Critical limits define the conditions separating control from loss of control at a critical control point (CCP). They translate a food safety requirement into a measured or observed condition that personnel can assess during production. Each limit needs evidence that the selected control can prevent or eliminate the identified hazard, or reduce it to the specified acceptable level, under the conditions in which the limit will be used.
The Hazard Analysis and Critical Control Point (HACCP) team develops the limits, validates each CCP and its limits, and maintains the approved criteria and supporting evidence. These are connected activities with different outputs. A proposed value needs technical substantiation; a validated arrangement needs controlled implementation; an active limit needs continuing confirmation that its basis still applies. Reliable monitoring can then establish whether the required conditions were achieved during a particular production period.
Defining critical limits
Defining a critical limit establishes the complete acceptance condition for a designated CCP. The HACCP team is responsible for linking the significant hazard and required outcome to the parameters that demonstrate control. The result should be a candidate criterion with a defined scope and technical rationale, ready for validation. Completing a cell in the HACCP plan does not establish that the proposed condition is safe or ready for use.
Starting from the hazard and the actual process
Begin with the significant hazard, the selected control measure and the reason the step is a CCP. Identify the outcome required: prevention, elimination or reduction to a specified acceptable level. The acceptable hazard level concerns the food. A critical limit concerns the condition used to achieve that outcome, such as product exposure to a treatment or demonstrated separation performance. A finished-product specification may define that outcome without providing information soon enough to decide whether the process is under control.
Obtain the current product description, intended use and consumer information, verified process flow and relevant specifications. Identify the products, recipes, lines, equipment and operating range covered. Examine formulation tolerances, unit dimensions, packaging, loading, throughput and storage or distribution conditions where they affect the control. Resolve conflicting or missing information before treating the criterion as complete. Use the guidance on determining critical control points for the preceding designation assessment; defining a limit does not replace it.
Group products under one limit only where evidence supports the grouping. A shared commercial name or production line does not establish equivalence. Identify the differences affecting the mechanism and justify a representative or challenging product. Where a required safety control cannot be provided, return the product or process for modification, select another supported control or restrict the intended scope. A nominal limit cannot remedy the absence of an effective control measure.
The HACCP team leader is responsible for coordinating input from product and process specialists, operations, engineering and the function responsible for measurement control. Obtain specialist explanations of hazard behaviour and formulation effects, operational information on actual variation and access for monitoring, and engineering information on equipment behaviour, configurations and failure conditions. The measurement-control function is responsible for assessing how reliably the proposed condition can be determined. General HACCP knowledge may need support from a specialist in the particular process or hazard.
Selecting all the conditions needed for control
Identify how the control achieves its effect before choosing its parameters. Heating can reduce a target microorganism; a time and temperature restriction can limit growth; formulation can restrict growth through acidity or water availability; separation can remove a defined physical contaminant. These outcomes differ. A growth-inhibiting condition does not by itself demonstrate destruction, and evidence for a removal device establishes its capability for particular contaminants and conditions.
For a thermal control, assess the exposure received by the least adequately treated product. An equipment temperature reading may give no indication of a cold product centre, excessive product thickness or insufficient residence time. Establish through validation how the monitored condition relates to the product condition. For continuous processing, the shortest time in which product could realistically pass through the process may determine minimum exposure. Where capability depends on load, flow, product dimensions or equipment configuration, retain those conditions in the controlled specification and identify which require critical limits themselves.
Acidification needs the relevant product portion to reach its required condition within the supported time. A liquid sample may not establish the condition inside solid pieces. Choose the sampling location, preparation method and timing to examine the condition on which safety depends. Water activity concerns the water available for microbial behaviour and differs from total moisture content. Moisture percentage or formulation can serve as a substitute measurement only where a supported relationship has been established for that product. The organism, formulation, temperature and subsequent storage conditions affect the conclusion.
