LANDFILL : A CASE STUDY ON SAMPLING, PROCESSING AND CHARACTERIZATION OF EXCAVATED FROM AN AUSTRIAN

*C. GARCIA LOPEZ, 1A. CLAUSEN, 1T. PRETZ

*Department of Processing and , RWTH Aachen University, Aachen, 52062, Germany

SUMMARY: The following case study belongs to the NEW-MINE project and the objective of the project is to develop a new scenario “Enhanced Landfill Mining” (ELFM) for a combined resource-recovery and remediation strategy. This strategy could reduce future remediation costs and reclaim valuable land while simultaneously unlocking valuable resources. In the past, insufficiently reliable data about the composition of , overestimation of the quality of excavated material and poor product marketing of the possible recyclables result nowadays in a bad reputation for landfill-mining projects. The ongoing research project of NEW-MINE wants to show that there are possibilities to create valuable outputs from landfills with enhanced treatment processes, like with a better distributed mechanical processing. The European Commission supports the idea of the project with funding the project. In order to create mechanical routes to recover valuable materials from old landfills, it is important to characterize the material, as a basis for the research. The aim of this case study, executed in November 2016 at the landfill site in Halbenrain, was to study the effect of certain mechanical treatment processes on the characteristics of old landfill material. The excavated waste was transported and supplied to an already configured Mechanical-Biological Treatment plant (MBT) located next to the landfill. During the mechanical processing, metals and high calorific fractions were sorted out with different aggregates. Samples of material in different flows of the plant (input and outputs) were sieved to obtain the particle size distribution. By manual sorting, the material was classified into plastics, wood, paper, textile, inerts, Fe metals, NF metals, glass/ceramic and residuals. In addition, the fines <40mm, from the sample materials, were sent to a laboratory to analyze the moisture content, ash content, calorific value and the concentration of heavy metals. As a result of this paper, the characterization of five material flows derived from the mechanical treatment is given together with the mass balance of the MBT. Although every landfill has its own characteristics, the results and experience obtained from this case study can help to understand general potentials, contribute to develop methodologies for characterization of old landfill material and identify problematic fields that require further research.

1. INTRODUCTION

As fast as the societies grow, also the generation of MSW is increasing in most of the cases and landfills continue to be filled with recyclables that could be used otherwise as raw materials or for energy recovery. The situation is even more critical if we look back in time. Before the Sardinia 2017 / Sixteenth International and Landfill Symposium / 2 - 6 October 2017

European Directive 1999/31/CE was implemented and defined the different categories of waste, municipalities could mix their waste and bury it without previous treatment/sorting. Besides, insufficient disposing charges for landfilling did not prevent the negative impacts of landfills either.

In the 1970s, there was a period of rapidly increasing raw material prices and rising concern about finite natural resources (Ferrati 1985). Many studies were forecasting serious shortages by the end of the century. Recycling of household waste was considered a partial solution of the problem. There was also rising awareness regarding the negative impact of simple dumping without appropriate barrier systems or pre-treatment prior to landfilling. Since then, many technological advances have been developed to produce refused-derived fuels (RDF) and to separate recyclable materials from residual . However, in spite of the efforts of the last 40 years to improve the situation of waste, landfilling is still the most common method of waste disposal in Europe, according to Eurostat.

In 2014, Europe treated 2.319 mill tons of waste by five-treatment operations, defined in the Waste Framework Directive 75/442/EEC: 41% disposal on land; 36% recovery (other than energy recovery- except backfilling); 10% recovery (backfilling); 7% land treatment/release into water; 5% energy recovery, 1% (Eurostat, 2014). Figure 1 shows the main waste management systems in Europe by countries.

Figure 1 Relevance of the main waste management systems in the EU-28 in 2012 (Eurostat, 2014).

Currently, Europe accumulates between 150.000 - 500.000 old landfill sites of which just 10% meet the EU requirements and are considered sanitary landfills (NEW- MINE, 2016). In most cases, non-sanitary landfills lack the required environmental protection technologies and will eventually require costly remediation to avoid future problems.

Due to the existence of those non-sanitary and sanitary landfills, valuable resources are being lost and concurrently the environment and human health damaged. Therefore, it is clear that remediation strategies for existing landfills are fundamental in the direction to preserve resources, environment and human health. A further argument to recover the valuable resources is the crisis that concerns the economic situation and the energetic matrix, which is mainly based on fossil fuels and water energy, where prices for energy and secondary resources are increasing steadily. As a technical solution and officially approved in March 2017, Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 the EU Landfill Directive is now to add a specific reference to Enhanced Landfill Mining (ELFM): “The Commission shall further examine the feasibility of proposing a regulatory framework for enhanced landfill mining to permit the retrieval of secondary raw materials that are present in existing landfills. By 31 December 2025 Member States shall map existing landfills and indicate their potential for enhanced landfill mining and share information” (European Parliament 2017).

The present work belongs to the NEW-MINE project, which is supported by the European Commission since 2016. The scope of the project is to transform a large fraction of old excavated landfilled material into higher-added-value products. The project is designed to combine a remediation strategy with the recovery of resources, as seen in Figure 2.

