Final Report Summary - POLYMARK (Novel Identification Technology for High-value Plastics Waste Stream)
Executive Summary:
OBJECTIVE.
The objective of POLYMARK was the development of a food container marking and identification system that allows a reliable separation of food contact approved from non food contact approved post consumer plastic waste. The novelty of this system was to code the plastic itself for easy identification. For ready industry take-up, the system should be easy to integrate, reliable, rapid and low cost. It should also be applicable to other plastic identification challenges.
PROBLEM TO BE SOLVED.
The POLYMARK consortium represent the SME communities of Polymer Recyclers. Current EU legislation states (EC/282/2008) that plastic containers that have been used in a non-food application cannot, subsequently, be recycled and used in an application where it will come into contact with food. While this regulation is intuitively meaningful, it poses a practical problem: At the start of the POLYMARK project there existed no technology to reliably separate food contact approved from non food contact approved post consumer plastic waste. POLYMARK set out to develop a food container marking and identification system for this purpose.
As well as benefiting the recycling community and the food packaging industry, such a technology can have a positive environmental impact. It will help to increase the overall content of re-used recycled material by improving access to higher volumes of food contact approved recyclate with a superior plastics purity.
RESULTS.
The main result of POLYMARK is a method of marking plastics containers combined with an optical identification technology. To demonstrate the practicality of the technology, the POLYMARK project focused on the identification and separation of food contact approved PET (polyethylene terephthalate) as PET occurs in high volume and value in the European market with 60 billion bottles collected and recycled in 2012. In POLYMARK, a prototype industrial scale sorting machine was built and successfully tested at industrial speed on a large number of marked bottles mixed with standard PET bottles. The marking method is transferable to other types of plastic and the resulting equipment is easy adapt to any industrial recycling facility.
WORKPLAN.
The project included an initial stage, where the marking method was identified. After that point, the detection equipment was developed and the marking technology was implemented. With these results, three prototypes of the detection setup were built at IPMS and integrated into an industrial scale sorting machine at Sesotec. First large-scale tests of this machine were performed using PET bottles custom-marked at HERI. Training in the technology is now available from the POLYMARK website.
PARTNERSHIP.
In addition to Petcore Europe (Belgium), who is the project leader, the project has 10 European partners: EUPR, EFBW and EPRO (Belgium), Closed Loop Recycling, Colormatrix and HERI (UK), 4PET (Netherlands), Mikrolin (Hungary),the Fraunhofer IPMS and Sesotec (Germany).
FUNDING.
The POLYMARK project has received funding from the European Union’s Seventh Framework Programme managed by the REA – Research Executive Agency FP7/2007_2013 under Grant Agreement №FP7-SME-2012-311777.
Project Context and Objectives:
PROJECT CONTEXT.
The POLYMARK consortium represents the SME communities of Polymer Recyclers. The EU polymer recycling sector forms a vital part of the polymer sector and has an estimated annual turnover of €2,500 million generated by over 1,000 SMEs. However, our members have a major problem. Within the EU legislation exists (EC/282/2008) that states that plastic containers that have been used in a non-food application cannot, subsequently, be recycled and used in an application where it will come into contact with food. This is not true for other major world markets (including the US) putting Europe at a significant disadvantage and therefore a solution is urgently required.
Additionally, the food packaging industry is under pressure to use more recycled content but has difficulty accessing sufficient volumes of food contact approved recyclate. In our industry, recycling employs a combination of sorting techniques that have been used for particular applications. However, there exists no technology to distinguish containers of food from non-food applications. Moreover, in order to be implemented in practice, such a technology most be efficient, easy, reliable, rapid and low cost.
Therefore an opportunity exists to capture this, to date, untapped market by undertaking the project we propose. Our solution is different. Our idea is to develop a food container marking and identification system that will allow reliable separation of food contact approved from nonfood contact approved post-consumer plastic waste. The novelty is that the plastic itself is coded for easy identification. The technology will also be applicable to other plastic identification challenges. As well as benefiting the recycling community and the food packaging industry, it will also have a positive environmental impact.
MAIN OBJECTIVES.
The overall technological aim of the POLYMARK project was to develop an innovative and industrially practical system to discriminate polymer packaging according to type. To demonstrate the practicality of the technology the POLYMARK project focused on the most commercially significant barrier to improved performance within the plastics recycling industry – namely the identification and separation of food-contact approved and non food-contact approved packaging.
The following technological specific objectives were defined to reach the above stated main objective:
1. Develop a range of marker substances and method(s) of deactivation to encode information regarding the properties and structure of waste plastic and coating the markers onto polymers.
2. Developing an identification system that can detect the markers and decode the information in order to subsequently separate out the plastics by mechanical means. The detection system should meet the requirements of an industrial environment and be designed to be compatible with and easily integrated into existing sorting lines within recycling plants.
3. To validate the full system in the laboratory and on industrial scale.
WORK PLAN.
In order to achieve the aforementioned objectives, the following work packages were carried out:
WP1. New scientific knowledge & initial specifications.
The main objective of this work package was to determine the overall specification of the process / system as well as performance and legislative requirements. The information that a marker will carry was defined in terms of a system of product coding. The most appropriate chemical marker families were determined.
WP2. Development of chemical markers.
The second step of the project was to research appropriate marker chemistries and validate their functionality. Sources of markers were identified, either via synthesis or purchase. A method of removing/deactivating markers was developed. The effects of the markers on the polymer properties were tested for the given range of polymers.
WP3. Development of Spectral Identification technology.
The requirements of the POLYMARK marker detection system were translated into technical specifications to guide the development of the prototype. Then the hardware part of the system was designed and manufactured following these specifications, resulting in the construction of a fully functional, laboratory scale marker detection system. The performance of this system was characterised. The design of the detection system was then adapted for integration into an industrial optical sorter and three detector heads were built.
WP4. Integration of proto-type industrial-scale process.
The main objective of this work package was to integrate the detector and lighting units into an industrial scale optical sorting machine. To allow testing at an industrial scale, appropriate recycling trial systems were selected. The marker system was validated at an intermediate scale prior to full industrialisation.
WP5. Real world recycling proto-type trials.
Marked PET samples were produced to perform industrial-scale trials of the discrimination system. Thereby, the functioning of the detection system was tested under industrial conditions. Further tests were carried out to investigate whether the conditions expected during the normal lifetime of the trial systems will adversely affect the performance of a marker.
WP6. Validation of the industrial-scale recycling system.
The industrial-scale trials were evaluated in order to determine the effectiveness of the marker identification and sorting process. Furthermore, the techno-economic benefits of the recycled materials were assessed, a life cycle analysis of the entire system was performed and issues relating to management / infrastructure were examined.
WP7. Exploitation and dissemination activities.
The objective of this WP was on the one hand, to disseminate the results of the project amongst the consortium and, on the other, to facilitate access to market for future exploitation of the developed system to the European plastics recycling sector. Dissemination activities included personal contacts to several like-minded projects and activities on a European scale and participation or organisation of joint events in order to encourage standardisation.