For a detector or separation step, define the demonstrated performance and conditions of use. A detector response to a specified test object does not establish absence of every metal fragment. Contaminant material and orientation, product effect, pack format, aperture and speed can affect detection or rejection. Ensure the criterion and supporting process specification address the operation of the complete detection and rejection arrangement and the passage of all intended product through it. An unsupported statement such as “no foreign bodies” gives the monitor no usable test of the actual control.
Several parameters may jointly determine control. A time and temperature combination requires both conditions, at the relevant location and with a defined start of timing. Where alternative schedules are validated, identify each complete combination and how personnel select and record it. Values from different schedules cannot be combined without supporting assessment. An average is unsuitable where the validated criterion requires every relevant portion to achieve a minimum exposure or stay below a maximum condition.
Establishing the technical basis and applicable requirements
The HACCP team is responsible for obtaining credible evidence appropriate to the control. This may include technical publications, authoritative guidance, applicable processing requirements, studies by competent specialists or equipment manufacturers, and relevant evidence from the site's operations. Obtain enough of the original information to understand its methods, results and conditions. A brochure, source title or search summary may leave important assumptions unexamined.
Compare the evidence with the proposed use. Identify the hazard or organism, required reduction or prevention, product matrix, initial conditions, treatment and all factors essential to the result. A reported value cannot be transferred safely by ignoring the conditions under which it worked. Historical records can demonstrate that the plant repeatedly achieved particular conditions; a separate technical basis is needed to establish that those conditions control the hazard.
Where sources suggest different values, compare their endpoints, assumptions and operating conditions before choosing. They may answer different questions. Where the difference could affect safety, the team leader is responsible for obtaining competent technical interpretation and recording the basis for selecting the condition. Choosing the least demanding reported value without that comparison leaves the decision unsupported.
Have a person competent to interpret compliance requirements establish which legally prescribed processing conditions or customer requirements apply to the product, activity and market. This assessment should examine the current provision, its effective date and permitted alternatives. Scientific capability does not authorise departure from a binding condition. A customer's quality requirement may be stricter than the food safety criterion without automatically becoming the CCP critical limit. Keep these requirements distinguishable and coordinate them in the process specification.
Assess feasibility alongside the safety requirement. More severe processing can harm quality or introduce another risk, while inadequate equipment capability can make a sound limit impracticable. The response is a supported change to capability, product, process or scope. Production convenience and commercial timing do not justify weakening the required safety outcome.
Writing a criterion that can be applied consistently
State the parameter and unit, its minimum, maximum or permitted range, and the measurement location and timing. Define all combined conditions and identify the products and processes covered. Specify whether equality is acceptable and how corrections, measurement uncertainty and display precision affect the decision. Identify excluded conditions where personnel might otherwise extend the criterion beyond its evidence.
Use measurable terms wherever possible. Where an observable or subjective assessment is necessary, provide a precise description and controlled examples of acceptable, unacceptable and borderline conditions. Specify examination conditions such as lighting, access or a reference standard when they affect classification. Establish separately that the observed characteristic represents hazard control. Clearer photographs or more training cannot create a technical relationship unsupported by evidence.
Ask personnel who will use the instruction to apply it to representative observations and boundary results. Have them demonstrate the method and explain when they would report a failure or uncertain result. Compare their decisions with the intended criterion. If classifications differ, examine whether the inconsistency arises from the instruction, reference examples, examination conditions, access or competence. If the method cannot distinguish control reliably, improve the method or redesign the control. Specify whom monitors should refer ambiguous observations to and how they should protect potentially affected product.
The method must provide information in time for the required action. Delayed microbiological testing can support validation or verification while failing to supply the timely production decision required at a CCP. Examine the time between sampling, result and possible intervention, including the product produced or moved during that period. Where compliance cannot be established in time, the arrangement needs a suitable control and defined restrictions or holding conditions. Detection of a problem after release does not itself control the process.
