Voltage Control for Unbalanced Low Voltage Grids Using a Decoupled-Phase On-Load Tap-Changer Transformer and Photovoltaic Inverters

Transcription

1 Downloded from orit.dtu.dk on: Ot 06, 2018 Voltge Control for Unlned Low Voltge Grids Using Deoupled-Phse On-Lod Tp-Chnger Trnsformer nd Photovolti Inverters Zehino, Antonio; Mrinelli, Mtti; Hu, Junjie; Coppo, Mssimilino ; Turri, Roerto Pulished in: Proeedings of th Interntionl Universities Power Engineering Conferene (UPEC) Link to rtile, DOI: /UPEC Pulition dte: 2015 Doument Version Peer reviewed version Link k to DTU Orit Cittion (APA): Zehino, A., Mrinelli, M., Hu, J., Coppo, M., & Turri, R. (2015). Voltge Control for Unlned Low Voltge Grids Using Deoupled-Phse On-Lod Tp-Chnger Trnsformer nd Photovolti Inverters. In Proeedings of th Interntionl Universities Power Engineering Conferene (UPEC) IEEE. DOI: /UPEC Generl rights Copyright nd morl rights for the pulitions mde essile in the puli portl re retined y the uthors nd/or other opyright owners nd it is ondition of essing pulitions tht users reognise nd ide y the legl requirements ssoited with these rights. Users my downlod nd print one opy of ny pulition from the puli portl for the purpose of privte study or reserh. You my not further distriute the mteril or use it for ny profit-mking tivity or ommeril gin You my freely distriute the URL identifying the pulition in the puli portl If you elieve tht this doument rehes opyright plese ontt us providing detils, nd we will remove ess to the work immeditely nd investigte your lim.

2 Voltge Control for Unlned Low Voltge Grids Using Deoupled-Phse On-Lod Tp-Chnger Trnsformer nd Photovolti Inverters Antonio Zehino, Mtti Mrinelli, Junjie Hu Deprtment of Eletril Engineering (CEE) DTU Tehnil University of Denmrk Contt: Astrt- This pper presents modeling nd nlysis of the potentil enefits of joint tions of MV/LV three-phse power distriution trnsformer with independent on-lod tp-hnger ontrol on eh phse nd photovolti inverters provided with retive power ontrol pility, in terms of ommodting more renewle genertions in the LV grid. The potentil enefits re investigted in terms of voltge unlne redution nd lol voltge regultion. 24-hours root-men-squre dynmis simultion studies hve een rried out with timestep of 1 seond using 10-mins resolution onsumption nd prodution profiles. A totlly pssive rel Dnish low voltge distriution network is used for the grid topology s well s for the hrteriztion of lods profiles, while the prodution ones re empirilly defined under ssumptions in senrios with different level of photovolti penetrtion nd grde of unlne. Index Terms- On-Lod Tp-Chnger, Unlned Low Voltge Grid, Photovolti Hosting Cpity, Photovolti Inverter, Power Trnsformer, Retive Power Control, Voltge Control I. INTRODUCTION Trditionlly, Dnish low voltge grids re hrterized y unlned lyout of the household eletril instlltion. The voltge unlne effet is even higher due to the users single-phse onnetions of smll generting units whih nowdys re prtilly repling onventionl genertors, with onsequent redution of the regultion pility in the overll eletril system. In the speifi se of the inresing penetrtion of flututing photovolti genertion, espeilly if unlned, distriution system opertors (DSO) re further hllenged y possile lol overvoltges unevenly distriuted on the three phses with the risk of violting the llowed voltge nd [1], [2]. Therefore, with the im to void expensive expnsion investments y network opertors, severl studies hve een performed proposing solutions to redue the mentioned overvoltge phenomen. The proposed strtegies in this regrd inlude voltge ontrol using retive power provision from PV inverters [2], [3], tive power derting of the PV prodution in se of overvoltge onditions [4] nd voltge ontrol t the LV side of the MV/LV trnsformer y on-lod tp hngers () [5]. Mssimilino Coppo, Roerto Turri Deprtment of Industril Engineering University of Pdov, Itly Speifilly, in [6] [9] the voltge ontrol method