= 0.989904 - jO.0669962

The results of these iterations is given in Table 2.3 (a)

In the polar form, all the voltages at the end of the 14th iteration are given
in Table E2.3(b).

Downloaded From : www.EasyEngineering.net

ww
Table E2.3(a)

It.No Bus 2 Bus 3 Bus 4 Bus 5 Bus 6

w.E
0 1+ jO.O I + jO.O 1+ jO.O 1+ jO.O 1+ jO.O

I 1.016789 + jO.0557924 1.089511 + jO.3885233 0.992808 - jO.0658069 1.02669 - jO.O 1879179 0.989901 - jO.0669962

3
1.05306 + jO. 1018735

I 043568 + jO.089733
asy
1.014855 + jO.2323309

1.054321 + jO.3276035
1.013552 - jO.0577213

1.021136 - jO.0352727
1.042189 + jO.O 177322

1.034181 + jO.00258192
1.041933 + jO.0192121

1.014571-jO.02625271

E
4 1.047155 + jO.101896 1.02297 + jO.02763564 1.012207 - jO.0500558 1.035391 + jO.00526437 1.02209 + jO.00643566

6
1.040005 + jO.093791

1.04212 + jO.097843I
1.03515 + jO.3050814

1.027151 +':0.2901358
ngi
1.61576 -jO.04258692

1.013044 - jO.04646546
0.033319 + jO.003697056 1.014416 - jO.01319787

10.33985 + jO.004504417 1.01821-jO.001752973

nee
7 I 040509 + jO.0963405 1.031063 + jO.2994083 1014418-jO.0453101 1.033845 + jO.00430454 1.016182 - jO.00770669

8 1.041414 + jO.097518 1.028816 + jO.294465 1.013687 - jO.04561 0 I 1.033845 + jO.004558826 1.017353 -jO 0048398

rin
9 1040914+ jO.097002 1.030042 + jO.2973287 1-014148 - jO.04487629 1.033711 + jO.004413647 1.016743 - jO.0060342

\0 1.041203 + jO.0972818 1.02935 + jO.2973287 1013881-jO.04511174 1.03381 + jO.004495542 I 017089 - jO.00498989

12
1.041036 + jO.097164 1.029739 + jO.296598 1.01403 - jO.04498312

1.041127 + jO.0971998 1.029518 + jO.2960784 1.013943 - jO 04506212

1.03374 + jO.004439559

1.033761 + jO.00447096
g.n
1.016877 - jO.00558081

\.016997 - jO.00524855

14
1.041075 + jO.0971451

1.041104 + jO.0971777
1.029642 + jO.2963715 1019331-jO.04501488

1.02571 + jO.2962084 1.0013965 - jO.04504223

e
1.033749 + jO.004454002 1.016927 - jO.00543323

1.033756 + jO.004463 713 1.016967 - jO.00053283

t
Downloaded From : www.EasyEngineering.net
Downloaded From : www.EasyEngineering.net

56 Operatinn and Control in Power Systems

Table E2.3(b)

Bus Voltage magnitude (p.u.) Phase angle (0)

I 1.05 0
2 0.045629 5.3326
3 1.071334 16.05058
4 1.014964 -2.543515
5 1.033765 2.473992

ww 6 1.016981 -3.001928

w.E
E2.4 For the given sample find load flow solution using N-R polar coordinates, decoupled
method and fast decoupled method.

asy
En 5

gi nee
r ing
3 4 . net
Bus Code Line impedance Zpq Line charging

1-2 0.02 +jO.24 j 0.02

2-3 0.04 +jO.02 j 0.02

3-5 0.15 + jO.04 j 0.025

3-4 0.02 + jO.06 j 0.01

4-5 0.02 + jO.04 jO.OI

5-1 0.08 + jO.02 jO.2

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 57

Bus Code (Slack) Generation Load

Mw Mvar MW Mvar
1 0 0 0 0
2 50 25 15 10
3 0 0 45 20
~
4 0 0 40 15
5 0 0 50 25

ww
1 11.724 - j24.27
~
- \0 + j20
3
0+ jO
4
0+ jO
5
-1.724 + j4.31
2
3
-10+ j20
0+ jO
w.E10.962- j24.768 - 0.962 + j4.808
- 0.962 + j4.808 6.783 - j21.944'-
0+ jO
- 5 + j15
0+ jO
- 0.822 + j2.192
4 0+ jO
asy 0+ jO - 5+ j15 15 - j34.98 -10+ j20

En
5 -1.724 + j4.31 0+ jO - 0.82 + j2.192 - \0 + j20 12.546 - j26.447

Solution: gi E2.4 Bus admittance matrix

The Residual or Mismatch vector for iteration no: 1 is nee

dp[2] = 0.04944
dp[3] = -0.041583 r ing
dp[4] = -0.067349
dp[5] = -0.047486
. net
dQ[2] = -0.038605
dQ(3] = -0.046259
dQ[4] = -0.003703
....,
dQ[5] = -0.058334
T~e New voltage vector after iteration I is :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 1.984591 F :- 0.008285
Bus no 3 E : 0.882096 F :- 0.142226
Bus no 4 E : 0.86991 F :- 0.153423
Bus no 5 E : 0:~7581 0 F :- 0.142707

Downloaded From : www.EasyEngineering.net

S8 Operation and Control in Power Systems

The residual or mismatch vector for iteration no : 2 is

dp[2] = 0.002406
dp[3] = -0.001177
dp[4] = -0.004219
dp[5] = -0.000953
dQ[2] = -0.00 I087
dQ[3] = -0.002261
dQ[4] = -0.000502

ww dQ[5] = -0.002888
The New voltage vector after iteration 2 is :

w.E
Bus no IE: 1.000000 F : 0.000000
Bus no 2 E : 0.984357 F :- 0.008219

asy
Bus no 3 E : 0.880951 F :- 0.142953

En
Bus no 4 E : 0.868709 F :- 0.154322
Bus no 5 E : 0.874651 F :- 0.143439

dp[2] = 0.000005
gi
The residual or mismatch vector for iteration no : 3 is

nee
dp[3] = -0.000001
rin
dp[4] = -0.000013
dp[5] = -0.000001 g.n
dQ[2] = -0.000002
dQ[3] = -0.000005
et
dQ[4] = -0.000003
dQ[5] = -0.000007
The final load flow solution (for allowable error.OOOI) :
bus no I Slack P = 1.089093 Q = 0.556063 E = 1.000000 F = 0.000000
bus no 2 pq P = 0.349995 Q = 0.150002 E = 0.984357 F = -0.008219
bus no 3 pq P = -0.449999 Q = -0.199995 E = 0.880951 F = -0.1429531
bus no 4 pq P = -0.399987 Q = -0.150003 E = 0.868709 F = -0.154322
bus no 5 pq P = -0.50000 I Q = -0.249993 E = 0.874651 F = -0.143439

Downloaded From : www.EasyEngineering.net

Load Flow Analysis

Decoupled load flow solution (polar coordinate method)

The residual or mismatch vector for iteration no : 0 is
dp[2] = 0.350000
dp[3] = -0.450000
dp[ 4] = -0.400000
dp[5] = -0.500000
dQ[2] = -0.190000
dQ[3] = -0.145000

ww
dQ[4] = -0.130000

w.E
dQ[5] = -0.195000
The new voltage vector after iteration 0 :

asy
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E ; 0.997385 F ;- 0.014700

E
Bus no 3 E: 0.947017 F:- 0.148655
Bus no 4 E: 0.941403 F:- 0.161282 ngi
Bus no 5 E : 0.943803 F :- 0.150753
nee
The residual or mismatch vector for iteration no: I is
dp [2] = 0.005323 rin
dp[3] = -0.008207
dp[4] = -0.004139
g. net
dp[5] = -0.019702
dQ[2] = -0.067713
dQ[3] = -0.112987
dQ[4] = -0.159696
dQ[5] = -0.210557
The new voltage vector after iteration 1 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.982082 F :- 0.013556
Bus no 3 E : 0.882750 F ;- 0.143760
Bus no 4 E : 0.870666 F :- 0.154900
Bus no 5 E : 0.876161 F :- 0.143484

Downloaded From : www.EasyEngineering.net

60 Operation and Control in Power Systems

------\~-------------------------------------------------
S
The residual or mismatch vector for iteration no:2 is
dp[2] = 0.149314
dp[3] = -0.017905
dp[4] = -0.002305
dp[5] = -0.006964
dQ[2] = -0.009525
dQ[3] = -0.009927
dQ[4] = -0.012938

ww dQ[5] = 0.007721

w The new voltage vector after iteration 2 :

.Ea
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.981985 F:- 0.007091

syE
Bus no 3 E : 0.880269 F :- 0.142767
Bus no 4 E: 0.868132 F:- 0.154172

ngi
Bus no 5 E : 0.874339 F :- 0.143109

nee
The residual or mismatch vector for iteration no:3 is
dp[2] == 0.000138
dp[3] = 0.001304
dp[4] = 0.004522
r ing
dp[5] = -0.006315
dQ[2] = 0.066286
. net
dQ[3] = 0.006182-
dQ[4] = -0.001652
dQ[5] == -0.002233
The new voltage vector after iteration 3 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984866 F :- 0.007075
Bus no 3 E : 0.881111 F :- 0.142710
Bus no 4 E : 0.868848 F :- 0.154159
Bus no 5 E : 0.874862 F :- 0.143429

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 61

The residual or mismatch vector for iteration no:4 is

dp[2] = -0.031844
dp[3] = 0.002894
dp[4] = -0.000570
dp[5] = 0.001807
dQ[2] = -0.000046
dQ[3] = 0.000463
dQ[4] = 0.002409

ww dQ[5] = -0.003361
The new voltage vector after iteration 4 :

w.E
Bus no 1 E: 1.000000 F : 0.000000

asy
Bus no 2 E : 0.984866 F :- 0.008460
Bus no 3 E: 0.881121 F:- 0.142985

E
Bus no 4 E : 0.868849 F :- 0.1546330
Bus no 5 E: 0.874717 F:- 0.143484
ngi
nee
The residual or mismatch vector for iteration no:5 is
dp[2] = 0.006789
dp[3] = -0.000528
rin
dp[4] = -0.000217
dp[5] -0.0000561 g.n
et
=

dQ[2] = -0.000059
dQ[3] = -0.000059
dQI4] = -0.000635
dQ[S] = -0.000721
The new voltage vector after iteration 5 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984246 F :- 0.008169
Bus no 3 E : 0.880907 F :- 0.142947
Bus no 4 E : 0.868671 F :- 0.154323
Bus no 5 E : 0.874633 F :- 0.143431

Downloaded From : www.EasyEngineering.net

62 Operation and Control in Power Systems

The residual or mismatch vector for iteration no : 6 is

dp[2] = 0.000056
dp[3] = 0.000010
dp[4] = 0.000305
dp[5] = -0.000320
dQ[2] = 0.003032
dQ[3] = -0.000186
dQ[4] = -0.000160

ww dQ[5] = -0.000267

w.E
The new voltage vector after iteration 6 :
Bus no 1 E : 1.000000 F : 0.000000

asy
Bus no 2 E : 0.984379 F :- 0.008165
Bus no 3 E : 0.886954 F :- 0.142941

En
Bus no 4 E : 0.868710 F :- 0.154314

gi
Bus no 5 E : 0.874655 F :- 0.143441

nee
The residual or mismatch vector for iteration no:7 is

r
dp[2] = - 0.001466
dp[3] = 0.000106
ing
dp[4] = -0.000073
dp[5] = 0.000156
dQ[2] = 0.000033
. net
dQ[3] = 0.000005
dQ[4] = 0.000152
dQ[5] = -0.000166
The new voltage vector after iteration 7 :
Bus no 1 E: 1.000000 F : 0.000000
Bus no 2 E : 0.954381 F :- 0.008230
Bus no 3 E : 0.880958 F :- 0.142957
Bus no 4 E : 0.868714 F :- 0.154325
Bus no 5 E : 0.874651 F :- 0.143442

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 63

The residual or mismatch vector for iteration no : 8 is

dp[2] = --0.000022
dp[3] = 0.000001
dp[ 4] = --0.000072
dp[5] = --0.000074
dQ[2] = --0.000656
dQ[3]:;= 0.000037
dQ[4] = --0.000048

ww dQ[5] = --0.000074
The new vc;>ltage vector after iteration 8 :

w.E
Bus no 1 E : 1.000000 F : 0.000000
Bus, no 2 E : 0.984352 F :- 0.008231

a syE
Bus no 3 E : 0.880947 F :- 0.142958
Bus no 4 E : 0.868706 F :- 0.154327

ngi
Bus no 5 E : 0.874647 F :- 0.143440

nee
The residual or mismatch vector for iteration no:9 is
dp[2] = 0.000318
dp[3] = --0.000022
rin
dp[4] = 0.000023
dp[ 5] = --0.000041 g.n
dQ[2] = --0.000012
dQ[3] = --0.000000
et
dQ[4] = 0.000036
dQ[5] = --0.000038
The new voltage vector after iteration 9 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984352 F :- 0.008217
Bus no 3 E : 0.880946 F :- 0.142954
Bus no 4 E : 0.868705 F :- 0.154324
Bus no 5 E : 0.874648 F :- 0.143440

Downloaded From : www.EasyEngineering.net

64 Operation and Control in Power Systems

The residual or mismatch vector for iteration no: lOis

dp[2] = 0.000001
dp[3] = -0.000001
dp[4] = 0.000017
dp[5] = -0.000017
dQ[2] = 0.000143
dQ[3] = -0.000008
dQ[4] = 0.000014

ww dQ[5] = -0.000020
The new voltage vector after iteration 10 :

w.E
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984658 F :- 0.008216

a
Bus no 3 E : 0.880949 F :- 0.142954

syE
Bus no 4 E : 0.868707 F :- 0.154324

ngi
Bus no 5 E : 0.874648 F :- 0.143440
The residual or mismatch vector for iteration no: 11 is
dp[2] = -0.000069
dp[3] = 0.000005 nee
dp[4] = -0.000006
rin
dp[5] = 0.000011
dQ[2]
,
= 0.000004 g.n
dQ[3].= -0.000000
dQ[4] = 0.000008
et
dQ[5] = -0.000009
The final load flow solution after 11 iterations
(for allowable arror.OOO 1)
The final load flow solution (for allowable error.OOOl) :
Bus no 1 Slack P = 1.089043 Q = 0.556088 E = 1.000000 F = 0.000000
Bus no 2 pq P = 0.350069 Q = 0.150002 E = 0.984658 F = -0.008216
Bus no 3 pq P = -0.450005 Q = -0.199995 E = 0.880949 F = -0.142954
Bus no 4 pq P = -0.399994 Q = -0.150003 E = 0.868707 F = -0.154324
Bus no 5 pq P = -0.500011 Q = -0.249991 E ::;0.874648 F = -0.143440

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 6$!

Fast decoupled load flow solution (polar coordinate method)

The residual or mismatch vector for iteration no:O is
dp[2] = 0.350000
dp[3] = ....{).450000
dp[4] = 0.400000
dp[5] = ....{).500000
dQ[2] = 0.190000
dQ[3] = ....{).145000
dQ[4] = 0.130000

wwdQ[5] = ....{). 195000

w.E
The new voltage vector after iteration 0 :
Bus no 1 E : 1.000000 F : 0.000000

asy
Bus no 2 E : 0.997563 F :- 0.015222
Bus no 3 E: 0.947912 F:- 0.151220

En
Bus no 4 E: 0.942331 F:- 0.163946

gi
Bus no 5 E : 0.944696F :- 0.153327

nee
The residual or mismatch vector for iteration no: I is
dp[2] = 0.004466
dp[3] = ....{).00075I
dp[4] = 0.007299
r ing
dp[5]
dQ[2]
= ....{).012407
= 0.072548
. net
dQ[3] = ....{).118299
dQ[4] = 0.162227
dQ[5] = ....{).218309
The new voltage vector after iteration 1 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.981909 F :- 0.013636
Bus no 3 E : 0.882397 F :- 0.143602
Bus no 4 E : 0.869896 F :- 0.154684
Bus no 5 E : 0.875752 F :- 0.143312

Downloaded From : www.EasyEngineering.net

66 Operation and Control in Power Syste"is

The residual or mismatch vector for iteration no: 2 is

dp[2] = 0.153661
dp[3] = -0.020063
dp[ 4] = 0.005460
dp[5] = -0.009505
dQ[2] = 0.011198
dQ[3] = -0.014792
dQ[4] = -0.000732

ww dQ[5] = -0.002874
The new voltage vector after iteration 2 :

w.E
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.982004 F :- 0.007026

asy
Bus no 3 E : 0.880515 F :- 0.142597

En
Bus no 4 E : 0.868400 F :- 0.153884

gi
Bus no 5 E : 0.874588 F :- 0.143038

nee
The residual or mismatch vector for iteration no: 3 is
dp[2] = -0.000850
dp[3] = -0.002093
dp[4] = 0.000155
r ing
dp[5] = -0.003219
dQ[2] = 0.067612
. net
dQ[3] = -0.007004
dQ[4] = -0.003236
dQ[5] = -0.004296
The new voltage vector after iteration 3 :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984926 F :- 0.007086
Bus no 3 E : 0.881246 F :- 0.142740
Bus no 4 E: 0.869014 F :;-,0.154193
Bus no 5 E : 0.874928 F :- 0.143458

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 67

The residual or mismatch vector for iteration no: 4 is
dp[2] = -0.032384
dp[3] = 0.003011
dp[4] = -0.001336
dp[5] = -0.002671
dQ[2] = -0.000966
dQ[3] = -0.000430

wwdQ[4] = -0.000232
dQ[5] = -0.001698

w.E
The new voltage vector after iteration 4 :
Bus no 1 E : 1.000000 F : 0.000000

asy
Bus no 2 E : 0.984862 F :- 0.008488
Bus no 3 E: 0.881119 F:- 0.143053

En
Bus no 4 E : 0.868847 F :- 0.154405
Bus no 5 E: 0.874717 F:- 0.143501
gi nee
The residual or mismatch vector for iteration no: 5 is
dp[2] = 0.000433
dp[3] = 0.000006 r ing
dp[4] = -0.000288
dp[5] = 0.000450 . net
dQ[2] = -0.014315
dQ[3] = -0.000936
dQ[4] = -0.000909
dQ[5] = -0.001265
The new voltage vector after iteration 6 :
~no 1 E : 1.000000 F : 0.000000
/Bus no 2 E : 0.984230 F :- 0.008463
Bus no 3 E : 0.881246 F :- 0.143008
Bus no 4 E : 0.869014 F :- 0.154357
Bus no 5 E : 0.874607 F :- 0.143433

Downloaded From : www.EasyEngineering.net

68 Operation and Control in Power Systems

The residual or mismatch vector for iteration no: 6 is

dp[2] = 0.006981
dp[3] = -{).000528
dp[4] = 0.000384
dp[51 = -{).000792
dQ[2] = 0.000331
~

dQ[3] = 0.000039
dQ[4] = -{).000155

ww dQ[5] = 0.000247

w.E
The residual or mismatch vector for iteration no: 7 is
dp[2] = -{).000144

asy
dp[3] = -{).000050
_ dp[4] = 0.000080

dQ[2] = 0.003107
E
dp[5] = -{).000068

ngi
dQ[3] = -{).000162
dQ[4] = -0.000255 nee
dQ[5] = -{).000375
rin
The new voltage vector after iteration 7 :
Bus no 1 E : 1.000000 F : 0.000000 g.n
Bus no 2 E : 0.984386 F :- 0.008166
Bus no 3 E : 0.880963 F :- 0.142943
et
-. Bus no 4 E : 0.868718 F :- 0.154316
Bus no 5 E : 0.874656F :- 0.143442
-
The residual or mismatch vector for iteration no: 8 is
dp[2] = -{).001523
dp[3] = -{).OOO 105
dp[4] =-{).000115
dp[5] = -{).000215
dQ[2] = 0.000098
dQ[3] = -{).000024
dQ[4] = -{).000037
dQ[5] = -{).000038

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 69

The new voltage vector after iteration 8 :

Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984380 F :- 0.008233
Bus no 3 E : 0.880957 F :- 0.142961
Bus no 4 E : 0.868714 F :- 0.154329
Bus no 5 E : 0.874651 F :- 0.143442
The residual or mismatch vector for iteration no: 9 is
dp[2] = -0.000045

ww dp[3] = 0.000015

wdp[ 4] = -0.000017
dp[5] = 0.000008
dQ[2] = .Ea
0.000679
dQ[3] = 0.000031
syE
dQ[ 4] = -0.000072
dQ[5] = -0.000105 ngi
The new voltage vector after iteration 9 :
Bus no 1 E : 1.000000 F : 0.000000
nee
Bus no 2 E : 0.984350 F :- 0.008230 r ing
Bus no 3 E : 0,880945 F :- 0.142958
Bus no 4 E : 0.868704 F :- 0.154326
Bus no 5 E : 0.874646 F :- 0.143440
. net
The residual or mismatch vector for iteration no: lOis
dp[2] = 0.000334
dp[3] = -0.000022
dp[4] = 0.000033
dp[5] = -0.000056
dQ[2] = 0.000028
dQ[3] = 0.000007
dQ[4] = -0.000007
dQ[5] = 0.000005

Downloaded From : www.EasyEngineering.net

70 Operation and Control in Power Systems

The new voltage vector after iteration 10 :

Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E: 0.984352 F:- 0 .. 008216
Bus no 3 E : 0.880946 F :- 0.142953
Bus no 4 E : 0.898705 F :- 0.154323
Bus no 5 E : 0.874648 F :- 0.143440
The residual or mismatch vector for iteration no: 11 is
dp[2] = -0.000013

ww
dp[3] = -0.000004

w.E
dp[4]
dp[5]
=

=
0.000003
-0.000000

dQ[3]asy
dQ[2] = 0.000149
= -0.000007
dQ[4] = 0.000020
En
dQ[5] = -0.000027,
gin
eer
The new voltage vector after iteration II :
Bus no 1 E : 1.000000 F : 0.000000
Bus no 2 E : 0.984358 F :- 0.008216
Bus no 3 E : 0.880949 F :- 0.142954
i ng.
Bus no 4 E : 0.868707 F :- 0.154324
Bus no 5 E : 0.874648 F :- 0.143440
n et
The residual or mismatch vector for iteration no: 12 is
dp[2] = -0.000074
dp[3] = 0.000005
dp[4] = -0.000009
dp[5] = -0.000014
dQ[2] = 0.000008
dQ[3] = -0.000002
dQ[4] = -0.000001
dQ[5] = -0.000000

Downloaded From : www.EasyEngineering.net

Load Flow Analysis / 71

The load flow solution

Bus no 1 Slack P = 1.089040 Q = 0.556076 E = 1.000000 F = 0.000000
Bus no 2 pq P = 0.350074 Q = 0.150008 E = 0.984358 F = -0.008216
Bus no 3 pq P = -0.450005 Q = -0.199995 E = 0.880949 F = -0.142954
Bus no 4 pq P = -0.399991 Q = -0.150001 E = 0.868707 F = -0.154324
Bus no 5 pq P = -0.500014 Q = -0.250000 E = 0.874648 F = -0.143440

ww Questions
w.E
I. Explain why load flow studies are performed?

asy
2. Discuss the classification of buses.

En
3. With a neat flow chart explain the load flow solution by Gauss - Seidel method.
4. Explain the principle and method of solution of the load flow problem by Newton-

gi
Raphson (a) rectangular coordinates and (b) polar coordinates methods.

nee
5. Compare Gauss - Seidel method and Newton Raphson method for load flow solution
6. Explain
(a) rin
(i) Decoupled load flow and
g.n
(ii) Fast decoupled load flow methods
(b) What are the application of the methods in (a) et

• Downloaded From : www.EasyEngineering.net

Downloaded From : www.EasyEngineering.net

72 Operation and Control in Power Systems

Problems

P2.t Obtain a load flow solution for the system shown in Fig. P2.1 use
(i) Gauss - Seidel method

ww (ii) N-R polar coordinator method

w.E
asy
E ngi
nee
rin
g.n
et
Bus code p-q Impedance Zpq Line charges Y pq/s
1-2 0.02 + jO.2 0.0

2-3 0.01 + jO.025 0.0

3-4 0.02 +jO.4 0.0

3-5 0.02 + 0.05 0.0

4-5 0.015 + jO.04 0.0

1-5 0.015+jO.04 0.0

Values are given in p.u. on a base of lOOMva.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 73

The scheduled powers are as follows

Bus Code (P) Generation Load

Mw Mvar MW Mvar
I (slack bus) 0 0 0 0
2 80 35 25 15
3 0 0 0 0
4 0 (I 45 15

ww 5

Take voltage at bus I as ILOo p.u.

0 0 55 20

w.E
Pl.2 Repeat problem P2.1 with line charging capacitance Yp/2 = jO.025 for each line

a
P2.3 Obtain the decoupled and fast decouple load flow solution for the system in P2.1 and

syE
compare the results with the exact solution.
- P2.4 For the 51 bus system shown in Fig. P2.1, the system data is given as follows in p.u.

ngi
Perform load flow analysis for the system

Line data
nee
Resistance Reactance Capacitance

2-3

3-4
0.0287

0.0028
0.0747

0.0036 rin
0.0322

0.0015

3-6

3-7
0.0614

0.0247
0.1400

0.0560
g
0.0558

0.0397 .ne
7-8

8-9
0.0098

0.0190
0.0224

0.0431
0.0091

0.0174
t
9-10 0.0182 0.0413 0.0167

10-11 0.0205 0.0468 0.0190

11- 12 0.0660 0.0150 0.0060

12 -13 0.0455 0.0642 0.0058

13 -14 0.1182 0.2360 0.0213

14-15 0.0214 0.2743 0.0267

15 - 16 0.1336 0.0525 0.0059

16 -17 0.0580 0.3532 0.0367

Contd. ....
Downloaded From : www.EasyEngineering.net
Downloaded From : www.EasyEngineering.net

74 Operation and Control in Power Systems

Line data Resistance Reactance Capacitance

17-18 0.1550 0.1532 0.0168

18 -19 0.1550 0.3639 0.0350

19-20 0.1640 0.3815 0.0371

20-21 0.1136 0.3060 0.0300

20-23 0.0781 0.2000 0.0210

23 -24 0.1033 0.2606 0.0282

ww 12 - 25

25 -26
0.0866

0.0159
0.2847

0.0508
0.0283

0.0060

w.E
26-27 0.0872 0.2870 0.0296

27-28

28- 29 asy 0.0136

0.0136
0.0436

0.0436
0.0045

0.0045

29-30

30- 31
E 0.0125

0.0136 ngi
0.0400

0.0436
0.0041

0.0045

27 - 31 0.0136
nee
0.0436 0.0045

rin
30- 32 0.0533 0.1636 0.0712

32 - 33 0.0311 0.1000 0.0420

32 - 34 0.0471 0.1511
g.n
0.0650

30- 51

51-33
0.0667

0.0230
0.1765

0.0622
0.0734

0.0256 et
35 -50 0.0240 0.1326 0.0954

35 - 36 0.0266 0.1418 0.1146

39-49 0.0168 0.0899 0.0726

36- 38 0.0252 0.1336 0.1078

38-1 0.0200 0.1107 0.0794

38- 47 0.0202 0.1076 0.0869

47-43 0.0250 0.1336 0.1078

42-43 0.0298 0.1584 0.1281

40-41 0.0254 0.1400 0.1008

Contd. ••.•

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 75

Line data Resistance Reactance Capacitance

41-43 0.0326 0.1807 0.1297

43 -45 0.0236 0.1252 0.1011

43 -44 0.0129 0.0715 0.0513

45 -46 0.0054 0.0292 0.0236

44-1 0.0330 0.1818 0.1306

ww 46-1

1-49
0.0343

0.01 \0
0.2087

0.0597
0.1686

0.1752

w.E
49-50

37 - 38
0.0071

0.0014
0.0400

0.0077
0.0272

0.0246

47-39
asy 0.0203 0.\093 0.0879

48-2
En 0.0426 0.1100 0.0460

gi
3 -35 0.0000 0.0500 0.0000

7-36

11-37
0.0000

0.0000
0.0450

0.0500 nee 0.0000

0.0000

14-47

16- 39
0.0000

0.0000
0.0900

0.0900
r 0.0000

0.0000ing
18-40

20-42
0.0000

0.0000
0.0400

0.0800
0.0000

0.0000
. net
24-43 0.0000 0.0900 0.0000

27-45 o.ooeo 0.0900 0.0000

26-44 0.0000 0.0500 0.0000

30-46 0.0000 0.0450 0.0000

1-34 0.0000 0.0630 0.0000

21-2 0.0000 0.2500 0.0000

4-5 0.0000 0.2085 0.0000

-
19-41 0.0000 0.0800 0.0000

Downloaded From : www.EasyEngineering.net

76 Operation and Control in Power Systems

ww
w.E
asy
E ngi
nee
rin
Fig. E2.1 51 Bus Power System

Bus P-Q
3 -35
TAP
1.0450 g.n
7-36
11-37
14-47
1.0450
1.0500
1.0600
et
16-39 1.0600
18 -40 1.0900
19-41 1.0750
20-42 1.0600
24-43 1.0750
30-46 1.0750
1-34 1.0875
21-22 1.0600
5-4 1.0800
27-45 1.0600
26-44 1.0750

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 77

Bus Data - Voltage and Scheduled Powers

Bus no Voltage magnitude Voltage phase angle Real power Reactive power
(p.u.) (p.u.) (p.u.)

I 1.0800 0.0000 0.0000 0.0000

2 1.0000 0.0000 - 0.5000 - 0.2000
3 1.0000 0.0000 -0.9000 -0.5000
4 1.0000 0.0000 0.0000 0.0000
5 1.0000 0.0000 -0.1190 0.0000

ww
6
7
1.0000
1.0000
0.0000
0.0000
-0.1900
-0.3300
-0.1000
- 0.1800
8
9
w.E1.0000
1.0000
0.0000
0.0000
-0.4400
-0.2200
-0.2400
-0.1200
10
asy
1.0000 0.0000 -0.2100 -0.1200
II
12
13
E
1.0000
1.0000
1.0000
0.0000

ngi
0.0000
0.0000
- 0.3400
- 0.2400
- 0.1900
-0.0500
- 0.1360
-0.1100
14 1.0000 0.0000
nee -0.1900 -0.0400
15
16
1.0000
1.0000
0.0000
0.0000
0.2400
-0.5400 rin 0.0000
-0.3000

17
18
1.0000
1.0000
0.0000
0.0000
-0.4600
-0.3700
g.
-0.2100

net
- 0.2200
19 1.0000 0.0000 -0.3100 -0.0200
20 1.0000 0.0000 - 0.3400 -0.1600
21 1.0000 0.0000 0.0000 0.0000
22 1.0000 0.0000 - 0.1700 -0.0800

23 1.0000 0.0000 -0.4200 - 0.2300

24 1.0000 0.0000 - 0.0800 - 0.0200

25 1.0000 0.0000 -0.1100 - 0.0600

26 1.0000 0.0000 - 0.2800 -0.1400

27 1.0000 0.0000 - 0.7600 -0.2500
28 1.0000 0.0000 - 0.8000 -0.3600

Contd. •.•.

Downloaded From : www.EasyEngineering.net

78 Operation and Control in Power Systems

Bus no Voltage magnitude Voltage phase angle Real power Reactive power
(p.u.) (p.u.) (p.u.)
29 1.0000 0.0000 - 0.2500 -0.1300

30 1.0000 0.0000 - 0.4700 0.0000

31 1.0000 0.0000 - 0.4200 -Q.I800

ww 32
33
1.0000
1.0000
0.0000
0.0000
- 0.3000

0.5000
- 0.1700
0.0000

34
35
w.E 1.0000
1.0000
0.0000
0.0000
- 0.5800
0.0000
-0.2600

0.0000
.-
36
asy1.0000 0.0000 0.0000 0.0000

37
En
1.0000 0.0000 0.0000 0.0000

gi
38 1.0000 0.0000 1.7000 0.0000

39
40
1.0000
1.0000
0.0000
0.0000 nee 0.0000
0.0000
0.0000
0.0000

42
1.0000

1.0000
0.0000
0.0000 r ing
0.0000

0.0000
0.0000

0.0000

44
1.0000

1.0000
0.0000

0.0000
0.0000

1.7500
. net
0.0000

0.0000

45 1.0000 0.0000 0.0000 0.0000

46 1.0000 0.0000 0.0000 0.0000

47 1.0000 0.0000 0.0000 0.0000

48 1.0000 0.0000 0.5500 0.0000

49 1.0000 0.0000 3.5000 0.0000

50 1.0000 0.0000 1.2000 0.0000

51 1.0000 0.0000 -0.5000 - 0.3000

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 79

Bus No. Voltage at VeB Reactive power limit

15 1.0300 0.1800
30 1.0000 0.0400
33 1.0000 0.4800
38 1.0600 0.9000
44 1.0500 0.4500
48 1.0600 0.2000
49 1.0700 0.5600

ww 50 1.0700 1.500

w.E
P2.4 The data for a 13 machine, 71 bus, 94 line system is given. Obtain the load flow
solution.
Data:
asy
En No. of buses
No. of lines
71
94

giBase power (MVA)

No. of machines
No. of shunt loads
200
13
23 nee
rin
BUS NO GENERATION
. .
LOAD POWER
g.n
I
2
3
0.0
506.0
0.0
150.0
0.0
0.0
0.0
0.0
0.0
0.0
et
4 0.0 0.0 0.0 0.0
5 0.0 0.0 0.0 0.0
6 100.0 32.0 0.0 0.0
7 0.0 0.0 12.8 8.3
8 300.0 125.0 0.0 0.0
9 0.0 0.0 185.0 130.0
10 0.0 0.0 80.0 50.0
,
II 0.0 0.0 155.0 96.0
12 0.0 0.0 0.0 0.0

Contd. ....

Downloaded From : www.EasyEngineering.net

80 Operation and Control in Power Systems

BUS NO GENERATION LOAD POWER

13 0.0 0.0 100.0 62.0
14 0.0 0.0 0.0 0.0
15 180.0 110.0 0.0 0.0
16 0.0 0.0 73.0 45.5
17 0.0 0.0 36.0 22.4
18 0.0 0.0 16.0 9.0
19 0.0 0.0 32.0 19.8

ww 20 0.0 0.0 27.0 16.8

w.E
21 0.0 0.0 32.0 19.8
22 0.0 0.0 0.0 0.0
23 0.0 0.0 75.0 46.6

a
24
25 syE0.0
0.0
0.0
0.0
0.0
133.0
0.0
82.5
26
27 ngi
0.0
300.0
0.0
75.0
• 0.0
0.0
0.0
0.0
28 0.0 0.0
nee30.0 20.0
29
30
260.0
0.0
70.0
0.0
0.0
120.0 rin 0.0
0.0
31 0.0 0.0 160.0
g.n
74.5
32
33
34
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
112.0
99.4
0.0
69.5
et
35 0.0 0.0 0.0 0.0
36 0.0 0.0 50.0 32.0
37 0.0 0.0 147.0 92.0
38 0.0 0.0 93.5 88.0
39 25.0 30.0 0.0 0.0
40 0.0 0.0 0.0 0.0
41 0.0 0.0 225.0 123.0
42 0.0 0.0 0.0 0.0
43 0.0 0.0 0.0 0.0

Contd. ....

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 81

BUS NO GENERATION LOAD POWER

44 180.0 55.0 0.0 0.0
45 0.0 0.0 0.0 0.0
46 0.0 0.0 78.0 38.6
47 0.0 0.0 234.0 145.0
48 340.0 250.0 0.0 0.0
49 0.0 0.0 295.0 183.0

ww 50
51
0.0
0.0
0.0
0.0
40.0
227.0
24.6
142.0

w.E 52
53
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

asy
54 0.0 0.0 \08.0 68.0
55
56 En 0.0
0.0
0.0
0.0
25.5
0.0
48.0
0.0
57
58
gi 0.0
0.0
0.0
0.0
55.6

nee
42.0
35.6
27.0
59 0.0 0.0 57.0 27.4
60 0.0 0.0
rin
0.0 0.0
61
62
0.0
0.0
0.0
0.0
0.0
40.0
0.0
27.0 g.n
63
64
0.0
300.0
0.0
75.0
33.2
0.0
20.6
0.0
et
65 0.0 0.0 0.0 0.0
66 96.0 25.0 0.0 0.0
.'
67 0.0 0.0 14.0 6.5
68 90.0 25.0 0.0 0.0
69 0.0 0.0 0.0 0.0
70 0.0 0.0 11.4 7.0
71 0.0 0.0 0.0 0.0

Downloaded From : www.EasyEngineering.net

82 Operation and Control in Power Systems

LINE DATA

Line No From Bus To Bus Line impedance 112 Y charge Turns Ratio

1 9 8 0.0000 0.0570 0.0000 1.05

2 9 7 0.3200 0.0780 0.0090 1.00
3 9 5 0.0660 0.1600 0.0047 1.00
4 9 10 0.0520 0.1270 0.0140 1.00
5 10 II 0.0660 0.1610 0.0180 1.00
6 7 10 0.2700 0.0700 0.0070 1.00

ww 7
8
12
11
II
13
0.0000
0.0600
0.0530
0.1480
0.0000
0.0300
0.95
1.00
9
10
w.E14
13
13
16
0.0000
0.9700
0.0800
0.2380
0.0000
0.0270
1.00
1.00
11
asy
17 15 0.0000 0.0920 0.0000 1.05

En
12 7 6 0.0000 0.2220 0.0000 1.05
13 7 4 0.0000 0.0800 0.0000 1.00
14
15
16
4
4
4
3
5
12
gi
0.0000
0.0000
0.0160
0.0330

nee
0.1600
0.0790
0.0000
0.0000
0.0710
1.05
1.00
1.00
17
18
12
17
14
16
0.0160
0.0000
0.0790
0.0800
r 0.0710
0.0000 ing 1.00
0.95
19
20
2
4
4
26
0.0000
0.0190
0.0620
0.0950
0.0000
0.1930
. 1.00
0.00 net
21 2 1 0.0000 0.0340 0.0000 1.05
22 31 26 0.0340 0.1670 0.1500 1.00
23 26 25 0.0000 0.0800 0.0000 0.95
24 25 23 0.2400 0.5200 0.1300 1.00
25 22 23 0.0000 0.0800 0.0000 0.95
26 24 22 f).0000 0.0840 O.OOOG 0.95
27 22 17 0.0480 0.2500 0.0505 1.00
28 2 24 0.0\00 0.1020 0.3353 1.00
29 23 21 0.0366 0.1412 0.0140 1.00
30 21 20 0.7200 0.1860 0.0050 1.00
31 20 19 0.1460 0.3740 0.0\00 1.00

Contd. .••.

Downloaded From : www.EasyEngineering.net

Load Flow Analysis 83

Line No From Bus To Bus Line impedance 1/2 Y charge Turns Ratio
32 19 18 0.0590 0.1500 0.0040 1.00
33 18 16 0.0300 0.0755 0.0080 1.00
34 28 27 0.0000 0.0810 0.0000 1.05
35 30 29 0.0000 0.06\0 0.0000 1.05
36 32 31 0.0000 0.0930 0.0000 0.95
37 31 30 0.0000 0.0800 0.0000 0.95
38 28 32 0.0051 0.0510 0.6706 1.00

ww 39
40
3
31
33
47
0.0130
0.0110
0.0640
0.0790
0.0580
0.1770
1.00
100
41
42
w.E 2
33
32
34
0.0158
0.0000
01570
0.0800
0.5100
0.0000
100
0.95
43
asy
35 33 0.0000 0.0840 0.0000 0.95

En
44 35 24 0.0062 0.0612 0.2120 1.00
45 34 36 0.0790 0.2010 0.0220 1.00
46
47
48
36
37
40
gi 37
38
0.1690
0.0840
0.4310

nee
0.1880
0.3800
0.0110
00210
0.0000
1.00
1.00
1.05
39 0.0000
49
50
40
38
38
41
0.0890
0.1090
r
0.2170
0.1960 ing
0.0250
0.2200
1.00
1.00
51
52
41
42
51
41
0.2350
0.0000
0.6000
0.0530
0.0160
0.0000
. net
1.00
0.95
53 45 42 0.0000 0.0840 0.0000 0.95
54 47 49 0.2100 0.1030 0.9200 1.00
55 49 48 0.0000 0.0460 0.0000 1.05
56 49 50 0.0170 0.0840 0.0760 1.00
57 49 42 0.0370 0.1950 0.0390 1.00
58 50 51 0.0000 00530 0.0000 0.95
59 52 50 0.0000 0.0840 0.0000 0.95
60 50 55 0.0290 0.1520· 0.0300 1.00
61 50 53 0.0100 0.0520 0.0390 1.00
62 53 54 0.0000 0.0800 0.0000 0.95
63 57 54 0.0220 0.0540 0.0060 1.00

Contd•....

Downloaded From : www.EasyEngineering.net

84 Operation and Control in Power Systems

Line No From Bus To Bus Line impedance 1/2 Y charge Turns Ratio
64 55 56 0.0160 0.0850 0.0170 1.00
65 56 57 0.0000 0.0800 0.0000 1.00
66 57 59 0.0280 o.ono 0.0070 1.00
67 59 58 0.0480 0.1240 0.0120 1.00
68 60 59 0.0000 0.0800 0.0000 1.00
69 53 60 O.OJ60 0.1840 0.3700 1.00

ww70
71
45
45
44
46
0.0000
0.0370
0.1200
0.0900
0.0000
0.0100
1.05
1.00
n
73 w.E46
46
41
59
0.0830
0.10'"70
0.1540
0.1970
0.0170
0.0210
1.00
1.00
74
75 asy
60
61
61
62
0.0160
0.0000
0.0830
0.0800
0.0160
0.0000
1.00
0.95
76
77
58
62 E 62
63
0.0420

ngi
00350
0.1080
0.0890
0.0020
0.0090
1.00
1.00
78
79
69
69
68
61
0.0000
0.0230 nee 0.2220
0.1160
0.0000
0.1040
1.05
1.00
80 67 66 0.0000
rin
0.1880 0.0000 1.05

g.n
81 65 64 0.0000 0.0630 0.0000 1.05
82 65 5'6 0.0280 0.1440 0.0290 1.00
83
84
85
65
65
67
61
67
63
0.0230
0.0240
0.0390
0.1140
0.0600
0.0990
0.0240
0.0950
0.0100
et
1.00
1.00
1.00
86 61 42 0.0230 0.2293 0.0695 1.00
87 57 67 0.0550 0.2910 0.0070 1.00
88 45 70 0.1840 0.4680 0.0120 1.00
89 70 38 0.1650 0.4220 0.0110 1.00
90 33 71 0.0570 0.2960 0.0590 1.00
91 71 37 0.00(,)0 0.0800 0.0000 0.95
92 45 41 0.1530 0.3880 0.1000 1.00
93 35 43 0.0131 0.1306 0.4293 1.00
94 52 52 0.0164 0.1632 0.5360 1.00

Downloaded From : www.EasyEngineering.net

Load Flow ,Analysis 85

Shunt Load Data

S.No Bus No Shunt Load

I 2 0.00 - 0.4275
2 13 0.00 0.1500
3 20 0.00 0.0800
4 24 0.00 - 0.2700

ww 5
6
28
31
0.00
0.00
- 0.3375
0.2000

w.E 7
8
32
34
0.00
0.00
-0.8700
0.2250

asy 9
10
35
36
0.00
0.00
- 0.3220
0.1000

EnII 37 0.00 0.3500

gi
12
13
14
38
41
43
0.00
0.00
0.00 nee
0.2000
0.2000
-0.2170 .
15
16
46
47
0.00
0.00 r0.1000
0.3000 ing
17
18
19
50
51
52
0.00
0.00
0.00
0.1000
0.1750
-0.2700
. net
20 54 0.00 0.1500
21 57 0.00 0.1000
22 59 0.00 0..0750
23 21 0.00 0.0500

Downloaded From : www.EasyEngineering.net

ww
w.E
3 ECONOMIC
asy OPERATION OF
En
POWER SYSTEMS
gin
eer
ing
Planning operation and control of interconnected power systems presents a variety of challenging
problems. An important problem in this area is the economic operation of the system, which

.
means, that every step in planning, scheduling and operation of the system, unit-wise, plant

net
wise and inter connection-wise must be optimal, leading to absolute economy. In this, the
transmission losses too play an important role. In this chapter on econOl)1ic operation of
power system, both thermal and hydro system will be dealt with using suitable analytical
models that result in meaningful savings.

3.1 Characteristics of Steam Plants

In analyzing the economics ot operation of thermal systems, modelling of input output
characteristics assumes significance. For this purpose a single unit comprising of boiler,
turbine, and generator may be considered. The unit has to supply power to the local needs for
the auxiliaries in the station. This later component may be around 2-5%. The station auxiliaries
includes boiler feed pumps, condenser circulating water pumps, fans etc. The total input to
the unit could be either K-cal/hr in terms of heat supplied or Rs/hr in terms of the cost of the
fuel such as coal, gas or fossil fuel of any other form. The net output of the unit that is
supplied into the system at the generator bus will be in KW or MW. Scheduling is the process
of allocation of generation among various operating generator units. Economic scheduling is

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 87

the cost effective mode of generation allocation among the various units. This can also be
termed as optimal operation. The analytical solution proposed in general for optimal operation
depends on incremental cost concept.

Boiler I------} Electrical

output

ww Turbine

Fig. 3.1 (a) Boiler turbine generator unit

3.2 w.E
Input Output Cunres

asy
As has been already stated the input output characteristics for arty thermal unit or units that
comprise the plant can be obtained from the operating data. The input can be in kilo calories

En
per hour and the output may be in kilowatts or preferably in maga watts. Typical characteristic

gi
is shown in Fig. 3.I(b) for the unit shown in Fig. 3.I(a).

nee
1...
.eOJ
r ing
~
~
'r

'5
0-
. net
.:
-.:;
----
~ 'Min
Power output (P) MW •

Fig. 3.1 (b) Thermal unit input-output characteristic

The characteristic in practice may not be such a smooth idealized curve and from the
practical data such an idealized curve can be interpolated.
Steam turbine generating unit characteristics may have minimum and maximum limits
in operation. \
They may be determined by factors such as steam cycle used, operating tempera~ures,
material thermal characteristics.

Downloaded From : www.EasyEngineering.net

88 Operation and Control in Power Systems

3.3 The Incremental Heat Rate Characteristic

From the input output characteristic the incremental heat rate characteristic can be
obtained which is the ratio of the differentials.
d(in put)
Incremental fuel rate or heat rate = d( ) ..... (3.1 )
output
By calculating the slope of the characteristic in Fig. 3.1 (b) at every point the incremental
fuel rate characteristic can be plotted as shown in Fig. 3.2. This characteristic infact tells
about the thermal efficacy of the unit under consideration that can be used for comparison
with other units in performance.

ww
w .Ea ,,
I

syE, ,,
I

,,
I

,n
gi 'Min

nee
Output (MW) P --+
I
,Max

r ing
Fig. 3.2 Incremental fuel rate characteristic for thermal unit

3.4 The Incremental Fuel Cost Characteristic

. net
The incremental fuel rate in K-caI/K-whr can be multiplied by cost-of the fuel in terms of
Rs per K-cal. In any case the ordinates are in Rs/Kw-hr. or Rs/MW-hr. The calorific value of
the fuel is required in these calculations. This characteristic is shown in Fig. 3.3 ..

,,,
I

,,
,,
,, ,,
'Min , Max
Power output (MW) --+
Fig. 3.3 Incremental fuel cost characteristic for thermal unit

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 89

3.5 -Heat Rate Characteristic

Some times the unit net heat rate characteristic is also considered important. To obtain this
characteristic the net heat rate in k.caI/K-whr is plotted against the power output (Fig. 3.4).
The thermal efficiency of the unit is influenced by factors like steam condition, reheat
stages, condenser pressure and the steam cycle used. The efficiency of the units in practice is
around 30%.

ww I \
\

w.E
asy
En
Min
gin-.
Power output (MW)
Rated Max

eer
Fig. 3.4 Heat rate characteristic

3.6 Incremental Production Cost Characteristics

in g.n
The production cost of the power generated actually depends on several items such as fuel
cost, labour charges, cost of i,tems such as oil, water and other supplies needed and also the
cost of maintenance. It is well known that in thermal generation the fuel cost is by far the
et
largest cost head and is directly related to the power generated. Even the other charges, that is
the additional running expenses too are, more or less, related to the amount of generation.
Thus, it is a simple practical proposition to assume that, all the additional costs as a fixed
percentage of the incremental fuel cost.
The sum of incremental fuel cost and other incremental running expenses is called
incremental production cost. The ordinates of Fig. 3.3 will also represent incremental
production cost to some other scale. An explicit mathematical relationship involving all the
factors involved in power generation to the total power generated is infact a very difficult task.
In general, the input-output data is fitted into a quadratic characteristic even though it is
also possible to fit into any polynomial curve for ea~e of mathematical manipulation.

Downloaded From : www.EasyEngineering.net

90 Operation and Control in Power Systems

Large turbine-generator units may have several steam admission valves which are opened
in a sequence to meet the increasing steam demand. The input output characteristic for such
a unit with two valves may show discontinuity as shown in Fig. 3.5.

ww
w.E I
I

asy
I
I Min Max
Power output (MW) --.

En
Fig. 3.5 Input-output characteristic for multiple admission valves (for two valves)

3.7 gi
Characteristics of Hydro Plants
nee
The input output characteristics for hydro units can be obtained in the same way as for thermal

rin
units on the assumption of constant water head. The input-output characteristic may be as
shown in Fig. 3.6. The ordinates are water input or discharge in cubic meters per second

g.n
shown against power output in megawatts. While the water requirement is nearly linear till
rated load, after that the efficiency decreases and greater discharge is required to meet the
increased load demand.
et
t

Power output (MW)--.

Fig. 3.6 Input-output characteristic for hydro unit

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 91

It may be noted that, if the head varies the input-output characteristics change. 11 will
move vertically upwards, as head falls and vice versa since the hydro power generatbd is
directly related to the head of the water level and as head falls, higher water discharge is
required for the same power generation. Similarly as head rises lesser discharge is needed.
The characteristic moves downwards. This is shown in Fig. 3.7.

ww
w.E
asy
En Power output (MW) ~

3.8 gi
Fig. 3.7 Effect of head on discharge for hydro unit

Incremental Water Rate Characteristics

nee
The incremental water rate curve is obtained from the water input-output characteristic in the

rin
same way as for thermal units. A typical characteristic is shown in Fig. 3.8.

g.n
As the input-o.utput curve is linear for a greater part, the incremental water rate
characteristic is a horizontal line over this region indicating constant slope, and thereafter it
rises rapidly. With increase in load, more and more units will have to be brought into service .
et

.---------t~
I :
I I
Power output (MW)

Fig. 3.8 Incremental water-rate characteristic

Downloaded From : www.EasyEngineering.net

92 Operation and Control in Power Systems

There will be discontinuity in the characteristics in such a case, as showriinFig. 3.9.

The discontinuity generally can be neglected so that the characteristic will still have the same
shape as in Fig. 3.8.

ww
w.E Power output (MW) ~

asy
Fig. 3.9 Multiple unit operation incremental water-rate characteristic

En
The conversion of incremental water rate into incremental production cost requires
considerations of agriculture, navigation, drinking needs of water, and other riparian rights etc.

gi
even through water is available freely in nature. Further the cost of dam~ canals, conduits,

nee
gates, penstocks and other parts of hydro development are also involved.

3.9 Incremental Production Cost Characteristic

rin
The incremental water rate characteristic can be converted into incremental production cost

g.n
characteristic by multiplying the incremental water rate characteristic by water rate or cost of
water in rupees per cubic meter C\~. The incremental production cost characteristic is shown
in Fig. 3.10.
et
-
t il

.~t..ai CI
:::l I

"8:::
~E
til til
cCX:
Q)

E
...
Q)

u
.5
0 PHI P H2
Power output (MW)

Fig. 3.10 Incremental production cost representation

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 93

The incremental production cost characteristc can be expressed analytically as follows .

IPC = C I (0 ::: P ::: PHI) ..... (3.2)
= a PH + C I (PHI::: P ::: PH2 ) ..... (3.3)
where a is the slope ofthe· characteristic between PHI and PH2 .

3.10 Generating Costs at Thermal Plants

Consider any ith plant among n thermal plants supplying active power. CI is the cost per unit
of PI in the neighborhood of PI then the generating cost is
..... (3.4)

wwThe total cost of operating a system with N g generating sets can be represented by
Ng

w.E F = IC,p,
,=1
..... (3.5)

If the system operates over a time period T then the total expenditure involved will be

aS!sy TN"
c, P, (ndt
En
FT =
0,=1
..... (3.6)

gi
Steam plants with partial admission nozzle governing give better performance at partial

nee
loads since the cost coefficient increases with increasing megawatt loading. However, such
units cannot be shut down, frequently, because of the complexities of steam chest. But, steam
plants with throttle governing are more suitable for periodic shut down due to their simpler

r ing
steam chest. Such units are more suitable for rapid starting and loading. From minimum to
maximum permissible limits of operation, the cost coefficients of the units are substantially
constant.

3.11 Analytical Form for Input-Output Characteristics of Thermal Units

. net
It is already pointed out that the input-output characteristic may be fitted into a polynomial of
the form
Cost F = A + BP + Cp2 -t ......... . . .... (3.7)
where A, B, C ...... are constants
If a quadratic representation is made then denoting the cost
F = Y2 a p2 + bP + C ..... (3.8)
will lead to an incremental cost characteristic of the form
dF
- =aP+b ..... (3.9)
dP
a linear relationship around any operating point P.

Downloaded From : www.EasyEngineering.net

94 Operation and Control in Power Systems

3.12 Constraints in Operation

The power system has to satisfy several constraints while in operation. These may be broadly
divided into two types. The first of these arises out of the necessity for the system to satisfy
load balance and are called equality constraints. At any bus i. Ps, and Os, correspond to the
scheduled generation. PD, and O[), are the load demands then the following equations must be
satisfied at the bus i :
Ps, - PD , - P, = G, = 0
Os, - O[), - 0, = H, = 0 ..... (3.10)
P, and Q, are given by

ww 11

P, = L:V,V1Y"Cos(O, -Ol-e ll )
1=1
..... (3.11 )

w.E ..... (3.12)

asy
with usual notation and G, and H, are the residuals at bus i which should become zero at the
point of solution.
E ngi
In addition, a number of other constrains due to physical and operational limitations of

nee
the units and components will arise in economic scheduling. These are in the form of inequality
constraints.

rin
Each generator in operation will have a minimum and maximum permissible output and
the production must be constrained to ensure that
p,mm:s P, :s p,ma" i = I, 2 .... , ng
g.n ..... (3.13)
Similarly limits may also have to be considered over the range of reactive power
capabilities of the generator units requiring that
Q,mm :s Q, :s Q,ma" i = 1,2 .... , nq
et ..... (3.14)
where nq is the total number of reactive sources in the system.
Further, the constraint
P,:! + O,1:s (S,fated)2 ..... (3.15)
must be satisfied, where S is the MVA capacity of the generating unit for limiting stator heating.
Dynamic limits may also have to be considered when fast changes in generation are
envisaged for picking up or for shedding down of loads. These limits put additional constraints
of the form.

.. ... (3.16)

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 95

The maximum and minimum operating conditions for a group of generators within a
power station may be different from the respective sum of the maximum and minimum operating
levels of turbines that are supplied by a single boiler. The extremes ofboiler operating conditions
will determine these limits. Thus. groups of generators from individllal boiler units may have
to be subjected to additional constraints of the nature

Pkgill'" <
-
Pkl <
-
plIl,l\ k ~ I'"
kg'
(' R
- , - ..... J ..... (3.17)
1f.(1

where GR is the total number of generator groups. the outputs of which are to be
separately limited.

ww Spare capacity is required to account for the errors in load prediction, sudden and fast
changes in load demand and the inadvertent loss of scheduled generation. Thus. the total

w.E
generation G available at any time must be in excess of the total anticipated load demand and
system losses by an amount not less than a specified minimum spare capacity Psp'

asy G~
ng
L P, + PSP ..... (3.18)

En 1=1

In a similar manner, constrains may be required to be associated with groups of

gi
generators, where all plants are not equally operationally suitable for taking up additional load.
If TG is the total number of groups then
nee
LP
Gk ~
If.Ci
kl + p~G
rin ..... (3.19)

where k = I, 2 ..... , TG.

The summation is over the set G over which group constraints are applied. g.n
Thermal considerations may require that the transmission lines be subjected to branch
transfer constraints of the type.
et
- S:na, ::; Sbl ::; S~,"x • i = 1,2, ...... , nb .. ... (3.20)
where nb is the number of branches and S bl is the branch transfer MVA.
In addition, constraints are to be imposed for bus voltage magnitudes and for phase
displacements between them for maintaining voltage profile and for limiting overloading
respectively.
Thus we have
V'Imill <
-
V'I -< V'Imax ,1-,
. - I 2
...... ,n ..... (3.21)

s: mill
011
< I: < s: max ,1--
-0'l-0'l
. - I 2
....... ,n ..... (3.22)

where j = I. 2 ...... m and .i:t: i

Downloaded From : www.EasyEngineering.net

96 Operation and Control in Power Systems

where n is the total number of nodes and m is the number of nodes neighbouring each node
with interconnecting branches.
In case transformer tap positions are to be included for optimization, then the tap
positions T) must lie within the range available, i.e.

T)1Il1l1 ~ T) ~ Tinax . .... (3 .23)

Sometimes, phase shifting transformers are made available in the system. If such
equipment exists then constraints of the type

ww ..... (3.24 )

w.E
must be reckoned where PSI is the phase shift obtained from the ith phase shifting
transformer.

a syE
If power system security is also required to be considered in the formulation for economic
operation then power flows between certain important buses may also have to be considered
for the final solution. It may be mentioned that consideration of each and every possible

ngi
branch for outage will not be a feasible proposition.

3.13 Plant Scheduling Methods

nee
rin
At the plant level, several operating procedures were adopted in the past leading to efficient
operation resulting in economy

(i) Base loading to capacity

g.n
et
The turbo generators are successively loaded to their rated capacities in the order
of their efficiencies. That is to say, that the most efficient unit will get greater
share in load allocation which is a natural solution to the problem.

(ii) Base loading to most efficient load

In this case the heat rate characteristics are considered and the turho-generator
units are successively loaded to their most efficient loads in increasing order of
their heat rates.
In both the above methods thermodynamic considerations assumed importance
and the schedules will not differ from each other much.

(iii) Proportional loading to capacity

A third method that was considered as a thumb rule in the absence of any technical
data is to load the generating units in proportion to their rated capacities as stated
on the name plates.

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems .97

3.14 Merit Order Method

If the incremental cost characteristics are fairly constant over a wide range in operation then
neglecting the transmission losses and running reserve requirements, it is possible to prepare
schedules for load allocation using incremental efficiencies. Merit tables based upon incremental
efficiencies are prepared and each unit is loaded to its rated capacity in order of the highest
incremental efficiency. Changes in fuel costs, plant cycle efficiencies, plant availabilities etc
require the merit tables to be revised regularly to reflect these factors. Then, it is possible to
look at the tables so prepared and schedule the generation to different units.

3.15 Equal Incremental Cost Method: Transmission Losses Neglected

ww
Method of Lagrange multipliers:
It is but natural that for a given load to be allocated between several generating units, the most

w.E
efficient unit identified by incremental cost of production should be the one to get priority.
When this is applied repeatedly to all the units the load allocation will become complete when
all of them, that are involved in operation, are all working at the same incremental cost of
production.
asy
The above can be proved mathematically as follows:

total load demand Po' E

Consider ng generating units supplying PI' Pc' ............ Png active powers to supply a

ngi
hour.
nee
The objective function for minimization is the total input to the system in rupees per

F(P I,P2 ,······,Png )=F(P)= IC,(p,)

rin ..... (3.25)
,=1

g. net
where C,(P) is the generation cost for the ith unit and ng is the total number of generating
units.
The equality constraint is given by
ng
G (PI' P"
_ ... , Png ) = G(P)" Po - LP, =-0 ..... (3.26)
"I
i.e., total supply = total demand neglecting losses and reserve. Using the method of Lagrange
multipliers for equality constraints. the Lagrange function is defined as
L(P, A) == F(P) + AG(P) ..... (3.27)
Where A is a Langrange multiplier.
The necessary conditions are given by
aL
-=0, and ..... (3.28)
OP,

..... (3.29)

Downloaded From : www.EasyEngineering.net

98 Operation and Control in Power Systems

ng (ng )
since L= ~ C, (P, ) + Al Po - ~ P, ..... (3 .30)

elL ?
we obtain - = -- C (P ) + A( -I) = 0 ..... (3.31)
(lP, oP, ' ,

(0L ng
and - ' = PD - LP =0 ..... (3.32)
CA ,=1 '

The later equation is the load demand constraint only; while the former gives

ww ac,(p,) = (lC2(P Z ) = ......... = acng(Png ) =A ..... (3.33)

w.E ap, oP1 cPng

The above equation states that at the optimum all the generating stations operate at the

asy
same incremental cost for optimum economy and their incremental production cost is equal to
the Lagrange multiplier A at the optimum.

En
Eqn. (3.32) gives

gi(PI + Po- +, ........ + PnoJ = Po

nee
..... (3.34)
The principle underlying the mathematical treatment is that load should be taken up
always at the lowest incremental cost. It must be ensured that the generations so determined

r ing
are within their capacities. Otherwise, the generation has to be kept constant at the capacity
limit for that unit and eliminated from further optimum calculations.

. net
It can be seen that in this method at the optimum the incremental cost of production is
also the incremental cost of the power received. Eqn. (;;.33) and (3.34) can be solved for
economic scheduling analytically neglecting the effect of transmission losses.
The size of the power systems increased enormously, with long transmission lines
connecting several power generating stations extending over large geographical areas transferring
power to several load centers. With this development, it has become necessary to consider not
only the incremental fuel costs but also incremental transmission losses incurred in these lines
while power is transmitted. Initial attempts in this direction involved development of a
comprehensive formula involving the generating powers and line parameters. The most important
work has come from Kirchmayer and others in 1951. Their method is based on a set of
coefficients called B. coefficients. Determination of these coefficients is based on several
assumptions and is mathematically quite involved requiring transformations.
However, due to the elegance of the B-coefficient formula and simplicity in application
these coefficients are widely used by a number of power companies in the past in USA and
elsewhere for economic scheduling for including the effect of transmission losses.