Accounting for measurement uncertainty
The measurement-control specialist and HACCP team are responsible for specifying the actual quantity to be measured, where and when it is measured, and the method used. Examine range, resolution, response time, calibration corrections and uncertainty at the operating point. Measurement uncertainty describes the quantified doubt associated with a result on a stated basis. It differs from a correction for a known bias and from the actual variation between products. A satisfactory calibration result establishes conformity with the stated checks and tolerances. The installed method also depends on sensor placement, sampling, product variation and delay. A precise instrument at an irrelevant location can give an inadequate answer.
For a minimum limit, a reading that overstates the true value of the product quantity being measured can lead to false acceptance. For a maximum limit, a reading that understates it can do the same. Apply known corrections in their documented direction. Agree the decision rule before measurements are interpreted, meeting any prescribed legal or customer rule. A guard band can make the measurement acceptance threshold more demanding to account for uncertainty; its selection needs a justified basis. Avoid adding another margin automatically where the prescribed rule already addresses it.
For illustration only, suppose the validated minimum for a hypothetical quantity is 100 units. The corrected result is 100.4 units and the method's expanded measurement uncertainty is 0.8 units on its stated basis. Under an approved rule requiring the corrected result minus that uncertainty to be at least 100 units, the comparison is 99.6 against 100. Compliance has not been demonstrated. A corrected result of 100.8 would meet that comparison at the stated precision. This arithmetic illustrates a decision rule, rather than a food processing value, universal margin or guarantee that the true value lies within an absolute bound. Any required time, location and other parameters still need their own assessment.
State the correction and resulting acceptance threshold explicitly in the monitor's instruction. Define the action for a result that leaves compliance unproven. Do not choose a more favourable uncertainty interpretation after seeing the result, round a failing value into acceptance or combine separate results through an unvalidated average. If the method cannot resolve the decision reliably, improve it, use an assessed margin or change the control. The wider arrangements for maintaining instrument suitability are developed in the guidance on calibration and control of measuring and monitoring devices.
Separating operating targets from the critical boundary
The normal operating target is the condition that production personnel aim to achieve. A warning or action threshold can prompt intervention before a critical-limit failure. Keep these values distinguishable from the safety boundary and from any measurement acceptance threshold. A controller set point describes an instruction to equipment; it does not establish the condition actually received by the product.
Select the operating margin using observed process variation, measurement capability, intervention delay and the consequences of more severe treatment. An arbitrary percentage or the instrument's display resolution alone does not establish an adequate margin. Operations and engineering are responsible for assessing whether the process can remain sufficiently stable within the conditions supported by validation and whether adjustments can take effect before failure. If that capability is lacking, improve the arrangement rather than redefining safety around its current variability.
Record the candidate limit, scope, rationale, proposed measurement rule and outstanding validation questions in the HACCP documentation. The candidate remains a proposal until validation establishes capability and controlled implementation establishes readiness for use. Revise the definition if validation or preparation for implementation exposes a material gap.
Validating CCPs and critical limits
Validation establishes that each selected CCP and its critical limits can consistently control the relevant hazard to the specified acceptable level under the claimed conditions. The HACCP team is responsible for defining the validation question, obtaining appropriate evidence and recording the conclusion it supports. The work combines scientific justification of the control with evidence that the site's arrangement delivers its required conditions. A completed study may support acceptance, a limited scope, further investigation or rejection.
Defining the question and choosing the evidence
Identify the CCP, significant hazard, required outcome, candidate limits and proposed range of operating conditions. List the conditions on which the control depends and the production variants the conclusion is intended to cover. Establish the acceptance criteria before collecting results. The relevant specialist is responsible for determining what evidence can demonstrate the hazard-control effect and which measurements can establish whether the required conditions are achieved at the site.
Review existing evidence first. Credible literature, authoritative guidance or a third-party study can establish a suitable basis without commissioning a new study of an already demonstrated mechanism. Compare the product, organism, process, equipment and critical conditions with the site's proposal. Obtain the original methods and findings where an abstract or summary leaves material uncertainty. Record why relevant differences do not undermine the conclusion or identify the additional work needed.
