implementation of biowindows for degasification of an
IMPLEMENTATION OF BIOWINDOWS FOR DEGASIFICATION OF AN OLDER MUNICIPAL SOLID WASTE LANDFILL AFTER REMOVAL OF THE ACTIVE GAS EXTRACTION SYSTEM M. HUBER-HUMER*, M. HRAD*, K. SCHLOFFER**, G. KAMMERER** * Institute of Waste Management, University of Natural Resources and Life Sciences, Muthgasse 107/III, 1190 Vienna, Austria ** Institute of Hydraulics and Rural Water Management, University of Natural Resources and Life Sciences, Vienna, Austria
SUMMARY: Engineered systems enhancing microbial methane oxidation, like biocovers, biowindows or biofilters, have been investigated and applied as a low-tech and low-cost measure to mitigate methane emissions from municipal solid waste landfills, in particular on older landfill sites with decreasing methane production. Important for the development of an optimal technical design of methane oxidation systems is the understanding of the fundamental processes, both the microbial process of oxidation as well as the behaviour of gas transport and emission fluxes. Based on this, an essential issue is the technical design of such systems, particularly the homogeneous distribution and supply of methane is still the most challenging factor for the optimal (longterm) performance of a methane oxidation system. On an older municipal solid waste landfill in Austria with an already strongly decreasing landfill gas generation the gas extraction has been shut-down and a pilot project to test passive degasification via methane oxidising biowindows has been started in 2014. For the pilot trial two former gas wells in an older section of the landfill were removed and replaced by two biowindows. In this paper the pre-investigations, the construction design, first experiences and lessons learned are presented.
1 INTRODUCTION Technical gas extraction and collection systems are regarded as important measures to reduce methane emissions from landfills, particularly in industrialised countries. Field studies have shown that >90% recovery rates can be achieved at well designed cells with final impermeable covers and an efficient gas extraction system (Spokas et al., 2006). However, some sites may have less efficiency or only partial gas extraction systems, and there are remaining fugitive emissions from disposed waste prior to and even after the implementation of active gas extraction; thus estimates of ‘life-time’ recovery efficiencies may be as low as 20% (Oonk and Boom, 1995) or on an average about 50% (Fischedick et al., 2014). Particularly at older sites, where the gas generation is decreasing and collecting and flaring the gas becomes inefficient and intricate from an operational point of view, other options have to be
Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
considered, like the technical support of the biological methane oxidation process. During the past decade engineered systems enhancing microbial methane oxidation, like biocovers, biowindows or biofilters, have been investigated and applied as a low-tech and lowcost measure to mitigate methane emissions from municipal solid waste landfills, in particular on older landfill sites, on landfills with mechanically-biologically pretreated waste or as a practical measure for developing countries during the past decades. Optimising microbial methane oxidation on landfills is also one key-technology listed in the 4th IPCC Assessment Report as a promising mitigation strategy to reduce greenhouse gas emissions from the waste sector (Bogner et al., 2007). However, depending on the technical design, the material selection, the operation, and climate impacts, methane oxidation rates achieved in such technically enhanced systems can vary over several orders of magnitude and range from negligible to 100% of the applied methane flux to the oxidising system. Important for the development of an optimal technical design of methane oxidation systems is the understanding of the fundamental processes (microbial process of oxidation as well as behaviour of gas transport and emission fluxes), the impacting environmental conditions, and the needs of the microorganisms responsible for the oxidation process. Environmental factors have a decisive impact on the activity of methanotrophic bacteria, and thereby control the oxidation performance; e.g., mainly temperature, moisture content, methane and oxygen supply, as well as availability of nutrients or inhibiting substances. Most of these conditions are affected again by material properties, like soil texture, porosity, water holding capacity, as well as content and quality of organic matter. The selection of suitable materials represents, therefore, one key-issue in constructing biotic methane oxidation systems. Another essential issue is the technical design of such systems, particularly the homogeneous distribution and supply of methane is still the most challenging factor for the optimal (longterm) performance of a methane oxidation system. Especially in the case of switching the gas wells (point sources) of an active gas extraction system into passively supplied methane oxidation windows (biowindows), the distribution and homogenuous supply of methane to each biowindow seems to be an immense challenge. In this paper the pre-investigations, the construction design and first experiences with pilotbiowindows, that replace former gas extraction wells on an Austrian landfill, are described.
