2004 IEEE International Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004) April 2004 Hong

Kong

Evolutionary Programming Based Economic

Power Dispatch Solutions With Independent
Power Producers
Narayana Prasad Padhy Latha Kumari

of the government bodies. This new environment creates

Abstract-In this paper, economic dispatch solutions using significant challenges. There has been no technological
evolutionary programming has been presented with the breakthrough or in other words successful re-regulated models
presence of independent power producers(1PPs). The are not in use in the Indian power industry [9].
performance of the proposed algorithm has been tested for The forces behind electric power sector deregulation taking
synthetic IEEE-30 bus system having both smooth and place worldwide has either been political reform, regulatory
non-smooth generator cost characteristics with both failure, high tariffs, managerial insufficiency or global
bilateral and multilateral transactions. It has also been economic crises. The reason for deregulation is different for
observed that the proposed algorithm can be applied to different countries. Many countries made the changes as a
larger systems and do not suffer with mathematical result of the failure of the state to adequately manage
complexity. electricity companies or the force behind this has been the lack
of public resources to finance the required investment for the
Index Terms- Economic Dispatch, Evolutionary development.
Programming and IPPs.
1.1 PRIVATE POWER DEVELOPMENT IN INDIA
I. INTRODUCTION
HE organization of the electric power sector in the Energy development in India commenced towards the end
world has been changing dramatically to allow for of the nineteenth century with the commissioning of an electric
competition among private sector generators and to create supply company in Assam in 1897. This was followed by the
market condition in the power sector, seen as necessary launching of a hydro power station at Sivasamudram in
conditions for increasing the efficiency of electric energy Kamataka in 1902. During the pre independence era, the
production and distribution, offering a lower price, higher power generation and its distribution was dominated by the
quality and secure products. private sector and restricted to urban areas. After
In both privatized and deregulated environment, all power independence (1947), the govemment of India felt the need for
transactions are made based on price rather than cost and also
restructuring the energy sector with a view to rationalize
in a deregulated energy marketplace, participants are interested
growth throughout the country. The electricity (supply) Act
in maximizing their own profits, regardless of the system-wide
profits. It is perceived that the competition will reduce the was enacted in 1948, and the industry was nationalized with
price of electricity for retail customers, however, the key issue state electricity boards leading the effort in each state. In the
for participants is related with the price definition to remain quest for enhanced capacity, the country adapted a blend of
competitive [ 121. thermal, hydro, and nuclear sources. In 1975, the government
Deregulation and market competition was introduced in decided to b h e r augment the power generation program with
most of the countries, for e.g. Norway, Chile, Peru, Bolivia, the creation of two public sector corporations owned wholly by
England, U.S. and partially initiated now in India also. Many the federal government, namely National Thermal Power
developing countries have chosen to deregulate their power Corporation (NTPC) and National Hydro Power Corporation
sectors, creating conditions for competition in generation and ( N H W [81.
increasing participation by the private sector. In doing so, they Since achieving independence in 1947, India has multiplied
have not only left in past traditional vertically integrated its power generating capacity more than 50 times. Increases in
publicly owned utilities, but are also leaving back to capacity, however, have been modest as compared to the
government centrally determined generation-transmission increase in demand resulting from India’s rapid
extension plans. Private investors are taking independent industrialization and its growing population. The country has
decisions according to their own assessments rather than those one of the lowest per capita electricity consumption in the
world and is currently among the world’s leaders in its need
Narayana Prasad Padhy IS currently with the Department of Electncal for power generation.
Engineenng, Indian Institute of Technology, Roorkee - 247667, India With the objective of improving the power situation in the
(nppeefee6iitr cmet in ) country, the Government of India (GOI), decided in 1991 to

172
2004 IEEE Intemational Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004) April 2004 Hong Kong

