How does pyrolysis work
Pyrolysis converts plastic, tyres and oil sludge into fuel oil, carbon black and a combustible gas. This guide explains what happens inside the reactor, what limits the output, and what to test before you buy a plant.
Pyrolysis in one minute
Pyrolysis is thermal decomposition without oxygen. That single condition separates it from burning, and it is why a pyrolysis plant produces oil and char rather than ash and flue gas. Everything else in the design — reactor type, heating method, condenser train, discharge system — exists to hold that condition steady while the material is heated.
When the material is heated past the point where its chemical bonds start to break, the long chains split into shorter ones. Those shorter fragments leave the reactor as vapour, and the vapour is condensed into oil. What does not vaporise stays behind as solid char, and the part that will not condense at all is burned to keep the reactor hot.
Pyrolysis, gasification and incineration are not the same process
Buyers routinely compare quotations from suppliers who mean different things by the same word. The three processes differ mainly in how much oxygen is present and what leaves the plant.
| Process | Oxygen | Main output | What it is for |
|---|---|---|---|
| Pyrolysis | None, or a controlled deficit | Oil and wax, char, combustible gas | Recovering fuel and carbon |
| Gasification | Limited, deliberately added | Syngas (CO and H₂), ash | Making heat or power |
| Incineration | Excess | Heat, flue gas, ash | Destroying volume |
| Thermal desorption | None, lower temperature | Recovered oil, treated solids | Sludge and soil clean-up |
The commercial consequence is direct. A unit sold as pyrolysis but fed with air is gasifying. You will get gas instead of oil, and the revenue case you were shown no longer applies. Ask what the burner does and what the reactor atmosphere is, before you compare prices.
The four stages inside the reactor
The stages are always the same, and their order matters because each one sets up the next.
1. Drying
Water leaves first and takes heat with it. A feed at 30 % moisture spends a large part of the heat budget evaporating water before any cracking begins. This is why moisture is the first number to measure on any feedstock, and why drying before the reactor is usually cheaper than drying inside it.
2. Melting and devolatilisation
Thermoplastics soften, and the volatile fraction starts to leave the solid. In a batch retort you can watch this happen: reactor pressure climbs before the first oil appears at the condenser.
3. Cracking
Long chains break into shorter ones, and this is the stage that decides what your oil is worth. Temperature and residence time set the balance between a heavy waxy fraction and a lighter oil. Push the temperature up and more of the product ends up as gas. Hold it lower and you keep more liquid, but wax deposition starts to block lines and the condenser.
4. Char and gas
The solid residue keeps most of the fixed carbon plus whatever inert material came in with the feed. The non-condensable gas — hydrogen, methane and the C2–C4 fraction — is normally piped back to the burner so the plant heats itself instead of consuming purchased fuel.
In a continuous kiln these stages overlap along the length of the reactor. In a batch retort they happen one after another, which makes batch plants easier to observe and harder to run efficiently.
Why oxygen exclusion decides the product
Carbon that meets oxygen leaves as CO₂. That carbon was the char you intended to sell, and the heat released is heat the process did not need. So the reactor is kept sealed and held slightly above atmospheric pressure, and the feed is purged of air before heat-up.
Where air actually gets in
- the feed inlet, particularly on batch plants that open every cycle
- the discharge, when hot char is dumped into open air
- flange and seal faces on a rotating kiln
- the condensation train, when it is run under vacuum
What a leak looks like before it becomes a failure
Oil yield drifts down over a few weeks. Flue gas CO₂ rises. Reactor pressure stops repeating from batch to batch, and the oil darkens. In practice an operator smells it before the trend is visible on a chart, which is one argument for keeping a person on the floor rather than running the line entirely from a screen.
If the oil yield falls and nothing else changed, look for air before you look for the catalyst, the temperature or the feed. Seal and purge faults account for more lost yield than any other single cause.
From vapour to saleable oil
The condenser train is where a plant either makes sellable oil or quietly loses it. Three decisions matter more than the rest.
Total or fractional condensation
Total condensation gives you one mixed oil stream: simpler to operate and simpler to sell as boiler fuel. Fractional condensation splits the heavy and light cuts, which is worth the extra complexity only if you have a buyer for each cut.
