A fleet depot buildout fails in one of two predictable ways. Either the operator sizes the service and transformer for the fleet they hope to have in five years, sinks the capital before a single van is ordered, and pays interest on capacity that sits idle for years. Or they build for exactly the vans on the lot today, then discover eighteen months later that adding the next ten vehicles means a second trench, a second utility application, and a second round of the same lead times they just finished waiting out. Both mistakes come from the same error: treating the electrical build as one decision instead of two.
The fix is to split it. Some parts of a depot's electrical infrastructure are cheap to size for the end state now and expensive to enlarge later: conduit runs, panel and switchboard space, the transformer pad. Other parts should track actual vehicle arrivals, not a five-year forecast: the chargers themselves, and often the transformer's actual kVA rating. Get that split right and a depot can add vans in phases without a re-trench every time the fleet grows.
Why depots are a different electrical problem than a fixed site
A retail lot or an office park adds chargers to serve drivers who show up on their own schedule. A fleet depot is different in a way that changes the electrical math: the operator controls both the vehicle count and when those vehicles charge, and that count is usually going to grow. A pilot of 5 to 10 vans this year is rarely the plan; it is usually a step toward electrifying 25 percent to 40 percent of the fleet within a few years, then the rest over a longer horizon. An electrical infrastructure assessment done for a fixed retail site can reasonably assume the load it studies is close to the load that gets built. A depot assessment has to study a moving target on purpose, because the whole point of the buildout is that it keeps moving.
That has a direct electrical consequence: depot charging is pushing commercial sites toward the highest power levels the industry uses. Overnight Level 2 charging still covers most light-duty depot fleets running predictable local routes, but medium- and heavy-duty vehicles on tighter schedules are increasingly served by DC fast charging in the 150 kW to 400 kW range so a truck or van can recover most of its range during a shift break rather than overnight (industry coverage from charging-equipment manufacturer Ekoenergetyka and charging-software vendor JointCharging both name depot and fleet charging as the fastest-growing segment of commercial charging in 2026, driven by that shift toward medium- and heavy-duty electrification; treat the growth framing as vendor-reported context, not an independent market study). A single 150 kW port asks more of a service and a transformer than a dozen Level 2 ports combined, which is exactly why the phasing decision matters more for a depot than for almost any other commercial property type.
Decision 1: size the assessment and the application for the end state
Start the electrical infrastructure assessment and the utility service application around the fleet you expect to have at full buildout, not the fleet you are ordering this quarter. Two reasons make this the opposite of over-engineering.
The utility needs both near-term and full-buildout load. A service upgrade or new transformer can require will-serve review, design, pricing, and construction, but the scope and lead time can change materially with the requested load. As of mid-2026 the country is still in a multi-year power and distribution transformer shortage driven by data-center demand and constrained production of grain-oriented electrical steel, with larger units running many months to multiple years and even routine distribution transformers stretching long in some territories (already covered in detail, with sourcing, in the electrical infrastructure assessment article linked above). If you size only for today's 8 vans, a later increase may trigger another capacity study, redesigned equipment, added cost, or a new place in the construction queue. Give the utility the current connected load, the full-buildout load, and the phase timeline so it can tell you which work can be designed once and which must be revisited.
Distribution and conduit work do not shrink to fit a smaller ask, but they do get much more expensive to redo. The same logic that applies to a future-proofed retail installation applies harder at a depot: trenching a parking lot, running conduit to every charging stall, and building out panel or switchboard sections all cost meaningfully more as a second mobilization than as extra scope on the first one. The assessment should therefore answer two separate questions rather than one: what does the depot need today, and what does the site plan call for at full buildout, including stall count and the power level (Level 2 versus DC fast) each stall is meant to carry.
Decision 2: phase the equipment that is cheap to add later
The corresponding rule for the second half of the split: do not buy or energize charger hardware for vehicles you do not have yet. Chargers depreciate, warranties run on a clock from installation, and network and software subscriptions bill whether or not a vehicle is plugged in. The parts that are genuinely expensive to add later, mainly the conduit and the panel or switchboard capacity, should be sized for the end state; the parts that are comparatively cheap to add later, the actual charge points, should track vehicle arrivals.
In practice this usually means:
- Trench and conduit to every planned stall in the first mobilization, even the stalls that will not have a charger for two or three years. Leaving empty conduit in the ground costs a fraction of what it costs to reopen the same trench later.
- Build the panel, switchboard, or switchgear with spare breaker positions and bus capacity for the full stall count, not just the stalls being energized now. This is the same principle covered for any commercial site in switchgear and service equipment, and it matters more at a depot because the phased plan assumes you will come back to the same gear repeatedly.
- Install chargers only for the vehicles arriving in the current phase, leaving the remaining conduit runs capped and the remaining panel positions empty until the next batch of vans is ordered.
- Reassess the transformer at each phase rather than assuming the pad-sized unit from day one. The transformer pad and the primary conductors feeding it are worth sizing for the end state up front, since a pad and primary run are civil work with the same re-trench problem as conduit. The transformer itself, the actual kVA unit sitting on that pad, is a different story: it can often be right-sized to the current phase and upgraded later, since replacing a transformer on an already-built pad with existing primary service is a materially smaller job than the site work that would otherwise need to be redone. Confirm this sequencing with your utility during the service application, since some utilities prefer to set the full-size unit once rather than swap it later.
