Received: 26 September 2018 Revised: 20 December 2018 Accepted: 6 January 2019

DOI: 10.1002/er.4388

R E V I E W PA P E R

A comprehensive state-of-the-art survey on power

generation expansion planning with intermittent
renewable energy source and energy storage

Olubayo M. Babatunde1 Josiah L. Munda1 Yskandar Hamam1,2

1
Department of Electrical Engineering,
French South African Institute of Summary
Technology, Tshwane University of Generation expansion planning (GEP) is a power plant mix problem that iden-
Technology, Pretoria, South Africa
2
tifies what, where, when, and how new generating facilities should be installed
ESIEE - Paris, Noisy-le-Grand, France
and when old units be retired over a specific planning horizon. GEP ensures that
Correspondence the quantity of electricity generated matches the electricity demand through-
Olubayo M. Babatunde, Department of
out the planning horizon. This kind of planning is of importance because most
Electrical Engineering, French South
African Institute of Technology, Tshwane production and service delivery is dependent on availability of electricity. Over
University of Technology, Pretoria, South the years, the traditional GEP approaches have evolved to produce more real-
Africa.
Email: olubayobabatunde@gmail.com
istic models and new solution algorithms. For example, with the agitation for
green environment, the inclusion of renewable energy plants and energy stor-
Funding information
age in the traditional GEP model is gradually gaining attention. In this regards, a
NRF-TWAS, South Africa, Grant/Award
Number: UID: 105469 handful of research has been conducted to identify the optimal expansion plans
based on various energy-related perspectives. The appraisal and classification of
studies under these topics are necessary to provide insights for further works in
GEP studies. This article therefore presents a comprehensive up-to-date review
of GEP studies. Result from the survey shows that the integration of demand
side management, energy storage systems (ESSs), and short-term operational
characteristics of power plants in GEP models can significantly improve flexi-
bility of power system networks and cause a change in energy production and
the optimal capacity mix. Furthermore, this article was able to identify that to
effectively integrate ESS into the generation expansion plan, a high temporal res-
olution dimension is essential. It also provides a policy discussion with regard to
the implementation of GEP. This survey provides a broad background to explore
new research areas in order to improve the presently available GEP models.

K E Y WO R D S
energy storage system, generation expansion planning, intermittent renewable energy, operational
flexibility, optimisation methods, power plant mix problem, renewable energy share

1 I N T RO DU CT ION with gross domestic product (GDP).1 Since its product is

defined in terms of goods and services, it provides the
Access to a reliable power system is crucial to the devel- impetus for economic upturn. The power sector is there-
opment of a nation's economy due to its interrelationship fore a substructure on which many economies are built.

Int J Energy Res. 2019;1–30. wileyonlinelibrary.com/journal/er © 2019 Wiley Publishing Ltd. 1

2 BABATUNDE ET AL.

As a result, electricity is essential for the socio-economic demand (planning).5 On the other hand, to ensure sus-
development of any society. However, population and eco- tainability, energy demands are expected to be met using
nomic growth, improvement in social status, and techno- an environmentally friendly, cost-effective, modular, and
logical advancement contribute to increase in electricity reliable means. These criteria result in a conflicting optimi-
demand. In order to maintain a prescribed level of comfort, sation problem involving engineering, management, and
electricity demand must be met in an economical man- economics.
ner. This has made electricity planning a relevant problem In order to adequately manage the power system indus-
among various researchers. Electricity demand is expected try, the capacity planning problem has been divided
to be supplied efficiently and reliably in a sustainable man- into demand forecasting, distribution expansion plan-
ner. Apart from this, it is necessary that demands are ning (DEP), transmission expansion planning (TEP), and
supplied by the most economical and flexible scheme that generation expansion planning (GEP).6 For each capac-
can adequately accommodate unexpected peak demands ity planning problem, the time horizon can be divided
or withstand loss of some electricity facilities on the grid.2 into long-term, medium-term, or short-term studies7
In order to achieve the aforementioned, the electric- (Figure 1). The aim of TEP is to identify the time, loca-
ity industry is traditionally divided into the distribution, tion, and the number of transmission facilities to be
transmission, and the generation company. The distri- included in the power system network to reliably satisfy
bution system comprises low-voltage lines that feed the the consumer demand growth under different conflict-
residential and commercial loads while the generation ing criteria with minimum investment, operation, and
consists of different generators that feed the transmis- interruption costs.6 In DEP, the distribution network is
sion lines. This complex components all work together designed and planned to accommodate the energy needs
under various conflicting objectives and constraints to of the consumers without violating the standard network
provide reliable electricity for consumers.3,4 The com- constraints.8 GEP is a power plant mix problem that iden-
plex nature of the power system structure necessitates tifies what, where, and when new generating facilities
the need to plan the electricity industry in order to iden- should be installed and when old units should be retired
tity the needs for network expansion to forestall system over a specific planning horizon.6,9 The what estimates
collapse as a result of overloading and to avoid overin- the capacity of the generating facilities to be built across
vestments. Power system collapse can lead to massive the planning horizon, while the how evaluates the opti-
blackouts, load shedding, and huge economic losses. Over- mal generating plant combination. The where identifies
investment in capacity planning can lead to mothballing of the site where the new generating unit will be located
power system facility. In the process of avoiding these chal- and when determines the appropriate time when the gen-
lenges, a power system industry is expected to satisfy the erating facility will commissioned into a power system
present energy demand (operations) and the future energy network.

FIGURE 1 Power system capacity planning categories2 [Colour figure can be viewed at wileyonlinelibrary.com]
BABATUNDE ET AL. 3

These planning problems are strategic for any electric-

This review
ity industry. A slight mistake in any of them can affect the
rest. For instance, the energy demand is dependent on the
economic growth and the population growth; hence, eco-

✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
nomic and population growth results in increase in energy
demand.10 This could affect the input into a GEP model

Sadeghi et al36
and consequently result in capital loss in terms of invest-
ments, operational environmental, and social costs. From
the foregoing, after load forecasting, GEP is the first major
step in utility expansion planning process.5 The study of

✓
✓
✓
✓

✓
GEP has been carried out for over 7 decades when it was

Oree et al35
modelled as a linear programming (LP) problem by Masse
and Gibrat.11
Although GEP can be grouped into different aspects,

✓
✓
✓
✓
✓

✓
the prominent and basic concern is the decision on the

Hemmati et al6
power plant mix and size of candidate generating units
to be constructed by utilities over a planning horizon.7
Solution to GEP problems will aid decision makers (DMs)
and planners in taking the right decisions on the optimal

✓
✓
✓

✓
✓
power plant mix, generation technology alternative, capac-
ity, location, and the time of construction under varying

Zhu and Chow31

constraints and techno-economic policies. Such policies
are generally subject to retirement of aging generating
units as well as investment and production cost of candi-
date plants. ✓
Recently, there are new and emerging trends in GEP.
Traditionally, the focus of the GEP problem is on finding
Kagiannas et al30

the most economical generation expansion schedule to be

installed in specific years over a planning horizon.12 This
is done so as to achieve a high level of reliability. Gener-
ally, for GEP problems, the determination of only the type
and size of candidate plant is acceptable for a network
✓
✓

whose transmission line is not congested.13 Whereas, when

Kannan et al29

a transmission network experiences congestion (due to

inclusion of renewable energy [RE]) it is usually required
that the effects of network constraint be included in the
GEP model. Inclusion of the network constraints will
TABLE 1 Previous review/survey studies related to GEP

✓
✓

also facilitate the determination of the appropriate loca-

References

tions where new generation capacities will be constructed.

Hobbs28

In order to obtain a more realistic model, it is there-

fore of importance to represent the network in GEP.14,15
✓

Furthermore, the single time period (a relatively simpler

Distribution of research (by continent)

approach) of GEP is no longer enough.16,17 The use of

multiple time periods along the planning horizon in now
Intermittent renewable energy
Model formulation discussion

gaining more attention.17-20 More recently, the inclusion of

Demand side management

demand side management (DSM) and RE in the genera-

Environmental concerns
Multicriteria approach
Optimisation methods

Comparison of studies

tion mix has been identified as an effective approach to

Model classification

addressing energy efficiency, conservations, CO2 emission,

Unit commitment
Policy discussion

and environmental concerns.12,21-23 This is aimed at limit-

Energy storage

Sustainability
Uncertainty

ing emissions while ensuring the conservation of conven-

Reliability
Theme

tional energy sources.24 Apart from these issues, the effect

of uncertainty is also important. Ideally, all the decisions
are made under uncertainty. Some of these uncertainties
4 BABATUNDE ET AL.

may be technical, economic, climatic or political (govern- another study presented a state-of-the-art survey on TEP
ment policies).25 All the aforementioned attributes makes that focused not only on the issues discussed in Hemmati
the formulation of the GEP problem non-linear and dis- et al6 but also on integration of wind technology, security,
crete in nature, which is usually difficult to solve using and flexible AC transmission system (FACTS) devices.34 A
simple mathematical approaches. Hence, when long term review study conducted by Oree et al35 presented a survey
planning is required, the sizes of such problems become on the evolution of GEP methods considering environ-
large and as such the memory size required to carry out mental policies and the presence of uncertainty. A survey
the solution. Therefore, the use of an efficient solution of GEP problem based on key events in electrical power
algorithm is inevitable.26,27 sector over the last 30 years has also been presented.36 A
summary of the previous review/survey paper related to
GEP is presented in Table 1. It can be seen from the sum-
mary that the existing survey is deficient in the areas of
1.1 Existing reviews
model classification, energy storage, sustainability, flexi-
In the past, various researchers have contributed to the bility, and policy discussion. The review of these themes
review on some aspects of GEP problem. The classifica- and some other emerging issues in GEP is the focal point
tion of solution approaches and emerging technologies of this paper.
in GEP has been previously presented.28-32 Hobbs28 pre-
sented a survey of a range of models applicable in GEP
while Zhu and Carolina31 reviewed both heuristic and clas-
sical approach of solving GEP model. Kagiannas et al30 1.2 This review
focused on a comprehensive review of generation plan- Despite the availability of literature survey on GEP stud-
ning methods as it apply to competitive electric power ies, there are various aspects that either are not adequately
generation market. A taxonomical survey of distributed investigated or have not been reviewed. These issues are
generation panning has also been presented.33 Further- comprehensively reviewed in this paper. Some of these
more, a study6 focused on the modelling, reactive power issues include model formulation, solution methodology,
planning, solution approaches, electricity market, reliabil- uncertainty, RE integration, addition of emissions cur-
ity, DSM, uncertainties, line congestion, and distributed tailment in GEP model, and application of multicriteria
generation, as it relates to GEP and TEP. However, their approach in GEP. Some other emerging subjects that are
study focused more on the TEP problem. The authors in reviewed in this paper include inclusion of energy storage

FIGURE 2 Methodological approach adopted for the review. GEP, generation expansion planning [Colour figure can be viewed at
wileyonlinelibrary.com]
BABATUNDE ET AL. 5

system (ESS) in GEP models to mitigate intermittency of 1.2.1 Methodology

RE resources, classification of GEP models, the inclusion For this survey, a comprehensive literature search strategy
of unit commitment (UC) and economic dispatch, policy was adopted. An initial search of related studies was con-
discussion, sustainability, and DSM. As a contribution to ducted using Google scholar under the keywords power
the body of knowledge on the review of GEP, this paper GEP, power plant mix problem, intermittent RE in capacity
presents a comprehensive up-to-date survey of the GEP planning, energy storage in capacity planning, and sus-
problems, which addresses the aforementioned research tainable energy. Furthermore, the database of journals and
gaps. It is essential to state that the survey is not extended conferences were explored for relevant materials. Details
to other important elements of power system planning of the methodology adopted in the selection of the litera-
such as TEP and load forecasting. ture are shown in Figure 2.

FIGURE 3 Graphical summary of works compared in Section 8 [Colour figure can be viewed at wileyonlinelibrary.com]
6 BABATUNDE ET AL.

This survey included studies based on the current 1.2.2 Contribution of the study
research themes and the novelty of the papers. The papers This review presents three contributions to the existing
were classified and centred on themes such as objective studies on the survey of GEP. For the first time, we present
function, solution method, network topology, reliability, a classification and synopsis of GEP models. The GEP
intermittent RE, uncertainty, ESS, DSM, and UC. Papers model classifications are based on time frame, uncer-
were included from 1973 to date and included in the main tainty, market structure, and network topology. Secondly,
summary (Section 8). a STEEP-cum-SWOT framework for GEP sustainability
Figure 3 presents the graphical summary of the studies is proposed. As an additional contribution, this study
compared in Section 8. Due to the fact that some stud- presents a summary of different studies in GEP based on
ies included various themes, the total number of studies various themes that has been discussed and presented
is not the same for all charts. There has been remark- in the literature. The rest of this paper is organised as
able increase in the annual quantum of researches inter- follows: Section 2 discusses the classification of GEP mod-
ested in GEP over the last few decades from a handful els, Section 3 explores the formulation of GEP models,
to more than 60 in the 2010 (Figure 3A). The major- Section 4 reviews the various solution approaches used in
ity of the studies (54%) has the first author affiliated GEP models, Section 5 presents a review on the environ-
to Asia (Figure 3B). Europe (22%) comes second with mental consideration in GEP studies, Section 6 presents a
respect to the quantum of publications related to GEP policy discussion on the implementation of GEP, Section
studies. This is followed by North America (14%), South 7 compares and contrasts some researches presented in
America (8%), and Africa (2%). The summary also shows the reviewed literature, and Section 8 finally presents the
that the issue of flexibility (DSM, UC, and energy storage) conclusion of the study.
has received little attention when compared with the other
themes (Figure 3C).
Likewise, a substantial number of studies in this sur- 2 CLASSIFICATION O F G EP
vey focused on representing the GEP as a centralised MODELS
node model (81%) (Figure 3D). This type of model is
simple to formulate and solve when compared with This section gives the classification of GEP models accord-
the network-constrained model, which constitutes 19% ing to time horizon, handling of uncertainties, network
of the studies compared in Section 8. It further shows topology, and market structure (Figure 4). The classifica-
that the classical optimisation technique has gained tions and subclassification include time horizon—static
more popularity as compared with the heuristic methods and dynamic; network topology-centralised and network
(Figure 3E). Furthermore, the majority of the studies mod- constrained; uncertainties handling- deterministic and
elled the GEP as a single-objective optimisation problem stochastic and market structure-regulated and deregu-
(Figure 3F). lated.

FIGURE 4 Classification of generation expansion planning (GEP) models [Colour figure can be viewed at wileyonlinelibrary.com]
BABATUNDE ET AL. 7

2.1 Time horizon likely that the demand at the reference year will be far
greater than that of the short term during the planning
Based on the time framework adopted in the planning
horizon. This will result in overcapacity at the short-term
horizon, a GEP problem can be classified into two types:
level. Another challenge of this model is that it is possi-
static16,17 and dynamic planning.19,20 In the static model,
ble for the energy demand to reduce along the planning
decisions about the generation expansion are only made
horizon.13 When this happens it becomes impossible to
at the onset of the planning horizon. As such, the optimal
adapt changes through modularity since all the facilities
plan is made at a single point in time on the planning hori-
would have been constructed at the onset of the planning
zon. When the static model is adopted, the entire planning
horizon. However, these challenges can be mitigated by
horizon is represented by a year that is typically regarded
using the dynamic GEP model.13 It allows capacity to be
as the reference or target year. Usually, the energy demand
added as the need arises, thereby encouraging flexibility
for the target year is used in the determination of the size
in the construction of generation units along the planning
of generating units to be constructed. Since the genera-
horizon.13 By doing so, changes along the planning hori-
tion expansion is based on the projected demand, which
zon are easily adopted into decision making. The dynamic
increases over the planning horizon, the reference year
model is usually more complex than its static model equiv-
is usually taken as last year on the planning horizon so
alent. It may result in the involvement of large number
as to accommodate all possible load growth. Hence, this
of variables, constraints, and enormous computational
model only determines the size of the generator units to be
efforts to obtain the optimal plan. These models have been
added.37 The static model does not determine when.
applied by some researchers. For example, Ramos et al38
As opposed to the static model, the dynamic planning
presented a static optimal mix problem in generation plan-
model splits the planning horizon to various time periods
ning using a non-linear programming (NLP) approach,
with each period consisting of specific number of years.
while Pozo et al39 formulated a three-level static equilib-
The time periods are therefore a subset of the entire plan-
rium model based on mixed integer linear programming
ning horizon. As done in the static model, the last year of
(MILP). Studies on dynamic model in GEP have also been
each time period is usually specified as the target year on
presented.18,40-48
which the periods planning will be based. Dynamic model
does not only determine what but also when.37 The differ-
ence between the dynamic and static models is shown in
Figure 5. One major advantage of the static model is its rel- 2.2 Network topology
ative simplicity in terms of formulation as well as solution Based on network topology, GEP problem may be classified
approach.37 It can therefore be used for preliminary stud- into two categories: the single-node (centralised) model
ies in GEP studies. However, it has some drawbacks. One and the network-constrained model. In the single-node
of this is that the decision for the whole planning period model, network constraints are neglected, and as such a
is based on the demand projection for the last year.13 Since single-node virtual network is assumed. It assumes that all
the GEP problem is for long-term planning horizon, it is generators and load points are connected to a single node.

