Recycling of plastic-rich streams from waste electrical and electronic equipment (WEEE) sorting plants: An in-depth study of pyrolysis potential through product characterization and life cycle assessment (Lca)

Energy Conversion and Management 329 (2025) 119633 Available online 16 February 2025 0196-8904/© 2025 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies. Research Paper Recycling of plastic-rich streams from waste electrical and electronic equipment (WEEE) sorting plants: An in-depth study of pyrolysis potential through product characterization and life cycle assessment (LCA) Borja B. Perez-Martinez a,* , Alexander Lopez-Urionabarrenechea a , Adriana Serras-Malillos a , Esther Acha a , Miren Martínez-Santos a , Blanca M. Caballero a , Maider Iturrondobeitia b , Hugo Afonso b a Department of Chemical and Environmental Engineering, Bilbao School of Engineering, University of the Basque Country (UPV/EHU), Plaza Ingeniero Torres Quevedo, 1, 48013 Bilbao, Spain b Department of Graphical Design and Project Engineering, Bilbao School of Engineering, University of the Basque Country (UPV/EHU), Plaza Ingeniero Torres Quevedo, 1, 48013 Bilbao, Spain ARTICLE INFO Keywords: WEEE Pyrolysis Plastic waste Secondary raw materials Life cycle assessment Pyrolysis oil ABSTRACT Thermochemical recycling is emerging as a viable alternative for the recycling of complex plastic waste streams, such as those originated in waste electrical and electronic equipment (WEEE) sorting facilities. In this work, three different WEEE plastic-rich samples are subjected to pyrolysis and the resulting products are thoroughly analyzed to assess their potential industrial utilization. The pyrolysis processes are conducted in a 3 L non-stirred tank reactor at a heating rate of 15 ◦C/min, reaching a final temperature of 500 ◦C with a dwell time of 30 min, using 1 L/min of N 2 as carrier gas. The pyrolysis products are characterized and evaluated for their potential applications, including petrochemical feedstock, refuse derived fuel (RDF), and solid adsorbent. The results indicate that pyrolysis liquids can be used as RDF in cement kilns, provided the halogen content is below acceptance limits. The gases produced could be used as refinery gases after pollutant removal. Additionally, the solid fraction casts promising results in preliminary tests as a drug adsorbent in water, suggesting a new and very interesting path of research. Life cycle assessment (LCA) shows that the pyrolysis of sample C, which has the worst chemical properties, gives the lowest environmental impact, since the solid fraction from this sample is the most effective adsorbent, achieving almost 100 % removal efficiency for the tested drugs. The findings suggest that pyrolysis of plastic-rich streams should not always be focused on the oil production, as it can yield other valuable products. 1. Introduction The amount of electrical and electronic equipment (EEE) placed on the global market is increasing by almost 2.8 Mt per year [1]. The consumption of these materials is strongly influenced by the rapid development of some of the world’s most populous countries and the resulting increase in the gross domestic product (GDP), but also due to the throw-away philosophy surrounding these products in the developed countries [2]. With an average lifetime of 4 to 10 years [3] and the relatively low cost of some of the devices, which makes them easily replaceable, the amount of waste from electrical and electronic equipment (WEEE) generated rises to 62 Mt per year worldwide [1]. WEEE is one of the most complex waste streams to manage and recycle due to its material heterogeneity and the presence of hazardous substances. As a consequence, in the European Union, its collection, treatment and recycling is established by a specific Directive since 2012 (Directive 2012/19/EU) [4]. However, from the total WEEE generated in Europe in the year 2018, roughly the 50 % was collected for proper treatment, losing the other half in other waste flows as mixed residual waste or mixed metal scrap [5]. This collection rate is still far away from the 85 % that the European Directive established for the year 2019 [3]. The first step to recycle WEEE is the separation and sorting of its constitutive materials through a mechanical treatment that contains different steps such as depollution, sorting, size reduction, separation and concentration into useful final fractions [6]. Hazardous substances and components, for instance the refrigerants in “temperature exchange * Corresponding author. E-mail address: [email protected] (B.B. Perez-Martinez). Contents lists available at ScienceDirect Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman https://doi.org/10.1016/j.enconman.2025.119633 Received 2 December 2024; Received in revised form 4 February 2025; Accepted 9 February 2025 Energy Conversion and Management 329 (2025) 119633 2 equipment” or mercury containing batteries in “small equipment”, are separated beforehand during depollution step to avoid the release of pollutants to both the environment and the subsequent sorting process. Large metal and plastic parts, e.g. steel, aluminum and plastic cases, and frames of the appliances, are also recovered in this step. Afterwards, consecutive size reduction steps are carried out to progressively release the metallic materials from the complex components of the appliances and subsequently separate them by means of magnetic, eddy current and induction separators. The process at this point is mainly focused on recovering steel, aluminum and copper rich fractions, which make up 40 wt% of WEEE [7]. At the same time that these metals are separated, plastics are concentrated, generating plastic-rich streams that also contain small pieces of metals that are not easy to separate, such as those present in printed circuit boards (PCB) and wires, and other unsorted materials such as rubbers, glass or wood [8]. The main plastics in these streams are: (1) Styrene-based plastics, mainly acrylonitrile butadiene styrene (ABS), butadiene styrene (SB), high impact polystyrene (HIPS) or blends of ABS and polycarbonate (PC) from housings. (2) Thermoset resins (mainly epoxy/phenolic resins) from electronic boards. (3) Other thermoplastics such as polyvinyl chloride (PVC) from wire jackets, polybutylene terephthalate (PBT) and polypropylene (PP), among others [9]. The heterogeneity of these streams makes them unsuitable for direct mechanical recycling and further separation is also challenging due to the small particle size [10]. Therefore, they are usually incinerated or landfilled. Alternatives for recycling these plastic-rich complex streams must be studied following a circular economy approach, which is being promoted by the European Union in the new European Green Deal strategy, in order to