fouses on the oordintion of opertion nd retive power exhnge etween the DSO nd the PV inverters. All the reserh tivities disussed ove onerning the pplitions fous on the use of trditionl synhronized tp-hnger tions on the three phses, sine they do not onsider the different voltge onditions in low voltge networks due to the single-phse onnetion of DGs. This kind of onnetion, typil of PV inverters, ould worsen the power flow unlne lredy existing in LV distriution networks. In [10] n nlysis of the effetiveness of distriution trnsformer with deoupled phse on-lod tp-hnge pility is performed, onsidering the sme pssive low voltge network used in this study. The novelty of this work is the investigtion of the potentil enefits of omining the tions of distriution trnsformers provided with single-phse-independent tpping pility nd of utomti retive power mngement systems y the PV inverters in rel Dnish low voltge grid. Anlyses re performed in senrios with different PV penetrtion levels, ompring situtions with or without tp tions or retive power regultion y PVs. II. METHODOLOGY All the elements of the rel referene low voltge network, the nd the PV plnts s well hve een modeled in DIgSILENT PowerFtory softwre environment, whih llowed 24-hours RMS dynmis simultion studies with time-step of 1 seond using 10-mins resolution onsumption nd prodution profiles. A. Dnish LV grid The rel network onsidered for the nlysis is DONG Energy network loted in Bistrup, villge round 20 km out of Copenhgen. It is 6- us totlly pssive low voltge feeder onneted to the MV network through 10/0.4 kv trnsformer, s shown in Fig. 1. The short iruit power of the min network is 20 MVA [11].

3 MV usr LV usr Bus1 Bus2 Bus3 Bus4 Bus5 Bus6 Controlled Bus Rel mesurements on voltges, urrents nd powers on the three phses of the usrs hve een onsidered in order to hrterize the 24-hour-long onsumption profiles in terms of tive nd retive power of the 33 single-phse lods (Fig. 2). The totl energy sorption resulted to e out 775 kwh, with men power of 32.3 kw. The dily energy losses mounted to 8.9 kwh, i.e. the 1.14% of the totl energy sored from the MV grid. The totl lod energy nd men power mounts for eh phse re reported in Tle I. TABLE I TOTAL LOAD ENERGY ABSORBED IN THE 24 HOURS AND MEAN POWER Phse Phse Phse Energy [kwh] Men Power [kw] Fig. 1. Single line digrm representtion of the rel Dnish LV network. Sine the rel grid is totlly pssive, the PV units re mnully dded ording to speifi ssumptions. B. modeling In order to perform independent single-phse hnges of the trnsformtion rtios, the three-phse Delt-Wye trnsformer hs een modeled through three single-phse units independently ontrolled, whose seondry sides re onneted etween n erthed neutrl point nd different phse of the LV grid under exm. Eh single-phse trnsformer hs een set with rted power P n of 210 kva, short-iruit voltge relted to the positive sequene impedne V % equl to 4% of the primry side nominl one, rtio of positive sequene impedne nd resistne X/R of 10 nd oth the off-lod urrent i 0% nd offlod power P 0 set to zero. The deision of negleting the inner iron losses is justified y the ft tht the results Fig. 2. Ative nd retive power sorption on the three phses. nlysis do not present ny influene in terms of omprisons of different senrios, sine ll of them re hrterized y the sme mount of off-lod inner losses. For exmple, onsidering typil inditive vlues of i 0% nd P 0 respetively of 1.4% nd 0.42 kw for eh single-phse trnsformer, the totl dily mount of off-lod would hve een 30 kwh. To model the three-phse trnsformer tions, eh single-phse tp-hnger is ssumed to e ssoited to the sme ontrol sheme. Speifilly, in this study the rtio vrition is otined through the proportionl ontroller shown in Fig. 3: the tp position is hnged ontinuously ording to the instntneous mesurement of the phse voltge t the ontrolled us. The urve hs een defined setting the output voltge vrition rnge to ±5% of the rted vlue V n nd the voltge dependeny sensitivity to p.u. it represents the preision of the voltge mesurements. Fig. 3. Proportionl Tpping Logi Lw. In order to otin the ove-sid output voltge vrition rnge, it hs een set tht eh tp unit orresponds to n output voltge vrition of 2.5% of V n, sine ±2 hs een onsidered s limit tp positions. C. Lods nd PV modeling Lod tive nd retive power profiles re hrterized y rel mesurement dt with 10 minutes resolution during 24-hours intervl. With the sme time disretiztion, the PV

