Battery Cell Inspection: Before Shipment and After the Container Lands
A pack can pass a visual inspection and still be built from mismatched cells. This page covers the cell-level checks that predict field failures, and what to measure in the first 72 hours after arrival if you want a claim to hold.
Why Cell Level Is Where the Risk Sits
Pack-level inspection tells you the unit works today. It does not tell you how it will behave in month six.
A LiFePO4 pack is only as good as its weakest cell. Grading and matching happen at cell level, before assembly — which means a pack built from cells of different capacity or internal resistance will pass a functional test on the day it is made and drift out of balance later. The BMS will mask it for a while by shunting current, then it stops masking it.
This is why the two inspection timings answer different questions. Pre-shipment asks “was this built properly”. Post-shipment asks “did it arrive as the thing that was approved”. Neither one substitutes for the other, and a post-shipment claim with no pre-shipment record is a complaint rather than evidence.
What Each Timing Can and Cannot Catch
| Timing | What it can establish | What it cannot |
|---|---|---|
| Pre-shipment at the factory | Cell grade actually used, internal resistance spread inside each pack, weld quality, every BMS protection function, measured capacity, packaging and documents | How the pack behaves after weeks in a container. Self-discharge and transit damage only show up later. |
| Post-shipment on arrival | Transit damage, open-circuit voltage drift against the shipping record, capacity shortfall on a sample, fault history stored by the BMS | It cannot restore capacity that was never there, and without a pre-shipment baseline it cannot prove the shortfall happened in transit. |
If you only get one of the two: take pre-shipment. Arrival testing without a baseline tells you something is wrong but not when it went wrong, and the cost of fixing it after the freight is paid sits on your side of the invoice.
Four Cell Tests That Predict Field Failures
Each needs an instrument. None can be replaced by looking at the unit.
| Test | How it is done | Common pass criteria | What a fail means |
|---|---|---|---|
| Capacity grading | Full discharge at 0.2C on a cell tester, measured Wh against nameplate | Within ±5% of nameplate; cells within a pack grouped to a narrow band | Mixed capacity means the lowest cell sets the pack ceiling, and balance drift starts early |
| Internal resistance matching | AC internal resistance meter at 1kHz across every cell in the pack | Spread under 10% inside a pack; tighter for high-current packs | Uneven resistance means uneven heat under load — the hot cell ages fastest |
| OCV and self-discharge | Record open-circuit voltage, rest, re-measure after a set interval, compare cells against each other | Cells tracking together; no single cell falling away from the group | One cell dropping faster than its neighbours suggests a soft short or contamination — the classic slow field failure |
| Weld and mechanical | Pull test on tabs and busbars, visual check of weld penetration, X-ray where the order justifies it | No cold welds, consistent penetration, no loose tab | A cold weld becomes resistance and heat in service, not a failed test on the line |
The criteria above are the ranges used in our own checks; they are standard practice rather than a standard. What matters is that the criteria are written down before the batch runs, so “pass” means the same thing to you and to the line.
The First 72 Hours After the Container Lands
Six steps, in this order. Skipping step one is what loses claims.
1. Photograph before you open. Container seals, then pallet and carton condition, before any carton is cut. Damage found after the cartons are open is very hard to attribute.
2. Check by pallet, not by unit. Crush and drop damage clusters: one bad stack in a container usually means the whole pallet is suspect, and the units at the bottom take the load.
3. Measure OCV on every unit and write it down. This takes an hour and gives you a baseline you can compare against the pre-shipment sheet. A unit that reads noticeably lower than its neighbours is the first thing to pull.
4. Run a capacity test on a sample. At minimum one unit per pallet, at 0.2C, with the discharge rate recorded. A capacity number without the rate is not comparable to the factory's number.
5. Read the BMS fault history where the model exposes it through an app or PC tool. Over-current and over-temperature events during transit leave a trace, and a long container journey in summer is exactly when they happen.
6. File the claim against the record, not the impression. Photos, measured values, sample size. A supplier can act on “unit 7 of pallet 2 measured 41% low at 0.2C”. There is nothing to act on in “the batteries seem weak”.
BESS and Module-Level Checks
Rack and container systems need three extra measurements on arrival.
For a rack or container system, add insulation resistance, cell block voltage spread and a torque check on the busbar connections — transit vibration loosens what the factory tightened. Read the state of charge on arrival against the shipping SOC: a module that arrived well below its shipping SOC has been self-discharging, and that is a cell-level question, not a logistics one.
Pull the thermal log if the system stores one. Temperature excursions inside a container are often the first visible sign that something in the pack is not behaving, well before a capacity test shows it.
What Post-Shipment Inspection Cannot Do
Three limits worth knowing before you plan around it. It cannot create capacity that was never in the cells — a shortfall measured on arrival was almost certainly there at loading, and the arrival test only proves it. It cannot replace pre-shipment control: a claim after landing costs you the freight in both directions and the time, which is usually more than the units are worth. And it cannot expose a design problem: a pack with inadequate thermal design tests fine on arrival and starts failing in month three, long after the inspection window closed.
Related: Full inspection checklist · Jobsite power · RV & campervan power
Questions Buyers Ask
What is the difference between pre-shipment and post-shipment inspection?
Pre-shipment happens at the factory before the goods are packed, and it is where you can still reject a lot and have it rebuilt. Post-shipment happens after the container arrives and establishes whether what arrived matches what was approved. Only the first one can prevent a bad batch from shipping; the second one documents it after the freight is already paid.
Can a cell problem be found after the container arrives?
Partly. Arrival testing finds cells that have drifted away from the group, transit damage and capacity shortfall on a sample. It will not find a cell that was marginal when it left but is still inside tolerance, and it cannot distinguish “bad from the start” from “damaged in transit” unless you have the pre-shipment measurements to compare against.
How is cell capacity graded?
By a full charge followed by a controlled discharge at a defined rate — 0.2C is the usual reference — with the measured watt-hours compared to nameplate. A tolerance of about 5 percent is common practice. Grading also means grouping cells of similar measured capacity into the same pack, because mixing them is what causes early imbalance.
What internal resistance spread is acceptable?
Under 10 percent spread within a pack is the common working limit, and high-current packs are usually held tighter. Spread matters more than the absolute value: two packs can both read normally on average and still behave very differently if one has a single cell at twice the resistance of its neighbours.
Do I need to test every unit?
No. Use a sampling plan — AQL General Inspection Level II with Major 2.5 is the usual starting point for power stations — and treat safety functions as zero-tolerance regardless of the sampling number. For cell-level checks, the samples come from the same production batch, so a bad batch shows up in a sample almost immediately.
Ask for the cell-level record
Tell me the model and the market, and I will send the inspection record format we use per batch — the measurements that get written down, and the documents that travel with the goods.