How much smart charging is smart?

The threat of climate change motivates the expansion of variable renewable energy sources (VRES) as well as the transition to plug-in electric vehicles (PEV). Both technologies present unique challenges to electricity systems, potentially leading to high grid expansion cost. In this paper, we present a detailed assessment on how smart-charging of PEVs can mitigate these problems and reduce cost on the distribution grid level. To this end, we propose a heuristic policy that dynamically decides on charging of PEVs, curtailment of VRES, and subsequently on infrastructure investments trading off fixed investments with variable operational costs. The main inputs are modeled as stochastic and the proposed policy is non-anticipative. We conduct a comprehensive case study for Germany in the year 2035 using detailed descriptions of existing distribution grids, realistic driving patterns, and real-world VRES feed-in data. Potential savings of €6.2 billion in investment costs lead to a reduction in total distribution grid cost of around 19%. This result is achieved by upgrading 7 million (21%) PEV chargers all over Germany. Upgrading 100% of chargers to smart-chargers as is proposed in the extant literature is clearly sub-optimal as it leads to significantly higher total cost. A closer investigation for single grids reveals that savings as well as the optimal share of smart chargers varies widely between grids. In particular, the potential of smart charging is much greater in rural areas than in urban centers. Furthermore, the results suggest that curtailment of VRES production is economically only in rare circumstances.