1

New Efficient Reserve Rate Index of Power System Including Renewable Energy Generators
Jeongje Park, Student Member, Jaeseok Choi, Senior Member, Mohammad Shahidehpour, Fellow, Kwang Y. Lee, Fellow and Roy Billinton, Life Fellow, IEEE
2) Resources and power control of REG are difficult relatively than the conventional generators as like as the nuclear, coal, LNG, etc. power plants. 3) Outage of renewable energy may give the same effect with forced outage of conventional generators to demand in viewpoint of reliability. Fig.1 shows differences between the uncertainties of conventional generator plant and these of the renewable energy resource power plant[3,4].

Abstract This paper proposes a new effective installed reserve rate in order to evaluate reliability of power system considering renewable energy generators(REGs), which include uncertainty of resource supply. It is called EIRR(effective installed reserve rate) in this paper. It is developed with considering capacity credit based on ELCC( (Effective Load Carrying Capability) by using LOLE reliability criterion. While the conventional installed reserve rate index yields overevaluation reliability of renewable generators, the proposed EIRR describes actual effective installed reserve rate. However, it is not the probabilistic reliability index as like as LOLE or EENS but another deterministic effective reliability index. The proposed EIRR is able to evaluate the realistic contribution to the reliability level for power system considering wind turbine generators and solar cell generators with high uncertainty in resource supply. The case study in model system as like as Jeju power system size in South Korea presents a possibility that the proposed EIRR can be used practically as a new deterministic reliability index for generation expansion planning or operational planning in future. Index TermsInstalled reserve rate, REG(Renewable energy generator), ELCC, Probabilistic reliability evaluation

I. INTRODUCTION he utilization of renewable resources such as wind and solar, tidal to generate electric power has been receiving considerable attention in recent years [1],[2]. Wind energy in particular has been fast growing and is recognized as the most successful energy source of all available sources. A most of renewable energy generators(REGs) have following characteristics. 1) Location of REG is distributed because the successful operation of the REG depends on their renewable resource supply very strongly.
This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant funded by the Korea government (MEST) (No. R01-2008000-10567-0). Jeongje Park, and Jaeseok Choi are with Dep.of Electrical Engineering, ERI, Gyeongsang National University, Jinju, GN, Korea (e-mail: pakjan81@gmail.com, jschoi@gnu.ac.kr). Mohammad Shahidehpour is with Department of Electrical and Computer Engineering, IIT, USA (e-mail: ms@iit.edu). Kwang Y. Lee is with Dept. of Elect. & Comp. Eng., Baylor University, Waco, TX, USA (e-mail: kwang_y_lee@baylor.edu) Roy Billinton is with Department of Electrical Engineering, Univ. of Saskatchewan, Saskatoon, Canada (e-mail: roy250@mail .usask.ca).

Fig. 1 Uncertainties of the renewable energy resource power and conventional plants

From the Fig.1, There are two kinds of aleatory uncertainty and epistemic uncertainty in power system[3,4]. While uncertainties of the forced outage of generator, cooling system circular pump and boiler system at power plant are included to former, the uncertainties of load forecasting or supply of resource are included to latter. Therefore, as the latter gets more detail and accuracy information system, the uncertainty will be decreased. By the way, however, demands will have no idea without an experience and identifying same outage of electrical energy in view point of no supply of electrical energy no matter whether it comes from forced outage of plant and grid or no supply of renewable resource. And so, in recent, planners and operators of power system meet with an important problem about how should the REG with high uncertainty level be handled and solved in order to operate successful system and make a reasonable planning? In order to evaluate the actual contribute level of the REGs in view point of system reliability, the capacity credit has been developed. They are based on ELCC or capacity factor. But, conventional installed reserve rate is more familiar index than capacity credit index etc. to non-expert as like as citizen.

978-1-4244-6266-7/10/$26.00 2010 IEEE

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.