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 99

3.16 Transmission Loss Formula - B. Coefficients

An expression for the transmission loss was derived by Kirchamayer with several assumptions
made. In his derivation he used Kron's tensorial transformation. However a simplified procedure
for the derivation of the B-coefficients will be presented here.
Consider a power system supplying nlloads. Let the load currents be ill' iL2 , .•....• iLnl .
These loads are supplied by ng generators. Let the generator currents be igl • ig2 ,····· .. igng .
This is shown in the Fig. 3.11.

ww Power Network

w.E Network Element

Loads

asy
En
Generators

gi
Fig. 3.11 Power system with generator and load currents

nee
Consider a network element K (an inter connected line in the system) carrying current
IK. Let generator I alone supply the entire load current IL where
rin
'L = iLl + iL2 + .......... + iLnl
nl g.n
= fi
1=1
L.!

under this condition let the current in K be iKI

et
..... (3.35)

In a similar manner if each of the ng generators operating alone also supply the total load
current IL while the rest of the generators are disconnected the current carried by the network
element K changes from iKI to iK2 ' iK3 to iKng .
Let the ratio of iK I to iL be dK1

..... (3.36)

Also

and so on.

Downloaded From : www.EasyEngineering.net

100 Operation and Control in Power Systems

Now, if all the generators are connected to the power system simultaneously to supply
the same load, by the principle of super position.
..... (3.37)
Let the individual load currents remain a constant complex ratio of the total load current
Iv It is assumed that (X/R) ratio for all the line elements or branches in the network remains
the same. The factors dK1 will then be real and not complex.
The individual generator currents may have phase angles 8 1, 82, .•.•• 8ng with respect to
a reference axis. The generator currents can be expressed as :

ww igl =lig1lcos8 1+li gl lsin8 1

w.E ..... (3.38)

a syE
igng =li~'11!!lcos8ng +lignglsin81l!!
For simplicity to derive the formula

ngi
Let ng = 3 so that eqn. (3.37) becomes

nee ..... (3.39)

rin
g.n
et

..... (3.41)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 101

..... (3.42)

d
+ 2d u d k i g3!!i gl! cos(8~ - 8 1 )

Eliminating currents in terms of powers supplied by the generators

ww
w.E .
Ig2
P?
= .J3IV21~os~:> and

asy
En
gi
where PI' P2 and P3 are the active power supplied by the generators 1,2 and 3 at voltages IV II,

nee
IV 21 and IV 31 and power factors at the generator buses being COS ~I' COS ~2 and Cos <1>3
respectively.

rin
g.n
31VIIIV21cos<l>I COS~2 3IV21IV~lcOS~2 cos~\
et
+ 2dkldk2PIP2 cos(8 1 - ( 2) + 2d U d Io. J P:>PJ cos(8 2 -8J

The power losses in the network comprising of nb network elements or branches PLOSS
is given by

where RK is the resistance of the element K.

Downloaded From : www.EasyEngineering.net

102 Operation and Control in Power Systems

PI2~d~IRK pr~~d~2R", pr~:d~3RK

k=1 . + k=1 + k=I _ _

IVd2(COS<pI)~ IVi(COS<P2)2 IVl(cOS<P3)2

2PIP2Ldkldk2RK cos(8 1 - ( 2 )
k=1
+

ww +
ng
2P2 P3L dk2 d U R Kcos(8 2
-
_---;'k'-'=I'-::--:-_ _ _ __
8J

w.E IV 2 II V, ICOS <P2 COS <P3

asy ng
2P3 PI Ld k3 d KI R Kcos(8 3 -( 1 )
+
En -----7"k=-'.I-,c--.,.------
Iv31 vd cos <P3 COS<PI
..... (3.43)

Let
gi nee ..... (3.44)

rin
g.n ..... (3.45)

et
..... (3.46)

..... (3.47)

..... (3.48)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 103

..... (3.49)

PLOSS = PI2 BII +P22B 22 + P2B

3 33

+ 2P I P2 B I2 + 2P 2 P,B 23 + 2P,PI B,1 + ..... . ..... (3.50)

, 3
= L LPmBmnPn
111=1 m=1

ww
In general the formula for Bmn coefficients can be expressed

= cos( 8 III - 8 11 )
a~

w.E B
mil I I I(
Y m Vn Cos'!'m X
A. Cos'!'n
"

A.) L-k
d d R
~m kn I<,.
..... (3.51 )

asy
In the matrix form, the loss formula is expressed for an n generator system as :

En .... Pn ]
BII
B21
BI2
B22
Bin
B 2n
PI
P2

gi
[PI P2

Bnl Bn2
nee
Bnn Pn
..... (3.52)

r
The coefficients can be considered constant, if in addition to the assumptions already

ing
made we further assume that the generator voltages V I' V2' .••. etc remain constant in magnitude

.
and generator bus power factors cos <PI' cos <P2 ....... also remains constant.

3.17 Active Power Scheduling

Economic scheduling of thermal plants considering effect of transmission losses:
net
The objective function is
F(P) = F (PI' Pl' ......... Png)
ng
= LC,(P,) ..... (3.53)
,=1
which has to be minimized over a given period of time. As only active power is scheduled the
equality constraint G(P) is given by
ng
G(P) = Po - PL - L P, =0 ..... (3.54)
,=1

Downloaded From : www.EasyEngineering.net

104 Operation and Control in Power Systems

must be satisfied at every generator bus where PI is the generation at bus i ; PD' the total load
demand and PL the total transmission loss in all the lines. It is desired to minimize
eqn. (3.53) subject to the constraint eqn. (3.54)
The Lagrange function L is formed as

..... (3.55)

Applying the necessary conditions for the minimum of L

a ~l =0
o[
ww oL
-(P,A)=-F(pJ+-A
oPI oPI OPI
Po + PL - LP,
1=1
..... (3.56)

i.e., w.E a
-C[PI ....... ,PI,. .. Png]+-A ~ I =0
c [ PD+PL-LP 1 ..... (3.57)

asy oPI oPI 1=1

E ng1;i
..... (3.58)

Further,
oPI nee
OC j (PI) = 1.[1 - oL
oPI
i == 1,2 ....... ng ..... (3.59)

Also, it can be expressed as rin

g.n ..... (3.60)

et
The sum of the incremental production cost of power at any plant i and the incremental
transmission losses incurred due to generation PI at bus i charged at the rate of A must be
constant for all generators and equal to A. This constant A is equal to the incremental cost of
the received power.
In section (3.16) the loss formula is derived as

..... (3.61)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems lOS

Differentiating

oPL = "'2B P ..... (3.62)

oP £... IJ J
I J

Also ..... (3.63)

dCI(PI ) ~
--'-'-'-'- + L 2BIli = A J'= I, 2, ..... , ng ..... (3.64)
dPI 1=1

If the incremental costs are represented by a linear relationship following a quadratic

ww
input-output characteristic.
Then, denoting

w.E dCI(PI ) _ P b
dPI
-a l I + I ..... (3.65)

Eqn. (3.64) will become

asy ng

Further,
E
al PI + b l + AL2BIli = A
J=I

ngi
..... (3.66)

I.2B I / J = 2B 11 PJ + I.2 B I , P, nee ..... (3.67)

J=1 J=1

rin
Eqn. (3.66) can be rewritten as
alPI + b l + A2B11 PI + ALBIJP, = A g.n
alPj +b l + A2B11 PI +ALBlJ =A
ng

1=1
et
)"1

solving for PI

ng
A- bl - AL2B I l,
1=1
)"1 ..... (3.68)
PI = - - - - ' - - - -
a l + 2AB11

b
1- -L - ~' 2B P
~ I) 1
II- 1=1
PI = _ _ _)::-"_1_ _ ; i = 1,2, ..... , ng ..... (3.69)
i +2BI1

Downloaded From : www.EasyEngineering.net

106 Operation and Control in Power Systems

There are ng equations to be solved for ng powers (P) for which Gauss's or
Gauss - Seidel method is well suited. Knowing ai' bl , and BIJ coefficients for any assumed
value of A, PI values may converge to a solution, giving the generator scheduled powers. It is
very important that a suitable value is assumed for A, so that a quick convergence of the
equations is obtained.

3.18 Penalty Factor

Consider the eqn. (3.59)

dC I (PI) + A oP L -= A
dPI oPI

ww It can be rewritten as

A[I- 1
w.E
in other words
OCI(P) -=
oPI
oPL
oPI
..... (3.70)

asy
OC8~~I) [1-E~ln=A ..... (3.71)

gin OPI

eer
When transmission losses are included, the incremental production cost at each plant i

must be multiplied by a factor (1 _~] which then will be equal to the incremental cost of
i ng.
oPI
n et
power delivered. Hence, the factor [
1--
~PL 1 is called penalty factor:

oPI

·
Smce -oPL .
IS muc h Iess th · some times
an umty . .It IS
. approximate
. d by I + -oPL so th at
~ ~

OC I (PI) (1 + oPL ) =A ..... (3.72)

oPI oPI

oP
The term 1 + _L is called approximate penalty factor.
oPI

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 107

3.19 Evaluation of A. for Computation

Consider eqn. (3.69)

b ng
1- --;- - L2BIlJ
fI. J=I
P = _ _ _J,-",_I- - ; i =1, 2, ..... ,ng
I ai
i+2BIIPj

ww
solution to PI values depends upon the value of l. chosen. The value of A determines a set of

w.E
generations for a particular received load.

It is established that for scheduling purposes it is necessary to start the computation

a syE
with two different values of A. As the arbitrarity assigned generations are improved from
iteration to iteration, a new value of A may be computed for each new iteration using the

ngi
following algorithm for a specified total received load power PRo

nee ..... (3.73)

rin
where p~I-I) = received power with A(I-I)
g.n
p t 2) = received power with

A(I), A(I-I), A(I-2)

A(I-2)

are the values of A during iterations (i), (i - 1) and (i - 2)

et
P~ = the desired total power to be received by the loads.

= ~PI -PL
i=1

When two values of PR calculated successively during the iterations converge to a

single value with reasonable accuracy, the latest value of PR calculated will also converge

to P:.

Downloaded From : www.EasyEngineering.net

108 Operation and Control in Power Systems

E 3.1 -Consider a power system with two generating stations. The incremental production
cost characteristics for the two stations are
of
_I = (27.5 + 0.15P I )Rs / MwHr
oPI
of
_2 = (19.5 + 0.26P 2 )Rs / MwHr
OP2
Given that the minimum and maximum powers are 10 MW and 100 MW -at each plant
schedule the generation at each plant to supply a system load given by the load curve
shown in figure.

ww 200 -------------,.---,

w.E 150

1 asyE 125 - - - - - - - - - - -r----l

ngi
100 ----...---~
I

nee
I
I I
I I

50 t----+----r------
L -+---,
I I

rin
o~-----~----~--~----~~--~
C 6
Time (t) ----+ 12g.n
12
Mid night
a.rn 12 3 6 10
noon p.rn p.rn p.rn

Fig. E 3.1 Load curve

Mid
night
et
Solution:
Case (i) : PR = PI + P2; PR = 50 MW; PI == PR - P2 == 50 - P2

oFI = oF2 =A
OPI OP2
27.5 + 0.15 PI == 19.5 + 0.26 P2
27.5 + 0.15 (50 - P2)== 19.5 + 0.26 P2
27.5 - 19.5 + 7.50 == 0.26 P2 + 0.15P 2
15.5 == 0.41P2; P2 == 37.8
PI = 50 - 37.8 = 12.2 MW

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 109

Case (ii) : PI + P2 = 100MW

PI = 100 - P2
27.5 + 0.15 (100 - P2) = 19.5 + 0.26 P2
27.5 + 15 - 19.5 = 0.26 P2 + 0.15 P2
23 = 0.41 P~
P2 = 23/0.41 = 56.01 MW
PI = 100 - 56.1 = 43.90 MW
Case (iii) : PI + P2 = 125MW

ww 27.5 + (0.15) (125 - P2) = 19.5 + 0.26 P2

w.E
27.5 + 18.75 - 19.5 = 0.15 P2 + 0.26 P2 = 0.41 P2
26.75 = 0.4IP 2

asy
therefore and
P2 = 65.24MW

Case (iv) :
E
PI = 59.76MW
P I +P 2 =150MW
ngi
nee
27.5 + 0.15 (150 - PI) = 19.5 + 0.26 P2
27.5 + 22.5 - 19.5 = 0.26 P2 + 0.15 P2
30.5 = 0.41 P2; P2 = 74.39
rin
Case (v) :
PI = 75.61
PI + P2 = 200 MW g. net
27.5 + (0.15) (200 - P2) = 19.5 + 0.26 P2
27.5 + 30 - 19.5 = 0.41 P2
38 = 0.41 P2 ; P2 = 92.68 ; PI = 107.32
The value of A can be computed now. The results are tabulated.

Case PI P2 PD A.
Case (i) 12.2 37.8 50 29.33

Case (ii) 43.9 56.1 \00 34.08

Case (iii) 59.76 65.24 125 36.55

Case (iv) 75.61 74.39 150 38.84

Case (v) 107.32 92.68 200 43.598

Downloaded From : www.EasyEngineering.net

110 Operation and Control in Power Systems

EJ.2 The fuel input characteristics for two thermal plants are given by

F2 = (6P 2 + 0.004 pi + 120) 106 K-cal/hr

Where P I and P2 are in megawatts
(i) plot the input-output characteristic for each plant
(ii) plot the heat rate characteristic for each plant

ww (iii) assuming the cost of fuel as Rs. 100lton

calculate the incremental production cost characteristic in Rs/MWhr at each plant. Plot

w.E
the same against power produced in MW.
Solution:
(i)
asy
Input-output characteristic
Consider the power in steps of 10MW
Given that
En
and gi
F I=(8P,+0.024+80)10 6 K-cal/hr
F2=(6P I+0.004+120)106 K-cal/hr
nee
Substituting PI and P2 in steps of 10 MW the values of F, and F2 are obtained and
tabulated.
rin
Table E 3.2.1 Input Output Calculations
g.n
P1(MW)
10
Fl x 106 K-cal/hr
162.4
P1(MW)
10
Fl x 106K-callhr
184
et
20 250 20 256
30 342 30 336
40 438 40 424
50 540 50 520
60 646 60 624
70 758 70 736
80 874 80 856
90 994 90 984
100 1.220 100 II 20

Downloaded From : www.EasyEngineering.net

Economic Operation of PoWh Systems 111

the input output characteristics are plotted as shown in figure.

1(0)

ww
w.E
asy Fig . E 3.2.1 Input - Out put curves

(ii) En
Heat rate characteristic

gi
Let P I = 10 MW at plant I
· p. 162.4
Heat rate at t h IS I IS - - x 10
10
6 K - cal / hr
MW nee
At plant 2 for P2 = 10 MW
rin
Heat rate = -184 x 106 =.
10
18 4 105
x
K - cal / hr
MW g.n
100 MW at both the plants. The results are tabulated and curves are plotted.
Table E 3.2.2 Heat Rate Calculations
et
In a similar manner the heat rate is computed for powers in steps of 10 MW till

OutputMW Heat rate x 106 at plant 1 K-caVMW-h Heat rate x 106 at plant 2 K-caVMW-hr
10 1625 18.4
~ 12.48 12.8
30 11.39 11.2
40 10.96 10.6
50 10.80 10.4
ro 10.77 10.4 '
'Xl 10.82 10.,5
ro 10.92 10.7
~ II.OS" 10.93
100 112 11.2

Downloaded From : www.EasyEngineering.net

112 Operation and Control in Power Systems

ww
w.E
asy
En Fig. E 3.2.2 Heat rate curves

(iii)
gi nee
Calorific values of fuel at plant t = 4000 K-cal/hr
'--.
Cost of fuel
tOORs I kg rin
= Rs.t 00 per ton = Rs I 001 I 000 per kg
100 Rs
=
1000 kg
x-----:::
4000 k - cal 6
--
4 x t 0 k - cal
g.n
dFI ::: (8 + 0.048P )1 0 6 K - cal
dPI
1
MwHr
et
Incremental fuel cost

6 100 Rs
= (8 + 0.048 PI) 10 x -- 6 - -
4x 10 MwHr
= (2 + 0.012 PI) x 100 = (200 + 1.2 PI) Rs/MWHr
At 100 MW generation, the cost is
(200 + 12) = 212 Rs/MWHr
The incremental production cost is obtained by adding the maintanance costs
of 10%
Hence, the incremental production cost at plant 1
::: (200 + 1.2 PI) 1.I = (220 + 1.32 PI)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 113

At PI = 10MW ;IPC 1 = 220 + 13.2 = 233.2 RslMWHr

Similarly at Plant 2

dF2 = (6 + 0.08P ) x 106 K - cal

2
dP2 Mwhr
o
incremental fuel cost
100
= (6 + 0.08 P2) x 10 6 x - - = (120 + 1.6 P2) RslMWHr
5 x 10 6
Considering maintenance costs at 10%

ww The incremental production cost is

(120 + 1.6 P2) x l.l (132.0 + 1.76 P2) RslMWHr
=

w.E
At P2 = 10MW, the incremental production cost
= (120 + 1.6 = (136 = 149.6 Rs/
MWHr
asy x 10) x 1.1 x 1.1) Rs/MWHr

E
Like wise, the incremental production costs are calculated at all power levels and

ngi
tabulated. Finally the value are plotted as shown in table (E3.2).

nee
Table E 3.2.3 Incremental prodOlction costs

Power(P) fPC at plant 1

Rs/MWHr. rin fPC at plant 2
RsIMWHr.
10 233.2 149.6
g.n
20
30
246.4
259.6
167.2
184.8
et
40 272.8 202.4
50 286.0 220.0
60 299.2 237.2
70 312.4 255.2
80 325.6 272.4
90 338.8 290.4
100 352.0 . 308

Downloaded From : www.EasyEngineering.net

114 Operation and Control in Power Systems

ww
w.E
asy
En
gin
Fig . E 3.2.3 Incremental production cost curves

eer
E 3.3 For the plants in example E3.2 obtain the economic generation schedule. The load
curve is given in Fig. (E3 .3)
i ng.
200
n et
i ISO

100

SO -------

I I 1
12 6 12 6 12
a.m noon p.m
Time (t) - - - .

Fig. E 3.3 Load curve

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 115

Solution:
For economic schedule

dCI(PI) = dC 2 (P2 ) =A.

dPI dP2

(220 + 1.32P I) = (132 + 1.762P2)

Case (i) PI + P2 = 50MW
PI = 50 - P2

ww 220 + 1.32 (50 - P2) = 132 + 1.76 P2

220 + 66 - 132 = 1.76 P2 + 1.32 P2 = 3.082 P2

w.E 154 = 3.082 P2

P2 = 49.967554 ~ 50MW

a syE
:. PI=O
If a minimum generation limit is imposed then P I= IOMW and P2 = 40MW
Case (ii) P = 200MW
ngi
200 + 1.32 (200 - P2) = 1.32 + 1.76P2

nee
220 - 132 + 264 = 176 P2 + 1.32 P2 = 3.08 P2
352 = 3.08 P2
rin
P2 = 114.2857 ~

PI = 200 - 114.3 = 85.7MW

114.3MW

g.n
If a maximum generation limit of 100MW is imposed then
P2 = 100MW and hence PI = 100MW
et
E 3.4 Given a two bus system as shown in Fig. (E 3.4)

2
200MW 75 MW
Line
N~~----------------------~~N

Fig. E 3.4

Downloaded From : www.EasyEngineering.net

Downloaded From : www.EasyEngineering.net
o

116 Operation and Control in Power Systems

It is observed that when a power of 75MW is imported to bus I, the loss amounted to
5MW. Find the generation needed from each plant and also the power received by the
oload, if the system A. is given by Rs.20/MWHr. The incremental fuel cost at the two
plants are given by 0

o
dC I (PI) = 0.03PI + 15Rs I Mwh
odPI

ww
Solution:

Thereforew.E
The load is at bus I. Hence, P, will not have any effect on the line losses
B I , = B'2 =; B 2 , =0

a syE
PL = B22
5 = B22 75 2
pi

B22 ngi
=-~- =8.9 X 10-4
At station I
5~250

nee
rin
g.n
At station 2
et (i)

dC 2 (P 2 ) +A. aPL =A.

dP2 ap2
0.05 P2 + 18 + A. (2(8.9) x 10-4) P2 = A. (ii)
At A. = 20 from eqn. (i)
0.03 P, + 15 = 20
0.03P,=20-15=5
P, = 510.03 = 166.67MW

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 117

Again, from eqn. (ii)

0.05 P 2 + 18 + 20 (2 x 8.9 x 10-4) P2 = 20
0.05 P 2 + 0.0356 P2 = 20 - 18 = 2
0.0856 P2 = 2; P2 = 23.3644MW
Total transmission loss = B22 P 22
= 8.9 x 10-4 x 23.36442 = 0.4858MW
The load demand at bus I is
166.67 + 23.3644 x 0.4858 = 178.Q2MW

ww
E 3.5 A power system with two generating stations supplied a total load of300MW. Neglecting
transmission losses the economic schedule for the plant generation is 175MW and

w.E
125MW. Find the saving in the production cost in Rs/hr. due to this economic schedule
as compared to equal distribution of the same load between the two units.
The incremental cost characteristics are

asy dCI (PI) = 30 + 0.3P and

En
t
dP I

Solution:
gi dC 2(P 2) = 32.5 + 0.4P
dP2
2

nee
The cost of generation at Plant 1

I dCIdP(PI) .dPI = I(O.3PI + 30)dPI

r ing
Ct =

=
I

(0.3 P~ + 30PI + x) Rs/Hr

.net
dC,(P,) I
C2 = I dP - .dP2 = (32.5 + 0.4P 2 )dP 2
2
= (0.4 Pi + 32.5P2 + y) Rs/Hr.
x and yare constants of integration which need not be evaluated.
For equal distribution of generation
P t = 150MW and P2 = 150MW
The increase in cost at plant 1 by generating 175MW instead of 150MW is

(0.3xI752 +30xI75+x)-(0.3xI502 +30 x I50+x)

= 0.3(175 2-150 2)+30(175-150)
= (30625 - 22500)0.3
= 0.3 x 8125 + 750 = 3187.5 Rs/hr

Downloaded From : www.EasyEngineering.net

118 Operation and Control in Power Systems

The reduction in cost at plant 2 by generating 125MW only instead of 150MW is

[0.4 (150)2 + 32.5 x (150 + y)] - [0.4 (125)2 + 32.5 (125) + y]
= 0.4 (- 15625 + 22500) + 32.5 (-125 + 150)
= 0.4 x 6875 + 812.5 = 2750 + 812.5 = 3562.5
Savings = 3562.5 - 3187.5 = 372Rs/hr
Annaul Savings = 375 x 24 x 365 = Rs.32,85.000
E 3.6 Consider two steam power plants operating with incremental production costs

dCI(P I ) =(0.08PI +16)Rs/Mwhr and

ww dPI

w.E dC z(P 2 ) = (0.08P 2 + 12)Rs I Mwhr

dP z
Given the loss coefficients

asy BII = 0.001 per MW

EnBI2 = B21 = - 0.0005 per MW

B22 = 0.0024 per MW

Solution:
gi
Find the economic schedule of generation for", = 20Rs/MWhr

nee
=
16
r
-~:-:------ ing
1- 20 -2(-0.0005)P2 0.2 + 0.001P
2
= ----..::..
0.08 + 2(0.001)
20
. 0.006

net
1 __
2
b -2B I2 PI 1-~~-2(-0.0005)PI 0.4+0.00IP
P2 = _--'' '-0..-___ == == 2

_a~ + 2B22 0~~8 + 2(0.0024) 0.0088

starting with P2 = O. PI = 33.3

then the value of P2 is computed as 49.2
continuing the iterations the values obtained are tabulated
Iteration PI P2
I 33.3 49.2
2 41.5 50.2
-
3 41.7 50.2
4 41.7 50.2

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 119

The converged values are PI = 41.7MW; and P2 = 50.2MW

With these generations, the total transmission losses are
2 2
PL =L L Pi Bij Pj
i=1 j=1

B l1 P12 + 2B 12 P1 P2 + B22 Pi
= (0.001) x (41.7)2 + 2 (-0.0005) (41.7) (50.2)
+ (0.0024).(50.2)2 = 5.7MW
2
Total generation = PT = LP =41.7 + 50.2 =919Mw
wwTotal power received
i=l
2

w.E PR = PT - PL = 91.9 - 5.7 = 86.2MW.

3.20 Hydro Electric Plant Models

asy
Conventional hydroelectric plants are classified as run - of - river plants, run - of - river
plants with pondage and storage type phmts. In the former type water is utilized as is available

En
in the stream, as there is no provision for storage. Where there is pondage provision, hourly
fluctuations in load can be met. In the later type, where storage is provided the water stored

gi
during the excess water period can be utilized during the lean season or when power demand

nee
is high. The plant may be a single development on a river or there may be several plants
constructed and cascaded on the same river (Fig. 3.13). In some cases inter connection of

r
plants on different streams is also possible.

ing
. net

Power
Tail
_-.::- ...race
'-- ............
Fig. 3.13 Cascaded hydro electric plants

Downloaded From : www.EasyEngineering.net

120 Operation and Control in Power Systems

3.21 Pumped Storage Plant

The pumped storage hydroelectric plants generally store water to supply peak load oemands,
so that fuel is saved at thermal plants. For this purpose, at light load periods water is pumped
to the reservoir back from the tail water pond using power from the grid. The power house
may either have both turbines and pumps separately or reversible pump turbines. The operating
characteristics of a pumped storage plant are shown in Fig. 3.14.

ww
w.E
asy
En r-----~~-r----------70--------~ P
12gi I------rl;=rl
Power
nee
am

12
r ing
p.rn
Load Curve
Power
supplied . net
·12

Fig. 3.14 Dumped storage plant characteristics

Let ee be the energy spent to pump water to the reservoir. By releasing this water at
peak load times the energy supplied to the load is es' The ratio (eglee) is usually of the order
60-70%. Pumped storage plants are to be operated in such a manner that due to the p~ load
chipping on the load curve, the saving in fuel cost thus achieved should exceed the pumping of
water charges.

3.22 Hydro Thermal Scheduling

Most of the power systems are a mixof different modes of generating stations of which the
thermal and hydro generating units are predominant. While in some systems hydro generation

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 121

may be more than thermal generation in some other cases it may be the other way. The
operating cost of thermal plants is high even though their capital cost is low. In case of hydro
electric plants, the running costs are very low, but the capital cost is high as construction of
dams, canals, penstocks, surge tanks and other elements of development are involved in addition
to the power house. The hydro plants can be started easily and can be assigned load in very
short time. This is not so in case of thermal plants, as it requires several hours to bring the
boiler, super heater, and turbine system ready to take the load allotment. For the reason
mentioned, the hydro plants can handle fast changing loads effectively. The thermal plants in
contrast are slow in response. For this reason, the thermal plants are more suitable to operate
as base load plants, leaving hydro plants to operate as peak load plants.

ww
However, the exact mode of operation depends upon the type of the development, and
factors such as storage and pondage, and the amount of water that is available is the most

w
important consideration. A plant may be run - off river, run - off river with pondage, storage
or pumped storage type.
.Ea
Whatever, may be the type of plant, it is necessary to utilize the total quantity of water

syE
available in hydro development so that maximum economy is achieved. The economic scheduling
in the integrated operation is however, made difficult as water release policy for power is

ngi
subject to a variety of constraints. There are multiple water usages which are to be satisfied.
Determination of the so called pseudo - fuel cost or cost for water usage for use in conjunction

nee
with incremental water rate characteristic is a formidable exercise. Nevertheless, hydro thermal
economic scheduling is possible with assumptions made wherever necessary.

r
In systems where there is close balance between hydro and thermal generation and in

ing
systems where the hydro capacity is only a fraction of the total capacity, it is generally desired

.
to schedule generation such that thermal generating costs are minimized.

3.23 Energy Scheduling Method

Consider two plants, one of hydroelectric and the other thermal. Let these both supply a
net
common load.
Let for any time period K the maximum hydro power available.

P~ ~ Pload,k k = 1, 2, .............. K ..... (3.74)

where Pload, k is the load power during the time period k. It is presumed that the energy
available from the hydro plant is not sufficient to meet the load demand.
K K
LPHKn K ::;; LPload,knK ..... (3.75)
k=! k=!

wHere nk is the number of hours in period k

Downloaded From : www.EasyEngineering.net

.122 Operation and Control in Power Systems

Let ..... (3.76)

= total period of time over which energy schedule is required

It is proposed to utilize the hydro energy completely in such a way that the operating
cost of the thermal plant is minimized.
K
Load energy to be supplied = LP1oad,k n K ..... (3.77)
k=!

ww K
Hydro energy to be utilized = LPHk n K
k=l
..... (3.78)

w.E
Thermal energy required

asy ..... (3.79)

intervals Ks
E
Let the thermal plant be operated for time period less than Tmax and for number of

ngi
Hence,
nee ..... (3.80)

where Ps denotes steam power

The problem can then, be stated as rin
~F(PSK)nk g.n ..... (3.81)

et
MinFT =
k=!

subject to ..... (3.82)

The Lagrange function is given by

L = %F(Psk)n k +{ Es - %Pskn k ) ..... (3.83)

aL = dF(PSk) _ A = 0 for k = 1, 2, ...... , Ks ..... {3.84)

ap Sk dPSk
~

dF(PSk )
(i.e.) -.:-=:.;.. =A for k = 1,2, ...... , Ks ..... (3.85)
dP Sk

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 123

Eqn. (3.85) indicates that the steam plant must be run at constant incremental production
cost for the entire time period of its operation. Denoting this value as P~k' the schedule is
depicted in Fig. (3.15) graphically.

ww
w.E
o
k >1
asy Fig. 3.15 Hydro thermal co-ordination

En
gi
Let the steam plant cost characteristic be expressed by
F(Ps ) = aP~ + bPs + C
nee
The total cost of running the steam plant over the interval Ts
..... (3.86)

FT = (apt +bP~ +C)Ts r ing ..... (3.87)

Infact FT is also given by

FT = !F(p~)nk
. net
k=l

..... (3.88)
Further

..... (3.89)

E
T =_s
Hence, ..... (3.90)
s P~

Downloaded From : www.EasyEngineering.net

124 Operation and Control in Power Systems

So, ..... (3.91)

...... (3.92)

ww
w.E
asy ..... (3.93)

En
The steam unit is operated at its maximum efficiency through out the time period Ts'

gi
This can be proved as follows:
Let fc be the fuel cost.
nee
F(P s) = a
=fcH(Ps)
pi + bPs + c
r ing ..... (3.94)
where function H denotes the heat value
The heat rate is then given by
. net
H(Ps )
Ps
= ~[aPi + bPs +
fc Ps
c] ..... (3.95)

~[H(PS)] - ~[_1 (aPs + b + ~)]

dP s Ps dP s fc Ps

..... (3.96)

P =
s k-=Pso
a ..... (3.97)

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 125

3.24 Short Term Hydro Thermal Scheduling

3.24.1 Method of Lagrange Multipliers (losses neglected)
Consider a power system which contains both hydro and steam power generating stations.
Consider that the entire hydro generati~s equal to PH and the total steam power is Ps. With
such equivalent power generating station let a lo-ad- PL be supplied as shown in Fig. (3.16).

Hydro
POWER
SYSTEM Power demand

ww Steam

w .Ea Fig. 3.16 Hydro Thermal scheduling

Let the combined operation be over a period of time T. Let this time period be divided

syE
into intervals 1, 2, ..... J to suit the load curve so that
J
Ln J =T
J=1 ngi ..... (3.98)

nee
The total volume of water available for discharge over this time period.

r
Where Wj is the water rate for interval j. The fuel cost required to be minimized over ing . .... (3.99)

the time period T is given as

FT =Lnl(Ps)
J
. net
..... (3.100)
j=l

For load balance, the equality constraint is

P10adj - PSJ - PHj = 0 j = 1, 2, ........ , J ..... (3.101)

The loads are assumed to remain constant during time intervals considered. The total
value of water at the beginning and at the vend of the interval T, in the reservoir are Wi and Wf
respectively.
During this period of scheduling the head of water is assumed to remain constant.
The input - output characteristic for the equivalent hydro plant is given by
w = w(PH) •...• (3.102)

Downloaded From : www.EasyEngineering.net

126 Operation and Control in Power Systems

The Lagrange function for minimization of eqn. (3.100) subject to the constraints (3.99)
and (3.101) is \

L= tnJ F(ps;l+AJP\.... - PSj - PHJ+ {t njW 1

j(plij)- W ... (3.103)

For any specific value of j = k, the necessary conditions are

8L 8L
--=0 and - - = 0 ..... (3.104)
8PSk OPHk
giving

ww
and
w.E ..... (3.105)

asy
solution to above two equations gives the economic generations at steam and hydro plants over
any time interval. The incremental production cost at the steam plants must be the same as

En
incremental production cost at the hydro plants. For simplicity nk may be taken one unit.
So that

dFs =A
dPs gi nee
and
dw
y.-=A
dPH rin ... :.(3.106)

3.24.2 Lagrange Multipliers Method Transmission Losses Considered g.n

If the transmission losses are considered then the equality constraint includes PL' the loss
terms.
[P IOadJ + PLJ - PSj - PHj] = 0
et
.. ... (3.107)
The Lagrange function changes now into

..... (3.108)

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 127

dF(PSk ) '\ aPLk '\

i.e. nk +lI.k--=lI.k
dPHk aPSk

dWk aPLk
and ynk --(PHk)+Yk - - = A.k
dPHk aPHk
since k is chosen arbitrarily, and by considering the time period nk = 1
The equations reduce to

dF(PS) + A. aPL = A. ..... (3.109)

ww dPs aPs

Y dw(Pk ) + A. aPL = A.
and

w.E dP H aPH
..... (3.110)

It can be shown that the above equations are valid for any number of steam plants ns

Hence a
and for any number of hydro plants nH.

syE
dF(PSi )
ngi aPL
--'---"''"'-+A.--=A.
dPsl aPSi
.
for 1-1, 2, ......... , ns ..... (3.111)

and
dw(PHk )
Y dP
aPL
nee
+A. aP = A. for k = 1,2, ........ , nH ••••• (3.112)
Hk Hk
rin
the above equations are called coordination equations, the solution to which will give tlle
economic schedule for Psi and PHk.
g.n
3.24.3 Short Term Hydro Thermal Scheduling using B-Coemcients for
Transmission losses et
Let the total number of generating stations be 'n'. Out of these n stations, S stations are steam
power stations and the remaining H stations are hydro electric generating stations so that
S+H = n ..... (3.113)
Let the water input rate at the jth hydro plant be assumed as Wj m3/sec
The total transmission losses in the lines of the system are given by
s s n n
PL =:£ :£ P B, PsJ· +:£ :£ PH' B .. I1IJ.
i=l j=l 51 1J i=s+l j=s+l 1 1J

s n
+2:£ :£ P, B ..
i=l j=s+1 51 1J
l1Ii~ ..... (3.114)

Where the subscripts S and H refer to steam and hydro plants.

Downloaded From : www.EasyEngineering.net

128 Operation and Control in Power Systems

The total input hourly rate to all the thermal plants is

F = iF RSI ..... (3.115)

T i=l I Ihr
The incremental cost of production at ith steam plant

_ dFI Rsl
- dP. IMwhr ..... (3.116)
51

The incremental transmission losses at steam and hydro plants are given respectively by

8PL and 8PL

ww 8Ps! 8PHJ
The incremental rate of water flow at jth hydro plant

w.E%
__ J m
_
Ow·
- OP sec ..... (3.117)

asy Hj
The total power received by the loads,
s
P=LP+ En n
LP.-I\. ..... (3.118)
R

gi i=l 51 j=s+l HJ

nee
It is desired to make the total input to the system over a period T a minimum.
From Calculus of variations for FT to be a minimum over the time period T, the first
variation of eq. (3.118) must be set equal to zero.
rin
s
L op·+
i=l SI
L
n
j=s+l
OP.- L
HJ
s· 8PL
--
i=l apsi
0 p. -
SI
n
~
oPL
--8P
j=s+lOPHj g.n
HJ = 0 ..... (3.119)

The quantity of water available over the period T is assumed constant.

T
et
i.e., Jw jdt = W for j = s + 1, ... , n ..... (3.120)
o

Then, it is further required to make

T
f
[ FTdt + J=~+l Y
n
J f 1
T
wjdt ..... (3.121)

a minimum; where y} are constant multipliers.

The first variation of the quantity in equation must vanish. Hence

..... (3.122)

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 129

s 17FT n Owj
i.e., L - ap L r·--aPH· - 0 ..... (3.123)
i=1 aPsi si + j=s+1 J aPHj J - ......

Where OPsi and aPHj are the variations in steam power generation and hydro power
generation respectively.
From equation (3.119)

~i=1(1- aPsi - -1
aPL 1Opsi + Ln [ 1 -171\ a~j=o ... (3.124)
j=s+l a~j

wwFor a small variation of oPHm at the mth hydro plant eqn. (3.124) can be split into the
following form.

w~l.E ±
[ 1 - ~ ap11m = - [1 - oPL ] ap -.
oP.
~
J=s+l
(1 - 0It. )aPHi"
OllI ..... (3.125)

asy
1=1 51 J
j*m
Again, eqn. (3.123) can be rewritten as

En
Hence gi nee
..... (3.126)

Ow aF Ow

r
C1 n
Ym-_m apHm =- L ap -Y L aPH

ing
_T _J
ap'Hm 1=1 ap.. 51 J J=5+1 8P, ~ ..... (3.127)
p'm HJ

Multiplying Eq. 3.127 on both sides by (1- 8~~ J .net

-y. ~
l ~=s+1
Ow; aP (ll- 8PL ")
8P,Hl. Hi~ 8P ...... (3.128)
l"m m

In the above eqn. (3.128) if

(1- a~ J8PHm is replaced by the quantity on the right hand side of eqn. (3.124) we
~
obtam.

.. .... (3.129)

Downloaded From : www.EasyEngineering.net

130 Operation and Control in Power Systems

soFT
=-I:-oP
I=IOP
( OPL 1
51 ll--)-Y
oRlim JJ"m
J=s+1
Ow J
I:n -oR [OPL
ORHJ. H C 1:iD.._.
-- 1 ....... (3.130)
Sl V~ l1J

Rearranging the tenn

soFT ( OPL 1 n Ow ( OPL 1

I:-ll--)OP
1=1 OP51 oRlim -I:y _ m ll--)OP
'I 1=1 m oP.lim oRlim ..

OItl
ww + I:n
~=s+ 1
J~m
J
Owj
r· - - [ 1 - - - oPJij-
OPHj Ol\Ij

w.E ~ Ow m ( OPL 1
~=s+l Ym ~ll- oP. ) oPHj = 0 ..... (3.131)

asy J~m

Eqn. (3.131) can be put as

lim Hj

J _ El
.~
l=lOPsi ngi
aFT (1- OPL
OPHm
Ym Ow m (1- oPL 1 oP,;
oHm apsi )

+ f Ow j l(l- oPL 1)
YJ Ym OR -A. m OP. nee
Ow m l(l- ap.
OPL 1)op'Hj =0 ..•••(3.132)
1=5+1
J"m Hj
OR
lim lim

rin
lim

g.n
Note that Opsi and OpHj are the variations in powers at steam and hydro plants respectively
and hence are finite. Then, to make the equation (3.128) satisfied, each of the coefficients of
the variations must be zero.

:. OFT l(l- OPL )1_ y Ow m (1- oPL 1=0

.
et
..... (3.133)
oP.. aPHj m aplim l a p.. )

and Yj
Ow, ( _ aPL
oP II 1 Ow m (
OR ) - Ym - - 1 - - - =0
0It) ..... (3.134)
Hi lim OPHm 0I\un
In other words,
OF
_T
oW m
Owi
ap OP aPHj
(1- ~PJ) = Ym (ll _ ;PL 1) = Yj (ll _ OPL 1)
l op.. OPHj oPHj
= constant ..... (3.135)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 131

The partial derivatives in the above equation are also the total derivatives. It may also be
recognized that each of the term in the eq. 3.13 5 is also the incremental cost of the received

power in Rs/ MW. If A. denotes their incremental cost, then :T =!i

51 51
is the cost of generation

at each plant i and it depends on the generation at that plant only

( ')
~l_l_j=A.
d
dP 1- OP
51 cp..
L

. .... (3.136)

ww r -)
Ow (
) apH,j
l-j
1-~
=A.
1 ')

Also w.E
asy ..... (3.137)

'Y) dw, +A. oPL =A- En

and
dPHi OPH)
gi nee
The above equations are called coordination equations for hydro thermal combined operation.
Solution to these equations results in minimizing the input costs to supply the given load.

Let
dF·
_1
dPsi
= a. P . + b.
1 s1
r ing . .... (3.138)

.
1

and
dWj
-dP = c· PH· + d.
Hj J J J
net
..... (3.139)

where a and c are the slopes of the incremental cost curves and b and d are the intercepts.

Further, PL = ~k Pk BIG PI ..... (3.140)

Partially differentiating PL in eqn. (3.140)
and substituting in eqn. (3.137)

a·I P51. + b.1 + 2 A. (±

K=I'
Bk Psi< ~ BilHi) =A-
+ )=1+1 ..... (3.141)

..... (3.142)

Downloaded From : www.EasyEngineering.net

132 Operation and Control in Power Systems

Rewriting the third term on the left side of both the above equations, as
S
L Bki P k = BIZ',PSI' + ..... (3.143)
K=1 S

n n
and L BjkI1Ik =B. Ph' + L BkiP k ..... (3.144)
K=s+1 1) ~ K=s+1 s .
K~j

combining the terms containing Psi and PHj for solution

ww
w.E ..... (3.145)

and a syE
ngi
d) n s
l-"i - L 2B)kPHk - L 2B)IPsl
k=s+1 1=1
..... (3.146)

nee
These two equations (3.145) and (3,146) can be solved iteratively till convergence is obtained
assuming suitable values for A. and Yj .
rin
E 3.7 The input - output data for a particular hydro - plant is given in Fig. E 3.7.1. Find the
incremental water rate characteristic and convert it into an equivalent incremental g.n
production cost characteristic taking the cost of water as 1Q-3Rs.lm3.
et
r--
""'
'"
1600
1400
1200
'"'e- ' 1000
800
~
.c 600
~ 400
i5
~
200

0 50 100 150 200 250 300 350 400 450

P Plant generation (MW) ----+

Fig. E3.7.1 Input - Output Curve

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 133

Solution: The incremental water rate is obtained by finding the slope of the given input -
output curve at different points. A straight line segment approximation of incremental water
rate is obtained as in Fig. E 3.7.2.

4.0

I
~
3.5
3.0
:::E
----
...,rIl
2.5
5 2.0
c:G

ww ~ 1.5
I
1.0
~ Q.I
w.E
"0"0
0.5
-l--+--I--+----l'---+-4--L..+--I---- M W
0 100 200 300 400 500 600 700 800

a
Incremental water rate
syE
P ---.
Fig. 3.7.2 Incremental Water rate

ngi
1.7m3/s/MW for P = 0 to 563 MW

for P
(0.0233P - 11.4) m3/s/MW
563 to 640 MW nee
Incremental production cost
rin
Incremental water rate
1.7 x 3600 x 10-3 Rs.IMWh
x

g.n
cost of water

6.2 RslMW h for P = 0 to 563 MW

Incremental production cost for the range
et
P = 563 to 640 MW is given by
(0.0876 P - 41.2) Rs/MW h
E 3.8 Consider the following 6-bus system with three generating plants and three load buses.
G, and· G2 are steam power plants and G3 is hydro - electric plant.
Generating 1YPe No. of Units Capacity of each Total capacity
Station unit(MW) (MW)
G, Thermal 3 12.5 37.5
~ Thermal 2 60.0 120.0
G3 Hydro 2 60.0 120.0
Total Capacity 277.5

Downloaded From : www.EasyEngineering.net

134 .Operation and Control in Power Systems

Loads:
L1 160 MW
L2 35 MW
L3 72.5 MW
Total Load = 267.5 MW
Base Voltage = 220kV
Base MVA = 100
The B - coefficient matrix is given as

ww B = G 1 [0.0210
G1 G2
0.0034
G3
0181
1
w.E G2 0.0034 0.02497 0.0.003

G3 0.0181 0.003 0.050

GI - asy
Characteristics of steam and hydro plants
thermal station :

En
Calorific value of coal 5125 kcal/kg
Cost of coal
gi
Incremental fuel cost at near no - load
Incremental fuel cost at 40MW nee
Rs.45 per ton
Rs.21/MWh
Rs.23/MWh
Slope of incremental cost curve
rin
0.05
Intercept on y-axis 21
g.n
260
260 et
240 236 230
215
220

1
~
200
178
187
~ 180

o 6 12 16 20 22 24
Time (h) ----.
Fig. E 3.8 Load Curve

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 135

G2 - thermal station :
Calorific value of coal 4600 kcallkg
Cost of coal Rs.48 per ton
Incremental fuel cost at near no - load Rs.25/MWh
Incremental fuel cost at 40MW Rs.31.6/MWh
Slope of incremental cost curve 0.094
Intercept on y-axis 21
dF2
- = 0.094PS2 + 21
dP 2

ww
G 3 - Hydro electric plant
Assuming Y3 = 1.0 paise/IOO m3 of water at 104 MW output, the incremental plant

w
cost = Rs.19.5 per MW h and at 120 MW output the incremental plant cost = Rs.36

.Ea
per MWh. Slope of the incremental cost curve

Intercept on y-axis
y

= -
-16
88
syE
= 16.5 =1.03

Y1
dW
=--'l ngi
= 19.5;0 $ PH3 $104MW
- dP H3
= 1.03PH3 - 88; nee
104 $ PH3 $120MW
r ing
The system daily load curve is given in Fig. E 3.8 as shown. Obtain the economic
schedules.
Solution: . net
The scheduling equations are given from eqs. (6. 102) and (6.103) as

1-~-0.000068P2 -0.000362P
PI = ___A~___________________3
0.05 + 0.00042
A
21
1-- - 0.000068 PI - 0.000060 P3
A
P2=--~--------------------
-
0.094 + 0.0004994
A
88
I + - - 0.000362 PI - 0.000060 P2
P3 = A
1.03 + 0.00 I 0
A

Downloaded From : www.EasyEngineering.net

136 Operation and Control in Power Systems

PI + P2 + P3 - (0.00021 P~ + 0.0002497 P~ + 0.0005 P~)

+ (0.000068 PI P2 + 0.000362PIP3 + 0.OOO06P2P3) PR =

in order to schedule for a specified received load, the values of I are computed from
equation as

+ (p(d)
1.(1-1) 1.(1-2) 1
1.(1) = 1.(1-1)
R
_ p(I-I)
R

where the superscript i indicates the iteration being started.

l -
p(l-I) _ p(I-2)
R R

PR is the received load and P(d)Ris the desired received load.

To start the computation, two initial values for A are assumed, namely

ww AI = 22.0

w.E 1.2 = 30.0

Economic schedules are then computed for different water rates over a wide range

a dW 1
Y3 dP'
syE
3
= 0.515P3 -44atY3 =0.5paiseIl00m

ngi= 0.618P3 - 52.8 at Y:1 = 0.6 paise II 00 m

=0.742P:1 -63.5atY:1 =0.7paiseIl00m

nee
= 0.824P3 - 70.4 at Y3 = 0.8 paise / 100 m
= 0.927P3 -79.2atY3 = 0.9paisell00m
rin
= 1.03P3
= 1.24P3
-

-
88 at Y3 = 1.0 paise / 100 m
105.5 at Y3 = 1.2 paise /I 00 m g .ne
the results of the economic schedule are shown in tables E 3.8 (a) to E 3.8 (g)
corresponding to different water rates.
Table E 3.8 (a)
t
Y = 0.5 PslI 00m 3
p(d) P2 PR A
R PI P3 Ptotal Ploss
(MW) (MW) (MW) (MW) (MW) (MW) (MW)
178.000 342786 28.1089 125.870 188.168 102033 177.965 242133
187.000 37.5000 33.4225 126.731 197.674 10.6672 187.006 24.8071
215.000 37.5000 58.5480 131.085 227.133 12.1321 215.001 27.5978
230.000 37.5000 72.1076 133.447 243.055 13.0705 229.984 29.1365
\

i 236.000 37.5000 77.5726 134.404 249.477 13.4780 235.999 29.7608

260.000 37.5000 99.5099 138274 275.284 15.2841 259.999 32.3101

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 137

Table E 3.8 (b)

y = 0.6 Ps!1 00m 3
p(d) P2 P3 PR A.
R PI Plolal Ploss
(MW) (MW) (MW) (MW) (MW) (MW) (MW)

178.000 37.5000 30.4325 119.685 187.617 9.60962 178.007 24.4699

187.000 37.5000 38.6243 120.872 196.997 9.9923 187.004 25.3683
215.000 37.5000 64.2988 124.632 226.431 11.4309 215.000 28.2338
230.000 37.5000 78.1710 126.688 242.359 12.3596 229.999 29.8147

ww
236.000
260.000
37.5000
37.5000
83.7402
106.168
127.517
130.888
248.758
274.556
12.7620
14.5569
235.996
259.999
30.4564
33.0778

w .Ea Table E 3.8 (c)

y = 0.7 Ps/I00m 3
p(d)
R PI P2
syE
P3 Plolal Ploss PR A.

ngi
(MW) (MW) (MW) (MW) (MW) (MW) (MW)
178.000 37.5000 34.7763 114.802 187.080 9.07387 172.006 24.9369
187.000
215.000
37.5000
37.5000
43.1259
69.3002
115.827
119.072
196.453
225.872 nee
9.44968
10.8717
187.003
215.000
25.8570
28.7397
230.000
236.000
37.5000
37.5000
83.4432
89.1077
120.848
121.564
241-791
248.171
r
11.7948
12.1942 ing
229.997
235.977
30.4096
31.0668
260.000 37.5000 112.001 124.484 I 273.985 13.9878 259.998
. 33.7546
net
Table E 3.8 (d)
y = 0.8 Psll 00m 3
p(d) P2 P3 PR A.
R PI Ptolal Ploss
(MW) (MW) (MW) (MW) (MW) (MW) (MW)

178.000 37.5000 37.1910 112.105 186.796 8.79129 178.005 25.1972

187.000 37.5000 45.6237 113.044 196.167 9.16186 187.003 26.1275
215.000 37.5000 72.0465 116.019 225.565 10.5818 214.984 29.0978
230.000 37.5000 86.4328 117.651 241.494 11.5045 229.989 30.7381
236.000 37.5000 92.0947 118.312 247.907 11.9066 236.000 31.4043
260.000 37.5000 115.205 120.994 273.699 13.7017 259.997 34.1278

Downloaded From : www.EasyEngineering.net

138 Operation and Control in Power Systems

Table E 3.8 (e)

y = 0.9 Ps/IOOm 3
p(d) P2 P3 PR A.
R PI Ptotal Ploss
(MW) (MW) (MW) (MW) (MW) (MW) (MW)
178.000 37.5000 39.5740 109.453 186.527 8.52298 178.004 25.4547
187.000 37.5000 38.0938 110.303 195.897 8.89451 187.002 26.3966
215.000 37.5000 74.8023 112.999 225.302 10.3088 214.993 29.4050
230.000 37.5000 892306 114.447 241.208 112304 229.977 31.0671

ww 236.000
260.000
37.5000
37.5000
95.0556
118.416
115.078
117.513
247.634
273.429
11.6339
13.4335
236.000
259.996
31.7422
34.5033

w.E Table E 3.8 (t)

asy y = 1.0 Ps/100m 3

(MW)
p(d)
R PI
(MW)
P2
(MW)E P3
(MW)
ngi
Ptotal
(MW)
Ploss
(MW)
PR
(MW)
A.

178.000
187.000
37.5000
35.3000
41.5314
50.1235
107232
109.682
186.314
195.682 nee
8.31038
8.68070
178.004
187.002
25.666
262182
215.000 37.5000 77.0672 110.525 225.048
rin
10.0942 214.998 29.6583

g.n
230.000 35.3000 91.6397 111.878 241.018 11.0182 230.000 31.3386
236.000 37.5000 97.4943 112.426 247.421 11.4210 236.000 32.0212
260.000 37.5000 121.065 114.658 273221

Table E 3.8 (g)

132262 259.995 34.8137
et
y = 1.2 Ps/IOOm3
p(d) P2 P3 PR A.
R PI Ptotal Ploss
(MW) (MW) (MW) (MW) (MW) (MW) (MW)
178.000 37.5000 44.9039 103.559 185.963 7.96039 178.003 26.0326
187.000 37.5000 53.6119 104.220 195.332 8.33065 187.001 27.0000
215.000 37.5000 80.9275 106.321 224.749 9.74867 215.000 30.0921
230.000 37.5000 95.7020 107.476 240.678 10.6780 230.000 31.8021
236.000 37.5000 101.639 107.913 247.083 11.0835 235.999 32.4970
260.000 37.5000 125.548 109.849 272.897 12.9033 259.993 35.3412

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 139

E 3.9 A system consists of two generators with the following characteristics

FI = (7 PI + 0.03 p/ + 70) 106

F2 = (5 P2 + 0.05 pi + 100) '106

Where F and P are fuel in put in K-cal/hr and unit output in MW respectively. The daily
load cycle is given as follows.

Time Load
12 midnight 6 am 50MW

ww 6 am to 6 pm 150MW

w.E 6 pm to 12 midnight 50MW

asy
Give the economic schedule for the three periods of the day
Solution:

E
Let the cost of the fuel be the same at both the plants.
dF
ngi
_I = (7 + 0.06PI )10 6 K-cal/MWhr
dPI
If C is cost of fuel in Rs/K-cal. nee
The incremental production cost at plant 1
rin
dC I 6
=-=(7+0.06PI )10 xC Rs per MWhr
dPI g.n
similarly
dC 2
=-
dP 2
6
= (5 + 0.IP 2 )10 xC Rs per MWhr
et
For economy
dP I dP 2
and P I +P 2 = 50 MW (i)
(7 + 0.06 PI) \06 xC = (5 + 0.1 P2 )106 xC (ii)
solving (i) and (ii) for PI and P2
PI = 18.75 MW
P2 = 31.25 MW
For PI + P2 = 150MW

Downloaded From : www.EasyEngineering.net

140 Operation and Control in Power Systems

We obtain (7 + 0.06 PI) = (5 + 0.IP 2)

7 + 0.06 PI =:5 -+- 0.1 (150 - PI)
0.06 PI + O. 1 PI = 5- 7 -+- I5
0.16 PI = 13
PI = 81.25 MW
P2 = 68.75 MW

E3.10 The incremental production cost data of two plants are dFI = 2 + PI and dF2 = 1.5 + P2
dPI dP2

ww where PI and P2 are expressed in per unit on I OOMVA base. Assume that both the units
are in operation and that the maximum loading of each unit is I OOMW and the minimum

w loading of each unit is 10MW. The loss coefficients on a 100MVA base are given by

.Ea 8=[0.10 -0.05]

syE
- 0.05 0.2
for A = 2.5 solve the coordination equations, by the iterative method.
Solution:
ngi
[PC I = a I PI + b I
nee
I-~
A -28 21 P2
a
2
r ing
I - - - 2 x (-0.05) x p,
2.5
I
-
l
-+28
A
11
- + 2x (0.10)
2.5
. net
1-0.8+0.1P2 = 0.2 =0.333 at P, =0
0.4 + 0.2 0.6 -

I- ~- 2 x (-0.05) X P,
1-0.6+0.1PI
2.5 -
I 0.4 + 0.4
- + 2x(0.02)
2.5

At PI = 0.333,

P = 0.4 + 0.0333 = 0.4333 = 0.541625

2
0.8 0.8

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 141

performing iterations further

PI = 0.2 + 0.1(0.5416) = 0.2 + 0.5416 = 0.25416 = 0.4236

0.6 0.6 0.6

P = 0.4 + 0.1(0.4236) = 0.4 + 0.04236 = 0.55295

2
0.8 0.8

P, = 0.2 + 0.055295 = 0.42549 p) = 0.4 + 0.042549 = 0.5531865

. 0.6 - 0.8

ww The results are

w.E Iteration
I
PI
0.333
P2
0.541625

asy 2
3
0.4236
0.42546
0.55295
0.55318

At the end of 3 iterations En

PI = 0.4255
P2 = 0.5532
gi nee
If one more iteration is carried out

PI = 0.2 + 0.05532 = 0.4255

r ing
p)
-
0.6

= 0.4 + 0.04255 =0.5532

0.5
. net
No change in the values.
E3.1l The following incremental costs pertain to a 2 plant system.

dFI = 0.03P I + 14 Rs/MWhr

dP I

dF2 = 0.04P) + 10 Rs/MWhr

dP 2 -

The loss coefficient are 8 11 =0.001 (MWt I

8 12 = 8 22 = O. If A for the system is Rs.30 ! MWhr compute the required generation at
the 2 plants and the loss in the system.

Downloaded From : www.EasyEngineering.net

142 Operation and Control in Power Systems

Solution:
For economy

I - -bl - 28 21 P2 I -- ~- 0
A. 30
PI::: a ::: 003 ::: 177.78MW
1
-+28 _ .-+2xO.001
11
A. 30

2 b 10
1- --
A. - 28 12 P1 1- ---
P2 -.lQ
ww
::: = = 500MW
a2 0.04
-+28 22
A. 30

w.E
The transmission losses are

asyL PL =
2

i=1
2
LP,8,lJ
J=1

E ngi
nee
rin
E3.12 Consider a steam station with two units the input - output characteristics being specified
by

g.n
F2 = 120 + 6P 2 + 0.04 pi
In scheduling a load of 1OOMW by equal incremental cost method, the incremental cost
et
of unit 1 is specified wrongly by 10% more than the true value while that of unit 2 is
specified by 6% less than the true value
Find (i) The change in generation schedules and
(ii) The change in the total cost of generation.
Solution:
The incremental production cost at plant I is 8 + 0.048 PI
It is specified 10% more.
That is (8 + 0.048 PI) 1.1 = 8.8 + 0.0528 Pl.
The incremental production cost at plant 2 is 6 + 0.08 P2
\

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 143

It is specified 6% les
That is as (6 + 0.08P 2) 0.94 = 5.64 + 0.0752 P2
Schedule with correct incremental production costs:
8 + 0.048 PI = 6 + 0.08P 2
PI + P2 = 100MW
Solving PI = 46.875MW
P2 = 53.125 MW
Cost of production at plant I

ww C I = 80 + 8
Case of production at plant 2
x 46.875 + 0.024 x 46.875 2 = 507.73438

Total costw.E C 2 = 120 + 6 x 53.125 + 0.04 x 53.125 2 = 551.64063

asy
Schedule with incorrect specification:
8.8 + 0.0528P I = 5.64 + 0.752P 2

En PI + P2 = 100
Solving
gi
PI = 34.0625
P2 = 65.9375 nee
Cost of production at plant 2
r ing
C:
Cost of production plant 1
=80 + 8 x 34.0625 + 0.024 x 34.0625 2 = 380.34609
. net
C~ 2
= 120 + 6 x 65.9375 + 6.04 x 65.9375 = 569.53528
Total cost CI = C: +C~ = 949.88218
Change in total cost of generation
= 1029.375 - 949.8821=109.49
E3.t3 A two bus system shown in figure supplies a load at bus 2. If 50MW is transmitted
from plant I to load at bus 2 overthe line, the loss is 2.5MW. The incremental production
costs at both the plants are given with usual notation by
dC)
- = 0.03P) + IS and
dP)

dC 2 = 0.05P + 20
dP2 2

Downloaded From : www.EasyEngineering.net

144 Operation and Control in Power Systems

< <

The value of A. is 23 Rs/MWhr

Determine the generation schedule for economy with losses coordinated. What will be
the generation schedule if the losses are not coordinated but considered. What will be
the savings by coordination of losses into economic schedule?

2
Line

ww Plant 1 Plant 2

w.E Fig. E 3.13

Load

Solution:
asy
En
The load is at bus 2 only. There is transmission loss only from the power supplied by plant I.
No power is transmitted over the line from plant 2.

Hence, gi nee
B = 2.5 =O.OOI(MW)-1 r ing
Case (i)
II

Losser coordinated
SOx 50

dC I + A. dPL = A.
. net
dPI dPI

dC 2 + A. dP L = A.
dP2 dP 2
0.03P I + 15 + 2 x O.OOIP I = 23
PI = 258.06MW
0.05 P2 + 20 + 0 = 23
P2 = 3/0.05 = 60MW
Total generation = PI + P2 = 318.06MW
Total load demand = PI + P2 - PLOSS
= 318.06 - 0.001 x 258.06 2 = 60.595MW
Po = 318.06 - 66.595 = 251.465MW

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 145

Case (ii) Economic Schedule for PD=251.465MW with losses not coordinated.
0.03P I + 15 = 0.05P 2 + 20 (i)

PI + P2 = B II PI2 +251.465 (ii)

0.05 P2 = 0.03 PI + 15 - 20

= 0.03PI - 5 == 0.6P _ 100

P2 0.05 . I
Substituting this P2 value in equation (ii)
PI + 0.6 P I - 100 = 0.001 p?
+ 251.465

ww i.e., 0.001 P? - 1.6 PI + 351.465 = 0

w.E
Solving PI = 262.88 (valid answer)
Then, P2 = 157.728 - 100 = 57.728MW

asy
Total generation PI + P2 = 320.608MW
The savings at plant 1 due to loss coordination.