Scientific support and plant performance answer distinct questions. A study may establish that a particular treatment controls the hazard while leaving unanswered whether the least adequately treated product on this line receives that treatment. Conversely, consistent plant readings do not establish the hazard-control effect if their safety basis is missing. Choose further work to answer the unresolved question. The whole-plan validation and verification arrangements place this CCP evidence within the broader HACCP system.
Select specialist methods according to the question. Heat-distribution and product heat-penetration studies can examine different parts of a thermal process. A controlled trial may demonstrate that conditions supported by existing evidence are actually achieved. A challenge study may be needed to establish a biological effect. Predictive models can inform the assessment within their validated domain, with the organism, product assumptions, input ranges and endpoint identified. Predictions outside that domain require competent assessment and additional support.
A laboratory result or model output must be interpreted through its design. A negative product sample does not establish universal absence or prove adequate treatment of every unit. A convenient indicator or surrogate does not automatically represent the target hazard. The specialist is responsible for justifying the chosen organism or surrogate, method and endpoint, and explaining the conclusion that the results can support. Microbial challenge work needs competent specialist facilities and controls so the experiment does not introduce a hazard into saleable production.
Planning representative studies and providing resources
Prepare a protocol stating the question, pre-established criteria, products and recipes, equipment configuration, proposed range and conditions to be challenged. Identify measurement locations, instruments, recording intervals, analytical methods and how data will be evaluated. Explain why the selected trials address the important variation and limitations. There is no universal replicate count or study duration that demonstrates every CCP's capability.
Select challenging conditions by examining how the control works. Product size, load, initial condition, formulation or maximum throughput may determine the least favourable treatment, but do not assume which factor is worst without technical reasoning. Establish the least adequately treated location or fastest passage where relevant. Cover relevant differences in equipment and production conditions. One successful trial on an easy product or lightly loaded shift supports only a limited conclusion.
The HACCP team is responsible for identifying the expertise needed and commissioning specialist work where the necessary competence is unavailable within the team. Agree the scope, methods, expected report, access to raw data and assumptions, delivery date and treatment of limitations. Assess the provider's relevant competence and any required accreditation scope. A general certificate or an equipment supplier's assurance does not establish capability for the particular study. Maintain enough internal understanding to evaluate the delivered advice and recognise a later change affecting it.
Secure plant access, appropriate product, calibrated instruments, engineering assistance, analytical facilities and time for evaluation. Verify that the available sensors and data intervals can resolve the condition being examined. Identify devices, installed locations and data channels so results can be traced to their measurement. Synchronise time records where timing affects the comparison. Frequent data logging cannot remedy a sensor at an unrepresentative location.
Plan how trial product will be identified, segregated and assessed or disposed of. Production and engineering are responsible for confirming the study configuration and how the line will return to its approved operating condition. Conduct a proposed safety-affecting change as a controlled trial where plant data are needed; do not treat its output as accepted routine commercial production before capability and product status have been established. Existing saleable production remains under its current approved controls.
Senior management is responsible for providing resources and resolving competing production or development priorities. If specialist support or plant access is delayed, identify the proposed use that remains unsupported. Maintain the existing validated arrangement or restrict the affected activity. Budget approval, an order placed or a study date booked does not complete the required technical work.
Conducting the study and preserving actual results
Before the study begins, the assigned study coordinator is responsible for checking readiness against the protocol and confirming that personnel can perform the required measurements and record the conditions. During execution, record actual product and process conditions, dates, times, configuration and results. Retain unsuccessful trials and departures from the planned method. Preserve the original data, calibration information, calculations and corrections in readable form.
Compare the conditions actually achieved with every factor necessary in the supporting scientific evidence. A favourable endpoint does not automatically overcome a trial departure that makes the result unrepresentative. The study specialist is responsible for identifying which observations remain interpretable, explaining any data exclusions and specifying what must be repeated. Keep the exclusion and its technical justification visible. A final average or favourable graph must not conceal individual failures relevant to safety.
