MULTI-OBJECTIVE OPTIMIZATION OF A BATTERY ENERGY MANAGEMENT FOR AN OFF-GRID SMART HOUSE. University of the Ryukyus, Okinawa, Japan

Transcription

1 Proceedings of BS: h Conference of Inernaional Building Performance Simulaion Associaion, Hyderabad, India, Dec. 79,. MULTIOBJECTIVE OPTIMIZATION OF A BATTERY ENERGY MANAGEMENT FOR AN OFFGRID SMART HOUSE Tsubasa Shimoji, Hayao hahara, Harun Or Rashid Howlader, Sharma ADITYA, Hidehio Maayoshi, Asushi Yona and Tomonobu Senjyu Deparmen of Elecrical and Elecronics Engineering, Universiy of he Ryukyus, Okinawa, Japan ABSTRACT Recenly in japan, an offgrid smar house is araced for effecive energy uilizaion and measure of a blackou a he ime of an acciden. In he offgrid smar house, i is effecive ha a phoovolaic (PV) generaor, a solar collecor (SC), which use as a renewable energy, a hea pump (HP) which is energy saving ho waer supply equipmen, and a fixed baery for saving elecriciy are inroduced. Also, if an elecric vehicle () is inroduced ino his smar house, domesic power consumpion can be suppored by using a baery of he. However, here is necessary o deermine opimal capaciy of he fixed baery and opimal number of PV panels, when he offgrid smar house is designed. So, his paper proposes an offgrid smar house, uilized by he PV generaor, SC, HP, fixed baery, and, which is independen from an elecric power sysem. The HP, he fixed baery, and are used as conrollable loads. This paper plans o disclose he capaciy of fixed baery and number of PV panels which are se, for geing opimal operaional mehod of conrollable loads, and showing operaion of he offgrid smar house wihou shorage of elecriciy. INTRODUCTION In recen years, he demand reducion from fossil fuels is recommended o produce energy. Increase use of fossil fuels causes depleion of fossil fuels as well as environmenal problems like global warming. Therefore, in japan, i is considered abou life which does no use supplied elecriciy from an elecriciy company. So, a smar house consiss of a phoovolaic (PV) generaor and a solar collecor (SC), which are use renewable energy, a Hea pump which is energy saving ho waer supply equipmen, and a fixed baery which saves elecriciy. Also, if an elecric vehicle () is inroduced ino his smar house, domesic power consumpion can be suppored by using a baery of he. Furhermore, developmen of charging infrasrucure in Japan is progressing hrough quick charging sysems ha charge o he baery of he by quickly. So ha, elecriciy ino he smar house can be supplied by discharging from he baery of he which is charged oudoor adequaely, and inroducion coss are reduced, because of he capaciy of a fixed baery and number of PV panels which are inroduced ino he smar house are reduced. Therefore, a self DC DC Load P L Ba P P DC DC DC DC DC DC PV Baery DC AC Oneuni house SC DC bus Hea Pump Elecric Hea Figure Offgrid Smar house model Sorage Tank sufficien life of elecriciy wihou elecriciy supply from he elecriciy company, by living he smar house which have he PV sysem, SC, HP, fixed baery, and. So ha, i is expeced ha energy uilizaion became effecive and measure of a blackou a a ime of he acciden. So, his paper proposes an offgrid smar house consiss of a PV generaor, a SC, a HP, a fixed baery, and, which is independen from an elecric power sysem. The HP, fixed baery, and are used as conrollable loads. The capaciy of he fixed baery and number of PV panels which are se, his paper plans opimal operaional mehod of conrollable loads by using muliobjecive opimizaion mehod. This paper plans o display he capaciy of fixed baery and number of PV