WP8. Training.
For quick and effective technology transfer and to allow stake holders to understand how the technology functions and its benefits, training on the marker/detection technology was provided.
WP9. Consortium management and administration.
The main objective of this work package was to monitor and coordinate all activities within the project to ensure its optimum development.
Project Results:
WP1. NEW SCIENTIFIC KNOWLEDGE AND INITIAL SPECIFICATIONS.
At the start of the POLYMARK project, a number of ideas existed how to go about distinguishing food approved from non-food approved polymer material during the industrial value chain of plastics production, packaging manufacture and filling, post consumer collection, sorting and recycling. Scientific facts and measurable criteria were required to single out the most promising approach. Three topics were subjected to intensive desk-based studies of the RTD partners and close discussions with the entire consortium:
a) A system of product coding was developed that is carried by the marker and decoded by a suitable detector. The pieces of information worthy of coding included food contact / non-food contact; simple structure / barrier structure; homogeneous polymer / polymer containing additives; homogeneous polymer / blend of polymers. The code is based upon specific combinations of markers to produce a unique fingerprint. Naturally, the implementation of the code was subject to the availability of a suitable marking system.
b) The entire range of chemical markers was thoroughly investigated: UV markers, UV blockers, IR markers and IR blockers. In addition, non-chemical methods of marking such as holography or RFID tracking were considered. More than 400 chemical markers were screened and listed how accurately and repeatably the marker can be detected, its performance-cost ratio, interaction with barrier technologies, impact on the colour-appearance of the container, likelihood of food approval, thermal and light stability, compatibility with and the potential for removing a marker at the end-of-use. Very conscious that a successfully established marking system might use many tons of marker per year, the consortium extensively discussed marker cost, application and removal as these are issues concerning both consumer acceptance, practicality of the process and impact on the environment. In particular, marker cost is a very sensitive parameter in the beverage industry, marker application interested producers of raw polymer and bottle pre-forms, while marker removal and potential accumulation was a concern of the recycling industry. As members of the entire value chain of food packaging participated in our discussions via their link to our SME association partners, we were able to issue recommendations for all of these aspects in our specification document and narrow down the number of suitable markers.
c) Closely linked to each type of marker is the method by which it can be detected. Therefore, a desk-based study of marker detection instrumentation was carried out in close communication to the above discussions. Ultraviolet (UV), visible, near-infrared, infrared and combined optical detection techniques were evaluated for their speed, size and cost of equipment, ease of operation and maintenance, requirements to marker use as well as reliability and accuracy of detection. Again the entire consortium was involved in the discussion of the collected information, finally resulting in the selection of a marking-detection system based on UV illumination and collection of fluorescence light emitted from the marker substance. In terms of minimal marker use, detection cost, speed, and reliability, this method was superior to all others. This choice, in turn, further narrowed down the number of suitable markers to below 20.
The main result of WP1 was the system specification. The requirements of the POLYMARK optical detector system and of the marking procedure were established in consensus of the consortium. In particular, we chose to focus on PET containers out of all polymers because of the high volume and value of this material in the recycling chain. The marker was to be applied as a fluorescent but transparent and colourless coating on the container, which makes it easier to remove after the end-of-use because no heat is involved in this particular deposition process. The thickness of the coating was to be significantly below the wall thickness of a PET bottle to reduce visibility and cost. Fluorescence was to be detected in the visible spectral range from bottle-sized containers with this thin coating at industrial speed (min 2m/s belt speed) under UV illumination. In addition, through our extensive network, we contacted other existing or past consortia of similar projects in order to learn from their experience and avoid repeating mistakes. Thereby, WP1 laid the foundation for all subsequent work by fixing development targets that are meaningful in the industrial context of the food packaging value chain.
WP2. DEVELOPMENT OF CHEMICAL MARKERS.
The main objective of this WP was to develop marker chemistries, validate their functionality and develop a method of removing the marker.
The main results of this WP are summarised below:
A predominantly food contact approved, coating-based approach for the addition of a fluorescent marker to PET bottles has been developed with the following benefits:
a) Use of commercially available, near UV-excitable markers with strong fluorescence in the visible region of the electromagnetic spectrum to allow development of a detection system with minimal UV/ozone generation hazards
-The most suitable markers were found to be the food contact approved fluorescent brighteners 4,4’-bis(benzoxazol-2-yl)stilbene and 2,5-bis(5-tert-butyl-benzoxazol-2-yl)thiophene
-Stable dispersions of the preferred marker candidates above were achieved. The hyperdispersant was the only formulation ingredient lacking food contact approval.
-In particular, the use of 4,4’-bis(benzoxazol-2-yl)stilbene as a dispersed pigment gave a unique fluorescence fingerprint that can be distinguished from the same chemical when used in a molecularly dissolved form (e.g. as an optical brightener in plastics) or from other background fluorescence
-Low levels of marker were required for detection due to their strong fluorescence
b) Water-based, sprayable coating formulation to minimise use of VOCs with associated flammability hazards
- Good water resistance of coating following drying
- Sufficient robustness of mechanical properties to retain the coating during the bottle life cycle
c) The coating approach to marking plastic articles provides flexibility for:
- Addition of further, dispersible marker pigments to the coating formulation in combination to allow development of a coding system in the future
- Extension of the system to alternative polymer types, such as polypropylene (subject to adhesion, which may require surface treatment)
d) Coating removability under alkaline wash conditions already in use within recycling plants
- To allow previous food contact status of plastic to be erased in case it subsequently goes into a non-food contact application
- Temperature and sodium hydroxide concentration minimised to reduce costs without impacting on removal efficiency
WP3. DEVELOPMENT OF THE SPECTRAL IDENTIFICATION TECHNOLOGY.
The objective of this work package was to develop an identification system for distinguishing marked from unmarked polymer samples according to the marking regime defined above. In particular, this required:
(a) translating the system specifications of WP1 into technical specifications by the RTD partners IPMS and Sesotec. Thereby, we ensured that the harsh environmental conditions in an industrial sorting plant are considered as well as the challenging sensor demands for measuring weak fluorescence signals at high throughput rates as well as the cost restrictions towards the marker coating and the sensing equipment. The parameters fixed in the technical specifications included: working distance, detection line width, spectral and spatial resolution, sensitivity of detection, and overall setup geometry. These specifications were then used to design the prototype.
In addition, samples of the remaining ~20 marker substances were tested in the laboratory. In particular, their efficiency of fluorescence emission was evaluated and only two highly efficient marker substances remained that were considered suitable for high-speed detection in a thin coating. After further laboratory tests with realistic samples including labels and glue residues, it was found that one of these markers could be mistaken for a certain common type of label and thereby give rise to false identification. Thus, the detection system was designed for just a single marker substance. As a side effect, the spectral range of detection could be narrowed down to enable a quick, reliable and efficient yes/no identification during sorting. Should further marker substances become available in the future, the detection system is readily extendable without overly increasing its cost.
