Same Part Number, Two Dispatch Lines: What the 2026 IATA Data Logger Exception Actually Covers
Write the same part number on two lines of one air consignment. Line one: forty cold chain temperature loggers, fixed to the cartons they monitor. They record from the moment the pallet leaves the cold room. Line two: forty of the same loggers, boxed and switched off, topping up the consignee’s pool.
Same factory. Same cells. Same firmware build. Two different sets of air transport rules — and the first thing that decides them is not what the device is, but what it is doing on that flight. That distinction is the whole of the 2026 IATA data logger exception. It is also the part that gets missed.
One part number, two transport statuses
The governing text is the IATA Dangerous Goods Regulations, 67th Edition, effective 1 January 2026, at 1.2.7.1(i). The addendum posted on 1 January 2026 revises that paragraph. It covers data loggers and cargo tracking devices with installed lithium batteries, attached to or placed in packages. Those devices fall outside every provision of the Regulations, provided every one of a set of conditions is met. Falling outside the Regulations is not the same as permission to fly. The operator still applies its own acceptance rules.
Read the sentence that closes the paragraph before you read the conditions. It is the one that splits the dispatch note. In the addendum’s own words: “This exception does not apply where the data loggers or cargo tracking devices are offered for transport as a consignment in accordance with Packing Instruction 967, or 970 or 978.”
Movement 1 · monitoring
Movement 2 · stock and spares
Movement 3 · return
State definition — in use. A device earns the exception by monitoring the consignment it travels with. That status belongs to the shipment, not to the model, the certificate or the datasheet. Power the same unit down, put it in a carton with thirty-nine others, and on that flight it is no longer a monitoring device. It is a battery-containing product being shipped.
Where line two actually lands
Those three packing instructions deserve naming, because they are where line two lands. PI 967 covers lithium ion batteries contained in equipment, UN 3481. PI 970 covers lithium metal batteries contained in equipment, UN 3091. PI 978 covers sodium ion batteries contained in equipment, UN 3552. One instruction per chemistry, all of them the “contained in equipment” route. None of them is a disaster. They are simply a different paperwork and packaging path — not the one your forwarder assumed when they read “IATA compliant” on the quotation.
What does the IATA 2026 data logger battery exception actually cover?
In summary: it covers data loggers and cargo tracking devices with installed lithium batteries, attached to or placed in a package, in use or intended for use during that transport. The cells and batteries must meet the 3.9.2.6 conditions. Energy limits are 20 Wh for a lithium ion cell or battery, 1 g of lithium for a lithium metal cell, 1 g aggregate for a lithium metal battery. Quantity, shock and heat conditions apply too. Devices offered as a consignment in their own right are excluded.
IATA’s own 2026 guidance document for lithium and sodium ion batteries sets the conditions out as a list. That document describes itself as guidance only, so 1.2.7.1 remains the governing text. The list is still a useful inventory, and it is longer than the energy limits everyone quotes.
The devices must be in use, or intended for use, during transport. The cells or batteries must meet the conditions referenced in 3.9.2.6. The energy limits apply. The number of devices in or on any package or overpack must be no more than the number required to track the consignment. And the devices must withstand the shocks and loadings normally encountered during transport, without generating a dangerous evolution of heat.
That last pair is a design requirement written in plain language, and the exception attaches no test number to it. How a cell is retained inside the enclosure is exactly the kind of thing the wording reaches and a datasheet does not. So the question to put to a supplier is not which construction is better; it is what design verification exists against that condition, and what document records it.
State definition — required to track. The quantity clause asks for a justification. It is not a licence, and it is not a fixed allowance. A monitoring protocol may genuinely need a second sensor at the door end of a pallet. That is an argument to make and document — not an assumption to make quietly on the packing bench.
20 Wh is not the number most datasheets were designed against
Here is the part that catches people. The general lithium thresholds are the ones most of us can recite, because they decide which section of which packing instruction a shipment lands in. The data logger exception does not use them. It caps a lithium ion cell or battery at 20 Wh. It caps a lithium metal cell at 1 g of lithium, and a lithium metal battery at 1 g aggregate. For a battery, that is a fifth of the general watt-hour ceiling and half the general lithium ceiling.
| Cell or battery | General packing-instruction threshold | DGR 1.2.7.1(i) logger exception |
|---|---|---|
| Lithium ion cell | 20 Wh | 20 Wh |
| Lithium ion battery | 100 Wh | 20 Wh |
| Lithium metal cell | 1 g lithium | 1 g lithium |
| Lithium metal battery | 2 g aggregate | 1 g aggregate |
| Sodium ion device | 20 Wh cell / 100 Wh battery | not covered in 2026 |
The arithmetic is not hard.