J En m~
2
(0.03P I + 15)dPI == 0.03P1 + 15PI I
m~
==

gi
25806

nee
[69105.9 - 66594.9] == 2511.0 Rs / Hr
25806

The extra expenditure at plant 2 due to increased production

57 728
f (0.05P
r
2
+ 20)dP2 == 0. 05P 2 + 20P 2 160
ing
57728

.
2 =
60

[180 + 1154.56] = 1334.56 Rs / Hr net

Net savings = 2511 - 1334.56 = I I 76.44Rs/Hr
E3.14 Consider the two bus system shown in figure. The incremental production costs at the
two generating stations are given by

-
dC I == 0.005P + 5 an d
I
dPI

dC 2 = 0.004P J + 7
dP 2 -

The B coefficients are given in matrix form as

B == [0.0002 - 0.00005]
- 0.00005 0.0003

Downloaded From : www.EasyEngineering.net

146 Operation and Control in Power Systems

Determine the penalty factors at both the buses and also the approximate penalty factors.
Given A. == 8

Line

Fig. E 3.14

ww
Solution:

w.E BII == 0.0002; B22 = 0.0003; BI1 == B11 == -0.00005

asy
En
giaP L
-==2P
ap 2 B 22 +2P I B 21 =2
2

The coordination equations are

nee
0.0003P 2 +2xO.00005P I

r
(0.005 PI + 5) + 8 (0.0004 PI + 0.0001 P2) == 8
(0.004 P2 + 6) + 8 (0.0006 P2 + 0.0001 PI) ing
== 8
Simplifying
0.005 PI + 0.0032 PI + 0.0008 P2 = 3
. net
0.004 PI + 0.0048 PI + 0.0008 P2 = 2
(i.e.) 0.0082 PI + 0.0008 P2 == 3
0.0008 PI + 0.0088 P2 == 2
Solving for PI and P2
PI == 346.75 and P2 == 195.75

I - aP L ) = 1- 2 x 346.75 x 0.0002 + 0.0001 x 195.75 == 0.8417

[ aPI

1- aP L ) == 1- 2 x 0.0003 x 195.75 + 0.0001 x 346.75 == 0.847875

[ aP2

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 147

Penalty factor at plant 1 == == 1.17942

0.847875
1
Penalty factor at plant 2 == 0.8417 == 1.18807
Approximate penalty factors
At plant 1 the approximate penalty factor ==

( ap)
J
1+ aP L == I + 0.1387 + 0.019575 == 1.158275

ww
At plant 2 the approximate penalty factor ==

w.E J (
1+ aP L == 1+ 0.11745 + 0.034675 == 1.152125
ap2

asy
E3.15 Given the network in the figure shown along with the currents flowing in the lines and
the impedances of the lines in per unit on a 100MVA base. Compute the B. coefficients

En
for the network when the voltage at bus 1 is 1.0 LO o p.u.

gi
GEN I

nee
GEN2

rin
---.--'-+-- 3

g.n
et
4

Fig. E 3.15 (a)

Current in line a == 1.5 - jO.3 p.u.

Current in line b == 2 - jO.3 p.u.
Current in line c == 1 - jO.2 p.u.
Load current at bus2 == 2.5 - jO.5 p.u.
Impedance of line a == 0.01 + jO.06 p.u.
Impedance of line b == 0.01 + jO.05 p.u.
Impedance of line c == 0.0 I + jO.04 p.u.

Downloaded From : www.EasyEngineering.net

148 Operation and Control in Power Systems

Solution:
GEN I GEN2

1
ww
w.E 2 --'---r'-- 4

asy Fig. E 3.15 (b)

En
The current distribution factors

gidal == 1

d
bl
== 2 - jO.2
4.5 - jO.7
== 0.44135
nee
rin
del
== 2 - jO.2
4.5 - jO.7
== 0.44135

g.n
GEN I GEN2 et
Open

Fig. E 3.15 (c)

Downloaded From : www.EasyEngineering.net

Economic Operation of Power Systems 149

da2 = 0

= 2 - jO.2 = 0.44135
d b2 4.5 - jO.7

= 2.5 - jO.5 = 0.559825

de2 4.5 - jO.7
Voltages
Voltage at bus I = V I = 1.0 LOo
Voltage at bus 2 = (I + jO) + (0.0 I + jO.06) (1.5 - jO.3)
= 1.033 + jO.087 = 1.03366571 L4.814°

wwVoltage at bus 3 = (1.033 + jO.087) + (I - jO.2) (0.01 + jO.04)

= 1.051 +jO.0125= 1.0584L6.78257°

w.E
Voltages at bus 4 V 4 = (1.051 + jO.125) - (0.0 I + jO.05) (2 - jO.2) =
1.03 + jO.027 = 1.03035 L 1.500

asy
0 1 = Tan -I ( - 0.3) = _11.3 0

E
2 -I( - n
1.5

~.4) = -7.595 u
O = Tan

Cos (d l
gi - nee
d 2) = [- 11.3° - (-7.595 0 )] = Cos (-3.705 0 ) = 0.9979
Power factor at plant I (COS~I) = Cos (0 + 11.3°) = 0.9806
0

rin
Power factor at plant 2 (COS~2) = Cos (6.78257 + 7.595) = 0.96868
g.n
= 2 I
2
2 =[lxO.01+0.44135 xO.01+0.44135 xO.01]
2
et
I x 0.9806
= 0.01445

(1.036571)2 X (0.96868)2

= [0.44135 2 X 0.0 I + 0.559825 2 X 0.0 I + 0]

= 0.00464292

Downloaded From : www.EasyEngineering.net

150 Operation and Control in Power Systems

---------------------------

0.9979[0 + 0.44135 x 0.44135 x 0.0 I + 0.44135 x 0.559825 x 0.0 I]

(I x 1.036657)(0.96868 x 0.9806)

= 0.0044778857
on 100 MVA base the loss coefficients are

ww E" = 0.01445 x IO-~(MW)·'

w B22 = 0.0046429

.Ea
x

B'2 = B2, = 0.00447788

10-2(MWt'
x 10- 7(MWt'

syE
ngi
nee
r ing
. net

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 151

Questions

3.1 Explain the following terms with reference to thermal plants

(i) heat rate curve
(ii) incremental fuel rate cure
(iii) incremental production cost curve
3.2
ww
Explain the following terms with reference to hydro plants
(i) input out curve
(ii)
(iii)
w.E
incremental water rate curve
incremental production cost curve
3.3
a syE
How is generation scheduled among various generators when transmission losses are
neglected in a thermal system? Explain
3.4
ngi
Derive the transmission loss formula for a system consistmg of n-generating plants
supplying several loads inter connected through a transmission network.
3.5
nee
Derive expressions for economic distribution ofload between generating units considering
the effect of transmission losses.
3.6 What are coordination equation? Give their physical significance.
rin
3.7
3.8 g.n
What is a penalty factor in economic scheduling. Explain its significance.
Explain how the incremental production cost of a thermal power station can be

3.9
determined.
Discuss the general problem of economic operation of large inter connected areas.
et
3.10 State what is meant by base - load and peak - load stations. Discuss the combined
operation of hydro electric and steam power stations.
3.11 Derive an expression for the hourly loss in economy due to error in the representation of
input data.
3.12 What are the assumptions made in deriving transmission loss coefficients? Enumerate
them.
3.13 Explain hydro thermal economic load scheduling. Derive the necessary equations.

Downloaded From : www.EasyEngineering.net

152 Operation and Control in Power Systems

Problems
"
P3.1 (a) A power system consists of two, 120 MW units whose input cost data are
represented by the equations:
C, = 0.04 P? = + 22 P, + 800 Rupees / hour
C2 = 0.04 p22 = + 22 P2 + 1000 Rupees / hour
If the total received power PR = 200MW Determine the load sharing between
units for most economic operation.
(b) Discuss the general problem of economic operation oflarge interconnected areas.
P 3.2 (a) Derive an expression for the hourly loss in economy due to error in the

ww (b)
representation of input data.
The incremental fuel costs for two plants are given by

w.E dc
- ' = O.lP, + 20Rs./ Mw - Hr
dP,
dC 2
-
dP 2
= 0.15P 2 + 22.5Rs./ Mw - Hr

asy
The system is operating at the optimum condition with P, = P2 = 100MW and

q"
ap2 En
= 0.2

P 3.3 (a)
gi
Find the penalty factor at plant I and the incremental cost of received power.

nee
The incremental fuel costs for the two plants are given by
dc,
-
dP,
= 0.2P, + 45
dc,
rin
- - = 0.25P, + 34
dP2 -

g.n
where C is in Rs/hr and P is in MW. If both units operate at all times and

et
maximum and minimum loads on each are 100MW and 20MW respectively.
determine the economic load schedule of the plants for the loads of 80MW and
180MW Neglect the line losses.
(b) write short notes on physical interpretation of co ordination equation.
P 3.4 (a) Derive the conditions to be satisfied for economic operation of a loss less power
system.
(b) 150 MW, 220MW and 220MW are the ratings of three units located in a thermal
power station. Their respective incremental costs are given by the following
equations:

dc, = Rs(O.IIP, + 12) dc, (

- - = Rs 0.095P, + 14
) dc, = Rs(O.IP, +13)
dP, dP 2 - dP, .
where PI' P2 and P3 are the loads in MW Determine the economical load allocation
between the three units. when the total load on the station is (i) 350MW (ii)
500MW

Downloaded From : www.EasyEngineering.net

Economic Operation 0/ Power Systems 153

P 3.5 The equations of the input costs of three power plants operating in conjunction and
supply power to a system network are obtained as follows:
C I = 0.06 + 15P I + ISO Rupees I hour
C2 = 0.08 + 13P2 + 180 Rupees I hour
C3 = 0.10 + IOP 3 + 200 Rupees I hour
The incremental loss - rates of the network with respect to the plants 1,2 and 3 are
0.06, 0.09 and 1.0 per MW of generation, respectively. Determine the most economical
share of a total load of 120MW which each of the plants would take up for minimum
input cost of received power is Rupees per MWH.

ww
P 3.6 (a) The incremental costs for two generating plants are
IC I =0.1 PI + 20 Rupees/MW hour

w.E
IC 2 =0.1 P2 + 15 Rupees/MW hour
Where PI and P2 are in MW. The loss coefficients (Bmn) expressed in MW-I unit
are
a syE
BII = 0.001. B22 = 0.0024, BI2 = N21 = -0.0005. Compute the economical

ngi
generation scheduling corresponding to the Lagrangian multiplier 1.25 Rs/MW hr
and the corresponding system load that can be met with. If the total load is

nee
150MW, taking 5% change in the value of I, what should be the value of I in the
next iteration?
(b)
rin
What are the assumptions made in deriving the loss coefficients? "\ ""'\

P 3.7 The fuel inputs to two plants are given by

FI = 0.015 + 16P I + 50 F2 = 002 + 12P2 + 30 g.n
The load to be met is 200MW. Determine the economic operating schedule and the
corresponding cost of generation if the transmission line losses are coordinated.
et
The loss coefficients of the system are given by BII = 0.005; BI2 =-0.0012 and B21 =0.002.

P 3.8 A constant load of 300MW is supplied by two 200MW generators. , and 2 for which
the respective incremental fuel costs are

with powers PG in MW and costs C is Rs/h, determine

(a) The most economical division of load between the generators and
(b) The saving in Rs/day thereby obtained compared to equal load sharing between
the machines.

Downloaded From : www.EasyEngineering.net

154 Operation and Control in Power Systems

P 3.9 A system consists of two plants connected by a tie line and a load is located at plant 2.
When IOOMW are transmitted from plant I, a loss of IOMW takes place on the tie -
line. Determine the generation schedule at both the plants and the power received by the
load when I for the system is RS.25 per megawatt hour and the incremental fuel costs
are given by the equations:
dF dF
- ' = O.03P1 + 17Rs I MWr --2-=O.06P2 +19Rs/MWr
dP, dP 2
P3.10 The incremental fuel cost curves of generators A and B are shown below in the figure.
How would a load (i) more than 2 PG (ii) less than 2PG and (iii) equal to 2PG be shared

ww between A and B if both generators are running.

w.E I.e

as rI
:
· I
i

I
:
Gen. 8

Gen. A

yE
I I

ngi
--~----~----~---)~MW
MW
min
MW
nee
max

rin
P3.11 (a)
(b) g.n
Explain heat rate curve and cost curve. Bring out the differences between them
Determine the economic operating point of three units supplying a load of800MW.
The incremental fuel costs of the three units are:
dF
_I
dP,
= 6.48 + O.00256P1Rs. I MWh
dF
_2
dP 2
et
=7.85+0.00388PJRs'/MWh
-

dF
_3 = 7.97 + O.00964P, Rs.I MWh
dP 3 .

Downloaded From : www.EasyEngineering.net

w ww LOAD FLOW
4 OPTIMAL
.Ea
syE
Optimal load flow studies are concerned with economic operation of the system in all aspects.

ngi
These aspects include consideration of all constraints while satisfying the load balance. The
various types of inequality constraints have been already discussed in Chapter 3. Satisfying

nee
the load constraints itself is the load flow solution. Hence, in optimal operation, the cost
minimization is implemented in addition to aforcesaid considerations. It may be noted that

ri
even the slack bus voltage also has to be optimized for economy while the choice of slack bus
may be based on other considerations.
ng.
4.1 Reactive Power Control for Loss Minimization
n
Transmission losses are quadratic functions of bus-bar voltage magnitudes at all buses. With
usual notation. et
PLOSS = ± ±Iv, IIY'JIIVJI cos(8, -8J-8,J
,=1 J=1
..... (4.1 )

At each voltage controlled bus the best voltage that minimized the loss is determined.
Consider the bus impedance matrix ZBUS. Let R be the real part of the ZBUS matrix.
At each bus i, the current

P, - jQ, (C ·S· s: )
I, = A, + jB, = IV,I s:
osu, - J IOu, ..... (4.2)

Downloaded From : www.EasyEngineering.net

156 Operation and Control in Power Systems

The components AI and BI can be identified. The power loss.

. .... (4.3)

The necessary condition for PLOSS to be a minimum is dP LOss = 0 at all the reactive
dQ
generation buses. A gradient algorithm can be used for iteration

Qnew = QoJd +
, ,
k( ap80,
LOSS )OJd ..... ( 4.4)

ww
4.2
where k is a scalar parameter.

Gradient Method for Optimal Load Flow

w.E
Dommel and Tinney (1968) proposed a method for optimal load flow based on Lagrange and
Kuhn-Tucker conditions. In this method the variables are divided into two types X and Y.
X vector
a syE
includes the voltage magnitudes IVII and the phase angle of the voltage 8 1 at
each bus i for all P-Q type buses. At each P-V type bus 8 1 is considered as X - Variable .

ngi
.Y vector includes Pk and Q k at each P, Q bus k.
At the slack bus 8 k and IV kl are also y-vector components.

nee
It can be recognized that .Y is the control vector and X is the vector of ,lependent
variables.
At every bus the equality constraint h(X) =
rin
0 must be satisfied.
(i.e) Pgl
Qg, -
- P'I - PI = 0
0'1 - 0, = 0 g.n ...... (4.5)
Where the suffix g stands for generation bus and f stands for load bus.
The Lagrange function is formed as :
et
L = f + A' h ..... (4.6)
The necessary conditions for minimum of L are

..... (4.7)

aL =af + ah I .A = 0 ..... (4.8)

ay- cy- cy-

and aL = h = 0 ..... (4.9)

aA -

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 157

eh
Since X vector contains IVI and 8, eV is the Jacobian matrix of N-R method in load

flow solution.

Ie= _[aht )-1 ~ ..... (4.10)

ax- ax-
Of
where ax is known from incremental costs. Knowing the Jacobian and incremental

ww aL
w
cost curves. Ie can be computed. With Ie known -;- is determined using the gradient method.

.Ea
aL
L'y

syE
old
y new = y + k ay where k is a scalar parameter ..... (4.11 )

ngi
The computation is repeated till convergence is obtained. In order to consider the inequality
constraints, they are set and held at their limits, in case they are violated for all control variables

nee
't.. If the dependent variables are violated, then the penalty function method is lIsed to force the
variables back to their limits. It is said that the method offers good convergence in general.

4.3 Non - Linear Programming

r ing
Given the objective function, that is, the function for maximization or minimization

The function is subject to constraints of the type

. net
..... (4.12)

i = 1,2, ..... N ..... (4.13)

and j = 1,2, ..... m ..... (4.14)

If the functions f, g and h are all linear the optimization of the functional f can be
achieved through linear programming methods. However, if anyone of them is non-linear,
then non-linear programming methods are to be applied. The general conditions for the solution
of non-linear programming problem are given by Kuhn-Tucker conditions.
Forming the composite L - function

..... (4.15)

where a, are arbitrary

Downloaded From : www.EasyEngineering.net

158 Operation and Control in Power Systems

and ..... ( 4.16)

It is to be noted that the set of active constraints is not known in advance. The variables
u, are the usual Lagrange - multiplies used in case of equality constraints and PI are the
Kuhn-Tucker mUltipliers. Together, they are called dual variables.
In 1962 Carpentier formulated the general problem of minimizing the instantaneous
operating costs of a power system subject to both, the network constraints and inequality
constraints applying Kuhn - Tucker conditions. The variables of vector K contain components
of independent variables P (active power), Q (reactive power), T (in-phase transformer tap

ww
position) TS (phase shift transformer tap position) together with dependent variables voltage
magnitude IVI and bus voltage phase angle D.

w.E x = [P,Q,T,TS, V,DJ ..... ( 4.17)

4.4
asy
Lagrange Function for Optimal Load Flow
The general Lagrange function with all the constraints incorporated is written as

L= Ef(K)
+ Ai' hP (K ) + IlQ hQ (X)
ngi
(Scalar cost function for optimal evaluation.)

+ pP gP(K) nee
(equality constraints for load balance)

(reactive power constraints)

+ pRgR (x) rin
(generator rating constrains)
+

+
pBgB(K)
pRSgRS~)
(boiler constraints, if any) g. net
(running reserve constraints)
+ pBRgBR~) (branch flow constraints)
+ pVgV~) (bus voltage constraints)

+ pT gT~) (in phase tap range constraints)

+ pTSgTS~) (phase shift transformer constraints)
+ pSgS~) (line outage or security constraints) ..... (4.18)
Note that the constraints h are of the type
h(K) = PgI - P'I - PI =0 .... (4.19)

Q gI -Q'I -QI =0 at all buses i = 1,2, ..... n ..... (4.20)

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 159

\
while constraints g. for instance gv is of the type \
\

V,m," ::s; V, ::s; V,max at any bus i

4.5 Computational Procedures

Solution to a set of non linear equations of the form

oL _ 0 ._
::l - for I - I, 2, ..... n ..... (4.21 )
('X,

ww
is very difficult to obtain in practice. However, there are several techniques, which are iterative
in nature. that can be applied to find the optimal solution around the operating point. There are

w.E
sequential unconstrained minimization techniques which transform the constrained problem
into an unconstrained problem. Penalty function methods are important in this category. In
the next section a method that utilizes the jacobian of the Newton - Raphson method will be

asy
detailed. In this method neither the penalty functions are used nor the Kuhn - Tucker variables.
No effort will be made to limit the controllable variables initially.

4.6 E
Conditions for Optimal Load Flow
ngi
n
n
Taking into consideration only the equality constraints, the Lagrange function becomes

L Aqgq (x) +
L eer n
L = F(x) +
q=!
in
q=!
Il qg q (x)

g.n
..... ( 4.22)

At the optimum
et ..... (4.23)

and ..... ( 4.24)

considering the voltage phase angle and its magnitude only as variables
Multiplying equations (4.23) by d8 p and equation (4.24) by dlV pi

::IF
::lL
_u d8 p =
08
~d8p
08 p
+" n
L-
q=! 08 p
og
Aq - q oOp +"
n
L-
q=!
oil
r"q - s:
oUp
q o8 p ..... ( 4.25)

and

..... (4.26)

Downloaded From : www.EasyEngineering.net

160 Operation and Control in Power Systems

Combining the two equations

n OPgq n oP1q n oPq
dL=dF+ ~).q - d D p -IAq - d 8 p - IAq - d 8 p
q~1 oP gP q~1 oP gP q~1 oPgP

~ aQ gq ~ aQlq ~ oOq
+ + ..:::...Ilq - - d 8 p - ..:::...Ilq --d8 p - ..:::...Ilq --d8 p ..... (4.27)
q~1 oP gP q~1 oPgP q~1 oP gP

" OPgq n oP1q " O Pq

q~1 q 0 I VP I I P I ~ q 0 I VP I I P I ~ Aq -
+ I A --d V - A --d V - -d V

ww 0 I VP I I p I

w.E
asy
The load powers are generally assumed to remain constant

Hence,
En oP1q aQ Iq
-=--=--=--=0
ODp o8 p
oP1q aQ Iq
olVpl clVpl
..... (4.28)

gi
Also, the total derivatives for Pgq and Qgq are given by
nee
~ ~ oP gq ~ oP gq
ar
..:::...AqdP gq = ..:::...Aq --d8P + ..:::...Aq - - d I Vp I
q~1
n
q~1
-"
o8P
aQ
q~1
n ing
I Vp I
aQ .....(4.29)
andLllqdO gq =
q~1
Lll
q=1
q ~d8P
ODP
+ Ill
q=1 aq

Hence equation (4.27) reduces to the following with the substitution of (4.28)
. gq
I Vp I
d I

net
Vp I

and (4.29)

..... (4.29)

The incremental production costs are given by

dF
- - = Yq = a q + bqPgq for all q = 1,2, ..... ng ..... {4.31)
dPgq
where ng = number of generator buses.
d F= - Yq dP gq = - (aq + bq Pgq) d Pgq

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 161

for all dF to have a minimum. Substituting this in equation (4.30) for dF

..... (4.32)

For the above equations to be zero, since, Pgq' Qgq' op and IV pi are finite changes their

ww
coefficients have to be zero.
Hence Yq = Aq ..... (4.33)

w.E 11q = 0 ..... (4.34)

asy
at all generating buses

En oPq
Further
gi oop
oPq
81Vp l nee
q = I, 2 .... n P = I, 2, ..... n rin
g.n
..... (4.35)
The above equations are the conditions for optimal load flow.

4.7 Implementation of optimal conditions

Method 0/ El-Abiad and Jaimes
et
It can be recognized that the coefficient matrix of equation (4.35) is transpose of the Jacobian
ofN-R method. Partitioning of this matrix is proposed to solve for the optimal solution.

Algorithm:
1. Any bus is assumed as reference bus I. The incremental production cost at this bus
from any feasible load flow solution is computed and set equal to AI.
IPC I = YI = AI
2. The coefficient matrix (4.35) is expanded and is shown in equation (4.36). Separating
out the first column it has the order reduced as follows; to enable the computation of
all remaining Als and Ills.

Downloaded From : www.EasyEngineering.net

162 Operation and Control in Power Systems

aPI aP2 apng aPno +1 oPn 00 1 OOng OOng+1 OO n

--"'- AI
alii ali 2 alii iJo l 0°1 iJo l iJo l iJo l 0° 1

oPI aP2 oPng aPng +1 oPn 001 OOng OOng+1 OO n

OOng OOng OOng OOng DOng iJo ng DOng aO ng aO ng Ang

oP2 oPpg oPng +1

~ oPn 001 cOng OOng+1 OO n
aOn olin oOn oOn oOn cOn olin aO n 0O n An

wwoPI
olvd
OP2
alv21
oPng
aivil
oPng +1

lYJ
oPn
clv~l
00 1
civIl
OOng
alv,1
aO ng +1

(11 VI I
aQn
chi
•
~ll
=0

aPI aP2w.E oPng oPng +1 aPn 001 ilO ng OOng+1 OO n

Il ng
alvngl olVngl
asy a/vngl

apng
olvngl alvngl alvngl alvngl illvngl alvngl

En
aPI OP2 aPng +1 OPn cOl OOng OOng+1 OO n
.. Il n
~Vnl alvnl alvnl alvnl alvnl alvnl alvnl lYJ alvnl J

gi nee . .... (4.36)

ap2 aPn CO ng +1
r ing
0° 2
aPn
0° 2
aPn
0° 2
CO ng +1
CO"
DOl

OOn
Al

aPI
.
?PI
2° 2 net
aO n aO n DOn aO n An
aO n
ap2 = (-AI ..... ( 4.37)
oPn OOng+1 OOn ~ng+1 oPI
alVng+11 alvng+d alvng+,1 alVng+11 81 Vng+11

ap2 aPn OOng+1 OOn ~n oPI

alvnl alVnl alvnl alvnl 01 Vn 1

Knowing the right hand side vector the Als and ~I s are computed for all on the left hand side.
The real power generations are computed lIsing these A values for y, the incremental
production cost using the relations.
Yq = k Aq; q = 1,2, ..... ng ..... (4.38)

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 163

ng
" P
[ ~ uq
q=i'
+" __batL_1
q=i
ng
~
pq
and K == --'-------;-----::--'--'-'- ..... (4.39)
I[~l
b
q=i q

An improved generation schedule is obtained from

Yq == aq + b q P gq .... (4.40)

ww
At the optimum k becomes unity.
From the voltage magnitude derivatives the voltage magnitudes are computed:

oP
-,-,II.I
i 1oP2
avp +-,-,11.2
oVp w.E 1 eP"
+...... +-;1'
eVp-'lI. +-;1-,
1 ?On
n (V p-,l1n +1 +...... +-;1,-,l1n
OO"g+i
eVp 0g ==

a
P == I, 2, ..... ng
syE
The above equations do not appear in (4.37) as they are partitioned out from equn.( 4.36).
..... (4.41)

Separating the diagonal self terms for bus P.

ngi
nee ..... (4.42)

rin
The influence of () on reactive power can be neglected while scheduling real power.
Voltages IV pi are calculated during every iteration, until convergence is obtained.
g.n
Computation of real power generation and voltages is continued till the solution is obtained.
Both Q and IVI are monitored at every iteration and whichever exceeds the limit is set at
that value and the node is treated as P, Q bus and that voltage equation drops from Eqn.(4.42).
If both are violated, load flows are performed till they lie within the limits.
et
E.4.1 For the network shown in Fig. E4.1 the following data is given:
Incremental cost curves
Generating station I
dF
_I == (24 + 8P I )Rs./ p.u.MWh
dFI

Generating station 2

dF)
- - == (19.6 + 8P2 )Rs. / p.u.MWh
dF2

Downloaded From : www.EasyEngineering.net

164 Qperation and Control in Power Systems

4 3

ww 6

w.E
asy Fig. E4.1

En
The restrictions on real powers at the stations are
0.1
0.1 gi~ PI ~
~ P2 ~
1.0 p.u. MW
1.0 p.u. MW
nee
The load powers are given in Table E4.1.

rin
Bus
Table E4.1
Real Power (p.u. MW) Reactive Power (p.u. MW) g.n
3
4
5
0.55
0.0
0.50
0.05
0.0
0.05
et
6 0.30 0.Q3

Find the most economical generation at stations I and 2. Find also the annual savings,
if both the plants operate for 7000h in an year.
Solution:
Let bus 1 be the slack bus. Its voltage is assumed (I + jO.O). Let the generation at bus 2 be
arbitrarily assigned as
P = 0.80 p.u. MW and Q = 0.05 p.u. MW
Performing load flow computation following the procedure given in chapter 2, the
following results are obtained.

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 165

Table E4.2 Bus Powers

Bus Real Power (p.u. MW) Reactive Power (p.u. MW)

1. 0.70618 (Slack bus) 0.27735 (slack bus)
2. 0.80 0.05
3. 0.55 0.05
4. 0.00 0.0
5. 0.50 0.05
6. 030 0.03
Losses 0.15618 0.19735

ww The voltages are given in Table E4.2.

w.E Bus Symbol

TableE4.3
Voltage p.u.

asy I. VI
V2
1 +jO.O
1.04454+ jO.1891O

En
2.
3. V3 0.95948 + jO. 17580
4.
5.
6.
gi V4
V~
Vb
0.90461 +jO.I2718

nee
0.88958 + jO.15143
0.89768 + jO.11 0 1

The production costs at plant I are : r ing

(24 + 8 x 0.70618) Rs.lp.u. MW
The production cost is Rs.5564.944h-
1 . net
The load flow solution is optimized using the procedure indicated in Sec. 4.7. The
following results are obtained.
PI = 0.80367 p.u. MW
P2 == 0.65662 p.u. MW
Total real power generation == 1.46029 p.u. MW
Q1 == 0.21841 p.u. MW
Q2 == 0.21841 p.u. MW
Total reactive power generation == 0.24741 p.u. MW

Downloaded From : www.EasyEngineering.net

166 Operation and Control in Power Systems

Table E4.4

Iteration Al Az A3 A4 As A6
1. 3.706 2.679 4.037 4.000 4.000 3.897
2. 3.981 3.546 4.444 4.414 4.494 4.444
3. 3.808 2.918 4.193 4.158 4.186 4.095
4. 3.850 3.099 4.244 4.213 4.252 4.173
5. 3.845 3.099 4.4232 4.201 4.239 4.160
6. 3.835 3.084 4.212 4.181 4.215 4.136
7. 3.833 3.094 4.202 4.172 4.204 4.127
8. 3.828 3.096 4.191 4.161 4.192 4.115

ww 9.
10.
II.
3.824
3.821
3.818
3098
3.100
3.101
4.180
4.171
4.163
4.152
4.143
4.128
4.180
4.170
4.162
4.105
4.095
4.087

w.E
12.
13.
3.815
3.812
3.103
3.104
4.156
4.150
4.135
4.122
4.154
4.146
4.080
4.073

asy
14. 3.810 3.106 4.144 4.116 4.140 4.067
15. 3.808 3.107 4.138 4.111 4.134 4.061
16. 3.804 3.108 4.133 4.106 4.128 4.056

En
Iteration
I.
gi J.l 3
0.0178
Table E4.S
J.l 4
0.0430
neeJ.l s
0.1153
J.l 6
0.1433

r
2. 0.0315 0.0547 0.1189 0.1235
3.
4.
0.0207
0.0202
0.0501
0.0473
0.1244
0.1162 ing 0.1404
0.1292
5.
6.
7.
0.1783
0.0150
0.0126
0.0447
0.0417
0.0390
0.1125
0.1086
0.1047
.
0.1260
0.1230
0.1193
net
8. 0.0104 0.0366 0.1013 0.1163
9.. 0.0084 0.0345 0.0983 0.1137
10. 0.0068 0.0326 0.0956 0.1114
II. 0.0053 0.D31O 0.0933 0.1093
12. 0.0040 0.0295 0.0912 0.1075
13. 0.0027 0.0282 0.0896 0.1058
14. 0.0016 0.0270 0.0876 0.1043
15. 0.0006 0.0260 0.0861 0.1030
16. -0.0003 0.0249 0.0846 0.1017

Note: III = 112 = 0.0

Total real power losses = 0.11029 p. u. MW
Total reactive power losses = 0.24741 p.u. MW
I

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 167

The voltages at the buses are

YI = 1.04676 + jO.O p.lI.

Y2 = 1.10437 + jO.0774 p.lI.

Y 3 = 1.02699 - jO.19679 p.lI.
Y 4 = 0.95996 - jO.1431 0 p.lI.
Y5 = 0.94362 - jO.17032 p.u.
Y 6 = 0.960066 - jO.15036 p.u.
1
The total production costs are Rs.5528.232h-
I

ww
The savings per hour are Rs. 5564.944 - 5528-232 = Rs.36.71 h-
Annual savings with 7000 working hOllrs is

w.E
The variation of
7000 36.71
Lagrall~\?
= Rs. 258.370
Illultipliers during iterations is shown in Tables E4.4
and E4.5.
asy
En
gi nee
r ing
. net

Downloaded From : www.EasyEngineering.net

168 Operation and Control in Power Systems

Questions

4.1 ~ring out the salient features in optimal operation of generators in thermal and
hydro plants.
4.2 Discuss the various constraints to be considered for economic load dispatch
problem.
4.3 What is an optimal power flow? Explain the problem in detail.
4.4 What are Kuhn - Tucker conditions? How are they employed
4.5 Explain how Kuhn - Tucker conditions are useful in optimal load flow.
4.6 Explain the non - linear programming method for optimal power flow

ww 4.7
4.8
Explain reactive power control loss minimization
Explain gradient method for optimal power flow.

w.E
asy
E ngi
nee
rin
g.n
et

Downloaded From : www.EasyEngineering.net

Optimal Load Flow 169

Problems
P4.1 Consider the Six bus system shown in Fig. P4.1.

2 3

ww
5

w.E 6

asy Fig. P4.1

En
The bus admittance matrix is given as follows

14.45 - j32,50
0,0
0,0
0,0
7,33 - j28,25
-1,0+j5,50
gi 0,0
-1,0+ 5.50
1,0-j5,50
- 5,48 + j8. \0
0,0
0,0
nee
- 9,0 + j24,40
-5,05+ j16.30
0,0
0,0
- 1,25 + j6,44
0,0
-5,48+ j8,10 0,0 0,0 10,D3 - j22,70
rin
- 4,55 + j 14,60 0,0
- 9,0 + j24,60

g.n
-5,05+ j16.30 0,0 - 4,55 + j 14,60 18,6- j55,30 0,0
0,0 - 1,28 + j6.44 0,0 0,0 0,0 1.28 - j6,44

The incremental cost characteristics are given by et

dC, =3.85+P
dP , ,

dC 2 = 3.25 + P,
dP2 -

dC 3 = 2.75 + P,
dP3 .

The constraints are

0.1 .:::: P, .:::: 1.0
0.15.:::: P2 .:::: 1.5

Downloaded From : www.EasyEngineering.net

170 Operation and Control in Power Systems

0.15 .::: P3 .::: 1.5

0.99.::: IV,I .::: 1.15
0.99.::: iV21 .::: 1.15
0.99.::: IV31 .::: 1.15
P4 = 1.4 p.u MW; 0 4 = +0.55 p.u. MW
Ps = 0.8 p.u MW; OS =0.033 p.u. MW
P6 = 0.7 p.u MW; 0 6 = 0.016 p.u. MW
Obtain the optimal load flow

ww An answer is provided for reference for fixed IV". However,

P, = 0.64644 p.u.
IV" also can be optimized.

w.E
P2 = 1.07780 p.u.
P3 = 1.30267 p.u.
0, = - 0.05844

asy
O2 = - 0.43422
0 3 = 0.2351

En
V, = 1.5 + jO.O
V2 = 1.08667 + jO.67867

gi
V3 = 1.10884 + jO.28434
V4 = 1.02292 - jO.04 75
V s = 1.04923 - jO.00235 nee
V6 = 1.06258 - jO.O 1875
The Lagrange multipliers rin
A, = 4.49644
g.n
1..2 = 4.327741
1..3 = 4.052448
1..4 = 4.724052
et
As = 4.536256
1..6 = 4.489106
Il, = 112 = 113 = 0
114 = 0.018764
Ils= 0.009117
1l6= 0.01 1654

Downloaded From : www.EasyEngineering.net

ww
5 UNIT
w.COMMITMENT
Ea
syE
ngi
The life style of a modern man follows regular habits and hence the present society also
follows regularly repeated cycles or pattern in daily life. Therefore, the consumption of electrical

nee
energy also follows a predictable daily, weekly and seasonal pattern. There are periods of high
power consumption as well as low power consumption. It is therefore possible to commit the

rin
generating units from the available capacity into service to meet the demand. The previous
discussions all deal with the computational aspects for allocating load to a plant in the most

g
economical manner. For a given combination of plants the determination of optimal combination

.ne
of plants for operation at anyone time is also desired for carrying out the aforesaid task. The
plant commitment and unit ordering schedules extend the period of optimization from a few
minutes to several hours. From daily schedules weekly patterns can be developed. Likewise,
monthly, seasonal and annual schedules can be prepared taking into consideration the repetitive
nature of the load demand and seasonal variations. Unit commitment schedules are thus required
t
for economically committing the units in plants to service with the time at which individual
units should be taken out from or returned to service.

5.1 Cost Function Formulation

Let F 1 be the cost of operating the ith unit when supplying an output of PI. Then, if C I is the
running cost coefficient
FI == C I PI
where C j may vary depending on the loading condition.

Downloaded From : www.EasyEngineering.net

172 Operation and Control in Power Systems

We may thus define a variable cost coefficient CIJ for ith unit when operating at jth load
for which the corresponding active power is PII. Since the level of operation is a function of
time, the cost coefficient may be described with yet another index to denote the time of
operation, so that it becomes CII ! for the sub interval t corresponding to a power output of PIll"
If each unit is capable of operation at k discrete levels, then the running cost FII of unit
i in the time interval t is given by
~

FII = LCIIIPIiI
1=1
If there are N units available for service in the interval t then the total running cost of

ww
N units during the period t is
N •

w.E FINI = L L C III Pili

1=1 J=1

asy
For the entire time period of optimization, having T such sub-intervals of time, the
overall running cost for all the units may be put in the form

En FT =
TN.
L L L C 111 Pili
gi I~I I~I J=I

nee
Suppose that for a plant to he brought into service, an additional expenditure CSI has to
be incurred in addition to the running cost, then the start up cost of the unit i, the cost of

r
starting X units during any subinterval t is given by

Fsc ::=
x
L C 'CI<\ ing
where 011 =
I~I

.
I if the ith unit is started in sub-interval t and is zero other wise
In a similar manner, if a plant is taken out of service during scheduling period, the cost
net
of shutting down also must be taken into account. If Y units are to be shutdown during the
sub-interval t, the shutdown cost may be represented by
y

Fsd = L C Sdl • 0"11

1=1
where 0"11 = I when unit i is taken from service in sub-interval t and is zero otherwise.
The start up and shut down costs for X and Y units over the complete scheduling period
containing T sub intervals are given respectively by
T x
Fscf = L L CSCIO'I
1=1 1=1

Downloaded From : www.EasyEngineering.net

t}nit Commitment 173

T Y
and FsdT = L L Csdlcrll
1=1 1=1

The total expression for the cost function including funning, start up and shut down
costs will then have the form,

..... (5.1 )

For each sub-interval of time t, the number of generating units to be committed to

ww
service, the generators to be shut-down and the quantized power loading levels that minimize
the total cost have to be determined.

5.2
w
Constraints for Plant Commitment Schedules

.Ea
As in the optimal point generation scheduling the output of each generator must be within the
minimum and maximum value of loading
pmm
I
<
- syE
p <
1)1-
p max
I
..... (5.2)
where i = 1,2, .... N
ngi
j = 1, 2, ..... k
t = 1,2, ..... T
nee
The total available plant capacity from which schedules are to be prepared must be

r
in excess of the plant generating capacity required in meeting the predicted load demand and

ing
in satisfying the requirements for minimum running reserve capacity during the entire period
of scheduling.

STAC -> '"

N
L.. pmaxa
I
1=1
II
+ Smm
r
. net
..... (5.3 )

where STAC is the total available capacity in any sub-interval t and Srmm is the minimum
running reserve capacity.
all = 1, if generator i is in operation during the sub-interval t and is zero otherwise.
In addition, for a predicted total load demand PD' the total generation operated in sub-interval
t must be in excess of the load demand by an amount not less than the minimum running
reserve capacity Srmm.
N k
LL Pijl ~ PD +S~m ..... (5.4)
1=1 )=1

Where the transmission losses PL are also considered eqn.(5.4) then becomes
N k
L L--PI) I ~ PD + S~m + PL ...... (5.5)
1=1 )=1

Downloaded From : www.EasyEngineering.net

174 Operation and Control in Power Systems

The generator start-up and shut down logic indicators bit and a'l respectively must be
unity during the corresponding sub-intervals of operation, or
a'l - a'(l _ I) = b'l - a'l
i = 1. ..... N
t = I, ..... T

5.3 Priority - List Method

In this method full load average production cost is computed for every unit. Then in the order
of ascending costs, the units are arranged for commitment. The method will be explained

ww
through an example. Consider three thermal plants operating with the following incremental
fuel rate characteristics.
Plant J :
w.E
FI = (7 + 0.003 PI) 10 3 k-cal/Mw-hr.

asy
Plmax = 500 MW and p,mm = 50MW
Plant 2 :

E
F2 = (7.5 + 0.002 P2 ) 103 k-cal/Mw-hr
P2max = 400 MW and Pdnm = 40 MW ngi
Plant 3 :
F3 = (8 + 0.004 P3) 103 k-caI/Mw-hr
nee
P3max = 200 MW and P3mm = 20 MW
rin
The fuel costs at the plants are given by
CP I = I. I Rs/kcal
g. net
CP 2 = 1.05 Rs/kcal
CP 3 = 1.2 Rs/kcal
The full load average production cost at plant I
= 103 (7 + 0.003 x 500) 1.1 = 10 3 (7 + 1.5) 1.1
= 9.35 x 10 3 Rs/MwHr
The full load average production cost at plant 2
= 10 3 (7.5 + 0.002 x 400) 1.05
= 103 (7+0.8) 1.05 = 8.715 x 10 3 Rs/MwHr

The full load average production cost at plant 3

= 103 (8 + 0.004 x 200) 1.2 = 10 3 (8 + 0.8) 1.2
= 10 3 (8.8) = 10.56 x 10 3 Rs/MwHr

Downloaded From : www.EasyEngineering.net

Unit Commitment 175

In the order of increasing costs the following table is constructed.

--- -
Unit RslMwHr Max(MW) \lin(MW)
2 8.715 x IQl 400 40
I 9.35 X 10\ 500 50
3 ,10.56 x IQl 200 20

The priority list for supplying a load upto 1000 Mw is prepared as follows which is :.clf
explanatory.
Units Max(MW) Min(MW)

ww 2.1 and 3
2,1
1000
900
110
90

w.E 2 400 40

Table 5.1 Priority list for supply of IOOMW

asy
The general procedure can be then staled as .

En
I. Check at the end of every hour of operation. if the load demand has fallen. If the
demand has decreased check if the last unit in the priority list is dropped, the load

gi
demand can be met, satisfying the spinning reserve requirement. Status quo is maintained
if the demand cannot be met.
nee
2. If it is possible to drop the unit in step I, then determine the number of hours '"h" before

r
the unit is required again for service. If this "h" is less than the shut down and start up
times for the unit, it has to be left in service without removal.
ing
.
3. Then, calculate the cost of floating the unit within the system without supplying any
generation and the cost of shut down and start up processes and if there is sufficient
savings from shutting it down, and starting it again for service it can be removed. net
4. The process is to be n:peated for th~ next unit on the priority list and continued.

5.4 Dynamic Programming

This method can be applied to problems in which many sequential decisions are required to be
taken in defining the optimum operation of a system or process composed of a distinct number
of stages. However, it is suitable only when the decisions at the later stages do not affect the
operation at the earlier stages. Dynamic programming is based on the principle of optimality
enunciated by Bellman in 1957. It states that "an optimal policy has the property, that, whatever
the initial state and the initial decisions are, the remaining decisions must constitute an optimal
policy with regard to the state resulting from the first decision".
Hence, the method is suitable only to multistage decision processes where a decision taken
at one stage does not influence the preceding events in the sequence but affect only what
tVllows later. The principle and methodology will be explained through a routing problem.

Downloaded From : www.EasyEngineering.net

176 Operation and Control in Power Systems

Consider the route map shown in Fig. 5.1. The values on the routes are the cost of transport
along the route. Let G be the goal to be reached from any of the locations A, 8, C, D, E, F.
Traffic is allowed only along the direction shown on the routes.
E l F
~

p ;/
A

~w
ww
H

"
,/
G

.Ea B 3 2 D

syE
Fig. 5.1 The Routing Problem
It is desired to reach the goal G or destination G from the various locations. The costs
are as indicated.
From D to reach G ngi
CnG = 3 = C~G nee
From H to reach G via D
rin
C~G :::: C HD + C~(j :::: I + 3:::: 4
From C to reach G via D g.n
In a similar manner
C~G =Ceo + C~G =2 + 3 =5 et
C~G = Min[CFG,C FH +C~G]:::: Min[6,1+4]::::5
C~G :::: [C EF + C~G' C EC + C~G]:::: [I + 5,4+ 5] =6 along EFHDG

C~G = [CBC + C~G ] = [3 + 5] :::: 8 along 8 C D G

C~G == Min(C AB + C~G,CAE + C~G)== Min[2 + 8,1 + 6] =7 along A E F H D G

. We can thus conclude that the minimum cost to move from one location P to another
location R via Q is given by
CPR:::: C pQ + C~R

where CpO is the cost to move from P to Q and C~)R is the optimal cost to move from Q
to R.

Downloaded From : www.EasyEngineering.net

Unit Commitment 177

The optimal decision at Q is taken based on

C~R := Min[C Q1R , CQ2R , ......... , CQNR ]

where QIR, Q2R ..... QNR are the n alternative paths available to reach R from Q.

5.5 Unit Commitment by Dynamic Programming

In general load changes in small steps. For the complete load cycle, it is possible to apply
dynamic programming method for unit commitment. For this, the number of units available,
the cost characteristics of these units and the load cycle are the data required. Assuming that
this information is made available, the functional recurrence relations of the method will be

ww
applied now
N
L P"
w.E
Let N units in service supply a total load of

(~PI
1=1

asy
The total operating cost for this generation = FN )

En
This cost can be divided into two components as follows:
(i)
gin
The first component is the cost of operating (N-I) Llnits to supply a reduced

demand of N-I
eer
~ P" The operating cost for this generation is FN_1 ~ P,
(N )

(ii)
output of PN is fN(P N)' in
The second component is the cost of operating the Nth unit at an active power

g.n
et
N N-I
The cost of supplying the total load L P, = P + L P,N is given now by
1=1 1=1

Hence

subject to constraints.
Thus, the problem of unit commitment is converted into a multistage decision process.
To start with the solution to the problem, consider the single stage decision process of selecting
one unit that will provide the lowest cost of generation up to the maximum capacity of the unit.
Once such a unit is picked up, the first stage of optimization is completed.

Downloaded From : www.EasyEngineering.net

178 Operation and Control in Power Systems

Let the cost of operating this first unit be fl(P I) when supplying PI
FI(P I) = fl(P I)
The optimal combination of this unit with a second unit can be derived next to supply a
load ofP I+P 2 for which the optimal cost is
FZ<P I + P2) = min [fiP2) + FI(P I)]
Since FI(P I) is already the optimal value, the best unit to supply P2 from the rest of
(N-l) units will be picked up during this second stage of optimization for lowest possible cost
of production.
In a similar way, the best 3 unit combination is found from

ww FiPI + P2 + P3) = min [f3(P 3) + F2 (PI + P2)]

Subject to constraints and so on, so that finally we arrive at
w.EFN(~Pl) FN_{~lpl)= minfN(PN)+

asy
which is the required solution.

En
gin
eer
ing
. net

Downloaded From : www.EasyEngineering.net

Unit Commitment 179

E 5.1 Obtain the economic schedule for the two units, the production costs of which are
given as follows, to supply a load of 3MW, in steps of 1MW
CI = 0.8 p~ + 25 PI

C 2 = 1.2 p~ + 22 P2
Use dynamic programming method.
Solution: With usual notation
F I(3) = fP) = 0.8 x 32 + 25 x 3 = 7.2 + 75 = 82.2
F2(3) = min [fiO) + fP), f 2(1) + FP), Fi2) + FI(I), F2(3) + FI(O)]

ww = min [0 + 82.2, 1.2 + 22 + 3.2 + 50,4.8 + 44 + 0.8 + 25, 10.8 + 66 + 0]

= min [82.2, 76.4, 75.6, 76.8] = 74.6

w.E
The most economical combination is unit 1 supplying I MW and unit 2 supplying 2 MW.

asy
E 5.2 If only a load of 2MW is to be supplied, how are the units to be committed in E5. I.
Solution:
F I(2)
En
= f l (2) = 0.8 x 22 + 25 x 2 = 53.2
Min [ fiO) + F I(2), fiO + FI(I), fi2) + FI(O)]
F2(2)

gi
=
= min [ 53.2, 1.2 + 22 + 0.8 + 25, 1.2 x 4 + 22 x 2 + 0]
= min [ 53.2, 49, 48.8]
nee
F2(2) = fi2) + FI(O). Unit 2 supplies the total load for economy

r
From the above two examples unit commitment schedule may be now prepared based
on the two results. ing
Load range Unit in service
. net
IMW 2
2 MW 2
3MW 2 and 3

For more accuracy the step size can be reduced to 0.5 MW, 0.2Mw or even 0.1 Mw.
The smaller the step, size, the larger the calculations needed.

Downloaded From : www.EasyEngineering.net

180 Operation and Control in Power Systems

Questions

5.1 Explain the unit commitment problem.

5.2 What are the various constraints in unit commitment problem? List them.
5.3 Explain the priority - list method for unit commitment.
5.4 Discuss the dynamic programming method to solve unit commitment problem in a
power system.
5.5 What is the cost function for unit commitment? Explain

ww
w.E Problems

asy
P S.l Write down the recurrence functional relation for scheduling generation among 5
units to meet a load demand of IOMW in steps of 2MW.

E
PS.2 Use dynamic programming method to determine the most economical units to be

ngi
committed to supply a load of 6MW. There are three units with the following data

C 1 = O.8 P1 2 + 22P1
C2 = O.85Pi + 21P2 nee
C3 = O. 8P32 + 20P3
rin
g.n
The maximum and minimum capacities of each unit are 5MW and I MW respectively.

et
PS.3 Use dynamic programming method to determine the most economical units to be
committed to supply a load of 9MW. There are four units with the following data.

The maximum and minimum limits for each unit are 6 MW and I MW respectively.

Downloaded From : www.EasyEngineering.net

Unit Commitment 181

PS.4 Given the input output cost curves

Fl (PI) = 21 + 6P1 + 0.00 I2P12 Rs/kcal

F2 (P2 )=2I+7P2 +O.OOIPi Rs/kcal

FJ (PJ ) = 17 + 8PJ + 0.00 II pJ2 Rs/kcal

The maximum and minimum values are
30:s P1:S 300 MW
40:s P2:S 400 MW

ww 25:s P3:S 250 MW

Obtain unit commitment schedule for a load demand of 600MW.

w.E
asy
E ngi
nee
rin
g.n
et

Downloaded From : www.EasyEngineering.net

ww
w.E
asyE CONTROL
6 LOAD FREQUENCY
ngi
nee
In a power system the load demand is continuously changing. In accordance with it the power

rin
input has also to vary. If the input - output balance is not maintained a change in frequency
will occur. The control offrequency is achieved primarily through speed governor mechanism

g.n
aided by supplementary means for precise control. At the outset, the speed governor mechanisms

et
and its operation will be presented.

Governor: The power system is basically dependent upon the synchronous generator
and its satisfactory performance. The important control loops in the system are :
(i) frequency control, and
(ii) automatic voltage co"trol.
In this chapter the frequency control will be discussed. Frequency control is achieved
through generator control mechanism. The governing systems for thermal and hydro generating
plants are different in nature since, the inertia of water that flows into the turbine presents
additional constrains which are not present with steam flow in a thermal plant. However, the
basic principle is still the same; i.e. the speed of the shaft is sensed and compared with a
reference, and the feed back signal is utilized to increase or decrease the power generated by
controlling the inlet valve to turbine of steam or water as the case may be.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 183

6.1 Speed Governing Mechanism

The speed governing mechanism includes the following parts.
Speed Governor: It is an error sensing d'evice in load frequency control. It includes all
the elements that are directly responsive to speed and influence other elements of the
system to initiate action.
Governor Controlled Valves: They control the input to the turbine and are actuated by
the speed control mechanism.
Speed Control Mechanism: It includes all equipment such as levers and linkages,
servomotors, amplifying devices and relays that are placed between the speed governor
and the governor controlled valves.
Speed Changer: It enables the speed governor system to adjust the speed of the turbo

6.2 ww
generator unit while in operation.
Speed Governor

w.E
A simple schematic representation of the governor is shown in Fig. 6.1.

a syE t XB
txc
n
--.J gi
B c

J nee
rin
g.n
et
Fig. 6.1 Speed Governor
The pilot valve v operates to increase or decrease the opening of the steam inlet valve V.
Let XB and Xc be the changes in the position of the pilot valve v and control valve V
responding to a change in governor position. XA due to load.
When the pilot valve is closed XB= 0 and Xc == 0, (Le.,) the control valve is not
completely closed, as the unit has to supply its no-load losses. Let roo be the no-load angular
speed of the 9-'rbine. As load is applied, the speed falls and through the linkages the governor
operates to move the piston P downwards along with points A and B. The pilot valve v admits
oil under n and lifts it up so that the input is increased and speed raised. If the link Be is
removed then the pilot valve comes to rest only when the speed returns to its original value. An
"isochronous" characteristic will be obtained with such an arrangement where speed is restored
to its preload-disturbance value. This is shown in Fig. 6.2 and Fig. 6.3.

Downloaded From : www.EasyEngineering.net

184 Operation and Control in/Power Systems

8 c

ww
w.E
asy
E ngi
Fig. 6.2 Isochronous Governor

nee
IOO%t-==============(A) Isochronous

Speed
t I
Drooping rin (8)

o
g. 50% 100%
net
Power
Fig. 6.3 Governor Characteristics

With the link Be, the steady state is reached at a speed slightly lower than the no load
speed giving a drooping characteristic for the governor system. A finite. value of the steady-
state speed regulation is obtained with this arrangement. For a given speed changer position,
the per unit steady state speed regulation is defined by
N -N
Steady state speed regulation = ~
r

Where No = Speed at no - load

N r = Rated speed
N = Speed at rated load
The isochronous and drooping characteristics are shown in Fig. 6.3.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 185

6.3 Steady State Speed Regulation

The opening of the control valve is directly'related to the power input to the unit so that
P
-=-
Xc
..... (6.1 )
P, C
Where C is the opening of the control valve corresponding to rated output Pr and Xc is
the opening while the turbine delivers an output ofP. The displacement at A, xA is proportional
to the change in speed so that
xA = K (No - N)
=K~N ..... (6.2)

ww
Defining the lever ratios

w.E AB _I
BC - , ..... (6.3)

asy BC _I
AC - 2 ..... (6.4)

and En
AB _I
AC - 3 ..... (6.5)

-, Xc
gi
~_ AB_I
-
BC nee ..... (6.6)

so that XA = I, XC
P
= II -
P,
.C r ing
~N=~=!i.~.C= CII.~
k k P, k P,
. net
..... (6.7)

The steady state speed regulation

..... (6.8)

thus, the steady state speed regulation is directly proportional to the output power.
R has the dimension of HzlMW.

6.4 Adjustment of Governor Characteristics

For a given governor R can be changed by varying the lever ratio II and the no load speed No
The droop of the governor characteristics can be changed from curve B to B I as in Fig. 6.3 by
changing the lever ratio II' Generally in order to coordinate a generating unit with the rest of the
system, the lever ratio II is adjusted. This is done occasionally and when the machine is cold.

Downloaded From : www.EasyEngineering.net

186 Operation and Control in Power Systems

When the machine is in operation the no load speed No can be adjusted by varying an
external force acting on the governor thrust sleeve at S so that the characteristic is shifted
parallel to itself as shown by curves C and D. This is achieved by the speed changer.
See Fig. 6.4.

--- --- --- ---

I
Speed --- --- --- ---
C
8
--- --- --- 81
---
ww D

w.E Power

Fig. 6.4 Action of speed changer

6.5 asy
Transfer Function of Speed Control Mechanism
From Eqn. (6.2).
E
xA = K (N o- N)
ngi
= K I (roo - ro)

If the point C is fixed, then

= k ! flro
nee ..... (6.9)

rin
g.n ..... (6.10)

A 8 c
et
..... (6.11)

Fig. 6.5
If C were to move with A fixed then from Fig. 6.6.

~_ AS_I
Xc - AC - 3

Xs = 13 , Xc Z

Downloaded From : www.EasyEngineering.net

Load Frequency Control 187

A B c
Fig. 6.6

ww
As the point C will be shifted upwards by Xc due to the load increment, the actual
movement of B would be
.,
w.E xB=12xA-13xc ..... (6.12)
The opening of the pilot valve, xBdetermines the rate of oil flow into the cylinder. The

asy
movement of the control valve, Xc is proportional to the total oil admitted under the piston P.
Therefore,
Xc = k fXB ; En where k is a constant ..... (6.13)
Taking Laplace transform on both sides.
k
Xc (S) = -XB(S)
gi nee ..... (6.14)
s
Taking Laplace transform of Eqn. (6.12),

XB(S) = 12 X A(S) -1)Xc(S)

r ing
.
..... (6.15)
substituting for XB(s) in (6.14),

SXc(s) = 1 X (s)-I X (S)

net
k 2 A ) C ..... (6.16)

..... (6.17)

..... (6.18)

..... (6.19)

Downloaded From : www.EasyEngineering.net

188 Operation and Control in Power Systems

I
--=Ts
k 13
is the time constant of the speed control mechanism and is of the order of several
milliseconds. The transfer function of the speed control mechanism

G (S) = XC<S) = Ks
..... (6.21 )
sc XA(S) I+STs

6.6 Transfer Function of a Power System

ww
Under steady state operating conditions the synchronous angular frequency

O)s = 21tfs where fs is the synchronous frequency (i.e.) the rated

w.E ilo =
frequency.
Change in angular position of the generator rotor

asy corresponding to an increased load demand il PD on

the generator.

E 0= Os + ilo radians
d d ngi d
0) = -(D) = -(os +ilo) = 0)5 +-ilo
. .... (6.22)

..... (6.23)
dt dt

0) = O)s + ilO) rad / sec

nee dt

..... (6.24)
Let f = fs + ilf rin ..... (6.25)

Where
I d
M= --iloHz
2n dt
g. net ..... (6.26)

The kinetic energy stored in the machine

W= ~ 10)2 = ~ 1(21tf)2 MJ ..... (6.27)

2 2
The kinetic energy at synchronous speed is
I ,
Ws. = -1(2nfs)-
2 . MJ ..... (6.28)

Then W=WS[f:)' MJ ..... (6.29)

..... (6.30)

Downloaded From : www.EasyEngineering.net

Load Frequency Control 189

1+2M) M
== Ws ( fs for smal fsI ..... (6.31 )

The rate of change of kinetic energy is the increase in power

~(W) = 2Ws ~(M) ..... (6.32)

dt fs dt

W
Defining per unit inertia constant H = _s
Pr

ww d 2H d
- (W) = -
dt
.
- (M)p.u. takIng Pr = 1.0 pu
fs dt
..... (6.33)

w.E
Further, all types of composite loads experience a change in power consumption with
frequency. Defining the load damping factor.

asy D - oPo
- i\f p.u. MW/Hz ..... (6.34)

E ngi
Where PD is the load demand in p.u.; the change in load demand in this case is
then D. M p.u.M.W
For a small step change in load demand by
written as
~
nee
PD the power balance equation can be

2H d rin
~PG -~Po = DM + - -(M)
fs dt
g.n ..... (6.35)

Taking Laplace transform of the above equation

[~PG (S) - ~Po (S)]= D~F(S) + 2H .S.~F(S)

et..... (6.36)
fs

..... (6.37)

..... (6.38)

Downloaded From : www.EasyEngineering.net

190 Operation and Control in Power Systems

..... (6.39)

1
where K = -Hz/p.u.M.W. . .... (6.40)
P D

and ..... (6.41 )

The transfer function relating the frequency change to the change in input-output power
may be then written as

ww Gp(S)= I+STp
Kp
..... (6.42)

w.E
Fig. (6.7) shows the block schematic of Eqn. (6.39)

asy
En
gi
/ - - - -....... dF(S)

nee
Fig. 6.7
rin
The complete block schematic of speed governing system is shown in Fig. (6.8).

6.7 Transfer Function of the Speed Governor g.n

Let m be the mass of the flying masses
c f be the friction control and C s be the spring control of the governor system.
et
-dP D

Fig. 6.8 The Block Diagram of Speed Governing System

The equation of motion of the governor is

Downloaded From : www.EasyEngineering.net

Load Frequency Control 191

..... (6.43)

Taking the Laplace transform on both sides of Eqn. (6.43).

(mS 2 + crS + cs) xA(S) == ks~w(S)
Rearranging

x A (S) ks
~(j)(S) = mS + crS + Cs ..... (6.44)

ww
2

== Gg(S) ..... (6.44)

w.E
Where Gg(S) is the transfer function of the speed governor.

a~s~y v= == natural frequency of oscillation of flying masses with

E ngi
no damping

s== 2~ksm nee

== damping constant ..... (6.45)

XA(S)== G(S) . ~ (j) (S)rin

6.8 Governing of Hydro Units g.n
The governing system of hydro turbines has to meet complex requirements because of the
destabilising effect of the inertia of water. Hydro-governors are provided with temporary
et
droop compensation to prevent over travel of the gate motion. The setting of temporary droop
and decay time constant Tr of the compensator determine the dynamics of response following
changes in load. The speed drop for a hydro unit is defined by

p.u. speed drop == ..... (6.46)

,
where N I == extrapolated speed corresponding to zero gate opening
N2 == extrapolated speed corresponding to 100% gate opening
and NR = rated speed.

Downloaded From : www.EasyEngineering.net

192 Operation and Control in Power Systems

The temporary droop provided by a dash-pot mechanism consisting of a spring, needle

valve and piston is shown in Fig. (6.9).

Governor permanent droop

r-------~~------------------__.c
Needle
valve

LO--1
/ I
A'------I
ww o
Temporary droop
mechanism

w.E ~p"otvaJve D

asy Fig. 6.9 Hydro turbine speed governing mechanism

En
The dash pot (temporary droop mechanism) is represented by the transfer function .

gin
e eri
..... (6.47)

where s is the complex frequency 0' + j{().

n g.n
The value of the gain 0 ranges from 0.2 to 1.0. The rest or decay time constant Tr lies
between 2.5 and 25s. The block schematic of a typical hydro governing system is shown in
Fig. 6.10. The value of k lies between 0.03 to 0.06.
et
Pilot valve & Rate Gate Position Penstock
PRef limits turbine
servomotor limits servo

+ Feedback for permanent droop

K
+

Dash pot for temporary droop

Fig. 6.10 Block schematic of Hydroturbine speed governing system

Downloaded From : www.EasyEngineering.net

Load Frequency Control 193

6.9 Penstock Turbine Model

Consider the penstock-turbine system shown in Fig. 6.11, having an effective length of I:, m
discharging water to the turbine at a velocity V mls operating at a head of H m. Following a
load change, consider the following p.u. changes in the variables:

ww L

w.E ________ J
H

asy Fig.6.11 Penstock Turbine mode

En l\H = change in head

gin
l\N = change in speed
l\X = change in turbine gate opening

eer
l\Q = change in flow
l\T
It can be proved that, in = change in turbine torque

p.u.change in head(l\H) = -2 1- exp[-2JeS] g.n

where
p.u.change in flow(l\Q)
Te = elastic limit of the penstock
1 + exp[ -2TeS]
et ..... (6.48)

Z = normalized penstock impedance

And s = (J + jro, the complex frequency.
Using first order Pade's approximation

x
1--
2
e -x = - -
x
1+-
2
Eqn. (6.48) becomes.

l\H
l\Q
= _Te Zs = - T w S ..... (6.49)

where T w is defined as water starting time.