2 MATERIALS AND METHODS 2.1 Landfill site and biowindow construction On a municipal solid waste landfill in Austria an active gas extraction system has been operated between 1994 and 2014 to collect and flare the generated landfill gas in order to minimise uncontrolled gas emissions. The total area of the landfill is about 10 ha in size and the total waste filling capacity is about 540,000 m³. The operation of the landfill started in 1978 with residual, pretreated municipal solid waste (residues after mechanical treatment and sorting). The site is divided in four sections (older closed sections (I- III) and one still operating section (IV)) and was adapted to the technical state of the art according to the Austrian landfill directive in the years 1992-1999. The gas collection system was subsequently installed within this timeframe in the older sections (I-III). About 20 vertical gas wells were subsequently placed by excavating the temporary cover and the landfilled waste down to a depth of 1 meter below landfill surface, filling the hole with gravel and installing vertical steel pipes (80 cm diameter). The gas wells were connected via horizontal slotted gas collection pipes made of HDPE. The pipes were placed in trenches filled with gravel directly beneath the temporary cover (soil-compost mixture)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
and lead to 8 gas collection/regulation stations, where they were connected to the main gas collection pipelines (one pipe optional for rich gas going to the flar, and one for weak gas which was supplied to a biofilter located beside the landfill). Only in the new, still operating section IV the gas collection system was implemented from the beginning of waste disposal. In this section 14 vertical gas wells were connected to the main gas collection pipes via a gas drainange system (slotted HDPE-pipes (DN 63 mm) embedded in a 50 cm gravel layer (32/64 mm)), which has been placed at the bottom of the waste fill. Two compressors, each with an extraction rate of 300 Nm³/h, were in operation between 1994 and 2014 (one for rich and one for weak gas lines). Due to the fact that mainly pretreated municipal waste has been disposed on this site, the gas formation is comparable low and the active gas extraction and flaring system did not work well during the past years. The landfill operator calculated that in 2013 about 19,700 kg CH4/a has been produced, that means allocated to the total area about 0.2 kg CH4/m²a (= 0.547 g CH4/m²d or 0.766 Nl CH4/m²d). Therefore, the active gas extraction has been shut-down in the year 2014. Screenings of methane concentrations using a Flame-Ionisation-Detector (FID) at the landfill surface after shutting down the gas extraction systems on the older sections (I-III), however, indicated emissions of landfill gas at some locations, mainly connected to the still existing vertical gas wells. Thus, the landfill operator decided to replace the gas wells by biowindows to minimise the methane escape into the atmosphere, and initiated a pilot project to test the passive degasification via methane oxidising biowindows in 2014. At a first step only two former gas wells in the older landfill sections (sections I and II) were removed and replaced by two biowindows. Each biowindow is about 8 x 8 m in size, consists of a gas distribution layer (0.5 m basalt gravel with a particle size between 30 and 60 mm) and an oxidation layer made of 1.4 m bio-compost mixed with wooden chips (70:30 vol%). The scheme of such a biowindow is shown in figure 1.
