Where to put temperature loggers, and what to do with the data

Placement is the whole exercise

The rule that governs every other decision in this guide: place the logger where the product is warmest, not where it is easiest to reach. For most loads that means away from the reefer unit, away from the door where sun and dock heat hit hardest during loading, and buried inside the pallet stack rather than taped to the outside of a box. A logger clipped to the outer wrap of the last pallet loaded is recording ambient dock air through a thin layer of cardboard, not the produce three pallets in.

The single most common placement error is mounting the logger in or near the supply air stream, the cold air coming directly off the evaporator coil. That air is close to setpoint by design, because it is the machine's own output, and a logger sitting in it will show a flat line at setpoint for the entire voyage regardless of what is happening to the cargo. Return air, which has already passed through the load, is a somewhat better proxy, but it can still mask a warm pocket in the middle of a dense pallet stack where airflow barely reaches.

The test to apply before fixing a logger anywhere: if this container lost power for six hours, where would the product warm up fastest and by how much? Put the logger there, usually the geometric center of the densest part of the load, not the point nearest the technician's hand.

Pallet level versus container level

A single logger on a container tells you about the container's worst point, not about any specific pallet inside it. If the shipment is one lot from one grower going to one buyer, container-level monitoring is usually enough, because a claim or a rejection applies to the whole load anyway. If the container is consolidating multiple lots or multiple buyers who each want their own proof of custody, pallet-level logging becomes the only way to defend an individual pallet's condition when a dispute is about one buyer's portion and not the whole box.

Pallet-level loggers should sit inside the stretch wrap, roughly at the vertical mid-point of the stack rather than at the top or bottom. The top layer of a pallet sees the most radiant heat during any dock exposure; the bottom layer sits closest to the truck bed or container floor, which carries its own thermal load from road heat during drayage. The middle of the stack best represents the pallet as a whole, and it is also least disturbed by someone bumping the logger while restacking.

The trade-off is cost and retrieval. A logger buried inside stretch wrap has to be cut out or the pallet broken down to recover it, fine for a claim investigation but friction on every routine shipment. Most exporters land on a hybrid: two or three container-level loggers on every load as a baseline, with pallet-level loggers added only on lots that are higher value, split across multiple buyers, or already flagged as a quality risk.

The three-point pattern for a full container load, and why the door end matters most

For a standard 40 foot reefer container carrying a full load, the working default among cold chain auditors is three loggers in a fixed triangle: one near the door, one at the mid-point of the load, and one at the nose, the end furthest from the door where the reefer unit sits. Some programs add a fourth logger at floor level near the T-bar flooring, because floor-level air can run warmer or colder than the bulk of the load depending on whether the floor channels are clear enough for airflow to return properly.

The door end deserves the most attention of the three, for two compounding reasons. It is the point most exposed to ambient heat during every door-opening event, loading, a customs inspection, a partial unload at a transshipment point, each one lets outside air in and the door-end pallets absorb that heat fastest. And it is furthest from the reefer unit's most direct airflow path in many container designs, so it recovers slower after each exposure. A trace that shows the nose holding steady at setpoint while the door end spikes by several degrees on every stop is not describing three different container conditions, it is describing the one condition that actually matters for the pallets nearest the door, which are very often the ones unloaded and inspected first at destination.

Worked example: a mixed citrus load crossing three ports of call showed the nose logger holding a tight band around setpoint throughout, the mid-container logger drifting up roughly a degree during each port call and recovering within hours, and the door-end logger spiking well above setpoint at every call and taking most of a day to recover each time. Read alone, the nose trace would have cleared the shipment. Read as a full pattern, the door-end trace was the one that mattered, because it was the door-end pallets the receiver's quality team sampled on arrival, and it was the door-end trace that explained the softening they found.

Cold room versus in-transit placement

A cold room's temperature is rarely uniform, and the same warmest-point logic applies indoors. Air near the evaporator coils and the door is typically the most volatile part of a cold room: coil air cycles with the defrost schedule, and door air spikes every time a forklift comes through. The most representative point in a cold room is usually the geometric center of the storage area, at a height roughly in the middle of the stored product rather than at floor or ceiling level, because heat stratifies and ceiling air can run several degrees warmer than floor air in a room with imperfect air circulation.

In transit, placement has to account for the specific failure modes of that leg. On a reefer container, that is the door-end pattern described above. On a refrigerated truck making multiple stops, place the logger nearest the load that goes out last, because that is the product sitting in a partially opened, partially unloaded vehicle the longest across a multi-stop route. On break-bulk or less-than-container-load shipments moving through a consolidation warehouse, the highest-risk point is often not any single vehicle but the dwell time on a dock between legs, where product can sit outside active refrigeration entirely; a logger that only records once cargo is back inside a reefer will miss that gap.