(b) optical design. In order for the entire system to meet the technical specifications, optical design development was carried out at this stage. Among these are the conception of optical systems, the corresponding simulations and calculations with established software tools such as ZEMAX. Suitable collection optics and sensor components were chosen as a result.
(c) mechanical design. A camera-like system was designed around the selected optical and sensor components considering methods for adjusting and assembling the optical system, stability of mounting and minimisation of stray light entering the detector. Thereby, the resulting laboratory demonstrator was already designed with its future application in mind, e.g. robust enough to withstand temperature changes in an expected range and machine vibrations.
(d) manufacturing or otherwise sourcing components. The parts designed above were custom made and the optical and sensing components were purchased. The readout electronics of the sensor was also custom-made. It had to ensure fast access to the spectral data and the data interface had to be compatible with the prospect of integration into a sorting machine.
(e) building a bench-top detection system to allow for evaluation & optimization. All parts including electronics were assembled in an optics laboratory and a software for reading out sensor data was written and implemented on a lab PC. Standard black/white targets were used to confirm the basic functions of the laboratory system.
(f) testing the spectrometer system to verify its performance. In close collaboration with WP2, the detection system was tested using marked and unmarked PET bottles produced in WP2. Feedback about sample quality was given to WP2. At the same time, the performance of the detection system was evaluated and its capabilities of analysing and identifying marked PET bottles was tested. This bench-top version successfully fulfilled all key features such as optical requirements and the detection speed specified above so that it could enter the next stage of development.
(g) adapting the detection unit for integration with an optical sorter. After performance verification, necessary changes in the design were performed in order to increase robustness and to prepare the detector unit for mounting at the sorting machine. In particular, a robust housing was designed for protecting the sensitive optics and electronics from dust and scratches. Furthermore, the light source previously used in the laboratory tests was equipped with specially designed and custom-made collimation optics to further boost the sorting efficiency during WP5. As a side consideration, the cost of the overall detection system was estimated and found to be in the range typical for a sorting system sold to industrial recyclers. In summary, the detection system was designed to be cost-effective in terms of manufacture and suitable for operation in an industrial environment.
(h) manufacturing three prototype spectrometer systems. Altogether three prototype detection units were built including optics, sensor, electronics and dust-proof housing. All units were satisfactorily tested for compliance with the technical specifications ready to be integrated as detection modules into an optical sorter at Sesotec for WP4, 5 and 6. One unit was later kept as a reference at IPMS.
The main results of this WP are summarised below:
- Specifications of the prototype were fixed.
- The optics of the detector unit were designed.
- The mechanical holder (lens-detector adapter) was designed.
- All components were purchased or manufactured and an initial version of the detector unit was constructed as a laboratory scale demonstrator.
- The laboratory demonstrator was successfully tested with standard targets and marked/unmarked polymer samples.
- Collimation optics for a linear light source were designed, manufactured and assembled.
- A housing was designed to protect the sensor and optics in the industrial environment and to provide means for mounting at the sorting machine.
- Three sets of components and housings were assembled into three detector units. One unit remains at IPMS for reference, two were transferred for industrial-scale integration.
WP4. INTEGRATION OF PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 4 was to select a trial system to allow testing at an industrial scale and to integrate the necessary detector and lighting unit into an industrial scale optical sorting machine. Appropriate recycling trial systems were selected and the marker system was validated at an intermediate scale prior to full industrialisation.
To demonstrate the practicality of the technology, the Polymark project focused on the identification and separation of food contact approved PET (polyethylene terephthalate) from non-food contact material. The project aims to increase the availability of recycled PET from used bottle to new bottle and to contribute to the EU’s circular economy.
In order to perform small scale sorting trails a laboratory benchtop prototype was developed and built. With this small detection and separation unit the basic detection concept was validated and the suitability of the marker coating on the first samples produced was verified.
The first detection and sorting trails were carried out in the laboratory of S+S (Sesotec) and at the laboratory of IPMS. Bottles with different marker concentration levels were applied to find the necessary amount for a safe detection of the marker.
These first sorting trails showed a marker concentration of at least 0.1% is required to achieve a signal to noise ratio which allows a safe differentiation between a marker coated bottle and normal PET bottle.
After successful trials with the small laboratory benchtop prototype, the marker detection and sorting machine on an industrial scale was designed. This machine was based on the adaption of existing state of the art optical sorters, which are currently used in the recycling industry.
By basing the industrial scale Polymark sorting system on existing recycling industry processes, we made sure it is fitting in commonly used recycling lines in terms of processing speed and in terms of the dimensions or footprint of the system.
The complete Polymark sorting machine and the fine tuning of the machine settings was validated with the first marker coated samples provided by HERI. All validation tests were successfully performed at a speed of 3 m/s by feeding a defined test set of 100 bottles several times until the result could be based on a total quantity of at least 1000 bottles.
WP5. INTEGRATION OF PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 5 was to produce marked samples on an industrial scale and to perform industrial-scale trials of the detection and sorting system.
Finally demonstrate that the conditions expected during the normal lifetime of the trial systems will not adversely affect the performance of a marker.
Within this Work Package the RTD partner HERI produced batches of marked PET bottles to test the sorting efficiency of the integrated system (marker, detector, sorting machine) on an industrial scale. Bottles with varying degrees of marking have been prepared by HERI, who produced 250 marker coated 0.5 litre bottles and shipped them in March 2016 to S+S (Sesotec):
- Fully marker coated bottles (150 pcs)
- Partly marker coated bottles (50 pcs, app. 50% area coated)
- Bottles with marker coating on a transparent label (45 pcs)
- Non coated reference bottles (50 pcs)
This range of samples with different coated areas were sufficient to test the resolution and limits of the detection/sorting system. Scale-up of marker dispersion was successfully achieved at HERI and scale-up of coating mixing was achieved at Colormatrix.
For industrial scale trails on the marked samples form were carried out by S+S (Sesotec), and were supported by IPMS, HERI and SMEs.
The overall efficiency of the recycling process was tested. This was achieved by using different trial composition batches of marked bottles with non-marked bottles and original bottles from the general waste stream.
The Polymark sorting machine designed and manufactured in WP4 of the project was built into a demo recycling line which allowed testing of the sorting process in a way which is reflecting the typical setup in the recycling industry.
During the various trial runs the detection and separation system has been be monitored and the machine settings were gradually readjusted and optimized.
The overall outcome shows that this marker concept and detection system is sufficient for separating non-food contact from food contact PET bottles.
The lifetime simulation showed that the marker and coating are robust in the designated environment and withstand the anticipated wearing and abrasion caused by normal usage and during anticipated recycling procedures. The effects of ageing and of UV or environmental exposure did adversely affect the marker coating in respect to visible changes or detectability.
WP6. INTEGRATION OF THE PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 6 was to determine the effectiveness of the sorting process as a whole and evaluate the techno-economic benefits of the process. Another objective was a life cycle analysis of the entire system and an assessment of the management and infrastructure for the adoption of the system.
The main emphasis of this work package was on evaluating the effectiveness of the detection and sorting process, as well as on the transferability of this process.
