Nitrogen Dosing for Glass Bottle Filling Lines

Nitrogen dosing on a glass bottle line means one of two different things, and the specification has to say which. A drip of liquid nitrogen placed just ahead of the capper evaporates inside the sealed bottle, pushing out air and leaving the container under internal pressure. A sweep of gaseous nitrogen across the open neck strips oxygen from the headspace but adds almost no pressure. Pick the route from the effect the product needs, size the dose from the headspace volume and the capping temperature, and keep the station as close to the filler and the capper as the layout allows.

Liquid drip or gaseous purge

The two routes are often presented as a cheaper and a dearer version of the same thing. They are not. Each changes the headspace in its own way, each has its own failure pattern, and the bottle leaves the capper in a different state.

A liquid nitrogen doser opens a nozzle for a few milliseconds and drops a small mass of very cold liquid into the open container. Within a second or so the liquid has boiled off, expanding to many hundreds of times the volume it had as a liquid. Air is pushed out of the neck, and once the cap is on, the gas still forming raises the pressure inside. That pressure is the reason the route exists. A bottle closed under positive pressure is stiffer than one closed at atmospheric, and it remains stiffer through cooling.

A gaseous purge feeds nitrogen gas through a nozzle or a short lance for somewhere between tenths of a second and a few seconds. The gas arrives at or close to atmospheric pressure, so the sealed bottle ends up with essentially no added pressure. Its strength is oxygen removal. A properly designed purge reaches a low residual oxygen figure deliberately, whereas a drip gets there as a side effect and less precisely. A purge will not stiffen a container and will not stop a hot-filled bottle from panelling.

The drip has an awkward tolerance structure that should be understood before it is written into a specification. A fixed mass of liquid expands into whatever volume the fill level leaves, so the resulting pressure moves with headspace volume and with temperature at capping. A slight shift in fill level gives a proportionally bigger shift in pressure. Fill level and product temperature happen to be the two variables a line holds least tightly. The specification has to absorb that, by tightening the fill level, by aiming the dose at a slightly conservative target, or by accepting a broader pressure band and qualifying the bottle against it.

Many lines run both routes in sequence. A light purge strips oxygen, then a small drip supplies pressure. The arrangement works on two conditions: the drip station comes after the purge, never before, and the purge flow is gentle enough not to blow product or liquid nitrogen out of the bottle.

Where a project has an oxygen target and a pressure target together, write them separately, give each to one route and give each its own tolerance. One dose figure asked to deliver both tends to hit one at the expense of the other, and not consistently.

Five quantities the specification has to define

Headspace is the part of the container the product does not fill. Suppliers use the word loosely, which is how two of them can each promise to treat it and then argue at commissioning. A geometric definition avoids the argument: the headspace is everything above the fill level when the closure goes on, neck included, plus the space under the cap once it is seated. Five quantities describe what happens to that volume, and each is measured somewhere different.

• Headspace volume, in millilitres, usually also given as a percentage of nominal capacity. Fill level and internal bottle shape fix it. A long, narrow neck gives a large headspace for the size of the bottle; a wide jar filled near the rim gives a small one. The neck deserves most attention because it is the narrowest, coldest and most exposed section, the place where gas actually moves, and the surface the closure seals on.
• Residual oxygen, as a percentage of the headspace gas or as milligrams of oxygen per container. It is distinct from dissolved oxygen, which is settled at the filler and deaerator, and from total package oxygen, which adds whatever migrates through product and closure over time. A shelf life target and a residual oxygen figure only compare if they are stated on the same basis.
• Headspace pressure, in bar gauge at a named temperature. It matters right after capping, when a hot product cools and the gas contracts, and again in storage and transit as temperature swings. Pressure-rated and non-pressure-rated bottles are specified against entirely different values.
• Dose, the nitrogen delivered to each container in each cycle. For a drip it is a mass that turns into a gas volume. For a purge it is a flow, a pressure and a time. The two units do not convert into one another, and buyers who try to carry a specification across from one route to the other usually stumble here.
• Dwell and timing, the interval from dosing to a fully seated closure, together with the time the gas has to mix. Dwell decides whether the gas is still in place when the bottle is sealed. It also changes on its own whenever line speed changes, so it is the quantity most often altered without anyone deciding to.

