World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:4, 2012

Generation Expansion Planning Strategies on

Power System: A Review
V. Phupha, T. Lantharthong, N. Rugthaicharoencheep

Abstract—The problem of generation expansion planning (GEP) Choices are left to each player, which is made according to
has been extensively studied for many years. This paper presents their own estimation on the prospect for the price level and
three topics in GEP as follow: statistical model, models for market environment in the future. Therefore, it is strongly
generation expansion, and expansion problem. In the topic of required to develop a new practical expansion planning
statistical model, the main stages of the statistical modeling are methodology which is applicable to the changed electric power
briefly explained. Some works on models for GEP are reviewed in industry environment where individual generation firms seek to
the topic of models for generation expansion. Finally for the topic of
expansion problem, the major issues in the development of a long-
maximize the returns on their expansion planning decision. In this
term expansion plan are summarized. section, new generation expansion planning formulations in
competitive markets are explored [5].
Keywords—Generation expansion planning, strategies, power We can represent the optimum status of the set of power plants,
International Science Index, Industrial and Manufacturing Engineering Vol:6, No:4, 2012 waset.org/Publication/10548

system by the equality between the operation marginal cost and the
expansion marginal cost [6], [7].
I. INTRODUCTION This condition can be understood if we imagine the two
possibilities to meet one part of market in the system:
G ENERATION expansion planning (GEP) is one of the
most important decision-marking activities in electric
utilities. Least-cost GEP is to determine the minimum-cost
1) Through the existing means, incurring in operational costs
increase.
2) Through incorporating a new power plant to the system
capacity addition plan (the type and number of candidate
and/or reinforce the transfer ability among the subsystems.
plants) that meets forecasted demand within a prespecified
reliability criterion over a planning horizon [1]. The GEP
II. STATISTICAL MODEL
problem has been one of the most studied problems. GEP is a
challenge topic for several reasons. The first reason since there Because of the importance of the demand and price in
is uncertainty associated with the input data. The second one planning procedure, only these two data, i.e. the demand of the
since difficulty arises as a result of a need to consider several whole system and the corresponding market energy price,
conflicting objectives simultaneously. With the aid of a from the historical information of the market are used in this
multiobjective model, decision makers may grasp the statistical study. These information are hourly available in any
conflicting nature and the tradeoffs among the different of the real life market in the world. The main stages of the
statistical modeling are as follows [8].
objectives in order to select satisfactory compromise solutions
for the GEP problem. In the past, there have been many attempts A. Data Justification
to deal with GEP problem [2]. Some of them used deterministic After choosing a base year, the demand and prices are
criteria whereas other incorporated analysis of uncertainties. justified according to annual demand growth and discount
These uncertainties could be either technical, such as hydrological rates. By this, the whole system load duration curve can be
conditions and generators stoppage, or economical, like fuel constructed and with a good approximation, it can be assumed
prices and interest rates [3]. There are difficulties, however, in that the shape of this curve remains unchanged.
taking into account too many aspects of the problem because of
the overwhelming complexity that rapidly arises. The classical B. Price Grouping
formulation of the least-cost GEP is not a suitable approach to After choosing the equivalent load levels for the load
modeling the behavior of generation firm, seeking to maximize duration curve of the system, the corresponding prices data are
profits, in competitive environments. Supposed that the revenue is devoted to the suitable groups. Then, for each demand level,
based on the marginal price, some generation firms will choose to the maximum price, the minimum price and the median of
retain higher priced units to continue to set marginal price high prices as the most realistic price for that load level are
and increase their revenue and profit by higher marginal price determined.
while others replace the higher priced units with efficient and C. Generalization
cheap ones and increase revenue by higher utilization factor [4]. The last stage in the statistical study is data generalization
to the planning horizon. For this, It is assumed that the shape
Vallop Phupha is with the Department of Industrial Engineering, Faculty of
Engineering, Rajamangala University of Technology Phra Nakhon, Bangkok,
of the load duration curve remain unchanged. For each of the
10800, Thailand. (phone: 662 913-2424; fax: 662 913-2424; e-mail: planning horizon years, after applying the annual load growth
v.phupha@hotmail.com). rate to the load levels, the new load levels are formed and then
Thong Lantharthong is with the Department of Electrical Engineering, Faculty of
Engineering, Rajamangala University of Technology Phra Nakhon, Bangkok, the new duration time for each load level is determined by
10800, Thailand. (phone: 662 913-2424 ext. 150; fax: 662 913-2424 ext. 151; e-mail: rearranging the historical data.
thong.l@rmutp.ac.th)
Nattachote Rugthaicharoencheep is with the Department of Electrical Engineering,
Faculty of Engineering, Rajamangala University of Technology Phra Nakhon, Bangkok,
10800, Thailand. (phone: 662 913-2424 ext. 150; fax: 662 913-2424 ext. 151; e-mail:
nattachote@ieee.org).