Downloaded From : www.EasyEngineering.net

194 Operation and Control in Power Systems

The changes in flow (discharge) and torque can be represented by

AQ = allAH + a l2 AN + a l3 AX ..... (6.50)
AT = ~IAH + ~2AN + ~3AX ..... (6.51 )
where all' a 12, a 13 , ~I' ~2 and ~3 are constants.
The turbine torque change due to speed changes can be neglected in comparison to
other changes since speed change is relatively small. Therefore, Eqn. (6.50) becomes
AQ = allAH + a 13 AX
AT = a21 AH + a23 AX

ww Using Eqn. (6.49) and (6.50).

w.E
asy
En .. ... (6.52)

gi
For an ideal, loss less turbine with valve opening Xo
all = 0.5 Xo nee
a 21 = 1.5Xo
rin
a 13 = 1.0
~3 = 1.0 g.n
so that
AT
AX
=
1- (1.5X o + O.5Xo)TwS
I+O.5XoTwS
=
1- Xo TwS
1+0.5XoTwS
et
At full load in p.u. Xo= 1.0

AT = I-Tw s == I-Tw s ==GPT(S)

Therefore, ..... (6.53)
AX I+O. 5Tws l+(T;S)

Eqn. (6.53), is the classical turbine penstock transfer function. The water starting time,
Tw' varies from 0.5 to 4 s and can be calculated from the relation.

T == LV
w Hg ..... (6.54)

where g is the acceleration due to gravity (6.8 m/s2).

Downloaded From : www.EasyEngineering.net

Load Frequency Control 195

,--_ _ _ _ _ _--, Change in gate

Change in speed KG (I +STr ) position
(I + STI )(\ + ST2 )

Fig. 6.12 Approximate transfer function for hydro-turbine speed governor

The block diagram in Fig. 6.10 excepting penstock turbine model can be represented by
the linear approximate model shown in figure 6.12 where KG=11k and TI and T2 are given by

ww 1
T ,T
2
Ts
="2± (T2
k
)l
2

w.E .; -TA

and
a syE
1 ngi
T =-
G Ks
nee
Typical values for Tr and 0 can be obtained using the relations
Tr
5Tw = rin
and
0= 2. 5Tw
2H g.n
where H is the inertia constant of the turbine-generator on the machine MVA base. et
change in gate
Reference position
6F(S)
Ka tl+STrJ
(I +STI )(1 + ST 2 )

Turbine & Penstock

Fig. 6.13 Block diagram of a hydro turbine speed governing system

A complete block diagram for representing the speed - governing system, turbine and
penstock for dynamic studies is shown in Fig. (6.13).

Downloaded From : www.EasyEngineering.net

196 Operation and Control in Power Systems

6.10 Modal for a Steam Vessel

Steam inlet piping or steam chest can be considered as a steam vessel with steam input and
steam output Fig. 6.14.
For changes in steam valve position the dynamics can be represented by

- - - - . . . Steam outlet

ww Steam Vessel

Steam
w.E
Input asy
E ngi
Fig. 6.14 Steam Vessel

nee
Where W is weight of steam in kg and Om and OOlIt are the rate of flow of steam in
kg/sec at inlet and outlet of the steam vessel respectively. Assuming that the flow rate is
proportional to the pressure.
p rin
Where Po and
Oout --0
- P 0
o
g. ..... (6.55)

net
0 0 are the steady state pressure and rate of flow of steam in the vessel.

..... (6.56)

Assuming constant temperaLure in the vessel

..... (6.57)

d(V) dp
= dp ~ dt
Where V is the volume of the vessel in m3 and v is the specific volume of steam (m 3/kg)
in the vessel.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 197

= V.~(~)dP
dp v dt

= V ~(~). dQout .~
dp v dt Qo

= V ~(~)~.~Qout
dp v Q dt ..... (6.58)

ww T= ~.V~)
o

Let
w.E Qo dp v ..... (6.59)

asy Q. _ Q = T dQ out ..... (6.60)

In out dt

Taking Laplace transforms of Eqn. (6.60).

En
Qout (S)
Qin (S)
=--
I
I +ST
gi
Qin (S) - Qout (S) = T S Qout (S)

nee ..... (6.61 )

The transfer function of the steam vessel is

I
r ing
6.11 Steam Turbine Model
G
Sv
(S)---
- 1+ ST
. ..... (6.62)
net
In normal steady state the turbine power Pr keeps balance with the electromechanical air-gap
power PG' resulting in zero acceleration and a constant speed or frequency. Changes in Pr or
PG i.e., ~PT or ~P G or both will upset the above balance. If the difference ~PT - ~PG is
positive the turbine generator unit will accelerate, if negative it will decelerate.
The Turbine - Generator power increment ~PT depends entirely upon the valve power
increment ~P v and the response characteristics of reheat and non-reheat type of turbines are
different. However, it is possible to express the turbine - generator dynamics in terms of
turbine transfer function. The turbine - generator transfer function is given by

G (S) = ~PTG (S)

TG ~Pv (S)
For a simple non-reheat type turbine the model is given by a single time constant

Downloaded From : www.EasyEngineering.net

198 Operation and Control in Power Systems

K TO
G TO (S) = 1 + ST ..... (6.63)
TO

Where KTO is the gain constant and TTG is the time constant of turbine - generator unit.,

6.12 Reheat Type Steam Turbine Model

Reheat turbines have more than one time constant. Fig. 6.15 shows a tandem compound single
reheat turbine system configuration. Its block diagram is shown in Fig. 6.16.
The time constants TCh' TRh and TCo represent the delays due to the steam chest and
inlet piping reheater and cross over piping respectively Fig. 6.16.

ww
w Valve .Ea
syE
Control
Position Valve &
Steam
Stage
Chest
ngi Stage
condenser
Legentd H.P. = High pressure stage turbine
nee
I.P. = Intermediate pressure stage turbine
L.P. = Low pressure stage turbine
r
Fig. 6.15 Tandem Compound Single Reheat System ing
+ +
. net

l+sTCo

Fig. 6.16 Block Diagram Representations

KHP• KIP and K LP are the fractions of the portions of the total turbine power developed
in various stages. The steam chest and inlet piping to the first turbine cylinder, the reheater and
the cross over piping down introduce time delays between valve movement and the change in
steam flow. The time constants are associated with the entrained steam in the components
mentioned.

Downloaded From : www.EasyEngineering.net

LOlld Frequency Control 199

If a two stage steam turbine with a reheat unit is considered then the dynamic response
will be influenced by (i) the entrained steam between the steam inlet valve and first stage of
turbine and (ii) the storage action in the reheater which causes the outlet of the L.P. stage to
lag behind that of the H.P. stage.

Inlet valve

ww (a)
Fig. 6.17 Two Stage Reheat Unit
(b)

w.E
The turbine transfer function is characterized by two time constants. For ease of
analysis it can be assumed that the turbine is modeled by a single equivalent time constant. The

asy
value of T T lies between 0.2 to 2.5 sec in most of the cases. This is shown in Fig. 6. I 7(b).

6.13 Single Control Area

En
In the previous sections models for turbine - generator, power system and speed governing

gi
systems are obtained. In practice, rarely a single generator feeds a large area. Several generators

nee
connected in parallel, located also, at different places will supply the power needs of a
geographical area. Quite normally, all these generators may have the same resl'onse characteristics
for changes in load demand.
r ing
In such a case, it is possible to define a control area, grouping all the generators in the area

. net
together and treating them as a single equivalent generator. For small load changes all these
generators swing in unison. Putting together, the various models derived so far a single control
area or simply an area can be conceived as shown in Fig. 6.IS.
From Eqn. (6.20),

From Eqn. (6.42),

Kp
OpeS) = I +STp

and from eqn. (9.39) L1F(S) = [L1 Po(S) - LI PD(S)] [1 +K;TP ]

Note that ~Po(S) = ~PTO(S) as generator power change which is also the turbine
power change and also the turbine generator unit power change.

Downloaded From : www.EasyEngineering.net

200 Operation and Control in Power Systems

Also, from eqn. (6.63) GTG = KTG /1 + STTG

1
R

-Ll PD(S)
Xd S)
Ks K TG + Kp

LlPref 1+ STs I+STTG + - I +STp Ll/lS)

Ll PTG
Ll PG

ww Fig. 6.18 Block diagram of a single area system

w
6.14 The basics of Load Frequency Control

.Ea
The following basic requirements are to be fulfilled for successful operation of the system:
1.
2. syE
The generation must be adequate to meet all the load demand
The system frequency must be maintained within narrow and rigid limits.
3.
4. ngi
The system voltage profile must be maintained within reasonable limits and
In case of interconnected operation, the tie line power flows must be maintained at
the specified values.
nee
When real power balance between generation and demand is achieved the frequency

r ing
specification is automatically satisfied. Similarly, with a balance between reactive power
generation and demand, voltage profile is also maintained within the prescribed limits. Under

.
steady state conditions, the total real power generation in the system equals the total MW
net
demand plus real power losses. Any difference is immediately indicated by a change in speed
or frequency. Generators are fitted with speed governors which wi II have varying characteristics:
different sensitivities, dead bands response times and droops. They adjust the input to match
the demand within their limits. Any change in local demand within permissible limits is absorbed
by generators in the system in a random fashion.
An independent aim of the automatic generation control is to reschedule the generation
changes to preselected machines in the system after the governors have accommodated the
load change in a random manner. Thus, additional or supplementary regulation devices are
needed along with governors for proper regulation.
The control of generation in this manner is termed load-frequency control. For
interconnected operation, the last of the four requirements mentioned earlier is fulfilled by
deriving an error signal from the deviations in the specified tie-line power flows to the neighboring
utilities and adding this signal to the control signal of the load-frequency control system. This
last requirement will be discllssed in detail in the next chapter.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 201

Should the generation be not adequate to balance the load demand, it is imperative that
one of the following alternatives be considered for keeping the system in operating condition:
I. Starting fast peaking units.
2. Load shedding for unimportant loads, and
3. Generation rescheduling.
It is apparent from the above that since the voltage specifications are not stringent. load
frequency control is by far the most important in power system control. The block schematic
for such a control is shown. in Fig. 6.19.

ww
w.E 2

asy
En
gi nee
Actual output Sensor

rin
Fig. 6.19 Block diagram for load frequency control

g.n
In order to understand the mechanism of frequency control, consider a small step-
increase in load. The initial distribution of the load increment is determined by the system
impedance; and the instantaneous relative generator rotor positions. The energy required to
supply the load increment is drawn from the kinetic energy of the rotating machines. As a
result, the system frequency drops. The distribution of load during this period among the
et
various machines is determined by the intertias of the rotors of the generators partaking in the
process. This problem is studied in stability analysis of the system.
After the speed or frequency fall due to reduction in stored energy in the rotors has
taken place, the drop is sensed by the governors and they divide the load increment between
the machines as determined by the droops of the respective governor characteristics.
Subsequently, secondary control restores the system frequency to its normal value by
readjusting the governor characteristics.

6.15 Flat Frequency Control

In this method, a master machine is charged with the task of mainta,ining the system frequency.
All the remaining machines carry constant loads. The capacity of the machine which controls

Downloaded From : www.EasyEngineering.net

202 Operation and Control in Power Systems

the frequency should be between 5 and 10 % of the entire generating capacity of the system
for effective frequency control. In small, independent systems, old and inefficient machines
may be assigned to frequency regulation leaving new and efficient machines to supply the load
demand.
Modern power systems are so large that it is impossible to design a single central control
system that would handle the overall control job. It is extremely useful to take into account the
weak links in the system and then apply control through decomposition. The demarcation of
load frequency control and Mvar voltage control characteristics is one such decomposition.
Geographical and functional decomposition are successfully applied to power systems and this
leads to the concept of area control. A modern power system can be divided into several areas

ww
for load frequency control. Each control area fulfils the following:
I. The area is a geographically contiguous portion of a large interconnected area,

w.E
which adjusts its own generation to accommodate load changes within its precincts.
2. Under normal conditions of operation, it exchanges bulk power with neighbouring
areas.

asy
3. Under abnormal conditions of operation, it may deviate from predetermined schedules
and provide assistance to any neighbouring control area in the system.
En
4. It is expected, in addition, to partake with the other areas in the system in a suitable

gi
manner in the system frequency regulation.

nee
The rotors of all generators in a control area swing together for load changes. Thus, a
coherent group of generators within a geographical region may constitute a control area which

rin
is connected to other similar areas by weak tie lines as shown in Fig. 6.20.
Inter tie with
area I
g.n
Inter tie with
area 3 et
Inter tie with
area n

Inter tie with

area 2

Fig. 6.20 Single control area inter ties

6.16 Real Power Balance for Load Changes

It has been already m.entioned that when an imbalance occurs between real power generation
and load demand, frequency deviation takes place. Consider a single area operating in steady-
state and let it undergo a step change ~PD in load demand. Assuming that the generation

Downloaded From : www.EasyEngineering.net

Load Frequency Control 203

changes by an amount ~p G following the load change due to governor action, the power
imbalance is ~p G - ~p D' There are three possible modes in which this power can be absorbed.
For positive power imbalance:
1. The area kinetic energy increases (i.e. the kinetic energy of all the generator rotors in
the area increases),
2. The load demand increases (due to load characteristics) and
3. The power out-flow from the area to other interconnected areas increases, if such
interconnections exist.

6.17 Transfer Function of a Single Area System

ww
Under steady state operating conditions. Transfer functions for a single control area can be
derived exactly in the same way as for the system in sec. (6.6).

w.E IDS = 27tfs

where IDs and fs are synchronous angular speed and rated frequency respectively.

asy
If ~() is the change in angular position of the equivalent generator representing the area
corresponding to an increased load ~PD in the area.
Then,
E 0 = Os + ~o rad
d d
C=-(o)=-(L +CL)=C s + - ngi
. d~o
dt dt s
IDs + ~ID
=
. dt
nee ..... (6.64)
Let f =fs+M
rin
Where
I d
M= - -
27t dt
~oHz
g.n ..... (6.65)

The area kinetic energy

W = \J2 I ID2 = Y2 I (2 7t t)2 MJ et
Where I is the moment of intertia of the area.
Also, the kinetic energy at synchronous speed IDs is
Ws = \J2 I (2 7t t)2 MJ

Thus, W ~ Ws (~ J MJ

r:::.
( M)
fs
for small -M
Ws I + 2 -
fs
..... (6.66)

Downloaded From : www.EasyEngineering.net

204 Operation and Control in Power Systems

The rate of change of kinetic energy is the increase in area power which is

~(W)= 2Ws ~(M)

dt fs dt
Defining per unit inertia constant M = W/Pr
d 2M d
-- (W) = - - (M)p.u.
dt fs dt

All types of composite loads experience a change in power consumption with frequency.

ww
Defining the load damping factor D.
_ oPD
8f" p.u. MW/HZ
w.E 0- ..... (6.67)

Where PD is the load demand in p.u; the increase in load demand in this case is then

asy
D.6f p.u. MW, For a small step change in load demand, 6f p.u. MW. For a small step change
in load demand, 6P D the power balance equation takes the form.

En .. ... (6.68)

gin
eer
Taking the Laplace transform of the above equation

[6PG (S) - 6PD (S)]= D6F(S) + 2H S6F(S)

fs
i ng.
or
n et

..... (6.69)

Where Kp = 110 HzJp.u. MW

2H
Tp == f 0
s

Downloaded From : www.EasyEngineering.net

Load Frequency Control 205

The transfer function relating the frequency change to the change in input/output powers
may be designed by Gp(S) so that
Kp
Gp(S) = ..... (6.70)
I +STp
The block schematic for a single area system is shown in Fig. 6.21 (a) An entire control
area may be represented as in Fig. 6.21(b).
~PD(S)

ww .PGe,j-J 1~f-I--~-F·~S) (a)

w.E
asy 1--"--+ ~F(s)

Governing
En
EqUivalent turbme ~P o(s) PO\\ier

gi
system generator system
(b)

nee
Fig. 6.21(a, b) Block diagram of a single area system

6.18 Analysis of Single Area System

r ing
Consider'the single area system shown in Fig. 6.21 (b). For a step load change in the system,

.
the following equations can be written
[L\P o(S) - L\PG(S)] Gp(S)

I _ L\P (S)
= L\ F(S)
net
RG ST (S)L\F(S) - G

where GST = ( KTG x Ks )

1+ STrG 1+ STs

Solving for L\F(S) L\F(S) = - G p (S) L\P (S)

o
1+ ( ~ )G p(S)G ST (T)

L\P
For a step load change, L\Po(S) = __ 0
S
Substituting the value of L\PoCS)

Downloaded From : www.EasyEngineering.net

206 Operation and Control in Power Systems

Applying the final value theorem, the steady state error Mss would be

Lim ~ ] M'o
~fss = S~O LSM'(S) = - D K K
+ TG S
R

ww The product KrG Ks can be made unity by properly selecting the units for the input-
output quantities to the combined block G ST(S) so that

w.E ~fss
-~p
= _ _0_

asy D+-
1
R

E
The quantity (D + llR) is defined as area frequency response characteristic or area

ngi
frequency regulation characteristic (AFRC) and denoted by ~. It may be noticed that as R
becomes smaller ~ fss approaches zero.

Reference Power Setting nee

rin
In Fig. (6.18) reference power is indicated at the input summer to the system. This determines

We obtain g.
the starting point of the governor characteristic. Under static conditions setting KsKrG = 1.

net
Changing the reference power setting will also change the turbine generator output in a
proportional manner. For instance if the machine connected to an infinite bus in which case
M = 0 we have the direct relationship.
M'ref= M'TG
For a fixed setting of the speed changer the steady state increase in power output
from the turbine generator M'TG is directly proportional to the frequency drop. As in this
case M'ref = 0
1
M' =- -~f
TG R

R has the units hertz per MW or hertz pr unit MW as the case may be.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 207

106

Set to 50% load

20 40 60 80 100 120

ww Percent of rated output

If 0
w.E
0, then AFRC
Fig. 6.22 Speed Changer Setting

P= I/R = the area speed regulation coefficient.

asy
= =

The time constants Ts and T T are much smaIler compared to T p and as such,

En
an approximate solution to the response of the system could be obtained by neglecting
them. Then

gi
~F(S) =
1+(
Gp(S)

~)Gp(S)
~PD

S nee
Kp
I +STp ~PD
r ing
1+(~)I+K;Tp S . net
Where A = Kp + R
B =Tp R
Partial fractions for the expression in the bracket can be obtained as follows:
C 0
. =-+---
S A + BS S A + BS
I
At S = O·, C = -A

A B
And at S=-_· 0=--
B' A

Downloaded From : www.EasyEngineering.net

208 Operation and Control in Power Systems

So that

M(t) == VI [~F(S)]

-~PDP [2. - I
KpP JI
ww S S+
Tp

w.E ..... (6.71 )

asy
The response is shown in Fig. 6.23.

E L\f( Hz)

""T"+------------.--.. ngi
nee
F(s)

L\PD (Hz)
J3

rin
g.n
Fig. 6.23 Uncontrolled single area system-response to step load change

6.19 Dynamic Response of Load Frequency Control Loop

et
It has been shown that the load frequency control system posses inherently steady state error
for a step input. Applying the usual procedure, the dynamic response of the control loop can
be evaluated so that the initial response also can be seen for any overshoot.
For this purpose considering the relatively larger time constant of the power system the
governor action can be neglected, treating it as instantaneous action. Further the turbine-
generator dynamics also may be neglected at the first instant to derive a simple expression for
the time response.
It has been proved that

~F(S)==- I G
p
~PD(S)
l+-G s G TG G p
R

Downloaded From : www.EasyEngineering.net

L(}ad Frequency Control 209

For a step load change of magnitude k.

. -k
~PD(S) =S
Neglecting the governor action and turbine dynamics

~F(S) = _ Gp k
I+~G S
R p

ww (
Kp )
=- I+STp ( I Kp
1+----'--
I
)S k

w.E
Applying partial fractions
R 1+ STp

asy
E ngi
(-~ -~-"-ll
S[S + Tp + RTp
Kpk
---

nee
6.20 Control Strategy
rin
The uncontrolled system is subject to a steady state error for step load changes. To reduce
this error, consider first the introduction of a negative feedback signal from the frequency
deviation, i.e.
g. net
Let ~Pc = -K) ~ F(S) .... (6.72)
where K) is the gain for the proportional control.
The response would then be

..... (6.73)

In the steady state the frequency error would be

M = -Kp ~P =_ ~PD
ss I( K+- I)
R)
0 I (I
-- + -- + K
Kp R I
) ..... (6.74)

Downloaded From : www.EasyEngineering.net

210 Operation and Control in Power Systems

If K] is made very large then only ~fss reduces to zero. This, in other words, means that
R should be made equal to zero, which is not desirable. Proportional control is not suitable for
reducing the steady state error to zero.
It is a well known fact in control theory that integral control will improve the steady
state performance.

Let ..... (6.75)

So that ..... (6.76)

ww The negative sign signifies that the control signal ~ Pc has to be increased for reduction
in frequency. The equations for the system in Fig. 6.24 are:

w.E L\F(s)

asy
En
gi nee
r
L\F(s)

Fig. 6.24 Proportional and Intelgral Control

ing
. net
..... (6.77)

Also, if Ks KTG is adjusted to be unity

( &+lfR)~F(S)(
S
I )( 1
l+STs I+STTG
J=~PG ..... (6.78)

Substituting eqn. (6.78) in eqn. (6.77) and solving for ~F(S)

~F(S) == ~PD(S)Kp(l + STT)(1 +STs)S

(l + STs)(l+ STT)(1 + STp)S + Kp(K2 + Sf R)

for a step load change ~PD(S) = - ~PDfS, and the steady state error obtained as

Lim ]
~fss == [S.~F(S)
S~O

Downloaded From : www.EasyEngineering.net

Load Frequency Control 211

= -SKp ~PD (\ + STT )(S + STs )S = 0

..... (6.79)
S (\ + STs)(1 + STT )(1 + STp)S + Kp(K2 + SI R)

By making use of integral control strategy, the steady state error can be eliminated.
For the reasons explained before, TT and T s cam be neglected for an approximate
analysis for the response. For a step load disturbance.

F(S) =
S(~pp ) . -~PD
..... (6.80)

ww S 2 + S(I +Kp)
- -I +K2Kp
R Tp
--
Tp
S

w.E
The response ~f(t) depends on the nature of the denominator expression, i.e. the
characteristic equation.

a S2 + 2 80J n S + OJ n2

sJ yE
Where OJ =
n Tp
ngi
K 2 K P is the natural frequency

and il =
1+~
R J Tp nee
=damping ratio
2Tp K2Kp
rin
g.n
The gain for critical damping can be obtained by setting 8 equal to unity.

et
or . .... (6.81)

For values of K2 greater than that given by eqn. (6.81), the response is oscillatory due to
under damping. For values ofK2 less then the critical value response is monotonically decreasing
without oscillations or as in Fig. 6.25.
The selection of the gain for the controller should be such that the following specifications
are satisfied.

Downloaded From : www.EasyEngineering.net

212 Operation and Control in Power Systems

Without integral control

ww Fig. 6.25 Response to integral control with different gain settings

w.E
1. The control loop must be stable.
2. The frequency error should return to zero following a step load change. The deviation

asy
in the transient state must also be minimized.
3. The integral of the frequency error should not exceed a certain value (say 150Hz or
3 seconds)
En
6.21 PID Controllers
gin
eer
From the above analysis, it is clear that proportional integral and derivative control strategy can
be applied for load frequency control. While proportional control is inherent in the feedback

i ng.
through the governor mechanism itself, derivative control when introduced improves transient
performance and ensures better margin of stability for the system.

n et
where K3 is the gain of derivative controller.
The response to such a control is

By proper choice of K J' K~ and K3 all the specifications for the system performance can
be satisfied.

Downloaded From : www.EasyEngineering.net

Load Frequency Control 213

E 6.1 A \00 MW generator has a regulation parameter R of 5%. By how much will the
turbine power increase if the frequency drops by 0.1 Hz with the reference unchanged.
Solution:
Actual change in frequency = 5% of 50Hz = 0.05 x 50 = 2.5Hz
R = 2.5Hzl100Mw = 0.025 HzlMw
If M = -0.1 Hz, the increase in turbine power
I I
~P= -- M= ---x(-O.I) =4 MW
2 0.025

ww The turbine power increase = 4MW.

E 6.2 A 100 MW generator with R = 0.02 HzlMW has its frequency fallen by 0.1 Hz. If the

w.E
turbine power remains unchanged by how much the reference power setting be changed.
Solution:

asy
The signal to increase the generation is blocked. Thus at the input summing point the reference

En
power setting must be changed. Such that

~ Pref ~R
gi -

I
M= 0

1 nee
(i.e .. ) ~ Pref - R M= 0.025 xO.l = 4 Mw

r ing
E 6.3 Two generators with ratings 100 MW and 300 MW operate at 50Hz frequency. The

. net
-s.ystem load increases by 100 MW when both the generators are operating at about half
of their capacity. The frequency then falls to 49.5Hz.
If the generators are to share the increased load in proportion to their ratings
what should be the individual regulations? What should be regulations if expressed in
, per unit Hertz/per unit megawatt?

1
Solution: AP= --M
R

I
~P = --M
2 R2
J
M=0.5 Hz
Power is shared in proportional to their ratings

Downloaded From : www.EasyEngineering.net

214 Operation and Control in Power Systems

100
~P = 100 x - = 25 MW
I 400

300
~P = 100 x - = 75 MW
2 400

Hence R = - 0.5 = -0.00667

I 75

If regulation is expressed in p.u. then with f = 50Hz

ww
w.E R2 = -
0.00667
50
300
x-
I
= 0.04 pu HzlPu MW

a syE
Both have the same value, even though based on their individual ratings, they have
different regulation.

ngi
E 6.4 Determine the primary load frequency control loop parameters for a control area having
the following data:
nee
Total rated area capacity Pr = 1000 Mw
Normal operating load
Inertia constant H = 4.0 sec
= 500 MW
rin
Solution:
Regulation R = 2.5 Hzlpu MW g .ne
Load damping
OP 500MW
D = Of = 50Hz = 10 MW/Hz
t
[Here, the load damping is assumed linear and percentage change is assumed to be
the same]
10
In per unit D =- = 0.01 pu MW/Hz
1000

2H 2x4.0
Tp = fsD = 50x 0.01 = 16 sec

1 1
K
p
= -
D
= -O.oI~
- = 100 Hzlpu MW

Downloaded From : www.EasyEngineering.net

Load Frequency Control 215

E 6.5 Determine the area frequency response characteristic and the static frequency error for
a system with the following data, when I % load change occurs?
D = 0.0 I pu Mw/Hz
R = 2.5Hzlpu MW
Tp = 16 sec
Kp = 100 Hzlpu MW
Solution:
1
Area frequency response characteristic ~ =D+ R

ww 1

w.E = 0.01 + 2.5 = 0.41 MW/Hz

asy M=- M =-
~ 100 x 0.41
=0.02439Hz

En
E 6.6 In the example E6.5, the governor is blocked so that it does not change the generation.

gi
In that case what would be the steady state frequency error?
Solution:
nee
When the governor is not acting, the feedback loop is not existing. In such a case R is infinite.

I
~=D+ R =D=O.OI puMW/Hz r ing
Hence, ~f = - -
M
~
=- -0.01 = -I Hz
0.01
. net
Frequency falls by 1 Hz (i.e.) f = 50 - I = 49 Hz.
It may be noted that with the generator acting the frequency from E6.5 is
50 - 0.02439 = 49.9756 Hz
The importance of feedback through governor mechanism can be understood from the
above.

E 6.7 A 100 MVA synchronous generator operates initially at 3000 rpm, 50Hz. A 25MW load
is suddenly applied to the machine and the steam valve to the turbine opens only after
0.5 sec due to the time lag in the generator action. Calculate the frequency to which the
generated voltage drops before the steam flow commences to increase to meet the new
load. The value of the stored energy for the machine is 5kW-sec per KVA of generator
energy. Also calculate the value of H constant for the generator.

Downloaded From : www.EasyEngineering.net

216 Operation and Control in Power Systems

Solution:
Stored energy = 5KW per/kVA
(i.e.,) = 500 MW sec pr 100 MVA
Load increase =25 MW
Energy required to supply this load for 0.5 sec = 25 MW sec
Frequency at 500 MW sec stored energy = 50Hz
Frequency fall = M

ww M
-=-=
f
i1f
50
25xO.5 MW
500
sec

w.E i1f = 50 x 25 x 0.5 = 1.25Hz

500

asy
Frequency falls to 50 - 1.25 = 48.75 Hz

H constant =
E stored kinetic energy at rated frequency = 5MW - sec x I OOMV A = 5 sec
machine rating
ngi MV A I OOMV A

nee
E 6.S Given the following parameters, obtain the frequency error. Plot it when a step load
disturbance of (i) I % and (ii) 2% occur in the system.
Tp = 22 sec
rin
Solution:
R = 2.5
Kp= 100 g. net
Gp
The expression for i1F(s) = --I-~--i1PD(S)
l+-GsGToG p
R

Neglecting the turbine dynamics and governor action (G s GTO :::: 1.0)

k
i1F(S) =
I +STp I +~_~ s
R 1+ STp

k] = 0.01 and k2 = 0.02

simulating the transfer functions in .MATLAB the response is obtained and shown
in Fig. E(6.8).

Downloaded From : www.EasyEngineering.net

Load Frequency Control 217

ww
w.E
asy
En
gi
Fig. E6.8 Step load frequency error characteristic without supplementary control 1. k = 0.01
=
nee
2. k 0.02 (governor action and turbine dynamics are neglected

E 6.9 Show the effect of governor action and turbine dynamics, if they are not to be neglected
in E6.8 given Ts = 100 msec and TTG = 0.5 sec.
rin
Solution:
For this the exact frequency error is used g.n
LlF(S) = I Gp . LlPD (S)
et
1+ RGSGTGG p

_ Ks
LlPD(S) = O.Olpu; G s --'"""-- G TG -- K TG
1 +STs I+STTG

Ks KTG ::::: 1
The response is shown in Fig. E(6.9).
It can be seen that greater the step load change, larger the error and the governor action
and turbine dynamics does not cause any change in the response in the steady state, except for
a transient deviation at the beginning of the distribution.

Downloaded From : www.EasyEngineering.net

218 Operation and Control in Power Systems

ww
w .Ea
syE
ngi
nee
r ing
Fig. E6.9 Step load frequency error characteristic without supplementary control 1. k
= = =
=0.01

.
2. k 0.02 Ts 100 msec and TTG 0.5 sec

E 6.10 An isolated power station has the following parameters

Turbine time constant T t = O.5sec
net
Governor time constant Ts = 0.2sec
Governor inertia constant H = 4 sec
Governor speed regulation R = R per unit
The load varies by 0.8 percent for a I percent change in frequency, i.e. D = 0.8 . The
governor ~peed regulation is set to R = 0.05 pu. The turbine rated output is 250 MW at nominal
frequency \>f 50Hz. A sudden load change of 50MW(dP L = 0.2 pu) occurs.
(a) Construct the SIMULINK block diagram and obtain the frequency deviation
response.
(b) Set integral gain to 7 and obtain the frequency deviation response and compare
both the responses.

Downloaded From : www.EasyEngineering.net

- - - - - - - - - - - -

Downloaded From : www.EasyEngineering.net

Load Frequency Control 219

Solution:
To create a SIMULINK block diagram presentation select new (model) file from FILE menu.
This provides an untitled blank window for designing and simulating a dynamic system. Copy
different blocks from the simqIink libraries or other previously opened windows into the new
window by depressing the mouse button and dragging.
I. Open the continuous library and drag the transfer function block to the window.
Double click on transfer function to open the dialog box. Enter the numerator and
denominator values (the coefficients in the descending powers of s, if any power of
s is missing, enter zero) of the transfer function .
2. Open the math library and drag the sum block in to the window. Open the sum dialog

3. ww box and enter + - under list of signs.

From the math library drag the Gain block into the model file right click on the gain

4. w.E
block and click on the Flip option to rotate the gain block by 180degrees.
Open the source library and drag the step input block to window. Double click on it

asy
to open its dialog box and set up the step time(step duration), initial and final
values(which will be same) to represent the step input.
5.
En
Open the sink library and drag scope to window to observe the response.

gin
By using the left mouse button, connect all the blocks.
Before starting simulation, set the simulation parameters. Pull down the simulation

eer
dialog box and select parameters. Set the start time, stop time and for a more accurate integration,
set the maximum step size.

ing
In this example the parameters for all the blocks for the system in the fig. are initialized.
Open m-fiIe and enter the parameter values. The following m-fiIe has to be run prior to the
simulation.( of model file)
.ne
Open new m-fiIe and enter the parameter values as shown below.
Tg = 0.2;
T t = 0.5;
t
H= 5;
D = 0.6;
R = 0.05;
%for integral control
Kj = 7;
Save the m-fiIe under parameters and run the file. SIMULINK block diagram and results
for LFC
By using above procedure construct the simulink block diagram for the load frequency
control of isolated power system as shown in Fig. E 6.1O(a).

Downloaded From : www.EasyEngineering.net

220 Operation and Control in Power Systems

Constant Governer Turbine Inertial.load

Step

l/R

ww '-----------<-K-I4-----------'

w.E
Fig. E 6.1 O(a) Block diagram Model of Load Frequency Control (Isolated Power System)

asy
Pull down the file menu and use save as to save the model under AFC. Start the
simulation. Double click on the scope, click on the auto scale, the result is displayed as shown
in Fig. E 6.1 O(b).
En
Add integral controIIer block for above system and save the file under LFC 1.

gin
Start the simulation. Double click on the scope, click on the auto scale, the result is
displayed as shown in figure below.

eer
i ng.
n et

Fig . E6.10(b) Response of LFC

1. Uncontrolled response
2. Integral control response

Downloaded From : www.EasyEngineering.net

Load Frequency Control 221

E 6.11 A single control area system with the following data experiences a sudden load change
of3%.
Kp= 100
Tp= 25
R = 2 Hzlp.u. MW
with integral control using a gain of 10 obtain the frequency plot with time and show that the
frequency deviation is reduced to zero.
Solution:
Using MATLAB, the solution is obtained and shown in Fig. (E 6.11).

ww Kj=IO Kp = IOO
d l = (I1Tp) + (K/ (R
Tp = 25
* Tp))
R =2
d2 = Kj * K/Tp
num = Kp/Tp
den = [I d l d 2]

w.E
asy
En Transfer Fan
gin
eer
ing
.ne
t

Fig. (E 6.11)

Downloaded From : www.EasyEngineering.net

222 Operation and Control i!!.!Qwer Systems

6.22 The optimal Control Problem

A dynamic system is described mathematically by
x = f(x,u, t) ..... (6.82)
Where x denotes the state vector and u the control vector
Let it be desired to minimize the performance index or cost functional
I

J = fL(x,u,t)dt ..... (6.83)

°
Pontriagyn's method for optimal control involves in creating an augmented cost function

ww
for minimization. .
L ' (x,x,u, t) = L(x,u, t) + At[f(x,u, t) - x ..... (6.84)

w.E
Applying Euler-Lagrange equations, we obtain
.au
asy A.=--
ax ..... (6.85)

E . au
x=--
aA
ngi ..... (6.86)

0= au
au nee ..... (6.87)

Where u = L (x, u, t) + At [f(x, U, t)]

rin ..... (6.88)

g.
and A. denotes Lagrange multipliers. Simultaneous solution to eqn. (6.85, 6.86, 6.87)
subject to boundary conditions gives optimal control Uo. But, the equations are non linear and
time varying and hence even for simple systems the solution will be difficult to obtain.
net
6.23 The Linear Regulator Problem
Consider the linear dynamic system

~(t) = [A](t)X(t) + [B][t]!:! (t) ..... (6.89)

Let the cost functional be

I I 12
J = _XT (t 2 )HX(t 2 ) +- fX(t)Q(t)X(t) + u T(t)P(t)u(t)]dt ..... (6.90)
2 21J

Where Hand Q are real symmetric positive semi definite matries and P is a real symmetric
positive definite matrix. Again t2 is fixed, but X(t 2) is free. For simplicity, omitting the arguments .

..... (6.91 )

Downloaded From : www.EasyEngineering.net

Load Frequency Control 223

For optimality ..... (6.92)

..... (6.93)

..... (6.94)

From the above VO =-P-I(t)B T(t)90(t) ..... (6.95)

ww substituting this VO in equation (6.92)

XO (t) = A(t)X(t) - B(t)p-I (t)B T(t)9° (t)}

w.E XO (t) = -Q(t)X(t) - AT (t)9° (t)

..... (6.96)

asy
En ..... (6.97)

or
gi nee ..... (6.98)

where + is called state transition matrix. r ing

From the boundary conditions we obtain )..0 (t 2) = HXO (t 2)
.
and hence,
..... (6.99)

net
H XO(t 2) = +21 XO(t)+ +12 AO(t) ..... (6.100)
It can be derived from the above that

A° (t) = [+22 - H+ 12 ]-1 [H+ 12 - +22]x° (t)

= R(t) XO (t) ..... (6.101 )
From eqn. (6.95) VO(t) = _P- 1B T(t)Ao(t)

= _p-l(t) BT(t) R(t) XO(t)

= _[Kt (t) XO(t)] ..... (6.102)
The optimal control is a time varying linear combination of system states.

Downloaded From : www.EasyEngineering.net

224 Operation and Control in Power Systems

6.24 Matrix Riccati Equation

Consider eqn (6.10 1)
°
A (t) == R(t)Xo (t)

Differentiating i Get) == R(t)Xo(t) + R(t)Xo (t) ..... (6.103)

Substituting XO(t) and AO(t) from eqn. (6.92) and (6.93)

== Q(t) X(t) - AT(t) AGet)

== R(t) X(t) + R(t) A(t) X(t) - R(t) B(t) p-l(t) BT(t) AO(t) ..... (6.104)

ww Eliminating A(t) and rearranging the terms

R(t) B(t) p-l(t) BT(t) R(t) XO(t) - R(t) A(t) X(t) ..... (6.105)

w.E - R (t) XO(t) - Q(t) XO(t) - AT(t) R(t) X (t) == 0 ..... (6.106)

asy
since the above eqn. must be satisfied for at XO(t)
R(t) + R(t)A(t) + AT (t)R(t) - R(t)B(t)p-l(t)B T(t)R(t) + Q(t) ..... (6.107)

En
• The above equation is called matrix Riccati equation. R is an n x n symmetric matrix.
n(n + 1)
2
gi
first order differential equations only need be solved to get the elements ofR matrix,

nee
so that the elements of feed back gain K(t) can be computed further from equation (6.102)

rin
For controllable dynamic systems H = 0, A, B, P and Q are time invariant matrices and as
S ~ ex) R(t) ~ RO, a constant matrix. The matrix Riccatti equation becomes
ROA + ATRo+ Q - ROBp-l BT RO = 0
g.n ..... (6.108)

In the steady state R (t) ~ o.

The gain elements from K matrix can be determined as constants after the elements
of RO matrix are determined. This involves solution to simultaneous non-linear algebraic
et
equations. There are several methods available for the solution of which Klineman's method is
simple and elegant.

6.25 Application of Modern Control Theory

The load frequency control (LFC) problem can be viewed as a dynamic optimization problem.
Integral square error criterion can be applied and controllers can be designed using optimal
control theory. Controller design using this theory was first proposed by Fosha and Elgerd.
The results obtained by using this theory contradict the existing practice in the selection of the
bias parameters and the integral controller gains.
The system equations are put in the state variable form denoted compactly by
X==AX+Bu+Fd ..... (6.109)

Downloaded From : www.EasyEngineering.net

Load Frequency Control 225

where A.B and F are the system, input distribution and disturbance distribution matrices
respectively and X, U and d are the state, control and disturbance vectors respectively.
By a suitable transformation, eqn. (6.24) is transformed into the form.
X' =A'X' +B'U' ..... (6.110)
The LFC problem is then treated as a linear regulator problem and the solution is obtained
by solving algebraic Riccatic equation.

6.26 Optimal Load Frequency Control- Single Area System

The states to be minimized are the frequency deviation Mand the time integral of the frequency

ww
deviation f~f.dt. The contrl signal is to be weighted and included in the cost function so that
too large a control is not required. The cost function selected is

w.E('T )
J= f 2"1
00
~ Qx + UT Pu dt ..... (6.111)

asyo
Where Q and P are the weighting matrices for the state and control vectors and X and u.

En
Consider the block diagram for the single area system shown in Fig. 6.26.

gi
r---II/R~--------------.....,

nee
r ~...-+t.F(s)

ing
Fig. 6.26 Block diagram for single area system.

As indicated in section (6.18) Ks' KrG is set equal to one.

. net
The following states are defined

XI = filf(t)dt = integral of frequency error

~ = ~ f(t) = frequency deviation

x3 = ilPTdt) = change in turbo generator output

x4 = ~X v(t) = valve displacement at speed governing system.

Thus, xT = [J~f ilf LlPTG ~Xv ]

Let u be the control vector and d = LlPd be the disturbance vector.
Here, the control is a single input scalar u, and the disturbance d is the change in the load
demand ilPd

Downloaded From : www.EasyEngineering.net

226 Operation and Control in Power Systems

The perfonnance index is

qll
XT Qx = {x j x_}
--2 { q2l ..... (6.112)

Since only ~f and fMdt are to be minimized.

Identifying the states and the state variable system model, we obtain
Xl (t) = X2 ( t) ..... (6.113)

Since Xj(t) = fM{t) dtand ~(t) = ~f(t)

ww
From the block diagram

M(s) M>TG (s) -M>p {s)J (I

w.E = [

Cross multiplying and rearranging

+KiT )
p
..... (6.114)

a
S ~ F(s) 1
= __
Tp
syE~F(s) + Kp ~PTG (s) -
Tp
Kp M>D (s)
Tp
..... (6.115)

ngi
In the time domain, using the state variables

. () 1 Kp Kp
x2 t =--x2+-x3-~' d

Also, from the block diagram

Tp Tp. Tp
nee ..... (6.116)

~PTG(S)=( 1+ST1 l~Xv (S) rin

TG
g.n ..... (6.117)

As before,
STTG ~PTG(S) = ~xv(s) - ~PrG(s)
and in time domain
et
..... (6.118)

X2 (t) =_I-x4(t) __I_ X 3 (t) ..... (6.119)

TTG J-ro
Finally, from the block diagram

~X v (s) = (_1) [U(S)-~M(S)]

I+ST5 R
..... (6.120)

1
i.e., STs ~(s) = -~ ~ (s) + U(s) - - M(s) ..... (6.121)
R
In the time domain

..... (6.122)

Downloaded From : www.EasyEngineering.net

Load Frequency Control 227

Putting eqn. (6.113, 6.116, 6.119 and 6.122) in matrix from, the state variable model is
obtained as

0 0 0
XI (t) 0 0
-I Kp
0 0 0 kp
x2 (t) Tp 1p [ x2
Xl
-T
= __1_ _1_ + 0 + p d
x3 (t) 0 0 x3 0 ..... (6.123)
TTG TTG I
x 4 (t) __1_ I
x4 Ts
0 0 0
RTs -Ts

redefining the state and control variables.

ww XI = Xl

w.E
"2 = x 2
x3 =
,
+ LlPD
asy
X3

x4 = X4 + LlPD

En
U + u' + LlPD

Eqn. (6.112) can be reduced to

..... (6.124)

.'
gi 0 1
1
0
Kp
0
nee 0

[~} rin
Xl 0 0
Tp 1Tp 0
X2
= 0 I u' ..... (6.125)
0 0
X3
x4
.'
0
Tr
0
Tr
1
I
Ts g.n
RTs Ts
The optimal control to minimize the performance index given in equation (6.111) can be
determined using the solution technique described for the linear regulator problem. It is required
et
to solve the algebric matrix Ricoati equation.
Q + ATR + RA- RBp-l BT R =0 ..... (6.126)
for the elements of the R matrix which is positive defmite and symmetric. Kleinman's method
may be used to solve the matrix Riccati equation. An initial feedback gain vector KI is selected
such that the matrix ~ = [A - B K I] has eigenvalues with negative real parts. Then the matrix
equations.
Ai RT + ~ Al + Q + ~T P KI = 0
are to be solved for the elements of RI" The new gains are computed using
K(l) = p-l BT R
I I
The procedure is repeated till convergence is obtained for the elements of R I .

Downloaded From : www.EasyEngineering.net

228 Operation and Control in Power Systems

The optimal control

UO =- p-l BT RX(t)
= - KT X(t) ..... (6.127)
can be calculated.
E 6.12 For the single area system with the following data, determine the optimal control.
T p = 0.04s R = 2Hzlp.u. MW
TT = O.Ss Kp = 100 Hzlp.u. MW
Ts=O.IS Pn = 0.01 p.u.

ww Q[~ ~1
w.E
Assume

Solution:

a syE
Substituting the parameters in the algebraic matrix Riccati equation and solving the equation
using, Kleinman's method with an initial vector of K j = [1 1 1 1]
The values of Ki converge to
K = ngi
[1.0000 1.1368 1.7092 0.2976]
The optimal control is shown in Fig. E 6.12.
nee
8
7 r ing
6

5 . net
r
';'
4

::: 2
~
0
::s
0

-1

1.0 2.0 3.0 4.0

o
Time (s) ---.

Fig. E 6.12 Optimal control

Downloaded From : www.EasyEngineering.net

Load Frequency Control 229

6.27 Optimal Control for Tandem Compound Single Reheat Turbine- Generator
System
The model for tandem compound single reheat turbine is discussed in
Chapter 3 : The system is shown in Fig. 6.27.

ww
w.E
asy
Fig. 6.27 Tandem compound single reheat turbine system

En
The state - variable model of this system is given by

x = AX+Bu+ Fd
with the initial conditions X(O) = 0
gi nee
where XT =
= (f~f M
[Xl X2 X3
~PG
X4

~PRH
Xs
r
~PCH
X6]
~XV ] ing .... (6.128)

0
0
1
1 Kp
0 0
0
0
0
0
0
. net
Tp Tp
1 FHP
0 0
Tco TJ T2 TCH
A= 0 1 1
0 0 0
TRH TRH
1
0 0 0 0
TCH TCH
1
0 0 0 0
Ts

The constant T 1 and T2 are given in chapter 3.

1 F FIP FIP
- = -LP
-+-----
Tl Tco Tco TRH

Downloaded From : www.EasyEngineering.net

230 Operation and Control in Power Systems

BT = [0 0 0 0 0 ;s 1
FT = [0 - ~: 0 0 0 0 1
u=APc and d=AP D

ww The above equations are transformed into the form

Xl =AX I +Bu l

w.E
using the transformation

asy 0
0

En X=X'+
d

gi d
d
d nee
and u~1 + d
A quadratic cost function r ing
= r ±(X'T.QX· +u'TRu'~t . net
is selected and the algebraic matrix Riccati equation is solved as before to obtain the
optimal gains using the equations.
Q-RB p-l BTR + RA +ATR= 0
and u opt = _p-l BT RX
=_LTX ..... (6.129)
The results are plotted in Fig. E.6.
E 6.13 Compute the optimal load frequency control for a single thermal power system with
tandem compound single-reheat steam turbine with the following data
R = 20s TRH = 7.5s Kp = 2 p.u. Hz / p.u. MW
Ts = O.ls Tco = 0.45s Tp = 20s
TCH = 0.25s FHP = 0.3 D = 0.01 p.u.
FJP = 0.4 FLP = 0.3

Downloaded From : www.EasyEngineering.net

Load Frequency Control 231

Plot the variations of the controlled states with time

Solution:
Q and P matrices are selected as follows

1 0 0 0 0 0
0 1 0 0 0 0
0 0 0 0 0 0
Q= ; P=1
0 0 0 0 0 0
0 0 0 0 0 0

ww 0 0 0 0 0 0
The solution is obtained using a computer program for the algorithm explained

w.E
in 6.26. The feedback gains, starting from initial values of unity for each, converge to the
following solution.

asy LT = [1.0 0.6617 0.0269 0.3037 0.0047 0.0177]

Plots of fM and ,M, the two optimally controlled states are shown in Fig. E. 6.13 (a)
and (b)
En
I x 10-3
gi o 5

Time (5)
10

nee
15

....
..::: -2 x I0- 3 r ing
1 -3xI0-3
. net
Fig. E6,13 Optimal response
(a) Optimal response of time error to step load change

-4xI0-4 ,.-------------------

....~t 5
Or------+-~---_r-------+_--~-
10

+
-4xI0-4
15

-8x 10-4

-12x 10-4

Fig. E6.13(b) optimal response of frequency error to step load change

Downloaded From : www.EasyEngineering.net

232 Operation and Control in Power Systems

6.28 Optimal Control of Hydro Speed Governing System

Consider the system shown in Pig. 6.20. The state variable mode for the system is

X =AX +Bu+ Cu + P.d ..... (6.130)

with the initial conditions XeO) = 0
where XT = [XI X 2 X3 X 4 Xs X6 ]
= U~f ~f ~PG ~PGV ~Ppv ~ppv]
0 1 0 0 0 0
1 Kps

ww A= 0
0
Tps
0
Tps
---
1
0
1
0

---
0

w.E 0 0
O.5Tw
0
0. 5Tw
0
0. 5Tw
1 0

asy 0
0
0
-KI
0
-K2 -K3
0 0
-K4 -Ks

E BT =[0 0 0 0 0
ngiTpT~TG 1
T
C =[0 0 0 0 0 nee
Tp~Gl
_ Kps r1in
pT = [0

u = ~ Pc = control input
Tps
0 0 0 Kps
Tp TR Tps
g. net
and d = ~ PD = step-load disturbance.
The constants Kl' K:z, K3 K4 Ks are given by
- Tps -TR
K1-
TpTGTRTpS
_ Kps
K2 -
Tp TG Tps
cr
K 3 =---
TpTGTR
K4 = TR(cr+o)+TG
. TpTGTR
1 1
Ks =-+-
Tp TR

Downloaded From : www.EasyEngineering.net

Load Frequency Control 233

The state and control variable are transfonned using the relation,

Xl
0
Xl
X 0
X2 2
X d
X3 3
= + d ..... (6.131)
X4 X
4
0
X5 X5 u'
X6 X6 TpTG

ww U = u' + d
The transfonned equations are
(j

w.E j( = A'X' + B' .u'

where
a[0sy0 0 0 TP~J
A'=A

and
En
B'T =

gin
The cost functional to be minimized is selected as

J=
eer
~ r(X'TQX' +u'Tpu')dt

in
Selecting Q and R matrices are :

1 0 0 0 0 0
g.n
Q=
0 1 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
et
0 0 0 0 0 0
0 0 0 0 0 0

andP=l
the algebraic matrix Ricccati equation
Q = R B' p-l B'T R + R A' +A'T R = 0
is solved for the elements of R.
The optimal control is obtained using
u = _p-l B'T R X'
opt
=_LTX ..... (6.120)

Downloaded From : www.EasyEngineering.net

234 Operation and Control in Power Systems

The states with optimal control are determined from

..... (6.133)

E 6.14 For the hydro system shown in Fig. 6.14.1; using the following data, compute the
optimal control

Distributor volve and

Pilot volve and
servomotor

ww Vet min Turbine and Power system

w
penstock

.Ea
syE
Permanent speed droop

ngi Speed governing system

nee
Fig. ES.14.1 Hydro speed governing system

o 4 8
Time (s)
12
--+ r 16 20 ing
. net
~ -0.004
~
5 (i) Optimal response
..,l: -0.008
.., (ii) Uncontrolled response
E
E=
-0.012

1 -0.016

(a)

Fig. ES.14.2 Uncontrolled and optimal response of hydro speed governing system
(a) uncontrolled and optimal response of time. error

Downloaded From : www.EasyEngineering.net

Load Frequency Control 235

TG = 0.2 s KPS = 2 p.u. Hz! p.u. MW

Tp = 0.04s Ll PD = 0.01 p.u.
Tw = 0.6 s
Tps = 20s
TR = 3s
(J = 0.05

8 = 0.2
The 100 % load conditions on the plant.
.,

----+
ww
Time (s)
0.0004

w.E
0.0004

t
E
0.0008

0.0012 asy
:::-
'-
<l 0.0016
En (i) Optimal response
(ii) Uncontrolled response

+ 0.0020

0.0024
gi nee
r ing
Solution:
Fig. E6.14.2 (b) uncontrolled and optimal response of frequency error
. net
Following the procedure outlined in Sec. 6.20, the optimal feed-back gain components are
computed as
LT =[1.0 3.4061 0.078 0.5393 0.1656 0.0063]
The uncontrolled and optimal response are shown in figure E6.14.2 (a) and (b).

6.29 A Review of Optimal Control

The optimal control determined by eqn. (6.116) is quite often impractical due to the following
reasons:
1. The optimal control is a function of all the states of the system. In practice, all the
states may not be available. The inaccessible states or missing states are required to be
estimated.
2. It may not be economical to transfer all the information over long distances.

Downloaded From : www.EasyEngineering.net

236 Operation and Control in Power Systems

3. The control which is a function of the states in turn is dependent on the load demand.
Accurate prediction of the load demand may be essential for realizing the optimal controller.
4. The optimal control is also dependent on the weighting matrices in equation (6.111) and
is thus not unique.

6.30 Load Frequency Control with Restrictions on the Rate of Power Generation
The optimal load frequency control discussed in section 6.25 does not include the effect of the
limits on the rate of change of power generation. If these limits were not included in the
control, there will be a tendency for the system to chase large momentary disturbances.

ww
This causes undue wear and tear of the controller. Even a telemetering outage may
cause problems to the system in operation. Several methods are proposed to include the effect
of generation rate constraints in the computation of the control.

w.E
If the generation rate constraints denoted by P are introduced into the state vector, the
system order will be changed. Instead of augmenting them, it is suggested that while solving

asy
the state equations, it may be verified at each step whether or not the generation rate constraints
are violated.

E ngi
Another simple way of considering the generation rate constraints is dual mode control
strategy. Separating the state vector X into those states which are to be rate limited as XI and

nee
the rest by xr, the state space representation of the system becomes

rin ..... (6.134)

the generation rate constraints are given by

g. net
..... (6.135)

max
That is IAn Xd+AIr Xr +BI u~ XI

The choice of U(t) may be made using

max
Blu(t) = KXI -AnXI -AJrX r ..... (6.136)

where K is a ( m X m) diagonal matrix with sgn. (Xj ), j = 1, 2, ..... m and m is the

number of rate limited state variables out of n state variables. The dual mode control may be
represented by eqn. (6.134) and (6.136)

Downloaded From : www.EasyEngineering.net

Load Frequency Control 237

max
UO(t) = -K,X; ifx'J < Xlj j = 1,2, ............ ,n ..... (6.137)

= -Bi' [ Kx,max -LX 1 ;ifxiJ >

max
X'j ;j = 1,2,...........n ..... (6.l38)

The generation rate constraints result in larger deviations in area control errors. As the
rate at which generation can change in the area is constrained by the limits power import via tie
lines becomes imperative. Under generation rate constrained conditions, the selection of governor

ww
speed regulation cofficient R requires careful consideration. In practice, a low value, of the
order of2 to 4%, is chosen for R. With ,a proper, supplementary control, the steady state error

w.E
can be reduced to zero, whatever may be the value ofR. However, it is desirable that a proper
value of R be selected so as to give the best dynamic response. Improper selection of R may
lead to instability whatever may be the integral controller gain settings.

asy
In systems with hydro-thermal combination, the generation rate in the hydro area

En
generally remains below the safe, permissible generation rate, and as such the rate constraints
for generation at all the hydro plants can be omitted.

gi
The presence of governor dead-band introduces oscillations in the dynamic response. It

nee
has been reported that the governor dead band does not influence the selection of integral
controller gain settings in the presence of generation rate constraints.

6.31 Load Frequency Control using Output Feedback r ing

.
The optimal control derived in Sec. 6.22 requires the availability of all the states. However, if
some of them are not available, reconstruction of those states requires either the Lueneberger
observer or Kalman filter. This may not be feasible for various reasons. Controllers based on
net
output feedback are proposed to overcome this problem. It has been pointed out later in chapter
7 that area control error for each area is used to implement the tie- line bias control strategy.
Correspo_~ding to this practice, if the area control error is taken as the output, then the load
frequency control problem may be considered as an output zeroing problem in the presence of
persistent disturbances.
An additional output equation
Y=HX
is required to be considered in conjunction with the system Eqn. (6.134). The design of output
dependent controller is not an easy task as it leads to an equivalent parametric optimization
problem. Nevertheless, making use of matrix minimum principle, the controller can be designed.
To simplify the computation, minimum norm methods are proposed.

Downloaded From : www.EasyEngineering.net

238 Operation and Control in Power Systems

6.32 Load frequency control ana J!.conomlc dispatch

Whenever load changes the initial frequency correction is achieved by the speed governor and
this control can be treated as primary load frequency control. The adjustment offrequency by
the primary control loop may take a few seconds.
After the speed governor response is over, the steady state frequency error is reduced
to zero so that system frequency is maintained constant by the integral controller action. This
will be after the primary control action is over and can take a time period up to one minute.
This i~the secondary control. The adjustment of frequency error to zero by changing the
gener ion schedules that are determined earlier by economic criterion again requires

ww
readju tment of generation. This tertiary control can be implemented by using economic
dispat h computer which works on the cost characteristics of various generating units in the
area. The speed changer settings are once again operated in accordance with economic dispatch

w.E
computer programme.

asy
En
gi nee
rin
g.n
et

Downloaded From : www.EasyEngineering.net

Load Frequency Control 239

QUESTIONS

6.1 Explain the necessity of maintaining a constant frequency in power system operation.

6.2 With a neat diagram, explain briefly different parts of a turbine speed governing system.

6.3 Derive the model of a speed governing system and represent it by a block diagram.

6.4 With a block diagram explain the load frequency control for a single area system.

6.5 Derive the model of a speed governing system and represent it by a block diagram.

ww
6.6 With first order approximation explain the dynamic response of an isolated area for load
frequency control.

w.E
6.7 Discuss the importance of combined load frequency control and economic dispatch
control with a neat block diagram

asy
6.8 Discuss in detail the importance of load frequency problem.

E
6.9 Distinguish between load frequency control and economic dispatch control.

ngi
6.10 A synchronous generator supplies power to a synchronous motor via a transmission

nee
network. Find equivalent inertia constant of a machine connected to infinite bus.

6.11 Explain how modem control theory can be applied to load frequency control

rin
6.12 Describe how optimal control can be determined in case of LFC problem.

6.13 What are the limitations of optimal control theory ? g. net

Downloaded From : www.EasyEngineering.net

240 Operation and Control in Power Systems

PROBLEMS

P 6.1 Two generators rated 200MW and 400MW are operating in parallel. The droop
characteristics of their governors are 4% and 5% respectively from no load to full
load. Assuming that the generators are operating at 50Hz at no load, how would a
load of 600Mw be shared between them? What will be the system frequency at this
load? Assume free governor operation. Repeat the problem if both the governor
have a droop of 4%.
P 6.2 A 100MVA asynchronous generator operates on full load at a frequency of 50Hz.
The load is suddenly reduced to 50MW. Due to time lag in the governor system, the

ww steam valve beings to close after 0.4 secs. Determine the change in frequency that
occurs in this time. Given H = 5 KW -s/KYA of generator capacity.

w.E
P 6.3 Two generators rated 200Mw and 400MW are operating in parallel. The droop
characteristics of their governors are 4% and 5% respectively from no load to full

asy
load. The speed changes are so set that the generators operate at 50Hz sharing the
load of600Mw in the ratio of their ratings. If the load reduces to 400Mw, how will

E
it be shared among the generators and what will the system frequency? Assume

ngi
free governor operation. The speed changers of the governors are reset so that the
load of 400MW is shared among the generators at 50Hz in the ratio of their ratings.

nee
What are the no load frequencies of the generators?
P 6.4 In the single area system shown below determine

rin
(a) The steady state frequency error with t.P c = 0
(b) Critical gain K of the integral control of t.P c = - fKM
g.n
et

P 6.5 A 500Mw generator is operating at ~ load of20Mw. A load change of I % causes the
frequency to change by 1%. If the system frequency is 50Hz determine the value of
load damping factor in per unit.

• Downloaded From : www.EasyEngineering.net

Downloaded From : www.EasyEngineering.net

ww
7 w.E
CONTROLOF
asyE SYSTEMS
INTERCONNECTED
ngi
nee
rin
Power Systems came into existence in 1880s and from that time onwards the systems have
grown enormously in both size and complexity. For better performance and reliability of

g.n
operation and control, there were many significant developments in generation, transmission
and distribution. The concept of energy control centers emerged in 1970's. Computer aided
analysis and computer based control have been proposed in this context.

7.1 Interconnected Operation

et
Power systems are interconnected for economy and continuity of power supply. For the
interconnected operation incremental efficiencies, fuel costs. water availability, generation limits,
tie line capacities, spinning reserve allocation and area commitmen'ts are important considerations
in preparing load dispatch schedules.
In this chapter the power control of interconnected system is presented.

7.2 Flat Frequency Control oflnterconnected Stations

Consider two generating stations connected by a tie line as in Fig. 7.I(a). For a load increment
on station B, the kinetic energy of the generators reduces to absorb the same. Generation
increases in both the stations A and B, and frequency will be less than normal at the end of the
governor response period Fig. 7.1(b). The load increment will be supplied partly by

Downloaded From : www.EasyEngineering.net

242 Operation and Control in Power Systems

A and partly by B. The tie line power flow will change thereby. If a frequency controller is
placed at B, then it will shift the governor characteristic at B parallel to itself as shown in
Fig. 7.1(c) and the frequency will be restored to its normal value fs' reducing the change in
generation in A to zero.

feN)
f(N)

ww A
• B

~
1
,iPA
1
wE
I.- . -J
1
,iPB
1
14-
1

I...
~_..J.-_ P
asy-+---'--~

o
~---t-- P -t---+---t~

E
(b)
ngi
{}

(c)

(a)
(b)
Two interconnected stations
nee
Uncontrolled system with load increment on Station B
(c) Frequency controller located at Station B
rin
Fig. 7.1 Two station system
g.n
et
If the load increment comes on station A, then as before, initially the generation in both
A and B changes to absorb the additional load, while finally the additional load is absorbed by B
only. Station A absorbs none of its load changes in the steady state. It is possible that, in
interconnected operation, a given station can be made to absorb the load changes occurring
elsewhere in the system so long as the controlling station has capability to absorb the change.
The same analysis can be extended to a two area system.
Assumption in Analysis:
The following assumptions are made in the analysis of the two area system:
1. The overall governing characteristic of the operating units in any area can be
represented by a linear curve of frequency versus generation.
2. The governors in both the areas start acting simultaneously to changes in their
respective areas.
3. Supplementary control devices act after the initial governor response is over.

