As the interest in electric heavy-duty vehicles has grown, so has the need for route planning tools to coordinate fleets of electric vehicles. This problem is called the Electric Vehicle Routing Problem (EVRP) and it can be solved using a Branch-Price-and-Cut method, where routes for individual vehicles are iteratively generated using information from the coordinated problem. These routes are computed in a pricing problem, which is a Resource Constrained Elementary Shortest Path Problem (RCESPP). Because of its iterative nature, the Branch-Price-and-Cut method is dependent on a fast solver for this RCSPP to get a good computational performance. In this thesis, we have implemented a labeling algorithm for the RCESSP for electric vehicles with state-of-the-art acceleration strategies. We further suggest a new bounding method that exploits the electric aspects of the problem. The algorithm's performance and the effect of the different acceleration strategies are evaluated on benchmark instances for the EVRP, and we report significantly improved computational times when using our bounding method for all types of instances. We find that route relaxation methods (ng-routes) were particularly advantageous in test instances with a combination of clustered and randomly distributed customers. Interestingly, for test instances with only randomly distributed customers, ng-relaxation required longer processing time to achieve elementary optimal routes and for these instances, the bounding methods gave better computational performance.