Fix them in that order. The bottle and the fill level give the headspace volume, and nothing at the dosing head can alter it. Residual oxygen follows from volume and dose, pressure from dose and temperature, and dwell from the layout. Specifications that begin with a dose figure instead of the headspace volume generally get rewritten at the trial.

Products that benefit, and products that do not

More products are said to need headspace treatment than actually gain from it. What usually separates the two groups is whether a shelf life claim exists, not what the product is.

Strong case on oxidation

Fruit juices and juice drinks that claim a colour or a vitamin content lead the list, since vitamin C oxidises readily and the loss can be measured. Ready-to-drink teas, pale ones especially, darken and change flavour in ways consumers pick up. Cold-pressed and unrefined edible oils turn rancid, and in a small bottle the headspace is a meaningful share of the oxygen present. Nut, seed and grain products follow the same pattern. Any label carrying an oxidation-sensitive nutrient claim is a candidate, because that claim is a commercial commitment.

Case that is really about pressure or shape

Hot-filled products in light containers need headspace pressure to balance the contraction of the contents as they cool, or the walls panel. Thin-walled bottles on fast lines gain stiffness from a small positive pressure and suffer less damage downstream. Neither case involves oxidation. Writing a residual oxygen target for them pulls attention away from the pressure figure that governs the outcome.

Weak or absent case

Four situations rarely repay the effort:

• products sold within days of filling, before oxidation has time to show;
• products whose own dissolved oxygen dominates, leaving the headspace a minor contributor;
• container and closure combinations that already hold oxygen ingress very low;
• products whose flavour rests on a component that oxygen does not affect, where the figure improves easily and nobody can taste the difference.

For these, work on the closure, the fill temperature or product deaeration first and return to the headspace later.

A control batch settles the question before any equipment is committed. Fill untreated bottles next to treated ones, store both sets under matched conditions, and measure the attribute behind the shelf life claim at identical intervals. Curves that diverge show the product can see the treatment. Curves that stay together mean any justification has to come from container stiffness or from a regulatory or customer requirement, not from an unproven shelf life gain.

What a residual oxygen figure says about shelf life

Buyers often state a shelf life and expect a residual oxygen number to drop out of it by conversion. The link is real, but it runs through the product, and it differs from one product to the next.

Oxidation rate depends on how much oxygen is available, how sensitive the product is, and how quickly oxygen reaches the sensitive fraction. Headspace oxygen speaks only to the first of those, and only to the oxygen present in the headspace at sealing. Three other sources lie beyond anything a dosing head can touch: oxygen dissolved in the product, oxygen entering through the closure during storage, and oxygen already bound in the raw materials. Leave them out, and the specification passes on the line and fails on the shelf.

What headspace treatment does dependably is take a fixed amount of oxygen out of the system on day one, so the product never has to consume it. Whether that buys the claimed shelf life depends on the share of the total oxygen load it represents. If the headspace is a large share, the gain is large. If it is a small share, the same reduction hardly registers, and effort is better directed at the closure, at deaerating the product or at storage temperature.

Three rules follow for writing the target:

1. Derive it from a stability trial on your own product, not from a supplier's general recommendation.
2. State the time of measurement alongside the figure. Headspace oxygen climbs after capping as oxygen leaves the product and enters the gas.
3. Pair it with a closure specification. A closure that admits oxygen cancels the dosing step during storage, and the dosing machine takes the blame.

When stiffness is the aim, substitute pressure for oxygen and the logic holds. Capping pressure is only where the bottle starts; what counts is the pressure left once the liner has crept and the temperature has cycled several times. In both cases, write the specification around what the product experiences at the customer.