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 243

The following time instants are defined to explain the control sequence:
to is the instant when both the areas are operating at the scheduled frequency and
tie line interchange and load change takes place.
t l is the instant when governor action is initiated at both A and B.
t2 is the instant when governor action ceases.
t3 is the instant when regulator action begins.
t4 is the instant when regulator action ceases.
Consider a load increment in area B. From Fig. 7.2(a), it is clear that at the end of the

ww
governor response, the tie-line schedule is upset and frequency is less than normal. If now a
frequency controller is provided in area B, which shifts the governor characteristic upwards,

w.E
parallel to itself, so as to provide the required control action, generation in B meets its own load
change. Tie line schedule is maintained. Change in generation in A is also reduced to zero.

a
Consider now the controller action to load change in area A with the controller located in

syE
area B as before. The response is shown in Fig. 7.2(c).
f f

ngi
Load change
inB
Load change
inB

f nee
rin
g.n
PA 0 PB PA 0
I I I
I I I
I I
et
PB

t4 - -
I I I
t3 - -
--- --1-
t2 - -
t3 - -I-
tI - - -
t2 - -I-
--- -'"'-
to ~
To A ~ To B
To A ~ ~To B
(a) (b)
Unregulated case Controller at B

Fig. 7.2 Flat frequency control

(a) Load increment in area B - no controller

(b) Load increment in area B and frequency controller in area B

Downloaded From : www.EasyEngineering.net

244 Operation and Control in Power Systems

--------------------------~~----
f Load change
inB

o I PB
I I I
ww -,- --j-
I I I

w.E -,-
asy To A ~ 0 ~ To B

En The line flows

gin
While the initial governor response is the same as for the previous case, the action of the
controller in B will force the generation in area B to absorb the load increment in area A. When

eer
the controller begins to act at t 3, the governor characteristic is shifted parallel to itself in B ti II
the entire load increment in A is absorbed by B and the frequency is restored to normal. Thus,

ing
in this case while the frequency is regulate<i on one hand, the tie-line schedule is not maintained
on the other hand.

.
If area B, which is in charge of frequency regulation, is much larger than A, then load
changes in A will not appreciably affect the frequency of the system. Consequently, it can be
said that flat frequency control is useful only when a small system is'connected to a much
net
larger system.

7.3 Flat Tie-Line and Flat Frequency Control

Since a flat frequency controller cannot provide the required control, let us consider another
controller, a tie line power flow controller located in the area.
As an example, let a frequency controller be provided in A and a flat tie-line controller be
located in area B. The following four possible cases are considered according to the sequence
of operation of the controllers in areas A and B.
1. Load change in A: Sequence of operation A, B, A, .. .
2. Load change in B: Sequence of operation B, A, B, .. .
3. Load change in A: Sequence of operation B, A, B, .. .
4. Load change in B: Sequence of operation A, B, A, .. .

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 245

Case 1 : Let a load increment take place in area A. It is assumed that following the initial
governor response, the frequency controller in A acts first at t3 and its control action
ceases at t4 . By the time the tie-line controller at B becomes effective at t5, it finds the
system in normal condition. both frequency and tie line deviations are reduced to zero,
and further regulating action is unnecessary. This action is shown in Fig. 7.3(a).
f

ww
w.E
asy
PAl Flat 0 Tie -line 1 PB
I frequency frequency 1
controller controller

ts-llii-l------Engi
t4 -
I 1 1

-l--+-----
______ L __
----- +---
-----1-
1 1

t3- ,
t2 -T
t1 ----
-,---
-T---
-----
-
- nee
-----1-
to--~~--~------------~----------~--~----

To A ~ 0 ----I.~ r
To 8
ing
.
Fig. 7.3 Flat time -line and frequency control- frequency controller in area A and tie-line
controller in area B.
(a) Load increment in area A - sequence of operation A - B - A - etc.
net
Case 2: Let the load increment occur in area B. The tie - line controller in B acts first
following the initial governor response at t3 and increases the generation in B ti II the load
increment is completely absorbed. At the instant t5, the frequency controller in A finds
that the conditions are normal and no regulating action is needed. The load increment in
B is absorbed by the generation in B itself. The case is illustrated in Fig. 7 .3(b).
Case 3 : Load change occurs in A and the controller in B acts first. The initial governor
response which begins at the instant t1 ceases at t2. The tie line controller in B begins
and completes its action before the frequency regulator at A takes control. During this
period of control action, i.e. t3-t 4 by the tie line regulator the frequency decreases further
than it was at the end of the governor action period while the tie line power schedule is
maintained.

Downloaded From : www.EasyEngineering.net

246 Operation and Cont-:ol in Power Systems

Flat o Tie -line I

I frequency frequency

ww I I controller controller
I

~:==~I
I I I
1======
w ::J------ =====llS'
t3 - -
t2- -
t)--+- .Ea -i------
~-----
-
- ----+
-----
---.1 _L
-I-
~-t-

syE
to----~-4------------~------------~--+--+
To A ~ 0 ------i~~ To B
Tie-line flows

ngi
Fig. 7.3 (b) Load increment in area B - sequence of operation B-A-B-etc.

nee
At t5 the frequency controller in A takes over and restores the frequency to its normal
value. It can be seen from Fig. 7.3(c) that the tie line schedule is again upset but by a
larger margin than before. Considering subsequent cycles of control action by the

rin
controllers in B and A, it can be seen by similar reasoning that stable operating conditions
cannot be attained for the sequence of regulator action conceived in this case.
f
g. net
®

Fig. 7.3 (c) Load increment in area A - sequence of operation B-A-B-etc.

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 247

Case 4 : Finally consider the load increment to have taken place in area B. Let us assume that
the frequency controller in A starts and completes its action the tie-line - controller in B starts
functioning. The response to governor and controller action is showf.l in Fig. 7.3(d).
f

ww
w.E
a syE
ngi
nee
To A . - - 0 ----i.~ To B
rin
Fig. 7.3 (d) Load increment in area 8 - sequence of operation A-8-A etc

g.n
The flat frequency controller at A incre?,ses the generation in A to correct the frequency
by shifting the characteristic parallel to itself upwards; The generation in A and hence the tie-
line power from A to B increases. At ts' when the tie-line controller takes over the control, it
will find excessive power flow from A to B and to keep it at the scheduled value, it increases
et
the generation in B. While doing so, the generation in A gets reduced and the tie line power flow
to B is also reduced. Finally, when the frequency controller begins to act in A at t8 neither the
tie-line power schedule nor the frequency is normal. The system starts oscillating and no
regulation is feasible.
It can be seen from the above analysis that if one of the regulators is sluggish in any area
where changes in power demand take place, than other regulators, flat tie-line flat frequency
regulator strategy cannot assist in getting the desired control.

7.4 Tie-Line Bias Control

The regulators described in Sec. 7.3 can attain the desired control in so far as the local load
changes are considered. However, they fail to respond to changes in remote areas. One way to
overcome this problem is to make the flat tie-line controller responsive to frequency changes
also. That is, the flat tie-line controller is to be biased so that we obtain a tie-line bias controller.

Downloaded From : www.EasyEngineering.net

248 Operation and Control in Power Systems

With this type of controller, the regulator does not act to hold a constant tie-line interchange
regardless of system frequency, but instead allows the tie line schedule to deviate from the
normal in adjustable proportion to frequency deviations from the rated value. In principle, the
tie line bias controller in an area endeavours to take action for changes in load in its own
precincts, while taking no cognizance of the same for other interconnected areas. I

The characteristics of tie-line bias controller and flat frequency controller are shown in
Fig. 7.4.

i Lower i Lower generation

ww
generation
Raise
Raise generation
generation

w.E 0 p~ 0 p~

asy
(a) Tie-line bias controller (b) Flat frequency controller

Fig. 7.4 Tie - line bias and flat frequency controller characteristics

En
The point of intersection of both the characteristics is the only point of operation where

gi
the system, frequency and tie line schedule can be maintained at the normal values. As in the

nee
previous control strategy, all the four possible cases of operation are discussed.
Let area A be equipped with a flat frequency controller and area B be provided with a
tie-line bias controller.
r ing
Case 1 : Consider a small load increment in Area A. Let the flat frequency controller in
area A act first.
. net
The response to governor and controller action is same as that shown in Fig. 7.5(a).
Following the initial governor response, the frequency controller in A begins and completes
its action before the tie-line bias controller at B comes into operation. Both the frequency
and tie-line power schedules are maintained at scheduled levels by the controller in A.
No further corrective action is needed by the tie-line bias controller in B.
Case 2 : Let the load increment take place in B with the tie-line bias controller in B acting
first. The response to governors and controllers would again be the same as given in
Fig. 7.5(b). Following the initial governor response, the tie-line bias controller in B acts
during t3 - t4 an9, by shifting the governor characteristic in B, both the tie line power
and frequency deviations are corrected. No further regulating action is required from
the controller in A.
Case 3: Let the load change occur in area A. The controller in area B acts first, followed
by the frequency controller in area A. If the governor characteristic is prevented from
being shifted by the action of the tie-line regulator in B during the interval t3 - t4 the

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 249

frequency controller in A can restore the frequency to its normal value during the period
t5 - t6 so that the controller at A acts for changes in load in area A only and absorbs all of
them. This can happen if the slope of the tic-line bias controller in B is same as the slope
of the governing characteristic at A at the point of operation under consideration. Under
such a condition, the response to controller action is shown in Fig. 7.S(a).
f

ww I I o
w.E -l-~---
-~-

asy ---1-
--1-
t2--

E-,-
tl ----
to ngi
To A ~ 0 --i~~ To B

(a) Load increment

nee
Fig. 7.5 Tie-line bias control in area A and tie-line bias controller in area B
area A - controller A acts first

rin
In In

g.n
Case 4 : Finally, let the load increment occur in B while the controller in A initiates action
first following the governor action. It r.an be seen from Fig. 7.S(b) that during the

et
period tl - t2, the governors in A act following the load increment in area B. The frequency
falls and the tie line power flows from A to B. During the period t3 - t4 , the flat frequency
controller in A acts, generation in A increases and more power flows to B via the tie-line.
The frequency is normal. During t5 - t6, the tie-lil}e bias controller in B acts to set the tie-
line deviation to zero. The governor characteristic is raised in B. Frequency increases
but tie-line flow from A to B is reduced. Neither the frequency nor the tie-line schedule
are normal at this stage. However, during the period t7 - t8, the flat frequency controller
in A again takes over the control and the governor is set at a lower value in A. Generation
in A is reduced until the frequency is normal and tie-line power flow deviation is zero.
From the above analysis it can be inferred that each controller operates effectively to
control load changes in its own area. For load changes in other interconnected areas,
the tie-line bias controller can be set to refrain from taking improper action while the
frequency regulator still persists in its tendency to sacrifice the local area requirement to
overall requirement.

Downloaded From : www.EasyEngineering.net

250 OP.eration and Control in Power Systems

Since the frequency controller in A responds unnecessarily to changes in load in B, it may

be replaced by another tie-line bias controller. This results in complete tie-line bias control.
o--- __ - f
®

ww
w .Ea
syE
To A ~ ---t~~ To B
7.5 (b) Load increment in area 8 - controller in A acts first

7.5 Complete Tie-Line Bias Control ngi

nee
With tie-line bias controllers located in both the areas the control action is complete in all the cases.
For the case w1lere the tie-line bias controller in A initiates action for local load changes,

r ing
the response is shown in Fig. 7.6(a). The response to controller action is similar for load
changes in B when the controller in B initiates regulating action.
f
. net
o
I I

t4_
t3__
+_1. __ _
I I

--1----
t 2 __ --t---
tJ ----
to----~~------~-----~-~--
~ 0 ~ To 8
Tie-line flows
(a)
Fig. 7.6 Complete tie-line bias control - tie line controllers in both the areas A and 8
(a) controller action in A for local load changes

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 251

For the case when a controller initiates action for non-local load changes the controllers
will not act if the slope of the controller characteristic is matched to that of the governing
characteristic. The response characteristics are illustrated for a load change in area A while the
controller in B is the first to initiate action Fig. 7.6(b).
During this period the controller in B is prevented from regulating the system. The
. incremental changes are corrected by the controller in A only, during t5 - t6 . The control action
is quite satisfactory. From the above analysis, it is seen that the governor action in an area is
not opposed by supplementary control action in the same area.
f

®
ww
w.E
asy
E ngi
to----~~------~----~~---
nee
A"-- 0

rin~B

Tie-line flows
Fig. 7.6(b) Load change in A - controller in B acts first
g.n
et
The governor action is not changed till the controller in the area, where a load change
has occurred, becomes effective. This avoids unnecessary change in generation, frequency or
tie-line power. The controller in the area, where load change occurs, acts in such a manner that
the area absorbs its own load change. Only a single shift is necessary to the governor
characteristic to restore both frequency and tie-line power to normal. A smooth, cooperative
regulation is thus achieved with a tie-line bias-tie-I ine bias control scheme for the two area system.
For successful operation, it is imperative that the control characteristic of each tie-line
bias controller must have the same slope as the governor characteristic of its own area.
Fig. 7.7 shows the three types of controller characteristics for comparison.
It can be seen from Fig. 7.7 that flat tie-line control characteristic and flat frequency
control characteristic are only limiting cases of tie-line bias control characteristic with zero
bias and infinite bias respectively. A tie- line bias controller with larger bias (less slope) compared
to its area governor charlatanistic will be more sensitive to load changes in the outside system
than the governor and acts like a flat frequency controller.

Downloaded From : www.EasyEngineering.net

252 Operation and Control in Power Systems

I
f f f I Lower gen
Raise Lower Lower gen
gen gen
Raisegen Governor
gen characteristic
'----..:...----. p ' - - - - - - - -.. P
A+-- ~B ~p A~ ~B

(a) Flat tie-line controller (b) Flat frequency controller (c) Tie-line bias controller

Fig. 7.7 A comparison of controller characteristics

ww If the tie-line bias controller has larger slope (smaller bias) than the governor

w
characteristic, it acts like a flat tie-line controller. It is therefore necessary to match the speed

.Ea
governor characteristic and the tie-line bias control characteristic for satisfactory control.

syE
ngi
nee
r ing
.
In I' I"

Fig. 7.8 Mismatching of tie-line bias control characteristics

Further, as the tie-line bias controller acts to correct the deviations in the tie-line power
net
flow schedules, the tie-line interchange must be metered accurately and then a correcting
strategy be applied.
If the tie-line bias controller characteristic is not matched to the concerned area governor
characteristic, the effect could be seen from the following analysis.
Let the governor characteristics of areas A and B be represented by curves I and 2 as
shown in Fig. 7.8. The tie-line bias controller characteristic is assumed to coincide with the
governor characteristic in area A (i.e. curve I) but the characteristic of the controller in area B
is assumed to be curve 3.10 is the actual tie-line power flow. Due to mismatching of the curves
in B, the interchange increases to " and the frequency falls from the synchronous frequency
fs to f '. In order to control the frequency, the characteristic' is shifted upwards to I' and the
curve 3 to 3'. The inter-change increases to I" while the frequency is regulated. After this
discussion mathematical analysis of the two area system will b~ now presented.

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 253

In chapter 6, single control area and its response to changes in load, both uncontrolled
and controlled is discussed. Let us consider now two control areas connected by a tie-line.
Each control area will have the same frequency and the same frequency deviations for the
entire area. The tie-line is considered weak in the sense that both the areas can have frequency
deviations different from each other in uncontrolled mode (i.e) M, and f1f2 respectively.

7.6 Two Area System - Tie-Line Power Model

Tie line (X)

Area Area
I 2

ww I'D
12

w .Ea
Fig. 7.9

Consider two inter connected areas as shown in figure operating at the same frequency
fl while a power
syE
P,~ flow from area I to area 2 let IV,I and IV 21 be the voltage magnitudes at

8? and 8~ voltage phase angles at the two ends of the tie-line while P,~ flows from area I to
area 2 then,
ngi
pO
12 -
_IVlllv2ISin(~o
X \u,
-(0)
2
nee ..... (7.1 )

where X is the reactance of the line.

r ing
If the angles change by f1o,and f10 2 due to load changes in areas I and 2 respectively.
Then, the tie-line power changes by

f1P12 =
Iv,ollv~1
X
0 0
COS(OI - oJ (f1P, - f1P 2 )
. net
..... (7.2)

f1P 12 f1P 11
Defining --'-=--- = ---
MW/radian ..... (7.3)
f10 1 -M2 M
as synchronizing coefficient of the tie-line or "stiffness coefficient" of the line, denoted
by~
..... (7.4)

consider f1(j) = ~f18

(i.e.) 21tM = ~f1o

dt
I d
M=--f18Hz
21t dt

Downloaded From : www.EasyEngineering.net

254 Operation and Control in Power Systems

In otherwords

L\8 = 21t fM dt rad ..... (7.5)

Hence L\8 1 = 21t fMI dt rad

and L\8 2 = 21t fM2 dt rad

From equation (7.4)

LlPI2 = 21tT°{JM1dt - fM 2dt} MW ..... (7.6)

Taking Laplace transform on both sides

ww 21tT o
L\P12 (S) = - - [L\FI (S) - L\F2 (S)] ..... (7.7)

w .Ea
S
Block schematic of the above equation is shown in Fig. (7. I 0)
If the two areas are rated at Prl and Pr2

then if
P
_rl
P
syE
=a
12
r2

ngi
nee
r ing
_---:--,I-_ _ _ _ -~--.--:--I~~
.L\P 1(s)
net
. . t------.,li..r--I
.
L\P 12(S) ~-
2n:TO

_...J.t__---.~ L\P 2(S)

Fig. 7.10 Block diagram for tie-line power

7.7 Block Diagram for Two Area System

The block diagram for a two area system can now be developed. Each area can be represented
by a block diagram as in the case of a single area system, but with suffixes I and 2. the block
diagram for the tie-line power deviation can be used to inter connect both the areas as shown
in Fig. (7:1 I).

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 255

I
-6T1(s)
RI
6PRefl

ww
w.E
6P Ref2
asy
I
-6T2 (s)
R2
+
E ngi
nee
Fig. 7.11 Block diagram for 2-area systems

7.8 Analysis of Two Area System

Steady State Response: rin
Consider the speed changer positions as fixed so that
g.n
Ll P ref, I = Ll P ref, 2 = 0
et
let the loads in both the areas change by LlP DI = kl and LlP D1 = k2, step changes in the
two area respectively. Let MV be the final or steady state frequency deviation due to load
changes. Simialary LlP102 let be the change in the tie-line power flow.
From the Fig. 7.11

..... (7.8)

..... (7.9)

where LlP~G,1 and LlP~G,2 are the steady state changes in turbine-generator outputs.
Also, from the same Fig. 7.11, at the summing point of load and tie-line powers, we get

°
( -~M -kl -LlP12 Kpi = M
I 0) ° ..... (7.10)

Downloaded From : www.EasyEngineering.net

256 Operation and Control in Power Systems

(i.e.)

since

(-t MO -kl = D1M o J +,1PI~

I
.... b.ll)
I .
I

similarly for area 2 (- ~2 M O

- k 2) =D 2M + ,1PI~
O
..... ~7.12)

ww
solving the above equations (7.11) and (7.12)

we obtain
w.E ,1fo = -( kl + k2
~I + ~2
JHz ..... (7.13)

and
a ApO __ ApO _ ~lk2 -~2kl
Ll

syE
12 - Ll 21 -
~I +~2
..... (7.14)

ngi
1
where ~I =D 1 + - ..... (7.15)
RI

and ~2 = D2 + -
1
R2 nee ..... (7.16)

which are defined as the area frequency response characteristics.

rin
Let both the areas be identical
Then, DI = D2 = D g.n
The deviations become
RI = R2 = R
~I = ~2 = ~ et
,1fo=k l +k 2 Hz ...... (7.17)
2~ ,

and ,1pO __ ,1po _ k2 - kl MW ..... (7.18)

12 - 21 - 2 p.u.

If the area ratings are different and ~ = a 12 then,

Pr2

and
..

Downloaded From : www.EasyEngineering.net

Control 0/ Interconnected Systems 1.57

If a load disturbance occurs in only one of the areas, it is clear that with kl = 0 or
k2 = 0 the frequency derivation ~fl is only half of the steady state error that would have
occurred had there been no inter connection. Thus, with several systems inter connected, the
steady state frequency error would be reduced.
Also from the tie-line power derivation it can be observed that half of the load change in
either area will be supplied by the other area, which demonstrates the importance of emergency
assistance in inter connected or pool operation.

7.9 Dynamic Response

ww
From the block diagram of Fig. (7.11) the following equation can be written

w.E ..... (7.19)

asy
E ngi
nee
rin
~P21 (S) = -~P21 (S)
g.n
There are four equations with four variables, ~fl' ~f2' ~P12 and ~P21 to be determined
for given
involved.
~PDl and ~PD2.
et
The dynamic response can be obtained; even though it is a little bit

For simplicity assume that the two areas are equal. Neglect the governor and turbine
dynamics, which means that the dynamics of the system under study is much slower than the
fast acting turbine-governor system in a relative sense. Also assume that the load does not
change with frequency (D, = D2 = D = 0).
We obtain under these assumptions the following relations

Downloaded From : www.EasyEngineering.net

258 Operation and Control in Power Systems

since D = 0

equations (7.19) and (7.20) simplify to

O
1 ~P12 (S) ] - f =~FI (S)
ww [ - -R ~FI (S) - ~PDl (S) -
2SH
..... (7.21)

w.E1
[ - -R ~F2 (S) - ~PD2 (S) - ~P21 (S) ] - f
2SH
O
= ~F2 (S) ..... (7.22)

a
subtracting (7.22) from (7.21)
syE
~ ~FI (S) + ~R ~F2 (S) -
n
~PDl ~PD2 g~P12i
[- R (S) + (S) -
ne~P21e ~ ~F ~F
(S) + (S)]
2SH
=
I
(S) -
2
(S)

rin
~P12 (S)
° g.
O
1 ] f
[ -R ~F2 (S) + ~FI (S) + ~PD2 (S) - ~PDI (S) - 2~PI2 (S) -2SH = 2nT
net ..... (7.23)

since

Therefore,

~P (S)[ Sfo + 2fo _S_] =[~PD2 (S) - ~PDI (S)]f O

12 2nRT0 2SH 2SH -+I 2nTO 2SH

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 259

ww
w.E ..... (7.~4)

a syE
The denominator is of the form

ngi
(S2 +2KS+ro 2)=(S+K)2 +(ro 2 _K2)

nee
where
rin
setting
g.n
et
..... (7.25)

Note that both K and ro 2 are positive. From the roots of the characteristic equation we
notice that the system is stable and damped. The frequency of the damped oscillations is given
by roo.
Since Hand fO are constant, the frequency of oscillations depends upon the regulation
parameter R. Low R gives high K and high damping and vice versa. IfR ~ a; K ~ 0 is the
condition for no-governor action and there will be undamped oscillations.
We thus conclude from the preceding analysis that the two area system, just as in the
case of a single area system in the uncontrolled mode, has a steady state error but to a lesser
extent and the tie line power deviation and frequency deviation exhibit oscillations that are
damped out later.

Downloaded From : www.EasyEngineering.net

260 Operation and Control in Power Systems

E 7.1 Two control areas have the following characteristics

Area I; Rl = 0.011 p.U
0 1= 0.85 p.u.
Base MVA = 1000
Area 2; R2 = 0.018 p.u
O2 = 0.95 p.u.
Base MVA = 1000

ww A load change of200 MW occurs in area I. Determine the new steady state frequency.
1lution:

w.E
asy kl = 0.2pu ; k2 = 0
I
En
I
PI =0 1 +-=0.85+--=90.9091+0.85=91.7591
Rl 0.011

P2 = O 2 + -
1
R2
gi 1
nee
= 0.95 + - - = 55.55 + 0.95 = 56.405
0.Ql8

~fO = kl = 0.2
(131 +P2) 91.7591+56.405
r ing
=~=0.0013683
148.16 p.u. Hz

f= 50 - (0.0013683 x 50) = 49.93Hz

E 7.2 In example E. 7.1, determine the tie-line power flow deviation.
. net
Solution:

-k 1P2
~Ptle-line = A A
PI +P2
= (~fl) P2 = -0.0013683 x 56.405
= -0.0771789 p.u MW

= -77.1789 MW

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 261

E 7.3 In E 7.1 if the load disturbance occurs simultaneously also in area 2 by 100 MW determine
the frequency and tie-line power changes.
Solution:
K2 = 0.1 pu

MO = kl +k2 = 0.2+0.1 =0.0020248puHz

~I + ~ . 91.7591 + 56.405
~f>= 0.10124 Hz
f= 50 - 0.10124 = 49.89876 == 49.9 Hz
~p. = ~lk2 - kl~2

ww tIe - hne rt rt
PI +P2

w.E 91.7591 X 0.1- 0.2 X 56.405

148.16
=
9.17591 -11.2810
148.16

asy -2.10509 _ 0.014208 u Mw

148.16 P

En = - 14.2082 MW
E7.4 Two interconnected areas A and B have capacities 2000 MW and 750 MW respectively.

gi
The speed regulation coefficients are 0.1 p.u. for both the areas on their own area

nee
ratings. The damping torque coefficients are 1.0 p.u. also on their own base.
Find the steady state change in system frequency when a load increment of 50 MW

rin
occurs in area A. Find also the tie line power deviation. System frequency is 50 Hz.
Solution:

aI2 - - .!A -- 2000 -- 2.67 g.n

Pr2 750
I
D = 1.0 p.u. MW/puHz = 50 p.u. MW/Hz
et
R = O.lpu = 0.\ x 50 = 5Hz/p.u. MW
1 1 \
~I = ~2 = ~ = D + R= 50 + 5' = 0.02 + 0.02 = 0.22

~
o k2 + a l2 k l
f = --=---'-"--'-
~2 + a12~1
50Mw 1
kl = 2000puMW = 40 puMw;k2 =0

1
2.67x-
MO= 40 = 0.06675 = 0.08267Hz
0.22 + 2.67 x 0.22 0.8074

Downloaded From : www.EasyEngineering.net

262 Operation and Control in Power Systems

== -0.00681 x 2000 == 13.62MW

E7.5 In problem E 7.4, if the tie line connection is lost while it is carrying 50MW load, what
will be frequency in system A.
Solution:
The system A will be now an isolated area.

ww
The frequency drop

w.E ~f
°
== - ~PDA
~
== - 40 == -0.1 13636Hz
0.22

asy
E7.6 For a two identical area system the following data is given. Determine the frequency of
oscillations when a step load disturbance occurs.

En
Speed regulation coefficient == R == 4 Hz Ip.u. MW
Damping coefficient D == 0.03 pu MW1Hz
System frequency = 50Hz gi nee
The tie - line has a capacity of 0.1 p.u.

r
The power angle is 30° just before the occurrence of the load disturbance.
Solution: ing
fo
K == - - ==.
50
4RH 4x4x5
== 0.625 . net
TO == 0.1 cos 30° == 0.0866

0)2 == 2nTofo == 2n x 0.0866 x 50 =5.443

H 5
0) = 2.333 K 2 = 0.3906

0)0 == ~5.443 - 0.3906 == ~5.052375 == 2.2477489rad I sec

However, if the damping coefficient is not to be neglected

K == ~ [D + ~] == 0.625 + ~ x 0.03 == 0.625 + 0.075 == 0.7; k 2 == 0.49

4H R 4x5
In this case
ro = ~5.443 - 0.49 = ~0.4953 == 2.2255rad I sec

Downloaded From : www.EasyEngineering.net

Control 0/ Interconnected Systems 263

E7.7 A power system consists of two areas interconnected by a tie line which has a capacity
of 500MW and is operating at a power angle of 35°. If each area has a capacity of
5000MW and the speed regulation coefficients for both the areas are also the same and
are equal to R = 2HzJpuMW determine the frequency of oscillation of the power for step
change in load. The intertia constants are also the same for both the areas and are equal
to H = 5sec.
Solution:
500
TO = --cos35° = 0.819 x 0.1 = 0.0819
5000

ww 0)0 = [21t X50XO.0819]_( 50 )2

5 4x2x5

w.E =./ 5.148 -

1.5625 = ./3.5855 =1.8935rad / sec

asy
E7.8 In the above problem, if a step load change of85MW occurs in one of the areas determine
the tie line power deviation.
Solution:
En
A power of 85/2
change occurs.
= gi
The two areas are equal, the load change will be shared equally by both the areas.

nee
42.5 MW will flow from the other area into the area where a load

r ing
E7.9 Two power stations A and B operate in parallel. They are inter connected by a short
transmission line. The station capacities are 100MW and 200MW respectively. The

output of each station and the load on the interconnector if,

(a) the load on each station is 125MW,
.
generators at A and B have speed regulation 3% and 2% respectively. Calculate the

net
(b) the loads on respective bus bars are 60MW and 190MW and
(c) the load is 150MW at the station A bus bar only.
Solution:
Case: (a) P'+P 2 = 125+125 = 250MW
No-N fo-f
Regulation ---"--=--

PI 1-f
-=--
100 0.03

P 1- f
- 2= - -
200 0.02

Downloaded From : www.EasyEngineering.net

264 Operation and Control in Power Systems

0.0003P I = (I - f) = 0.0001P 2
that is, 3P I = P2
P I +P2 =2S0
Solving PI = 62.S MW and P2 = 187.SMW

A B

62.5MW ~1---t-~ 62'_5

__ --0-+---1~ 187.5MW

ww 125MW 125MW

w.E
Case (b) : the loads on each bus bar is determined by the speed regulation characteristics.

asy 3 PI = P2
PI + P2 = 60 + 190 = 2S0MW
Solving
E PI = 62.S MW and P2 = 187.S MW

ngi A B

62.5MW nee
~f--+-~_2'5-0--+----i~ 187.5MW

rin
60MW

Case (e) : There is load of ISOMW at A only

g. 190MW

net
3 PI = P2
PI + P2 = ISO
Solving PI = 37.S
P2 = 112.SMW

A B

37.5MW
r;:::'\ ~ I 12.5 M W I r;:::'\
'01--I~I---------+---1'0 112.5MW

150MW

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 265

E7.10 The turbines in a power station A have a ulliform speed regulation of 2.5% from no -
load to full load. The rated capacity of the! generators connected to the bus bars total
60MW and the frequency is 50Hz. Station B has total generating capacity of 35MW at
the bus bars and has a speed regulation of 3.0% connected through an induction motor
generation set rated 10MW. Which has1a fupload slip of 3.0%. There are loads of
32MW and 24.5MW connected to station A ~nd B respectively. Find the load on the
inter connecter cable at this operating condition.
Solution:

ww 60MW
A -+p IOMW

6?-+--~~M_G---,set1---~---+---l6?
B

35MW

w.E
asy
Let P MW flow from A to B
32MW 24.5MW

Total load on A
Total load on B
E
=

=
(32+P)MW
(24.5 -P) ngi
Percentage drop
.
In speed at A =
nee
(32 + P)
2.5 -=----~
60
rin
Percentage drop
.
In
d
spee at B =
3 (24.5 - P)
.5 -'-----'-
35 g.n
3P
Percentage drop in speed of induction motor generator set = 100
et
Percentage drop in speed at B - percentage drop in speed at A = percentage drop in
speed of induction motor generator set

3.5(24.5 - P) 2.5(32 + P) 3P
=-
35 60 10

(24.5 - P) 3P 32 + P
-'-----'- - - =--
10 10 24
solving 106P = 268

268
P = 106 = 2.5283MW

Downloaded From : www.EasyEngineering.net

266 Operation and Control in Power Systems

P flows from A to B

Total load as A = 32 + 2.5283 = 34.5283MW

Total load as 8 = 24.5 - 2.5283 = 21.9717MW

7.10 Tie-Line Bias Control- Implementation

From the preceding discussion it is clear that in interconnected operation each area in normal
steady state must control the changes in such a fashion that it absorbs its own load change.
The tie line bias control strategy presented can be summarized: all inter connected

ww
areas must contribute their share to frequency control in addition to taking care of their own
net interchange.

w.E
The control error for each area can be now defined as a linear combination offrequency

asy
and tie line power errors.

Area control error for area I; ACE 1 = ~P12 + 8 1~f1 . .... (7.26)

E
and area control error for area 2; ACE 2 =

ngi
Hence in the block diagram of Fig. 7.11 now the
~P21 + 8 2 ~f2
~p ref commands
..... (7.27)

can be defined .

nee ..... (7.28)

and rin ..... (7.29)

g.
where 8 1 and 8 2 are the frequency bias parameters for areas I and 2 while K·1 and K;
net
are the integral controller gain constants.

Note that under steady state conditions when ~p 12 and ~fl, ~f2 become zero, from
eqn. (7.26) ~P12 + 8 1~f1 = 0 and ~P21 + 8 2 M2 = 0 convey that the controller action in the final
stage is independent of 8 1 and 8 2 .

In fact, even one of the 8's, either 8 1 or 8 2 can be zero. It is suggested that only one 8

I
can be selected as equal to the area frequency response characteristic P= 0 + - to give
R
satisfactory performance.

The complete tie-line bias control of a two area system is shown in Fig. 7.12.

p
If area ratings are different (-I) block is replaced by (- a 12 ) = (f-
r2

Downloaded From : www.EasyEngineering.net

Control 0/ Interconnected Systems 267

.-----~B,~--.---------~
I
-~F,(s)
R,

ww
w.E
asy
E ngi
Fig. 7.12 Tie-line bias control of a two area system

nee
The selection of integral controller gains K'I and K; must be such that too large a value

rin
should not lead to chasing of minor deviations of no-consequence.

7.11 The Effect of Bias Factor on System Regulation

g.n
Let a and b be the ratios of generation of areas A and B in terms of the total generation, i.e.
a + b = I p.u. et
Let Psc be the total system generating capacity per cycle and correspondingly, let PAC
and PBC be the generating capacity per cycle of A and B. BA is the tie-line bias setting at area
A expressed in generation capacity / cycle and Bs is the imposed regulating characteristic for
the whole system with a tie-line bias controller in A only, with no controller in B, the ratio of
bias setting at A ~o the generation characteristic at A, r, is given by

Psc = PAC a + PBC (I - a) = PAC . a + PBC . b ..... (7.30)

Bs = BA a + PBC (I - a) = BA . a + PBC . b ..... (7.31)

Downloaded From : www.EasyEngineering.net

268 Operation and Control in Power Systems

The change in natural frequency ~ fn following the load change as dictated by the
governor characteristic is

M=~PD
Pn
sc
If ~ fb is the system frequency deviation due to the controller, then
M =~PD
S B
s
from eqn. (7.30) and (7.31)

ww M = _ _ _~_P=-D_ _
n PACa + PBC (1- a)
..... (7.32)

and
w.E ..... (7.33)

asy
The improvement change in frequency deviation, as a percentage of the initial
disturbance is

En ..... (7.34)

gi nee
r ing
(PAC - BA)a
(BA - Pnc)a + PBC
x 100 . net
..... (7.35)

Since, r = BA / PAC and assuming PAC = P BC '

L\f =- a(r -1) xl 00% ..... (7.36)

1 + a(r -I)

or ~f = (b -I)(r -I) x 100% ..... (7.37)

r - b(r-I)
The percentage change in frequency is independent of ~ PD' and depends only on a, b
and r. When r = I; ~ f = 0, therefore
~fn=~fs
The initial ungoverned frequency deviation is equal to the subsequent governed frequency
deviation.

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 269

7.12 Scope for Supplementary Control

During operation, the system, frequency, tie-line flows and other derived quantities will be
continuously swinging with periods generally of a few seconds. Considering the time lags in
measurements, the prime mover governor characteristics, etc. supplementary controllers cannot
respond to rapid swings offrequency and tie-line power. Neither can they effectively regulate
these changes nor is it expected of them. It is desired that supplementary controllers should
regulate the frequency and tie-line deviations about a base that corresponds to the prescheduled
values. In a reactive sense, the supplementary control is a steady-state control. A high speed
dynamic analysis of the system "and area characteristics are required only in connection with
improving system stability. .

ww
7.13 State Variable Model for a Three Area System
In chapter 6 state variable mode for a single area system and application of optimal control has

w.E
been discussed. Modern control theory can be applied very elegantly for multi area system.
The block diagram representation of a three area system is shown in Fig. 7.13.

a syE
The folIowing equations are written directly from the block diagram.

ngi
nee
rin
g.n
et

Fig. 7.13 Three area system

Downloaded From : www.EasyEngineering.net

270 Operation and Control in Power Systems

ww
w.E
asy
E ngi ..... (7.38)

nee
rin
g.n
et

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 271

Taking the inverse Laplace transform of equation (7.38)

~fl =_I_(-~fl +Kpl~PGI -Kpl~PDI -Kpl~Pne.l)

Tpi

~f3 = }-(- M3 + Kp3~PG3 - Kp3~PD3 - Kp3~Ptle,3)

ww ~PGl =-l-(-~PGl +~XVl)

TTl

w.E ..... (7.39)

asy
E ngi
nee
rin
g.n
~Ptlel =21tTI02(MI-M2)+21tTI03(MI-MJ
et
~Ptie.2 = 21tT~1 (M 2 - Ml ) + 21tT~3 (~f2 - ~fJ
~Pl1e,3 = 21tT~1 (M3 - M 1 )+ 21tT302 (M) - M 2 )
The above equations are put in the form

X= AX+Bu+Fd ...... (7.40)

where
XT = [M1 M2 M3 ~PGI ~PG2 ~PG3 ~PVl ~PV2 ~PV3
~Ptie, I ~Ptle, 2 ~Pl1e, 3]
= [~P CI ~PC2 ~PC3]
T
u
dT
= (~PDl ~PD2 ~PD3)

Downloaded From : www.EasyEngineering.net

The matrices A, Band F are given by

ww
1
0 0 0 0 0
Kpl
0 0
_ Kpl
0 0
TPl
0

0
w 1

Tp2
0

T .Ea0 0 0
TPl
0
KPl
TP2
0

Kp3
TPl
0
Kp2
TPl
0

Kp3

syE
0 0 0 0 0 0 0 0
Tp2 Tp3 Tp3
1 1
0 0 0 0 0 0 0 0 0 0
Tn Tn

A=
0 0 0 0
TT2
0

1
0

ngi
1
TT2
0

1
0 0 0

..... (7.41)
0
"='
.,1=
~

....
...
---
0

1
0

0
0

0
Tn
0
0

1
0

0
Tn
0nee 0

0
0

0
=
=
1=
=
rin
RITSl TSl =-
0
1
0 0
1 n
R2 TS2
0 0 0
~~2
0 0 0 0
=
.,=
....
0 0 ---
1

R3 TS3
0 0 0 0 0
TS3
1
0 0 0
g.n -
I
=
...
=
j et
A P Q 0 0 0 0 0 0 0 0 0 ~
R B s 0 0 0 0 0 0 0 0 0 =
~
T U C 0 0 0 0 0 0 0 0 0 .,
~

00
'<
....
r.I
~
a
r.I

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 273

Where,

A= 21t(TI02 + TI03)
B= 21t(Tfl + Tf3)
c= 21t(T~1 + T3 02 )

-P = -21tTI2°

ww Q- = - 21tT °
I3

w.E- -
R = -21tT2I°
°
a S = -21tT23

-
syE
T = -21tT31
0

- ° n
and
gi
U = -21tT32

nee
0 0 0
rin
0
0
0
0
0
0 g.n
0
0
0
0
0
0
0 et
0 0
1
B= 0 0 ..... (7.42)
TSI
0 0
TS2
0 0
TS3
0 0 0
0 0 0
0 0 0

Downloaded From : www.EasyEngineering.net

274 Operation and Control in Power Systems

Kpi
0 0
Tpi
_ Kp2
0 0
Tp2
Kp3
0 0
Tp3
0 0 0
0 0 0
F-~
..... (7.43)

ww 0
0
0
0
0
0

w.E 0
0
0
0
0
0

a 0
0
0 syE
0
0
0
0
0
0

ngi
The equations can be written down for any number of interconnected areas in a similar
manner.

7.14 State Variable Model for a Two Area System nee

In the case of two area system
rin
~Ptle,2 = a 12 Plle,1

P g.n
and a I2 =-_ll
Pr2
The system matrices time become
et
1 Kpi _ Kp2
0 0 0 0
TpI Tpi Tp2
Kp2 Kp2
0 0 0 0 -a 12 -
T-
Tp2 Tp2 P2
Kpi
0 0 0 0
Tst TSI Tpi
A= I
0 0 0 0 0
TS2 TS2
--- 0 0 0 0 0
RITSI TSI
0 --- 0 0 0 0
R2 TS2 TS2
21tTl~ - 21tTl~ 0 0 0 0 0
Downloaded From : www.EasyEngineering.net
Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 275

0 0
BT oro 0 0 TSI
000 0 0
TS2
:]
0 0 0 0
FT =
r KPI
:PI
o 0]
_ Kp2
0 0 0 o 0
Tp2
The static variables, control variations are disturbance variables reduce to

ww XT = [M1 M2 LlPGI LlPG2 LlX vl LlX V2 LlPtie,d

u = [LlPCI ]

w.E LlPC2

d = [LlPDI ] respectively.
a syE
LlPD2

7.15 State Variable Model for a Single Area System

ngi
The state variable model for the single area system is a single input, single output model with
no tie-line power. The matrices of the system equations are

Tp
0 ~
Tp nee
A= 0
I
rin
Ts
0
Ts
1
g.n
0
RTs Ts
et
B= 0
1
Ts
_ Kp
Tp
F= 0
0

XT = [M LlPG LlX v]
u = LlPc
d = LlP D

Downloaded From : www.EasyEngineering.net

276 Operation and Control in Power Systems

---------------------------------
E 7.11 Solve the state space equations for a three area system with the following parameters
and plot the frequency, generation and tie-line power deviations for a step load disturbance
of 0.01 p.ll. in area 1.
RI = R2 = R3 = 2.SH:zJp.u. MW
TpI = Tp2 = Tp3 = 20s
TTl = TT2 = TT3 = 0.3s
TSI = TS2 = TS3 = 0.08s
KpI = Kp2 = Kp3 = 120H:zJp.ll. MW
TI2 = TI3 = T23 = T21 = T32 = T31 = 0.07 p.ll. MW/rad
ww PDI = O.Olp.ll. MW

Solution: w.E PD2 = PD3 = 0.0

a
The state variable mode for the system using eqn. (7.27) is given by

~1\ = syE
20 [- MI + 120(~PGI )-120(0.01) -120(~Ptle,I)]
1

~f2
1 ngi 120(~PG2) - 120(0.0) - 120(~Ptle,2 )]

~f3
= 20 [- M 2 +

nee
= 20 [- M3 + 120(~PG3 )-120(0.0) -120(~Ptie,3)]
1

rin
g.n
et
·
~XVI =--- -~XVI
0.08
1[ I]
+O.O--MI
2.5

· I[
~XV2 =--- -~XV2 +0.0--~f2
0.08
I]
2.5

· . 0.081[
~XV3 =--- -~XV3 +0.0--~f3
.
I]
2.5

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 277

~t( s)

0 2 3 4 5 6 7 8 9 10 II 12 13 14

- 0.004

N - 0.008
~
'-
<l
-0.012

l - 0.016

ww - 0.020

w.E (a)

asy
;-\j}v Eng I,
I •

(",
. ~ \~
.~p G3 ine
eri
/,,
.,...............,. ~PG2

0
,,~PG,

2 3 4 5 6 7 8
n
9 10 11
g.n
12 13 14

(b)

Fig . E 7.4.1 Response of a three area system to step load change in one area:
~t( s)

et
(a) variation of frequency error (b) variation of generation changes

~Ptie' = 27t(O.07)(M, - M 2 ) - 27t(O.07)(M, - M2 )

~Ptie2 = 27t(O.07)(M2 - M, ) - 27t(O.07)(M2 - M 3 )

~Prie3 = 27t(O.07)(M3 - M,) - 27t(O.07)(M, - M 2 )

The equations are solved on digital computer using the Runge - Kutta method. The
results are plotted: The riot is silm,vn in Fig. E 7.4.I(c).

Downloaded From : www.EasyEngineering.net

278 Operation and Control in Power Systems

----------------------------------
0.007
0.006
0.005
0.004
0.003

r
.!:!
0.002
0.001

0
c:
<l
2 3 4 5 6 7 8 9 10 II 12 13 14
-D.OOI

lww
~t(s)
-D.002
-D.003
-D.OO4
-D.OOS
w\ I(\.\E;1
\
I ,'a "
II\! \! sy
-D.006
I I\ iI '\ // \
I
I", /
r'\'-.. . //"-......,~- ~ -~.~---
-D.007
-D.008
'\j
En -
-D.009
-D.OIO
\ 1
,)
'\
6Plie 1
gi
E 7.4.1 (c) Variation of tie line power nee
E7.12 Given the following parameters pertaining to a two area system in appropriate units
Ttl 0.28
r ing
Tt2

Tpi
0.3
18
. net
Tp2 20
RI 2.4
R2 2.5

TSI 0.1

TS2 0.1
Kpi 120

KP2 100

TI2 0.06
Obtain the frequency deviations in both the areas and tie power deviation when a load
change of llP DI == 0.0 I p.u. MW occurs.

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 279

Solution:
The system is simulated in MATLAB the frequency deviation is shown in Fig. E 7.S (a) and the
tie-line power deviation in Fig. E 7.S (b).

ww
w.E
asy
E ngi
nee
Fig . E 7.5(a)

rin
g.n
et

Fig. E 7.5(b)

Downloaded From : www.EasyEngineering.net

280 Operation and Control in Power Systems

E7.13 For the system given in E 7.5 obtain the controlled response of the system with tie-line
bias control.
Solution:
For tie line bias control the bias coefficients are computed form
I I I
8 1 =°1 + - = -- + -
RI KPI RI
I 1 1
and 8 2 = 0 2 + - = -- + -
R2 KP2 R2
An integral control gain of KI = 0.5 is used for both the areas.
The tie-line power deviation is shown in Fig. E 7.6(a) and the frequency response in

ww
Fig. E 7 .6(b).

w.E
asy A):OO5'
\
\

E \
\
ngi
nee
rin
Fig . E 7.6 (a) g.n
et

Fig. E 7 . ~ (b)

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 281

E7.14 In the system shown in example 7.5, if load disturbance of L1PDI = L1P D2 = occur
simultaneously obtain the freciuency response and tie line power deviation.
Solution:
The response of tie-line deviations and frequency deviation are shown in Fig. E 7.7(a)
and E 7.7(b).

;: ;,

"::0.005

-0,01

ww :..().Qj~
_·it

w.E
asy
-0.025

-0.03
En
-0.035 '----'---'-
0
gin
2 ---'-3--4'----'5---'-6--7"---8'-----'9----"0

Fig . E 7.7 (a)

eer
8 x 10'"
in g.n
~
6

I \ /''\,
et
2

0
I \ /' -/-----
MI -2 : \ I ~.

\)
-4

-6

~ v

-10
0 2 3 4 5 6 7 8 .9 fO

Fig . E 7.7 (b)

Downloaded From : www.EasyEngineering.net

282 Operation and Control in Power Systems

E7.1S For the system in E (7.5) obtain tie-line bias controlled frequency and tie-line power
deviation with integral controller gains Kfl = K'2 = 0.5.
Solution:
The responses are shown in Fig. E 7.8(a) and E 7.8(b).

ww
w.E
asy
E ngi
nee
Fig. E 7.8 (a)

rin
g. net

Fig. E 7.8 (b)

Downloaded From : www.EasyEngineering.net

Control of lnterc~nnected Systems 283

The MATLAB programme and the simulation setup are given below.
Ttl 0.28;
Tt2 0.3;
Tpl 18;
Tp2 20;
RI 2.4;
R2 2.5;

ww Tsl
Ts2
0.08;
0.1;

w.E Kpl
Kp2
120;
100;

asy Tl2 0.06;

E Gain
-K- ngi
nee
rin
g.n
Transfer Fon

simout I
Transfer Fon I

To Workspace I
et
Scope 3

Gain I
~------------------~-K-~------~~~--------~

Fig. Uncontrolled Case

Downloaded From : www.EasyEngineering.net

284 Operation and Control in Power Systems

Gain
~----------------<-K-~------------------~,
Bl Gain 3

ww simout I

w.E To Workspace 1

a syE
ngi Gain I
~------------------C-K-~------~~~--------~,

nee
Fig. With tie-line bias control rin
7.16 Model Reduction and Decentralised Control g.n
et
Decentralized control has been developed to avoid the necessity of too many feedback loops in
the controller structure. Optimal controller design requires that every sensor output affect
every actuator input. In decentralized control, restrictions are placed on information transmission
between sensors and actuators.

Fig. 7.14 Decentralized control

Downloaded From : www.EasyEngineering.net

Control of Interconnected Systems 285

l 2
Considering Fig. (7.14) XI is used to determine the vector u while X alone is used to
determine vector u2 •
This is complete decentralized control.
Model Reduction by Aggregation:
Consider
X(t) = AX(t)+ Bu(t) ..... (7.44)

Let 2:(t) = F2(t)+ Gu(t) ..... (7.45)

It is required that eqn. (7.45) represent the same dynamic system rep'resented by

ww
eqn. (7.44), given that the number of variables in 2 is much less that those in X.
For dynamic exactness

w.E 2(t) = CX(t)

It is necessary then that the equations
..... (7.46)

asy 'FC = CA ..... (7.47)

and
EnG=CB ..... (7 .48)

gi
are satisfied.

nee
The aggregate state vector 2(t) must be a linear combination of certain modes of X(t).
The eigenvalues of F are the eigenvalues of A corresponding to those modes of X(t) which are
retained in 2(t). In other words, a model reduction is achieved by retaining dominant modes.

r ing
Many a time, it is advantageous to define 2(t) corresponding to physical variables.
Then
2(t) = HX(t)
This may be termed as modal approximation.
. net
..... (7.49)

Eqn. (7.44) may be rewritten as

XI(t)] [All ..... (7.50)

[ X2 (t) - lOA 21

where 10 is a small, positive parameter. For 10 = 0, the system decomposes into two
completely independent sub systems. The two sub stems thus obtained may be treated each
separately for simulation, analysis and control design. The dimensionality of the problem is
then reduced.
In multi area operation, the decision makers in each area may use a detailed model of his
area and a dynamic equivalent of the remainder of the system. Thus, the same system appears
in different forms to different controllers. By singular perturbation technique, it is possible to
devise control strategy for fast and slow dynamic models of the system. Multi parameter
perturbation method may be used for periodic coordination of all the sub-system controllers
using Pareto control strategy.

Downloaded From : www.EasyEngineering.net

286 Operation and Control in Power Systems

Decentralized suboptimal controller

Consider a system defined by an n-dimensional state vector X. It is desired to replace it by a
reduced order model with m-dimensional measurement vector Z. Then each Z may be considered
to be a linear combination of the states X so that
Z = MX ..... (7.51)
If UO is an appropriate control for the original system, it is structured from
UO=KX ..... (7.52)
In view of the relation (7.51) UO must also be a linear combination of Z. Therefore

ww Uo= FZ

w.E
The elements ofF can be obtained using various concepts in control theory. For example,
if minimum norm concept is utilized then

asy
Fj = K M; (M) M; tl M) j = 1,2, ...... , m ..... (7.53)

En
In eqn. (7.53) K is the full state variable feedback gain matrix.
For a two area power system each area may be controls u l and u2 which depend upon
the vectors ZI and Z2 given by
gi nee
ZJ = (LlfILlPoILlXEILlP"e,1 fACE)] ..... (7.54)
rin
zi = [Llf2LlPG2LlXE2LlP"e,dACE2] ..... (7.5~)

g.n
Remote measurements are thus avoided. In fact, area controllers are influenced only a
little by remote measurements.

Downloaded From : www.EasyEngineering.net

Control 0/ Interconnected Systems 287

Questions

7.1. Explain how the tie-line power deviation can be incorporated in two-area system
block diagram?
7.2. What is a tie-line? Explain
7.3. Derive the relation between steady state frequency error and tie-line derivation
for step load disturbances in both areas
7.4. What is area frequency response characteristic. Explain it in the context of two

ww 7.5.
area systems.
What are the advantages of inter connected operation of power systems? Explain.
7.6.
w.E
Explain how state variable representation can be obtained for a two area system.
Write down the equations.
7.7.
asy
Sketch and explain the block schematic of a two-area system.
7.8.
En
What are the features of the dynamic response of a two area system for step load
disturbances?
7.9.
7.10.
gi
Explain tie-line bias control applied to a two-area system.

nee
What are the considerations in selecting the frequency bias parameter.

r ing
. net

Downloaded From : www.EasyEngineering.net

288 Operation and Control in Power Systems

Problems

P 7.1 Two interconnected control areas have the foIlowing characteristics

Area I Rl = O.Olpu
0 1 = 0.08 p.u.
Base MVA = 500
Area 2 R3 = 0.02pu
O2 = 1.0 p.u.

ww Base MVA = 1000

w.E
A load change of 100MW occurs in area l. Determine the new steady state
frequency and tie-line power derivation.
P 7.2

asy
In P 7.1 if a load change of 200MW occurs in area two determine the frequency
derivation and tie line power derivation.
P 7.3
En
In P 7.1 if a load change of 150MW in area I and a load change of 250MW in
area 2 occur simultaneously determine the frequency and tie line power derivation.
P 7.4
gi
Given Rl = 2.4Hzlpu MW
R2 2.6Hzlpu MW
nee
TTl
TT2
0.28s
0.31s r ing
TSI
TS2
Kpl
0.075 S
0.08s
100Hz / pu MW
. net
Kpl 120Hz / pu MW
Tl2 0.07 p.u. MW / rad
T21 0.07 p.u. MW / rad
Solve the state equations for a two area system and plot the frequency, generation
and tie line power deviations for a step load disturbance.
P 7.5 Repeat problem 7.4 using MATLAB compare the result.
P 7.6 Repeat problem 7.5 with PDI = O.Olpu MW in area I and PD2 = 0.01 pu MW in
area 2.
P 7.7 Repeat problem 7.3 with PD2 = 0.03 p.u. MW. What are your conclusions from
the study of the results of problem P7.6 and P7.7.

Downloaded From : www.EasyEngineering.net

ww
w .Ea AND REACTIVE
8 VOLTAGE
syE
POWER CONTROL
ngi
nee
Industrial and domestic loads, both, require real and reactive power. Hence, generators have

rin
to produce both real and reactive power. Reactive power is required to excite various types of
electrical equipment as well as transmission network. The reactive power requirement of

g.n
consumers arises mostly from the lagging vars n.eeded to supply magnetizing current to
transformers and induction motors.
8.1 Impedance and Reactive Power et
In the transmission network, requirement is the difference between that absorbed in the series
inductance (J2X) and that produced in the shunt capacitance (y2B). There is a level of loading
at which the leading vars of charging current balance the lagging vars of the inductive lines;
called the natural or surge impedance loading of the system. This natural load for a transmission

line is given approxirtlately as (~) p.u., In the case of cables the shunt capacitance is higher

and series inductance is lower. As a consequence of which the natural loading is higher. However,
this limit is generally above the thermal limit of the cable. The natural impedance loading for
overhead lines is shown in Fig. 8.1 as a function of system voltage on logarithmic scale. The
conductors are bundled two together for 220kY and four for 380kY.. TaSte 8.1 shows the
reactive power requirements of overhead lines, cables and transformers. The reactive power
compensation of the transmission system depends on the load and its power factor. When the

Downloaded From : www.EasyEngineering.net

290 Operation and Control in Power Systems

ro' 104 I-----r--r-t---t-r--r-----,----,--

ww 10 I 1-----'---'---'-------'-'------'-----'--'-

w.E 10 20 30 100 220

(kV) ----.
1000

a
Fig. 8.1 Natural impedance loadings as a function of system voltage

syE
line is operated at no-load, the full charging power occurs and would result in considerable
increase in voltage unless some compensating device is used. With full compensation at

ngi
no-load, the line may be operated at any partial load between no-load and full-load with the
voltage not exceeding the permissible limits:

nee
Table 8.1 Var requirements of overhead lines, cables and power transformers

Equipment Mvar requirements

no-load
rin
Inductive, - Capacitive
Full load
Over head lines:
400kV,4 x 0.4 sq. inch -105.0 g.n
+1251.0
100 miles
275 kY. 2 x 0.4 sq. inch
50 miles
-22.0 +174.0 et
132 kV. 1 x 0.175 sq. inch -1.6 +10.0
20 miles
Under taken Cables:
400kV,2 x 3.0 sq. inch -540 -487.0
/
10 miles
275kV, 1 x 3.0 sq. inch -170 -155.0
10 miles
132kV,1 x 0.5 sq. inch -14.5 -13.6
5 miles
Transformers:
400t275kV,750MVA +0.5 +90.0
275/132kV, 240MVA +OJ +48.0
132133kV,'9omVA +0.2 +20.5

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 291

The loss in a line is given by,

2
P =K p £
L
y2 COS2 q,

where P is the power transmitted at voltage Y over a line of length e at a power factor
cosq,. Higher voltages are selected for transmission to keep the losses in an economically
justifiable relationship to the power P. In view of the inverse square relationship, reduction of
reactive power becomes an essential factor for obtaining efficient operation of the high voltage
lines. The reactive power requirement of overhead lines are shown in Fig. 8.2(a) & (b).

ww Inductive /

wi...
600 /
700 kv/
400 / ./

.-- Ea -
/ ./

E 200 /-:::....-- "'--400kv

0.5 ~-
0
0

syE
0
_:;::=;:;:::P'" 1.0 1.5 2.0
0 200
p

ngi
Capacitive

(a)

nee
1.0
0.8 rin
0.6
0.4 g.n
r
Cl..
PSIt).2

-0.2
-0.4
0
2.0 et
-0.6

(b)

Fig. 8.2 Reactive power requirement of overhead lines (4 conductor bundle)

(a) Variation with line voltage (b) Variation with line length

In order to supply quality service to customers reliably and economically, voltage and
(or) var control play (s) a leading role. Such a control has to be exercised at all the power
system. i.e. right from the generating point to the consumer terminals.

Downloaded From : www.EasyEngineering.net

292 Operation and Control in Power Systems

Rapid changes in voltage (flicker) can result due to some industrial loads such as a*
furnaces, arc welders, and wood chippers. Fig. 8.3 shows the real and reactive power demand
following the daily load cycle supplying a composite system load which may create relatively
large variations in voltage if control is not exercised. Also, there may be cyclic and non-cycli¢
loads that create voltage disturbance at both transmission and distribution levels.

5 5

r 4 r 4

ww 3

2
3

w.E
asy 10 20
Real power (MW)
30
~
10
Reactive power
20 30
(Mvar)~

En Fig. 8.3 Arc furnace consumption

gi
In addition, events such as planned line switching, un planned line trips, planned and

nee
unforeseen generator trips and equipment failure may produce voltage and var variations.
Unless proper voltage support is given at strategic locations in the system, the aforesaid events

r
may result in loss of stability and possible loss of service to a large number of consumers.

ing
For long distance transmission of power, the use of HVDC transmission has proved

.
economical in certain cases. The var demand of DC terminals varies usually from 0-60% of

net
the MW rating of the DC lines as power transfer is varied over its full range. When a fault
takes place on the nearby AC system, the var demand of the DC link may reach a high value
and unless compensated may produce large AC voltage variations.
Continuous control can be achieved by means of synchronous compensators installed
at line ends and/or in the intermediate substations.
The use of shunt connected controllable var compensation to improve the power transfer
capability and stability is an acknowledged fact. From Fig. 8.4 it can be seen that the theoretical
maximum power transfer takes place at a power angle of 0'2 = 90.

P'2= - -
xl2
Fig. 8.4 Power transfer with ,,0 intermediate voltage support

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control .. 293

With an intermediate, controllable, shunt var compensator, the angle could be increased,
in principle to 1800 across the line (Fig. 8.5).

controllable shunt
var compensator

ww Fig. 8.5 Power transfer with intermediate var compensator

8.2
w.E
System Voltage and Reactive Power
Consider the system shown in Fig. 8.6. The voltage at bus 2 is related to the voltage at bus I
by the relation.
asy
Bus I Bus 2

.. En
I giP + 1Q

nee
I

rin
Fig. 8.6 Power flow across a short line

g.n
Also,
V 2 = VI-IZ
VI' I· = P + jQ

I = P - jQ =P- jQ
et
..... (8.1 )
..... (8.2)

So that ..... (8.3)

V; VI
since V I is the reference phasor.
From eqn. (8.1) and (8.2)

V2 =VI-(p~~Qlz
Neglecting the line resistance
P Q
V 2 =V I -J' -XL--X
V V L
..... (8.4)
I I

Downloaded From : www.EasyEngineering.net

294 Operation and Control in Power Systems

Eqn. (8.4) is illustrated by the phasor diagram in Fig. 8.7. It can be inferred from

eqn. (8.3) or Fig. 8.7 the voltage level is influenced largely by the reactive power drop QX L ,
VJ

PX L
since the quadrature component - - does not materially affect the voltage profile (both the
VJ
drops are only small fractions of the bus voltage magnitudes).

ww
w.E V2
Fig. 8.7 Phasor representation of eqn. (8.3)

8.3 asy
Reactive Power Generation by Synchronous Machines

E
Synchronous generators are able to produce both lagging and leading vars. Over-excitation of

ngi
a generator field produces vars while under excited field causes vars to be absorbed. At lagging
power factors, the limit on var generation is imposed by either rotor heating (due to maximum

nee
excitation current limit) or by stator heating (thermal MVA loading limit of the stator) consideration.

rin
Turbine output
limit

g.n
et
Rotor
heating
limit

0.4 0.2 0 0.2 0.4 0.6 0.8 1.0

Lead p.u. Mvar Lag
Fig. 8.8 (a) Operating chart for non-salient pole synchronous machine

Generators are invariably fitted with automatic voltage regulators which maintain the
thermal voltage at its normal value by adjustment of excitation. The operating charts for salient
and non salient pole synchronous machines are shown in Fig. 8.8.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 295

Theoretical Turbine output Stator heating

stability limit __ !...<?..( limit
limit

Rotor heating
limit

...

