Temperature mapping is the study that proves a store, fridge, cold room or vehicle holds its labelled range everywhere you keep stock, and it tells you where the permanent monitoring probes belong. The one thing to remember: no UK or EU source sets a fixed re-mapping interval, so your frequency, triggers and method are yours to justify in a written risk assessment.
Why it matters
A probe only tells you about the air around it
A monitoring probe reports on one point. A warehouse, cold room or large fridge is not one temperature: stock high in the racking, by a heater or near a loading bay door can sit well away from the reading on the wall, so the record can look perfect while product elsewhere goes out of range. That is why EU GDP 3.2.1 asks for permanent monitors to be placed according to the mapping results, where the extremes of fluctuation are.
The risk runs both ways. The MHRA Inspectorate blog points out that vaccines, insulin and blood products are vulnerable to freezing as well as heat, so a cold spot matters as much as a warm one. Mapping is how you find both before your stock does.
What the rules say
From the licence condition to the mapping clause
For a wholesaler the duty starts in law. Regulation 43 of the Human Medicines Regulations 2012 makes compliance with the GDP guidelines a condition of the licence and requires premises and equipment that maintain the quality of the products. For Great Britain those guidelines are, in practice, the EU GDP Guidelines 2013/C 343/01 as they stood at the end of the transition period. The MHRA GDP Inspectorate enforces them; in Ireland the HPRA does.
Section 3.2.1 is short. Map the storage area before use, under representative conditions; put the permanent monitors where mapping shows the extremes; and repeat according to a risk assessment or after significant modifications to the facility or its temperature-controlling equipment. Small premises of a few square metres at room temperature may use a risk assessment (of heaters, for example) instead. Section 9.4 applies the same idea to temperature-controlled vehicles, taking account of seasonal variations.
If you are applying for a licence, MHRA Guidance Note 6 says you must be inspection-ready at the point of application, which includes a mapping exercise of the proposed wholesale area with a clear methodology.
How the inspectors and WHO fill the gaps
The guidelines do not say how to map, how many sensors to use or how long to run, and the MHRA mapping blog says so. Its advice is practical: map before stock goes in, using dummy product in empty fridges and cold stores; map again when fully stocked; repeat to cover seasonal variation; and include the highest storage point, heaters, lighting, windows, doors, loading bays, mezzanines and the controlled drug, packing, returns and quarantine areas. The RP should be involved throughout.
The European Commission GDP Q&A adds that probe numbers follow a risk analysis, not a fixed count, and that mapping should be done in different seasons. For numbers, most people turn to WHO Technical Supplement 8: loggers every 5 to 10 metres, at low, middle and high level where ceilings are 3.6 m or lower, at least seven consecutive days for an ambient warehouse (five working days and a weekend) and 24 to 72 hours for cold rooms and freezer rooms. WHO presents these as guidance, not rules.
What good looks like
A mapping programme, not a one-off report
A good study starts with a protocol approved before any logger goes in: the area, the conditions tested (empty, loaded, summer, winter), logger positions with reasons, run length and acceptance criteria. WHO Supplement 8 expects one logger type per study, sampling at the same interval of 1 to 15 minutes, each with a valid 3-point calibration within ±0.5 °C. If a contractor runs the study, your organisation still reviews and approves it: MHRA's 2015 GMP deficiency data cite contractor mapping the company had not approved.
The report should end in decisions: where the permanent probes go, where stock must not be stored, and when to re-map. If your monitoring went in before the mapping, or the two were never reconciled, see how to justify each probe position. WHO's guidance adds that if a spot near a door causes alarms that cannot be fixed, stop storing product there rather than moving the probe. Triggers worth writing into your procedure:
- significant modifications to the facility or its temperature-controlling equipment (EU GDP 3.2.1);
- reconfiguring the storage area (MHRA mapping blog);
- changes to loading, airflow, refrigeration equipment or set point, and unexplained variability in monitoring data (WHO Supplement 8);
- repair or replacement of a refrigeration unit, changes to how duty and back-up condensers run, and moving a fridge (HPRA IA-G0011, for Irish sites, and a useful checklist anywhere).
Fridges, cold rooms and freezers
The HPRA expects large fridges and walk-in cold rooms to be mapped on installation, empty and full, with an ambient probe outside the unit. WHO Supplement 8 adds that a room with duplicate refrigeration units should be mapped with each unit running on its own. WHO Supplement 7 asks qualification to show how long a new store holds temperature after a power failure and how quickly it recovers, tested while the store is empty because the test can cause an excursion. The result feeds your contingency plan.
The MHRA blog says the effect on temperature distribution should be investigated if a fridge is filled to capacity, and the freezer blog says freezers must be qualified, monitored and mapped like refrigerators. Vans and shippers are covered on the transport page.
