PET flake is post-consumer bottle material that has been sorted, crushed to 12–20 mm, washed and dried, then sold by grade rather than simply by the tonne. Food, fibre and strapping grade all come off the same kind of line — what separates them is how much contamination survives it.
That distinction matters more than most first-time buyers expect. Two plants can run the same PET bottle washing line on the same baled input and sell their output into completely different markets at completely different prices. The gap is not the machine list. It is input control, and it is the handful of steps where contamination is either caught or locked in for good.
The Three Grades PET Flake Is Sold In
Flake is not a single product. Before you plan a line, decide which of these three you are actually selling into, because it sets the specification everything else has to meet.
Food grade — the strictest, and not a claim to make lightly
Food-grade PET is destined to become bottles or containers again. Washing alone does not get you there. The process itself has to be assessed by a regulator: in the United States that means a favourable opinion from the FDA on recycled plastics for food packaging, and in Europe an evaluation of the recycling process by EFSA under the food contact materials framework.
What that means in practice is traceability of the incoming bales, tighter decontamination than a standard hot wash, and documentation. A washing line can produce flake clean enough to be a candidate for food-grade use, but the line does not certify itself. Any supplier telling you a machine is “food grade” out of the box is describing something that does not exist. The demand behind this grade is worth understanding separately, because it is driven by recycled-content rules rather than by price — see PCR resin and the regulations behind it.
Fibre grade — the largest market by volume
Most PET flake in the world goes into polyester staple fibre: textiles, carpet, wadding and non-woven. This market tolerates colour variation that food grade cannot, which is why mixed-colour and light-blue flake still sells. What it does not tolerate is moisture and inconsistent viscosity, because both cause breaks during spinning. Fibre buyers care far more about batch-to-batch consistency than about the flake looking bright. If you have not yet identified who takes your output, the article on where to sell recycled plastic granules and flake maps the buyer types before you commit to a grade.
Strapping grade — the lowest threshold to enter
PET strapping band takes the widest range of input. Colour is largely irrelevant, and the tolerance for minor contamination is higher than in the other two. It is usually where a new plant starts while it proves out its feedstock supply and its washing consistency. The trade-off is price: strapping grade sits at the bottom of the ladder, so it is a place to begin rather than a place to stay.
What Separates Grade A Flake From Grade C
Buyers do not judge flake by looking at it. They judge it against a spec sheet, and it is worth knowing which numbers they check before you build a line to hit them. The figures below are the output specification for a PET bottle-to-flake line of this type.
| What the buyer measures | Typical requirement | Why it decides the grade |
|---|---|---|
| PVC content | Below 100 ppm | Degrades PET during heating and cannot be washed out |
| PE / PP content | Below 100 ppm | Caps and label residue that survived separation |
| Total impurity | Below 200 ppm | Everything else: paper, sand, metal fines, glue |
| Moisture | Below 1–2% | Causes hydrolysis and viscosity loss during extrusion |
| Flake size | Below 14–16 mm | Uneven size feeds unevenly and melts unevenly |
| Colour consistency | Batch dependent | Decides whether the flake can go to fibre or only to strapping |
Two things about that table are worth saying out loud. First, the tolerances are in parts per million, not percent — a handful of the wrong bottles in a bale is enough to move the number. Second, the real quality of any batch also depends on the incoming material, so a specification describes what a line is capable of, not a guarantee attached to whatever arrives on the truck.
Why PVC Is the One Contaminant You Cannot Wash Out
Every other contaminant on that list has a stage that removes it. PVC does not, and the reason is density.
A sink-float separation tank works by density and nothing else. Polypropylene caps at 0.90–0.91 g/cm³ and polyethylene label film at 0.92–0.96 float; PET at 1.38–1.40 sinks; the two streams part company in the water. PVC sits at 1.30–1.45, which overlaps PET. The stage that removes almost everything else is blind to the one contaminant that does the most damage.
That is why PVC has to be caught upstream, on the sorting belt with metal detection, before the material reaches the crusher. Once a PVC bottle has been crushed into the same 12–20 mm flake as the PET, it is no longer separable by any stage in the line. During extrusion it releases hydrogen chloride that attacks the PET chain and pulls down the viscosity of the entire batch.
The practical rule is simple: whatever you cannot sort out before the crusher, you will be selling.
The Steps That Decide Your Grade
Running through a plastic recycling washing line in the order the material meets it, these are the stages where the grade is actually won or lost.
