Polypropylene recovery from post consumer carpets

Transcription

1 Final report Polypropylene recovery from post consumer carpets Report into the recovery of polypropylene from post consumer carpets

2 Executive summary In the UK, up to 500,000 tonnes per annum of post consumer carpets are sent to landfill. Currently, there are very few recycling processes in existence to handle this waste stream. One option is to size reduce synthetic post consumer carpets for use in equestrian surface applications. However, this is a relatively low value application and it should be possible to gain a higher value material from recovered post consumer carpets. Axion Consulting were commissioned by Envirolink Northwest to investigate a recycling option for post consumer carpets, in response to an application for technical support made by Carpet Recycling UK. The aim of the project was to research and develop a recycling process for post consumer carpets made from polypropylene (PP). Working closely with Carpet Recycling UK, the project was to assess whether a recycled PP product could be produced from post consumer carpets. The project consisted of a combination of demonstration trials and laboratory work, all of which took place at Axion Polymers in Salford. The project involved the following stages: Size reduction of the carpets using a granulator fitted with a 15mm screen; Screening of the carpets with a 12mm flip-flop screen; Laboratory scale extrusion of the clean granulated material; and Physical properties testing of the extruded polymer. There were three samples of carpet tested: PP post industrial carpet; and PP post consumer carpet mix sourced from a household waste recycling centre (HWRC) in Salford and from a carpet recycler in Leicestershire. PP post consumer tufted carpet sourced from a household waste recycling centre (HWRC) in Salford. An initial sample of post consumer carpet was tested with a portable near infrared (NIR) machine to identify which carpets were made from PP. Only the PP carpets were granulated and screened. Subsequent preparations for the laboratory extrusion work indicated that the granulated material was not entirely PP and contained contamination. Assessment of the sample showed fibres from the carpets, which were identified as nylon and polyethylene terephthalate (PET). An attempt was made to extrude the material but it was not possible to maintain the extrusion strand. This issue highlighted the problem of carpets containing a range of fibres made from different polymers. A second sample of post consumer carpets was sourced from the same HWRC as the first sample, but this time only carpets of a tufted construction were included. All the carpets were thoroughly tested with a Fourier Transform InfraRed (FTIR) machine. Only carpet pieces, where both the pile and backing fibres had been confidently identified as PP were processed. The granulation trial, in terms of size reduction, was a success. The carpets required manual size reduction to approximately 20x20cm pieces before granulation, to ensure the carpets did not tangle around the granulator shaft and block the machine. Additionally the machine was fed slowly to allow plenty of time for the material to be processed and to prevent blockages from occurring. The granulator size reduced the material and successfully released the pile fibres from the backing fibres. The granulated material was then processed over the flip-flop screen. The results of the trial showed that 38% of the material was oversized and 59% was undersized with a 3% loss. The oversize fraction was the desired product and assessment of the material showed it was significantly cleaner than the feed material. The undersize fraction contained a high proportion of small fibres and dust/dirt contamination. The yield loss in the screening trial was high and this was due to the combination of the granulator and flip-flop screen sizes. For a 12mm flip-flop screen the carpet should ideally only be granulated to approximately 20-25mm. If a smaller granulated fraction is produced, as in this trial, the flip-flop screen should then be smaller, at 6-8mm, in order to reduce yield losses. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 2

3 Samples of both post industrial PP carpet and the clean granulated post consumer carpet were extruded, moulded into plaques and bars and physical properties tested. The post industrial sample was used as the benchmark against which the post consumer fraction was compared. The physical tests showed that both the fractions were good but there were measurable differences between the two samples. Physical properties test Units Post industrial sample Post consumer sample Melt Flow Rate (at 230 C, 2.16kg) 6 52 Tensile Mpa Yield % Break % Density g/cm Ash % The differences between the two samples are not unexpected as the post industrial material was manufactured recently, where as some of the post consumer samples may be 10 to 15 years old and made with different techniques and grades of PP. Not only is the age of the material likely to cause variations, different carpet manufacturers are likely to use different grades of PP causing more variations. Therefore it is anticipated that the physical properties of the extruded fibres from post consumer carpets is likely to change between batches of carpets. A direct comparison with virgin PP is difficult due to the presence of fillers in the carpets samples. However, in general terms the physical properties of the post industrial sample are good compared to what would be expected from a filled PP product. It is thought that the post industrial extruded polymer could be used in injection moulding applications, although further research with moulding companies would be required to verify the suitability of the material. The post consumer polymer could potentially be used in medium to lower grade injection moulding applications, such as plant pots or buckets. The main conclusions from the project are: Post industrial carpet can be processed into a product with good physical properties which has potential for use in a range of applications; Post consumer carpet, consisting of entirely PP fibres, can be recycled into a product with physical properties making it suitable for use in medium grade applications; However, the extruded product from the post consumer carpet is only acceptable if the carpets can be correctly identified and sorted. Correct identification and segregation of post consumer carpets into an entirely PP fraction is critical to the viability of a commercial process to recycle carpets; and Carpets which contain a mix of polymers, including PP, produced a fraction which was not suitable for extrusion and hence the end market for mixed synthetic carpets is still likely to be equestrian surfaces. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 3