International Scholarly and Scientific Research & Innovation 6(4) 2012 784 scholar.waset.org/1999.8/10548
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:4, 2012

D. Basic Assumptions The model in [17], objectives are the minimization of

GEP problems in competitive market environments are investment, operation and transmission costs, environment
basically modeled as games among competing generation impact, imports of fuel and fuel prices risks of the whole
firms [9], [10]. Electricity industry environment in which the system. Multiobjective linear programming and analytical
generation expansion games should be solved has been so hierarchy process are made use of to solve this problem. The
changing that it is very difficult to incorporate practical model of Antunes et al.[19] is presented to provide decision
considerations of electric power industries in the formulation support in the evaluation of power generation capacity
of GEP games. In order to properly address the generation expansion policies. The objective functions are the total
expansion games in competitive markets, therefore, we make expansion cost, the environmental impact associated with the
some basic assumptions about the market environments and installed power capacity and the environmental impact
game variables. In our proposed games, electric markets are associated with the energy output. Ahmed et al. [18] use a
modeled by uniform pricing auctions. We assume that there scenario tree approach to model the evolution of uncertain
exists a single gross power pool where buying and selling of demand and cost parameters, and fixed-charge cost function to
energy among firms happen in the form of auctions. The model the economies of scale in expansion costs. Several
market price is based on the operating cost of the marginal works have focused on government regulations on energy
generation firm. Since we do not consider the bilateral markets.
International Science Index, Industrial and Manufacturing Engineering Vol:6, No:4, 2012 waset.org/Publication/10548

contract between firms in this model, the profit of every The study in [20] provides a framework for representing
generation firm can be simplified revenues from energy selected price-oriented government regulations in
market modeled by auctions and construction and/or operation mathematical programming model of a market. The work of
costs of plants [11]. [21] compares effects of R&D and demand subsidies on the
After defining the quality criteria of meeting the market, the future costs of purely organic photovoltaics (PV) which is not
optimum status of system expansion is given by minimizing currently commercially available. They combine an expert
the total cost of meeting the market, which includes elicitation and a manufacturing cost model to compare the
investment costs, maintenance and operational costs, expected outcomes of policy choices over various scenarios.
costs with thermal plant fuel and the cost of non-meeting a
market share (energy deficit cost) [12], [13]. Naturally, some IV. EXPANSION PLANNING PROBLEM
restrictions associated to the hydrothermal coordination shall This section examines the major issues in the development
be considered, as meeting the market, hydrological balance, of a long-term expansion plan for the generating system [22].
interconnection ability, assured energy balance, among others.
A. Demand
III. MODELS FOR GENERATION EXPANSION The forecast of electrical demand is clearly one of the most
important components of a generating system analysis. The
In this section, some works on models for GEP are
forecast typically must be for power (kW), energy (kW-h) and
reviewed, government incentive policy and inverse
load variation for time intervals within a year, such as a month
optimization. [14] and [15] apply the Cournot model of
or season, for all years of the study. If a great deal of effort is
oligopoly to model GEP. The model of [14] incorporates plant
to be devoted to analyzing the alternative expansion
capacity limitations and energy balance constraint in
possibilities, the demand forecast should also receive a
competitive environments dominated by auction markets.
significant effort. There are two distinct types of uncertainty in
They present an analytical formulation of the generation
demand forecasting. First, there is the uncertainty that results
planning process involving decisions on new plant
from the randomness of the load at any time because of, for
construction at a single point in time with multiple technology
example, weather conditions. This type of uncertainty is, of
options available. In [15], three capacity expansion models are
course, a major concern for the load dispatcher. The other type
analyzed in the context of a restructured electricity industry.
of uncertainty is associated with the estimate of future
The first model assumes a perfect, competitive equilibrium.
demand, i.e. the estimate may be too high or too low.
The second model (open loop Cournot game) extends the
Underestimating future demand can create serious difficulties
Cournot model to include investments in new generation
because service dates for new facilities can seldom be
capacities. The third model (closed-loop Cournot game)
advanced appreciably. The result may be a generating system
separates the investment and sales decision with investment in
with low reliability and the inability to serve some portion of
the first stage and sale in the second stage. The study of [16]
demand.
presents 3-Tier game theoretic model to obtain multi-period,
Overestimating the demand is also undesirable because
multi-player equilibrium capacity expansion plans. This model
excess generating equipment imposes increased costs on the
provides an excellent means of conducting analysis of
system. Service dates for new facilities under construction can
sensitivity of the expansion and the corresponding market
be delayed if load growth has been overestimated, but such
power indices with respect to several restructuring policy
delays can be very expensive.
related parameters, such as varying levels of price caps,
arbitrage, pricing and settlement mechanisms, demand side
B. Technology Options
bidding, and capacity payment. The works in [17], [18]
propose a multiobjective model for generation expansion Various technologies are currently available as candidates
planning (MGEP). for expanding electrical generating systems. Each has a unique
set of characteristics that must be considered from a system

