What is package leak testing?
Package leak testing checks whether a finished pack has an unintended leakage path. Select a method by the package construction, contents, headspace and defects that matter to product protection, then demonstrate detection on representative filled packs. A method name or advertised minimum hole size is insufficient to specify an acceptance test.
This guide covers vacuum decay, force decay, internal-pressurization bubble testing and CO₂-based leak detection for equipment buyers. It separates leak detection from seal-strength and visual checks and provides an RFQ and trial framework. The framework is editorial procurement guidance; it does not replace an approved quality procedure or the full applicable test standard.
Why revisit leak-test selection now?
At Cibus Tec, running 27–30 October 2026 in Parma, Antares Vision Group’s official exhibition page lists combined leak/vision and leak/X-ray/checkweighing configurations. These are manufacturer-described configurations, not independently verified production results.
The purchasing question is which defect each module can establish on your pack. A shared enclosure does not make leak detection, seal-area imaging and weight measurement interchangeable. The comparison and acceptance checks below help turn a combined-machine quotation into separate, testable requirements.
Vacuum decay, force decay and bubble testing: what differs?
In PTI’s vacuum-decay description, the pack sits in an evacuated chamber and pressure measurement tracks changes during the test. Fixture design, package characteristics, pressure and timing affect detection. ASTM F2338-24’s public scope identifies vacuum decay as a nondestructive leak-test method. Porous surfaces require appropriate treatment; an unmodified method for a non-porous pouch is not automatically suitable for porous lidding.
| Method | Measurement or observation | Buyer proof required |
|---|---|---|
| Vacuum decay | Pressure change in an evacuated chamber containing the pack. | Fixture, test settings and detection on the filled product-package system. |
| Force decay | Change in measured force during the configured package test. | Pack restraint, load-cell response and method suitability for the format. |
| Internal-pressurization bubble test | Visible escaping bubbles from a pressurized pack. | Method scope, access to the pack, operator observation and gross-defect detection. |
| Seal-strength peel test | Force required to separate a seal specimen. | Consistent specimen preparation and test configuration; separate leak evidence. |
Sepha describes its force-decay approach as nondestructive for passing packs. Ask for the actual measurement sequence rather than assuming force decay and vacuum decay are interchangeable. ASTM F2096’s public scope describes a destructive internal-pressurization test for gross leaks. Other water-bath methods use different pressure arrangements; do not treat every bubble test as F2096.
A strong seal is not the same as a leak-free pack
Instron’s F88 overview describes a peel-strength measurement. That measurement answers a mechanical-strength question. It does not directly demonstrate the absence of a leak elsewhere in the finished package. Write distinct acceptance requirements for seal strength, leak detection and any visual defect checks.
A camera may identify visible contamination, folds or seal geometry. Its ability to find a particular leakage path must be demonstrated rather than inferred from a clear image. Use the machine vision inspection guide for image-based acceptance tests and pair it with package-integrity evidence where required.
CO₂ leak detection versus seal-area imaging: specify separate evidence
CO₂ leak detection measures escaping tracer gas; seal-area imaging examines the sealing region. They answer different questions. A camera finding product in a seal is useful process evidence, but it does not by itself quantify leakage. Equally, a passing leak result does not establish that a seal meets its separate appearance or strength requirement.
Antares Vision Group’s LDS documentation describes applying controlled pressure to horizontal MAP packs, sampling released gas and measuring CO₂ by IR absorption spectroscopy. Its SIS 700 description identifies hyperspectral imaging of organic material in the seal area. Confirm the supplied sensor, film compatibility and inspection coverage; these are product-family descriptions, not proof for every tray or film.
| Inspection | Evidence it provides | What buyers must establish separately |
|---|---|---|
| CO₂ tracer-gas leak detection | Escaping CO₂ under the configured test conditions. | Suitable gas content, defect detection, pack handling and background-gas tolerance. |
| Seal-area imaging | Image-based detection of the specified seal-region defects. | Finished-pack leakage and mechanical seal strength where required. |
| Vacuum or force decay | The configured pressure or force response of the product-package system. | Detection across filled-pack variation, fixtures and approved recipes. |
| Dynamic checkweighing | Package mass against configured limits. | Leakage, seal quality and foreign-body detection. |
Confirm the gas and package operating window
WITT’s MAPMAX documentation describes a different implementation: a chamber pressure difference encourages gas escape and the instrument detects a CO₂ concentration increase. A CO₂-based method needs an appropriate CO₂ signal. Do not assume a nitrogen-only pack qualifies, or transfer a result between chamber and pressure-assisted conveyor systems without a trial.
