R.M.G.D. Ranathunga and Lalith A. Samaliarachchi

Transcription

1 - Vl. L, N. 4, pp. [1-11], 217 Vl. L, N. 4, pp. [page range], 217 The Institutin f Engineers, Sri Lanka The Institutin f Engineers, Sri Lanka Impact f Electric Vehicle Lads n the System Lad Prfile f Sri Lanka R.M.G.D. Ranathunga and Lalith A. Samaliarachchi Abstract: Electric vehicles (EVs) are being prmted wrldwide because f the ptential they have t address atmspheric pllutin issues and relieve cuntries frm the burdens assciated with the use f liquid petrleum based transprt fuels. Ever increasing envirnmental cncerns, imprvements t battery technlgies, entry f new manufacturers and new vehicle mdels, and intrductin f favurable fiscal plicies have all cntributed t the increase in EV penetratin rates wrldwide including Sri Lanka. Hwever, it is nt clear hw the pwer system in Sri Lanka wuld face the challenge f this new and unknwn demand which wuld get added t the already exiting demand prfile and which wuld be a result f the stchastic nature f the battery charging behaviur f EVs. In this research study, EV charging mdes and charger types available in the market were cnsidered and a mdel was established t ascertain the relatinship between the charging demand and the ther cntributry factrs such as EV battery sizes, charge remaining at the cmmencement f recharging, charging rates at different places f the chrnlgical lad prfile, charging habits f custmers and time-f-use pricing (TOU) plicies. The prbability distributin f variables such as the time f cmmencement f charging and battery charge duratin was cnsidered at a significantly acceptable level. By cmbining the prbability distributin curves f the said variables, several EV charge demand curves were established using Mnte Carl methd and the charging demand curves were subsequently superimpsed n the system lad prfile. While the prpsed methdlgy gives an insight int the impact f the EV lad n the system lad prfile, it als shws hw an effective cntrl f EV charging culd bring dwn the peratinal csts and investment cst f a pwer system. Keywrds: 1. Intrductin Electric vehicle, charging demand, lad prfile, Mnte Carl simulatin 1.1 Backgrund Because f the emphasis placed n lw carbn emissins and the advantages f electricity when used in transprt in place f liquid fuels, the develpment f EVs is being prmted in the transprt sectr at an accelerated pace. Cmpared t ther electrical lads, EV lads have the ability t play a better rle in lad shifting and valley filling in the system lad prfile. With the increasing EV ppulatin, the evaluatin f the impact f aggregated EV lads with assciated battery charging characteristics n the system lad prfile becmes imprtant. The factrs influencing the charging characteristics can be categrized as internal and external [1]. Internal factrs include EV battery size, charging rates that differ depending n the amunt f battery charge remaining at the cmmencement f charging, gegraphical distributin f the charging infrastructure and charging habits. The nly factr that can be cnsidered as external is the time-f-use (TOU) pricing f electrical energy. The charging behaviurs f EVs have their wn randmness and intermittency and their impact n the pwer system is high, especially when all EV users start charging their batteries at the same time during lw tariff and peak lad hurs, requiring mre pwer frm the grid t balance the resulting increase in the generatin and peratin csts. Thus, there will be a grwing demand fr balancing measures such as the intrductin f pwer plants fr unprecedented peak lads and strage systems r demand respnse measures [1]. Since the EV charging lad is cntrllable, by studying it and cntrlling the charging prcess, it wuld be pssible t cnsiderably imprve the security and ecnmy f pwer system peratins. EVs wuld be a very attractive ptin fr the transprtatin industry in Sri Lanka due t the nature f its daily lad curve pattern which is skewed twards night with a high peak value. Mrs. R.M.G.D. Ranathunga, B. Tech. Eng, Open University f Sri Lanka. gdranathunga@gmail.cm. Eng. Lalith A. Samaliarachchi, C.Eng, MIE(Sri Lanka), B.Sc. Eng. (Hns) (Mratuwa), M. Eng.(AIT), MIEE(UK), Senir Lecturer, Department f Electrical & Cmputer Engineering, Open University f Sri Lanka. lasam@u.ac.lk 1

