What mean kinetic temperature (MKT) is, and why QA teams reach for it
Mean kinetic temperature (MKT) is the single temperature that would cause the same amount of chemical degradation in a drug product as the fluctuating temperatures it actually experienced over a period. ICH Q1A(R2) defines it as "a single derived temperature that, if maintained over a defined period of time, affords the same thermal challenge to a drug substance or drug product as would be experienced over a range of both higher and lower temperatures for an equivalent defined period," adds that MKT "is higher than the arithmetic mean temperature and takes into account the Arrhenius equation," and points to the formula of J. D. Haynes (1971)[1]. USP General Chapter <1150> describes it as "the single calculated temperature at which the total amount of degradation over a particular period is equal to the sum of the individual degradations that would occur at various temperatures"[2].
Why does a warehouse manager or pharmacist care? Because temperature excursions are common and expensive: 44.6% of biopharma cold-chain professionals report multiple excursions a year, and the IQVIA Institute puts the industry's losses from temperature-control failures at roughly US$35 billion annually[3]. When a controlled-room-temperature store drifts to 29 °C for an afternoon, MKT is the tool that tells you whether the month's cumulative thermal stress still sits inside the product's stability envelope, and it is the number USP-labeled storage conditions are written around. This article gives you the formula, a step-by-step worked calculation, the difference between MKT and a plain average, and, most importantly, where regulators accept MKT and where they do not.
Photo: David Trinks / Unsplash
The MKT formula (USP <1150> / Haynes) and what each term means
MKT comes from the Arrhenius equation, which says the rate of a chemical reaction rises exponentially with absolute temperature. Haynes derived the "virtual temperature" concept in 1971 for climatic-zone stability testing[4], and USP adopted it in <1150> Pharmaceutical Stability[2]. The formula is:
MKT = (ΔH / R) / ( −ln [ (e−ΔH/RT1 + e−ΔH/RT2 + … + e−ΔH/RTn) / n ] )
| Term | Meaning | Value used |
|---|---|---|
| ΔH | Heat of activation (activation energy) of the degradation reaction | 83.144 kJ/mol by default, "unless more accurate information is available from experimental studies"[2][5] |
| R | Universal gas constant | 8.3144 J/(mol·K); ΔH/R therefore equals 10,000 K, which is exactly why USP chose 83.144[2][5] |
| T1 … Tn | Each recorded temperature converted to kelvin (°C + 273.15) | From your data logger at equal intervals |
| n | Number of readings | All available readings in the period; USP has stated that "all available data must be used"[6] |
| e−ΔH/RT | The Arrhenius factor for each reading; hotter readings produce disproportionately larger factors | Computed per reading |
Two practical consequences follow. First, MKT is always equal to or higher than the arithmetic mean, and the gap widens as the temperature swings get larger. Second, for minor excursions the result is "remarkably insensitive to the value assumed for Ea," so the default 83.144 kJ/mol is defensible unless you have product-specific stability data[2]. The standard formula assumes equally spaced readings; loggers that speed up sampling during an alarm need a time-weighted (trapezoidal) variant[2], which is one reason to keep a fixed logging interval on compliance loggers.
Step-by-step MKT calculation with a worked example
Suppose a CRT store logs every three hours over one day and an air-conditioning fault pushes the afternoon readings up. Here is the calculation done by hand, using ΔH/R = 10,000 K.
| Reading | °C | K | ΔH/RT = 10,000 / K | e−ΔH/RT |
|---|---|---|---|---|
| 1 | 21.0 | 294.15 | 33.9963 | 1.7203 × 10−15 |
| 2 | 21.0 | 294.15 | 33.9963 | 1.7203 × 10−15 |
| 3 | 22.0 | 295.15 | 33.8811 | 1.9303 × 10−15 |
| 4 | 26.0 | 299.15 | 33.4280 | 3.0366 × 10−15 |
| 5 | 29.0 | 302.15 | 33.0961 | 4.2318 × 10−15 |
| 6 | 27.0 | 300.15 | 33.3167 | 3.3943 × 10−15 |
| 7 | 23.0 | 296.15 | 33.7667 | 2.1643 × 10−15 |
| 8 | 21.0 | 294.15 | 33.9963 | 1.7203 × 10−15 |
| Sum of factors | 1.9918 × 10−14 | |||
| Average factor (÷ 8) | 2.4898 × 10−15 | |||
| −ln(average) | 33.6266 | |||
| MKT = 10,000 / 33.6266 − 273.15 | 24.23 °C | |||
| Arithmetic mean for comparison | 23.75 °C |
- Convert each reading to kelvin by adding 273.15.
