Sayyed Ali Akbar Shahriari 1, 1,* Haidar Samet ,
J. Electrical Systems 10-2 (2014): 149-155 Regular paper
JES Journal of Electrical Systems
Optimal Allocation of Fault Current Limiters and Distributed Generations in the Presence of Remote Controllable Switches In this paper allocating of DGs and optimal number, locations and size of solid State Fault Current Limiters (SSFCLs) is performed for a case study contains modern protective devices. Genetic algorithm (GA) is employed for optimal allocating of DGs and SSFCLs. The distribution system contains remote controllable switches in addition to conventional tools such as fuses, relays and circuit breakers. The results demonstrate that minimization of distribution protection system cost, is still valid for a distribution system, which contains modern protective devices.
Keywords: distribution protection system, fault current limiter, genetic algorithm , remote controllable switches
1. Introduction
The electricity demand has increased a lot from the early 19th century. But large generators cannot be over loaded to provide this demand. Beside that the transmission and distribution costs also increased a lot from the 19th century. So, new technologies such as distributed generations (DGs) have come up to solve the mentioned problems. These sources are cost efficient and easy to install [1]. Connecting these generations to the grid increases the level of fault current. This fault current may be more than the rating of the existing protective devices such as fuses, circuit breakers and remote controllable switches. Furthermore, due to connecting distributed generation to the distribution system the system are not operated in the radial configuration yet. This change causes the protective coordination to be lost and the unacceptable operation of protective devices may occur. Several ideas have been introduced as possible solutions to overcome these problems. Some of them suggest to replace protection devices and to employ the intelligent and microprocessors based method [2-5]. But these methods are expensive and increase complexity. Some references propose to decrees the size of DGs [6, 7]. But it is not costefficient, because the same price as introducing larger DGs should be paid. Employing fault current limiters (FCLs) has been introduced as one of the solution to limit the fault current produced by DG and prevent changing the protective devices [8, 9]. Due to the cost of FCLs, it is necessary to determine the minimum number of FCLs. The optimal sizes and locations of FCLs were determined in [10]. But this reference finds the optimal sizes and location of FCLs when the locations of DGs are known. *
Corresponding author: H. Samet, School of Electrical and Computer Engineering, Shiraz University, Engineering Faculty No. 1, Zand St., Shiraz, IRAN. E-mail:
[email protected] 1 School of Electrical and Computer Engineering, Shiraz University. Copyright © JES 2014 on-line : journal/esrgroups.org/jes
S. A. A. Shahriari et al: Optimal Allocation of FCL & DG in the Presence of RCS
This paper determines the optimal locations of DGs and optimal locations and sizes of FCLs simultaneously. What's more this paper considers a case study that contains modern protective devises. That means the distribution system studies in this paper contains remote controllable switches in addition to conventional tools such as fuses, relays and circuit breakers. 2. Installing solid state FCL in series with DG Connecting DGs to the distribution system introduces several problems to distribution networks protection such as false tripping of feeders, exceeding the interruption capacity of circuit breaker and fuses from its rating, fuse-relay miss-coordination and relay - relay miss-coordination [10, 11]. At present paper, Solid State FCL (SSFCL) designed in [12] is used to minimize the mentioned problems. In Fig. 1, a modern distribution network containing remote controllable switches is considered. The three remote controllable switches are installed at feeder 1 to 3. Relays and the circuit breakers are installed at the beginning of network and branch 3 (b3) and 4 (b4). Table 1 and 2 shows the distribution system parameters and Load parameters, respectively. It is planned to connect three DGs to the system. By connecting DGs to the system, the DGs inject fault current and the mentioned problems are occurred. To lessen these problems, the SSFCL is installed in distribution system. In this case, during the fault, the SSFCL inserts series impedance into the network and minimizes the effects of DGs on the fault current [10].