FIGURE 5 Classification of generation expansion planning (GEP) models based on time horizon: A, static GEP model; B, dynamic GEP
model [Colour figure can be viewed at wileyonlinelibrary.com]
8 BABATUNDE ET AL.

Therefore, the single-node modelling approach identifies environment, and as such, a more accurate decision is
the sizes of the generators to be added over the horizon informed. This kind of model is developed based on past
and when to be added them. It does not determine the experience and future expectations and considers all pos-
locations of the generators. The formulation and solution sible scenarios that may occur during the course of the
approach for the single-node model is relatively simple as planning horizon.
compared with the network-constrained approach. This is
because the practical electricity network consists of thou-
sands of nodes and transmission lines whose modelling 2.3.1 Uncertainty handling
and solution approach may be complex.6 It is however In long-term planning, the order and nature of occurrence
generally acceptable to use the single-node model if the of events are unusually unknown. Furthermore, when
targeted network is not congested. It is worth noting that these events occur, the magnitude of their effect cannot be
inclusion of the network constraint is important in a net- easily predicted. Uncertainty usually occurs due to lack of
work whose transmission line is congested or consider- access to perfect information of key input factors involved
able share of RE is included.13 In the network-constrained in the planning. Hence, uncertainty cannot be neglected in
model, the network constraints are considered. The out- power system expansion planning.35 It is therefore impor-
put of this model gives the optimal size, as well as the tant to incorporate a reasonable level of uncertainty in GEP
best location of the candidate plants on the network. Con- model, since they involve a long term planning so as to
sidering the literature, many studies have focused on the mitigate risks.66 According to Bracht et al,67 uncertainty in
single-node model,7,9,18,24,26,43,49-59 while60-65 focused on the GEP emanates from decisions on dispatch decision affect-
network-constrained GEP. ing investment decisions, which in itself are dependent on
long-term uncertainties. Table 2 gives a general idea on the
relevant uncertainties that substantially impact GEP. Some
of these includes demand growth, fuel cost, delay in project
2.3 Uncertainty point of view completion, and financial constraints.66 Addition of uncer-
Based on uncertainty representation, GEP problem may tainties increases the complexity of the resulting model
also be formulated in two ways—deterministic and as well as the computational efforts required to obtain
stochastic approach. The deterministic approach assumes optimal output for such models.13 Over the years, method-
that all information as at the time of decision making ologies that can handle uncertainties in GEP have been
is accurate. The deterministic model can only take into proposed and developed. These include scenario analy-
consideration the extreme case. It does not accommo- sis, sensitivity analysis, portfolio analysis, and probabilistic
date changes that can occur during the planning horizon. analysis.68 These approaches are based on the determin-
Some of such changes may include future demand, wind istic model and have been adapted to incorporate uncer-
speed, fuel cost, and DSM. The deterministic model has tainty handling characteristics. In order to ascertain the
the advantage of simplicity in formulation and solution effectiveness of these techniques to accommodate uncer-
approach. Conversely, since practical GEP is over a long tainties, two metrics are usually used: robustness and flexi-
range planning horizon, obtaining a perfect forecast of bility. Robustness investigates the extent to which the plan
the system characteristics may be difficult. The stochas- is affected by changes in parameters while flexibility is the
tic model has the ability to handle this drawback. In the ability of the plan to accommodate and effectively adjust
stochastic approach, decisions are made within a fuzzy to changes.69,70

TABLE 2 Uncertainty sources in generation expansion planning (GEP)

Uncertainty Source Examples
Cost Fuel cost, cost of plant construction, operation
and maintenance cost, environmental cost, energy
price, regulatory cost, inflation rate, interest rate,
DSM incentives
Load growth Population growth, economic growth, improvement
in social status, weather conditions, demand
response (DR), DSM, demand uncertainty
Facility performance Line outage, unit outage, renewable energy system
inclusion, reserve margins, new plant technologies,
power plant characteristics, emission
Socio-political Social status, policies, regulation and standards,
public perception, market, change in government
BABATUNDE ET AL. 9

In scenario analysis, a set of possible futures known as In the deregulated model, different sections of the indus-
scenarios are generated by making various assumptions try are owned and controlled by different entities.90 The
about the future occurrences.68 A combination of impor- transmission company purchases energy from the gener-
tant variables that constitute uncertainties are simulated ation companies (GenCos) and sells to the distribution
to create expansion plans over the planning horizon. This companies (DisCos) who in turn sell to the consumers. In
technique offer DMs the ability to choose from a vast set the deregulated model, the goal of a utility is to maximise
of realistic futures that can offers promising generation its own profit, while ensuring profits for the participating
expansion options. One of the challenges of this method IPPs. Its main objective is to provide a competitive environ-
is that difficulty arises in performing a robust uncertainty ment for the participating companies. Furthermore, the
analysis because of the need to assign weights to individ- utility is saddled with the responsibility of checking for
ual scenarios to evaluate the probability of occurrence and the possibilities of underinvestment or overinvestment, in
scenarios need to be mutually exclusive and exhaustive.71 order to ensure system security, national security (fuel mix
Sensitivity analysis technique develops a set of preferred ratio), social welfare, and overall system reliability.
combination of plans and different values are assumed Since the deregulation of the electricity market, vari-
for potentially important variables that can contribute ous changes have been experienced in GEP studies. In the
to uncertainties.68 The performance of the initial plan is regulated market structure, all the capacity planning-GEP,
then examined based on the changes in the conditions. TEP, DEP, and load forecasting are centrally performed
This process allows DMs to identify which factor causes and closely monitored. Whereas, in the deregulated elec-
the largest changes and the alternatives that are most tricity market, choices and assessments are implemented
sensitive.68 individually by the companies involved.90 This can lead
In portfolio analysis, DMs identify multiple plans with to inadequate harmonization in planning process in the
each attached to different groups of objectives.68 The industry if not well managed. Furthermore, the deregu-
resulting plans are usually subjected to sensitivity analy- lated electricity market is susceptible to changes in fuel
sis. According to Seddighi and Ahmadi-Javid,71 scenario, prices, which considerably escalates risk as it concerns
sensitivity, and portfolio techniques cannot provide infor- power plant pay-off.90 All these has considerably intro-
mation on the likelihood of changes as well as detailed duced changes in the representation and solution method-
information on the robustness and flexibility of associated ology in GEP studies. While majority of studies reviewed
plans. To overcome these shortcomings, the probabilistic in this work have considered the GEP model with respect
approach assigns probability values to various uncertain to regulated market, the following literature16,19,61,90-100 con-
variables. It further identifies the optimal output using sidered the GEP problem in a competitive market.
tools such as stochastic programming and Monte Carlo
simulation.35
From the foregoing, uncertainty analysis gives the free- 3 MODEL FORMULATION
dom to make creative decision along the planning horizon.
A handful of literature have considered uncertainties in The GEP problem is basically an optimisation problem
GEP models. These includes emission,72 energy price,73,74 in which the aim is to establish the optimal size, type of
intermittency of RE sources,61,75-77 fuel price,58,72,74,78 generating unit, and construction time of new generat-
market,15,79,80 unit outage,77,78,81-87 line outage,17,77,83,84,86 and ing facilities so as to satisfy the electric power demand
load.19,39,73-75,80,88,89 at least cost over a planning horizon.28 The goal is either
to minimise or maximise a single- or multiple-objective
functions subject to some constraints (Figure 6). The GEP
model can be made flexible by the consideration and addi-
2.4 Market structure tion of various objectives and constraints. Usually, the
Based on market structure, the GEP problem is classified GEP objective function consists of two main parts namely
either as regulated or as deregulated/structured. In the the investment and the operation vector.7,15,22,26,31,101 This
traditional (regulated) GEP model, the overall goal of the property of GEP problem makes the application of the
industry is to identify the technology, size, location, and traditional branch and cut algorithms such as Benders
the time horizon to install candidate plants so as to ade- algorithm relevant for solution purposes.26,31 In order to
quately meet the predicted demand at minimum cost. In ascertain the optimal expansion plan, it is required that dif-
the regulated GEP model, the generation, transmission, ferent constraints that affect the expansion plan be consid-
and distribution in the electricity industry are under the ered. The constraints can also be classified into two types:
control of a sole owner.90 However, if under the owner- mandatory constraints and optional constraints. Some
ship of separate entities, it is managed by a single agency. mandatory constraints include balance of power demand,
10 BABATUNDE ET AL.

plant capacity limits, and non-negativity constraints. A Traditionally, GEP models have represented the plan-
non-negative constraint and integer constraints are usu- ing problem with a single-objective function, basically
ally added depending on the structure of the problem. The minimisation of the overall cost as shown by prvious
mandatory constraints must be considered when formu- studies.7,9,12,26 This is done so as to attain an acceptable
lating the GEP problem. On the other hand, the optional degree of reliability and energy demand at minimum
constraints may not be necessarily considered in the GEP cost.12 Up till recently, the minimisation of overall cost only
model. They are usually added to make the model robust has been seen as adequate for the GEP model.
and flexible. Some of these constraints are related to envi- However, the minimisation of the total cost of a specific
ronment, DSM, fuel import limit, reliability, investment GEP project may not be the optimum objective function
limit, renewable fraction, maintenance, and so on (see especially if more than one conflicting factors is of inter-
Table 3).7,9,51,102,103 Based on this, the GEP problem can be est. Hence, recent GEP problems have been represented
formulated as as a multiobjective optimisation problem. A multiobjec-
tive GEP optimisation problem can handle a compromise
Min or Max ∶ Ob𝑗ective𝑓 unction(s)
{ among various capacity planning objective functions to
mandator𝑦 constraint(s), (1)
sub𝑗ect to ∶ optional constraint(s). ascertain an optimal alternative. Some of these objectives
are related to the following: the inclusion of DSM, RE

FIGURE 6 Pictorial representative of generation expansion planning (GEP) problem [Colour figure can be viewed at
wileyonlinelibrary.com]

TABLE 3 Objectives and constraints available in the literature

Objectives Constraints
minimisation Total cost Emission level
Emission cost Reliability margin
Emission quantity Budget
Imported fuel/ electricity Peak demand
Cost of unserved demand Spin reserve
Fossil fuel used Generator capacity (upper and lower)
Fuel cost Number of units
Energy price risk Renewable penetration level
maximisation Renewable energy generation Energy storage participation
Feed-in-tariff Fuel import
DSM application Total electricity generated
Profit/revenue Resource allocation
Reliability Energy balance
Employment Installation time
Social acceptance Retirement/ lifespan
Network constraints
BABATUNDE ET AL. 11

FIGURE 7 General structure of generation expansion planning (GEP) problem formulation and solution [Colour figure can be viewed at
wileyonlinelibrary.com]

in the generation mix , minimisation of emissions, min- single-objective optimisation problem. Such model accom-
imisation of cost of RE intermittency, fuel consumption, modates salient factors such as reliability, risks, and emis-
energy price risk, reliability, and so on (Table 3).104 All sion into a single-cost model with simplified operating
these are aimed at improving the GEP models in terms of constraints to describe the operating characteristics of the
flexibility.6,24 The general structure for the formulation and power system.105 From the foregoing, the classical GEP
solution approach for GEP problem is given in Figure 7. model either minimises cost of generation in a regulated
electricity market or maximises profit in a competitive elec-
3.1 Single-objective GEP model tricity market. However, total cost minimisation or profit
maximisation of a particular GEP problem may be a less
The single-objective GEP model combines various objec-
preferred criterion especially when considering emissions
tives into one. This approach does not allow DMs to
reduction. Hence, the GEP problem has been modelled to
access solutions that present trade-off among various
incorporate distinct evaluation attributes as incommensu-
objectives. In this model, planners usually present all
rable objective functions instead of combining them into
objectives in terms of economic criteria such as genera-
a single-cost objective function.105 This will provide DMs
tion cost, investment costs, DSM cost, cost of RE inter-
with an in-depth understanding of the trade-off among
mittency, operation and maintenance cost, outage cost,
conflicting objectives, so as to establish a compromise
emission cost, and salvage value. To complete the model,
solution. One way of handling such conflicting multiob-
mandatory and optional constraints are also included
jective functions simultaneously is the use of multicriteria
to achieve the desired flexibility. Based on the litera-
decision-making (MCDM) approach.35 The MCDM tech-
ture, the single-objective GEP has been applied in both
nique makes decision based on multiple and conflicting
the regulated and deregulated electricity market struc-
criteria. It provides a stepwise approach based on which a
ture. The regulated electricity system aims to minimise
consensus decision is made by DMs from a set of feasible
total cost while the deregulated market is competitive
alternatives. The MCDM approach applied in GEP stud-
and maximisation of profit is the goal of the DMs.61 The
ies is divided into two, viz, multiattribute decision making
following studies7,9,12,15,26,41,49-52,61 presented the GEP as a
and multiobjective decision making (MODM).
single-objective optimisation problem.
In the MADM, DMs make preference decision (eg,
evaluation, prioritisation, and selection), based on avail-
3.2 Multiobjective GEP models able alternatives, which is characterised by multiple
For simplicity, the traditional formulation of the GEP and conflicting attributes or criteria.106 In MADM, the
problem usually involves presenting the model as a number of predetermined alternatives are typically
12 BABATUNDE ET AL.

constrained and are linked with a level of the attainment solving the resulting problem. According to Climaco et al,
of the attributes.106 With the help of interattribute and “TRIMAP combines three main procedures: weight space
intra-attribute appraisals, the final selection of the best decomposition, introduction of constraints on the objec-
alternative is carried out. Based on the features and type tive function space, and introduction of constraints on the
of information received by DMs, an attempt was made weight space.”112 Kim and Ahn113 presented a multicriteria
by Hwang and Masud107 to classify MADM methods into GEP procedure based on a preference-order dynamic pro-
17 techniques. In another study, Yoon and Hwang108 pro- gramming (DP). The multiattribute objective GEP model
posed a modified classification of 13 MADM methods. In was implemented using Wien Automatic System Planning
a similar study, Løken109 gave a classification of MADM Package (WASP). The study considered the following
into four broad categories (Figure 8). MODM is a con- objectives: minimisation of discounted total investment
tinuous form of MCDM. In MODM problems, DMs are cost, minimisation of cumulative CO2 emission from
required to meet various nonproportionate conflicting fossil fuels, and the minimisation of nuclear hazards.
objectives.35,110 A MODM model consists of a vector of Aghaei et al,104 formulated a multiperiod multiobjective
objective functions, decision variables, and constraints. mixed-integer linear programming (MILP) GEP optimi-
While attempting to either minimise or maximise the sation problem that incorporated RE source as part of
objective functions, DMs usually choose a solution from a the candidate plants. The study considered five objective
set of best alternatives because this kind of problem does functions that include maximisation of system reliability,
not have a unique solution. The major difference between minimisation of energy consumption, total costs, port-
the MADM and the MODM is that the former focuses on folio investment risk, and emissions. In order to obtain
problems with discrete spaces while the latter focuses on the nondominated solution for the proposed model, a
continuous decision spaces (basically on mathematical hybrid augmented weighted epsilon constraint and lexi-
programming with multiple-objective functions).110 cographic optimisation was employed. Coit et al114 pre-
One of the earliest efforts in addressing GEP problem sented a GEP problem with two objectives (minimisation
from a multiobjective perspective was presented as a of cost and minimisation of emissions) under uncertain-
two-stage procedure for evaluation of power expan- ties. Saboori and Hemmati115 proposed a multiobjective
sion planning.111 In a similar study, Climaco et al GEP model that simultaneously minimises the invest-
expressed a power GEP problem using multiobjective LP ment and operational costs of the generating units, and
optimisation.112 The study considered the optimisation of the cost of carbon capture and storage and maximises
three objective functions: minimisation of the total expan- the revenue cost over a long-term planning horizon. The
sion cost, maximisation of the generation scheme relia- resulting model is solved using particle swarm optimisa-
bility, and the minimisation of environmental impacts. tion (PSO) technique. Furthermore, Mavalizadeh et al116
The constraint considered by their study is also cate- presented a mixed-integer multiobjective linear model
gorised into three (load requirements, operational restric- for a coordinated generation and TEP based on two
tions, and budget). The TRIMAP method was used in sources of uncertainty: load forecast and prices. Pereira