Where the plant cannot deliver the required conditions, examine equipment capability, execution, the proposed criterion and study design. These require different responses. Correcting a measurement error may justify repeating the study, while a genuine capability failure requires process change, another supported control or narrower scope. Preserve the original result and the reasoning for subsequent work.
Evaluating and accepting the conclusion
The technical reviewer is responsible for examining the protocol, actual configuration, measurement capability, raw results, calculations and departures, comparing them with the pre-established criteria and scientific evidence. This review includes assessing uncertainty, unexplained variation, excluded products and untested conditions that could change the judgement. Arrange competent challenge proportionate to technical complexity and consequence, including external expertise when needed. The reviewer must be able to require more work and reject an unsupported conclusion.
The report should state which products, processes and operating conditions the findings support, the conditions essential to achieving control, and the conclusion for each material question. Distinguish evidence that the control achieves the required effect on the hazard from evidence that the necessary operating conditions are achieved at the site. Explain how the limitations affect permitted operation: identify the product, line, range or variant excluded and prevent its use under the approval. A conditional conclusion is usable only where its conditions can be maintained. Safety-critical uncertainty remains unresolved work and prevents acceptance for the unsupported use.
The HACCP team is responsible for accepting the relevance of commissioned evidence. The designated technical authority is responsible for deciding and recording whether that evidence demonstrates the required hazard control under the proposed conditions. Preserve approval identity, date, accepted version, supporting material and required conditions. Managerial approval may authorise implementation within the site's assigned authority; it cannot create missing scientific support or waive a binding requirement. Record who has authority to approve implementation, accept operational readiness and decide the disposition of affected product. Keep these decisions distinct even when one person holds more than one role.
An accepted conclusion provides the basis for the final monitoring and response instructions. The feasibility of measurement and the study's measurement arrangements are addressed earlier, so the team may need to revisit design and validation as the work progresses. The approved production procedure nevertheless relies on validated controls and limits. If the evidence requires a changed parameter or monitoring relationship, revise the candidate and reassess the affected capability before use.
Responding to failed validation and renewed questions
When validation criteria are not met or evidence remains insufficient, the HACCP team is responsible for recording what was established, which proposed uses remain unsupported and what further decision is needed. Options include obtaining missing information, improving the study, redesigning the control, narrowing the permitted operating range or rejecting the proposed use. A study can be complete with a conclusion that the proposed control failed the criteria; a signed report does not make that control acceptable.
Escalate unresolved technical questions to someone competent to answer them and resource or priority constraints to management able to resolve them. Explain which products and processes are affected, the hazard, available evidence and uncertainty, feasible options, resources needed and the deadline for a useful decision. The study coordinator is responsible for confirming that the recipient has received the referral and following it through until the recipient accepts the assigned work. Keep dependent production restricted while necessary capability remains unsupported.
Revisit validation when a safety-affecting change or contrary evidence from operations challenges its assumptions. A new formulation, increased throughput or replacement component may require additional evidence even if the nominal limit remains unchanged. Compare the specific change with the conditions covered by the accepted validation through change assessment and revalidation. A routine annual review does not require repeating every study without a technical question, and it does not justify deferring a material current question until the next meeting.
Maintaining approved limits and validation evidence
Maintaining limits keeps the approved acceptance condition, the process in use and its supporting evidence aligned. The HACCP team retains technical ownership. Operations is responsible for applying the approved arrangement. Document control supports the issue of approved instructions, while record control supports the protection and retrieval of evidence. This continuing activity begins with implementation of an accepted limit and persists through routine use, review, change and retirement.
Authorising and introducing the approved version
Before issue, the person responsible for implementation compares the proposed production instruction with the technically accepted limit and validation conclusion. Confirm the covered products, line, complete schedule, measurement rule and essential supporting conditions. That person is also responsible for checking that the required equipment and instruments are available and suitable for their tasks, and obtaining the approvals specified in the site's arrangements, including senior-management approval where reserved to it.