The composition inside a landfill generally depends on different parameters like waste legislation, differences in the waste management system, recycling systems, standard of living and on society and culture of the setting (Quaghebeur et. al, 2013). Proper investigations of each site including operation history, waste type dumped, dimensions of the landfill, topography and physical-chemical analysis are necessary in order to make a careful feasibility analysis about the material potential inside the landfill (Eugene Salerni, 1995). Apart from considering the material potential, a critical factor to take in consideration for ELFM is the market prices of the materials to recover, which vary over time.

Figure 2. Comparison of different scenarios for the EU’s landfills, Do-Nothing (only acceptable for well-monitored sanitary landfills), Classic Remediation (where the materials are excavated and re-landfilled), and Classic Landfill Mining coupled with (co)incineration, and the NEW-MINE, ELFM Scenario. Source: www.new-mine.eu. Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Since the beginning of the project, in September 2016, the Department of Processing and Recycling (I.A.R.) of RWTH Aachen University in Germany is part of the consortium. The enrollment of I.A.R. is to develop an enhanced mechanical processing that allows to evaluate the potential of landfills as a source of secondary raw materials.

2. MATERIAL AND METHODS

The sampling campaign presented in this paper was performed in order to obtain reliable information about the effect of mechanical treatment on the characterization of the landfilled waste. This chapter presents the methodology of characterizing different material flows (Figure 3) followed by the site description, the excavation process, the sampling campaign, the sieving classification of five material flows (SP1, SP2, SP3, SP4 and SP5) and the analysis of the fines (<10mm) from SP1 at the laboratory.

Figure 3. Scheme flowchart of the methodology for the characterization of landfilled material.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

2.1. Site description

The landfill site in Halbenrain belongs to FCC Halbenrain Centre and it is located 75 km southeast of Graz (Austria). The landfill was established in 1978 and received municipal solid waste and . Currently, the examined area is in the post-closure phase and it has an extension of 16 ha, with a total volume of 2.4 million m3 of waste. The site has developed over the years to a waste disposal facility with treatment, conversion of into electricity, composting, sorting and mechanical-biological waste treatment.

2.2. Excavation and processing at the site

In June 2016, FCC initiated a landfill-mining project on the site with the aim of recovering metals disposed between 1997 and 1999. Based on records about the landfill composition, eight areas of interest were estimated to contain a relatively high percentage of recyclable materials, being of great interest for the ELFM; method studies (see black circles, Figure 4). The material examined during the case study was taken out of the projected area marked in red, 20x20x10 m approx., in a depth of 6m.

Figure 4. Overview of landfill site including division in sectors; „A” (blue) was filled between 1979 and 1990; “B” (yellow) was filled after 1990; black circles mark especially interesting areas; projected area marked in red.

The mining activity included the following steps: 1) excavation (Figure 5), 2) transportation to a MBT plant, 3) biological treatment and 4) mechanical treatment with potential Refused Derived Fuels (RDF) separation and metal recovery. During the case study, two batches (batch 1: 220 tons, batch 2: 280 tons) were excavated, treated and characterized.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 5. Excavation of landfilled material (from late 1990s) at the landfill site in Halbenrain (Austria).

As it is known, MBT plants stand at the beginning of an efficient waste treatment process. By using a selective treatment process, unsorted waste can be separated into different fractions, which can undergo further treatment (e.g. potential RDF) or be used for material recovery (metals). The design of these processes must adapted to national regulations and market situations to not fail.

After the excavation, the landfilled material was sent to an already existing and configured MBT plant, which usually treats fresh household waste. As a first step of the MBT process, the material was dried by aerobic activity (rotting boxes for 3-4 weeks). During this treatment, a loss of water of approx. 10% and a possible organic matter reduction was achieved. Once the material was stabilized, it was sent to the full-scale mechanical process. Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 6. Flowchart of mechanical process at the MBT and sampling points (SP) and selection area of the characterized flows.

According to the flowchart, Figure 6, the semi-dry waste was fed to a single shaft shredder that reduced the particle size of the material down to 250mm. The shredder also brought the input material to a uniform size and eliminated overlengths that could interfere with successive processes. Afterwards, an overbelt magnet separated the large pieces of Fe metals. After the magnet, the first sampling point (SP1) was established.

Due to the plant design, the input material (prior the magnetic separator) was not accessible and the Fe metals sorted before the SP1 was calculated measuring the weight of the pile after processing the batch of excavated waste. After passing the magnetic separator, the material was screened by a flip-flop screen 60 mm (S1). The screen enabled the enrichment of the high calorific fraction -referred as potential RDF in the following- and metals in the coarses and also to concentrate the biogenic material in the underflow.

The material flows 250-60 mm and <60mm were further processed and characterized:

1) The overflow, 250 – 60 mm, was sorted with an additional overbelt magnetic separator (MS2) and screened at a diameter of 200 mm (S2). The fines from the screen 200mm Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

(fraction: 200 – 60 mm) went through a zigzag windsifter to separate the light from the heavy fraction of the flow. The coarse fraction, 250 - 200 mm, was directly balled.

