The global shift towards electric vehicles (BEVs) to reduce carbon emissions necessitates a robust charging infrastructure. Inadequate charging infrastructure significantly impacts BEV adoption rates. Urban flooding, exacerbated by climate change, poses a significant threat to this infrastructure. Previous research has focused on the impact of flooding on traffic, but not on the specific implications for BEV charging networks. This study addresses this gap by examining how geographically correlated charger outages caused by flooding affect the public EV charging network in Greater London. This region was selected because of its high EV sales, significant flood vulnerability, and high reliance on public parking (which is most at risk during floods). The study aims to quantify the impact of flooding on charging patterns and propose strategies to improve infrastructure resilience, ultimately contributing to a smoother transition to widespread BEV adoption.

Existing literature extensively studies the effects of flooding on traffic flow and general urban infrastructure. However, studies specifically analyzing the impact of flooding on the rapidly expanding EV charging infrastructure are lacking. Research on multi-infrastructure cascading effects examines the simultaneous impacts of floods on various networks, but usually doesn't consider EV charging infrastructure. Studies assessing charging infrastructure adequacy, given a certain BEV penetration level, do not explore the effects of flooding. Although some research advocates against siting chargers in flood-prone areas, this may be impractical in densely populated areas. This study bridges this gap by focusing on the unique aspect of geographically correlated charger outages during flooding, providing crucial insights into how flooding affects the charging infrastructure and subsequently, BEV adoption.

The researchers simulated BEV trips in Greater London using publicly available data on vehicle driving patterns and the locations of 5925 public chargers. They divided Greater London into uniform grids and identified flood-risk grids using data from Climate Central. Three flooding scenarios (representing increasing flood intensity) were simulated, with each at-risk grid having a probability of 0.5, 0.7, or 0.9 of being flooded. When a grid flooded, the chargers within it were considered out of service. The simulation modeled BEV trips based on real-world driving patterns, considering factors such as trip purpose, departure time, and vehicle speed. The researchers tracked two key metrics: charger utilization (percentage of time a charger is in use) and distance to the nearest available charger. They ran 100 simulations for each scenario, varying driving patterns and flooded chargers to obtain statistically robust results. Four mitigation strategies (ring-fencing, usage-dependent placement, distance-based placement, and random placement) were evaluated by adding 5% of additional chargers to the network, assessing their impact on charger utilization and accessibility.

The study found that even with over 34% of public chargers out of service due to flooding, the success rate of BEV trips remained high (over 99.7%). However, flooding caused disproportionate stress on the remaining charging infrastructure. While mean charger utilization decreased, maximum utilization increased significantly, indicating that the remaining chargers bore a heavier load. Surprisingly, the impact was felt not just near flooded areas, but also in regions 10-13 km away. These distant areas experienced up to a 50.9% increase in charger utilization and a 269.9% increase in distance to the nearest charger. This effect was most pronounced in areas already experiencing high charger utilization, highlighting the vulnerability of stressed parts of the network. The four mitigation strategies all improved charger accessibility. The random placement strategy showed the highest city-wide impact, while ring-fencing and distance-based placement were more effective in reducing stress specifically near and far from flooded areas, respectively. The usage-dependent strategy performed best in mitigating charger utilization stress. The usage-dependent strategy showed the best results overall, indicating that adding chargers in areas of high demand is the most effective strategy.

The findings demonstrate that while flooding might not immediately prevent BEV usage, it creates significant stress points in the charging infrastructure, disproportionately impacting already-burdened areas. The observation that effects propagate far from the flood zones indicates the need for strategic planning beyond immediate flood risk areas. The success of the usage-dependent strategy emphasizes the importance of considering current demand when planning flood mitigation measures. The study's results have implications for urban planners and policymakers, urging them to prioritize flood resilience when expanding EV charging infrastructure.

This research highlights the vulnerability of urban EV charging infrastructure to flooding. The disproportionate impact on already stressed sections of the network underscores the need for proactive mitigation strategies. The study's findings suggest that usage-dependent placement of additional chargers is an effective strategy for improving flood resilience. Future research could explore more sophisticated optimization models incorporating detailed cost-benefit analyses and high-resolution data to fine-tune charger placement strategies. Further work could investigate the impact of different flood scenarios and their temporal dynamics on charging infrastructure.

The study assumes that flooded chargers remain unavailable throughout the day. It also doesn't consider dynamic flood changes or interoperability issues between BEVs and chargers. The simulation relies on existing travel patterns, which may not fully reflect future BEV usage changes. Finally, the four mitigation strategies are qualitative; a quantitative approach considering cost and other factors would strengthen the recommendations.