Fleet Charging Service Levels: Prove Departure Readiness

Fleet charging service levels should measure whether the agreed vehicles have enough verified energy at their departure deadlines—not merely whether chargers are online. For a Vietnam factory moving staff transport, service vehicles or delivery operations to electric power, a charger purchase and an operating-service commitment answer different questions. The owner needs an auditable definition of success, a feasible operating envelope and a process for investigating missed departures.

This guide develops an owner-side service schedule for a return-to-base fleet. It is not a standard contract, a legal opinion, a charger recommendation or an electrical design. Every numerical example is hypothetical; no actual customer, fleet, tariff, warranty or QuangAnhcons project result is implied. The physical depot connection, protection and civil design remain separate work packages.

Prepared for QuangAnhcons · English FDI owner and procurement perspective · Primary sources checked: 17 October 2026.

1. Define the unit of service as one vehicle departure

Start with a controlled departure register. Each row identifies a pseudonymous vehicle, route or duty group, arrival and plug-in deadlines, required departure time, energy or state-of-charge target, evidence source and responsible dispatcher. Freeze the reporting cohort before the service period. Otherwise, changing the denominator after an incident can turn an operational failure into an apparently successful month.

Distinguish three times: the vehicle entering the depot, becoming physically available at a suitable bay, and being connected and authorised for charging. A vehicle parked behind another vehicle is not necessarily chargeable. Record who moves vehicles, who inserts the connector, who provides the next route and who confirms a target change. For a leased site, also identify who can restrict power during factory production.

The target must have a defensible meaning. A battery state-of-charge percentage is a vehicle estimate; its relationship to usable energy changes with vehicle condition and operating limits. Energy recorded at an EVSE meter is not automatically energy stored in the battery. Require the fleet and vehicle supplier to agree the appropriate readiness evidence for the exact vehicles. Do not invent a universal conversion between charger kWh, battery percentage and route kilometres.

The US Department of Energy’s smart-charge implementation guidance identifies operating schedules, dwell periods, fleet information and electrical capacity as relevant inputs. Its US federal context is not Vietnamese law or a local incentive entitlement. The service register and metrics below are original procurement proposals derived from the owner decision, not requirements quoted from that guidance.

2. Report departure readiness, charging access and equipment availability separately

A backend can show a charger as available while a particular vehicle cannot complete its route. Conversely, one unavailable connector may have no departure impact when another compatible bay has enough time and energy. Three separate measures prevent an equipment dashboard from being mistaken for an operating result.

For example, assume ten eligible departures: eight have valid records showing the target met, one has valid records showing an energy shortfall, and one lacks trustworthy departure telemetry. The verified-ready rate is 8 / 10 = 80%. The measured-not-ready share is 10%, and the unverified share is 10%. This does not establish that the unverified vehicle physically failed; it establishes that readiness was not demonstrated. Report all three categories rather than publishing a misleading 100% success rate from the eight known successes alone.

Commercial attribution can use a separate, pre-agreed calculation. If one missed target is demonstrably outside the supplier’s responsibility, retain it in the gross operating outcome and show the adjustment transparently. The engineering report should not decide liability, service credits or damages. Those require the actual agreement and legal review; no commercial threshold or penalty amount is prescribed here.

3. An adequate overnight total can still miss the morning departures

Hypothetical deadline-feasibility example: four vehicles depart at 06:00. Vehicles A and B are chargeable from 22:00 and each need 54 kWh added at the battery-energy boundary. Vehicles C and D become chargeable only at 04:00 and each need 36 kWh. Assume a constant aggregate 30 kW AC charging-input cap, a simplified constant AC-input-to-battery efficiency of 0.90, sufficient compatible connectors, no auxiliary demand, no taper and no interruption. These are illustrative assumptions, not design values or OEM performance claims.

Total required battery energy is 2 × 54 + 2 × 36 = 180 kWh. At the assumed efficiency, required AC energy is 180 / 0.90 = 200 kWh. The eight-hour AC envelope is 30 × 8 = 240 kWh. A total-energy-only screen therefore appears comfortable: required average input is 25 kW, below the 30 kW cap.