For an RFQ, document gas recipe, normal headspace, fill, temperature, pack orientation and elapsed time between sealing and inspection. Ask the supplier to state its approved range for each, including treatment of background gas and packs outside that range. Do not change a food-protection gas recipe merely to suit a detector without the responsible quality team’s approval. A quoted hole size is not a universal leak-rate or shelf-life threshold.
Four combined-system acceptance challenges
The following are PackLine IQ editorial purchasing checks. Agree sample preparation, quantities, acceptance limits and safe execution with the quality team and supplier; they are not a replacement for a validation procedure.
- Separate the defect classes. Run traceable good packs, agreed leak challenges and agreed seal-area contamination challenges. Document which module is expected to respond to each sample. A sample can contain more than one defect; confirm its status rather than assuming contamination always creates a measurable leak.
- Challenge normal variation. Repeat across the approved film, gas, fill, headspace and orientation range. Include the required production mix and changeover. Record missed defects and good-pack rejects separately for each module and the complete machine.
- Follow each pack to disposition. Use an agreed mixed sequence to verify that the right result follows the right pack through transfers. Demonstrate a pack failing one module and a pack failing multiple modules. Confirm rejection and the treatment of missing or inconclusive results.
- Test unavailable inspection. Through authorized fault simulations, demonstrate sensor unavailability, lost result communication and a reject-path fault. Confirm the agreed hold, stop or diversion behavior and recovery records. A running conveyor is not evidence that every inspection is active.
Compare the complete installed duty
Evaluate combined and separate stations against the same detection, sustained output and recovery requirements. Request itemized format tooling, sensor access, cleaning, calibration, software support and spare parts. Ask what production can continue when one module needs service, and which inspections would then be absent. Count false rejects and investigation labor in operating cost; do not assume a shared frame guarantees lower total cost.
Use the machine-vision acceptance guide for image challenges, the checkweigher combination guide for weighing and reject coordination, and the line-integration guide for interface ownership. The buying decision should end with a documented coverage matrix, approved operating window and verified pack disposition.
Specify the product-package system before the tester
For fresh-food trays, the vacuum skin versus MAP tray sealing guide explains where headspace analysis fits alongside seal integrity checks and how to compare the two formats in a supplier trial.
Prepare a format matrix with material construction, permeability, dimensions, closure or seal design, filled mass, product state, headspace and normal temperature range. Identify the smallest and largest packs and any production conditions likely to change the measurement. SealTick identifies headspace as an application requirement for its vacuum-decay approach; confirm how your chosen system handles low-headspace or liquid-filled packs through trials.
Define relevant defects with the quality team: where leakage matters, the required detection objective and how challenge samples will represent the risk. Ask the supplier to explain differences between an artificial hole, a seal channel and a calibrated leak. Avoid converting a marketing hole diameter into a universal product-release threshold.
Repeat relevant qualification when packaging changes. The material compatibility guide provides the wider line-trial context, while the filling machine guide helps define the upstream product and cleaning duty.
Five buyer tests to agree before ordering
Set trial duration, sample numbers, statistical treatment and acceptance limits with your quality team and supplier before testing. The following is a procurement framework, not a universal validation protocol.
- Separate good-pack variation from defects. Challenge normal variation in fill, temperature, headspace and pack size. Retain a traceable set of reference and known-defect samples with documented preparation.
- Blind the sample sequence. Where practical, run an agreed randomized sequence whose status is concealed from the operator. Record missed defects and good packs rejected; repeat across active positions.
- Demonstrate the complete method. Record fixture, recipe, pressure or force settings, timing and result rules. Establish which parameters are locked and which changes require review.
- Measure usable throughput. Include loading, stabilization, measurement, unloading and handling of failed or inconclusive results. Demonstrate required output using the actual product mix.
- Challenge faults and records. Demonstrate recipe mismatch, sensor or utility faults and communication loss through approved procedures. Verify pack/result association, access permissions and recovery.
