Performance indicators for a circular economy: A case study on post-industrial plastic waste

Resources, Conservation and Recycling 120 (2017) 46–54

Contents lists available at ScienceDirect

Resources, Conservation and Recycling journal homepage: www.elsevier.com/locate/resconrec

Full length article

Performance indicators for a circular economy: A case study on post-industrial plastic waste Sofie Huysman a , Jonas De Schaepmeester a , Kim Ragaert b , Jo Dewulf a , Steven De Meester a,c,∗ a

Research Group ENVOC, Ghent University, Coupure Links 653, 9000 Ghent, Belgium Research Group Applied Material Sciences, Ghent University, Valentin Vaerwyckweg 1, 9000 Gent, Belgium c IBW Department of Industrial Biological Sciences, Ghent University, Graaf Karel de Goedelaan 5, 8500 Kortrijk, Belgium b

a r t i c l e

i n f o

Article history: Received 28 June 2016 Received in revised form 23 December 2016 Accepted 23 January 2017 Available online 5 February 2017 Keywords: Indicators Circular economy Compatibility Plastic waste LCA

a b s t r a c t A linear economy approach results in many environmental challenges: resources become depleted and end up as waste and emissions. One of the key strategies to overcome these problems is using waste as a resource, i.e. evolving toward a circular economy. To monitor this transition, suitable indicators are needed that focus on sustainability issues whilst taking into account the technical reality. In this paper, we develop such an indicator to quantify the circular economy performance of different plastic waste treatment options. This indicator is based on the technical quality of the plastic waste stream and evaluates resource consumption by using the Cumulative Exergy Extraction from the Natural Environment (CEENE) method. To illustrate the use of this new indicator, it was applied in a case study on post-industrial plastic waste treatment. The results show that the indicator can be a very useful approach to guide waste streams towards their optimal valorization option, based on quality of the waste flow and the environmental benefit of the different options. © 2017 Elsevier B.V. All rights reserved.

1. Introduction The transition towards a more sustainable society is a complex task. One of the key strategies to manage this transition is the circular economy concept. Preston (2012) defined the idea of a circular economy as follows: “open production systems − in which resources are extracted, used to make products and become waste after the product is consumed − should be replaced by systems that reuse and recycle resources and conserve energy”. This idea has been implemented in several governmental policies, with Japan and Europe at the forefront. The Japanese government introduced the material-cycle society vision in the year 2000. This vision involves a number of laws based on the 3R (reduce, reuse, recycle) principle (Government of Japan, 2010). Recent strategies in the European Union (EU) are the ‘Zero waste programme for Europe’ (EC (European Commission), 2014) and ‘Closing the loop action plan for the Circular Economy’ (EC (European Commission), 2015). An important material that still can be improved within the circular economy is plastic, as also confirmed in the recent report

∗ Corresponding author at: IBW Department of Industrial Biological Sciences, Ghent University, Graaf Karel de Goedelaand 5, 8500 Kortrijk, Belgium. E-mail address: [email protected] (S. De Meester). http://dx.doi.org/10.1016/j.resconrec.2017.01.013 0921-3449/© 2017 Elsevier B.V. All rights reserved.

The New Plastics Economy: Rethinking the future of plastics (World Economic Forum, 2016). Indeed, the role of plastics in our daily life cannot be underestimated. Ever since the production of Bakelite in 1907, the importance of plastics in society kept growing. In 2014, the global production of plastic was 311 million tonnes. Europe is the second largest producer of plastic materials, responsible for 20% of the world production. Packaging applications are the largest application sector, representing 39.6% of the total plastics demand (Plastics Europe, 2015). However, the problem is that all these plastics end up as waste. In 2014, Europe produced 25.8 million tonnes of post-consumer plastics waste: 29.7% was recycled, 39.5% was incinerated with energy recovery, and 30.8% was landfilled. Landfill of plastic waste may cause environmental problems, as plastics are often not biodegradable. Further, there is also the problem of resource conservation. The production of plastics consumes yearly 4 to 8 % of the global crude oil extraction (Kreiger et al., 2014). If plastics are disposed instead of being recycled, these resources are lost (EC (European Commission), 2013). Hence, the role of plastic waste is a major issue in circular economy strategies. To monitor plastic waste treatment management, suitable indicators are needed. In the current policies, most indicators are situated at the macro-economic level (countries, regions),