Reference conditions by headspace state

The table is arranged by headspace condition, since that is what product, container and fill level actually select; machine type comes afterwards. It gives no dose values. The right figure depends on headspace volume, capping temperature, line speed and closure, and has to be confirmed by a trial on your own bottle.

Pressure-rated and non-pressure-rated bottles

An identical dose does a different job depending on whether the bottle was designed for internal pressure. A specification that ignores the distinction ends up applied to the wrong container family.

A non-pressure-rated bottle is made for still product at roughly atmospheric pressure. Handling and stacking loads set its wall thickness distribution, and the base is normally simple. Any pressure from nitrogen has to stay below the point of permanent deformation and must not alter how the bottle sits on a conveyor or in a pack. The gain is modest and specific: slight positive pressure that stiffens the bottle for handling, quietens rattling contents and offsets a little cooling contraction. The dose is small and the tolerance narrow, because not much room separates no effect from deformation.

A pressure-rated bottle is designed against a stated internal pressure. Its base is shaped to resist that load and its walls are thicker where the load concentrates. Here nitrogen can maintain a properly pressurised headspace, as a carbonated product requires, and the designed margin permits a larger dose. The constraint shifts from the glass to the closure and finish, which become the weakest point in the pressure path.

Deciding which family applies takes one comparison. If the pressure the process needs falls inside the container's design range, the glass is not the limit and attention belongs on the closure. If the process needs more, do not lower closure torque or loosen the dose tolerance. Review the container, because a bottle deforming under its own pressure fails invisibly on the line and obviously at the customer.

Strength under pressure is also not one number. Body diameter, wall thickness distribution, base design, the presence of a punt and local shoulder geometry all contribute. A narrow, thick-walled bottle copes far better than a wide, thin-walled jar, and wall thickness alone does not explain the gap. So the specification has to name the actual bottle, not a container family, and a pressure target borrowed from a different bottle is only a place to start. When we check a bottle drawing for a dosed project, these are the features we look at.

Destination catches projects late. Pressure in the bottle tracks the temperature it reaches in storage and transit. A bottle that is comfortable at the filling plant can be marginal in a hot climate or in a shipping container standing in the sun. A band written for one market does not automatically hold for another, so the destination is a legitimate input.

What the dose asks of the closure and finish

The decision concerns gas, yet it is the seal that carries it out, and the seal most often caps the result. This article covers only what the headspace requires of the closure; head geometry, cap presentation and applied torque are dealt with in the guide to capping machines for glass bottles.

A dose that raises internal pressure asks the closure to retain it. Whether it can depends on liner material, on how far application torque compresses the liner, and on how regular the finish really is as opposed to what the drawing says. A liner that serves an atmospheric fill may relax or creep under sustained pressure. The symptom is pressure fading over days, not a leak at the capper.

The sealing basis differs by closure type. On a metal lug or twist-off lid it comes from lug engagement and liner compression. On a plastic screw cap it comes from thread engagement and resistance to liner creep. Name the closure and understand how it seals before any pressure target is written. Otherwise the target is missed at the customer, where nobody is watching for it.

Headspace height belongs in the same part of the specification. The gap between product surface and closure is the space a purge must sweep and a drip must fill. Fill nozzle, fill temperature and internal bottle geometry all move it. Since it enters the dose calculation directly, treat it as a controlled dimension with a tolerance, exactly as fill level is treated.

Hot fill and the pressure the bottle keeps after cooling

A hot-fill line changes conditions inside the bottle twice, so the headspace step must be specified for the full temperature excursion and not just for the instant of capping.

At capping the product is hot, and the headspace gas is expanded and saturated with vapour. Cooling then shrinks the product and condenses the vapour, and both pull pressure down. With nothing to make up the lost volume, pressure drops below atmospheric and the walls move inward. A light container with a large flat panel can distort visibly, and the original appearance does not come back. A liquid nitrogen dose ahead of the capper starts the bottle above atmospheric, so cooling uses up a surplus.