4 Table of contents 1 Introduction Types of carpets Existing recycling techniques Existing end markets for post consumer carpets Demonstration trials Sourcing of the carpets for the trials Initial demonstration work Size reduction Granulation Results of granulation trial Discussion of results Shredding Results of shredding trial Discussion of results Alternative size reduction options Screening and cleaning Flip-flop screen Results of flip-flop trial Discussion of results Alternative cleaning options Extrusion Laboratory extrusion Physical properties comparison of extruded products Discussion of demonstration trial results Waste carpets volumes in Northwest England Carpet specification Composition of mixed carpet waste Source of waste carpets Industrial sources Commercial waste sources Domestic waste sources Total carpet waste Proposed recycling flowsheet for waste PP carpets Mass balance for recycling PP carpets Commercial viability of recycling post consumer carpets Conclusions Appendices Appendix 1 - Carpet Recycling UK Tufted PP carpet waste arising in NW, List of figures Figure 1 Schematic of woven carpets... 6 Figure 2 Schematic of tufted carpets... 7 Figure 3 Flow diagram of demonstration trials Figure 4 Samples of carpets which were granulated Figure 5 Sample of carpet showing backing and pile fibres Figure 6 Second backing layer peeled back to show primary backing layer and latex Figure 7 Sample of contaminated carpets Figure 8 Granulator set up Figure 9 Granulated carpet Figure 10 Shredded post consumer carpets Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 4

5 Figure 11 Flip-flop screen at Axion Polymers Figure 12 Flip-flop screen processing the granulated carpets Figure 13 Oversize output from flip-flop Figure 14 Undersize output from flip-flop Figure 15 PP test bars moulded from post industrial carpets Figure 16 PP test bars moulded from post consumer carpets Figure 17 recycling process for waste PP carpets Figure 18 Option 1 mass balance for the recycling of PP carpets Figure 19 Option 2 mass balance for the recycling of PP carpets List of tables Table 1 Granulator mass balance Table 2 Mass balance for flip-flop trial Table 3 Physical properties results from the carpet samples Table 4 HWRC case studies Table 5 Economic payback calculation for option Table 6 Economic payback calculation for option Glossary FTIR NIR PET PP Fourier Transform Infra Red Spectroscopy Near Infra Red Spectroscopy Polyethylene Terephthalate Polypropylene Acknowledgements Axion Consulting would like to thank Carpet Recycling UK for their input and assistance with the project. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 5

6 1 Introduction Carpets have been in use for many years and the quantities entering the UK waste stream are in excess of half a million tonnes per year. Currently in the UK the recycling rate for carpets is less than 2%, with only a few companies involved and recycling processing only in development stages. Carpets are considered to have high carbon footprints as they contain virgin plastic manufactured from petrochemicals. The manufacturing process to produce the virgin raw materials for carpets can be energy and carbon intensive. The aim of this project was to research and develop a recycling process for carpets made from polypropylene (PP). Working closely with Carpet Recycling UK the project was to assess whether a PP product could be produced from the carpets. The two types of carpet considered by this project were: Post industrial carpet which has never left the manufacturing site and is therefore clean; and Post consumer carpets, including off-cuts from installation and uplifted flooring which is dirty. Recycling of the post consumer carpets was the main focus of work for the project. 1.1 Types of carpets There are various types of carpets in the domestic market. The two most common are: Woven; and Tufted. Figure 1 and Figure 2 show diagrams of tufted and woven carpets to illustrate the differences between the two styles. Figure 1 Schematic of woven carpets 1 Woven carpets are manufactured by weaving the carpet pile onto other fibres with a loom. Woven carpets tend to contain a range of coloured yarns to produce intricate patterns. The manufacturing process also tends to be quite slow. The fibres onto which the carpet pile is woven can be polyester, whilst the pile itself tends to be PP. 1 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 6

7 Figure 2 Schematic of tufted carpets 2 Tufted carpets are the most common type for use in domestic applications. They are manufactured by attaching the pile fibres to a backing layer. The pile fibres can be natural materials such as wool or synthetic materials such as PP. A latex compound is then used to bond the fibres to the backing layer. A second backing layer is added to the carpet, sometimes this can be made of hessian but it can also be manmade fibres such as PP. The second backing layer provides stability to the carpet. Various types of materials can be found in carpets including: Plastics and rubber: o PP; o Polyester/PET; o Nylon; Natural fibres: o Wool; o Hessian/Jute; There are a wide variety of carpet manufacturers and the age of post consumer carpet found in the waste stream can vary from a few years to over 15 years. End of life carpets can contain significant quantities of dust/dirt, which need to be removed during the recycling process. 2 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 7

8 1.2 Existing recycling techniques Research undertaken by Axion into carpet recycling indicates that recycling processes do exist abroad and significant work has already been conducted into suitable processes. In particular, both the United States and Europe have been carrying out research into carpet recycling for a number of years. Founded in 2002, the Carpet America Recovery Effort (CARE) is a joint industry-government non-profit organisation. Their mission is to develop market based solutions for recovering value from post consumer carpet. However in America over 80% of the carpets handled by CARE were made from nylon and only 8% were PP. 3 Carpets in America also tend to have deeper piles than UK carpets and the recycling technologies developed in the USA typically involve shearing of the pile from the carpet backing, rather than shredding or granulation. For example, InterfaceFLOR in Georgia, America, has a recycling process which shears off the pile fibres from the carpets. The main focus for the process is the recovery of nylon 6 and 66 for use in new carpet tiles, but a PP fraction is created by the process which is then sold on 4. In Europe, in the late 1990 s, extensive work was done by the European carpet industry to develop a recycling route for post consumer carpets. The project involved the following stages 5 : The RECAM (Recovery of Carpet Materials) Project. The aim of RECAM, which was funded by the European Community, was to develop an economically feasible, closed loop system for post consumer and post industrial carpet waste. The project studied the collection, identification and sorting of carpets as well as the recovery of high grade materials and energy from residual fractions. The project was completed in 1999; Development of the technology. This involved finding or developing the necessary technical equipment and tools for the recycling process. One element of this was COCARE, a coding system to allow for easy identification of carpets; and Pilot plant. Carpet Recycling Europe (CRE) was established in 1998 with the aim to implement the findings from the RECAM project. The project involved the building of the first automated sorting plant for carpets in Mainz, Germany. It was the intention of the project to use the pilot plant to gather data to allow for an economic, technical and environmental assessment of the carpet recycling process. The plant was able to process approximately 3.5 tonnes per hour. Although the plant demonstrated technical feasibility, the economic conditions were unfavourable, due to landfill being the cheaper option at the time. Because of this the plant closed in August The project concluded that although technically feasible, carpet recycling would require more favourable economic conditions for the process to become viable. The work completed by RECAM and CRE provided useful know how into the recycling of carpets. In Germany in 1999 Polyamid , the largest nylon 6 carpet recycling facility, was built. The plant reclaimed post consumer carpet from all over Europe and utilised chemical depolymerisation and re-polymerisation of nylon 6 and extrusion compounding of nylon 66. The plant also recovered PP which was sold for use in fibre reinforced concrete panels. Again due to difficult economic conditions and a lack of suitable feed material the plant shut in June , 8. Currently in the UK, there is no well developed recycling sector for post consumer carpets. The UK's first carpet recycler was Swindon based Greenback Recycling which was set up only a few years ago. However Greenback does not produce an extruded product from the carpets they receive. 3 CARE 2008 Annual Report - accessible as Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 8

