Short answer: Test frequency should be set out in your food safety plan, which for most sites is built on HACCP, and meet the requirements of the certification standard you are audited against, such as BRCGS or SQF, and any retailer or customer code of practice. Many sites test at start-up, at product changeovers, at regular intervals during production, and at the end of a run. Testing more often than the minimum reduces how much product is at risk if a check fails.
Routine testing only means something if the metal detection system has first been validated on the real product, and if every check is recorded clearly enough to stand up to an audit. This guide covers how often to test, what each test must prove, the difference between validation and verification, what to record, and what to do when a test fails.
A typical test routine may include:
There is no single interval that applies to every line. The frequency should be set out in your food safety plan and meet the requirements of the certification standard you are audited against, such as BRCGS or SQF, and any retailer or customer code of practice.
Treat that as a minimum, not a target. If a check fails, everything made since the last successful check has to be held and evaluated, so the time between checks decides how much product is at risk and how long a problem can go unnoticed. A minimum frequency may satisfy an audit, but testing more often shortens that window and lowers the risk of producing non-conforming product before a fault is found.
Where the detector is used as part of a critical control point, records must be clear, complete, and available for audit.
Each test should confirm that the detector finds the agreed test pieces and that the reject system removes them under normal working conditions.
Under BRCGS, test pieces contain a metal sphere of known diameter, chosen on the basis of risk, and are marked with their size and type. Separate ferrous, stainless steel and, typically, non-ferrous test pieces are used. Where possible, the test piece goes inside a clearly identified sample pack of the product being run, passes through the least sensitive part of the aperture (usually the center), and is run in successive packs at normal line speed. Failsafe systems on the detector and reject should be checked as well. In-line detectors are tested at least at line start-up and at the end of production.
Sensitivity targets are often set by the food manufacturer's customer. A retailer or brand owner may require a product to achieve defined ferrous, non-ferrous, and stainless steel test piece sizes before approval. The exact sizes depend on the product, the customer, and the site's risk assessment.
Stainless steel is usually the most difficult to detect, especially when the fragment is small or when the product itself creates a strong background signal.
Test with the product, not just the test piece. A test piece passed through the detector on its own may be detected easily. Paired with a product that has a strong product effect, the same test piece can push the detector much closer to its detection limit. Passing the test piece within or alongside the normal product stream means the detector sees the same product effect as it does in production, which gives a far more accurate picture of real-world performance. This applies to validation as well as routine checks: on the actual product, in the actual packaging, at the actual production speed.
The detector type depends on how product moves through the line:
Validation, routine checks, and verification are often used together, but they are not the same. Food safety standards treat them as separate steps.
Proves that the metal detection system can detect the agreed test pieces under real production conditions, with the test piece passed through within or alongside the actual product rather than on its own.
When: Before the system is put into use, for each new product, and before changes to the line, product, or settings are implemented.
Routine checks with certified test pieces confirm the detector is still performing during production. Depending on your plan, these are recorded as monitoring or as verification. Verification also covers calibration and review of the check records by a responsible person.
When: At every scheduled check through the production run, with records reviewed on a set schedule.
In simple terms, validation asks whether the system can detect the required metal contaminants. Routine checks ask whether it is still detecting them today, on this product, during production. Verification confirms that the whole system, including the records, is working as planned.
In many food plants, metal detection forms part of a documented food safety plan and may be defined as a critical control point, quality control point, or prerequisite control depending on the site's risk assessment and customer requirements. To support compliance, the system must be specified, installed, validated, tested, maintained, and documented correctly.
BRCGS requires the testing procedure to cover, at a minimum, who is responsible for testing, the sensitivity required and any variation for particular products, how and how often the detector is checked, and how results are recorded. Most sites also define:
Auditors, customers, and retailers often expect evidence that the system works under real production conditions. BRCGS requires either an automatic reject that diverts contaminated product out of the product flow or into a secure unit that only authorized staff can access, or a belt stop with an alarm where product cannot be rejected automatically. Retailer codes of practice may also require locked or password-protected settings, reject confirmation, bin full detection, air pressure monitoring, audit trails, and event logs.
A metal detector test record should provide enough detail to prove that the system was working correctly at the time of inspection. Records may include:
| Record | What to capture |
|---|---|
| When and where | Date and time of the test, line number or detector identification |
| Product | Product name or product code, pack size |
| Who | Name and signature or confirmation of the person doing the check, supervisor sign-off where required |
| Test | Test piece type and size, detector settings or product profile |
| Result | Pass or fail result, reject confirmation result |
| Rejects | Product rejected or removed by the detector |
| Follow-up | Corrective action taken |
| Review | Records reviewed and signed off by a responsible person. In the US, FSMA requires this within 7 working days unless a longer period is justified. |
Keep records beyond the shelf life of the product, or longer if your certification standard or customer requires it. Electronic records can help reduce paperwork and improve traceability. However, the system must still be easy for operators to use correctly during busy production periods.
A failed metal detector test must be treated seriously. If a detector fails to identify a test piece, fails to reject correctly, or behaves inconsistently, production should be stopped or controlled according to the site procedure.
Product made since the last successful check should be held, evaluated for safety, and its disposition decided and recorded before any of it is released. The detector should then be investigated, corrected, and re-tested to confirm it is working before production restarts. If the failure points to a wider problem, the plan itself should be reviewed. Typical corrective actions include:
A strong procedure protects the consumer, supports audit compliance, and helps the site identify recurring issues before they become larger problems. For the wider picture, including product effect and where to install detectors, see How Industrial Metal Detectors Support Food Safety and Compliance.
Test frequency should be set out in your food safety plan, which for most sites is built on HACCP, and meet your certification standard, such as BRCGS or SQF, and any retailer or customer code of practice. Many sites test at start-up, at product changeovers, at regular intervals during production, and at the end of a run. Testing more often than the minimum reduces how much product is at risk if a check fails.
Validation proves that the metal detection system can detect the agreed test pieces under real production conditions, with the test piece passed through within or alongside the actual product, before the system is used and before changes are made. Routine checks with certified test pieces confirm it is still working during production, and verification confirms that the whole system, including calibration and records, is working as planned.
Sensitivity requirements are often set by the food manufacturer's customer, retailer, or brand owner. These specifications normally define separate test piece sizes for ferrous, non-ferrous, and stainless steel contamination.
Production should be stopped or controlled according to the site procedure. Product made since the last successful check should be held, evaluated for safety, and its disposition decided and recorded. The detector should be investigated, corrected, and re-tested to confirm it is working before production restarts.
Bunting can help assess your product, line layout, inspection point, and compliance requirements to identify a practical metal detection solution.
On this page