Starting a plastic recycling business is a sequence, not a purchase. The order that works is: confirm who buys your output, pick the feedstock you can source every week, secure the site and the utilities, then specify the line. Plants that reverse those steps usually discover the problem after the machines arrive.
That ordering is not a matter of taste. Your buyer decides your grade, your grade decides your process, your process decides your power and water demand, and all of that decides the machine list. Work forwards and each decision narrows the next. Work backwards from a machine you liked at an exhibition and you will be looking for a market for whatever it happens to produce.
Stage 1 — Find the Buyer Before You Find the Machine
A recycling plant fails from unsold output far more often than from a missing machine feature. Material piles up, cash stops, and the equipment is fine the whole time.
So the first work is commercial, not technical. Identify two or three real buyers for the specific output you intend to make, and get from each of them the thing that matters most: their written specification. Not “good quality PET flake” but the actual limits — moisture, contamination in parts per million, melt flow index, colour tolerance. Ask what volume they take per month and how consistent it must be: a buyer needing 200 tonnes every month is a different plant from one who buys a container when the price suits them.
Those specifications become your engineering brief. The article on where recycled granules and flake are sold maps the buyer types by downstream industry, and it is worth going through it before any supplier conversation, because the same tonnage sold into fibre, into film and into pipe implies three different lines.
Stage 2 — Choose the Feedstock, Because It Chooses the Plant
The second decision is what you will process, and it is more binding than most first-time operators expect. Four streams dominate, and they need genuinely different plants.
PET bottles
The most standardised stream and the one with the clearest grade ladder. Bales are widely traded, the process is well understood, and a PET bottle-to-flake washing line is close to a known quantity. The catch is that the specification is tight: PVC and metal have to be removed on a sorting belt before crushing, because after that nothing in the line can separate them. Where the output ends up depends on which flake grade you can hold consistently.
HDPE drums and containers
Thick-walled, robust and rewarding, but the residue inside is the whole problem. Drums that held chemicals need to be cut open, emptied and pre-washed before anything else, and they need a twin-shaft shredder at the front because a crusher will not take them whole. A rigid plastics washing line handles the rest. HDPE recycling by input form covers why drums, bottles and crates behave so differently.
LDPE film
The cheapest feedstock to buy and the most expensive to process. Film arrives light, bulky and carrying soil, and it holds water stubbornly, so a soft plastics washing line needs a mechanical squeezer to drive moisture out before thermal drying is affordable. Agricultural film is the hardest version of it, for the reasons set out in LDPE film recycling by contamination type.
PP woven bags
Cheap, plentiful and full of things that are not PP: sewing thread, inner liners, printed labels and whatever the bag was carrying. Density separation struggles here because PP and PE are too close together, so sorting has to happen before washing rather than during it. See PP woven bag recycling for the contaminant breakdown.
Set the four side by side on the figures that decide the budget, and the choice stops being about which material is interesting:
| Stream | Line power and consumption | Front end it forces | The thing that catches people |
|---|---|---|---|
| PET bottles | 150–500 kW for 500–3,000 kg/h; 166–300 kWh/t | Sorting belt and metal detection before the crusher | Most labour-intensive of the four, 5.3–6.7 people per t/h at 1,500 kg/h |
| HDPE drums and rigid | 100–250 kW for 200–2,000 kg/h; 128–400 kWh/t | Twin-shaft shredder; a crusher will not take a whole drum | Whatever the drum contained, which is a permitting question as much as a washing one |
| LDPE film | 110–600 kW for 200–1,500 kg/h; 253–640 kWh/t | Squeezer before drying, compactor before the extruder | Cheapest to buy, dearest to run: two to three times the power and water of rigid |
| PP woven bags | As film, on the same soft line | Ram-fed shredding and sorting before washing | Sink-float cannot separate PP from PE, so liners must come out by hand |
One rule cuts across all four: choose the stream you can source consistently, not the one with the highest headline price. A plant designed around a feedstock you can only get intermittently will spend most of its life idle, and an idle line still costs rent, wages and loan repayments. The second rule is visible in the table — film and woven bags are the cheapest feedstock to buy and by some distance the most expensive to convert, so a low bale price is not the saving it appears to be.