4 prodution profiles re defined, ording to the different senrios onsidered. Lods need to e represented s onstnt impedne units, while PVs s onstnt power ones. When running RMS simultions in DIgSILENT PowerFtory softwre environment, the lod model refers to onstnt impedne unit. Therefore, in order to reprodue the rel ehvior of PVs, it hs een neessry to refer to the ZIP Theory [12], y introduing sort of orretion to the PV tive power input vlues. Eqution (1) shows the proedure, where P ref is the tive power red from the text file nd P mod is the one whih will effetively e pplied to the PV unit model: The implemented retive power ontrol pility (RPC pility) from PVs is depited in Fig. 5. Sine in the simultion tool PVs re implemented with the sme model s the lod units, they thus need to e onsidered s tive lod. Positive vlues of retive power orrespond to n indutive nture, whih mens tht retive power gets sored y the inverter; on the other hnd if negtive it ehves like pitor nd Q is injeted into the grid. P mod = P ref (1/V) 2 (1) Of ourse the sme onsidertions re vlid lso with referene to the retive power. A referene 24-hours tive power prodution profile of 1 kwp PV plnt is used to llow the definition of ll the input vlues for eh generting unit in eh senrio onsidered (Fig. 4). Fig. 5. Retive Power Control pility for PV inverter. Fig. 4. Typil dily PV prodution of 1 kwp PV plnt in ler-sky ondition with pnels pointing South nd 30 of inlintion. Sine nowdys the Dnish grid ode does not provide for ny tehnil guidelines out the retive power mngement y the smll distriuted genertion plnts onneted to the LV network, n ritrry lgorithm whih ould e effiient nd prtilly onform to the Europen guideline is implemented. Considering tht Denmrk, Itly nd Germny elong to the sme synhronous region, the funtion of the ontroller is defined ording to tehnil rules for low voltge tive users stted y the Itlin nd Germn Tehnil Stndrds respetively CEI 0-21 [13] nd VDE-AR-N 4105 [14]. These stndrds set different requirements on the retive power prodution y the PV inverters with rted power ove 3 kw nd define severl vritions depending on the size of the plnt together with speifi DSO-users greements. Strting from these guidelines nd with referene to [3], new regultion funtion is therefore reted, with oth voltge nd tive power dependene (2): Aording to the Europen Stndrd EN [15], voltge limits hve een set to ±10% the nominl voltge V n, i.e. V min =0.9 p.u. nd V mx =1.1 p.u. The green re etween 0.99 V n nd 1.01 V n represents ded nd without ny retive power ontrols regrdless the instntneous produed tive power. The red re identifies the opertion in overvoltge onditions, when the inverter sors retive power up to 0.5 p.u. in order to lower the voltges. Symmetrilly, the inverter injets up to 0.5 p.u. of retive power when opertion onditions re in the lue undervoltge re. III. SIMULATIONS A. Senrios nd opertive ses In order to evlute the ontrollers effetiveness under different operting onditions, severl PV penetrtion senrios nd opertive ses re onsidered nd strightwy introdued. As previously sid, sine the referene rel LV grid is merely pssive, PV prodution profiles need to e empirilly defined: eh senrio is therefore hrterized y ertin instlled pek power nd its distriution mong the three phses in terms of grde of unlne, s shown in Tle II. TABLE II CHARACTERIZATION OF DIFFERENT ANALYZED SCENARIOS Senrio Totl instlled PV power Instlled power distriution [kwp] mong the three phses [%] 1 140, (50, 50) 2 210, (50, 50) 3 245,, (50, 30, 20) 4 280,, (50, 30, 20) Q=f(V,P) (2)