9 1.3 Existing end markets for post consumer carpets There are various end markets which currently exist for material from post consumer carpets. The main, relatively low value, end market in the UK is the equestrian or the horticultural market. In the equestrian market the material recovered from the carpets is used as an additive to sand based all-weather surfaces. surface material and laid on the ground. Once the carpet has been used as an equestrian surface the plastics cannot be recovered further. Although not a well developed option in the UK, the materials recovered from carpets have the potential for use in a range of every day products including: Plastic park benches; Road cones; Compost bins; Plastic car parts; Soil erosion protection; Underlay; Insulation; Building products; and Sports surfaces. The use of materials recovered from carpets in the manufacture of new carpets, closed loop recycling, is currently uncommon. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 9

10 2 Demonstration trials The overall project plan was to demonstrate a possible recycling process for post consumer PP carpets. The flow diagram in Figure 3 shows the various demonstration trials completed as part of the project. The three main stages were: Size reduction by granulation; Screening of the granulated material with a flip-flop screen to remove fines; and Laboratory scale extrusion trials and physical properties testing. Figure 3 Flow diagram of demonstration trials Carpet Recycling Trials Post-consumer PP carpets Granulation of carpets at Axion Polymers Screening of granulated carpets with flip-flop screen at Axion Polymers Sample of granulated carpet Sample of screened granulated carpet Extrusion and physical properties testing of samples 2.1 Sourcing of the carpets for the trials The project involved collecting post consumer carpets from the Cobden Street HWRC in Salford and a carpet recycler in Theddingworth, Leicestershire. The initial trial used carpets which were tested with a handheld Near Infrared (NIR) Spectrometer to identify the polymers the carpets were made from. The sample comprised of both woven and tufted carpets. Most of the carpet pieces were large and had clearly been uplifted and rolled up. There were a few small off-cut pieces which were clean and appeared to have not been used. A second quantity of carpets was also sourced from the same facility and was tested with a Fourier Transform Infrared (FTIR) machine to identify the polymers present. The second sample consisted of only tufted style carpets. The photographs in the figures below show the various pieces of carpet. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 10

11 Figure 4 Samples of carpets which were granulated Figure 5 Sample of carpet showing backing and pile fibres Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 11

12 Figure 6 Second backing layer peeled back to show primary backing layer and latex Figure 5 and Figure 6 show tufted carpets but with subtle differences. Tufted carpets have tufts individually inserted into a lattice backing layer by a needling technique. A latex compound is then coated onto the backing to anchor the tufts in place. An additional backing layer is then added to provide stability and strength. In Figure 5 the additional backing layer is also a scrim or lattice i.e. the same as the lattice layer onto which the tufts were initially attached, which gives the backing a mesh like appearance. In Figure 6 the additional backing layer is an artificial felt like fabric made from PP. Peeling the felt layer back shows the latex compound and lattice backing layer onto which the tufts were inserted. The tufts, lattice layer and felt in both of the above samples are all made from PP. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 12

13 2.2 Initial demonstration work Prior to any trials being conducted on post consumer carpet a small amount of initial test work was conducted with post industrial carpet. The post industrial carpet sample would form the benchmark against which the post consumer carpet could be compared. It was anticipated that the results from the post industrial carpet would be as good as could be obtained for recycled carpet, as the carpet has never been laid and therefore was not contaminated with dirt or dust. Testing of the post industrial carpet comprised of: Melting small pieces of the carpet - this was necessary in order to be able to granulate the material using the laboratory scale granulator; Granulation of the carpet with a 3mm screen; Extrusion of the granulated carpet; and Moulding of test plaques and bars to conduct physical properties testing on the extruded polymer. The results from the initial post industrial carpet trial were very promising and hence the project proceeded according to the plan in Figure 3. Prior to granulation the carpets were assessed with a hand held NIR scanner and only the carpets with a PP signal were processed. During the preparations for the extrusion stage of the trial it was discovered that the carpets were not entirely PP and the granulated material contained contamination in the form of PET and nylon, see Figure 7. Figure 7 Sample of contaminated carpets In Figure 7 there are visible fibres which have not melted, which were identified as PET and nylon. Both PET and nylon have higher melting points than PP and hence did not melt in the extrusion preparation stage. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 13

14 The contamination meant that the extrusion process would need to operate in excess of 300 C to ensure all the fibres melted and extruded correctly. In turn this would have a significant detrimental effect on the physical properties of the PP component. Based on this a second sample of post consumer carpets was sourced which were visually identified as of a tufted construction. These pieces were thoroughly tested with an FTIR machine to ensure all fibres within the carpets were PP. The following sections report on the processing of the second sample of post consumer PP carpets. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 14

