A new optimization model and solution approach for the Electric Vehicle Routing Problem (EVRP) is presented in this research, where the electric vehicles are used to deliver the products from manufacturing plant to multiple customers through a split delivery system which allows vehicles to partially recharged and return to a depot to get more loads if needed. The proposed model is a Mixed Integer Linear Programming (MILP) model used to determine the optimal route achieving the desired objective function for each vehicle which ensures maximization of customer satisfaction by minimizing total amount deviations and minimization of total distribution time. CPLEX solver is used to solving the different sizes of data sets to test the performance of the model, while a new heuristic approach is constructed to solve the same data sets. Then the results of the heuristic are compared with those obtained by CPLEX. The results of the small size data sets show that the CPLEX and heuristic obtain a competitive solution in terms of both computation time and solution quality. For medium-size data sets, the proposed heuristic can find the same or better solution for some of them compared with the result produced by CPLEX.