Pasteurization Line Bottles: Thermal Shock and Closure Audit

A glass bottle is fit for a pasteurization line when three things have been shown against the tunnel's real profile: the glass withstands the worst temperature difference the line delivers, at the rate it delivers it; the closure holds through the pressure rise on heating and the vacuum on cooling; and the decoration survives the hot water. A heat rating in degrees alone proves none of this. The filler has to state the worst case, including a stop and restart, and the container supplier has to show test results against that case before the order is placed.

Two neighbouring subjects are handled elsewhere. The test method and the wording of a pass criterion are covered in our note on thermal shock testing of glass bottles, and removing soil, scale and residues before filling is covered under cleaning glass bottles. What follows is the thermal process side only. Any value you take from it should be treated as a reference approach and confirmed on your own line.

What the audit has to establish

Pasteurization is the only stage in a filling room that heats and pressurises a package after it has been sealed. Before that point a bottle slightly out of tolerance may still fill and cap. In the tunnel, glass and closure are loaded together, and the load keeps changing from the first heating zone to the last cooling stage.

An audit therefore answers four questions:

• Process target. What the product needs, in pasteurization units or as a time and temperature pair agreed with a food technologist.
• Delivered profile. The temperature and pressure the tunnel really applies, with ramp rates and the worst case when the belt stops.
• Package survival. Which closure, liner and decoration come through that profile and are still acceptable on the shelf.
• Records. What documents will exist afterwards that an outside party could read, re-check and argue from.

Ownership is usually split. The pasteurizer stands at the filler, a co-packer or the brand's own plant, and the glass comes from a supplier who never sees it. One party documents the process, another documents the container, and the two papers are seldom written with the other in view. If filler and glass supplier have never exchanged the worst-case numbers, nobody has yet been made responsible for the risk.

Questions that show whether a line has really been characterised

Short, precise answers matter more than long ones. Start with the equipment: tunnel or batch vessel, and how many temperature zones. A tunnel with two heating zones and no separate cooling stage behaves very differently from one with controlled preheat, two hold zones and three cooling stages, and the bottle feels that difference as ramp rate. Then ask how heat reaches the bottle: nozzle sprays, a water curtain, or full immersion.

Next comes the process calculation. Ask how the process value is derived and which thermocouple controls it. A pasteurization unit figure given without dwell time and reference temperature is not a specification.

The question answered worst is what happens at a stop. Steady running is rarely the harshest event. When the belt halts, packages in a hold zone go on receiving process and packages in a cooling zone do not; both are deviations, with different consequences. Follow up with:

• the belt speed range and the pitch for your container. Dwell is zone length divided by belt speed, so speeding up the tunnel shortens a process that food safety has fixed;
• whether pressure outside the package is controlled, or the tunnel is simply an open bath;
• the source of the cooling water and whether it is softened;
• which other products run through the same tunnel and how they are kept apart;
• who signs the daily process record, and at what point in the shift it is filled in.

Weak answers have a pattern. "We can pasteurize anything" expresses willingness, not capability. A temperature with no ramp, a process figure with no dwell, nothing on stops, nothing on pressure, traceability that ends at the whole order: the line behind such answers may be fine, but its evidence is not.

Line data to request in writing

These figures become the inputs to the bottle requirement, so ask for them on paper:

• zone temperatures and the dwell time planned for your product;
• heating and cooling ramp rates per minute, not just an overall figure;
• spray flow rate, nozzle type and nozzle spacing, or for a bath the immersion depth and circulation pattern;
• temperature and flow of the cooling water in the final stage;
• any counter-pressure or zone pressure control;
• belt speed range and the pitch used for your format;
• the process target, in pasteurization units or as an agreed time and temperature pair.

Reading a capacity figure

A tunnel is a serial process. Dwell, not filler speed, sets its output, and because the thermal process sets the dwell, the tunnel tends to be the bottleneck of the whole line. Throughput is belt speed divided by bottle pitch. A bottles-per-hour number therefore means something only alongside the format, the dwell and the hours the tunnel is truly available.

Changeovers, stoppages, cleaning cycles and the time packages spend in the cooling stages all cut into the usable rate. A tunnel serving several products gives each one only part of the schedule. Ask for the rate on your bottle, not on the reference bottle in a brochure, and take the reply as a range to confirm.