15 2.3 Size reduction The first stage of the recycling of post consumer carpets is a size reduction process. The two main size reduction techniques are: Granulation; and Shredding Granulation A small scale, hand fed granulator with a 15mm screen at Axion Polymers was used for the granulation trial. Figure 8 shows a photograph of the granulator used for the trial. The aim of the trial was to test the ability of the granulator at size reducing the carpets and to assess the output product from the granulation process. The carpets had to be manually size reduced prior to being fed into the granulator to pieces approximately 20-30mm in size. This was to prevent the carpet pieces from becoming entangled in the granulator shaft and blocking the machine. Figure 8 Granulator set up The pieces of carpets were fed by hand into the top of the granulator. The granulated material comes out at the bottom of the granulator and is sucked through to the collection bag. The material passes through a cyclone, prior to the collection bag, which removes some of the dust from the granulated material. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 15

16 2.3.2 Results of granulation trial Due to the manual way in which the granulator was fed it was not possible to measure the throughput. The material was purposefully fed slowly to ensure the system would not block and hence the time taken to process the material would not reflect the actual throughput rate which could be achieved in a commercial scale process. Table 1 Granulator mass balance kg % kg % Granulation Feed % Granulated material % kg % Losses % Figure 9 Granulated carpet Discussion of results The granulator trial was successful. The granulator did not block and appeared to be able to process the carpets with relative ease. However it was necessary to manually size reduce the carpets prior to granulation to prevent tangling around the granulator shaft. In a commercial process a pre size reduction stage is likely to be required. It is probable that the pre size reduction stage would require some manual input, albeit the size reduction itself could be done with a cutting machine. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 16

17 The granulator successfully size reduced the material. A good degree of separation was achieved with a significant quantity of the pile fibres being released from the backing fibres. The granulation action was also sufficient to remove some of the latex from the pile fibres Shredding As a comparison to the granulation trial a shredding trial was conducted by Carpet Recycling UK at MachTech Service in Rochdale, UK Results of shredding trial Figure 10 Shredded post consumer carpets The carpet material used in the shredder trial was different to that used in the second granulator trial, hence the different coloured fibres seen in the photographs. The shredded carpets were taken from the initial sample of carpets which were not exclusively PP. The carpets had been identified as having both PP pile and PP backing but the sample had not been segregated into woven and tufted carpets. It is thought that the non-pp warp fibres in the woven carpets were the source of the contamination seen in the partially melted material from the granulation trial Discussion of results In contrast to the granulated carpets the shredded material contained more pieces of small complete carpet. The size of the fibres in the shredded material is smaller than in the granulated material. There are some noticeable pieces of latex within the shredded sample, where as in the granulated sample there are not many latex pieces visible. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 17

18 The pile in the granulated sample pulls apart to create a cotton wool like material but this is not the case with the shredded carpet, where the fibres have remained as short twisted piles Alternative size reduction options The main alternative technique for size reducing carpets is the shearing technique used in America. The shearing technique works by shaving the pile fibres from the backing layers of the carpet. This produces two fractions; one consisting of the pile fibres and one of the backing layers. Typically the latex backing is disposed of, whilst the carpet pile fibre is processed through a screen or dry cleaner, then extruded and pelletised. In the past shearing has been considered an unsuitable technique for the recycling of carpets in the UK, as the pile on UK carpets is generally too short. Development work on shearing of short-pile carpet has now started but it is likely that the yield of pile fibres recovered by shearing would be too low for the process to be economically viable. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 18

19 2.4 Screening and cleaning Post consumer carpet can contain significant quantities of dirt and other forms of contamination. In order to have a clean fraction for the extrusion stage the granulated material was processed over a flip-flop screen to remove the fines and dust. The flip-flop screen was chosen as the technique to clean the granulated carpet because of the action it creates. The movement of the flip-flop screen back and forth causes the material to jump up and down as it travels down the screen. This motion results in a high G-force screening action. It was hoped that this would cause the dust, dirt and possibly latex compound to be liberated from the granulated carpet Flip-flop screen Figure 11 shows the flip-flop screen used for the trial, which has a 12x12mm deck. Figure 11 Flip-flop screen at Axion Polymers Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 19

20 2.4.2 Results of flip-flop trial Figure 12 Flip-flop screen processing the granulated carpets Table 2 Mass balance for flip-flop trial kg % Losses % Granulated material kg % Oversize kg % Screening % fraction % Undersize fraction kg % % Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 20

21 Figure 13 Oversize output from flip-flop Figure 14 Undersize output from flip-flop Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 21

22 2.4.3 Discussion of results The flip-flop screen was able to process the granulated material. The results of the trial showed that 38% of the material was oversized and 59% was undersized, with a 3% loss of material. The oversize fraction was the desired product and a visual assessment of the material showed it was significantly cleaner than the feed material. The undersize fraction contained a high proportion of small fibres and dust/dirt particles. The yield loss in the screening trial was high, which was due to the screen sizes used in the two trials (granulation and flip-flop screen). For a 12mm flip-flop screen the carpet should ideally only be granulated to approximately 20-25mm. If a smaller granulated fraction is produced, as in this trial, ideally the flip-flop screen should then be smaller, at 6-8mm, in order to reduce yield losses. During the trial it was observed that the material tended to get caught as it travelled down the screen and rolled like tumble weed. On occasions manual intervention with a broom was required to keep the material moving down the screen. Although the flip-flop screen was able to remove the dust and dirt, it would be advisable to conduct further work with other types of screens and screen sizes in order to select the most suitable screen for processing of carpets Alternative cleaning options A Pla.to dry cleaning unit could be used as an alternative method to remove the dust and fines from the size reduced carpet. The Pla.to dry cleaning unit has a screen with beaters which knocks the material against itself and the screen, with the dust being removed from the carpet through the screen. The Pla.to dry cleaner does not require water to clean the material which is an advantage. However the cost of Pla.to dry cleaner is in the region of 200,000, which is significantly more than a basic flip-flop screen. As the results from the flip-flop screen trial were promising the decision was made that a trial with a Pla.to dry cleaner was unnecessary. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 22

