A package can pass a visual check, pass a burst test, ship clean, and still fail in the field three weeks later. A channel defect in the seal, invisible to the eye and invisible to a strength test, lets oxygen creep in over time. By the time the complaint comes back, the whole run has shipped.
That gap between “looks sealed” and “actually sealed” is what seal integrity testing closes. It verifies that a package stays sealed under real conditions, not just on the bench, and it points to where the seal is failing so you can fix the line. Most standardized methods give a visual result in about 30 seconds.
This guide covers what integrity testing catches, the ASTM methods worth knowing, how to pick the right one for your package, and how to build it into a QA program.
Seal Integrity vs. Seal Strength
These get used interchangeably, and they measure different things. Seal strength is the force needed to pull a seal apart. Seal integrity is whether the seal actually keeps leaks out under expected conditions. A seal can test strong and still carry a channel defect that lets oxygen in. You need both, for different reasons. The full breakdown is in our guide on seal strength vs. seal integrity.
What Seal Integrity Testing Catches
The defects that cause recalls are rarely dramatic. They are the small ones that pass visual inspection and fail in distribution.
- Channel defects. A hairline path through the seal, usually at a corner, a wrinkle, or where product contaminated the seal zone. It leaks too slowly to see, which is how a shipment clears receiving and then spoils on the shelf.
- Pinholes. Small punctures from sharp product, flex-cracking in transit, or a film defect. A fraction of a millimeter is enough to shorten shelf life.
- Weak or open seals. A seal that looks closed but peels under light pressure or shipping stress.
- Microleaks. Too small to see, still large enough to let oxygen in or moisture out. Sterile pharma and medical packaging need more sensitive methods for these.
- Closure failures. Threaded caps, lug closures, and press-on lids that did not fully seat or torque, then work loose under vibration.
- Seal-bar drift. Temperature or dwell-time changes that produce seals at the edge of spec. The first samples pass, mid-shift samples start failing, and the drift stays invisible without regular testing.
Where the leak shows up matters as much as the pass or fail, because it tells you what the sealing process is doing wrong.
Reading the Result: What the Bubbles Tell You
In a bubble-based test, the pattern identifies the defect:
- A fine, steady stream from one point: a pinhole or puncture.
- A line of bubbles along part of the seal: a channel leak.
- Wide, spreading bubbles along a seal edge: a weak or partially open seal.
- Bubbles that appear briefly, drift, and stop: trapped air, not a leak. A real leak repeats from the same spot through the full dwell time.
The ASTM Methods FlexPak Runs
Each method fits a specific package type and defect. These are the methods the FlexPak Leak Detector performs.
Bubble emission, ASTM D3078 (external vacuum). The package is submerged in water inside a vacuum chamber. As vacuum is drawn, air escapes through any leak as a visible bubble stream. Best for flexible packaging with headspace gas (snacks, dairy pouches, pet treats). It detects gross leaks; actual sensitivity depends on vacuum level, headspace, product, and material, and small leaks may be missed at lower vacuum. Fast, visual, easy to train. Step by step: how to perform a bubble leak test.
Internal pressurization, ASTM F2096 (FPIPA). Air is introduced inside the package through a septum while it is submerged, and bubbles reveal the defect. Better for low-headspace, vacuum-sealed, and porous sterile-barrier packs (including Tyvek). The standard states sensitivity to 250 µm at roughly 81% probability from round-robin data, so detection at that size is not guaranteed (about 1 in 5 specimens were missed). It is a pass/fail method, not a leak-rate measurement, and it is destructive because the package is punctured. More: ASTM F2096 practical uses.
Altitude simulation, ASTM D6653. Recreates the pressure drop of air shipment or mountain transport so marginal seals reveal themselves. This is a conditioning method, not a detection method: it stresses the package, then you pair it with bubble emission to see what failed. More: ASTM D6653 altitude testing.
Dry chamber, ASTM D5094. A water-free vacuum test for liquid-filled rigid or semi-rigid containers with threaded or lug closures (up to 4L). Leaks show as liquid escaping onto absorbent material, which keeps the product clean and simplifies cleanup. More: dry chamber testing.
Low-headspace and vacuum-sealed packs. Vacuum-sealed products have almost no internal gas, so they will not bubble until inflated. The FlexPak VAC Attachment adds a measured shot of air before the cycle so D3078 works on vacuum-sealed and liquid-filled formats. Skipping this is one of the most common false-pass causes on the floor.
A note on sterile pharma: for primary container-closure integrity validation on sterile injectables, USP <1207> favors deterministic methods such as vacuum decay. FlexPak focuses on bubble emission and internal pressurization for flexible sterile-barrier and complementary checks. Background on the deterministic side is in our vacuum decay testing guide.