Why the curve matters more than the peak temperature

Thermal shock is a temperature difference across the container wall, applied faster than the glass can even it out. Glass is brittle and cannot yield. The surface that changes temperature first strains against the remainder of the wall, and that tension either passes through harmlessly or starts a crack. The shape of the curve decides which: how fast the wall is driven, how long it is held, and most of all how fast it comes back down.

Heating is normally the kinder direction, because the contents warm at a similar pace and back the wall up from inside. Cooling pulls the outer surface straight into tension while the inside is still expanding. This is why most field breakage happens late in a spray tunnel, where a cold jet hitting one side of a hot bottle creates a local gradient much steeper than the planned average. A bath is more even, but entry is abrupt, and leaving it for a cooling stage is a second shock in the reverse direction.

Two container properties reduce the margin before the tunnel is even switched on. Residual stress from a poorly controlled annealing curve uses up part of the glass's resistance. Uneven wall distribution focuses the gradient on the thinnest section. The glass supplier controls both, which makes the evidence below a container matter, not a process matter.

How to write the requirement

State a temperature difference the bottle must survive without fracture, applied at the rate the line delivers, covering the stop-and-restart worst case, plus a margin that filler and supplier agree. State the direction of the change and the method used to check it.

Low-acid foods carry a further duty that has nothing to do with packaging: a scheduled process with documented time and temperature control. In the United States the reference is FDA 21 CFR 113, and juice products fall under the juice HACCP framework in 21 CFR 120. We cite these as public references only.

Thermal shock evidence to ask the glass supplier for

The useful evidence is a short list of specific documents. A sentence saying the bottles resist thermal shock does not count, and neither does a result with no method named. Results from another size, another wall thickness or another production run carry partial weight. Results from a heavier development sample carry almost none.

Request these items:

• Thermal shock test result naming the method, the temperature difference passed, the direction of change, the sample count and the mould or moulds sampled, with a statement on whether the samples belong to the production run meant for your order. Our page on measuring thermal shock resistance explains the set-up and an inspectable pass criterion.
• Annealing records proving the bottles passed through a controlled Lehr curve and were not left to cool freely.
• Residual stress readings by polariscope on the batch concerned.
• Wall thickness measurements, since resistance drops as the wall gets thicker and less evenly distributed.
• Finish dimensions from the actual mould. A finish out of tolerance weakens the closure exactly when the closure is loaded.
• Filler hold tests on aggressive processes: which samples the filler has held and tested at the worst-case difference, not a nominal one.

Treat three things with suspicion: a temperature difference with no direction, because hot-to-cold and cold-to-hot are not equivalent; a pass achieved on a bottle with a much thicker wall than the one offered; and report dates that do not match the batch. Remember too that a test report is not a certificate and not an inspection report. If the real doubt is whether a document is authentic or what a certificate should contain, settle that as its own exercise with its own evidence chain.

Pressure balance and closure retention

Once sealed, the package is a pressure vessel, and the closure is its weakest part. Heat expands the headspace gas and drives dissolved carbon dioxide out of solution, so internal pressure climbs. At the same moment the liner softens and the closure's engagement relaxes. Retention depends on internal pressure set against the pressure outside the package, together with the mechanical grip of the closure system. An audit that stops at thermal shock misses this, and it is among the commonest ways a sound batch of filled bottles turns into a pallet of leakers.

How the route changes the outside pressure

In a bath, the package sits beneath a head of water. External pressure is above atmospheric, rises with depth, and helps hold the closure down. In a spray tunnel the package is surrounded by air and thin films of water at roughly atmospheric pressure, so the closure takes the full internal pressure unaided. The route that seems milder because it uses less water can thus be the harder one for the cap. Counter-pressure tunnels control the variable on purpose, which explains their use for carbonated and nitrogen-dosed drinks.

Failure on heating and on cooling

On heating, internal pressure greater than the closure's retention lifts it and breaks the seal. You see leakage in the tunnel, a crown that has turned or popped, a lug cap backed off, or a permanently deformed liner. On cooling, internal pressure drops below atmospheric and the package goes under vacuum. Some liners distort, and where the seal is already damaged, cooling water can be sucked into the headspace. That is a microbiological hazard, not just a packaging fault.

Carbonated products start with a pressurised headspace and so have less room in either direction. An overfilled bottle has less headspace to take up expansion than one filled to the intended level.