International Scholarly and Scientific Research & Innovation 6(4) 2012 785 scholar.waset.org/1999.8/10548
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:4, 2012

viewpoint to determine the mix of future additions that E. Constraints

provides the best outcome for the stated objectives for Mathematically, the problem referring to power plant
expansion. In addition to existing technologies, long-term expansion is formulated as a problem of integer linear
studies of generation expansion must consider whether programming difficult to solve and which present some
advanced technologies will become available and, if so, what particularities, such as:
their costs and characteristics will be. Power generation 1). non-convex solution region, which allows existing several
technologies may be classified into existing major options and solutions.
potential future options. 2). Combinatorial nature of expansion process, which leads
C. Economic Evaluation to the combinatorial explosion phenomena related to
generating alternatives, causing increased computing time.
A fundamental aspect of any economic evaluation is the
Considering that for problems of discrete nature, Applications
time element, since implementation time and economic
as optimization method is being greatly surveyed and
lifetime of a generating unit require a certain number of years:
disseminated, mainly by obtained result quality. Therefore, a
a particularly large number in the case of nuclear units, where
new algorithm was developed in this work, based on
the overall period to be considered usually varies between
optimization technique, in order to establish an optimal
30 and 45 years (typical values are 10 years for
strategy of generation expansion [23].
International Science Index, Industrial and Manufacturing Engineering Vol:6, No:4, 2012 waset.org/Publication/10548

implementation and 30 years for economic lifetime).

In a general context, the solution of the expansion problem
A key concept in understanding the basic principles of
is the schedule of plant additions and network development
economic evaluation is the time value of money, i.e. how
over a certain period of time which yields the optimum
streams of costs or incomes (or alternatively of produced
benefits while satisfying the projected electricity demand with
electricity since it generates an income) occurring through
a certain margin of reserve and respecting certain foreseeable
time can be compared on an equivalent basis.
constraints.
The relationship between time and money is affected by:
In other words, the expansion program must include [24]:
1) Inflation (or deflation) which changes the buying power
1) The year-by-year capacity additions needed to satisfy the
of money.
projected electricity demand with a satisfactory level of
2) The value given to possession of money now rather than
reliability with due regard to the characteristics of generating
later, since the former allows this amount of money to be
units in the existing system.
invested for an interval of time to earn a real return (i.e. in
2) The timely reinforcement of the transmission system so
addition to inflation).
that the proposed network is capable of meeting power flow
Alternatively, raising capital through a financial market
requirements under any foreseeable condition with due regard
implies the payment of a cost of capital for years to come
to load flows, power station siting, circuit and switchgear
(again independent of inflation). The annual factor that
ratings, and transient stability limits.
accounts for the time value of money independently of
Definition of the benefits to be optimized is of paramount
inflation is called the real discount rate (or real present worth
importance since it leads to selection of the economic criteria
rate).
to be used for evaluating and comparing all alternative
D.Reliability expansion policies for the power system.
The objective of electric system planning was stated as On the other hand, definition of the constraints is perhaps
adequately meeting the demand for electrical power at the more complex since this requires the resolution of important
minimum cost. Adequately meeting the demand can be issues, such as [25]:
interpreted in various ways with major implications for the • Adequate reserve margins or level of reliability.
generation planning effort. Typically, a technical constraint is • Required quality of service in terms of continuity of
used as the minimum acceptable level of generating system supply, frequency and voltage.
performance, or an economic criterion is introduced in an • Availability of resources (manpower, fuel, funds).
attempt to include the generating system reliability • Technical considerations.
considerations directly in the determination of minimum cost. • Infrastructure needs.
The generation planner must design the future generating • Environmental consideration.
system to be responsive to such problems as: • The country's policies concerning new units for
• Random breakdowns of generating equipment (forced electricity generation.
outages.
• Variations in demand to be met by the generating system V. CONCLUSION
(including random variations). The aim of GEP requires minimum-cost capacity addition
• Variations in hydraulic conditions which affect plan that meets the forecasted demand within a reliability
hydroelectric capacity and energy available to the criterion. This work presents an overview of some issues in
generating system. GEP. It is clear from the previous studies that GEP is one of
• Scheduled maintenance of generating equipment and the most important decision-marking activities in electric
refuelling of nuclear units. utilities.
• Changes in anticipated new capacity scheduled to come
on line, e.g. delays or cancellations because of financial
and other constraints.