recover secondary raw materials instead of the current disposal at landfill or energy recovery routes [11]. However, the recycling of such a complicated mixture of materials is a challenge in itself. In this sense, thermochemical recycling is an alternative to overcome the problem of heterogeneity and incompatibility of plastics in mechanical recycling. Thermochemical processes, such as gasification and pyrolysis, have proven their efficiency in the treatment of complex plastics streams, being capable of generating a variety of valuable products. Due to the intrinsic nature of the gasification process, it focuses on the production of valuable gases, mainly syngas (H 2 +CO), for the synthesis of chemicals. This process has already proven its efficiency converting waste plastic into gaseous streams in which syngas (up to 50 vol%) [12,13] or hydrogen (up to 67 vol%) [14] were the major compounds. Pyrolysis, on the other hand, is a more flexible thermochemical treatment technique, able to produce a wide range of products with many applications, normally divided into 3 phases (solid, liquid and gas), whose quantity and quality can be modified by adjusting the operating conditions. For instance, when considering pyrolysis of packaging plastic waste, the process is normally oriented towards the production of pyrolysis oils, the liquid fraction, which have demonstrated great potential for use in oil refining and petrochemical industry [15]. Conversely, when considering the pyrolysis of polymers with a more complex structure, e.g. thermosetting resins, gas production is the preferred option over the production of liquid hydrocarbons due to the high complexity of the latter [16]. In this respect, recent investigations carried out by the authors of this work on the decomposition of epoxy and polyester matrixes of composite materials exhibit the possibility of achieving gaseous streams rich in syngas (over 50 vol%) by pyrolysis [17,18]. In the case of plastic-rich streams generated in WEEE sorting plants, the pyrolysis process usually focuses on the pyrolysis oils. However, their industrial utilization is completely conditioned by the pollutants they may contain, mainly halogens from PVC and plastics containing brominated flame-retardants (BFR), and heavy metals. The concentration of both halogens and heavy metals is normally limited and under surveillance in industrial processes due to the severe consequences they can generate for processing, the environment and human health [19]. However, this kind of information is not normally reported in literature, generating the wrong belief that pyrolysis oils from WEEE plastic-rich streams could be directly used in industry. Therefore, it is necessary to conduct an exhaustive and rigorous study on the possible contamination of pyrolysis oils from WEEE plastic-rich streams and the influence this may have on their potential applications. The objective of this work is to study the pyrolysis of three different plastic-rich streams rejected at different separation steps of a WEEE sorting plant. The samples will be deeply analyzed and subjected to pyrolysis. Then, the products will be thoroughly characterized and their feasibility as alternative feedstock or product for the industry will be established. At last, the environmental performance of the pyrolysis of the three samples will be compared by means of life cycle assessment (LCA) methodology, based on the previously identified applications for each pyrolysis product. This assessment will help to estimate which is the rejected stream with the best possibilities to be treated by pyrolysis among these generated in common WEEE sorting plants. The entire work as a whole represents a significant novelty in the field, since it reports the nature of different plastic-rich streams generated in WEEE sorting plants and the possibilities that these streams have to be valorized through pyrolysis. The comparison of the chemical properties of the products with the limit values used in the industry is very rigorous, resulting in a realistic LCA analysis and a trustworthy snapshot of the real possibilities of these pyrolysis products. 2. Materials and methods 2.1. Preparation of WEEE samples Three plastic-rich samples (A, B, and C) from different process locations of a WEEE sorting, separation and classification plant were used for the experiments. The appearance of the samples can be seen in Fig. 1. The original A and B samples presented more than 93 wt% of particles bigger than 10 mm. On the contrary, more than half of sample C was composed of particles smaller than 1 mm in diameter, since it came from a milling stage downstream of the point where streams A and B were generated. The samples were left to dry under atmospheric conditions until the equilibrium moisture was reached. Subsequently, a quartering process was applied to get representative smaller samples of 100 g for both material composition measurement and pyrolysis tests. The original samples, except for sample C, were further milled below 2 mm for characterization analyses. These analyses included proximate analysis, ultimate analysis (C, H, N, S, O, Cl, Br and metals), higher and lower heating value (HHV and LHV) and thermogravimetric analysis (TGA). 2.2. Pyrolysis experiments Pyrolysis experiments were carried out in the laboratory scale plant shown in Fig. 2. It comprises, as main elements: an unstirred 3.0 L stainless steel tank reactor operating in semi-batch regime at atmospheric pressure; a water-cooled condensation system for gas–liquid separation and liquid collection; an activated carbon column for the removal of pollutants and condensable compounds still present in the gaseous stream; and a 25 L Tedlar plastic bag for the collection of gases for analysis. Pyrolysis experiments were designed in order to maximize oil production. When liquids are the target, it is usual to work at the lowest temperature at which complete degradation of the plastic residue is guaranteed. If that is not the case, lower temperatures (e.g. 400 ◦C) will produce less liquid due to lack of decomposition [20] and higher temperatures (e.g. 600 ◦C) will start to enhance cracking reactions to produce gas to the detriment of the liquid fraction [21]. As far as the other operating parameters are concerned, although they may be more dependent on the plant and reactor design, it can be said that liquid production is favored by high heating rates and short residence times in the reactor. These are, at least, the conclusions drawn by the authors after several years of experience working with complex plastic waste B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 3 [21–25]. Consequently, the operating conditions of the pyrolysis tests performed in this work were 100 g of sample, 15 ◦C/min heating rate, 500 ◦C final temperature, 30 min retention time. A pure nitrogen flow of 1.0 L/min is introduced from the bottom of the reactor and passes through a pierced distribution plate before contacting the sample and the rest of the reactor in its way to the outlet, flushing the vapors generated during the process. The temperature is controlled by a thermocouple placed in the center of the reactor, close to the sample. Pyrolysis liquid and solid yields were calculated by the weight difference in the pipelines, condensers and reactor before and after the pyrolysis, and were expressed as wt.