2 1.2 Types f Electric Vehicles EVs require highly efficient rechargeable battery packs. The electric mtr installed inside an EV replaces the internal cmbustin (IC) engine f a cnventinal vehicle. The energy strage capacity f an EV is almst equivalent t the energy stred in the fuel tank f a cnventinal vehicle. There are tw main types f EVs [1,1] available in the market, namely: Plug in Electric Vehicle (PEV) Plug in Hybrid Electric Vehicle (PHEV) Plug in Electric Vehicle (PEV) A PEV uses nly electricity fr prpulsin. The cst f peratin f PEVs is much lwer than that f cnventinal IC engines because f their higher energy efficiency [1]. Als, it has a great ptential t reduce city pllutin with zer emissins and lw nise Plug in Hybrid Electric Vehicle (PHEV) A PHEV can stre a significant amunt f energy within an n-bard battery fr use during daily driving and the battery can be recharged frm the electric grid as and when required. PHEVs have IC engines that can be used fr prpulsin when the battery is depleted and when higher acceleratin is required and this will increase the near-term marketability f PHEVs [2]. This research study was fcused n PEVs because f their ecnmic and technlgical impacts PEV Characteristics PEVs have several characteristics which differ frm thse f IC engine vehicles, and can be categrised accrding t their ecnmic and technlgical impacts: Ecnmic Impacts [2]: Withut any ecnmic benefits, custmers wuld nt pt t buy PEVs because f their higher investment cst and lw cnvenience due t their lng charging times. When there are substantial tax credits fr PEVs, the latter becme mre attractive and cnsequently, a higher penetratin rate f PEVs culd be expected. Technlgical Impacts [2]: The average daily cmmute distance in Sri Lanka can be assumed t be arund 5 km fr each EV (at a 5% charge level), which culd be translated int an energy cnsumptin f abut 12 kwh. By integrating EVs, the ttal demand n the pwer system will increase especially during the night peak, which will lead t a higher current flw in the netwrk, thereby creating higher transmissin and distributin lsses. Electricity generatin will als incur higher incremental csts and hence the system as a whle may als incur larger perating csts. 1.3 Electrical Vehicle Chargers An EV battery can be charged at different pwer levels, namely frm level-1(l1) t level- 3(L3). L1 and L2 charging are cnsidered as slw charging and fast charging respectively, and are cmmn in residential husehlds [4, 9]. Level-3 is cnsidered as fast direct current charging (CHAdeMO/Cmb) which is typically used in public areas because f its high pwer requirement and high installatin cst. With a standard electric cable, an EV can be plugged in t a standard electric utlet t charge the battery at L1. Hwever, L2 and L3 charging require additinal equipment. Therefre, L1 seems t be mre attractive since it des nt require additinal investment. Charging times will depend n bth the capacity f the charger and the EV battery [4], which as shwn in Table 1 will vary depending n the mdel f the car. Table 1 - Charger Level Classificatins Charging mde L1 Slw charging Husehld L2 Fast charging Husehld + Wrks+ Shpping malls L3 (Rapid charging) Recharging statin 1.4 Aim and Objectives The aim f this research was t investigate, develp and apply a reasnably acceptable mathematical mdel fr the predictin f the envisaged EV lads and t study the impact f same n the chrnlgical lad prfile f the pwer system in Sri Lanka [3]. The bjectives were: AC charging DC charging Charging type Single Phase Single phase r three Phase Three phase 63A Three phase 125A Rated Charging Lad current time Up t 13A Hurs kw 32A Up t 7-22 kw Up t 43 kw Up t 43 kw 3-4 Hurs 3-6min 8% Charging 3 min 8% Charging t identify and evaluate factrs that impact n the pwer system because f the aggregated EV lads and t establish a mdel t capture the cmbined effect f EV charging lads n the system lad prfile. 2

