Composite generation and transmission expansion planning considering distributed generation

Electrical Power and Energy Systems 62 (2014) 792–805

Contents lists available at ScienceDirect

Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes

Composite generation and transmission expansion planning considering distributed generation Ahmad Rouhani a,⇑, Seyyed Hadi Hosseini b, Mahdi Raoofat c a

Young Researchers and Elite Club, Beyza Branch, Islamic Azad University, Beyza, Iran Department of Electrical Engineering, Zanjan University, Zanjan, Iran c Department of Power and Control Engineering, School of Electrical and Computer Engineering, Shiraz University, Shiraz, Iran b

a r t i c l e

i n f o

Article history: Received 5 May 2013 Received in revised form 20 May 2014 Accepted 22 May 2014

Keywords: Composite expansion planning Distributed generation Generation expansion planning Mixed integer linear programming Transmission expansion planning

a b s t r a c t This paper presents a model for use in the problem of composite generation and transmission expansion planning considering distributed generation. Generation expansion planning is defined as the problem of determining what capacity, which, and when new generating units should be constructed over a long range planning horizon, to satisfy the expected energy demand using single nodal generation planning model. Then, the place of every planned generating units and distributed generation is determined simultaneous with transmission expansion planning considering nonuniform geographical fuel supply cost and potential of distributed generation technology. The problem is formulated as a Mixed-Integer Linear Programming. By allocating the overall generation capacity among the grid nodes and determining the new transmission element additions along the planning horizon, the overall cost of the system is minimized. To assess the capabilities of the proposed approach, the Iranian Power Grid as a large scale system is considered. The effectiveness of the proposed modifications is illustrated in detail. Ó 2014 Elsevier Ltd. All rights reserved.

Introduction Rapid growth in consumer demand, along with other technical and economical reasons may cause inadequacy in the available electric network. So, electric utilities face the challenge to serve electricity demand for the coming years with acceptable reliability, safety and quality through the expansion in generation, transmission and distribution systems [1]. Therefore planning for the electric power sector encompasses generation, transmission and distribution systems [2]. Generation expansion planning (GEP) is considered one of major parts of power system planning issues. The aim of GEP is to seek the most economical generation expansion scheme achieving an acceptable reliability level according to the forecast of demand increase in a certain period of time. (Long-term planning can cover more than 30 years) [2]. The feasibility of the generation structure, the cost of primary energy resources and fuel for the scheme, and the reliability indices of electricity supply, make generation planning a very complicated optimization mathematically [3]. Some of these restrictions have been applied in GEP in the recent literature [4–7]. But applying transmission line restrictions is not simply possible without transmission expansion planning (TEP). On the ⇑ Corresponding author. Tel.: +98 7127623257. E-mail address: [email protected] (A. Rouhani). http://dx.doi.org/10.1016/j.ijepes.2014.05.041 0142-0615/Ó 2014 Elsevier Ltd. All rights reserved.

other hand, not applying this constraint may led to non-optimal response. WASP-IV is powerful software developed by International Atomic Energy Agency (IAEA) in which a dynamic programming approach is employed to find an overall optimal required generation capacity for the network so that an index, such as LOLP, is minimized [8]. In using WASP-IV, it is assumed that the fuel cost throughout the geographical distribution of the network is uniform. This assumption is invalid in real life, as allocation of a power plant far from a fuel resource supply center results in high fuel transmission costs. Moreover, in using WASP-IV, a single-node load center is assumed which is not obviously a valid assumption [8]. In other words, while WASP-IV is capable of predicting the overall generation capacity requirements for the grid, it is unable to geographic-ally distribute and allocate the capacities among the areas [9]. As mentioned before, the investment in a power plant is greatly influenced by the environment in which the power plant is situated, e.g. water supply, dissipation conditions, and the cost of the land. In addition the geographical locality also has a bearing on the entire system investment and operational cost. For example there will be additional transmission investment cost if the power plant is far away from the load centers and the fuel cost will be greater if the power plant is far from a fuel source [2]. However, the locations of generating units and costs of transmission lines

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

793

Nomenclature

Acronyms and abbreviation DP dynamic programming LDC load duration curve LINGO linear interactive and general optimizer LOLP Loss of Load Probability MATLAB matrix laboratory WASP wien automatic system planning

CFdb Dct Ddb Db PGct PGhc PGdb

Indices and sets T number of years in a planning horizon t year index along the planning horizon d load duration curve level index i existing or candidate thermal plants index j existing or candidate DG technologies index si existing and candidate plants index h existing or candidate hydroelectric plants index c hydrological conditions index b bus index l existing and candidate transmission line index le existing transmission line index lc candidate transmission line index lb existing and candidate transmission lines index connected to bus b m transmission lines type index f fuel type index for thermal plants e emission type index of generating units Variables LOLPct LOLP index of critical period in year t e standard level of LOLP index c spinning reserve ratio bjb geographical potential of DG technology j in bus b a allowed penetration of DG in network at lower bound of reserve margin in year t bt upper bound of reserve margin in year t td duration of subperiod d Itsib present value for investment cost of generating unit si in year t in bus b Ftsib present value for fuel cost of generating unit si in year t in bus b Mtsib present value for maintenance cost of generating in bus b unit si in year t. Otsib present value for operating cost of generating in bus b unit si in year t Stsib salvage value for investment cost of generating unit si in year t in bus b CDGjb investment cost of DG technology j in bus b CTlcm investment cost of a circuit of candidate transmission line lc of type m CGib Investment cost of generating unit i in bus b

are usually neglected. It is customary to assume that there are adequate transmission lines to achieve any generation expansion plans. To solve separately the TEP problem, different methodologies have been presented [10–16]. Actually the practical situations are often different from above premise, especially in some developing countries where transmission networks are very large and weak. So, TEP and GEP have considerable effects on each other. Therefore, considering GEP & TEP as two separate optimization problems, results in reducing the genuine optimization point [2]. Therefore generation and transmission expansion planning is a key factor in a long term power system operation [17].