%. The gas yield was calculated by difference to 100. 2.3. Analytical techniques 2.3.1. Waste samples characterization The determination of the composition of the waste samples was made as follows. A representative sample of 100 g was separated from the rest following the coning and quartering method. Then, each piece of these 100 g was taken by hand and the nature of the material was determined based on the author’s experience. In case of doubt (mainly with some plastics), the piece (only that piece) was introduced in a KUSTA 4004 M multiplexed NIR spectrometer to determine its nature. Once all the pieces were identified, they were separated into families and weighted. The thermogravimetric profiles of the plastic-rich samples were obtained on a Mettler-Toledo TGA/SDTA851 thermobalance by means of dynamic analysis from 30 to 900 ◦C under a heating rate of 10 ◦C/min in a 50 mL/min nitrogen atmosphere. Proximate analysis was performed on a LECO TGA-701 thermobalance following the EN-ISO 21660–3:2021 standard for moisture, the EN-ISO 22167:2022 standard for volatile matter and the EN-ISO 21656:2021 for ash determination. The higher heating value (HHV) determination was conducted in a LECO AC-500 automatic calorimeter following the EN-ISO 21645:2022 standard. Then, lower heating value (LHV) was calculated based on the HHV and the hydrogen content. For the determination of carbon, hydrogen, nitrogen, and sulfur content a LECO TruSpec CHNS automatic analyzer was used according to the steps dictated by the ENISO 21663:2021 standard, while oxygen was analyzed in a Eurovector Euro EA elemental analyzer. The chlorine and bromine contents were quantified following the UNE-EN 15408:2011 standard, which consists in the combustion of the sample in an automatic calorimeter, the same one used for the HHV determination. The produced gaseous halides were adsorbed in a basic solution for their later analysis by ion chromatography (Dionex ICS 3000). The content of metals was determined via XRF analysis on an ED-XRF Rigaku NEX QC +QuantEZ Analyzer by means of standard-less semi-quantitative Fundamental Parameters (FP) Fig. 1. Appearance of the as received WEEE samples. Fig. 2. Schematic representation of the laboratory scale pyrolysis plant. B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 4 method. 2.3.2. Oils characterization The elemental composition and HHV of the pyrolysis liquids were determined using the same techniques and procedures as those explained for raw samples. The main components in the oils were measured by gas chromatography (GC) AGILENT 6890 using a J&W DBFFAP capillary column (60 m ×0.320 mm ×0.25 µm), coupled to a mass spectroscopy detector (MS) AGILENT 5973, employing a chromatographic method developed by the authors and reported and discussed in a previous work. To this regard, it must be taken into account that GC–MS technique is able to “see” only the 90 wt% approximately of the total oil when analyzing this kind of pyrolysis oils [26]. Besides, the results in this article are given in peak area percentage, i.e., without calibration. Therefore, they cannot directly extrapolate into wt.% results, because this can lead to errors in the range 50–200 % for the substances typically found in these oils [26]. At last, those chemicals presenting a spectra quality match with NIST08 spectra library lower than 85 were classified as “not identified”. The oils derived from sample A were further analyzed to determine the presence of heavy metals (Ni, Pb, Cd, Cr, As, Cu, Co, Tl, Sb, Sn, Hg, Mn, Zn) and the content of polychlorinated biphenyls (PCBs 18, 20, 28, 31, 44, 52, 101, 105, 118, 138, 149, 153, 170, 180, 194, 209). Acid digestion followed by inductively coupled plasma atomic emission spectroscopy (ICP-AES) was used for metals determination. PCBs were determined by GC–MS, according to a method based on the EN 12766 standard. 2.3.3. Gases characterization The composition of the gases was measured by an Agilent 7890 series gas chromatograph (GC), coupled with a thermal conductivity detector (TCD) and a flame ionization detector (FID). Column setup was composed of a first column, Agilent HP-Plot Q (30 m x 320 µm x 20 µm), for the separation of the main hydrocarbons and a second column, Agilent HP-PLOT Molesieve (30 m x 530 µm x 25 µm), place in series for the separation of the permanent gases; the flow between the columns was controlled by an electronic valve. The quantitative characterization of the gases was performed by calibrating the main compounds usually found in this type of gaseous stream using a standard gas mixture. The composition of the standard was: CO 2 15 vol%, H 2 S 5 vol%, H 2 20 vol%, CO 20 vol%, CH 4 20 vol%, C 2 H 4 5 vol%, C 2 H 6 5 vol%, C 3 H 6 5 vol%, C 3 H 8 3 vol%, 1-Butene 1 vol%, and 1-Pentene 1 vol%. Both HHV and LHV of the produced gases were theoretically calculated based on the combustion enthalpy of each compound [27], by attributing to each compound percentage its own HHV or LHV. 2.3.4. Solids characterization and adsorption tests Ultimate and proximate analysis of pyrolysis solids were determined by the same means as the ones explained in section 2.3.1. for the original waste samples. The measurement of surface area and pore volume was carried out by adsorption isotherms of CO 2 at 0 ◦C in a Microtrac BESORP-max II BEL, which is the standard procedure for the measurement of coal and coal like surfaces when small and narrow micropores or ultramicropores are expected [28]. For the adsorption tests, six types of penicillin drugs (β-lactam antibiotics) were selected as main representatives of the most common drugs found in wastewater. Their description and the given codification are shown in Table 1. Stock solutions of the six penicillin types (AMOX, AMPI, CARB, CLOX, PEN-G and PEN-V) were diluted with high-performance liquid chromatography (HPLC) grade water (Scharlau, Spain) and mixed in a single solution containing 1200 µg/mL of each pharmaceutical compound. The adsorption experiments were carried out in triplicate and in batch mode in 50 mL polypropylene Falcon tubes, containing 0.5 g of pyrolysis solid and 10 mL of the abovementioned penicillin solution. This high adsorbent loading was selected under the assumption that the solids maybe