Figure 1. Scheme of a biowindow which replaces the gas extraction well (GDB = gas distribution layer)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Figure 2. Site plan of the landfill showing the location of the two pilot-biowindows (red squares) still connected via the former horizontal gas collection pipes 2.2 Pre-investigations To find the optimal locations for the two pilot-biowindows FID-screenings of the surface methane releases in sector I and II, and the gas composition at diverse gas wells were measured several times in April 2014. Moreover, sucking tests at selected gas wells took place at the same time. Based on these measurments and with respect to the daily landfill operation measures the two placements for the biowindows have been selected (see figure 2). In addition to the pre-checks in the field diverse materials and mixtures have been tested on their oxidation performance in lab-test. Counter-gradient column test (air supply at the top and landfill gas supply at the bottom) werde conducted at controlled temperature and moisture conditions in a climate chamber. Three different materials (mechanically-biologically pretreated municpial solid waste (mbpmaterial), sewage-sludge compost and bio-waste compost) have been tested in the column experiments. While the two compost materials have been sieved to a particle-size < 25 mm for the tests, the mbp-material was shreddered to a size < 20 mm. To achieve a suitable pore volume, wooden structure material has been added in a ratio of 70 : 30 vol% (compost : wood). In order to test the maximum oxidation capacity the tests were run for 110 days, divided into three phases with different methane supply (phase 1: 100 Nl CH4/m²d; phase 2: 200 Nl CH4/m²d and phase 3: 300 Nl CH4/m²d). All three mixtures showed 100% oxidation rate after an adaptation phase of 4 to 5 days in phase 1, but only the mbp-material and the bio-compost achieved a full (100%) oxidation in phase 2 (200 Nl CH4/m²d), and about 90% oxidation rate in phase 3 (300 Nl/m²d). The mbpmaterial exhibited a clear generation of exopolymeric substances at the end of the lab-scale experiment and performed a quite high respiration activity, indicating that the long-term performance and stability of this material was not guaranted. Thus, the biocompost performed best in the lab-scale tests and was choosen as substrate for the pilot-biowindows. Main characteristics of the compost material are shown in table 1. Measurements of respiration activity was done according to ÖN S2027-1. Parameters as TOC, Ntotal, loss on ignition, pHvalue und conductivity have been done according to ÖN S2023. Ammonium (photometric), nitrate and sulphate (using an ion chromatograph) were analysed in an eluat according to ÖN S2023.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Table 1. Some characteristics of the applied bio-compost mixture (RA: respiration activity; DM: dry matter)
Parameter pH conductivity NH4-N NO3-N SO4-S Phosphorus Organic matter content (loss on ignition) TOC Ntotal RA4 RA7 Water holding capacity
unit mS/cm mg/kg DM mg/kg DM mg/kg DM % DM % DM
Bio-compost with wood admixture (70:30vol%) 7.6 0.7 50 425 89 1.0 41
% DM % DM mg O2/g DM mg O2/g DM % DM
19 1.7 1.8 2.9 115
The maximum methane oxidation capacity of the bio-compost has been defined as about 275 Nl CH4/m² d. Based on these measurements and including a safety margin for sub-optimal conditions in the field, the specific methane load for the field was defined to be not higher than 5.7 l CH4/m²h (= approx. 136 l CH4/m²d in the field).
2.3 Monitoring measures After the pre-tests in the laboratory (material selection and characterisation), the check of the landfill site (mapping of gas composition and emissions) and the construction of the biowindows a continuous monitoring of the biowindows has been conducted between summer 2014 and the end of 2016. The monitoring program included quarterly emission flux measurements (using both a closed chamber and a dynamic open chamber method), FID-mappings, gas concentration and temperature profiling of the biowindows, and a yearly FID-check of the entire landfill surface. The FID-screenings were done using a Thermo Scientific FID (TVA-100; accuracy ±2.5 ppm in the range of 1.0 to 10,000 ppm) in a minimum grid of 1,5 x 1,5 m on each biowindow, visible disturbances (lack in vegetation etc.) on the surface were specifically measured. In each biowindow four measuring points for the gas profils were installed; each measuring point (nest) included 4 steel probes reaching into different depths (20, 50, 100 and 140 cm). The concentrations of oxygen, methane and carbon dioxide were directly measured at the probes with a portable gas analyzer (Multigas analyser LMSx). Temperature were measured in parallel in the probes using resistor-type thermometers (PT-100 sensors, class B; Fa. Testo). Moreover, in summer 2016 an additional sample taking campaign of the compost material from the biowindows (for e.g. water content, respiration activity according to the measurement methods defined above) as well as in-situ field tests (permeability, soil physical parameters, etc.) have been conducted. Due to difficulties with the air permeameter equipment, mainly soil hydraulic parameters describing water retention and storage have been determined and evaluated. On both pilotwindows spots with closed grass cover and “hotspots” were chosen and the biocompost layer was surveyed in three depths – at the surface, in 40 cm and in 80 cm – comprising following in-
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
situ experiments: single-ring infiltrometer (Reynolds, Elrick and Youngs, 2002), constant head well permeameter (Guelph permeameter according to Reynolds and Elrick, 2002), tension infiltrometer (Soil Measurement Systems, 2008). Undisturbed soil samples of 200 cm3 volume gained with the core method were analysed in the laboratory for dry bulk density and the only one status quantity, volumetric water content θ. For comparison θ had been measured in-situ with a two-rod sensor based on uncalibrated time domain reflectometry (Trase System, Soil Moisture Company). The underlying gas distribution layer was not investigated.