Stage 3 — The Building Comes Before the Machines
Site work is the stage most commonly compressed, and it is the one that cannot be compressed.
The floor. A washing line sits in water. That means a slab that takes both the static load and the vibration, sealed so it does not degrade, and graded with drainage falls to a collection point. Retrofitting drainage into a finished floor means breaking it up.
Ceiling height and layout. Silos, platforms and the discharge end of a dryer need vertical space. A line laid out to fit a building, rather than a building chosen to fit the line, usually ends up with a bottleneck at a conveyor angle nobody predicted.
Truck access. Bales arrive on trucks and product leaves on trucks, so you need a turning circle, a loading area and covered storage for several weeks of input. Wet bales are heavier and dirtier, and outdoor storage in a wet climate quietly raises your drying cost every day.
Room to grow. The most common second-year regret is a building sized exactly to the first line. Space for one more washing stage or a pelletizing section costs nothing at the planning stage and is close to impossible to add later.
Stage 4 — Power and Water Have the Longest Lead Times
Machines have a delivery date you can negotiate. A transformer upgrade has a queue you cannot.
Installed power depends on the stream as much as the tonnage, and the spread is wide enough to change which connection you need. A PET bottle line runs about 150 kW at 500 kg/h and 500 kW at 3,000. A rigid line is the most modest of the three, 100 kW at 200–300 kg/h rising to 250 kW at 1,800–2,000. A soft line is the heaviest by a distance: 110–160 kW at 200–300 kg/h and 380–600 kW at only 1,500. If film is your stream, size the supply against the top of that band rather than the middle.
Actual running draw is roughly 60–70% of nameplate, but your supply has to be sized for the installed figure. The moment that number exceeds what your existing connection can carry you are in a conversation with the utility that takes months rather than weeks. Start it in the same week you start talking to equipment suppliers.
Water is the parallel question. A washing line consumes continuously, and there are two choices: draw and discharge, or recirculate through a treatment loop. Discharge is cheaper to build and depends entirely on whether local regulation permits the effluent. Closed-loop treatment costs more up front and removes both the permitting risk and most of the water bill. Either way the sludge has to go somewhere, and that route needs arranging before commissioning rather than after the first skip fills.
Both of these belong in the budget from the beginning; they are the classic entries in the layers of plastic recycling plant cost that never appear on a machine quotation.
Stage 5 — The First Three Months
Commissioning is not a switch. It is a period during which a line that works in principle is made to work on your specific material, and planning for it honestly is what separates a smooth start from a stressful one.
Feeding and flow. The first week goes on feed rate. Overfeed the shredder and it stalls; underfeed the washer and it wastes water and heat on nothing. Every stage has a rate at which it runs best, and finding it takes real material rather than a manual.
Blade changes come sooner than expected. New operators run knives until output degrades rather than to a schedule, and the material teaches them the interval. Grit is the driver: sandy agricultural film dulls tooling far faster than clean bottle bales. Budget a spare set from day one, because a line waiting for blades to arrive is a line not earning.
Moisture is the most common early failure. A centrifugal dryer at 1500 rpm takes moisture to 0.5–2%, and a hot-air stage finishes it below 1%. If the first buyer sample comes back rejected, moisture and fines are the usual reasons, not anything visible in the bag.
Operator training is a real line item. Somebody has to know why the hot wash runs at 85–95 °C with 1–3% alkali rather than treating the setpoint as a dial for saving energy. Understanding what each stage is for turns operators into people who can diagnose a problem instead of reporting it.
Then send the sample. The last step of commissioning is the first step of the business: a representative sample to the buyer you identified in stage one, with the test results attached. Their feedback, not the supplier’s acceptance test, is what tells you the plant is finished.
What Actually Changes When You Scale Up
The question behind most sizing decisions is whether a bigger line is proportionally better. It is not proportional in either direction, and the two halves of the answer pull against each other.
