Electric vehicles (EV) load model is an important parameter to solve power flow of the grid. In addition, size and position of EVs load can be considered in optimal requirement for minimizing impact on the grid. Further, power flow analysis under various electric vehicle load models are presented, these models are based on modified backward and forward sweep method^{ }^{2}. An impact of PEVs on distribution network voltage unbalance is investigated. Additionally, voltage unbalance factor (VUF) is reduced optimally by choosing state (charging or discharging) of PEVs. Further, effect of coordinated or uncoordinated charging of electrical vehicle is examined ^{3}. A decentralized method is proposed to avoid overloading on feeders. By applying the suggested technique proposed solution converges and satisfies the limit of the feeder lines^{ }^{5}. To- efficiently use vehicle to grid technique by adding project based battery reduction cost to embed into micro grid, two energy management planning is suggested for addition of vehicle to grid into micro grid on basis of forecast accurate power supply and demand, these strategies are further implemented on multi-agent system ^{6}. The process of charging introduce oscillations in power system that reduces stability, this process can be examined by comparing EV load and normal power system load ^{7}. Therefore problem need to be solved of EV that affect various buses of electric power system, charging process of EV create voltage imbalance, voltage difference between two phases ^{8}. Another aspect to consider EV’s in future is that oil reserves are depleting day by day therefore electric vehicles have been introduced that can be recharged at recharging stations, but process of recharging can affect electric power system. Other benefits to use EV’s are that they are relatively cheaper to run and their maintenance cost is also low, beneficial for environment ^{9}. Bidirectional vehicle to grid method is introduced to use the energy from EV in a parking lot, this system is checked in real power distribution network load, from simulated results it was known that this technique provide savings and reducing peak demand of connected grids ^{10}. A composed distribution location marginal pricing method to make less severe hindrance due to EV load in power system, in this method distribution system operator (DSO) find distribution location based marginal prices, function of DSO is to manage distribution network, operating at low and medium voltage ^{11}.

A backward forward sweep technique is used for simulation, standard IEEE 33-bus RDS is considered for simulation, IEEE 33-bus system is designed with its standard electrical parameter by using ETAP software. Simulation is designed in such a way that PEV(plug in electric vehicle) is connected with recharging stations (RS) to charge its battery, in this study one recharging station contains of 8 number of PEVs, means maximum 8 PEVs can be connected to single RS. Generally battery of electrical vehicles is rated in kWh, kWh is divided with number of hours to get kW rating of single PEV, by getting kW rating further simulation is carried out in ETAP software in which main components of RDS are buses, branches, utility grid and load. It consists of 33 buses, 32 branches. Load of single RS is combination of load of 8 number of PEVs, further RS are connected to different number of buses, as the load on system is increased system active and reactive power losses are increased. In addition of management of load, distributed generators (PV array) are installed at certain buses to compensate system losses. The methodology for simulation of IEEE 33-bus system with EV load model consists of following steps.

The methodology for simulation of IEEE 33-bus system with EV load model consists of the following steps:

Step1. Preparation of the network that can be used as branch numbering method.

Step2. Deciding load of single PEV unit and complete load of single recharging station.

Step3. Integrating recharging stations at certain buses of system.

Step4. Integrating distributed generations (PV cells) within same system for compensation of losses and load management.

Step5. Run load flow analysis and to obtain result.

Forward Sweep (FS) method is based on Kirchhoff’s voltage law (KVL) that is used to determine voltage with most recent branch currents, with this method, nodal voltage is computed at each node, the value is updated in FS by starting from slack bus towards the end of network. Node voltage is computed by making difference of the last node V_{L1}^{K }, from voltage drop in transmission line _{L2}^{K}:

Where V_{L1}^{K }is sending end node voltage, _{ L2K is receiving end node voltage, ILZLK is voltage drop in transmission.}

Backward Sweep is based on KCL, it is used to determine current with most recent voltage updates. The direction of current in branch has a backward direction, that starts from backward direction towards starting slack bus, with the help of equations can be explained as.

Equation (equ (2)) is load on power system, where S_{i }apparent power, P_{i }is actual power and Qi is reactive power of system.

When EVs are connected, active power losses on network is shown in equ (3). Where _{i, EV }is_{ }total active power losses of network load and EV, _{ } P_{i }active power losses due to network load_{ ,}P_{EV }is active power losses due to EV load.

When EVs are connected, reactive power losses on network is shown in equ (4). Where _{i, EV }is total reactive power losses due to network load and EV load,_{ }Q_{i }is reactive power losses due to network load,_{ }Q_{EV }is reactive power

Total load on power system can be rewritten as equ (5).

In this research type of PEV load model is invariable current load, it consists of active power and reactive power achieved by exponential load, which is founded on typical type of general load exponential indices

Eqs (6) and (7) shows active and reactive power of constant current load model respectively.

This paper carries 5 case studies, including the size of DG, the location of DG with different electrical vehicle load levels. Each case study has been designed by observing the power losses on that bus, the maximum loading at that bus or minimum voltage level. At final, it was observed which case will be more beneficial practically.

RDS with 1 DG: 70% of load size connected at bus 06.

RDS with 1 DG: 30% of load size connected bus at 18.

RDS with 1 DG: 30% of load size connected at bus 32.

RDS with 2 DGs: DG 1 with 70% of load size connected at bus 06, DG 2 with 30% of load size connected at bus 32.

RDS with 2 DGs: DG 1 with 30% of load size connected at bus 18, DG 2 with 30% of load size connected at bus 32.

Total load of IEEE 33-bus RDN 3.72MW for active power, 2.3MVAR for reactive power losses.

By integrating PEVs load in RDN, load of 1.08MW is increased. Therefore, %DG of total load size is…..

70% of 4.80MW is 3.36MW.

50% of 4.80MW is 2.4MW.

30% of 4.80MW is 1.44MW.

In

In

In

In

In

In

In

In

In

In

Case
Active power losses (kW)
Reactive power losses (KVAR)
%improved active power losses
%improved reactive power losses
Base
211
143
--------------------
---------------------
With PEV load
308
210
--------------------
---------------------
Case 1
143
108
53%
48%
Case 2
189
134
38%
36%
Case 3
191
138
37%
34%
Case 4
139
106
54%
49%
Case 5
122
92
60%
56%

MinimumVoltage (p.u) | Base Case | With PEV load | Case 1 | Case 2 | Case 3 | Case 4 | Case 5 |

0.9038p.u(18thbus) | 0.8804p.u(18th bus) | 0.9249p.u(18thbus) | 0.9204p.u(33thbus) | 0.9004p.u(18th bus) | 0.9438p.u(18th bus) | 0.9685p.u(29th bus) |

In this study, two cases are suggested to limit power losses and improve voltage profile within acceptable limit. First case is RDS with 1 DG: 30% of load size connected at bus#18, in this case 38.43% active power, 36.42% reactive power losses are improved, further voltage is improved from 0.8804 P.U to 0.9204 P.U at weakest bus. Second suggested case is RDS with 2 DGs: DG 1 with 30% of load size connected at bus #18, DG 2 with 30% of load size connected at bus #32. In this case 60.1% active power losses, 56.27% reactive power losses are improved and voltage is improved from 0.8804 P.U to 0.9685 P.U. Additionally, this study provides solution to improve system losses and voltage profile optimally by using backward-forward sweep method with minimum size of DG units that ultimately reduces the cost of distribution system.