3 RESULTS AND DISCUSSION The previous data show that the landfill gas supply to the biowindows is very heterogeneous, showing a high temporal and spatial variability. During the first months of the pilot phase diverse technical adaptions have been implemented to optimise the gas distribution and to increase the oxidation performance of the windows, since already in the first weeks after the initial installation of the biowindows a very heterogeneous spatial gas distribution was detectable with the FID at the surface. Particularly in the edges, where the gas collection pipes entered the biowindows, hotspots occured and were visible due to a lack of vegetation (see figure 3). Moreover, due to the fact that the landfill operator wanted to built just two bio-windows (with a total biofiltration area of about 128 m²) in the pilot phase, and under the theoretical assumption (worst case) that all of the generated gas may penetrate through these two permeable windows (expecting a homogeneous gas distribution within the biowindows) and not through the more cohesive temporary cover material, a theoretical methane load of about 590 Nl CH4/m²d (at a filter area of 128 m²) could maximally occur, which would strongly exceed the maximal oxidation capacity of the filter material investigated in the lab-tests.
Figure 3. Hotspot occuring at the edge of the biowindow where the old gas collection pipe enters the gas distribution layer of the biowindow (gas probe nest M1 close to the visible hotspot)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
3.1 Gas concentration profiles and gas variability over time Gas concentration and temperature profils have been measured at least quarterly. In figure 4 a concentration profile from a probe nest located at the hotspot of biowindow 1 (left side) and one located on an area with adequate methane load (right side) are shown. The profiles on the right side show typical gas curves indicating methane oxidation (shift in CH4 to CO2 ratio, depletion (consumtion) of oxygen, methane no longer detectable in the upper part of the cover); while the profiles measured on the hotspot clearly indicate the overload with landfill gas (methane) at this spot. Oxygen does not penetrate deeply into the cover (affected by the high landfill gas flow from below and due to an intensive consumption in the upper 20 cm), there is no clear shift in the ratio of CH4 to CO2. The curves indicate dilution of the landfill gas in the upper part due to air infiltration, and only limited oxidation activity can be expected in the upper 20 cm. methane concentration (vol%)
depth (cm)
methane concentration (vol%) and temperature (°C)
depth (cm)
Figure 4. Exemplary gas concentration profiles measured on/close to a visible hotspot (left) and on an “undisturbed” spot (right) at biowindow 1
In addition to the strong spatial variability within the biowindows, also temporal variability in the methane load of the biowindows have been observed, indicated by the changing methane concentrations in the depth of 140 cm in the gas distribution layer made of coarse gravel (see figure 5 exemplary for biowindow 1). The methane concentrations measured within the landfill body (prior to the installation of the pilot windows and sporadically checked in still existing gas wells; data not shown) usually ranged between 15 and 35 %vol. The methane concentrations in 140 cm depth were quite similar in all probes at the first measurement dates in 2014, meaning the first weeks after biowindow installation (April (not shown) and June). Afterwards, the methane concentrations drifted apart. In probe nests M1 and M3 (closer to the inlet of the old gas collection pipe) the methane contents were higher compared to M2 and M4. The temporal variability was on the one hand obviously influenced by seasonal (climatic) changes (most probably impact on gas generation), and on the other hand due to operational measures, like the opening and closing of still existing gas wells and gates, which were all still connected to the main gas collection pipes, and thus consequently impacting the gas flow to the biowindows. Some specific test periods regarding the controlling of the gas flows (and loads to the
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
biowindows) were conducted during 2016. The whole corresponding data set and final results are still under evaluation, but the impact on the variability is indicated in figure 5 (monitoring period 2016).
Figure 5. Temporal variation in methane concentrations measured in 140 cm depth (in gas distribution layer) in the four different measuring points installed in biowindow 1 3.2 FID-screenings and spatial gas distribution The intensity of diffuse and preferential methane escapes at the surface of the biowindows were checked with the FID at least quaterly. The data indicate that particularly in the edges, where the old (slotted) gas collection pipes enter the biowindows, methane preferentially escaped. This was even visible with the naked eye (see figure 3). In figure 6 an exemplary FID-mapping of biowindow 1 is presented, showing the high spatial variability of the methane releases (methane concentrations up to 2,000 ppm at the visible hot spot and zero emissions in most of the remaining (grass covered) area of the biowindow.