International Scholarly and Scientific Research & Innovation 6(4) 2012 786 scholar.waset.org/1999.8/10548
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:4, 2012

ACKNOWLEDGMENT [21] L. Wang, “Cutting plane algorithms for the inverse mixed integer linear
The authors would like to express his gratitude to programming problem,” Operations Research Letters, vol. 37, no.2, pp.
114-117, Mar. 2009.
Rajamangala University of Technology Phra Nakhon, [22] Y. M. Park, J. B. Park, and J. R. Won, “A hybrid genetic
Thailand for support. algorithm/dynamic programming approach to optimal long-term
generation expansion planning,” Electrical Power and Energy Systems,
vol. 20, no. 4, pp. 295-303, May 1998.
REFERENCES [23] K. J. Ho, P. J. Keun, and K. Sung. “A market based analysis on the
[1] J J. B. Park, Y. M. Park, J. R. Won, and K. Y. Lee, “An improved generation expansion planning strategies,” in Proc. IEEE Conf. power
genetic algorithm for generation expansion planning,” IEEE Trans. system technology, 2005.
PowerSystems, vol. 15, no. 3, pp. 916-922, Aug. 2000. [24] A. G. Kagiannas, D. Th. Askounis, and J. Psarras, “Power generation
[2] S. Kannan, S. Baskar, J. D. McCalley, and P. Murugan, “Application of planning : a survey from monopoly to competition,” Electrical Power
NSGA-II algorithm to generation expansion planning,” IEEE Trans. and Energy Systems, vol. 26, no. 6, pp. 413-421, July 2004.
Power Systems, vol. 24, no. 1, pp. 454-461, Feb. 2009. [25] S. Kannan, S. M. R. Slochanal, and N. P. Padhy, “Application of
[3] M. S. Sepasian, H. Seifi, A. A. Foroud, and A. R. Hatami, “A multiyear evolutionary computation techniques for generation expansion
security constrained hybrid generation-transmission expansion planning planning,” in Proc. IEEE Conf. and Exposition, Transmission and
algorithm including fuel supply costs,” IEEE Trans. Power Systems, vol. Distribution, 2003, pp. 120-125.
24, no. 3, pp. 1609-1681, Aug. 2009.
[4] Y. M. Park, K. Y. Lee, and L. T. 0. Youn, “New analytical approach for Vallop Phupha received his Ph.D. in Industrial Engineering from King
long-term generation expansion planning based on maximum principle Mongkut’s University of Technology North Bangkok (KMUTNB), Thailand
International Science Index, Industrial and Manufacturing Engineering Vol:6, No:4, 2012 waset.org/Publication/10548