That changes three lines of the specification:

• Size the dose against the contraction cooling will cause. Fill temperature, the thermal expansion of the product and final storage temperature determine it.
• Tighten the tolerance compared with a non-cooling application. Too low a start and the bottle panels; too high and closure and base are stressed.
• Verify after cooling. Pressure at the capper is not the pressure the bottle will hold.

The cooling method counts too. Spray and immersion cooling load the container differently and reach the headspace at different rates. Under aggressive cooling the peak internal pressure may occur during cooling, not at capping, since the contents lose heat more slowly than the gas above them. Check this on the real bottle before the dose is fixed.

Thermal steps after capping alter the headspace again, because the bottle is heated and cooled with the gas sealed inside. Temperature and time windows, and the pressure balance between container and process, are covered in our guides to the pasteurization line and the sterilizer tunnel.

Where the dosing head sits between filler and capper

The correct position is described by distance and time, not by a machine number: directly after the filler discharge and directly before the capper infeed, with the shortest workable transfer on either side. Both routes depend on it. Liquid nitrogen has to boil off inside a closed bottle to build pressure, and an open bottle that waits too long simply vents the nitrogen. A purge displaces air only while the neck is open, and a long run to the capper lets air diffuse back, so the oxygen target is missed even though the machine delivered its set quantity.

The layout has to settle three dimensions.

Filler discharge to dosing head

This run must be long enough for the fill to settle and for foam from a fast fill to collapse, yet short enough that the product surface is where the calculation assumes. Dose a foaming product while its surface is still rising and the headspace at capping will differ from the one the dose was sized for.

Dosing head to capper

This sets the flash time and decides whether a drip is usable at all. The liquid needs time to evaporate and mix before the cap seats, but not so long that gas is lost or a pressure peak lifts the cap from its thread. With the capper mechanically close to the filler, a drip is easy to place. With a checkweigher, a reject station or an accumulation loop in between, the transfer lengthens, and either the route becomes a purge or the layout is reviewed before any machine is ordered.

Head height over the opening

A full bottle has its product surface nearer the rim than a short-filled one. A head set for the target fill level is therefore too close or too far once fill level drifts. This ties the dosing step to filler accuracy: a wide fill-level spread forces a wider dose tolerance, and the dosing head is blamed for variation it did not create. Fill level, valve behaviour and the rinse and purge sequence inside the filler are treated in the article on the glass bottle filling line.

Treat filler, dosing head and capper as a single mechanical group, joined by transfers as short and as positively controlled as the line can manage. Where that cannot be done, a gaseous purge with a longer, gentler window is frequently the sounder choice than a drip whose dwell the line cannot hold. Everything after the cap is on, from forming the tray to building the pallet, is a separate subject covered under tray forming for glass bottle packs.

Setting the dose and verifying it in production

A dose is inferred, not read off a dial. The machine delivers a mass or a flow, whereas the specification concerns the headspace at sealing and the bottle a day later. Neither can be observed at the nozzle.

Work the calculation backwards from the requirement to the gas quantity. For a purge, headspace volume and target residual oxygen give the nitrogen volume that must pass through the bottle to replace the air. For a drip, the required pressure rise and headspace volume give the mass of liquid that must evaporate. Both need the temperature at capping, since gas laws take no notice of a panel set point. Run the numbers before ordering equipment. Now and then they show that the bottle and fill level already chosen cannot reach the target, and paper is the easiest place to learn that.

On the machine, the sequence is:

1. Start from a dose the verification method can clearly detect.
2. Run a defined number of containers.
3. Measure those same containers.
4. Adjust in steps until readings sit in the middle of the tolerance band.

Centring matters. A dose left at the edge of the band passes commissioning and then fails soon after, because every later drift carries it outward.

Choose the verification method before the machine. No other decision constrains the nozzle and instrumentation as much.