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

for example in the Japanese 3R-policy (Takiguchi and Takemoto, 2008). Fewer indicators are situated at the micro-level (products, companies). One example is the recyclability benefit rate (RBR) indicator, developed by the European Commission’s Joint Research Centre (JRC) (EC-JRC, 2012; Ardente and Mathieux, 2014), which is based on an LCA-approach (Life Cycle Assessment). It is defined as the ratio of the environmental benefits that can be obtained from recycling a product, over the environmental burdens related to production from virgin resources followed by disposal. These benefits and burdens are expressed in terms of environmental impacts, calculated through LCA. In our previous work, this indicator was further developed for open-loop recycling systems (Huysman et al., 2015). The report of JRC (EC-JRC, 2012) suggested also an alternative version of the RBR indicator, attempting to take the quality loss occurring during recycling into account. Indeed, once plastics have gone through a recycling process, they most often have no longer the properties they had in their original virgin state. This is due either to thermo-mechanical degradation during (re-)processing or to the fact that the plastics get mixed with other types during the recycling process (Ignatyev et al., 2014). To bring this potential quality loss into account, the report proposes to use a quality factor that is defined as the ratio of the quality of the recycled material over the quality of the virgin material. The measurement of quality however is a difficult issue, which has no common understanding in the scientific community. The report suggests that this quality can be measured through physical parameters (e.g. the tensile strength) or economic parameters (e.g. market price) (EC-JRC, 2012). In most cases, the price of the recycled material versus the virgin material is used, as described in the work of Villalba et al. (2002). Nonetheless, the use of monetary values has its disadvantages, as market values and prices fluctuate heavily over time. Problems may also arise when prices are missing or distorted, e.g. monopolies, or when there are government interventions, e.g. subsidies (Ardente and Cellura, 2011). Physical parameters on the other hand are independent from changes in the economy. However, they are rarely applied, as is it difficult to determine a suitable physical parameter for each material type, and it is another research field. Another issue is the implementation of the quality factor in the formula of the RBR indicator. With this indicator, treatment of lower quality waste always has a lower benefit, regardless of how it is valorized. As a result industries that process waste of lower quality would always get a low result whereas the responsibility of the quality of the waste is mainly determined by the preceding production and application. From a perspective of waste valorization benefits, it would be more adequate to use the quality factor as a classification tool, to select the most suitable waste treatment option according to the quality of the plastic waste flow. Therefore, the objective of this paper is to develop a circular economy performance indicator, defined as the ratio of the actual obtained environmental benefit (i.e. of the currently applied waste treatment option) over the ideal environmental benefit according to quality for this flow. Similar to the RBR indicator, these benefits can be quantified in terms of environmental impacts, calculated through LCA. From a historical perspective, the focus in LCA is on impacts related to emissions. In this paper, the focus is shifted to natural resources, as these are more relevant in the context of a circular economy. Therefore, we selected the CEENE (Cumulative Exergy Extraction from the Natural Environment) method, quantifying resource consumption (Dewulf et al., 2007). The possible waste treatment options (closed-loop recycling, semi closed-loop recycling, open-loop recycling and incineration) are discussed more detailed in the Materials & Methods sections. To determine the most suitable waste treatment option from a technical point of view, we developed a quality factor for plastic

47

waste, based on a physical parameter: the compatibility between the composing polymers in the mix, which plays a major role in the resulting mechanical properties of the polymer blend (Koning et al., 1998). Finally, the use of the indicator is illustrated with a case study on plastic waste treatment. There are broadly two types of plastic waste to be dealt with: post-consumer waste, which is generated by end-users, e.g. households, and post-industrial (or pre-consumer) waste, which is generated during the manufacturing phase (Reynolds and Pharaoh, 2010). This is similar to the distinction made in steel industry: old scrap consists of used goods (e.g. vehicles, machine parts), while new scrap is generated during steel production (Bilitewski et al., 1997). Most research is focused on post-consumer waste. Examples are the studies on packaging waste recycling systems in Portugal (Ferreira et al., 2014) and in Istanbul (Yıldız-Geyhan et al., 2016). Other examples are the studies of Simon et al. (2016) and Tonioli et al. (2013), which are focused on the recycling of beverage bottles. However, case studies on the recycling of post-industrial plastic waste are more limited. Therefore, the indicator will be demonstrated in a case study on post-industrial plastic waste treatment. 2. Materials and methods 2.1. Development of the indicator 2.1.1. Possible waste treatment options The ISO 14044 standard makes a distinction between two types of recycling: closed-loop recycling occurs when ‘a material from a product is recycled in the same product system’, open-loop recycling occurs when ‘a material from one product system is recycled in a different product system’. However, in this classification, the link with the material quality is missing. Therefore, we propose the following classification for the possible waste treatment options: if the plastic is of high quality, it can substitute the virgin original material in a 1:1 ratio (closed-loop recycling, option I). If the quality is lower, there are two possibilities: (1) the recycled material can still substitute the original virgin material, but not in a 1:1 ratio, as additional virgin material has to be added to meet the same quality requirements (semi closed-loop recycling, option II); (2) the recycled plastic can only be used in low-grade applications, in which it substitutes different types of materials (open-loop recycling, option III). In the worst case scenario, if the quality is extremely low, the waste can only be incinerated for energy recovery (incineration, option IV) (Fig. 1). 2.1.2. Calculating the performance indicator For each of these waste treatment options, it is possible to calculate the ‘circular economy performance indicator’ (CPI). This paper defines the CPI as the ratio of the actual obtained environmental benefit (i.e. of the currently applied waste treatment option) over the ideal environmental benefit according to quality, the latter being the benefit of the waste treatment option to which the stream should be directed according to its composition/quality with a minimal required effort, assuming option I (closed-loop recycling) is better and option IV (incineration) is less preferable: CPI =

actual benefit ideal benefit according to quality

These environmental benefits are expressed in terms of natural resource consumption, which can be calculated by Life Cycle Assessment, for example by using the CEENE method as LCIA. In option I (closed-loop recycling), the recycled material has the potential to substitute the original virgin material (␣) in a 1:1 ratio. For example, 1 kg recycled PE substitutes 1 kg virgin PE. However,

48

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

Fig. 1. different waste treatment options.