Fig.1: Real 13 bus distribution network
150
J. Electrical Systems 10-2 (2014): 149-155
Table 1: Distribution System parameters From 1 2 3 4 5 5 7 8 8 7 11 7
To 2 3 4 5 6 7 8 9 10 11 12 13
R[ohm] 0.176 0.176 0.045 0.089 0.045 0.116 0.073 0.074 0.093 0.063 0.068 0.062
X[ohm] 0.138 0.138 0.035 0.069 0.035 0.091 0.073 0.058 0.093 0.05 0.053 0.053
Table 2: System Load parameters L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12
Active[kW] 890 628 112 636 474 1342 920 766 662 690 1292 1124
Reactive[kVAR] 468 470 764 376 344 1078 292 498 480 186 554 480
3. Cost function and constraints
When DG is installed in distribution system, fuse-relay coordination and relay-relay coordination are likely to face problems. What's more, the fault currents in the system may be greater than interrupt capability of fuses, circuit breakers and remote controllable switches. By reducing fault current by SSFCL the problems arisen by the increase of the fault current level are alleviated. In this paper, GA is used to find optimal DGs' locations and optimal number, locations and sizes of SSFCLs to minimize the total cost of fuses, circuit breakers and remote controllable switches that must be replaced and also the cost needed to change relay setting to re-coordinate relays. Because of SSFCL's cost, it is necessary to considered SSFCL's cost in the cost function to minimize the cost of distribution protection system in presence of DGs.
151
S. A. A. Shahriari et al: Optimal Allocation of FCL & DG in the Presence of RCS
The cost function is defined as follows:
CF = N f × C fuse + N CB × CCB + N rcs × Crcs + Ccor + N FCL × CFCL + C p (US $)
(1)
13
CP =
∑ Pi i =1
(2)
If i > If min 0 Pi = 100,000 otherwise
(3)
where i = 1, 2 ,.., 13, corresponds to a fault at buses 1 to 13, respectively; Nf, is the number of fuses which must be replaced to meet the increased fault current level and coordination criteria; Cfuse, is the cost of fuse that must be changed; NCB, is the number of circuit breakers that must be changed to meet the increased fault current level in the presence of DGs; CCB, is the cost of circuit breakers that must be changed; Nrcs, is the number of remote controllable switches which must be replaced to meet the increased fault current level; Crcs, is the cost of remote controllable switches that must be changed; Ccor, is the cost needed to re-coordinate relays to establish relay-relay coordination and fuse-relay coordination again ; NFCL, is the number of SSFCLs to be used; CFCL, is the cost of SSFCL that must be employed; Cp, is a penalty value for fault current level. Ifi and Ifmin stand for fault current level at bus i and minimum detectable fault current, respectively [13-20]. The CFCL is calculated based on the number of thyristors used in each SSFCL (SSFCL size) and the cost of each thyristor, reactor and MOV [14-20] were considered. 4. Numerical results A distribution network which is shown in Fig. 1 is considered to analyze results of genetic algorithm. It is supposed to connect three DGs to the network. Their locations are unknown, and hence, they may be located in any bus. To assume the worst case in this study, DGs are considered as ideal sources. Since when DGs are considered as ideal sources, they can produce current without any limitation during the fault. The DGs' parameters considered in this study are shown in Table 3. Table 3: Distributed Generators parameters Parameter Rated Voltage / kv Resistance (Ra) / pu Synchronous Reactance ( Xd )/ pu Transient Reactance ( X’d )/ pu Sub transient Reactance ( X’’d )/ pu
Value 20 0.01 0.9 0.04 0.03
At the first step, after connecting DGs to arbitrary buses, three-phase short circuit fault is applied to all lateral feeders of the system. It can be expected that integrating DGs to the network results in increasing fault current level and in these situations, the interrupting capabilities of all protection devices may not be enough to operate under this fault situation. 152
J. Electrical Systems 10-2 (2014): 149-155