and gaussian distribution function,” IEEE Trans. Power Apparatus and in 2011. He is an assistant professor at the Department of Industrial
Systems, vol. PAS-104, no. 2, pp. 390-397. Feb.1985. Engineering, Faculty of Engineering, Rajamangala University of Technology
[5] A. Sadegheih, “New methods for system planning,” Applied Soft Phr Nakhon, His research interests are energy and planning.
Computing, vol. 11, no. 1, pp. 1098-1107, Jan. 2011. Thong Lantharthong received his M.Eng in Electrical Engineering from
[6] S. H. Sevilgen, H. H. Erdem, B. Cetin, A. V. Akkaya, and A. Dagdas, Rajamangala University of Technology Thanyaburi, Pathumthani, Thailand in
“Effect of economic parameters on power generation expansion 2010. He is currently a lecturer at the Department of Electrical Engineering,
planning,” Energy Conversion and Management, vol. 46, no. 11-12, pp. Faculty of Engineering Rajamangala University of Technology Phra Nakhon
1780-1789, July 2005. (RMUTP), Bangkok, Thailand. His research interests include power system
[7] J. Wang, M. Shahidehpour, Z. Li, and A. Botterud, “Strategic generation operation, optimization technique, and distributed generation.
capacity expansion planning with incomplete information,” IEEE Trans. Nattachote Rugthaicharoencheep (M’10) received his Ph.D. in Electrical
Power Systems, vol. 24, no. 2, pp. 1002-1010, May 2009. Engineering from King Mongkut’s University of Technology North Bangkok
[8] M. K. Sheik-El-Eslami, M. P. Moghaddam, and S. Jadid “Expansion (KMUTNB), Thailand in 2010. He is currently a lecturer at the Department
planning in private generation companies: A practical method,” in IEEE of Electrical Engineering, Faculty of Engineering Rajamangala University of
Proc. Conf. Power Engineering Society General Meeting, 2006, pp. 1-5. Technology Phra Nakhon (RMUTP), Bangkok, Thailand. His research
[9] A. Ramos, I. J. Percz-Arriaga, and J. Bogas, “A nonlinear programming interests include power system operation, optimization technique, and
approach to optimal static generation expansion planning,” IEEE Trans. distributed generation.
Power Systems, vol. 4, no. 3, pp. 1140-1146, Aug. 1989.
[10] Y. Q. He, and A. K. David, “Advances in global optimisation for
generation expansion planning,” IEE Proc.-Gener. Transm. Disfrib.,vol.
142, no. 4, pp. 423-428, July 1995.
[11] I. Dyner, and E. R. Larsen, "From planning to strategy in the electricity
industry”, Energy Policy, vol. 26, pp. 1145-1154, Nov. 2001.
[12] S. M. M. Tafreshi, A. S. Lahiji, J. Aghaei , and A. Rabiee, “Reliable
generation expansion planning in pool market considering power system
security,” Energy Conversion and Management, vol. 54, no. 1, pp. 162-
168, Feb. 2012.
[13] Y. Zhou, L. Wang, and J. D. McCalley, “Effective incentives design for
renewable energy generation expansion planning: An inverse
optimization approach,” in Proc. IEEE Conf. Power and Energy Society
General Meeting, 2011, pp. 1-7.
[14] A. S. Chuang, F. Wu, and P. Varaiya, “A game-theoretic model for
generation expansion planning: problem formulation and numerical
comparisons,” IEEE Trans. Power Systems, vol. 16, no. 4, pp. 885-891,
Nov. 2001.
[15] F. H. Murphy, and Y. Smeers, “Generation capacity expansion in
imperfectly competitive restructured electricity markets,” Operations
Research, vol. 53, no.4, pp.646-661, July- August 2005.
[16] V. Nanduri, and T. K. Das, “Game theoretic approach for generation
capacity expansion in restructured power markets,” in IEEE Proc. Conf.
Power and Energy General Meeting-Conversion and Delivery of
Electrical Energy, 2008, pp. 1-3.
[17] J. L. C. Meza, M. B. Yildirim, and A. S. M. Masud, “A model for the
multiperiod multiobjective power generation expansion problem,” IEEE
Trans. Power Systems, vol.22, no.2, pp.871-878, 2007.
[18] S. Ahmed, A. J. King, and G. Parija, “A multi-stage stochastic integer
programming approach for capacity expansion under uncertainty,”
Journal of Global Optimization, vol.26, pp.3-24, May 2003.
[19] C. H. Antunes, A. G. Martins, and I. S. Brito, “A multiple objective
mixed integer linear programming model for power generation
expansion planning,” Energy, vol.29, no.4, pp. 613-627, Mar. 2004.
[20] H. J. Greenberg, and F. H. Murphy, “Computing market equilibria with
price regulations using mathematical programming,” Operations
Research, vol. 33, no.5, pp. 935-954, 1985.

International Scholarly and Scientific Research & Innovation 6(4) 2012 787 scholar.waset.org/1999.8/10548