one must also take into account the recycling rate r and the impact of the recycling process R. The recycling rate r is the amount of effectively recycled material produced per kg waste input when considering that part of the materials are lost during recycling. The environmental benefit of option I is thus the avoided impact of virgin production V˛ multiplied with recycling rate r, minus impact R. The latter is the resource consumption or CEENE required for the recycling process. In option II, the recycled material can substitute the original virgin material ␣ only partially. For example, the final product consists of 80% recycled PE and 20% virgin PE (in terms of mass). The virgin material is added to fulfill the quality requirements. To calculate the environmental benefit of option II, the avoided impact of the virgin production of material ␣ (V˛ ) has to be multiplied with the maximal percentage of substituted virgin material (p) and the recycling rate r. Next, the impact of the recycling process R has to be subtracted from this result. In the example above, the substitution percentage is 80%. Option I is in fact a special case of option II, in which p equals 100%, and the product application remains the same. In option III, the recycled material can only be used in low-grade applications, in which it substitutes another type of virgin material (ˇ), be it another type of plastic or a different material type entirely. For example, 95.5 kg recycled PE is used to produce a street bench. Normally, this street bench is made of 63 kg cast iron, thus 1 kg recycled PE substitutes 0.65 (63/95.5) kg cast iron (Huysman et al., 2015). To calculate the benefits, the substitution percentage has to be taken into account: mV,ˇ kg virgin material is substituted





by mR,ˇ kg recycled material m = mV,ˇ /mR,ˇ . The environmental benefit of option III is thus the avoided impact of the virgin production of material ˇ (Vˇ ), multiplied with the maximal substitution percentage m, minus the impact of the recycling process R. In option IV, the waste can only be incinerated, as recycling is not possible. Hence, the environmental benefit of option IV is the avoided impact for the virgin production of the obtained amount of energy (E), including both heat and electricity, minus the impact of the incineration process I. These environmental benefits are presented in Table 1. For example, based on the quality factor, the waste should go

to waste treatment option I. If this waste is actually valorized in waste I, the CPI indicator becomes  treatment option  (rab .V˛ − Rab ) / ribaq .V˛ − Ribaq in which rab and Rab are the actual recycling rate and the actual impact of the recycling process respectively, whereas ribaq and Ribaq are the recycling rate and the impact of the recycling process according to the quality of the waste with a minimal environmental impact and state-of- the-art technologies, e.g. determined by BAT studies. For reasons of simplification, in the following these factors will be considered as 1 and 0 respectively, assuming that no material is lost during recycling and the ideal recycling process has a minimal impact. When in reality, the waste goes to waste treatment option III, the actual obtained environmental benefit is r.m.Vˇ − R. Thus, the circular economy performance   indicator (CPI)  of the considered system is rab .m.Vˇ − Rab / ribaq .V˛ − Ribaq . With ribaq and Ribaq assumed





to be respectively 1 and 0 this becomes rab .m.Vˇ − Rab / (V˛ ). A good circular economy has a CPI value equal to one. In that case, the actual environmental benefit (obtained by the currently applied waste treatment) is equal to the ideal environmental benefit according to quality (obtained by treating the waste according to its technical quality), meaning that the currently applied recycling process indeed has an insignificant material loss and impact. Usually, the CPI value will be smaller than 1, meaning that the actual environmental benefit is lower than the ideal environmental benefit according to quality. This can mean that or: • The waste treatment/recycling scenario has a significant material loss or environmental burden • The waste is not treated according to its scenario based on technical quality Occasionally, the CPI can also be larger than 1 meaning that the actual environmental benefit is larger than the ideal environmental benefit according to quality. This means that from an environmental point of view, the actually chosen waste treatment is the best option. This is because the aforementioned assumption that option I (closed-loop recycling) is better and option IV (incineration) is less preferable. In some cases this can be the case, e.g. a very efficient

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

49

Table 1 Ideal environmental benefit according to quality and actual environmental benefit for each waste treatment option. Option I is closed-loop recycling, option II is semi closed-loop recycling, option III is open-loop recycling and option IV is incineration.

Actual environmental benefit Ideal environmental benefit according to quality

Option I

Option II

Option III

Option IV

r.V˛ − R V˛

r.p.V˛ − R p.V˛

r.m.Vˇ − R m.Vˇ

E−I E

incineration compared to open loop recycling in which materials are substituted with a low environmental impact. This would mean that it should be analyzed if the state-of-the art recycling processes can be improved. Another example is the case of upcycling, which could be considered as a special case of open loop recycling, e.g. when the plastic is of good quality (for option I), but it is used to substitute another material with a higher environmental burden compared to the virgin plastic. Key in this approach is to find an approach that allows to determine to which waste treatment option the waste should be directed. Such a first approach is discussed below.

2.1.3. Most suitable option based on the technical quality In many cases, such as the multilayer films used in the case study of the current research, it is physically impossible (or economically not viable) to fully separate recycled polymers into their composing mono-streams (Hopewell et al., 2009). However, different polymer types are inherently immiscible due to their difference in chemical nature (Van Krevelen, 2009). An untreated blend will typically end up as a two-phase morphology with poor interfacial adhesion, which in turn leads to poor mechanical properties. One such mechanical property is the impact strength, which decreases significantly after blending two incompatible polymers (Lei et al., 2009). Therefore, we propose an approach to predict this quality based on the compatibility between the composing polymers in a mix, as compatibility is considered to be one of the most crucial parameters (Tall et al., 1998). In the current preliminary stage, this includes the following simplifications: (i) the mix is supposed as being binary in nature, (ii) the effect of thermo-mechanical degradation of the polymer chains is considered negligible in relation to the adverse effect of mixing and therefore not taken into account, and (iii) the mix in considered as untreated, meaning no compatibilizing agents have been added. All of these aspects are part of the future refinement of the model. Several accepted models to determine the miscibility of a polymer blend are available, all of them related to the interfacial activity (or lack thereof) between the components in the molten state (Manias and Utracki, 2014): (i) the Flory-Huggins parameter, which determines the enthalpy-of-mixing as a complex function of nine parameters and varies with concentration and temperature, (ii) the Hildebrand solubility parameter and (iii) the interfacial tension method. As the latter relates to interfacial adhesion and therefore bears a strong relevance to the solid-state mechanical behavior of the resulting material, we have selected the interfacial tension method. The interfacial tension between the polymers is related to the components of the respective polymers’ surface energy. The surface energy , expressed in energy per unit area (mJ/m2 , milli Joules per square meter) or force per unit length (mN/m, milli Newtons per meter), can be seen as the result of two contributing intermolecular forces: dispersion forces and polar forces (Fowkes, 1964). The surface energy is temperature-dependent, and decreases with increasing temperature (Falsafi et al., 2007). Wu (1971) proposed a formula for the interfacial tension  12 (mN/m) between two poly-