PGb PDGdjb Pdl Pdle Pmax le Pdlcm Pmax lcm Pblbl Pmax blbl W max hcdt Fid Fifd W max fd Eiedt Emax edt NGit NGmax it NGmax i NDGmax j NDGjb NDGmax jb NGib max NGib NTlcm NTlc NT max lc nisland S Z

cost of fuel consumed in subperiod d in bus b forecast peak demand in the critical period of year t total demand in bus b in subperiod d total demand in bus b installed generation capacity in the critical period of year t installed generation capacity of hydroelectric plant h in hydrological condition c total installed generation capacity in bus b in subperiod d total installed generation capacity in bus b total installed generation capacity of DG type j in bus b in subperiod d Power flow capacity in existing or candidate transmission line l in subperiod d power flow capacity in existing transmission line le in subperiod d maximum power flow capacity of a circuit in existing transmission line le power flow capacity of a circuit in candidate transmission line lc of type m in subperiod d maximum power flow capacity of a circuit in candidate transmission line lc of type m power flow capacity from bus b to lb in transmission line l maximum power flow capacity in transmission line l maximum energy enhanced from hydroelectric plant h in hydrological condition c in subperiod d in year t fuel consumption of thermal unit i in subperiod d fuel consumption type f of thermal unit i in subperiod d maximum fuel type f available in subperiod d total emission type e of generating unit i in subperiod d in year t maximum emission type e in subperiod d in year t number of new generating unit i constructed in year t maximum number of allowed generating unit i constructed in year t number of determined generating unit i by WASP number of determined DG technology j by WASP number of installed DG technology j in bus b. maximum number of allowed DG technology j in bus b number of installed generating unit i in bus b maximum number of allowed generating unit i in bus b Number of candidate transmission line lc of type m number of constructed circuit of candidate transmission line lc maximum number of constructed circuit of candidate transmission line lc number of island detection feasible solution domain of GEP objective function of the expansion plan

After 1990, the locations of generating units and costs of transmission lines are receiving more attention [18,19]. An algorithm is proposed in [20] for generating unit location optimization. The transmission congestion and competition on power generation expansion was studied in [20]. The market-based coordination of transmission and generation capacity planning is proposed in [21]. On the other hand, distributed generation (DG) is one new option being promoted for solving distribution system capacity problems [22–24]. DG is a feasible alternative for developing new capacity, especially in competitive electricity networks, from an economic, technical and environmental point of view [25–29]. Power system

794

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

Problem definition and remarks The problem to be solved is a composite generation and transmission expansion planning one. The aim is to allocate the overall generation capacity among the grid nodes and to determine the transmission elements requirements, considering the distributed generation. The problem is supposed to be solved for several years within a specified planning horizon. In doing so, the following points are worth mentioning:  In general the composite expansion planning (CEP) is MixedInteger Non-Linear Programming (MINLP) problem [2], however, a Mixed-Integer Linear Programming (MILP) model is applied in this part by making some approximation.  The planning problem to be solved is of a dynamic type. In other words, the planning horizon is divided into several stages and subperiods of known duration, so the elements to be installed in each stage should be determined. In addition, it is assumed that the predicted load is known for each stage.  Load Duration Curves (LDCs) represent the operating conditions of power systems over the time; they are obtained from hourly data of demand over a period of time. It can be used in generation expansion planning when all the load and all the generating units are assumed to be connected at the same node (single

Load Level (%)

d1 d2 d3

...

deregulation and the shortage of transmission capacities have led to increase interest in DG sources [30,31]. Also, it is known that renewable energy such as wind, hydro, solar, and geothermal are relatively expensive and limited in availability. Anyway, to mitigate the environmental impacts to the planet and the risk of depending only on few sources of energy, there is an increasing interest in renewable energy sources [4]. A multistage model for distribu-tion expansion planning with DG is proposed in [32–34]. However, for large scale practical systems, more effective approaches are still in need to solve the dimension problems. Also none of the mentioned papers look at the DG from CEP view point and consider its effects on the expansion planning. In this study, a new constructive heuristic approach is presented for use in the problem of composite generation and transmission expansion planning with DG consideration. First, GEP problem is defined as the problem of determining what capacity, which, and when new generating units should be constructed over a long range planning horizon. To achieve this aim, the WASP software using single nodal generation planning model is employed to satisfy the expected energy demand. Then the place of every generating unit and distributed generation is determined simultaneous with TEP. Considering the constraints related to the reliability of power plants and transmission networks, maintenance of generating units, fuel supply constraints, determination of the type of new power plants, and distributed generation consideration are some features of the proposed approach. This paper is organized as follows: In Section ‘Problem definition and remarks’, the problem definition and the basic assumptions are provided. Section ‘Problem formulation’ shows how the optimization problem is modeled, with details of the objective function and constraints imposed. The proposed solution algorithm is presented in Section ‘Solution algorithm’. To assess the capabilities of the proposed approach, the Iranian Power Grid as a large scale system is considered over long range planning horizon. Section ‘Case study’ describes the Iranian Power Grid as a largescale practical case used to implement on the proposed model. Section ‘Specific details’ presents the specific details of the system. The results obtained using the proposed approach are described in Section ‘Numerical results’. Finally, this paper ends with a summary of conclusions.

dn td1

td2

t1

td3

t2

tdn

t3

tn Time (hour)

Fig. 1. Linearly approximated load duration curve.











 







nodal point generation planning). LDC consists of several levels, as shown in Fig. 1. It is a linear approximation to practical load during curve. Spare or redundant capacities in generation and network facilities have been inbuilt in order to ensure adequate and acceptable continuity of supply in the event of failures and Forced Outage Rate (FOR) of plant, and the removal of facilities for regular scheduled maintenance. Therefore, the total outage in the failure events may be due to a forced outrage or a maintenance outage [35]. Those are not neglected in the proposed approach. The reliability of generation system configuration is evaluated by WASP in terms of the Loss of Load Probability index (LOLP). This index is calculated in WASP for each period of the year. The capacity factor of a generating unit is determined by the load it serves. The capacity factor can be defined as the percentage of hours of the year that a generating unit serves a load [17]. In each stage to the planning horizon, nodes may be modified by increasing the capacity and installing new generating units; branches may be modified by conductors replacing or by adding branch connecting nodes not the one previously connected to. It is noted that all generators of the same generation technology in one bus are expressed as an equivalent one. Its capacity is the sum of the generators’. All transmission lines between a bus pair are also treated as an equivalent one. The candidate transmission elements, generating units, and DGs from which some would be selected, are assumed to be known. Limits on generation capacities, transmission capacities, and availability of distributed generation are taken into account at each stage. The allocated generation capacity of each node can vary from 0 to maximum possible capacity. The load and the network model used in this paper are adapted from the transporting load-flow linearized network model. During the planning horizon, the transmission network may have to be expanded sufficiently so that the generation (both existing and newly added) can be transmitted to the load nodes. So any overload violation on any transmission element should be avoided. For the checking of this point, transporting load flow formulation is employed. This is assumed to make the convergence possible and to reduce the search space. The reserve rates of transmission capacity are introduced to the model as security or stability margins of transmission lines. In practical power systems, some transmission lines cannot be operated on its rated transmission capacity because of the system stability or security level limitation, which can be taken in account in this planning model by using the reserve rate. The node fuel supply costs are not the same and can be different. In other words, it consists of two components: the cost of fuel; and the cost of transmission of fuel which is different and depends on the geographical distances between the node and the refinery.

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

Power balance at each bus:

Problem formulation The problem as defined in Section ‘Problem definition and remarks’ is, in fact, an optimization one to be solved by a proposed solution algorithm. The formulation of the composite expansion planning of generation and transmission line is presented in this section. When GEP is combined with TEP, the main objective is to decide whether to invest in new generation, new transmission, or a combination of both to ensure no load curtailment while minimizing the variable generating cost. The objective function terms as well as the various constraints will be discussed in the following subsections.