What a phased depot service application actually says
When you sit down with the utility's service planner, per the workflow in utility service applications and interconnection, a depot's load letter should show both numbers explicitly: the connected load for the phase you are building now, and the full build-out load the site plan anticipates, along with the timeline you expect between phases. That lets the utility evaluate the near-term service request and understand the long-term plan, but it does not reserve future feeder or transformer capacity. The utility may require a new capacity study, application, or construction payment at a later phase, especially if the load, charger mix, or schedule changes. Treat future phases as planned expansions, not pre-approved permit-and-connect work, unless the utility documents otherwise.
Two things commonly complicate this conversation, and it is worth raising both during the earliest planning call rather than after a load letter is submitted.
A distribution capacity study becomes likely once DC fast charging enters the plan. As covered in the interconnection article, a load in the hundreds of kW is far more likely to trigger a utility capacity study than a bank of Level 2 ports. A depot planning to add even a handful of 150 kW to 400 kW chargers in a later phase should expect that study to apply to the full buildout number, not just the phase in front of it, so budget the time for it up front.
Ask specifically about the utility's fleet or make-ready programs before defaulting to a standard line extension. Many utilities run a make-ready or new-business pathway aimed specifically at fleet and depot charging that changes who pays for the transformer and feeder work, similar in spirit to the utility make-ready programs covered elsewhere on this site. Availability, dollar amounts, and eligibility change often and vary by utility and by state, so confirm what is currently open in your territory rather than assuming a program you read about elsewhere still applies.
Load management changes the phasing math, and the code around it is moving
Load management is the single biggest lever for stretching a fixed service across a growing fleet, and it matters even more at a depot than at a retail site because a depot operator controls when vehicles charge. Overnight depot charging on predictable routes is close to the ideal case for a managed system: most vans return within a similar window, need to be full by roughly the same departure time, and do not all need full power simultaneously. Under NEC Article 625's energy management provisions, a listed system can stagger or cap the aggregate load below what the nameplate sum of every charger would otherwise require, which is often the difference between the service you have and the service a naive full-power calculation says you need.
NFPA's official 2026 NEC development materials place EVSE load-management requirements in Section 625.42(A), not Section 625.48. Section 625.42(A) allows a qualifying energy management system to limit the maximum EVSE load used to size the service, feeder, and branch circuit; the same materials describe Section 625.48 as applying to interactive equipment that exports power. Load management can reduce the capacity assigned to charging, but it does not by itself eliminate every continuous-load or overcurrent-sizing requirement. Confirm the final rule in the adopted code book with your electrician and local authority having jurisdiction, because NEC adoption happens state by state and local amendments may apply.
A worked illustration
Take a light-duty depot phasing from 10 vans to a planned 40 vans over three years, all on Level 2 charging at 7.2 kW per port with no DC fast charging in the plan.
- Nameplate load at full buildout: 40 ports at 7.2 kW is 288 kW, before any demand diversity.
- Managed load at full buildout: with a hypothetical energy-management cap of 40 to 60 percent of nameplate, the site would be limited to roughly 115 kW to 173 kW. That percentage is not a general benchmark. It is viable only if route-level energy needs, arrival times, departure deadlines, charging losses, and contingency margin all fit inside the cap. Use a fleet charging model and the NEC load calculation for the actual vehicles before sizing service or transformer capacity.
- What gets sized now: conduit and panel positions for all 40 stalls, and a transformer pad and primary run sized for the managed full-buildout figure above (with margin, since a distribution transformer under-sized for its final load is the more expensive mistake to fix).
- What gets phased: the transformer's actual installed kVA rating and the chargers themselves, added in step with the 10, then 25, then 40 van milestones, so the site never carries charger hardware or a fully-rated transformer for vans that have not arrived.
Swap in your own port power, route patterns, and phase timeline before using any of this for a budget; the shape of the split, not the specific kW figures, is the part that generalizes.
The bottom line
A fleet depot's electrical build is not one sizing decision, it is two, and conflating them is what produces both classic failure modes. Size the conduit, the panel and switchboard capacity, and the transformer pad for the fleet you actually expect to reach, because civil and utility-side work does not get cheaper to redo later. Phase the transformer's installed rating and the charger hardware itself to match real vehicle arrivals, because that equipment is comparatively cheap to add in the next round and expensive to carry idle in this one. Bring both numbers, current phase and full buildout, to the utility in the same application, and ask on day one whether load management or a fleet-specific make-ready pathway changes what the utility will require you to build.
Last factually verified: 2026-08-31 against NFPA's official 2026 NEC development materials, including Section 625.42's EVSE energy-management provisions and Section 625.48's interactive-equipment provisions; 2026 industry coverage of depot and fleet charging as the fastest-growing commercial charging segment and its push toward 150 kW to 400 kW DC fast charging (Ekoenergetyka, JointCharging, both vendor sources); and this site's own previously verified reporting on the 2026 U.S. power and distribution transformer shortage and utility interconnection workflow (see the linked electrical infrastructure assessment and utility service application articles for full sourcing on those points).
evcharginghelp.com is editorially independent and receives no compensation from any company mentioned.