Going from 500 to 3,000 kg/h on a PET line, installed power rises from 150 to 500 kW — six times the output on three and a third times the connected load. Energy per tonne falls from 300 to 166 kWh, a drop of 44%. Headcount goes from 4–6 to 12–16, which sounds like a lot until you divide it: labour per tonne falls by about 55%, from 8–12 person-hours to 4–5.3.
Two cautions attach to that, and both are the reason a large line is not automatically the right one.
The gains are not evenly spread across the tiers. On a PET line the step from 1,500 to 2,000 kg/h holds energy at 200 kWh per tonne on both sides — you buy throughput and no efficiency. On a rigid line almost the whole energy improvement is finished by about 900 kg/h, and the tier above that is marginally worse per tonne. Water barely improves anywhere past mid-range on any of the three streams.
Every one of those per-tonne figures assumes the line is fed. They are calculated against rated throughput, so a 3,000 kg/h line running on 1,000 kg/h of available bales has the energy bill and the payroll of the large line and the output of a small one. This is the arithmetic that turns a sensible-looking scale-up into the first entry on the list below.
Five Ways New Plants Lose Money
These recur often enough to be predictable, and every one of them is cheaper to avoid than to fix.
1. Starting without a confirmed buyer. The single biggest cause of failure. Production without a route to market converts cash into a warehouse of material.
2. Under-investing in sorting. Sorting is manual, unglamorous and the only place several contaminants can be removed at all. Whatever passes the sorting station is in your product permanently.
3. Undersizing the dryer. The easiest way to make a quotation look cheap and the fastest way to fail a buyer specification. Drying capacity should be matched to your wettest material, not your average.
4. Buying with no spare parts and no local service. Blades, screens, seals and bearings are consumables. A cheaper line with a three-week parts lead time is more expensive than a dearer one with a shelf of spares in the building. Both are checkable in advance rather than after delivery, and the most direct check is to see the plant before you place the order instead of taking a specification on trust.
5. Choosing the machine before the market. Which is stage one again, and worth repeating, because it is the mistake that produces all the others.
It is also worth watching what drives demand for your output. Recovery data such as the US EPA figures on plastics shows how much material a market actually collects, while recycled-content rules like the EU packaging and packaging waste framework are creating obligated buyers who need certified material rather than the cheapest available — the demand side covered in PCR resin and recycled content requirements.
Frequently Asked Questions
Is plastic recycling profitable?
It can be, and the margin comes from four variables: what you pay for feedstock, what your output grade sells for, how much energy and labour each tonne consumes, and how many hours a year the line runs. Uptime is the one most often underestimated — a correctly sized plant fed irregularly will lose to a smaller one that runs every shift.
What plastic is best to recycle first?
The one you can source consistently and sell reliably, not the one with the highest theoretical price. In practice that is usually PET bottles or rigid HDPE, because both have established bale markets and established buyers. Film is cheaper to buy and considerably harder to process, so it is a better second stream than a first one.
How long does it take to get a plastic recycling plant running?
The machinery is rarely the constraint. Building work, an electrical supply upgrade and effluent permitting have the longest and least controllable lead times, and they should run in parallel with manufacturing rather than after it. Commissioning and training then take weeks rather than days, because the settings have to be found on your material.
Should I start with washed flake or go straight to pellets?
Most new plants should stop at flake. It is a shorter line, a smaller investment and a saleable product, and it lets you prove out feedstock supply and washing consistency before adding an extrusion stage. Adding a pelletizing system later is a normal upgrade path, and by then you will know from your buyers whether the higher price justifies it.
How many people does a recycling line need?
Roughly 4–6 for a 500 kg/h line and 12–16 for a 3,000 kg/h line, covering sorting, feeding, operation and packing. Per tonne that is a large difference — labour falls by about 55% across that range — but sorting is the part that scales least, because it stays partly manual whatever the capacity.
What should I have ready before contacting an equipment supplier?
Your buyer’s written specification, photographs and a sample of the feedstock you can actually source, your target tonnes per hour, floor area, voltage and frequency. With that we can respond with a configuration, installed power and manpower rather than a catalogue — send us the details, and include the dirty material as well as the clean, because the worst input is what the line has to be designed for.