Table 2 Lower and upper boundaries for the compatibility classes. range

compatibility class

0 < 12 ≤ 0.1 0.1 < 12 ≤ 1 1 < 12 ≤ 10 10 < 12

(i) Perfectly compatible (ii) Reasonably compatible (iii) Limited compatible (iv) Incompatible

Table 3 Lower and upper boundaries for the waste treatment options. Range

waste treatment option

0.9 < Q ≤ 1 0.7 < Q ≤ 0.9 0.2 < Q ≤ 0.7 0 < Q ≤ 0.2

I. Closed-loop II. Semi closed-loop III. Open-loop IV. Incineration

mers, with ␥id and ␥ip the dispersion and polar components of the surface energy ␥i of these two polymers (i = 1,2). 12 = 1 + 2 − 4

1p 2p 1d 2d −4 1d + 2d 1p + 2p

The calculation of the interfacial tension between the most abundant polymers is included in the Supporting Information. Based on the interfacial tension  12 , four compatibility classes are defined: perfectly compatible, reasonably compatible, limited compatible and incompatible. This is presented in Table 2. Typical examples of blends related to this classes are LPE-PE for ‘perfectly compatible’, PE-PP for ‘reasonably compatible’, PP-PET for ‘limited compatible’ and PE-PA for ‘incompatible’. Per compatibility class, the quality of the recycled plastic Q varies (between 0 and 1) in function of the mass percentage of the added polymer. These graphs are included in the supporting information. Finally, each waste treatment option is linked with a certain range of Q-values, see Table 3. This first model of Q as a function of compatibility class and composition, as well as the definition of the boundaries for Q is based on a literature review (Bertin and Robin, 2002; Borovanska et al., 2012; Brandalise et al., 2009; Haider et al., 2007; Inoya et al., 2012; Kukaleva et al., 2003; Navarro et al., 2012) and own experimental experience (Van Bruggen et al., 2016; Delva et al., 2016; Hubo et al., 2015a; Hubo et al., 2015b; Ragaert et al., 2014) with blends from the different classes. Obviously, this is a preliminary approach and future (experimental) research could be performed to validate and improve this model. 2.2. Illustration with a case study 2.2.1. Description of the system An overview of the case study is presented in Fig. 2. The company producing the post-industrial plastic waste is company A, located in Belgium. This company is specialized in the production of flexible packaging. It consists of a co-extrusion department, where the plastic films are produced, and a conversion department, where the plastics films are printed, laminated and slitted. In this case study, the focus is on the plastic waste produced by the co-extrusion department in the year 2013, more specifically polyethylene (PE) waste. There are two types of PE waste: the smallest fraction (25%,

50

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

Fig. 2. Overview of the case study. The first block shows the actual waste treatment scenario, the second block shows the alternative waste treatment scenarios for waste stream 1.

stream 1) consists of pure PE. The largest fraction (75%, stream 2) consists of PE with a small amount of contamination, i.e. <5% ethylene vinyl alcohol (EVOH). Stream 1 is transported by train to company B, located in England. This company is also a producer of flexible packaging. Company A could not use recycled materials because of the high requirements for the end-products, which are used in the food and medical sector. Company B on the other hand is allowed to use recycled materials because the requirements for its end-products are lower. Company B uses two different types of recycling machines: the more recent machine has a closed water cooling circuit, while the older machine has an open water cooling circuit. In a closed water circuit, water that evaporates during the cooling process is condensed again afterwards. The older machine is only used when the more recent machine has an insufficient capacity. The majority of the films produced by company B consist of 20% recycled PE and 80% virgin PE. This is because the recycled material is only used for the middle layer, otherwise all the impurities and stripes would be visible, which is mostly not wanted by the clients. Hence, the waste goes to waste treatment option II (semi closed-loop recycling). Nonetheless, some clients specifically ask for films made entirely out of recycled material. In that particular case, the waste stream goes to waste treatment option I (closed-loop recycling). Stream 2 is first transported by truck to a waste collection site. From this site, the waste is again transported by truck to company C, located in Belgium. Company C is a manufacturer of garbage bags and industrial films. The produced garbage bags are composed of 80% recycled PE and 20% virgin material, which is linear low density polyethylene (LLDPE). This is necessary because the bags need to fulfill certain requirements that are federal regulated: drop impact resistance and resistance to leakage. This means that stream 2 undergoes waste treatment option II.

