What the Cell Lot Code Reveals That Most Receiving Inspections Miss

What the Cell Lot Code Reveals That Most Receiving Inspections Miss

The lot code printed on the cell can is the last identifier still standing after the packaging is gone, and it is the one most receiving inspections never actually read. Cell makers mark the can with a code that encodes at minimum the manufacturing date and the plant, and sometimes the line and shift as well. It is not the supplier's part number, and it is not on the carton. If incoming inspection stops at the carton label and the part number, nothing about the cells themselves has been verified — only that a correctly labeled box arrived on the dock.

I keep a loupe in my receiving kit. Not because I am fussy about optics, but because the code on a 21700 is roughly the height of a pinhead, and the first time I tried to read one with bare eyes I logged a character that was not on the can. The supplier's paperwork said one thing. The carton label said another. The can said a third. That day is why I stopped treating the box as the evidence.

The moment the packaging stops being the answer

Here is the setup, and if you have stood at this bench you already know it. A carton of cells lands in incoming inspection. The carton label is clean. Part number matches the PO, quantity matches, the country of origin line reads what you expect. Inside, the cells sit in trays or shrink-wrapped blocks, and the assembler or the distributor has put its own label on the inner box. Everybody signs off. Nothing electrical is wrong. Voltage reads fine on the handful of cells somebody pulled and checked.

What nobody did was pick up a single cell, turn it under the lamp, and find the code the cell maker burned into the can. That code is the only mark in the whole shipment the distributor, the assembler, and the freight forwarder had no hand in applying — and the only one still there in two years when the pack comes back from the field.

Sit with that for a second. Everything else in the shipment is a chain of labels applied by parties with an interest in the shipment moving. The can code is the cell maker's own record, put there at the point of manufacture. It is not perfect, and I will get to that, but it is a different category of evidence.

What the code actually is

Three things get conflated constantly, so let me separate them.

The model or part number is the cell's product designation — the thing you specified on the PO and the thing the datasheet describes. It is usually printed somewhere, sometimes on the wrap, sometimes on the can, sometimes only on the packaging.

The supplier's lot number is what your distributor or assembler assigns internally so they can find the shipment again. It lives in their system. It tells you nothing about when the cell was made.

The cell maker's lot code is the mark on the can itself. It is stamped, laser-etched, or inkjet-printed at the plant, and it typically encodes the production date — year plus week, or year plus month — along with the plant or line. Some makers add shift or a sequence character. The format is not standardized across the industry by any stretch; two plants under the same brand can use different schemes, and it changes when a maker swaps out marking equipment.

That last part is what trips people up. There is no universal decoder ring. No standard says the fourth character is the week. You cannot reason your way from characters to a date the way you can with a batch of food. You need the maker's key, and without it you are reading a string you cannot interpret, which puts you right back to accepting cells on faith.

Getting the decode key before the cells arrive

I ask for it at the PO stage, not at receiving, because at receiving it is too late to be anything but a complaint.

The ask goes to one of two places. If you buy directly from the cell maker, you ask your account contact for the lot code format in writing — a one-page document that maps character positions to meaning. Most will send it. If you buy cells inside a finished pack from an assembler, you ask the assembler, and you ask them to push the request up to whichever cell maker they build with. That second path is slower and sometimes comes back with a shrug, because the assembler may not have the key either. They may be buying through a distributor who buys through someone else.

When the key never arrives, you have a decision to make, and I will tell you how I make it. I do not stop the shipment over a missing decode sheet. I write the gap into the receiving record — lot code present, decode key not provided, date verification not performed — so that whoever opens that file after a field failure knows exactly what was and was not checked. An unverified field is a liability I want visible on the form, not hidden behind a clean-looking signature.

What a real marking check looks like

The check itself takes about ninety seconds per lot once it is set up.