3 2. Methdlgy The EV battery charging prcess is subject t uncertainties such as charging schedule, lcatin preference and energy requirement, which cmplicate the charging scenari. These effects result in an uncertain ttal demand pattern which has t be catered t by the utility since PEV wners charge their vehicles in an unpredictable manner [2, 3, and 8]. Therefre, the randm variables that are t be identified and cnsidered wuld be: battery charging time (plug in time) battery charge duratin battery state f charge charging mde (L1, L2 r L3) number f EVs Given the circumstances, Mnte Carl simulatin methd will be well-suited t calculate the prbabilities f the events f deterministic systems with randm inputs. It will especially be useful when the system cnsists f mre than ne randm variable [2]. The generic algrithm f the Mnte Carl simulatin is as fllws: Determine the randm variables Generate randm numbers fr each randm variable Prcess the deterministic system using all f the randm variables There is n standard structure fr the Mnte Carl simulatin because f its dependency n the prblem. A flw chart that describes the structure f this study is shwn in Figure 1 [2, 1]. The first step is t initialize the fllwing tw types f parameters: a. EV Parameters Number f EVs and their charge preferences and lcatins with respect t the time f day Charging pwer Battery capacity (kwh) b. System Netwrk Parameters Surce vltage V Standard deviatin f the lad variatin develped and superimpsed n the system lad prfile. 2.1 Data Prcessing and Initializatin The first step f the simulatin was t arrange the data prperly and initialize the required parameters. The infrmatin abut the behaviur f the EV drivers was included as the data set f drivers. N N START Read data & initialize parameters Numbers f EVs Generate EVs Plug in time Battery state f charge in terms f hurs Generate EVs charging prfile Generate system lad prfile End f number f battery SOC cnsidered Yes End f all trials? END Yes Figure 1 - Flw chart The infrmatin abut the behaviur f the EV drivers had t be assumed since a realistic data set was nt readily available. 2.2 Existing EV Usage in Sri Lanka The details f EV usage in Sri Lanka btained frm the Mtr Traffic Department are shwn in Table 2. Fr each trial, the assumed arrival time (plug in time) and battery state f charge (charge duratin) were cnsidered; and this allwed fr the cmputing f the charge prfile fr each EV and fr the whle day. By cmbining the prbability distributin curves f EVs plug in time, battery charging duratin and the number f times f EV charging, demand curves were 3

4 Table 2 - Electric Vehicle Data Brand Name electric Battery mtr capacity capacity (kwh) (kw) NISSAN (ZAA-ZEO) NISSAN (LEAF) NISSAN (ZAA-ZEO) NISSAN (LEAF) MITSUBISHI (i-miev) *Other vehicles Ttal *Ppular brands and ther types Withut TOU: The mean plug in time was assumed t be 19 hrs. (µ=19 hrs. and σ=1.67 hrs.) as shwn in Figure 2a With Time f use tariff (TOU): The mean plug in time was assumed t be 223 hrs. (µ=223 hrs. and σ=.5 hrs.) as shwn in Figure 2b The standard deviatin f the randm number distributin is as per 99.7% cnfidence level f the standard nrmal distributin. 