OEE for Carbonated Soft Drink Lines: What to Measure and Where

A CSD line is eight or more machines tied together by conveyors, and it only makes money while the filler turns.

This page shows how to set up OEE and downtime tracking for a high-speed soft drink line, and what to ask of any software you put on it.

We sell Fabrico, so test what we say like any other vendor's claims.

How a CSD line is built: the V-curve

A typical PET line runs in this order: blow moulder, rinser, counter-pressure filler, capper, warmer, labeller, packer, palletiser. A can line starts with a depalletiser in place of the blow moulder, and a seamer does the capper's job.

The mixer and carbonator sit beside the line and feed the filler bowl. On many lines the rinser, filler and capper are one block at one speed, and some PET lines join the blow moulder to that block and drop the rinser.

The filler is rated slowest on purpose. Machines before it and after it are rated faster, and the further from the filler, the faster they run.

Plot the rated speeds along the line and you get a V with the filler at the bottom. Line builders call this the V-curve.

The extra speed has one job. After a short stop, the labeller or packer can catch up and clear the conveyor before the filler has to stop.

The conveyors between machines hold bottles for the same reason. Our page on sizing conveyor accumulation covers how many minutes that buys.

A blow moulder joined to the filler has no such buffer. Every blow moulder stop is a filler stop in the same second.

Which machine defines line OEE

The filler does. It is the constraint, so a minute lost at the filler is a minute of output the line never gets back.

A minute lost at the packer is different: if the accumulation holds and the packer catches up, the line lost nothing.

So line OEE is the OEE of the filler, with the good count taken at the end of the line. Labeller and packer figures help you find causes, but they are not the line's score.

Check that the filler really is your constraint. On an old line with a tired labeller, the slowest machine in practice may be somewhere else.

Who owns a filler stop on a CSD line

When the filler stops, it is in one of four states, and each points to a different owner.

• Down: the filler or capper itself has a fault. Think of a cap chute jam, a burst bottle, or a filling valve that will not seal.
• Starved: the filler is ready, but no bottles arrive. The cause is upstream: the blow moulder, the air conveyor, the depalletiser or the rinser.
• Blocked: the filler is ready, but the discharge conveyor is full. The cause is downstream: the warmer, labeller, packer or palletiser.
• Waiting for product: the filler is ready and bottles are there, but the mixer or the syrup room has not released the batch.

All of these are lost minutes for the line, but if they all land on the filler, the filler crew gets blamed for the packer's film breaks. Keep the minutes in line OEE and charge them to the machine that caused them.

The general method is on our page about filler OEE on a bottling line.

Where a CSD line loses time, speed and bottles

This is the typical picture, from our own engineering view, not a measured ranking.

Foaming, CO2 and temperature

A counter-pressure filler first brings the bottle up to bowl pressure, then fills, then lets the pressure off in a controlled step called the snift. Each step takes time, and that time sets the top speed.

Cold product holds CO2 well. When product arrives warmer, or with more CO2 than the recipe says, it foams at the snift.

Foam means low fills and wet bottles. The usual answer on the floor is to turn the filler speed down until the foaming stops.

That is the classic CSD speed loss: the line is running, nobody logs a stop, and it makes fewer bottles a minute than it should. Crews often find that some products, sugar-free ones among them, foam more than others at the same settings.

So log product temperature and CO2 next to the filler speed if you can. A speed loss with no cause beside it will never get fixed.

Micro-stops at the labeller and packer

The labeller and packer stop often and briefly: a label splice that did not take, a fallen bottle at the infeed, a film break. The accumulation hides these stops, so the filler keeps running and nobody writes them down.

Then the tenth short stop in an hour fills the conveyor, and the filler goes blocked. The log says "filler blocked, 3 minutes", and the real cause is spread over the last hour.

That is why micro-stops on these two machines need to be caught from the machine signal, not from a paper sheet.

Planned time and ideal rate for a CSD line

Decide once where cleaning goes. Many plants take the scheduled full CIP out of planned time, and keep rinses between flavours in as changeover.

The ideal rate is the filler's rated speed for that bottle size, not the budget rate and not last year's average. A big bottle takes longer to fill, so a table like this, with your own figures, is the minimum.

If the software holds one rate for the whole line, a day of 2 L bottles looks like a bad day. It was not.

Do not lower the ideal rate for a product that foams. Keep the rated speed, and let the foaming show as a performance loss you can see and work on.

Flavour, bottle size and pack changes on a CSD line

A CSD line has three kinds of changeover, and they differ a lot in length.

• Flavour only: a rinse or a clean of the mixer and filler, a new syrup batch, new labels.
• Bottle size: blow moulds, filler and capper handling parts, labeller parts, packer and palletiser settings. This is the long one.
• Pack format only: the same bottle goes from a 6 pack to a 12 pack. Only the packer and palletiser change, and the filler can often keep running into the accumulation for a while.