5 Eh senrio is studied performing omprisons etween three distint opertive ses hrterized y s mny omintions of the two ontrollers: 1. Bse Cse ( Bse ): neither the nor the retive power ontrol system y PVs is tivted: this se is used s solute referene for the omprisons; 2. 1-Phse Cse ( ): the tions re operted independently on the three phses, referring eh single-phse tp tion to orresponding singlephse phse-neutrl voltge mesurements. This opertive se does not inlude the retive power regultion y PVs; 3. 1-Phse Cse with Q reg. y PVs ( + Qreg ): in ddition to the independent tp djustment tivities y s, now lso the lol retive power provision ontrol y PV inverters is tivted. B. Results For eh simultion, grphil results re given in terms of phse-neutrl voltge profiles over the 24 hours, with prtiulr ttention to the devitions from the nominl vlue t us 6 (i.e. the ontrolled us). In ddition, numeril results re reported in Tle III, where vlues of the mximum devition of the phse-neutrl voltges t us 6 nd mximum nd men vlues of oth the Voltge Unlne Ftor (VUF, defined in (3) [16]) nd the neutrl ondutor potentil t the sme point of the grid re presented. VUF % = V negtive seq / V positive seq 100 (3) Fig. 6 nd Fig. 7 show the phse-neutrl voltge profiles t us 6 over the 24 hours, ompring the three opertive ses in the senrios #1 nd #2, whih refer respetively to the sitution of PV instlled power of 140 nd 210 kwp, evenly distriuted mong the phses nd, without onsidering ny power injetion in phse. Fousing on the first plot, y ompring the se se to the se, it n e seen tht due to independent phse regultions the three phse-neutrl voltges get loser to 1 p.u.. The dditionl retive power ontrol y PVs llows further derese of the devition from the rted vlue, even though the differenes from the previous opertive se do not pper so relevnt. Similr onsidertions re deduile from the results of senrio #2, whih speifilly present more ppreile redution of the phse-neutrl voltge devition when the lol retive power provision ontrol is inluded, llowing the voltges to sty within the eptle rnge of ±10% of V n. Compring senrios #1 nd #2, it n e onluded tht, due to the joint tions of s nd lol Q provision ontrol, the PV hosting pity of the LV grid for this prtiulr grde of unlne grows from 140 kw to 210 kw. The simultion results of senrios #3 nd #4, whih refer to PV instlled power of 245 nd 280 kwp, re reported respetively in Fig. 8 nd Fig. 9 in terms of phse-neutrl voltges. Both re hrterized y the sme grde of unlne, defined y the following distriution mong the Fig. 6. Phse-neutrl voltges t us 6 for the three different opertive ses in the Senrio #1. Fig. 7. Phse-neutrl voltges t us 6 for the three different opertive ses in the Senrio #2. three phses of the totl instlled power: 50% is onneted to phse, 30% to phse nd 20% to phse. It n e seen tht in senrio #3 the phse-independent tions llow redution of the phse-neutrl voltge devitions ompred to the se se, nd gin the sitution is even etter if the dditionl Q regultion ontrol y PVs is performed. The sme onsidertions re vlid in senrio #4 too, in whih the ddition of the lol Q provision ontrol llows to keep the voltges within the eptle rnge of ±10% of V n, thing tht otherwise would not hve een possile. This spet leds to the onlusion tht, for this Fig. 8. Phse-neutrl voltges t us 6 for the three different opertive ses in the Senrio #3.