2.3 Arrival Times f PEVs The charging preferences f EV drivers are practically uncntrllable which means that EV drivers will try t charge their vehicles at any time at their wn cnvenience [5]. In this study, tw cases were studied in rder t enable the cmparisn f the impacts. In ne case, it was assumed that EV drivers wuld start charging their vehicles when they arrive hme at 19 hrs near peak time and in the ther case, it was assumed that they will start charging their vehicles at 223 hrs. which was during ff peak hurs. Figure 2a - Graph f N. f EVs withut TOU (µ=19 and σ=1.67) 2.4 Randm Number Generatin fr the Prbability Distributin As described belw, a set f unifrmly distributed randm numbers was generated fr the EVs with a given prbability distributin fr the EV plug in time and EV charge duratin. Generatin f a set f unifrmly distributed randm EV numbers fr tw EV mean plug in times (Prbability density curve f Time (24 hrs.) vs. N. f EVs). Tw plug in times were cnsidered with tw different tariff structures fr a better cmparisn since the Public Utilities Cmmissin f Sri Lanka (PUCSL) had intrduced a new ptinal tariff structure with effect frm 1 th September 215. The tw tariff structures used were with and withut time-f-use (TOU) tariff (wrst case scenari). The ttal number f EVs taken t bserve the sensitivity f the grwing impact were 1, 3, 5, 1, 25, 5, and 1. Figure 2b - Graph f N. f EVs with TOU (µ=223 and σ=.5) Figures 2a and 2b shw the randm number distributin f 1 EVs fr the tw plug in times. Generatin f a set f unifrmly distributed randm numbers fr the EV charge duratins. (prbability density curve f charge duratin vs. N. f EVs). This study assumed that t get fully charged, an EV battery required 3 kw fr 8 hurs and thus the average charge duratins were cnsidered as 2, 3, 4, 5 and 6 hrs. 4

5 Table 3 - Charge duratin data (Surce: 214 LEAF wners manual) Figure 3a - Charge duratin fr 3 hurs Charge methd Level 1 (Nrmal) Level 2 (Trickle) Level 3 (Quick) Charging vltage 22-24V 11-12V Public charging statin Charging time fr an empty 24 kwh battery Apprximately 4-7 hrs. Apprximately 21 hrs. Apprximately 3 min If an EV cmmences charging in the last few hurs f the day, it s end time is likely t g beynd hrs. In the circumstances, this prtin f the charge prfile will mve t the initial hurs f the next day. 2.7 Custmer Lad Prfile The daily ttal lad variatin btained frm the System Cntrl Branch f the Ceyln Electricity Bard is shwn in Figure 4. The data is prvided hurly fr bth weekdays and weekends. Figure 3b - Charge duratin fr 5 hurs 25 Figures 3a and 3b shw the prbability distributins f EV charge duratin fr 1 EVs fr a duratin f 3 and 5 hurs respectively. A fully discharged battery f an EV will require up t 8 hrs. t get fully charged since the charging time will als depend n the battery state f charge. Therefre, the X-axis f each f the tw graphs was limited t 8 hrs Pwer (kw) week day weekend 2.5 EV Battery State f Charge In practice, the battery state f charge (SOC) als has a strng impact n the EV charge prfile. The battery SOC f an EV can be estimated frm the PEV cmmute distance, energy efficiency which is based n the driver s behaviur, traffic cnditins, climatic cnditins etc. Hwever, the study assumed that all the EVs that were cnsidered used Liin batteries each with a capacity f 24 kwh. 2.6 EV Charge Prfile The ttal charge prfile f each EV was first cmputed individually based n the arrival time, energy requirement, and charging methd. The first rw in Table 3 shws the types f charging methds cnsidered in this study, as well as the charging pwer and charging time f a fully discharged 24 kwh Liin battery. 5 Hurs :3 1:3 2:3 3:3 4:3 5:3 6:3 7:3 8:3 9:3 1:3 11:3 12:3 13:3 14:3 15:3 16:3 17:3 18:3 19:3 2:3 21:3 22:3 23:3 Figure 4 - Daily lad prfiles f Sri Lanka When the tw lad curves shwn in Figure 4 are cmpared, it can be bserved that bth week-day and week-end lads start decreasing at 23 hrs. It can als be bserved that the lads between 223 hrs. and 63 hrs. (8 hrs.) n a weekday are belw the 15 MW bench mark and that they are still less during the weekend. 3. Test Results and Analysis The plug in times and charge duratins f the EVs had t be cnsidered as randm variables. Furthermre, because f uncertainty, distributins f randm variables five in 5

