Login
Blog
OEE for Juice Lines: Where PET, Glass and Carton Lines Lose Time
In short: measure OEE for a juice line at the filler, with an ideal rate set per product and format, and decide once, in writing, how flavour changeovers and their CIP are classed.
- Pulp, viscosity and foam change the filler speed, so one ideal rate for all juices gives a wrong performance figure.
- The pasteuriser feeds the filler and the cooling tunnel buffers it. Stops on both sides must be split into starved, blocked and own stops.
- How to class a flavour changeover with CIP is the plant's choice, not a law of OEE. Our default is to keep it inside planned time as an availability loss with a standard time, and the rule must be the same on every line and shift.
- Count filled bottles at the filler and good bottles after the last reject point.
A juice line looks like any bottling line until you run one. Then the product starts to matter: it is hot or sterile, it foams, it carries pulp, and it must not sit still for long.
This page is for the person who runs a juice line in PET, glass or carton and wants an OEE number the shift will believe. It covers where the time goes, how to set the rate, and what to ask of OEE software.
We sell Fabrico, so test what we say here like any other vendor's claims.
Where a juice line loses time, speed and product
Here is a hot fill PET line in sequence, with the loss each station typically causes. A glass line swaps the blow moulder for a depalletiser, and a carton line forms the pack inside the filler.
Availability goes mostly to flavour changes with CIP, cap and closure jams, and waiting: for juice from the pasteuriser, or for room in the cooling tunnel.
Performance goes to pulp and foam. A pulpy orange juice or a thick nectar often fills slower than clear apple juice on the same valves, and operators turn the speed down to stop foam and underfills.
Quality loss clusters around restarts. The first bottles after a flavour change or a long stop are the ones that fail on fill level, mixed product, cap seal or temperature.
Why the pasteuriser and the cooling tunnel decide the stop pattern
On most bottling lines a short downstream stop costs only its own minutes. On a hot fill juice line it can cost more, because hot juice cannot wait.
When the labeller or packer stops, the cooling tunnel fills up. Once it is full, the filler is blocked and the pasteuriser has nowhere to send product.
The juice then typically goes round in recirculation or is diverted. Hold it too long and the taste and colour suffer, so some of it may be dumped.
The filler has the same limit. After a longer stop the juice in the filler bowl can fall below the fill temperature, and the first bottles after restart are rejected for it.
So one labeller stop can show up three times: as blocked time at the filler, as lost product at the pasteuriser, and as rejects at restart. Your OEE setup must be able to tie those three to one cause.
The reverse case is starvation. A pasteuriser fault, a late blend or an empty tank stops the filler with nothing wrong at the filler.
If you run a tunnel pasteuriser on glass, read our page on tunnel pasteuriser OEE for that machine on its own.
PET, glass and carton: what changes for OEE
The method is the same on all three. The machines, the count points and the usual losses are not.
| Line type | What is different | Where to count | Loss to watch |
|---|---|---|---|
| PET, hot fill | Blow moulder up front, bottle tilt and cooling tunnel after the capper | Filler out, then after the last reject point | Blocked filler when the tunnel is full, cap jams |
| PET, cold or aseptic fill | Bottles and caps are sterilised, no cooling tunnel | Filler out, then packer infeed | Long sterile restarts after a stop, so short stops cost more |
| Glass, hot fill | Depalletiser and rinser, twist-off capper, tunnel for cooling or pasteurising | Filler out, then after the tunnel and the inspector | Broken glass stops and the clean-up after them |
| Carton | The filler forms, fills and seals the pack. Straw or cap applicator follows | Filler pack counter, then tray packer | Packaging material splices, applicator misfeeds |
On an aseptic line, write down what a restart costs after sterility is lost. That one event can take longer than all the cap jams of the week.
How to set the ideal rate when pulp changes the speed
Do not use the filler nameplate as the ideal rate for all products. The nameplate is usually given for a thin, clear liquid in one bottle size.