FIGURE 8 Multiattribute decision making classification35 [Colour figure can be viewed at wileyonlinelibrary.com]
BABATUNDE ET AL. 13

et al23 also proposed a multiobjective GEP that incor- deterministic or probabilistic.2 The deterministic indices
porates the effects of renewables variable generation on reflect postulated conditions while the probabilistic
thermal power plants efficiency as well as hourly UC. indices evaluate the uncertainty present in power sys-
Promjiraprawat and Limmeechokchai,24 in another study tem operation. The deterministic indices though limited
proposed a hybrid multiobjective GEP model. The model are simpler to evaluate and requires little or no data.
minimised the carbon dioxide emissions and external cost. Probabilistic indices allow the numerical assessment of
Analytic hierarchy process is utilised to determine the system alternatives by taking directly into consideration
optimal Pareto. Tekiner et al43 also proposed a GEP model the features that affects reliability.2 Some of these features
that simultaneously minimises cost and air emissions over include the capacities of each power generating units
a long-term planning horizon. Aghaei et al48 presented and the forced outage rate of individual unit. Some of the
a multiperiod, multiobjective GEP model with the inclu- deterministic indices include largest unit (LU), dry year,
sion of RE source. The proposed model simultaneously and reserve margin (RM).2 Some other technical indices
optimises three objective functions, which include total such as loss-of-load probability (LOLP),4,7,9,12,29,48-51,71,121,122
investment costs minimisation, environmental impact loss-of-load expectation (LOLE),54,74 energy expected
minimisation, and system reliability maximisation. The not supplied (EENS),48-51,122,123 and value of lost load
model evaluates the optimal capacity and time as well (VOLL)84,121,124 have been used to benchmark minimal
as the plant mix along the planning horizon. A multi- reliability performance levels. Some other reliability
period, multiobjective GEP problem that simultaneously indices relevant in GEP are probability of positive margin
minimises total investment costs, imported fuel, environ- (POPM), expected unserved energy (EUE), loss of energy
mental impact, and fuel price risks of electric systems was probability (LOEP), expected loss of load (XLOL), emer-
formulated by Meza et al.40 gency operating procedure expectation (EOPE), frequency
and duration of failures to meet the load (FD), effec-
tive load-carrying capability (ELCC), and firm capacity
3.3 Reliability criteria in GEP studies
equivalent (FCE).2
The aim of power system planning is to satisfactorily meet From the current GEP literature, generator reliability is
the projected electricity demand at a minimum cost and often expressed in terms of technical constraint in GEP
minimum environmental impact. Satisfactorily meeting models thereby eliminating the needs of expressing the
the consumer demand can be viewed from various per- economic effect of different levels of reliability.
spectives. Construction of generation facilities above the
required capacity increases the average cost of generation
4 OPTIMISATION TECHNIQUES
and consequently the tariff because the consumers bear
the cost of excess capacity. Conversely, undercapacity will
IN GEP
result in unmet demand. Therefore, there must be an To obtain an optimal generation expansion plan, it is
optimum reliability level for the generating system based required that diverse constraints that affect the expansion
on the system's operational attributes. Usually, a technical plan be considered. Some of these include plant availabil-
index is used as a minimum standard for the evaluation ity, power demand, plant capacity limits, reserve capacity
of the generation system performance. This at times can limit, power flow limit, voltage angle limits, and power
be represented as an economic criterion, which is usually balance at nodes.9,12,18,26,48,61 While considering the objec-
introduced in an effort to embed the generating system tive function(s) and these constraints, the GEP problem
reliability considerations directly in the minimum cost takes the form of a multiple constrained mathematical
model. Hence, reliability of the system can be formulated programming problem with binary variables (generation
as part of the objective function(s)104,117 or included as unit/transmission line added or not). It could also include
part of the constraints.7,9,12,29,118-120 A typical generation continuous variables such as costs and emissions. As a
expansion plan is expected to satisfy the reliability crite- result of this, the GEP problem is usually formulated as
rion and ensure that the reliability requirements are met a mixed-integer non-LP problem (MINLP). It is however
as specified by appropriate standards. As such, DMs must possible to reformulate the MINLP as a mixed-integer lin-
ensure that designed system is responsive to challenges ear programming problem (MILP), which can be solved
such as forced outages, demand variation, scheduled using the classical or heuristic methods. In order to solve
maintenance, variations in operational conditions of the resulting problems, an efficient solution algorithm
power plants etc. It is therefore important to specifically that reduces the complexities of the problem, improves
consider the required level of reliability for system plan- computational efficiency, and saves computational time is
ning. The reliability indices applied in power systems needed. Several optimisation methods have been applied
operation and planning can be catagorised into either in GEP studies. These include the classical optimisation
14 BABATUNDE ET AL.

techniques and the meta-heuristic optimisation approach. ing, evaluating, and selecting expansion options, with or
There are also commercially available software whose without the users help (interactive or noninteractive).149
solution principle is based on either or both classical and To achieve this, the heuristic methods comprehensively
meta-heuristic methods. In general, the proposed solution preform a search of the solution space of the problem
approach applied to any GEP optimisation problem must to find near optimal solutions using a set of logical or
be able to obtain a solution within the feasible region. empirical rules based on natural occurrences, physical
principles, social behaviour, etc. The search process con-
tinues until the algorithm is unable to find a better solution
4.1 Classical optimisation based on the selection criteria. Some selection criteria that
have been considered by researchers are overall invest-
In this section, a summary of the available classical opti-
ment cost, reliability, renewable fraction, etc. Examples of
misation techniques that has been used in GEP studies
the meta-heuristic techniques in GEP literature include
is discussed. Basically, classical optimisation techniques
genetic algorithm,7,12,29,45,74,91,95,118,120,150-157 PSO,89,95,136,158-162
can obtain the optimum solution of differentiable and
simulated annealing,29,95,118 differential evolution,49-52,55,163
continuous functions. To locate optimum points, clas-
ant colony optimisation,29 artificial immune system,87,122
sical optimisation techniques uses the methods of dif-
frog leaping algorithm,119,121,164 bee algorithm,165,166 fuzzy
ferential calculus. Classical optimisation techniques are
and expert system,128,129,167-170 and tabu search29,95,154 ; meth-
useful in obtaining the optimum solution to uncon-
ods such as harmony search, biogeography, and hill climb-
strained maxima or minima of continuous and differ-
ing have found application in TEP and offer promising
entiable functions.36 This approach has the tendency to
applications in GEP problems.
handle three types of problems: single variable function,
multivariable functions with both equality and inequal-
ity constraints, and multivariable functions with no con-
straints. On the other hand, Lagrange multiplier is usually 4.3 Commercially available GEP software
used in problems with equality constraints while prob- packages
lems with inequality constraints are handled using the
This section presents some of the commercially available
Kuhn-Tucker conditions to identify the optimum solution.
software developed to help DMs in GEP. The two main
Some of the classical techniques that have been applied
functions of these tools are electricity production simula-
to GEP problems include LP,40,125,126 DP,19,30,37,101,113,123,127-130
tion and capacity expansion optimisation.2 The software
system dynamic method (SDM),131,132 hierarchical decom-
packages are developed to evaluate the investment cost,
position (HD),24,40,133 game theory (GT),16,120,134-136 NLP,38,137
installation periods of candidate plants, and to obtain the
branch and bound (BB),110,138-140 mixed-integer linear pro-
optimum plan.
gramming (MILP),43,58,83,105,141 Dantzig-Wolfe decomposi-
The WASP is the most widely used commercially avail-
tion (DWD),22,142 quadratic programming (QP),143,144 cross
able software for GEP. It was developed by the Tennessee
entropy (CE),9 and the Benders decomposition method
Valley Authority (TVA) and Oak Ridge National Labora-
(BD).7,26,63,65,101,103,145-147 Although easy to use, the classi-
tory (ORNL) for the International Atomic Energy Agency
cal techniques present a challenge in real-world applica-
(IAEA) and has found application in some studies.171-177
tions because practical problems consist of noncontinuous
With WASP, a DM can determine the optimal expansion
and/or differentiable objective functions.36
plan for a power generation network over a long planning
period based on the constraints specified by the planner.178
In WASP the optimum expansion plan is specified based
4.2 Meta-heuristic techniques on the minimum discounted total costs.179 The most recent
The heuristic techniques are alternatives to the classical version of the software (WASP-IV) includes environmen-
optimisation methods for solving optimisation problems. tal constraints, reliability criteria, and fuel availability.85
It is a higher-level method designed to obtain, gener- SPPA-M3000 Generation Planning and Monitoring opti-
ate, or select a partial search algorithm that may give mises asset utilization for the entire power plant fleet by
a sufficiently good solution to an optimisation problem real-time integration of financial and GEP information.180
especially those with insufficient information or limited It also performs the availability, load capability, economic
computation resource. Compared with classical optimisa- dispatch, load scheduling, and generation monitoring.
tion techniques, and numerical methods, meta-heuristics SPPA-M3000 Generation Planning and Monitoring deter-
approaches do not guarantee a global optimality of the mines the optimum cost condition by drawing on a set
generated solution.7,148 When applied in GEP problems, of linear equations with n variables, it takes into account
these methods provide a stepwise approach for generat- environmental constraints.180 Electric Generation Expan-
BABATUNDE ET AL. 15

sion Analysis System (EGEAS) is another state-of-the-art and energy storage into the long-term power plant mix
GEP software package originally developed by Electric is essential.35,104 Generally, environmental impact in GEP
Power Research Institute (EPRI) in 1983.181 This software studies are included by invoking an emission constraint on
has modules that handle dispatchable generation sources a GEP model and setting a maximum acceptable emission
DSM and RE sources. In order to improve the software, limit. In other studies, the environmental impact is asso-
the version released in 2015 included the dump energy ciated with external costs related to emission penalty and
penalty and State Renewable Portfolio Standard (RPS) carbon taxes.
constraints. Some other commercially available software
that have found application with respect to energy expan-
sion include Long-range Energy Alternatives Planning 5.1 Emission consideration in GEP
(LEAP),21,182,182-184 PLEXOS Integrated Energy Model,185
Emission considerations in the available GEP models take
URBS,186 MARKAL/TIMES.187-196
either the form of an objective function (cost or emis-
In this section, the optimisation methods applied in GEP
sion quantity) to be minimised or as a constraint to set
studies have been presented and discussed. The classical
a maximum allowable limit. GEP literature also contains
optimisation techniques have been reported to exhibit lim-
various methods and mathematical functions to evaluate
ited scope in practical applications. Consequently, in some
the quantity of pollutants released by each fossil-fuel based
cases, meta-heuristic optimisation approaches are consid-
generation technology. Some of these include the linear
ered more efficient when compared with classical opti-
model,201 the quadratic polynomial model,202 hybrid poly-
misation techniques.7,29,159 Furthermore, in order to make
nomial and exponential emission function,87,122,162,203 and
planning easier for DMs and other stake holders, several
emission coefficients method.116,147
user-friendly commercial GEP tools have also been devel-
oped. This discussion could provide a guide in the selection
of the most effective method in the modelling and solution
5.2 Intermittent RE integration in GEP
approach of a specific GEP problem based on its complex-
ity and structure. It also stairs up prospects for original Concerns about climate change and environmental impact
research areas in the development and utilization of new of fossil fuel-powered generation have tremendously
heuristic and hybrid optimisation techniques in GEP. The increased and as a result, many nations have set targets and
efficacy of such optimisation techniques will be subject to policies for reducing emissions. One approach of achieving
a trade-off between the efficiency, quality of the solution, this target is by increasing the RE penetration on a util-
simplicity, and speed of execution. ity grid. The European commission has set a milestone to
reduce GHGs by 40% (in comparison to 1990) through a
27% base increment on the RE share and also 27% improve-
5 ENVIRONMENTAL ment in energy efficiency.204 RE integration into modern
CO NSIDERATIONS power system is an emerging strategy to provide green
energy and ensure sustainability of resources. It has been
More recently, the inclusion of DSM and RE in the gen- a major driver for a green environment in the power sector
eration mix has been identified as an effective approach during the last decade due to favourable government poli-
to addressing energy efficiency, conservations, CO2 emis- cies, advancement in RE technologies (RETs), as well as a
sion , and environmental concerns.12,21-23 This is aimed rapid decline in the procurement cost of RETs (especially
at limiting emissions while ensuring the conservation of solar PV). The integration of RETs in GEP has a dual ben-
conventional energy sources24 for sustainable future. At efit of long-term energy security and emission-free elec-
present, a large percentage of the electricity is generated tricity generation. A report released by IEA in 2015 shows
using conventional means (gas, coal hydro, and nuclear). that investments in RETs recorded a significant increase in
Most generating units worldwide, especially in developing 2014 by up to 85% year on year.205 The report further pro-
nations, are powered by conventional energy sources.197 jected that by 2040, renewables will be the largest source
These energy sources have increased the carbon foot- of electricity globally.
print, thereby causing climate change.1 Thus, there have RE such as PV, off-shore and on-shore wind, biomass,
been several governmental policies, researches, and pro- solar thermal, and geothermal have found commercial
grammes that are targeted at reducing greenhouse gases applications in the generation of electricity worldwide.1,206
(GHGs) released into the atmosphere.198,199 Beyond emis- Meanwhile, intermittent RE sources such as wind and
sions reduction, the desire to preserve our planet for future solar are expected to take the most share of it. This antic-
generations has made interest in RE sources a global ipated large-scale RE penetration makes it necessary to
issue.200 To achieve this aim, including RE generating units consider them in long-term generation planning. In order
16 BABATUNDE ET AL.