6 number had t be cnsidered. In all cases, the parameters f the distributins were such that they all had different means and variances. Hwever, all the distributins were based n the fact that 3 kw f pwer were required by each EV during the charging perid [8]. The ttal pwer f each EV after adding the transmissin and distributin lsses wuld be mre than 1.85% f the required 3 kw. Therefre, the ttal charging pwer required by an EV can be cnsidered as kw. The expected daily pwer demand f a typical EV culd be prbabilistically predicted. Hwever, in the ttal distributin, the pwer demand will differ frm ne anther. This gives us a useful insight int the expected charging pwer demands f EVs thrughut the day. Here, 1, 3, 5, 1, 25, 5 and 1 EVs were cnsidered fr a scheduling hrizn set fr a 24 hur interval [6, 7]. Fr a better analysis, as described in Sectin 2.4, tw mean times fr the plug-in times were cnsidered. The demand prfiles f the tw plug in times had five different patterns depending n the charge duratins since charging times differ frm EV t EV. In the absence f realistic figures, it was reasnable t assume that the EV driving behaviur and the battery state f charge are indirectly linked t sme degree. Therefre, the average battery charge duratins were cnsidered t be 2, 3, 4, 5 and 6 hrs. The standard deviatin f the randm number distributin was calculated fr the cnfidence level f the distributin (i.e. 99.7% f the data were within three standard deviatins (µ+3σ) f the mean). Mean (µ 1) = 2, standard deviatin (σ 1) = 2 Mean (µ 2)= 3, standard deviatin (σ 2) = 1.67 Mean (µ 3)= 4, standard deviatin (σ 3) = 1.33 Mean (µ 4)= 5, standard deviatin (σ 4) = 1 Mean (µ 5)= 6, standard deviatin (σ 5) =.67 Five sets f EV charge demand curves were prduced fr the tw plug-in times using the abve mentined values f the means and the standard deviatins. 3.1 EV Charge Prfile (withut TOU) Figures 5 and 6 shw the EV lad prfiles fr 1 EVs at 19 hrs. fr a plug-in time and battery charge duratin f 3 hrs. and 4 hrs. respectively. Mean (µ 2) = 3, standard deviatin (σ 2) = 1.67 Figure 5 - EV lad prfile fr charge duratin f 3 hrs. Mean (µ 3) = 4, standard deviatin (σ 3) = 1.33 Figure 6 - EV lad prfile fr a charge duratin f 4 hrs. 3.2 Charge Demand Curves (withut TOU) f all EVs Cmbined Table 4 - EV charge demand data (MW) Time Charge Duratin (hrs.) µ = 2 µ = 3 µ = 4 µ = 5 µ =

7 DEMAND (MW) EVs charge Demand Curves µ = 2 µ = 3 µ = 4 µ = 5 µ = 6 TIME OF DAY Figure 7 - EV charge demand curves (withut TOU) Figure 7 shws five different demand curves fr 1 EVs with different charge duratins. Plug-in time was cnsidered as 19 hrs. 3.3 Daily Demand Curves (withut TOU) Ttal Demand = [Nrmal daily lad prfile + EV charge lads fr different hurs f charge duratins (µ= 2, 3, 4, 5, 6)] Table 5 - Ttal demand values fr different charge duratins Time Ttal Demand (MW) (hrs.) µ = 2 µ = 3 µ = 4 µ = 5 µ = Table 6 - EV charge demand data Time Charge Duratin (hrs.) µ = 2 µ = 3 µ = 4 µ = 5 µ = DAILY DEMAND CURVES WITH DIFFERENT CHARGING DURATIONS µ = 2 µ = 3 µ = 4 µ = 5 µ = Figure 8 - Daily demand prfiles fr different charge duratins Figure 8 shws the nrmal daily demand curve fr different charge duratins superimpsed n 1 numbers f EV lads plugged in 19 hrs. mean time (wrst case scenari). Nte that all f the charge duratin curves are falling in line since 1 EVs will nt have much f an impact. 3.4 EV Charge Prfile (with TOU) Figures 9 and 1 shw the EV lad prfiles fr 1 EVs with the plug-in time at 223 hrs. and with a battery charge duratin f 2 hrs. and 4 hrs. respectively. Mean (µ 1) = 2, standard deviatin (σ 1) = 2 Figure 9 - EV lad prfile fr a charge duratin f 2 hrs 7