Reconcile the HACCP plan, work instructions, monitoring forms, reference examples, controller settings, warning thresholds and relevant software configuration. Inspect the actual settings and copies at points of use, including other affected lines and shifts. A document issue record alone does not establish that a controller or local instruction changed. Resolve discrepancies before introducing the revised arrangement.
Supervisors are responsible for arranging instruction, practical demonstration and assessment for monitors and staff implementing the response. Include absence cover, temporary personnel and relevant language or access needs. Ask personnel to measure or classify representative conditions and explain the action for warning, failure and unproven compliance. Resolve unclear instructions or gaps in competence before unsupervised work. Training attendance establishes delivery of instruction; observed application establishes more about ability to perform the task.
Identify the effective date and production boundary, such as the first batch using the new version. If old and new arrangements coexist, distinguish their products, equipment and records explicitly. Withdraw obsolete operational copies at the appropriate point and retain them as historical evidence. A limit may be technically approved yet await implementation readiness; record it as such until the required conditions are established.
The site's allocation of responsibilities should include deputies for technical coordination, monitoring supervision and urgent decisions. Confirm their competence, availability, authority limits and record access. Before planned absence, transfer active issues, pending changes, study findings and due reviews. Where necessary competence or authority cannot be obtained, restrict the dependent activity and use the authorised escalation route. Appointing a deputy does not by itself establish that the person can make every technical or product decision.
Applying the limit and distinguishing the response
Monitors are responsible for using the active criterion at the specified location, time and frequency, recording the actual date, time, result and relevant product and line identity as the work occurs. Include the responsible monitor's signature or secure electronic equivalent in each record. Where checking and verification are required, they must be performed by a competent authorised person. Electronic records also retain evidence of that checking and verification. The monitoring-system guidance develops the wider design and recording arrangements.
When a result approaches the operating warning threshold, follow the planned intervention before failure. A result that fails the critical-limit decision rule calls for the defined deviation response. A missing, unreliable or ambiguous result leaves compliance unproven and requires protection and assessment of affected production. Define each response and the authority for further action. A satisfactory current result cannot establish a condition that was not reliably measured earlier.
Use a monitoring method and frequency capable of supporting those decisions. Continuous information can reveal events between checks; discontinuous measurements need representative selection and timing sufficient to identify and control potentially affected product. Establish the arrangement for the actual hazard, process variation and intervention time. Increasing check frequency alone cannot make an unsuitable method or unsupported limit valid.
The HACCP team is responsible for maintaining documented actions for limit failure and trends towards loss of control. Operations is responsible for initiating the prescribed correction, stop or restriction, identifying and protecting product, and informing the authorised assessor responsible for technical or product-control decisions. The corrective-action-plan guidance develops the full response. Protect product promptly while technical investigation and management escalation continue in parallel.
Controlling affected product and restoring capability
For a failed or unproven result, establish the potentially affected period from reliable monitoring evidence, event timing, product flow and traceability. Include product since the last demonstrated control where the failure could have occurred unnoticed, and expand the scope when the evidence is unreliable. Keep that product identified and on hold pending an authorised assessment. If released product may be unsafe, activate incident, withdrawal and recall arrangements promptly.
The authorised product assessor is responsible for deciding what evidence is needed to justify release, reprocessing, another permitted disposition or disposal. Restoring a setting or obtaining a passing repeat result does not prove earlier product complied. A repeat measurement may investigate an anomaly, but preserve both results and state what each establishes. Approval of a new limit cannot retrospectively change the criterion or release held product.
When an instrument fails calibration or its suitability is questioned, the measurement specialist is responsible for examining the magnitude and direction of error, the last reliable check, relevant uses and other evidence. The specialist determines which earlier acceptance decisions may have been unreliable and supplies that assessment to the authority responsible for affected product. A device now passing a check does not establish that its earlier readings were valid.