were not good adsorbents. All Falcon tubes were covered with aluminum foil to avoid photodegradation. Then, all samples were stirred at 150 rpm on an orbital shaker under ambient laboratory conditions (~20 ◦C) during 1 h. The control sample was also spiked with the antibiotic solution, but without adding any pyrolysis solid. This control was included to investigate the natural degradation and possible contamination of antibiotics. After the experiments, the collected samples were filtered (with 0.22 µm polypropylene filters) and analyzed by HPLC-MS/MS chromatograph (Agilent 1290 Infinity II) coupled to a triple quadrupole mass detector (Agilent Technologies 6430 Triple Quad). The antibiotics removal efficiency was calculated using Eq. (1), where C 0 and C t are the concentrations of antibiotics (µg/mL) in the control sample and in each sample at the end of the experiment, respectively. Removalefficiency(%) = C0−Ct C0 (1) 2.4. Life cycle assessment (LCA) The LCA methodology was carried out following the ISO 14040 standards. The main goal of the LCA was to assess the environmental impacts of the pyrolysis process applied to each of the plastic-rich WEEE samples. The functional unit (FU) selected for the analysis, 1 t of WEEE plastics, was set based on the amount of different WEEE plastic fractions that a conventional sized WEEE treatment plant would produce per hour. The inventory of the LCA is shown in Table 2, which includes the inputs and outputs of the pyrolysis process in each case. Except for the treated sample, the material and energy inputs in the three cases were the same, as there was no variation in the pyrolysis process itself. These were: (1) nitrogen used as inert and carrier gas, (2) energy necessary to heat an industrial pyrolysis reactor using natural gas as fuel, (3) activated carbon used for gas cleaning and (4) water used for cooling the condensers. The quantities employed for the LCA analysis were directly measured from the laboratory experiments. The nitrogen input was considered liquid nitrogen to avoid the zero impact that the software attributes to the flow “nitrogen gas”. The selection of the output flows will be explained in “Results and discussion” section, since it was done taking into account the chemical properties of the products. Environmental impacts were calculated by using OpenLCA software coupled with the Ecoinvent v3.9 database. The ReCiPe 2016 Midpoint (H) method was used to categorize the results into eighteen environmental impact indicators [29]. The wide range of environmental impacts calculated with the selected method enables a deeper insight of the process analysis and makes easier the comparison with results from other research works available in the literature. 3. Results and discussion 3.1. Characterization of the original waste samples The material composition of the waste samples is presented in Table 3. Samples A and B showed a predominant styrenic nature (PS, HIPS and ABS), along with a high content of polyolefins (PP and PE) and other thermoplastics. This is well in accordance with the compositions reported in other studies developed with similar waste streams, where styrenic plastics were present in the greatest percentage and polyolefins Table 1 Description of the six different penicillin drugs used in the adsorption tests. Penicillin type Codification Purity (%) Origin Amoxicillin AMOX 99 Sigma-Aldrich Ampicillin AMPI 95 Sigma-Aldrich Carbenicillin CARB 99 Sigma-Aldrich Cloxacillin CLOX 99 Sigma-Aldrich Penicillin G PEN-G 99 USP Penicillin V PEN-V 89.9 USP B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 5 were also found in significant quantities [30,31]. The composition of C sample, on the contrary, presented a large amount of fines, which did not allow a complete determination of its composition. In any case, the materials that could be identified in C sample showed a much smaller amount of styrenic plastics, polyolefins and other thermoplastics, being PVC the most abundant material. As it can be seen, there are clear differences in composition (and also particle size) among the plastic rich fractions that are generated in a WEEE treatment plant. This is an issue that must be highlighted and taken into consideration when it comes to project a pyrolysis plant to treat “WEEE plastics” without further specification. It is common to think that those WEEE plastics will have a composition similar to that of sample A because this is the typical composition reported in literature. However, this could lead to undesirable surprises in the exploitation period of the plant. Table 4 shows the ultimate analysis, the proximate analysis and the HHV of the three waste samples. The results of the ultimate analysis were quite similar for A and B samples, being both rich in carbon. The main difference between them was the chlorine content, which was higher in sample B due to the higher quantity of PVC present in this Table 2 Inventory used to perform the LCA. Sample Input/ output Flow Amount Unit Avoided waste Provider A I WEEE plastics-A 1 t   nitrogen, liquid 0.75 t market for nitrogen, liquid | nitrogen, liquid | Cutoff, U −RER energy, industrial level natural gas 1965 MJ natural gas, burned in gas turbine | natural gas, burned in gas turbine | Cutoff, U activated carbon, granular 0.2 t activated carbon production, granular from hard coal | activated carbon, granular | Cutoff, U −RoW tap water 20 t tap water production, conventional treatment | tap water | Cutoff, U −RoW O biogas 92 m 3 TRUE anaerobic digestion of manure | biogas | Cutoff, U −RoW hazardous waste, for incineration 0.209 t treatment of hazardous waste, hazardous waste incineration, with energy recovery | hazardous waste, for incineration | Cutoff, U −RoW waste mineral oil 0.686 t clinker production | waste mineral oil | Cutoff, U −Europe without Switzerland B I WEEE plastics-B 1 t   nitrogen, liquid 0.75 t market for nitrogen, liquid | nitrogen, liquid | Cutoff, U −RER energy, industrial level natural gas 1965 MJ natural gas, burned in gas turbine | natural gas, burned in gas turbine | Cutoff, U activated carbon, granular 0.2 t activated carbon production, granular from hard coal | activated carbon, granular | Cutoff, U −RoW tap water 20 t tap water production, conventional treatment | tap water | Cutoff, U −RoW O refinery gas 0.131 t TRUE refinery gas production, petroleum refinery operation | refinery gas | Cutoff, U −RoW hazardous waste, for incineration 0,27 t treatment of hazardous waste, hazardous waste incineration, with energy recovery | hazardous waste, for incineration | Cutoff, U −RoW waste mineral oil 0.599 t treatment of waste mineral oil, hazardous waste incineration, with energy recovery | waste mineral oil | Cutoff, U −RoW C I WEEE plastics-C 1 t   nitrogen, liquid 0.75 t market for nitrogen, liquid | nitrogen, liquid | Cutoff, U −RER energy, industrial level natural gas 1965 MJ natural gas, burned in gas turbine | natural gas, burned in gas turbine | Cutoff, U activated carbon, granular 0.2 t activated carbon production, granular from hard coal | activated carbon, granular | Cutoff, U −RoW tap water 20 t tap water production, conventional treatment | tap water | Cutoff, U −RoW O activated silica 0,525 t TRUE activated silica production | activated silica | Cutoff, U −GLO waste mineral oil 0.36 t treatment of waste mineral oil, hazardous waste incineration, with energy recovery | waste mineral oil | Cutoff, U −RoW refinery gas 0.115 t TRUE refinery gas production, petroleum refinery operation | refinery gas | Cutoff, U −RoW Table 3 Material composition of the waste samples (wt.