8 Mean (µ 3) = 4, standard deviatin (σ 3) = 1.33 Figure 1 - EV lad prfile fr a charge duratin f 4 hrs 3.5 EV charge demand curves (with TOU) cmbined EV Charge Demand Curves Pwer (MW) µ=2 µ=3 µ=4 µ=5 µ=6 Time f a day Figure 11 - EV charge demand (with TOU) The chart in Figure 11 shws EV charge demand prfiles fr different charge duratins. 3.6 Ttal Demand Curves (with TOU) Ttal Demand = [Nrmal Daily Lad Prfile + EVs Charging lads with different charging duratins. (µ= 2, 3, 4, 5, 6)] TOTAL DAILY DEMAND CURVES WITH DIFFERENT CHARGING DURATIONS µ = 2 µ = 3 µ = 4 µ = 5 µ = 6 Table 7 - Ttal demand values fr different charge duratins Time Ttal Demand (MW) (hrs.) µ = 2 µ = 3 µ = 4 µ = 5 µ = Figure 12 shws the nrmal daily demand curve superimpsed n the 1 EV charge lads fr different charge duratins. Since the vehicle number is small, the charge duratin curves fall in line and are nt significantly visible. 3.7 Ttal Demand Curves fr Different Numbers f Vehicles The EV lads and their cnsumptin behaviurs can be cntrlled with the intrductin f the TOU by PUSL. Since EV lads will depend n the number f vehicles, different numbers f EVs were then cnsidered t ascertain the sensitivity f their impact. The number f EVs was taken as 1, 3, 5,1,25,5 and 1 with an average battery charge duratin f 5 hrs. t make it simple Figure 12 - Ttal daily demand curves fr different charge duratins 8

9 3.8 EV Charge Demand (withut TOU) Table 8 - EV demand (withut TOU) in MW Time EV# EV# EV# EV# EV# EV# EV# (hrs.) ELECTRIC VEHICLES CHARGING DEMAND CURVES EVs n: 1 EVs n: 3 EVs n: 5 EVs n: 1 EVs n: 25 EVs n: 5 EVs n: Figure 13 - EV charge demand (withut TOU) Figure 13 shws seven demand curves fr a 5 hr. charge duratin and fr several numbers f electric vehicle lads. Plug-in time was cnsidered as 19 hrs. 3.9 EV Charge Demand (with TOU) Table 9 - EV demand (with TOU) in MW Time (hrs) 1 EV# EV# EV# EV# EV# EV# EV# ELECTRIC VEHICLE CHARGING DEMAND CURVES EVs n: 1 EVs n: 3 EVs n: 5 EVs n: 1 EVs n: 25 EVs n: 5 EVs n: Figure 14 - EV charge demand (with TOU) Figure 14 als gives a cmparisn f lad prfiles f different numbers f electric vehicle lads. It shws that when the number f EVs increases, the energy cnsumptin f the EVs will als increase. The plug-in time was cnsidered as 19 hrs. 3.1 Daily Demand Curve (withut TOU) If nt fr the intrductin f the TOU tariff, the ttal demand curves fr different numbers f vehicle fleets wuld have been as shwn in Figure 1A f Appendix-1. It als shws that there are several peaks that ccur during day time, with the highest values generally ccurring in late evening, with significant spikes in the mrnings and in early afternns. All demand curves reach their peaks during the same time interval. The highest peak lad is nted between 19 hrs. and 21 hrs. while mrning spikes are fund between 5 hrs. and 7 hrs. As can be seen frm the curves in Figure 1A f Appendix-1, the cntributin f EV vehicles t the peak demand wuld be 2.74, 7.842, , , 65.82, and MW fr 1, 3, 5, 1, 25, 5 and 1 EVs respectively Ttal Daily Demand Curves (with TOU) Figure 1B in Appendix-1 shws that the ttal demand prfiles with the TOU tariff are generally different frm ne ther. All demand curves reach the peak lad at the same time. The highest peak lad is nted between 183 hrs. and 213 hrs. while mrning spikes are nted between 43 hrs. and 73 hrs. It is interesting t see that the effect n the system peak is minimal with TOU. 9