to increase the RE share, some government incentives and reinforcement of the transmission infrastructure. The
such as carbon taxes,subsidies on RETs, feed-in-tariffs, and inclusion of RETs with economical ESS can mitigate the
tax credits for RE generation have been put in place.35,197 challenges of uncertainty and variability.217 ESS is required
Traditional GEP models have mostly focused on the con- at various locations on the power system network because
ventional plant whose operation and planning can be eas- both the demand and the grid connected intermittent
ily carried out by varying fuel inputs to match variability RES are usually characterised by hourly, daily, and sea-
on the load side. Meanwhile, the addition of intermittent sonal variations. Its inclusion will cancel out the mismatch
renewable sources in power generation capacity planning between the generators and consumption and/or to store
creates challenges with respect to variability on the supply the excess energy from renewable sources for future use
side. This is because intermittent RE sources are charac- during the period of generation deficient.213 ESSs are intro-
terised by fluctuations in power output due to frequent duced to the utility to achieve peak shaving, provide sys-
changes in wind speed and solar irradiance, which may tem reserve, enhance system load factor, and to effectively
not be anticipated and controlled. As a result, the addi- reduce total generation cost.218 Other advantages of ESS
tion of RE into the power plant mix increases system include power quality improvement, mitigation of envi-
insecurity and uncertainty. According to Moreira et al207 ronmental impact, congestion management, and system
to address the variability of such resources under tight balancing.219 ESS is either classified based on storage dura-
security criteria, the system might require additional lev- tion (short and long term), or the storage medium.115,206,216
els of quick-response reserves to circumvent contingen- The long-term ESS is a centralised relatively large storage
cies under the uncertainty of intermittent injections of installation suitable for storing large amounts of energy
the renewable sources. Another way to mitigate this chal- in the range of a few MW to hundreds of MW. Such ESS
lenge is the adoption of storage units.1 Various authors can supply energy with more than 8 hours discharge dura-
have included the integration of RE sources in GEP tions. Examples of such systems are pumped hydraulic
problem.44,49-51,59,61,97,99,104,147,170,208-214 From the reviewed lit- storage, large-scale batteries, compressed air energy stor-
erature, the use of wind and solar PV in GEP studies age, and pumped hydro storage.115,206,216 Short-term ESS
has received considerable attention. This is as a result of can only store energy for fewer hours with smaller energy
their widespread resource availability and state-of-the-art storage capacity. Examples of short term ESS technolo-
technological advancements. gies include flywheel energy storage, battery ESS, super-
conducting magnetic energy storage, supercapacitor, and
ultracapacitor energy storage.115,206,216 Depending on the
5.2.1 Application of ESS in GEP storage medium, ESSs are categorised either as mechani-
In order to decarbonize and ensure resource sustainabil- cal, chemical, or electrical storage systems.220 A survey of
ity in the power sector across the globe, the integration the literature shows that few studies have integrated ESS
of higher variable RE and minimisation of conventional in GEP models.51,54,62,65,67,115,211,213,218,221-224
fossil fuel offers a promising solution. RE sources such The inclusion of ESS technologies in GEP models cre-
as wind and PV have offered promising applications in ates a challenge because most GEP models make use of
large-scale electricity generation, due to advancement of load duration curves. However, in order to capture the
the technologies and the drop in investment costs. They operational details of the ESS, a chronological load data
have enormous tendency to shrink the reliance on fos- is inevitable.218 Hence, the fundamental challenge with
sil fuels and GHG emissions in the electricity industry.215 the use of load duration curves involves the determina-
However, the inclusion of such quantity of intermittent tion of the actual or approximate chronological time when
source of RE into the power system network presents sev- energy is scheduled to be stored in or withdrawn from
eral technical and operational challenges. RES such as the storage facility. According to Kandil et al,218 “ adjacent
wind and solar are intermittent, difficult to predict, and points on a load duration curve could represent an hour
deliver fluctuating power output. For example, the wind from the middle of the night on a weekday and a daytime
speed and solar irradiance are usually spatially distributed hour at the weekend.” Furthermore, if a ESS is assumed
and most times does not correlate in time with the load to store energy daily, it may store energy at one point of
profiles. Consequently, this presents a high level of uncer- the load duration curve and generate at the next. As a gen-
tainty and variability in GEP problem formulation and eral rule, the ESS is used at the times of moderately high
solution. Hence, in order to effectively incorporate large energy demand while the ESS is expected to store energy at
scale intermittent renewable, the issues of flexibility must times of moderately low energy demand. However, these
be considered.216 According to Haas et al some of the ways assumptions must be adopted with cautiousness when
to add the required flexibilities is by using flexible genera- estimating the benefits of energy storage facilities adoption
tion technologies such as gas turbines, integration of ESSs, over a long-term period. This is because such assumptions
BABATUNDE ET AL. 17

may lead to overestimation of the size of new energy stor- integrates UC and GEP as a mixed-integer optimisation
age facilities or the operation of the existing storage facil- problem. It was shown that the neglect of UC can result
ities beyond their feasible operating region. By so doing, in suboptimal capacity mix with significant increase in
the operational costs are underestimated. Hence, to effec- the operating cost and carbon emission. Koltsaklis and
tively integrate ESS into the generation expansion plan, a Georgiadis18 also proposed a generic mixed-integer pro-
high temporal resolution dimensioning is essential. gramming GEP model that incorporates UC framework.
The model was validated using the Greek power network.
Results show that improvement in the optimal power
5.2.2 Short-term consideration in GEP mix and the system marginal price evolution could be
With an increased environmental concerns, more RE is achieved. Furthermore, Pereira et al23 presented a GEP
being integrated into the electricity grid. Due to the fact model that integrated the effects of RES variability on
that most RE sources are intermittent in nature, it is impor- thermal plant efficiency. The model also takes into con-
tant to study and integrate the operational features of sideration an hourly UC problem. A novel model that
generating plants (dispatch and UC). This will help in combines both GEP and short-term operation (UC and
designing a flexible protocol for the short-term dispatch of economic dispatch) of power system was presented by
generators to limit unavailability that may be caused by the Wierzbowski et al.229 The study examined the effect of
intermittent nature of the RE sources. Most of the avail- both UC and economic dispatch on the optimal power
able GEP models did not consider the effect of short-term plant mix. The results revealed the importance of gas
operation of generators on long term GEP problems. This technology in the future energy mix. In a similar study,
is evident in the quantum of research paper available on Flores-Quiroz et al22 also proposed the integration of UC
the subject matter. Most traditional GEP models oversim- constraint in the GEP model. Because of the complex and
plified short-term operational details of generating plants large nature of the resulting problem, the authors proposed
as well as emission dispatch. Dynamic details such as a novel DWD and a column generation approach to reduce
generator ramping rates, start-up and short-down times, the computational burden and overcome intractability.
minimum up/down time, spinning reserves, synchroni-
sation soak, and desynchronisation have been neglected
in most studies.26,35,52,61,169 These details and other opera-
tional characteristics are deemed to be predictable with 5.3 Integration of DSM in GEP models
unit production cost and generation mix, experiencing no A report compiled by the International Energy Agency in
significant change from 1 year to another. Hence, tra- 2007 identified fossil fuels as the major source of energy
ditional GEP models adopt a cost model presented as a consumed globally.230 It accounted for at least 80% of the
product of a discounted linear operational cost and the energy mix globally, making it rank first.230 Apart from
power output of the generators.225 This is only applica- being a source of GHG that causes climate change, the sus-
ble in cases where the output of the generators can be tainability of these sources of energy is a threat that needs
controlled and the demand profiles can be accurately pre- to be addressed. Although the role of energy in economic
dicted based on historical data.35 With a gradual rise in growth has been widely recognized, it is essential to man-
RE penetration in the energy mix and the growing adop- age the energy consumption in a sustainable manner.231,232
tion of DSR programmes, the GEP problem is expected To reduce emissions, it is anticipated that by 2100, the
to incorporate UC and economic dispatch strategies in major share of global energy mix will come from nuclear
a computationally traceable manner. This will encourage and RE and less oil and coal.233 Apart from this, several
flexibility and controllability. A review of related GEP lit- frameworks have been proposed to combat GHG emis-
eratures indicates that only few of them have incorporated sions. Some of these include REDD+ (reducing emissions
short-term operational characteristics of generating units from deforestation and degradation),234 setting emission
in GEP problem.18,22,23,102,202,226-229 Hua et al226 presented limits, investment in clean energy economy (DSM and
the operational flexibility in a GEP problem using con- other energy efficiency techniques). Through the imple-
vex relaxation of the UC problem to ensure tightness and mentation of DSM activities, reduction in electricity use,
tractability of the proposed model. The model embedded a electricity cost, and emissions related to the energy use
convex operational problem, which was later solved using can be achieved. DSM can also play a significant role in
Benders decomposition algorithm. The work of Palmintier the integration of intermittent RE sources, such as wind
and Webster227 was able to show that the inclusion of and solar, into the power system network. This approach
short-term operational characteristics of power plant in will be vital in the decarboonisation of the power system
GEP problem can significantly change the energy produc- industry. Examples of DSM are the application of energy
tion and the optimal capacity mix. Their model efficiently efficiency and conservation. The enthusiasm behind the
18 BABATUNDE ET AL.

adoption of DSM is differ for the various stakeholders these targets offer clear strategies and goals for all coun-
involved. The consumers see DSM as a means to reduce tries to embrace based on their own priorities and in the
the amount of energy they consume. This consequently wake of the prevailing environmental challenges of the
reduces the money spent on offsetting electricity bills. On world at large. It is an inclusive agenda that attempts to
the other hand, for utility companies, the reduction or shift address the sources of poverty. Goal number 7 identifies
of consumer demand results in shelving additional gener- the need to “ensure access to affordable, reliable, sus-
ation capacity. Thus, both financial and energy resources tainable and modern energy for all.”236 The aim of this
(fuel) are maximised. DSM therefore has the potential to goal is to ensure a decline in energy poverty as well as
provide substantial environmental, economic, and tech- environmental pollution caused by conventional means
nical (increased reliability) benefits. DSM activities are of power generation. To accomplish this, efforts must be
categorized as either energy reduction or load manage- geared towards embracing energy efficiency and increas-
ment programmes.21 According to Karunanithi et al,21 “a ing international collaboration to enable more open access
realistic power system planning needs integration of both to green energy technology and increased investment in
DSM and Supply Side Management (SSM) which involve RE. Based on the aforementioned goal, the theme of sus-
simultaneous consideration of both quantitative and qual- tainability is essential in GEP. Sustainability with respect
itative issues like plant mix, costs and reliability of power to GEP ensures that the energy demands of the present
supply.” Martins et al235 presented a multiple-objective LP are met without encroaching into the resource meant for
model for power GEP, which incorporated DSM. the future. A sustainable GEP scheme is expected to offer
Their study modelled DSM in form of a new gener- both social and economic opportunities and guarantee that
ating plant along with the other generating alternatives the energy demand is adequately satisfied with little or
from the supply side. Antunes et al105 also presented no environmental consequences. Hence, sustainability in
a multiple-objective mixed-integer LP (MOMILP) model GEP will ensure an optimal Pareto is achieved among the
that provides decision support in the assessment of social, technical, economic, environmental, and the policy
power generation capacity expansion policies. The model aspects of capacity planning.237-239
included the effect of DSM on the power plant mix. Their According to Aghaei et al104 “no single Pareto optimal
model embedded investment and operation costs associ- solution is adequate to be claimed as an optimum solu-
ated with DSM and the costs of lost revenues due to DSM tion of a problem with multiple conflicting objectives.”
actions. Some other studies have also included the aspect In literature, MCDM tools have been extensively engaged
of DSM in GEP model.12,21,93,104 in the selection of optimal generation mix from multiple
energy mix alternatives in the presence of several conflict-
ing criteria.106 Some of the criteria and subattributes used
6 SUSTA INABILITY ISSUES IN G EP in the literature have been presented in Table 4. Accord-
ing to sustainable planning framework, the assessment of
The United Nations have highlighted 17 sustainable devel- the five aspects Social (S), Technical (T), Economic (E),
opmental goals with 169 targets to make the right choices Environment (E), and Policy (P) are very important.240
for the present needs of this generation. These goals have While this method (STEEP) is effective for ensuring sus-
the capacities to improve life in a sustainable manner tainability in GEP studies, we propose the integration of
and to secure the future generation. On the other hand, the SWOT framework (strength, weakness, opportunity,

TABLE 4 Sustainability evaluation criteria for energy studies

Criteria Subcriteria
Social Political acceptance, Social acceptance, Energy
justice, Social benefits, Job opportunities, etc.
Technical Feasibility, Technical know-how, Uncertainties,
Duration of implementation, Reliability, etc.
Economic Interest rates, Inflation, Investment cost, Business
model, Availability of funds, Profitability, Taxation
regime, Electricity demand growth rates,
Economic value, etc.
Environmental Noise, Extreme weather/climate, Natural disaster,
Emission, Need for waste disposal, Land
requirements, etc.
Policy National, regional and global policies, Legislations,
Government support, etc.
BABATUNDE ET AL. 19

FIGURE 9 Proposed STEEP-cum SWOT framework for generation expansion planning (GEP) sustainability [Colour figure can be viewed
at wileyonlinelibrary.com]

and threats) with the STEEP criteria (Figure 9) so as to ity shortages in the past, Eskom's inflexible construction
provide a detailed analysis of different GEP alternatives in programme has caused a substantial and upward sur-
order to arrive at a best compromise choice. plus of costly generation capacity. Hence, a well-executed
implementation of GEP plays a crucial role in assuring
energy security for a country. For example, the electric-
7 P O L I C Y D ISCU SSION S O N ity crisis that occurred in South Africa in 2008, which
IMPLEMENTAT ION O F G EP: SOUTH led to enormous blackouts, load shedding, and massive
A F RI C A C A SE ST U DY financial losses,242,243 was occasioned by policy inadequa-
cies and lack of government will to identify projected
Based on the summary presented in Section 1, the quan-
demand beforehand when projected demand was going
tum of studies as it pertains to GEP has received very
to exceed supply capacity. This is also at variance with
little research attention in Africa. In order to contribute to
faulty expansion policies in the 1990s that led to the
the literature related to GEP studies in Africa, this study
mothballing of the Komati, Camden, and Grootvlei power
chooses a case study on GEP policy discussion as it relates
stations.244
to South Africa. This case study was selected because South
Africa possesses a world class power utility whose sustain-
ability is currently being threatened with issues bordering
around expansion planning.241 This section thus briefly 7.1 Policy discussion on pricing
highlights the implications of a well-executed GEP on elec- Electricity pricing plays a crucial role for any power
tricity pricing, sizing of RMs, minimisation of operations project. Pricing is usually subject to anticipated costs and
and maintenance costs, supply capacity utilization, and expenditure of the utility in expanding supply capacity,
energy poverty mitigation as it pertains to South Africa. improving transmission network and associated costs like
Presently, Eskom, is facing a record period of decrease insurance and overhead. For example, Eskom's electric-
in electricity consumption, while having concurrently ity pricing is majorly influenced by its estimated capital
invested (huge amount) in coal-fired power generation expenditure on new constructions, maintenance, oper-
facilities (Kusile 4 800 and MW Medupi 4 764 MW).These ations, overhead, insurance, and other related running
investments have been afflicted with delays, interrup- costs. According to Eskom,245 there was a revenue shortage
tions, and sunk investments. Consequently, Eskom has of about R35 billion for 2014/2015 due to low tariff. This
been forced to effect the country's highest electricity tariff implies that while Eskom aims at maximising revenue
increases in recorded history. Having experienced capac- accumulated through higher tariffs, the resulting surge
20 BABATUNDE ET AL.