This paper proposes a new effective installed reserve rate in order to evaluate reliability of power system considering renewable generators, which include uncertainty of resource supply. It is called EIRR(effective installed reserve rate) in this paper. It is developed with considering capacity credit based on ELCC(Effective Load Carrying Capability) by using LOLE reliability criterion. The proposed method is applied to Jeju system in South Korea. II. ELCC AND CAPACITY CREDIT A. ELCC The ELCC(Effective Load Carrying Capability) was proposed by L.L. Garver in 1966[5]. The eventual purpose is to evaluate the actual reliability contribution level of the generator with uncertainty for generation system expansion planning. It is proposed usefully in order to assess how much does a new generator cover future load with considering uncertainty of the generator?[5]. It is defined as difference of increasing system loads between before and after the new generator penetration in looking for same target risk level as Fig. 2 [5].

system added wind farms, due to the nature of wind, the reliability of wind farms is lower than conventional power plants. All the extra added wind power can not supply load demand fully. The ELCC, specially, is more important and reasonable approach model for reliability evaluation of wind farms because wind turbine generator has usually high epistemic uncertainty. Additionally, the method which evaluates the LOLE of power system including wind turbine generators is different from it of conventional power plants. As it is, a two-state model used popularly for conventional generator can be used no more for wind turbine generator. A multi-state model should be used. B. Capacity Credit in View Point of Reliability Two kinds of methods are developed for assessing capacity credit. One uses ELCC in viewpoint of reliability as previous comments. Anther uses capacity factor in view point of economics. They have a little bit difference. The ELCC is used in this paper because it is focused on development of reliability new index rather than economics viewpoint. Using above ELCC, the capacity Credit can be formulated as Eq. (4) by the extra load divided new generator capacity. It means how much is the actual reliability contribution of new generator and it is one of the actual contribution assessment methods [6-8].
Capacity Credit ( ) = L 100 CA [%]

(4)

III. NEW INDEX OF EFFECTIVE INSTALLED SUPPLY RATE CONSIDERING RENEWABLE GENERATORS
Fig. 2 ELCC (Effective Load Carrying Capability)

Therefore, the effective load-carrying capability (ELCC) is a way to measure a power plants capacity credit based on its influence on overall system reliability. When a new generating unit is added into a power system, the ELCC of the unit is the amount of extra load that can be served while keeping a suitable level of reliability. The suitable level is the LOLE of the system before addition of the new generating unit. Thus, The LOLEs of existing and potential systems are equal. The concept of ELCC is Introduced by Eq. 1~4[6-8] LOLEE=LOLEp 1 2 ELCC=L 3

A most of the conventional deterministic reliability index, the Installed Reserve Rate (IRR) can be formulated as Eq. (5)
IRR = g (C i , L p ) =

Lp

Lp

100 [%]

(5)

Where, P(XE>CE-Li) and P(XP>(CE+CA)-(LI+L)) are the loss of load probabilities(LOLPS) of the existing and potential systems. The LOLP is the LOLE dividing the period. L is the extra load that can be served by the additional generation. CA is the added generators capacity. For example, in a power

It is one of very simple, practical and useful deterministic reliability indices. But, it can not consider uncertainty as like as the outage rate of generators. And so, when it is very low, the index is fit well to actual but it is no more practical index for power system including REG of high uncertainty of resources supply. For example, the present install capacity is 550MW and peak load is 500MW. Assume that new WTG of the capacity 100MW is penetrated. The conventional install reserve rate yields IRR=30%. But, ELCC range of 100MW of WTG has from 20~30MW and so the percent capacity credit is 0.2~0.3. It is more reasonable, therefore, that the actual effective installed reserve rate should be assessed 14~16%. This paper proposes Eq (6) for a new effective installed reserve rate in order to evaluate reliability of power system considering renewable energy generators(REGs), which include uncertainty of resource supply by using new

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.

parameter(), which is capacity credit. It is called EIRR (effective installed reserve rate) in this paper.
HNM C/S

EIRR = g (C k , k , L P ) C C k + C Rk k L P k GR 100 = k GC LP

HVDC JEJU-MS

Jeju T/P

G3 : 40MW
G3

JCN
D D/P : 40MW
T2 T3

HWN-WF
10MVAr

Filter : 27.5MVAr X4 ShR : 1375X2.275

G1 G2

T1

SNJ SJJ

DJJ

C1, C2 : Syn, Con

T/P : 10MW X1 75MW X2

SSN
20MVAr

(6)

