Plug-in Electric Vehicle Planning Toward DDPP Constrained by Electricity Grid Limitation Ali Hajebrahimi 1 and Innocent Kamwa 2 Abstract— Electric vehicle (EV) has alluded as a solution for CO 2 emission reduction in the transportation sector. However, uncontrolled penetration of EVs considering power grid limita- tion will increase CO 2 emission in the electricity sector. Hence, in this paper, a decomposed model of EVs planning is proposed to obtain the optimal penetration of EVs considering associated uncertainties. Moreover, a new bi-level charging/discharging control which considers both desires of the PEVs and the system operator is addressed in this paper. The results demonstrate that it is possible to increase the penetration of EVs up to 30% by 2025 while reducing the total load curtailment by 37% and the total emission by 28% compared to the baseline case with no supervisory EV charging/discharging control. The proposed planning problem is applied to Ontario’s grid considering existing and projected plans of transmission and generation expansion. I. INTRODUCTION Deep Decarbonization Pathway Projects (DPPP) are struc- tured on three pillars [1]: 1) Energy efficiency and conservation. 2) Low carbon electricity. 3) Electrification of end-use sector . It is not possible to reach deep decarbonization targets if any of these pillars aren’t accomplished at sufficient level [1]. Amongst the three mentioned pillars, the electrification of the end-use sector includes the efforts for shifting end-use energy consumption from fossil fuels to zero-carbon fuels. In summary [1]: • Shifting from coal to natural gas. • In the longer run, shifting to decarbonized energy car- riers. • Adoption of electric, biofuel or hydrogen vehicles. Therefore, the electric vehicle (EV) in transportation sector can be considered as a solution to reduce the CO 2 emission in this sector [2], [3]. However, it is necessary to mention that the required energy of EVs should be provided through a clean electricity sector [2]. Otherwise, increasing the pen- etration of EVs would cause a shift from a decentralized emission production to a centralized one. Therefore, the optimal penetration of electric vehicle in the electricity grid has turned to a controversial issue in power system studies [4], [5], [6]. The majority of previous studies have disregarded the impact of transmission constraints on PEVs penetration [4], *This work was not supported by any organization 1 A. Hajebrahimi , Laval University, Quebec City, QC G1V 0A6, Canadaa (e-mail: ali.hajebrahimi.1@ulaval.ca) 2 I. Kamwa is with Hydro-Quebec/IREQ, Power Systems and Mathemat- ics, Varennes, QC J3X 1S1, Canada (e-mail: kamwa.innocent@ireq.ca) [5]. In this context, the authors in [6], [7] have investigated the optimal transition to hybrid EVs considering electricity grid limitation. However, the uncertainties related to EVs mobility and renewable generations are overlooked. In [6], the EVs are considered as simple loads which appear only on weekends, and there is no charging/discharging control pol- icy. In order to obtain the maximum penetration of PEVs in the electricity sector in this paper, a Monte Carlo simulation is applied to investigate the impact of EVs uncertainties as well as the renewable generation on the optimal transition to the electric vehicle. The required energy of EVs to be fully charged, arrival and departure time and wind generation are considered as uncertain variables in this paper. Herein, the EVs can participate in different charging programs proposed by the independent system operator (ISO). Three charging control strategies are investigated here. In the first control strategy, the EVs start to charge and discharge as soon as they arrive homes and there is no control policy for charging. At the second charging control strategy, a home-based charg- ing/discharging control which obtains the desired charging profile of EVs is applied [8]. A bi-level charging control as the third control strategy is proposed in this paper in which the EVs participate in different programs satisfying both interests of end-user and the ISO for charging EVs. More- over, a benders decomposition is developed here for EVs planning such that the optimal number of electric vehicle fleets are determined through the designated feasibility and infeasibility benders cuts. The proposed planning problem is applied on simplified Ontario’s electricity grid. It is shown that the proposed model increases the penetration of EVs up to 30% by 2025. The rest of the paper is organized as follows: Section II represents the planning problem, charging and discharging management, feasibility check subproblem and operation subproblem, respectively. The results are explored in Section III and the concluding remarks are drawn in Section IV. II. FORMULATION The planning problem is decomposed into one master problem and three subproblems. Monte Carlo simulation is implemented here to generate the scenarios corresponding to required energy of EVs (E re ), arrival time (t a ), departure time (t d ) and wind farm generation. Afterward, a scenario reduction using SCENRED tool in GAMS is applied to extract the most probable scenarios. Then, the generated scenarios are submitted to home-based charging manage- ment and two subproblems. In the next step, the master problem specifies the number of the electric vehicle fleet 978-1-5386-5583-2/18/$31.00 ©2018 IEEE