Bale opening and the trommel screen. The bale is broken and the rotary drum takes out stones, sand and loose caps. Nothing sophisticated, but the material that leaves here is what every later stage has to cope with.
Label removal, before crushing. A dry friction system removes 80–90% of labels and caps while the bottles are still whole. The sequence matters: a label attached to a whole bottle is a large object that friction can strip, whereas the same label crushed into flake becomes thousands of fragments the size of the flake itself.
Sorting belt with metal detection. This is where PVC, coloured bottles and metal come out. It is the last point at which they can be removed at all.
Wet crushing. A crusher with a force feeder reduces the bottles to 12–20 mm flake with water already in the chamber, which keeps dust down and starts the cleaning early.
Friction washing. A high-speed friction washer scrubs off what is stuck on by mechanical action rather than chemistry.
Hot washing at 85–95 °C with 1–3% alkali. This is the stage that removes glue, oil and adhesive residue. It is also the most expensive stage to run, which is why it is worth understanding when hot washing is actually necessary and when cold is enough before specifying it.
Dewatering and drying. A horizontal centrifugal dryer at 1,500 rpm brings moisture down to 0.5–2%, and a hot-air dryer finishes below 1%. A zig-zag air classifier takes out the last of the label dust.
What a PET Line Consumes per Tonne
The specification tells you what the line can produce. This tells you what it costs to keep producing it, and the figures move enough across the capacity range to change which projects are viable.
These are installed power, operators and water draw for a PET bottle-to-flake line at each of the five capacities it is built in, divided by rated throughput.
| Throughput | Installed power | kWh per tonne | People per t/h | Water, m³ per tonne |
|---|---|---|---|---|
| 500 kg/h | 150 kW | 300 | 8.0–12.0 | 8.0–10.0 |
| 1,000 kg/h | 250 kW | 250 | 6.0–8.0 | 8.0–10.0 |
| 1,500 kg/h | 300 kW | 200 | 5.3–6.7 | 5.3–8.0 |
| 2,000 kg/h | 400 kW | 200 | 5.0–6.0 | 5.0–7.5 |
| 3,000 kg/h | 500 kW | 166 | 4.0–5.3 | 6.7–10.0 |
Three things in that table are worth reading carefully.
Energy per tonne falls by 44% across the range, but not evenly. The drop from 500 to 1,500 kg/h is steep — 300 down to 200 kWh. Between 1,500 and 2,000 it does not move at all. The gain resumes only at 3,000. If you are sizing a line on running cost, 2,000 kg/h is the tier that buys you throughput without buying you efficiency.
Water per tonne does not scale, and at the top it goes back up. It improves from 8.0–10.0 down to 5.3–8.0 m³ in the middle of the range, then returns to 6.7–10.0 at 3,000 kg/h. Planning to dilute a water bill by building bigger does not work on a PET line. If water supply or effluent capacity is the constraint at your site, it constrains you at every tier.
Installed power is not metered consumption. It is the sum of every motor’s nameplate rating, and a running line typically draws 60–70% of it, because motors are sized with margin and not everything is loaded at once. The per-tonne figures are correct for comparing tiers against each other and too high if read as an electricity bill.
Why PET Lines Need More People Than Rigid Lines
The labour column above is the one that catches people out, because it is the only consumption figure on a PET line that is worse than every alternative.
Compare like for like at 1,500 kg/h. A PET bottle line needs 5.3–6.7 people per tonne per hour. A rigid line running drums and crates at the same throughput needs 2.1–3.6. In absolute terms that is eight to ten operators against three to five, for the same tonnage out of the door.
The difference is not spread across the line. It is concentrated at one stage: the sorting belt. A rigid stream arriving as drums and crates is largely self-identifying — the form tells you the polymer, and the contaminants are dirt rather than other plastics. A bottle bale arrives with PVC, PLA, coloured PET, metal and mixed polymers all in the same shape and the same size, and separating them is a judgement rather than a measurement. Optical sorters help, but no plant selling into fibre or food grade relies on them alone.
Two consequences follow, and they point in opposite directions depending on where you are.
