Sizing a battery-buffered station
The expensive part of a fast-charging site is rarely the charger. It is the grid connection: the transformer, the trenching, the utility study, the demand charges and the eighteen-month queue. A battery behind the dispenser lets you serve a 350 kW session from a service that could never deliver 350 kW, by buffering energy between vehicles.
The question that decides the design is simple: how many charges does one full battery deliver?
Work it in energy, not power
Start with what is actually usable:
usable kWh = size x stack x depth of dischargeA 250 kWh cabinet at 90% depth of discharge holds 225 kWh. If you start at 100% and stop at a 10% reserve floor, the session pool is 202.5 kWh. Reserve floors exist because deep discharge costs cycle life and because a station that hits zero cannot serve the car already plugged in.
Then subtract what each vehicle costs you. Two efficiencies stand between the cells and the pack:
energy from the cells = pack kWh / vehicle-side losses / discharge efficiency
= 52.5 / 0.94 / 0.94
= 59.4 kWhfor a Model 3 taking 52.5 kWh from 10 to 80%. So 202.5 / 59.4 gives three full charges, with 24 kWh left over, which takes a fourth vehicle from 10% to 39% before the floor stops it. That is the headline number in section 2 of the calculator.
The inverter changes the clock, not the count
Here is the result that surprises people. Run the same 250 kWh station with a 125 kW output limit and then with a 500 kW output limit:
| Output limit | Full charges | Session time |
|---|---|---|
| 125 kW | 3 | 28 min |
| 500 kW | 3 | 24 min |
Identical charge counts. The output limit governs how fast each session goes, not how much energy the battery holds. If your problem is throughput per hour, buy inverter. If your problem is vehicles per visit, buy cells. They are different purchases and it is worth knowing which one you have.
Grid assist changes the question entirely
Turn on grid assist and the service feeds the dispenser alongside the battery, so available power becomes:
min(plug rating, battery output + grid supply)Now add a gap between vehicles, during which the battery recharges. If the energy recovered in the gap matches what a session consumes, the station stops being buffer-limited and becomes sustained: it runs indefinitely on the grid, and the battery is there for burst power rather than total energy. In the calculator, a 250 kWh station with a 65 kW service and 30-minute gaps stops draining altogether.
That is the real design target for most sites. You are not trying to store a day’s energy. You are trying to turn a small continuous connection into a large intermittent one.
Where the output limit does bind
If the battery’s output ceiling sits below the car’s curve, the session simply takes longer and the calculator names the limit as BESS output limit. This is not a failure mode, it is a design choice: a 0.5C inverter on a 250 kWh cabinet gives 125 kW, which is fine for a fleet of commuter cars arriving at 20% and leaving at 80%, and inadequate for a truck.
The default in the calculator is 0.5C of the selected size, because that is a common commercial ratio, and it is adjustable because the right answer depends entirely on the vehicles you expect. Which is the point of the tool: put your actual fleet in the queue and read the count.
Once you know the count, the other half of the problem is how long the battery takes to refill, which is the service math.
Next
Root three and the 80% rule
The two bits of electrical arithmetic that decide what a 480 V service can actually deliver, and the moment when a bigger breaker stops buying anything.