Therefore the performance of the industrial prototype was quantified by exploiting the sorting process results achieved during the various separation trials. In order to do that, a set of different test scenarios were developed to cover the typical and anticipated use cases in the recycling industry. These test scenarios were then used in several well defined sorting trails on a demo recycling at an industrial scale.
The quantified results from these trials proofed the fitness and effectiveness of this process for the anticipated recycling sorting applications like separating non-food-contact PET from food-contact PET.
The general design concept facilitates an easy transferability of this process to other sorting lines in the recycling industry along with the cross validation trials which confirmed that there should be no interference with existing sorting technology.
The economic feasibility of the sorting process was demonstrated on two implementation cases, for both implementation cases the total cost of ownership was evaluated.
Potential Impact:
In 2012, 25 million tonnes of plastic waste were created in the European Union. The fate of this waste was split between recycling (26%), energy recovery (36%) and landfill (38%). Closer inspection of the collected data shows a wide variation on plastics landfill waste between EU members, with Germany sending less than 10% of plastic waste to landfill (due to a landfill ban in 2005) and the UK sending more than 66% of plastic waste to landfill (PlasticsEurope, 2015). Much of this waste constitutes a valuable resource in terms of material reuse, chemical building blocks or (in the worst case) energy recovery. The positive news is that the proportion of plastic waste going to landfill reduces year on year due to improvements in collection schemes and recycling technology, and investment in energy recovery. However, clearly there is still significant room for improvement.
Poly(ethylene terephthalate) (PET) is a key target for further improvements to recycling technology. PET accounted for 6.9% of the total EU plastics demand in 2013 (PlasticsEurope, 2015), with approximately 30% of that arising from single use, food contact bottles (Petcore Europe, 2014). In addition, its thermoplastic nature and chemical stability lend well to recycling processes. The majority of recycled PET (rPET) comes from bottles (food contact and non-food contact) and other packaging derived from thermoformed sheet/injection moulded products since these are the easiest to mark for identification as recyclable. Recycling of PET has been extensively discussed in the literature and therefore will only be discussed in broad terms here (Al-Salem et al., 2009; Awaja and Pavel, 2005; Hopewell et al., 2009; Luijsterburg and Goossens, 2014; Shen et al., 2010; Welle, 2011). It can be subdivided into mechanical systems (where the articles are cleaned, flaked and re-extruded) and chemical systems (where the articles are returned to their original chemical feedstocks to make new, virgin, PET). Key challenges associated with mechanical recycling are; contamination with other polymers (particularly PVC) that cause discolouration and molecular weight reduction caused by thermal degradation that leads to changes in PET properties. Mechanical systems can be further sub-divided into “open loop” systems (where bottles become other, lower value products such as textile fibres) (Shen et al., 2010) and “closed loop” systems (where bottles are turned into new bottles) (Welle, 2011).
Use of rPET in fibre and sheet applications was around 50/50 with virgin PET in 2013, but use of rPET in food contact bottles, while growing slowly, remains low at around 10% with respect to virgin PET. The reasons for this will be discussed shortly, but it is clear that there is a great opportunity to increase rPET use in higher value, blowmoulding applications.
In addition to the recycling challenges already described, to allow fully closed-loop recycling of PET, the process must fulfil the requirements of EC legislation “on recycled plastic materials and articles intended to come into contact with food” (EC, 2008). This legislation seeks to protect consumers from contamination of food-safe plastics during the recycling process: once a food-safe PET bottle goes to the consumer, it may be used to hold alternative (non-food) materials, or it may become contaminated with non-food products/plastics during the collection/recycling process (Welle, 2013). This may be a particular challenge in countries where recycling collections are of mixed waste, or could arise from counterfeit materials (Puype et al., 2015). An important exemption to the regulation is for recycled plastics used behind a “functional barrier” (subject to migration testing). Thus, rPET may already be used in food packaging provided it is separated from the food product behind a layer of virgin PET. Such products reduce the demand for virgin polymer, but cannot eliminate it and add complexity to the manufacturing process.
The following points are key to the European Food Standards Agency (EFSA) approval of material/articles from a recycling process (EC, 2008):
1. Input material must originate from articles manufactured in accordance with regulations for food contact plastics (EC, 2011, 2006)
2. Either...
- Input arises from a closed and controlled product loop
or
- The recycling process is able to demonstrate reduction of contamination to a concentration that does not pose a risk to human health
3. Existance of an appropriate quality assurance system to ensure reproducibility of the product
Points 2 and 3 can be demonstrated in a straightforward manner using a combination of cleaning processes, vacuum stripping, migration testing, chemical analysis and management system audits (EFSA, 2012a, 2012b; Welle, 2011). However, point 1 is more difficult to achieve where food contact plastics are collected as part of a mixed recycling waste stream. In addition, plastic materials become steadily more complex as greater shelf-life for package contents is sought through multilayer/ additivated barrier materials and through active packaging: such materials may be labelled as, or identify as PET in current sorting processes, but will lead to poor quality rPET products. The challenge is therefore for sorting and separation of these materials firstly, to meet EFSA requirements and secondly, to ensure the future quality of rPET (Dvorak et al., 2013). Improvements in rPET quality and acceptance of a greater proportion for food contact use should enhance demand for this material and subsequently lead to greater penetration rates of rPET into the food contact bottles market. In principle, up to approximately 20% further reduction in demand for virgin PET (with associated carbon emission reductions (Packham, 2014)) is possible if all produced food contact PET bottles are captured into a fully closed-loop recycling system. In practice, material “leakage” through increasing complexity will make this reduction unachievable unless a practical method is found to manage it (World Economic Forum, 2014). A resolution to the material identification and sorting issue is the portion of the overall picture addressed in this project.
Further information can be found on the document « Removable Identification Technology to Differentiate Food Contact PET in Mixed Waste Streams: Interim Report” (http://polymark.org/system/files/generated/files/Polymark%20report_preliminary%20technical%20results.pdf) which is one of the main dissemination activities of the project. This scientific article is available on the website of the project and explained the results of the first part of the project (WP1, 2, and 3).
Another main dissemination activity is linked to the training activities to the members and also the wider public. These reports are also uploaded on the website (http://polymark.org/project-results ) and explained the whole outcome of the project and the benefits for the industry. Also as introduction to this training, it has been produced a video where the Polymark Technology working on the demo line of S+S Sesotec is shown.
List of Websites:
PETCORE EUROPE (Belgium). Contact: info @petcore-europe.org - http://www.petcore-europe.org/ - Project Coordinator: maria.estela.izquierdo@eupc.org
THE FRAUNHOFER INSTITUTE FOR PHOTONIC MICROSYSTEMS (Germany). Contact person: Susanne Hintschich (susanne.hintschich@ipms.fraunhofer.de +49 3518823362). http://www.ipms.fraunhofer.de
SESOTEC S+S (Germany). Contact person: Hans Eder (Hans.Eder@sesotec.com) http://www.sesotec.com
FURTHER INFORMATION.