So, it is possible that the interrupting capability of fuse 1 to 9 and circuit breaker 2 and 3 and remote controllable switches 1 to 3 are not enough to interrupt the fault current. Then, they must be replaced. On the other hand, the protection coordination between relay 1 and relay 2, relay 2 and 3, and relay 3 and fuses may be lost, which result to re-coordinate relays. To reduce these costs, SSFCLs are used. Further cost reduction is achieved using optimal locations of DGs and locations and sizes of FCLs. Studies are performed in 3 cases. In first case, the optimal locations of three DGs and optimal number and places and sizes of SSFCLs are found to minimize the cost of distribution protection system, based on real prices of protection devices and SSFCL devices [13-20]. It is mentioned that no more than one DG can be connected to each bus. Numerical results for this case are shown in table 4. As shown in table 4, when only one FCL is used, cost of protection system is minimized. Also, the best locations for FCL and three DGs are branch 8 (b8) and bus 8, 9 and 10, respectively. Furthermore, the forth column of this table shows the protection cost when no FCL is used. The role of FCL to reduce the protection cost is obvious when the third and forth column of the table are compared. Table 4: Total protection cost of different DGs' locations and different sizes and locations of FCL DGs' locations Bus number 8 , 9 , 10 4 , 9 , 12 4 , 6 , 12 4 , 6 , 10 6 , 11 , 12 4,5,9
Case 1 FCLs' locations [branch , L(mH)] [8 , 3] [8 , 9] [5 , 62] [11 , 12] [5 , 64] [5 , 92] [8 , 8] [6 , 8] [11 , 7] [5 , 36] [7 , 11] [3 , 1]
Cost(US$) 5850.39 6980.30 6985.95 7013.09 8314.96 10104.28
Cost (US$) (without using FCL) 80891 80891 80891 80891 80891 80891
Cases 2 and 3 are studied to investigate the effect of FCL cost on the numerical results. In case 2 cost of FCL is considered to be 50 percent higher than the real estimated FCL cost. Table 5 shows the numerical results achieved. As shown in table 5, that when FCL is located in b8 and three DGs are connected to bus 8, 9 and 10, the cost needed for protection system is minimized. What's more, the cost difference between the cases of using two FCLs and using one FCL increases since the FCL's cost is increased. Further more, it is clear to see the effect of FCL to reduce the protection cost by comparing third and forth column of the table 5. In case 3, the FCL cost is assumed to be 30 percent of the FCLs' prices considered in case 1. Table 6 illustrates the obtained values for optimal DGs' locations and optimal locations and sizes of FCLs. The results show that the best option is to locate DGs at bus 2, 3 and 4, and use three FCLs in branch 3, 2 and 4.
153
S. A. A. Shahriari et al: Optimal Allocation of FCL & DG in the Presence of RCS
Table 5: Total protection cost of different DGs' locations and different sizes and locations of FCL DGs' locations Bus number 8 , 9 , 10 8 , 9 , 10 10 , 11 , 12
Case 2 FCLs' locations Cost(US$) [branch , L(mH)] [8 , 3] 7427.09 [8 , 3] [10 , 9] 11468.36 [11 , 5] [8 , 8] 11470.05
Cost (US$) (without using FCL) 80891 80891 80891
Table 6: Total protection cost of different DGs' locations and different sizes and locations of FCL DGs' locations Bus number 2,3,4 3,4,5 6 , 10 , 12 8 , 9 , 10
Case 3 FCLs' locations [branch , L(mH)] [3 , 7] [2 , 4] [4 , 5] [3 , 4] [4 , 7] [5 , 9] [8 , 9] [6 , 10] [12 , 10] [10 , 9] [8 , 3]
Cost(US$) 3440.43 3441.79 3444.84 3892.07
Cost (US$) (without using FCL) 80192 80891 80891 80891