Pull three to five cells at random, ideally from different positions in different trays rather than three from the same bundle. Lay them under a bright bench lamp at an angle. Raking light is what makes a laser etch readable; straight-on light washes it out. Look at the can, not the wrap. On a cylindrical cell the code usually sits on the side of the can, sometimes near the positive terminal, sometimes near the bottom, occasionally on the base. On a prismatic cell it tends to be on the top face or the narrow side. On a pouch cell you are usually dealing with a small printed area near the tabs, and pouch marking is the least consistent of the three, which is its own thing to raise with the supplier.

Photograph each code with a phone at macro range. Date the photo and log the cell's position in the tray. If you can decode it, write the decoded date and plant next to the photo. If you cannot, the photo still becomes your evidence that the code exists, and comparing this shipment's photos against the last four shipments' photos is where most of the value actually lives.

Because the failure mode is rarely a code that reads clearly as the wrong date. It is a code that does not match the pattern of the previous shipments from the same maker. A different character count. A different font weight. An etch that looks printed rather than burned. A shrink-wrap seam in a place you have not seen it before. Cells supposedly from one lot coming out with two visibly different markings.

That last one should stop the line. Mixed markings inside a single shipment is the most useful tell I have found, and it shows up at the marking and packaging level long before anything electrical notices. Voltage on a rested cell will not separate a matched lot from a mixed one. Neither will a quick internal resistance spot check. The variation that matters — capacity spread, self-discharge spread — is what bites you eighteen months into a series string when one cell starts dragging the stack out of balance, and by then the packaging is long gone and so is your chance to catch it at the dock.

The distributor layer and where it hides the code

If your cells come through a distributor, expect the cell maker's code to be the deepest thing in the box. The outer carton carries the distributor's label. The inner box carries the assembler's or the distributor's label. The cells may be in trays the distributor repacked.

That means the marking check has to happen after you break the inner packaging, and some suppliers push back, because they do not want opened packaging coming back at them. That is a legitimate concern, and it is better solved with language than with stubbornness. Write into the PO that a defined sample quantity per lot may be opened for marking inspection, that opened cells go back into the same tray, and that the sample is documented photographically. Put a number in it — three to five cells per lot is what I ask for — so nobody has to argue about it on the dock while the truck waits.

If a supplier refuses even that, and I have had two over the years, treat it as a data point about how much of their own supply chain they can actually see. There is a version of this problem where your supplier genuinely does not know what lot is in the box. No amount of rigor on your dock fixes a supplier who is reading a label they did not write.

The origin statement and the can have to agree

If your file carries a country of origin statement, the marking check is where you test it against something physical. The plant code, if you can decode it, points to a plant in a country. The shipping documents name a point of export. The can itself may carry origin marking depending on the maker and the market.

I am not going to walk you through how origin is legally determined, because that is a different conversation with a different kind of expert, and it is one you should have with someone who does it for a living. What I will say is narrower. In incoming inspection, the job is to notice when the paperwork and the physical object tell different stories, and to hold the shipment until somebody reconciles them. That is a receiving function, not a customs function, and it belongs on the inspection form.

Where this check stops being useful

I want to be straight about the limits, because a marking check that gets oversold becomes a marking check that gets ignored.

It does not catch a good counterfeit. A determined relabeler can print a plausible code in a plausible font on a plausible wrap, and if you are comparing against nothing, you will pass it. What makes the check work is the archive — the photos from the last several shipments of the same cell. Pattern matching against your own history is what turns a loupe from a prop into a tool, and it takes a few months of records before it is worth much.

It does not replace electrical testing. It runs alongside it. The lot code check is cheap, fast, and catches a category of problem electrical testing is bad at catching, but it will never tell you a cell is out of capacity spec.

And it does not replace a supplier audit. If you have never stood inside the assembler's receiving area to see whether they check lot codes themselves, the code on your cells may be the only verification in the entire chain. That is worth knowing, and worth fixing upstream whether you fix it at receiving or not.

The reason I keep coming back to the can code is simple. Everywhere else in a battery procurement file, the evidence was produced by someone with a reason to want the shipment to pass. The code on the can is the one piece of evidence the cell maker generated for its own purposes, at the moment of manufacture, before anyone knew who was going to buy it. That is a different kind of witness. It is worth the ninety seconds and the loupe.