0.8 0.6 0.4 0.2 0 0.2 0.4 0.6

ww Lead p.u.MVAr Lag

Fig. 8.8 (b) Operation chart for salient pole machine

8.4
w.E
Effect of Excitation Control
Consider a synchronous machine with terminal voltage VI' The direct axis rotor angle with

a
respect to a synchronously revolving axis is 8. The voltage due to excitation acting along the

syE
quadrature axis is Eq and Eq' is the transient voltage along this axis. If a load change occurs and
the field current, If' is not changed then the various quantities mer..ioned change with P, the real
power as shown in Fig. 8.9(a).
ngi
i
00
,
8
Eq nee
>"'cr
'""
E~ rin
""
'cr

--r
VI
g.n
Fig.8.9(a)
In case the field current If is changed such that the transient flux linkages along the
et
q-axis Eq' proportional to 'P f' the field flux linkages is maintained constant, the power transfer
could be increased by 30-60% greater than in case (a) and the quantities are plotted for this
case in Fig. 8.9(b).

8
1
00
q

E~
>"'CT
VI
'""
'iT

"" --p

Fig.8.9(b)

Downloaded From : www.EasyEngineering.net

296 Operation and Control in Power Systems

If the field current If is changed along with P simultaneously so that Vt is maintained

constant then it is possible to increase power d~livery by 50-80% more than case (a). This is
shown in Fig. 8.9(c).

00
i E'
q

;;:- Vt
.i:T
.,.
L.1.l

L.1.l
--P

ww Fig.8.9(c) Variation of Voltages

w.E
It can be concluded that excitation control has a great role to play in power system
operation.

8.5
asy
Voltage Regulation and Power Transfer

En
If the fall of terminal voltage is assumed to be linear, then the graphs of machine terminal
voltage with load P can be represented by the relation.

gi E = Eo (\ - KP)

nee
Where Eo is the non-load terminal voltage and K is a coefficient of regulation (Fig.8.1 0).

r ing
t . net
o P(MW)--+

Fig. 8.10 Variation regulation

EV. s:
Since P =-smu
X
..... (8.5)
= Eo V sin8- Eo V KPsin8
X X
"
where 0 is the angle between E and V. Substituting Pm for T
E V
and solving for P

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 297

P= Pm sino
1+ PmKsino
which is a maximum at 0 = 90° having the value

P = Pm
max 1 K
+ Pm
under ideal conditions, where there is perfect control of excitation, K is zero.
If the machine is represented by a certain voltage E behind a certain reactance Xe then
Fig. 8.11 illustrates the effect of various types of regulators on its performance.
It can be proved for an ideal, voltage-actuated automatic regulator that the maximum

ww
value of the gain is

Xd - xd

w.E
K vmax = , ..... (8.6)
Xd
where xd and x~ are direct axis synchronous and transient reactances of the machine
respectively.
asy
There is also a minimum value for the gain to ensure stability at large load angles. For

En
better performance, forced regulation using the derivation of current and / or voltage is
recommended.

2.0
gi
The effect of different types of regulations is shown in Fig. 8.11.

nee
1.8
1.6
1.4
r ing
1.2
1.0
0.8
. net
0.6
0.4
0.2
Forced regulator xe g 0

Fig. 8.11 Effect of regulators

8.6 Exciter and Voltage Regulator
The function of an exciter is to increase the excitation current for voltage drop and decrease
the same for voltage rise. The voltage change is defined
C1 V~(Vl - Vref )
where VI is the terminal voltage and Vref is the reference voltage.

Downloaded From : www.EasyEngineering.net

298 Operation and Control in Power Systems

Exciter ceiling voltage: It is defined as the maximum voltage that may be attained by
an exciter \yith specified conditions of load.
Exciter response: It is the rate of increase or decrease of the exciter voltage.
When a change in this voltage is demanded. As an example
consider the response curve shown in Fig. 8.12.

Response = IOOV = 250V / sec

OAsec
Exciter build up : The exciter build up depends upon the field resistance and
the charging of its value by cutting or adding. The greatest
possible control effort is the complete shorting of the

ww field rheostat when maximum current value is reached in the

field circuit. This can be done by closing the contactor shown

w.E in Fig. 8.13.

asy
En
200

gi nee
r ing
o 0.2
Seconds -+

Fig. 8.12
0.4
. net

tt: Stator

Alternatorfield
Pi/ot Exciter Main Exciter

Fig. 8.13

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 299

When the exciter is operated at rated speed at no load, the record of voltage as function
of time with a step change that drives the exciter to its ceiling voltage is called the exciter build
up curve. Such a response curve is show in Fig. 8.14.

0 To ceiling
C voltage

EFD

i
ww a d

w.E 0'----------------' e

asy ----. seconds

Fig. 8.14 Exciter Build-up curve

Response ratio
E
Line ac represents the excitation system voltage response.
Cd
---p.u. V /sec ngi ..... (8.7)
Oa(O.5)
nee
Table 8.2 Typical ceiling voltages
rin
Response ratio
0.5
1.0
Conventional exciter
1.25
1.4
-
-
1.35
1.5
SCRexciter
12
1.2 - 1.25
g. net
1.5 1.55 - 1.65 1.3-1.4
2.0 1.7 - 1.8 1.45 -1.55
4.0 ... 2.0-2.1

In general the present day practice is to use 125V excitation up to IOMVA units
and 250V systems up to 100MVA units. Units generating power beyond IOOMVA have
excitation system voltages variedly. Some use 350V and 375V system while some go up to
500V excitation system.

8.7 Block Schematic of Excitation Control

A typical excitation control system is shown in Fig. 8.15. The terminal voltage of the alternator
is sampled, rectified and compared with a reference voltage, the difference is amplified and fed
back to the exciter field winding to change the excitation current. An excitation system using
ampJidyne is shown in Fig. 8.16.

Downloaded From : www.EasyEngineering.net

300 Operati'on and Control in Power Systems

Exciter generator

P.T

ww
w.E
Voltage
Amplifier

Fig. 8.15 Excitation control

asy
Pilot exciter Ac Main Exciter Slip Rings Synchronous
Generator

E ngi
nee
rin
Magnetic
.--------1 Voltage
Regulator
g.
Automatic 1------'
net
Amplifier

Fig. 8.16 DC-Type Excitation Systems

8.8 Static Excitation System

In the static excitation system, the generator field is fed from a thyristor network shown in
Fig. 8.17. It is just sufficient to adjust the thyristor firing angle to vary the excitation level. A
major advantage of such a system is that, when required the field voltage can be varied through
a full range of positive to negative values very rapidly with the ultimate benefit of generator
voltage regulation during transient disturbances. The thyristor network consists of either
3-phase fully controlled or semi controlled bridge rectifers. Field suppression resistor dissipates
energy in the field circuit while the field breaker ensures field isolation during generator faults.

Downloaded From : www.EasyEngineering.net

,
Voltage and Reactive Power Control 301

A compact AC static excitation scheme is shown in Fig. 8.18.

Field
breaker CT

Field
suppression
resistor

control
Reference

ww Thyristor network

w.E Fig. 8.17 Static Excitation Scheme

asy Thyristor
Bridge
Synchronous Generator

En
gi nee
r ing
Fig. 8.18 AC Static-Type Excitation Systems
. net
8.9 Brushless Excitation Scheme
Generator

Field

'1000' PT

Gate control

Fig. 8.19 Brushless excitation scheme

Downloaded From : www.EasyEngineering.net

302 Operation and Control in Power Systems

In the brushless excitation system of Fig. 8.19, an alternator with rotating armature and
stationary field is employed as the main exciter. Direct voltage for the generator excitation is
obtained by rectification through a rotating, semiconductor diode network which is mounted
on the generator shaft itself.
Thus, the excited armature, the diode network and the generator field are rigidly connected
in series. The advantage of this method of excitation is that the moving contacts such as slip
rings and brushes are completely eliminated thus offering smooth and maintenance-free operation.
A permanent-magnet generator serves as the power source for the excitor field. The output of
the permanent magnet generator is rectified with thyristor network and is applied to the excitor

ww
field. The voltage regulator measures the output or terminal voltage, compares it with a set
reference and utilizes the error signal, if any, to control the gate pulses of the thyristor network.

w.E
A detailed brush less excitation scheme is shown in Fig. 8.20.
Ac main exciter Synchronous Generator

asy
E ngi
nee
rin
g.n
Automatic
Voltage
Regulator

Fig. 8.20 AC Alternator-Type Excitation System

8.10 Automatic Voltage Regulators for Alternators

et
For an isolated generator feeding a load the automatic voltage regulator (AVR) functions to
maintain the bus bar voltage constant. However, on dynamic, interconnected systems the AVR
has the following objectives
I. To keep the system voltage constant so that the connected equipment operates
satisfactori Iy
2. to obtain a suitable distribution of reactive load between machines working in parallel
and
3. to improve stability.
The last of the above functions is not within the scope of this book.
Under normal working conditions, it is not difficult to maintain generator terminal voltage
at any specified level. However, with sudden disturbances like load fluctuations, the AVR is
required to reduce the magnitude and duration of voltage variations as effectively as possible.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 303

The magnitude of the voltage dip is primarily determined by the transient reactance of
the machine while the duration of the dip depends on the time constant of the generator
and rapidity of regulation. Static regulators are more useful in this context as they have lesser
time delays.
The AVR which senses the terminal voltage and adjusts the excitation to maintain a
constant terminal voltage also maintains the reactive output at the required level since the latter
depends on the effective voltage difference between generator terminals and its point of
connection to the main system. Thus. while AVR is allowed to maintain the voltage on the low
voltage side, a change in reactive output to suit a change in system conditions is obtained
by tap changing a generator transformer. With several generators synchronized on a single

ww
low impedance bus section, one generator with the AVR may be permitted to maintain bus
bar voltage while the excitation of the rest are used to distribute properly the vars among
the generators.
w.E
8.11 Analysis of Generator Voltage Control

asy
+o-----------~~
En ",--...-----------0 R
Reference
voltage gi +
Ampli-
tier

nee
~--_r_-r_O

'-.L---r--'"-----I-O
Y
B

Exciter
rin
Alternator

g.n
Rectitier t - - - - - - - - - - - - - - - '
' - - - - - - - - - - - 1 and tilter t - - - - - - - - - - - '

Fig.8.21 Excitation System et

Consider the excitation system shown in Fig. 8.21 the block diagram representation is given
in Fig. 8.22.

Load

+
v
Vref
v Amplltier Exciter Generator

Fig. 8.22 Excitation System Block Schematic

Downloaded From : www.EasyEngineering.net

304 Operation and Control in Power Systems

------------------------------
From the block diagram

The amplifier is generally assumed to be instantaneous in action so that Ae A, the output

from the amplifier is KAAe where KA, is the amplifier gain. The input to the exciter field will be
absorbed by the resistance and inductance of the exciter field Re and Le respectively so that

Ae A= ReAie + Le ~(Aie)
dt

ww where Aie is the change in exciter field current. If I ampere change in field current
produces K) volt change in the output, then

w.E Ae r = K)Aie
The transfer function of the exciter can be obtained as follows:

asy Ae r K)Aie

En Ae A AieRe + Le ~(Aie)
dt

gi
In the frequency domain, taking Laplace transform

nee
_~L
AEr{s) K)
=
Re
rin~ ..... (8.8)
AEA (s) Re + Le S I+S~
Re
I +STe
g.n
Hence Ge =
Ke
I +STe
is the transfer function of the exciter with the time constant
et
The value of the amplifier time constant may be of the order of 0.02 to 0.1 sec while Te
may be 0.5 to 1.0 sec for conventional machines.
At a later stage it will be shown that the exciter block is more conveniently represented
I
by in accordance with IEEE recommendation.
KE + STE

The input voltage signal Ae r to the generator field, when applied to the circuit results in
the following Kirchoff's voltage equation.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 305

where Rf and Lff are the alternator field resistance and self inductance respectively and
flif is the change in the field c~rrent. Taking Laplace transform.

flEr(S) = (R f + LffS)fllf(S)
If the output voltage changes by fllVI then

.fi
ww L\I I' (s) =
ffiL fa
fllVI where Lfa is the mutual inductance between the field and stator

phase winding. Hence, the transfer function for the generator block will be

w.E
asy ..... (8.9)

En
where Tgf is the direct axis open circuit time constant also denoted more commonly by
TdO = Rf seconds.
LIT /
gi nee
The voltage regulator loop can be represented by the block diagram shown in Fig. 8.23.

Amplitier Exciter rinGenerator Load

g.n
et
Fig. 8.23 Voltage regulator block diagram

The three cascaded transfer function blocks GA' Ge and Ggf can be combined into a
single block.

so that the feed back control loop can be further simplified as in Fig. 8.24.

Downloaded From : www.EasyEngineering.net

306 Operation and Control in Power Systems

~IVI

Fig. 8.24 Simplified Block Diagram

8.12 Steady State Performance Evaluation

ww
The control loop must regulate the output voltage VI so that the error is made equal to zero. It
is also imperative that the response must be reasonably fast, yet not cause any instability
problem.
w.E
The steady state voltage error ,1e ss is given by

a Mss =
syE L\lvl~ef

= ,1I Vlo _
- L\ Vss

ngi
ref
G(O) ,1IVIO
1+ G(O) ref

where G(O) is the value of G(s) as s

= 1+ G(O) - G(O) ,1IVlo

nee
0 (i.e.) the steady state value

1+ G(O) ref

rin
= 1
1+ G(O)
,1IVlo
ref g.n
G(O)=~~ KgI' =K
(I + 0) (1 + 0) (1 + 0)
et
1
,1e ss = - - ..... (8.10)
I+K
Larger the overall gain of the forward block gain K smaller is the steady state error. But
too large a gain K can calise instability.

8.13 Dynamic Response of Voltage Regulation Control

Consider

,1V(t)=L- 1[,1V (S) G(S) ]

ref 1+ G(S)

The response depends upon the roots of the characteristic eqn. 1 + G(S) = o.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 307

As there are three time constants, we may write the three roots as SI' S2 and S3' A
typical root locus plot is shown in Fig. 8.25.
jro

-<J S3 S2
0
SI

ww
w.E Fig. 8.25 Root - Locus Plot
From the plot, it can be observed that at gain higher than Kc the control loop
becomes llnstable.
asy
8.14 Stability Compensation for Voltage Control
En
Since at higher gains the voltage control loop tends to oscillate and even may become unstable.

gi
it requires phase lead compensation or derivative control. As there are three time constants

nee
contributing to phase lag, a phase lead compensator with transfer function.
Gc (S) = I + STc
will alter the open loop transfer function to

_ K(I + STc)
r ing
G VR (S ) - -------=----
(I + STA)(1 + STe)(1 + STgf )

It will not affect the steady state accuracy as K is kept constant

. net
..... (8.11 )

8.15 Stabilizing Transformer

Consider the excitation system with a stabilizing tralJSformer as shown in Fig. 8.26.
KA = Error amplifier gain
Ksp TST ~ stabilizing transformer gain and time constants.
A stabilizing transformer is introduced in the exciter circuit as shown in Fig.8.26(a, b).
The input voltage Ve causes a current i ST to flow through the resistance and inductance of the
feed back loop.

Taking Laplace transform

Downloaded From : www.EasyEngineering.net

308 Operation and Control in Power Systems

Field

Error amplifier
I t - - - - j Power
L
amplifier ,.........~-r---.- 0
,-....; r--t---.--t- a
'-"'----t---t-_t_ d

Main excitor

ww Fig. 8.26 (a) Circuit

w.E
V Ref(S)

asy
+

En
gi Fig. 8.26 (b) Block diagram
nee
Ve(S) = R f IsT(S) + L f S IsT(S)
rin
= (R f + S Lf ) IsT(S)
If the mutual inductance is M for the stabilizing transformer then, g.n
"
Taking Laplace transform
VST (t) = M ~ i,t (t)
dt . et
VST = MS 1st (S)
Hence
SM
VST(S) SM Rf
Ve(s) R f + LfS 1+ lLs
Rf
L M
Letting Rf = Tst and -R = K st
f f

Vst(S) SK
- - = - - - =st - ..... (8.12)
Ve(S) 1 + TstS

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 309

It may be noted that since the secondary of the stabilizing transformer is connected to
the input terminals of an amplifier, it draws zero current. Thus, the stabilizing transformer
improves the response of the excitation system giving derivative control.

8.16 Voltage Regulators

The heart of the excitation system lies in the voltage regulator. It is a device that serves
the output voltage change (some times current or both) and provid~s corrective action to
take place.
The early voltage regulators were all of electromechanical type. Electronic voltage
regulators came in 1930's. They provide smoother and faster voltage control than the

ww
electromechanical type. Another development resulted in the use of rotating amplifiers. Magnetic
amplifiers were also used for voltage regulators.

w.E
Later versions of voltage controllers used solid state active circuits. Whatever may be
the hardware used, there was consistent effort to provide faster response, without dead band,
backlash, and absolutely reliable. Excitation systems which include voltage regulators also are

asy
classified into continuously acting type and non continuously acting type.

En
All the systems have a potential transformer and rectifier to sample the output a.c.
voltage and provide d.c. signal to the control loop.

gi
Another component is a comparator that compares the d.c. signal with reference value
and provides an error signal.
nee
An amplifier amplifies the error signal and brings it to the operational level.

rin
The exciter output is manipulated by the amplified error signal to suitably modify the
output voltage of the generator.
g.n
As a simple proportional feed back results in a steady state error, a rate feedback or
derivative control is proposed by Kron in 1954 and this was discussed in section (8.15).

Computer representation of excitation system

et
For synchronous machines, several models were developed and linear representations of the
machine using state space methods needed suitable models for excitation system for computer
simulation to perform dynamic studies. IEEE has suggested certain standard models for different
types of excitation system.
I. Type 1 System: It is a continuously acting regulator and exciter system. It will be
explained in detail later.
2. Type 1 S system: This is a special case of continuously acting system. In this
system excitation is obtained through rectification of the terminal voltage while in
type I system regulator voltage has maximum and minimum limits, in this case, the
maximum regulator voltage is made proportional to Vt' the terminal voltage of the
generator.

Downloaded From : www.EasyEngineering.net

310 Operation and Control in Power Systems

3. Type 2 System: In this system rotating rectifiers are used and is brush less. Altemator-
exciter and diode rectifiers are rotating on the same shaft and hence has no slip/
rings. Being brush less, the excitation voltage (E FD) is not available for feed back. In
this there are two damping loops giving good performance.
4. Type 3 System: This type of model is evolved for systems where both current and
voltage signals are used for feed back.
5. Type 4 System: All the previously described systems are continuously acting type
with high gain and fast action. Several of the earlier systems are non continuous
acting type. [n such cases presence of dead zone results in open loop operation.
They are also slow due to friction and inertia. Rheostat systems are examples for

ww such an operation. [n type 4 systems now there are two speeds of operation
depending upon the magnitude of the voltage error. Type 4 system hence is

w.E
non-linear and as such is not represented in state variable form. Westinghouse and
General electric have developed such excitation systems.

a
8.17 IEEE Type 1 Excitation System
syE
Consider Fig. 8.27. Let V f be the voltage across the field winding and if the current through it

ngi
Afand also the flux linkages in the field winding.

nee . .... (8.13)

consider the non-load magnetization curve shown in Fig. 8.27 for the exciter voltage
build up.
rin
I
slope of airgap
line = Rag
g.n
VexO
et

Fig. 8.27 In Engineering Units

· t he saturatm
Se IS . coe ffi' A-B
JClent -B-

.
The excIter voltage v 0ex == f("Jf ' Je\
0 )
..... (8. [4)

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 311

o
. Vex S 0
If = - + eVex ..... (8.15)
Rag

where Rag is the resistance given by air gap line.

The saturation may be represented by

..... (8.16)

an exponential relation, A and B to be determined.

ww . .... (8.17)

, ,
w.E
from eqn. (8.2)

asy ..... (8.18)

En rf [ . 10
= -1+RagSe(lf)JVex +-o-lI.f Vex - -
d ~ ( 0) dv~x

gi Rag

In per unit system (see Fig. 8.28)

dVex dt

nee
( p.u. ) _- ~ [1 R S (. )10 dAf{V ex ) dv ex (p.lI.)

r
Vf + ag e If JVex p.U.+
Rq ~~

ing ~

. net
A-B
Let ..... (8.19)
B

Downloaded From : www.EasyEngineering.net

III Operation and Co.tro. in Power Systems

V xpu slope of airgap

A ,
line= I /
/
8
V'pu I+---.--.:'----i~

, ,
~epu (V xpu~
It .,

ww Fig. 8.28 In Per Unit

w.E +-____....,..._.. Em

asy
E ngi
n
Rag nee
SePLI ( \ex
Pu )
Ero=v ex p.u.

Fig. 8.29 rin

IEEE type 1 excitation system representation is shown in Fig. 8.30.
g.
Then the transfer function for the exciter block can be represented by Fig. (8.29).

net

Fig. 8.30 IEEE Type I Excitation System

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 313

8.18 Power System Stabilizer

It can be proved that an voltage regulator in the forward path of the exciter generator systems
will introduce a damping torque and under heavy loading conditions this damping torque becomes
negative. This is a situation where dynamic instability may calise concern. Further, it has
been also pointed out that in the forward path of excitation control loop, the three time constants
introduce a large phase lag at low frequencies just above the natural frequency of the
excitation systems. To overcome this effect and improve damping, compensating networks
are introduced to produce torque in phase with the speed. Such a network is called "power
system stabilizer" (PSS)
Fig. 8.31 shows the block schematic of PSS at the appropriate point in the control loop.

ww
w.E vs
~--- ~(lJ()r~N

asy I---~EFD
L...-_ _....J

E Regulator

ngi
and Exciter

nee
Fig 8.31 Connecting Power System Stabilizer

Power system stabilizer is generally shown as a feedback element from the shaft speed
and has a transfer function of the form
rin
g. net
..... (8.20)

A washout circuit for reset action is used to eliminate steady effect after a time lag To
(4 sec to 30 sec). This control circuit ensures that there is no permanent offset in the terminal
voltage due to a prolonged error in frequency. The later may occur due to prolonged over load
or islanding situation. A lead compensation pair is used at two stages with center frequencies

21n ~TIT2 and 2~ ~T3 T4 . This part of the stabilizer circuit will improve the phase lag.

A filter section is generally added as shown in Fig. 8.32 so that undesirable frequency
components are suppressed and thus eliminate the possibility of undesirable interaction. Limits
are placed so that the output signal of the PSS is prevented from driving the excitation into
heavy saturation.
The output signal of the PSS is fed as a supplementary input signal V pss for the regulation
of the excitation system. Design of the PSS circuit is beyond the scope of this book.

Downloaded From : www.EasyEngineering.net

314 Operation and Control in Power Systems

Limiter Phase-Compensation Washout Filter

~
1+ sTI l+sTl K,s
1+ ST2 I + sT. I +sT"

VPSS
V x = V'l'li
}----l~ Regulator and Exciter 1----.---.

~________~ f-------~

ww
KSF

TSF

w.E Fig. 8.32 Power System Stabilizer

8.19 Reactive Power Generation by Turbo Generator ~"--

asy
The reactive power generation capability of a steam turbo-generator is dependent on the type

E
of construction.

ngi
Unit construction: In this type, where there is no steam interconnection, the generator
can absorb and produce reactive power while supplying the rated real power to the load. A

nee
reduction of active power load on the generator increases both its reactive absorption and
production capability. However, the active power scheduling is quite often based on an optimal

rin
criterion as discussed in Chapter 6, overlooking which may add to the costs. The level to
which real power loading may be reduced is mostly determined by considerations of boiler

g.
performance. Minimum permissible loadings usually fall in the range of 50 to 70% of rating.

net
Steam range construction: In the steam range construction having steam interconnection
between generators, the minimum loading constraint imposed by the boiler is of less significance
since steam from one boiler can be used to supply several generators.
As an example, iffour sets are supplied by a single boiler, then with three sets drawing
steam just for gland ceiling, maintaining vacuum and cooling while the fourth set at the partial
load, a wide range of reactive generation or absorption capability can be obtained. However,
this arrangement is not in use now.

8.20 Synchronous Compensators

There are rotating machines which are connected to the system at appropriate places only for
the purpose of controll ing vars. They are synchronous motors operating on load with adjustable
excitation. They are normally fitted with voltage regulators which operate when the voltage
deviates by a certain predetermined percentage of system voltage.
These machines are also capable of giving dynamic reactive power compensation, i.e.
they make available the required.reactive power within a short time. The prime consideration
for installing synchronous compensator is its flexibility of operation under all load conditions.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 315

This aspect justifies its use in certain systems even though the initial cost is high. Another
important consideration for using the synchronous compensator is the occurrence of sudden
voltage dips due to sudden short circuits or overloads.
Under such conditions the synchronous compensator performs better than other
apparatus. With additional excitation windings through which excitation can be forced
momentarily to counteract the transient effects, the synchronous compensator can even improve
stability. The synchronous compensator may be supplied from the tertiary winding of the
transformer as shown in Fig. 8.31, so that additional cost on transformer is not incurred.

8.21 Reactors

ww
Inductive reactors absorb reactive power and may be used in circuits, series or shunt connected,
While series connected reactors are used to limit fault currents, shunt reactors are used for var

w.E
control. Reactors installed at line ends and intermediate substations can compensate upto 70%
of charging power while the remaining 30% power at no-load can be provided by the under

asy
excited operation of the generator.

En 7
~IY~ Ll.y
Bus
giI

nee
~ I Bus

(a) Shunt reactor

r ing
(b) Shunt reactors
in parallel
with taps

. net
Fig. 8.33 Shunt Reactors

With increase in load, generator excitation may be increased with reactors gradually
cut-out. Fig. 8.33 shows some typical shunt reactor arrangements.

S.22 Capacitors
Capacitors produce vars and may be connected in series or shunt in the system. Series capacitors
compensate the line reactance in long overhead lines and thus improve the stability limit.
;However, they give rise to additional problems like high voltage transients, sub-synchronous
resonance, etc. Shunt capacitors are used for reactive compensation. Simplicity and low cost
are the chief considerations for using shunt capacitor.
Further, for expanding systems additions can be made. Fig. 8.34 shows the connected
of shunt capacitors through the tertiary of a transformer.

Downloaded From : www.EasyEngineering.net

316 Operation and Control in Power Systems

HV~ ~p-----11 LV
Bu~ I ~ ~ Bus

Shunt capacitors In

I I I
parallel

Fig. 8.34 Shunt Capacitors

ww Criteria
Table 8.3 Comparision of Shunt Capactor and Synchronous Condensers
Shunt capacitors Synchronous condenser
Initial cost
Losses w.E 1.0
Less than 1.5%
1.0 to 1.2
1.5 %t04 %

a
Variation of reactive
components of current
Over load
In steps only with possible

syE
surges of voltage and current
Not possible -
Continuous

Possible with reduced reacting

Installation
Increase of rating ngi
Simple
-.--.~

Possible by adding more units

Relatively difficult
1-7:-. -
Not possible unless another

Effect ofhannonic voltage Possibility of resonance nee unit is installed

No effect.

rin
8.24 Tap-Changing Transformers
g.n
Tap-changing transformers with variable transformation ratio can calise substantial change in
the flow of vars. The tap-changing transformers when used in radial lines maintain voltage at
their secondary terminals or at load terminals within limits. When used in tie lines, the et
tap-changer can regulate vars substantially. In case of weaker tie lines, active power may also
change to some extent. Lastly, when used in networks or loops, circulating vars can be controlled
by tap-changing transformer.
Consider a tie line with sending end and receiving end, both having tap-changing
transformer connected to them.
Let Vs and Vr be the voltages at either end at nominal conditions.
Let ts Vs = actual voltage at sending end
and tr Vr = actual voltage at receiving end
Vs - Vr = VI _. V2 == (lR cos<P r t-IX sin<p r)

Downloaded From : www.EasyEngineering.net

Voltage turd Reactive Power Control 311

PR+QX
Hence
Vr
PR+QX
Hence ts Vs == tr Vr + V
r

To ensure that the same overall voltage prevails allover the line minimum ranac of taps
are used so that ts tr = I
Substituting this condition and solving for ts

ww
w.E
asy
t; EnII ~
gi [1 - ( PR + Q X
VI Vs
=
Vs

nee
r ing ..... (8.21 )

8.25 Tap-Staggering Method

. net
At low loads. the reactive losses in transforme.rs are also low. The surplus ge~tjd by the
system at such low loads can be absorbed by increased reactive power losses in transformers
using tap staggering. Consider the pair of transformers shown in Fig. 8.35 connected between
HV and LV buses.

IN,~

LV~
Bus

Fig 8.35 Tap Staggering

Downloaded From : www.EasyEngineering.net

318 Operation and Control in Power Systems

If the taps on the transformers are staggered, a quadrature current circulates around the
transformers and carries additional reactive (I2X) losses. With more number of transformers
operating, the losses could be increased by this method.

8.25 Voltage Regulation and Short Circuit Capacity

Consider the short line voltage regulation
IR IX .
I1Y=-cos<l>, +-sm<l>,
Yr Y,
where Yr is the receiving end voltage and <l>r the angle between Y r and Ir' Ir being the

ww
current in the line. This voltage drop can be obtained alternatively from the short circuit capacity
of the bus. Since R + JX is the line impedance, the voltage drop is also equal to

w.E 11 Y = (R + JX). Ir

asy = (R + JX) P - jQ
Y

= (PR En
~QX)+ {PX ~QR)
gi nee
when a three phase fault occurs at the load bus, the short circuit volt amperes into
the bus

Sse = Y . Isc = Y .
Y
Zse
r
(R + jX) ing
R+jX= -
Sse
Y" . net
y2
R cos <l>sc
Sse

y2
and X sin <l>sc
Sse

y2 y2.]1
I1Y = P-cos<l>se +Q-sm<l>se -
[ Sse Sse Y

+ J.[ P-SIl1<1>sc
y2 . Y"
-Q-cos<Pse - ]1
Sse Sse Y

Downloaded From : www.EasyEngineering.net

Voltllge and Reactive Power Control 319

It can be shown that the magnitude of voltage regulation given by I R coscp + I X sincp
PR+QX
or V as the imaginary component gives only phase shift. Hence, the real part of

equation for d V above gives the change in voltage magnitude

Hence,

d V = [Pcoscpsc + Psin CPsc ]V

Sse Sse
ww
w.E
asy
For small changes in P and Q

E ngi
much nee
Once again, remembering that the real power does not affect the voltage magnitude

dV= -IdQ' rin

g.
h.
- sm'l'~c
V S<;c .

Hence, sin CPse = I.

net
Under short circuit conditions the current is almost lagging by 90° to the voltage

Thus
dV =dQ ..... (8.22)
V Sse
Per unit change in voltage magnitude is equal to the rate of change in reactive power to
the short circuit capacity of the bus.
It may be noted in this context that for a system with X » R, the inphase voltage drop
would be approximately equal to (I sin cpr)X, An in-phase voltage compensator or booster
would control the reactive power flow in the system. The quadrature voltage drop would be
approximately equal to (I sin CPr)' The quadrature voltage compensator would control the active
power flow in the system.
X
The ratio of R is very high for EHV and UHV lines.

Downloaded From : www.EasyEngineering.net

320 Operation and Control in Power Systems

8.27 Loading Capability of a Line

There are four kinds of limitations identified for the loading of a transmission line.
(i) thermal limit
(ii) dielectric limit
(iii) stability limit
(iv) natural loading limit.
Of all the above limits natural loading limit is the lowest limit and thermal limit is
the highest.

ww (i) Therma/limit: This limit depends upon

Atmospheric condition including ambient temperature and wind conditions

w Condition of the conductor and

.Ea
Ground clearance
Normal loading at the design and planning level is determined by loss evaluation. Any

syE
increase in line capacity with thermal consideration should also consider transformers connected
at either end also.

ngi
(ii) Dielectric limit: Line design on the basis of insulation requirement is always on the
conservative side. Line voltage can always be increased by 10% for increased

nee
power transfer. This requires proper consideration of transients and arresters in
operation.

r ing
Stability limit: There are several stability considerations for increased transmission
capability.
They are
Transient stability
Dynamic stability
. net
Steady state stability
Frequency collapse
. Voltage collapse
• Sub synchronous resonance

8.28 Compensation in Power Systems

Electrical power demand is growing at a great rate day by day and the generation is, in general,
not able to cope up with the demand. Several ways of increasing the power demand. Several
w~ys of increasing the power generation are investigated including many non conventional
modes. Again, transmission of increased power over the existing lines is considered to meet
the increasing demand as laying of new Iines and acquisition of right of way are too expentive
in developed areas. This necessitated implementation of compensation in power systems. For
many years series and shunt compensators are in use. Electrical energy cannot be stored in
bulk quantities. There must always exist a balance between generation and demand.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 321

8.29 Load Compensation

Utilization of reactive power to improve voltage profile and power factor is termed as load
compensation. Load compensation thus refers to improving power quality. Reactive power
can be injected by installing various sources such as reactors and capacitors in shunt or in
series with the system at appropriate places. This has been discussed in section (8). Another
aspect to be considered is that as far as possible negative sequence currents are to be avoided
so that power loss is reduced. This is achieved by operating the system in balanced condition.
Line compensation: Surge impedance loading (SIL) of a line is the power delivered by
a line to a purely resistive load equal to its surge impedance.

ww SIL = Vc MW =:0 VC
w.E J¥ Rc

asy
Power transmitted by a line is usually expressed in terms of this power. When surge
impedance is the load impedance the voltage profile of the line is flat. This is not practicable.

En
However, it is possible to modify the characteristics of the line by using such elements as
capacitors, inductors or synchronous machines. These elements can be connected in series or

gi
in parallel. Shunt compensation is similar to load compensation. Shunt connected capacitors

nee
improve the power factor, reduce the reactive power and thus increase the real power
transmission. However, this method has limitations and the size of the condenser banks increase

r
to a prohibitive level. Further they are to be switched off under Iight load conditions as otherwise

ing
the voltage rise may not be acceptable. Series compensations using capacitors cancel a part of
the line reactance and increase the maximum'power transfer. Such a compensation win reduce

.
the power angle for a given power transfer and enables increased loading. Excessive series
compensation using capacitors may result in resonance. Sub synchronous resonance at net
frequencies less than 50Hz occur with series compensation if care is not taken. This is discussed
later in the chapter.
Line compensation results in
(i) minimization of Ferranti effect
(ii) elimination of the need for under excited operation of generators and
(iii) enhanced power transfer capability
From the receiving end power circle diagram, we have

P
'IVsllVRI
= IAIIV~I
cos(/3 - 8) - - - cos(/3 - a) ..... (8.23)
R IBI IBI
with usual notation.
PR is a maximum when 0 = /3

Downloaded From : www.EasyEngineering.net

322 Operation and Control in Power Systems

..... (8.24)

similarly

..... (8.25)

ww For a tie-line with series impedance

B=Z=R+jX= ~R2+X"
w.E D = 1.0
(1=0

asy
,1=0
f3 Tan- i X/R

En
and =

gi nee
and p.s
.. max
=
~
VSVR
" R2 + X "
vi R 1
+ 0
R - + X-rin
differentiating PR.max with respective to X g.n
et

4( VR)" _I
Vs

when vs =V R''X= "J'3R ..... (8.26)

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 323

X
In overhead lines R ratio is much greater than this optimum value needed to maintain

X
the same voltage at both the ends. For 132kV lines R ratio" is from 2.S to 3.S; for 27S kv lines

/ X
it is about 8 and for 400kv lines it is as high as 16. Hence to maintain voltage profile R ratio

has to be decreased. This can be effectively done by series capacitor compensator so that line
inductance is compensated.

ww
Nevertheless, series capacitors are to be protected against over currents and over voltages
during fault conditions.

w.E
E 8.1 Two substations A and B are interconnected by a line having an impedance of
(0.03 + jO.12) p.u the substation voltages are 33 Lt) kv and 33 LOo kv respectively. In

a
phase and quadrature boosters are installed at A. Determine their output-voltage ratings

syE
and MVA ratings in order to supply SMVA at 0.8p.f lagging at substation B.
Solution:
Let Base MVA
Base KV ngi =

33kv
SOO

nee
=

VA = 1.0 L2° p.u and VB = 1.0 LOo p.u

Let current through the inter connector
Let load current at B = IB
= I
rin
SMVA g.n
IB = SMVA (0.8 - jO.6) = (0.8 - jO.6) p.lI.

Without boosting voltage drop in the line

= Z IB = (0.03 + jO.12) (0.8 - jO.6)
et
= (0.096 + jO.078)p.u.
voltage drop in the line with boosters = VA- VB
= 1.0 L20- 1.0 LOo

= 0.99939 + jO.0348994 - 1.0

= (-0.00061 + jO.0348997) p.lI.
Hence, in phase component of boosted voltage
= + 0.096 - (-0.00061)
= 0.09661 p.u.
Quadratic component of boosted voltage

Downloaded From : www.EasyEngineering.net

324 Operation and Control in Power Systems

= 0.078 - 0.3349 = 0.0431 p.u.

Rating of the in phase booster = 0.09661 x 1.00
= 0.09661 p.u. = 0.09661 x 500 = 48.3 MVA
Rating of the quadrature booster = 0.0431 x 1.00
= 0.0431 p.u. = 0.0431 x 500 = 21.55 MVA

E 8.2 Two substations A and B operating at II kv 3-phase are connected by two parallel lines.
1 and 2. Each line has a 11/132 kv transformer and a 132/11 kv substation. Each line has
an equivalent impedance of ZI = 0.2 + .i0.4 and Z2 = 0.2 + jO.6 ohms per phase which
includes both the transformers and the line, referred to II kv side.

ww (a) If the bus bar A is at II kv and is sending 30Mw at 0.8 p.f leading, find the
individual currents into each transformer and the powers at the station A.

w.E
(b) If the transformer at A in line 2 is fitted with tappings on the II kv side. what
percentage tapping would be required to make each line carry equal reactive powers?

Solution: asy
What would be the power sent by line 2 in this case?

~
Engi 0.2 + j 0.4
Line I ~~
1II132KV

~
Line2
nee~~ 132111KV

A
0.2 + j 0.6
rin B

Let base voltage be II kv

Fig. E 8.2
g. net
Let Base MVA = 30.0 = 37.5MV A
0.8

Total current flowing through both the lines

= 37.5 (0.8 + jO.6) - (0.8 + jO.6)p.u.

37.5
= 100 L36.87° (since II kv = 1p.u.)

Ilkv x Ilkv = 3.2270hms

Base impedance
37.5

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 315

2 = 0.2 + j0.40 = (0.0627)0.124d u

, 3.227 P
= 0.1386 L63°.43

22 = 0.2 + jO.60 = 0.0627 + jO.186 = 0.1961 L71.S6 op.u.

3.227
Hence,

Z· ,+ Z· ,= 0.5+ jl.O = 0.1 24+ JO.3101

.
p.u.
- 3227
= 0.334 L68.20 p.u.

ww I, = 0.1961 x 1.0 L(71 0 .S6+36 0 .87-68 0 .2)=0.S84L40 o.23p.u.

w.E 0.334

12 = 0.1386 x 1.0 L (0 ° °)
68 .43 + 36 .87 - 68 .2 = 0.41 SL32 .Ip.u. °
asy 0.334

Base current
E 37.5
r::;
'\"3 x II
3
x 10 = 1968A mpers

ngi
nee
I, =11SSL40.23° (I968xO.S87L400.23)
12 =S17L32.1O° (1968x0.41SL32°.IO)
The per unit MVA line I rin
g.n
S, = P, + jQ, = 1.00 LOo x 0.587 L-40.23° = (0.448 - jO.379) p.u.
The per unit MVA from line 2
S2
P,
=

= 0.448 x 37.5 = 16.8 MW

et
P2 + jQ 2 = 1.00 LO O x 0.415 L-32.1 0 = (0.352 - jO.220S) p.u.

P2 = 0.352 x 37.5 = 13.2 MW

Q, = 0.379 x 37.5 = 14.21 MVAR (leading)
Q 2 = 0.2205 x 37.5 = 8.27 MVAR (leading)
(b) The tap changer at station A on the sending end transformer in line 2 will alter the
reactive power loading.
Total load = (P, + P2 + jQ, + jQ2) = (30 - j22.S) MVA
Equal reactive powers means

Q, "" Q 2 = - 2~.S = -11.25MVAR

Downloaded From : www.EasyEngineering.net

326 Operation and Control in Power Systems

P 1 = 16.8 MW
and P2 = 13.2 MW
The change in voltage ~V is given by the approximate expression,

~ V = ZI = PR + QX
V
= tV _ V = PR + QX
V

where t is the tap changer setting ratio

ww (t - I) V =
PR+QX
V

w.E (t - I) =
PR+QX
V2

asy t = 1 + PR+QX

E V
2

ngi
P = P2 = 0.352 p.u. (= 13.2 MW/37.5)

X2 = 0.186 p.u. nee

Q 2 = -0.30 p.u. (= I I .25MVAR/37.5)

R] = 0.062 p.u.
rin
t = 1+ 0.352 x 0.062 + (- 0.3)0.186
1.00 x 1.00
g.
= 1- 0.034 = 0.966
net
percentage tap change = 96.6 - 100 = -3.4 %
Power sent through line 2
S2 = P2 + j Q 2 = (0.352 - jO.30)pu
S1 = S - S2 = (0.8 - jO.6) - (0.352 -j0.3)
= (0.448 - jO.3)p.u.

S1 = P 1 + j Q 1 = 16.8 - j 11.25 MVA

S2 = P2 + j Q 2 = 13.2 -j 11.25 MVA
P 1 = 16.8 MW
P2 = 13.2 MW
Q 1 = -I 1.25 MVAR
Q;! = -11.25 MVAR

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 327

E 8.3 A load of (15 + j 10) MVA is supplied with power from a generating station from a line
at II Okv 3-phase 50Hz. The line is of 100km length. The line is represented by 1t model
with the parameters
R = 26.4 ohms
X = 33.9 ohm
B = 2!-9 x 10-6
Voltage at t~e generated in 116kv. Determine the power supplied by the generating station.

ww B/2

w.E
a syE
SR = PR + j OR
Fig. E 8.3

=
IS + jlO
3 MVA/Phase

ngi
~ = 63.5KV
VR =

nee
Reactive power through
B
2" at receiving end
rin
OCR = 63.5 x 63.5 x
219 X 10- 6
6
x 10 var g.n
= 883062.75
2
= 441531.37VAR
. .
2
= 0.44151MVAR et
Receiving end power

IS. + jl 0'_ .0.441S3

S = PR + OR = . 3 . J

= (S + j 2.8918) MVA
Real power loss in the line = 12 R •
6
SR (S + j2.89I8)I0
I= - = -'--......;;,.----
VR (63.5)2 x 10 3

JS2 +2.8918 2 xl0 6

------- = 90.96
63.Sx10 3

Downloaded From : www.EasyEngineering.net

328 Operation and Control in Power Systems

12R == (90.96)2 x 26.4YA

==0.2184313MW
3
12X == reactive power loss == 52 +2.89: xl0 x33.9==(90.96)2 x33.9YA
63.5
Reactive power in the sending end capacitance 8/2
Oes == y2 Xes

116
Ys == .J3 == 66.9746

ww 6
Oes == 66.9746 2 x 219x 10- x 10: == 0.491 I 72.87MYAR

w.E
Power at the sending end
2 10

asy Ss == S + 12 R + 12 X + OCR + Qes

En == (5 + j2.89) + 0.218 + jO.279 - jO.04415

== (5.218 + j 1.9574) MYA

gi
power consumed from the station
3 x Ss == 3 [5.218
nee
-tl j 1.9574] MYA == (15.54 + j5.8722)MYA

rin
E 8.4 A short line having an impedance of (2 + j3) ohm interconnects two power stations A
and 8 both operating at II kv; equal in magnitude and phase. To transfer 25MW at 0.8

g.n
power factor lagging from A to 8 determine the voltage boost required at plant A.
Solution:

I == 25 x 10
.J3
3

x II x 0.8
== 1640.2476L36.870
et
== 1640.2976 (0.8 - jO.6) == (1312 - j984.1486)A
voltage drop in the short line == (1312.2 - j984.1486) (2+j3)
== 2624.4 - j 1968.297 + j3936.6 + 2952-446
== 5576.8458 + j 1968.3
This voltage drop must be compensated by the booster so that the voltages are maintain
the same at both the ends
Voltage boost == (5576.85 + j 1968.3) volts.

8.29 Static Compensators

In Recent years reactive compensation of charging power is made feasible with the application
of 3-phase, thyristor, power controller circuits with automatic control functions

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 329

(a) (b) (c)

Fig. 8.36 Idealized var compensators

ww The term static var compensator is applied to a number of static var compensation
devices for use in shunt reactive control.

w.E
These devices consist of shunt connected, static reactive element (linear or nonlinear
reactors and capacitors) configured into a var compensating system. Some possible

asy
configurations are shown in Fig. 8.36.

En
Even though the capacitors and reactors in Fig. 8.36 are shown connected to the low
voltage side of a down transformer, the capacitor banks may be distributed between high and
low voltage buses.
gi nee
The capacitor bank often includes, in part, harmonic filters which prevent tharmonic
currents from flowing in the transformer and the high voltage system. Filters for the 5th and

r
7th harmonics are generally provided. Fig. 8.37 shows one type of static var compensator.

ing
The thyristor controlled reactor (TCR) is operated on the low voltage bus.

connected to the secondary of a transformt"T.

.
In another form of the compensator illustrated in Fig. 8.38 the reactor compensator is

net
~ HYBus
Step down
transformer

Capacitor
and filter
bank

Fig. 8.37 Thyristor controlled reactor compensator system

Downloaded From : www.EasyEngineering.net

330 Operation and Control in Power Systems

With this transformer, the reactive power can be adjusted to anywhere between 10% to
the rated value. With a capacitor bank provided with steps, a full control range from capacitive
to inductive power can be obtained. The reactor's transformer is directly connected to the
line, so that no circuit breaker is needed.
The primary winding is star connected with neutral grollnded, suitable to the thyristor
network. The secondary reactor is normally nonexistent, as it is more economical to design
the reactor transformer with 200% leakage impedance between primary and secondary windings.
The delta connected tertiary winding will effectively compensate the triplen harmonics.
The capacitor bank is normally subdivided and connected to the substation bus bar via
one circuit breaker per subbank. The regulator generates firing pulses for the thyristor network

ww
in such a way that the reactive power required to meet the control objective at the primary side
of the compensator is obtained. The reactor transformer has a practically linear characteristic

w.E
from no load to full load condition. Thus, even under sllstained over voltages, hardly any
harmonic content is generated due to saturation. The transformer core has non ferromagnetic
·gaps to gil'e the required linearity.

asy
The following requirements are to be borne in mind while designing a compensator.

En
1. Reaction should be possible, fast or slow, whenever demanded. No switching of
capacitor should take place at that time to avoid additional transients in the system.

gi
Commutation from capacitor to reactor and vice versa should be fast.

nee
2. No switching of the capacitors at the high voltage bus bar, so that no higher frequency
transients are produced at EHV level.

r ing
3. Elimination of higher harmonics on the secondary side and blocking them from
entering the system.

.
In a three phase system the thyristor controlled inductors are normally delta
net
connected as shown in Fig. 8.39 to compensate unbalanced loads and the capacitors may be
star or delta connected.
Transmission line

HV
Bus

r rJMY~
Reactor ,
transformer

Fig. 8.38 Reactor - transformer compensator

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 331

8.31 Steady State Performance of Static var compensators

In the thyristor controlled reactor, the inductive reactance is controlled by the thyristors. For a
limited range of operation the relationship between the inductive current iL and the applied
voltage V is represented in Fig. 8.40. As the inductance is varied, the susceptance varies over
a range within the limits BLmln and BLmax (corresponding to XLmax and XLmln ) while the voltage
changes by v volts.

ww
w.E
asy
E
Fig. 8.39 Fixed capacitor, thyristor controlled inductor type var compensator

ngi
The use of a fixed capacitor in parallel with the controlled reactor results in the

nee
chaltacteristics shown in Fig. 8.41. The currents can be controlled to have any desired value
over the range selected from leading to lagging.

rin v

g.n
et
o

Fig. 8.40 Thyristor controlled reactor charactenstics

Corresponding vars are produced or absorbed. Fig. 8.41 shows the effect of changing
the oonduction angle 0 from 1800 to a small value 0c-
The current flowing in the inductance would be different in each half cycle, varying
with the .conduction angle such that each successive half cycle is a smaller segment of a sine
wave. The fundamental component of inductor current is then reduced to each case. Quick
control can be exercised within one half cycle, just by giving a proper step input to the firing
angle control

Downloaded From : www.EasyEngineering.net

332 ~
Operation and Control in Power
62
Systems

Block diagram representation of a static var compensator is given in Fig. 8.42. The net
change in the reactor current IR drawn by the static var compensator supplied through the
i1 V 1
system impedance Zs =
i1! I

v v

~--
/

ww
/
/

w.E +-(Lead)

Applied voltage
(Lag) -+ +-(Lead) (Lag) -+

asy
E ngi
nee
Reactor currents
rin
Fig. 8.41 Effect of conduction angle on reactor currents
g.
Following an instantaneous change in firing angle e, the admittance BL of the inductor
and therefore the current, 'I will settle down to a new value in about one cycle or less. The net
impedance Zs is the entire system impedance viewed by the compensator. The thyristor
controlled reactor can be represented by the transfer function
\ e- ST1
GR(S)=--
1 +T2S
where TI and T] are of the order I ms and 4 ms. For practical purposes, the compensator
can be assumed to have a pure gain of unity with the present thyristor technology. The AVR
can be represented by

Transformer
Fig. 8.42 Block diagram of static var compensator

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 333

Gv(S) = _k_ (I +T4 S) ..... (8.27)

I+T3S (I+T5S)
The time constant T) lies between 50 and ISO ms.
The compensator block diagram for dynamic studies is shown in Fig. 8.43.

ww V-E (2a-sin2a) E

w.E
j = -jX- : E=CY. C = --2-n---'-

Fig. 8.43 Compensator block diagram for dynamic studies

asy
Static var compensators when installed reduce the voltage swings at the rolling mill and

En
power system buses in drive system applications. They compensate for the average reactive
power requirements and improve power factor.

gi
Electric arc furnaces impose extremely difficult service requirements on electrical power

nee
systems since the changes in arc furnace load impedance are rapid. random and non symmetrical.
The three phases of a static var compensator can be located independently so that

r ing
it compensates for the unbalanced reactive load of the furnace and the thyristor controller
will respond quickly in order to minimize the voltage fluctuations or voltage flicker seen by

.
the system.

Lme net
Furnace

Input ;=~=::!E~=~::; Light pulses

~oltageand
current)

Magnetic interface

Fig. 8.44 Arc furnace applications

Downloaded From : www.EasyEngineering.net

334 Operation and Control in Power Systems

Thus, the furnace characteristics are made more acceptable to the power system by the
static var compensator. Fig. 8.44 shows the application of the static var compensator to an arc
furnace installation for reactive power compensation at the HV bus level.

8.32 Overvoltages on Sudden Loss of Load

Over voltages caused by load rejection are basically caused by a voltage rise across the source
reactance which may be aggravated by over speed causing a decrease in the line capacitive
reactance and increase in inductive reactance. There are two types of over voltages following
sudden loss of load.

ww The first occurs with the first cycle or two after the load rejection takes place. The
second type, dynamic in nature, may last a few seconds and is caused by higher machine
internal voltages and over speed. This type is, to a large extent influenced by the existing

w
conditions before the load rejection takes place, the type of voltage regulator and the

.Ea
compensating devices. The magnitude of the first type can be determined from:
I. The constant flux linkage voltage in the equivalent circuit of the rotor direct and
quadrature axes.
syE
2. The voltage rise components across the transient reactances of the two axes due to
line-charging current.
ngi
The second type of over voltage can be computed from the relation.

nee .. ... (8.28)

r ing
where
Es is the sending end voltage
E'(t) the voltage behind transient reactance as a function of time
. net
Qc' the reactive MVA generation of the line at rated voltage and frequency and,
SC' the short circuit MVA for fault at sending end (based on transient reactance).
If the rejected load forms a large portion of the total load there will be a rise in frequency,
Qc and E!(t) and decrease of Sc with a cumulative effect to increase the overvoltage.
To avoid self excitation: the line reactive generation should be within
_ O.9Pn
Qc - ? ..... (8.29)
n-(Xd+X t )
where n is the over speed in p.u.,
Xt the transformer reactance in p.lI.
Xd the direct axis reactance in p.u. and
Pn the nominal MVA of the generating capacity at the sending end.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 335

If the excitation system has negative field current capability and AVR is provided then

where Xq is the quadrature axis reactance in p.u.

The over voltages resulting from self excitation are dangerous not only for the system
but also to machines and transformers connected to the sending end. This conditions must be
avoided by providing the excess ofline reactive vars with shunt compensator or rapid switching
I
off of lines. Further, over voltages caused by load rejection prove to be worse when switching

ww
is performed on the transformer secondary at the receiving end. Due to voltage increase, the
transformer gets over excited and then saturated with subsequent injection of harmonic voltages

w.E
into the system. Then, with some combination of system inductance and line capacitance, due
to Ferro resonance, extremely high voltage will be produced for 0.5 to LOs, resulting in
equipment damage.

a syE
Shunt reactor compensation is quite effective in limitings the over voltages due to
load rejection. Such compensation has given 50% reduction in over voltages for line to
ground faults.
ngi
When synchronous compensator is installed, it automatically absorbs the reactive power

nee
from the system which is a function of the compensator reactance and interval voltage.

rin
The time constant of the machine, however, limits the rate at which the compensator
inductive current can be increased. The static var compensator on the other hand can change

g.n
over to its maximum current in the inductive range very rapidly since the compensator output
increases as the square of the voltage. Further, the overload capacity of the static var

et
compensator in the inductive range can be increased at additional costs much lower than those
required for synchronous compensators.

8.33 Voltage Dips

Voltage dips take place either due to faults or switching in of other motors connected to the
same bus. A voltage dip causes the synchronous condenser to act in accordance with its
internal voltage behind the sub transient and transient reactance and immediately starts feeding
the corresponding capacitive reactive power into the system. In fact, with modern excitation
control equipment with an appropriate ceiling voltage, it is possible to maintain a high capacitive
current for a few seconds to support the system voltage. Static var compensator, on the other
hand, as it consists of passive elements only, while reacting very quickly fails to support the
system voltage since the capacitive current decreases as the square of the voltage. At large
voltage dips, it is observed that the synchronous compensator is superior in performance to
static var compensator.

Downloaded From : www.EasyEngineering.net

336 Operation and Control in Power Systems

Design 0/ Filter Circuit

The maximum harmonic content permitted in the delta connected system as per IEEE
specifications are
for the 5th harmonics - 5% of fundamental
and for the 7th harmonics - 2.5% of fundamental
Since the compensator circuit is delta connected, third harmonic is not considered.
I fundamental = 260A

(~)
ww
15th harmonic = 260 = 13.0 A
100

w.E 17th harmonic =

2.5 )
260 ( 100 = 6.5 A

asy
Filter for fifth harmonic
I
ro ==---

En
At resonance
n ~LFCF

gi
The impedance offered by filter circuit at power frequency is

l-ro 2 L FC F
nee
roC F
rin
1
and this should be much greater than the impedance offered by CM which is -CM, i.e.
0)
g.n
et
we require
l-ro 2 L FC F > _1_
roC F - roC M
Let this be greater be 20 times the value
1
C F == (nroY LF

2
LF = 1316 xOIO- 6 x31421x24 ==0.OO643H

1
C F == , ~ == 0.00006303 == 63.03J.LF
(5 x 314)- x 6.43 x 10-'

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 337

Filter for seventh harmonic:

The current for 7th harmonic = 6.5A
nf = 350Hz

I
nro = ----;::==
~LFCF

ww Again assuming a factor of 20

w.E LF =
20
6 7
1316xlO- (314)- x48
= 0.000 16H = O. 16mH

Therefore, asy CF =
I
,=0.001284F=1284f.lF

En
7'
(7x314t xO.16IxI0-

gi
8.34 Subsynclrronous Resonance

nee
Series capacitor compensation is the most technically feasible and economically viable choice

rin
for increasing the transmission capability ofEHV power systems. However, series compensation
brings in certain problems related to dynamic performance of the system. One such problem

g.n
is sub synchronous resonance. It is an electrical system condition where the electrical system
exchanges energy with a turbine-generator at one or more frequencies below the synchronous
frequency of the system. This interchange may be a Iittle damped or undamped.
et
The exchange of energies takes place at the natural frequencies of the respective systems.
For the electrical system, the natural frequencies are higher compared to the mechanical system.
The range lies between 15Hz and 45Hz for steam turbo generator whereas it is 10Hz and below
for a hydro generator. For the electrical system, the modes of oscillation depend upon the
number of circuit configurations that can be made through switching. Depending on the system
operating conditions and mechanical moment of inertia, the electromechanical system has a
frequency of oscillation between I and 2 Hz.
There are three types of sub synchronous oscillations that have been identified. The
apparent negative resistance (rotor resistance viewed from armature terminals) characteristic
of the generator at sub synchronous frequencies causes the generator to act as an induction
generator and this effect is called the induction generator effect. If this negative resistance
exceeds the network resistance for a specific combination of inductance and capacitance, self
excitation may result anhe resonant frequency.

Downloaded From : www.EasyEngineering.net

338 Operation and Control in Power Systems

Resonance may take place due to the torsional interaction between electrical and
mechanical systems. This may occur when the electrical resonance frequency is near the
complement of the torsional resonant frequency of the turbo-generator shaft system (the
difference of synchronous frequency and the turbine mechanical natural frequency is the
complement of the torsional resonant frequency).
A small voltage induced in the armature by rotor oscillation under these conditions can
result in large sub synchronous currents. This current will produce oscillatory component of
rotor torque and if the mechanical damping is small, the electromechanical system may
experience growing oscillations.

ww Finally, when the transmission system contains a series capacitor, the transient electrical
torque, under fault or switching operations, may contain large amplitude frequency components

w.E
close to resonant frequency of the shaft system. This can result in high shaft torques which
may be seriously damage the turbo generator units.

asy
Consider a series compensated transmission system, let L1, Lg and Lt be the inductances
of the line, generator and transformers respectively and let Cs be the series capacitance then
the resonant frequency.
En
gi nee
otherwise,
r ing
=(j)
s
~c
X
L . net..... (8.30)

where fs is the rated frequency; Xc and XL are the capacitive and inductive reactances
calculated at the rated frequency. Since the compensation is generally less than 100% from
eqn. (8.30) it can be seen that the resonant frequency is always less than the rated frequency.
Several methods are suggested to prevent sub synchronous resonance. One technique
is the use of filtering and damping. A blocking filter may be inserted in series with the generator
step-up transformer.
Otherwise, an appropriately designed reactor may be connected in parallel with an existing
series capacitor to form a blocking filter. Other types of filters are also suggested in literature.
Relays are also used to detect excessive mechanical system. System switching method can
also be implemented to avoid sub synchronous resonance. When a system configurations
occurs which can the generator can be isolated from the series capacitances that create
problems. This is accomplished by switching the machine off from the compensated system
and switching on to an uncompensated line. Unit tripping also can be implemented if all
conditions that cause transient torques are completely predetermined.

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 339

E 8.5 A load of (66 + j60) MVA at the receiving end is being transmitted via a single circuit
220KV line having a resistance of 21 ohm and reactance of 34 ohm. The sending end
" voltage is maintained at 220kV. The operating conditions of power consumers require
that at this load voltage drop across the line should not exceed 5%. In order to reduce
voltage drop standard single phase 660V, 40 kVAR capacitors are to be switched in
series in each phase of the line. Determine the required number of capacitors, and rated
voltage neglect the losses.
Solution:
Three phase load = (66 + j60) MVA

66 + j60

wwPer phase load 3 = (22 + j20) MVA

w.E
The impedance Z = R + jX = (21 + j34) ohm

Phase voltage v = 220 == 127.02kV

asy
Without series capacitor
P fj

En = 22 x 21 + 20 x 34 = 8.9907KV
gi
6. V == PRR + OR X

Permitted drop in voltage =

5% of 127.02
127.02

nee
== 462 + 680
127.02

== -
5
x 127.02 = 6.351kV
rin
g.n
100

Let Xc be the capacitive reactance to bring down the drop from 8.9907kV to 6.351 kV

6.351 = PRR+(X-XC)QR
VR
22x21+20(34- X d
127.02
et
Solving for Xc
Xc = 16.65 ohm
Given that 40kVAR capacitor at 0.66kV are to be switched.

I' = 40kVAR = 60.606A

c 0.66kV

V R IR ~ (22 +j20) MVA

I = ~222 +20~ xl0

6
.J884 x 10
1
= 234.07 Am s
R 127.02x10 1 127.02 p

Downloaded From : www.EasyEngineering.net

340 Operation and Control in Power Systems

capacitors per phase in parallel

= 234.07 = 3.862
n 60.606

The number of capacitors must be either 4 or more than 4.

Vc 660V
Xc =- = = 10.890hm
Ie 60.606A

Each capacitor has 10.89 ohm reactance.

ww Four in parallel makes Xc parallel =

10.89
-4- ohm

w.E
Since total Xe has to be 16.65 ohm

asy
The series units ns will be

10.89 x ns = 16.65
4
En
Hence n
s 10.89 gi
= 16.65 x 4 = 6.1157
nee
As ns must be a whole number n,
Installed capacity of the capacitor bank
= 7
r ing
Qc
=
= Rated voltage of the capacitor bank
0.66 kV x 7 == 4.62 kV . net
E 8.6 Find the capacity of a static VAR compensator to be installed at a bus with ± 5% voltage
fluctuation. The short circuit capacity == 7000MVA.
Solution:
i1Q == capacity of the compensator
Sse == Short circuit capacity
i1 V == Voltage fluctuation

i1 V = i1Q
Sse
i1Q == i1 V . Sse == ± 0.05 x 7000 == ± 350 MVAR
Capacity of the static VAR compensator = ± 350 MVAR

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 341

E 8.8 A 400kV line is fed through a 132!400kV transformer from a constant 132kV supply. At
the load end of the line another transformer of nominal ratio 400!132kV is used to
reduce the voltage. The total impedance of the line and transformers at 400kV is (50+j 100)
ohm. Both, transformers are equipped with tap changing so arranged that the product
of the off nominal setting is unity. If the load on the systems is 250MW at 0.8 p.f
lagging. Calculate the settings of the tap changer required to maintain the voltage of the
load bus at 132kY.

ww 132/400

w.E Fig. E 8.8

250MW
0.81g

Solution:
a syE
ngi
Power per phase (P) = 250 = 83.33MW
3
nee
rin
250 0.6
Reactive power per phase = - x - = 62.5MVAR
0.8 3

132
VR = VS = -Jj = 76.21247 kV 76.21247 g.n 103 V

et
= x

we have

132
V =V = -kV
R S Jj

when the secondary is maintained at 132kV the primary is kept at 400kY.