2) The underflow, <60 mm, was also sorted with an overbelt magnet (MS3), and non- ferrous metals where separated by the first eddy current separator (ES1). After recovering the Fe and NF metals, another screen of 14 mm (S3) classified the material. The fines, <14 mm, went directly to the landfill and the coarse fraction was treated with a second windsifter (MS2). The heavy fraction, 60 - 14 mm, was sorted with a fourth magnetic separator (MS4) and a second eddy current separator (ES2) and Fe and NFe metals were sorted out, again, of the flow. The light fraction was mixed together with the coarse fraction of the sieve 200 mm and balled. The analyses of the underflow (<60 mm) are still in process and they will be published in future papers.

2.3. Sampling campaign

The selection of a method for representative sampling is one of the most difficult decisions of waste flow analysis due to the material’s heterogeneity. Normally, the material inside the landfill is variable depending on the digging point chosen for the excavation. In figure 6 is possible to see the sampling points chosen in this case study, referred as SPx.

The number (n) of the single samples in each SPx was based on the German directive LAGA PN 98 – procedures for physical, chemical and biological testing in connection with the recovery/disposal of waste. LAGA PN 98 defines that the number of samples depends on the total quantity (m3) of the material flow, (see Table 1). The mass of the sample depends on the maximum diameter, Dmax, of the particle size (Formula 1) according to LAGA PN 78;

SP2

SP1

Figure 7. Input material after a shredder and an overbelt magnet, SP1 (left) and fines from the first screen 60 mm, SP2 (right).

(1)

After taking several single samples in the same sampling point (example, figure 7), 3 or 4 of these single samples were mixed together to generate composite samples. During the sampling campaign, two batches of ~230 tons/each of biologically semi-dried old landfill material were characterized.

Table 1. Minimum number of individual/composite samples according to LAGA PN 98. Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Theoretically Sampling campaign Name Volume Individual Composite Sample Individual Composite Sample sampling samples samples weight samples samples weight point m3 n n kg n n kg SP1 400 32 8 15 18* 6* 21

SP2 30 8 2 15 8 2 12*

SP3 30 8 2 15 8 2 15

SP4 30 8 2 12 8 2 18

SP5 30 8 2 12 8 2 21 *For technical reasons the number and the amount of the samples were not possible to reach in accordance to the guideline.

2.4. Mass balance

The mass balance of the MBT plant in Halbenrain is calculated based on the batches that fed the MBT plant. The weight of both input and outputs were measured.

2.5. Characterization of landfilled waste

The outcome of the characterization helps to study the potential of landfills for raw materials. In addition, it provides data about the effect of MBT technology with old landfilled waste. It has to be noted that the facility used during this case study was not built for handling landfilled material. Therefore, the choice of the technology in the process was not perfectly optimized for the treatment of the landfill material and the results from the sorting should be interpreted with care.

The quantity and quality of the landfilled material after each mechanical process was determined by particle size distributions in each material flow, classifying each sample (table 2) and analyzing the fines chemically.

2.5.1. Bulk density

The bulk density of each material flow at SP1, SP2, SP3, SP4, SP5 was defined by filling a recipient of 90L directly on the flow and measuring its weight.

2.5.2. Particle size distribution analysis

Each composite sample from each sampling point was individually sieved for 60 seconds by a drum sieve. In order to find the fractions of the flow in which desired materials concentrate, an exhaustive particle size distribution study was made with the following mesh sizes: 200, 100, 80, 60, 40, 20, 10 mm.

2.5.3. Manual Sorting

The outcome of the drum sieve were eight fractions: >200 mm, 200 - 100mm, 100 - 80mm, 80 - 60mm, 60 – 40 mm, 40 – 20 mm, 20 – 10 mm, <10 mm. Subsequently, all fractions >10 mm were sorted manually into 10 categories, Table 2.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Table 2. Classification by categories. Category Material Wood All types of wood Paper Paper/cardboard/composite carton Textile All types of textile Plastic 2D Aluminum package/bags (transparent/white/colored) Plastic 3D PP/PET/PET Oil/PEAD/PEBD/PVC/PS/Others Fe metals Iron NF metals Copper/Aluminum can/steel Inerts Mineral fraction (stones) Glass Colorless glass/green glass/brown glass/others Residual Sanitary material, rubber, foam, silicone, melted plastics, sandpaper, electronic plates, hazards, undefined

2.6. Characterization at the laboratory: physical-chemical analysis

The fractions <40 mm (40 - 20 mm; 20 - 10 mm; <10 mm) were reduced in mass based on the German standard given in the previous chapter 2.3 and delivered to the RWTH Aachen University. There, further analysis were carried out in order to figure out the waste-to-energy characteristics (moisture content, ash content, calorific value and heavy metal concentration). To be able to analyze the last three parameters, the particle size of the samples had to be reduced down to <2 mm mechanically (see Figure 9).

Figure 9. Flowchart of the fine fraction characterization (<40 mm).