TABLE 5 Comparison of researches

Optimisation Network Unit
Reference Objective Approach Topology Reliability RET Uncertainty Storage DSM Commitment
Ozcan et al208 SO HUO SNT ✓ ✓
Farghal and Aziz170 SO CLO SNT ✓
Ramos et al38 SO CLO SNT
Mo et al37 SO CLO SNT ✓
Kandil et al218 SO CLO SNT ✓ ✓
Climaco et al111 MO CLO SNT ✓
Sasaki et al249 SO HUO SNT
Kim and Ahu113 MO CM SNT ✓
Handke et al250 MO CLO SNT
Fukuyama and Chiang151 SO HUO SNT
Yasuda et al221 SO CLO SNT ✓
Park et al127 SO HUO SNT
Chuang et al16 SO CLO SNT ✓
Firmo and Legey152 SO Both SNT ✓ ✓
Park et al91 SO HUO SNT
Min et al251 SO CLO SNT ✓
Kannan et al159 SO HUO SNT ✓
Lin et al252 SO HUO SNT
Kannan et al29 SO HUO SNT ✓
Slochanal et al92 SO HUO SNT
Yamgar et al209 SO CM SNT ✓
Moghaddam et al93 MO CLO SNT ✓
Kim et al94 SO CLO SNT ✓
Oatman and Hamant222 SO CLO SNT ✓
Kannan et al95 SO HUO SNT ✓
Nualhong et al210 SO HUO SNT ✓ ✓
Marcato et al253 SO HUO SNT
Sirikum et al7 SO HUO SNT ✓
Nordlund et al211 SO CLO SNT ✓ ✓
Kothari and Kroese9 SO CLO SNT ✓
Parsaeifard et al96 SO CLO SNT ✓
Kamalinia and SO CLO NCT ✓ ✓ ✓ ✓
Shahidehpour212
Phuc et al254 SO CLO SNT
Jalilzadeh et al45 SO HUO NCT
Pereira and Saraiva97 SO HUO SNT ✓ ✓ ✓
Careri et al147 SO CLO SNT ✓
Yaghooti et al255 SO CLO NCT ✓
Alizadeh and Jadid81 SO CLO NCT ✓
Palmintier and Webster227 SO CLO SNT ✓
Aghaei et al104 MO CLO SNT ✓ ✓ ✓
Hu and Jewell213 SO CLO NCT ✓ ✓
Dehghan et al256 SO CLO SNT ✓
Zhang et al214 MO HUO SNT ✓
Aghaei et al86 SO CLO NCT ✓ ✓
Sadeghi et al99 SO HUO SNT ✓
Khan et al257 SO HUO SNT ✓
Gil et al258 SO CLO SNT ✓
Palmintier102 SO CLO SNT ✓ ✓
Bhuvanesh et al163 SO HUO SNT ✓ ✓
Kim et al63 SO CLO NCT
Coit et al114 MO CLO SNT ✓
Shengyu et al259 SO CLO SNT ✓ ✓
Park et al150 SO HUO SNT
Dang et al260 SO CLO SNT ✓ ✓ ✓
Saboori and Hemmati115 MO HUO SNT ✓ ✓ ✓ ✓
BABATUNDE ET AL. 21

TABLE 5 Continued
Reference objective Optimisation Network Reliability RET Uncertainty Storage DSM Unit
approach Topology commitment
van Bracht et al67 SO CLO SNT ✓ ✓
Poncelet et al223 SO Both SNT ✓
Manabe et al100 SO CLO SNT
Tohidi and Hesamzadeh98 SO CLO NCT
Hu et al261 SO CLO NTC ✓
Mavalizadeh et al116 MO CLO NCT ✓ ✓
Hinojosa and SO CLO SNT ✓
Gonzalez-Longatt262
Kendziorski et al263 SO CLO SNT ✓ ✓
Hua et al226 SO CLO SNT ✓
Rajabani et al4 MO CLO SNT
Sirikum and SO HUO SNT ✓ ✓
Techanitisawad12
Kazempour et al15 SO CLO NCT
Wang et al17 SO HUO NCT
Koltsaklis and Georgiadis18 SO CLO SNT ✓ ✓ ✓
Botterud et al19 SO CLO SNT ✓
Karunanithi et al21 SO CM SNT ✓ ✓ ✓ ✓
Flores-Quiroz et al22 SO CLO SNT ✓ ✓
Pereira et al23 MO CLO SNT ✓ ✓
Promjiraprawat and MO CLO SNT ✓ ✓
Limmeechokchai24
Hamam et al26 SO CLO SNT
Pozo et al39 SO CLO NCT
Meza et al40 MO CLO NCT ✓
Chen et al41 SO CLO SNT ✓
Aghaei et al42 SO Both SNT ✓
Tekiner et al43 MO CLO SNT ✓ ✓
Park and Baldick44 SO CLO SNT ✓
Mirzaesmaeeli et al46 SO CLO SNT ✓
Aghaei et al48 MO CLO SNT ✓ ✓
Rajesh et al49 SO HUO SNT ✓ ✓
Rajesh et al50 SO HUO SNT ✓ ✓
Rajesh et al51 SO HUO SNT ✓ ✓ ✓
Hemmati et al54 SO HUO SNT ✓ ✓
Karthikeyan et al55 SO HUO SNT ✓
Deossa et al56 SO CLO SNT
Zhang et al57 SO CLO SNT ✓
Bakirtzis et al58 SO CLO SNT ✓
Li et al59 SO CLO SNT ✓
Sharan and SO CLO NCT
Balasubramanian60
Hemmati et al61 SO HUO NCT ✓ ✓
Kim and Kim62 SO CLO NCT
Kim et al63 SO CLO NCT
Barati et al64 SO CLO NCT
Thomé et al65 SO CLO SNT ✓
Gorenstin et al66 SO CLO SNT
Wierzbowski et al229 SO CLO SNT ✓ ✓ ✓

Abbreviations: CLO, classical optimisation; CM, commercial software; HUO, heuristic method; MO, multiple objective; NCT, network constrained; SNT, single
node; SO, single objective; Both means CLO and HUO.

in electricity tariff is capable of triggering energy poverty and precipitate poverty (declining electricity consumption
as a result of households expending more money in pur- has been established).244 GEP execution must thus con-
chasing lesser energy units. This has the propensity to sider the effect of expansion costs and resulting pricing
lower electricity consumption, impede economic growth, scenarios on households especially vulnerable households.
22 BABATUNDE ET AL.

7.2 Policy discussion on reserve sizing 8 CO MPARISON OF RESEARCHE S

In South Africa, the projected addition to supply capac-
Table 5 shows a summary of the issues considered in
ity between 2017 and 2024 is over 500% of net demand
some of the reviewed literature. The evaluation is based
increase to the grid within the same period.244 As a result,
on various perspectives, and a complete comparison is
Eskom risks incurring more revenue shortfall due to a
carried out. The various perspectives considered include
rise in operational losses caused by the underutilization of
objective function, solution methodology, network topol-
installed capacity, increasing operations and maintenance
ogy, reliability, RE inclusion, uncertainty, inclusion of
costs and reduced revenue owing to decreased electricity
storage facility, DSM, and UC. From Table 5, it is seen
units purchases. An efficient GEP must thus incorporate
that the inclusion of energy storage, UC, and DSM is
initiatives like demand side management that allow for
yet to receive adequate attention. Most models available
some operational control of end-use load. With increas-
from the reviewed literature were presented as a single-
ing participation of flexible loads, the utility is able to
objective optimisation problem without the inclusion of
minimise RMs without compromising on security.
RE resources.

7.3 Policy discussion on operations cost

minimisation 9 CO NCLUSION AND FUTURE
DIRECTIONS
In addition to traditional associated costs such as mainte-
nance, fuel, and operations costs, emissions cost is playing In the past 7 decades, the GEP problem has been studied
a prominent role in the incurred costs for power sta- by various researchers and DMs. It has since evolved from
tions. In South Africa, for example, a proposed carbon tax a simple single-objective optimisation problem of total cost
of ZAR120/tCO2 energy equivalent by National Energy minimisation to a more complex multiobjective optimisa-
Regulator of South Africa (NERSA) was expected to add tion formulation. This has resulted in new trends. Recent
about R11 billion to Eskom's expenses246 from 2015. A trends in GEP studies have taken into consideration uncer-
well-executed GEP thus provides avenue to exploit RE tainties, RE penetration, energy storage, deregulation, and
sources such as wind and solar with increasing partici- sustainability issues. The combined inclusion of these top-
pation of flexible loads. An increase in the utilisation of ics in GEP studies has increased the complexities of its
RE sources would lead to a reduction in overall plant's formulation and hence its solution algorithm. Reviewing
operations costs. up-to-date research in this subject can substantially lead to
the identification of research gaps to unlock future studies.
This study has presented a state-of-the-art comprehensive
7.4 Policy discussion on supply capacity review of GEP from various thematic perspectives. It also
utilisation and energy poverty mitigation made attempt to categorise the GEP models based on mar-
ket structure, time horizon, uncertainties, and network
With increasing flexible loads, GEP can allow for higher
topology. It is observed that over the years, the available
utilisation of plant's installed capacity. This is due to its
GEP models have constantly evolved and will continue to
ability to dispatch flexible loads during periods of low base
evolve because of the constant changes in technological
loads to achieve a smooth operations profile. The poverty
advancements and government policies.
rate in South Africa is over 50%.247 A direct inference on
Although outstanding achievements have been recorded
this can be gleaned from the declining electricity consump-
in the formulation and solution approaches for GEP
tion in households despite increasing investments in elec-
problem, further improvements could still be achieved
tricity capacity expansion.244,248 However, with increased
in the face of a constantly evolving RE penetration and
flexible loads and increased utilisation of RE sources, the
environmental policies. In the context of energy demand,
utility can incorporate dynamic pricing schemes that can
flexibility, optimisation approach, and sustainability, the
offer households considerable reduction in monthly elec-
following suggestions can be considered as future studies.
tricity bills (lowering household energy poverty). Any sav-
ings accrued through participation in demand response 1. Future studies can explore the formulation of a
can be used in either extending the operation time of elec- robust GEP model that takes its annual energy
trical appliances that can contribute to households quality demand input from a dynamic energy demand fore-
of life (lighting, entertainment, heating, and cooking) or casting model.
engage in other activities that are capable of improving 2. Due to the inclusion of intermittent RE sources, the
households quality of life (QoL). traditional GEP model is no longer sufficient to han-
BABATUNDE ET AL. 23

dle the issues of operational flexibility with respect to Algorithms and Solutions. Springer-Verlag Berlin Heidelberg:
supply and demand variability. Models that included Springer Science & Business Media; 2011.

details of uncertainty margins, short-term demand 6. Hemmati R, Hooshmand R-A, Khodabakhshian A. Compre-
hensive review of generation and transmission expansion plan-
response, power plant decommissioning , and energy ning. ET Gener Transm Distrib. 2013;7(9):955-964.
storage will provide the prospect of handling opera- 7. Sirikum J, Techanitisawad A, Kachitvichyanukul V. A new
tional flexibility challenges. efficient GA-benders' decomposition method: for power gener-
3. Due to the inclusion of flexibility in a GEP model, ation expansion planning with emission controls. IEEE Trans
Power Syst. 2007;22(3):1092-1100.
the sizes of such problem become large and as
8. Hamam YM, Hindi KS. Optimized design of low-voltage distri-
such the memory size and computational efforts bution networks a comprehensive algorithm. IEEE Trans Syst
required. Therefore, a computationally efficient solu- Man Cybern. 1987;17(3):502-507.
tion algorithm is inevitable. Hybridisation of opti- 9. Kothari RP, Kroese DP. Optimal generation expansion plan-
misation techniques to ease computational burden ning via the cross-entropy method. In: Winter Simulation Con-
ference Winter Simulation Conference. Austin, Texas, USA;
could be investigated. Further studies can also con- 2009:1482-1491.
sider investigating the impact of linearisation tech- 10. Olanrewaju O, Jimoh A, Kholopane P. Integrated IDA–
niques on GEP optimisation problems. ANN–DEA for assessment and optimization of energy con-
4. Finally, the subject of sustainability in GEP is also sumption in industrial sectors. Energy. 2012;46(1):629-635.
essential. It will ensure that a balance is achieved 11. Masse P, Gibrat R. Application of linear programming to
investments in the electric power industry. Manage Sci.
among the technical, economic, environmental,
1957;3(2):149-166.
resource, and the socio-political aspects of GEP.
12. Sirikum J, Techanitisawad A. Power generation expan-
Subsequent research on sustainability issues in gen- sion planning with emission control: a nonlinear model
eration expansion will provide more insights into and a GA-based heuristic approach. Int J Energy Res.
alternatives for achieving sustainable capacity 2006;30(2):81-99.
expansion. 13. Conejo AJ, Morales L, Kazempour SJ, Siddiqui AS. Invest-
ment in Electricity Generation and Transmission. New York:
Springer; 2016.
14. Kaymaz P, Valenzuela J, Park CS. Transmission congestion and
ACKNOWLEDGEMENT competition on power generation expansion. IEEE Trans Power
The financial assistance of the National Research Foun- Syst. 2007;22(1):156-163.
dation (NRF) through the DST-NRF-TWAS doctoral fel- 15. Kazempour SJ, Conejo AJ, Ruiz C. Strategic generation invest-
ment using a complementarity approach. IEEE Trans Power
lowship towards this research is hereby acknowledged. Syst. 2011;26(2):940-948.
Opinions expressed and conclusions arrived at are those 16. Chuang AS, Wu F, Varaiya P. A game-theoretic model for
of the authors and are not necessarily to be attributed to generation expansion planning: Problem formulation and
the NRF. numerical comparisons. IEEE Trans Power Syst. 2001;16(4):
885-891.
17. Wang J, Shahidehpour M, Li Z, Botterud A. Strategic gener-
ORCID ation capacity expansion planning with incomplete informa-
tion. IEEE Trans Power Syst. 2009;24(2):1002-1010.
Olubayo M. Babatunde https://orcid.org/ 18. Koltsaklis NE, Georgiadis MC. A multi-period, multi-regional
0000-0001-6533-7171 generation expansion planning model incorporating unit com-
mitment constraints. Appl Energy. 2015;158:310-331.
19. Botterud A, Ilic MD, Wangensteen I. Optimal investments in
REFERENCES power generation under centralized and decentralized decision
1. Akinbulire TO, Oluseyi PO, Babatunde OM. Techno-economic making. IEEE Trans Power Syst. 2005;20(1):254-263.
and environmental evaluation of demand side management 20. Baringo L, Conejo AJ. Risk-constrained multi-stage wind
techniques for rural electrification in ibadan, nigeria. Int J power investment. IEEE Trans Power Syst. 2013;28(1):401-411.
Energy Environ Eng. 2014;5(4):375-385. 21. Karunanithi K, Saravanan S, Prabakar B, Kannan S, Thangaraj
2. IAEA. Expansion planning for electrical generating systems: A C. Integration of demand and supply side management strate-
guidebook. Tech. rep., Vienna, International Atomic Energy gies in generation expansion planning. Renewable Sustainable
Agency–IAEA; 1984. STI/DOC/10/241 88. Energy Rev. 2017;73:966-982.
3. Sen R, Bhattacharyya SC. Renewable energy-based mini-grid 22. Flores-Quiroz A, Palma-Behnke R, Zakeri G, Moreno R. A col-
for rural electrification: case study of an indian village. umn generation approach for solving generation expansion
Mini-Grids for Rural Electrification of Developing Countries. planning problems with high renewable energy penetration.
Cham Vietnam: Springer; 2014:203-232. Electric Power Syst Res. 2016;136:232-241.
4. Rajabani N, Mehregan MR, Taghizadeh MR. Developing a 23. Pereira S, Ferreira P, Vaz A. Generation expansion planning
multi-objective mathematical model for power generation with high share of renewables of variable output. Appl Energy.
expansion planning in iran. Manage Administrative Sci Rev. 2017;190:1275-1288.
2014;3(7):1116-1125. 24. Promjiraprawat K, Limmeechokchai B. Multi-objective and
5. Seifi H, Sepasian MS. Electric Power System Planning: Issues, multi-criteria optimization for power generation expansion
24 BABATUNDE ET AL.