- Divide 10,000 by each kelvin value (this is ΔH/RT with the USP default).
- Take e to the power of the negative of that number for each reading. In a spreadsheet:
=EXP(-10000/(C2+273.15)). - Average the factors (sum, then divide by n).
- Take the natural log of the average and change its sign:
=-LN(AVERAGE(...)). - Divide 10,000 by that result and subtract 273.15 to get MKT in °C.
The 29 °C spike lifted MKT to 24.23 °C against a plain average of 23.75 °C. Both are below the 25 °C MKT ceiling that USP attaches to controlled room temperature (see next section), so this single day would not, by itself, threaten the product, but it must still be logged and investigated as an excursion. Cloud monitoring platforms run exactly this arithmetic continuously per sensor and per period, which is why the MKT column appears in 21 CFR Part 11-style reports.
Photo: Luke Chesser / Unsplash
MKT vs arithmetic mean: three examples that change the verdict
People sometimes assume the average temperature is "good enough." It is not, because degradation is exponential in temperature. Consider these 24-hour profiles, all with the same arithmetic mean of 25.0 °C:
| Profile (24 h) | Arithmetic mean | MKT (ΔH = 83.144 kJ/mol) | Verdict against CRT MKT ≤ 25 °C |
|---|---|---|---|
| Steady 25 °C all day | 25.0 °C | 25.0 °C | At the limit |
| 12 h at 20 °C, 12 h at 30 °C | 25.0 °C | 26.3 °C | Fails |
| 12 h at 18 °C, 12 h at 32 °C | 25.0 °C | 27.4 °C | Fails badly, and 32 °C is outside the 15–30 °C excursion band anyway |
| 30-day month: 27 days at 24 °C, 3 days at 30 °C | 24.6 °C | 24.8 °C | Passes on MKT, but three days at 30 °C are individual excursions to document |
The lesson: cool nights do not "cancel out" hot afternoons. A store that averages 25 °C by swinging between 18 °C and 32 °C is exposing product to the equivalent of a constant 27.4 °C. That is the whole point of MKT and the reason David Trew's worked climatic-zone example shows an MKT of 25.1 °C being "equivalent to long term stability testing at 25 ± 2 °C" even though ambient ranged from 15 to 30 °C[5].
Where MKT is accepted, and where it is not
Accepted: controlled room temperature and controlled cold temperature, as a storage-condition metric
USP <659> Packaging and Storage Requirements defines controlled room temperature as 20–25 °C with excursions permitted between 15 °C and 30 °C, provided the MKT does not exceed 25 °C, and controlled cold temperature as 2–8 °C with an MKT not exceeding 8 °C and transient spikes up to 15 °C allowed under specified conditions[7]. USP <1079> states that excursions outside labeled storage conditions "for brief periods, may be acceptable provided that stability data and scientific/technical justification exist"[8]. MKT is the accepted way to express that cumulative exposure for CRT stock and, within its stated limits, for 2–8 °C stock. USP's dedicated chapter, <1079.2> Mean Kinetic Temperature in the Evaluation of Temperature Excursions During Storage and Transportation of Drug Products, was revised in USP–NF 2025 Issue 2 and became official on 1 August 2025; it describes MKT calculation windows for CRT and controlled cold temperature and is written "as a global document"[6][9][10].
Not accepted: as a way to hide excursions or a system out of control
- Repeated excursions. USP's summary of comments on <1079.2> is blunt: "Repeated deviations indicate a lack of system control"[6]. As one compliance explainer puts it, "MKT cannot be used to justify poor control"[9].
- Retroactive normalization. "You cannot justify an out-of-spec event by showing that the temperature later dropped back within limits. Product degradation is cumulative and not reversible"[9]. Every excursion is still logged, investigated and closed with CAPA; MKT informs the impact assessment, it does not replace it. See our guide on what to document when a temperature excursion happens.