6 prtiulr grde of unlne of distriuted genertion, the PV hosting pity hs grown from 245 to 280 kw thnks to the omined tions of the s nd the lol Q ontrol. Fig. 11. Phse-neutrl voltges t us 6 for the three different opertive ses in the Senrio #4 using the new RPC pility. Fig. 9. Phse-neutrl voltges t us 6 for the three different opertive ses in the Senrio #4. As it n e notied in Fig. 8 nd Fig. 9, the phse-neutrl voltges t us 6 re hrterized y mny undesired flututions: the use is relted to the high gin of the retive power regultion lw (Fig. 5), whih mkes the ontrol system unstle. Due to this, it hs een onsidered different Q=f(V,P) urve (Fig. 10), whih, ompred to the previous one, presents the sme ded nd ut different gin, hving now ontinuous growth/derese from the limit voltge vlues of the ded nd to the extreme voltge vlues of the regultion lgorithm. lerly e identified upon the neutrl-ground voltges: even though it inreses euse of the phse independent tions, the dditionl Q provision system y PVs mkes it deresing down to vlues even elow the one relted to the se se, oth in terms of men nd mximum vlues. On the other hnd, onerning the VUF, ritilities re found in terms of oth men nd mximum vlues when the two ontrol lgorithms work simultneously, espeilly when the mounts of totl PV instlled power is higher i.e. in senrios #3 nd #4. The reson is relted to the positive nd negtive sequenes mounts: sine the Q provision is just sed on tive power nd phse-neutrl voltge mesurements, the implemented retive power ontrol lw does not tke into ount ny sequene-relted index. Therefore, sine the VUF is not supposed to e ontrolled y the tp logi system neither y the retive power ontrol, the DSO is supposed to onsider this spet nd look for other regulting tions elsewhere. TABLE III RESULT OVERVIEW FOR ANALYZED SCENARIOS Senrio Opertive Cse Mx V devition t us 6 VUF t us 6 Men Vl. Bse Fig. 10. New Retive Power Control pility for PV inverter. Thnks to this modifition, the undesired flututions re voided: ording to the new regultion lw, the phseneutrl voltges trends t us 6 in the senrio #4 re replotted in Fig. 11. In Tle III n overll result overview is reported: vlues of the mximum devition of the phse-neutrl voltges t us 6 nd mximum nd men vlues of oth the VUF nd the neutrl ondutor potentil t the sme us re presented for eh opertive se in eh senrio. The ove-desried onsidertions out the phseneutrl voltges devitions re herey numerilly explited in the third olumn. Moreover, further positive effets n #1 #2 #3 #4 Mx Vl. Neutrl potentil t us 6 Men Vl. Mx Vl /+10.0% 0.71% 1.72% 1.96% 5.05% -6.5/+5.0% 1.04% 2.49% 2.03% 5.31% +Qreg -4.0/+4.5% 1.08% 2.59% 1.88% 4.54% Bse -15.0/+15.0% 1.03% 2.32% 3.11% 7.84% -11.0/+11.2% 1.45% 3.24% 3.29% 8.32% +Qreg -7.0/+8.4% Bse -4.0/+13.0% 0.56% 1.63% 1.48% 4.08% -2.0/+8.8% 0.79% 2.01% 1.61% 4.42% +Qreg -2.0/+4.5% 1.17% 4.05% 1.41% 3.95% Bse -6.0/+16.2% 0.72% 1.89% 2.08% 5.47% -3.0/+12.0% 0.95% 2.03% 2.23% 5.89% +Qreg.mod -2.0/+7.0% 1.75% 4.14% 2.75% 6.56% 1.41% 4.47% 1.89% 4.86%

7 IV. CONCLUSIONS AND FUTURE WORKS This work presented the simultion studies to develop nd test the fesiility of deoupled three-phse MV/LV trnsformer with the ojetive of improving the distriution network power qulity. Anlysis hve een performed onsidering the simultneous tions of n dditionl ontrol system, sed on the mngement of the retive power provided y the PV plnts on the sis of lol rel time voltge nd tive power mesurements, ording to speifi Q=f(V,P) regultion lw. At first, very unlned PV onnetion senrios hve een onsidered: the joint tions of the two ontrollers enled n inrese of the PV hosting pity of the grid, sine eptle phse-neutrl nd neutrl-ground voltges hve een gurnteed. Afterwrds, for the next senrios, more relisti distriution of the instlled PV power mong the three phses hs een ontemplted: results showed further enhnement of the mximum eptle PV hosting pity. The results showed tht, with signifintly lrge mount of PV instlled, mny undesired flututions my pper, mking the ontrol system unstle. The use is due to the gin of the Q-regultion lw, therefore different Q=f(V,P) lw with the sme ded nd nd lower gin hs een onsidered in order to mke the system stle gin. A onlusion to this is tht the DSO is supposed to tke into ount tht, whether Q-regultion lws with too high gin re issued, flututions ould tke ple if onsiderle PV power is instlled in the LV grid. In this se it should provide regultion lws with smller gin, mking it possile to ept suh high PV penetrtion. In ll the