We will also make a comparison with two alternative scenarios. In the first alternative scenario, stream 1 is transported to company D, located in Belgium, where it is subjected to waste treatment option III (open-loop recycling). This waste treatment system has been described in an earlier study (Huysman et al., 2015). We consider the case in which PE waste is used to produce a street bench. Normally, this street bench would be made of 63 kg cast iron. If the street bench is made of recycled plastic, 95.5 kg recycled PE is required. This means that 1 kg recycled PE substitutes 0.65 (63/95.5) kg cast iron (m = 0.65). Or, with a recycling rate of 80%, 1 kg PE waste is used to substitute 0.52 (0.65*0.8) kg cast iron. In the second alternative scenario, stream 1 is transported to company E, where it undergoes waste treatment option III (incineration). This company, which is also located in Belgium, treats industrial waste and sludge that cannot be recycled or landfilled. Per kg waste incinerated in the fluidized bed incinerator, 2.74 kWh of electricity is produced. 2.2.2. LCA in terms of resource consumption The LCA is performed according to the ISO 14040/14044 guidelines (ISO (International Organization for Standardization), 2006a; ISO (International Organization for Standardization), 2006b). Foreground data were collected in close collaboration with the companies. To model the background system and the avoided virgin production, the Ecoinvent v2.2 database (Swiss Centre for Life Cycle Inventories, 2016) was used. The environmental impact will be expressed in terms of resource consumption by applying the Cumulative Exergy Extraction from the Natural Environment (CEENE) v.2013 as impact assessment method (Alvarenga et al., 2013; Dewulf et al., 2007). This method quantifies the total exergy that is contained in the various resources extracted from the natural environment over the life cycle of a product. The exergy of a natural resource is

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

defined as the minimum energy required to produce it with a specific structure and concentration from common materials in the reference environment (Valero et al., 2013). Compared to other exergy-based impact methods, e.g. the Cumulative Exergy Demand (CExD) (Bösch et al., 2007), the CEENE method covers a more complete resource range: fossil energy, nuclear energy, metal ores, minerals, water resources, land use, abiotic renewable resources (including wind power, geothermal energy and hydropower) and atmospheric resources (Liao et al., 2012). All the extracted resources are quantified and expressed in megajoules of exergy (MJex ). The CEENE method has already been applied in several case studies, for example in the resource use assessment of a dairy farm (Huysveld et al., 2015) and pharmaceutical tablet manufacturing (De Soete et al., 2013).

3. Results and discussion 3.1. Resource consumption As mentioned in the introduction, the environmental impact is expressed in terms of resource consumption (‘resource footprints’) by using the CEENE LCIA method. The functional unit is 1 kg plastic waste. Fig. 3 presents the resource footprint of the waste treatment process itself (R value), and Fig. 4 shows the resource footprint of the avoided virgin production. For stream 1 going to options I & II (company B), we made a distinction between the two recycling machines. If the more recent recycling machine is used, the resource footprint of the recycling process is 3.93 MJex . If the older recycling machine is used, the resource footprint is 11.21 MJex . In the current situation, the older (10%) and recent recycling machine (90%) are combined. This combination results in a resource footprint of 4.42 MJex per kg waste, see Fig. 3. The type of recycling machine has thus a large influence on the final results. The avoided impact is the resource footprint of the virgin production of 1 kg PE (V␣ = 80.95 MJex ) multiplied with the p-value (100% or 20%) and the recycling rate r (99.56%). The result is thus either 80.59 MJex (p = 100%) or 16.12 MJex (p = 20%) per kg waste, see Fig. 4. For stream 2 going to option II (company C), the resource footprint of the recycling process is 2.23 MJex per kg waste, see Fig. 3. It can be noticed that the process consumes a large amount of renewable resources. This is because a green electricity mix is used. If these renewable resources are considered as freely available, the resource footprint of the recycling process is 0.64 MJex . Fig. 3 also shows that the footprint of company C is lower than the one of company B. This is mainly because less transport is needed, and the recycling machine of company C requires less electricity. The avoided impact is the resource footprint of the virgin production of 1 kg PE (V␣ = 80.95 MJex ) multiplied with the p-value (80%) and the recycling rate r (99.27%). The result is 64.29 MJex per kg waste, see Fig. 4. For the first alternative scenario, in which stream 1 goes to option III (company D), the resource footprint of the recycling process is 6.25 MJex per kg waste. Also here, the process consumes a large amount of renewable resources. If these renewable resources are again considered as freely available, the resource footprint of the recycling process is 1.71 MJex . The avoided impact is the resource footprint of the virgin production of 1 kg cast iron (Vˇ = 27.56 MJex ) multiplied with the m-value (66%) and the recycling rate r (80%). The result is 14.55 MJex per kg waste, see Fig. 4. For the second alternative scenario, in which stream 1 goes to option IV (company E), the resource footprint of the incineration process is 1.62 MJex per kg waste. This incineration generates 2.74 kWh of electricity. The avoided impact is the resource foot-