23 2.5 Extrusion The following two samples were extruded using a laboratory scale extruder: Post industrial material; and Post consumer granulated and flip-flopped carpet Laboratory extrusion In order to be able to process the carpets with the laboratory equipment at Axion Polymers the carpets required partial melting. The partial melting was achieved by placing small samples of granulated carpet in a vacuum forming machine for a short period of time. After partial melting the carpets were granulated to 3mm, the particle size required for the extruder feed system. Both samples were extruded using the laboratory scale extruder without any complications. An attempt was also made to extrude the initial sample of carpet which contained PET and nylon contamination. However, the contamination caused the strand from the extruder to blow and the material spluttered from the nozzle. The material had been dried prior to extrusion and so its inability to extrude was not due to moisture content Physical properties comparison of extruded products The samples of post industrial and post consumer carpet underwent a number of physical properties tests: Melt Flow Rate (MFR). The MFR is the number of grams of polymer that can be pushed out of a capillary die of standard dimensions (diameter mm, length 8.0 mm) under the action of standard weight (2.16 kg for PP, at 230 C) in ten minutes (ASTM Standard 1238). The usual melt index range is from less than 1.0 (called fractional) to up to 100 for injection moulding. The higher the MFR, the easier the PP fills the plastic mould and the easier the injection or blow moulding process. As the MFR increases, some of the physical properties, such as impact strength decrease; Tensile strength (Ultimate) (MPa). The ultimate tensile strength is the maximum stress a material can withstand when subjected to tension, compression or shearing. Typical PP tensile strengths range from MPa 9 ; Elongation at point of yield (%). The elongation at point of yield measures the elongation to the point where the maximum stress is applied; Elongation at break (%). The elongation at break measures the elongation at the point of rupture; Density (g/cm 3 ); and Ash (%). The ash test indicates the content of filler (typically chalk) within the polymer. Table 3 Physical properties results from the carpet samples Physical properties test Units Post industrial sample Post consumer sample MFR (at 230 C, 2.16kg) 6 52 Tensile Mpa Yield % Break % Density g/cm Ash % Table 3 shows the results of the physical properties testing of the two samples. The post industrial sample formed the base case against which the post consumer material was compared. The physical property tests showed that both the fractions were good but there were measurable differences between the samples. 9 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 23

24 The post industrial sample has a MFR of 6, whilst the post consumer sample MRF was 52. This means the post consumer sample may be easier to mould, but is likely to have a lower impact strength. The tensile strengths of the two samples are similar. The post industrial sample has a lower elongation at yield than the post consumer sample, but a higher elongation at break. The post industrial sample stretched 7.3% before the maximum force was reached, whilst the post consumer sample stretched 10.1% before maximum stress was achieved. However the post industrial sample was stretched 35% before breaking, whereas the post consumer sample was only stretched 29%. The post industrial sample had 17% filler content, while the post consumer material had 10%. The differences between the samples are not unexpected as the post industrial material has been manufactured recently, where as some of the post consumer samples may be 10 to 15 years old and made with different manufacturing techniques and polymers. Different manufacturers are likely to use different grades of PP with different quantities of fillers. The melt flow index of the PP may change during manufacturing of the carpets due to the sheer forces created in the manufacturing process. Therefore it is highly probable that different batches of post consumer carpet will demonstrate different physical properties to those stated above due to variations in the age, type and manufacturer of the carpets. A direct comparison of either result with virgin PP is difficult due to the presence of fillers in the carpet samples. In general terms the post industrial sample had good physical properties compared to what would be expected from a virgin filled PP. The colour of the extruded PP is influenced by the colour of the carpets forming the feed material. Figure 15 and Figure 16 shows the test bars which were made from the post industrial and post consumer carpets. Both extruded polymers are brown in colour, with the post consumer sample being darker than the post industrial. It is likely that the PP produced from post consumer carpets would require the addition of master batch to make the final product black. Black PP will be more attractive to potential customers and hence should be sold for a higher price. Figure 15 PP test bars moulded from post industrial carpets Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 24

25 Figure 16 PP test bars moulded from post consumer carpets It is thought that the post industrial extruded polymer could be used in injection moulding applications, although further research with moulding companies would be required to verify the suitability of the material. With the post consumer polymer it is expected that there would still be a market for the material, likely to be in medium to lower grade injection moulding applications such as plant pots or buckets. 2.6 Discussion of demonstration trial results Overall the demonstration trials were successful and the project showed that an extruded PP product could be produced from the post consumer carpets. The granulator was able to size reduce the carpets without any problems, but care was taken to ensure the feed rate was slow enough to prevent blockages. The granulator was a suitable size reduction technique for carpets and produced a slightly different style of output to the shredded material. The shredded particles were smaller and less of the fibres were released. The granulator output consisted of a cotton wool like material and more of the fibres had been liberated. The flip-flop screen produced an oversize fraction which was noticeably cleaner, with significantly less dirt/dust contamination than the feed material. However, too much of the granulated material was captured in the undersize fraction. This was due the screen size of the flip-flop screen being slightly too large for the particle size of the granulated material, hence the loss to the undersize fraction. The trial proved that a screening technique, such as a flipflop, could be used to remove the dust and dirt. The extrusion trial was very successful. The results of the physical properties testing showed that both the post industrial and post consumer samples of extruded PP were of a good standard and quality; the post industrial sample had slightly better physical properties than the post consumer sample. Both samples appeared suitable for injection moulding applications, with the post consumer material likely to be used in medium to low grade injection moulding applications such as plant pots or buckets. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 25