2.6.1. Moisture content

The water content is one of the most important parameters in the degradation of any waste, as it plays a vital role in the metabolism of microorganisms (Braumler and Kögel-Knabaner, 2008). The moisture content of waste is closely related to the amount of organic matter, and it differs with the habits of the population. In the EU and the USA it ranges from 20-30%, while values in China are between 30-60% due to the higher content of kitchen garbage (Jani, 2016).

Looking at existing landfills, different factors affect the moisture content, e.g. the the type, the composition and properties of the waste, the climatic conditions, the landfill operating system Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 and the soil cover layer (Hull et al., 2005). It is important to take into consideration the moisture content when considering the recycling of waste to produce energy through biological and/or (Brunner and Rechberger, 2015).

The standard DIN EN 14346:2007 “Characterization of waste - Calculation of dry matter by determination of dry residue or water content” suggests to dry the samples at 105 °C. However, volatile fractions would also evaporate at this temperature making plastic particles melt, thus resulting in a less precise analysis. The moisture content was determined by drying the samples of each material flow in a ventilated oven at 75 ºC until reaching a constant weight (Formula 2).

(2)

Where: MC: moisture content [%], mwet: weight wet [kg], mdry: weight dry [kg].

2.6.2. Particle size reduction

After the drying process, a reduction of the particle size of the three fractions (40 - 20 mm, 20 - 10 mm and <10 mm) was necessary to analyze the calorific value, the ash content and to determine the heavy metal content (with a XRF analyzer).

As shown in Figure 10, the sample was classified with the air sifter into a light fraction (LF) and a heavy fraction (HF). The metals contained in the sample were sorted using a magnet to avoid damages at further processing machines. The LF went to a sieve to separate remaining inerts in the fraction. In general, the LF had bigger particle sizes than the remaining inerts in the flow, and by sieving using a mesh size of 2 mm, the LF could be freed of the majority of the inerts. The HF was crushed with a hammer mill and afterwards grinded in a disk mill. In the case of LF with a particle size >2 mm, a cryogenic comminution method, using liquid nitrogen, had to be used to reduce flexible materials like plastics in size. The result of comminution was a powder with a size <1 mm for the HF and <2 mm for the LF.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 10. Flowchart of the pre-treatment to reduce the particle size of the sample down to <2mm, prior to the laboratory analysis.

The pretreatment to reduce the particle size provided relatively homogenic material fractions in comparison to the initial material. The weight of each output fraction (LF, HF/rest and metals) was measured (See Figure 9) to consider the relevance of each in % for further tests (ash content, calorific value and XRF analysis)

2.6.3. Ash content

After reducing the particle size of the samples from the fractions 40 - 20 mm, 20 – 10 mm and <10 mm down to <2 mm, 1 g of each sample was heated to calculate the ash content (AC) by using a muffle furnace (Formula 3), according to DIN EN 14775.

(3)

(4)

Where: AC: ash content [%], m1: initial sample weight [kg], m2: weight of ash + crucible [kg], mc: weight of crucible [kg], OC: organic content [%].

The volatile compounds calculated by the difference of the initial weight of the test and the remaining solids after the incineration is an indication of the organic matter content (Formula 4).

2.6.4. Net calorific value

The calorific value was determined for the same fractions than in the ash content analysis. In this case, a bomb calorimeter was used according to the standard DIN 51900. The test consisted of a complete combustion with oxygen of about 0.5 g of sample (dried weight) in a bomb a pressure of 40 bars. The heat transmitted to the surrounding water was measured and the net calorific value calculated accordingly.

2.6.5. XRF analysis

The content of heavy metals contained in the sample from the SP1, fractions 40 - 20 mm, 20 - 10 mm and <10 mm, was determined with a Niton™ XL3t, X-ray fluorescence (XRF) analyzer. Five tests were analyzed for average results and to validate the analysis.

3. RESULTS AND DISCUSSION

3.1. Mass balance of the MBT plant with landfilled material

In total, 2786 tons of excavated material were treated mechanically and ~3.4% of Fe metal concentrate was recovered. The rest of the output flows with possible recoverable materials were landfilled again. Figure 11 depicts the mass balance of the selected area of the MBT plant, during the sampling campaign. Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 11. Mass balance of the selected area at the MBT plant in Halbenrain.

3.2. Characterization of the material flows

3.2.1. Bulk density

The bulk density depends on both the density and arrangement (compaction) of the particles in the flow. For instance, the bulk density in the input material flow (SP1), with a particle size 250 – 0 mm, was 0,25 tons/m3 (see Table 3). Higher bulk density material was found in the fine fraction (<60 mm) of the screen 60 mm with 0.62 tons/m3. The coarses of the screen 60 mm went to a second screen 200 mm and the coarses (250 – 200 mm) had a very low bulk density in comparison to the fines (200 – 60 mm), 0.05 tons/m3 and 0.80 tons/m3, respectively. The fines from the second sieve (200 mm) had the highest density, 0.27tons/m3, due to its composition, which is examined in the next chapter 3.2.2. Composition.