HLM-WF

20MVAr 20MVAr

SSN-WF SSN-MS

Where, CCk: capacity of # k- conventional generator [MW] CRk: capacity of # k- REG [MW] : capacity credit [pu] GC: a set of conventional generators GR: a set of REG The eventual purpose of this study is focused on REGs actual reliability contribution problem. And so, the conventional generators are still on not consideration of their capacity credit even if they can be considered. IV. FLOW CHART

HLM C/C G2
S 35MW X3 10MVAr 10MVAr

G1

GNS-MS

AND NMJ
T4 T3 T/P : 100MW X2

HNR SSG

SGP-MS

D1 D2 D3 D4 D/P : 10MW X4

Fig. 4. A Jeju power system in Korea (WF: Wind Farm, MS: Measuring station) TABLE 1 THE GENERATORS DATA OF JEJU ISLAND POWER SYSTEM
Name Type Capacity Num. [Gcal/ [MW] MW2h] 50 1 30 1 20 1 75/ 2 0.004 150* 100 2 0.004 10 1 0.062 75 2 0.003 35 2 0.004 35 1 0.004 40 1 0.025 10 4 0.006 945 18 CO2 Fuel cost [Gcal/ [Gcal emission FOR [$/Gcal] MWh] /hr] [Ton/MWh] 0.010/ 1.512 45.207 43.300 0.65 0.028* 1.512 45.207 43.300 0.65 0.012 2.100 5.971 43.599 0.96 0.015 1.832 30.231 43.599 0.70 0.012 2.401 20.320 77.909 0.95 0.013 2.401 20.320 77.909 0.95 0.013 0.364 28.484 43.599 0.59 0.018 1.999 1.360 43.300 0.62 0.018 -

The step-by-step process for assessing new EIRR reliability index of power systems including REGs is shown in Fig. 3.

1 HWN WTG 2 SSN WTG 3 HLM WTG 4* HVDC DC 5 6 7 8 9 10 11 NMJ3 JJU1 JJU2 HLM1 HLM1 JJU3 NMJ1 Total T/P T/P T/P G/T S/T D/P D/P

(*, HVDC is modeled as three states.)

(a) Daily peak load variation curve Fig. 3. A flow chart describing the proposed method.

V. CASE STUDY A test system as size as Jeju Island power system including WTG growing rapidly in recent in South Korea as shown in Fig 4 is used to demonstrate the usefulness of the proposed new index method. Therefore, WTG is only considered for REG in test system. The data of Jeju island power system is are shown in Table 1. The system load duration curve is given in Fig. 5. Data related to three wind farms(WF) are constructed in three different locations; Hangwon(HWN), Sungsan (SSN) and Hanlim (HLM) are given in Table 2.

(b) Daily peak load duration curve Fig. 5. Load variation and daily peak load duration curves of model system

Table 2 and Table 3 show the wind speed characteristics of three wind farms and specification of WTG in Test system

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.

respectively. Table 4 shows coefficients of WTG power characteristics functions of the test system obtained by using the Table 2 and Table 3. The calculation process is introduced in detail and well in appendix A1 and References[9],[10] and [14]. Fig 6 shows the outage capacity probability density functions for three wind farm. Therefore, the probabilistic reliability indices and production cost can be assessed using Eq. (A1), Eq. (A2), Eq. (A3) and Eq. (A4). Table 5 shows the comparative results for W/O(without) WTGs case and W(with) WTGs case.
TABLE 2 WIND SPEED OF THREE WIND FARMS IN TEST SYSTEM (c) Outage capacity density function(fo3) of WTG at HLM-WF Fig. 6. Outage capacity PDF(foWTG) of equivalent WTGs of wind farms in model system TABLE 5 RELIABILITY INDICES OF TEST SYSTEM FOR TWP CASES

