How to Evaluate Energy Storage System Manufacturers: An 8-Step Checklist from a Quality Manager

If you're shortlisting suppliers for a BESS project — whether you're stocking a distributor catalog or developing a 50MWh site — this checklist is for you. I run the same 8 steps every time, and it usually eliminates 60–70% of candidates before we get to pricing.

Quick background: I'm a quality and brand compliance manager at a renewable energy distribution company. I review somewhere between 40 and 60 supplier deliveries a year. In 2024, I rejected 18% of first deliveries for incoming inspection misses — every one of which was preventable at the PO stage. (Should mention: the rejections weren't about failures. They were about documentation and spec drift.)

Who This Checklist Is For (and Who It Isn't)

This applies to B2B buyers, wholesale distributors, EPCs, and project developers evaluating lithium battery suppliers for commercial or utility-scale BESS. It's not for evaluating consumer products, and it's overkill for sub-20kWh single-unit purchases. But if you're signing six-figure POs, this is the process.

Step 1: Write Your Own Spec Sheet First

The first move isn't to open the vendor's datasheet. It's to write down what you actually need. Capacity in kWh. Power in kW. C-rate. Operating temperature range. Cycle life. And the installation standard the system has to satisfy.

One thing most buyers skip: state the level at which you will parallel-expand, and the level at which you won't. A lot of lithium battery specifications arguments late in a project trace back to a mismatch — the supplier was expanding at the rack level, the owner was planning at the PCS level. Same numbers, completely different architectures.

(I learned this one the hard way. It's now item #1 in our procurement template.)

Step 2: Trace Every Cell From Origin to Enclosure

If a supplier won't disclose cell origin, that's a pass. Full stop. Cell sourcing is where most BESS performance and safety issues live — you're entitled to know where they come from.

Ask for:

• Cell manufacturer name and factory location
• Cell model and IEC 62619 certification number
• Batch/lot number and manufacturing date
• Traceability documentation from cell to module to rack

Here's something suppliers won't tell you: many manufacturers run different production lines for sample builds and volume builds. The specs are the same, the consistency isn't. If you don't trace per batch, you won't catch it.

Step 3: Demand UL 9540A Reports — and Read Them at the Unit Level

UL 9540A is the test methodology for thermal runaway propagation in energy storage systems. It's what NFPA 855 and most local fire codes reference for installation approval.

Set a filter for yourself: only accept unit-level and module-level test reports. System-level reports tell you what the enclosure survived. Unit-level reports tell you how the propagation was contained and what mitigation was active. If a report doesn't list the testing lab, the test date, and the exact product model, ask for the full report.

Per NFPA 855 (2023 edition), energy storage systems must be listed and installed per their UL 9540 listing, with UL 9540A test data supporting the installation location.

(Side note: some suppliers conflate UL 9540 with UL 9540A. They aren't the same thing. UL 9540 is the certification of the system. UL 9540A is the test method. Don't let the terminology pass.)

Step 4: Unpack the Cycle Life Claim

A supplier says 8,000 cycles and 10 years. That's a good headline. Now ask:

• At what C-rate was it tested? (0.5C and 1C are not the same number.)
• At what temperature? (25°C vs 40°C is a different test.)
• At what depth of discharge?
• To what retained capacity — 80%? 70%?
• Cell-level or system-level data?

If a number is cell-level and being presented as system-level, it's not comparable. We now standardize on IEC 62933 system-level cycle life and require every claim to be mapped back to the test conditions. If they can't map it, we don't count it.

Self-correction: a few suppliers do publish third-party cell test data and use it honestly, labeling it as cell-level. That's fine. The problem is when the label falls off by the time it reaches your RFP response.

Step 5: Audit a Production Line, Not Just a Sample

Samples are curated. Production is not. The only way to catch drift is to go see it.

When I visit a factory, I look for:

• End-of-line (EOL) test data — is it 100% tested, or batch-sampled?
• Torque verification records for module assembly
• Any MES-style traceability log linking cell batch to finished unit
• Rework rate for the most recent production run

If a supplier can produce Powin BESS-level factory audit documentation, half your job is done — several public filings and technical documents list the audit requirements their systems meet, which you can use as a template. For most mid-size lithium battery suppliers, you'll have to bring your own auditor. Budget for it.

Step 6: Translate the Warranty Into Cost

A warranty is both a sales tool and a line item. Until you unbundle it, it's just words on a page.

Get these in writing, per system:

• What's covered — capacity degradation, outright failure, or both?
• What threshold triggers a claim?
• Who pays freight during the warranty window?
• What date does the clock start — ship date, delivery date, or commissioning date?

The worst clause I've seen: a 10-year warranty with the clock starting on ship date. If the systems sit in your warehouse for 18 months, your warranty is really 8.5 years. That changes the math on your resale pricing.

Step 7: Test Compatibility With Your Own Inverter and PCS

This is the step most buyers skip. It's also where a surprising number of projects lose weeks.

Confirm at minimum:

• Protocol support — Modbus TCP, CAN, DNP3, or vendor-proprietary
• Communication test script with your EMS/SCADA platform
• Firmware update policy — how often, and how issues get patched

On a 22-unit project we ran into a firmware polling mismatch that kept everything offline for six weeks. Since then, compatibility sits before the PO, not after.

Step 8: Evaluate Whether the Supplier Outlives the Warranty

This isn't electrochemical, but it belongs on the list. A warranty from a company that won't be around in year eight is worth roughly zero.

Look at:

• Public filings, if listed
• Concentration of key component suppliers
• Utilization of manufacturing facilities
• Ownership structure — including any recent acquisitions or restructuring

(Adding here: a restructuring or ownership change isn't automatically a red flag. But you need to confirm which legal entity stands behind the warranty when a claim is filed. In large cloud/utility projects this is sometimes a separate SPV — check.)

Financial due diligence isn't a box-ticking exercise. It's the answer to "who pays when half the racks underperform in year six."

Common Mistakes Before You Sign

1. Letting the acceptance criteria live in an email. Put kWh/kVA tolerance limits, EOL tests, and incoming inspection protocol into the contract itself.
2. Relying on sample testing as the quality signal. A sample proves capability. It doesn't prove consistency.
3. Starting the warranty at ship date. Ask for either delivery or commissioning. This alone can cost you a year of coverage.
4. Skipping the first-article plant visit. Yes, it's a flight. Yes, it's worth it. I want to say the last one saved us six figures, but don't quote me on the exact figure — it was enough that we institutionalized the practice.

Everything I'd read going into this said supplier evaluation was mostly about price and lead time. In practice, on the systems that actually matter, price and lead time finished third and fourth behind documentation quality and process consistency. Five minutes of verification beats five days of correction — every time.