Figure 6. Spatial variation in methane concentration at the surface of biowindow 1 indicating the visible hotspot
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Due to the convective gas escape the hotspot dried off more and more over time (less moisture content compared to the other material of methane oxidation layer were measured; data not presented). Thus, the gas permeability at the hotspot increased as well (lower water content => more air filled pores), which again lead to an enhanced gas flow and an “overdose” of the methanotrophic microorganisms.
3.3 Gas flux measurements Based on the results of the FID-measurements the location for the flux-measurements using two different types of chambers (open wind tunnel and closed dynamic chamber) was selected. The bars shown in figure 7 present the mean value of at least 3 measurements at the same spot and date (range shown as well). The orange bars indicate the measurements directly on a visible hotspot, the grey bars show emission rates measured on grass covered spots (with small and without pre-detected methane concentrations using the FID).
Figure 7. Temporal variation in methane fluxes (g/m²d) on hotspots (orange bars) and on placements in the middle (no visible hot-spot; grey bars) of biowindow 1 (G19) and 2 (G23)
The emission rates of the (non-hotspots) placements of the windows ranged between 0 – 11.0 g CH4/m²d (0 – 4.0 kg CH4/m²a) which is below the current limit value for gaseous emissions from temporary landfill covers according to the Austrian landfill directive (DVO 2008, appendix 3 lit. C (BGBl. II 39/2008)) which is set to 5 kg CH4/m²a. On the visible hot-spots the maximum emissions ranged between 14.9 up to extreme values of 562.4 g CH4/m²d (5.4 to 205 kg CH4/m²a). This means the current limit value for maximum hot-spot emissions on temporary covers for municipal solid waste landfills (containing a high amount of biodegradable waste) set in the Austrian landfill directive (< 10 kg CH4/m²a) was strongly exceeded on some measurement dates at the visible hot-spot.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
3.4 In-situ tests on water and gas permeability In additon to the monitoring of the gas flows and profiles, in summer 2016 field test have been performed in order to analyse soil physical parameters. The best information about soil properties governing vertical fluid movement in the soil matrix and in the macropores is given by ring infiltrometer methods. Field saturated hydraulic conductivity for tap water, corrected for 10 °C was kfs = 2.2×10−5 m·s−1 and 4.3×10−5 m·s−1 at the surface (measured on two dates), 5.1×10−6 m·s−1 at 40 cm depth and 1.3×10−5 m·s−1 at 80 cm depth at plot 1 (biowindow 1, grass covered spot without detected methane release at the surface). The values for plot 2 (biowindow 1, hotspot) at the same depths are 1.1×10−4 m·s−1, 2.0×10−7 m·s−1 and 4.0×10−5 m·s−1. These results reveal no clear distribution over profile depth or distinct properties related to one of the plots. The values were confirmed by similar results determined with guelph permeameter experiments, especially the small value at 40 cm depth on plot 1. This is not a matter of course considering a probable small-scale heterogeneity and anisotropy and different dependency on the latter by the two methods. Soil was moderately wet on both field days. Dry bulk density deduced from cylinder samples was very small and ranged for both plots from ρd = 0.4 g·cm−3 to 0.6 g·cm−3. Under not to wet conditions it can be stated that vertical gas convection will not be restricted reasonably in the biocompost layer. Close to water saturation at the bottom of the layer a barrier effect may be speculated, leading to a horizontal redistribution of the gas in the GDB, an upward breakthrough in few macropores or preferential flowpaths and reaching the surface, e.g. at the hot spots visible in figure 3. Validation would require knowledge of the water retention behaviour of both layers at the transition and either simulation or monitoring of soil water movement up to a full parameterization for a 3D two-phase flow simulation.