Time a changeover from the last good bottle of the old SKU to the first good bottle of the new one at rated speed. If you stop the clock at the first bottle out, the slow start and the start-up rejects vanish from the record.

Keep changeover time inside planned time, as an availability loss, and log the three kinds apart. One average for the line tells you nothing about any of them.

Worked example: one shift on a PET line

These are invented round numbers, describing no plant. The line fills 0.5 L PET at an ideal rate of 500 bottles a minute, so the ideal cycle time is 0.002 minutes a bottle.

• Planned time: 480 minutes.
• Filler stopped: 96 minutes. That is one flavour changeover of 40, filler and capper faults of 20, and 36 starved or blocked.
• Filler running: 384 minutes.
• Bottles filled: 180,000, counted at the filler discharge.
• Good bottles: 175,500, counted at the palletiser. Rejects are 4,500.

Availability is 384 of 480, or 80%. Performance is 180,000 of the 192,000 bottles that 384 minutes could give, or 93.75%.

Quality is 175,500 of 180,000, or 97.5%.

Multiply them: 0.80 x 0.9375 x 0.975 = 0.731, or 73.1%.

Now the short way: 175,500 good bottles x 0.002 minutes is 351 good minutes, and 351 of 480 is 73.1%. Both routes agree.

The 36 starved or blocked minutes are bigger than the filler's own faults, and they belong to other machines.

The 24 minutes of slow running and stops too short to log are 12,000 bottles that no stop log will ever show. The full method is in our guide to OEE calculation.

What to count, and where

• Filled bottles: at the filler discharge. This is the total count for performance.
• Good bottles: at the end of the line, after fill level, cap and label checks. On most lines that means packs at the packer or palletiser, times bottles in a pack.
• Filler state: running, down, starved, blocked. The starved and blocked states usually come from the lack and back-up sensors on the conveyors.
• Run signals: from the blow moulder or depalletiser, labeller, packer and palletiser, so a blocked filler can be tied to a cause.

The pack count is where it often goes wrong. The multiplier changes with the pack format, and an older packer may give no count you can use.

That is one place for a camera that counts parts: packs at the packer discharge, when there is no usable PLC count.

Bottles thrown out before the filler, such as bad preforms, are a material loss. Track them, but keep them out of the filler's quality figure.

What to check in OEE software for a CSD line

Run these in the demo with every vendor, including us. Bring one real shift with a changeover in it.

1. Starved and blocked: ask to see a filler stop marked as starved or blocked, and the machine it is charged to.
2. Line OEE: ask which machine sets the line's OEE, where the good count is taken, and if you can change both.
3. Ideal rate per SKU: set up a 0.5 L and a 2 L product with different rates. Check that the performance figure follows the SKU that ran.
4. Speed loss: run the filler at 90% for ten minutes. Ask where that shows up, and if it is split from short stops.
5. Short stops: ask what the shortest stop is that the system records at the labeller and packer, and who can change that limit.
6. Changeovers: ask how one starts and ends, and if flavour, bottle size and pack format can be logged apart.
7. Units: ask how packs turn into bottles when the pack format changes in the middle of a shift.
8. Waiting for product: ask where minutes go when the filler waits on the mixer or the syrup room.

For a wider view of the market, see our reviews of OEE software and downtime tracking 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: detected from the machine signal.
• Losses in money: losses can be shown in money, not only in minutes.
• AI camera: an add-on that counts parts. You can open the video of a stop. In the standard setup, video is streamed, not stored.
• Your data: REST API and webhooks, and any list exports to Excel or CSV.
• Maintenance: a stop reason can be set up to open a maintenance task.

Production planning and scheduling are not part of Fabrico today. Your flavour and pack sequence stays in your planning tool.

The eight checks above apply to us too. Book a demo or Contact us.

What people ask next

Should we measure OEE on the filler or on the whole line?

Both, for different jobs: line OEE is taken at the filler with the good count from the end of the line. Machine figures for the labeller and packer help you find causes.

Is starved or blocked time an availability loss?

Yes, for the line: the filler was planned to run and did not, so the minutes count. Charge them to the machine that caused them, not to the filler crew.

Why does the filler stop when the packer has a short stop?

It should not, if the accumulation is big enough and the packer can catch up. When short stops come too close together, the conveyor fills and the filler goes blocked.

Can we track a CSD line without a PLC count?

Often, yes: a run signal and a sensor count at the filler discharge cover the basics. At the packer, a camera that counts packs is one option when there is no usable count.

Want to run these checks on your own line? Book a demo or Contact us.