For the long range of planning, the present value of the total cost of engineering project is usually taken as an objective function. Therefore, the objective function to be minimized is the present value of investment and operational costs. The objective function comprises two terms: the costs of generation plants and the costs of transmission enhancement requirements along the planning horizon. This objective function is defined as the total present value sum of the investment cost for new units, the generation costs, and the cost for additional capacity of transmission. T XX X XX ðItsib þ F tsib þ M tsib þ Otsib  Stsib Þ þ CT lcm t¼1 si

b

PGb  Db 

XX Pblbl ¼ 0 lb

lc

m

 NT lcm

ð1Þ

Constraints

Transmission lines: as mentioned before, in most cases, the structures of transmission networks are mainly dependent on the geographical distribution of installed capacity of generators and electrical demand. It means that the expanded transmission network would have no overload lines when all generators are run at full load states. In constraint (12) the reserve rates of transmission capacity are introduced to the model. They can be seen as security or stability margin of transmission lines.

ð2Þ

Reliability: the reliability index LOLP is used to evaluate adequacy of generating units.

LOLP ct 6 e

ð3Þ

The presence of hydro power plants: this constraint expresses the maximum energy obtained from a hydro power plant in the different periods of the planning horizon at different climatic conditions.

PGhc  t d 6 W max hcdt

ð4Þ

Fuel constraint: maximum fuel supply of different fuel types of thermal plants.

X F ifd 6 F max fd

ð5Þ

i

Emission constraint: maximum production rate of pollution.

X Eiedt 6 Emax edt

ð6Þ

i

Repairing time of different types of generating units:

Repairing time for each unit P Required maintenance time

ð11Þ

Islanding condition: the grid should be so designed that no islanding would happen in normal or contingency conditions. As a result:

nisland ¼ 0

ð12Þ

Solution algorithm The optimization problem formed by the objective function terms, defined by (1), and the constraint terms, defined by (2)–(6), (8)–(13), is a MILP problem, which for a large scale system is very difficult to solve. In this subsection, a new constructive heuristic algorithm is proposed which can find an acceptable optimal solution for the complex problem, even for a large system. As shown in Fig. 2, the proposed approach for the planning horizon can be divided into two phases that is described as follows: Phase one

The constraints to be observed during the optimization process are as follows: Generation capacity: the capacity sum of newly installed and existing generating units are more than or equal to the load demand plus reserve in each year within planning period.

ð1 þ at ÞDct 6 PGct 6 ð1 þ bt ÞDct

ð10Þ

l

max Pmax blbl 6 P blbl 6 P blbl

Objective function

z¼

795

ð7Þ

Maximum number of generating units and constructed transmission lines in each period throughout the planning horizon.

NGit 6 NGmax it

ð8Þ

NT lc 6 NT max lc

ð9Þ

In phase I, a preliminary study is carried out to determine the suitable solution of GEP problem. As mentioned before, the aim is to find the type of power plant (steam, gas, combined cycle, nuclear, hydro, renewable, and so forth) that minimizes the total cost of producing electrical power during the planning horizon. The generation model and configuration are determined in planning horizon based on DP using WASP. The response space and constraints for solving GEP problem using WASP-IV is shown in Fig. 3. In this figure, PGcon represents the installed generation capacity curve of under construction and downtime generating unit in the network in planning horizon. Increasing the curve in some periods, expresses the increase of the capacity of the network due to installed generating units under construction. On the other hand, reducing the curve meant the outage of generating units due to end of their useful life. According to the mentioned issues, the production should be within the range of S (the feasible solution domain). Therefore, with the loss of the network adequacy since t0, by doing an optimal GEP, PGsch curve is obtained. It represents the installed generation capacity of new scheduled generating units to restore the generation network adequacy. It should be noted that in using WASP, a single-node load center is assumed. So, in the first phase of the proposed method, as shown in Fig. 4, the whole power network is considered as a single-node. Considering the large scale non-linear optimization, in order to solve this planning program, a method has been developed by interfacing MATLAB as shown in Fig. 5. Combination of the power plants determined by the WASP software in each period (that is obtained from the first phase), technical and economic data of transmission network, and the generation and demand of each region in each period read by MATLAB software, then, the LINGO software code is created to apply in the next phase of the proposed approach.

796

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

WASP Determining optimum generation model and distributed generations in planning horizon

Refinery

Load, generating Input 1 units and DGs data

G1

Nuclear energy

DG1

Gi

DGj

...

...

Hydro energy

No New configuration of DG determined?

Phase I

Wind energy

Yes Solar energy

LINGO

1

Allocation of DGs for each period in planning horizon

Fig. 4. A single-node sample network. Geographical data

Input data*

Network demand correction according to allocation of DGs in every region

2

WASP (DP) For planning Horizon

LINGO

3

Allocation of generating unit simultaneous with TEP for each period of planning horizon

4

Network optimization results

LINGO (MILP) For each periods of planning horizon

Generation planning solution

Transmission network data

Input data**

MATLAB

Correction of generating units and transmission lines according to GEP and TEP

* Generating units and DGs technical and economic data, Load forecast, ** In format of applicable with LINGO

t=t+1

Fig. 5. WASP–LINGO interface.

t
Yes

Phase two

No No

end

5

Phase II

New Transmission line(s) determined?

Yes Considering the transmission cost in construction costs of centralized power plants Considering original demands of each region

Fig. 2. Flowchart of proposed composite expansion planning algorithm.

PG(MW) S

bt PGsch

In phase II, All DGs and generating units are allocated simultaneously with TEP for each period of planning horizon using the proposed heuristic approach. To explain the proposed approach, first a single-year planning horizon is considered. Then the algorithm of single year is extended for the case of multiyear planning horizon. In this phase, the network is simplified. A node in the simplified system represents an area, rather than a substation. Thus the expansion scheme of transmission lines between the main substations cannot be obtained in detail. Fig. 6 represents the equivalent of the above network that is composed of several regions. In the equivalent network, each region will be shown with a single-node that all load and generating units are located on it. Also, different regions are connected by lines, which are equal to the total capacity of transmission lines between the regions of the original

ε

at

Reg.1

Reg.2

Dt DG1

G1 DGj

..

DGj

..

.. Gi G1

t0

DG1

Gi

..

PGcon

..

Gi

..

G1

DGj DG1

Reg.3

t (year)

Fig. 3. The feasible solution domain and constraints for solving GEP problem.

Fig. 6. The equivalent of the sample network composed of several regions.