However, C and D cannot use the energy available before they arrive. Their combined 72 kWh battery requirement has only two hours remaining. Even if A and B are already complete and the full cap serves C and D, the maximum battery energy in that interval is 30 × 2 × 0.90 = 54 kWh. Their combined shortfall is at least 72 − 54 = 18 kWh. The ideal minimum AC cap for this late cohort alone would be 72 / (2 × 0.90) = 40 kW, or the ideal minimum charging window at 30 kW would be 72 / (30 × 0.90) = 2.667 h, approximately 2 hours 40 minutes.

These are lower bounds, not a proposed connection increase. The owner must change an actual constraint—earlier availability, a qualified different energy requirement, another feasible charging opportunity, or an independently approved supply arrangement. A scheduling algorithm cannot create time or energy. Raising a software limit without approved upstream capacity is not a solution. Vehicle limits, temperature, taper, occupied bays and auxiliaries can only make this simplified screen less optimistic.

A useful general necessary check is: for any group whose charging cannot start before time t and whose deadline is d, required battery energy must not exceed the integral of usable battery-side charging power between those times. Passing that check alone does not prove a feasible allocation when vehicle, connector and overlapping-window constraints remain. Request a time-resolved schedule and inspect the critical cohorts, not just one daily kWh sum.

Repeat the screen for explicit degraded configurations. If an independently specified shared-cabinet failure reduces the hypothetical aggregate AC cap to 15 kW throughout the late window, the available battery energy falls to 27 kWh and the shortfall becomes 45 kWh. This is an assumed failure boundary, not a statement about any real cabinet architecture. Two sockets on one failed power stage are not two independent charging paths.

4. Make the service measurable without collecting unnecessary personal data

The minimum event record should join the departure plan, vehicle identity, connector/cabinet identity, plug-in and authorisation events, energy readings with units and boundary, power-limit changes, alarms, target revisions and the final readiness observation. Record the timezone, clock source, sampling interval and data-quality flags. Preserve original events and corrections with a revision trail; do not overwrite the underlying record to make a monthly report agree.

Define which source is authoritative when vehicle telematics, charger meter and charging-management system disagree. Agree tolerances and reconciliation rules from the actual equipment and measurement purpose before testing. A dashboard estimate is not automatically a settlement-grade energy reading or evidence of battery energy. A transaction ending after midnight must not be assigned to a different departure cohort merely because a billing report groups by calendar day.

The Open Charge Alliance describes OCPP as the communication protocol between charging stations and their management systems, with distinct versions including 1.6, 2.0.1 and 2.1. Protocol support is relevant to integration, but the procurement inference is narrower: require demonstrated features on the actual EVSE, firmware and backend combination. A version label alone does not prove that a particular vehicle supplies usable state-of-charge data, that every event is retained, or that a departure target is achieved.

Ask for a sample export with documented field names, units, missing-value behaviour and identifiers. Keep driver names, route details, credentials and customer delivery information out of general contractor reports unless actually necessary and authorised. The owner should set access roles, retention and permitted cross-border/cloud handling with its privacy and IT teams. This article makes no finding that a particular cloud or data transfer is legally approved.

5. Give each exception an owner and a visible consequence

Maintain a causal exception register, not an unlimited list of exclusions. For each event, record the affected departures, start and end times, evidence, constraint, mitigation, responsible party and unresolved disagreement. Distinguish late vehicle arrival from late plug-in, failed authorisation from electrical loss, and an imposed facility cap from an unexplained charging derate.

Agree what happens when the factory requests a temporary power restriction. Does the fleet dispatcher accept a revised target, move a vehicle, arrange another charging location or declare an operating shortfall? The decision and time must be recorded. An unapproved reduction cannot quietly become a new contractual baseline, and an energy-management command must not override protection or electrical safety limits.

A generic response-time promise is also insufficient. Separate acknowledgement, remote diagnosis, safe access, arrival of a competent technician, availability of a compatible spare, restoration and confirmation that the next departure remains achievable. A quick acknowledgement can coexist with a missed route. Link response reporting to the affected operating window without implying that every failure can be repaired within it.