Worked example
Mapping a new ambient store before go-live
This is an illustrative example; the site and readings are invented. A small wholesaler has fitted out an ambient store with racking to 5 m ahead of its licence application, for stock labelled 'do not store above 25 °C'.
Week one, the protocol. Using WHO Supplement 8 as the reference, the QA lead plans a grid of loggers, vertical arrays over the full racking height, and extra loggers by the roller door, the south-facing windows, the heaters and the returns cage. Logger calibration certificates are checked and filed, and the RP approves the protocol.
Monday 09:00 in July, the summer run. It runs seven days including a weekend. On Saturday afternoon the logger at the top of the rack by the windows peaks at 26.4 °C; every other position stays in range. What they do: mark the top beam in those bays 'not for use', raise a change control for solar film, put that zone's permanent probe at the top of the rack, and book a winter run for January.
What they record: the protocol, raw data, calibration certificates, the report with the deviation and its investigation, the reason for each permanent probe position, the re-mapping triggers and the RP's sign-off. That file is the evidence Guidance Note 6 asks applicants to show.
Where it is unclear
The worst case, not the door threshold
Mapping often finds the biggest swings by a door or a loading bay, and some sites put the alarm probe right there. It then alarms on every delivery and the team stops listening. That is not what the guidance asks for. WHO Annex 9 puts monitoring sensors where the greatest variability is expected, but positioned so as to be minimally affected by transient events such as door opening. WHO Supplement 6 says monitors should not sit where door openings generate an abnormally high number of alarms, and the HPRA puts probes in a load simulator so door openings during picking do not trigger alarms.
So the alarm probe goes at the mapped worst case that represents stock, and it is buffered so a door opening does not trip it. There are three ways to buffer it:
- A sensor whose housing is the buffer. A sealed device with its sensing element inside, such as our Node+ Embedded, where the housing's thermal mass damps brief air changes. There is no liquid to top up or spill, no cable to snag and nothing to replace, which suits warehouses and ambient or cold stores. Confirm its response against your mapping data before setting delays.
- An external probe in a simulant: water, glycol or glycerol, glass beads or a block, with the container and fill volume fixed and written down.
- A product probe: a probe in a real or mock unit of product. For blood, that means a 200 mL water-equivalent simulant with no delay on the product alarm (blood product simulants).
Interval, seasons and small rooms
Three questions come up again and again. How often to re-map: EU GDP, the MHRA, the HPRA and SPS set no interval (our recommendation is below), and WHO's 'every three years' is an example (re-mapping frequency). Whether cold rooms need summer and winter runs: the Commission's Q&A says different seasons with no exception, WHO says typically not (seasonal mapping of cold rooms). And how small 'a few square metres' is: nobody defines it (small premises).
Two more sit at the edges. A 2016 MHRA blog suggests a central fridge probe while EU GDP points to the mapped extremes (probe position), and no UK source gives a fridge fill limit (fill limits). WHO's Supplement 8 excludes fridges from mapping, while its 2020 storage guideline includes them.
| Control | EU GDP | MHRA blog | WHO Supp. 8 | HPRA |
|---|---|---|---|---|
| Map before first use | ✓ | ✓ | ✓ | ✓ |
| Map to cover the seasons (ambient) | · | ✓ | ✓ | ✓ |
| Re-map after significant change | ✓ | ✓ | ✓ | ✓ |
| A fixed re-mapping interval | · | · | · | · |
A tick means the source states it. WHO says a second seasonal study 'may be needed'; its three-year interval is an example, not a requirement.
Inspection findings
What inspectors have found
The MHRA GDP deficiency data for 2016, a sample of 148 of 1,428 inspections, record major findings where mapping was not adequate to show the areas were suitable or to place thermometers at the extremes. They also cite a warehouse thermometer not calibrated across its operating range, and an ambient thermometer installed so its readings did not come from the calibrated probe.
On the manufacturing side, the 2015 GMP deficiency data include a storage room reconfigured without assessing the impact on the existing mapping. The 2016 report is the only GDP-specific deficiency dataset found on GOV.UK, so patterns may have changed since. For manufacturing sites, where Annex 15 qualification applies as well as or instead of the GDP mapping clause, see GMP storage, mapping and monitoring.
In summary
The short version
- Map every storage area before use, under representative conditions, and keep the protocol, data and report.
- Place permanent probes where mapping shows the extremes, and record why each one is there.
- Test the conditions that change the answer: empty and full, summer and winter, each refrigeration unit alone.
- Use calibrated loggers of one type per study, and check the certificates first.
- Approve any contractor's protocol and report yourself, with the RP involved.
- Write down your re-mapping interval and triggers, and review them at self-inspection.
- Where the sources differ, assess the risk and record your reasoning.