Tradeoffs in Remote Dispenser EVSE Architectures for EV Fleets

For electric vehicle (EV) fleet operators, especially those operating medium- and heavy-duty (M/HD) EVs, EV Support Equipment (EVSE) approaches using remote dispensers (those where the dispenser is physically separate from the AC/DC conversion and power conditioning electronics) are extremely popular. Remote dispensers allow fleet operators to maximize vehicle density in vehicle lots by minimizing the EVSE footprint near the vehicles, while clustering the Power Control Systems (PCSs - the AC/DC conversion and power conditioning electronics) near the electric power inflow location at vehicle lot. However, there are a number of different architectures for these deployments, affecting the ratio of dispensers to PCSs, and the way that the dispensers and PCSs are connected to each other.

Why Remote Dispensers are Popular

For fleets of any sort, maximizing the number of vehicles that can be fit into a vehicle yard is critical. This is especially true for medium- and heavy-duty (M/HD) vehicles such as public transit buses, school buses, and municipal vehicles. These yards often have one hundred vehicles spread over a couple of acres. It is typical to see these vehicles parked end-to-end with little more than a foot of space on the sides between them to maximize parking density. Integrated high-power AC/DC dispensers have fairly large footprints (think the size of a commercial freezer), and putting an integrated charger next to M/HD vehicles significantly impacts vehicle parking density. This is especially true when you considering the need to protect the integrated chargers with bollards or similar barriers. In contrast, the typical remote dispenser has a footprint of roughly 1-2 ft² (with many less than this), allowing them to be placed in the space on the side of the buses. This eliminates any impact on vehicle lot parking density.

Remote Dispenser Architectures

There are typically three approaches that are used for remote dispenser deployments:

• Dedicated PCS per Dispenser Approach: Also known as the "1:1" approach, as each dispenser is paired with a dedicated PCS, which is typically located at the entry of the utility grid power feed into the vehicle yard.
• Parallel Dispensers Approach: This approach powers multiple remote dispensers from a single PCS, with a dedicated connection (e.g., high-voltage DC power feed) from the PCS to each dispenser.
• Serial Dispensers Approach: This approach also powers multiple remote dispensers from a single PCS, but utilizes only one high-voltage DC power feed to connect the PCS and all of the dispensers (the feed runs from the PCS to the first dispenser, and then runs serially to each remaining dispenser). 

The primary advantage of the serial approach over the parallel approach is that it significantly reduces the number of trenches that need to be dug and cable that needs to be laid during installation. At prices that can reach $10,000-$20,000 per 1000 feet of trenching, there can be significant cost savings from the serial approach.

In both parallel and serial dispenser approaches today, only a single remote dispenser is generally powered at a time. In the parallel case, this is typically accomplished through "1 to X" high-voltage DC switching circuitry, which is either incorporated into the PCS or is located next to it ("X" being the number of remote dispensers per PCS). In the serial case, a two-way high-voltage DC switch is incorporated into each dispenser - it either switches the power from the PCS to the dispenser port (and into the vehicle), or sends it downstream to the next dispenser. While it is possible to have multiple remote dispensers powered at the same time from a single PCS, this requires isolation circuitry to be incorporated into each dispenser to isolate the vehicles from each other. If the battery voltages of the vehicles are different, the dispensers would also have to have DC-DC voltage stepdown capabilities. Both of these features would significantly increase the cost of the dispensers, and the entire system would still be limited by the power capacity of the PCS. 

Which of these approaches makes sense for your EVSE depends on your use case. For situations where the amount of charge that each vehicle requires approaches the total power the PCS can output during the vehicle charging window, a dedicated approach is best - the savings achieved by buying fewer PCSs will likely be offset by the impact of possibly not charging all of your vehicles completely. On the other hand, if the amount of charge required by each vehicle is significantly less than the amount of power that your PCS can output during the charging window (say 33 percent or less), then a multiple dispensers per PCS approach probably makes sense. Let's look at two examples to illustrate when each approach makes more sense:

• A 2-acre vehicle yard must support 100 electric school buses, each of which has a 150kWh battery. Each bus typically uses 50kWh of power on its morning run and 50kWh of power on its afternoon run. The charging window for the buses is 5 hours long between the morning and afternoon runs (9am-2pm), and 9 hours long between the afternoon run and the morning run (9pm-6am, with no charging during the 4pm-9pm peak hours). Each bus has 14 hours to replace 100kWh of energy, requiring an inflow of roughly 7kW each. In this case, one 60kW charger could easily support four or more buses, reducing the number of PCSs required by 75 percent. Note that some of this savings would be consumed by switching gear either in/near that PCS (parallel case) or in the dispensers (serial case). If serial dispensers were used, the trenching cost reduction could be on the order of $250K-$500K vs the parallel approach (eliminating 1000' of cable/trenches per PCS times 25 PCSs).
• The same 2-acre vehicle yard must support 100 long-range electric transit buses, each of which has a 600kWh battery. Each bus typically uses 540kWh of power (90 percent of capacity) on its daily route. The charging window for the buses is 9 hours long (9pm-6am, with no charging during the 4pm-9pm peak hours). Each bus has 14 hours to replace 540kWh of energy, requiring an inflow of 60kW each. In this case, one 60kW charger could only support a single bus, so a dedicated approach would make sense. While a parallel or serial approach could make sense with a much larger PCS (say at least 180kW), this would require very expensive switching gear and (in the case of the serial approach) cable that could handle the amperage. 

Of course, most vehicle yards are likely to have a variety of vehicles, each with different energy usage profiles. It is entirely possible that a single vehicle yard could use multiple architectures to maximize the EVSE investment. Weighing the tradeoffs for your particular situation is the best way to make the most of your time and money.

Mike Heumann is VP of Marketing, and Joseph Gottlieb is CTO at Rhombus Energy Solutions. Rhombus Energy Solutions designs and develops products for electric vehicle DC fast charging, electrical energy conversion, and energy management system software.

Rhombus Energy Solutions | rhombusenergysolutions.com