Table 3. Bulk density of the material flows taken in different sampling points of the MBT plant Bulk density Sampling point tons/m3 SP1 0,25 SP2 0,62 SP3 0,05 SP4 0,18 SP5 0,27

3.2.2. Composition of the material flows

• Input flow of the mechanical treatment (SP1). The composition of the input flow is decisive for the rest of the flows in the plant, since this is the raw material that mechanical treatments try to sort by particle, classify by heavy and light fraction and pick the Fe/NF metals from it. Figure 12 shows the average composition (Ma.-%), where the largest proportion is the fine fraction (<10 mm) with about 50% of the total mass. The fines (<10 mm) are mainly soil, Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

glass shards and mineralized organic matter. In the next chapter 2.3., “Physical-chemical analysis”, there is a detailed characterization of this fraction from the input flow. The content of ferrous metals (Fe) in the input flow, as presented here, has to be considered with care, since it is not representative for the input of the mechanical treatment, but rather for the input after the first magnet separator (MS1).

Figure 12. Average composition (Ma. %) of the material supplied in the MBT plant during the sampling campaign after a shredder and a magnet separator, SP1.

In order to give an impression of the variability of the results, Figure 13 illustrates the fluctuation of 12 samples taken in the same sampling point (SP1).

Figure 13. Range of the share of different material groups in the input flow in both batches.

• Output flow of the screen 60 mm - fine fraction <60 mm (SP2). The fines from the output flow of the screen 60 mm is characterized by a high portion of fines (<10mm), which account for 71%. Impurities are also found, plastic particles (6%), glass (3%) and metals (1%). The content of glass in this material flow is the higher than in the rest of the examined flows.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 14. Average composition (Ma. %) of the fine fraction of the screen 60mm, SP2 and picture of the material flow (right).

• Output flow of the screen 200 mm - coarse fraction > 200 mm (SP3). The coarses of the screen 60 mm (250 – 60 mm) were sorted by a second magnet separator and sieved by a screen 200 mm. The composition of the coarses of the screen 200 mm (250 – 200 mm) are represented in Figure 15 (left). The amount of fines in the flow is reduced, but still there is a remaining portion, 7.1%, which indicate that screening efficiency was not 100%. In the next chapter, particle size distribution, the efficiency of the screen can be assess. This flow is characterized by its light fraction of 2D plastics (28.9%), followed by 3D plastics (21.2%), residuals (16.6%) and textiles (13,7%). This is an example for cleaning the material flow from impurities (mineral fraction) and concentrate the potential RDF (plastics, wood, textile, paper) by screening.

Figure 15. Average composition (Ma. %) of the coarse flow of the screen 200 mm, SP3 and picture of the material flow (right) .

• Output flow of the screen 200mm - fine fraction < 200 mm (SP4). The fine fraction from the screen 200 mm (200 – 60 mm) consists of a large share of inerts (20.2%) that could have been part of the covering layer of the landfill, back in the 1990s. MSW rarely has medium stones in its composition in that amount. Together with the inerts, 3D and 2D plastics have the biggest share in the composition of the flow. Once again, a significant amount of fines (<10 mm), 8.9%, were found.

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 16. Average composition (Ma. %) of the fine flow of the screen 200mm, SP4 and picture of the material flow (right).

• Output flow of the windsifter I - heavy fraction (HF) (SP5). The input material of the windsifter had a particle size of 60 – 200 mm. The major categories inside the heavy fraction flow are inerts (34.5%), wood (19.6%) and 3D-plastics (24.1%). The windsifter enriched the LF flow with 2D plastics, paper, textile and fines, which were almost missed in the heavy fraction flow of the windsifter. In addition, a big share of metals are found in the HF flow in comparison to the rest of the material flows. Hypothetically, a magnetic separator and an eddy current separator, after the windsifter in the heavy fraction, could have recovered 3,2% of Fe metals and 1.7% of NF metals. Instead, these valuable metals were landfilled again.

Figure 17. Average composition (Ma. %) of the heavy fraction of the windsifter 1, SP6 (left) and picture of the material flow (right)

3.2.3. Particle size distribution of the material flows

The results of the particle size analysis demonstrate the efficiency of the screens and the sorting aggregates by the grain sizes: <10 mm, 10 – 20 mm, 20 – 40 mm, 40 – 60 mm, 60 – 80 mm, 80 – 100 mm, 100 – 200 mm and 200 – 250 mm. The share of the main groups inside the landfilled waste is presented (Tables 4 – 12) by particles sizes in each material flow and Figures Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

18 – 22 gives an overview of the distribution of the particle sizes in each sampling point. The outcome of the SP2, SP3, SP4 and SP5 are discussed in the conclusion.

• Input material flow (SP1): The particle size distribution of the input material shows that the potential RDF (plastics, textiles and wood), with supposedly high calorific value, is dominant in the coarse fraction (>40mm), while the opposite is true for inerts, metals and glass (see Table 4). NF metals have a bigger share in the fraction 40 – 100 mm than in other particle sizes. Figure 18 provides information about the percentage (Ma.-%) of the particle sizes in the total mass and a line with the cumulative screening throughput.

Figure 18. Particle size distribution of the input flow (SP1).