26 3 Waste carpets volumes in Northwest England The following section is based on information provided by Carpet Recycling UK. The full report by Carpet Recycling UK can be found in Appendix 1. The aim of the task was to calculate the volume of post consumer carpet waste available in the Northwest of England. 3.1 Carpet specification Based on the results obtained from the demonstration trials in this project, the waste carpet material must meet the following specification: Each piece of carpet to be positively identified as having PP pile fibre; Carpet type to be of a tufted construction; All the backing fibres to be positively identified as PP; The carpet to be dry; No heavy contamination (for example paint, nails); and Acceptable for manual handling (for example no pet urine or odours). 3.2 Composition of mixed carpet waste Research indicates that 45% of mixed carpet waste is identified as having a pile fibre made from PP. Of this 45%, 70% has a synthetic backing. Based on sales the split between tufted and woven carpets is expected to be approximately 85:15 respectively. Typically, woven carpets have backing fibres containing polyester so these carpets can be discounted. This means that 27% of mixed carpet waste meets the PP fibre and tufted construction criteria. 3.3 Source of waste carpets There are three key sources of carpet waste: Industrial; Commercial; and Domestic Industrial sources In the Northwest of England there is one carpet manufacturing site which produces post industrial carpet waste meeting the specification. The quantity generated is approximately 20 tonnes per month Commercial waste sources The following results are based on a case study with Flooring UK, a flooring contractor based in Southport and serving mainly public sector housing contracts. Flooring UK generates approximately ten bales of tufted PP carpets per month, weighing 3.33 tonnes in total, which gives an annual quantity of approximately 40 tonnes. Assuming that it would be feasible to collect similar amounts from eleven flooring retailers or contractors in the region, this would give 450 tonnes per annum. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 26

27 3.3.3 Domestic waste sources The main route for the disposal of waste carpets from domestic sources is through Household Waste Recycling Centres (HWRC). Compositional analysis of the residual waste collected at HWRCs indicates that the carpet content can range from 5%- 15%. A conservative estimate of 8% has been used in the following calculations. In order for carpet to be collected at HWRCs there must be space on site for a suitable storage container to keep the carpets dry. Table 4 shows two case studies for HWRCs in the Northwest of England. Table 4 HWRC case studies Area Merseyside municipal waste HWRC sites Sites which could collect carpets Residual waste to landfill from suitable sites Carpet waste (8%) Assuming a 75% availability rate for carpet waste ,000 tonnes 7,200 tonnes 5,400 tonnes Cheshire ,750 tonnes 1,260 tonnes 945 tonnes Using the information in the above table an estimate of the carpet waste airings in the whole of the Northwest of England was completed. It was assumed that 60% of HWRC sites could take a suitable container for carpet collections, meaning 63 of the 103 HWRCs in the Northwest region could be considered. The residual waste from Merseyside and Cheshire HWRCs accounts for 30% of the total waste in the Northwest region. Extrapolating the case study results of 6,345 tonnes per annum of carpet waste from Merseyside and Cheshire to the whole of the Northwest region gives 21,150 tonnes per annum of waste carpets. However this figure is for all carpet waste and, as stated previously, the composition of mixed carpet waste is such that only 27% of the material will meet the specification. Hence 5,710 tonnes of post consumer tufted PP carpets could be collected from HWRCs in the region Total carpet waste The calculations show the following volumes of waste: Industrial waste tonnes; Commercial waste tonnes; and Domestic waste 5,710 tonnes. Total waste arisings of carpet which meets the specification is estimated at 6,400 tonnes per annum. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 27

28 4 Proposed recycling flowsheet for waste PP carpets Figure 17 shows the proposed process for the recycling of post consumer carpets. One of the main aspects of the recycling process is the correct identification of the carpets to ensure that only PP carpets are processed through the recycling plant. The recycling plant requires three main processing stages: Size reduction; Screening/cleaning; and Extrusion. Figure 17 recycling process for waste PP carpets Overall stages required for recycling for PP carpets Collection of carpet waste from a range of sources Delivery of carpet waste to central processing location Polymer identification and testing of carpets Segregation of PP and mixed polymer carpets Size reduction of carpets Screening of size reduced carpets Extrusion of cleaned carpets PP pellets Waste fraction Non-PP carpets and PP woven carpets Fines fraction Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 28

29 4.1 Mass balance for recycling PP carpets Two options for the recycling of post consumer PP carpets have been proposed: Option 1 involves buying in bales of pre-sorted tufted PP carpet at 80/te; and Option 2 involves buying in size reduced tufted PP carpets at 250/te. Figure 18 Option 1 mass balance for the recycling of PP carpets PP carpet recycling mass balance Option 1 Power cost Stage operating 8 hours, 5 days a week Power cost Stage operating 8 hours, 5 days a week Power cost Stage operating 8 hours, 5 days a week Power cost Stage operating 24 hours, 5 days a week 5000 te/year of Baled sorted tufted PP carpet Cost 80/te De-baling stage 5000 te Size reduction 5000 te Screening 4000 te Extrusion 3880 te of PP pellet stage Sell for 500/te stage Yield = 80% 1000 te stage Yield = 97% 120 te Waste to landfill 1120 te Disposal cost 60/te Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 29

30 Figure 19 Option 2 mass balance for the recycling of PP carpets PP carpet recycling mass balance Option 2 Power cost Stage operating 8 hours, 5 days a week Power cost Stage operating 24 hours, 5 days a week 5000 te/year of Sorted and size reduced tufted PP carpet Cost 250/te Screening stage Yield = 80% 1000 te 4000 te Extrusion 3880 te of PP pellet stage Sell for 500/te Yield = 97% 120 te Waste to landfill 1220 te Disposal cost 60/te Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 30