Controls and verification

The controls sit with the process, not with claims about the bottle: fill level and headspace volume, fill temperature, deaeration or nitrogen dosing where the recipe allows, a controlled cooling ramp in place of a cold shock, zone pressure control or a counter-pressure tunnel for sensitive products, and a closure system with the right liner and engagement.

To verify, run filled and closed packages through the complete curve at process temperature. After cooling, check removal torque, seal integrity and vacuum retention, and repeat at the worst case, not the average. A supplier who replies to a pressure-balance question by talking about thermal shock has mixed up two subjects, and it is worth pressing for clarification.

Spray versus immersion for decoration and cooling water

The two methods are often compared on water use alone. Three other points tell a buyer more: temperature uniformity across the bottle, how much water reaches the closure, and how much hot water the decoration must endure.

Immersion heats more evenly and is easier on glass. It also wets the entire outside of the bottle, and because bottles float, hold-down rails are needed that bring mechanical contact of their own. Spray consumes much less water and lets each zone be tuned. In exchange, nozzle condition becomes a quality variable. One blocked or worn nozzle leaves a cold patch on a hot bottle, the very gradient that cracks glass, and that package is under-processed with no alarm.

Labels and printing

The first decision is timing. Decoration put on after the tunnel is exposed to nothing; decoration put on before takes the whole load. For decoration applied beforehand, behaviour differs by type:

Cooling water deposits

Water drying on a hot bottle leaves its dissolved solids behind. Hard cooling water therefore lays calcium carbonate on the body and, worse, on the sealing surface. Consumers see the marks, but the functional harm is greater: the deposit alters friction and surface condition where the closure seats, and the result shows up later as erratic cap application torque or leakage in storage.

Prevention is water treatment and maintenance, not glass specification. It covers softening or reverse osmosis for the final rinse, filtration, nozzle inspection, blow-off after the last stage, scheduled draining and cleaning of the tunnel, and control of biological growth in warm recirculated water. Removing deposits already formed is a job for the cleaning routines upstream of the filler. Bottles that arrive clean and dry also give the tunnel much less material to deposit.

Pasteurization routes compared

Use the table as the outline of an audit note. Entries are qualitative on purpose, because real values come from your product and your tunnel. The last column lists the reply that suggests a route was never audited.

The sequence is the same for every row: identify the route, get its profile, turn the profile into a bottle and closure requirement, then verify with records that stand up to scrutiny. Write the requirement before ordering.

Products that need in-container pasteurization, and the alternatives

Pasteurizing in the container makes sense in three situations: stabilising the product before filling would harm its quality, the format makes aseptic filling impractical, or the flavour calls for a lower temperature over a longer period than hot fill allows.

Typical cases are beer and craft beer stabilised in the package instead of by flash treatment, juices and nectars including pulpy ones, lemonades and functional drinks with heat-sensitive flavours or particulates, fermented drinks destined for ambient distribution, cider, and dairy drinks in small glass. Small milk bottles illustrate the last group well. The container is small and thin-walled, so thermal load per unit of product is high next to a large format. Sizing and sales channels for that format are discussed under small glass milk bottles, and the wider choice of container starts from our glass bottle collections.

The other routes each trade something:

• Hot fill and hold. The bottle must take a fill temperature typically in the high eighties or low nineties Celsius, then a hold and a controlled cool. Less equipment is needed than for a tunnel. The thermal load arrives in one sudden event at the filler head, and the product is cooked harder, so delicate flavours suffer.
• Tunnel pasteurization. Lower temperature for longer is gentler on the product. Pressure balance, decoration exposure and cooling water quality all become live variables.
• Cold fill with preservatives, or aseptic filling in another package. No thermal load at all, at the cost of limits on recipe and equipment.

This is a product decision that has consequences for the container, not the other way round. Make it with the food technologist before freezing the bottle specification. Approval for hot fill does not carry over to a tunnel, nor tunnel approval to hot fill.

Records to audit on site

The record is where suppliers tend to be least prepared, and a thermal process is only as credible as what was written down. Ask for:

• The tunnel's daily process record: zone temperatures against time, belt speed, calculated dwell, process value achieved, cooling water temperature, calibration status and interval for the controlling thermocouples, and each stop or speed change with the position of packages at that moment.
• The handling of a thirty-minute stoppage. Packages stuck in a hold zone have had more process than scheduled and those in a cooling zone less. Each case needs a recorded decision.
• Traceability from a batch of filled bottles to its tunnel run, so a leakage complaint narrows to a process window and not a month of output.
• Container evidence filed under the same batch reference.
• The deviation procedure: whether affected packages are quarantined, re-processed or released by a named decision, and who has authority to decide.

