Load Management Basics for EV Depot Charging

Before I joined the energy software side of this industry, I spent several years running fleet operations for a mid-size logistics company. When the decision came down to transition a portion of the fleet to EVs, my first instinct was to treat it like a fuel logistics problem: work out how much energy the fleet needs, buy enough capacity to deliver it, and make sure everything is in place by the time the first vehicles arrive. What I got wrong was assuming that more capacity automatically meant more efficient delivery of energy. It does not. It often means exactly the opposite.

Load management exists because the timing of when electricity is drawn matters as much as the total amount drawn. A depot that needs 2,000 kWh per night for its fleet can deliver that over 12 hours at an average of 167 kW, or it can attempt to deliver half of it in the first two hours at 500 kW and the rest overnight at much lower rates. The first approach works within a 200 kVA grid connection. The second approach does not, and if you try it, either your circuit protection trips or your demand charges become punishing.

What Load Management Actually Does

At its core, load management for EV charging does three things. It monitors the aggregate power draw from all active charging sessions in real time. It maintains that aggregate below a configured maximum, which is typically set to match available grid connection headroom after accounting for other site loads. And it distributes that budget across individual chargers by adjusting their maximum current setpoints.

The mechanism for adjusting individual charger current is OCPP, the Open Charge Point Protocol. When a load management system sends a SetChargingProfile command over OCPP to a charger, it specifies a maximum current limit in amperes per phase. The charger enforces that limit regardless of what the connected vehicle's onboard charger would accept. This allows the load management system to, for example, tell charger bay 12 to operate at a maximum of 8 A per phase even though the charger is capable of 16 A and the van would happily accept it.

At the aggregate level, the load manager continuously sums the active power draws from all chargers and adjusts setpoints to keep the total below the configured threshold. When a new vehicle plugs in and starts drawing current, it does not immediately get full power. It receives an initial allocation that keeps the system total within bounds, and its allocation increases as other vehicles finish charging and their sessions end.

Static Load Management Versus Dynamic Optimization

There are two broad approaches to load management for fleet charging. Understanding the difference between them helps clarify why the simpler approach, while adequate in some cases, leaves significant money on the table.

Static load management divides the available power budget equally across all connected chargers and does not change allocations during the night. If you have 150 kW of headroom and 30 vehicles connected, each gets 5 kW. Simple, predictable, and reasonably effective at preventing the worst overload scenarios. The problem is that equal distribution is rarely the right distribution. Some vehicles need less charge because they returned with a higher SoC. Some vehicles depart earlier and need priority. And the best time to charge, from a tariff perspective, is typically not the time when you divide capacity equally across all 30 vehicles.

Dynamic optimization goes further. It incorporates departure times, vehicle SoC, tariff pricing by hour, and real-time grid conditions to continuously recompute optimal current allocations. Vehicles with early departures and low SoC get higher allocations. Vehicles with late departures and already-high SoC get lower allocations or have charging paused entirely during peak tariff hours. When a vehicle finishes charging and disconnects, its budget is immediately redistributed to vehicles still in progress.

The practical difference in electricity cost between these two approaches is not trivial. Static management prevents peak demand charges but does not shift charging toward cheaper hours. Dynamic optimization does both. In the Netherlands, where daytime and evening electricity prices in a typical time-of-use commercial contract can be 25 to 40 percent higher than off-peak rates between 23:00 and 07:00, the tariff optimization component alone can reduce energy costs significantly for a fleet that previously just started charging when vans plugged in after the evening shift.

The Role of the Grid Connection Limit

The grid connection maximum is the hard boundary that load management respects. Set it correctly and the system will never allow aggregate demand to breach it. Set it incorrectly and you either waste available capacity or risk exceeding the connection threshold and triggering demand penalties.

For a typical Dutch commercial depot, the grid connection capacity is specified in your transport contract with the DSO as a kVA value. The load manager needs a kW value, not kVA. If you assume a power factor of 0.95 (reasonable for EV charging loads, which are predominantly resistive and capacitive through the charger's power electronics), converting 200 kVA gives you approximately 190 kW. Subtract whatever your non-EV site loads draw during overnight hours, typically 30 to 60 kW for lighting, refrigeration, and monitoring systems, and your effective charging budget is 130 to 160 kW.

Most operators we have worked with set their load management threshold at 90 to 95 percent of that available charging budget, not 100 percent. That margin provides buffer for measurement uncertainty, charger current overshoot during ramp-up phases, and brief spikes when multiple vehicles start sessions within the same minute. The cost of that conservatism is roughly 5 to 10 percent of theoretical charging throughput. The benefit is never accidentally triggering a demand charge.

What Happens When a Charger or Connector Fails

Load management systems need to handle hardware failures gracefully. If a charger loses its OCPP connection to the management system and continues to charge at full power, the aggregate load calculation becomes inaccurate. The management system no longer knows what that charger is consuming and cannot account for it in its budget.

Good charger hardware defaults to a safe current limit if it loses contact with the management system. This is sometimes called "failsafe mode" or a configurable standalone charge rate. The management system, for its part, should detect the lost OCPP connection and either alert the operator or reduce allocations to other chargers to compensate for the unknown load contribution from the disconnected unit.

This is one of the reasons we emphasize thorough OCPP testing during charger commissioning. A charger that claims OCPP 1.6 compliance but fails to implement the ChargingProfile functionality correctly will accept SetChargingProfile commands and acknowledge them, but may not actually enforce the specified limits. The only way to confirm correct behavior is to send a current limit command and verify at the circuit level that the charger is drawing the specified amount.

Monitoring and Visibility

Load management without visibility into what is actually happening is a system you cannot trust. The minimum useful dashboard shows the current aggregate load from all chargers against the configured maximum, the SoC and charging status of each connected vehicle, and any chargers that have lost connectivity or are in a fault state.

More useful is a view that shows whether each vehicle is on track to reach its target SoC by its departure time, given current and projected charging rates. This lets fleet managers spot problems early, before they result in a vehicle leaving short of charge. If van number 8 is at 52 percent SoC at 03:00 and needs to be at 90 percent by 05:30, the system should flag that as an issue and automatically increase that vehicle's priority, but the manager should also be able to see it and override if needed.

We built our monitoring layer specifically around this departure readiness view because it is the question that fleet managers actually need answered every morning: will every van be ready to go? Everything else in the system is in service of getting that answer to green.

Getting Started Without Overhauling Your Infrastructure

Fleet operators who are in the early stages of electrification often ask whether they need to solve load management completely before adding their first EVs. The answer is no. A small initial fleet of 5 to 10 vehicles rarely creates load management problems, because the aggregate charge demand is low enough to fit comfortably within any standard commercial grid connection.

What matters is that the chargers you install from the beginning support OCPP properly, and that you have a load management system in place before you reach 15 to 20 vehicles or before your fleet represents more than 30 to 40 percent of your grid connection's headroom. At that point, simultaneous charging during evening peak hours will start to create visible impacts on your electricity bill, and the cost of retrofitting load management intelligence is much lower when the infrastructure was designed for it from the start.

Plan for it early, install OCPP-capable hardware throughout, and treat the load management software as infrastructure rather than an optional add-on. The operators who make that decision at the outset spend considerably less on their grid connection and their electricity contract than those who discover the need for it two years and 30 vehicles later.