A roaster in Colombia switched to nitrogen flushing last year for his retail coffee bags. He bought a flow meter, set the regulator to 2 bar, and watched the bags puff up on the line. Six months later, importers in Spain started sending back lots with a stale, papery taste.
His residual oxygen reading was 3.8%. Target was under 2%.
The machine was flushing. The nitrogen was flowing. What was missing was the part nobody sells you with the machine: knowing where the oxygen comes in, and measuring it at the right place.
Oxygen does not only come from the headspace
Most buyers assume residual oxygen equals the air left in the bag at sealing. That is one source out of five, and on a well-set line it is rarely the biggest.
Air carried in by the film. A laminate roll holds a boundary layer of air against both surfaces. At 60 bags a minute, that air goes into the bag ahead of your nitrogen. Slow, wide nozzles positioned after the forming tube push most of it out; a single narrow tube aimed at the seal pushes almost none.
Air dissolved in the product. Ground coffee, milk powder, and cocoa hold air in the particle structure. It migrates into the headspace over the first 48 hours after sealing, which is why a bag measured at the factory reads 1.2% and the same bag reads 2.6% a week later. Measure again after 24 hours, not only on the line.
Film permeability. OPP/PE has an oxygen transmission rate around 1,500–2,500 cc/m²/day. A metallised or foil laminate is 1–5. If your product needs 12 months, the film choice matters more than the flushing pressure, and no amount of nitrogen compensates for the wrong structure.
Micro-channels at the seal. A seal with product in it leaks slowly. This is the link between gas flushing and sealing quality: a bag that passes a squeeze test can still exchange gas through a 0.1 mm channel in the seal. We covered the mechanical causes in [why bags leak](https://jingrui-packingmachine.com/bag-seal-failure-causes/).
Exhaust from the product itself. Freshly roasted coffee releases CO2 for days. That is why roasters use one-way degassing valves and why the residual oxygen reading climbs in the first few days regardless of flushing quality.
What residual oxygen actually buys you
The numbers are product-specific, and it is worth knowing them before you spend money on gas equipment.
Roasted coffee holds up best under 2% residual oxygen, with 1% giving a noticeably longer window on the shelf. Milk powder and infant formula targets under 2%. Nuts and oily snacks need under 1–2% because fat oxidation is what turns them rancid. Dried fruit tolerates 3–5%. Bread and pastries gain almost nothing — they fail from moisture, not oxygen.
If your product sits in the 3–5% tolerant group and sells within eight weeks, flushing equipment is a cost with no return. This is the case where we tell buyers to spend the money on seal consistency instead.
How to measure it properly
The reading on the line is usually wrong, in one of two directions.
Sampling at the seal with a needle lets outside air in along the needle, giving a low reading. Sampling the top of the bag gives a high reading because CO2 from the product has risen there. The reliable method: let the bag rest 20–30 minutes, insert the needle through a septum patch in the middle of the bag face, draw 20 ml, and discard the first draw before reading the second.
Then repeat on a bag that has sat 24 hours. The gap between the two readings is your product’s dissolved-air release, and it tells you whether you need to flush harder or simply accept a higher starting number.
What to change, in order of cost
Cheapest first, because buyers often start with the expensive option.
Reposition the nozzle. A wide, low-velocity nozzle sitting just above the product level, with the fill sequence slowed by 100 ms, often drops residual oxygen by 1–1.5 points at zero cost.
Add a second flush point on the film path. On a [VFFS powder line](https://jingrui-packingmachine.com/product/vffs-powder-packing-machine-flour-packing-machine/) this is a small bracket and a tee, and it clears the boundary layer before the bag forms.
Switch to a foil or metallised laminate. This changes permeability by two or three orders of magnitude, and on a 12-month shelf-life product it matters more than the flushing hardware.
Add vacuum-assisted flushing. Vacuum-then-gas machines pull the headspace down before filling, reaching under 0.5%. On [pre-made pouch lines](https://jingrui-packingmachine.com/product-category/pre-made-pouch-packing-machines/) with a vacuum station this is standard; on a VFFS it needs a different sealing sequence and roughly doubles cycle time.
For stick packs, gas flushing is usually pointless — the headspace is too small to matter, and the film is the limiting factor. That is why [multilane stick lines](https://jingrui-packingmachine.com/product-category/multilane-stick-packing-machines/) rarely carry nitrogen hardware.
Send us your product, target shelf life, and current film structure, and we will tell you which of the five sources is costing you the most before you buy another regulator.
FAQ
Why is residual oxygen higher a week after packing?
Because air dissolved in the product migrates into the headspace after sealing. Ground coffee, cocoa, and milk powder hold air in the particle structure, so a bag measured at 1.2% on the line can read 2.5% a week later. Measure 24 hours after packing to get the number the customer will actually see.
How much nitrogen pressure should I use?
Enough to displace the air, not enough to disturb the product. Most VFFS lines run 0.3 to 0.8 bar at a wide, low-velocity nozzle positioned just above the fill level. Higher pressure creates turbulence at the seal and can cause product to be blown into the seal zone, which causes leaking bags rather than fresher ones.
Is gas flushing worth it for stick packs?
Rarely. The headspace in a stick pack is a few millilitres, so the absolute oxygen volume is tiny, and the film’s permeability dominates shelf life. Money spent on flushing hardware is usually better spent on a foil laminate for stick formats.