Downloaded From : www.EasyEngineering.net

342 Operation and Control in Power Systems

Since total impedance is referred to 400kY side we obtain

ww
w.E
------=-----
1 _ 4166.5 + 6250 1_~6.5
53333.333 53333.333
1 - 0.1953093
- - - = 1.2427136
0.8046906

asy ts = JI.2427 136 = 1.1 147706

ts = I
En
1.1147706
= 0.8970455

gi nee
r ing
. net

Downloaded From : www.EasyEngineering.net

Voltage and Reactive Power Control 343

Questions

I. Discuss in detail about the generation and absorption of reactive power in power system
components.

2. Explain reason for variations of voltages in power systems and explain anyone method
to improve voltage profile.

3. Explain clearly what do you mean by compensation ofline and discuss briefly different

wwmethods of compensation.

4. What i~ a static compensator? Explain with diagrams working principle of various

w.E
types of static compensators.

asy
5. Explain with diagrams, the operation of a fixed capacitor and thyristor controlled reactor.

6. Discuss in detail about the generation and absorption of reactive power in power system
components.
E ngi
7. Discuss in detail about the generation and absorption of reactive power system
components.
nee
8. Describe the effect of connecting series capacitors in the transmission system.

9. (a) What does on mean by load compensation? rin

(b) With neat diagrams discuss shunt and series compensation
g.n
(c) What are the specifications of lead compensator?
et
10. Discuss the advantages and disadvantages of different types of compensating equipment
for transmission systems.

11. Discuss in detail the voltage stability problem in power systems.

Downloaded From : www.EasyEngineering.net

344 Operation and Control in Power Systems

Problems

P 8.1 Find the regulation and efficiency of an SO-lan, 3-phase, 50-cis transmission line delivering
24,000 kVA at a power factor ofO.Slagging and 66kV to a balanced load. The conductors
are of copper, each having resistance 0.12 ohm per km, 1.5 c, outside diameter, spaced
equiIaterally 2.5m between centers. Neglect leakance and use the nominal-7t method.

P 8.2 A 3-phase, overhead line has resistance and reactance of 6 and 20 ohm respectively per
phase. The sending end voltage is 66kV while the receiving end voltage is maintained at
66kV by a synchronous phase-modifier. Determine the kVAr of the modifier when the
load at the receiving end is 75MW at power factor 0.8 lagging; also the maximum load

ww that can be transmitted.

w.E
P 8.3 A long line from a hydroelectric station operating at 132kV feeds a 66kV system through
a transformer. The load taken from the 66kV windings of the transformer is 50MVA at
power factor O.S lagging. A tertiary winding on the transformer feeds a synchronous

asy
condenser at II kV. If the power factor at the receiving end of the line is to be unity,
calculate the rating of each of the three windings. Neglect losses.

En
P 8.4 Find the rating of synchronous compensator connected to the tertiary winding of a

gi
132kV star connected, 33kV star connected, Ilk V delta connected three winding

nee
transformer to supply a load of 66MW at O.S power factor lagging at 33kV across the
secondary. Equivalent primary and tertiary winding reactances are 32 ohms and 0.16

r
ohm respectively while the secondary winding reactance is negligible. Assume that the

ing
primary side voltage is essentially constant at 132kV and maximum of nominal setting
between transformer primary and secondary is I: I.

. net
P 8.S A single circuit three phase 220kV line runs at no load. Voltage at the receiving end of
the line is 21 OkV. Find the sending end voltage, if the line has resistance of 20.5 ohms,
reactance of 81.3 ohms and the total susceptance as 5.45 x 10-4 mho. The transmission
line is to be represented by 7t-model.

P 8.6 A )ower system is operating at 1000Mw, 132kV, 50Hz, with O.S p.f. lagging in parallel
with another line at 750MW, I 32kV, 50Hz with 0.707 p.f. lagging. Both are interconnected
at the station and when the compensating device is on, the overall power factor is
improved to 0.9 lagging. Suggest suitable capacitors, shunt and series. Individual loads
and combined load are to improve power factor to 0.9 in all c~ses.

Downloaded From : www.EasyEngineering.net

ww
9 w .Ea TO
INTRODUCTION
ADVANCEDsTOPICS
yE
ngi
nee
Modern power systems with super power plants are highly complex and sophisticated with

r
huge capital investment. Further, advanced technologies are used in the design and operation

ing
of these systems. For efficient and reliable operation of these systems, control action is
envisaged with the following organs:
I.
2.
3.
Data acquisition and control
Computers,
Man - machine interface, and
. net
4. Software and human operators.
During the earlier days, control action was appl ied to generation dispatch and supervisory
control. Modern systems demand a comprehensive and integrated approach to monitoring and
controlling of power flows for economic and secure operation. This brought into the picture
a third and equally important aspect of control, viz, security control. The human operator is
finding it more and more difficult to take instantaneous decisions in cases of serious and
complex situations.
Generation control has slowly shifted from analog to digital methods during the sixties.
Station and system supervisory control too has shifted to digital computer masters from hard-
wired masters. The man - machine interface consisted of strip chart records, loggers, indicating
lights, anunciators, console push button panels, etc. Black and white or colour CRT display is

Downloaded From : www.EasyEngineering.net

·346 Operation and Control in Power Systems

already in vogue. However, the entire control configuration has undergone a radical change
with the introduction of system security requirements.

9.1 FACTS Controllers

In chapter 8 static var compensators are discussed. Subsequent developments lead to the
emergence of Flexible AC Transmission System (FACTS) controller. Hingorani and Gyugi are
pioneers in this area. FACTS controllers is a fast developing area with potential applications.
The basic principle of the flexibility in the transmission of power can be understood
easily from the following example Fig. 9.1.

ww Line I

w.E 1----+ Load

asy Line 2

En
gi Fig. 9.1 Transfer of Power

nee
The power flow from the generator to the load through the parallel lines is dependent

r
upon the impedances or reactances when resistances are neglected. If Xl = X2 ' the transfer

ing
of power is 50% on each of the two lines. If X2 = 2X I then the p.ower transfer will be in the
ratio of 2 : I. Thus the higher impedance line carries less load and the lower impedance line

. net
may even get over loaded. If one of the lines, say line 1 is a HYDC line then power flow is
electrically controlled and the above problem is eliminated. Further, with HYDC the stability
problem also is controlled because of the speed of control. However, HYDC is expensive in a
relative sense.
A FACTS controller can over come the problem discussed earlier by controlling the
impedance, or phase angle. A FACTS controller can control the power flow in any manner
that is desired. It is possible to inject desired voltage in series with the line.

Types of FACTS Controllers

Generally FACTS controllers are divided into four categories. They are briefly discussed.

9.1.1 Series Controllers

They are either variable reactor or capacitor or a power electronic based variable source so as
to serve the need. The drop of voltage across the variable impedance can be considered as
series injected voltage (Fig. 9.2).

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 347

H~
Line

Fig. 9.2 Series Controller

9.1.2 Shunt Controller

The shunt controller, also, may be either a variables source or a variable impedance or a

ww
combination connected in shunt as shown and they all inject current into the line (Fig. 9.3).

w.E Line

fasy
En
gi nee
Fig. 9.3 Shunt Controller

9.1.3 Series - Series Controllers

r ing
Two different types of configurations are feasible. A combination of separate series controllers
operated in a coordinated manner in the network or other wise, it could be a unified controller

. net
wherein series controllers provide independent series reactive power control as well provide
real power transfer through a power link. This is shown in Fig. 9.4.

I l A. C. Line 1
~
I I
D. C. Power link for real
power exchange

A. C. Line :2
I ~
l
I I

Fig. 9.4 Series - Series Controllers

In FACTS terminology the term unified is used to indicate that the D.C. terminals of all controller
converters are connected together fm real power transfer.

Downloaded From : www.EasyEngineering.net

348 Operation and Control in Power Systems

9.1.4 Series - Shunt Controllers

In the same way as series - series, this could be also a combination of separate series and
shunt controller - controlled in a coordinated manner or a unified power controller with series
and shunt elements. The basic scheme is shown in Fig. 9.5.

-v--.
Line

D. C Power link for

ww f Real power exchange

w.E Coordinating unit

asy
Fig. 9.5 Series - Shunt Controllers

9.1.5 Power Flow Control

En
There are two types of converters - voltage sourced and current sourced, using gate turn off

gi
devices. The voltage sourced converter consists of a gate turn off device paralleled with a
reverse diode. A D.C. capacitor is used as a voltage source. The current sourced converter

nee
has a gate turn off device in series with a diode. It has a D.C. reactor as a current source.
Voltage sourced converters are preferred in FACTS controllers. Sequential switching

r ing
of the device will give A.C. voltage from the D.C. source. They are called thus static synchronous
compensator. It is operated as a shunt connected static var compensator whose output voltage

. net
can be controlled independent of A.C. system voltage. Such a device is also called STATCOM.
The structure is shown in Fig. 9.6.

o
Fig. 9.6 Static Synchronous Compensator

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 349

9.1.6 Static Var Compensator(SVC)

It is a device that supplies capacitive or inductive current so as to maintain, voltage at any bus.
This has been already discussed in Chapter 8. The device is shunt connected.
In a similar manner a thyristor controlled reactor (TCR), shunt connected has the effective
reactance varied in a continuous manner by conduction control of the thyristor. TCR naturally
belongs to the class of SVC.
If the reactance is varied in a stepped manner instead of continuous mode, the device
becomes thyristor switched reactor TSR. Several shunt connected reactors may be switched
in or out.

ww Likewise, thyristor switched capacitor (TSC) is operated such that shunt capacitor
units are switched in - and out depending upon the need. It should be noted that in this case

w.E
the thyristors are operated without firing angle control. They operate with full or zero conduction.
A thyristor controlled resistor (TCBR) with shunt connection can be use to brake a

asy
generator during accelerating period when a fault occurs. They are called then thyristor controlled
braking resistors.

En
In the same way as STATCOM a static synchronous series compensator (SSSC) operates
with a voltage sourced converter or current sourced converter giving output voltage in series

gi
with the line. Thyristor controlled series reactor (TCSR) thyristor controlled series capacitor

nee
(TCSC) thyristor controlled switched series reactor (TSSR) are similar devices in series mode
of operation.

9.1. 7 Unified Power Flow Controller

r ing
.
This is a combination of series and shunt connected static compensators (STATCOM and

flow of active power between both the series output terminals of the SSSC and the shunt net
SSSC). They are coupled through a common D.C. link. The controller permits bi-directional

output terminals of the STATCOM. Without any external electric energy source they provide
simultaneously real and reactive series line compensation.
UPFC is also capable of providing independently controllable shunt reactive compensation.
This is schematically shown in Fig. 9.7.
There are several more devices such as thyristor controlled phase shifting transformer
(TCPST), inter phase power controller (lPC) Thyristor controlled voltage limiter, thyristor
\
controlled voltage regulator etc.

9.1.8 Advantages due to FACTS devices

The following are some of the advantages that can be obtained by using FACTS controller.
I. Power in lines is controlled in any desired manner.
2. Line capacity is increased, practically upto thermal limits
3. By raising transient stabilitv limit, system security is enhanced

Downloaded From : www.EasyEngineering.net

350 Operation and Control in Power Systems

----~~------------~~~------
Line
~ ~
I

sssc
0
ww
w.E D. C. Link
=}=
0 Statcom

asy Fig. 9.7 United Power Flow Controller

En
4. Up gradation of lines is easier.
5. Reduced reactive power flow, thereby permitting greater active power flow.

9.2 Voltage Stability

gi
6. Reduced cost of energy received due to enhanced line capacity.

nee
r
Voltage stability is an integral part of the power system response and is an important aspect of

ing
system stability and security. Voltage instability has been detected well before the onset of
angle instability in many cases. If the problem is not corrected it can lead to voltage collapse
and system wide disturbance.
. net
The loss of synchronism of generators is called angle instability. Voltage stability also
called load stability is a subset of overall stability of a power system and is a dynamic problem
that occurs due to monotonically changing voltages.
In industrialized areas increase of load demand is met without a corresponding increase
in transmission capacity leading to severe problem including voltage stability.
Voltage stability is the ability to maintain the voltage so that when load is increased load
power will increase and so both power and voltage are controllable.
A power system at a given operating state and subjected to a given disturbance is voltage
stable ifvoltage near loads approaches post disturbance equilibrium values. Following voltage
stability, a system undergoes voltage collapse if the post disturbance equilibrium voltages near
loads are below acceptance limits. Both voltage stability and collapse may occur in a time
period of a fraction of a second to a few minutes.
Consider the plot of voltage V at a load bus as a function of load power P shown in
Fig. 9.8 as ABCDE.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 351

v
A _" B
v I '1---- ---,
, C
- - --i - - )Nose pomt

1/1
1/ 1
v
)/iu
Yo 1
1

//
~------~--~~-------- P

ww E PeR
Fig. 9.8 P. V. Diagram at the Load Bus

w.E
It can be observed that there are two voltages Y I and Y2 for a constant power load. The
high voltage solution is stable while the other is unstable. The maximum loadability is determined

a
by point C and the part CDE is uncontrollable.

VS~-----,---VR~
R+j:-.
syE
ngi
I ...

Line l.oad
P +.IQ

nee
Fig. 9.9 Transmission Line with load

r i +Q2)]g
Mathematically the voltage YR can be expressed by (see Fig. 9.9).

V =[ -2;Y ±1~(2QX-Y(r _4X 2(p2 n

.n
s ..... (9.1 )
R
/
from whi6h it is clear that Y R is a double valued function.
For differe~ power factors the PY characteristic can have the shape shown in Fig. 9.10.
et
v
/' Nose of the
curve

/
08 (load)

L -_ _ _ _ _ _ _ _~~_ _ _ _ _ _ _ _ _ _ P
E Power

Fig. 9.10 P. V. Curves for Different Power Factors

Downloaded From : www.EasyEngineering.net

352 Operation and Control in Power Systems

Static analysis is sufficient to assess voltage stability, but for accurate prediction dynamic
analysis must be carried out.
The probability of voltage instability increases as the system is operated close to its
maximum loadability limit. Environmental and economic expansion of transmission network
and for obvious reasons distant location of the generators from the load centers all result in
over loading of the existing networks. The present trend is to optimally utilize the inher\!nt
margins available with flexible A.C. transmission system controllers.
Reactive power compensation close to the load centers as well as at the critical buses in
the network is essential to over come voltage instability.

ww The location size and speed of control of FACTS controllers determine the maximum
benefit that can be obtained without getting into voltage instability.

9.3
w.E
Power Quality
Any amount of power can be drawn from and delivered to an infinite bus without affecting

a syE
(i) the magnitude of bus voltage
(ii) the frequency of bus voltage.

ngi
The short circuit ratio at a load bus (SCR) is defined as
SCR = Short circuit current at rated voltage with load short circuited

nee
rated load current
It can be seen that SCR is iaversely proportional to the Thevenin's impedance of the

rin
network as seen from the bus. If SCR is low the load bus is said to be connected to weak
power grid and vice versa. Power quality (PO) problem is likely to occur at that bus. One
aspect of PO can be selected in terms of phase voltages V R' V) and V B as

1
g.n
[
V
V: =
sin rot
J2 V sin( rot - 231£ ) et
..... (9.2)

VB (2 )
sin rot + 31£

under all conditions, V and ro remaining constant. Similarly the harmonic content of the
bus voltage is also an indication of PO.
Electric power quality can be loosely defined as a measure of how well electric power
can be utilized by customers.
Power utilization is degraded when wave shapes are irregular, voltage regulation is
poor, harmonics and flicker are present or even when there are momentary events that distort
the usually sinusoidal voltage wave. All the above conditions result in degradation of
power quality. The introduction and the wide spread use of high power semi conduction
devices at all levels of power system operation, and utilization have made non sinusoidal load
currents quite co~mon.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 353

9.3.1 Power Quality Index

The harmonic components of voltages and current are used as power quality indices.
Let i(t) be the load current that is periodic
11' 12, ..... are the fundamental and harmonics of i(t).
For all general purposes, the standard is taken as total harmonic distortion (THO) defined
by
a I
THO = L--'-
1=2 II
..... (9.3)

ww In regard to frequency, high frequency phenomenon is termed noise and the low frequency
phenomenon is named flicker. As has been already mentioned rectifier loads are the main

w.E
source of harmonic load current.
Oifferent rectifier connections such as single phase bridge and 3-phase bridge rectifiers
contribute to harmonic distortion and various factors such as THO, displacement factor, power

asy
factor are clearly defined. It may be noted that displacement factor is the cosine of the angle
between fundamental voltage and fundamental current while, power factor is the ratio of real
or active power to apparent power V RMS JRMS '
En
9.3.2 Voltage Sags
gi nee
A short duration magnitude reduction in voltage is called voltage sag. The time period may be
of the order of few seconds. The cause for a sag is sudden, short duration increase in current.

r
Such an event occurs during motor starting, transformer energizing, faults etc.'
ing
For example, the transient voltage when a big motor is started will have the shape of all

.
the phase voltages shown in Fig. 9.11. The voltage falls by a large amount with gradual recovery
over a few cycles.
200
net
215
.,
~
210
'"Cl
C>()

~
.0
~
205
~
~
195

190

0 2 3 11
Time (Cycles)

- Fig. 9.~1 Voltage sag due to motor starting

Downloaded From : www.EasyEngineering.net

354 Operation and Control in Power Systems

It can be seen that the voltage drops are practically the same in all the three phases. In
case of magnetizing inrush current of transformers, the inrush current is not the same in all the
phases. This is shown in Fig. 9.12 second harmonic and fourth harmonics contribute more to
harmonic distortion.

II
10.9
10.8
10.7
~
ww ~
(l)

~
10.6
10.5

w.E
[/J
:::E
ex:
10.4
10.3

asy 10.2 -
10.1 .

En
10
0 5 10 15- 20 25

gi
Time (Cycles)

nee
Fig. 9.12 Voltage sag due to transformer energizing

rin
In case of faults, the most common type of fault is single line to ground fault. The
voltage falls more in two phases and recovers sharply after a few cycles. This is shown in
Fig. 9.13 the most severe faults are due to short circuits and faults involving earth.
g.n
11.5

II
et
~
10.5
>
(l)

~ 10
~[/J
:::E 9.5
ex:

8.5
0 2 4 6 8 10 12 14 16
Time (Cycles)

Fig. 9.13 Voltage sag due to a fault

Downloaded From : www.EasyEngineering.net

•
Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 355

Transformer c6nnections affect the impedance of fault. For example a phase to

phase fault for a star connected load gives a similar sag as a single-phase fault for a delta
connected load.
Further a voltage sag is not exactly recorded the way it occurs at the load due to
intervening components such as transformers which effect the voltage sags.

9.3.3 Rectifier Loads

Most electronic equipment (such as personal computers) with power rating less that I kw. Use
single phase diode bridge rectifiers. The output of the bridge rectifier is filtered using a capacitor
filter or an LC - filter. The diode bridge rectifiers are often followed by switch mode power

ww
supplies. The switch mode power supply converts the unregulated D.C. link voltage to a well
regulated output voltage. For higher power applications (greater than 1 KW) either 3-phase
bridge of six pulse type or twelve pulse type are used. The single phase case is shown in Fig.

w.E
A.C asy
mains
En SMPS Output

gi nee
Fig. 9.14 Single phase rectifier load

9.14. rin
g.n
The voltage sag depends upon the energy storage device used and the switch mode
power supply (SMPS). If the sag in the A.C. main voltage is of considerable magnitude and
duration, the D.C. link voltage may fall below the minimum value for which it is desired. This
results in the shut down of the .electronic load connected to the output termals of SMPS. et
9.3.4 Flicker
Light flicker perception is said to be a physiological process in which the eye and the brain
participate. Voltage fluctuations occur due to mainly electric arc. furnace loads. These occurs
because of voltage and current harmonics due to inherent non -linearity of the arc characteristic.
Light flicker is caused mainly due to reactive power flow from the system at the instant of
short circuit created by the arc. Any step taken in regard to control of reactive power will
control light flicker. Furnaces' of ratings about I OOMVA and above cause the problem.
Again when wind turbines are connected to system they are frequently switched on and
off around the cut - in speed then voltage fluctuations occurs.
Another major source for harmonics in addition to electronic loads is adjustable speed
drives as they use diodes, SCRs, power transistors, switches etc to chop wave forms to

Downloaded From : www.EasyEngineering.net

356 Operation and Control in Power Systems

control power.
Total harmonic distortion of voltage less than 5%' is acceptable but THO above 10%
requires, positive steps to reduce the harmonics.

9.3.5 Power Acceptability or Voltage Tolerance

The voltage tolerance or more commonly power acceptability curves are loci dra»,n in the bus
voltage duration tim~ plane. There curves indicate the tolerance ofa load to withstand momentary
low and high voltage events. The International Electrochemical Commission (IEC) calls the
same curves as equipment immunity curves. Computer business equipment manufactures
association (CBEMA) has defined its power acceptability curve but later in 1996 the Information

ww
Technology Industry Council (ITIC) announced its power acceptability curve which superceded
CBEMA curve. This is shown in Fig. 9.15.

w.E 250r---~~--~--------------------~

- asy200 ................................. ;................................................................. .

: Overvoltaga Conditions
~
~

~
Q)
:0) En
f50 ............................. '1' ...... "' .......................... ,............................

·t·. · · · . . . . . ;'0j.(5'....................................,.................. ,'_.

~
m
~
100 ·..........

50 ............................. =0
:!O gi nee
·T..........· · . · ·. . ·....·........ · ........ . ·'·· . · . ·.
~
.!:
~
g>
tI1
o Acceptable
Power
~;BR~at~ed!L======::!:===~
]Volta~ge::...._.]""":;~--'
r ing
..c::
U
I'-----r---' (Jj;
E:
-50 ............................ ~; ....... ·....r·..""'.. ;,,;,,:
~i
.. ·.:..:.;·"~
.. ·;...
....;
.
.. ;..::·~"'--......;.;.;.;..:..:..:..;.;:.;,
.. ·.;...·..""'
Undervoltage Conditions net
-100L-O-.O-O~01~O.-OO-1-=O~.O-1~~O.=1===1====1=O===1=OO==~1000
Time (Seconds)

, Fig, 9.15 ITIC Power acceptability curve

9.3.6 Solutions to Power Quality problem

There are several power electronic solutions to power quality problem.
A dynamic voltage restorer, which is a voltage source inverter CVSt) injects a voltage in
series with the distribution line. This is shown in Fig. 9.16. Three phase voltages generated by
the VSI can be injected into the line through a special coupling transformer. The terminals of
each of the three primary windings are brought out individually and are connected in series
with the l-phase lines. On the secondary side the thre'e windings are connected either in star
or in delta and the three terminals of the, VSI are connected'to the three secondary terminals.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 357

VSource

ww
w.E Fig. 9.16 Dynamic Voltage Restorer

Point of Common Coupling (PCC) is the point in the inter connected power system

asy
where loads are connected to the network. It is the point at which the load interacts with loads
and the network itself. For example the pec for typical residential applications is the distribution

En
transformer secondary. Where the distribution transformer serves a single customer, the

gi
primary of the transformer is the PCe.
vpee = V, (t) + VDVR (t)
nee
The DVR will maintain near perfect power quality conditions at the PCe. The DVR is
..... (9.4)

useful for sags, harmonics and flicker.

r ing
There are many other devices such as unified power quality conditioner (UPQC) which
is a combination of STATCOM and DVR for even better performance.

9.4 Data Base for Control

. net
Computer control of modern power systems is proposed to improve economy, to maintain
quality and for better security. Such a control is feasible, primarily if all meter readings and
other information pertaining to the operating state of the system are processed in real time into
a more useful form so that control decisions are made using it. For power systems in normal
operating state, bus voltage magnitude and phase angles are the state variables as they completely
determine the state of the system in steady-state (quasi-static state, infact). The vector of
voltage magnitudes and phase angles is called the static state vector. State estimation is the
mathematical process of obtaining the state vector from the variable information.
A knowledge of the state vector enables the system operator to take important decisions.
The state estimator, an algorithm based on load flow and statistical estimation theory, is of
immense use in calculating security and contingency evaluation. At an advanced level, it can
also be used for load frequency control and economic dispatch.

Downloaded From : www.EasyEngineering.net

358 Operation and Control in Power Systems

9.5 State Estimation

Since loads and generation change continuously, the power system never attains a steady
state. For the purpose of mathematical analysis, however, a steady state operating point over
a short interval of time may be considered where the changes in system variables are within
reasonable limits.

Power System Variables

The variables of interest for network monitoring are (i) voltage magnitudes and phase angles,
real and reactive power flows at each network bus and (ii) real and reactive power flows in
each line.

ww Meters for measuring watts, vars and volts can be placed at generator buses, EHV lines,
tie - lines and load buses and the readings can be telemetered in real time to the control centers.

w.E
However, only some of the measurements are available immediately while the rest are
either delayed as is the case ofinformation from a neighbouring, interconnected system or not

asy
available (like data from the low voltage distribution side). Further, the available data may not
be perfect due to errors in the meters. Some of the meter readings may also be missing.

En
Let n be the dimension of the state vector X and m be the dimension of the measurement
vector Z. Since the slack bus is fixed, we have Nb voltage magnitudes and (Nb-I) voltage

gi
phase angles where Nb is the number of buses.
XT = [Vi' V 2' ..... VNb ' 82
nee
..... 8 Nb ] ..... (9.5)

r
Knowing X, all other quantities like active and reactive power flows, line flows, etc can
be computed. In practice, all the values of X cannot be directly metered. The measurement
vector Z and the state vector X are related by the nonlinear equation. ing
Z = h(X) + 11
where h is the measurement noise. By exercising option in the selection of variables for
metering. Five types of measurement vectors can be identified.
. net
.. ... (9.6)

Case I Measuring real an,d reactive powers PI and Q I at all nodes, except P, at the
swing bus.
ZT = [P 2, P3, ..... PNb' Q, ..... QNb]
and it is a (2 Nb - I) dimensional vector.
PI and Q I are given by equation (2.11) and (2.12)
Case 2.' Real and reactive powers at all buses except Pi at slack bus and voltage
_ magnitudes at all buses may be measured in which case
ZT = PNb ' Q I ..... QNb Vi .... · V Nb ]
[P 2 .....

Here ZT is a (3Nb - I) dimensional vector.

Case 3: Real and reactive power flow in the lines Pij and Qij at both ends ot'each line
may be measured.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 359

ZT - [P
IJ'
0]
IJ

and it is a 4 N e dimensional vector where N e is the number of lines.

Case 4: All the variables of case 3 and voltage magnitudes at all buses may be measured
yielding.
ZT = [PII' 011' VI' V~ ..... V Nh ]
an (N I + N h) dimensional vector.
Case 5: All variables may be measured giving
ZT = [PI 0 1 PI~ Ol~ VI ~\]

ww Which is (4N h - I + 4 N () dimensional.

Modelling of uncertain(v:

w.E
Gaussian distribution is assumed for mathematical modelling of the uncertainty in measurements .

asy ..... (9.7)

The uncertainty is due to .

En
I. Instrumentation error,
2. Operational uncertainty. and
3. Incompleteness of the model
gi nee
Least Squares Estimation r ing
Using Eqn. (9.6)

The optimal estimate

n = Z - h (X)

X is that value which minimizes the cost functional

. net
..... (9.8)

J = nT R- I n = [Z - h(X)T] R- I {Z - heX)] ..... (9.9)

Where R is the covariance matrix of the observation noise n. The minimization may be carried
out using any of the standard techniques including steepest descent method, Fletcher - Powell
method, etc.

Linear Estimation Theory

In this case, the measured variables are assumed to be linear functions of the state variables.
The measurement equation (9. I) is then written as
Z=AX+n ..... (9.10)
where A is the m x n measurement matrix
Z is the m - dimensional vector of noise measurements comprising real and reactive
powers, and voltage

Downloaded From : www.EasyEngineering.net

360 Operation and Control in Power Systems

X is the n-dimensional state vector and

n is the noise vector
It is desired to minimize
J = (Z - AX? R-I (Z - AX) ..... (9.11 )

so that the optimal estimates X is obtained. In practice, three cases may arise.
Case I : m > n; the estimated equations are

..... (9.12)
The estimate covariance matrix is

ww Rx=(ATN-IAt l ..... (9.13)

w.E
The noise covariance matrix
Nx = E (1111 T) ..... (9.14)

asy
The number of basic arithmetic operations required for the computation of the optimum
estimate area of the optimum estimate are of the order ofmn'. The difficulty in this case is that

En
the maximum Rx tends to be singular making inversion difficult.
Case 2: m = n: The estimation equations are

gi
X = A-I Z

nee
..... (9.15)
..... (9.16)

r
In this case, the computational problems are fewer since A is a square non singular
matrix and Rx is determined from
R, = A-I N(ATt l ing
. net
Case 3: m < n; while an estimate of 11 can be determined in principle, he estimate is
neither unique nor does it have much relevance to state estimation in power systems.

9.6 Power System Security

The operating modes of a power system have already been discussed in Chapter I. In the
'normal' mode, both the load and the operating constraints are satisfied and adequate spinning
reserve is available. In the 'alert' state the load requirements are met with, but an adequate
spinning reserve is not available. A system is in 'emergency' mode when the operating constraints
are not completely satisfied. There may be two types of emergency modes. The system may
be in 'steady state emergency' when the steady state operating constraints like bus voltage
limits are violated. On the other hand, the system may reach 'dynamic emergency mode of
operation when, for example, stability operating constraints are violated. In either case, the
normal load requirements cannot be satisfied. In the 'restorative mode' control action is initiated,
but the load constraints are not completely satisfied. Under this condition, the system may be
in a completely or partially shut-down state.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 361

The above classification of the states of the power system will facilitate in distinguishing
between the different types of controls to be applied to the system. It can now be stated that
the objective of system security control is to keep the power system in normal mode of
operation and prevent it from entering into either emergency or restorative mode.
The system state is continuously assessed via data acquisition systems making use of
telemetered data, redundant data and by predicting the missing or doubtful data whenever a
metering channel becomes erroneous or its accuracy is beyond the normal specified range.
This job is referred to as security monitoring.
During security assessment, a series of fast computations are made to examine the
effect of various credible contingencies and those that are not so. If some of the assumed

ww
contingencies result in unsatisfactory performance from the point of view of security, a corrective
strategy is called for, to determine the best corrective action either by special calculations
within the computer or from results obtained from off-line studies.

w.E
Continuous monitoring for security and sending command signal for corrective action
whenever necessary is 'referred to as security control.

asy
Steady state security analysis is performed to determine whether there will be a new
steady state operating point at which the system will settle after the post-fault oscillations have

En
been damped out. The post contingency steady-state solution will have to be checked for
violation of operational constraints. If it violates the constraints, the contingency is declared to

gi
be insecure and necessary corrective action is to be initiated.

nee
Steady state security may be defined as the ability of the system to operate steady-
state-wise within the specified limits of safety and supply, following a contingency in the time

r ing
period, after the fast acting automatic Control devices have restored the system balance, but
before the slow-acting controls like transformer tappings, human decisions, etc, have responded.

.
For steady state security analysis the following contingencies may be considered:
I. Loss of a generating unit
2. Sudden loss of a load,
net
3. Sudden change in flow in an inter-tie,
4. Outage of a transmission line,
5. Outage of a transformer, and
6. Outage of a shunt capacitor or reactor.
The outage may be either network outage or power outage. Fast and decoupled load
flows are employed for security analysis of systems.

9.7 Steady State Security Assessment

The steady state security assessor is a computer program that uses real time data for analyzing
the current state of the system for steady state security and is thus essentially an on-line
process. For this, an approximate load flow algorithm is needed. Consider the input-output
model given by

Downloaded From : www.EasyEngineering.net

362 Operation and Control in Power Systems

F (X, u, d) = 0 ..... (9.17)

where X is the vector of dependent variables
u is the vector of control variables and
d is the vector of disturbance variables
For a given disturbance vector, it is required to determine the output X corresponding to
a given input u as shown [n Fig. 9.17.

ww u

·1
F (X. R d)

f
• X

w.E Fig. 9.17 System representation

asy
Eqn. (9.14) and
bus system. There are
Fig. 9.17 correspond to 2n real nonlinear power equations for any n-
2n bus power injections (n-real power and n-reactive power) which
may be denoted by the
bus impedance matrix, Envector S. The network parameters are given by the bus admittance or
designated by P.

gi
The steady state security assessment problem is concerned with the analysis of the

nee
system described by the vector triplet (X, S, P). Given a base triplet (Xo, so, pO) along with
contingency details,
pI, i = I, 2, ...... , nw;
SI, i = nw + I, ....... nw + pO r ing
is then required to determine the post-contingency states.
XI' i = I, 2, ....... , (nw + po)
. net
where nw is the number of network outages and po is the number of power outages, it

The steady state security assessor is a computer program that determines the post-
contingency states. The program is also interlinked with several other programs at the computer
control center.

9.8 Application to Outage Studies

Referring to the fast decoupled load flow algorithm given in chapter 2.

B'~b == ~p
IVI

The above equations may be represented as

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 363

..... (9.18)
so that ..... (9.19)
The outage of a line (neglecting charging capacitance) joining buses i and j can be
reflected in K by modifying two elements in row i and two in row j. The new outage matril'
can be presented as
K' = [K + BA AT] ..... (9.20)
where B is the line susceptance and
A is a column vector with AI = + I and AI = -I and zeroes in other locations.

ww It can be shown that

K'-I = K -I - h L AT K- I ..... (9.21 )
with
and w.E h=[B-'+ATLr l
L = K- I A
..... (9.22)
..... (9.23)

asy
The solution vector
yl = K'-IC ..... (9.24)

En
gives the post-outage state of the system.

gi
Complete system security involves the following:

nee
I. System prediction: This is done by computer programs that predict the system load
conditions in advance for a predetermined time interval (say, an hour or day),

r
2. System contingency evaluation: This involves programs that detertninethe state of

ing
the system consequent to generation, line or transformer outages.
3. System corrective strategy This is aimed at taking all possible decisions to keep the

. net
system in steady state secure condition. It also involves the necessary steps to be
taken to overcome both unexpected and anticipated problems, either in the present
state of operation or in future:
4. Automatic control: This eliminates the necessity of the operator taking initiative for
all corrective actions.

9.9 Pattern Recognition Methods

System security can be evaluated through pattern recognition techniques also. The approach is
quite different from conventional methods. Pattern recognition methods essentially involve the
formation of a training set with pattern vectors, feature extraction and defining a security
function.

Pattern Vector
Each operating condition ofa power system is determined by real and reactive power generation,
line flows, loads voltage and voltage magnitudes and phase angles. All these variables from
measurements constitute the compollents of a pattern vectur X = (XI' x~ ..... , xn ).

Downloaded From : www.EasyEngineering.net

364 Operation and Control in Power Systems

Training Set
Every conceivable operating condition of the system gives rise to a corresponding pattern
vector. The set of all possible pattern vectors constitutes the training set.

Pattern Classification
Each of the patterns belonging to the training set are classified as either secure or insecure
through off-line studies. Secure patterns contain all operating conditions that are secure while
insecure ones contain such conditions in system operation that are insecure.

Feature Extraction

ww
The actual number of variables used for representing an operating condition may be too many.

w.E
But in practice, only a few of them may considerably influence the security of the pattern. It is
desired to select a relatively small number of such variables called feature variables and the
process of selection is called feature extraction. Feature extraction can be simplified by applying

asy
operational experience and engineering judgment in addition to any statistical criterion.

E
The feature vector is defined by
F = (fp f2' .......
ngi
~n); m< n ..... (9.25)
All variables that are redundant may be removed from the pattern. Variables with high
correlation coefficients may also be eliminated.
nee
Security Function
rin
g.n
A security function may be defined as a linear function of the feature variables,
S (F) = Sjl + SI fl + S2 f2 + .. :.. + Sm fm ..... (9.26)

et
The security function S (F) > 0 if F is secure and S (F) < 0 if F is insecure. Different
methods are available to determine the constants So' SI' ..... Sm' For example, least squares
method or optimal search method can be used successfully to determine the coefficients
So' Sm' More sophisticated and mathematically complex functions may be defined in place of
Eqn. (9.26).
After the security function is identified, its validity may be checked using patterns of
known class.

Security Evaluation
Once a valid security function is determined, then any operating condition represented by the
corresponding feature vector may be used, and from Eqn. (9.23) the operating state may be
classified either as secure or insecure.
The success of the method depends on the proper selection of the feature variables and
security function.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 365

Pattern recognition methods may not have much relevance for steady state security
evaluation. However, it appears that for transient security evaluation, pattern recognition
techniques may provide the answer in future for situations where the fast acting automatic
control devices have restored the system balance and the slow acting controls like transformer
tappings, human decisions, etc. have not yet responded.

9.10 Power System Control Centres

Increase in unit sizes, growth of interconnections and the need to maintain the system in
normal mode require sophisticated control, instrumentation and protection: The multiplicity of
monitoring instruments in the control room and their distance apart make the observation of

ww
more than a few vital ones almost impossible, especially during the intense activity of plant
start-up. The operators are called upon to visualize the implications of a variety of changing

w.E
plant parameters and take critical decisions. These requirements led to the development and
application of more advanced solid-state modular electronic instruments, computer based direct

asy
digital control and data processing systems.
A modern power system control centre has the following functions to perform:
1. Automatic generation control,
2. Economic load despatch En
4. On-line load flow,
gi
3. Automatic voltage (reactive power) control,

nee
5. On-line short circuit.
6. State estimation, r ing
7. Security monitoring,
8. Steady state security analysis,
9. Supervisory control,
. net
10. Automatic trouble analysis,
11. Emergency control like load shedding, generator t ripping, and
12. Automatic circuit restoration, etc.
The computer system involves dual configuration with external interfaces to monitor
the data. The first one is a process computer linked by telechannels to various generating and
sub-stations for data acquisition. The second one is a larger one where major calculations are
carried out and is linked to the process computer. For real-time computer control of power
systems, the following basic components are needed:
1. System wide instrumentation,
2. High speed digital telemetry,
3. Central processing unit (CPU),
4. Memory and bulk storage,

Downloaded From : www.EasyEngineering.net

366 Operation and Control in Power Systems

5. Interactive display, and

6. Software (operating and application).
The real time computer is designed to perform data acquisition, storage and retrieval, data
processing, interactive display and remote signaling and control. It consists of modems and
interfaces, CPU, memory and bulk storage-and input-output devices like display devices, card
reader, printer, etc. Fig. 9.18 shows a functional block diagram of a real-time computer.

ww nrsPI A Y 1<1111111-----1 r-I -M-OD-E-.M-S---'~ •

DATA
FROM
THE

w.E r----4~ BULKSTORAGE

SYSTEM

asy Fig. 9.18 Functional Block Diagram of a real time computer

En
The data base consists of static data, dynamic data and software. The static data consists

gi
of the details of lines, transformers. generators, etc. The dynamic data includes line flows.

nee
voltage levels, breaker condition, generation and demand. The software includes operating
software, application software and support software.

r ing
The operating software is the system software for real time operation compiling routines,
file management, etc. The application software consists of programs written for power system

debugging, maintenance and testing programs.

Computer control can take any of the following forms:
.
operation and control, which were listed earlier. The support software consists of diagnostics:

net
I. Off-line computer control,
2. computer-assisted control. and
3. on-line computer control.
For planning and operating computation, which are carried out at infrequent intervals,
off-line computer calculations suffice. The results are to be updated at regular time intervals. ;
In the computer-assisted control scheme, necessary data is transmitted at regular intervals
to the computer located in a central control station and its decisions are communicated to the
human operator.
For short-time processes extending up to a few minutes, the system is under continuous
control of direct acting devices. Under emergency conditions. the on-line computer which
processes the data continuously sends commands to the direct acting devices according to a
pre-planned strategy to prevent danger to the system.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 367

The type of computer control, the configuration and the peripherals used are ultimately
decided by the cost that the utility industry is willing to invest.

9.11 Level Decomposition in Power Systems

Power systems are characterized by strong hierarchical order. Accordingly, control strategy
can be devised to fit into the hierarchical structure advantageously as indicated in table 9.1.

Table 9.1 Hierarchical Structure in Power System

Level System Monitoring and control

ww
First Level Generating stations and sub stations Local control center
Second Level Sub transmission and transmission networks Area load dispatch center
Third level
Fourth level
or top level
w.E
Transmission system

[nterconnected power systems

State load dispatch center

Regional control centers

asy
Local Control Centre
En
A number of control functions can be performed locally at power generating stations and sub-

gi
stations using local equipment and automatic devices.
Some of the typical control applications are :
nee
1. Local monitoring and control,
2. Protection, r ing
3. Auto reclosure,
4. Voltage regulation,
5. Capacitor switching,
. net
6. Feeder synchronization;
7. Load shedding in the event of necessity, and
8. Network restoration.

Area Load Dispatch Centre

A group of generating stations and sub-stations, along with the associated network and loads
may be considered as a unit for control under an area load dispatch centre. The area control
centre receives information and processes it for appropriate control action.

State Load Dispatch Centre

The minute-to-minute operation of the power system at the state level may be carried out by
the stale load dispatch centre. It may have the following functions:

Downloaded From : www.EasyEngineering.net

368 Operation and Control in Power Systems

I. System generation and load monitoring and control,

2. System wide, state monitoring and control,
3. Circuit breaker state monitoring and control,
4. Load shedding and load restoration,
5. Supervisory control for lines and equipment states,
6. System alarm monitoring and corrective action, and
7. Planning and monitoring of system operations.

ww
Regional Load Dispatch Centre
The regional load dispatch centre may be regarded as a coordinating and monitoring centre for

w.E
state level dispatch centres with the following main objectives:
I. Integrated operation of state level dispatch centres,

asy
2. Operation and maintenance schedules for the generating plant,
3. Operation and maintenance schedules for maximum capacity utilization,

En
4. Monitor and control inter-state power transactions, and

9.12 Network Automation

gi
5. Monitor and control inter-regional power transactions.

nee
rin
Network automation reduces the job of the human operator to some extent. It involves
monitoring the process and remote control. Whatever may be the level of automation, all
important emergency decisions are required to be taken by the human operator.
g.n
The input data for the network automation may be obtained from important load centres

may perform the following functions:

1. Obtain the system data in analog form,
et
through microprocessor or computer-based remote terminal units. The remote terminal units

2. Convert the analog data into digital form, and

3. Transmit the digital data to area load dispatch centre.
The data so transmitted will contain information like circuit breaker status, isolator
position, etc. which will be indicated only when the condition changes as well as regularly
sampled information about the network variables such as line flows (MW, Mvar), frequency
(Hz), voltage levels (kV), currents (Amps), etc. The information processed will also include
generator or transformer tripping, battery failure, etc. Also other functions that the remote
•
terminal units may perform include limit or threshold value monitoring, fluid level processing,
data logging, etc. .
The data may be used for taking appropriate control action with the system of computer
control planned.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 369

@ Di5k~ Oisk~

~
ww Con!iOl~ Console

w.E o If-lone
.dollng
laclhly
Visual di~plQy unll
Drive
tor
dlspbys

asy
E ngi
nee
Fig. 9.19 Computer control in Power system

The data may be classified as real time, static and long term data. The real time data

rin
consists of measurands, alarms, etc and is stored in core memory. Static data consists of the
background picture, text data files, etc. and is normally stored on a secondary, fast access
memory.
g.n
for postmortem analysis, measurands, etc
The man-machine interface may consists of the following:
et
Long term data may be stored on secondary, slow-access memory and consists of data

I, Alphanumeric - technological key boards.

2. Visual display units.
3. Mosaic boards
4. Digital displays
5. Loggers, and
6. Strip chart recorders
A typical computer control set up is shown in Fig. 9.19.

9.13 Load Prediction

Accurate prediction of load demand is required for service reliability and efficient operating
performance of a power system. If the system is predominantly thermal, this is all the more

Downloaded From : www.EasyEngineering.net

370 Operation and Control in Power Systems

important in order that adequate generating capacity to supply the load demand, maintain system
security and supply the necessary spinning reserve is economically scheduled. In practice,
spinning reserve is provided as a back-up for either the loss of the largest unit or for the loss of
transmission capacity which renders unavailable the greatest amount of generating capacity.
But, one should not lose sight of the possibil ity ofinaccuracy in system load prediction consequent
to which deficit operating reserve capacity may result.
For plant loading schedules in thermal systems, load prediction up to two hours in
advance is necessary while for unit commitment schedules prediction up to 24 hours is sufficient.
Also, at all stations and control centres, short-time prediction is needed for storage and. display

ww
of advance information. Based on this information, predictive security assessment of the system
is made. This also helps to contain the rates of change of generator outputs within their permissible
limits.
w.E
For the implementation of economic scheduling of generation using digital computers,

a
detailed estimates of the future load demands are essential In order to allow sufficient time for

syE
the calculation and implementation of the generator schedules. Whatever method is envisaged
for the calculation of such economic schedules consistent with the security and spare

ngi
requirements of the system, the schedules should be calculated every 15 or 30 minutes and
each economic schedule should be a predictive one, for at least about 30 minutes ahead of the

nee
event. It is then obvious that the predictions are to be revised frequently in the light of any
fresh information so as to minimize the estimation errors.

rin
Peak load demand forecasts are useful in determ'ining the investment required for

g.n
additional generating and transmission capacities required. Forecasts for planning require data
extending over several previous years. Meaningful forecasts can be obtained with lead time of
3 to 5 years.
Generally, load demand is assumed to have two salient features helpful in prediction.
These are the long-term trend component dependent on economic growth, seasonal changes,
et
etc. and the weather sensitive fluctuations in daily and hourly load demands.
Thus, the overall load demand has a predictable trend component superimposed on
which is an erratic variation attributable to weather fluctuations. Summer, winter and monsoon
periods will have definite patterns of load consumption. Also, the load demand varies during
the day continuously, following a pattern with definite peaks at certain times of the day.
The weather sensitive component depends on the following meteorological factors:
1. Temperature,
2. cloudiness,
3. wind velocity,
4. visibility, and
5. precipitation.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 371

9.14 Load Prediction using Matereoiogicai Data

Weather Weighting Method
In this method, the load is separated into two main components. The first component is a base
load which is of fixed value and the second a variable component which is a function of the
weather conditions.
Estimates can be made 24 hours ahead, using the weather, forecast. The temperature
base for weighting the effect of the predicted temperature on the load is the normal, mean
temperature of the month. The normal, mean temperature of the month has zero weight.
Similarly the change in consumers demand due to cloudy weather may be assumed to vary in
direct proportion to the degree of cloudiness. This in turn may be expressed by an illumination

ww
index with fair, clear sky corresponding to zero weight.
The base load is determined from past records, proper weighting of the elements of the

w.E
weather will be attained only after several trials. The method of prediction stabilizes after this
trial period. It may be noted that the base loads for week days and weekend will generally be
different for any hour.

asy
Using these base loads, a load estimate based on the best available weather forecast can

En
be made using proper weighting of meteorological factors like temperature, cloudiness, wind,
velocity; etc.

gi
In thermal stations, a boiler may take about 3 hours of time to reach full operating

nee
condition from the cold. Thus, peak load estimates In systems with predominantly thermal
generation are more critical, since they are required in time to bring any necessary generating

rin
capacity on-line. The feasibility of this method lies in the consistency of the calculations of the
base load. The base loads are to be revised every year as load demand changes continuously.

Multiple Regression Methods

g.n
The meteorological information can be reduced to a number of spec!fic factors like:
I. Temperature (T)
2. Cooling power of the wind(W),
et
3. Illumination index(L)
4. Rate of precipitation (P), etc
In deriving the functional relationship between the variation of demand and the specific
metrological factors to which it is sensitive, it may be assumed that the weather -sensitive
component of the demand may be expressed as the sum of functions of the respective
meteorological factors.

Linear Regression Techniques

The first obvious approximation which can be made is to assume a linear functional relationship.
It may be assumed that the basic demand curve is closely approximately by a step function
with a step length of one week.

Downloaded From : www.EasyEngineering.net

372 Operation and Control in Power Systems

The data may be fitted into a regression equation of the form

Y = a + b l T + b2 W + b) L + b4 P + F(t) + d
where Y is the demand at a fixed time each day; T, W,L, P are the corresponding specific
meteorological factors; a, b 1, b2, b3, b4 are constants d is the day of the week correction and
F(t) is a polynomial function of the ith week. The regression coefficients b I, b2, b3, b4 and d
are to be estimated. In the above equation, more factors may be added or some may be
removed, if necessary. In the linear regression technique the data points are fitted into a linear
equation of the type.
Y = ao + a l X
where Y is the dependent and X the independent variable. In the least squares curve fitting

ww
method, we minimize the square of the deviation of the dependent variable from its actual or
observed value.
Let
w.E Y I = ao + al XI
If n is the number of observations or data points

asy
Dividing throughout by n gives ELX,n
LY
n
I
--=ao+a,--
n
gi nee
since
n
Y
L-'
X
r
and L-' are the mean values ofY and X respectively
n

X are the average values.

ing
- -
Y = a o + a, X and

Y and
y, = Y, - Y . net
and

Then, subtracting Y from the both sides of the basic equation

Y, - Y = a o X, - Y
=a o +a,X, -au -a,X,
i.e. YI =ao - ao + a l XI = a, X.
i.e. a coordinate transformation is effected. The individual deviations become
Id, I= Iy, - a ,x ,I
and D= Ld~ = L(YI -alx}
,;\

is the error term.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 373

To minimize D we get the condition

dD
-=O=-2L>,(Y, -a,x,)
da, .

or LX, Y, - a, I x~ = 0

form which

i.e.,

having found a" from the equation

ww Y = a 0 + a, X we get ao as

w.E
Also
asy
En
gi nee
For the case where Y is a function of several sets of variables XI' X2,
multiple regression coefficients ai' a~, ..... ak can be derived.
........ , X k the

Let
rin
where Y is the most probable value of Y.
then if, Y= Y - Y g.n
we have
x =

Y = a,X,
X- X
+a 2 X 2 + ..... +a k Y. so that ao is eliminated.
et
Let D = (Y- yf
= I[y-(ao+a,X, +a 2 X 2 + ..... +a k x k ))2
Ifwe differentiate this with reference to ai' a2, ak

Downloaded From : www.EasyEngineering.net

374 Operation and Control in Power Systems

- L x IY+ a IL x ~ + a 2 L x I X2 +........... + a k L x I Xk = 0
- Lx 2 y+a l Lxl X2+a2Lx~ +........... +akLx l xk=O
ie

Rewriting as
ajLX~ +a 2 Lx jx 2 +a 3Lx 1X3 +................. + akLxlXk = LXIY
alLx l X2 +a2Lx~ +a 3Lx 2 X3 +................. +a kLx 2 Xk = LX Y2

ww aiL XI Xk +a 2L x 2 Xk +a 3
or in matrix equation form
L X3X +................. + a kL X~ = LXkY
k

w.E IX XI IAI = I XY I
the matrix equation is to be solved for A.

asy
Using the technique indicated above, the coefficients in eqn. (9.1) can be determined.

E
The terms of F(t) can be determined using the method of polynomials of least squares.

ngi
It was found that a good fit to a year's data could be obtained by fitting polynomials up to the
sixth degree.
Least Squares Polynomials
If we write
nee
~
Y=a O +a IX+a 2 X +a3 X
2
rin
3

for a cubic form, the individual deviations d, can be expressed as

?, g.n
d 1= YI -aO-aIX I -a2X~ -a 3 X;
and D the sum of the squared deviation is
~ d I2 = (Y
D = L.. - a0 - a IX I - a 2 X I2 - a, X I3)2
et
I
I~I

Minimizing with respect to ao' ai' a2 and a3

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 375

which can be rewritten as

LY, ::: nao +aILX, +a~LX~ +a,LX;

LX,Y, =aoLX, +aILX~ +a 2 LX;' +a,LX~

I
In the matrix form

n LX, LX;LX~ LY,

ww LX, LX~ LX; LX;

LX,2 LX;' LX~ LX;
al
a,
a,
LX,Y,
LX;Y,

w.E LX; LX~ LX~ LX~ a, , ,

LX3y

from which the values ofao, ai' a 2 and a 3 can be calculated. Extending the equation to

asy
sixth order will yield a good fit for F(t) for an year's data.

En
Nonlinear Regression Methods
If the relationship between demand and the specific meteorological factors on which it depends

gi
are nonlinear, the linear functions introduced in the regression analysis earlier will give only

nee
average effects over the range covered by the data. We cannot assume any specific mathematical
model for nonlinear functional relationship in advance, since its nature is not known in advance.
We can write
rin
S = fl(T) + fiW) + f3 (L) + f4 (P) + F(t) + d

g.n
Where the constant a has been included in the basic demand curve F(t). As a first
approximation, the form of the unspecified functional relationships fl(T), fiW), f 3(L), f4(P)
et
can be taken as straight lines whose slopes have the values of the corresponding linear regression
coefficients. First, all the weather sensitive components are completely removed.
F(t) + d = x - fl (T) - f2 (W) - f3 (L) - f.j (P)
Plotting a graph of the above relations, we can make an estimate of F(t) +d. Grouping
these estimates according to the day of the week and noting the difference between their
respective means and the grand mean, approximation to the day - of the week adjustments are
to be made. Then F(t) is known, giving the basic demand curve.
From this initial estimate and the data of the meteorological variables. T. W etc. the linear
approximations are successively replaced by curvilinear relationships and the proces~ is carried
on until no further improvement is obtained. For a daily load estimation, the basic demand for
the following day can be read from the final graph of F(t) and to this can be added the day-of-
the week correction. A further weather correction can be made by applying the weight
determined from the meteorological response curves using the best available weather forecasts.

Downloaded From : www.EasyEngineering.net

376 Operation and Control in Power Systems

9.15 Spetral Expansion Method

In this method, only past load data is required. It is well - known that the pattern of load
demand tends to repeat every 24 hours. The load demand curves may be divided to given
ensemble of time series. The problem then becomes one of predicting a non-stationary process
given by an ensemble of sample functions.
Consider the load curve for any working day, divided into the three 8 hour periods or
four 6 hour periods. These parts may be made to have an overlap of, say, 2 hours for continuity
in prediction. Then the load on the mth working day at the nth instant of time can be denoted
by pmn' For example, we can have six part load curves for the time period of, say, 12 noon to

ww
6 pm for the six working days of a week, this forms an ensemble of six functions. For
Sundays a similar ensemble can be constructed. It is possible to represent the load by the
relation.
w.E Pmn = Amn + fj(T m) 8 mn

asy + f2(L m) Cmn

En + f3 (W m) Dmn +...... . . .... (9.27)

Where fj(T m)' f 2(L m) fiW m) are functions of temperature Tm' illumination Lm, and

gi
wind velocity Wm respectively. 8 mn , Cmn and Dmn are coefficients used to given weightage to

nee
the weather parameters depending on the time of the day. Amn is the base load. The sample
functions Pmn are thus assumed to belong to a linear manifold generated by the weighting
functions of their linear combinations.
r ing
As a preliminary step, it is required that from the load all trend components are eliminated.

. net
Finally, the residual component can be expanded into time series using spectral theory of
stochastic processes.
Equation (9.27) may be rewritten as
..... (9.28)
where Pwet) and Pit) represent the trend components of weekly and daily patterns of load
demand. Pmn(t) is the residual load component. The component Pw(t) is given by

..... (9.29)

where nd is the number of days in a week.

Pct<t) is computed from

..... (9.30)

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 377

where nw is the number of weeks for which data is available. The residual component
Pmn(t) can be expressed as a time series. If N is the total number of days for which data is
available, the auto correlation function for the residual component is given by
I N
R(t:r) = -Ly,(t)y,(t) ..... (9.3 I)
N ,=1
A characteristic function fit) is defined by the integral equation

l' R(t, t)f. (t)dt = b k f, (t) ..... {9.32)

where T is the total time period over which the prediction is valid.

wwThe residual load component is then expressed as

w.E Pmn (t)

K
= L a k bUk (t) + e(t)
k=1

a
The coefficients ak can be determined for the best fit using any standard criterion. The

syE
coefficients bk are determined by eqn. (9.32) and e(t) is the error term.

ngi
The number of terms in the time series, K, can be determined using the minimum, mean
- square error criterion given by

e 2 (t) = R(t.t) - b k fk (t)

where e{t) is the average error.
nee
9.16 Prediction by Scaling a Standard Load rin
g.n
In this method, a particular day is taken as reference and the load curve for the reference day
is taken as standard. The load on the day o~ prediction is compared with the standard load
curve. Each measured load of the day is divided by the corresponding value of the load on the
standard load curve at that instant.
et
Consider the load curve for the previous hours. Let this be divided into four quarter -
hourly periods. The load for the next one hour can be predicted as exponentially weighed
average of the previous four ratios. The ratio at the (n + k)th instant is estimated from the
relation

I-x 3 n
r(n+k) = --4 Lrn_mx ..... (9.33)
1- x 111=0

where x is a weighting parameter determined empirically, the value of which varies

between zero and unity. The predicted load can be obtained by mUltiplying the predicted ratio
r(n+kJ by the corresponding value of the load on the standard load curve.

Downloaded From : www.EasyEngineering.net

378 Operation and Control in Power Systems

9.17 Short - Term Load Forecasting Using Exponential Smoothing

From the available data, the seasonal and weekly variations are removed. The remaining load
may be represented as a linear combination of known functions of time and noise component.
pet) = AT F(t) + net) ..... (9.34)
The elements A are to be assumed constant during the period offorecasting even though
they change slowly. Using known fitting functions F(t) and current coefficients A(T). forecasted
loads can be computed by extrapolating equation
pet + t) = AT(T) F(t + t) ..... (9.35)
Where pet + t) is the forecast with the lead time t. The elements of A(T) can be

ww
computed using weighted least squires criterion. For this, the expression

r
w.E f J=
j=O
Pj [peT - j) - A T (T). F(T - j)

asy
is minimized. The constant ~ has value between zero and unity.
As an example, the time function may be a Fourier series given by

mEn
pet) = A +
,=1 gi
L aj sin ro, t + b, sin ro, t
If the time period is taken as one week then. nee
ro·=
,
27t
N
(7)(24)-'
r ing
.
where N is a positive integer. The selection of the actual number of frequencies is
dependent on the auto - correlation function and power spectrum for the observed hourly
net
dat,. The coefficients "a,", "b," and Ao are to be revised each hour

9.18 Peak Power Demand Prediction

The fluctuating values of peak power demand may be assumed to be controIled by probabilistic
laws in time. The monthly peak demand series, for example, contains three additive components\
the trend Pp seasonal Ps and noise Pw
Thus the peak demand is given by
PMO (T) = PT (t) + Ps (t) + PN (t)
The trend component shows a continuous change in the demand, normaIly an increase.
The seasonal component consists of periodic variations, the pattern of which increases in
amplitude year-by-year. Using the stochastic theory of stationary and nonstationary processes,
the peak demand can be predicted.

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 379

9.19 State Estimation in Load Forecasting

Real time control of power systems requires accurate short term forecasting of load demand.
State estimation in forecasting will be quite useful for such a purpose.
The model for forecasting demands upon the lead time. Fer a very short lead time, less
than 10 minutes or so, the system load and the observations may be represented by
\
pet + ~t) = pet) + net)
Z(t) = pet) + K(t) ..... (9.36)
Where pet) is the load at the instant t

ww n is the system noise

Z(t) is the observation at the instant t

w.E
and k(t) is the observation noise.
For short term load forecasting, with lead times of about ten minutes to an hour, the

asy
load may contain an increment ~P(t). The weather conditions may not play any role, since in
such a short time they may not vary so much as to change the consumption pattern.

En
The model could be represented in thts case by
pet + ~t) = pet) + ~P(t) + n 1(t)

gi
~P(t

Z(t) =
+ ~t) = pet) +
pet) + K(t)
n2 (t)

nee
..... (9.37)

The expected or average values of the noises is zero.

E[n(t)] = E[n 1(t)] = E[nlt)] = ErK(t)] = 0 r ing
Further E [net) K(t)] = 0

Rewriting eqn. (9.37) for the forecasted model P

.. .
. ..... (9.38)

net
pet + ~t) = pet) + K(t)[Z(t + ~t) - pet) . .... (9.39)

where 0 < K(t) < I

The value of K that minimizes the variance of forecasting error is
K(t + ~t) = R(t) K(t) +Q(t) ..... (9.40)
Where R(t) and Q(t) are the variance of observation noise and system noise respectively.
The forecasted load demand is
.. .
pet + ~t) = pet) + K(t + ~t)[Z(t + ~t) - pet) . .... (9.41 )

for the short term model

. . ..... (9.42)
pet + ~t) = [I - K(t + ~t)][P(t)] + K(t + ~t)Z(t + ~t)
the procedure is basically not different from the exponential smoothing method, but the
correcting gain K can be sequentially evaluated if the noise variances are known in advance.