According to tables 4-6, it is obvious that the cost needed for protection system changes significantly with respect to the DGs' locations and the number, locations and sizes of SSFCLs. So, the allocation and sizing of SSFCLs are necessary. 5. Conclusion Fault current level in a distribution system may be greater than designed capability of fuses and breakers while DGs are connected. In order to adjust the fault current level to desired value, SSFCL is a good choice. The main advantages of SSFCL are fast response and simplicity in structure and control system. This type of FCL has no effect on utility voltage and current in normal system operation. But cost disadvantage of the studied current limiter causes employing GA algorithm to find the optimal DGs' locations and optimal number, locations and sizes of FCLs to be used in the network. Recently distribution systems are equipped with modern protective devices such as remote controllable switches. A case study of such system was used in this paper, and the system simulated in three cases, while DGs and SSFCLs were optimally allocated. The results of simulation demonstrated the GA results for minimization of total protection cost is valid for a distribution system which uses modern protective devices. References [1] [2]
[3] [4]
154
H. Lee Willis and Walter G. Scott, Distributed Power Generation Planning and Evaluation, Marcel Dekker, Inc. 1st edition 2000. F. A. Viawan, D. Karlsson, A. Sannino & J. Daalde, Protection scheme for meshed distribution systems with high penetration of distributed generation, PS '06 Power Systems Conference: Advanced Metering, Protection, Control, Communication, and Distributed Resources, 14-17 March 2006, pp. 99-104. S. M. Brahma & A. A. Girgis, Development of adaptive protection scheme for distribution systems with high penetration of distributed generation, IEEE Transactions on Power Delivery, Vol. 19, No. 1, pp. 56–63, Jan. 2004. S. M. Brahma & A. A. Girgis, Microprocessor - based reclosing to coordinate fuse and recloser in a system with high penetration of distributed generation, IEEE Power Engineering Society Winter Meeting, 2002, Vol. 1, pp. 453–458.
J. Electrical Systems 10-2 (2014): 149-155 [5] [6]
[7] [8] [9] [10] [11] [12] [13]
[14] [15] [16] [17] [18] [19] [20]
K. Tuitemwong & S. Premrudeepreechancharn, Expert System for Protection coordination of Distriution System with Distributed Generators, International Journal of Electrical Power and Energy System, 2010; 33; 466-71. S. Chaitusaney & A. Yokoyama, Impact of Protection Coordination on Sizes of Several Distributed Generation Sources, The 7th International Power Engineering Conference, Vol. 2, 29 Nov.-2 Dec 2005, page(s): 669- 674. S. Chaitusaney & A. Yokoyama, An Appropriate Distributed Generation Sizing Considering Recloser – Fuse Coordination, IEEE PES Transmission Distribution conference Exhibition, Asia Pacific, page(s): 1-6, 2005. G. Tang & M. R. Iravani, Application of a fault current limiter to minimize distributed generation impact on coordinated relay protection, International Conference on Power Systems Transients in Montreal, Canada on Jun. 19-23, 2005. T. Sato, et al., Study on the effect of fault current limiter in power system with dispersed generators, IEEE Transactions on Applied Superconductivity, Vol. 17, No. 2, pp. 2331 – 2334, Jun. 2007. S. A. A. Shahriari, A. Yazdian Varjani & M. R. Haghifam, Cost reduction of distribution network protection in presence of distributed generation using optimized fault current limiter allocation, International Journal of Electrical Power and Energy Systems, 43:1453-9. 2012. A. Girgis & S. Brahma, Effect of distributed generation on protective device coordination in distribution system, Large Engineering Systems Conference on Power Engineering, 11-13 July 2001, pp. 115 – 119. T. Ueda, Solid-state current limiter for power distribution system, IEEE Transaction on power Delivery, October 1993, Vol.8, no.4, pp.1796-1801. M. Raoofat, Simultaneous allocation of DGs and remote controllable switches in distribution networks considering multilevel load. International Journal of Electrical Power and Energy Systems 33, page(s): 1429–1436. 2011. Yueqing City Yiyuan Power Electronics Company, (www.fivefine.com). Yueqing Aoda Electric & Electronic Factory, (http://www.made-in-china.com/showroom /topist). Chinazenli Rectifier, (http://www.chinazenli.com). ANDELI Group Company, (http://www. andeligroup.com) Quality Power Electrical Equipments Pvt. Ltd, (www.qualitypower.co.in). Pars Electric Device Company, (http://www.parsete.com). Payesh Niroo, (www.profuse.ir).
155