51

print for the virgin production of the same amount of electricity (E = 30.49 MJex ), see Fig. 4. 3.2. Circular economy performance First, we will determine the quality factor of both waste streams. Stream 1 consists of 100% pure PE. As there is no contamination, the quality factor is 1, and thus the most suitable waste treatment is option I (closed-loop recycling). Stream 2 consists of PE with a small amount of contamination, i.e. <5% EVOH. The interfacial tension between PE and EVOH is 7.89 mN/m, see Supporting Information S1. This interfacial tension corresponds with compatibility class 3, see Table 1. Next, we consider the graph for compatibility class 3, see Supporting Information S2. For a contamination of <5%, the quality factor is on average 0.87. This value corresponds with waste treatment option II (semi closed loop-recycling), see Table 2. In reality, stream 1 is subjected to waste treatment option II, when the end-product consists of 80% virgin material and 20% recycled material (p = 20), and waste treatment option I, when the end-product consists of 100% recycled material (p = 100). The latter depends on the demands of the client. Stream 2 goes also in reality to waste treatment option II: the end-product of company C consists of 20% virgin material and 80% recycled material (p = 80). The calculation of the indicators is given in Table 4. The results are also compared with the two alternative waste treatment scenarios for stream 1: open-loop recycling (option III) and incineration with energy recovery (option IV). The indicator results are presented in Fig. 5. Stream 2, which goes to waste treatment option II (semi closed-loop recycling by company B) achieves an overall highest circular economy performance, 98.3%. Stream 1 has a circular economy performance of 94.1% for waste treatment option I (closed-loop recycling by company B), 72.3% for waste treatment option II (semi closed-loop recycling by company B), 15.9% for waste treatment option III (open-loop recycling by company C) and 35.7% for waste treatment option IV (incineration by company E). This implies that if waste stream 1 undergoes the waste treatment option that corresponds with its quality factor, i.e. closed-loop recycling, the highest circular economy performance can be obtained. Further, in case of stream 2, the indicator demonstrates that option 2 leads to excellent results. If the waste is treated in an optimal way (impact of the recycling process R is low, recycling rate r is high) through a waste treatment option corresponding with its quality, this is close to the current best practice in the circular economy. 4. Conclusions In this study, we developed an indicator capable of measuring the circular economy performance of plastic waste treatments. This indicator can be a valuable addition to circular economy strategies, as disposal of plastic waste causes many environmental problems. The quality of the recycled material is used as starting point. Based on the compatibility between polymer blends, a quality factor is determined. Based on this quality factor, the most suitable waste treatment option can be selected: if the plastic is of high quality, the recycled material can substitute the virgin material entirely (closed-loop recycling, option I). If the recycled plastic is of lower quality, extra virgin material has to be added (semi closed-loop recycling, option II), or a different material is substituted (openloop recycling, option III). If the quality is too low, the waste can be incinerated (incineration, option IV). It must be remarked that the calculation of the quality factor is preliminary, as it is based on several assumptions: the waste

52

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

Fig. 3. Resource footprint of the treatment of 1 kg waste (R-value). For stream 1 to closed-loop and semi closed-loop recycling, a distinction is made between the recent recycling machine (left bar), the older machine (middle bar) and a combination of both (right bar). For stream 2 to semi closed-loop recycling and stream 1 to open-loop recycling, the results are presented inclusive (left bar) and exclusive abiotic renewable resources (right bar).

Fig. 4. Resource footprint of the avoided production per kg waste. For stream 1 (to closed-loop and semi closed-loop recycling), a distinction is made between films made of 100% recycled plastic, i.e. closed-loop recycling (left bar) and 20% recycled plastic, i.e. semi closed-loop recycling (right bar).

Table 4 Resource efficiencies for each waste treatment scenario. Stream 1

V˛ (MJex ) Vˇ (MJex ) E (MJex ) R (MJex ) I (MJex ) r (%) p (%) m (−) CPI (%)

Stream 2

Stream 1

Closed-loop (option I)

Semi closed-loop (option II)

Semi closed-loop (option II)

Open-loop (option III)

Incineration (option IV)

80.95 – – 4.42 – 99.56 100 – 94.1

80.95 – – 4.42 – 99.56 20 – 72.3

80.95 – – 0.64 – 99.27 80 – 98.3

– 27.56 – 1.71 – 80 – 0.66 15.9

–

consists of two polymers, no compatibilizers are added, no chain degradation occurs. If the influence of each of these parameters is investigated, the quality factor could be further improved, which is a challenge for future research. Also, the current classification based on interfacial tension, is only valid for thermoplastics, and not for thermosets or other material types. Therefore if a similar principle

30.49 – 1.62 – – – 35.7

is to be used for other material types, a quality factor based on other parameters should be developed. The circular economy performance indicator (CPI) can be calculated for each of these waste treatment options for thermoplastics. The CPI is defined as the ratio of the actual environmental benefit over the ideal environmental benefit according to quality. These

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

53

Fig. 5. Circular economy performance for each waste treatment options.

benefits are expressed in terms of resource footprints (CEENE), calculated through LCA. To illustrate the use of the indicator, it was applied on a case study about post-industrial plastic waste recycling. As most LCA case studies consider post-consumer plastic waste, this case study is also a relevant addition to existing literature. Company A, a producer of plastic films, generates polyethylene (PE) plastic waste. The purest PE-waste stream is transported to company B, also a producer of plastic films. Based on the quality factor, this stream should go to closed-loop recycling. In reality, the waste undergoes either closed-loop recycling (CPI = 94.1%) or semi closed-loop recycling (CPI = 72.3%), depending on the demands of the client. The PE-waste stream with a small amount of impurities is transported to company C, which is a producer of waste bags. Based on the quality factor, this stream should go to a semi closed-loop recycling treatment. In reality, the waste indeed undergoes semi closed-loop recycling (CPI = 98.3%). The results are also compared to two hypothetical scenarios: the PE-waste goes to company D for open-loop recycling into street benches (15.9%), or to company E for incineration with energy recovery (35.7%). This indicator and its results can be useful for policies aiming toward a circular economy. The plastic waste could be assigned to the most suitable waste treatment option according to quality, in order to obtain a higher environmental benefit in terms of resource consumption. Occasionally a ‘lower’ treatment option can result in a higher environmental benefit, meaning in most cases that the ‘higher’ recycling scenario might still need technical optimization or that this is a case of so-called upcycling. For now, legislation is based on the classical waste hierarchy, which puts recycling on top, followed energy recovery and landfill. This hierarchy makes no distinction between what is technically feasible, while our circular economy performance indicator does. Further, the compatibility between polymers could be taken into account when ‘designing for recycling’, as this is the most important parameter for the quality of recycled plastics. These considerations could for example be implement in environmental legislation on subsidies and taxes to steer the waste flows into the best suited treatment option. Notes and disclaimers The authors declare no competing financial interests. Acknowledgments S.H. was funded by the Flemish Policy Research Center for Sustainable Materials Management (Steunpunt SuMMa). The authors

