Análise do impacto da infraestrutura de recarga de veículos elétricos no planejamento da expansão dos sistemas de distribuição

Abstract

With the increasing penetration of electric vehicles (EVs) the necessary charging infrastructure deployment becomes imperative. Thus, it is necessary to consider the increase in the demand associated with the EVs charging in the distribution systems expansion planning (DSEP), determining the necessary investments to meet the demand with competitive costs. In this way, methodologies to support the distribution system must be developed, considering the uncertainties related to both the conventional and the EV demand. Furthermore, other characteristics must be observed in relation to the insertion of EVs on the EDSs such as their location of connection and the charging type. In this context, the present work shows a two-stage stochastic model of mixed-integer linear programming (MILP) to solve the static expansion planning of distribution power system problem. The expansion planning assesses the construction and/or reinforcement of substations and circuits as well as the allocation of distributed generation units, capacitors banks, and electric vehicles public recharging stations (EVPS). The construction of scenarios considers a set of historic data to represent the stochastic features and the correlation between conventional and EV demand, wind and solar generation. To properly represent the historic EV demand data the Gaussian mixture models (GMM) methodology his used, modeling the main metrics of EVs recharge. To reduce the number of scenarios the present work makes use of the k-means clustering method generates. Finally, the importance of considering the uncertainties associated with the EVs demand in the DSEP is evaluated. As the recharge infrastructure development for EVs in public places is strategic some governments have embraced actions to install PSEVs. So, the present work also address the impact of the implementation profile of PSEVs on the DSEP. In this regard, two new constraints are proposed to compose the mathematical model of sizing the PSEVs inside the multistage expansion planning of distribution systems. This approach considers a deterministic MILP model that assesses the investment decisions over the considered planning horizon. To evaluate the results of the inclusion of those constraints a case of study is proposed. Both deterministic and stochastic models use the 18-bus test system in the simulations. The results show the importance of considering the characteristics concerning the insertion of EVs in the distribution systems, aiming to obtain expansion plans more adapted to the real conditions.