add more than 14 GW of new electric capacity by the year step must balance the total extraction, in respect of each
2000. Acute shortage of funds required to finance the individual bilateral and multilateral transaction.
procurement of incremental capacity and the inefficient In a deregulated environment, the number of bilateral
management of existing power plants by the state electricity transactions will grow rapidly and therefore, new methods and
boards (SEBs) forced the government to liberalize the power tools will be required to help system operators to evaluate their
sector and open it to private and foreign investors pursuant to impacts on the operation of the systems. These impacts can be
important regulatory and policy changes. Since then, numerous measured in terms of changes in system performance measures
reforms in existing rules and new policies have been such as reliability, transmission losses, production costs, etc.
announced to facilitate private investment in the power sector The power injections associated with different bilateral
PI. transactions can influence the loading on transmission
A large number of transmission transactions are expected to facilities. As a result, power flow on lines and transmission
take place with introduction of privatization of power system. interfaces can increase or decrease depending on the system
These transaction need to be evaluated ahead of their operating conditions, transaction size, and the direction of
scheduling time to check their feasibility with regard to the power transfer and the number of transactions considered [ 181.
system conditions at the time of scheduling. Transmission So the utility is required to take an optimal decision in
system operator would have to honor and execute only selecting the location of IPPs(invo1ved in transaction) and
transactions as far as the system design and system operating the magnitude of feasible transactions such that the
conditions permit. economic dispatch solutions leads to either further optimized
Transactions will be classified into feasible and unfeasible or unchanged.
transaction. Feasible transactions can be accommodated
without violating the system economic dispatch and 11 PROBLEM FORMULATION
transmission network constraints. Unfeasible transactions
violate transmission network constraints and cannot be The ED problem studied in this report can be described
accommodated fully without altering the system economic as an optimization process with the following objective
dispatch. In order to accommodate an unfeasible transaction, function and constraints:
the system economic dispatch may be altered and there will be R
penalty (additional fuel cost) associated with the transaction
~41.
subject to power balance constraint:
1.2BILATERAL AND MULTILATERAL TRANSACTIONS N

Many large buyers of power such as municipalities, large

industries, large housing estates and other complexes, may not and generating capability constraints:
wish to buy their electricity through the pool, and may prefer
to make their own power purchase agreements with selected Pgimin for = 1,......,N
5 P gi 5 Pgimax (3)
suppliers. There could be the following motivation behind it.
Pool prices are described above are spot prices, hence they Real and reactive powers of the system can be determined as
fluctuate throughout the day and between days and seasons.
Some buyers, especially those who cannot, or do not wish to
adjust their consumption at short notice to take advantage of
price fluctuations, may prefer to ignore the pool price and i=1,2,...., n
establish stable fixed price contracts with their suppliers. This
provides motivation for bilateral and multilateral contracts
[151.
These transactions could be of two types, either involving
just one buyer-seller pair (even if their physical injection and
utilization points are multiple) and known as bilateral
i=1,2 ,....,n
transactions, or bringing together a multiple of buyers and (5)
sellers who group themselves together to enter into a Subject to following operating constraints :
multilateral transaction. From the commercial point of view a
power broker may put a multilateral transaction, or indeed a
pil& qvi1 5lY.l" (6)

bilateral one, together, and it may be a short or long term deal. (7)
From the system operator's point of view what is required is a
schedule of how much power is to be injected or extracted by
group members at each point in the grid during each time Qgi,min 5 Qgi 5 Qg;,max (9)
period of the day. Bilateral and multilateral transactions must
Both the bilateral and multilateral transactions
maintain their own intemal power balances, that is,
injection power must be equal to the load power.
transmission losses aside, the total injection during each time

173
2004 IEEE International Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004) April 2004 Hong Kong

111 EVOLUTIONARY
PROGRAMMING n-l

Where q=p * - Pgimin)