Headspace oxygen

This is the most direct check and the one most specifications cite. A sample bottle is pierced or fitted with a probe, and headspace gas is drawn off and analysed. The test is destructive or nearly so, which calls for a defined sampling plan, and the sample must be handled so that room air does not mix in while it is taken. A reading an hour after capping will differ from an immediate one, and the later value is what the product lives with.

Internal pressure

Used on the drip route, it shows whether the dose is producing the mechanical effect it was chosen for. Sealed bottles are measured at a defined temperature and judged against a band, not a single value. As pressure depends on headspace volume, the bottles tested must be at the fill level the line really achieves.

Fill level and weight

Fill level belongs in the plan even though no gas is measured. On a drip route it feeds the dose calculation, so a periodic fill-level check is in effect a check on headspace volume and on the pressure to come. Leaving it to the filler team alone is a dependable way to be surprised by a pressure result.

Weighing the sealed bottle is an inexpensive extra. A checkweigher after the capper cannot distinguish dose mass from product mass with useful precision on a normal specification band. It does catch gross faults: no dose on every bottle, a head that has stopped firing, a systematic drift in fill level. The verification plan should state what the checkweigher is expected to detect and what it is not, so a passing weight is never read as a correct dose.

Retrofitting a dosing head to an existing line

Most dosing heads go onto lines commissioned without one. Five conditions decide whether the retrofit works, and the machine itself is not among them.

• Space and access. This rejects more retrofits than anything else. The head needs a mounting point with short transfers both ways, and room to adjust it, replace nozzles and clean. A head nobody can reach goes out of adjustment and stays there.
• Gas supply. A drip needs liquid nitrogen, a suitable delivery arrangement and a controlled evaporation path to the nozzle. A purge needs a steady gas flow at stable pressure. Pressure and flow must be present where the gas is used, since line losses and other consumers on site set the reading at the nozzle. Purity matters to a purge with a low oxygen target, because the impurity is the very thing being removed.
• Safety. Nitrogen is non-toxic, and that is the hazard: it displaces air with no warning. A leak in an enclosed space produces an oxygen-deficient atmosphere nobody can smell or see. The retrofit must include ventilation or oxygen monitoring suited to the room, a defined working procedure, and training for operators and maintenance staff. Agree this with the plant safety function before ordering.
• Line control. The head must fire on a bottle, which requires a reliable container-present signal and a line speed held within a range. If speed varies widely, either the dose window widens and the result spreads, or the line runs in a narrower band while dosing. When a machine is offered with a wide speed range, ask what happens to dose tolerance across it.
• Measurement. The plant needs a headspace oxygen probe and a pressure gauge, or an arrangement with a laboratory, before the installation can be qualified. The sampling plan is agreed ahead of the trial. With no measurement, the only verdict is an absence of complaints, which is slow and unreliable.

A line that meets all five can normally take a head without disturbing filler or capper, since the head is simply a station placed between them. A line that misses one tends to end with a transfer longer than the route tolerates or a headspace specification nobody can verify. Either outcome is worse than not dosing.

Fault-finding order when the target is missed

A dosing step that stops meeting its target nearly always has one of a short list of causes. Checking them in a fixed order yields a named cause, where random adjustment yields a second fault.

First: timing, speed, fill level, temperature

These explain most cases and are quick to check.

• Timing drift is by far the most frequent. Nozzle opening duration, the trigger point of the container-present signal and transfer time to the capper all move with wear, belt tension and speed changes. Each alters the effective dose while the setting stays untouched. Compare timing with the commissioning record before altering any dose parameter.
• Line speed. A head set at one speed gives another effective dose at a different speed, since dwell and flash time both shift. On a multi-speed line the specification must name the speed the dose is valid at and say what applies elsewhere.
• Fill level. On a drip route a fill-level drift presents as a pressure defect with nothing changed at the head. No fault is more often wrongly pinned on the dosing machine.
• Product temperature. A different fill temperature alters vapour pressure in the headspace and the way a liquid dose evaporates. Seasonal movement in incoming product temperature is a common, unnoticed source of a wandering result.

