Resilience of Urban Public Electric Vehicle Charging Infrastructure to Flooding

 

Research Fellows: Gururaghav Raman and Gurupraanesh Raman (equally contributing authors) | Advisor: Jimmy C.-H. Peng | Project Duration: 2021 - 2022

Cite this work as:

G. Raman, G. Raman, and J. C.-H. Peng, "Resilience of urban public electric vehicle charging infrastructure to flooding", Nature Communications, vol. 13, no. 3213, June 2022. DOI: 10.1038/s41467-022-30848-w.

An adequate charging infrastructure is key to enabling high personal electric vehicle (EV) adoption rates. Public EV chargers are particularly important in contexts such as Greater London, where over 40% of car owners depend on on-street parking and will therefore be dependent on public chargers for their charging needs. 

Shocks such as flooding present an interesting scenario because a significant fraction of public EV chargers will become simultaneously unusable, either due to flood-related damage, or when parking spaces become flooded, deterring EV users. Arguably, under normal circumstances, it is unlikely that a such a large section of chargers will be taken offline. In this study, we examine the impact of flooding on the public EV charging infrastructure in Greater London.

Urban flooding has disproportionate impact on EV charging network

 

We simulated personal battery EV (BEV) rides in Greater London using publicly available data on vehicle driving patterns. Considering the 5,925 public EV chargers currently located in the region, we studied how the stress on the charging network changes when flooding occurs.

Referring to Fig. 1, we observe that a significant fraction of the chargers lie in flood-vulnerable areas, which constitute the region around the Thames. We consider two metrics to quantify the stress on the charging infrastructure: (i) the charger utilization, which is the fraction of the day when a charger is utilized, and (ii) the distance to the nearest available charger from each building in the city. The first metric measures the availability of the charging network to EV users that require them, and the second, the comfort level of EV users in finding an available charger.

 
Vulnerability_to_flooding.PNG

Figure 1. Vulnerability of EV chargers in Greater London to flooding. (a) Charger utilization of public chargers. (b) Regions at risk from flooding. (c) Fraction of chargers affected by flooding in three flooding scenarios.

Plotting the changes in the two metrics between the baseline case (no flooding) and flood scenarios of increasing intensity, we observe that the impact of the flooding worsens with its intensity as more chargers are taken out of service (see Fig. 2).

Within the regions at risk from flooding, the chargers that become unavailable manifest a reduction in utilization, while the destinations in these areas may also experience significant increase in the distance to the nearest available charger. Outside these regions, both metrics experience overwhelming increases (see Fig. 2(a) and (c)).

To study how the flood impact propagates through the city, we plot in Fig. 2(b) and (d) the changes in the stress metrics as a function of the distance from the nearest flooded grid. Here, in addition to the increased stresses just outside the flooded zones, we find that the stress propagates significantly farther away, peaking some 10-13 km away. These regions can experience a disproportionate increase of over 50% in their charger utilization values and over 250% increase in the distance to the nearest available charger. For each metric, we also observed a strong positive correlation between the baseline values and the subsequent change due to flooding. This indicates that those regions in the city that are already stressed, are stressed even more when flooding occurs.

Figure 2. How flooding impacts battery electric vehicle charging behaviors.

Mitigating the impacts of flooding

 

We studied four different strategies for placing additional EV chargers to reduce the stresses brought about by flooding. These are:

  1. Ring-fencing flooded areas to potentially prevent the propagation of impact into the rest of the city;

  2. Usage-dependent placement, where chargers are preferentially placed in areas of high baseline demand;

  3. Distance-based placement, to specifically mitigate peak stresses in areas farther from the flooded regions;

  4. Random placement across the city.

Broadly, we find that all the strategies increase access to chargers and improve user comfort, observing the reduction in the distance to the nearest available charger (see Fig. 3). Particularly, the usage-dependent and random-placement strategies exhibit the best performance in reducing the flood impacts city-wide on charger utilization and distance to the nearest available charger, respectively. The ring-fencing and distance-based strategies exhibit more targeted improvements, near and farther from the flooded regions respectively.

Figure 3. Exploring four strategies for placing additional chargers to mitigate the impact of flooding on the charging network: ring-fencing flooded areas, usage-dependent placement, distance-based placement, and random placement.

Summary

In our study, we did not find any appreciable impact of flooding on the ability of BEVs to serve typical driving patterns for intra-city travel. In fact, the BEVs retained a ride-success rate of over 99.7% despite over 34% of the public chargers taken offline due to flooding.

We also did not find any chargers currently exhibiting over 100% utilization in Greater London even when flooding occurs. However, this may change when more EVs are added in the future, and if the EV charging network expansion does not keep pace.

With over 30% of the total vehicle stock expected to become electric by 2030, it is key that the charging network be resilient to flooding-related shocks. Otherwise, EV users who may repeatedly face such disruptions could be discouraged from adopting or continuing their use of BEVs.

Arguably, we cannot avoid placing chargers in areas that are at risk from flooding, as this would inconvenience riders under normal circumstances. Therefore, city planners need to explore what combination of the mitigation strategies that we propose suits their city the best.

Supplementary data file

Supplementary Information is available here. The codes used in this study are available for download here.