-
Who this checklist is for
-
Step 1: Lock the specs on paper before you talk price
-
Step 2: Verify cell-level traceability
-
Step 3: Review the BMS, not just the cells
-
Step 4: Visit the factory — the testing floor, not the assembly line
-
Step 5: Ask for raw test data, not certificates
-
Step 6: Match certifications to your actual use case
-
Step 7: Sample from the shipment, not the sample cabinet
-
Step 8: Define the after-sales terms in writing
-
Common mistakes
Who this checklist is for
If you're sourcing lithium battery cells in bulk, or specifying a BESS for a distribution channel or an integrated project, this list is for you. Eight steps. Same order, every time.
I'm not here to tell you which brand to buy. I work in quality and brand compliance at Gotion Inc., and I review roughly 200+ bulk shipments a year before they go out the door. Some come in clean. Some get bounced back to the line. The difference is rarely just price.
Here's the process we've been running since a 2022 incident I'd rather not repeat.
Step 1: Lock the specs on paper before you talk price
Most buyers start with unit cost. Understandable. Also backwards.
Write down first: cell chemistry, nominal capacity and voltage, usable capacity window, cycle life test conditions, operating temperature range, BMS communication protocol, enclosure IP rating.
If you don't nail the test conditions, your supplier's "cheaper" quote isn't cheaper — it's different. Two 280Ah cells, one rated at 0.5C continuous discharge, one at 0.25C. Price gap of 15%. Not the same product.
I learned this the hard way. In 2021 we took two pallets of cells from a new supplier, spec sheet looked clean. Three weeks later, field testing showed the capacity retention didn't hold. Reading the fine print, the test condition was 0.1C discharge at 25°C. Our actual use case was nothing like that.
Now every contract includes the test method itself, not just the number.
Step 2: Verify cell-level traceability
Ask the question: who actually made this cell?
Not who assembled the pack. Not who relabeled the brand. Who produced the cell?
"We use tier-1 cells" is not an answer. It's a category. You want batch-level or cell-level traceability — to the factory, the production date, ideally the specific Gotion lithium battery production line if that's the source.
Suppliers who can't produce that traceability get vague. The vagueness is the signal.
Step 3: Review the BMS, not just the cells
This is where most people stop paying attention.
The cells get all the attention. The BMS is the quiet part. It decides how the system behaves in the field. Ask: what firmware version is running? What's the communication protocol — CAN, RS485, Modbus? What's the process if firmware needs updating? How does the BMS handle cell-level imbalance?
"Standard BMS" is not a spec. It's a request to stop asking.
Step 4: Visit the factory — the testing floor, not the assembly line
Everyone gets walked past the assembly line. It's clean, it's orderly, and it tells you almost nothing about what leaves the building.
Ask to see the testing floor. Ask: how does this plant sample cells for inbound QC? How often are batches pulled? What happens when a batch fails?
A testing floor that looks in-use — logs being written, equipment running, staff who can answer on the spot — is its own data point. A quiet testing floor means either they haven't started, or the testing is theater.
Step 5: Ask for raw test data, not certificates
Certificates tell you a sample passed at some point. They don't tell you about this batch.
Ask for BMS logs, factory test records, cell-level test data on the actual shipment. When you get it, look at the pattern, not the peak.
Now — I don't have hard data on industry-wide defect rates across bulk energy storage system suppliers. What I can say from our own inbound QC over roughly five years of orders is that somewhere around 8–12% of first deliveries on new supplier relationships come with issues that need sorting. Some serious, some cosmetic. Wide range, because our categories are wide.
The point stands: if you're only looking at a certificate, you're only looking at the part the supplier chose to show you.
Step 6: Match certifications to your actual use case
Certifications matter. But not every certificate applies to every project. You want to know the mapping, not just the alphabet soup.
- IEC 62619 — safety requirements for secondary lithium cells and batteries for industrial applications. Table stakes for many stationary storage projects.
- UL 1973 — batteries for stationary, vehicle auxiliary, and light rail applications. Hard to skip for grid-tied storage in the US.
- UL 9540A — the test methodology for evaluating thermal runaway propagation in BESS. If you're putting a system inside a building, this may be the gatekeeping document for fire review.
- UN 38.3 — transportation testing for lithium batteries. If you move cells across borders, you need this on every shipment.
One more thing: check the certificate expiry and the production date range it covers. A certificate from 2021 saying "this product line passed" tells you nothing about a cell produced in 2025 after a supplier change.
Step 7: Sample from the shipment, not the sample cabinet
"Sample cabinet" is a trap. The unit you're shown is not the unit you'll get. This is an open secret in the industry, and it still catches people.
What I do: pull random units from already-bagged production. Not the ones the factory hands you. Photograph them, seal them, log them, and attach the log to the contract as an addendum.
I forgot this step once. The batch that followed didn't match what we'd approved — different housing material, different terminal finish. Redoing it cost us several weeks. Several weeks on a project timeline is often the whole window.
Step 8: Define the after-sales terms in writing
This is the step most teams push to post-signature. Too late by then.
Agree upfront on: RMA process. Spare parts availability and pricing. What triggers a warranty claim — is it a capacity fade threshold, a specific failure mode, or a field incident? Get the trigger definition in writing.
On cycle life: when a datasheet says 6,000 cycles at 80% retention, that's a specification. It's a number you check against data during acceptance testing. It is not a promise. Specifications live in writing. Promises live in sales conversations and evaporate when you need them.
Draw that line clearly before you draw the purchase order.
Common mistakes
Don't shop on the price sheet alone. I've watched buyers switch suppliers over a 3% unit-cost delta and then eat far more than 3% in returns, rework, and schedule slip. That's not theoretical. That's our own purchasing record. The savings don't cover the freight on returns plus the downtime.
"One-stop-shop" sounds great. It usually means every stage is done at 70%. A vendor who focuses on cells is not the same as a vendor who also bolts together packs, does BMS, and offers consulting. I'd rather work with a specialist who knows the edges of their lane than a generalist who promises every lane.
When a supplier says "that part isn't our strength — here's who does it better," that usually earns a point from me. Boundaries named upfront let me decide what to rely on and what to source elsewhere. That's useful information. "We do everything" is not.
Define what you won't buy before you define what you will. True for suppliers. Also true for your own spec. The tighter your requirements, the fewer places a vendor can substitute something else without telling you.
Last thing. Data without a source is noise. If someone quotes you an "industry-average cycle life" or "industry defect rate" — first question is: what's the source, what's the sample size, what's the test condition? If any of those three can't be answered on the spot, the number isn't a number. It's a vibe.
Build a cost guide from verified line items, not vibes. It's the only version that survives the next procurement cycle.