%, n =1). Polyolefins Styrenic plastics PVC PVC +metal Other thermoplastics Rubbers Foams PCB A 13.2 45.0 1.7 2.0 23.2 1.6 0.1 2.7 B 13.4 43.6 11.8 1.0 24.8 3.8 0.1 0.0 C 7.2 7.0 16.5 0.8 15.7 5.4 0.1 3.9 Textile Paper / Cardboard Magnetic metals Non-magnetic metals Petrous Multimaterial Fines (Ø <1 mm) Others A 0.3 0.1 0.0 3.2 0.1 1.5 2.5 2.9 B 0.0 0.2 0.0 0.3 0.0 0.4 0.0 0.5 C 0.0 0.3 0.0 0.6 0.3 1.0 36.0 5.2 Table 4 Ultimate and proximate analysis of samples and high heating value (HHV). Elemental organic analysis (wt.%, as received basis, mean value ± STD, n ¼3) C H N O S Cl Br (ppm) Others 1 A 65.4 ±2.4 6.8 ±0.3 1.9 ±0.2 8.1 ± 2.6 0.2 ±0.1 0.8 ±0.1 369 ± 37 4.9 B 63.3 ±2.8 7.5 ±0.3 2.6 ±0.4 4.8 ± 1.4 0.1 ±0.1 3.2 ±0.3 592 ± 59 4.9 C 43.4 ±3.2 6.2 ±0.4 1.4 ±0.2 16.0 ±2.1 0.4 ±0.1 2.6 ±0.3 211 ± 36 −- Proximate analysis (wt.%, as received basis, mean value ± STD, n ¼3) and HHV (MJ/kg, as received basis, mean value ± STD, n ¼3) Moisture Volatile Matter Ash Fixed Carbon 1 HHV (MJ/kg) A 0.4 ±0.1 83.3 ±1.5 11.9 ±1.4 4.5 ±0.4 30.2 ±0.3 B 0.9 ±0.1 79.8 ±0.5 13.5 ±0.5 5.9 ±0.3 29.1 ±0.6 C 0.8 ±0.1 53.0 ±5.4 40.8 ±5.7 5.4 ±0.4 15.0 ±1.0 1 by difference. B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 6 sample. When compared to similar waste streams studied in other works, the major elements (C, H and N) present in both A and B samples were within the reported ranges, while their oxygen concentration fell in intermediate values [32,33]. C sample showed a very different nature, presenting lower carbon content and a very high quantity of oxygen. This elemental composition is similar to that of printed circuit boards (PCB), where oxygen become more present due to the formulation of their constitutive resins, thus lowering the carbon plus hydrogen concentration [34,35]. Therefore, it could be thought that PCB concentration in this C sample was greater than that observed in the compositional analysis. In this sense, PCB could be hidden in the “fines” fraction due to their brittleness, which make them more prone to yield small particles when a crushing or reducing step is applied. With respect to the pollutants found in the three samples (S, Cl and Br), chlorine for both B and C samples appeared in a greater concentration, while its concentration in sample A was below those values. This is in accordance with the PVC content of sample A, which was significantly lower than in samples B and C. However, the highest chlorine content was reported for sample B, while the highest PVC content was found in sample C. This could be explained by the fact that the chlorine content of PVC plastics can be very different depending on the specific formulation of the PVC in question, since the proportion of additives in some commercial PVCs has been reported to be as high as 80 wt% [36]. Thus, it is possible that if sample C came from the sinking fraction of a float-sink process in the WEEE sorting plant, the PVC it contained would have a lower chlorine content, as it is known that the higher the amount of additives, the lower the flotation of PVC [37]. Bromine, on the other hand, appeared in concentrations two orders of magnitude lower than the values described in other works [9,32]. The reason for finding these small quantities of bromine in the samples is that the regulation on persistent organic pollutants (POP) implies that plastics with a total bromine concentration higher than 2 wt% must be separated in WEEE sorting plants and sent to elimination, because some brominated flame retardants (BFR) are classified as POP and the 2 wt% of total bromine statistically ensures that the concentration of these regulated BFR will be below the threshold concentration [38]. Consequently, the plastic-rich streams generated after this separation step have low bromine contents, although its presence can still be attributed to plastics with BFR in low concentrations. Finally, sulfur in the samples appeared at low values, casting a slightly greater concentration that the ones found in other reported works dealing with WEEE streams, and maybe related to some polymerization initiators such as sodium persulphate [32,33]. The proximate analysis and HHV are also shown in Table 4, where a high organic matter content (volatiles +fixed carbon) is reported for samples A and B, above 85 wt% in both cases. These samples also showed very low moisture content (bellow 1 wt%) and the typical ash content values for plastic-rich rejects of WEEE sorting plants, around the 10 wt% [32,39]. This is in agreement with the material composition, where a main polymeric nature was observed for both samples. The high content of volatile matter made these samples especially interesting for the production of pyrolysis oils and gases. C sample, otherwise, contained a significant quantity of ash (40.8 wt%), probably related to such a high content of fines, which caused the volatile matter content to decrease considerably. A and B samples showed acceptable HHV that would a priori allow them to be used directly as fuels, with values close to those of bituminous coals [40]. This was not the case for sample C, whose HHV was highly conditioned by its content of ash and oxygen, and consequently providing a HHV that was half that of the other two samples. Table 5 shows the content of metals of the three samples. In general terms, it can be noted that the metal content of sample C was higher than that of samples A and B. This result is in direct agreement with its higher ash content and indirectly with the large amount of fines in this sample, which at first sight appeared to be metallic powder. However, it is important to remember that this analysis was semi-quantitative, based on fundamental parameters and without the use of standards, so the values in the table do not have the accuracy of a quantitative analysis. The metals found in the highest proportion in all samples were aluminum, silicon, copper, calcium and titanium. The first three are major metals in the composition of EEE, so it is logical to some extent that they