10 4. Cnclusin Electric vehicles have an increasing ptential in the vehicle market while issues related t energy crises are still ging n. Plug-in EVs create pprtunities t reduce the reliance n fssil fuels thereby lwering the perating csts. EVs can be regarded as distributed energy strage units and have a great ptential t cntribute t lad shifting in lad management. The incnclusive nature f EV charging behaviurs brings in great challenges t pwer grid peratin and peak-valley regulatin. EV charging lads are mainly determined by EV types, charging duratins, charging mdes, and battery characteristics. The charging f a large number f EVs will have an impact n the transmissin and distributin f a pwer system. The ttal pwer required frm the system by an EV was apprximately kw which included its transmissin and distributin lsses. The EV charge demand curves were prepared by cmbining the prbability distributin curves f bth variables, and charge demand curves were subsequently superimpsed n the system lad prfile. Because f its stchastic nature, the prbabilistic distributin f the variables such as the number f EVs, plug-in times and battery charge duratins were cnsidered using the Mnte Carl simulatin technique. The impact due t the plug-in time was cnsidered with and withut the TOU tariff structure fr a better cmparisn. Since the EV usage in Sri Lanka is n a psitive trend, if nt fr the TOU f electrical tariff system, the EV s charging behaviur in time t cme wuld have a significant effect n the system demand during the peak lad perid. As can be seen, the prpsed methdlgy gives an idea abut the impact f EVs n the system lad prfile. The impact f the EV lads after the intrductin f the present TOU tariff seems t be minimal. Acknwledgement The authrs wuld like t ffer their appreciatin t Dr. Narendra de Silva and Eng. Sahan Edirisinghe f LECO, Eng. M. Lakshitha Weerasinghe f CEB and Mr. I. A. Premaratne f OUSL fr the technical guidance prvided. References 1. Nanjun, L. U., Javad, L., Electric Vehicles Charging Lad Calculatins Based n Mnte Carl Methd, Faculty f Eng., The University f Clumbia, 15 th May 214, pp Au, T. K., Assessment f Plug in Electric Vehicle Charging n Distributin Netwrks, M.Sc. in Electrical Engineering. University f Washingtn, 212, pp Gu, Q., Wang, y., Sun, H., Zhengshu Li., Xin, S., Zhang, B., Factr Analysis f the Aggregated Electric Vehicle Lad Based n Data Mining, University f Beijing, 21th June 212, Energies 212,5, pp Zeng, D., Wang, K., Yaping, Li., Xiarui, G., Jiang, X., Lad Cluster Characteristic Analysis and Mdelling f Electric Vehicles, Electric Pwer Research Institute, China, June 213, pp Alns, M., Amaris, H., Germain, J. G., Galan, J. M., Optimal Charging Scheduling f Electric Vehicles in Smart Grids by Huristic Algrithms, Dept. f Elecrical Eng.,University f Carl,Spain,17 th April 214,Energy 214,7, Putrus, G. A., Suwanapingakarl, P., Jhnstn, D., Bentley, E. C., Narayana, M., Impact f Electric Vehicles n Pwer Distributin Netwrks, Schl f Eng. and Infrmatin Science, The University f Nrthumbria, UK, pp Hung, S., Infield, D., The ptential f Dmestic Electric Vehicles t Cntribute t Pwer System Operatin thrugh Vehicle t Grid Technlgy, Dept. f Electrnic & Electrical Engineering, The University f Strachclycle, UK. 8. Huang, S., Infield, D., The Impact f Dmestic Plug-in Hybrid Electric Vehicles n Pwer Distributin System Lads, Dept. f Electrnic Electrical Eng., University f Strathclyde, UK. IEEE. 9. Taylr, J., Maitra, A., Alexander, M., Brks, D., Duvall, M., Evaluatin f The Impact f Plug in Electric Vehicle. Lading n Distributin System Operatin, pp Rsell P., Rbels R.V., Sumper A., Jane J.B., Prbabilistic Agent Based Mdel f Electric Vehicle Charging Demand t Analysis the Impact n Distributin Netwrk, 11 th May 215, Energies 215,8, pp

11 Appendix ELECTRIC VEHICLE CHARGING DEMAND CURVES (WITHOUT TOU TARIFF) Withut Evs EVs n: 1 EVs n: 3 EVs n: 5 EVs n: 1 EVs n: 25 EVs n: 5 EVs n: Figure 1A - EV charge demand curves fr different number f EVs (withut TOU tariff) 24 ELECTRIC VEHICLE CHARGING DEMAND CURVES (WITH TOU TARIFF) Withut Evs EVs n: 1 EVs n: 3 EVs n: 5 EVs n: 1 EVs n: 25 EVs n: 5 EVs n: Figure 1B - EV charge demand curves fr different number f EVs (with TOU tariff) 11