TL;DR: A poorly scored FMEA that underweights detection difficulty is the most common reason a packaging line ships a latent defect — scoring severity and occurrence without honest detection ratings produces false confidence.
For a related specification, compare Automated Inspection Systems — Safety & Risk Assessment before finalising the packaging brief.
TL;DR: In our production environment, any FMEA Risk Priority Number above 120 triggers a mandatory process hold until corrective action is verified — we’ve applied this threshold across 14 active SKUs in the past 18 months.
How We Score Hazard Risk in Packaging Production: FMEA Methodology and Thresholds #
Risk Priority Numbers are only as reliable as the scoring discipline behind them. Our FMEA process follows the IEC 60812 functional failure analysis framework, adapted to packaging line conditions. Each failure mode receives three scores — Severity (S), Occurrence (O), and Detection (D) — each on a 1–10 scale, producing an RPN of S × O × D with a maximum of 1,000.
Where many production teams go wrong is treating Detection as a formality. A defect that a downstream camera system catches reliably scores D = 2 or 3. A defect that only manifests after consumer opening — a delaminating inner layer, an adhesive bond that passes cold pull-test but fails at 40°C — scores D = 8 or 9. That difference alone can swing an RPN from 48 to 144.
Our production team runs FMEA reviews on three trigger points: new product introduction, any raw material supplier change flagged in our MRQ-04 Material Requalification Checklist, and after any incident classified Category B or above in our non-conformance tracking log.
| Failure Mode | Severity (S) | Occurrence (O) | Detection (D) | RPN | Action Threshold |
|---|---|---|---|---|---|
| Ink bleed on food-contact inner liner | 9 | 2 | 6 | 108 | Monitor — approach threshold |
| Greyboard caliper below 1.8mm (rigid lid) | 7 | 3 | 4 | 84 | Acceptable — inline gauge active |
| Magnetic closure misalignment >1.5mm | 5 | 4 | 3 | 60 | Acceptable |
| Adhesive cure incomplete (UV flexo) | 8 | 3 | 8 | 192 | Process hold — RPN >120 |
| Folding carton glue flap bond failure | 7 | 4 | 5 | 140 | Process hold — RPN >120 |
The table above shows actual failure modes from our folding carton and rigid box lines, scored in Q2 2024. The adhesive cure entry scored high on Detection because our in-line UV energy meters verify output wattage but not cure depth — a substrate with unexpected UV-blocking pigment can pass wattage check and still carry under-cured adhesive. That scenario required a revised test protocol: we now run ASTM D1876 T-peel tests on 5 units per 2,000-unit batch as a secondary confirmation for UV-bonded structures.
Any RPN above 120 initiates a documented process hold, a root cause analysis within 48 hours, and a re-score after corrective action. The 120 threshold is not arbitrary — it reflects the point at which at least two of the three scoring dimensions are elevated simultaneously, indicating a systemic risk rather than an isolated variable.
What Goes Wrong: Three Failure Scenarios We Track Closely #
Scenario 1: Ink migration on food-adjacent packaging. We produce folding cartons that sit adjacent to (not in direct contact with) food products — think retail bakery boxes and outer sleeves for beverage multipacks. The relevant standard here is EU Regulation 10/2011 on plastic materials in food contact, and while our paperboard isn’t a plastic film, migration chemistry from printing inks still requires evaluation under similar principles. The hazard: photoinitiator residues from UV offset printing can migrate through a paperboard wall if the barrier layer is thinner than specified. Our standard barrier spec for food-adjacent cartons is a 15 gsm PE-coated kraft liner on the inner face. Below 12 gsm, we’ve measured toluene-equivalent migration rates exceeding the 0.01 mg/kg food simulant threshold in our internal testing. What to check: incoming caliper on the PE-coated liner, and whether your ink supplier has provided a migration compliance declaration under GB/T 36787 (China’s food packaging ink standard).
