Depot charging, whole-home energy, grid flexibility and vehicle-to-grid, in service.

What the pilot has delivered

• Three data sources reconciled into one depot view: telematics, settlement-grade half-hourly metering and charger sessions, with no site works.
• A daily charging plan for every vehicle, worked back from its next departure and held inside the site’s import capacity, compared each day with what actually happened.
• An evidence pack on a tamper-evident audit trail from day one, so the measured phase starts from a baseline the council has agreed.

The challenge

Fleet electrification turns the depot’s grid connection into the limiting asset. Charging that starts when vehicles return lands on the evening peak and uses most of the available capacity, and any plan that risks a missed round is not a saving. The council needed evidence it could take through procurement before it changed anything on site.

The challenge

Vans return around 17:30 and plug in together. Charging lands on the evening peak at the highest prices and takes most of the depot’s connection, so adding vehicles looks like it needs a bigger supply and a long wait for the network operator.

The approach

Each charge is planned back from the van’s 07:00 departure, inside the connection limit, and moved into the cheapest overnight hours. Readiness is a hard limit in the engine: a van that needs charge gets it, and the schedule flexes around it.

What we proved

• A utility-to-charger control chain over an open standard, with no proprietary link between the utility server and the charging equipment.
• Every control tracked through each stage on its own: received, acknowledged, executed and reported, so the evidence holds up at each step.
• Status and metering reported back to the utility server and accepted in full.

The challenge

Most vehicle-to-grid projects prove the control chain with one vendor’s equipment end to end. Utilities planning programmes at scale need the same chain to work across manufacturers, with evidence they can audit, before they can trust it.

What we delivered

• A constrained optimisation that turns the day’s 96 prices into one schedule, allowing for charge taper and the driver’s departure requirement.
• Exports in the forms operators use: setpoints, a full 96-interval file and a JSON audit record, so every setpoint traces back to a price.
• Charging in the cheapest intervals and discharging into the dearest, presented to the Task 53 participants at ElaadNL.

Why it matters

A tariff that changes every 15 minutes is only useful to a fleet if the resulting schedule can be explained to the people who run it. The same explainability is what AmpVerve’s monthly statements rest on: each instruction traceable to the input that caused it.

What the pilot delivered

• One half-hourly picture of the house from two independent sources, the supplier’s meter feed and the charger’s own telemetry, reconciled interval by interval.
• A control gap in the charger’s load-matching worth about £244 a year, invisible in the charger’s app, documented and taken to the manufacturer.
• A re-timed charge and discharge schedule that keeps the van ready for its departures, and a measured ceiling for further gains: up to £286 a year more with half-hourly closed-loop control.

The challenge

A house with solar, a heat pump, three vehicles and a bidirectional charger has four schedules and four apps, none of which can see the others. The owner wanted to know whether the charger’s discharge window, the tariff windows and the household’s real load lined up, and what a better plan was actually worth before anything changed.

What the pilot delivered

• One whole-house view across five manufacturers, each previously visible only in its own app.
• A single daily plan: solar into the battery and the vehicle first, the heat pump kept inside the owner’s comfort limits, grid import moved into the cheapest hours, export held inside the connection’s limit.
• A weekly statement in the AmpVerve app showing what was done and what it was worth, against the supplier’s own half-hourly data.

The challenge

The battery, the charger and the heat pump each optimise alone, and the vehicle charges whenever it is plugged in. Four controllers pulling in different directions on one 100 A supply means the cheapest hours are shared badly and the export limit is reached while the vehicle still needs charge.