Use an alternative monitoring or control method only where evidence supports its capability for the products and conditions concerned and its use is authorised. Define its scope, owner, required competence, resources, operating conditions, monitoring, duration, review date and stop criteria. An internal temporary approval cannot waive a legal condition or postpone necessary validation until after use. Restrict the activity where no supported alternative exists. Repeated extensions require renewed assessment of the remaining risk and feasibility of the permanent remedy.
Before restart, engineering and the technical function establish that the relevant defect has been corrected and control and measurement capability restored. Operations is responsible for confirming that instructions are current, resources are available and staff are competent, and obtaining approval from the person authorised to accept readiness. Assess operation immediately after restart against the approved conditions. Keep the decision on previously affected product separate. Refer serious, recurring or systemic findings for corrective action, linking the original event and product decisions to the case.
Preserving the evidence and version history
Maintain a retrievable file linking each active limit to its hazard, the products and processes covered, scientific evidence, validation protocol, raw data, calculations, report and technical acceptance. Identify the equipment configuration, instruments and essential conditions. Keep the source editions and supporting material needed to assess the decision; a web link alone may become inaccessible. An active approval must remain substantiated when a provider or source changes.
Link implementation approval and the effective production boundary to the limit version. Historical batch and monitoring records need to remain interpretable against the schedule and instruction in force at the time. Retain superseded versions, study failures, review decisions and authorised changes. Distinguish a blank form defining future recording from a completed record of what occurred. Changing the form does not change historical evidence.
Create records as work occurs and distinguish observation from later interpretation. Check that required entries, dates and references agree with related records before using them for a consequential decision. Investigate discrepancies with the person responsible and state any remaining limitation. For a reconstructed account, identify when it was created, its basis and uncertainty. It cannot be presented as a contemporaneous CCP measurement.
When correcting a record, preserve the original information and identify the change, reason, person making it and date. Control electronic permissions and attributable amendments, maintain backups and test recovery. Keep paper records legible and protected from loss or deterioration. Provide authorised technical reviewers access while restricting sensitive information to appropriate recipients. Before changing software or providers, preserve readable data and the means of interpreting them.
Determine how long to retain evidence from applicable legal and customer requirements, product life and the purpose the evidence continues to serve. Consider relevant extended consumer use. Where customer or external audit conditions require at least shelf life plus twelve months, apply that minimum and assess whether longer retention is needed. Preserve an active limit's technical basis throughout reliance on it, and retain historical evidence for related production, investigations and unresolved obligations. A routine production-record disposal date must not remove the only substantiation of a current limit.
Test retrieval with an actual active limit and a historical production record. Obtain the applicable version, technical support, implementation evidence, monitoring results and any deviation decision. Resolve inaccessible or broken links. Wider record arrangements are developed in the guidance on HACCP documentation and record-keeping.
Verifying current use and reviewing continuing applicability
A competent verifier is responsible for tracing an active limit through the HACCP plan, issued instruction, actual settings, monitoring practice and records. The verifier compares the product and process now used with the conditions covered by validation. Include relevant differences such as different shifts, changed recipes, challenging loads and equipment configurations where these could alter the judgement. Record what was examined and the limits of the conclusion; one convenient record cannot establish performance across all uses.
Review monitoring values, dates, times, monitor signatures and appropriate authorised verification, including electronic checking evidence. Examine whether warnings prompted timely action and failures or data gaps resulted in protection and product decisions supported by evidence. Consider calibration and maintenance findings alongside apparently favourable monitoring results. An absence of recorded deviations can reflect stable control, inadequate monitoring or missing records, so establish which explanation the evidence supports.
Evaluate trends on a compatible basis. Examine proximity to limits, variability and failures alongside production volume, product mix, throughput, shifts and measurement changes. A lower deviation count after a threshold or recording method changes is not automatically improvement. Preserve original criteria and explain the comparison. Favourable product tests, particularly for rare hazards or limited samples, cannot replace evidence of the control mechanism and its execution.