planning with CO2 mitigation in thailand. Songklanakarin J Sci 43. Tekiner H, Coit DW, Felder FA. Multi-period multi-objective
Technol. 2013;35(3). electricity generation expansion planning problem
25. Phupha V, Lantharthong T, Rugthaicharoencheep N. Genera- with Monte-Carlo simulation. Electric Power Syst Res.
tion expansion planning strategies on power system: a review. 2010;80(12):1394-1405.
Int J Ind Manuf Eng. 2012;6(4):436-439. 44. Park H, Baldick R. Multi-year stochastic generation capacity
26. Hamam Y, Renders M, Trecat J. Partitioning algorithm for expansion planning under environmental energy policy. Appl
the solution of long-term power-plant mix problems. Proc Inst Energy. 2016;183:737-745.
Electr Eng. 1979;126(9):837-839. 45. Jalilzadeh S, Shabani A, Azadru A. Multi-period generation
27. Hamam Y, Renders M, Trecat J. Fast algorithms for solving expansion planning using genetic algorithm. In: 2010 Inter-
long term power plant mix problems. In: Proceedings of the national Congress on Ultra Modern Telecommunications and
Seventh Power Systems Computation Conference: Lausanne Control Systems and Workshops (ICUMT). Moscow, Russia:
12-17 July 1981. Lausanne, Swizerland: Westbury House; IEEE; 2010:358-363.
1981:110. 46. Mirzaesmaeeli H, Elkamel A, Douglas PL, Croiset E, Gupta
28. Hobbs BF. Optimization methods for electric utility resource M. A multi-period optimization model for energy plan-
planning. Eur J Operational Res. 1995;83(1):1-20. ning with co 2 emission consideration. J Environ Manage.
2010;91(5):1063-1070.
29. Kannan S, Slochanal SMR, Padhy NP. Application and com-
parison of metaheuristic techniques to generation expansion 47. Kalika V, Frant S. Methodology of power generation sys-
planning problem. IEEE Trans Power Syst. 2005;20(1):466-475. tem planning: multicriteria optimization accounting for uncer-
tainty factors. In: Nineteenth Convention of Electrical and
30. Kagiannas AG, Askounis DT, Psarras J. Power generation plan-
Electronics Engineers in Israel. Jerusalem, Israel, Israel: IEEE;
ning: a survey from monopoly to competition. Int J Electr Power
1996:91-93.
Energy Syst. 2004;26(6):413-421.
48. Aghaei J, Akbari MA, Roosta A, Baharvandi A. Multiobjec-
31. Zhu J, Chow Mo-yuen. A review of emerging techniques
tive generation expansion planning considering power system
on generation expansion planning. IEEE Trans Power Syst.
adequacy. Electric Power Syst Res. 2013;102:8-19.
1997;12(4):1722-1728.
49. Rajesh K, Kannan S, Thangaraj C. Least cost generation expan-
32. Babatunde O, Munda J, Hamam Y. Generation expansion plan-
sion planning with wind power plant incorporating emission
ning: a survey. In: 2018 IEEE PES/IAS PowerAfrica. Cape
using differential evolution algorithm. Int J Electr Power Energy
Town, South Africa: IEEE; 2018:307-312.
Syst. 2016;80:275-286.
33. Singh B, Pal C, Mukherjee V, Tiwari P, Yadav MK, Member
IS. Distributed generation planning from power system perfor- 50. Rajesh K, Bhuvanesh A, Kannan S, Thangaraj C. Least
mances viewpoints: a taxonomical survey. Renewable Sustain- cost generation expansion planning with solar power plant
able Energy Rev. 2017;75:1472-1492. using differential evolution algorithm. Renewable Energy.
2016;85:677-686.
34. Hemmati R, Hooshmand R-A, Khodabakhshian A.
State-of-the-art of transmission expansion planning: com- 51. Rajesh K, Karthikeyan K, Kannan S, Thangaraj C. Genera-
prehensive review. Renewable Sustainable Energy Rev. tion expansion planning based on solar plants with storage.
2013;23:312-319. Renewable Sustainable Energy Rev. 2016;57:953-964.
35. Oree V, Hassen SZS, Fleming PJ. Generation expansion plan- 52. Rajesh SKK, Kannan K, Karuppasamypandian M. Generation
ning optimisation with renewable energy integration: a review. capacity expansion planning with solar power plant incorpo-
Renewable Sustainable Energy Rev. 2017;69:790-803. rating emission. VFSTR J STEM. 2015;1(2):2062-2455.
36. Sadeghi H, Rashidinejad M, Abdollahi A. A compre- 53. Karapidakis ES, Katsigiannis YA, Georgilakis PS, Thalassi-
hensive sequential review study through the generation nakis E. Generation expansion planning of crete power system
expansion planning. Renewable Sustainable Energy Rev. for high penetration of renewable energy sources. Mater Sci
2017;67:1369-1394. Forum. 2011;670:407-414.
37. Mo B, Hegge J, Wangensteen I. Stochastic generation expan- 54. Hemmati R, Saboori H, Jirdehi MA. Multistage generation
sion planning by means of stochastic dynamic programming. expansion planning incorporating large scale energy stor-
IEEE Trans Power Syst. 1991;6(2):662-668. age systems and environmental pollution. Renewable Energy.
2016;97:636-645.
38. Ramos A, Perez-Arriaga IJ, Bogas J. A nonlinear programming
approach to optimal static generation expansion planning. 55. Karthikeyan K, Kannan S, Baskar S, Thangaraj C. Application
IEEE Trans Power Syst. 1989;4(3):1140-1146. of opposition-based differential evolution algorithm to genera-
tion expansion planning problem. J Electr Eng Technol (JEET).
39. Pozo D, Sauma EE, Contreras J. A three-level static MILP
2013;8(4):686-693.
model for generation and transmission expansion planning.
IEEE Trans Power Syst. 2013;28(1):202-210. 56. Deossa P, De Vos K, Deconinck G, Espinosa J. Generation
expansion models including technical constraints and demand
40. Meza JLC, Yildirim MB, Masud AS. A model for the multi-
uncertainty. J Appl Math. 2017;2017.
period multiobjective power generation expansion problem.
IEEE Trans Power Syst. 2007;22(2):871-878. 57. Zhang T, Baldick R, Deetjen T. Optimized generation capacity
41. Chen Q, Kang C, Xia Q, Zhong J. Power generation expan- expansion using a further improved screening curve method.
sion planning model towards low-carbon economy and its Electric Power Syst Res. 2015;124:47-54.
application in China. IEEE Trans Power Syst. 2010;25(2): 58. Bakirtzis GA, Biskas PN, Chatziathanasiou V. Generation
1117-1125. expansion planning by MILP considering mid-term scheduling
42. Aghaei J, Roosta A, Akbari MA, Rabiee A, Gitizadeh M. Relia- decisions. Electric Power Syst Res. 2012;86:98-112.
bility constrained multi-period generation expansion planning 59. Li S, Coit DW, Felder F. Stochastic optimization for electric
of electrical energy resources using MILP. Int Trans Electr power generation expansion planning with discrete climate
Energy Syst. 2013;23(7):961-974. change scenarios. Electric Power Syst Res. 2016;140:401-412.
BABATUNDE ET AL. 25

60. Sharan I, Balasubramanian R. Integrated generation and trans- 78. Pereira AJ, Saraiva JT. A decision support system for generation
mission expansion planning including power and fuel trans- expansion planning in competitive electricity markets. Electric
portation constraints. Energy Policy. 2012;43:275-284. Power Syst Res. 2010;80(7):778-787.
61. Hemmati R, Hooshmand R-A, Khodabakhshian A. Coordi- 79. Su J, Wu FF. Evaluation of generation expansion investment
nated generation and transmission expansion planning in under competitive market environment. In: Power Engineer-
deregulated electricity market considering wind farms. Renew- ing Society General Meeting, 2005. San Francisco, USA: IEEE;
able Energy. 2016;85:620-630. 2005:2136-2140.
62. Kim H, Kim W. Integrated optimization of combined 80. Gnansounou E, Dong J, Pierre S, Quintero A. Market oriented
generation and transmission expansion planning consid- planning of power generation expansion using agent-based
ering bus voltage limits. J Electr Eng Technol. 2014;9(4): model. In: 2004 IEEE PES Power Systems Conference & Expo-
1202-1209. sition. New York, NY, USA: IEEE; 2004:1306-1311.
63. Kim H, Lee S, Han S, Kim W, Ok K, Cho S. Integrated gener- 81. Alizadeh B, Jadid S. Reliability constrained coordination of
ation and transmission expansion planning using generalized generation and transmission expansion planning in power sys-
Bender's decomposition method. In: 2015 IEEE International tems using mixed integer programming. IET Gener Transm
Conference on Computational Intelligence & Communication Distrib. 2011;5(9):948-960.
Technology (CICT). Ghaziabad, India: IEEE; 2015:493-497. 82. Khodaei A, Shahidehpour M, Wu L, Li Z. Coordination of
64. Barati F, Nateghi A, Seifi H, Sepasian MS. Generation and short-term operation constraints in multi-area expansion plan-
transmission expansion planning with considering natural gas ning. IEEE Trans Power Syst. 2012;27(4):2242-2250.
network. In: 2013 21st Iranian Conference on Electrical engi- 83. Khodaei A, Shahidehpour M. Microgrid-based co-optimization
neering (ICEE). Mashhad, Iran: IEEE; 2013:1-7. of generation and transmission planning in power systems.
65. Thomé FS, Binato S, Pereira MV, Campodónico N, Fampa IEEE Trans Power Syst. 2013;28(2):1582-1590.
MH, Costa Jr LCd. Decomposition approach for generation 84. Tohidi Y, Aminifar F, Fotuhi-Firuzabad M. Generation expan-
and transmission expansion planning with implicit multipliers sion and retirement planning based on the stochastic program-
evaluation. Pesquisa Oper. 2013;33(3):343-359. ming. Electric Power Syst Res. 2013;104:138-145.
66. Gorenstin B, Campodonico N, Costa J, Pereira M. Power system 85. Rouhani A, Hosseini SH, Raoofat M. Composite genera-
expansion planning under uncertainty. IEEE Trans Power Syst. tion and transmission expansion planning considering dis-
1993;8(1):129-136. tributed generation. Int J Electr Power Energy Syst. 2014;62:
67. van Bracht N, Grote F, Fehler A, Moser A. Incorporating 792-805.
long-term uncertainties in generation expansion planning. In: 86. Aghaei J, Amjady N, Baharvandi A, Akbari M-A. Generation
2016 13th International Conference on the European Energy and transmission expansion planning: MILP–based probabilis-
Market (EEM). Porto, Portugal: IEEE; 2016:1-5. tic model. IEEE Trans Power Syst. 2014;29(4):1592-1601.
68. Hirst E, Schweitzer M. Electric-utility resource planning and 87. Zhan TS, Kao CC, Yang CD, Tsai JI. Optimal generation expan-
decision-making: the importance of uncertainty. Risk Anal. sion planning strategy with carbon trading. World Acad Sci,
1990;10(1):137-146. Eng Technol, Int J Comput, Electr, Autom, Control Inf Eng.
2010;4(5):993-998.
69. Ku A. Modelling uncertainty in electricity capacity planning.
Ph.D. Thesis: University of London; 1995. 88. Gnansounou E, Dong J. Agent-based model for market ori-
ented planning of electricity generation expansion. Tech.
70. Husdal J. Robustness and flexibility as options to reduce uncer-
Rep, Swiss Federal Institute of Technology. Citeseer; 2004.
tainty and risk. Unpublished Course paper. Molde University
College, Molde, Norway; 2004. 89. Nanduri V, Das TK. Game theoretic approach for generation
capacity expansion in restructured power markets. In: 2008
71. Kann A, Weyant JP. Approaches for performing uncertainty IEEE Power and Energy Society General Meeting-Conversion
analysis in large-scale energy/economic policy models. Environ and Delivery of Electrical Energy in the 21st Century.
Model Assess. 2000;5(1):29-46. Pittsburgh, PA, USA: IEEE; 2008:1-3.
72. Seddighi AH, Ahmadi-Javid A. Integrated multiperiod power 90. Zeljko M. Generation expansion planning in the open electric-
generation and transmission expansion planning with sus- ity market. In: International Conference on Deregulated Elac-
tainability aspects in a stochastic environment. Energy. tricity Market Issues in South-Eastern Europe, 22-23 Septem-
2015;86:9-18. ber 2008. Lefkosia, Cyprus; 2008:1-9.
73. Botterud A, Korpås M. A stochastic dynamic model for opti- 91. Park JB, Kim JH, Lee KY. Generation expansion planning in a
mal timing of investments in new generation capacity in competitive environment using a genetic algorithm. In: Power
restructured power systems. Int J Electr Power Energy Syst. Engineering Society Summer Meeting. Chicago, IL, USA, USA:
2007;29(2):163-174. IEEE; 2002:1169-1172.
74. Pereira AJ, Saraiva JT. Generation expansion planning 92. Slochanal SMR, Kannan S, Rengaraj R. Generation expansion
(GEP)—a long-term approach using system dynamics and planning in the competitive environment. In: 2004 Interna-
genetic algorithms (GAs). Energy. 2011;36(8):5180-5199. tional Conference on Power System Technology, 2004. Power-
75. Botterud A, Mahalik MR, Veselka TD, Ryu HS, Sohn KW. con 2004, Vol. 2. Singapore, Singapore: IEEE; 2004:1546-1549.
Multi-agent simulation of generation expansion in electricity 93. Moghaddam MParsa, Sheikh-El-Eslam MK, Jadid S. A MADM
markets. In: 2007. IEEE Power Engineering Society General framework for generation expansion planning in small electric-
Meeting. Tampa, Florida, USA: IEEE; 2007:1-8. ity firms. In: 2005. IEEE Power Engineering Society General
76. Ma X, Zhou Y. Coordination of generation and transmission Meeting. San Francisco, CA, USA: IEEE; 2005:185-189.
planning for power system with large wind farms. Energy Pro- 94. Kim JH, Park JB, Park JK, Joo SK. A market-based analysis
cedia. 2012;16:1979-1985. on the generation expansion planning strategies. In: Proceed-
77. Pineda S, Morales JM, Ding Y, Østergaard J. Impact of equip- ings of the 13th International Conference on Intelligent Sys-
ment failures and wind correlation on generation expansion tems Application to Power Systems. Arlington, VA, USA: IEEE;
planning. Electric Power Syst Res. 2014;116:451-458. 2005:6-pp.
26 BABATUNDE ET AL.