were detected mixed with the plastic materials of these devices. Calcium and titanium, on the other hand, are part of additives that are widely used in plastic formulations, such as calcium carbonate, calcium sulfate and titanium dioxide, and can be present in concentrations close to 50 % of the total weight of the plastic [41]. It is worth mentioning that the amount of calcium present in sample C, significantly higher than that of sample B, could corroborate the hypothesis that the PVC present in sample C was much more additivated than that of sample B, since calcium is a very common element among PVC additives [42]. Iron, also a major metal in the composition of EEE, was present in lower concentrations, corresponding to traces that can still be found mixed along the discarded streams despite having an effective separation step in WEEE sorting plants. Other metals typical of plastics additives such as phosphorus, antimony, barium, lead or zinc were also found, having their origin in the composition of flame-retardants, thermal stabilizers, pigments or lubricants [43]. Finally, it is worth noting the low concentration of mercury and the absence of cadmium in the samples, as a consequence of the progressive prohibition of the use of these two substances in household applications. The thermal decomposition pattern of the three samples, shown in Fig. 3, reveals two main degradation phases. The first one occurs between 250 ◦C and 350 ◦C, while the second spans from 350 ◦C to 510 ◦C. The materials commonly present in polymeric waste streams that can undergo decomposition steps around 250 ◦C are wood, polyurethane (PU) and PVC. In the case of wood, this is attributed to the decomposition of hemicellulose [44]; in the case of PVC, according to the literature, it corresponds mainly to the release of chlorine, derived from the cleavage of C-Cl bonds [45,46]; at last, PU itself decomposes at temperatures around 300 ◦C [47]. Additionally, some flame retardants usually found in polycarbonates (PC) such as ammonium polyphosphate can also thermally decompose at 300 ◦C [48]. In the case in question, in view of the composition of the samples in Table 3, it could be said that the main contributor to the first major thermal decomposition phenomenon is PVC. However, sample C showed a higher decomposition peak than sample B, when the latter contained more chlorine according to Table 4. The reason may be that sample C contained more additivated Table 5 XRF analysis of samples (ppm, as received basis, mean value ±STD, n =3). Element Sample A B C Al 7050 ±3521 6950 ±2369 15200 ±1374 Si 8250 ±2645 6700 ±801 52800 ±9354 P 1780 ±47 767 ±201 n.d. 1 K 344 ±0 n.d. 1 1720 ±482 Ca 6770 ±1398 7740 ±1601 25000 ±1739 Ti 3000 ±388 4140 ±281 3170 ±1646 Cr 124 ±113 19 ±8 106 ±68 Mn 18 ±2 22 ±2 135 ±22 Fe 666 ±23 781 ±33 2090 ±535 Co 80 ±58 n.d. 1 10 ±5 Ni 15 ±13 24 ±0 n.d. 1 Cu 3460 ±527 2090 ±894 5900 ±361 Zn 329 ±57 275 ±38 1040 ±55 Sr 56 ±24 39 ±5 495 ±205 Zr 6 ±2 6 ±2 151 ±53 Cd n.d. 1 n.d. 1 n.d. 1 Sn 141 ±53 87 ±25 137 ±18 Sb 1210 ±49 1650 ±251 920 ±258 Ba 463 ±136 320 ±75 3840 ±474 Hg 2 ±1 2 ±1 7 ±1 Pb 130 ±4 420 ±126 1260 ±199 Total metal concentration ≈34000 ≈32000 ≈114000 1 Not detected. B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 7 PC than sample B within the ‘other thermoplastics’ or it could even contain PU and/or wood within the ‘others’, the percentage of which is higher than that of sample B. Another explanation could be the lack of representativeness of the samples used for the analysis. This is a very typical problem when working with such heterogeneous granular waste, due to its fundamental variability that is enhanced in the determination of parameters that are not present in all the particles that form the waste, as is the case of halogens in WEEE plastics. This issue has been addressed by the authors in a recently published work [49]. The second degradation step corresponds to the release of hydrocarbons produced by the cracking of plastics via the breakdown of C–C and C–H bonds, a phenomenon widely reported in the thermal degradation studies concerning plastic-rich waste streams [50]. Other less relevant decomposition phenomena can be observed in the range of 650–850 ◦C, which could be related to the decomposition of carbonates and sulfates in the samples [51]. In any case, it can be seen in the figure that 500 ◦C is a temperature high enough to achieve the quantitative decomposition of the volatile matter in the three samples. 3.2. Pyrolysis yields The yields obtained in the pyrolysis of samples A, B and C are presented in Table 6. The data for the pyrolysis yields include the mean value of at least two equivalent experiments that did not differ more than 3 points in wt.% from one to another. In the case of the solid fraction, the metallic parts clearly visible to the naked eye were separated from the carbonaceous solid in samples A and B in order to analyze the char without interference from these metallic parts, which constituted the 26.6 and 10.8 wt% of the solid fraction of A and B respectively. This was not possible with sample C, due to its small particle size. Table 6 shows that the liquid product was the main fraction of samples A and B, ranging around 70 wt% and 60 wt%, respectively. The following product by weight was the solid fraction, being greater for B sample, due to its higher ash and fixed carbon contents. In any case, as it can also be seen in Table 6, the yields obtained with these samples are comparable to those obtained in different works carried out with similar waste streams and operating parameters, where the product generated to a larger extent was the liquid (60–70 wt%), followed by the solid fraction (20–30 wt%) and finally gases (10–15 wt%). C sample, however, was once again conditioned by the great presence of metals and other inorganic compounds in its composition, producing a solid fraction as its main product. Liquids were the second most abundant product obtained by this sample C, yielding 36 wt%, a value significantly low compared to the other two samples and the results from literature, which a priori discarded this sample as raw material for the generation of oils. 3.3. Liquid fraction characterization The composition of the pyrolysis oils, determined as percentage of GC–MS peak area, can be observed in Table 7. As it can be seen, the composition of the liquids changed significantly from samples A and B to sample C, as did the pyrolysis yields and the characterization of the samples themselves. The first two samples presented what can be considered a typical composition of a pyrolysis liquid coming from a styrenic-rich plastic stream, with a great presence of mono-aromatic hydrocarbons (MAH) such as styrene, ethylbenzene and toluene, together with a non-negligible concentration of phenol and its derivatives, related to the presence of oxygen in the samples. Such oil Fig. 3. Thermogravimetric behaviour of samples A, B and C by means of dynamic analysis. Table 6 Pyrolysis yields of the three samples: a comparison with literature. Own work Literature A B C [9] [52] [53] Waste Rejected WEEE fraction Rejected WEEE fraction Plastic from scrap computers Small WEEE Reactor Batch Semi-batch Batch Semibatch Operation temperature (◦C) 500 500 500 400 Solid yield (wt. %) 20.9 27.0 52.5 32.0 18.3 18.0 Liquid yield (wt.