panels which are se, for geing opimal operaion of conrollable loads by using a muliobjecive opimizaion mehod. Moreover, by simulaing he operaional condiions which can fullfill he elecriciy demand by selfsufficien supply of he elecriciy oo, his paper consider abou proposed offgrid smar house. ELECTRIC POWER SYSTEM Offgrid smar house model The offgrid smar house model assumed in his paper is shown in Fig.. The PV generaor, SC, HP, fixed baery, and sysems are inroduced ino his smar house. The HP, fixed baery, and are used as conrollable loads. In his paper, use of he howaer 7

2 Load P L [kw] Proceedings of BS: h Conference of Inernaional Building Performance Simulaion Associaion, Hyderabad, India, Dec. 79,. Solar Hea Collecor I a Eq.() A l Eq.() Q l Load Eq.(7).. kw/day. T w Eq.() Q a Q sw cρ Aw s αh s Ho Waer Tank T h + T Figure Power consumpion Q e Auxiliary Hea Source Eq.(8) Figure Model of solar collecor supply is assumed for he smar house a boh morning and evening. The ho waer emperaure of he sorage ank is se for a arge emperaure of C a 7 p.m. When he waer emperaure is less han he arge emperaure, he HP sars funcioning o boil supply waer and raises he sorage waer o he arge emperaure. The HP waer heaer assumes a sorage ank capaciy of 7L, a raed heaing capabiliy of kw/kw, and a COP of.. Furhermore, he is assumed for ouside use as a passenger vehicle from 8 a.m. o p.m. The capaciy of is se o.kw/kwh. Here, he capaciy of he fixed baery is deermined by simulaion. Phoovolaic sysem In his paper, he parameers of he PV are as follows: The conversion efficiency η PV is.%, n PV [panels] is he number of panel, and panel area S PV is. m. Moreover, PV oupu obained from amoun of insolaion I a [kw/m ] is calculaed from he following equaion. = η PV n PV S PV I a (.(T CR )) () Here, T CR is cell emperaure [ C]. Solar collecor sysem In his paper, he parameers of he SC are as follows: The conversion efficiency η SC is %, he number of panels n SC is shees, and panel area S SC is. m. The solar collecor sysem can be modeled by equaions () (8). Fig. shows he solar collecor sysem mahemaical model. The emperaure change characerisics and ime change characerisics of he sorage waer can be obained by he following equaions. cρa w dt h d = Q h () dq h = α h (T h T a ) () d Here, T h is he emperaure of he waer in he sorage ank [ C], A w is he capaciy of sorage ank [L], Q h is he hea capaciy of he waer in he ank [cal], c is he specific hea of he waer [cal/(g )] (=. cal/(g )), ρ is he densiy of he waer [g/l] (= g/l), and α h is he coefficien of hea ransfer, T a is he ambien air emperaure [ C]. The quaniy of hea colleced in hea collecion panel Q a [J] is expressed by he following equaion: Q a = η h I a na c () Here, η h is he efficiency of conversion o hea, I a is he solar radiaion [J], n is he number of panels, and A c is he hea collecion area per panel [m ]. Hea los o he howaer supply Q l [cal], hea added by he waer supply Q sw [cal], ho waer used from he ank a supply ime A l [L], quaniy of waer supplied o he ank A sw, and hea added from an auxiliary hea source Q e [cal] are found using following equaions: Q l = cρa l T h () Q sw = cρa sw T w () A l = A sw = T l T w A l (7) T h T w Q e = cρa w (T e T h ) (8) Here, T l is he emperaure of he howaer supply [ C], T w is ciy waer emperaure [ C], A l is he quaniy of ho waer a he ime of use of he howaer supply [L], and T e is goal hea emperaure [ C]. OPTIMIZATION METHOD Objecive funcion P L, P PV, P Ba, P, and P HP in Fig. are respecively power consumpion excluding conrollable loads, PV oupu, discharge and charge power of he fixed baery, discharge and charge power of he, and power of he HP in he smar house a a given ime. Equaion (9) expresses he load dispaching balance of he offgrid smar house in Fig.. P L = P PV + P Ba + P (9) Objecive funcions in his paper, minimize he numberofpvpanelsn PV [panels] and supplied elecric power o he offgrid smar house by he baery of he which is charged oudoor P D [kwh], and are expressed in equaions () and (), redpecively. F = min n PV T F = min = P D 7