WF Name Wind speed range Mean wind speed Standard deviation

Wind speed[m/sec] HWN-WF SSN-WF 0-35 8.5 7 0-40 7.6 6

HLM-WF 0-45 6.4 5

Cases W WTGs W/O WTGs

LOLE [days/year] 0.19 0.33

EENS [MWD/year] 7.01 12.88

EIR [pu] 0.99996 0.99994

TABLE 3 SPECIFICATION OF WTG IN TEST SYSTEM

WTG capacity Cut-in speed(Vci) Rated speed(VR) Cut-out speed(Vco)

50MW 5m/sec 16m/sec 25m/sec

30MW 5m/sec 15m/sec 25m/sec

20MW 5m/sec 14m/sec 25m/sec

Fig. 7 shows ELCCs of three wind farms in the test system. It can be obtained by assessing the extra load (ELCC) for With WTG which is still maintaining LOLE level as same as Without WTG even if the peak loads are increased. Fig. 8 shows ELCC of total capacity of all WTGs in test system. Table 6 shows the capacity credits of WTG at three wind farms.

TABLE 4 THE COEFFICIENTS OF WTG POWER FUNCTION OF THE TEST SYSTEM

coefficients A B [m/sec]-1 C [m/sec]-2

HLM-WF 0.09275 -0.06486 0.00926

SSN-WF 0.1111 -0.06296 0.00815

HWN-WF 0.12034 -0.06 0.00719

(a) ELCC at HWN-WF

(a) Outage capacity density function(fo1) of WTG at HWN-WF

(b) ELCC at SSN-WF (b) Outage capacity density function (fo2) of WTG at SSN-WF

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.

The results are summarized in Table 7.

TABLE 7 THE COMPARISON OF THE CONVENTIONAL INSTALLED RESERVE RATE AND THE
PROPOSED EFFECTIVE INSTALLED RESERVE RATE

IRR EIRR EIRRALL

Total Capacity [MW] 945 945 945

Installed Reserve Rate [%] 38.77 26,73 26.73

(c) ELCC at HLM-WF Fig. 7. ELCCs of wind farms in model system

In the test system, WTGs are only considered because the WTG are growing rapidly on the Jeju Island system. But, SCGs(Solar Cell Generators) are also evaluated as same method proposed in this paper[10]. VI. CONCLUSIONS This paper proposes a new effective installed reserve rate in order to evaluate actual reliability contribution level of renewable generators penetrated rapidly in power system in recent. The REGs include high uncertainty of resource supply relatively rather than conventional generators. A most of the resources is uncontrollable and no storage of itself. Therefore, they should be handled carefully for system reliability evaluation. In this paper, it is called, EIRR(effective installed reserve rate) was proposed newly in order to assess actual installed reserve contribution. It is developed with considering capacity credit based on ELCC(Effective Load Carrying Capability) by using LOLE reliability criterion. The proposed method is applied to Jeju system in South Korea. VII. ACKNOWLEDGMENT This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant funded by the Korea government (MEST) (No. 2009-0058632). The support of the Advanced Power Network Reliability Research Center (APRRC) is acknowledged. VIII. REFERENCES
1. 2. Nick Jenkins, Ron Allan, Peter Crossley, David Kirschen and Goran Strbac: EMBEDDED GENERATION, 2000, PP. 31-38. Wind Power Today, Federal Wind Program Highlights, U.S. Department of Energy, Energy Efficiency and Renewable Energy, April, 2005. Roy Billinton and Dange Huang, Aleatory and Epistemic Uncertainty Considerations in Power System Reliability Evaluation , PMAPS, May 25-29, 2008. Wu Liang, Jeongje Park, Jaeseok Choi, A. A. El-Keib, Mohammad Shahidehpour and Roy Billinton, Probabilistic Reliability Evaluation of Power Systems Including Wind Turbine Generators Using a Simplified Multi-State Model: A Case Study IEEE PES GM2009, July 26-30, 2009, Calgary, AB, Canada. L. L. Garver, Effective load carrying capability of generating units, IEEE Transactions on Power Apparatus and Systems, vol. PAS-85, no. 8, 1966, pp. 910-919. Claudine D Annunzio, Surya Santoso, "Noniterative Method to Approximate the Effective Load Carrying Capability of a Wind Plant", IEEE Transactions on Energy Conversion, vol. 23, no. 2, June, 2008, pp. 544-550. Lennart S oder and Mikael Amelin, "A review of different methodologies used for calculation of wind power capacity credit" IEEE GM2008, Pittsburg, PA, USA. Cornel Ensslin, Michael Milligan, Hannele Holttinen, Mark O'Malley, and Andrew Keane, "Current Method to Calculate Credit of Wind Power, IEA Collaboration" IEEE GM2008, Pittsburg, PA, USA.