- MKT alone. The European Compliance Academy's reading of the 2023 revision draft: "MKT should not be abused, as MKT alone is not enough to assess the impact of a temperature excursion"; you also need the product's stability data, the excursion's magnitude and duration, and its history[11].
- Freeze events and freeze-sensitive products. MKT models Arrhenius chemical kinetics. It says nothing about physical damage. WHO ranks freeze-sensitive vaccines such as hepatitis B, DTP-containing vaccines and IPV as irreversibly damaged by freezing at or below 0 °C[12]. A 2–8 °C unit that dipped to −2 °C for an hour will show a comfortable MKT and a ruined product. Treat any freeze event as a product-specific decision with the manufacturer, never an MKT calculation.
- Frozen and ultra-low storage, biologics. The default 83.144 kJ/mol was chosen as representative of common organic reactions in small-molecule products[2]. Proteins, cell-based products and mRNA formulations degrade by mechanisms (aggregation, denaturation, phase change on thawing) that Arrhenius averaging does not capture. Use manufacturer stability statements and time-out-of-refrigeration budgets instead.
Regulatory reference table
| Document | What it says about MKT |
|---|---|
| ICH Q1A(R2) Stability Testing (2003), glossary | Defines MKT; notes it exceeds the arithmetic mean; cites Haynes formula; used to derive climatic zones[1] |
| USP <1150> Pharmaceutical Stability | Formal definition and formula; default ΔH 83.144 kJ/mol[2] |
| USP <659> Packaging and Storage Requirements | CRT 20–25 °C, excursions 15–30 °C, MKT ≤ 25 °C; controlled cold 2–8 °C, MKT ≤ 8 °C, spikes ≤ 15 °C[7] |
| USP <1079> (2020 rev.) | Brief excursions may be acceptable with stability data and scientific justification[8] |
| USP <1079.2> (official 1 Aug 2025) | How to use MKT for excursions in storage and transport; all data must be used; repeated deviations mean lack of control[6][9][10] |
How monitoring software computes MKT (and what to check)
A cloud monitoring platform computes MKT the same way you did by hand, but continuously and per sensor. Points worth verifying with any vendor, including us:
- Interval and completeness. The formula assumes evenly spaced readings and USP wants all data used[6]. Ask what happens if the network drops: loggers should buffer locally and back-fill so the MKT window has no holes.
- Window definition. MKT is meaningless without a period. Reports should state the window (a day, a month, a shipment, a batch's residence time) and let QA choose it.
- ΔH setting. 83.144 kJ/mol default, with the ability to enter a product-specific value if your stability team has one[2].
- Traceability. The MKT figure must sit in a report with an audit trail, so an inspector can see the raw readings behind it. This is where 21 CFR Part 11 features matter; see understanding FDA 21 CFR Part 11 compliance.
- Alarms are separate. MKT is a retrospective statistic. Real-time high/low alarms with escalation are what stop the excursion in the first place, as we argue in why real-time monitoring is essential for pharmaceutical storage.
Photo: Stephen Dawson / Unsplash
How Ideabytes IoT helps: MKT built into every 21 CFR Part 11 report
Ideabytes IoT loggers report to the CMARS21 cloud platform, and the "MKT Value" is a standard field in the 21 CFR Part 11-compliant PDF reports alongside audit trail, alarm logs, trend graphs and CAPA notes. Scheduled and on-demand reports can be exported as PDF, CSV or Excel, so QA can attach the MKT for a zone or a month directly to an excursion investigation.
- IBI-WT120 Wi-Fi temperature data logger: −50 to +120 °C, ±0.5 °C, 15-minute default polling (configurable), 90 days of local storage and up to 20 hours of battery backup, so the readings behind your MKT survive a Wi-Fi or power outage.
- IBI-MT120 4G M2M temperature data logger: the same logger with a built-in M2M SIM and 2G fallback for warehouses, pharmacies and depots without site Wi-Fi.
- IBI-WTR8 Wi-Fi 8-channel PT100 logger: eight probes across −200 to +550 °C for cold rooms, stability chambers and mapping grids, with MKT, audit trail and alarm logs in its reports; see how it fits a mapping study in our pharmaceutical warehouse temperature mapping guide.
- IBI-UT70 USB temperature data logger: a coin-cell shipment logger (−30 to +70 °C, up to 90 days at 10-minute intervals, 14,400 readings) that generates an encrypted PDF on plug-in with no software, giving you the reading set for a shipment MKT calculation on receipt.