Warning signs repeat from plant to plant. Records filled in when the shift ends instead of at each reading. One thermocouple for a multi-zone tunnel. Calibration certificates that match no instrument. No stop log. Traceability that goes no finer than the whole order. In a dispute, missing evidence decides the outcome even if the process itself was sound.

Scheduling limits of a thermal process line

A tunnel restricts the schedule in ways a filler does not. Food safety fixes the dwell, so lost hours cannot be recovered by running faster. A stoppage costs a full dwell cycle for every package inside. Changing format means adjusting rails, pitch and sometimes belt speed. Cooling stages hold packages for a further spell after the process value is reached. Add a tunnel shared between products, and you have the usual reason a date that looked comfortable on a quotation slips in practice.

Direct your questions at the tunnel, not the order:

• What tunnel window is planned for this product in the week in question, and what else competes for it?
• Which spares are on site? Nozzles, thermocouples, seals and pump parts are the ones whose absence stops a line.
• How often is calibration done, and does it take the line out of service?
• What breakage or loss allowance is planned? A thermal process adds handling steps and a small loss to the yield.
• What does the record from last season show for those same weeks? History is stronger evidence than intention.

On the glass side, the matching questions are mould availability and when the packed empties will arrive, because a tunnel window is useless without bottles on the premises. Ask for every figure as a range and check it against your own schedule.

What to send us for a pasteurization-ready bottle specification

Three inputs let us define the container requirement:

• The product: category and character, including whether it is carbonated, carries pulp or particulates, or is low-acid.
• The route: which pasteurization path the package takes, with zone temperatures, dwell and cooling behaviour if already fixed.
• The fill method: hot fill, cold fill with or without nitrogen dosing, or filling ahead of a tunnel, plus the intended closure system.

Together these set the temperature difference the glass must take, its direction and rate, the pressure balance facing the closure, and the exposure of the decoration. From them we draft a thermal shock requirement worded for a bottle specification, closure and decoration checks to carry out at process temperature and not after cooling, and a short list of evidence to travel with the batch. These are reference approaches for confirmation on your line; we do not promise a process result or a delivery outcome in advance.

Frequently asked questions

How does tunnel pasteurization differ from batch pasteurization for glass bottles?

In a tunnel, packages move on a belt through successive temperature zones, and each gets an identical profile and dwell. In a batch vessel the load stands still, the coldest spot dictates the process, and everything else is over-processed as a result. Batch fits small volumes and many formats. A tunnel fits continuous output.

Does a bottle for a pasteurization line have to differ from a hot fill bottle?

The shape may be the same; the requirement is not. Hot fill is one sudden thermal event, governed mostly by vacuum on cooling. A tunnel adds a ramp, a hold, a cooling sequence, pressure balance at the closure and prolonged wetting of the decoration. Ask for evidence against the actual profile before assuming one approval covers both.

Can a pasteurization tunnel push caps off glass bottles?

Yes. Rising headspace pressure meets a softening liner, and a spray tunnel offers little outside pressure to help, whereas a bath gives hydrostatic support. The controls are fill level, headspace, fill temperature, cooling ramp, and the choice of closure and liner.

Should labels go on before or after pasteurization?

After, where the line permits, since the label then avoids the load altogether. If labelling has to come first, wet-glue paper is the weakest under immersion, a pressure-sensitive label on wash-resistant adhesive suits better, and fired-on ceramic lasts longest. Test shrink sleeves against the real detergent and temperature.

Why do bottles break in the cooling zone and not the heating zone?

Cooling puts the outside surface in tension while the inside is still expanding; on heating, the contents support the wall. Cold spray on one face adds a local gradient well above the average, so nozzle condition and spray coverage count as process variables. Uneven wall distribution aggravates it.

Does every product need a pasteurization line?

No. Acidic sauces, jams and similar products that tolerate being filled hot and held are normally handled that way, and suitable recipes can use cold fill with preservatives or aseptic filling with no thermal load. In-container pasteurization is for products that pre-fill stabilisation would damage, formats that exclude aseptic filling, or flavours that need gentler, longer heat.