797

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

network. Resistance and reactance of the lines are parallel equivalent between the lines. However, this phase is implemented in several parts of the planning zone with different climatic conditions and energy sources. In addition a regional primary energy attribute also includes in this phase to express the potential of various kinds of energy sources in the region under study. Also, to implement this phase, the average price of the fuel in different regions of the study area is used. As shown in Fig. 6, to find the location of plants, the study area is divided into several different areas. Demand forecast for each region in each period, the price of fuel in each region, and the potential of the various DG technologies for each region are known. For this purpose, the coefficients are determined for the fuel price and the potential of distributed generations for the different regions. In addition, the maximum electric power transfer between regions is determined. For example, Table 1 shows a sample of the required information to find the optimum answers at this stage. This phase can be divided into five steps described as follows:

of transmission system design. Furthermore, according to the specification of DGs, installation of these generations in downward grid, influenced the delivered power through the transmission system, and could have significant economic benefits for the distribution companies. In this regard, the DG capacity determined in each region, obtained from the previous step, is reduced from the demand of the proposed region. Step 3. Allocation of generating units simultaneously with TEP In step 3, a preliminary study is carried out to determine the suitable transmission elements candidates. For placement of the new centralized power plants simultaneous to the TEP, MILP is used. A. Objective Function

Z¼

i

d

Power balance:

PGdb  ð1 þ cÞDdb 

XX XXX P dlb þ NGib  PGdb P 0 d

lb

d

i

ð18Þ

b

Power flow capacity of existing transmission lines:

Pmax le

6 Pdle 6 P max le

ð19Þ

j

max NT lcm  Pmax lcm 6 P dlcm 6 NT lcm  P lcm

b

NT lcm 6 NT max lcm

B. Constraints

ð20Þ

Construction of new transmission lines:

ð13Þ

ð21Þ

Construction of power plants in different shins:

Allowed penetration of DG in network:

XX XX PGdb  a P PDGdj d

ð17Þ

b

Power flow capacity of candidate transmission lines:

XX XXX min : Z ¼ NDGjb  CDGjb þ bjb  PDGdjb  NDGjb  t d

b

b

B. Constraints

A. Objective Function

d

i

XXX þ F id  CF db  Pdb  td  NGib d

It is difficult to find a unique DG technology that takes into account multiple considerations, such as economic, technical, and environmental attributes [25]. In the first step, all DGs are allocated based on system analysis and some engineering judgments such as geographical potential of DG technologies using MILP method by LINGO software. So, DGs would be added to initial network. The objective function and constraints of this step is defined as follows:

b

m

l

Step 1. Allocation of distributed generation

j

XX XX NT lm  CT lm þ NGib  CGib

ð14Þ

NGib 6 NGmax ib

ð22Þ

j

Restrictions on construction of different types of DG on different shins:

NDGjb 6 NDGmax jb

ð15Þ

The maximum number of new installed DG:

X NDGjb ¼ NDGmax j

ð16Þ

b

Step 2. Network demand correction In this study, the problem of power plants and distributed generation allocation is taken into consideration from the perspective

Maximum new power plants:

X NGib ¼ NGmax i

ð23Þ

b

Step 4. Correction of generating units and transmission lines These candidates are usually determined based on system analysis and some engineering judgments. In this step of proposed method, all performed expansion, such as capacity of allocated centralized power plants and new circuits of transmission lines are added to the initial network. The updated network is considered as the base case and is fed into the proposed heuristic approach in the next period of planning horizon.

Table 1 Geographical potentials of each region. Fuel source

Nuclear energy

Hydro energy

Distributed renewable energy Wind energy

Region 1 Region 2

Region 3

Solar energy

798

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

Step 5. Considering the transmission cost in construction costs of centralized power plants

region is considered and the other regions are considered as a single-node network. So, the exact location of power plants and transmission lines required between the region and within the region are determined.

In step 5, according to the cost of determined transmission lines, throughout CEP, prorated costs of transmission network will be added to the cost of centralized power plants. The iterative process will continue until a violation happens. This means that the extended transmission network would have no overload lines when all generators are run at full load states; Or in the first phase the proposed structure by WASP does not change. In these cases it can be said that optimum investment plan is achieved. The Improvement of the proposed approach to determine more details of every region is best illustrated by a simple sample as shown in Fig. 7. This figure indicates how to apply the proposed method in the large scale networks. In this regard, it is not necessary to implement WASP. At this point, the real model of the proposed

Reg.1

DG1

Gi

..

..

DGj

..

..

To validate the mathematical model given in the first part of this paper [3], the Iranian Power Grid as a large scale system is considered. In recent years, Iran has put a great deal of effort into moving towards restructuring and privatizing the power industry by establishing 17 Regional Electricity Companies (RECs), 28 generation management companies and 42 distribution companies. Iran is a vast country that has extensive resources of fossil fuels. Major fuel resources are located in the southern part of the country. In the previous years, these resources have been transferred by the oil and gas pipelines to most parts of the country [1]. Iran Grid Management Company (IGMC) is responsible for the network ensuring performance, the electri-city market operations, and providing the electric transit in the network. Iran Electrical Development Company (IEDC) is responsible for following up the implementation of power generation and national transmission projects. Establishing reliable balances in power supply to all RECs of the country, is a pivotal policy of power sector development. In this regard, the establishment of a reasonable proportion between demand and production in each RECs, establishing sufficient transmis-sion and distribution capacity to facilitate exchanges among the RECs, are considered as a priority of the power sector. Thus, regarding the presence of massive reserves of energy in the country’s southern provinces, and consequently the development of the energy industry in these regions, special attention will be performed to the development of generation, transmission and distribution facilities of electrical energy in Khozestan (R10), Fars (R11), and Hormozgan (R14) [1].

Reg.2

G1

Gi G1

Case study

DGj

Reg.3

DG1

Fig. 7. Improvement of the proposed approach to determine more details.

Caspian Sea R06

R03 R02 R05 R15

R01

R04

R08 R07 R09 R12

R10

R13 R11

R16

R14 REC load center

Persian Gulf

Refinery Wind energy

Oman Sea

Fig. 8. Geographic map of RECs in Iran and location of primary energy resources.

799

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805 Table 2 RECs Information and potential of primary energy resources of Iranian Power Grid. REC No.

REC name.