In a low labour-cost economy this is a mild disadvantage that scale partly answers — per-tonne labour falls from 8.0–12.0 at 500 kg/h to 4.0–5.3 at 3,000, so a bigger line dilutes it better than it dilutes water. In a high labour-cost economy it is frequently the figure that decides the project, and it is not visible anywhere in an equipment quotation. A PET line and a rigid line of the same capacity can be within a few percent of each other on machine price and a long way apart on annual operating cost. It is worth running both numbers before choosing a stream — what adds up in plastic recycling plant cost covers the rest of the layers, and HDPE drum recycling describes the rigid alternative.
Bottle-to-Bottle and Bottle-to-Fibre Are Different Lines
The two routes share their front end and diverge at the back.
Bottle-to-fibre stops at clean, dry flake. The flake is the saleable product, and the plant’s job ends when it meets the fibre buyer’s specification. This is the shorter and cheaper configuration, and for many operations it is where the line should stop.
Bottle-to-bottle continues into extrusion, melt filtration through a 40–120 mesh screen changer, and a decontamination step, then into pellets. It requires a pelletizing machine for rigid plastics and it raises both the capital cost and the regulatory burden, because the decontamination process is the part a regulator assesses.
Deciding between them is not a machine question, it is a market question — it depends on who is buying and at what price, which is worth settling before any equipment is quoted. If you are still working out which stages your material actually needs, the plastic recycling machine line guide walks through the full sequence.
PET is not the only stream where the output grade decides the line: the same logic applies to plastic regrind, where buyers apply a comparable set of tests before they pay, and the stage-by-stage engineering behind every step above is in the plastic recycling process from bale to pellet.
Frequently Asked Questions
Which PET flake grade is the easiest to sell?
Strapping grade, because it accepts mixed colour and a wider contamination tolerance than the other two. Fibre grade is the largest market by volume and usually the better commercial target once washing is consistent. Food grade is the hardest, since it depends on a regulator assessing the recycling process rather than on the flake alone.
How much PVC contamination is too much in PET flake?
Buyers commonly specify below 100 ppm. PVC matters far out of proportion to its quantity because it releases hydrogen chloride when heated, which breaks down the PET chain and lowers the viscosity of the whole batch. It also cannot be removed after crushing, since at 1.30–1.45 g/cm³ it sinks alongside PET at 1.38–1.40, so the limit has to be met by sorting rather than by washing.
How much power does a PET bottle washing line use?
Installed power runs from 150 kW at 500 kg/h to 500 kW at 3,000 kg/h. Per tonne that is 300 kWh at the small end and 166 at the large, a fall of about 44%, though the improvement stalls between 1,500 and 2,000 kg/h. Metered draw is lower than installed power, typically 60–70% of nameplate, since motors carry sizing margin and are not all loaded at once.
How many people does it take to run a PET bottle line?
Between four and sixteen depending on capacity, which is 8.0–12.0 people per tonne per hour at 500 kg/h falling to 4.0–5.3 at 3,000. At any capacity that is roughly twice what a rigid line of the same throughput needs, and nearly all of the difference is manual picking on the sorting belt, where PVC and coloured bottles have to be judged rather than measured.
Can coloured PET bottles be recycled together with clear ones?
Physically yes, commercially it depends on where the flake is going. Mixed-colour flake is normal for strapping and acceptable for much of the fibre market. Anything aimed at clear packaging needs the coloured bottles taken out on the sorting belt, because colour cannot be washed out and it sets a ceiling on what the batch can be sold as.
Why is PET flake priced on intrinsic viscosity?
Intrinsic viscosity indicates how long the polymer chains still are, which decides what the material can be made into. Heat, moisture and contamination all shorten those chains. A batch that has been dried properly and kept free of PVC holds its viscosity, which is why two visually identical batches can be worth noticeably different amounts.
How is a batch of PET flake tested before a buyer accepts it?
Typically by sampling for moisture, intrinsic viscosity, and contamination counts for PVC, PE and PP measured in parts per million, plus a visual check on colour and flake size consistency. Most disputes come from moisture and from fines rather than from anything visible in the sample. The same checks are what we run on a machine before it ships, described on our quality control page.
Do I still need a metal detector if the line already has a sink-float tank?
Yes. The two stages catch different things. Sink-float separates by density and removes floating PE and PP, but it cannot separate PVC or metal fragments that sink with the PET. Metal detection sits before the crusher for a second reason as well: metal entering the crusher chamber damages the knives.
If you are working out which configuration suits the bottles you have, send us the material details — photographs of the bales and the output grade you intend to sell tell us more than a model number does.
