Turning mixed plastic wastes into a useable liquid fuel

S.L. Low, M.A. Connor and G.H. Covey

Click here to download

As landfill and incineration become more expensive and less accepted, the recycling of plastic wastes is gaining increasing importance. More emphasis is thus being given to new disposal options, which have high energy recovery values and are more environmentally attractive. Pyrolysis is one promising method for the treatment of mixed and contaminated plastic wastes. Plastics are thermally degraded to produce useful liquid hydrocarbons, which can then either be added to existing fuel or solvent product, or returned to a refinery where they can be added to the feedstocks. A simple pyrolysis reactor system is described. Results of pyrolysis tests showed that pure samples of polyolefinic and polystyrene resin can readily be pyrolysed to produce liquid yields in excess of 70%. However, liquid yields were affected by heating rates and heat loss patterns in the reactor system. Further experimental work suggests that when pyrolysed, mixed plastic wastes behave much like the resins from which they originate. In light of the results from the experiments, the technical feasibility of setting up a pyrolysis plant in Victoria to process waste plastics into liquid fuel was discussed. This study thus forms the ground work needed for the design of a small pyrolysis plant.


Recycling of plastics already occurs on a wide scale. Extensive recycling and reprocessing of plastics are performed on homogeneous and contaminant free plastic wastes. However, a substantial fraction of the plastics in municipal waste still ends up in landfills. Minimising the amount of otherwise unrecyclable waste plastics going to landfill is thus the motivation of this research project.

The main hindrance to the implementation of plastics recycling is the inhomogeneity of many plastic wastes. Most recycling schemes require a feedstock that is reasonably pure and contains only items made from a single polymer type, such as high density polyethylene (HDPE) commonly used to make milk bottles or polyethylene terephthalate (PET) soft drink bottles. Realistically, most post consumer wastes contain a mixture of plastic types, and are often contaminated with non-plastic items.

Early research on ways to reuse mixed and contaminated wastes concentrated on mechanical recycling schemes. Such schemes involved melting and then moulding or extruding mixtures of plastic components [1]. Typical products were simple and often bulky items such as walltiles, flowerpots, fenceposts or planks. The wastes processed in this way did not have to be segregated according to plastic type and the presence of limited amounts of small, non-plastic particles was acceptable. A waste processing plant of this kind was established in Melbourne in the late 1980’s. Its feedstock was mixed post-consumer plastic wastes and these wastes were converted by melting and extrusion into a material known as Syntal [2], which was marketed as a substitute for more usual construction materials such as concrete and timber.

The disadvantage of such an approach to plastics recycling is that the product no longer has the special characteristics of the plastics used to make it. Thus the product is less useful and ends up competing for markets with cheap construction materials. The presence of non-plastic contaminants leads to concerns from potential buyers about product quality and consistency. For these reasons, mechanical recycling of mixed plastic wastes appears to have only a limited future.

Another widely researched approach to dealing with mixed plastic wastes is incineration. Incineration can be classed as a form of recycling, if carried out with energy recovery. However, incineration is a costly process. Also, if polyvinyl chloride (PVC) is incinerated, HCl is formed and can cause significant corrosion in equipment. An alternative thermal approach to dealing with waste plastics is so-called chemical feedstock or chemical recycling. This term has been used to describe a diversity of techniques, including pyrolysis, hydrolysis, hydrogenation, methanolysis and gasification. Some of these techniques are suitable for use only with homogeneous polymer wastes but others can accept a feed of mixed wastes.

The most attractive technique of chemical feedstock recycling is pyrolysis. In its simplest form, this involves heating and decomposing mixed plastics in the absence of oxygen. Unlike mechanical recycling techniques, in which the long polymeric chains of the plastic are preserved intact, pyrolysis produces lower molecular weight fragments. By adjusting operating conditions, the rate and extent of decomposition can be controlled. In this way, it is possible to obtain a predominantly liquid hydrocarbon product with potential for use as a fuel or a refinery feedstock [3]. This approach can be applied to 80% of commodity plastics, so mixed plastic wastes can usually be handled without the need for segregation by polymer type.

Although pyrolysis and related chemical recycling processes have obvious potential, there are still a number of problems. The main one concerns economic viability since the cost of collection and separation of the plastics component in the post-consumer waste stream can be considerable. At a chemical recycling plant in Grangemouth, Scotland, it was found that two-thirds of the operating cost was associated with collection and sorting [4].

Another difficulty relates to plant size and economies of scale. The maximum practical size for a plastics processing plant is determined by the volume of suitable wastes that can be economically collected and transported to the plant. Since the bulk density of plastic wastes is low and the transport costs high, even wastes generated comparatively close to the plant may be uneconomic to recover. Frequently, the quantities that it is cost effective to recover, are well below those needed by an optimally sized plant. In 1995, BASF shelved plans to build a chemical recycling plant in Germany, on the grounds that the German recycling authorities could not provide enough waste plastics to make the project viable [5].

Pyrolysis processes are frequently classified according to their operating temperature [6]. When high yields of liquid hydrocarbons are desired, low temperature (below 550 oC) processes are employed; these yield mainly oils or tars with smaller amounts of gaseous and solid (char) products. As the operating temperature is increased, the gas yield increases and the liquid yield decreases.

Pyrolysis processes can also be classified according to the type of reactor used. Fluidised bed reactors are widely accepted and utilised because they have excellent heat and mass transfer characteristics and maintain a highly uniform temperature across the fluid bed [7]. Under such conditions it is much easier to achieve a desired product mix. A problem with fluidised beds is that they are only viable when run as large-scale units, which are totally unsuited for use in Australia with its widely dispersed population and sprawling urban areas. Waste collection and transport costs would be at least as high as those overseas, and the quantities of waste plastics economically recoverable would almost certainly be smaller.

For Australian conditions, it would be more appropriate to use smaller plants employing a more basic pyrolysis technology. A network of such plants, distributed across the cities and larger regional towns, should enable recovery of a sizeable fraction of the post-consumer waste plastic stream. Ideally, such plants should be able to achieve a good yield of the liquid hydrocarbon product at an acceptable cost. This would need to be suitable for use locally as a fuel or to be sold to the nearest refinery. In the latter case, backloading petrol tankers returning to the refinery could provide a cheap means of transporting the product over quite long distances.

This study is a follow on from the research undertaken by Low et al. (1998) [8] which investigated the feasibility of designing a small-scale waste plastics pyrolysis plant. The main problem encountered with the previous experimental setup employing a batch spherical reactor heated with a heating mantle, was the presence of significant charring due to localised heating at the reactor wall. In this study, a more sophisticated experimental design has been developed to better monitor and to allow for greater control of the heating process which influences the products obtained. This paper will report on the technical feasibility of turning mixed plastic waste into a useable liquid product.