Demand (MW)

Generation (MW)

Oil & Gas

Wind

R01 R02 R03 R04 R05 R06 R07 R08 R09 R10 R11 R12 R13 R14 R15 R16 Total

Tehran Mazandaran Gilan Semnan Zanjan Azerbaijan Bakhtar Gharb Isfahan Khozestan Fars Yazd Kerman Hormozgan Khorasan Systan

8960 2618 1222 542 1295 2696 3058 1517 4973 6804 3935 952 1973 2245 3582 1122 47,501

9946.2 2260 1770.5 12.5 0 3570.8 2375.6 1336.4 4412 11063.6 5928.6 994.6 2230.5 2337.6 5137.7 1137.2 54513.8

Unfit Unfit Unfit Unfit Unfit Unfit Unfit Unfit Good Excellent Excellent Good Intermediate Excellent Intermediate Intermediate

Unfit Unfit Excellent Excellent Excellent Good Unfit Unfit Unfit Unfit Intermediate Intermediate Intermediate Unfit Good Good

Table 3 General information. Parameter

Value

Study period Planning horizon Number of periods in year Annual rate (%) Annual rate (%) Minimum reserve margin (%) Maximum reserve margin (%) Critical LOLP (%)

2010–2025 2025 4 10 10 10 30 0.05

energy to produce electricity. The potential capacity of wind power is figured at about 6500 MW for the country, mostly in the eastern sections [7]. As it is observed from Table 2, the availability of wind energy is illustrated by excellent, good and intermediate statements according to average wind speed of each REC. Specific details The case study was introduced in Section ‘Case study’. Some of detailed descriptions are provided in this section. The planning horizon is 16 years and each year is divided into four subperiods and considers three load levels. The first stage starts at the base year. The annual rate of interest on capital was set at 10%, with present value factors for the costs of investment and operation. General information required to perform this study is presented in Table 3. Load model With rapid annual growth of 5–8% electric consumption, the grid is confronted by a challenging planning problem for the years Table 4 Peak load ratio for each subperiod. Period

Peak load ratio

1 2 3 4

0.8996 1 0.8936 0.8348

1 0.95 0.9 0.85

Demand (P.U)

One of the main results of this effort has been to incite the private sector to take part in power industry development. For this purpose, one of the appropriate methods is to develop DG technologies due to their low capital investment and other benefits. Since DG technologies are conceptually defined as local resources and also the prioritization of various technologies depends on the regional potential of primary energy availability (conventional or renewable), in this paper the Iranian Power Grid are considered as case studies to present a strategic policy making process for the prioritization of DG technologies in each region. Since Kish (R17) is not connected to the network, it is not considered in this study. So, the topology of 16 RECs in the network and the potential availabilities of the primary energy sources of these regions are shown in Fig. 8 and also presented in Table 2. Although conventional fossil fuel resources like oil and gas are mostly located in the south and southwest of country, these fuels are available throughout country based on established oil and gas pipelines. Therefore, as it is shown in Table 2, the availability of these resources is significant in all RECs. As it is observed from Table 2, the availability of oil and gas resources are illustrated by excellent, good and intermediate statements. Excellent is used for the regions that have enormous resources of oil and gas, good denotes to the regions that are adjacent to oil and gas resources. Although these fuels are available throughout the country based on the pipelines and transportation, remote areas which are placed on the end of pipelines, suffer shortages of fuel in some seasons of the year. Therefore the intermediate statement is issued for regions far from the resources. It is known that renewable energies such as wind, solar, and geothermal are relatively expensive and limited in availability. However, to mitigate the environmental impacts to the planet and the risk of depending only on few sources of energy, there is an increasing investment in renewable energy sources. On the other hand, Iran enjoys only a moderate supply of wind power, though some regions have continuous airflows with sufficient

0.8 0.75 0.7 0.65 0.6 0.55 0.5

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Time (P.U) Fig. 9. Linearly approximated load duration curve of the Iranian Power Grid.

800

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

Table 5 Forecasted REC loads [mw] for the Iranian Power Grid. REC No.

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

R01 R02 R03 R04 R05 R06 R07 R08 R09 R10 R11 R12 R13 R14 R15 R16 Total

8960 2618 1222 542 1295 2696 3058 1517 4973 6804 3935 952 1973 2245 3582 1122 47,501

9711 2837 1324 587 1404 2922 3314 1644 5390 7375 4265 1032 2138 2434 3883 1216 51,484

10,525 3075 1435 637 1522 3167 3592 1782 5842 7993 4622 1118 2317 2637 4208 1318 55,797

11,412 3334 1556 690 1650 3434 3894 1932 6334 8666 5012 1212 2513 2860 4563 1429 60,500

12,231 3573 1668 740 1769 3681 4174 2071 6789 9288 5372 1299 2693 3065 4890 1531 64,843

13,105 3829 1787 793 1895 3944 4472 2218 7274 9952 5755 1392 2886 3284 5240 1641 69,473

13,636 3984 1860 825 1972 4104 4654 2308 7569 10,356 5989 1449 3003 3417 5452 1707 72,293

14,385 4203 1962 870 2080 4329 4909 2435 7985 10,924 6317 1528 3168 3605 5752 1801 76,261

15,134 4422 2064 916 2188 4555 5165 2562 8400 11,493 6646 1608 3332 3793 6051 1895 80,230

15,882 4640 2166 961 2297 4780 5420 2689 8816 12,061 6975 1687 3497 3980 6350 1989 84,198

16,631 4859 2268 1006 2405 5005 5675 2815 9231 12,629 7304 1767 3662 4168 6650 2082 88,166

17,462 5102 2381 1057 2525 5255 5959 2956 9693 13,261 7669 1855 3845 4376 6982 2187 92,574

18,335 5357 2501 1109 2651 5518 6257 3104 10,178 13,924 8052 1948 4037 4595 7331 2296 97,203

19,252 5625 2626 1165 2784 5794 6570 3259 10,686 14,620 8455 2046 4239 4825 7698 2411 10,2063

20,215 5906 2757 1223 2923 6084 6899 3422 11,221 15,351 8878 2148 4451 5066 8083 2531 10,7166

21,225 6201 2895 1284 3069 6384 7244 3593 11,782 16,119 9322 2255 4674 5319 8487 2658 11,2524

Table 6 Technical and economic attributes of candidate generating units. Attributes

Generation range (MW) Installation lead time (yr) Life time (yr) F.O.R. (%) Maintenance (h/yr) Capacity factor (%) Efficiency (%) Investment cost ($/kw) Fix operation and Maintenance cost ($/kw-month) Variable operation and Maintenance cost ($/kw-month) Allowed installed generating units in RECs

Type of generating units S325

G130

CC40

DG30

325 5 30 12.9 56 92 38.5 800 0.28 0.36 5

130 2 15 10.2 40 62 33.4 500 0.11 1.23 10

400 5 30 13.67 43 76 50 850 0.11 0.90 3

30 1 23 4.95 7 69 51 713 0.01 0.018 20

WT

MT

CT

DE

12 20 3.2 40 30 40 1000 0.01 0.01

1 20 6.7 20 95 82 950 0.01 0.014

9 30 4.2 350 70 42 550 0.01 0.024

7 20 5.7 250 80 40 350 0.01 0.025

Table 7 Technical and economic attributes of candidate DGs.