31 The option 1 mass balance in Figure 18 has the following assumptions: The process is split into two stages: o Stage one is the de-baling, size reduction and screening stages. These operate 8 hours a day for 5 days a week; o Stage two is the extrusion stage which operates 24 hours a day for 5 days a week; A feed quantity of 5,000 tonnes per year of post consumer carpets pre-sorted into tufted PP carpets costing 80/te delivered; It has been assumed that 100% of the carpets delivered to the plant will meet the specification; All of the carpets are shredded and screened with a yield of 80%; The undersize fraction from the screen has no economic value and will be sent to landfill to be disposed of; and The oversize fraction from the screen is extruded. The extrusion process produces a small quantity of waste which also has to be landfilled. The option 2 mass balance in Figure 19 has to following assumptions: The process is split into two stages: o Stage one is the screening stages which operates for 8 hours a day for 5 days a week; o Stage two is the extrusion stage which operates 24 hours a day for 5 days a week; A feed quantity of 5,000 tonnes per year of sorted and size reduced post consumer tufted PP carpets costing 250/te; It has been assumed that 100% of the carpets delivered to plant will meet the specification; The size reduced carpets are screened, again with a yield of 80%; The undersize fraction from the screen has no economic value and will be sent to landfill for final disposal; and The oversize fraction from the screen is extruded. The extrusion process produces a small quantity of waste which also has to be landfilled. It should be noted that the landfill disposal cost will increase to at least 68/t in April 2010 when landfill tax increases to 48/t Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 31

32 5 Commercial viability of recycling post consumer carpets The project has shown that from a technical perspective post consumer carpets can be recycled into an extruded PP material, which has potential use in a range of end markets. However the commercially viability of the process is dependent on the following factors: A regular supply of feed material, which meets the specification; Rigorous identification techniques to ensure that only PP material is processed through to extrusion. If too many carpets containing non-pp fibres were processed the resultant product would be of too low a grade to be extruded and hence would not achieve the same market value as the PP pellets; and Research conducted as part of this project indicated that the key issue for many carpet recycling plants is the economics of the process. If landfill is a cheaper option than processing, which it has been and in some places may still be the case, then the economics of the recycling process can be unfavourable. Based on the information obtained during the project a payback calculation has been completed for each option to assess the commercial viability of the recycling process. The payback calculation for option 1 uses the following assumptions: The plant capacity is 7,500 tonnes per annum; The project did not select any specific processing equipment and therefore the capital cost estimate is based on Axion s own recycling equipment experience. The capital cost takes into account the purchasing of: o Carpet identification equipment; a hand held NIR sorter(s) or a bench top FTIR machine for quality control; o 4te per hour de-baler; o 4te per hour size reduction machine; o Screening machine; o Buffer storage silos before extrusion; o Three 400kg extrusion machines; o Conveyors; o Fire protection; The calculation assumes that a suitable site and building for the processing facility would be rented rather than purchased, hence the inclusion of an 80,000 rental cost; The power cost break down consists of: o Stage 1 of the process operates for 2,000 hours per annum and 5,000 tonnes of carpet can be de-baled, shredded and screened. The power cost for stage 1 is 14,000 per annum based on 10p/kW hr; o Stage 2 requires three 400kg extruders to meet the necessary capacity. This stage runs for 6,000 hours per annum and processes 4,000 tonnes of carpet which costs 252,000 per annum. o Total power cost is 266,000; The labour costs breakdown consists of: o Labour for stage 1 - de-baling, size reduction and screening: o Three operatives required for stage 1 working 8 hours a day, 5 days a week at a job cost of 18,000 per annum per person, gives a labour cost of 54,000; o Labour for stage 2 - extrusion: o Extrusion process will operate 24 hours a day, 5 days a week which requires three shifts with three operatives per shift, with an additional allowance of two personnel, at a job cost of 18,000 per annum per person, gives a labour cost of 198,000; o A technical manager at 25,000; o Total labour cost is 277,000; Other additional operating costs include compliance costs such as quality control, health and safety and environmental considerations; It has been assumed the cost to the processor for the waste carpet to be delivered to the processing site will be 80/te; The small quantity of material, from the screening and extrusion process, which requires disposal to landfill costs 38,100 per annum; Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 32

33 The extruded PP pellet has been estimated to sell for 500/te creating just under 2 million revenue per annum; and The calculation shows the margin per annum is over 800,000, which gives a payback time of just over one year. Table 5 Economic payback calculation for option 1 tpa 7,500 Capacity te/hr 3.57 Capital cost of plant 1,000,000 Operating costs Power Stage 1 Basis of operation for debaling, size reduction and screening hr/yr 2,000 De-baler kw hr/hr 20 Size reduction kw hr/hr 60 Screening kw hr/hr 20 Cost (assuming 10p/kW hr) /hr 10 Power costs /te of feed 2.80 Overall Equipment Effectiveness (OEE) % 70% Plant input for shredding te/yr 5,000 Stage 1 power costs /yr 14,000 Power Stage 2 Extruder capacity te/hr 1.0 Basis of operation for extrusion hr/yr 6,000 Extrusion (3 x 400kg extruders) kw hr/hr 600 Cost (assuming 10p/kW hr) /hr 60 Power costs /te of feed 63 Overall Equipment Effectiveness (OEE) % 70% Material for extrusion te/yr 4,000 Stage 2 power costs /yr 252,000 Total power costs /yr 266,000 Labour Labour costs stage 1 3 operators at 8 hours per day with job cost 18,000 /yr Labour costs stage 2 3 shifts with 3 people per shift + 2 spare with job cost 18,000 /yr Technical management /yr 25,000 Total labour costs /yr 277,000 Others Rent and rates 10,000 sq ft at 8/sq ft /yr 80,000 Compliance - quality, safety, environment /yr 30,000 Insurance /yr 15,000 Total Operating Costs 668,000 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 33

