EV Load Management and Smart Charging for Depot Peak Demand

What EV load management is at a depot (and is not)

EV load management is the set of rules and controls that decide how much charging power the yard draws at once so vehicles still hit departure state-of-charge without blowing site supply, demand charges, or transformer headroom. It sits between the charger hardware and the facility kW/kVA limit — the same problem fleet software vendors label smart charging software or charge management, and the same problem a depot planner must size before layout freeze.

It is not a residential wallbox timer, and it is not public retail session optimisation. Depot EV load management optimises a known fleet, fixed or semi-fixed bays, overnight or dwell windows, and a commercial three-phase supply with DNSP connection rules. If the policy cannot be enforced on the yard floor, the spreadsheet peak savings will not show up on the bill.

Uncontrolled vs smart vs peak-limited charging

Uncontrolled charging starts every vehicle at full available power when it plugs in. That is simple operations and creates a sharp evening peak when the fleet returns — often the worst night for demand charges and the first night a site limit is breached.

Smart charging spreads energy across the available window so vehicles still reach target state-of-charge by departure while fewer chargers run at full power simultaneously. Peak-limited strategies add an explicit site supply cap (kW or kVA band) so the depot does not exceed a nominated limit while still prioritising readiness.

In practice most Australian bus and truck yards evaluate all three on the same scoreboard: readiness by pull-out, P50/P90/P95 peak demand (kW), headroom vs existing supply, and indicative energy plus demand-related cost. EV load management is choosing — and enforcing — the strategy that keeps those numbers honest as the fleet grows.

Why P95 peak demand matters more than a single average curve

Network discussions, demand charges, and transformer headroom care about high but plausible peaks. A single average load profile hides the nights when late returns stack and many vehicles still need a deep charge.

Reporting P50, P90, and P95 peaks makes the risk visible. A load-management policy that looks fine on average can still breach a site limit one night in twenty — which is often unacceptable for operations and for early DNSP conversations in areas such as Ausgrid, Endeavour Energy, Essential Energy, Energex, CitiPower/Powercor, and SA Power Networks.

Design to high-percentile readiness and peak together. Average-night EV load management under-protects the connection case; worst-night-only design over-builds hardware that smart charging may never need.

Smart charging software vs facility power decisions

Teams often split EV load management into a software SKU and an electrical drawing set. That split fails when the product promise assumes a dwell window operations cannot give, or when the switchboard limit assumes a policy the CSMS never enforces.

Treat smart charging software as one control layer inside a larger plan: charger count and kW class, bay concurrency, overnight window, site supply (kVA/kW), and who owns exceptions when a vehicle returns late. The planning question is not "which vendor has load management" alone — it is whether any chosen policy keeps P95 peak inside the band you can afford while every bus or truck is ready at pull-out.

Pair software evaluation with quantitative depot simulation. If uncontrolled and smart strategies are not compared on the same fleet assumptions, vendor feature lists stay disconnected from demand charges and DNSP lead times.

Demand charges, opex, and grid connection

On many Australian commercial tariffs, demand charges ($/kW/month) and time-of-use energy rates reward lower peaks and more off-peak energy. Smart charging and broader EV load management are not only an infrastructure story; they are an annual opex story.

Lower simulated peaks can also keep a depot inside a cheaper DNSP connection band or defer augmentation while the fleet ramps. That is why charger strategy, load management policy, and grid strategy should be modelled together — not sequenced as vehicle → chargers → "we will figure out the peak later".

Use high-percentile load profiles as planning evidence in early network talks. Simulation is not a formal connection offer; engineers and the DNSP still own applications, protection, and works.

Trade-offs to freeze before you buy bays

1) Window length vs peak — longer overnight dwell usually allows lower concurrent kW for the same energy; short turnaround forces higher power or more bays.

2) Charger count vs kW class — more lower-power ports can finish the same energy with a flatter profile; fewer high-power bays raise per-session speed and often P95 peak if unmanaged.

3) Uncontrolled simplicity vs managed peak — uncontrolled is easier on day one and expensive on demand charges and connection risk; managed EV load management needs yard discipline and clear exception rules.

4) Capex now vs network later — buying extra supply headroom without testing smart or peak-limited strategies can lock in years of unused capacity; relying on software without readiness checks can strand vehicles on high-energy nights.

Run the free simulator before you lock load management policy

On chargingcapacity.io, pick a Small, Medium, or Large bus depot preset (or set truck-oriented assumptions), enter overnight hours and charger kW, then compare uncontrolled vs smart or peak-limited charging. Read readiness and P50/P90/P95 peak demand, export a load profile, and use it when operations, finance, and the network argue about the same night.

That pass turns EV load management from a slide feature into a shared quantitative baseline. Re-run when fleet size, block plans, or charger class change — and pair this guide with depot charging, charger sizing, fleet electrification planning, charging cost, and EV depot grid connection so strategy and facility stay linked.

Try it on your depot assumptions

Open the free simulator, pick a depot preset, and compare uncontrolled vs smart charging before you lock charger counts or a grid story.

Run the EV depot charging simulator →

FAQ