Installation lead time (yr) Life time (yr) F.O.R. (%) Maintenance (h/yr) Capacity factor (%) Efficiency (%) Investment cost ($/kw) Fix operation and Maintenance cost ($/kw-month) Variable operation and Maintenance cost ($/kw-month)

to come. Table 4 gives the peak load ratio for each subperiod. As noted earlier, there are 16 interconnected RECs throughout the country. Table 5 gives the forecast REC loads for the Iranian Power Grid in the planning horizon. The peak demand in summer of 2025 is forecast to be 112,524 MW. Linearly approximated load duration curve is shown in Fig. 9. In this study, LDC is considered as a threepiece linear approximation. Generation grid From 56,181 MW installed generation capacity in the Iran Power Grid at the end of 2009, thermal (86.2%), hydro (13.7%), and miscellaneous (1%) are distributed geographic-ally among 16 RECs [36]. Due to the complexity of the generation and electric power transmission facilities, the construction of new electrical facilities is very time consum-ing. If the country is faced with the

DG technologies

12000 11000

Gas Steam Combined Cycle Atomic Hydro

10000 9000

Power (MW)

Attributes

8000 7000 6000 5000 4000 3000 2000 1000 0

2010

2011

2012

2013

2014

2015

2016

Year Fig. 10. Approved projects of generation capacity of under construction units by 2016.

801

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

blackouts phenomenon due to lack of generation and transmission facilities, solving the problem in the short term, even with extra spending, is simply not possible. Therefore, new generating units must be added to the grid in addition to the existence units. For the desired network, four types of candidate power plants are considered. 130 MW gas generating units (G130), 325 MW steam generating units (S325), 400 MW combined cycle generating units (CC40), and 30 MW DG which are considered as a plant collection (i.e., 30 MW wind farm). The DG technologies that are considered as alternatives in this comparative assessment are: Wind Turbines (WT), Micro Turbines (MT), Combustion Turbine (CT) and Diesel Engines (DE). This

analysis is performed in several regions of Iran according to their potential of conventional and renewable energy resources. The comparative assessment of all the individual technologies with all of the possible options can provide an executive summary to the decision makers to allocate their total investment budget to various technologies. Tables 6 and 7 present various technical and economic attributes related to generating units and DG technologies, respectively. It should be noted that, in Table 6, DG data are derived from the average of technical and economic data of the WT, CT, MT, and DE in accordance with Table 7. The generating units approved by the IGMC [36] to expand the generation capacity by

160,000 150,000 140,000

Power (MW)

130,000

Forecast peak demand Lower bound of reserve margin Upper bound of reserve margin PG1

120,000 110,000 100,000 90,000 80,000 70,000 60,000 50,000 40,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025

Year Fig. 11. Iranian Power Grid situation in planning horizon.

Caspian Sea

R06

R03 R05 304 Km

326 Km

R02 R15

R01

R04

R08 R07

422 Km 439 Km

566 Km

R09

615 Km

R12 R10

800 Km

328 Km

R13 R11

R16

R14

REC load center New candidate branch Existing and candidate for expansion branch

Persian Gulf

Oman Sea

Fig. 12. Reduced existing and candidate for expansion and also new addition candidate branches of Iranian grid at the 400 and 230 kV levels.

802

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

2016, was introduced to the WASP software according to Fig. 10. Regarding the generation capacity of units under construction until the end of 2010, total generation capacity of the grid will increase to 65,800 MW. As shown in Fig. 11, total generation capacity of the existence and under construction units, would NOT provide the generation grid constraints until the end of the planning horizon. In this figure, PG1 represents the installed generation capacity curve of under construction and downtime generating unit in the network in the planning horizon.

Table 9 Technical and economic attributes of candidates for addition transmission Lines. Attributes

Type 1

Type 2

Line voltage (kV) Type Number of bundle Number of circuit Resistance (P.U./km) Reactance (P.U./km) Nominal transmission line capacity (MW) Variable investment cost (k$/km) Fix investment cost (k$/km)

230 CANARY 1 1 0.00012 0.000764 397 42 500

400 CURLEW 1 1 0.000035 0.00026 750 85 1600

Transmission grid The problem described is this paper addresses CEP for largescale transmission systems. It is specifically applied to 16 RECs of Iranian Power Grid for the summer electric power peak in 2025 without involving much in the details of downward systems (subtransmission and distribution). As mentioned before, the transmission grids are highly interconnected, managed by 16 RECs owned by a holder company (Tavanir). The transmission grid planning is monitored by Tavanir. The transmission backbone comprises 400 and 230 kV elements, in a totally interconnected network. This network is very complex, even at the 400 kV level. Therefore, for pictorial representation of the network in a simple form, the reduced network, at the 400 and 230 kV level, is shown in Fig. 12. In this figure, each node, in fact, is a cut-set and comprises two or more 400 and 230 kV buses, which are geographically close to one another. This form of representing omits the interconnections within each cut-set [4]. However, squares are the nodes where loads are concentrated, branches drawn as continuous lines denote the initial network (part of the fixed network and candidates for expansion), and branches drawn as dashed lines are candidates for addition (and are not part of the initial network). The 16 RECs, according to

Fig. 12, are connected to each other by 30 transmission lines (the equivalent of 400 and 230 kV transmission lines between regions). This figure also shows the new candidates of transmission lines, considered in Iran Power Grid, in addition to candidate for expansion transmission lines. The technical and economic attributes of 30 branches existing and candidate for expansion are given in Tables 8 and 9. In this table, the capacity of transmission lines and the equivalent resistance and reactance of the transmission lines are in per unit based on 100 MVA. It should be noted that the length of transmission lines in Table 8 and Fig. 12 are calculated based on the geographical coordinates of the gravity center of the different regions load. Numerical results In this section, according to the descriptions and information presented in the previous section, composite generation and transmission expansion planning with distributed generation consideration have been implemented to Iran Power Grid with respect to the proposed method. Therefore, WASP IV is employed to predict generation requirements based on single node analysis. Next, the place of every

Table 8 Existing and candidate for expansion transmission lines data. Line No.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

REC No. From

To

R01 R01 R01 R01 R01 R01 R01 R02 R02 R02 R03 R03 R04 R05 R05 R06 R07 R07 R07 R08 R09 R09 R09 R10 R11 R11 R12 R13 R13 R14

R02 R03 R04 R05 R06 R07 R09 R03 R04 R15 R05 R06 R15 R06 R07 R08 R08 R09 R10 R10 R10 R11 R12 R11 R13 R14 R13 R14 R16 R16

Capacity (MW)

Resistance (P.U.)

Reactance (P.U.)