34 Table 5 Economic payback calculation Table 5 cont d Cost of delivered feed /te 80 Feed cost /yr 400,000 Disposal cost (at 60/te) /yr 38,100 Revenue PP pellet - value of product /te 500 Quantity of PP pellet te/yr 3,880 Revenue from PP product /yr 1,940,000 Total revenue /yr 1,940,000 Margin /yr 833,900 Payback time (years) 1.2 The payback calculation for option 2 uses the following assumptions: The plant capacity is 7,500 tonnes per annum; The project did not select any specific processing equipment and therefore the capital cost is estimated based on Axion s own recycling equipment experience. The capital cost takes into account the purchasing of: o Carpet identification equipment; a hand held NIR sorter(s) or a bench top FTIR machine for quality control; o Screening machine; o Buffer storage silos before extrusion; o Three 400kg extrusion machines; o Conveyors; o Fire protection; The calculation assumes that a suitable site and building for the processing facility would be rented rather than purchased, hence the inclusion of an 64,000 rental cost; The power cost break down consists of: o Stage 1 of the process operates for 2,000 hours per annum and 5,000 tonnes of carpet can be screened. The power cost for stage 1 is 2,800 per annum based on 10p/kW hr; o Stage 2 requires three 400kg extruders to meet the necessary capacity. This stage runs for 6,000 hours per annum and processes 4,000 tonnes of carpet which costs 252,000 per annum; o Total power cost is 254,800; The labour costs breakdown consists of: o Labour for stage 1 - de-baling and screening; o One operative required for stage 1 working 8 hours a day, 5 days a week at a job cost of 18,000 per annum per person gives a labour cost of 18,000; o Labour for stage 2 - extrusion: o Extrusion process operating 24 hours a day, 5 days a week will require thee shifts with three operatives per shift with an additional allowance of two personnel, at a job cost of 18,000 per annum per person, gives a labour cost of 198,000; o A technical manager at 25,000; o Total labour cost is 241,000; Other additional operating costs include compliance costs such as quality control, health and safety and environmental considerations; It has been assumed the cost for the sorted and shredded carpet to be delivered to the processing site will be 250/te; The small quantity of material, from the screening and extrusion process, which requires disposal to landfill costs 38,100 per annum; The PP pellet has been estimated to sell for 500/te creating just less than 2 million revenue per annum; and Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 34

35 The calculation shows the margin per annum is 47,100 which gives a payback time of 17 years. Table 6 Economic payback calculation for option 2 tpa 7,500 Capacity te/hr 3.57 Capital cost of plant 800,000 Operating costs Power Stage 1 Basis of operation for screening hr/yr 2,000 Screening kw hr/hr 20 Cost (assuming 10p/kW hr) /hr 2 Power costs /te of feed 0.56 Overall Equipment Effectiveness (OEE) % 70% Plant input for screening te/yr 5,000 Stage 1 power costs /yr 2,800 Power Stage 2 Extruder capacity te/hr 1.0 Basis of operation for extrusion hr/yr 6,000 Extrusion (3 x 400kg extruders) kw hr/hr 600 Cost (assuming 10p/kW hr) /hr 60 Power costs /te of feed 63 Overall Equipment Effectiveness (OEE) % 70% Material for extrusion te/yr 4,000 Stage 2 power costs /yr 252,000 Total power costs /yr 254,800 Labour Labour costs stage 1 1 operators at 8 hours per day with job cost 18,000 /yr 18,000 Labour costs stage 2 3 shifts with 3 people per shift + 2 spare with job cost 18,000 /yr 198,000 Technical management /yr 25,000 Total labour costs /yr 241,000 Others Rent and rates 8,000 sq ft at 8/sq ft /yr 64,000 Compliance - quality, safety, environment /yr 30,000 Insurance /yr 15,000 Total Operating Costs 604,800 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 35

36 Table 6 Cont d Cost of delivered feed /te 250 Feed cost /yr 1,250,000 Disposal cost (at 60/te) /yr 38,100 Revenue PP pellet - value of product /te 500 Quantity of PP pellet te/yr 3,880 Revenue from PP product /yr 1,940,000 Total revenue /yr 1,940,000 Margin /yr 47,100 Payback time (years) 17.0 The two payback calculations show that the more economically favourable scenario is to purchase sorted bales of tufted PP carpet and to size reduce the material as part of the process rather than purchasing already size reduced material. The main issue with the second option is the high cost of the feed material. Both processes have similar overheads and revenue but the feed costs are significantly different, hence the difference in the payback times. If the cost of the feed material in option 1 is increased from 80/te to 150/te the payback time increases to 2.1 years but overall the economics are still promising. Therefore, there is flexibility in the option 1 model to cope with increased feed prices. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 36

37 6 Conclusions The main conclusions from the project are: Post industrial carpet can be processed into a product with good physical properties which has potential for use in a range of applications including injection moulding; Post consumer carpet, consisting of entirely PP fibres, can be recycled into a product with good physical properties making it suitable for use in medium to low grade applications, for example injection moulded plant pots or buckets. The processing stages required to achieve the product are straight forward and involve size reduction, screening and extrusion; Should a carpet recycling plant be of commercial interest further work would be required to select the correct equipment for each of the processing stages; However, a critical factor to the success and quality of the extruded polymer is the correct identification of all polymer fibres within the carpet. Carpets which contain polymers other than PP, for example nylon or PET, produce a granulated material which cannot be extruded. It is expected that carpets which contain mixed polymer blends are more likely to be of woven than tufted construction. The probable end market for mixed polymer carpets remains equestrian surface applications; and The economics of the recycling process look promising and the assessment considered two possible processing scenarios of which the option of buying pre-sorted whole carpets was favourable over pre-sorted and size reduced carpets. The payback period of the first option (buying pre-sorted whole carpets) was just over a year; whereas the payback for the second option (buying pre-sorted and size reduced carpets) was 17 years. In conclusion the project has proved that a good quality polymer pellet can be produced from post consumer carpets made entirely of PP, if all of the carpet fibres are correctly identified. Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 37

38 Appendices Appendix 1 - Carpet Recycling UK Tufted PP carpet waste arising in NW, 2009 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 38

39 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 39

40 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 40

41 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 41

42 Project Name: Envirolink WPC0102 PP recovery from post consumer carpets 42