Table 4. Comparison of the composition by particle sizes in the input material flow (SP1). Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 200-250 Fines <10mm 100,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% Potential RDF 0,0% 27,9% 42,6% 54,6% 60,0% 64,4% 62,8% 84,3% NF 0,0% 0,3% 0,7% 1,5% 1,1% 2,1% 0,3% 0,0% FE 0,0% 2,4% 3,7% 4,4% 1,3% 3,5% 1,5% 1,6% Inert 0,0% 22,2% 23,4% 23,0% 18,9% 8,2% 7,8% 0,0% Glass 0,0% 8,9% 5,8% 0,8% 1,1% 0,0% 0,0% 0,0% Residual 0,0% 38,2% 23,9% 15,7% 17,7% 21,8% 27,7% 14,1%

Table 5 shows the possible RDF materials (wood, paper, plastics, textile) contained in the different fractions. In this case, the coarses (>80 mm) contain a bigger share of plastics and textiles. Wood and paper have a greater share between 20 – 100 mm than in the rest of fractions.

Table 5. High calorific materials by particle sizes in the input material flow (SP2). Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 200-250 Wood 0,0% 9,1% 12,8% 13,2% 15,2% 10,1% 6,7% 0,3% Paper 0,0% 1,8% 3,7% 3,5% 3,0% 3,1% 1,4% 0,2% Textile 0,0% 0,3% 1,1% 2,3% 3,9% 6,8% 6,8% 30,1% Plastic 2D 0,0% 10,2% 13,1% 18,4% 20,6% 24,6% 31,9% 41,5% Plastic 3D 0,0% 6,5% 11,9% 17,1% 17,2% 19,8% 15,9% 12,2%

• Output flow of the screen 60 mm – fine fraction <60 mm (SP2): Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 19. Particle size distribution of the fine fraction (<60mm) of the screen 60 (SP2).

Table 6. Comparison of the composition by particle sizes in the fine fraction of the screen 60mm (SP2). Particle size (mm) <10 10-20 20-40 40-60 Fines <10mm 100,0% 0,0% 0,0% 0,0% RDF 0,0% 28,1% 37,1% 47,3% NF 0,0% 0,2% 1,1% 1,3% FE 0,0% 2,2% 1,9% 3,6% Inert 0,0% 37,6% 32,9% 30,5% Glass 0,0% 16,1% 11,0% 1,3% Residual 0,0% 15,8% 15,9% 16,1%

Table 7. High calorific materials by particle sizes in the fine fraction of the screen 60mm Particle size (mm) <10 10-20 20-40 40-60 Wood 0,0% 7,2% 9,8% 11,6% Paper 0,0% 4,5% 3,4% 3,8% Textile 0,0% 0,6% 1,2% 3,2% Plastic 2D 0,0% 6,5% 9,9% 14,6% Plastic 3D 0,0% 9,3% 12,8% 14,1%

• Output flow of the screen 200 mm - coarse fraction >200 mm (SP3)

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Figure 20. Particle size distribution of the coarse fraction (250 – 200 mm) of the screen 200 (SP3)

Table 8. Comparison of the composition by particle sizes in the output flow (>200mm) of the screen 200. Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 200-250 Fines <10mm 100,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% RDF 0,0% 75,5% 71,4% 69,5% 73,9% 71,1% 79,6% 81,4% NF 0,0% 0,0% 1,6% 1,2% 1,1% 1,8% 0,4% 0,0% FE 0,0% 0,0% 1,5% 0,6% 0,9% 0,1% 1,4% 3,4% Inert 0,0% 2,7% 3,1% 11,8% 0,6% 4,7% 0,0% 0,0% Glass 0,0% 2,9% 1,5% 0,0% 0,0% 0,0% 0,0% 0,0% Residual 0,0% 18,9% 21,0% 17,0% 23,5% 22,2% 18,6% 15,2%

Table 9. High calorific materials by particle sizes in the output flow (>200mm) of the screen 200. Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 200-250 Wood 0,0% 31,6% 15,5% 26,1% 17,3% 14,4% 10,5% 0,8% Paper 0,0% 9,1% 8,1% 3,7% 2,1% 1,2% 0,5% 0,0% Textile 0,0% 7,1% 6,2% 2,4% 5,3% 6,5% 8,7% 26,9% Plastic 2D 0,0% 16,8% 16,4% 18,5% 23,5% 23,4% 35,5% 32,9% Plastic 3D 0,0% 10,9% 25,1% 18,8% 25,6% 25,5% 24,4% 20,8%

• Output flow of the screen 200mm - fine fraction <200mm (SP4)

Figure 21. Particle size distribution of the fine fraction (<200 mm) of the screen 200 (SP4)

Table 10. Comparison of the composition by particle sizes in the output flow (<200 mm) of the screen 200. Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 Fines <10mm 100,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% RDF 0,0% 54,5% 64,2% 63,2% 39,2% 48,5% 64,5% NF 0,0% 1,9% 0,1% 1,7% 1,9% 0,2% 1,5% FE 0,0% 2,8% 2,6% 1,3% 0,2% 0,9% 5,7% Inert 0,0% 7,4% 5,0% 16,4% 39,4% 35,2% 12,8% Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Glass 0,0% 3,5% 1,4% 1,6% 0,3% 0,0% 0,0% Residual 0,0% 29,8% 26,8% 15,8% 18,9% 15,2% 15,5%