Set an ideal rate for each product and format pair: clear apple in 1 litre, pulpy orange in 1 litre, nectar in 0.33 litre. Use the best steady rate the line has held on that pair with good fills, not the average.
Keep the nameplate as a second reference. The gap between it and the product rate shows what pulp and viscosity cost you, and that is a question for the planner and the product team, not for the shift.
Check the rate after any change that affects flow: a new pulp supplier, a different fill temperature, new valves. An old rate makes performance drift above 100% or sit low for no visible reason.
Small speed dips and short jams hide in this factor. Our page on micro-stops explains how to separate them from slow running.
How to class CIP and flavour changeovers on a juice line
OEE has no single rule for this. It is a choice each plant makes, and it moves the number more than any other setting on a juice line.
There are two common ways. Either the changeover and its CIP stay inside planned time and count as an availability loss, or the standard clean is taken out of planned time and only the overrun counts.
Our default for juice is the first: a flavour or allergen changeover, with the CIP it needs, stays inside planned time as an availability loss. It gets its own category and a standard duration per changeover type.
The reason is that the line was scheduled to produce, and the flavour sequence decided how many of those minutes went to cleaning. Take them out of OEE and the cost of short runs and a poor sequence is no longer in the number.
The table shows how each event is classed if you take that default.
| Event | Class under this default | How to record it |
|---|---|---|
| Flavour change with rinse or CIP | Availability loss, changeover | Standard time per type, for example light to dark, or dark to light |
| Allergen changeover with full clean | Availability loss, changeover | Its own standard time, longer than a flavour change |
| CIP forced by run length | Availability loss, cleaning | Own reason, so the planner sees the run length rule at work |
| Minutes over the standard | Availability loss, overrun | Split from the standard part, with a reason |
| Bottle or pack format change | Availability loss, changeover | Separate from flavour, since the work is mechanical |
| Cleaning in a slot with no production planned | Outside planned time | For example an end of week clean. Not in OEE, but report the hours next to it |
| First bottles rejected after restart | Quality loss | Count them. Link to the changeover that came before |
The other way, standard CIP out of planned time and only the overrun counted, is also defensible. It gives a higher OEE, so you then need a second figure for total cleaning hours, or the flavour sequence never gets better.
Whichever you choose, write it down and keep it the same across lines, shifts and months. An OEE trend or a line comparison means nothing if the rule changed in between.
Our page on CIP time in OEE goes through the options in detail.
A worked OEE example for one juice shift
These are invented, round numbers for one shift. They describe no real plant.
A 1 litre PET hot fill line runs clear apple juice, then changes to a clear multifruit juice. This plant keeps the flavour change inside planned time.
The ideal rate for both is 100 bottles a minute, so the ideal cycle time is 0.01 minutes a bottle.
- Planned time: 480 minutes.
- Stops: 120 minutes. Flavour change with CIP 70, cap jams 20, labeller stops that blocked the filler 15, no product from the pasteuriser 15.
- Run time: 360 minutes.
- Bottles filled: 32,400. At the ideal rate, 360 minutes would give 36,000.
- Good bottles: 31,590. Rejects were 810, most of them in the first minutes after the flavour change.
Availability is 360 of 480 minutes, or 75%. Performance is 32,400 of 36,000 bottles, or 90%, and quality is 31,590 of 32,400, or 97.5%.
Multiply them: 0.75 x 0.90 x 0.975 = 0.658, so OEE is 65.8%.
Now the check: good bottles times ideal cycle time is 31,590 x 0.01 = 315.9 minutes. Divide by 480 planned minutes and you get 65.8% again, so the numbers close.
Look at what the split tells you. The flavour change took 70 of the 120 stop minutes, and the labeller's 15 minutes were lost at the filler, not at the labeller.
Now class the same shift the other way and take the 70 minutes out of planned time. Planned time is 410 minutes, availability is 360 of 410, or 87.8%, and OEE is 315.9 of 410, or 77.0%.
