Impact of Revel fast charging hub on NYC EVs

NOTE: all the work here was done by PhD candidate Chloe Bingqing Liu. I'm just the messenger.

Last week, I mentioned that we had developed a tool with NYC DCAS as part of our NYU VIP "NYC Clean Fleet" project last summer that could quantify the accessibility of electric vehicle charging infrastructure location configuration for a population, and could use it to analyze the impact of Revel's recent investment in the largest fast charging hub in the U.S. The tool is freely available on our BUILT lab's Github portal. Details of the methodology and application to DCAS charging stations and EV fleet are provided in a working paper in arXiv. The methodology assumes that any vehicle using a public charging station has three types of cost: there is the (1) travel time from EV home location to the charging station, the (2) queue delay waiting for charger availability (the first two combined is the "Access Time"), and (3) the charging time (which we assume to be mostly deterministic; if DCFC it's 1/2 hr to fully recharge and for Lvl2 we assume 4 hrs).

As promised, here are the results of our analysis.

Based on the information from the news, there will be a new hub with 30 DC fast chargers (DCFCs) at the Pfizer Building in Williamsburg. We assume that the public EV population will include non-Tesla users (who have their own proprietary charging infrastructure). From NYSERDA's site, we have the charging infrastructure in NYC as follows (only considering Level 2 and DCFCs):

The system without the 30 DCFC Revel chargers will be called the BASE SCENARIO, which beforehand only had 15 (non-Tesla) DCFCs located over 13 stations prior to Revel (which is a substantial difference). For the BASE SCENARIO EV population, NYSERDA provides the citywide total (from July 2020, this was ~ 6603 not counting Teslas). We assumed the spatial distribution of this citywide population is similar to the spatial distribution of the NYC EV fleet of 1484 vehicles that were shared to us from DCAS, and simply scaled up the numbers to the 6603 total. We also assume the same calibration of 7.6% average utilization ratio across the city (based on data from NYSERDA in 2018), leading to an average recharge trip rate of 1 every 3 days. We then apply our model to the EV population without and with the hub included. The results are shown below.

The high level observation is that access time in general increases while total time decreases. This is because the added availability of much faster charging stations leads to many EVs switching from low-demand, slow charging Level 2 chargers to high-demand (with high queue delay/access time) DCFCs at the Revel hub. But overall, people would save 21 minutes per EV recharge trip (8.3% improvement in system performance for the current EV population). We can visualize this graphically as well:

This graphic compares the total time (access time plus charging time) for each of the zones in NYC in which we scaled up NYC EV fleets to population level.

This second graphic compares the utilization ratio (how busy each station is on average). Revel's hub should substantially modify the charging patterns of the current EV population. We can look at the benefit of 21 minute savings another way. The average total time per EV (access plus charging time) without the hub is 252 minutes. With the added Revel hub, how much can the EV population in NYC grow before the average cost returns to the no-hub base scenario? In other words, how much growth might Revel's hub accommodate? We incrementally increased the total EV population at different rates and found that the hub could accommodate up to 62.7% growth in the non-Tesla EV population in NYC.

If the locations of the charging hubs were elsewhere, or if there was a sudden growth in the EV population in another part of NYC, then we would have to use the tool to recompute these results. As fleets start adopting as well (buses, taxis, maybe even trucks later?) the city might have to start thinking about "charging infrastructure management" in the same way they have to think about "curb management".