Downloaded From : www.EasyEngineering.net

380 Operation and Control in Power Systems

9.20 Generating Capacity Reliability and Qutage Probabilities

The availability of enough generating capacity to meet the load demand is the foremost priority
at the planning level that ensures satisfactory operation of the power system. The ap~ication
of probability methods enables quantitative prediction of the system reliability. The results
obtained by probability techniques will only provide a criterion for evaluation of the situation
rather than give a guarantied result.
Rules for combination of event probabilities
All events such that occurrence of anyone event is not influenced by the occurrence of any

ww
other event, are called independent events. The probabi Iity of occurrence of any two independent
events is given by the product of the probabilities of occurrence of each event separately.

w On the other hand, if two events cannot happen both at the same time, then they are said

.Ea
to be mutually exclusive. The probability of occurrence of mutually exclusive events is given
by the sum of the probabilities of the occurrence of each event in the mutually exclusive event
group.
syE
Binomial distribution may be used to calculate the probabilities for power generation

ngi
reserve capa;;ily calculations. For any event, if s is the probability of success and f is the
probability of failure, then for n trials

(s + f)n =
n
L nC, Sr fn-r nee
r
r~O

The binomial coefficients can be determined from Pascal's triangle

ing
I 2 I
. net
I 3 3 etc.
4 6 4

Application to Power Plant Capacity Calculations

Consider a power station with two units each of 60MW. Let the forced outage rate or the
probability of any set being out of service be 0.05. Then the availability rate or the probability
of the set being in service is obtained as (1-0.05) = 0.95. For the plant under consideration
three states of operation may be listed:
1. Both the units are in services
2. only one unit is in service, and
3. both the units are out of service

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 381

The probability for the existence of each of the above states can be computed using the
binomial expansion with n = 2. The results are shown in Table 9.2

Table 9.2 Two - unit System Outage Probabilities

State Capacity Capacity Probability

Outage MW Available MW
I 0 120 (0.95)(0.95) = 0.9029
2 60 60 2(0.95)(0.05) = 0 0950
"-
3 120 0 (0.05){0 05) = 0 0025

ww Total

It may be noted that the probabilities are computed from the relation
1.0000

w.E (s + f)~ = S2 + 2sf +f-

asy
In a similar manner. the probability of occurrence of various outages for a 3-unit power
plant can be computed. Consider a power station with three identical units. each of capacity

En
110MW. The forced outage rate for each unit is 0.03. the availability rate, therefore, for each
Ilnits is 0.97. The probability of occurrence of various states in operation are computed and
I. .... ted in Table 9.3

gi nee
Table 9.3 Three - Unit System Outage Probabilities

State Capacity
Outage MW
Capacity
Available MW
rin Probability

I 0 120
(0.97)3 g.n
(0.95)(0.95) = 0.9029
0.912673

et
I 0 330 =

2 110 220 3(0.97)2 (0.03) = 0.084681

3 220 110 3(0.97)2 (0.03)2 = 0.002619
4 330 0 (0.03)3 = 0.000027
Total 1.000000

The probabilities are computed from the relation

(S+f)3 = S3 + 3s2f + 3sf- + f3
Finally, taking the example of a 4-unit power plant, each with a rating of 65Mw and
forced outage rate of 0.04, the probabilities of occurrence of the five possible states of operation
are listed in Table 9.4.
(S+t)4 = S4 +4s 3f +6s 2(2+4s(3+[4
outage probabilities for dissimilar units in a plant or system with different forced outage rates
can be calculated extending the basic principle further.

Downloaded From : www.EasyEngineering.net

382 Operation and Control in Power Systems

Consider a power system with two similar units, each of 60MW capacity and another
generating unit and another generating unit of II OMw. The forced outage rate for the 60Mw
unit is 0.03, while the same for the II OMW unit is 0.1.
Table 9.4 Four unit System Outage Probabilities

State Capacity Capacity Probability

Outage MW Available MW
I 0 260 (0.%)4 = 0.84934000
2 65 195 4(0.96)3(0.04)=014155000

ww 3
4
130
195
130
65
6(0.96)2(004)2 = 0.00884730
4(0.96) (0.04)3 = 0.00024576

w.E
5 260 0 (0.04)4 = 0.00000250

asy Total 1.00000000

To begin with, the probabilities are computed separately as before. These are shown in
Table 9.S
En
gi
Table 9.5 Outage probabilities for different type of units

Two - 60 MW sets
nee One -1I0MW

r
Capacity out Mw Probability Capacity out MW Probability

0 (0.97)" = 0.9409 0
ing 090

120
Total
60 2(0.97)(0.03) = 0.0582
(0.03j2 = 0.0009
1.0000
110

Total
. 0.10

1.00
net
The results of both the calculations are combined now to find the outage probabilities
for the combined system (Table 9.6)
Table 9.6 Outage Probabilities for System with Dissimilar Units

Capacity out MW Probability

0 (0.9409) (0.9 = 0.846810
60 (0.0582) (0 9) = 0.523800
110 (0.1) (0.9409) = 0.940900
120 (00009) (0.9) = 0.000081
170 (0.0582) (D. I ) = 0.005~~9
230 (0.0009) (0.1) = 0.000090

Total 1.000000

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics 383

Loss of load probability

Loss ofload probability can be obtained by combining the system load duration curve with the
capacity outage probability. It may be pointed that capacity outage may not necessarily result
in loss of load. This will be demonstrated in the example that follows.
It can be seen from Fig. 9.20 that any capacity outage less than the reserve capacity will
not result in loss of load.
Installed capacity
.+ Reserve capacity

ww 1
~ Load duration curve

w.E ~:
'-'I
"0 '

'"
0'

asy
....ll L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~

~ Time (for which load exceeded the value) .....

E
I
Fig. '9.20 Application of probability to loss of load

ngi
,

Installed capacity

nee
Reserve capacity

rin
..... ..... g.n
Fig. 9.21
ime (for which load exceeded the value)-
dified model of capacity on maintenance
..... .....

et
Installed capacity
Capacity on maintenance

Reserve capacity

1
~
6r-------~==--~------~
Load duration curve
alo
....l

o Time (for which load exceeded the value) _

Fig. 9.22 Alternative model for capacity on maintenance

Downloaded From : www.EasyEngineering.net

384 Operation and Control in Power Systems

If Pk is the probability of the outage Lk and tk is the period of time units for which the
outage causes loss of load. The total expected loss of load for the time period is
n
E(t) = L Pk tk time units
k=l

All outages in excess of the reserve capacity will result in loss of load for different time
periods and each outage Pk contributes a system load loss of Pk tk time units. For a daily load
duration, curve, the time period is hours. For annual load curves, daily peaks may be used.
The effect of capacity on maintenance can be considered in two ways. This is shown in
Fig. 9.20 & 9.21.

ww Consider, as an example for the loss of load probability calculation, a system with four
generating units, each rated at 65 MW. The forced outage rate for each generator unit is 0.04.

w.E
the load duration curve is a straight line with a maximum of 190MW and a minimum of
40MW. Since the installed capacity is 260MW, the reserve capacity is 260- J 90=70MW. The

asy
outage probability table is shown in Table 9.7

Table 9.7 Cumulative Probabilities for a 4-unit System

State Capacity
MW E
Capacity available
MW
ngi
Individual probability Cumulative probability

I
2
0
65
260
195 nee
0.84934000
0.14155000
1.00000000
0.15064000
3
4
130
195
130
65 rin
0.00884730
0.00024576
0.00909560
0.00024830
5 260 0 0.00000256
g.n
0.00000256

As may be noted from Table 9.7, the cumulative probability decreases as the capacity
outage increases.
Installed capacity
et
130 195
MW
f MW
Load
~ duration
6
'0 t2 = 83.3 h
'"
.9 40
~
o
p...

o Time (for which load exceeded the value) -..100%

Fig. 9.22 Loss of load with time

Downloaded From : www.EasyEngineering.net

Introduction to Advanced Topics

Since the reserve capacity is 70MW, state 2 does not Involve any loss of load. State 3
results in a loss of load for a period of 40% of time and state 4 has loss of load for 83.3% of
time. This is shown in Fig. 9.22. The expected load calculations are shown in Table 9.8.

Table 9.8 Expected Load Loss calculations

State Capacity Capacity available Individual Time tk in Cumulative
MW MW probability per cent probability
I. 0 260 0.84934000 0 -

2. 65 195 0.14155000 0 -

ww3.
4.
130
195
130 0.00884730
0.00024576
40 0.353890
0.020466

w.E
65 83.3
Total 0.374350

asy
If the load duration curve has a time period of about 261 days excluding 52 weekends
(Sundays and Saturdays) from 365 days, then the expected load loss is

En (0.37435)(261) = 0.97705 days

gi 100 -

nee
the maintenance of the desired level of reliability on the operation of the system is
dependent 0 the spinning capacity reserve. In the case of static reserve capacity, the available
generation is the entire installation capacity; while for a spinning reserve capacity, it is generally

r ing
assumed that there is sufficient installed capacity available and only the time delay in bringing
in a replacement is considered. Thus, in spinning reserve capacity calculations, outage

.
replacement rate is used in place of forced outage rate for probability evaluation. The outage

net
replacement rate is defined as the product of unit failure rate and delay time for replacement
unit. However, this is valid only if the outage replacement rate is much less than unity.

Downloaded From : www.EasyEngineering.net

386 Operation and Control in Power Systems

Questions

I. What are FACTS controllers? Explain

2. What are the various types of FACTS devices? Explain their working principle briefly
3. Explain the working of a dynamic voltage restorer
4. What is state estimation. What is the significance of it in power systems operation
5. Explain the various methods of load prediction
6. Why load prediction is necessary in power system operation? Explain

ww
7.
8.
Explain the term voltage stability. How can it be over come? Discuss
What do you understand by the term "power quality"? Explain
9.
10.
w.E
Discuss the various factors involved in power quality problem
Explain any method for the computation of loss of load probability.
II. Ho~
asy
reliability studies are important in power system operation? Explain.
12.
13. En
Explain "Security" in relation to power system operation.
What is contingency evaluation is connection with power system security.
14.
gi
What are energy control centers? Explain.

nee
r ing
. net

Downloaded From : www.EasyEngineering.net

Objective Questions

I.
ww
Consider a power system with three identical generators. The transmission losses are

w.E
negligible. One generator (GI) has a speed governor which maintains its speed constant
at the rated value, while the other generators (G2 and G3) have governors with a droop

asy
of 5%. If the load of the system is increased, then in steady state []

E
(a) generation of G2 & G3 is increased equally while generation of G I is unchanged.

ngi
(b) generation ofGI alone is increased while generation ofG2 & G3 is unchanged
(c) generation of G I, G2 and G3 is increased equally

nee
(d) generation ofGI, G2 and G3 is increased in the ratio 0.5:0.25:0.25
2.
rin
Gausss - Seidel iterative method can be used for solving a set of
(a) linear differential equations only
(b) linear algebraic equations on Iy
g.n
3.
(c) both linear and nonlinear algebraic equations
(d) both linear and nonlinear differential equations et
A power station consists oftwo synchronous generators A and B of ratings 250 MVA and
500 MVA with inertia 1.6 p.u. and 1 p.u., respectively on their own base MVA ratings.
The equivalent p.u. inertia constant for the system on 100 MVA common base is
(a) 2.6 (b) 0.615 (c) 1.625 (d) 9.0
4. The incremental cost characteristics of two generators delivering 200MW are as follows

<if'
_I = 2.0 + O.OIPI
dPI

For economic operation, the generations PI and P2 should be:

(a) PI = P 2 = 100 Mw (b) PI = 80MW, P2 = 120Mw
(c) PI = 200MW, P2 = OMw (d) PI = 120MW, P2 = 80Mw

Downloaded From : www.EasyEngineering.net

388 Objective Questions

5. A power system has two synchronous generators. The Governer - turbine characteristics
corresponding to the generators are
P J = 50(50 - f), P2 = 100(51 - f)
Where f denotes the system frequency in Hz, and P J and P2 are respectively, the power
outputs (in MW) of turbines 1 and 2. Assuming the generators and transmission network
to be lossless, the system frequency for a total load of 400MW is []
(a) 4705Hz (b) 48.0Hz (c) 4805Hz (d) 49.0Hz
6. The bus impedance matrix of a 4-bus power system is given by

ww jO.3435 jO.2860 jO.2723 jo.2277]

jO.2860 j0.3408 jO.2586 jO.2414

w.E Z
BUS
-
- jO.2723 jO.2586 jO.2791
[
jO.2209
jO.2277 jO.2414 jO.2209 jO.2791

asy
A branch having an impedance ofjO.2 ohm is connected between bus 1 and the reference.

En
Then the values of Z22,new and Z 23,new of the bus impedance matrix of the modified
network are respectively
gi
(a) jO.5408 ohm and j0.4586 ohm
(c) jO.5408 ohm andjO.0956 ohm nee
(b) jO.1260 ohm and jO.0956 ohm
(d) jO.1260 ohm andjO.1630 ohm
[ ]

7.
r ing
A power system consists of 300 buses out of which 20 buses are generator buses, 25
buses are the ones with reactive power support and 15 buses are the ones with fixed

.
shunt capacitors. All the other buses are load buses. It is proposed to perform a load
flow analysis for the system using Newton - Raphson method. The size of the
Newton - Raphson Jacobian matrix is [
net ]
(a) 553 x 553 (b) 540 x 540 (c) 555 x 555 (d) 554 x 554
8. Ifa generator of250 MVA rating has an inertia constant of6MJIMVA, its inertia constant
on lOOMVA base is [ ]
(a) 15MJIMVA (b) lOo5MJ/MVA (c) 6MJ/MVA (d) 2.4MJ/MVA
9. In load - flow analysis, the load at a bus is represented as [
(a) a constant current drawn from the bus
(b) a constant impedance connected at the bus
(c) constant real and reactive powers drawn from the bus
(d) a voltage - dependent impedance at the bus

Downloaded From : www.EasyEngineering.net

Objective Questions 389

10. A transmission line has equal voltages at the two ends, maintained constant by two
sources. A third source is to be provided to maintain constant voltage (equal to end
voltages) at either the midpoint of the line or at 75% of the distance from the sending
end. Then the maximum power transfer capabilities of the line in the original case and
the other two cases respectively will be, in the following ration. []
(a) 1 : 1 : 1 (b) I : 2 : 1/0.75 (c)I:2:4 (d) I : 4: 16
11. The inertia constant ofa I OOMVA, 50Hz, 4-pole generator is 10MJ/MVA. If the mechanical
input to the machine is suddenly raised from 50MW to 75MW, the rotor acceleration
will be equal to [ ]

ww (a) 400kV (b) 260kV (c) 80kV

12. Economic scheduling of generation is performed for a period of
(d) 40kV

w.E
(a) Y:z hr (b) 24 hr (c) I sec (d) I week

asy
13. The input - output characteristic is plotted
(a) Fuel input Vs power output (b) Fuel input Vs time

E
(c) Fuel rate V s power output
14. Hear rate curve is plotted in
ngi
(d) Fuel rate Vs energy output

(a) k-cal/kwhr Vs KW
(c) k-callKW Vs Kw nee
(b) K-cal/hr Vs KW
(d) K-cal Vs KW

rin
15. Incremental production cost and incremental fuel cost are
(a) both the same (b) IPC > IFC
g.n
(c) IPC < IFC (d) They are not related
16. The incremental transmission loss for a 2-plant system is given by et

17. Consider the following constraints

(i) Pa-Po-P=O (ii) Qa - Q O- Q = 0
(iii) p2 + Q2 :s; (S rated? (iv) pmm :s; P :s; pmax

The equality constraints are

(a) (i) only (b) (i) (ii) and (iii) (c) (i), (iii) and (iv) (d) (i) and (ii)

Downloaded From : www.EasyEngineering.net

390 Objective Questions

18. Penalty factor is given by ]

(a) -dF ( 1 +
OPL-) (b) -dF ( 1 OPL
- -)
dPj oPi dP j oPI (c) (
1+_L
OP)
oPi
(d) (1- oP OPL )

19. IfR = 0.04 HzJMw by how much will the turbine power decrease, iffrequency rises by
O.IHZ [ ]
(a) 0.024Mw (b) 1I0.024MW (c) 0.0024Mw (d) 4.17MW
20. What is the load damping factor D for an area operated at 1000Mw, 50Hz

ww
(a) 100 (b) 50 (c) 25 (d) 20

w.E
21. If the power system has a transfer function
Kp
1 +STp
and given that the load damping

asy
D = 25MW/Hz. What is the value of Kp
(a) 0.04HzJMW (b) OAHzJMW (c) 0.25HzJMW (d) 25HzJMW

1
E
22. Area frequency response characteristic is

I ngi D D
[

(a) D+-
R
(b) R +-
D
nee
(c) -
R
(d) - -
R

increase if the area is rated for 2000MW

rin
23. A speed governing system has D = 20MWlHz. What is the frequency drop for 1% load
[ ]
(a) 1 Hz (b) 1.5 Hz (c) 0.5 Hz
g.
(d) 0.2Hz
24. Steady state frequency error in a single control area can be eliminated by
net
(a) proportional control (b) derivative control
(c) integral control (d) proportional plus derivative centro I
25. In a single uncontrolled area, area control error is defined by

d
(a) fddt (b) -M (c) df (d) - .!:df
dt R
26. When integral control is applied, if the control gain Kr is greater than critical gain, the
response is // [ ]
(a) over damped (b) damped oscillatory
(c) un damped oscillatory . I( d) unstable operation

Downloaded From : www.EasyEngineering.net

Objective Questions 391

-
27. Which of the following statements is correct?
(a) primary load frequency control is performed by integral control
(b) secondary load frequency control is performed by integral control
Cc)- speed governor action always keeps the speed constant with load changes
Cd) economic load scheduling is always required to be performed before load frequency
control
28. Turbine generator dynamics and speed governor actions affect the performance of the
frequency response with load [ ]

ww
(a) seriously
(b) only negligible action at t = 0+

w.E
(c) may cause trouble if governor is fast in action
(d) may cause trouble if the rating of the turbine is very high

asy
29. The time constant of the power system Tp is defined by

(a) Tp = -o-
2H
En( b) T = 2fo
HO
(c) Tp =-o-
20 f~
(d) 2HD
f D
gi
p f H
30. Which of the following correct?
nee
(a) load frequency control has serious affect on voltage profile
]

r ing
(b) if a generator is connected to an infinite bus the frequency change with load is
negligible

.
(c) turbine power increment for load change does not depend upon the response
characteristic of the turbine net
(d) all the generating stations in any country can be treated as belonging to a single
control area
31. Given til = 50Hz, 0 = 20Mw/Hz, area rated- power = 2000Mw and H =- 5 sec the area
time constant Tp is [ ]
(a) 5 sec (b) 10 sec (c) 15 sec (d) 20 sec
32. Unit commitment is
(a) Economic schedule among different units
(b) Planning of generating units for future load increase
Cc) Optimal combination of units for operation at anyone time
Cd) To choose proper units from thermal, hydro and nuclear plants

Downloaded From : www.EasyEngineering.net

392 Objective Questions

33. Which of the following is valid constraint for unit commItment

( c) I. pI-< pmax
pmlO :os;
I I -J I
(d) Vmtn < pm3x 2 + Qmax 2
I
34. Which of the following is a valid dynamic programming equation I
(b) Fn(~Pi)-fn(Pn)+Fn_{~PI)

ww (d) F,(t,P} m+N(PN)+ F,_.(~P,)]

w.E
35. Dual variables are
(a) Lagrange multipliers
[ ]

asy
(b) Kuhn - Tucker multipliers

(d) the problem variables E

ngi
36. The dependent variables in optimal load flow studies are ]
(a) P and Q (b) IVI and 0 (c) P andnee IVI (d) P,Q, IVI and 0
37. Given Bll = 0.02, B22 = 0.04; B12 - B21 = 0.001
rin
2 tv
g.n
et
load

What are the transmission loss when lOMw is supplied by plant 2 to the load [ ]
(a) 2Mw (b) 4Mw (c) O.OOIMW (d) zero
38. 0.04 regulation in speed for a 100Mw rated generator means [ ]
(i) 2.0Hz drop in frequency
(ii) 40Hz drop in frequency
(iii) R = 0.02 HZ/Mw
(a) (i) is only correct (b) (i) and (ii) are both correct
(c) (ii) and (iii) are both correct (d) (i) and (iii) are both correct

Downloaded From : www.EasyEngineering.net

Objective Questions 393

39. In a power system during load flow studies

(i) there are four variables at each bus
(ii) the voltage magnitude at one bus is specified and this bus is called voltage controlled
bus
(iii) at all the load buses P and Q are specified for the loads
(iv) power losses are 'assigned to swing bus.
Which of the above is I are correct?
(a) '(i) and (ii) (b) (ii) and (iii)
(c) (i) (iii) and (iv) (d) (i) (ii) (iii) and (iv)

ww
40. In load flow studies
(i) convergence of iterative method depends upon the diagonal dominance in YBUS

w.E
matrix
(ii) sparsity of Z - bus matrix is well exploited

asy
(iii) for large well conditioned system with n-buses the number of iteration required is
approximately n2 •

I are true,En
(iv) Z - matrix methods is very sensitive to choice of slack bus. Which of the above is
[ ]

gi
(a) (i) and (ii)
(c) (iii) and (iv)
(b) (ii) and (iv)

nee
(d) (i) and (iii)
41. In load flow solution by N-R method
r ing
(i) the rectangular coordinates method is slower compared to polar - coordinates method
(ii) the Jacobian in case of rectangular coordinates is not symmetric

.
(iii) the Jacobian with usual formulation in polar coordinates is symmetric
(iv) the N-R method converges in 2-5 iterations
net
which of the following is correct
(a) (i) and (iv) are true (b) (ii) and (iv) are true
(c) (i), (ii) and (iv) are true (d) (i), (ii) and (iii) are true
42. Economic dispatch problem is related to
(i) selection of generating units for loading
(ii) planning for generating system expansion
(iii) determine profit from energy sales
(iv) ascertain cost of energy production
(a) (i) and (iii) (b) (i) and (iv)
(c) (ii) and (iii) , (d) (iii) and (iv)

Downloaded From : www.EasyEngineering.net

394 Objective Questions

43. In the emergency mode of operation

(a) only system frequency is maintained
(b) only maximal load demand is being met
(c) both Ca) and (b) are maintained
(d) only system frequency and voltage are maintained
44. Electrical stiffness of the transmission line is given by

L\P L\P L\Q L\Q

M
ww (a) L\cS (b) -
M (c) L\cS (d)

w.E
45. Transmission capacity can be increased by
(i) reducing the effective reactance of the line

asy
(ii) increasing the voltage levels
(iii) by suing bundled conductors

E ngi
(iv) by increasing the load angle d to some extent
(a) (i) and (iii) are the same

nee
(b) (i) (ii) and (iii) increase the power transfer
(c) (i) and (iv) do not mean the same control
rin
(d) bundled conductors infact decrease the transmission capacity
46. The natural load of a transmission line can be expressed by g.n ]
(a)B.X\'- (b)B/X (c)X/B (d)B+X
et
where B is the line susceptance and X is the line reactance of the line being considered
47. The transmission losses in a line are
(a) directly proportional to voltage Y
,(b) inversely proportional to voltage Y
(c) directly proportional to y2
(d) inversely proportional toy2
48. Rapid changes in system voltage cause
(a) Flicker (b) Yoltage dips
(c) corona (d) loss of synchronism

Downloaded From : www.EasyEngineering.net

Objective Questions 395

49. Shunt connected controllable var compensators will

(i) improve power transfer capability
(ii) improve stability
(iii) improve power transfer but deteriorate stability
(iv) improve stability but reduced power transfer capacity
(a) (iii) is correct (b) (iv) is correct
(c) (i) and (ii) both are correct (d) (i) is only correct all other are wrong.
ww
50. Voltage response of an exciter expressed as
(a) Volts / sec

w.E
(b) volts / ampere
(c) field amperes per output amperes

asy
(d) change in field voltage to change in output voltage

(a) IOOV E
51. The exciter voltage for a 100MVA set is about
(b) 440V
ngi
(c) IKV
52. As per IEEE standard, the exciter block is represented by
(d) 250V

nee
~
(a) 1+ STe (b)
I +STe rin
53. The voltage regulator block diagram shows an
(i) inherent phase lag (ii) inherent phase lead
g. net ]

(iii) inherent steady state error (iv) inherent instability at larger gains
(a) (i) and (ii) are correct (b) (ii) and (iii) are correct
(c) (ii), (iii) and (iv) are correct (d) (i) iii) and (iv) are correct
54. Under sterdy state condition, if K is the over all forwar~ path gain. Then, the steady
state error .1ess is [ ]

I I
(a).1e =-- (b).1e ss =K-
ss 1+ K

I
(c) .1e ss is independent ofk (d) .:1e ss =1+ K

Downloaded From : www.EasyEngineering.net

396 Objective Questions

55. A stabilizing transformer

(i) introduces phase lead
(ii) introduces phase lag
(iii) increases stability margin
(iv) has its secondary drawing negligible current in a voltage regulator
(a) (ii) is correct (b) (i) and (ii) are both correct
(c) (ii) and (iv) are correct (d) (ii) (iii) and (iv) are correct

ww
56. A power system stabilizer improves

w.E
(i) phase lag
(ii) improves damping

asy
(iii) produces torque in phase with speed
(a) (i) and (ii) only are correct (b) (ii) and (iii) only are correct

57. Which of the following is false

E
(c) (i) and (iii) only are correct

ngi
(d) (i) (ii) and (iii) are correct
[

nee
(a) synchronous motors operate with adjustable excitation
(b) synchronous compensators cannot given dynamic reactive power compensation

rin
(c) by supplying a synchronous compensator from a tertiary winding, need for separate
transformer can be avoided.
(d) A synchronous compensator can even improve stability g.n
58. Per unit change in vol.tage magnitUde is equal to et
~Q S
(a) - (b) J£. (d) l:!.Q + Sse
·Sse ~Q

where ~Q is the change in reactive-power and Sse is the short circuit capacity.
59. Which of the following are true?
(i) the quadrature voltage compensator would control active power flow in the system
(ii) an in-phase voltage booster would control the reactive power flow in the system
(iii) the natural loading limit is the lowest limit for power transfer
(iv) thermal limit is the highest limit for power .transfer
(a) (i) and (ii) are only correct (b) (ii) only is correct
(c) (i) (ii) and (iii) only are correct (d) all are correct

Downloaded From : www.EasyEngineering.net

Objective Questions 397

60. Line compensation

(i) increases Ferranti effect
(ii) requires under excited operation of generators
(iii) reduces power transfer capabi Iity
(iv) is never used in power system operation
(a) (i) and (ii) are false (b) (ii) and (iii) are false
(c) (i) (iii) and (iv) are false (d) all are false
61. When sending voltage is equal to receiving end voltage, for maximum power at receiving

ww
end .
X
(c) R = -
[
R
]

(a) X = J'jR
w.E (b) R = J'jx

X
R
2
(d) X = -
2

62. For a 400kv line the ratio

(a) 1.5 (b) 6asy is about
(c) 10 (d) 16

En
63. In Var compensators using thyristors
(a) filters are not necessary
(b) the dominant harmonic is second
(c) filters are needed for 5th and
gi 7th harmonic nee
(d) capacitors are switched on the h.v. side only
rin
64. Static var compensators
(i) reduce voltage swings at the rolling mills g.n J

(ii) improve power factor

(iii) compensate the unbalanced reactive load of arc furnace loads
et
(iv) can control the fundamental component of current flowing in an inductor
(a) (i) and (ii) are correct (b) (ii) and (iv) are correct
(c) (i), (ii) (iii) and (iv) are correct (d) all are correct
65. Series - capacitor compensation
(i) produces undamped or slightly damped oscillations
(ii) produces oscillations the frequency of which are of the order of 103 to 106 K.Hz. -
which causes serious problem
(iii) may produce electromechanical oscillations of frequency I to 2 Hz
(iv) will not produce any self excitation.
Which of the above is true?
(a) (i) and (ii) (b) (ii) and (iv) (c) (iii) and (iv) (d) (i) and (iii)

Downloaded From : www.EasyEngineering.net

398 Objective Questions

66. The power flow from area 1 to area 2 is giv~n by with usual notation

67. Synchronizing coefficient of a line is given by

AP
(a) Ts = A8

ww (c) TsAP = [Ao l - Ao 2]

w.E
68. Which of the following is correct

(i) M = _I ~ A8 Hz (ii) M =21t fAfdt rad

asy
21t dt

21t
En
(iii) M = _I f Afdt (iv) M = 21t fMdt

(a) (i) and (ii)

gi (b) (i) only (c) (i) and (iv)

nee
(d) (i) (iii) and (iv)
69. The inertia consists of two groups of machines which do not swing together are MI and
M2• The equivalent inertia constant of the system is : .[ ]
(a) MI + M2
M,M 2 ~
r
(b) M, - M2 ifMI > M2
ing
(c) M +M
, 2
(d) "M I M 2
. net
70. If AP OJ = K\ and AP 02 = 0, in a two area system, the steady state frequency error is

71. What is the tie-I ine power deviation if AP D:! = Is and AP DI =0 for a two area system

72. For two equal areas, for a load disturbance of AP D in an area, the steady state frequency
error is [ ]
AP ~ 2~ (d) A~PD
(a) _D
(c) AP
2~
(b) 2AP 2
D D

Downloaded From : www.EasyEngineering.net

Objective Questions 399

73. If two equal areas with tie line power interchange of P12 experience a load change f1P D
in one of the areas, the tie line power deviation is [ ]
L\PD • n. L\PD n. L\PD
(a) - (b) 1-'1 -2- (c) 1-'_1 ") (d) L\P o
2 ....
74. Which of the following is true
(i) A synchrono\ls generator at sub synchronous frequencies acts as an induction
generator
(ii) At sub synchronous frequencies the rotor resistance viewed from the armature

ww ter.minals is negative.
(iii) At sub synchronous frequencies resonance may occur between electrical and

w.E
mechanical system
(iv) Series compensation causes sub synchronous resonance at very low frequencies

asy
(a) (i) is true
(c) (i), (ii) and (iii) are true
(b) (i) and (ii) are true
(d) (i), (ii), (iii) and (iv) are true

En
gi nee
r ing
. net

Downloaded From : www.EasyEngineering.net

Answers to Objective Questions

1. (b) 26. (b) 51. (d)

ww
2.
3.
(b)
(b)
27.
28.
(b)
(b)
52.
53.
(e)
(d)
4.
5.
w.E
(d)
(b)
"29.
30.
(a)
(b)
54.
55.
(a)
(d)
-6. (b)
(b)
asy 31.
32.
(d)
(e)
56.
57.
(d)
(b)
En
7.
8. (a) 33. (e) 58. (a)
9.
10.
(e)
(a)
(a)
34.
35.
36.
gi
(d)
(c)
(b) nee
59.
60.
(d).
(d)
11. 61. (a)
12.
13.
(a)
(a)
37.
38.
(d)
(~)
r 62.
63. ing(d)
(e)
14.
15.
(a)
(b)
39.
40.
(d)
(d)
64.
65.
.
(d)
{d) net
16. (d) 41. (d) 66. (a)
17. (d) 42. (d) 67. (a)
18. (d) 43. (c) 68. (a)
19. (d) 44. (a) 69. (e)
20. Cd) 45. (d) 70. (a)
21. (a) 46. (b) 71. (a)
22. (a) 47. (d) 72. (a)
23. (a) 48. (a) 73. (a)
24. (e) 49. (e) 74. (d)
25. (e) 50. Ca)

Downloaded From : www.EasyEngineering.net

References
Books
1. Beveridge G.S.G and R.S. Schechter, Optimisation: Theory and Practice, McGraw-

ww
2.
Hill Book Co., New York, 1970.
Rao, S.S., Optimisation: Theory and Application, Wiley Eastern Ltd., New Delhi.

3.
1978.
w.E
Kirk, Donald E., Optimal Control Theory: An Introduction, Prentice Hall Inc.,

asy
Englewood Cliffs, New Jersey, 1970.
4.
Book Co., New York, 1971.
En
Eigerd, Olle t, Electric Energy Steams Theory: An Introduction, McGraw Hill

5.
gi
Kirchmayer, L.K., Economic Control of Interconnected Systems, John Wiley and
Sons, New York, 1959.
nee
6. Zaborszky L. and J. W. Rittenhouse, Electric Power Transmission, The Ronald

7.
Press Co., New York, 1954.
r ing
Taylor, Openshaw (Ed.), Power System Plant, George Newnes Ltd .. London.

8.
1955.
Stagg Glenn w.,
. net
and A.H. EI-Abiad, Computer Methods in Power System Analysis
McGraw-Hill B09k Co. Ltd., New York, 1968.
9. Blindleistung, VDE Verlag, GmBH, Berlin 1963.
10. Handschin E. (Ed.), Real Time Control of Electric Power Systems, Elsevier
Publishing Co., Amsterdam 1972.
11. P.M. Anderson, Power System Control and Stability, A.A. Fauad, The Iowa State
University Press.
12. Paul.e. Krause, Analysis ofEleotrical Machines, McGraw Hill Book co.
13. V. Venikov, Transient Processes in Electrical Power System, Mir Publishers, Moscow
14.
15.
.
Hadi Saadat, Power System Analysis. McGraw Hill International Edition .
PSR Murty, Power System Operation and Control, Tata - McGraw HiU -1984.

Downloaded From : www.EasyEngineering.net

402 References

Periodicals and Conference Proceedings :

1. Albert M., 'Nonlinear programming solutions for load flow-Minimum loss and
Economic dispatching solutions, IEEE Trans., PAS-88, No.4. April 1969.
2. Funk, G. "Die Spannungsabhangigkeit Von Drehstrom lasten" Elektrizitats
wirtschafi,68,1969.
3. Murty, P.S.R., "Load modelling for power flow solution," Journal ofIE(India), 58,
Pt. EL3, Dec., 1977.
4. IEEE Committee Report, "Dynam ic Models for steam and hydro turbines in Power

ww system studies", IEEE Trans. PAS May 1969.

5.
w.E
Dsewenski, A.l., "Uber den Einfluss der spannungsenderung auf die wirk-und
Blindleistung der Verbraucher", Energietechnik, 12 Nov., 1953.
6.
asy
Anschutz, H., "Frequenz-und Spannungsschwankungen in Drehstromnetz und ihre
Auswirkungen," ETZ-13, Dec. 1949.
7.
En
Woodwards J.L. "Hydraul ie turbine transfer function for use in governing studies",
IEEE Proc, lIS No.3, Mar. 1968 (Correspondence).
8. gi nee
Ramey, R.<J. and J.w. Skooglund, "Detailed hydro-governor representation for
stability studies, IEEE Trans. PAS, Jan. 1970.
9.
rin
Hale, H.W. and J.W. Ward, "Digital computer solution of power flow problems",
AlEE Trans. PAS, 75, June 1956.
10. g.n
Glimn A.P. and G.W. Stagg, "Automatic calculation ofload flows", AlEE Trans.

11.
(PAS), 76, June 1956. '
et
Br.own H.E. et ai., "Power flow solution by impedance matrix iterative method",
IEEE Trans (PAS), 82, Apr. 1963.
12. H.E. Brown, Z-matrix algorithms in load flow programs, IEEE Trans (PAS), 87
March 1968.
13. Hart, C.E. and W.F. Tinney, "Power flow solution by Newton's method", IEEE
Trans (PAS), 86, 1967.
14. Stott. B. "Decoupled Newton-load flow", Trans (PAS), 91, Sep/Oct. 1972.
15. Peterson N .M. W.F. Tinney, and D. W: Bree, "Iterative linear ac power flow solution
for fast approximate outage studies", IEEE Trans. (PAS), 91, Sep./Oet. 1972.
16. Stott, Band O. Alsae, "Fast decoupled load flow", IEEE Trans. (PAS), 93,
May/ June 1974.

Downloaded From : www.EasyEngineering.net

References 403

17. Tinney W.F. and J. W. Walker, "Direct solution ot spares network equations by
optimally ordered triangular factorisation", Proc. IEEE, 55, Nov. 1967.
18. Stott, B, "Load flow calculation methods", Proc. IEEE, 62, July 1974.
19. Noakes, F. and A. Arismonandar, "A Bibliography on optimum operation of power
systems 1919-1959", IEEE Trans (PAS), 82,1963.
20. Baker, A.D. and S.O. Navarro "The. use of cubic cost curVes in the economic
loading of electric power systems", IEEE Trans. (PAS), 84,1965.
21. Smith, H.M. and S. Y. Tong, "Minimizing' power transmission losses by reactive
volt amperes control", IEEE Trans (PAS), 82,1963.

ww
22. Bauman, R."Power flow solution with optimal reactive flow"., Archiv fur Elektro-

23. w.E
techn ik, 48,1963.
Dopazo, J.F. O.A. Klitin, G. W. Stagg andM. Watson, "An optimization technique

24. asy
for real and reactive power allocation", Proc [EE; 45, 1967.
Carpentier, J. "Contribution a I'etude du dispatching economique'Bulletin de la

En
Societe fransaise des Electriciens, 3 Aug. 1962.
25.

26.
gi
Carpentier, J. and J. Siroux, "L'optimisation de la production a L'Electricite de

nee
France", Bulletin de la Societe Frallcaise des Electriciens, 1963.
Pescon, J.,D.S.Piercy,W.F. Tinney an~ O:J. Tveit, "Optimum control of reactive

27.
power flow," IEEE Trans. (PAS), 1968.
r ing
Shen, C.M. and M.A. Laughton, "Determination of optimum power system,

28.
.
operating conditions under constraints", Proc. IEEE, 116, Feb. 1969.
Dommel, H. W. and W.F.Tinney, "Optimal power flow solutions", IEEE Transnet
(PAS) 87, 1968.
29. EI- Abiad, A.H. and FJ. Jaimes, "A method fOT optimum scheduling of power
and voltage magnitude, IEEE Trans (PAS) 88, Apr. 1969.
30. Fletcher, R. and MJ.D. Powell, "A rapidly convergent descent method for
minimization," Computer journal, June 1963.
31. Bird, J.1.,' 'Load dispatching on the CEGB System," Paper presented at
Symposium on Optimal power system operation, Univ. of Manchester, Sep. 1969.
32. Lowery, P.G. "Generating unit commitment by dynamic programming," IEEE Trails
(PAS) 85,1966.
33. Udo, M., "Optimisation of start-up and shut-down of thermal units", Proe.
Third PSCC 1969 Rome paper 0.S.7.

Downloaded From : www.EasyEngineering.net

404 References

34. Kerr, R.H., J.L. Scheidt, A.J. Fontana and J.K. Wiley, "Unit Commitment" IEEE
Trans (PAS), 85, 1966.
35. Hara, K. Kimura M. and N. Honda, "A method for planning economic unit
commitment and maintenance of thermal power systems", IEEE Trans. (PAS)
85,1966.
36. Nagrath, I.J. and S.K. Agarwal,"Optimum scheduling of hydrothermal systems",
Proe. lEE, 1) 9, Feb. 1972.
37. Nagrath, I.J., G. Dayal and D.P. Kothari, "Optimum scheduling ofhydro-thennal
system taking transmission loss into account", Jour. IE (India), 52. Oct. 1971.
38.,ww Drake, J.H., L.K. Kirchmayer, R.B. Mayall, and H. Wood, "Optimum operation

39. w.E
ofa hydro-thermal system," Trans. AlEE (PAS), 82, Aug. 1962.
Bonaert, A.P. A-EI-Abiad and Koivo "Optimum scheduling of hydro-thermal power

40. a
systems", Proc. Seventh PICA Con/., Boston, 1971.

syE
Murty, P.S.R., "Effect of system loads on optimal load flow solutions," Jr.
(India), 58,.Pt. EL3, Dec. 1977.
[£

41.
ngi
Hochstetter Werner, "Properties of static compensators for power supply systems",

42.
Siemens Review, Aug. 1977.
nee
Achenbach, H., W. Hanke and W. Hochstetter, "Controllable static reactive power

rin
compensators in electrical supply systems", Second IEAC Symposium on Control
in Power Electronics and ElectricaJ Drive, Dusseldorf, Oct~ 1977.
43. g .ne
Tyll, Heinz, '''Use of static compensators in transmission system", Symp. On
Reactive Power Compensation, Rio de Janeiro, Brazil, May 1979.
44.
mark power system, IEEE Trans. (PAS) 97, May/June 1978.
t
Fouad, A.A., and K. T. Khu, "Subsynchronous resonance zones in the IEEE bench

45. Kilgore, L.A., L.e. Elioh and E.R. Taylor, "The prediction and control of self
excited oscillations due to series capacitors in power systems", IEEE Trans. (PAS),
90, May/June 1971.
46. Hergert, J.e. and e. W. Park, "Parameter identification and verification of low
order boiler models", IEEE Trans. on PAS. PAS-75, No.4, July/Aug. 1976.
47. Cohn, Nathan "Power flow control-Basic concepts for interconnected systems",
Mid West Power Conference, USA, 1950.
48. Elgerd, 0.1. "Trends and philosophies of power systems control", Alterton
Conference on Circuits and Systems, Oct. 1973.

Downloaded From : www.EasyEngineering.net

References 405

49. Cohn, Nathan, "Fundamentals for control of generati<;m and power flow on
interconnected systems", Invited Talk, North American Power Systems
Interconnection Committee Meeting, Las Vegaes, Navada, USA. Sep. 1970.
50. mgerd, 0.1. and C.E. Fosha, "Optimum megawatt-frequency control ofmulti-ar~
electric energy system", IEEE Trans (PAS), 89, Apr. 1970.
51. Fosha, C.E. and 0.1. Elgerd, "The megawatt-frequency control problem: A new
approach via optimal control theory", IEEE Trans (PAS) 89. Apr. 1970.
52. Murty, P.S.R. and P. V. Ramana Rao, "Optimal load frequency control and system

ww modelling". Jour IE (India), Pt. EL4, 57, Feb. 1977.

Ramana Rao P. V. and P.S.R Murty, "Identification an\t verification of reduced
53.
w.E
order steam reheat turbine model", Jour. (IE), Pt. ME6, May 1981.
54.
asy
Ramana Rao, P. V. and P.S.R Murty, "Modelling, Identification and optimal control
of hydro governor system", Jour. IE (India), 60, Pt. EL2, Oct. 1979.
55.
En
Venkateswarlu K. and A.K. Mahalanabis, "Optimum load frequency control with
. restrictions on the rate of power generation, Jour I.E. (India) Pt. EL4, 58, Feb.

56.
1978.
gi nee
Venkateswarlu K. and A.K. Mahalanabis. "Load frequency control using output

57. r
feedback", Jour. IE (India), Pt EL4 58, Feb. 1978.

ing
Nanda, J. M.L. Kothari, and P.S. Satsanghi, "Automatic. Generation control reheat

58.
Journal IE (India), Pt EL6, 63. June 1983. .
thermal system considering generation rate constraint and governor dead band'.,

net
Liacco Thomas E.Dy. et aI., "Automation of CEI system lor security", PICA
Boston, USA, 1971.
59. Bhaskaran K., "Technology trends in power plant operation", Second Symp. on
Power plant dynamics and control BHEL (R&D), Hyderabad, Feb. 1979.
60. Liacco, Thomas E. Dy "Real time computer control of power systems", Proc. A
IEEE, 162, July 1974.
61. Larson, R.E. et aI., "State estimation in power systems", IEEE Trans. (PAS)
89,1970.
62. Kuppurajulu, A., "Digital Computers for on line control", NPSC, Electrical Engg.
Dept., Osman ia University, Dec. 1981.
63. Choudhari, S.H. "Turbine instrumentation and automation", NPSC, Elec. Engg.,
Osmania University, Dec. 1981.

Downloaded From : www.EasyEngineering.net

406 References

64. Gupta, P.G., "Stochastic approach to peak power demand forecasting in Electric
Utility systems", IEEE Trans. on PAS PAS-90 No 2, Mar./Apr. 1971.
65. Christiannse, W.R., "Short-term load forecasting using exponential smoothing",
IEEE Trans on PAS PAS-90 No.2, Mar'! Apr. 1971.
66. Mathewan, P.D. and H. Nicholsan, "Techniques for load prediction in the electricity
supply industry", Proc. lEE (London), tiS (power) Oct. 1968.
67. Farmer, D.E. and MJ. Potton, "Development of on-line prediction techniques
with results from trials in the south West Region of the CEG", Proc lEE Londo!"!,

ww 115, Oct, 1968.

w 68. Toyoda. J, Mo-Shing Chev and Yukiyoshi Inoue, "An application of state to

.Ea
estimation to short-term load forecasting part-I "Forecasting modelling Part-II,
Implementation, IEEE Trans. On PAS, PAS - 89, No 7, Sept/Oct 1970.
69.
syE
Larson, Robert E, Willam, F, Tinney and John Peschan, "State estimation in power
systems part-I theory and feas.ibility", IEEE Trans. on PAS, PAS. March 1970.
70.
ngi
Larson, Robert E., W.F. Tinney, L.P. Hajdu and D.S. Piercy "State estimation in

nee
power systems Part-II: Implementation and applications", IEEE Trans on PAS 80,
March 1970.

Lecture Notes and Others: r ing

I.
. net
Real time control of power systems for security, Summer school notes, sponsored
by ISTE, College ofEngg. Guindy, Madras, May/June 1976.
2. General Electric's static var control, General Electric Brochure, Switchgear
Business Dept. USA.
3. Advances in power systems, Winter school notes, sponsored by ISTE, Regional
Engg. College, Warangal, 1967-77.
4. P.Y. Ramana Rao, Parameter estimation and model reduction of power plant
components using identification techniques for power system control, Ph.D. thesis,
Kakatia Un iv., Warangal, May 1981.
5. P.S.R. Murty., Optimal load flow studies in interconnected networks, (German),
Doctoral thesis, Technical University, Berlin.

Downloaded From : www.EasyEngineering.net

In.dex

A E
Active Power Scheduling 103 Effect of Bias Factor ,267

ww
Analysis of Single Area System 205
Arc furnace 293, 332,333
ElectricalStiffness 7
Emergency Mode 6

w.E
Area Load Dispatch Centre 367 Energy Control Cente~ 7
Energy Scheduling 121

asy B Equal Incremental Cost Metho<l 97

B. coefficients 98,99
Basics 200
Brushless excitation 300
E ngi
Evaluation oft for Computation 107
Excitation Control 295,296
Exciter 297,298
Brushless Excitation Scheme 301
Bus Impedance Formation 29 nee
Exciter BUjId Up 298

c rin
Exciter Ceiling Voltage 298
Exciter Response 298

Capacrto~ 7,315,321
Exponential Smoothing 378,379
g.n
Commitment schedules 171, 173
Compensation 289,292,320,321 FACTS Controlte~ 7,346
F
et
Complete Tie-Line Bias Control 250,251,266 Fast Decoupled Methods 27
Constraints 94,95 Flat Frequency Control 201, 241, 244, 251
Constraints for Plant 173 Flat Tie-Line and Flat Frequency Control 244
Cost Function 171,173, 180 Flicker 352, 355
Fuel ordering 5
D
Decentralised Control 284 G
Decoupled Methods 27
Gauses - Seidel Iterative Method 13
Dynamic Programming 175, 177, 179-, 180
Generating Capacity Reliability and Outage
Dynamic Response 199,208,237,257,287,306 Probabi 380

Downloaded From : www.EasyEngineering.net

408 Index

Generating Costs at Thermal Plants 93 LoadPrediction 7,370,371

Governor 182, 183 Loadability 351,352
Governor Controlled Valves 183 Loading Capability 320
Gradient Method for Optimal Load Flow 156 Loss of Load Probability 383,384

H M
Heat Rate Characteristic 88 Matrix Riccati Equation 224
Hydro Speed Governing System 232,235' Merit Order Method 97
Hydro Thermal Scheduling 5, 120, 125 Modal for a Steam Vessel 196

ww I
ModelReduction 284,285

w.E
Modem Control Theory 224,239
IEEE Type 1 Excitation System 310 MUltiple Regression Methods 371
Impedane of Reactive Power 289

asy
Incremental Production Cost Characteristics 89
Incremental Water Rate Characteristics 91 NetworkAutomation 368
N

E
Induction Regulators Static VarCompensators 7
Input Output Curves 87 ngi Newton Raphson method 16, 71
Ncn-linearProgramming 157
Interconnected Operation 241,242, 266
nee
Nonlinear Regression Methods 375
Normal mode 6

Kleinman's method 227

K
roin
L
Optimal Control 222,223
g.n
Lagrange Function 156, 158, 159
Lagrange Multipliers 97
Optimal Load Frequency Control 225
Optimal Ordering 25,26
Output Feedback 237
et
Least Squares Estimation 359 Overvoltages 334
Least Squares Polynomials 374
Level Decomposition in Power Systems 367
p
Linear Estimation Theory 359 Pattern Recognition Methods 363
Linear Regression Techniques 371 Penalty Factor 106
Load Compensation 321 Penstock Turbine Model 193, 195
Load flow analysis 8, 73 PID Controllers 212
Load Flow Solution Using Z Bus 29 Plant Ordering 5
Load Frequency Control Plant Scheduling Methods 96
183,200,201,202,208,212 Power Acceptability or Voltage Tolerance 356

Downloaded From : www.EasyEngineering.net

Index 409

Power Quality 7,352,353 Speed Changer 183,184,186,206

Power Quality Index 353 Speed Control Mechanism 183, 186, 188
Power system 182, 188,201, 208, 230 Speed Governing Mechanism 183, 192
Power System Control Centres 365 Speed Governor 182, 183
Power System Security 360,386 Spetral Expansion Method 376
Power System Stabilizer 313 ~tability Compensation 307
Power Transfer 292, 295, 296, 321 Stabilizing Transformer 307
Preventive mode 6 State Estimation 7,357, 358,360
Priority - List Method 174 State Estimation in Load Forecasting 379

ww
Pumped Storage Plant 120
PV-bus 14,22
State Variable 269,274,275
State Variable Model 269, 274, 275

w.E R
Static Compensators 328,343
Static Excitation System 300

asy
Rate of Power Generation 236
Reactive Power Control for Loss Minimization
Static Load Model 32, 33
Static Var Compensator (SVC) 349
155
Reactors 7,315,321
En Steady State Security Assessment 361, 362

gi
Real power balance 200, 202
Rectangular coordinates 16, 17, 20, 22
Rectifier loads 353,355
Steady State Security Assessor 361,362

nee
Steady State Speed Regulation 184, 185
Steam Turbine Model 197, 198

Regional Load Dispatch Centre 368

Reheat Type Steam Turbine Model 198 r
Subsynchronous Resonance 337

ing
Supplementary Control 242,251,269
Synchronizing Coefficient 7
Restorative mode 6

s .
Synchronous Compensators 292,314
System Voltage and Reactive Power 293net
Security Evaluation 364,365
Security Function 363,364
T
Series - Series Controllers 347 Tap Changers 7

Series - Shunt Controllers 348 Tap-changing transformers 3 16

Series Controllers 346, 347 Tap-Staggering Method 317
Short Circuit Capacity 318, 319, 340 The Incremental Fuel Cost Characteristic 88
Short Term Hydro Thermal Scheduling 125 The Incremental Heat Rate Characteristic 88
Shunt Controller 347 The Linear Regulator Problem 222
Single Area System The Optimal Control Problem 222
203,205,225,254,259,269,275 The Polar Coordinates Method 19
Single Control Area 199,203,221 Three Area System 269,276
Sparsity of Network 22 Thyristor Controlled Resistor (TCBR) 349

Downloaded From : www.EasyEngineering.net

410 Index

Thyristor Controlled Series Capacitor (TCSC)

349
u
Unit Commitment 171,171,177,
Thyristor Controlled Series Reactor (TCSR) 349
Unit Commitment by Dynamic Programming
Thyristor Controlled Switched Series ReaCl\Jr
177
(TSSR) 349
Thyristor Switched Capacitor (TSC) 349 v
Tie-Line Bias Control 247,250
Voltage Dips 315,335
Transfer Function 186, 188, 190, 192
Voltage instability 350,352
Treatment of Generator Buses 19
Voltage Regulation 296,300,'306, 318

ww
Triangular Decompostion 23
Two Area System 266, 288
Voltage Regulator 297,302,305
Voltage Regulators 294,302, 309

w
Type 1 S system 309
Type 1 System 309
Type 2 System 310 .Ea
Voltage Sags 353,355
Voltage Stability 7,350

Type 3 System 310

Type 4 System 310
syE w
ngi
Weather Weighting Method 371

nee
r ing
. net

Downloaded From : www.EasyEngineering.net

ww
w.E
asy
En
gin
eer
ing
.ne
t

**Note : Other Websites/Blogs Owners we requested you, Please do not Copy

Dynamic Analysis of IEEE 14 Bus System: Experiment No: 05
No ratings yet
Dynamic Analysis of IEEE 14 Bus System: Experiment No: 05
8 pages
Power System Analysis G Shrinivasan PDF
0% (1)
Power System Analysis G Shrinivasan PDF
509 pages
Manual Flare Sim Expro
100% (4)
Manual Flare Sim Expro
69 pages
Revisiting The Fundamentals of Control Surface Reversal Including Nonlinear Effects
No ratings yet
Revisiting The Fundamentals of Control Surface Reversal Including Nonlinear Effects
8 pages
Power Sytem Books List
No ratings yet
Power Sytem Books List
13 pages
Electric Power Systems
No ratings yet
Electric Power Systems
562 pages
Optimal Power Flow
No ratings yet
Optimal Power Flow
42 pages
Electrical Machines Vol 2 4thed Ua Bakshi MV Bakshi PDF
100% (4)
Electrical Machines Vol 2 4thed Ua Bakshi MV Bakshi PDF
547 pages
Eee-Viii-Power System Operation and Control (10ee82) - Notes PDF
81% (21)
Eee-Viii-Power System Operation and Control (10ee82) - Notes PDF
130 pages
Riddle of the Seventh Stone, The
From Everand
Riddle of the Seventh Stone, The
Monideepa Sahu
No ratings yet
Code For Hamming Distance Transmission & Reception
No ratings yet
Code For Hamming Distance Transmission & Reception
8 pages
Copula-Based Pairs Trading Strategy
No ratings yet
Copula-Based Pairs Trading Strategy
16 pages
M2 Difference Eqns
100% (1)
M2 Difference Eqns
36 pages
Psoc Textbook by Sivanagaraju
No ratings yet
Psoc Textbook by Sivanagaraju
925 pages
PSP Notes For 5 Questions
100% (3)
PSP Notes For 5 Questions
115 pages
AaPower System Analysis
No ratings yet
AaPower System Analysis
79 pages
Power System Analysis: S. Ramar S. Kuruseelan
100% (1)
Power System Analysis: S. Ramar S. Kuruseelan
11 pages
Economic Dispatch 1
100% (1)
Economic Dispatch 1
76 pages
Realtime Control of Power Systems PDF
No ratings yet
Realtime Control of Power Systems PDF
1 page
Methods of Improving Transient Stability
No ratings yet
Methods of Improving Transient Stability
2 pages
(Leon K. Kirchmayer) Economic Operation of Power System
88% (8)
(Leon K. Kirchmayer) Economic Operation of Power System
136 pages
SUBSTATIONS
No ratings yet
SUBSTATIONS
19 pages
Eee-Vi-Switchgear & Protection (10ee62) - Notes PDF
No ratings yet
Eee-Vi-Switchgear & Protection (10ee62) - Notes PDF
92 pages
Psoc Notes PDF
No ratings yet
Psoc Notes PDF
110 pages
Power System Voltage Stability
100% (2)
Power System Voltage Stability
31 pages
Module 5notes - PSOC
No ratings yet
Module 5notes - PSOC
14 pages
Power System Transients
100% (2)
Power System Transients
62 pages
Power System Control and Automation
No ratings yet
Power System Control and Automation
23 pages
2009 Kamaraju - Electrical Power Distribution Systems-McGraw Hill
No ratings yet
2009 Kamaraju - Electrical Power Distribution Systems-McGraw Hill
278 pages
Power System Analysis: Economic Dispatch
No ratings yet
Power System Analysis: Economic Dispatch
39 pages
Power System Operation and Control PDF
50% (4)
Power System Operation and Control PDF
86 pages
Economic Operation of Power Systems
No ratings yet
Economic Operation of Power Systems
21 pages
Badri Ram Power System Protection PDF
0% (1)
Badri Ram Power System Protection PDF
578 pages
Power Systems by Bakshi PDF
No ratings yet
Power Systems by Bakshi PDF
486 pages
GATE Power Systems Book
100% (1)
GATE Power Systems Book
12 pages
Power Systems
No ratings yet
Power Systems
76 pages
Automatic Generation Control (AGC) or Automatic Load Frequency Control (ALFC)
100% (3)
Automatic Generation Control (AGC) or Automatic Load Frequency Control (ALFC)
36 pages
Optimal Power System Operation
No ratings yet
Optimal Power System Operation
11 pages
Power System Dynamics and Control - Video Course
No ratings yet
Power System Dynamics and Control - Video Course
2 pages
1 - 3a Load Flow Calculations - Theory PDF
No ratings yet
1 - 3a Load Flow Calculations - Theory PDF
4 pages
Economic Operation of Power System Notes
No ratings yet
Economic Operation of Power System Notes
15 pages
Double Freq Transient
100% (1)
Double Freq Transient
22 pages
Load and Load Curves Gupta PDF
100% (1)
Load and Load Curves Gupta PDF
20 pages
Power System Operation and Control
83% (6)
Power System Operation and Control
175 pages
Genco-Transco Prev Papers
No ratings yet
Genco-Transco Prev Papers
257 pages
Power System Operation and Control Question Bank PDF
No ratings yet
Power System Operation and Control Question Bank PDF
26 pages
(Lecture #1) Transients in Power System
33% (3)
(Lecture #1) Transients in Power System
11 pages
Electric Drives
100% (1)
Electric Drives
2 pages
Load Flow - Autumn 2021 - PPT - PDF
No ratings yet
Load Flow - Autumn 2021 - PPT - PDF
68 pages
Electrical Interview Question
No ratings yet
Electrical Interview Question
74 pages
Psoc Important Questions 2019-20
No ratings yet
Psoc Important Questions 2019-20
2 pages
Project PPT 1
No ratings yet
Project PPT 1
11 pages
PSOC Question Bank
No ratings yet
PSOC Question Bank
17 pages
Load Equalisation
No ratings yet
Load Equalisation
4 pages
Load Flow Analysis
No ratings yet
Load Flow Analysis
56 pages
228 Power System Protection
100% (1)
228 Power System Protection
2 pages
Lecture 6 Synchronous Machine Modeling
No ratings yet
Lecture 6 Synchronous Machine Modeling
57 pages
EE6603 - Psoc Lesson Plan (III B Sec) 2017
0% (1)
EE6603 - Psoc Lesson Plan (III B Sec) 2017
6 pages
L14 Three Phase Half Controlled Converter
No ratings yet
L14 Three Phase Half Controlled Converter
18 pages
Simulation of Some Power System, Control System and Power Electronics Case Studies Using Matlab and PowerWorld Simulator
From Everand
Simulation of Some Power System, Control System and Power Electronics Case Studies Using Matlab and PowerWorld Simulator
Dr. Hedaya Mahmood Alasooly
No ratings yet
Computer Aided Design of Electrical Machines
From Everand
Computer Aided Design of Electrical Machines
K.M. Vishnu Murthy
No ratings yet
Computer Methods in Power Systems Analysis with MATLAB
From Everand
Computer Methods in Power Systems Analysis with MATLAB
Sekhar Chandra P.
No ratings yet
Distribution of Electrical Power
From Everand
Distribution of Electrical Power
Dr. Hidaia Mahmood Alassouli
No ratings yet
Reactive Power Compensation
From Everand
Reactive Power Compensation
Dr. Hidaia Mahmood Alassouli
No ratings yet
Operation and Control in Power Systems PDF
100% (1)
Operation and Control in Power Systems PDF
421 pages
Plot-30 Calculation
No ratings yet
Plot-30 Calculation
57 pages
Open Channel Lab
No ratings yet
Open Channel Lab
27 pages
Principles of Communication Systems by Taub and Schilling - Text PDF
No ratings yet
Principles of Communication Systems by Taub and Schilling - Text PDF
119 pages
Measuring Horizontal Distatnces
No ratings yet
Measuring Horizontal Distatnces
76 pages
January 2010 QP - Unit 1 OCR Physics a-level
No ratings yet
January 2010 QP - Unit 1 OCR Physics a-level
15 pages
Maths SyllabusGrade VIII 2024
No ratings yet
Maths SyllabusGrade VIII 2024
10 pages
Descriptive & Inferential Statistics
No ratings yet
Descriptive & Inferential Statistics
6 pages
Sim Write Up
No ratings yet
Sim Write Up
13 pages
Lampl - 1992 - Saltation and Stasis
No ratings yet
Lampl - 1992 - Saltation and Stasis
4 pages
CAT 2008 Question Paper by Cracku
No ratings yet
CAT 2008 Question Paper by Cracku
67 pages
Reliability of The Go No Go Association Task
No ratings yet
Reliability of The Go No Go Association Task
13 pages
Lampiran 1 Hasil Output Spss Regression: Variables Entered/Removed
No ratings yet
Lampiran 1 Hasil Output Spss Regression: Variables Entered/Removed
5 pages
Ch10 Lecture7 Error Detection LIU Fall 2021
No ratings yet
Ch10 Lecture7 Error Detection LIU Fall 2021
44 pages
Total Marks:100 Mathematics Time Allowed: 3 Hrs
No ratings yet
Total Marks:100 Mathematics Time Allowed: 3 Hrs
2 pages
Does The Length of The Short Chain Branch Affect The Mechanical Properties of Linear Low Density Polyethylenes
No ratings yet
Does The Length of The Short Chain Branch Affect The Mechanical Properties of Linear Low Density Polyethylenes
19 pages
Shapiro-Wilk Test - Wikipedia
No ratings yet
Shapiro-Wilk Test - Wikipedia
3 pages
Column Foundation Connection
No ratings yet
Column Foundation Connection
3 pages
Traffic Sign Detection Using You Only Look Once Framework
No ratings yet
Traffic Sign Detection Using You Only Look Once Framework
9 pages
CH13 Data Representation A2 QP 2
No ratings yet
CH13 Data Representation A2 QP 2
19 pages
POLYNOMIALS in Class IX, You Have Studied Polynomials in One Variable and
No ratings yet
POLYNOMIALS in Class IX, You Have Studied Polynomials in One Variable and
21 pages
A Dynamic Crashing Method For Project Management Using Simulation-Based Optimization
No ratings yet
A Dynamic Crashing Method For Project Management Using Simulation-Based Optimization
7 pages
5989 6639en
No ratings yet
5989 6639en
85 pages
On The Incompatibility of Accuracy and Equal Opportunity
No ratings yet
On The Incompatibility of Accuracy and Equal Opportunity
30 pages
IGCSE Files Telegram Channel: Thursday 08 October 2020
No ratings yet
IGCSE Files Telegram Channel: Thursday 08 October 2020
32 pages
Cat Paper 2
No ratings yet
Cat Paper 2
18 pages