Evaluate an EV charger manufacturer through product-specific safety documents, backend compatibility, measured charging performance, and repair support. For 40 ports rated at 11 kW, connected capacity reaches 440 kW; a 160 kW site allocation allows an average 4 kW when all ports charge together.
Request certificates matching the purchased configuration, port-level availability records, and software-version details. A 99% annual availability claim permits 87.6 hours of downtime per port.
Before ordering, test production-equivalent units with representative vehicles, payment methods, network interruptions, and power-sharing settings. Put acceptance thresholds, spare-parts availability, repair responsibilities, and data-access terms into the contract before approving volume deliveries.
The proposed equipment should first be checked against the site’s daily energy requirement. An illustrative depot with 20 vehicles needing 30 kWh each requires 600 kWh delivered to batteries before departure.
Across eight hours, that equals 75 kW of average battery-side delivery. Assuming 90% overall charging efficiency for planning purposes, average grid demand becomes approximately 83.3 kW. Replace that assumption with measured performance before final sizing.
Arrival times, interrupted sessions, and building demand can increase the required capacity. Give each bidder the same operating schedule so its proposal can be compared against the same conditions.
| Information supplied | Evidence requested |
|---|---|
| Vehicle models and departure times | Charging schedule and compatibility results |
| Available site power | Allocation settings and fallback behavior |
| Outdoor conditions | Temperature limits and supporting test reports |
| Paid or restricted access | Authorization, payment, and recovery workflow |
| Planned expansion | Additional equipment and software requirements |
The proposed configuration determines which safety documents need verification. UL identifies UL 2594 for AC electric vehicle supply equipment and UL 2202 for DC charging equipment among the standards used in its charging infrastructure services. ul.com
Compare the certificate’s model designation with the quotation, rating label, connector arrangement, and electrical ratings. Ask the certification body to clarify coverage where a supplied configuration differs from the documented product.
A factory’s ISO 9001 certificate does not establish product safety approval. Likewise, certification of one component does not establish approval of the complete assembled charger.
Documentation review should cover 100% of the model configurations being purchased. Physical performance assessment then needs to establish what those configurations can deliver.
For a hypothetical DC charger limited to 200 A, output at 400 V is approximately 80 kW. At 800 V, the same current corresponds to 160 kW, provided the charger’s power rating and operating conditions permit it.
Request voltage-current curves, continuous-output records, and temperature-related power reductions. For a two-connector charger, establish whether its advertised rating applies to the shared cabinet or each connector.
Ask for sustained-output records at the project’s specified temperatures, with timestamps, voltage, current, and firmware version. A brief peak reading leaves continuous performance unverified.
Performance records should lead into software testing because authorization and power allocation can prevent otherwise functional equipment from charging.
The Open Charge Alliance states that OCPP 1.6 and OCPP 2.0.1 are not compatible. Record the supported version, software release, and implemented functions rather than accepting “OCPP supported” as a complete specification. openchargealliance.org
Verify claimed certification in the alliance’s published product records. Then connect the proposed units to the intended charging management system and check authorization, meter readings, remote commands, fault reporting, and transaction recovery. openchargealliance.org
A proposed pilot could use four production-equivalent chargers and six representative vehicles. That is a buyer-defined sample, not an industry requirement; increase coverage when the fleet contains different charging architectures.
Record every valid start attempt. If 198 of 200 attempts succeed, the observed start-success rate is 99%. Keep the two failures in the report and identify whether equipment, payment, communications, or the vehicle caused them.
Start success should be assessed alongside availability. Over a 365-day observation period, 99.5% availability permits 43.8 hours of downtime per port; 99.9% permits 8.76 hours.
Neither percentage describes whether charging delivered the required energy. Request separate records for unavailable ports, failed starts, interrupted sessions, and vehicles missing their departure requirement.
| Measure | Calculation basis |
|---|---|
| Port availability | Available service time divided by assessed time |
| Start success | Successful starts divided by valid attempts |
| Session completion | Sessions meeting an agreed completion definition |
| Restoration time | Time from the defined incident start to verified recovery |
Use consistent exclusions across suppliers. Report utility interruptions, planned maintenance, and site damage separately, with reasons and durations. That allows the operating figures to be compared without removing inconvenient incidents.
Recovery testing should then examine what happens when communications fail. Disconnect the internet, restart the local controller, interrupt meter communication, and restore power using procedures approved by the electrical team.
For the 160 kW charging allocation, confirm that all active units remain within the agreed site limit during normal operation and communication failures. Document any permitted tolerance and response interval with the electrical designer.
Repeat each selected recovery scenario three times as a proposed acceptance procedure. Check whether transaction records return correctly, charging settings persist, and staff can identify the interruption.
Payment testing should follow the same recovery scenarios. Check declined payments, authorization without charging, interrupted sessions, duplicate requests, refunds, and receipt delivery.
For an illustrative 100 paid sessions, reconcile charger energy records, backend transactions, and payment settlement. Investigate every unexplained difference; apply metering tolerances appropriate to the equipment and destination market.
Operational access also needs review. Require individual administrator accounts, controlled permissions, encrypted communications, firmware authenticity checks, and a stated security-support period.
During the pilot, interrupt one firmware update under an approved procedure. Record whether the charger returns to service, needs local assistance, or requires component replacement.
Software recovery depends partly on the supplied hardware, so compare pilot units with production records. Obtain serial numbers, hardware revisions, firmware releases, final inspection records, and a list of approved components.
Require written notice before changes to meters, communication devices, connectors, power modules, or cooling components. Ask which changes require renewed certification review or integration testing.
Apply serial-number identification to 100% of delivered units. For physical inspection, agree a sampling plan based on shipment size, product risk, and previous quality results rather than inspecting an arbitrary percentage.
Manufacturing consistency still leaves repair arrangements to verify. Ask who diagnoses faults, stocks parts, sends technicians, and confirms restoration at each location.
DOE guidance notes that charging equipment requires maintenance and may need repairs, while warranty arrangements vary by manufacturer. Compare parts, labor, travel, freight, and on-site work separately. afdc.energy.gov
A proposed service agreement might require acknowledgment within two hours, diagnosis within four hours, and restoration within 48 hours for covered faults. Those are negotiating examples, not standard manufacturer performance.
Acknowledgment time and restoration time must be separate commitments. Ask for previous service records showing both, along with spare-part availability and geographic coverage.
Service terms should feed into a common five-year cost comparison. Include equipment, installation, commissioning, connectivity, software, maintenance, parts, and integration. DOE also identifies permitting and utility upgrades as infrastructure cost considerations. Alternative Fuels Data Center
At an assumed $6 per port per month, software for 40 ports totals $14,400 over five years. At an assumed $12, the total reaches $28,800. Use quoted fees and stated renewal terms in the purchase assessment.
Before volume approval, require 100% closure of agreed blocking defects. Attach the accepted hardware configuration, software release, test results, service terms, and change-notification requirements to the order, with payment stages linked to delivery and verified acceptance.