Mean kinetic temperature: how to adjudicate an excursion
Mean kinetic temperature is not an average. It is a single calculated temperature that represents the cumulative degradation effect of a varying temperature history — weighted so that time spent warm counts far more heavily than time spent cold. USP <1079.2> is the chapter that governs its use in evaluating storage and transport excursions.
Under the USP <659> controlled cold temperature definition, a refrigerated shipment may experience excursions between 2°C and 15°C for up to 24 hours provided calculated MKT stays at or below 8°C. MKT is how you determine whether a warm arrival fits inside that envelope.
It is also the most misused concept in pharmacy cold chain. MKT gets treated as a way to make an inconvenient logger reading go away, and USP anticipated exactly that. The chapter’s own guidance places a hard limit on the practice, and it is worth reading before the calculation:
“…may not be used to justify a storage or transportation system that has repeated excursions; such a system is not in control and needs to be corrected.”
USP guidance on the application of mean kinetic temperature
That sentence is the whole ethic of the tool. MKT adjudicates the individual shipment. It does not license a lane that keeps failing. If you are running the calculation routinely on the same route, the calculation is not your problem — the packout is.
Why an arithmetic average is the wrong tool
Chemical degradation rates rise exponentially with temperature, not linearly. Six hours at 14°C does considerably more damage than six hours at 2°C prevents. An arithmetic mean treats those symmetrically and therefore systematically understates the harm of a spike.
MKT corrects this by applying the Arrhenius relationship — converting each temperature reading into a rate-like quantity, averaging those, and converting back. The result is always at or above the arithmetic mean, and the gap widens as the temperature history gets spikier.
The calculation
MKT is computed with the Haynes equation:
MKT = (ΔH / R) ÷ −ln[ ( Σ ti · e−ΔH/(R·Ti) ) ÷ Σ ti ]
Where Ti is each temperature in kelvin, ti is the time spent at it, R is the gas constant (8.314 J·K−1·mol−1), and ΔH is the heat of activation. A value of 83.144 kJ/mol is conventionally used as the default in the absence of product-specific data, which makes ΔH/R conveniently equal to 10,000 K.
Two notes before you use it. First, the default heat of activation is a convention, not a property of your preparation — if you have product-specific stability data, that value should come from it. Second, MKT is calculated over the whole record you are evaluating, and the period you choose changes the answer. Choosing the window after seeing the result is not analysis.
Worked example: a 72-hour lane that passes
A refrigerated shipment on a three-day ground lane. The logger record shows 66 hours holding at 5°C, then a 6-hour excursion to 14°C during a warehouse hold on the final day.
| Segment | Temperature | Kelvin | Duration |
|---|---|---|---|
| Transit, in band | 5.0°C | 278.15 K | 66 h |
| Terminal hold | 14.0°C | 287.15 K | 6 h |
Working through the equation with ΔH/R = 10,000 K:
- e−10000/278.15 = 2.434 × 10−16
- e−10000/287.15 = 7.511 × 10−16
- Time-weighted mean = [(66 × 2.434e−16) + (6 × 7.511e−16)] ÷ 72 = 2.857 × 10−16
- −ln(2.857e−16) = 35.792
- MKT = 10,000 ÷ 35.792 = 279.39 K
Calculated MKT for Example A — at or below the 8°C ceiling, excursion under 24 hours, peak under 15°C. This shipment fits inside the USP <659> controlled cold temperature envelope.
Note that the arithmetic mean of this history is 5.75°C. MKT returns 6.2°C. The half-degree gap is the exponential weighting doing its job.
Worked example: a 72-hour lane that fails
Same lane, but the packout ran warmer throughout and the terminal hold lasted twice as long: 60 hours at 7°C, then 12 hours at 14°C.
- e−10000/280.15 = 3.146 × 10−16
- Time-weighted mean = [(60 × 3.146e−16) + (12 × 7.511e−16)] ÷ 72 = 3.874 × 10−16
- −ln(3.874e−16) = 35.487
- MKT = 10,000 ÷ 35.487 = 281.79 K
Calculated MKT for Example B — above the 8°C ceiling. Every individual reading stayed under 15°C and no single excursion exceeded 24 hours, but the shipment still falls outside the envelope.
This is the case that catches people. Nothing in Example B looks dramatic on the chart. No alarm threshold was crossed. The peak was identical to Example A. What failed was the cumulative thermal load — and only the calculation surfaces it.
What MKT does not do
- It does not override the hard bounds. A reading above 15°C or below 2°C puts the shipment outside the controlled cold temperature definition regardless of what MKT computes to.
- It does not extend the 24-hour excursion window. Both conditions must be satisfied.
- It does not substitute for stability data. MKT tells you whether the exposure fits a definition. Whether your specific preparation is still fit for dispensing is a stability question, and for a compounded preparation the stability data is yours to hold.
- It does not rescue a failing system. Per USP’s own language, repeated excursions mean the system is not in control and needs correcting. Serial MKT calculations on one lane are a finding waiting to happen.
Practical implementation
To run this at all you need a temperature record with enough resolution to be meaningful — USP <1079.3> covers monitoring devices, including calibration expectations. A single peak reading is not a temperature history and cannot support an MKT calculation.
Written procedures should specify, in advance: the sampling interval, the evaluation window, the heat of activation used and its basis, who performs the calculation, who dispositions the result, and what happens when the answer is unfavourable. Deciding any of that after seeing a number is the failure mode an inspector is looking for.
Frequently asked questions
What is mean kinetic temperature in simple terms?
A single temperature that would produce the same cumulative degradation effect as the varying temperature history a product actually experienced. Because degradation accelerates exponentially with heat, MKT weights warm periods far more heavily than an arithmetic average does, and always lands at or above the arithmetic mean.
What MKT value is acceptable for refrigerated products?
Under the USP <659> controlled cold temperature definition, calculated MKT must be not more than 8°C, with no excursion below 2°C or above 15°C and no excursion exceeding 24 hours. All of those conditions apply together.
Can MKT be used to release a shipment that went out of range?
It can be used to evaluate whether the exposure fits inside the permitted excursion envelope. It cannot be used to justify a lane with repeated excursions — USP states plainly that such a system is not in control and needs to be corrected. It also cannot substitute for stability data on your specific preparation.
What heat of activation should I use?
83.144 kJ/mol is the conventional default where product-specific data is unavailable, and it makes the ΔH/R term equal 10,000 K. If you hold stability data for your preparation, the value should be derived from it, and your procedure should record which value you used and why.
Which USP chapter covers MKT?
USP <1079.2>, Mean Kinetic Temperature in the Evaluation of Temperature Excursions During Storage and Transportation of Drug Products. It sits under <1079> on good storage and distribution practices; <1079.3> covers the monitoring devices that produce the data.
Do I need a data logger to calculate MKT?
You need a time-series temperature record, which in practice means a logging device rather than a single-point indicator. USP <1079.3> addresses device selection and calibration.
Running MKT on the same lane repeatedly?
That is the signal USP describes as a system out of control. Meridian builds packouts sized for pharmacy vial counts and chamber-tested at summer-peak ambient. Request a lane test on the route that keeps failing.
Request a lane testSources
- USP General Chapter <1079.2>, Mean Kinetic Temperature in the Evaluation of Temperature Excursions During Storage and Transportation of Drug Products. USP.org.
- USP General Chapter <1079>, Risks and Mitigation Strategies for the Storage and Transportation of Finished Drug Products. USP.org.
- USP General Chapter <1079.3>, Monitoring Devices — Time, Temperature, and Humidity.
- USP General Chapter <659>, Packaging and Storage Requirements.