Use the factory electrical interface register to allocate supply, controls and third-party dependencies. For the service schedule, add a named fleet decision owner; the electrical contractor cannot independently change route requirements or certify a vehicle’s roadworthiness.

6. Accept an evidenced operating service, not just a successful charging session

Installation verification and electrical release must precede an authorised operating trial. The owner’s test plan should identify the representative vehicles, software baseline, target data, approved power envelope, pass criteria, witnesses, stop conditions and evidence to retain. The owner’s electrical inspection and test plan can control the physical readiness evidence; it does not replace the fleet-service test.

Use simulations or vendor-supported test modes where fault injection would be unsafe or disruptive. Do not open energised enclosures, defeat interlocks, disconnect protective conductors or force protection trips to create a service demonstration. Any physical maintenance or isolation requires the actual site’s approved permit, lockout/tagout, stored-energy controls and competent authorised personnel. A remote stop command is not an isolation procedure.

Require retesting after a material vehicle, firmware, backend, dispatch-policy or power-envelope change. Keep a controlled acceptance baseline and distinguish acceptance of one tested configuration from a guarantee for every future fleet expansion. The factory operational-readiness checklist helps connect that baseline to custody, operating authority and controlled handover.

7. Technical standards do not supply a universal fleet SLA

IEC 61851-1:2017, Edition 3.0, with Corrigendum 1:2023, addresses general conductive EV supply-equipment requirements within its stated voltage scope. IEC 60364-7-722:2018, Edition 2.0, addresses installation circuits supplying EVs or receiving energy from them, terminating at the connecting point. The official catalogue scopes and editions were checked; licensed clauses and project test limits were not reproduced.

Neither catalogue scope establishes a fleet’s required departure success percentage, compensation formula or a specific product’s certification. Ask the responsible designer for the project-controlled IEC and TCVN editions, the exact OEM instructions and applicable Vietnamese QCVN/legal requirements, including current transitions and the actual installation/use classification. The official VSQI search result identifies TCVN 13078-1:2020 as the national general-requirements reference for conductive EV charging. Direct retrieval of that register page failed during this review; no full-text, clause-level or mandatory-applicability conclusion is drawn. Resolve the controlled national reference before project release.

For the upstream electrical network, include an applicability review of Circular 51/2025/TT-BCT issuing QCVN 26:2025/BCT, electrical technical requirements for power networks. The Government Gazette identifies its effective date as 1 June 2026. This is a network-regulation reference for the responsible project team to scope, not a fleet-service standard, a complete permitting checklist or evidence that a depot is compliant. No clearance, protection setting or operating procedure is taken from unread clauses.

No vehicle, EVSE or controller model is selected here, so no manufacturer charge curve, allowable test procedure, efficiency, warranty or compatibility certificate is invented. Bidders must supply model-specific manuals and evidence. Standards, serving-entity connection conditions, electrical design acceptance and the operating-service agreement belong in separate linked registers.

8. Ask bidders to return the same operating evidence

Issue the anonymised departure register, qualified vehicle requirements, approved supply envelope, access assumptions and reporting rules together. Require bidders to return their feasibility model, critical-window exceptions, exact supported hardware/software configuration, data export sample, test plan, support dependencies and priced inclusions/exclusions. Do not compare offers solely on charger kW or a headline availability percentage with different denominators.

Before committing, the owner should be able to answer: which departures are promised; what proves readiness; which party supplies each input; what happens when the envelope becomes infeasible; how unknown data are reported; who authorises a change; and how the service can be audited or transferred to another provider. The service schedule should remain usable even when the dashboard supplier changes.

For an owner-side discussion, send QuangAnhcons an anonymised arrival/departure pattern, energy-boundary explanation, proposed electrical cap and the unresolved interface list through the existing contact page. Keep private vehicle, driver and commercial records in an approved exchange route. The first useful output is a qualified scope and evidence gap register—not an unsupported price, uptime guarantee or promise that nameplate power will make every vehicle ready.

[contact-form-7 id="7239967" title="Contact form 1"]