**1. Introduction**

By definition a composite material is a material that is build up from two or more different components. For example, concrete is a composite material because it is build up from cement, sand and aggregates. However, with the name 'composites' often a category of materials is indicated that can be described as fibre reinforced plastics. In the following, 'composites' will refer to these fibre reinforced plastics. Fibres provide the reinforcement of the plastic, and the result is a strong but light material. These properties, when combined with the freedom to mould them into curved shapes has resulted in composites being used in many applications, notably: wind energy, aerospace structures, transportation, building and infrastructure, sport, yachting and various marine applications.

Industrial production of composite products started during the 1939–1945 World War and involved the use of epoxy or polyester resins reinforced with glass fibres. In the 1980's carbon and aramid fibres became available allowing the manufacture of stiffer, stronger and lighter structures. After the millennium, other reinforcement fibres became commercially available: high-oriented PE, basalt and natural fibres (e.g. flax and hemp). Despite these material innovations, glass fibre remains the volume material due to its excellent price to performance ratio.

Polyester and epoxy resins are thermosets. This means that they are initially liquid and solidify after a chemical cross-linking reaction. This liquid state enables the impregnation of the fibres and in this wet form they can be easily positioned into a mould without using pressure or elevated temperatures. The thermoset resin then solidifies and results in a strong fibre reinforced plastic product that can be subsequently removed from the mould. This combination of materials shows very good resistance to water or other corrosive environments. This results in composites having a long service-life with little or no maintenance. Even in outdoor applications a service-life of composite products is between 60 to 100 years [1].

Since the 1980's thermoplastic polymers have also been used for manufacturing composite products. Initially they were used only in short-glass fibre reinforcement for injection moulded parts (e.g. casings for tools). Typically the thermoplastic polymers PP, PA6, PA66, PBT and PET are used for these products. Thermoplastic composites require both a high temperature and pressure to achieve fibre impregnation. When producing composite parts using the injection moulding process, the high viscosity of the thermoplastic melt means that the impregnation of the fibres can only be achieved by a compounding step in an extrusion process. This results in a reinforcement with a short fibre and with such high pressures required, the size of the products is limited because of the need to use heavy steel moulds and high closing forces.

The commercial use of long-fibre thermoplastic composites has increased in the last twenty years. The production is a two-step process because of the difficulties in impregnation with the viscous thermoplastics, similar to the problem with the short-fibre thermoplastics. The long fibres are first impregnated with a melted thermoplastic polymer into plates or tapes, followed by a second step in which the pre-compounded materials are reheated and shaped into the desired form. This is achieved by hot-press moulding or laser-assisted tape laying. When cooled the product becomes solid. Long-fibre reinforced thermoplastics require PP or highperformance thermoplastics such as PEI, PPS and PEEK. Long-fibre thermoplastic composites are relatively small in volume compared to the total composite market but they are growing [2].

## **2. End-of-life thermoset composites**

Despite their longevity, thermoset composites do eventually come to an Endof-Life (EoL) stage. This can be because of esthetical reasons, damage, or the end of their *guaranteed* structural safety. Generally the composite material itself is still viable. Particularly in the case of rotor blades from modern windmills that are guaranteed for safe use for a period of 20–25 years. When rotor blades are decommissioned (**Figure 1**), the composite material still has very good properties.

The major volume of EoL thermoset composites consists of boat hulls and windmill rotor blades, and these waste streams are expected to increase in the coming years [4]. Boats are a fashion leisure product and as such are periodically replaced and, as mentioned, the dismantling of windmills occurs when the period of guaranteed of structural safety ends or, as rotor blade size increases, for reasons of efficiency.

### **3. The composite recycling challenge**

Thermoset composites are very hard to recycle into the original components (fibres, resin, fillers and core materials). To separate the components, the crosslinked resin must be decomposed because it cannot be melted. Decomposition can be achieved by burning or by dissolving in a chemical substance that can

**Figure 1.** *Obsolete windmill rotor blades [3].*

depolymerize the thermoset polymer. These methods have been extensively investigated since the 1990's, but to date there is no industrial method that is financially viable.

The burning method to regain the fibres is only a partial (caloric) recycling of the material because the resin and organic core materials are not being recycled. In addition, the fibres that are regained from this process are of a very limited value. The glass fibres experience a dramatic loss of fibre strength as reported by Thomason et al. [5]. The coupling agent on the glass fibres (binder) is also destroyed by the burning process. With regard to carbon fibres the situation is slightly more positive because the burning process does not affect their strength. In both cases, however, the after-burning result is not a suitable material for general industrial use.

Two development programmes were set up in the 1990's using the burning process to regain glass fibres from composites in automotive applications, mainly from Sheet Moulding Compound (SMC) and Bulk Moulding Compound (BMC) parts. One programme from the automotive industry in Germany was developed by ERCOM Composite Recycling GmbH that started in 1990 [6]. This development did not lead to a successful industrialisation and stopped in 2004. The other development was the VALCOR-process from the automotive industry PSA in France. This development also did not lead to success and was duly stopped.

Grinding the composite products into a filler was also investigated. Although the resulting filler can be re-used in new products, this method did not lead to a positive business case. This was because virgin fillers that are available on an industrial scale (e.g. limestone, talc, sand) have a very low price level in the range of € 0,10 to € 0,20 per kg. The processing costs to grind the composite products lead to a much higher price level, so competition with traditional fillers is not possible. Over the past years several companies have developed composite grinding methods but they have not been a lasting success.

The burning method was further developed into the 'cement-kiln route' [7]. In this method, EoL composite is fed into a cement oven and the organic components are burned off providing the caloric value to heat the oven. The inorganic components,

especially the glass fibre, remain as a filler in the cement. Although not technically a recycling of the composite, some useful components are retained in the form of energy and the fibre remainders that can serve as a filler. The cement kiln route has been accepted in Europe from 2012 to present day as a recognised method for the recycling of composites [8]. This method, however, is expensive: the EoL composite has to be processed into small pieces and to be brought to the cement oven in Bremen, Germany and then there, a gate fee of € 160, −per metric ton has to be paid.

To improve composite recycling, several initiatives in the last 20 years were undertaken to recycle composites into their original components but none have resulted in an industrial process yet. Comprehensive overviews have been given of these initiatives and have been described [9, 10] and presented by the ACMA [11].