Retain the results and unresolved exclusions with the purchase agreement. Repeat appropriate checks after installation; a supplier demonstration cannot establish the behavior of your complete production line.
Offline sampling versus inline inspection
Offline testing can support process development and an approved sampling plan. It does not individually test unsampled packs. Inline testing must demonstrate both the measurement and the handling system: correctly presenting each pack, retaining its result association and directing failed or untested packs to the agreed disposition.
Build a capacity model for every format. As an illustrative calculation, testing 60 packs per minute means completing one pack per second on average. A multi-position chamber might address that demand, but simultaneous positions do not by themselves prove the total cycle or availability. Ask for sustained output including transfers and interruptions.
Define what happens when the tester stops, when a result is absent and when downstream equipment is unavailable. Use the line integration guide for interface ownership. Coordinate reject gaps with the flexible-bag case packer so incomplete groups cannot become accepted cases.
Installed cost and an RFQ worth comparing
No comparable installed price has been independently verified for this guide. Request quotations against the same format matrix and acceptance protocol. Include chambers and fixtures, format parts, utilities, loading automation, reject handling, controls, method development, commissioning, training, calibration and ongoing service. Identify excluded formats and unproven defect classes.
Compare annual operating cost at the same required output. Include destructive sample losses where applicable, retained operator work, false rejects, downtime and changeover. Treat avoided complaints or product losses as scenarios using your own evidence rather than guaranteed savings. The automation ROI guide helps separate confirmed savings from assumptions.
A useful RFQ ends with an evidence request: show the full measurement and handling process on our filled packs, report both false-result types, and document the approved operating window. Buy demonstrated detection and a maintainable method.
Key takeaways
- Select a method for the filled product-package system.
- Keep leak, strength and visual tests distinct.
- Report missed defects as well as false rejects.
- Verify handling, records and installed performance.
Frequently asked questions
Is vacuum decay the same as force decay?
No. They use different measurement signals. Ask the supplier to demonstrate the complete method and detection on your filled package.
Does a seal-strength test prove a pack is leak-free?
No. A peel-strength result measures mechanical separation of a seal specimen. Obtain separate evidence for finished-package leakage.
Can low-headspace packs be leak tested?
Potentially, but suitability depends on the method, fixture, contents and settings. Demonstrate performance on representative filled packs.
Does a combined leak and vision machine prove every pack is defect-free?
No. Each module must demonstrate its specified defect detection within an approved operating window. Verify pack/result association, rejection and the handling of missing results; seal imaging, leakage and weight remain separate measurements.
How much does package leak-testing equipment cost?
No comparable installed price has been independently verified. Request matched quotations covering fixtures, handling, method development, installation and support.
When a film change produces contaminated seals, use the static-control and material trial checklist to investigate charge and contamination alongside finished-pack integrity.
For replacement films, qualify the production heat-seal operating window alongside leakage detection; laboratory strength data alone does not establish acceptable finished packs.
A leak test evaluates the package against its specified defect criteria; it does not establish equipment cleanliness. Keep the filler CIP and cleaning-validation scope separate in the purchase and acceptance plan.
Before selecting a leak detector, specify the upstream flow-wrapper sealing window and complete-pack acceptance trials so integrity testing is tied to the approved product and film.
For foil-lined bottles, pair finished-pack leak testing with the upstream induction cap sealing qualification, including changes in gap, cap torque and line speed.
Sources & editorial method
Source-checked 11 October 2026. Public manufacturer descriptions and standard scopes support the method distinctions. Use the full current standards and approved quality procedures for test execution. RFQ questions, capacity arithmetic and buyer tests are PackLine IQ editorial synthesis, not a certified design or universal validation procedure. The original Sepha reference was a preview for PACK EXPO, 18–21 October 2026; it is not a report of observed demonstration results. The combined-inspection section was source-checked on 29 October 2026 against the linked Antares Vision Group and WITT public documentation. Its coverage matrix and acceptance challenges are editorial synthesis; supplier claims are not independent validation.
- Sepha: latest PakScan and October 2026 PACK EXPO preview
- PTI: vacuum decay measurement and application variables
- ASTM F2338-24: public scope for vacuum decay package testing
- ASTM F2096-11(2019): public scope for internal-pressurization bubble testing
- Instron: F88 seal-strength testing overview
- SealTick: vacuum decay applications and headspace considerations