(Fi / fmin)*(Pgimax (1 1)
i=l
Although evolutionary computation techniques have proved
useful in general function optimization, they appear particular Where p is a scaling factor
apt for addressing non-linearly constrained optimization Fi is the fuel cost of the ith generator and
problems. Constrained optimization problems present the fminis the minimum value of the fuel cost in the generation
differences of potentially non-convex or even disjoint feasible The created off - spring vector must satisfy the
regions. Classical linear programming and non-linear minimum and maximum generation limits and other inequality
programming methods are often either unsuitable or limits of the units variables and finally line flow constraints.
impractical when applied to these constrained problems.
Difficulties arise because either the amount of computation I11 NUMERICAL RESULTS
required quickly becomes unmanageable as the size of the
problem increases or the constraints violate the required
assumptions, eg., differentiability or convexity. Unfortunately, The effectiveness of proposed approach has been illustrated
the real world often poses such problems [l 11. Evolutionary using the IEEE 30 bus test systems.
computation techniques can be constructed to cope effectively
with the above difficulties. There are, however, no well-
established guidelines on how to deal with unfeasible
solutions.
Evolutionary programming technique may provide more
rapid and robust convergence on many function optimization
problem. The rationale, behind the scheme is that an
evolutionary optimization algorithm is unlikely to be the best
optimization procedure for any specific function in terms of
efficiency convergence rate, solution accuracy, etc.; its
robustness comes at the sacrifice of domain specificity.
In this paper, a novel evolutionary programming (EP)
algorithm, is proposed to solve the ED problems for units with
non-smooth fuel cost units. The stochastic mechanics, which
combine offspring creation based on the performance of
current trial solutions and competition and selection based on
successive generations, from a considerably robust scheme for
large scale real valued combinational optimization. The
weaknesses of the algorithms mentioned above are
circumvented. The proposed EP approach is capable of not
only solving the ED with any type of fuel cost functions,
analytical or empirical curves, but also obtaining the global or Fig. 1. The IEEE 30-bus system
near global minimum solution considering transmission losses
within reasonable execution time. IEEE 30-Bus Svstem
Evolutionary programming is a probabilistic search Cases studies for IEEE 30 bus system [22] shown in Fig. 1.
technique, which generates the initial parent vectors distributed have been carried out to determine the effectiveness of the
uniformly in intervals within the limits and obtains global proposed model. The system has 6 generation, 4 LTC
optimum solution over number of iterations. The main stages transformers, and 41 transmission lines.
of this technique are initialization, creation of off - spring Five sets of generator cost curves were used to illustrate
vectors by mutation and competition and selection of best the robustness of the proposed algorithm. In the case 1, all the
vectors to evaluate best fitness solution. cost curves are considered to be quadratic [16], where as in the
A. Initialisation case 2, 3, 4 and 5 all the cost curves are quadratic with
exponential components [3], quadratic with sine components
The initial population (k number of parent vectors) is
[16], some of the cost curves are replaced with piece-wise
generated after satisfying the system constraints. The elements
quadratics and piece-wise linear [161 respectively. The cost
of parent vectors (Pgi)are the real power outputs of generating
coefficients of the above case studies are given in Table 1, 2,
units distributed uniformly between their minimum and
3, 4 and 5 respectively. the Therefore in cases 2, 3, 4 and 5 ,
maximum limits. Slack bus generator vector is calculated using
there are many local optimal solutions for the dispatch
Newton Raphson method for the above generations.
problem and as a result the conventional algorithm cannot
B. Mutation determine the global optimal solution. The problem is
An off - spring vector P,i’ is created from each parent vector therefore well suitable for validating the developed algorithm.
by adding gaussian random variable with zero mean and The EP based algorithms was implemented using the
standard deviation q, denoted as N(0, q2). MATLAB 6.1 Software. Further the performance of the
Pg,’=P,i + N(0, 0i2) for i=1,2,...,n-1 (10)

174
2004 IEEE International Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004) April 2004 Hong Kong

proposed model has also been tested by introducing the

following bilateral and multilateral transactions.
TABLE 4: GENERATORS WITH P-W QUADRATIC COST
T1: Bilateral Transaction of lOMW between bus 5 and 29. COEFFICIENTS
T t : Bilateral Transaction of lOMW between bus 29 and 5.
T3: Multilateral Transaction: with injection of 10 MW and 10
Bus MW MW
MW in the buses 5 and 8 respectively and load of 5 MW, 15
COST COEFRCIENTS
MW in the buses 20 and 29 respectively.
T4: Multilateral Transaction: with injection of 10 MW and 10 a b C
MW in the buses 20 and 29 respectively and load of 5 MW,
15 MW in the buses 5 and 8 respectively. 1 50 140 55.0 0.70 0.0050
140 200 82.5 1.05 0.0075
Solution to economic dispatch using the evolutionary
programming, with and without transactions(T1, T2, T3 and 2 20 55 40.0 0.30 0.0100
T4) have been presented in Table 6,7,8,9 and 10. It is now 55 80 80.0 0.60 0.0200
very significant to observe that the total generation cost of the
overall system reduces due to few transaction and similarly
increase due to few transactions. So the utility can optimal TABLE 5: GENERATORS WITH P-W LINEAR COST
decide how and where to encourage the IPPs participation COEFFICIENTS
through transactions using the proposed algorithm.