Length (km)

2443 1292 1408 1528 667 4932 1530 397 1581 878 2631 939 909 1098 1097 397 3632 1500 3643 397 5512 316 1292 3225 1293 1271 1868 2455 836 197

0.0013 0.0029 0.0029 0.0022 0.0085 0.0015 0.0030 0.0076 0.0021 0.0049 0.0014 0.0054 0.0045 0.0041 0.0032 0.0114 0.0011 0.0016 0.0011 0.0114 0.0006 0.0185 0.0024 0.0012 0.0016 0.0180 0.0018 0.0014 0.0105 0.0442

0.0126 0.0334 0.0280 0.0134 0.0968 0.0145 0.0246 0.0502 0.0203 0.0556 0.0121 0.0331 0.0515 0.0392 0.0371 0.0757 0.0105 0.0181 0.0113 0.0757 0.0074 0.1227 0.0345 0.0177 0.0283 0.1196 0.0239 0.0160 0.0597 0.2525

161 230 242 202 500 233 349 308 118 562 97 305 452 322 205 414 206 293 355 368 283 341 273 382 416 458 300 318 473 480

803

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805 Table 10 Generating units and DGs required for the planning horizon – Phase I of proposed approach. Generating units

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

G130 FL30 WT30 LOLP (%)

0 0 0 0.001

0 0 0 0.000

0 0 0 0.003

0 0 0 0.001

0 0 0 0.003

0 0 0 0.001

0 0 0 0.000

0 0 0 0.007

0 0 0 0.000

3 44 0 0.049

29 27 2 0.049

40 0 3 0.050

43 0 0 0.050

44 2 2 0.050

46 6 0 0.050

49 1 1 0.050

Table 11 Allocated generating units and DGs in different RECs for the planning horizon – Phase II of proposed approach. REC No.

2019

2020

2021

2022

2023

2024

2025

R01 R02 R03 R04 R05 R06 R07 R08 R09 R10 R11 R12 R13 R14 R15 R16 Total generating units Total generation (MW)

– – – – – – – – 5FL 10FL 10FL 9FL – 10FL 3G – 3G 44FL 1710

– – 1G + 2WT 2G 5G – – – – 10G + 10FL 1G + 7FL – – 10FL 10G – 29G 27FL + 2WT 4640

5G 8G 3WT – – – – – – 10G 10G – – 5G 2G – 40G 3WT 5290

– 9G 6G – – – – – 7G 10G – – – 8G 2G 1G 43G 5590

– 8G 5G + 2WT 1G – – – – 9G 10G + 2FL – – – 7G 3G 1G 44G 2FL + 2WT 5840

10G – 7G 7G 3G 8G – – 3G 6FL – 1G – – 5G 2G 46G 6FL 6160

5G 1G 1WT – – – – – 10G 10G + 1FL 10G 1G – 10G 1G 1G 49G 1FL + 1WT 6430

generating unit and distributed generation is determined simultaneous with TEP, in which the allocation of generating units, DGs, and transmission lines are specified. The results are presented in the following. For this purpose, justification of DGs in different RECs has been studied in the Iranian Power Grid. It is worth mentioning that this study aims to assess the impact of DG to reduce the cost of transmission network in CEP. As described in [3] according to the cost of

Table 12 Transmission addition elements required for the planning horizon.

1 1 1 1

Number

2 1 1 3

REC No.

Year of operation

From

To

R01 R04 R01 R09

R09 R12 R02 R10

Planned generation capacity (MW)

Line type

2019 2019 2024 2024

determined transmission lines, throu-ghout CEP, prorated costs of transmission network will be added to the cost of centralized power plants [1]. The results of the implementation of the WASP in planning horizon (Phase I of the proposed method), are shown in Table 10. It can be seen that in the first phase of the proposed method, WASP software has chosen 254 big gas generating units (G130) and 88 DGs, in addition to existing generating units under construction, with total capacity of 35,660 MW by 2025. So, the DG contribution in the total amount of installed capacity was 7.4%. According to Fig. 11, since the power system is adequate with respect to generating units under construction until 2025, no generation investment is expected to be proposed in the first 9 years. In the next step of proposed method, DG and centralized generating units has been allocated in 16 RECs of the network. The results of allocating justified centralized generating units and DGs simultaneous with TEP are shown in Tables 11 and 12. Due to over generation in some regions, new transmission lines to nearby regions (R01 and R02) are justified.

7000 6000

G130 FL30 WT30

5000 4000 3000 2000 1000 0

R01 R02 R03 R04 R05 R06 R07 R08 R09 R10 R11 R12 R13 R14 R15 R16

RECs of Iranian Power Grid Fig. 13. Allocated generation capacities of different RECs for the planning horizon.

804

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

Using DG technologies, in addition to technical and environmental benefits, has a great impact on reducing the cost of network expansion. As it was observed, considering techn-ical and economic features of distributed generations, in CEP and appropriate allocating for units construction, resulted in lack of additional costs to the transmission network, and thus reduced the cost of network expansion. Moreover, some factors such as fuel costs and proximity to the load centers have been great influence on allocating the justified generating units. Therefore, considering the coefficients of fuel costs in each region, leads to increasing tendency to establish generat-ing units in RECs with access to less expensive fuel. The allocated generation capacities of each REC is shown in Fig. 13 for the planning horizon. As expected, the most gas units are justified in the regions that have more regional demand and

cheaper fuel is available (i.e., R09, R10, R11, and R14). It also is observed that, conventional DGs (FL30) are justified in the regions where cheaper fuel is available (i.e., R10, R11, and R14), and also in the regions where there is more load and fewer transmission capacity with nearby regions (i.e., R09 and R12). In these cases, DGs are applied in terms of peak load shedding. It can be seen that, renewable DGs (WT30) are justified in the regions where there is appropriate geographical potential (i.e., R03). By doing the optimal GEP, as shown in Fig. 14, the total installed generation capacities of new scheduled generating units restore the generation network adequacy. It represents with PG2 in this figure. It is obvious that, with regard to the environmental impacts, feasibility of DGs based on renewable energy technologies, will considerably increase.

160,000

Forecast peak demand Lower bound of reserve margin Upper bound of reserve margin PG2

150,000 140,000 130,000

Power (MW)

120,000 110,000 100,000 90,000 80,000 70,000 60,000 50,000 40,000 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025

Year Fig. 14. Expanded Iranian Power Grid situation in planning horizon.

Caspian Sea

R06

R03 R02 R05 R15

R01

R04

R08 R07 R09 R12 R10

R13 R11

R16

R14

Persian Gulf REC load center Inter-region load center Inter-region branch Existence and planned branch

Oman Sea

Fig. 15. Improvement of the proposed approach to determine more details of typical REC No. 11 (Fars).