How to Choose a Method in Under a Minute
Most decisions come down to package type, headspace, and product sensitivity:
- Flexible pouch with headspace? D3078.
- Vacuum-sealed or low-headspace? D3078 with the VAC Attachment.
- Sterile or medical pack? F2096.
- Liquid in a threaded or lug-closure container? D5094 dry chamber.
- Ships by air or over mountain passes? D6653 conditioning, then D3078 to detect.
Quick examples: a snack maker testing nitrogen-flushed pouches runs D3078; a medical facility verifying sterile barriers runs F2096; a pet-food co-packer testing vacuum-sealed rolls runs D3078 plus the VAC Attachment.
Common Mistakes That Skew Results
- Wrong vacuum or pressure: too much ruptures good seals (false fail), too little hides real leaks (false pass).
- Not inflating vacuum-sealed packs before testing.
- Poor sample prep: dirty or wet seals, wrong temperature.
- Reading trapped air as a leak.
- Skipping calibration against a known reference defect.
Building It Into a QA Program
Integrity testing works when it is routine, not occasional.
- Match the method to the format (above).
- Set parameters per package type. Vacuum level, hold time, and fluid are not universal. A common practice is to find the package’s burst point and test at about 90% of it; many North American food lines run routine checks around 12 inHg.
- Set a cadence. Start of shift, once per hour, after any material changeover or machine adjustment, and end of shift. That catches seal-bar drift while there is still time to correct the line.
- Document every test. Date, shift, lot, operator, parameters, pass or fail, and the leak location on a failure. That record satisfies audits and lets you trend toward the root cause.
- Review the cadence every three to six months. Consistent clean results can justify less frequent testing; inconsistent results mean test more.
Across food safety frameworks like HACCP and GMP, leak testing often qualifies as a Critical Control Point for vacuum-sealed meat, MAP packaging, and any product where oxygen or moisture ingress is a hazard. Medical device packaging under ISO 11607 typically requires distribution-hazard qualification, which is where altitude simulation earns its place. Retailer and co-manufacturing audits increasingly expect documented results for every batch.
Why FlexPak Focuses on Visual Detection
The FlexPak Leak Detector does not just tell you a package leaks. It shows you exactly where. The core unit runs D3078 bubble emission and D6653 altitude simulation; the FPIPA attachment adds F2096 internal pressurization for trays, pouches, and Tyvek; the VAC Attachment makes vacuum-sealed and liquid packs testable under D3078. Cycles run in about 30 seconds, and the Fully Automatic Controller (FPFA-T) stores recipes and logs data for the audit trail. One platform covers the methods most flexible-packaging QA programs need.
If you are setting up a program or want help matching a method to your format, see the FlexPak Leak Detector or request a quote.
Frequently Asked Questions
What is the difference between seal integrity and package integrity testing?
Seal integrity testing focuses on the seal itself. Package integrity testing is the broader category covering the whole sealed package, including closures and the film or substrate. Both confirm the same thing: that what is sealed stays sealed.
What size leak can bubble emission (D3078) detect?
It is built for gross leaks. Real sensitivity depends on vacuum level, headspace, product, and material. Higher vacuum detects smaller leaks in most trials, but there is no single universal figure, and small leaks can be missed at lower vacuum.
Which test is right for vacuum-sealed packaging?
D3078 with the VAC Attachment. Vacuum-sealed packs have no headspace gas to form bubbles until air is added, so testing them without inflation is a common false-pass.
How often should we run integrity testing?
A typical cadence is start of shift, hourly during production, after any changeover or machine adjustment, and end of shift. Sampling rate depends on your ISO 2859-1 level and product risk.
Is integrity testing required by regulation?
Requirements vary. FDA guidance treats package integrity as part of good manufacturing practice, ISO 11607 covers medical device packaging, and USP <1207> covers sterile pharma. Even where it is not mandated, retailers and customers often require documented testing.
About the author
Gordon Bruce is a package integrity specialist at FlexPak Leak Detectors, where he works directly with QA managers, quality engineers, and plant teams across the food, beverage, pharmaceutical, and medical-device industries to match the right seal integrity test to their packaging. He advises on the methods covered here every day, including bubble emission (ASTM D3078), internal pressurization (ASTM F2096), altitude simulation (ASTM D6653), and dry chamber testing (ASTM D5094), and on the real-world seal failures, channel defects, pinholes, and seal-bar drift, that pass visual inspection and surface later in distribution.
FlexPak has built package leak detection and seal integrity testing equipment since 2009. Gordon’s focus is practical: helping teams see exactly where a package leaks in about 30 seconds, document results for audits, and trace failures back to the line.