Project website: http://www.polymark.org
OBJECTIVE.
The objective of POLYMARK was the development of a food container marking and identification system that allows a reliable separation of food contact approved from non food contact approved post consumer plastic waste. The novelty of this system was to code the plastic itself for easy identification. For ready industry take-up, the system should be easy to integrate, reliable, rapid and low cost. It should also be applicable to other plastic identification challenges.
PROBLEM TO BE SOLVED.
The POLYMARK consortium represent the SME communities of Polymer Recyclers. Current EU legislation states (EC/282/2008) that plastic containers that have been used in a non-food application cannot, subsequently, be recycled and used in an application where it will come into contact with food. While this regulation is intuitively meaningful, it poses a practical problem: At the start of the POLYMARK project there existed no technology to reliably separate food contact approved from non food contact approved post consumer plastic waste. POLYMARK set out to develop a food container marking and identification system for this purpose.
As well as benefiting the recycling community and the food packaging industry, such a technology can have a positive environmental impact. It will help to increase the overall content of re-used recycled material by improving access to higher volumes of food contact approved recyclate with a superior plastics purity.
RESULTS.
The main result of POLYMARK is a method of marking plastics containers combined with an optical identification technology. To demonstrate the practicality of the technology, the POLYMARK project focused on the identification and separation of food contact approved PET (polyethylene terephthalate) as PET occurs in high volume and value in the European market with 60 billion bottles collected and recycled in 2012. In POLYMARK, a prototype industrial scale sorting machine was built and successfully tested at industrial speed on a large number of marked bottles mixed with standard PET bottles. The marking method is transferable to other types of plastic and the resulting equipment is easy adapt to any industrial recycling facility.
WORKPLAN.
The project included an initial stage, where the marking method was identified. After that point, the detection equipment was developed and the marking technology was implemented. With these results, three prototypes of the detection setup were built at IPMS and integrated into an industrial scale sorting machine at Sesotec. First large-scale tests of this machine were performed using PET bottles custom-marked at HERI. Training in the technology is now available from the POLYMARK website.
PARTNERSHIP.
In addition to Petcore Europe (Belgium), who is the project leader, the project has 10 European partners: EUPR, EFBW and EPRO (Belgium), Closed Loop Recycling, Colormatrix and HERI (UK), 4PET (Netherlands), Mikrolin (Hungary),the Fraunhofer IPMS and Sesotec (Germany).
FUNDING.
The POLYMARK project has received funding from the European Union’s Seventh Framework Programme managed by the REA – Research Executive Agency FP7/2007_2013 under Grant Agreement №FP7-SME-2012-311777.
Project Context and Objectives:
PROJECT CONTEXT.
The POLYMARK consortium represents the SME communities of Polymer Recyclers. The EU polymer recycling sector forms a vital part of the polymer sector and has an estimated annual turnover of €2,500 million generated by over 1,000 SMEs. However, our members have a major problem. Within the EU legislation exists (EC/282/2008) that states that plastic containers that have been used in a non-food application cannot, subsequently, be recycled and used in an application where it will come into contact with food. This is not true for other major world markets (including the US) putting Europe at a significant disadvantage and therefore a solution is urgently required.
Additionally, the food packaging industry is under pressure to use more recycled content but has difficulty accessing sufficient volumes of food contact approved recyclate. In our industry, recycling employs a combination of sorting techniques that have been used for particular applications. However, there exists no technology to distinguish containers of food from non-food applications. Moreover, in order to be implemented in practice, such a technology most be efficient, easy, reliable, rapid and low cost.
Therefore an opportunity exists to capture this, to date, untapped market by undertaking the project we propose. Our solution is different. Our idea is to develop a food container marking and identification system that will allow reliable separation of food contact approved from nonfood contact approved post-consumer plastic waste. The novelty is that the plastic itself is coded for easy identification. The technology will also be applicable to other plastic identification challenges. As well as benefiting the recycling community and the food packaging industry, it will also have a positive environmental impact.
MAIN OBJECTIVES.
The overall technological aim of the POLYMARK project was to develop an innovative and industrially practical system to discriminate polymer packaging according to type. To demonstrate the practicality of the technology the POLYMARK project focused on the most commercially significant barrier to improved performance within the plastics recycling industry – namely the identification and separation of food-contact approved and non food-contact approved packaging.
The following technological specific objectives were defined to reach the above stated main objective:
1. Develop a range of marker substances and method(s) of deactivation to encode information regarding the properties and structure of waste plastic and coating the markers onto polymers.
2. Developing an identification system that can detect the markers and decode the information in order to subsequently separate out the plastics by mechanical means. The detection system should meet the requirements of an industrial environment and be designed to be compatible with and easily integrated into existing sorting lines within recycling plants.
3. To validate the full system in the laboratory and on industrial scale.
WORK PLAN.
In order to achieve the aforementioned objectives, the following work packages were carried out:
WP1. New scientific knowledge & initial specifications.
The main objective of this work package was to determine the overall specification of the process / system as well as performance and legislative requirements. The information that a marker will carry was defined in terms of a system of product coding. The most appropriate chemical marker families were determined.
WP2. Development of chemical markers.
The second step of the project was to research appropriate marker chemistries and validate their functionality. Sources of markers were identified, either via synthesis or purchase. A method of removing/deactivating markers was developed. The effects of the markers on the polymer properties were tested for the given range of polymers.
WP3. Development of Spectral Identification technology.
The requirements of the POLYMARK marker detection system were translated into technical specifications to guide the development of the prototype. Then the hardware part of the system was designed and manufactured following these specifications, resulting in the construction of a fully functional, laboratory scale marker detection system. The performance of this system was characterised. The design of the detection system was then adapted for integration into an industrial optical sorter and three detector heads were built.
WP4. Integration of proto-type industrial-scale process.
The main objective of this work package was to integrate the detector and lighting units into an industrial scale optical sorting machine. To allow testing at an industrial scale, appropriate recycling trial systems were selected. The marker system was validated at an intermediate scale prior to full industrialisation.
WP5. Real world recycling proto-type trials.
Marked PET samples were produced to perform industrial-scale trials of the discrimination system. Thereby, the functioning of the detection system was tested under industrial conditions. Further tests were carried out to investigate whether the conditions expected during the normal lifetime of the trial systems will adversely affect the performance of a marker.
WP6. Validation of the industrial-scale recycling system.
The industrial-scale trials were evaluated in order to determine the effectiveness of the marker identification and sorting process. Furthermore, the techno-economic benefits of the recycled materials were assessed, a life cycle analysis of the entire system was performed and issues relating to management / infrastructure were examined.
WP7. Exploitation and dissemination activities.
The objective of this WP was on the one hand, to disseminate the results of the project amongst the consortium and, on the other, to facilitate access to market for future exploitation of the developed system to the European plastics recycling sector. Dissemination activities included personal contacts to several like-minded projects and activities on a European scale and participation or organisation of joint events in order to encourage standardisation.
WP8. Training.
For quick and effective technology transfer and to allow stake holders to understand how the technology functions and its benefits, training on the marker/detection technology was provided.