A. Rouhani et al. / Electrical Power and Energy Systems 62 (2014) 792–805

The Improvement of the proposed approach to determine more details of every region (i.e., R11) is best illustrated by a simple sample as shown in Fig. 15. This figure indicates how to apply the proposed method in the large scale networks. Conclusion For a practical large-scale system, the problem of a compo-site generation and transmission expansion planning consider-ing distributed generation and nonuniform geographical fuel supply costs was addressed. The objective was to minimize the overall cost of the system by allocating the overall generation capacity among the grid nodes and determining the new trans-mitssion element additions. The problem was formulated as a mixed integer linear problem solved by a new constructive heuristic approach. The most important benefits of the proposed solution are possibility of considering the most constrains of power plants, the reliability constraints of power plants and transmission network, determination of the exact location of power plants and distributed generation, consider-ing nonuniform geographical fuel supply costs and the geographical potential of each region. The capability of the proposed approach was assessed in 16 interconnected RECs of Iran Power Grid as a large-scale network with different climatic conditions and energy resources. Four common DG technologies considered in this paper were: wind turbine (as a renewable technology), micro turbine, combustion turbine, and diesel engine (as conventional technologies). The proposed strategy can help governments to gain information about the preferred DG technologies for each REC of Iran in order to keep moving towards sustainable development. From the case study, it can be seen that the proposed method would be a powerful tool for the composite expansion planning for large scale power systems. References [1] Rouhani A. Composite generation and transmission expansion planning with distributed generation consideration, M.S. thesis, Dept. Elec. Eng., Zanjan Univ., Zanjan, Iran; 2010. [2] Wang X, McDonald JR. Modern power system planning. New York: McGrawHill; 1994. [3] Liu G, Sasaki H, Yorino N. Application of network topology to long range CEP of generation and transmission lines. Elect Power Syst Res 2001;57:157–62. [4] Meza JLC, Yildirim MB, Masud ASM. A model for the multiperiod multiobjective power generation expansion planning. IEEE Trans Power Syst 2007;22(2):871–8. [5] Kannan S, Raja Slochanal SM, Padhy NP. Application and comparison of metaheuristic techniques to generation expansion planni-ng problem. IEEE Trans Power Syst 2005;20:466–75. [6] Park JB, Park YM, Won JR, Lee KY. An improved genetic algorithm for generation expansion planning. IEEE Trans Power Syst 2000;15(3):916–22. [7] Wang J, Shahidehpour M, Li Z, Butterud A. Strategic generation capacity expansion planning with incomplete information. IEEE Trans Power Syst 2009;24(2):1002–10. [8] International Atomic Energy Agency (IAEA), Wien Automatic System Planning (WASP) Package, A Computer Code for Power Generation System Expansion Planning, Version WASP-IV User’s Manual. Vienna, Austria, IAEA; 2001. [9] Sepasian MS, Seifi H, Akbari Foroud A. A multiyear security constrained hybrid generation-transmission expansion planning algorithm including fuel supply costs’’. IEEE Trans Power Syst 2009;24(3):609–1618.

805

[10] Choi J, Mount TD, Thomas RJ. Transmission expansion planning using contingency criteria. IEEE Trans Power Syst Nov. 2007;22(4):2249–62. [11] Ruiz PA, Contreras J. An effective transmission network expansion cost allocation based on game theory. IEEE Trans Power Syst 2007;22(1):136–44. [12] Gajbhiye RK, Naik D, Dambhare S, Soman SA. An expert system approach for multi-year short-term transmission system expansion planning: an Indian experience. IEEE Trans Power Syst 2008;23(1):226–37. [13] Bayona GL, Arriaga IJP. Chopin, a heuristic model for long term transmission expansion planning. IEEE Trans Power Syst 1994;9(4):1886–94. [14] Bahiense L, Oliveira GC, Pereira M, Granville S. A mixed integer disjunctive model for transmission network expansion. IEEE Trans Power Syst 2001;16(3):560–5. [15] Sepasian MS, Seifi H, Akbari Foroud A, Hosseini SH, Mohseni Kabir E. A new approach for substation expansion planning. IEEE Trans Power Syst 2006;21(2):997–1004. [16] Delgado D, Claro J. Transmission network expansion planning under demand uncertainty and risk aversion. Electr Power Energy Syst 2013;44:696–702. [17] Alvarez Lopez J, Ponnambalam K, Quintana VH. Generation and transmission expansion under risk using stochastic programming. IEEE Trans Power Syst 2007;23(3):369–1378. [18] Wenyuan Li, Billinton R. A minimum cost assessment method for composite generation and transmission system expansion planning. IEEE Trans Power Syst 1993;8(2):628–35. [19] Xia Q, Song YH, Kang C, Xiang N. Novel models and algorithms for generation unit location optimization. IEEE Trans Power Syst 1997;12(4):1584–90. [20] Kaymaz P, Valenzuela J, Park CS. Transmission congestion and competition on power generation expansion. IEEE Trans Power Syst 2007;22(1):156–63. [21] Hyung Roh J, Shahidehpour M, Fu Y. Market-based coordination of transmission and generation capacity planning. IEEE Trans Power Syst 2007;22(4):1406–19. [22] Dugan RC, McDermott TE, Ball GJ. Planning for distributed generation. IEEE Ind Appl Mag 2001:80–8. [23] Moravej Z, Akhlaghi A. A novel approach based on cuckoo search for DG allocation in distribution network. Electr Power Energy Syst 2013;44:672–9. [24] Ebrahimi R, Ehsan M, Nouri H. A profit-centric strategy for distributed generation planning considering time varying voltage dependent load demand. Electr Power Energy Syst 2013;44:168–78. [25] Zangeneh A, Jadid S, Rahimi-Kian A. A hierarchical decision making model for the prioritization of distributed generation technologies-A case study for Iran. Energy Policy 2009;37(12):5752–63. [26] Chiradeja P, Ramakumar R. An approach to quantify the technical benefits of DG. IEEE Trans Energy Convers 2004;19(4):764–73. [27] Wang C, Nehrir MH. Analytical approaches for optimal placement of distributed generation sources in power systems. IEEE Trans Power Syst 2004;19(4):2068–76. [28] Javadian SAM, Haghifam M-R, Fotuhi Firoozabad M, Bathaee SMT. Analysis of protection system’s risk in distribution networks with DG. Electr Power Energy Syst 2013;44:688–95. [29] Javadian SAM, Haghifam M-R, Bathaee SMT, Fotuhi Firoozabad M. Adaptive centralized protection scheme for distribution systems with DG using risk analysis for protective devices placement. Electr Power Energy Syst 2013;44:337–45. [30] Ault GW, Mc Donald JR. Planning for distributed generation within distribution networks in restructured electricity markets. IEEE Power Eng Rev 2000;20:52–4. [31] Agalgaonkar AP, Kulkarni SV, Khaparde SA. Evaluation of configuration plans for DGs in developing countries using advanced planning techniques. IEEE Trans Power Syst 2006;21(2):973–83. [32] Haffner S, Alves Pereira LF, Pereira LA, Barreto LS. Multistage model for distribution expansion planning with distributed generation-Part I: Problem formulation. IEEE Trans Power Del 2008;23(2):915–23. [33] Haffner S, Alves Pereira LF, Pereira LA, Barreto LS. Multistage model for distribution expansion planning with distributed generation-Part II: numerical results. IEEE Trans Power Del 2008;23(2):915–23. [34] Gitizadeh M, Azizi Vahed A, Aghaei J. Multistage distribution system expansion planning considering distributed generation using hybrid evolutionary algorithms. Appl Energy 2013;101:655–66. [35] Billinton R, Allan RN. Reliability evaluation of power systems. 2nd ed. New York: Plenum Press; 1996. [36] Iran Grid Management Company official website: .