Loggers start from $99; software features, including MKT and Part 11 reporting, are provided in different plans, so contact sales for details.
Frequently asked questions
What is the difference between mean kinetic temperature and average temperature?
The average weights every reading equally. MKT weights hot readings exponentially more, following the Arrhenius equation, so it is always equal to or higher than the average and reflects true cumulative degradation[1]. A day split between 20 °C and 30 °C averages 25 °C but has an MKT of about 26.3 °C.
What ΔH value should I use in the MKT formula?
USP <1150> specifies 83.144 kJ/mol unless more accurate experimental data exist; with R = 8.3144 J/(mol·K) this makes ΔH/R = 10,000 K, which simplifies the calculation[2][5].
Can MKT justify a temperature excursion in a 2–8 °C refrigerator?
Only within the limits USP sets for controlled cold temperature (MKT ≤ 8 °C, transient spikes ≤ 15 °C)[7] and only for a one-off event with stability support[8]. Repeated excursions or any freeze event are outside what MKT can excuse[6][12].
Is MKT used for vaccines and biologics?
Rarely as the deciding metric. Freeze damage and protein degradation are not Arrhenius processes, so follow the manufacturer's stability guidance and, for vaccines, the relevant national storage program's excursion procedure[12].
How is MKT calculated in Excel?
Put temperatures in column C, use =EXP(-10000/(C2+273.15)) in column D, then =10000/(-LN(AVERAGE(D:D)))-273.15. That reproduces the USP <1150> formula with the default ΔH.
Conclusion
MKT turns a month of fluctuating temperature readings into one number that maps directly onto USP storage definitions and stability data. Used correctly, for CRT and controlled cold storage, on complete evenly logged data, and never to explain away repeated excursions, freeze events or biologic damage, it saves batches that a naive "it went out of range" reading would have discarded, and it condemns storage that a naive average would have passed. If you want MKT computed automatically per zone with an audit trail behind it, talk to Ideabytes IoT about CMARS21 and the loggers above.
References
- ICH. Q1A(R2) Stability Testing of New Drug Substances and Products, Glossary: "Mean kinetic temperature." Feb 2003. database.ich.org (PDF)
- Tong C, Lock A. "A Computational Procedure for Mean Kinetic Temperature Using Unequally Spaced Data." JSM 2015 Proceedings, Biopharmaceutical Section, American Statistical Association (quotes USP <1150> definition and default ΔH). amstat.org (PDF)
- Air Cargo News. "Failures in temperature-controlled logistics cost biopharma industry billions" (IQVIA Institute / Peli BioThermal survey), 26 Jul 2019. aircargonews.net
- Haynes JD. "Worldwide virtual temperatures for product stability testing." Journal of Pharmaceutical Sciences 1971;60(6):927–929. doi.org/10.1002/jps.2600600629
- Trew D. "Application of the Mean Kinetic Temperature Concept." David Trew Consulting. articles.davidtrew.co.uk
- ECA Academy. "USP Publishes Summary of Comments on Chapter <1079.2> Regarding Mean Kinetic Temperature (MKT)." GMP News, 11 Mar 2025. gmp-compliance.org
- United States Pharmacopeia. General Chapter <659> Packaging and Storage Requirements (definitions of controlled room temperature and controlled cold temperature). USP–NF Online. uspnf.com
- United States Pharmacopeia. General Chapter <1079> Risks and Mitigation Strategies for the Storage and Transportation of Finished Drug Products (APEC toolkit reprint). usp.org (PDF)
- Konradsen J. "USP <1079.2> explained: 7 updates for excursion evaluation in GxP." Eupry, updated 8 Oct 2025. eupry.com
- United States Pharmacopeia. General Chapter <1079.2> Mean Kinetic Temperature in the Evaluation of Temperature Excursions During Storage and Transportation of Drug Products (APEC toolkit reprint). usp.org (PDF)
- ECA Academy. "USP pre-posts Revision to General Chapter <1079.2> and Stimuli Article on Mean Kinetic Temperature (MKT)." GMP News, 21 Feb 2023. gmp-compliance.org
- World Health Organization. Temperature sensitivity of vaccines (WHO/IVB/06.10), 2006. iris.who.int