4 CONCLUSIONS As it is well-known from former studies and international literature the selection of suitable materials for methane oxidation systems and the technical design, particularly with respect to the homogeneous distribution and supply of methane, are the most important factors for a proper and effective performance of methane oxidation. Especially in the case of switching existing gas wells (point sources) of an (formerly operated) active gas extraction system into passively supplied methane oxidation windows (bio-windows), the distribution and homogenous supply of methane to each biowindow seems to be the biggest challenge - as indicated by ongoing investigations on an older Austrian municipal solid waste landfill with already strongly decreasing gas generation. So far the landfill owner has installed only two pilot-windows. It was expected from the very beginning, that the biofiltration area of two windows would be too small to catch and oxidise all of the still produced (calculated) methane, and this was confirmed during the first monitoring period. However, the pilot phase indicated first important information regarding the distribution of methane in the biowindows, and the behaviour of the landfill gas flows within the still existing gas collection pipes system. It is planned to remove further vertical gas wells and to install about 10 – 12 additional biowindows, with improved technical design and as far as possible controlled methane supply according to the lessons learned from the pilot phase. This project is currently still under evaluation by the local authorities, however, it is expected that it can be implemented in 2018 at the latest.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
REFERENCES Bogner, J., Abdelrafie Ahmed, M., Diaz, C., Faaij, A., Gao, Q., Hashimoto, S., Mareckova, K., Pipatti, R., Zhang, T. (2007). Waste Management. In: Metz, B., Davidson, O.R., Bosch, P.R., Dave, R., Meyer, L.A. (Eds.), Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA Fischedick, M., Roy, J., Abdel-Aziz, A., Acquaye, A., Allwood, J. M., Ceron, J.-P., Geng, Y., Kheshgi, H., Lanza, A., Perczyk, D., Price, L., Santalla, E., Sheinbaum, C., Tanaka, K. (2014). Industry. In: Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y. Sokona, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J. Savolainen, S. Schlömer, C. von Stechow, T. Zwickel and J.C. Minx (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.. Oonk, H., Boom, T. (1995). Landfill Gas Formation, Recovery and Emissions, TNO-report 95130, TNO, Apeldoorn, the Netherlands. Reynolds, W. D., Elrick, D. E. (2002). Constant Head Well Permeameter (Vadose Zone). In: Dane, J. H., Topp, C. G., (Co-Editors). (2002). Methods of Soil Analysis Part 4: Physical Methods. Madison, Wisconsin: Soil Science Society of America. 821-826 Reynolds, W. D., Elrick, D. E., Youngs, E. G. (2002). Single-Ring and Double- or ConcentricRing Infiltrometers. In: Dane, J. H., Topp, C. G., (Co-Editors). (2002). Methods of Soil Analysis Part 4: Physical Methods. Madison, Wisconsin: Soil Science Society of America. 821-826 Soil Measurement Systems. (2008). Tension Infiltrometer: Users Manual. Tucson, Arizona: Soil Measurement Systems, LLC Spokas, K., Bogner, J., Chanton, J., Morcet, M., Aran, C., Graff, C., Moreau-le-Golvan, Y., Bureau, N., Hebe, I. (2006). Methane mass balance at three landfill sites: what is the efficiency of capture by gas collection systems? Waste Management, 26, 516- 525.
SUMMARY: Engineered systems enhancing microbial methane oxidation, like biocovers, biowindows or biofilters, have been investigated and applied as a low-tech and low-cost measure to mitigate methane emissions from municipal solid waste landfills, in particular on older landfill sites with decreasing methane production. Important for the development of an optimal technical design of methane oxidation systems is the understanding of the fundamental processes, both the microbial process of oxidation as well as the behaviour of gas transport and emission fluxes. Based on this, an essential issue is the technical design of such systems, particularly the homogeneous distribution and supply of methane is still the most challenging factor for the optimal (longterm) performance of a methane oxidation system. On an older municipal solid waste landfill in Austria with an already strongly decreasing landfill gas generation the gas extraction has been shut-down and a pilot project to test passive degasification via methane oxidising biowindows has been started in 2014. For the pilot trial two former gas wells in an older section of the landfill were removed and replaced by two biowindows. In this paper the pre-investigations, the construction design, first experiences and lessons learned are presented.