%) 68.6 59.9 36.0 61.0 63.9 74.0 Gas yield (wt.%) 10.5 13.1 11.5 7.0 17.8 8.0 B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 8 compositions have already been reported by studies treating similar waste streams under thermal pyrolysis conditions, where BTEX and styrene were the predominant compounds present [54–56]. It is noteworthy the low concentration of poly-aromatic hydrocarbons (PAH) found in these two oil samples. Finally, the presence of benzenebutanenitrile was probably derived from the nitrile groups present in plastics such as ABS [57]. The results shown in Table 7 revealed that the higher concentration of oxygen in sample C was transferred to the liquid fraction during the process in the form of oxygenated organic compounds, such as phenols, caprolactam, water and others. As stated before, the origin of the higher oxygen concentration could be attributed to a greater complexity of the polymers comprising sample C, as the ones present in the resins of PCB. This is in line with the composition of the liquid fraction derived from PCB pyrolysis observed in the literature, where the main compounds detected happened to be phenolic and alkyl-phenolic compounds [34,58]. The oxygen concentration also affected the HHV of sample C derived oils, which, in comparison to samples A and B, is more than 10 MJ/kg lower, as it can be seen in Table 7. The composition of the liquids conditioned the elemental analysis shown in Table 8. As the liquids derived from samples A and B were mostly composed of hydrocarbons, they showed a carbon plus hydrogen content near 85 wt%. These results are expected when processing this type of streams under thermal pyrolysis conditions, since WEEE plastics, in general, tend to degrade into oils composed mainly of carbon and hydrogen, in weight percentages above 85 % [54]. However, only a 31.8 wt% of carbon content was found in the sample C derived liquids and the total percentage of quantified elements was 64.2 wt%, a rather low value. When calculating the elemental balance of the process, an under quantification of oxygen was detected in the products for sample C, so it is most likely that the oxygen analysis of these liquids did not give correct results. The reason could be that this oil was an organic liquid sample with a high water content, which generated water–oil interface zones in the form of small droplets that may have conditioned the oxygen analysis, giving smaller values than real ones. The presence of phosphorous in these liquids, not determined in this work, and part of some flame retardants, could also have influenced this unbalanced result. The pollutants present in the original samples were transferred to the generated liquid fractions, as can be seen in Table 8, and had a clear impact worsening the quality of the oils derived from the three samples. The oil from sample A had almost no sulfur and a reduced concentration of halogens compared to the original sample. The oil from sample B, although maintaining low sulfur levels, contained significantly higher quantities of halogens, which was a direct consequence of the greater concentration of these in the original sample. Finally, the oil from sample C presented a very high level of contamination, with values of 1 wt% of sulfur, 10 wt% of chlorine and 2 wt% of bromine, probably enhanced by the low liquid yield. It must be mentioned that the halogens content and heating values reported in Table 8 for sample C are based on a single measurement due to the difficulty of this sample to be combusted in the calorimeter. 3.4. Solid fraction characterization and adsorption tests The solid fractions obtained in the pyrolysis of samples A, B and C, as can be seen in Table 9, were mainly composed of ash (52.1, 44.5 and 77.3 wt%, respectively), whilst fixed carbon resulted to be the second most abundant component in their thermogravimetric analysis. These results correlate well with the ash content reported by the proximate analysis performed on the original samples and were to be expected. The reason is that the inorganic content of the initial samples is normally concentrated in the pyrolysis solid, which is also composed of a carbonaceous product commonly referred to as char, directly derived from the carbonization of some plastics and other materials such as wood or paper that could be present in the processed samples. Previous studies working with similar WEEE streams also reported results where half of the pyrolysis solid was composed of ash, while the volatile matter and the fixed carbon content casted values close to 15 and 30 wt% respectively [25,59]. As far as the ultimate analysis is concerned, the main element detected in all the three samples was carbon, coming directly from the fixed carbon present in the solid fraction, as can be seen by comparing both the proximate and elemental analyses. Sample C, significantly conditioned by its high ash content, revealed a lower carbon concentration. The second most abundant element in all three samples was oxygen, which seems to be evenly distributed among the three pyrolysis fraction. Smaller amounts of hydrogen, nitrogen and pollutants such as sulfur, chlorine and bromine were also detected in a similar trend reported by the authors when working on the pyrolysis of Table 7 Liquids composition by GC–MS analysis (area %) and HHV (ni: not identified). Chemical families and substances / samples A B C Hydrocarbons Olefins 1-Hexene 0.1 0.3 2.0 Monocyclic aromatic hydrocarbons  Toluene 7.8 8.7 4.3 p-Xylene 2.1 2.6 0.0 Ethylbenzene 8.6 10.3 0.0 Propylbenzene 0.7 1.5 0.0 Styrene 42.4 36.1 10.6 α -Methylstyrene 4.3 3.1 0.0 Polycyclic aromatic hydrocarbons  1,3-Diphenylpropane 3.3 3.7 7.9 Non-hydrocarbons Phenol 11.0 10.6 18.6 Alkylphenol 6.9 8.4 15.8 Benzenebutanenitrile 3.3 5.1 7.9 Caprolactam 0.0 1.7 7.6 Water 1.1 1.6 14.6 n.i. 8.0 6.4 10.5 HHV (MJ/kg) 36.2 32.7 20.9 Table 8 Organic elemental characterization (wt.% ±STD, as generated basis, n =3), HHV and LHV (MJ/kg ±STD, as generated basis, n =3) of pyrolysis oils. C H N O S Cl Br HHV LHV A 80.0 ±2.6 7.8 ± 0.2 2.0 ± 0.1 4.2 ± 0.5 0.1 ± 0.0 0.5 ± 0.2 0.1 ± 0.0 36.2 ±0.5 34.7 ±0.5 B 78.0 ±2.8 8.5 ± 0.2 2.5 ± 0.3 4.2 ± 0.2 0.1 ± 0.0 2.8 ± 1.8 0.3 ± 0.1 32.7 ±0.6 31.2 ±0.6 C 31.8 ±2.5 8.5 ± 0.2 2.5 ± 0.2 8.3 ± 0.7 1.0 ± 0.5 9.9 2.2 20.9 19.4 Table 9 Pyrolysis solids characterization. Elemental analysis (wt.