Scenario 2: Structural failure at cold-chain temperatures. A brand partner shipping refrigerated meal kits had box base panels delaminating during transit — not during production, not during our outgoing QC, but after 48 hours at 4°C. The mechanism: their lamination adhesive was a standard water-based PVA rated for ambient conditions. At 4°C with 85–90% relative humidity, the bond strength dropped below the 3.0 N/15mm peel minimum we specify for structural panel joints. Our QC team had signed off on ambient-temperature peel tests per ASTM D1002, which was the agreed spec. The lesson: when packaging is specified for cold-chain use, the agreed test temperature matters as much as the test method itself. We now flag cold-chain end use as a mandatory declaration field in our customer brief — if it’s checked, peel tests are conducted at both 23°C and 5°C before sampling sign-off.
Scenario 3: PPE gap during solvent-based coating operations. This one sits at the intersection of chemical safety and quality control. On our coating line, we run both water-based and solvent-based varnishes depending on the job. Solvent-based coatings involve toluene and MEK concentrations that require air supply respirators during closed-pan cleaning, not just half-face cartridge respirators. Our current PPE matrix, reviewed against ISO 45001:2018 Occupational Health and Safety requirements, specifies air-supply units at >50 ppm ambient solvent concentration, measured by continuous area monitors we installed in 2023. Before that installation, we relied on periodic sampling — which missed concentration spikes during shift changes when pan cleaning coincided with line startups. The consequence was not an acute incident, but two operators reported headaches in a 3-month window, which triggered a full air quality audit. The detection gap — periodic sampling versus continuous monitoring — mirrors the FMEA detection scoring issue: the hazard was real but intermittently visible.
Does FMEA Replace a Hazard Identification Matrix? #
No — they serve different scopes. An FMEA works at the process failure level: what can go wrong within a defined production step. A hazard identification matrix, structured per ISO 31000 risk management principles, operates at the system level: chemical exposures, ergonomic risks, equipment energy sources, emergency egress. Both are required and neither substitutes for the other.
For packaging production specifically, the hazard matrix categories we maintain cover chemical hazards (inks, solvents, adhesives), mechanical hazards (die-cutting, folding equipment, foil stamping press), fire hazards (UV lamp heat, solvent vapour), and ergonomic hazards (sheet-fed stack lifting). These are reviewed quarterly. FMEA reviews are event-triggered as described above.
Specification Notes for Brand Partners #
When you brief us on a new packaging project, our QC team uses that brief to pre-score the FMEA before tooling is cut. The information we need upfront: end-use environment (temperature, humidity, transit duration), any food-contact or food-adjacent classification, surface finish specification including any UV or solvent-based coatings, and whether the packaging will carry pharmaceutical, cosmetic, or child-safety regulatory requirements.
The most common brief gap we encounter is missing end-use environment data. A brand will specify the packaging structure in full detail but leave the distribution chain undefined. That omission forces us to default to ambient-condition testing, which may not cover cold-chain, high-humidity, or e-commerce drop-test requirements under ISTA 2A protocols. When that data arrives late, it typically causes one full sampling iteration.
Our standard sampling timeline from approved dieline to first physical sample is 12–15 working days for folding cartons and 18–22 working days for rigid boxes. If the project triggers a Category B FMEA item (RPN >120), we add 3–5 working days for the corrective action loop before samples are released. We’d rather hold the sample than ship one that hasn’t cleared its risk assessment.
Frequently Asked Questions #
What FMEA RPN threshold should I ask my packaging supplier to apply?
It depends on your product category and the regulatory environment it operates in. For general retail packaging, an RPN threshold of 100–125 is a reasonable control point. For pharmaceutical packaging, cosmetic ingestibles, or any food-contact application, I’d push the threshold down to 80 — at that level, you’re catching elevated Severity scores even when Occurrence and Detection appear manageable.
Do your production staff need specific certifications for hazardous material handling?
Yes. Our team members working with solvent-based coatings and adhesives hold certificates under China’s GBZ 1-2010 industrial hygiene standard, and our safety officer maintains an annual review schedule aligned with ISO 45001:2018. PPE training is documented per individual, not per department — each operator’s competency record is tied to the specific chemicals their station handles.
Can you share your FMEA documentation as part of a supplier audit?
We provide FMEA summaries for the specific product family under audit, including RPN scores and any open corrective actions. Full process FMEA documents covering proprietary line configurations are shared under NDA. For most brand partner audits, the summary level is sufficient to verify that the threshold discipline is in place.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