Second: nozzle and gas supply

• Blocked or iced nozzle. Ice or frost on a liquid nitrogen nozzle cuts delivery below the setting, and the shortfall grows gradually. Some frost is normal in operation; the machine's own provisions have to manage it.
• Supply condition. Look at pressure at the point of use, contamination, and delivered purity. On a purge with a low oxygen target, a minor impurity is enough to shift the reading.

Third: closure and container

• Closure change. A new liner, cap or application torque alters how much of the dose survives the following days. Capping pressure looks the same; later pressure does not. A failure that appears days after filling points here.
• Container change. A bottle from another mould or another supplier differs in internal volume and wall thickness distribution, which alters both headspace volume and tolerable pressure. Make a dose review mandatory whenever the bottle changes.

Last: the measurement

A new probe, a new sampling time or a new basis for the oxygen figure can make a sound process look faulty. Confirm that readings are taken as the specification describes before deciding the process has moved.

Keeping to this order converts a vague quality complaint into a documented cause. It also prevents a familiar mistake, in which the dose is raised to mask a fill-level drift and bottles reach the customer over-pressurised.

Frequently asked questions

What does nitrogen do in the headspace of a glass bottle?

Up to two things. Gaseous nitrogen swept over the neck replaces air and so lowers the oxygen the product meets in storage. Liquid nitrogen boils off inside the closed bottle, displacing air and also leaving positive pressure. One effect addresses oxidation; the other addresses stiffness and shape. The specification should name which is being bought.

Which route should I specify?

Let the required effect decide. For an oxygen target in a bottle that holds no pressure, a purge is direct and simpler to verify. For a bottle that must hold pressure, as with a carbonated product or a hot fill prone to panelling, only the drip delivers it. Lines needing both purge first and dose second, keeping the transfers to the capper short.

How much nitrogen does each bottle need?

No general figure exists. The quantity comes from headspace volume, the oxygen or pressure target, and capping temperature, so the calculation begins with the bottle and fill level, not the machine. The answer is a gas volume over a set duration for a purge, or a liquid mass per cycle for a drip.

Will dosing extend shelf life?

It removes one contributor to oxidation among several. The oxygen taken from the headspace is a large part of the total for some products and a minor part for others, and dissolved oxygen, closure ingress and oxygen bound in raw materials are untouched. A stability trial against an untreated control, stored identically, is the dependable way to find out for a given product.

Can an existing line take a dosing head?

In most cases, provided the five retrofit conditions above are met: room and access between filler and capper, gas of the right pressure and purity at the point of use, safety provisions for an oxygen-displacing gas, stable line speed, and a means of measuring the result under an agreed sampling plan.

Why does pressure fall after filling?

Cooling and closure behaviour account for most cases. A hot-filled bottle loses pressure as product contracts and vapour condenses; the dose exists to offset this, so a larger fall than expected suggests fill temperature or dose has shifted. Liner creep causes a slow decline over days. If the loss shows up after storage, review the closure and liner first.

Can pressure-rated and non-pressure-rated bottles share one pressure target?

No. A non-pressure-rated bottle is built for still product near atmospheric pressure, so its dose is capped at what the glass holds without permanent deformation, and the working window is narrow. A pressure-rated bottle has a design margin the other lacks, and on it the closure and finish, not the glass, usually set the limit.

What should a project brief include?

Send the product type, the headspace condition you want, line speed, and the bottle and closure you plan to use. Add whether the product is oxygen sensitive, whether the bottle must hold internal pressure, and the fill temperature, since those three inputs move the dose more than any others. With that we can indicate whether purge, drip or a combination fits, where the transfer allows the station to sit, and which verification method applies. Any figure that depends on the bottle run or the trial performed is given as a range to confirm on your own bottles.