95. Kannan S, Slochanal SMR, Baskar S, Murugan P. Applica- 112. Clímaco J, Henggeler Antunes C, Gomes Martins A, Traca
tion and comparison of metaheuristic techniques to generation Almeida A. A multiple objective linear programming model
expansion planning in the partially deregulated environment. for power generation expansion planning. Int J Energy Res.
IET Gener Transm Distrib. 2007;1(1):111-118. 1995;19(5):419-432.
96. Parsaeifard A, Manbachi M, Kopayi MA, Haghifam M. A 113. Kim YC, Ahu BH. Multicriteria generation-expansion plan-
market-based generation expansion planning in deregulated ning with global environmental considerations. IEEE Trans
environment based on distributed generations development. Eng Manage. 1993;40(2):154-161.
In: 2010 IEEE 11th International Conference on Probabilis- 114. Coit DW, Selcuklu S, Chatwattanasiri N, Wattanapongsakorn
tic Methods Applied to Power Systems (PMAPS). Singapore, N. Stochastic multiple objective electric generation expansion
Singapore: IEEE; 2010:677-684. planning. In: 2015 12th International Conference on Electrical
97. Pereira AJ, Saraiva JT. Building generation expansion plans-a Engineering/Electronics, Computer, Telecommunications and
decision aid approach to use in competitive electricity mar- Information Technology (ECTI-CON) IEEE; 2015; Hua Hin,
kets. In: 7th Mediterranean Conference and Exhibition on Thailand:1-6.
Power Generation, Transmission, Distribution and energy 115. Saboori H, Hemmati R. Considering carbon capture and stor-
Conversion (MedPower 2010); November 2010; Agia Napa, age in electricity generation expansion planning. IEEE Trans
Cyprus:1-9. Sustainable Energy. 2016;7(4):1371-1378.
98. Tohidi Y, Hesamzadeh MR. A mathematical model for strategic 116. Mavalizadeh H, Ahmadi A, Gandoman FH, Siano P, Shayanfar
generation expansion planning. In: Power and Energy Society HA. Multiobjective robust power system expansion planning
General Meeting (PESGM). Boston, MA, USA: IEEE; 2016:1-5. considering generation units retirement. IEEE Syst J. Septem-
99. Sadeghi H, Mohammadian M, Abdollahi A, Rashidinejad M, ber 2018;12(3):2664-2675.
Mahdavi SMahmoud. Renewable-based generation expansion
117. Farghal S, El-Dewieny R, Aziz MRA. Generation expansion
planning considering environmental issues using GSA. In:
planning using the decision tree technique. Electric Power Syst
2014 Iranian Conference on Intelligent Systems (ICIS). Bam,
Res. 1987;13(1):59-70.
Iran: IEEE; 2014:1-6.
118. Yildirim M, Erkan K, Ozturk S. Power generation expansion
100. Manabe Y, Funabashi T, Kato T, Kurimoto M, Suzuoki Y. Prob-
planning with adaptive simulated annealing genetic algorithm.
abilistic investment strategy modeling for generation expan-
Int J Energy Res. 2006;30(14):1188-1199.
sion planning. In: Power Systems Computation Conference
(PSCC). Genoa, Italy: IEEE; 2016:1-8. 119. Jadidoleslam M, Bijami E, Ebrahimi A. Generation expansion
planning by a modified SFL algorithm. Intell Syst Electr Eng.
101. Rebennack S. Generation expansion planning under uncer-
2011;2(1):27-44.
tainty with emissions quotas. Electric Power Syst Res.
2014;114:78-85. 120. Shayanfar H, Lahiji AS, Aghaei J, Rabiee A. Generation expan-
102. Palmintier B. Flexibility in generation planning: Identifying sion planning in pool market: a hybrid modified game the-
key operating constraints. In: Power Systems Computation ory and improved genetic algorithm. Energy Convers Manage.
Conference (PSCC). Wroclaw, Poland: IEEE; 2014:1-7. 2009;50(5):1149-1156.
103. Kazempour SJalal, Conejo AJ. Strategic generation investment 121. Jadidoleslam M, Ebrahimi A. Reliability constrained genera-
under uncertainty via benders decomposition. IEEE Trans tion expansion planning by a modified shuffled frog leaping
Power Syst. 2012;27(1):424-432. algorithm. Int J Electr Power Energy Syst. 2015;64:743-751.
104. Aghaei J, Akbari M, Roosta A, Gitizadeh M, Niknam T. Inte- 122. Chen SL, Zhan TS, Tsay MT. Generation expansion planning of
grated renewable–conventional generation expansion plan- the utility with refined immune algorithm. Electric Power Syst
ning using multiobjective framework. IET Gener Transm Dis- Res. 2006;76(4):251-258.
trib. 2012;6(8):773-784. 123. Karaki SH, Chaaban FB, Al-Nakhl N, Tarhini KA. Power gen-
105. Antunes CH, Martins AG, Brito IS. A multiple objective eration expansion planning with environmental consideration
mixed integer linear programming model for power generation for lebanon. Int J Electr Power Energy Syst. 2002;24(8):611-619.
expansion planning. Energy. 2004;29(4):613-627. 124. Javadi MS, Saniei M, Mashhadi HR, Gutiérrez-Alcaraz G.
106. Pohekar S, Ramachandran M. Application of multi-criteria Multi-objective expansion planning approach: distant wind
decision making to sustainable energy planninga review. farms and limited energy resources integration. IET Renew
Renewable Sustainable Energy Rev. 2004;8(4):365-381. Power Gener. 2013;7(6):652-668.
107. Hwang CL, Masud ASM. Multiple Objective Decision Making- 125. Delgado F, Ortiz A, Renedo C, Pérez S, Mañana M, Zobaa
methods and Applications: A State-of-the-Art Survey, Vol. 164. AF. The influence of nuclear generation on CO2 emissions
Berlin, Heidelberg: Springer Science & Business Media; and on the cost of the spanish system in long-term generation
2012. planning. Int J Electr Power Energy Syst. 2011;33(3):673-683.
108. Yoon KPaul, Hwang CL. Multiple Attribute Decision Making: An 126. Chaudry M, Jenkins N, Qadrdan M, Wu J. Combined gas
Introduction, Vol. 104. Thousand Oaks, CA: Sage publications; and electricity network expansion planning. Appl Energy.
1995. 2014;113:1171-1187.
109. Løken E. Use of multicriteria decision analysis methods for 127. Park YM, Won JR, Park JB, Kim DG. Generation expansion
energy planning problems. Renewable Sustainable Energy Rev. planning based on an advanced evolutionary programming.
2007;11(7):1584-1595. IEEE Trans Power Syst. 1999;14(1):299-305.
110. Lu J, Ruan D. Multi-Objective Group Decision Making: Methods, 128. David A, Zhao RD. Integrating expert systems with dynamic
Software and Applications with Fuzzy Set Techniques, Vol. 6. programming in generation expansion planning. IEEE Trans
London: Imperial College Press; 2007. Power Syst. 1989;4(3):1095-1101.
111. Climaco J, Antunes C, Martins A, Marta J, Almeida A. A 129. Su CT, Lii GR, Chen Jiannung. Long-term generation expan-
decision support system for power generation expansion plan- sion planning employing dynamic programming and fuzzy
ning with a case study. In: 1990 IEEE Proceedings of the techniques. In: Proceedings of IEEE International Conference
29th IEEE Conference on Decision and Control, 1990; Hawai, on Industrial Technology, Vol. 1 IEEE; January 2000; Goa,
USA:264-270. India, India:644-649.
BABATUNDE ET AL. 27

130. Tafreshi SM, Lahiji AS, Aghaei J, Rabiee A. Reliable generation 147. Careri F, Genesi C, Marannino P, Montagna M, Rossi S, Siviero
expansion planning in pool market considering power system I. Generation expansion planning in the age of green economy.
security. Energy Convers Manage. 2012;54(1):162-168. IEEE Trans Power Syst. 2011;26(4):2214-2223.
131. Fini AS, Moghaddam MP, Sheikh-El-Eslami M. A dynamic 148. Blum C, Roli A. Metaheuristics in combinatorial optimiza-
model for distributed energy resource expansion planning con- tion: overview and conceptual comparison. ACM Comput Surv
sidering multi-resource support schemes. Int J Electr Power (CSUR). 2003;35(3):268-308.
Energy Syst. 2014;60:357-366. 149. Latorre G, Cruz RD, Areiza JM, Villegas A. Classification of
132. Hasani-Marzooni M, Hosseini SH. Dynamic model for publications and models on transmission expansion planning.
market-based capacity investment decision considering IEEE Trans Power Syst. 2003;18(2):938-946.
stochastic characteristic of wind power. Renewable Energy. 150. Park JB, Park YM, Won JR, Lee KY. An improved genetic
2011;36(8):2205-2219. algorithm for generation expansion planning. IEEE Trans
133. Clarke RR. Choosing an integrated resource plan for electric Power Syst. 2000;15(3):916-922.
utilities: an analytic hierarchy approach. In: Proceedings of 151. Fukuyama Y, Chiang HD. A parallel genetic algorithm
the 31st Intersociety Energy Conversion Engineering Confer- for generation expansion planning. IEEE Trans Power Syst.
ence, 1996. IECEC 96, Vol. 3. Washington, DC, USA: IEEE; 1996;11(2):955-961.
1996:1592-1597.
152. Firmo HT, Legey LL. Generation expansion planning: an iter-
134. Xiaotong L, Yimei L, Xiaoli Z, Ming Z. Generation and trans- ative genetic algorithm approach. IEEE Trans Power Syst.
mission expansion planning based on game theory in power 2002;17(3):901-906.
engineering. Syst Eng Procedia. 2012;4:79-86.
153. Chung T, Li Y, Wang Z. Optimal generation expansion plan-
135. Ng SK, Lee C, Zhong J. A game-theoretic approach to ning via improved genetic algorithm approach. Int J Electr
study strategic interaction between transmission and genera- Power Energy Syst. 2004;26(8):655-659.
tion expansion planning. In: Power Symposium, 2006. NAPS
2006. 38th North American. Carbondale, IL, USA: IEEE; 154. Sadegheih A, Drake P. System network planning expan-
2006:115-120. sion using mathematical programming, genetic algorithms
and tabu search. Energy Convers Manage. 2008;49(6):
136. Moghddas-Tafreshi S, Shayanfar H, Lahiji AS, Rabiee A, 1557-1566.
Aghaei J. Generation expansion planning in pool market: a
hybrid modified game theory and particle swarm optimization. 155. Murugan P, Kannan S, Baskar S. NSGA-II algorithm for
Energy Convers Manage. 2011;52(2):1512-1519. multi-objective generation expansion planning problem. Elec-
tric Power Syst Res. 2009;79(4):622-628.
137. Yakin MZ, McFarland JW. Electric generating capacity plan-
ning: a nonlinear programming approach. Electric Power Syst 156. Saldarriaga CA, Hincapié RA, Salazar H. A holistic approach
Res. 1987;12(1):1-9. for planning natural gas and electricity distribution networks.
IEEE Trans Power Syst. 2013;28(4):4052-4063.
138. Zhou Y, Wang L, McCalley JD. Designing effective and effi-
cient incentive policies for renewable energy in generation 157. Pereira AJ, Saraiva JT. A long term generation expansion plan-
expansion planning. Appl Energy. 2011;88(6):2201-2209. ning model using system dynamics—case study using data
from the portuguese/spanish generation system. Electric Power
139. Unsihuay C, Marangon-Lima J, de Souza AZ. Integrated power Syst Res. 2013;97:41-50.
generation and natural gas expansion planning. In: Power
Tech, 2007 IEEE Lausanne. Lausanne, Switzerland: IEEE; 158. Kannan S, Slochanal SMR, Subbaraj P, Padhy NP. Applica-
2007:1404-1409. tion of particle swarm optimization technique and its variants
to generation expansion planning problem. Electric Power Syst
140. Khodr H, Gomez J, Barnique L, et al. A linear pro- Res. 2004;70(3):203-210.
gramming methodology for the optimization of electric
power-generation schemes. IEEE Trans Power Syst. 2002;17(3): 159. Kannan S, Slochanal SMR, Padhy NP. Application of evo-
864-869. lutionary computation techniques for generation expansion
planning. In: Transmission and Distribution Conference and
141. You S, Hadley SW, Shankar M, Liu Y. Co-optimizing gen- Exposition, 2003 IEEE PES, Vol. 1 IEEE; 2003; Dallas, TX, USA,
eration and transmission expansion with wind power in USA:120-125.
large-scale power gridsimplementation in the us eastern inter-
160. Neshat N, Amin-Naseri M. Cleaner power generation through
connection. Electric Power Syst Res. 2016;133:209-218.
market-driven generation expansion planning: an agent-based
142. Singh KJ, Philpott AB, Wood RKevin. Dantzig-Wolfe decom- hybrid framework of game theory and particle swarm opti-
position for solving multistage stochastic capacity-planning mization. J Cleaner Prod. 2015;105:206-217.
problems. Oper Res. 2009;57(5):1271-1286.
161. Abbasi AR, Seifi AR. Simultaneous integrated stochastic elec-
143. Ng SK, Zhong J, Lee CW. A game-theoretic study of the strate- trical and thermal energy expansion planning. IET Gener
gic interaction between generation and transmission expan- Transm Distrib. 2014;8(6):1017-1027.
sion planning. In: PSCE'09. IEEE/PES IEEE Power Systems
162. Shayeghi H, Pirayeshnegab A, Jalili A, Shayanfar H. Applica-
Conference and Exposition, 2009.2009; Seattle, WA, USA:1-10.
tion of PSO technique for GEP in restructured power systems.
144. Rogers J, Rolko M. A quadratic programming model for plan- Energy Convers Manage. 2009;50(9):2127-2135.
ning generation and inter-utility transmission. Int J Electr 163. Bhuvanesh A, Karunanithi K, Kannan S. Least cost generation
Power Energy Syst. 1992;14(1):18-22. expansion planning with wind plant using differential evo-
145. Lumbreras S, Ramos A, Banez-Chicharro F. Optimal trans- lution algorithm. In: IEEE 2014 International Conference on
mission network expansion planning in real-sized power sys- Circuit, Power and Computing Technologies (ICCPCT). 2014;
tems with high renewable penetration. Electric Power Syst Res. Nagercoil, India:540-547.
2017;149:76-88. 164. Jadidoleslam M, Bijami E, Amiri N, Ebrahimi A, Askari J.
146. Pantoš M. Stochastic generation-expansion planning and diver- Application of shuffled frog leaping algorithm to long term
sification of energy transmission paths. Electric Power Syst Res. generation expansion planning. Int J Comput Electr Eng.
2013;98:1-10. 2012;4(2):115.
28 BABATUNDE ET AL.

165. Jadidoleslam M, Ebrahimi A. Generation expansion planning 183. Habib MA, Chungpaibulpatana S. Electricity generation
considering demand side management programs using ABC expansion planning with environmental impact abatement:
algorithm. Energy Eng Manage. 2014;4(1):2-15. case study of Bangladesh. In: O-Thong S, Waewsak J,
166. Hejrati Z, Hejrati E, Taheri A. Optimization generation eds. 2013 International Conference on Alternative Energy
expansion planning by HBMO. Optimization. 2012;37(7): in Developing Countries and Emerging Economies (AED-
99-108. CEE), Vol. 52. Bangkok, Thailand: Elsevier; May 2014:
410-420.
167. Kandil M, El-Debeiky S, Hasanien N. Rule-based system for
determining unit locations of a developed generation expan- 184. Awopone AK, Zobaa AF, Banuenumah W. Techno-economic
sion plan for transmission planning. IEE Proc-Gener Transm and environmental analysis of power generation expansion
Distrib. 2000;147(1):62-68. plan of Ghana. Energy Policy. 2017;104:13-22.
168. Davis A, Rong-da Z. An expert system with fuzzy sets for 185. https://energyexemplar.com/software/plexos-desktop-edition/
optimal planning. IEEE Power Eng Rev. 1991;11(2):46. 186. https://wiki.openmod-initiative.org/wiki/urbs
169. Farghal SA, Kandil MS, Abdel-Aziz MR. Generation expan- 187. Wenying C, Jia L, Linwei M, Ulanowsky D, Burnard G. Role
sion planning: an expert system approach, In: IEE Proceed- for carbon capture and storage in China. Energy Procedia.
ings C - Generation, Transmission and Distribution; 1988: 2009;1(1):4209-4216.
261-267. 188. Contaldi M, Gracceva F, Mattucci A. Hydrogen perspectives in
170. Farghal S, Aziz MA. Generation expansion planning includ- Italy: analysis of possible deployment scenarios. Int J Hydrogen
ing the renewable energy sources. IEEE Trans Power Syst. Energy. 2008;33(6):1630-1642.
1988;3(3):816-822. 189. Tseng P, Lee J, Friley P. A hydrogen economy: opportunities
171. Santisirisomboon J, Limmeechokchai B, Chungpaibulpatana and challenges. Energy. 2005;30(14):2703-2720.
S. Impacts of biomass power generation and CO2 taxation on 190. Gielen D, Simbolotti G. Prospects for Hydrogen and Fuel Cells.
electricity generation expansion planning and environmental Paris: International Energy Agency; 2005.
emissions. Energy Policy. 2001;29(12):975-985.
191. Contreras A, Guervós E, Posso F. Market penetration anal-
172. Schenk K, Chan S. Incorporation and impact of a wind energy ysis of the use of hydrogen in the road transport sector of
conversion system in generation expansion planning. IEEE the madrid region, using MARKAL. Int J Hydrogen Energy.
Trans Power Apparatus Syst. 1981;12:4710-4718. 2009;34(1):13-20.
173. Becker N, Soloveitchik D, Olshansky M. Incorporating 192. Endo E. Market penetration analysis of fuel cell vehicles in
environmental externalities into the capacity expansion Japan by using the energy system model MARKAL. Int J Hydro-
planning: an israeli case study. Energy Convers Manage. gen Energy. 2007;32(10):1347-1354.
2011;52(7):2489-2494. 193. Vaillancourt K, Labriet M, Loulou R, Waaub JP. The role
174. Malik AS, Bouzguenda M. Effects of smart grid technologies of nuclear energy in long-term climate scenarios: an anal-
on capacity and energy savings—a case study of Oman. Energy. ysis with the world-times model. Energy Policy. 2008;36(7):
2013;54:365-371. 2296-2307.
175. Nakawiro T, Bhattacharyya SC, Limmeechokchai B. Electricity 194. Hamacher T, Lako P, Ybema J, et al. Can fusion help to mitigate
capacity expansion in thailand: an analysis of gas dependence greenhouse gas emissions? Fusion Eng Des. 2001;58:1087-1090.
and fuel import reliance. Energy. 2008;33(5):712-723. 195. Lechon Y, Cabal H, Varela M, et al. A global energy model with
176. Kabouris J, Contaxis G. Optimum expansion planning of fusion. Fusion Eng Des. 2005;75:1141-1144.
an unconventional generation system operating in parallel 196. Giannakis G. Monitoring and evaluation of the res directives
with a large scale network. IEEE Trans Energy Convers. implementation in eu27 and policy recommendations for 2020.
1991;6(3):394-400. In: RES2020 Workshop. Greece; 2007.
177. Limmeechokchai B, Chungpaibulpatana S. Application of 197. International Energy Agency. World energy outlook 2010.
cool storage air-conditioning in the commercial sector: an http://large.stanford.edu/courses/2010/ph240/riley2/docs/
integrated resource planning approach for power capacity EIA-0484-2010.pdf; Accessed November 20, 2018.
expansion planning and emission reduction. Appl Energy. 198. Bowen A, Rydge J. Climate-Change Policy in the United King-
2001;68(3):289-300. dom No. 886. Paris: OECD Publishing; 2011.
178. IAEA. IAEA Tools and methodologies for energy system plan- 199. Pahuja N, Pandey N, Mandal K, Bandyopadhyay C. GHG miti-
ning and nuclear energy system assessments. IAEA, Viena, gation in India: an overview of the current policy landscape, 1.
Austria., sustainable energy for the 21st century Edition. World Resource Institute (WRI) Working Paper. http://www.
https://www.iaea.org/OurWork/ST/NE/Pess; 2011. wri.org/sites/default/files/wri_workingpaper_india_final.pdf;
179. Jenkins RT, Joy D. Wein automatic system planning pack- 2014.
age (WASP): an electric utility optimal generation expansion 200. Suganthi L, Iniyan S, Samuel AA. Applications of fuzzy logic
planning computer code. Tech. rep., Tenn.(USA), Oak Ridge in renewable energy systems—a review. Renewable Sustainable
National Lab.; 1974. ORNL-4945. Energy Rev. 2015;48:585-607.
180. www.energy.siemens.com. 201. Xie J, Wei X, Lu Y, Gan D. Emission-constrained optimal gen-
181. Caramanis MC, Schweppe FC, Tabors RD. Electric- eration scheduling incorporating equity principles. IET Gener
generation-expansion analysis system. Volume 1: solu- Transm Distrib. 2010;4(2):201-210.
tion techniques, computing methods, and results. Final 202. Abdollahi A, Moghaddam MP, Rashidinejad M,
report.[EGEAS computer code]. Tech. rep., Massachusetts Sheikh-El-Eslami MK. Investigation of economic and
Inst. of Tech., Energy Lab: Cambridge (USA); 1982. environmental-driven demand response measures
182. Aliyu AS, Ramli AT, Saleh MA. Nigeria electricity crisis: power incorporating uc. IEEE Trans Smart Grid. 2012;3(1):12-25.
generation capacity expansion and environmental ramifica- 203. Zhan TS, Chen SJ, Tsay MT, Kang MS, Tsai JI, Liao BX. Opti-
tions. Energy. 2013;61:354-367. mal generation expansion planning strategy for the utility with
BABATUNDE ET AL. 29