3 Proceedings of BS: h Conference of Inernaional Building Performance Simulaion Associaion, Hyderabad, India, Dec. 79,. Soluion A Soluion B Baery [kwh] Baery [kwh] Baery [kwh] PV oupu. kwh/day. kwh/day 9. kwh/day. kwh/day. kwh/day Supplemened energy Σ Τ = P D [kwh] Soluion C PV panel n PV [panels] Figure Preo opimal soluions Table Evaluaion resul of he soluions Soluion T = P D n PV Baery A 9.8 [kwh] [modules] [kwh] B 7. [kwh] [modules] [kwh] C [kwh] [modules] [kwh] In his paper, wo opimizaion mehods are used. By using he abusearch which is a kind of singleobjecive opimizaion mehod and he objecive which minimze he shorage of elecriciy for he power consumpion, opimal operaional mehod of conrollable loads is planed. Nex, by using he nondominaed geneic algorihm (NSGA) which is a kind of muliobjecive opimizaion mehod, a se of opimal pareo soluions of he wo objecive funcion, which are radeoff relaion each oher, is searched. By he se of opimal pareo soluions, he opimal number of he PV panels and supplied elecric power of he for operaing he offgrid smar house. Consrains Operaion consrains of equipmen in he smar house are shown in equaions () (7).. C max < P Ba <Pmax Ba () P <Pmax (). Cmax Ba <CBa <.8 Cmax Ba (). Cmax <C s= T P S = P S = P L <C max () C (=7+(s )) <C max () P D = () P + () Temperure T h [ o C] 8 Power comsumpion HP P P S P + P D Sae of charge (a) PV oupu power (b) Waer emperaure of sorage ank kwh/day. kwh/day.7 kwh/day kwh/day kwh/day (c) Power consumpion of HP kwh/day.7kwh/day.kwh/day 9.9kWh/day kwh/day.kwh/day P S = P L + P P D =.7 [kwh] (d) Shorage of elecriciy P D =.[kwh] P D = 9.9 [kwh] P D = [kwh] P D =. [kw] (e) of (f) of fixed baery (g) Sae of charge for fixed baery and Figure Simulaion resul of soluion A baery 7

4 Proceedings of BS: h Conference of Inernaional Building Performance Simulaion Associaion, Hyderabad, India, Dec. 79,. Where, P Ba is he acive power of he fixed baery [kw], P is he acive power of he [kw], Pmax Ba is he maximum allowable value of discharge and charge power for he fixed baery [kw], Pmax is he maximum allowable value of discharge and charge power for he (kw), C Ba is he sae of charge of he fixed baery [kwh], C is he sae of charge for he [kwh], Cmax Ba is he maximum allowable sae of charge for he fixed baery, and Cmax is he maximum allowable sae of charge for he (kwh), s is he simulaion day [day]. Equaions () and () show he inverer consrains of he fixed baery and, respecively. Equaions () and () show he sae of charge consrains of he fixed baery and, respecively. Equaion () shows he sae of charge consrains for he a 7 a.m. every day. Equaion (7) is consrain no o occur shorage of supplied elecriciy for saisfying he power consumpion in he offgrid smar house. SIMULATION RESULT Simulaion condiions In his simulaion, operaion erm is days which weaher condiions are sunny, rainy, cloudy, sunny and sunny in order. Fig. shows power consumpion excluding conrollable loads. For volumes of ho waer supply used in he smar house, L was used as a shower during he hours, from 7 p.m. o p.m. If waer emperaures of sorage ank dropped lower