Cold-press juice plant: categorising control measures and proving validation
A cold-press juice and smoothie manufacturer (63 employees, two lines, one high-care filling room) has held ISO 22000:2018 certification for four years. Its hazard analysis for the green smoothie line identifies Listeria monocytogenes in raw leafy greens, Salmonella in fresh fruit, and undeclared milk carried over from a shared filler previously used for a yoghurt drink. The hazard control plan contains three entries: wash water at 80–100 ppm free chlorine, recorded every 2 hours, categorised as an OPRP; a changeover clean of the filler, verified by ATP swabs, categorised as a PRP and therefore excluded from the hazard control plan; and metal detection at 2.5 mm Fe, categorised as a CCP. The technical manager says categorisation was done 'with a decision tree in 2019 and nothing has changed'. Validation is a single 2018 trial report supplied by the wash-water chemical vendor. Verification is limited to chlorine logs and daily ATP results. No environmental Listeria monitoring exists; swabs are taken only after cleaning. Internal audit covers the line twice a year but has never sampled changeover records.
1. Evaluate the categorisation of the three control measures and identify what the auditor should conclude.
Read the worked answer
Work from clause 8.5.1.4, not from the organisation's decision tree. The standard requires a systematic approach, documented justification and consistency with the CCP/OPRP definitions; it does not require or favour a decision tree. Test each measure against the definitions rather than against tradition.
Metal detection. A CCP exists only for a hazard that the hazard analysis has identified as significant. Here metal is nowhere in the hazard list, yet a CCP with critical limits has been written for it. Either the hazard identification (8.5.1.2) is incomplete - no equipment wear, blade, sieve or glass hazard considered - or the control is miscategorised. Either way an inconsistency exists between the hazard analysis and the hazard control plan, and the documented justification is missing.
Allergen changeover clean. Undeclared milk is a significant hazard, and this clean is the only barrier for the filler. A measure that prevents or reduces a significant hazard to an acceptable level belongs in the hazard control plan as a CCP or OPRP, with action criteria (ATP pass/fail) and defined corrections if it fails. Calling it a PRP removes it from monitoring, correction and verification requirements and is not defensible.
Wash water. An OPRP with numeric action criteria is plausible and can be accepted in principle, but only if it is validated and monitored in a way that detects loss of control.
Conclusion: at minimum one nonconformity against 8.5.1.4 for categorisation not based on a systematic, documented hazard assessment, plus a related nonconformity for the hazard control plan content (8.5.3).
2. The technical manager claims 'we've validated the chlorine wash - the chemical supplier sent us a trial report'. How should the auditor test validation (8.5.2) and verification (8.8), and what records would be requested?
Read the worked answer
Separate the two concepts explicitly, because the client has merged them.
Validation (8.5.2) answers 'will this combination of control measures actually achieve the acceptable level?'. It is done before implementation and after any change, and must cover the actual conditions of use. A vendor trial report is supporting evidence, not validation, unless it covers the plant's real matrix, water temperature, organic load, contact time and the pathogen of concern. Ask for: the validation protocol and acceptance criteria (e.g. target log reduction for the relevant hazard), the data set, who approved it, and evidence it was repeated after any change of chemical, dose controller or produce supplier. A 2018 report for a process that has since changed does not validate the current process.
Verification (8.8) answers 'is the validated system working as intended?'. Evidence expected: calibration of chlorine and ATP meters (8.7), review of monitoring results against action criteria (9.1), product/environmental testing, internal audit of the line, and analysis of trends and complaints. Note that environmental Listeria swabbing after cleaning only is monitoring of cleaning, not verification of sustained control - ask what happens between cleans and at what frequency.
Auditor judgement: classify the gap in validation as a nonconformity against 8.5.2 only if no adequate validation exists for an identified significant hazard; if a partial study exists but does not cover current conditions, that is still a nonconformity because the requirement is to validate before implementation. Record precisely which controls, which hazard and which missing element - 'validation documents not seen' is not a finding.