IATA’s guidance gives it in one line: watt-hours are nominal voltage multiplied by capacity in ampere-hours. A lithium ion pack of 3.6 V nominal rated at 5.2 Ah is 18.7 Wh, and meets the exception’s energy limit. Specify 6.5 Ah at the same nominal voltage for a longer lane and it becomes 23.4 Wh. That remains well below the general 100 Wh lithium ion battery threshold. It also exceeds this exception’s 20 Wh limit, so the exception is gone.
Notice which device that removes. Not the cheap short-mission logger. The long-mission one — the unit somebody specified because the lane runs thirty days and nobody wants a flat battery at the consignee. Energy budget and regulatory status pull in opposite directions here. The pull gets stronger the more evidence you ask the device to produce, because sampling faster, adding channels and adding weeks all spend from the same 20 Wh.
What to ask a supplier, and what not to accept
“Meets IATA lithium battery requirements” answers a different question. So does a UN 38.3 test summary on its own, which is necessary and nowhere near sufficient. Start with the numbers that decide the energy condition. For a lithium ion device, the installed battery’s watt-hour rating. For a lithium metal one, the lithium content per cell and the aggregate per battery. In both cases, ask whether the figure describes the cell or the assembled pack. Ask in writing, against the exact hardware revision being quoted, and ask what keeps the installed cell tied to that revision. Then keep going: intended use, the quantity justification, the 3.9.2.6 evidence, and the design verification behind the shock and heat condition. The energy numbers rule a device out. They do not rule it in.
The units that are not monitoring anything
Stock shipments, spares and returns are where cold chain programmes actually get caught. Those movements get planned by people who were never in the conversation about the monitored lane. A cargo tracking device sitting in a box is an easy thing to wave through as ordinary freight.
A box of loggers sent to replenish a consignee’s pool is monitoring nothing. It is a consignment of equipment with batteries in it, and it goes down the contained-in-equipment route by chemistry. The return leg is the one people assume they already know the answer to. Whether the used units are still monitoring something on the way back is a question asked per shipment, not per SKU. The device does not carry its outbound status home with it. Condition matters on that leg too. A unit that comes back with a swollen cell is a damaged-battery question and a different conversation again; a cracked case is not automatically that, but it does put the mechanical conditions back in play.
One correction is worth making here, because it turns up in a lot of internal guidance. The 30% state-of-charge restriction has nothing to do with this exception. It belongs to lithium ion cells and batteries shipped by themselves, under UN 3480 and PI 965. Loggers with batteries installed in them go down the contained-in-equipment instructions instead: PI 967, 970 and 978. Those carry no such limit. Getting this backwards causes real trouble in both directions: loose cells shipped charged, or a pallet of loggers discharged for no reason.
And a date to put in the file
Sodium ion devices do not get this exception in 2026. Appendix H of the 67th Edition lists impending changes. It says the exception for data loggers will be expanded to include certain sodium ion batteries, effective 1 January 2027. That is an announced change, not a published set of conditions, and the word “certain” is doing real work in it. A sodium ion logger on a roadmap belongs in a 2027 plan, not in a 2026 workaround.
What the exception does not say
Everything above concerns one question: may this device fly without dangerous goods treatment. Read 1.2.7.1 again and notice what is absent. Nothing about calibration. Nothing about accuracy, or sampling interval, or whether the clock survives a flat battery, or how a gap shows up in the exported file. The paragraph grants a transport status. It never claimed to say anything about the quality of the record.
Those two questions fail independently, which is the uncomfortable part. A device can sit comfortably inside the exception and still hand a QA reviewer a CSV with an unexplained four-hour hole in it. A device can fall outside the exception — shipped correctly, fully declared, entirely legal — and produce a clean, defensible record. Approving a logger for air freight and approving it as an evidence instrument are two separate approvals. Reading the DGR settles only one of them.
Run the two approvals separately
Two energy numbers, the conditions in 1.2.7.1, and an honest answer about what the device will be doing on the aircraft: that closes the transport question. The evidence question closes somewhere else entirely. Disconnect the thing. Interrupt its power. Recover what it stored. Then look at whether the export makes the missing interval visible, or quietly smooths over it.
Once the shipment is accepted, the classification question is behind you. What lands at the consignee is the file. At that point one thing matters: whether it can say, on its own, which device produced it, when each reading was taken, and where it went quiet. None of the above is dangerous goods advice — classification for a specific shipment belongs to DG-trained staff and the operator’s acceptance rules, working from the DGR text itself.
