That ISO Certificate May Not Cover Your Battery Pack Assembly

A QMS certificate is evidence of one narrow thing: that an accredited certification body audited a defined management system, at a defined site, for a defined period, against a defined scope. It is not evidence that the battery pack assembly line you are standing in front of is inside that scope. The first move at a supplier audit is to read the scope statement, the legal entity name, and the site address on the certificate line by line, before anyone opens a process document, because the most common problem in battery pack sourcing is not a forged certificate. It is a completely genuine certificate that stops at the edge of the assembly floor.

The plant manager hands me the binder before I have taken my jacket off. Black three-ring, tabbed with masking tape labels, the certificates in plastic sleeves at the front. Quality management system. Environmental. Occupational health and safety. A product certificate with a hologram on it. He is proud of that binder, and he should be, because somebody in this building did the work to earn those documents. But I do not read certificates from the front. I read them from the scope line, and the scope line is exactly the part that gets photocopied at low resolution or left off the wall copy entirely.

So I ask for the original, not the frame. He brings it to the table. That is usually where it starts.

The scope statement is the whole document

Everything else on the certificate is decoration around one or two sentences. The border, the signature, the accreditation mark, the hologram. Those sentences are the scope of registration, and they describe what the certification body actually audited.

The wording matters more than people expect. "Manufacture of lithium-ion battery packs for industrial equipment" is not the same statement as "Assembly of electronic assemblies." And it is not the same as "Design and manufacture of portable power products." One of those covers the pack line. One of them covers the surface-mount line next door. One of them covers the design office and, depending on exactly how the certification body worded it, may or may not include the assembly operation at all.

What I am looking for when I read that line:

• The product or service described, in the supplier's own words as accepted by the certification body
• The standard and its revision year, so a current edition and not a superseded one
• The site address, exactly as it appears on the certificate
• The legal entity name
• The certification body that issued it
• The accreditation body, and whether an accreditation mark is present

I copy that scope sentence into my notes verbatim. Not a paraphrase. Verbatim. If I later write "scope covers battery pack assembly" in an audit report, I want to be able to point at the string of words that says so.

The address on the certificate is a real address

Battery pack assembly has a habit of moving. The corporate campus keeps the certificates; the pack line runs in a leased building across town, or at a contract manufacturer, or on a mezzanine that used to be the shipping office. A multi-site certificate can list additional sites on an annex, and that annex is part of the certificate. If it is not attached to the copy you were given, then you do not actually know what the certificate covers.

Then there is the entity. Your purchase order goes to one limited liability company. The certificate is in the name of a group company, or a holding company registered somewhere else entirely. That is not automatically a problem. Corporate structures are what they are. But it does mean the certificate in your hand is the wrong company's paper unless you can build a bridge: a legal relationship document, a separate certificate in the contracting entity's name, or the parent company explicitly listed as covering that site.

If I cannot build that bridge, the certificate goes into the file as "does not cover the contracting entity," and my audit plan gets heavier, not lighter.

Accreditation is checkable in about four minutes

The temptation is to trust the PDF. The certificate came from the supplier, and the supplier wants you to trust it. So do not start there. Every legitimate accreditation scheme publishes a searchable directory. You type in the certificate number, or the company name, and a record comes back showing the scope, the site, the issue date, the expiry date, and the current status.

I do this on my phone, in the room, usually while the plant manager is still standing there. Four things I am confirming:

1. The certificate exists in the directory at all.
2. The status is active. Suspensions do not always make it onto the wall copy, and a suspended certificate is one that failed a surveillance audit.
3. The scope recorded in the directory matches the scope printed on the paper. If they differ, the paper is old.
4. The accreditation mark corresponds to a body that is actually accredited for that standard.

That last one is where the odd documents live. A "certificate of registration" with no accreditation mark, issued by a firm that holds no accreditation for anything. It is not necessarily fraudulent. It just means no independent party has checked the issuer, which makes it close to worthless as evidence.

UN 38.3: you will get a summary, and you should ask what sits behind it

Almost everything you are handed for transport is a test summary. One page or so, in a standardized format, listing the cell and battery descriptions, the test items, and a pass or fail line. It is designed to be the travelling document. The full test report, with the actual data, the test conditions, the photographs of the sample as tested, and the equipment used, stays with the lab or with whoever paid for the testing.

Both documents are legitimate. They are not interchangeable.

The summary tells you that a test happened to something. The report tells you what that something was. When I have a configuration question, because the cell model changed or the capacity changed or the protection circuit supplier changed or the pack grew two cells, the summary usually will not answer it. The description line on a summary is brief, and it sometimes says something like "representative" or lists a generic capacity. The report is where the construction is spelled out in detail.