Bus MW MW Cost coefficients

TABLE 1 : GENERATOR DATA WITH QUADRATIC COST
COEFFICIENTS

B Pg. Pg. Qg. Qg. Cost Coefficients

U min max min max
S a I b l c

TABLE 6: SOLUTIONSOF CASE 1

TABLE 2: GENERATORS WITH EXPONENTIAL COST

COEFFICIENTS

I Bus I a
I b l c l d l e l
~

1 0.0409 -0.0555 0.0649 0.0002 0.0285

2 0.0254 -0.0604 0.0563 0.0005 0.0333
5 0.0425 -0.0509 0.0458 0.0001 0.0800
8 0.0532 -0.0355 0.0338 0.0002 0.0200
11 I 0.0425 I -0.0509 I 0.0458 I 0.0001 I 0.0800
13 I 0.0613 I -0.0555 I 0.0515 I 0.0001 I 0.0666

FCj(P@)=aj +bj Pi+ cj Pgi2+dj expej.pgj

TABLE 3: GENERATORS WITH SINUSOIDAL COST

COEFFICIENTS

1
I Bus I

25.00 2.50

FCJ(P,)= aJ+bJ P, + c, P,
Cost coefficients

0.0100

*+I
40.00

dJsin (e(P,
0.0980

Inln - P,))I
I

175
2004 IEEE Intemational Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004)April 2004 Hong Kong

TABLE 7: SOLUTIONSOF CASE 2

Generation TABLE 10: SOLUTIONS OF CASE 5

Schedule (MW)
introducing introducing Bus Generati Generation Generation
using bilateral multilateral No. on Schedule (MW) Schedule (MW)
transaction transaction Schedule introducing introducing
(MW) bilateral multilateral
using EP transaction transaction
T1 T2 T3 T4

I I I I I

1 I 139.41 I 137.41 I 139.41 I 139.69 I 139.91

Total 841.12 888.71 829.03 862.01 826.72

cost
($/hr) 13 I 26.355 I 27.668 I 26.082 I 26.738 I 25.945
Total I 454.74 I 466.39 I 452.03 I 458.74 I 451.02
cost
TABLE 8: SOLUTIONS OF CASE 3 ($/hr)
Bus Generati Generation Generation
No. on Schedule (MW) Schedule (MW)
Schedule introducing introducing
(MW) bilateral multilateral IV CONCLUSION
Privatization in India and Deregulation in West in the
electric power industry is expected to increase the benefits
associated with the operation of interconnected power system.
In this report, we introduced bilateral transactions concepts
and multilateral transactions to simulate the behavior of
participants in deregulated energy markets. The present
algorithm is able to solve the economic dispatch problem
under deregulated environment. The performance of the
developed algorithm has been demonstrated by its application
to the synthetic IEEE 30-bus test system. An evolutionary
Total I 842.87 I 865.5 I 837.04 I 852.44 I 835.79 programming approach has been proposed and demonstrated
cost to be a powerful optimization process in solving the highly
($/hr)
non-linear and discontinuous economic dispatch problem
where the classical Lagrange based algorithms cannot be
directly applied. Generators with input/output cost
characteristic curves such as exponential-wavelquadratic
TABLE 9: SOLUTIONS OF CASE 4
curve, sine-wavelquadratic curve, piece-wise quadratic and
piece-wise linear curve have been used providing a non-
Bus Generati Generation Generation
No. on Schedule (MW) Schedule (MW) convex solution surface. The algorithm has accurately and
Schedule introducing introducing reliably converged to the global optimum solution in each
(MW) bilateral multilateral case.
transaction transaction
T1 T2 T3 T4 V REFERENCES
1. 0.Alsac & B.Stott, ”Optimal load flow with steady state
security”, IEEE Trans. On Power Apparatus and systems, Vol.
PAS-93,1974, pp.745-75I.
2. Goldberg D.E,”Genetic Algorithm in Search, Optimization, and
Machine Learning”, Addision-Wesley, 1989.
3. Ahmed Farag, Samir AI-Baiyat, T.C.Cheng, ”Economic load
dispatch multiobjective optimization procedures using linear
programming techniques”, IEEE Transactions on Power
Systems, Vol. 10, No.2, pp.731-738, May1995.
4. K.P.Wong and Y.W.Wong,”Genetic and genetidsimulated-
annealing approaches to economic dispatch”, IEE Proc.-

176
2004 IEEE International Conference on Electric Utility Deregulation, Restructuring and Power Technologies (DRPT2004) April 2004 Hong Kong