Table 11. High calorific materials by particle sizes in the output flow (<200mm) of the screen 200. Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 Wood 0,0% 11,7% 24,5% 25,2% 10,5% 15,4% 12,9% Paper 0,0% 20,8% 11,6% 6,0% 3,0% 2,7% 2,4% Textile 0,0% 1,7% 1,3% 1,4% 2,0% 1,3% 6,4% Plastic 2D 0,0% 11,2% 11,3% 12,2% 10,6% 15,0% 21,6% Plastic 3D 0,0% 9,2% 15,5% 18,4% 13,1% 14,1% 21,2%

• Output flow of the windsifter I - Heavy fraction (SP5)

Figure 22. Particle size distribution of the heavy fraction (HF) windsifter I (SP5)

Table 12. Comparison of the composition by particle in the heavy fraction flow (HF) of the windsifter I. Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 Fines <10mm 100,0% 0,0% 0,0% 0,0% 0,0% 0,0% 0,0% RDF 0,0% 35,1% 45,6% 58,5% 38,4% 41,7% 55,2% NF 0,0% 3,1% 0,0% 0,9% 1,8% 0,9% 2,8% FE 0,0% 10,2% 9,5% 4,2% 4,8% 2,8% 1,3% Inert 0,0% 19,0% 7,6% 28,4% 45,1% 41,5% 24,3% Glass 0,0% 4,7% 2,8% 1,4% 0,6% 0,0% 0,0% Residual 0,0% 27,8% 34,5% 6,6% 9,4% 13,0% 16,4%

Table 13. High calorific materials by particle sizes in the heavy fraction flow (HF) of the windsifter I Particle size (mm) <10 10-20 20-40 40-60 60-80 80-100 100-200 Wood 0,0% 8,2% 18,3% 28,7% 17,5% 20,0% 19,6% Paper 0,0% 5,2% 4,3% 0,7% 0,5% 1,0% 0,7% Textile 0,0% 0,7% 0,3% 0,0% 0,1% 0,7% 0,0% Plastic 2D 0,0% 4,6% 2,2% 1,5% 0,4% 1,7% 1,1% Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

Plastic 3D 0,0% 16,4% 20,5% 27,6% 19,9% 18,3% 33,8%

3.3. Physical-chemical analysis

The results of the physical-chemical analysis of the fines (<40 mm) input material flow are presented in Table 14 and Table 15. The fraction (<40 mm) represents almost 75% of the entire flow. The moisture content of <10 mm was higher than in the other two fractions. Each fraction (<10 mm, 10 – 20 mm and 20 – 40 mm) was divided in three components (LF, metals and rest) and the organic/inorganic content, as well as, the calorific value was calculated. The calorific value of these fractions varies from 7200 to 11996 Kj/kg, depending very closely on the organic content (<10 mm: 34,6%, 10 – 20 mm: 32,3%, 20 – 40 mm: 40,0%). These parameters can specify the energy potential of the fines (Jani, Y. 2016), in case the properties of the fines (<40 mm) are not suitable for contruction material due to its percentage of metals (0,1%), glass shards (0,2%) and plastics (0,7%).

Table 14. Organic content and calorific value of the input material, fractions <10mm; 10-20 mm: 20-40 mm. Input Moisture Composition Organic content Calorific value Fraction (%) (%) (%) (Kj/kg) (mm) LF Rest Metal LF Rest Metal* LF Rest Metal* Total <10 32 7 91 2 78 32 0 25982 6018 0 7200 10 – 20 17 18 77 5 81 23 0 25006 5194 0 8558 20 – 40 20 23 73 4 84 29 0 32560 6113 0 11996 *It is assumed that the amount of organic content of the metals are 0, therefore its calorific value is also 0 Kj/kg.

The metal content (%) of the fines (<40 mm) of the input flow are shown in Table 14. There are no relevant differences between the three fractions, only higher concentrations of Pb, Zn, Ni, Cr and Cu in the smaller particles sizes. Table 15 compares the concentration of heavy metals in the fine fraction (<10 mm) in the different sampling points.

Table 15. Heavy metal content in the input material, fractions <10mm; 10-20 mm: 20-40 mm. Metals Fraction % <10 mm 10-20 mm 20-40 mm Pb 0,094 0,071 0,07 As 0,002 0,006 0,001 Zn 0,497 0,318 0,257 Ni 0,026 0,017 0,019 Cr 0,193 0,142 0,095 Cd 0,001 0,002 0,002 Cu 0,404 0,246 0,254

4. CONCLUSIONS

One of the major findings of this case study is the share of fines (<40 mm) and coarses in the input material. The fines (<40 mm), as it can be seen in Table 16, account for 75% in the total mass of the input flow, while the coarses only for 26%. Initially, the composition of the waste is not very positive in terms of discovering big amounts of recyclables, but by biological and mechanical treatment, this complex flow can be clean from impurities (particles <10mm) and Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 possible desirable materials, e.g. potential RDF (pRDF), metals, soil, can be sorted from the flow for next processes.