also wish to thank Bruno De koker and Bart Verbeeck from Amcor, Eric Goddaert and Ruud Bosch from Indaver, Walter Willekens from Powerpack and Jarne Van Meerbergen from Sleco. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.resconrec.2017. 01.013. References Alvarenga, R.A.F., Dewulf, J., Van Langenhove, H., Huijbregts, M.A.J., 2013. Exergy-based accounting for land as a natural resource in life cycle assessment. Int. J. Life Cycle Assess. 18 (5), 939–947. Ardente, F., Cellura, M., 2011. Economic allocation in life cycle assessment: the state of the art and discussion of examples. J. Ind. Ecol. 16 (3), 387–398. Ardente, F., Mathieux, F., 2014. Identification and assessment of product’s measures to improve resource efficiency: the case-study of an energy using product. J. Clean. Prod. 83, 126–141. Bösch, M.E., Hellweg, S., Huijbregts, M.A.J., Frischknecht, R., 2007. Applying cumulative exergy demand (CExD) indicators to the ecoinvent database. Int. J. Life Cycle Assess. 12, 181–190. Bertin, S., Robin, J.J., 2002. Study and characterization of virgin and recycled LDPE/PP blends. Eur. Polym. J. 38 (11), 2255–2264. Bilitewski, B., Härdtle, G., Marek, K., 1997. Waste Management. Springer-Verlag, Berlin. Borovanska, I., Dobreva, T., Benavente, R., Djoumaliisky, S., Kotzev, G., 2012. Quality assessment of recycled and modified LDPE/PP blends. J. Elastom. Plast. 44 (6), 479–497. Brandalise, R.N., Zeni, M., Martins, J., Forte, M., 2009. Morphology, mechanical and dynamic mechanical properties of recycled high density polyethylene and poly(vinyl alcohol) blends. Polym. Bull. 62 (1), 33–43. De Soete, W., Dewulf, J., Cappuyns, P., Van der Vorst, G., Heirman, B., Aelterman, W., Schoeters, K., Van Langenhove, H., 2013. Exergetic sustainability assessment of batch versus continuous wet granulation based pharmaceutical tablet manufacturing: a cohesive analysis at three different levels. Green Chem. 15 (11), 3039–3048. Delva, L., Cardon, L., Ragaert, K., 2016. Evaluation of post-consumer mixed polyolefines and their injection moulded blends with virgin poly-ethylene. Accepted for publication in Environ. Eng. Manag. J. Dewulf, J., Bösch, M.E., De Meester, B., Van Der Vorst, G., Van Langenhove, H., Hellweg, S., Huijbregts, M.A.J., 2007. Cumulative exergy extraction from the natural environment (CEENE): a comprehensive life cycle impact assessment method for resource accounting. Environ. Sci. Technol. 41 (24), 8477–8483. EC (European Commission), 2013. Green Paper on a European Strategy on Plastic Waste in the Environment. EC (European Commission), Brussels. EC (European Commission), 2014. Towards a Circular Economy: A Zero Waste Programme for Europe. EC (European Commission), Brussels. EC (European commission), 2015. Closing the Loop – An EU Action Plan for the Circular Economy. Swiss Centre for Life Cycle Inventories, 2010, Brussels www.ecoinvent.org. EC-JRC (European Commission – Joint Research Centre). Integration of resource efficiency and waste management criteria in European product policies – Second phase, report n(3. Luxembourg: Publications Office of the European Union; 2012. Available from http://publications.jrc.ec.europa.eu/repository/ handle/111111111/31086.

54

S. Huysman et al. / Resources, Conservation and Recycling 120 (2017) 46–54

Falsafi, A., Mangipudi, S., Owen, M.J., 2007. Physical Properties of Polymers Handbook. Springer, New York. Fowkes, F.M., 1964. Attractive forces at interfaces. Ind. Eng. Chem. 56, 40–52. Ferreira, S., Cabral, M., da Cruz, N.F., Simões, P., Marques, R.C., 2014. Life cycle assessment of a packaging waste recycling system in Portugal. Waste Manag. 34 (9), 1725–1735. Government of Japan, 2010. Establishing a Sound Material-Cycle Society: Milestone Toward a Sound Material-Cycle Society Through Changes in Business and Life Styles. Ministry of the Environment, Japan. Haider, S., Masood, S., Sbarski, I., 2007. An investigation of mechanical, thermal and creep behaviour of recycled industrial polyolefins progress in rubber. Plast. Recycl. Technol. 23 (2), 97–110. Hopewell, J., Dvorak, R., Kosior, E., 2009. Plastics recycling: challenges and opportunities. Philos. Trans. R. Soc. B 364 (1526), 2115–2126. Hubo, S., Leite, L., Martins, C., Ragaert, K., 2015a. Evaluation of post-industrial and post-consumer polyolefin-based polymer waste streams for injection moulding. Proceedings of the 6th Polymers & Mould Innovations International Conference, 201–206. Hubo, S., Delva, L., Van Damme, N., Ragaert, K., 2015b. Blending of recycled mixed polyolefins with recycled polypropylene: effect on physical and mechanical properties. In: Proceedings of the Polymer Processing Society Conference, Graz, Austria. Huysman, S., Debaveye, S., Schaubroeck, T., Ardente, F., Mathieux, F., Dewulf, J., 2015. Application and further development of the recyclability benefit rate indicator for closed-loop and open-loop systems: a case study on plastic recycling in Flanders. Resour. Conserv. Recycl. 101, 53–60. Huysveld, S., Van Linden, V., De Meester, S., Peiren, N., Muylle, H., Lauwers, L., Dewulf, J., 2015. Resource use assessment of an agricultural system from a life cycle perspective – a dairy farm as case study. Agric. Syst. 135, 77–89. ISO (International Organization for Standardization), 2006a. ISO International Standard 14044: Environmental Management – Life Cycle Assessment – Principles and Framework. International Organization for Standardization, Geneva, Switzerland. ISO (International Organization for Standardization), 2006b. ISO International Standard 14044: Environmental Management – Life Cycle Assessment – Requirements and Guidelines. International Organization for Standardization, Geneva, Switzerland. Ignatyev, I.A., Thielemans, W., Vander Beke, B., 2014. Recycling of polymers: a review. ChemSusChem 7 (6), 1579–1593. Inoya, H., Leong, Y.W., Klinklai, W., Takai, Y., Hamada, H., 2012. Compatibilization of recycled poly(ethylene terephthalate) and polypropylene blends: effect of compatibilization on blend toughness, dispersion of minor phase, and thermal stability. J. Appl. Polym. Sci. 124 (6), 5260–5269. Koning, C., van Duin, W., Pagnoulle, C., Jérôme, R., 1998. Strategies for compatibilization of polymer blends. Prog. Polym. Sci. 23 (4), 707–757. Kreiger, M.A., Mulder, M.L., Glover, A.G., Pearce, J.M., 2014. Life cycle analysis of distributed recycling of, post-consumer high density polyethylene for 3-D printing filament. J. Clean. Prod. 70, 90–96. Kukaleva, N., Simon, G.P., Kosior, E., 2003. Binary and ternary blends of recycled high-density polyethylene containing polypropylenes. Polym. Eng. Sci. 43 (2), 431–443. Lei, Y., Wu, Q., Clemons, C.M., Gu, W., 2009. Phase structure and properties of poly(ethylene terephthalate)/high-density polyethylene based on recycled materials. J. Appl. Polym. Sci. 113 (3), 1710–1719.