Gener.Transm. Distrib., Vol. 141, No.5, September 23. Optimization Toolbox for use with MATLAB”, users guide
1994,pp.507-5 IO. Version 2, PDF-Document, www.mathworks.com.
5. Po-Hung Chen and Hong-Chan Chang,” Large-scale Economic
Dispatch by genetic algorithm”, IEEE Transactions on Power
Systems, Vol.lO, No.4,November 1995,pp.1919-1921. VI BIOGRAPHIES
6. Sheble G.B. and Kristin Brittig, “Refined genetic algorithm Narayana Prasad Padhy received the Degree in Electrical
economic dispatch example”, IEEE Trans. on Power Systems, Engineering and Masters Degree in Power
Vol. 10, No.1, pp. 138-140, Feb. 1995. Systems Engineering with Distinction in 1990
7. J.M.William, “Open Transmission Access”, IEEE Power and 1993, respectively. He received the Ph.D.,
Engineering Review, Vo1.16, No.5, May 1996, pp.17-18. Degree in Electrical Engineering from Anna
8. M.Tahiliani, “Private Power Development in India”, IEEE University, Chennai, India in 1997. Then he
Power Engineering Review, Vol. 16, No.2, Feb 1996, pp. 16. joined Birla Institute of Technology &
9. Hugh Rudnick, “Planning in a deregulated Environment in Science(B1TS) as an Assistant Professor,
Developing Countries: Bolivia, Chile, Peru”, IEEE Power Electrical Engineering Department in 1997.
Engineering Review, Vol. 16, No. 7, July 1996, pp.18-19.
IO. Hong- Tzer Yang, Pai-Chung Yang, Ching-Lien Huang, Currently, he is Assistant Professor in the Department of Electrical
“Evolutionary Programming Based Economic dispatch for units Engineering, Indian Institute of Technology, Roorkee. He taught
with non-smooth fuel cost functions”, IEEE Transactions on course in Basic Electrical Engineering, Power Systems and Artificial
Power Systems, Vol. 1l,No.l,pp. 112-115, February, 1996. Intelligence. His field of interest is Power System Privatization,
11. Jong-Hwan Kim, and Hyun Myung, “Evolutionary programming Restructuring and Deregulation, Artificial Intelligence Applications
techniques for constrained optimization problems”, IEEE to Power System Operation and Optimization Problems, Unit
Transactions on Evolutionary Computation, Vol. 1, No.2” pp. Commitment, Power System Wheeling and FACTS.
129- 13O,.J~ly1997.
12. R.W. Ferrero, S.M. Shahidehpour and V.C. Ramesh,
“Transaction analysis in deregulated power systems using game
theory”, IEEE Transactions on power systems, Vol. 12, No. 3,
August 1997,pp 1340.
13. Kwang. Y. Lee and Frank F.Yang,”Optimal reactive power
planning using evolutionary algorithms: A comparative study for
evolutionary programming,evolutionary strategy, genetic
algorithm, and linear programming”, IEEE Transactions on
Power Systems, Vol. 13, No. 1, February 1998,pp.101-103.
14. John W.M.Cheng, Franscisco D. Galiana(F), Donald
T.McGillis,”Studies of bilateral contracts with respect to steady-
state security in a deregulated environment”, IEEE Transactions
on power systems, Vol. 13, No.3,August 1998,pp.I020.
15. R.S.Fang , A.K. David, “Transmission Congestion in an
Electricity Market”, IEEE Transmission on Power Systems,
Hong Kong Polytechnic University Kowloon, Hong Kong,
Vol. 14,NO.3, pp.877-878, August 1999.
16. Jason Yuryevich, Kit Po Wong, “Evolutionary programming
based optimal power flow algorithm”, IEEE Transactions on
Power Systems, Vol. 14, No. 4, November 1999,pp.1245-1249.

17. G.Hamoud, ”Feasibility Assessment of Simultaneous Bilateral

Transactions in a Deregulated Environment”, IEEE Transactions
on power systems, Vol. 15, No. 1, February 2000,pp.22.
18. Franscisco D. Galiana, ”Allocation of transmission losses to
bilateral contracts in a competitive environment”, IEEE
Transactions on power systems, Vol. 15, No.1, February
2000,pp. 143.
19. Whei-Min Lin,Fu-Sheung, and Ming-Tong Tsay, ” Nonconvex
economic dispatch by integrated artificial intelligence”, IEEE
Transactions on Power Systems, vol. 16, No. 2, May 2001,pp.
307.
20. Anastasios G. Bakirtzis, Christoforos E. Zoumas, ”Optimal
Power Flow by enhanced genetic algorithm”, IEEE Transactions
on Power Systems, Vol. 17, No.2,May 2002,pp.230-23 1.
2 1. Pathom Attaviriyanupap, Hiroyuki Kita, Eiichi Tanaka, and Jun
Hasegawa, “A Hybrid EP and SQP for Dynamic Economic
’ Dispatch With Non Smooth Fuel Cost Function”, ”, IEEE
Transactions on Power Systems, Vol. 17, No. 2, May
2002,pp.411-4 13.
22. Power Systems Test Case, The University of Washington
Archive, http:i/~~w.ee.washington .edu/researcli/pstca/,2000.

177