So the reasonable request is not "send me the full report" for every part number. It is this: send the summary always. Send the full report when the battery description on the summary does not match the configuration you are buying, when the summary is the only thread connecting the test to your product, or when the pack is the largest thing in the shipment and you want certainty. Suppliers do push back on full reports, and I understand why, because the report is more work and the lab sometimes charges to reissue it. In my experience they can usually produce it.

Product certifications attach to a model, not to a company

UL 2054 and the 62133 family, both the UL and IEC versions, get talked about as if they were a badge the company earned. They are not. They attach to a specific construction: a cell, a pack, a protection scheme, a rating.

This is the part that trips people up.

• The certificate or listing carries model designations. Those designations need to match the model on your drawing and the model on the label. "Close enough" model names are how you end up with a pack that is not covered by anything.
• The listing covers a construction. Change the cell and you may have a different construction. Change the protection circuit module and you may have a different construction. Substituting a functionally similar part is not a paperwork exercise, because it can put you outside the evaluated configuration.
• Listings typically include follow-up services, meaning periodic factory inspections by the listing organization to confirm the product still matches what was evaluated. If those inspections lapse, the listing can be affected even if the file number still returns a result in a search.

When I look at a product certificate, I write down the file or report number, the standard and edition, the model designations, and the date. Then I place the model designation next to the model on the purchase drawing. If they do not line up, that certificate is an interesting document about a different product.

The records behind the certificate

The certificate is a conclusion. The records are the reasoning. A surveillance audit is one day a year, maybe two, sampling a system that runs 250 days. So when I am on site with the scope confirmed, I ask for the paper trail. Not everything, but the parts that touch the pack line:

• The internal audit schedule and the reports from the last cycle, specifically any internal audit that covered pack assembly
• Management review minutes. Not the agenda. The minutes, with decisions and assigned actions in them
• The corrective action log, opened and closed, with a couple of recent entries traced back to root cause
• Calibration records for anything that measures on the pack line: torque drivers, insulation resistance or hi-pot testers, battery management system programming and test fixtures, capacity test equipment
• Incoming inspection records for cells, and the supplier qualification file for the cell manufacturer
• Nonconforming material records, showing what was quarantined, what was scrapped, and whether any of it was reworked into a shipped lot
• Training records for the operators on the pack line, and the work instructions they were trained against
• Control of externally provided processes, if any step of the pack, such as tab welding, potting, or harness assembly, is subcontracted

I am not reading all of this top to bottom. I am looking for whether the system touches the pack line at all. If the internal audit reports only ever mention the surface-mount line and the warehouse, then the management system is real, but the pack line is practically outside it, no matter what the scope sentence says.

That is the finding I write down. And it needs to be written in a way that does not overstate the certificate in either direction.

How this shows up in the report

The failure mode in audit reporting is compression. You write "supplier is certified to the quality management standard" and the reader downstream, the program manager, the contract, the customer's quality representative, reads that as "the manufacturing operation is under a certified quality system." Those are not the same sentence.

What I write instead is mechanical:

• Certificate number, standard and revision, certification body, accreditation body
• Legal entity and site address as stated on the certificate
• Scope statement, quoted verbatim
• Issue date and expiry date, plus the date and result of the directory lookup
• A coverage conclusion in plain language

The coverage conclusion is the part people skip, and it is the part that matters. Something like: "Scope covers assembly of lithium-ion battery packs at the listed site. Confirmed in directory on [date]." Or the other version: "Scope covers trading and distribution; does not cover pack assembly at the listed site. Pack assembly verified by process audit and records review. QMS coverage gap noted."

The second version is not a rejection. It is a description. It tells the next person exactly how much weight the certificate can carry, which is the entire point, and it protects you. Six months later, when somebody asks why you accepted a supplier whose certificate did not cover the line, the answer is already written down in the file.

What the certificate cannot do

Even when everything checks out, right scope, right entity, right site, active in the directory, product listing matched to your model, the certificate still only tells you that a management system was audited periodically against a standard. It does not tell you the pack line's first-pass yield. It does not tell you whether the tab welder is drifting. It does not tell you whether the operator on nights follows the same work instruction as the operator on days.

That is why I treat certification as one column in the supplier file, not as the file itself. It is a filter, not a finding. The scope check takes twenty minutes and eliminates a specific class of surprise, the surprise being that the thing you assumed was covered was never in scope to begin with. What it does not do is replace the process audit. Nothing does.

And there is a case that comes up more often than people admit: the scope genuinely does not cover the pack line, and the supplier cannot get it added before you need to place the order. That happens with a new line, a new site, or a contract manufacturer that is still ramping. The answer is not to pretend. It is to write the gap down, raise the audit frequency for the first year, and buy the first production run with your eyes open. That is a decision you can make on the facts. You just cannot make it on the certificate.