Liao, W., Heijungs, R., Huppes, G., 2012. Thermodynamic resource indicators in LCA: a case study on the titania produced in Panzhihua city, southwest China. Int. J. Life Cycle Assess. 17 (8), 951–961. Manias, E., Utracki, L., 2014. Thermodynamics in polymer blends. In: Utracki, L.A. (Ed.), Polymer Blends Handbook. Springer Science, The Nederlands, pp. 171–289. Navarro, R., Lopez, J., Parres, F., Ferrandiz, S., 2012. Process behavior of compatible polymer blends. J. Appl. Polym. Sci. 124 (3), 2485–2493. Plastics Europe, 2015. Available from www.plasticseurope.org. Preston, F., 2012. A Global Redesign? Shaping the Circular Economy. Chatham House, London. Ragaert, K., Maeyaert, G., Martins, C., Cardon, L., 2014. Bulk compounding of PCL-PEO blends for 3D plotting of scaffolds for cardiovascular tissue engineering. J. Mater. Sci. Eng. 3 (316). Reynolds, N., Pharaoh, M., 2010. An introduction to composite recycling. In: Goodship, V. (Ed.), Management, Recycling and Reuse of Waste Composites. Woodhead Publishing Limited, Cambridge, pp. 3–19. Simon, B., Ben Amor, M., Földenyi, R., 2016. Life cycle impact assessment of beverage packaging systems: focus on the collection of post-consumer bottles. J. Clean. Prod. 112, 238–248. Swiss Centre for Life Cycle Inventories, 2016. Available from: Swiss Centre for Life Cycle Inventories. Takiguchi, H., Takemoto, H., 2008. Japanese 3R policies based on material flow analysis. J. Ind. Ecol. 12 (5–6), 792–798. Tall, S., Albertsson, A.C., Karlsson, S., 1998. Recycling of mixed plastic fractions: mechanical properties of multicomponent extruded polyolefin blends using response surface methodology. J. Appl. Polym. Sci. 70 (12), 2381–2390. Tonioli, S., Mazzi, A., Niero, M., Zuliani, F., Scipioni, A., 2013. Comparative LCA to evaluate how much recycling is environmentally favourable for food packaging. Resour. Conserv. Recycl. 77, 61–68. Valero, A., Valero, A., Dominguez, A., 2013. Exergy replacement cost of mineral resources. J. Environ. Account. Manag. 1 (2), 147–159. Van Bruggen, E., Koster, R., Picken, S., Ragaert, K., 2016. Influence of processing parameters and composition on the effective compatibilization of polypropylene – poly(ethylene terephthalate) blends. Int. Polym. Proc. 31 (2), 179–187. Van Krevelen, D., 2009. Interfacial energy properties. In: van Krevelen, D., Te Nijenhuis, K. (Eds.), Properties of Polymers. Elsevier, Amsterdam, pp. 229–244. Villalba, G., Segarra, M., Fernandez, A.I., Chimenos, J.M., Espiell, F., 2002. A proposal for quantifying the recyclability of materials. Resour. Conserv. Recycl. 37 (1), 39–53. World Economic Forum, 2016. The New Plastics Economy Rethinking the future of plastics. Retrieved from http://www3.weforum.org/docs/WEF The New Plastics Economy.pdf. Wu, S., 1971. Calculation of interfacial tension in polymer systems. J. Polym. Sci. Polym. Symp. 34 (1), 19–30. Yıldız-Geyhan, E., Yılan-Ciftci, G., Altun-Ciftcioglu, A.G., Kadırgan, M.A.N., 2016. Environmental analysis of different packaging waste collection systems for Istanbul – Turkey case study. Resour. Conserv. Recycl. 107, 27–37.