Same shift, same bottles, about 11 points apart. That is why the rule has to be written down and kept the same.
The full method is on our OEE calculation page.
What to count, and where
Take the total count at the filler discharge or the capper. That is the machine that sets the rate, and it usually has a usable signal.
Take the good count after the last point where a bottle can be rejected: after the fill level and cap inspection, and after the cooling tunnel on a hot fill line.
If you count good product in cases or trays at the packer, convert with the pack size of the running order. A line that switches between 6 packs and 12 packs needs that factor to follow the order, or quality jumps around.
Some lines have no count you can use: an older glass line, a carton filler with a closed controller, or a manual packing step.
A camera that counts bottles or packs on the conveyor can fill that gap. See camera-based OEE for how that works.
Product dumped at the pasteuriser never becomes a bottle, so it is not in the OEE quality figure. Track it in litres next to OEE, with the stop that caused it.
What to check in OEE software for a juice line
Run these seven checks in a demo with any vendor, including us. Bring one real shift with a flavour change in it.
- Ideal rate per product and format. Ask to see two juices on the same line with different rates, and what happens to performance when the order changes mid shift.
- Starved, blocked and own stops at the filler. Ask how a full cooling tunnel is told apart from a filler fault, and what signals that needs.
- Changeover and CIP as their own categories. Ask to set a standard time and see the overrun split out, by changeover type.
- Planned time rules you can change. Ask to move a clean from inside planned time to outside, and see who is allowed to do that.
- Short stops. Ask what the shortest stop the system records is, and where cap jams of a few seconds end up.
- Count conversion. Ask how bottles at the filler and cases at the packer are matched when the pack size changes.
- Rejects after restart. Ask if rejects can be seen by time, so the ones after a changeover stand out from the steady ones.
For a wider view of the market, see our review of OEE software.
Where Fabrico fits
Fabrico is a cloud-based OEE and light MES platform. It runs on the web, iOS and Android.
- Stops: they come from the machine signal, through the PLC or a Fabrico IoT device. The operator adds the stop reason on a tablet or phone.
- OEE and downtime: availability, performance and quality by line and by shift. Micro-stops are detected from the machine signal.
- Losses in money: a flavour change can be shown as a cost, not only as minutes.
- Counting: the AI camera is an add-on that counts parts, for lines with no usable count. In the standard setup the video is streamed, not stored.
- Your data: REST API and webhooks, and any list exports to Excel or CSV.
- Not in Fabrico today: production planning and scheduling. Your flavour sequence stays in your planning tool.
A stop reason can also be set up to open a maintenance task, for the cap chute that keeps jamming.
How Fabrico handles each of the seven checks on your line is a fair question for the demo. If you are still deciding how much system you need, start with our page on light MES.
What people ask next
Should CIP count against OEE on a juice line?
It is the plant's choice, and our default is yes: a flavour or allergen changeover with CIP inside a planned shift is an availability loss with a standard time. Whatever you choose, keep it the same on every line and shift, and see CIP time in OEE for the other options.
Where do I measure OEE on a juice bottling line?
At the filler, because it sets the rate for the line. Record its stops as own, starved or blocked, and take the good count after the last reject point.
Why is juice filling line efficiency lower on pulpy products?
Pulp and thicker juice flow slower through the valves and foam more, so the filler runs slower to hold the fill level. Give each product and format its own ideal rate so the loss is shown fairly.
How do I cut juice production line downtime from flavour changes?
Start with the sequence: light to dark flavours and allergen products last usually need less cleaning between runs. Then time the changeover against its standard and work on the overrun.
Is hot fill or aseptic fill better for OEE?
Neither by default: hot fill lines lose time when the cooling tunnel backs up, and aseptic lines lose more when a stop breaks sterility and forces a long restart. Measure both the same way and compare the loss lists, not the single number.
Want to see your juice line's losses split this way? Book a demo or Contact us.
Related articles
Latest from our blog
Real Fabrico tiles, sample data.