Fig. 8. ELCC(Effective Load Carrying Capability) of all WTG TABLE 6 ELCC AND CAPACITY CREDITS OF WTG FARMS Capacity ELCC Capacity Credit WTG-WF [MW] [MW] [%] HWN SSN HLM Total All WTG 50 30 20 100 100 8 6 4 18 18 16 20 20 18.67* 18 0.18

[pu] 0.16 0.2 0.2

(where, *: average) Therefore, the proposed new reliability index, EIRR(Effective Installed Reserve Rate) is calculated as follows.
EIRR = (845 + 50 0.16 + 30 0.2 + 20 0.2 681) 100 = 26.73[%] 681

3. 4.

The traditional installed reserve rate of the test system is IRR= (945-681)x100/681 = 38.77%. And so, the difference value between new index EIRR and traditional index, IRR is 12.04%. It is identified that IRR may occur over reliability level than actual reliability contribution of WTG. The reason comes from high uncertainty of resource supply of REG. Additionally, it is an interesting result that EIRRALL using Total ELCC yields same value as follow.
EIRR ALL = (845 + 100 0.18 681) 100 = 26.73[%] 681

5. 6.

7. 8.

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.

6 9. 10. Jeongje Park, Wu Liang, Jaeseok Choi, and Junmin Cha, "Probabilistic Production Cost Credit Evaluation of Wind Turbine Generators", J. of KIEE, Vol. 57, No. 12, pp.2153-2160, Dec. 2008. (Written by Korean) Jeongje Park, Wu Liang, Jaeseok Choi, A. A. El-Keib, Mohammad Shahidehpour and Roy Billinton, Probabilistic Reliability Evaluation of Power System Including Solar/Photovoltaic Cell Generator IEEE PES GM2009, July 26-30, 2009, Calgary, AB, Canada. P. Giorsetto and K. F. Utsurogi, Development of a new procedure for reliability modeling of wind turbine generators, IEEE Trans. Power App. and Systems, vol. PAS-102, no. 1, 1983. X. Wang, H. T. Dai, and R. J. Thomas, Reliability modeling of large wind farms and associated electric utility interface systems, IEEE Trans. Power App. and Systems, vol. PAS-103, no. 3, 1984. C. Singh and A. Lago-Gonzalez, Reliability modeling of generation systems including unconventional energy sources, IEEE Trans. Power App. and Systems, vol. PAS-104, no. 5, May 1985. Rajesh Karki, PoHu, Roy Billinton, A Simplified Wind Power Generation Model for Reliability Evaluation, IEEE Trans. on Energy Conversion, Vol.21, No.2, June, 2006, pp.533-540. Mukund R. Patel: Wind and Solar Power Systems, CRC press, 1999, pp.125-145. M. K. C. Marwali, H. Ma, S. M. Shahidehpour, and K. H. AbdulRahman, "Short-term generation scheduling in photovoltaic-utility grid with battery storage" IEEE Trans. Power Syst., vol.13, no.3, Aug. 1998, pp.10571062. Ruey-Hsun Liang, and Jian-Hao Liao, "A Fuzzy-Optimization Approach for Generation Scheduling With Wind and Solar Energy Systems" IEEE Trans. Power Syst., vol.22, no.4, Nov. 2007 , pp.1665-1674.

11. 12. 13. 14. 15. 16.