WP9. Consortium management and administration.
The main objective of this work package was to monitor and coordinate all activities within the project to ensure its optimum development.
Project Results:
WP1. NEW SCIENTIFIC KNOWLEDGE AND INITIAL SPECIFICATIONS.
At the start of the POLYMARK project, a number of ideas existed how to go about distinguishing food approved from non-food approved polymer material during the industrial value chain of plastics production, packaging manufacture and filling, post consumer collection, sorting and recycling. Scientific facts and measurable criteria were required to single out the most promising approach. Three topics were subjected to intensive desk-based studies of the RTD partners and close discussions with the entire consortium:
a) A system of product coding was developed that is carried by the marker and decoded by a suitable detector. The pieces of information worthy of coding included food contact / non-food contact; simple structure / barrier structure; homogeneous polymer / polymer containing additives; homogeneous polymer / blend of polymers. The code is based upon specific combinations of markers to produce a unique fingerprint. Naturally, the implementation of the code was subject to the availability of a suitable marking system.
b) The entire range of chemical markers was thoroughly investigated: UV markers, UV blockers, IR markers and IR blockers. In addition, non-chemical methods of marking such as holography or RFID tracking were considered. More than 400 chemical markers were screened and listed how accurately and repeatably the marker can be detected, its performance-cost ratio, interaction with barrier technologies, impact on the colour-appearance of the container, likelihood of food approval, thermal and light stability, compatibility with and the potential for removing a marker at the end-of-use. Very conscious that a successfully established marking system might use many tons of marker per year, the consortium extensively discussed marker cost, application and removal as these are issues concerning both consumer acceptance, practicality of the process and impact on the environment. In particular, marker cost is a very sensitive parameter in the beverage industry, marker application interested producers of raw polymer and bottle pre-forms, while marker removal and potential accumulation was a concern of the recycling industry. As members of the entire value chain of food packaging participated in our discussions via their link to our SME association partners, we were able to issue recommendations for all of these aspects in our specification document and narrow down the number of suitable markers.
c) Closely linked to each type of marker is the method by which it can be detected. Therefore, a desk-based study of marker detection instrumentation was carried out in close communication to the above discussions. Ultraviolet (UV), visible, near-infrared, infrared and combined optical detection techniques were evaluated for their speed, size and cost of equipment, ease of operation and maintenance, requirements to marker use as well as reliability and accuracy of detection. Again the entire consortium was involved in the discussion of the collected information, finally resulting in the selection of a marking-detection system based on UV illumination and collection of fluorescence light emitted from the marker substance. In terms of minimal marker use, detection cost, speed, and reliability, this method was superior to all others. This choice, in turn, further narrowed down the number of suitable markers to below 20.
The main result of WP1 was the system specification. The requirements of the POLYMARK optical detector system and of the marking procedure were established in consensus of the consortium. In particular, we chose to focus on PET containers out of all polymers because of the high volume and value of this material in the recycling chain. The marker was to be applied as a fluorescent but transparent and colourless coating on the container, which makes it easier to remove after the end-of-use because no heat is involved in this particular deposition process. The thickness of the coating was to be significantly below the wall thickness of a PET bottle to reduce visibility and cost. Fluorescence was to be detected in the visible spectral range from bottle-sized containers with this thin coating at industrial speed (min 2m/s belt speed) under UV illumination. In addition, through our extensive network, we contacted other existing or past consortia of similar projects in order to learn from their experience and avoid repeating mistakes. Thereby, WP1 laid the foundation for all subsequent work by fixing development targets that are meaningful in the industrial context of the food packaging value chain.
WP2. DEVELOPMENT OF CHEMICAL MARKERS.
The main objective of this WP was to develop marker chemistries, validate their functionality and develop a method of removing the marker.
The main results of this WP are summarised below:
A predominantly food contact approved, coating-based approach for the addition of a fluorescent marker to PET bottles has been developed with the following benefits:
a) Use of commercially available, near UV-excitable markers with strong fluorescence in the visible region of the electromagnetic spectrum to allow development of a detection system with minimal UV/ozone generation hazards
-The most suitable markers were found to be the food contact approved fluorescent brighteners 4,4’-bis(benzoxazol-2-yl)stilbene and 2,5-bis(5-tert-butyl-benzoxazol-2-yl)thiophene
-Stable dispersions of the preferred marker candidates above were achieved. The hyperdispersant was the only formulation ingredient lacking food contact approval.
-In particular, the use of 4,4’-bis(benzoxazol-2-yl)stilbene as a dispersed pigment gave a unique fluorescence fingerprint that can be distinguished from the same chemical when used in a molecularly dissolved form (e.g. as an optical brightener in plastics) or from other background fluorescence
-Low levels of marker were required for detection due to their strong fluorescence
b) Water-based, sprayable coating formulation to minimise use of VOCs with associated flammability hazards
- Good water resistance of coating following drying
- Sufficient robustness of mechanical properties to retain the coating during the bottle life cycle
c) The coating approach to marking plastic articles provides flexibility for:
- Addition of further, dispersible marker pigments to the coating formulation in combination to allow development of a coding system in the future
- Extension of the system to alternative polymer types, such as polypropylene (subject to adhesion, which may require surface treatment)
d) Coating removability under alkaline wash conditions already in use within recycling plants
- To allow previous food contact status of plastic to be erased in case it subsequently goes into a non-food contact application
- Temperature and sodium hydroxide concentration minimised to reduce costs without impacting on removal efficiency
WP3. DEVELOPMENT OF THE SPECTRAL IDENTIFICATION TECHNOLOGY.
The objective of this work package was to develop an identification system for distinguishing marked from unmarked polymer samples according to the marking regime defined above. In particular, this required:
(a) translating the system specifications of WP1 into technical specifications by the RTD partners IPMS and Sesotec. Thereby, we ensured that the harsh environmental conditions in an industrial sorting plant are considered as well as the challenging sensor demands for measuring weak fluorescence signals at high throughput rates as well as the cost restrictions towards the marker coating and the sensing equipment. The parameters fixed in the technical specifications included: working distance, detection line width, spectral and spatial resolution, sensitivity of detection, and overall setup geometry. These specifications were then used to design the prototype.
In addition, samples of the remaining ~20 marker substances were tested in the laboratory. In particular, their efficiency of fluorescence emission was evaluated and only two highly efficient marker substances remained that were considered suitable for high-speed detection in a thin coating. After further laboratory tests with realistic samples including labels and glue residues, it was found that one of these markers could be mistaken for a certain common type of label and thereby give rise to false identification. Thus, the detection system was designed for just a single marker substance. As a side effect, the spectral range of detection could be narrowed down to enable a quick, reliable and efficient yes/no identification during sorting. Should further marker substances become available in the future, the detection system is readily extendable without overly increasing its cost.
(b) optical design. In order for the entire system to meet the technical specifications, optical design development was carried out at this stage. Among these are the conception of optical systems, the corresponding simulations and calculations with established software tools such as ZEMAX. Suitable collection optics and sensor components were chosen as a result.
