D-STATCOM Optimal Allocation Based On Investment Decision Theory

Transcription

1 Internatonal Conference on Computer Engneerng, Informaton Scence & Applcaton Technology (ICCIA 2016) D-STATCOM Optmal Allocaton Based On Investment Decson Theory Yongjun Zhang1, a, Yfu Mo1, b and Huazhen Cao2, c 1 School of Electrc Power, South Chna Unversty of Technology, Guangzhou , Chna; 2 Grd Plannng & Research Center, Guangdong Power Grd Corporaton, Guangzhou , Chna. a zhangjun@scut.edu.cn, b @qq.com, ccaohuazhen@gd.csg.cn Abstract. In order to avod certan types of voltage sags, ths paper proposed an optmal method to allocate D-STATCOM capacty, whch s based on Investment Decson Theory and a 10V dstrbuton lne equvalent model. In addton, the Investment Decson Theory was presented n ths paper, n order to establsh a D-STATCOM optmal allocaton model wth the nvestment paybac perod as man ndex. Some optmzaton suggestons were rased at the end of ths paper, consderng the hstorcal data of voltage sags and low voltage release allocaton. Fnally, the effectveness of the method s demonstrated by the smulaton on a 10V dstrbuton feeder n Shenzhen, Guangdong Provnce, Chna. Keywords: D-STATCOM; Investment Decson Theory; Voltage Sags; Optmal Allocaton. 1. Introducton A large area of low voltage trp n the cty caused by low voltage sags, whch leads to sudden loss of the load, s becomng a huge threat to the relablty of power grd operaton. There s no doubt that voltage sags have been a top ssue affects the power qualty [1]. STATCOM has obvous advantages n the cost, complexty of control and functon, s more applcable to the larger capacty user, and has thus become the man means and measures to solve the voltage sag and voltage nstablty [2]. A lot of researches have been made on the smulaton and analyss of the practcal result n usng D-STATCOM to suppress voltage sag. Subject to the prce, t s often necessary to study the method of D-STATCOM optmal confguraton n engneerng applcatons [3, 4]. Researchers abroad use a varety of algorthms to determne the optmal locaton and capacty of STATCOM n the system [5, 6]. Snce the above methods do not fully consder the actual settng value of the users under voltage trppng devce and the actual voltage sag n dstrbuton networ, the Cost estmaton of voltage sags shall have very bg dfference wth the actual condtons and cannot meet the requrements of the project. As the power qualty montor s wdely used, t s more feasble to collect the voltage sag data n the dstrbuton networ, whch allows us to confgurate D-STATCOM better. In order to gve the consderaton to the voltage sag compensatng capablty and economy of the dstrbuton networ, ths paper establshed a D-STATCOM optmal allocaton model wth the nvestment paybac perod as man ndex based on Investment Decson Theory, and demonstrated the effectveness of the method by the smulaton on a 10V dstrbuton feeder n Shenzhen, Guangdong Provnce, Chna V dstrbuton lne equvalent model After choosng a 10V feeder and dealng wth the rest of the grd equvalently, the 10V dstrbuton lne equvalent model s establshed as shown n Fg 1. VSH refers to power supply voltage; VH s voltage ampltude n dstrbutng sde; XSH refers to the system equvalent reactance; PL+jQL and PH+jQH s load apparent power and system apparent power, respectvely. Besdes, all of loads are equpped wth crcut breaer wth the functon of low-voltage trppng. D-STATCOM s connected to 10V bus, usng the rated voltage control method The authors - Publshed by Atlants Press 451

2 Generator Bus D-STATCOM Load 2 Load 4 Load 6 P H +jq H V H V SH X SH 10V Bus Load 1 Load 3 Load 5 Fg. 1 10V dstrbuton lne equvalent model All of the load parameters and the settng value of low-voltage trppng are set accordng to users nformaton. The voltage sags caused by the external fault are smulated by settng dfferent fault n the generator bus. In the most extreme case of one nd of the voltage sags, D-STATCOM allocaton capacty ncreases constantly untl there are no load losses caused by low voltage leaages so that the allocaton capacty wll be determned n order to suppress approprate nd of voltage sags. However, the allocaton capacty determned by above mean may be large, whch s necessary to come up wth an optmal model. 3. D-STATCOM Optmal Allocaton model Based On Investment Decson Theory Reference [6] optmzed D-STATCOM allocaton capacty by mnmzng the sum of the economc loss caused by the voltage sags and the cost of D-STATCOM. The total cost s contnuously reduced n the optmzaton used frefly algorthm. By contrast, Investment Decson Theory lays more stress on the transfer rate of funds and the effect of nvestment. Accordng to ths theory, ths paper establsh a D-STATCOM optmal allocaton model wth the nvestment paybac perod as man ndex. The ntal nvestment of D-STATCOM s defned as follows: C0 QC (1) Where the unt prce of D-STATCOM s δ RMB /Mvar (ncludng equpment cost and nstallaton cost); QC represents the capacty of D-STATCOM. The annual mantenance cost of D-STATCOM s defned as follows: Cm C0 QC (2) Normally, the annual mantenance cost equals the ntal nvestment multpled tmes a coeffcent α, whch s 0~10%. The outage loss that can be reduced every year after allocatng D-STATCOM s defned as follows: Cr PL fl (3) Where t represents the voltage sags can be ruduce f tmes per year after nstallng D-STATCOM. represents dfferent types of voltage sags, and the correspondng amount of load sheddng s PL MW. The economc loss of load sheddng s l RMB /MW. Assume the economc loss of nonconformng products caused by voltage sags can be reduced β RMB, and the servce lfe of D-STATCOM s T years. The nvestment s worthy when the reduced loss s bgger than the ntal nvestment and mantance cost of D-STATCOM durng ts servce lfe. T Cr C0 TCm (4) Plus nto relevant parameters and the followng equatons can be wrtten, where Pt s the nvestment paybac perod. QC T Pt (5) P f l Q L C 452