han C a 7 p.m., he waer was heaed by he HP. The baery of he is assumed used oudoor as passenger car a 7Wh from 8 a.m. o p.m. If shorage of supplied elecriciy occur, he baery of he is charged by quickly from 8 a.m. o p.m. of previous day. Simulaion resuls Simulaion resul of NSGA is shown in Fig.. This figure shows ha if inroducion capaciy of PV panels and he fixed baery is small, supplied elecriciy from he baery of he is large. Operaional mehods of conrollable loads in hree cases which are shown in Fig. are shown in Fig.,, and 7. Also, resuls of soluions A, B and C obained from Fig. are summarized in Table.. Figs. (a) of hese figures show PV oupu. Figs. (b) and (c) of hese figures show waer emperaures in he sorage ank and power of HP, respecively. On he day of cloudy and rainy, i can be observed ha he emperaures of sorage ank fulfill emperaure goals by HP and SC heaing due o waer emperaures in he sorage ank dropping lower han he goal emperaure a 7 p.m. Figs. (d) of hese figures show he shorage of elecriciy when supplied elecriciy from he is nohing. Figs. (e) of hese figures show charged and discharged power of he. These Figs. (e) show ha if shorage of elecriciy of he nex day is prediced, he is charged oudoor from p.m. o p.m. In Figs. (f) of hese figures, PV oupu Temperure T h [ o C]. kwh/day.9 kwh/day.9 kwh/day 8 Power comsumpion P S P + P D Sae of charge... kw/day. kw/day (a) PV oupu power (b) Waer emperaure of sorage ank kwh/day. kwh/day.7 kwh/day kwh/day kwh/day (c) Power consumpion of HP kwh/day kwh/day.kwh/day.9kwh/day kwh/day kwh/day P S = P L + P (d) Shorage of elecriciy 9 P D = P D = [kwh]. [kwh] P D =.9 [kwh] P D = [kwh] P D = [kw] (e) of (f) of fixed baery baery (g) Sae of charge for fixed baery and Figure Simulaion resul of soluion B 77

5 Proceedings of BS: h Conference of Inernaional Building Performance Simulaion Associaion, Hyderabad, India, Dec. 79,. PV oupu Temperure T h [ o C] 9. kwh/day.88 kwh/day 8. kwh/day 8 Power comsumpion P I [kw] P + P D [kw] Sae of charge.. 9. kw/day. kw/day (a) PV oupu power (b) Waer emperaure of sorage ank kwh/day. kwh/day.7 kwh/day kwh/day kwh/day (c) Power consumpion of HP kwh/day kwh/day kwh/day kwh/day kwh/day kwh/day P S = P L + P (d) Shorage of elecriciy P D = [kwh] P D = [kwh] P D = [kwh] P D = [kwh] P D = [kwh] (e) of (f) of fixed baery baery (g) Sae of charge for fixed baery and Figure 7 Simulaion resul of soluion C charged and discharged power of he fixed baery are shown and i is undersood ha he inverer consrains are saisfied. Figs.,, and 7 show ha a consumer which can inroduce he equipmen wih large capaciy does no have o charge he oudoor for operaing he offgrid smar house. However, a consumer which canno inroduce he equipmen wih large capaciy have o charge he oudoor for operaing he offgrid smar house. By simulaion resuls, he offgrid smar house on he day which amoun of PV oupu is low can be operaed by suplied elecric power of he. CONCLUSION In his paper, an offgrid smar house includes a PV generaor, a SC, a HP, a fixed baery, and, which is independen from an elecric power sysem is proposed. The capaciy of he fixed baery and number of PV panels, opimal equipmen capaciy of he offgrid smar house is considered oo. The simulaion resuls show he offgrid smar house can operae wiho shoage of elecriciy. REFERENCES Akihiro Yoza, Kosuke Uchida, A. Y. and Senjyu, T.. Opimal operaion mehod of smar house by conrollable loads based on smar grid opology. Inernaional Journal of Emerging Elecric Power Sysems, :. Sakai, S.. Problems and prospecs for a smar house. Seijyo universiy research. 78