0: the original Inverted Load Duration Curve (ILDC) x: random variable of NS: the total state number of the simplified multi-state model Ckj: outage capacity of state k of generator j qkj: the probability correspond the outage capacity of state k of generator j foj: the outage capacity probability distribution function (OCPDF) of generator j Therefore, the Loss of Load Expectation (LOLE), the Expected Energy Not Supplied (EENS) and the Energy Index of Reliability (EIR) which are the main reliability indices of power systems are described using ELDC () as (A2)~(A4).
LOLE = NG ( x ) |x = IC
EENS = NG ( x) dx
IC IC + Lp

[Days/year]

(A2)

17.

[MWD/year]

(A3) (A4)

IX. APPENDIX A1: The power output model of WTG The power output model of WTG is combined with the wind speed model as shown in Fig. A1 to yields the multi-state model. Each state has a pair of associated parameters; namely the power (Pi) and probability (PBi). Eventually, the reliability indices of WTG are evaluated by Effective Load Duration Curve (ELDC) using convolution integral method based on the multi-state model.
Power[MW ]
( Pn , PBn )

EENS EIR = 1ED

[pu]

Where, IC: the system installed capacity (=Cj) Cj : the capacity of generator j Lp: the peak load of demand NG: the total number of generators ED: total demand energy [MWD] NG: final ELDC convoluted with the foj of all generators X. BIOGRAPHIES
Jaeseok Choi (S88M91SM05) was born in Kyeongju, Korea, in 1958. He received the B.Sc., M.Sc., and Ph.D. degrees from Korea University, Seoul, in 1981, 1984, and 1990, respectively. He was a Postdoctoral at the University of Saskatchewan, Saskatoon, SK, Canada, in 1996. Since 1991, he has been on the faculty of Gyeongsang National University, Jinju, Korea, where he is a professor. Jeongje Park (S08) was born in Jinju, Korea in 1981. He received the B.Sc. and M.Sc. degrees from Gyeongsang National University, Jinju, in 2007 and 2008, respectively. His research interest includes Transmission Expansion Planning using Reliability Evaluation of Power Systems. He is now working forward a Ph.D. degree at Gyeongsang National University. Mohammad Shahidehpour (F01) is the Carl Bodine Distinguished Professor and Chairman in the Electrical and Computer Engineering Department at the Illinois Institute of Technology, Chicago. Dr. Shahidehpour is an IEEE Distinguished Lecturer and has lectured in 30 countries on electricity restructuring issues. He is the Vice President of Publications for IEEE/PES. Roy Billinton (S59M64SM73F78) received the B.Sc. and M.Sc. degrees from the University of Manitoba, Winnipeg, MB, Canada, and the Ph.D. and D.Sc. degrees in electrical engineering from the University of Saskatchewan, Saskatoon, SK, Canada. He was with the System Planning and Production Divisions of Manitoba Hydro. He joined the University of Saskatchewan in 1964. He is the author and coauthor of eight books on reliability evaluation and over 850 papers on power system reliability evaluation, economic system operation, and power system analysis. Dr. Billinton is a fellow of the EIC, the CAE, and the Royal Society of Canada, and a Registered Professional Engineer in the Province of Saskatchewan, Canada.

( Pi + 2 , PBi + 2 ) ( Pi +1 , PBi +1 ) ( Pi , PBi )

pdf

wind Velocity

wind
5 4 3
2

+ 2 + 3

+ 4

+ 5

Velocity

Fig. A1. The combination between wind speed model and WTG power output model for obtaining available capacity PDF of WTG

A2: Probabilistic reliability indices of WTG The probabilistic reliability indices have been used extensively for generation expansion. They can be calculated by using the Effective Load Duration Curve (ELDC, ) as described in (A1).
j = j -1 f oj = (1- q kj ) j -1 ( x ) + q kj j -1 ( x - C kj )
k =1 k =1 NS NS

(A.1)

Where,

Authorized licensed use limited to: Hindustan College of Engineering. Downloaded on July 31,2010 at 08:10:46 UTC from IEEE Xplore. Restrictions apply.