3 4. Survey on voltage sags and low voltage leaage n Shenzhen An annual statstcs of voltage sags n a 10V dstrbuton lne of Shenzheng s shown n Fg. 2. Fg. 2 Annual statstcs of voltage sags The parameters of voltage sags and low voltage leaages are descrped as shown n Table 1 and Table 2. Table 1 Duraton and ampltude of voltage sags Duraton(ms) Tmes Percentage(%) Ampltude( p.u.) Tmes Percentage(%) < ~ ~ ~ ~ ~ ~ ~ > < Table 2 Delay and operaton value of low voltage leaages Operaton value(p.u.) < ~ ~ ~ ~1 Sum Delay(ms) < ~ ~ ~ > Sum Smulaton Based on the 10V dstrbuton lne equvalent model, t s smulated usng the BPA software, where Load 1~6 are 0.3MW, 0.8 MW, 0.1 MW, 0.8 MW, 0.2 MW, 0.8MW, respectvely. And the load power factors are smlarly set up to 0.9. Accordng to the survey above, the parameters are set as shown n Table 3. Table 3 Parameters of the loads and low voltage leaages Load number Load capacty(mw) Capacty proporton Operaton value(p.u.) Delay(cycle) % ~ % ~ % By settng dfferent short-crcut faults n the generator bus, varous nds of voltage sags can be smulated as shown n Table 4. Table 4 Parameters of the voltage sags Number Type of voltage sags Probablty Ampltude n extreme case(p.u.) Duraton(cycle) Fault 1 Ampltude 0.7p.u. 79% Fault 2 Ampltude 0.6p.u. 87% Fault 3 Ampltude 0.3p.u. 92%

4 In order to determne the allocaton capacty needed to suppress approprate nd of voltage sags, the capacty of D-STATCOM ncreases slowly from 0 n the process of smulaton, untl there are no protecton actons. Accordng to the exstng survey, δ s 500,000RMB /Mvar; α s 0.1. Besdes, t wll tae ths factory 4h to restart ther producton lnes after an electrcty falure accdent, and the loss s 136,000RMB /MW as t s shown n [7], where l s 136,000RMB /MW. The loss of nonconformng product caused by voltage sags s gnored n the calculaton process, where β s 0. Then put them nto (5) and the calculaton results are shown as shown n Table 5. Table 5 Results of calculaton and smulaton Number Needed Investment paybac Load sheddng stuaton capacty(mvar) perod(year) Fault MW for 3 tmes 4.0 Fault MW for 2 tmes, 0.3MW for 7 tmes 3.7 Fault MW for 2 tmes, 1.2MW for 4 tmes, 0.3MW for 7 tmes 4.9 Assume the operaton value of Load 1 s 0.6p.u., calculate and smulate agan. We can get the results as shown n Table 6. Table 6 Results after optmzng the parameters of low voltage leaages Number Needed Investment paybac Load sheddng stuaton capacty(mvar) perod(year) Fault Fault MW for 2 tmes 0.5 Fault MW for 2 tmes, 1.2MW for 4 tmes Summary From the results of the smulaton and economc analyss, we cannot expect that D-STATCOM can suppress all nds of voltage sags, whch s unreasonable n technology and economy. Accordng to the parameters of low voltage leaages, t s more reasonable to avod power falure by optmzng the parameters when users demand low power qualty. D-STATCOM of a certan capacty s necessary to suppress most of voltage sags when users demand hgh power qualty. References [1]. Chapman D, Introducton to power qualty, n Power Qualty Applcaton Gude Copper Development Assocaton, May 2007 [Onlne].Avalable: [2]. Zhang, Y. J., Chen, C., L, Y., & Wu, G. B. (2010). Dynamc voltage support plannng for recevng end power systems based on evaluaton of state separatng and transferrng rss. Electrc Power Systems Research, 80 (12), [3]. Xubn Lu, Canbng L and Ka Sun, "Voltage support for ndustral dstrbuton networ by usng postve/negatve sequence passvty-based control", Power Electroncs and Moton Control Conference (IPEMC-ECCE Asa) 2016 IEEE 8th Internatonal, pp , [4]. X. Hongxa and Z. Yanyun, "Cooperatve Control and Applcaton of D-STATCOM and Swtched Capactor Branch," Intellgent Computaton Technology and Automaton (ICICTA), 2011 Internatonal Conference on, Shenzhen, Guangdong, 2011, pp [5]. Mlanovc J V, Zhang Yan, "Global Mnmzaton of Fnancal Losses Due to Voltage Sags Wth FACTS Based Devces," Power Delvery, IEEE Transactons on, vol.25, no.1, pp.298,306, Jan [6]. Farhoodnea M, Mohamed A., Shareef H, Zayandehrood H., "Optmum D-STATCOM placement usng frefly algorthm for power qualty enhancement," Power Engneerng and 454

5 Optmzaton Conference (PEOCO), 2013 IEEE 7th Internatonal, vol., no., pp.98, 102, 3-4 June 2013 [7]. Tollefson G, Bllnton R, Wacer G, Chan E, Aweya J, "A Canadan customer survey to assess power system relablty worth," Power Systems, IEEE Transactons on, vol.9, no.1, pp.443, 450, Feb