Schedule review from risk, operating experience and applicable duties. Where customer or external audit requirements call for at least annual review of HACCP and prerequisite programmes, include the limits and their continuing validation basis. Review sooner after relevant change, significant deviation, new science, changed obligations or an adverse finding. A routine procedure-review date does not defer assessment of a current threat to validity.
The HACCP team is responsible for recording the evidence considered and explaining why it continues to support the existing limit or what needs revision. Record verification results even when no action is needed and communicate them to the team. Record who is responsible for unresolved work, the dates for completing it and any interim controls needed. Refer significant performance and resource issues to management, while preserving the decisions assigned to the technical and product-control functions. Where findings change product safety commitments or objectives, send the implications to their owners for the appropriate review.
Assessing change and revalidating before use
The person responsible for a proposed change brings it to the HACCP team before implementation. Assess ingredients and suppliers, recipe tolerances, product dimensions, packaging, equipment, load, speed, flow, sensor position, software, intended use and distribution conditions where relevant. An increase in throughput or temporary recipe substitution can affect the safety basis even without a formal capital project.
Compare each changed condition with the accepted technical evidence. Determine whether it could change the hazard-control effect, delivered treatment, measurement relationship or decision. Record why existing support remains applicable or define the additional work required. Limit continued use to the products and conditions demonstrably covered by the evidence. When present capability is undermined, suspend the affected use or operate only under a separately substantiated temporary arrangement.
Validate safety-affecting changes before implementation. Coordinate study work, technical acceptance, authorisation, equipment readiness, document issue, settings and staff competence. Set an effective production boundary and preserve both versions where a transition is staged. After introduction, compare the actual configuration and practice with the approved change and examine performance under the conditions the change was meant to support. Reopen assessment if the intended result is not achieved.
When an external processor controls the CCP, agree the products and processes covered, required limits and supporting conditions, monitoring and deviation information, change notification and access to evidence. The commissioning site's HACCP team remains responsible for evaluating the control within its food safety plan and accepting the relevance of the external evidence. General supplier approval or receipt of a certificate does not establish the particular process's capability. Information about a change at the processor must reach the commissioning site in time for this assessment.
Improving reliability and transferring or retiring limits
Successful operating experience, operator feedback and new technical knowledge can justify improvement before failure. More representative measurement, stable equipment performance, clearer criteria, faster intervention and better evidence retrieval can strengthen an already functioning control. Define the problem or intended benefit, the proposed method, resources and acceptance evidence before making the change. Assess any effect on safety capability through the change and validation arrangements above.
Verify delivery and adoption separately from the effect of an improvement. Evidence of sensor installation and staff training demonstrates those implementation steps; evidence from representative production is needed to show better measurement or more reliable control. Allow relevant products, workloads and opportunities to occur before judging sustained effectiveness. Keep an action open pending evidence of effectiveness where those conditions have not occurred. If it fails, reassess its technical basis, implementation and resources rather than closing it on delivery alone.
Review recurrent difficulties to distinguish an unsuitable method from inadequate competence, access, capacity or unclear authority. More training cannot resolve an inaccessible measurement point, and extra checking cannot supply missing technical capability. Senior management is responsible for providing the resources and setting the priorities needed for a supported solution. The HACCP team retains responsibility for technical evaluation, and the operational team receiving the change is responsible for confirming readiness.
Before a role holder, provider, system or process is replaced, transfer active limits, original evidence, permissions, pending changes and unresolved actions to a competent recipient. That recipient confirms access to the transferred material and understanding of the responsibilities being taken over. Arrange usable data exports and specialist continuity where interpretation depends on a particular provider. Restrict dependent work if necessary capability cannot be sustained.
Retire a limit when its covered use ends or an approved replacement takes effect. Withdraw it from operational use, preserve its historical connections and assign continuing investigations or actions to an identified owner. Dispose of evidence only when applicable retention and continuing needs permit it. The end of one version or production activity does not close an unresolved product assessment or corrective-action case.