senrios, the Voltge Unlne Ftor hs grown ompred to the opertive ses without the retive power ontrol y PVs, s result of the opposite trends in the positive nd negtive sequenes mgnitudes. So, even though the dditionl Q-ontrol system rings enefits in terms of phse-neutrl voltge devitions nd neutrl potentil, it lso involves undesired inreses of the VUF, sine its ontrol logi does not onsider ny voltge sequenes nlysis. These ritilities need to e onsidered nd tken into ount y the DSO, whih hs to gurntee tht this index lies elow the limits issued y the ntionl nd Europen tehnil stndrds, looking for other regulting tions. The ojetive of future works is to investigte further enhnements of the oordinted tions of the ontrollers nd to perform prtil tests: the deoupled three-phse onlod tp-hnger trnsformer will e experimentlly tested using the SYSLAB-PowerL.DK experimentl fility t the DTU Risø Cmpus in The uthors thnk the tem of PSS Energy A/S nd Dong Energy for the interesting disussions. REFERENCES [1] M. Brun, T. Stetz, R. Bründlinger, C. Myr, K. Ogimoto, H. Htt, H. Koyshi, B. Kroposki, B. Mther, nd M. Coddington, Is the distriution grid redy to ept lrge sle photovolti deployment? Stte of the rt, progress, nd future prospets, Prog. photovoltis Res. Appl., vol. 20, no. 6, pp , [2] R. Cldon, M. Coppo, nd R. Turri, Distriuted voltge ontrol strtegy for LV networks with inverter-interfed genertors, Eletr. Power Syst. Res., vol. 107, no. 0, pp , Fe [3] K. Knezović, M. Mrinelli, R. J. Møller, P. B. Andersen, C. Træholt, nd F. Sossn, Anlysis of Voltge Support y Eletri Vehiles nd Photovolti in Rel Dnish Low Voltge Network, Universities Power Engineering Conferene (UPEC), 2014 Proeedings of the 49th Interntionl, pp.1-6, Cluj Npo, 2-5 Sep [4] A. Kehroud, P. F. Rieiro, nd W. L. Kling, Distriuted genertion support for voltge regultion: An dptive pproh, Eletr. Power Syst. Res., vol. 107, no. 0, pp , Fe [5] A. Einflt, F. Kupzog, H. Brunner, nd A. Lugmier, Control strtegies for smrt Low voltge grids the projet DG DemoNet smrt LV grid, in Integrtion of Renewles into the Distriution Grid, CIRED 2012 Workshop, 2012, no. 0238, pp [6] I. Leisse, O. Smuelsson, nd J. Svensson, Coordinted voltge ontrol in medium nd low voltge distriution networks with wind power nd photovoltis, in 2013 IEEE Grenole Conferene, 2013, pp [7] M. Stifter, B. Bletterie, H. Brunner, D. Burnier, H. Swsn, F. Andren, R. Shwle, A. Art, R. Nenning, nd F. Her, DG DemoNet vlidtion: Voltge ontrol from simultion to field test, in Innovtive Smrt Grid Tehnologies (ISGT Europe), nd IEEE PES Interntionl Conferene nd Exhiition on, 2011, pp [8] M. Stifter, R. Shwle, W. Tremmel, S. Henein, H. Brunner, B. Bletterie, A. Art, F. Her, nd R. Pointner, DG DemoNet: Experienes from volt/vr ontrol field trils nd ontrol lgorithm dvnements, in rd IEEE PES Innovtive Smrt Grid Tehnologies Europe (ISGT Europe), 2012, pp [9] M. Stifter, B. Bletterîe, nd F. Her, DG DemoNet: Impt of VOLT/VAR ontrol on inresing the voltge nd reserve Results from field tril vlidtions, in Eletriity Distriution (CIRED 2013), 22nd Interntionl Conferene nd Exhiition on, 2013, no. 0508, pp [10] M. Coppo, R. Turri, M. Mrinelli, nd X. Hn, Voltge Mngement in Unlned Low Voltge Networks Using Deoupled Phse-Tp-Chnger Trnsformer, Universities Power Engineering Conferene (UPEC), 2014 Proeedings of the 49th Interntionl, pp1-6, Cluj Npo, 2-5 Sep [11] DONG Energy. [Online]. Aville: [12] IEEE Tsk Fore on Lod Representtion for Dynmi Performne, Lod representtion for dynmi performne nlysis, IEEE Trns. Power Syst., vol. 8, no. 2, pp , [13] Itlin Tehnil Stndrd CEI Norm Itlin, [14] Germn Tehnil Stndrd VDE-AR-N FNN - Forum Netztehnik / Netzetrie im VDE, [15] Europen Tehnil Stndrd EN 50160, [16] P. Pilly nd M. Mnyge, Definitions of voltge unlne, IEEE Power Eng. Rev., vol. 22, no. 11, pp , ACKNOWLEDGEMENTS The uthors knowledge the finnil support to this work given y the Dnish EUPD progrm for the projet Energy Sving y Voltge Mngement, under the grnt numer