(c) mechanical design. A camera-like system was designed around the selected optical and sensor components considering methods for adjusting and assembling the optical system, stability of mounting and minimisation of stray light entering the detector. Thereby, the resulting laboratory demonstrator was already designed with its future application in mind, e.g. robust enough to withstand temperature changes in an expected range and machine vibrations.
(d) manufacturing or otherwise sourcing components. The parts designed above were custom made and the optical and sensing components were purchased. The readout electronics of the sensor was also custom-made. It had to ensure fast access to the spectral data and the data interface had to be compatible with the prospect of integration into a sorting machine.
(e) building a bench-top detection system to allow for evaluation & optimization. All parts including electronics were assembled in an optics laboratory and a software for reading out sensor data was written and implemented on a lab PC. Standard black/white targets were used to confirm the basic functions of the laboratory system.
(f) testing the spectrometer system to verify its performance. In close collaboration with WP2, the detection system was tested using marked and unmarked PET bottles produced in WP2. Feedback about sample quality was given to WP2. At the same time, the performance of the detection system was evaluated and its capabilities of analysing and identifying marked PET bottles was tested. This bench-top version successfully fulfilled all key features such as optical requirements and the detection speed specified above so that it could enter the next stage of development.
(g) adapting the detection unit for integration with an optical sorter. After performance verification, necessary changes in the design were performed in order to increase robustness and to prepare the detector unit for mounting at the sorting machine. In particular, a robust housing was designed for protecting the sensitive optics and electronics from dust and scratches. Furthermore, the light source previously used in the laboratory tests was equipped with specially designed and custom-made collimation optics to further boost the sorting efficiency during WP5. As a side consideration, the cost of the overall detection system was estimated and found to be in the range typical for a sorting system sold to industrial recyclers. In summary, the detection system was designed to be cost-effective in terms of manufacture and suitable for operation in an industrial environment.
(h) manufacturing three prototype spectrometer systems. Altogether three prototype detection units were built including optics, sensor, electronics and dust-proof housing. All units were satisfactorily tested for compliance with the technical specifications ready to be integrated as detection modules into an optical sorter at Sesotec for WP4, 5 and 6. One unit was later kept as a reference at IPMS.
The main results of this WP are summarised below:
- Specifications of the prototype were fixed.
- The optics of the detector unit were designed.
- The mechanical holder (lens-detector adapter) was designed.
- All components were purchased or manufactured and an initial version of the detector unit was constructed as a laboratory scale demonstrator.
- The laboratory demonstrator was successfully tested with standard targets and marked/unmarked polymer samples.
- Collimation optics for a linear light source were designed, manufactured and assembled.
- A housing was designed to protect the sensor and optics in the industrial environment and to provide means for mounting at the sorting machine.
- Three sets of components and housings were assembled into three detector units. One unit remains at IPMS for reference, two were transferred for industrial-scale integration.
WP4. INTEGRATION OF PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 4 was to select a trial system to allow testing at an industrial scale and to integrate the necessary detector and lighting unit into an industrial scale optical sorting machine. Appropriate recycling trial systems were selected and the marker system was validated at an intermediate scale prior to full industrialisation.
To demonstrate the practicality of the technology, the Polymark project focused on the identification and separation of food contact approved PET (polyethylene terephthalate) from non-food contact material. The project aims to increase the availability of recycled PET from used bottle to new bottle and to contribute to the EU’s circular economy.
In order to perform small scale sorting trails a laboratory benchtop prototype was developed and built. With this small detection and separation unit the basic detection concept was validated and the suitability of the marker coating on the first samples produced was verified.
The first detection and sorting trails were carried out in the laboratory of S+S (Sesotec) and at the laboratory of IPMS. Bottles with different marker concentration levels were applied to find the necessary amount for a safe detection of the marker.
These first sorting trails showed a marker concentration of at least 0.1% is required to achieve a signal to noise ratio which allows a safe differentiation between a marker coated bottle and normal PET bottle.
After successful trials with the small laboratory benchtop prototype, the marker detection and sorting machine on an industrial scale was designed. This machine was based on the adaption of existing state of the art optical sorters, which are currently used in the recycling industry.
By basing the industrial scale Polymark sorting system on existing recycling industry processes, we made sure it is fitting in commonly used recycling lines in terms of processing speed and in terms of the dimensions or footprint of the system.
The complete Polymark sorting machine and the fine tuning of the machine settings was validated with the first marker coated samples provided by HERI. All validation tests were successfully performed at a speed of 3 m/s by feeding a defined test set of 100 bottles several times until the result could be based on a total quantity of at least 1000 bottles.
WP5. INTEGRATION OF PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 5 was to produce marked samples on an industrial scale and to perform industrial-scale trials of the detection and sorting system.
Finally demonstrate that the conditions expected during the normal lifetime of the trial systems will not adversely affect the performance of a marker.
Within this Work Package the RTD partner HERI produced batches of marked PET bottles to test the sorting efficiency of the integrated system (marker, detector, sorting machine) on an industrial scale. Bottles with varying degrees of marking have been prepared by HERI, who produced 250 marker coated 0.5 litre bottles and shipped them in March 2016 to S+S (Sesotec):
- Fully marker coated bottles (150 pcs)
- Partly marker coated bottles (50 pcs, app. 50% area coated)
- Bottles with marker coating on a transparent label (45 pcs)
- Non coated reference bottles (50 pcs)
This range of samples with different coated areas were sufficient to test the resolution and limits of the detection/sorting system. Scale-up of marker dispersion was successfully achieved at HERI and scale-up of coating mixing was achieved at Colormatrix.
For industrial scale trails on the marked samples form were carried out by S+S (Sesotec), and were supported by IPMS, HERI and SMEs.
The overall efficiency of the recycling process was tested. This was achieved by using different trial composition batches of marked bottles with non-marked bottles and original bottles from the general waste stream.
The Polymark sorting machine designed and manufactured in WP4 of the project was built into a demo recycling line which allowed testing of the sorting process in a way which is reflecting the typical setup in the recycling industry.
During the various trial runs the detection and separation system has been be monitored and the machine settings were gradually readjusted and optimized.
The overall outcome shows that this marker concept and detection system is sufficient for separating non-food contact from food contact PET bottles.
The lifetime simulation showed that the marker and coating are robust in the designated environment and withstand the anticipated wearing and abrasion caused by normal usage and during anticipated recycling procedures. The effects of ageing and of UV or environmental exposure did adversely affect the marker coating in respect to visible changes or detectability.
WP6. INTEGRATION OF THE PROTOTYPE INDUSTRIAL-SCALE PROCESS.
The objective of Work Package 6 was to determine the effectiveness of the sorting process as a whole and evaluate the techno-economic benefits of the process. Another objective was a life cycle analysis of the entire system and an assessment of the management and infrastructure for the adoption of the system.
The main emphasis of this work package was on evaluating the effectiveness of the detection and sorting process, as well as on the transferability of this process.