% ± STD, as generated basis, n ¼3) C H N O S Cl Br A 36.8 ± 2.4 1.5 ± 0.1 1.4 ± 0.2 6.1 ± 0.5 0.1 ± 0.1 1.1 ± 0.3 0.4 ± 0.1 B 36.1 ± 3.4 1.7 ± 0.1 1.2 ± 0.2 11.3 ± 1.7 0.1 ± 0.0 3.6 ± 0.2 0.6 ± 0.1 C 13.6 ± 1.8 0.6 ± 0.1 0.5 ± 0.1 11.8 ± 0.0 0.2 ± 0.0 0.9 ± 0.3 0.1 ± 0.0 Proximate analysis (wt.% ± STD; as generated basis, n ¼3), and HHV (MJ/kg ± STD, as generated basis, n ¼3) Moisture Volatile Matter Ash Fixed Carbon HHV A 3.3 ±0.0 9.5 ±0.3 52.1 ±0.8 35.1 ±0.5 14.3 ±0.5 B 2.9 ±0.0 10.6 ±0.3 44.5 ±0.2 41.9 ±0.3 13.2 ±0.1 C 1.4 ±0.1 8.3 ±0.2 77.3 ±1.2 13.0 ±1.0 5.0 ±0.5 B.B. Perez-Martinez et al. Energy Conversion and Management 329 (2025) 119633 9 phones recycling rejected streams [25]. As far as the HHV is concerned, it was closely related to the ash content of the pyrolysis solids, so that the higher the ash content, the lower the HHV of the solids. In this sense, the pyrolysis solids of samples A and B presented HHV values slightly lower than those found for biomass fuels [60,61], while the solid of sample C could not be used competitively as a fuel. The surface area of the solid fractions was measured as a first step to determine their potential application as adsorbents of chemicals in water; the results are presented in Fig. 4. Again, inorganic materials (mainly metals) that were of sufficient particle size to be detected by the naked eye were removed by hand to the possible extent before the surface area determination. The specific surface areas of the solids derived from A, B and C samples were 133.4, 89.1 and 36.7 m 2 /g respectively, of which about 92 % was microporous area, as shown in Fig. 4. These values are similar to the ones found in the literature for the activated pyrolysis chars derived from the pyrolysis of plastics [62,63]. The adsorption efficiency of the pyrolysis solids for antibiotics in water was measured as established in section 2.3.4., and the results are presented in Table 10. In this case, the pyrolysis solid derived from sample C was the one that casted the best results, being capable of adsorbing near 100 % of the drugs present in the water for each antibiotic family tested. The results were also interesting when it comes to the adsorption efficiency of sample B derived solid, with a removal efficiency between 64 % and 81 % depending on the antibiotic family tested. The pyrolysis solid from sample A, on the other hand, did not perform well in the test, achieving poor removal efficiencies that could not reach 50 % efficiency in any case. The well-performing results of the pyrolysis solid of sample C are, at a glance, quite surprising. On the one hand, this pyrolysis solid had the smallest surface area (36.7 m 2 /g) and, normally, better adsorption efficiencies are expected for greater surface areas. This could be explained by the fact that the adsorption capacity does not only depend on the surface area, since other factors such as electrostatic forces, pore size, hydrogen bonds, hydrophobic effects or chemisorption must be also taken into account [64,65]. On the other hand, this solid product was very far from what is considered an activated carbon (typical industrial adsorbent), mainly because of its high ash content, which was 77.3 wt% versus 8–10 wt% of commercial activated carbons, e.g., Filtrasorb. In fact, the nature of the pyrolysis solids from the three samples was predominantly inorganic, with some carbonaceous content. Therefore, the high adsorption efficiency of the pyrolysis solid of sample C should perhaps try to be explained by its metal oxide content. Several works can be found in the literature concerning the successful adsorption of antibiotics in aqueous phase using adsorbents based on silica, alumina, iron, calcium oxides and mixtures of them [66–69]. When carrying out the mass balance of the metals quantified in the original sample C (considering that they quantitatively remain in the solid fraction in the form of oxides), it can be proved that the pyrolysis solid of this sample is composed of approximately 20 wt% silica, 7 wt% calcium oxide and 5 wt % alumina, between other active metal oxides. The presence of these oxides could explain the good adsorption performance of the pyrolysis solids. In fact, its carbon content (13.6 wt%) could even enhance the adsorption capacity of this material, since carbon–metal oxide composite adsorbents have been reported in literature as very efficient in the antibiotics adsorption [70–72]. In summary, it seems like this pyrolysis solid could act as a kind of multifunctional composite adsorbent material, especially efficient for antibiotics removal in water. This discovery will be studied and analyzed more deeply by the authors in the near future. One of the issues that need deeper investigation is the possible cross contamination of the liquid phase with metals desorbed from the pyrolysis solids. Lixiviation experiments carried out by the authors showed that this phenomenon is possible, since concentrations of Ni (≤218 µg/L), Pb (≤470 µ g/L), Cu (≤201 µg/L) and Cr (≤7 µg/L) were measured in the eluate. However, this issue must be studied in the same conditions as those of the adsorption experiments. 3.5. Gas fraction characterization Table 11 shows the composition of the gas fraction derived from the pyrolysis of the three samples, as well as its HHV and LHV. The composition of the gases is given in nitrogen, oxygen, water and pollutants free basis. Concerning pollutants, these gases were cleaned by activated carbon adsorption in the pyrolysis process itself, as described in section 2.2. However, it is likely that the cleaned gases still contained halides and maybe other pollutants not detected by the GC-TCD/FID 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 A B C Pore volume (cm3/g) Surface area (m2/g) Surface area Micropore surface area Total volume in pores Fig. 4. Solid fraction surface characterization. Table 10 Antibiotic removal efficiency in water of pyrolysis solids produced from samples A, B and C (% ±STD, n =3). A B C PEN-G 17 ±10 64 ±31 98 ±2 PEN-V 28 ±14 73 ±28 99 ±1 AMOX 26 ±8 64 ±32 99 ±1 AMP 34 ±11 68 ±30 99 ±1 CARB 10 ±9 65 ±33 99 ±1 CLOX 48 ±24 81 ±23 100 ±0 Table 11 Pyrolysis gas composition and heating values. Gases composition Compound A B C vol.% wt.% vol.% wt.% vol.% wt.% CO 2 28.8 43.1 10.0 18.8 7.1 14.5 H 2 13.6 0.9 22.3 1.9 27.3 2.6 CO 15.7 15.0 7.7 9.2 11.6 15.1 C 2 H 4 5.2 4.9 10.1 12.1 7.6 9.9 C 2 H 6 6.7 6.9 12.0 15.4 8.3 11.6 CH 4 17.0 9.3 24.7 16.9 25.4 18.9 C 3 H 6 9.1 13.6 8.1 15.2 6.1 12.3 C 3 H 8 2.3 3.3 2.6 4.6 3.1 6.1 C4 1.5 3.0 2.4 5.7 3.1 8.1 C5 0.0 0.0 0.1 0.2 0.3 0.9 Gases heating value A B C MJ/kg MJ/Nm 3 MJ/kg MJ/Nm 3 MJ/kg MJ/Nm 3 HHV 23.7 27.3 39.5 36.4 39.9 33.8 LHV 22.0 25.4 36.5 33.7 36.8 31.2 B.B. Perez-Martinez et al.