EV Charging

A DC fast charging site draws enormous power in short bursts and sits idle the rest of the time. That is the worst possible shape of load to buy from a utility, and the best possible case for a battery.

Why fast charging breaks the normal economics

Four 200 kW chargers is 800 kW of nameplate demand. Your utility sizes the service, the transformer and the bill around that number, whether you hit it twice a day or never.

Two things follow, and both hurt.

Demand charges dominate the bill. Commercial tariffs bill on the highest 15-minute peak in the month. At a representative North American rate of around $31/kW, a single 800 kW peak sets a five-figure monthly charge before you have sold a kilowatt-hour of energy. At low utilization, on a new site running ten sessions a day, that fixed cost is spread across very few charging sessions, and the site loses money on paper for years.

The service upgrade is the schedule. An 800 kVA service is a different conversation with the utility than a 300 kVA one: bigger transformer, longer lead time, sometimes primary work and a cost contribution. Many sites die at this step, not at the equipment step.

What battery buffering changes

A battery sits between the grid and the chargers. The grid trickles in at a modest, constant rate; the battery delivers the burst when a vehicle plugs in, then recharges slowly.

In one published four-charger analysis, the grid-only design required an 800+ kVA transformer to serve 4 × 200 kW, while the battery-buffered design served the same chargers behind a 300 kVA service using a 379 kWh battery and 250 kW of inverters. At ten sessions a day, the modelled utility cost difference exceeded $700,000 over ten years.

Those are US figures on a US tariff and we are not going to pretend they are your numbers. The mechanism is what transfers:

• Smaller service. Often the difference between a straightforward connection and a year-long utility project.
• Lower demand charges. You are billed on the grid-side peak, which the battery flattens.
• Faster build. Smaller transformers are easier to procure, which pulls the revenue date forward.
• More chargers per site. You can add dispensers without re-opening the service capacity question.

Sizing a buffered charging hub

Three inputs decide the design, and they are not the ones most people start with.

Session profile, not charger count

What matters is how many vehicles arrive in the busiest hour and how much energy each takes, not how many cables you have. A site with eight dispensers and four simultaneous sessions is a four-session site.

Recovery window

The battery has to refill between bursts. A site with steady all-day traffic needs more grid import than a site with a sharp commuter peak and a quiet afternoon. This is what sets the grid-side service, and it is where oversimplified sizing goes wrong.

Worst-case simultaneity

Design for the realistic bad day, a holiday weekend or a nearby site offline, not the average. The failure mode of an undersized buffer is a queue of drivers watching a charger throttle, which costs you the customer permanently.

For heavy-duty and transit depots the arithmetic shifts again: the loads are larger but far more predictable, because the vehicles are yours and the schedule is known. Depot charging is usually the easiest storage business case in this whole category.

Stacking the second and third revenue stream

A battery bought to avoid a service upgrade is idle most of the day. In Ontario that idle asset can also reduce your Global Adjustment exposure by discharging through the province’s five annual peak hours. The Ontario peak shaving page sets out how that is calculated.

A charging hub that shaves its own demand charge and its Global Adjustment allocation is running two business cases on one asset. That is usually what moves a marginal site into a fundable one.

Safety and permitting

Charging hubs are typically outdoor, public-access sites, which makes the fire-safety review more visible than it would be behind a fence at an industrial plant.

NFPA 855 is referenced in most stationary storage reviews, but it does not automatically govern a Canadian project. Which requirements apply, and which edition, comes from the adopted provincial and municipal codes, the insurer, the project specification and the authority having jurisdiction. Confirm that basis before the site plan is fixed. Then plan for three things: a site-specific hazard mitigation analysis, fire-test reports that name the standard edition, test level and sample configuration, and emergency response planning coordinated with the local fire service, with owners and dates in the contract.

Public siting also raises the practical questions early: setbacks, bollards, signage, and what the fire service expects to find when they arrive. Better to have those answers before the site plan goes in.

How BX approaches charging hub projects

BX Energy Systems supplies containerized LFP systems sized for charging hubs and depots across Canada, documented for Canadian AHJ review. If you have a site that stalled on a service upgrade quote, that is exactly the case worth modelling.

Book a 15-minute sizing call

Common questions

How big a battery does a fast charging site need?

Typically enough energy to cover the busiest hour of sessions plus a margin, with inverter power matched to peak simultaneous charging. For a four-dispenser 200 kW site that often lands in the few-hundred-kWh range, but the recovery window between bursts moves that number more than the charger count does.

Can we avoid the service upgrade entirely?

Often you can reduce it substantially rather than eliminate it. The battery still has to recharge, so there is a floor on grid import set by your daily throughput. The saving comes from sizing the service to average energy rather than instantaneous power.

Does this work for transit or truck depots?

Better than for public charging, usually. Known vehicles, known dwell times and a fixed schedule mean the buffer can be sized tightly, and overnight recharge is cheap.

What happens if the battery is empty when a driver arrives?

The site falls back to grid-limited charging, which is slower rather than offline. Good design sets that floor deliberately rather than discovering it on a long weekend.