Therefore the performance of the industrial prototype was quantified by exploiting the sorting process results achieved during the various separation trials. In order to do that, a set of different test scenarios were developed to cover the typical and anticipated use cases in the recycling industry. These test scenarios were then used in several well defined sorting trails on a demo recycling at an industrial scale.
The quantified results from these trials proofed the fitness and effectiveness of this process for the anticipated recycling sorting applications like separating non-food-contact PET from food-contact PET.
The general design concept facilitates an easy transferability of this process to other sorting lines in the recycling industry along with the cross validation trials which confirmed that there should be no interference with existing sorting technology.
The economic feasibility of the sorting process was demonstrated on two implementation cases, for both implementation cases the total cost of ownership was evaluated.
Potential Impact:
In 2012, 25 million tonnes of plastic waste were created in the European Union. The fate of this waste was split between recycling (26%), energy recovery (36%) and landfill (38%). Closer inspection of the collected data shows a wide variation on plastics landfill waste between EU members, with Germany sending less than 10% of plastic waste to landfill (due to a landfill ban in 2005) and the UK sending more than 66% of plastic waste to landfill (PlasticsEurope, 2015). Much of this waste constitutes a valuable resource in terms of material reuse, chemical building blocks or (in the worst case) energy recovery. The positive news is that the proportion of plastic waste going to landfill reduces year on year due to improvements in collection schemes and recycling technology, and investment in energy recovery. However, clearly there is still significant room for improvement.
Poly(ethylene terephthalate) (PET) is a key target for further improvements to recycling technology. PET accounted for 6.9% of the total EU plastics demand in 2013 (PlasticsEurope, 2015), with approximately 30% of that arising from single use, food contact bottles (Petcore Europe, 2014). In addition, its thermoplastic nature and chemical stability lend well to recycling processes. The majority of recycled PET (rPET) comes from bottles (food contact and non-food contact) and other packaging derived from thermoformed sheet/injection moulded products since these are the easiest to mark for identification as recyclable. Recycling of PET has been extensively discussed in the literature and therefore will only be discussed in broad terms here (Al-Salem et al., 2009; Awaja and Pavel, 2005; Hopewell et al., 2009; Luijsterburg and Goossens, 2014; Shen et al., 2010; Welle, 2011). It can be subdivided into mechanical systems (where the articles are cleaned, flaked and re-extruded) and chemical systems (where the articles are returned to their original chemical feedstocks to make new, virgin, PET). Key challenges associated with mechanical recycling are; contamination with other polymers (particularly PVC) that cause discolouration and molecular weight reduction caused by thermal degradation that leads to changes in PET properties. Mechanical systems can be further sub-divided into “open loop” systems (where bottles become other, lower value products such as textile fibres) (Shen et al., 2010) and “closed loop” systems (where bottles are turned into new bottles) (Welle, 2011).
Use of rPET in fibre and sheet applications was around 50/50 with virgin PET in 2013, but use of rPET in food contact bottles, while growing slowly, remains low at around 10% with respect to virgin PET. The reasons for this will be discussed shortly, but it is clear that there is a great opportunity to increase rPET use in higher value, blowmoulding applications.
In addition to the recycling challenges already described, to allow fully closed-loop recycling of PET, the process must fulfil the requirements of EC legislation “on recycled plastic materials and articles intended to come into contact with food” (EC, 2008). This legislation seeks to protect consumers from contamination of food-safe plastics during the recycling process: once a food-safe PET bottle goes to the consumer, it may be used to hold alternative (non-food) materials, or it may become contaminated with non-food products/plastics during the collection/recycling process (Welle, 2013). This may be a particular challenge in countries where recycling collections are of mixed waste, or could arise from counterfeit materials (Puype et al., 2015). An important exemption to the regulation is for recycled plastics used behind a “functional barrier” (subject to migration testing). Thus, rPET may already be used in food packaging provided it is separated from the food product behind a layer of virgin PET. Such products reduce the demand for virgin polymer, but cannot eliminate it and add complexity to the manufacturing process.
The following points are key to the European Food Standards Agency (EFSA) approval of material/articles from a recycling process (EC, 2008):
1. Input material must originate from articles manufactured in accordance with regulations for food contact plastics (EC, 2011, 2006)
2. Either...
- Input arises from a closed and controlled product loop
or
- The recycling process is able to demonstrate reduction of contamination to a concentration that does not pose a risk to human health
3. Existance of an appropriate quality assurance system to ensure reproducibility of the product
Points 2 and 3 can be demonstrated in a straightforward manner using a combination of cleaning processes, vacuum stripping, migration testing, chemical analysis and management system audits (EFSA, 2012a, 2012b; Welle, 2011). However, point 1 is more difficult to achieve where food contact plastics are collected as part of a mixed recycling waste stream. In addition, plastic materials become steadily more complex as greater shelf-life for package contents is sought through multilayer/ additivated barrier materials and through active packaging: such materials may be labelled as, or identify as PET in current sorting processes, but will lead to poor quality rPET products. The challenge is therefore for sorting and separation of these materials firstly, to meet EFSA requirements and secondly, to ensure the future quality of rPET (Dvorak et al., 2013). Improvements in rPET quality and acceptance of a greater proportion for food contact use should enhance demand for this material and subsequently lead to greater penetration rates of rPET into the food contact bottles market. In principle, up to approximately 20% further reduction in demand for virgin PET (with associated carbon emission reductions (Packham, 2014)) is possible if all produced food contact PET bottles are captured into a fully closed-loop recycling system. In practice, material “leakage” through increasing complexity will make this reduction unachievable unless a practical method is found to manage it (World Economic Forum, 2014). A resolution to the material identification and sorting issue is the portion of the overall picture addressed in this project.
Further information can be found on the document « Removable Identification Technology to Differentiate Food Contact PET in Mixed Waste Streams: Interim Report” (http://polymark.org/system/files/generated/files/Polymark%20report_preliminary%20technical%20results.pdf) which is one of the main dissemination activities of the project. This scientific article is available on the website of the project and explained the results of the first part of the project (WP1, 2, and 3).
Another main dissemination activity is linked to the training activities to the members and also the wider public. These reports are also uploaded on the website (http://polymark.org/project-results ) and explained the whole outcome of the project and the benefits for the industry. Also as introduction to this training, it has been produced a video where the Polymark Technology working on the demo line of S+S Sesotec is shown.
List of Websites:
PETCORE EUROPE (Belgium). Contact: info @petcore-europe.org - http://www.petcore-europe.org/ - Project Coordinator: maria.estela.izquierdo@eupc.org
THE FRAUNHOFER INSTITUTE FOR PHOTONIC MICROSYSTEMS (Germany). Contact person: Susanne Hintschich (susanne.hintschich@ipms.fraunhofer.de +49 3518823362). http://www.ipms.fraunhofer.de
SESOTEC S+S (Germany). Contact person: Hans Eder (Hans.Eder@sesotec.com) http://www.sesotec.com
FURTHER INFORMATION.
Project website: http://www.polymark.org