-
What this checklist is for, and who it's written for
-
Step 1: Verify document versions, not just document existence
-
Step 2: Read the nameplate, and don't mix nameplate with measured
-
Step 3: Confirm BMS communication against the actual inverter firmware
-
Step 4: Sample-test capacity—but record the test conditions
-
Step 5: The 12V auxiliary circuit and quick disconnect check
-
Step 6: Community solar projects need grid-tie and fire-code documents in the same packet
-
Step 7: The quote transparency review—ask what's not included before you ask the price
-
Common errors and things to watch
What this checklist is for, and who it's written for
I review incoming battery storage shipments for a system integrator. Roughly 15 to 20 shipments a month, maybe 200 a year—I'd have to check the log, the exact number isn't something I keep in my head. First-pass rejections run somewhere around 14 to 18%. More like 15%, if I'm honest, but the log is the source of truth on that.
Most of those rejections are not cell defects. They're label errors, stale documentation, modules shipped without a functional test, or a spec that got oversold in the sales conversation. The cells are usually fine. The paperwork and the wiring details are where things fall apart.
This is the checklist I run at the warehouse. Seven steps. It's written for system integrators, installers, and buyers sourcing pylontech us5000 battery units, pylontech 2 4 kwh lifepo4 modules for smaller builds, or anything headed into a community solar battery storage project.
One quick note before we start: if you landed here searching are lipids long term energy storage—that's a biochemistry question, not a battery question. I'll address it briefly at the end, because that query pulls a few installers off course every year and it's worth closing the loop.
Step 1: Verify document versions, not just document existence
Before anything gets unpacked, I put the paperwork on the table.
- UN 38.3 transport test summary—required for shipping lithium cells
- UL 1973 certificate and report for the battery pack
- IEC 62619 if the destination market is EU or Australia
- MSDS for the module chemistry
- The datasheet revision that matches the shipped batch
The trap here is datasheet revision. Pylontech has updated US-series datasheets more than once, and cycle life, depth of discharge, and operating temperature range have shifted between revisions. I had a shipment where the cycle-life figure in the supplier's cover letter didn't match the datasheet packed with the goods—off by something like 600 cycles. Not fabrication, just an old revision circulating. We re-papered the warranty terms against the current revision before anything got quoted to the customer.
A stale document is more dangerous than a missing one, because you assume it's valid.
Step 2: Read the nameplate, and don't mix nameplate with measured
The US2000 is 48V / 50Ah, nominally 2.4 kWh. The US3000 lands around 3.5 kWh. The US5000 is 48V / 100Ah, roughly 4.8 kWh. Side by side in a rack, they look almost identical, which is exactly why this step exists.
What I check:
- Model number per unit against the packing list—line by line, not spot checks
- Nameplate voltage, Ah, and kWh
- Serial numbers against the shipment batch and the warranty registration sheet
Boring work. It saves more downstream pain than any other step on this list. When capacity class or revision doesn't match the order, the inverter configuration and the string design downstream both inherit the error, and you find out at commissioning.
Step 3: Confirm BMS communication against the actual inverter firmware
Pylontech modules speak CAN and RS485, but the protocol version has to match the inverter's firmware—and I'm not going to pretend I remember every inverter brand's compatibility matrix. Check the vendor's compatibility list. The real check is whether the protocol revision shipped in this batch matches the inverter firmware revision on this project, not last quarter's project.
Procedure: pull one sample unit, connect it to a bench inverter from the actual project bill of materials, and run a handshake. Confirm you can read state of charge, voltage, and alarm flags. Do this before the modules go to site.
What I mean by "run a handshake" is: don't just confirm the cable fits and the LED turns green. Read values off the inverter screen and compare them against a multimeter reading on the terminals.
Step 4: Sample-test capacity—but record the test conditions
Pull one or two units and run a discharge test. The point is not whether the measured number equals the nameplate. It's what conditions the number was measured under. 0.5C versus 1C, 25°C versus 15°C, discharge to 80% DoD versus 90%—those swing the result meaningfully.
My acceptance threshold: sampled capacity at or above 95% of nameplate, with the test conditions written down on the inspection record. Verbal reassurances like "it's basically the same as spec" don't go on a record sheet.
Step 5: The 12V auxiliary circuit and quick disconnect check
This is the step most integrators skip, and it's where the real failures live. The 12V side usually feeds the BMS, monitoring, or external comms. If the build uses a 12 volt battery quick disconnect, check three things:
- Disconnect current rating has headroom over actual auxiliary load—not nameplate load, actual
- Terminal torque is set to the manufacturer's spec, not "tight enough"
- Contact resistance after repeated mating cycles, which is the one almost nobody measures
I know I should have specified a higher-rated quick disconnect on one build, but I thought—what are the odds the auxiliary draw matters? The odds caught up with me the following summer, when terminal heating caused the monitoring module to reboot on a loop. Full rework, new connectors, and a schedule slip we ate.
I have mixed feelings about how much inspection time this deserves. On one hand, it's a 12V circuit with a few amps on it and treating it like a high-voltage bus feels excessive. On the other, that's precisely the reasoning that cost us three weeks. Now it's a standing line item on the acceptance form, and it doesn't get skipped.
Step 6: Community solar projects need grid-tie and fire-code documents in the same packet
For community solar battery storage, the incoming inspection has to confirm that the documentation package can actually support your AHJ submittal. That means:
- UL 9540 for the complete energy storage system, not only the battery pack
- UL 9540A thermal runaway propagation test report, if the local authority requires it
- IEEE 1547-2018 interconnection compliance
- NFPA 855 installation compliance for the enclosure, spacing, and signage
- Local fire department specifics on separation distance, ventilation, and labeling
The battery vendor can't warrant your AHJ approval. What you can verify at receiving is whether the file set you have will support the application. Finding out at the permitting desk that a 9540A report is missing is an expensive way to learn this.
Step 7: The quote transparency review—ask what's not included before you ask the price
This step happens before the purchase order, not at the dock.
I've learned to ask what's not included before I ask what the price is. The recurring gaps: certification costs, freight insurance, BMS firmware upgrade service, logistics on in-warranty replacements, and whether consumables like quick disconnect connectors are on your account or theirs.
The vendor who itemizes all of that upfront—even when the total looks higher—usually costs less in the end. I've seen too many low initial quotes turn into a mid-project supplement where the final number landed close to the transparent bid, minus three weeks and a pile of email threads. So the rule is: line-item the exclusions on the quote, in writing, before signature. A total you can reconcile line by line is worth more than a number you can't.
Common errors and things to watch
- Don't reuse last shipment's checklist. Model, revision, and protocol can all change between orders.
- Don't sign off on a voltage reading alone. A BMS handshake and a capacity sample are separate tests.
- Expired paperwork beats missing paperwork at hiding problems, because you won't go looking for a replacement.
- 12V-side hardware gets treated as consumable during procurement. It shouldn't be—specify it like a real component.
- Cycle-life numbers must come from the datasheet revision that matches the shipped batch, not from a marketing sheet carried over from a prior year.
- Don't accept a verbal capacity result. Get the test conditions and the measurement on the inspection record.
If your team is publishing content around storage keywords and someone has queued up the lipid question—lipids are indeed a long-term energy storage form in living organisms, but they have nothing to do with LFP modules. Keep that query in a separate content track, or it will pull genuine storage buyers off the page.
