Validation and verification are different activities
Validation asks whether the control, operated as specified, is capable of achieving the intended effect. It is a scientific and technical question answered before or early in operation, and revisited when something changes.
Verification asks whether the control is actually operating as specified. Monitoring records, calibration, record review and periodic testing are verification. A plan can be perfectly verified and still invalid — every log completed correctly against a critical limit that was never adequate in the first place.
Confusing the two produces a specific and recognisable failure: a facility with immaculate records that cannot explain why its cook step delivers the reduction it claims.
Acceptable bases for a critical limit
A critical limit needs a source you can put in front of a reviewer. In practice, the defensible options are limited.
- A regulatory requirement or performance standard that specifies the parameter directly
- Published FDA or USDA guidance — for example the FDA Fish and Fishery Products Hazards and Controls Guidance for parasite destruction freezing parameters
- A letter or scheduled process from a recognised process authority, particularly for acidified and low-acid canned foods
- Peer-reviewed scientific literature applicable to your product matrix, not merely to a similar-sounding product
- A challenge study or validation study conducted on your product, by a competent laboratory, under conditions representative of production
Why the product matrix matters
Thermal and chemical inactivation are strongly influenced by the food itself: water activity, fat content, pH, salt, protective solutes and particle size all change the outcome. A time and temperature combination validated for a lean, high-moisture product may be inadequate for a high-fat or low-water-activity version of the same recipe.
This is why borrowing a critical limit from a competitor's plan, an equipment manufacturer's brochure, or a general industry rule of thumb is risky. The number may be right. The problem is that you cannot demonstrate it is right for what you actually make.
Practical criteria for the limit you choose
Beyond scientific adequacy, a critical limit has to be operable. It must be measurable in real time or near-real time by the person responsible, with an instrument that can be calibrated, at a frequency the operation can sustain.
A critical limit that requires a laboratory result available two days later is not functioning as a critical limit — it is a verification activity. Where the only true measure of a hazard is a lagging one, the control usually needs to be built around a measurable surrogate parameter that has itself been validated against the outcome.
Finally, build in an operating margin. Running the process at the critical limit guarantees regular deviations. Setting an operating target with headroom above the critical limit means normal variation does not generate a corrective action every shift, which in turn keeps the corrective action record meaningful.
Primary sources
Go to the source rather than relying on this summary. Regulation text and agency guidance are the authority; this page is an interpretation of them.
- FDA Fish and Fishery Products Hazards and Controls Guidance
- 21 CFR Part 117 Subpart C — validation and verification of preventive controls
- 21 CFR Part 113 and Part 114 — scheduled processes for low-acid canned and acidified foods
- NACMCF HACCP principles and application guidelines