Input material Categories Total flow Coarses Fines

250-0mm 250-40mm <40mm 100% 26% 74% Fines 49 0 67 Wood 5 9 4 Paper 1 2 1 Textile 3 9 0 Residuals 13 20 10 Plastic 2D 10 27 4 Plastic 3D 7 16 3 Fe 1 2 1 Metal Non Fe 3 5 3 Inerts 6 8 5 Glass 2 1 3

Going through the results of the mechanical processing, the coarses of the output of the screen 200 mm is an example of enrichment of the flow with pRDF, where they are concentrated for further valorization. In addition, it avoids that the volume of this material occupies a space inside the landfill again. The amount of pRDF in the particle size 100 – 200 mm of the input material is 7,3%, while in the coarses of the screen 200 mm, in the same particle sizes, is 28,5%. Another point of the sieving is to classify the material by sizes and achieve a higher efficiency in the next mechanical treatments. According to Pretz 2010, for an effective windsifter treatment, the ratio between the maximum and minimum particle size should not exceed 3:1. Therefore, the grain size fraction 60-200 mm was suitable for the sifting process. The windsifter helped to reduce the impurities contained in the fine fraction (200 – 60 mm) of the screen 200 mm. From a share of 8.9% of fines (<10mm) down to 1.3%. Furthermore, the windsifter concentrates the heavy fraction that commonly are 3D plastics, inerts and metals, which can be recovered. The composition of the HF material flow of the windsifter I contains 87% of coarses (200 - 60 mm) where 45% are pRDF and 37% are inerts, in the form of stones. It is also important to highlight the amount of residuals, between 200 – 60 mm, in the same material flow, which has a portion of 13%. From this analysis, it can be summarize that particle size distributions of the landfilled material can be used as a guidance for the mechanical treatment. Further chemical analysis are mandatory in order to find out if the quality of the excavated material fit the standards for its recuperation. One of the inconveniences of landfill mining is the discovery of hazards or contaminated soil, which reduces drastically the potential of the landfill as a source of resources. However, this inconvenience should not stop mining activities, since the environmental issue that landfills create cannot be ignored and remediation strategies are necessary to avoid future costs.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the financial support from the European Union’s with the Programme Horizon 2020. We are thankful for the help from the Chair of Waste Processing Technology and Waste Management of Montanuniversität Leoben (Austria) and the support given by FCC Halbenrain Abfall Service Gesellschaft m.b.H. & Co Nfg KG. A special gratitude Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 goes to Bastian Küppers for his time during the sampling campaing and his expertise in the field of waste management.

REFERENCES

Ferranti, M. P.; Ferrero, G. L. (1985): Sorting of household waste and thermal treatment of waste. London, New York, New York NY USA: Elsevier Applied Science Publishers; Sole distributor in USA and Canada Elsevier Applied Science Publishers.

Directive 75/442/EEC. Eurostat classification of treatments: http://ec.europa.eu/eurostat/web/waste/waste-generation-and-management/management

Eurostat (2017). Treatment of waste by waste category, hazardousness and waste operations. [ONLINE] Available at: http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=env_wastrt&lang=en [Accessed 4 June 2017].

Project NEW-MINE – EU Training Network for Through Enhanced Landfill Mining. 2016. Project. [ONLINE] Available at: http://new-mine.eu/ [Accessed 7 June 2017].

European Parliament (2017). Amendments adopted by the European Parliament on 14 March 2017 on the proposal for a directive of the European Parliament and of the Council amending Directive 1999/31/EC on the landfill of waste. [ONLINE] Available at: http://www.europarl.europa.eu/sides/getDoc.do?type=TA&language=EN&reference=P8-TA- 2017-0071 [Accessed 31 May 2017].

Quaghebeur, M., Laenen, B., Geysen, D., Nielsen, P., Pontikes, Y., Van Gerven, T., Spooren, J., 2013. Characterization of landilled materials: screening of the enhanced landill mining potential. Journal of , 55, 72-83.

Salerni, E.L. (1995) Landfill Reclamation Manual. Reclaim’95 Landfill Mining Conference, 28-29 September 1995, SWANA Landfill Reclamation Task Group, USA, p 16.

Bäumler R, Kögel-Knabaner I (2008) Spectroscopic and wet chemical characterization of solid waste organic matter of different age in landfill site, southeastern Germany. Journal of Environmental Quality 37: 146–153

Brunner PH and Rechberger H (2015) Waste to energy – key element for sustainable waste management. Waste Management 37: 3–12

Hull RM, Krogmann U, Asce M, et al. (2005) Composition and characteristics of excavated materials from New Jersey landfill. Journal Of Environmental Engineering 3: 478–490

Jani, Y., Kaczala, F., Charlotte, M., Hogland, M., Mait, K., et al. (2016). Characterisation of excavated fine fraction and waste composition from a Swedish landfill. Waste Management & Research. 34. 1292-1299.

Pretz, T., Julius, J., 2010. In: Kranert, M., Cord-Landwehr, K. (Eds.), Einführung in die Abfallwirtschaft. English: Introduction to Waste Management. Vieweg + Teubner Verlag, Wiesbaden, pp. 167–170.