How to Verify Battery Quality: A 5-Step Checklist from a Quality Manager Who Rejects Deliverables

If you’ve ever ordered lithium-ion cells for an ESS or EV project, you already know the sinking feeling of opening a batch that looks fine but behaves poorly. I’ve reviewed over 400 unique battery shipments in the last four years alone — Samsung SDI cells, CATL prismatics, LFP pouch packs — and I’ve rejected about 12% of first deliveries in 2023 due to things that wouldn’t show up on a spec sheet. Here’s the checklist I actually use. It’s not theory. It’s what I walk through before I approve anything.

When to Use This Checklist

This checklist is for anyone responsible for incoming quality inspection on lithium-based batteries — whether you are a procurement engineer, warehouse manager, or site lead for a utility-scale BESS installation. Use it when you receive a new lot from any supplier, including tier-1 manufacturers like Samsung SDI. It’s designed for intermediate to high-volume orders (500 to 50,000+ units) where a problem in 2% of cells can cost you weeks of field troubleshooting.

Step 1: Visual & Dimensional Check — Look for the Stuff That’s Wrong Before You Test

Honestly, so many problems are visible from two feet away if you train your eyes. I check every pallet — not a sample, but every outer container — for:

• Deformation or swelling: Even slight bulging on a prismatic cell means internal pressure issues. I once received a stack of Samsung SDI NMC cells where 3% had visible dome on the casing. Normal tolerance? Zero. We rejected the entire batch.
• Connector damage: For ESS modules with busbars or wire harnesses, look for bent pins or scratched insulation. I’ve seen a $12,000 power cart fail because a connector housing had a hairline crack.
• Label mismatch: Compare the date code, voltage, and capacity printed on the cell to your PO. It sounds obvious, but I caught a batch where a supplier had packed generation-1 cells into a generation-2 box. The cells inside were fine, but the paperwork was wrong. That’s a traceability nightmare.

Take photos of every opened box. I’ve learned this the hard way: if you find an issue later, having that timestamped image saves the argument.

Step 2: Open-Circuit Voltage (OCV) Sampling — But Don’t Sample the Way You Think

Standard incoming inspection says to measure OCV on 2% of units. That’s fine for production verification. For quality acceptance? I think you need to go heavier. Here’s what I do: measure OCV on 5% of the total lot, but I pull those samples from the first, middle, and last pallet of the shipment. The conventional wisdom is that cells from the same production batch are homogeneous. My experience with 200+ orders suggests that vibration during shipping can disturb connections in a way that looks like a cell voltage drift.

What you need to know: for a 48V ESS module, OCV should be within ±10 mV of the rated value across all cells. If you see more than 15 mV spread in the samples, flag the entire pallet for full inspection. I’ve seen a 30 mV mismatch turn into a balancing issue after 200 cycles on a UPS battery bank.

Step 3: Internal Resistance (IR) Measurement — The Metric Everyone Ignores Until It’s Too Late

This is the one most people skip. They check voltage, maybe capacity, but they don’t check DC internal resistance. I didn’t fully understand why IR mattered until a vendor failure in March 2022. We had cells that passed OCV and looked brand new, but after 50 cycles in an EV pack, the internal resistance had climbed by 35%. The pack overheated during fast charging. That investigation cost us $22,000 in testing and delayed our customer launch.

Measure IR on the same 5% sample you used for OCV. Use a dedicated AC milliohm meter or a DC pulse tester. For a Samsung SDI 94Ah NMC cell, typical fresh IR is around 0.5–0.8 mΩ. If any cell reads above 1.2 mΩ, set it aside. If more than 2% of your sample exceeds the threshold, request a full IR report from the vendor for the entire lot.

Step 4: Capacity Spot-Check — But Only If You Have the Right Equipment

Capacity testing every cell in a 10-MWh BESS is impossible unless you have a dedicated cycler. But you should spot-check at least one module per pallet. Here’s the tricky part: people assume that because a cell is from a tier-1 supplier, the capacity is guaranteed. The reality is that even premium manufacturers have variation. I saw a lot of Samsung SDI solid-state battery prototypes in 2023 that all read within 0.5% of nominal capacity — impressive. But for standard NMC cells, a 2% spread at 0.5C discharge is normal.

If your spot-check finds a module delivering 5% below rated capacity, reject that pallet. Everything I’d read about solid-state production said yields are lower, and that’s fine. But for commercial shipments, capacity guarantees are real obligations. The $200 savings on accepting a slightly low module turns into a $1,500 problem when your system underperforms and you owe penalties.

Step 5: Documentation & Traceability Audit — The Step That Feels Like Overhead Until It Saves You

I run a blind test with my team every quarter: same battery module with full documentation vs. minimal documentation. 87% identified the well-documented module as “professional” without knowing the difference. On a 10,000-unit ESS deployment, that perception matters for your brand reputation.

Check the following against your contract:

• MSDS / SDS sheets for the specific chemistry
• UN38.3 test summary (required for lithium battery transport)
• Cell and module test reports (date, tester ID, results)
• Batch and serial number traceability to the cell level

If a vendor can’t provide a UN38.3 summary with the exact transport classification, do not accept the shipment. It’s illegal to ship without it in many jurisdictions. I rejected a shipment in Q1 2024 because the paperwork listed “UN3480” (cell level) when we needed “UN3481” (battery pack). The vendor claimed it was within industry standard. It wasn’t. We redid the entire lot at their cost.

Common Mistakes and What to Watch For

1. Assuming “new” means “charged.” I’ve received cells at 10% SOC when the spec required 30%. This messes up your initial formation cycle if you’re assembling packs. Always measure before you store.

2. Only checking one cell per module. Module-level testing hides individual cell drift. If you measure module OCV but don’t open it up, you might miss a cell that’s shorted internally.

3. Ignoring storage conditions on site. If you store lithium batteries outside spec (temperature 15–25°C, humidity < 50%), you degrade them before they are commissioned. I’ve seen 8,000 units ruined in a warehouse with passive temperature control during a heatwave.

Bottom line: treat every shipment like it’s the one that could fail. A checklist like this isn’t bureaucracy — it’s the difference between a project that runs and one that eats your margin in rework fees. Take it from someone who’s rejected hundreds of thousands of dollars in batteries: the cheap inspection saves you little. The thorough one saves you everything.