Logger types, calibration, and start delay

Single-use loggers are the default for most export shipments: inexpensive, disposable, adequate for a one-way trip, and they remove the retrieval problem because nobody has to ship the device back. Their limitation is exactly that they are disposable, so there is no calibration history building across trips and no reusing the same known-good unit. Reusable loggers cost more per unit but carry a calibration certificate forward across shipments, which matters for a buyer that wants a documented calibration trail rather than a fresh, unverified device on every load. Real-time loggers, transmitting over cellular or satellite during the voyage rather than only on retrieval, cost the most and add a subscription, but they are the only option that lets someone act on an excursion while the product is still moving rather than discovering it after arrival.

Calibration is not optional paperwork. An uncalibrated logger has no defensive value in a dispute: a buyer, insurer, or claims adjuster can simply discount or reject a reading with no calibration certificate behind it, and they are right to, because an uncalibrated sensor's drift is unknown and unbounded. A logger should carry a calibration certificate traceable to a national metrology body or an accredited calibration lab, stating the date of calibration and the tolerance verified, typically within a fraction of a degree at the temperature range relevant to the cargo. Buy calibrated stock from a reputable supplier and keep the certificates on file.

Start delay is the setting that keeps the pre-load period from polluting the record. A logger started the moment it is switched on will record the warm ambient air of the packhouse or loading dock before the container door closes and cooling begins, and that pre-cooling period can look, on paper, like an early excursion that never actually touched the cargo inside a closed, refrigerated environment. Set a start delay long enough to cover realistic loading time, commonly one to a few hours depending on the facility, so the recorded trace begins once the load is sealed and the reefer unit is running, not while the door is still open.

Reading the data: what counts as an excursion, and mean kinetic temperature

An excursion is any period where the recorded temperature falls outside the specification set for that product, not simply any moment the trace ticks a fraction of a degree away from the target setpoint. The specification should define both a tolerance band and a duration threshold: a two-degree deviation for eleven minutes during a routine defrost cycle is normal reefer behavior, not an excursion worth acting on; the same deviation sustained for six hours is a different event entirely. Set the threshold before the shipment moves, not after a suspicious-looking trace comes back and someone has to decide in hindsight whether it counts.

Simple minimum and maximum readings are the wrong tool for judging whether a fluctuating trace still protected shelf life, because they treat one brief spike the same as a sustained one. Mean kinetic temperature is the better measure: a single calculated value that weights the trace by the chemistry of degradation, so time spent at higher temperatures counts for more than the same span near setpoint, reflecting how spoilage reactions accelerate faster than temperature rises linearly. A trace that spikes briefly but spends the overwhelming majority of the voyage tight to setpoint can still produce an acceptable mean kinetic temperature, while a trace that never breaches any single threshold but runs persistently a degree or two warm throughout can produce one that is actually worse for the product. Judge the whole trace by mean kinetic temperature against the product's stated tolerance, and treat raw min and max as a secondary check for genuine spikes, not the primary verdict.

The first hour of an alarm, and what to keep for a claim

When a real-time logger raises an alarm, the first hour is for containment, not paperwork. Confirm the alarm is real rather than a sensor fault by checking whether a second logger on the same load shows the same trend; identify whether the cause is still active, a reefer unit fault, a power loss, a door left open, or already resolved; and if it is still active, get whoever can act on it, the carrier, the terminal, the trucking company, moving on a fix immediately, because every additional hour at an elevated temperature both worsens the product and worsens the eventual claim. Document the timestamp of the alarm and the timestamp of any corrective action taken, because the gap between the two is exactly what an insurer or a buyer will ask about later.

For a claim, keep the full logger file, not just a summary printout, because an adjuster will want to see the complete trace including the periods that looked fine, not only the excursion itself. Keep the calibration certificate for that specific logger or lot. Keep a record of where the logger was physically placed, ideally a photo taken at loading showing the logger's position in the pallet or container, because a trace with no documented placement is open to the argument that it was not representative of the cargo. Keep the bill of lading or waybill showing the transit timeline, so the excursion can be tied to a specific leg and a specific responsible party. And keep the product specification the trace is being judged against, since an excursion is only meaningful relative to a stated tolerance: a trace with no accompanying specification is a chart with no verdict attached to it.