Abstract
In this article, we address the problem of how to consider load demand from fast-charging stations (FCS) of electric vehicles in distribution grid planning. As the electrification of the transport sector proceeds, connection of charging stations becomes relevant in an increasing number of grid planning cases. However, as many of the inputs and assumptions to distribution grid planning studies are associated with uncertainty, it is not clear how important the modelling of FCSs is for the estimated grid investment costs and the risk of over- and underinvestment in grid capacity. To address this, we apply global sensitivity analysis (GSA) to a grid investment optimization model, which allows us to rank the input factors according to their contribution to the output uncertainty. A voltage-constrained 22 kV distribution grid with expected connection of a FCS is used as case study. The results of the sensitivity analysis show that the uncertainty of the voltage limit planning criterion and the placement of the FCS in the grid, are responsible for most of the output variance of the calculated present value costs of grid investments. This indicates that, to improve grid investment decisions, it is more important to select an appropriate voltage limit and have knowledge about the location of new FCSs, than to improve the estimates of other FCS-related input parameters. In the paper we discuss the use of GSA for guiding data gathering and modelling efforts, leading to improved models and analyses for electric vehicle integration and distribution grid planning.