ipps participation and considering green house gas mitiga- 221. Yasuda K, Nishiya KC, Hasegawa J. Optimal generation expan-
tion. In: ICIEA 2009 IEEE 4th IEEE Conference on Indus- sion planning with electric energy storage systems. Trans Inst
trial Electronics and Applications, 2009. 2009; Xi'an, China: Electr Eng Japan. B. 1988;108(12):569-576.
2880-2885. 222. Oatman E, Hamant L. A dynamic approach to genera-
204. European Commission. Climate action, climate strategies tion expansion planning. IEEE Trans Power Apparatus Syst.
and targets. http://ec.europa.eu/clima/policies/strategies/; 1973;6:1888-1897.
Accessed November 19, 2018. 223. Poncelet K, Höschle H., Delarue E, Virag A, Dhaeseleer
205. Fatih B. World Energy Outlook 2015. Paris, France: Organisa- W. Selecting representative days for capturing the impli-
tion for Economic Co-Operation and Development (OECD); cations of integrating intermittent renewables in genera-
2015. tion expansion planning problems. IEEE Trans Power Syst.
206. Akinyele D, Rayudu R. Review of energy storage technologies 2017;32(3):1936-1948.
for sustainable power networks. Sustainable Energy Technol 224. Sobieski D, Bhavaraju M. An economic assessment of battery
Assess. 2014;8:74-91. storage in electric utility systems. IEEE Trans Power Apparatus
207. Moreira A, Pozo D, Street A, Sauma E. Reliable renewable gen- Syst. 1985;12:3453-3459.
eration and transmission expansion planning: Co-optimizing 225. Hamam K, Hamam YM, Hindi KS, Brameller A. Unit
system's resources for meeting renewable targets. IEEE Trans commitment of thermal generation. In: IEE Proceedings C
Power Syst. 2017;32(4):3246-3257. (Generation, Transmission and Distribution), Vol. 127 IET;
208. Ozcan M, Ozturk S, Yildirim M. Turkeys long-term generation 1980:3-8.
expansion planning with the inclusion of renewable-energy 226. Hua B, Baldick R, Wang J. Representing operational flexibility
sources. Comput Electr Eng. 2014;40(7):2050-2061. in generation expansion planning through convex relaxation of
209. Yamgar S, Nandaa G, Srivastavaa S, et al. Implications of energy unit commitment. IEEE Trans Power Syst. 2017.
tax on generation expansion plan & GHG emission: a case 227. Palmintier B, Webster M. Impact of unit commitment con-
study on indian power sector. In: 2004 International Con- straints on generation expansion planning with renewables. In:
ference on Power System Technology, 2004. Powercon 2004, IEEE Power and Energy Society General Meeting, 2011 IEEE;
Vol. 2. Singapore, Singapore: IEEE; 2004:1734-1739. 2011; Detroit, MI, USA, USA:1-7.
210. Nualhong D, Chusanapiputt S, Jantarang S, Pungprasert V. 228. Koltsaklis N, Georgiadis M. An integrated unit commit-
Generation expansion planning including biomass energy ment and generation expansion planning model. 12th Inter-
sources with global environmental consideration using national Symposium on Process Systems Engineering & 25th
improved tabu search. In: Tencon 2005 2005 IEEE Region 10 European Symposium on Computer Aided Process Engineering.
IEEE; 2005; Melbourne, Qld., Australia:1-5. Copenhagen, Denmark: Elsevier; 2015:2273-2278.
211. Nordlund P, Sjelvgren D, Pereira M, Bubenko J. Generation
229. Wierzbowski M, Lyzwa W, Musial I. MILP model for long-term
expansion planning for systems with a high share of hydro
energy mix planning with consideration of power system
power. IEEE Trans Power Syst. 1987;2(1):161-167.
reserves. Appl Energy. 2016;169:93-111.
212. Kamalinia S, Shahidehpour M. Generation expansion planning
230. Energy Information Administration. Annual Energy Outlook
in wind-thermal power systems. IET Gener Transm Distrib.
2007. http://large.stanford.edu/courses/2012/ph240/park1/
2010;4(8):940-951.
docs/0383-2007.pdf; Accessed November 20, 2018.
213. Hu Z, Jewell WT. Optimal generation expansion planning with
integration of variable renewables and bulk energy storage 231. Setlhaolo D, Xia X. Combined residential demand side man-
systems. In: 2013 1st IEEE Conference on Technologies for agement strategies with coordination and economic analysis.
Sustainability (SusTech). IEEE: Portland, OR, USA;2013:1-8. Int J Electr Power Energy Syst. 2016;79:150-160.

214. Zhang C, Ding Y, Kang C. Multi-objective generation expan- 232. Behrangrad M. A review of demand side management business
sion planning for integrating largescale wind generation. In: models in the electricity market. Renewable Sustainable Energy
AORC CIGRE Technical Meeting 2013. Guangzhou China; Rev. 2015;47:270-283.
2013. 233. Chefurka P. World energy and population: Trends to 2100.
215. Denholm P, Ela E, Kirby B, Milligan M. The role of energy stor- https://www.countercurrents.org/chefurka201109.htm;
age with renewable electricity generation. Tech. rep., National Accessed November 19, 2018.
Renewable Energy Laboratory; 2010. NREL/TP-6A2-47187. 234. Angelsen A, Brown S, Loisel C. Reducing emissions
216. Haas J, Cebulla F, Cao K, et al. Challenges and trends of from deforestation and forest degradation (REDD): an
energy storage expansion planning for flexibility provision in options assessment report. Tech. rep., Meridian Institute;
low-carbon power systems—a review. Renewable Sustainable 2009.
Energy Rev. 2017;80:603-619. 235. Martins AG, Coelho D, Antunes CH, Clímaco J. A multiple
217. Pandžić H, Wang Y, Qiu T, Dvorkin Y, Kirschen DS. objective linear programming approach to power generation
Near-optimal method for siting and sizing of distributed planning with demand-side management (DSM). Int Trans
storage in a transmission network. IEEE Trans Power Syst. Oper Res. 1996;3(3-4):305-317.
2015;30(5):2288-2300. 236. Babatunde OM, Munda JL, Hamam Y. Hybrid energy system
218. Kandil MS, Farghal SA, Hasanin NE. Economic assessment of for low-income households. In: IEEE AFRICON, 2017 IEEE;
energy storage options in generation expansion planning. In: 2017; Cape Town, South Africa:1038-1042.
IEE Proceedings C (Generation, Transmission and Distribu- 237. Omer AM. Energy, environment and sustainable development.
tion), Vol. 137 IET; 1990:298-306. Renewable Sustainable Energy Rev. 2008;12(9):2265-2300.
219. Acar C. A comprehensive evaluation of energy storage options 238. Yüksel I. Hydropower in turkey for a clean and sus-
for better sustainability. Int J Energy Res. 2018. tainable energy future. Renewable Sustainable Energy Rev.
220. Sorgulu F, Dincer I. Design and analysis of a solar tower power 2008;12(6):1622-1640.
plant integrated with thermal energy storage system for cogen- 239. Bilgen S, Keleş S, Kaygusuz A, Sarı A., Kaygusuz K. Global
eration. Int J Energy Res. 2018. warming and renewable energy sources for sustainable devel-
30 BABATUNDE ET AL.

opment: a case study in turkey. Renewable Sustainable Energy 254. Phuc NX, Marpaung C, Shrestha RM. Portfolio risk analysis
Rev. 2008;12(2):372-396. based generation expansion planning considering CO2 trad-
240. Akinyele D, Belikov J, Levron Y. Challenges of microgrids in ing. In: IEEE 2010 Proceedings of the International Conference
remote communities: a STEEP model application. Energies. onEnergy and Sustainable Development: Issues and Strategies
2018;11(2):432. (ESD); 2010; Chiang Mai, Thailand:1-8.
241. Bekker B, Eberhard A, Gaunt T, Marquard A. South africa's 255. Yaghooti A, Khanbeigi GA, Esmalifalak M. Generation expan-
rapid electrification programme: policy, institutional, plan- sion planning in IEEE power system using probabilistic
ning, financing and technical innovations. Energy Policy. production simulation. In: IEEE 2010 IEEE International
2008;36(8):3125-3137. Energy Conference and Exhibition (EnergyCon); 2010; Man-
ama, Bahrain:769-774.
242. Kohler M. Differential electricity pricing and energy efficiency
in South Africa. Energy. 2014;64:524-532. 256. Dehghan S, Amjady N, Kazemi A. Two-stage robust genera-
tion expansion planning: a mixed integer linear programming
243. Eskom Media Desk. Loadshedding to run from 8am to 10pm model. IEEE Trans Power Syst. 2014;29(2):584-597.
today. http://www.htxt.co.za/2015/05/21/loadshedding-to-
run-from-8am-to-10pm-today/; 257. Khan AZ, Yingyun S, Ashfaq A. Generation expansion plan-
Accessed November 30, 2017. ning considering externalities for large scale integration of
renewable energy. In: IEEE 2014 IEEE International Confer-
244. Monyei C, Adewumi A. Demand side management potentials ence on Intelligent Energy and Power Systems (IEPS); 2014;
for mitigating energy poverty in South Africa. Energy Policy. Kiev, Ukraine:135-140.
2017;111:298-311.
258. Gil E, Aravena I, Cárdenas R. Generation capacity expansion
245. Eskom. Eskom 2018/19 revenue application. http:// planning under hydro uncertainty using stochastic mixed inte-
www.eskom.co.za/OurCompany/stakeholderNews/ ger programming and scenario reduction. IEEE Trans Power
Documents/RevenueApplFY2018-19_20092017.pdf; Accessed Syst. 2015;30(4):1838-1847.
16 November 2017.
259. Shengyu W, Lu C, Xiaoqing Y, Bo Y. Long-term genera-
246. City-Express. Eskom may pass on new tax burden. https:// tion expansion planning under uncertainties and fluctuations
www.news24.com/Archives/City-Press/ of multi-type renewables. In: IEEE 2015 IEEE 5th Interna-
Eskom-may-pass-on-new-tax-burden-20150430; Accessed: tional Conference on Power Engineering, Energy and Electri-
November 14, 2017. cal Drives (POWERENG); 2015; Riga, Latvia:612-616.
247. Eastern Cape Socio Economic Consultative Council (ECSECC). 260. Dang C, Wang X, Wang X, Xiao Y, Teng W. Generation expan-
Poverty trends in South Africa 2006–2015: Highlights for the sion planning with decomposition-coordination: an improved
eastern cape. https://www.ecsecc.org/documentrepository/ iterative method. In: 2015 IEEE PES Asia-Pacific IEEE Power
informationcentre/poverty-trends-in-south-africa-ec_43745. and Energy Engineering Conference (APPEEC); 2015; Bris-
pdf; Accessed November 20, 2018. bane, QLD, Australia:1-5.
248. Monyei C, Jenkins K, Viriri S, Adewumi A. Policy discussion 261. Hu Y, Ding T, Bie Z, Lian H. Integrated generation and trans-
for sustainable integrated electricity expansion in south africa. mission expansion planning with carbon capture operating
Energy Policy. 2018;120:132-143. constraints. In: 2016 IEEE Power and Energy Society General
249. Sasaki H, Kubokawa J, Watanabe M, Yokoyama R, Tanabe R. Meeting (PESGM); 2016; Boston, MA, USA:1-5.
A solution of generation expansion problem by means of neu- 262. Hinojosa V, Gonzalez-Longatt F. Stochastic security-
tral network. In: Proceedings of the First International Forum constrained generation expansion planning methodology
on Applications of Neural Networks to Power Systems. Seattle, based on a generalized line outage distribution factors. In:
WA, USA, USA: IEEE; 1991:219-224. 2017 IEEE Manchester PowerTech, IEEE; 2017; Manchester,
250. Handke J, Handschin E, Linke K, Sanders H-H. Coordination UK:1-6.
of long-and short-term generation planning in thermal power 263. Kendziorski M, Setje-Eilers M, Kunz F. Generation expan-
systems. IEEE Trans Power Syst. 1995;10(2):803-809. sion planning under uncertainty: an application of stochas-
251. Min X, Jinfu C, Zhong DX. Generator maintenance schedul- tic methods to the german electricity system. In: IEEE 2017
ing in the generation expansion planning of interconnected 14th International Conference on the European Energy Market
power system. In: IEEE Transmission and Distribution Confer- (EEM).2017; Dresden, Germany:1-7.
ence and Exhibition 2002: Asia Pacific. IEEE/PES, Vol. 3. 2002;
Yokohama, Japan, Japan:1601-1605.
252. Lin WM, Zhan TS, Tsay MT, Hung WC. The generation expan-
sion planning of the utility in a deregulated environment.
In: IEEE Proceedings of the 2004 IEEE International Confer- How to cite this article: Babatunde OM, Munda
ence on Electric Utility Deregulation, Restructuring and Power JL, Hamam Y. A comprehensive state-of-the-art
Technologies, 2004. (DRPT 2004). Vol. 2. 2004; Hong Kong, survey on power generation expansion plan-
China, China:702-707.
ning with intermittent renewable energy source
253. Marcato A, Ivo Chaves SJr, Garcia P, et al. Genetic algorithm
approach applied to long term generation expansion planning. and energy storage. Int J Energy Res. 2019;1–30.
In: TDC'06. IEEE/PES IEEE Transmission & Distribution Con- https://doi.org/10.1002/er.4388
ference and Exposition: Latin America, 2006; 2006; Caracas,
Venezuela:1-7.