Sizing for the heaviest fraction
Heavy vapour condenses easily. The light end does not, and it is the part that escapes as haze at the vent when a condenser is undersized. If you see a permanent mist at the outlet, oil is leaving the plant as aerosol and the material balance will never close.
Keeping water out of the oil
Wet feed carries water into the condenser, and water plus oil gives an emulsion that separates slowly and sells badly. Dehydrating the feed is cheaper than resolving an emulsion downstream.
Reactor types, and what they actually change
| Mode | Feed consistency needed | Labour | Air ingress risk | Fits |
|---|---|---|---|---|
| Batch | Low — mixed and variable feeds are tolerated | High, one cycle at a time | High, the reactor opens every cycle | Small volume, mixed or changing feedstock |
| Semi-continuous | Medium | Medium | Medium | Steady feed with batch discharge |
| Continuous | High — size and moisture must stay in range | Low once stable | Low, sealed feed and discharge | Single feedstock at volume |
Heating method matters as much as mode. Direct heating puts hot gas through the reactor: fast, but it dilutes the off-gas with combustion products and makes the gas harder to reuse. Indirect heating through a jacket or kiln wall is slower and keeps the gas stream clean, which is why most continuous lines built for gas recycling use it.
What actually sets your yield
Any supplier who quotes a yield before seeing your material is quoting somebody else’s test. Yield moves with a short list of variables, and every one of them is measurable.
- Feedstock chemistry. Polyolefins crack to liquid. PET and PVC do not behave the same way, and mixing them in changes the product.
- Moisture. Water consumes heat and contaminates the oil.
- Inert content. Steel wire in tyres, grit in sludge, fillers in plastic. Inerts leave as char and ash, not as oil, so a high-inert feed lowers liquid yield per tonne fed.
- Particle size and bulk density. These set how fast heat reaches the centre of a particle, and therefore how much of the feed actually cracks.
- Temperature profile and residence time. The two levers that decide liquid versus gas.
- Condenser efficiency and seal tightness. Where a good design quietly loses several points without anything appearing to be broken.
Where pyrolysis is the wrong tool
This is the part most suppliers leave out, and it is the part that decides whether a project works.
Chlorine: PVC and mixed plastic
Chlorine becomes hydrogen chloride. It corrodes the condenser and the gas line, and at the wrong temperature it opens the door to dioxin formation. Either the feed is sorted to keep PVC out, or the gas train needs scrubbing. If a supplier tells you chlorine is not an issue, ask to see the gas-cleaning design.
PET and oxygen-rich plastics
High oxygen content in the polymer means more acids and more gas, and less of the oil you were planning to sell. PET is better removed than processed.
Very wet sludge
When water content is high enough, the drying energy can cost more than the recovered oil is worth. The cut-off depends on local fuel and disposal prices, not on the reactor.
High ash, low volatiles
With a feed that is mostly mineral, you are paying to heat sand. Pyrolysis still reduces volume, but the oil revenue case weakens quickly.
Mixed municipal waste
Sorting cost and output variability usually break the economics before the reactor does.
What to test before you commit
Four tests tell you most of what a design needs to know, and none of them is expensive.
- Proximate analysis — moisture, volatile matter, fixed carbon, ash.
- Ultimate analysis — carbon, hydrogen, nitrogen, sulphur, oxygen.
- Chlorine and sulphur content — the two numbers that decide how much gas cleaning you need.
- Thermogravimetric analysis — shows the temperature window where your material actually decomposes, which is where a realistic set-point comes from.
- A lab or pilot yield test on the same sample, so the mass balance is measured rather than assumed.
Sampling matters more than the method. A hand-picked sample from the top of a stockpile tells you nothing about the five hundred tonnes behind it. Composite samples taken across the pile, dried and quartered, are what make a yield test worth paying for.
How this shapes equipment choice
Once the feed is characterised, the equipment question usually answers itself. A consistent single feedstock at volume points to a continuous line. A mixed or seasonal feed points to batch, where the reactor tolerates variation and the capital cost is lower. High chlorine or high ash pushes the decision towards pre-sorting and a larger gas train rather than a bigger reactor.
Our equipment range is organised along those lines, and the plant FAQ covers capacity, delivery and after-sales questions that come up before a quotation.
Tell us what you are processing
Send a feedstock description, a moisture figure if you have one, and the daily tonnage you are planning for. We will come back with the process route that fits and the data we still need from you.