1 INTRODUCTION Technical gas extraction and collection systems are regarded as important measures to reduce methane emissions from landfills, particularly in industrialised countries. Field studies have shown that >90% recovery rates can be achieved at well designed cells with final impermeable covers and an efficient gas extraction system (Spokas et al., 2006). However, some sites may have less efficiency or only partial gas extraction systems, and there are remaining fugitive emissions from disposed waste prior to and even after the implementation of active gas extraction; thus estimates of ‘life-time’ recovery efficiencies may be as low as 20% (Oonk and Boom, 1995) or on an average about 50% (Fischedick et al., 2014). Particularly at older sites, where the gas generation is decreasing and collecting and flaring the gas becomes inefficient and intricate from an operational point of view, other options have to be
Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
considered, like the technical support of the biological methane oxidation process. During the past decade engineered systems enhancing microbial methane oxidation, like biocovers, biowindows or biofilters, have been investigated and applied as a low-tech and lowcost measure to mitigate methane emissions from municipal solid waste landfills, in particular on older landfill sites, on landfills with mechanically-biologically pretreated waste or as a practical measure for developing countries during the past decades. Optimising microbial methane oxidation on landfills is also one key-technology listed in the 4th IPCC Assessment Report as a promising mitigation strategy to reduce greenhouse gas emissions from the waste sector (Bogner et al., 2007). However, depending on the technical design, the material selection, the operation, and climate impacts, methane oxidation rates achieved in such technically enhanced systems can vary over several orders of magnitude and range from negligible to 100% of the applied methane flux to the oxidising system. Important for the development of an optimal technical design of methane oxidation systems is the understanding of the fundamental processes (microbial process of oxidation as well as behaviour of gas transport and emission fluxes), the impacting environmental conditions, and the needs of the microorganisms responsible for the oxidation process. Environmental factors have a decisive impact on the activity of methanotrophic bacteria, and thereby control the oxidation performance; e.g., mainly temperature, moisture content, methane and oxygen supply, as well as availability of nutrients or inhibiting substances. Most of these conditions are affected again by material properties, like soil texture, porosity, water holding capacity, as well as content and quality of organic matter. The selection of suitable materials represents, therefore, one key-issue in constructing biotic methane oxidation systems. Another essential issue is the technical design of such systems, particularly the homogeneous distribution and supply of methane is still the most challenging factor for the optimal (longterm) performance of a methane oxidation system. Especially in the case of switching the gas wells (point sources) of an active gas extraction system into passively supplied methane oxidation windows (biowindows), the distribution and homogenuous supply of methane to each biowindow seems to be an immense challenge. In this paper the pre-investigations, the construction design and first experiences with pilotbiowindows, that replace former gas extraction wells on an Austrian landfill, are described.
2 MATERIALS AND METHODS 2.1 Landfill site and biowindow construction On a municipal solid waste landfill in Austria an active gas extraction system has been operated between 1994 and 2014 to collect and flare the generated landfill gas in order to minimise uncontrolled gas emissions. The total area of the landfill is about 10 ha in size and the total waste filling capacity is about 540,000 m³. The operation of the landfill started in 1978 with residual, pretreated municipal solid waste (residues after mechanical treatment and sorting). The site is divided in four sections (older closed sections (I- III) and one still operating section (IV)) and was adapted to the technical state of the art according to the Austrian landfill directive in the years 1992-1999. The gas collection system was subsequently installed within this timeframe in the older sections (I-III). About 20 vertical gas wells were subsequently placed by excavating the temporary cover and the landfilled waste down to a depth of 1 meter below landfill surface, filling the hole with gravel and installing vertical steel pipes (80 cm diameter). The gas wells were connected via horizontal slotted gas collection pipes made of HDPE. The pipes were placed in trenches filled with gravel directly beneath the temporary cover (soil-compost mixture)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
and lead to 8 gas collection/regulation stations, where they were connected to the main gas collection pipelines (one pipe optional for rich gas going to the flar, and one for weak gas which was supplied to a biofilter located beside the landfill). Only in the new, still operating section IV the gas collection system was implemented from the beginning of waste disposal. In this section 14 vertical gas wells were connected to the main gas collection pipes via a gas drainange system (slotted HDPE-pipes (DN 63 mm) embedded in a 50 cm gravel layer (32/64 mm)), which has been placed at the bottom of the waste fill. Two compressors, each with an extraction rate of 300 Nm³/h, were in operation between 1994 and 2014 (one for rich and one for weak gas lines). Due to the fact that mainly pretreated municipal waste has been disposed on this site, the gas formation is comparable low and the active gas extraction and flaring system did not work well during the past years. The landfill operator calculated that in 2013 about 19,700 kg CH4/a has been produced, that means allocated to the total area about 0.2 kg CH4/m²a (= 0.547 g CH4/m²d or 0.766 Nl CH4/m²d). Therefore, the active gas extraction has been shut-down in the year 2014. Screenings of methane concentrations using a Flame-Ionisation-Detector (FID) at the landfill surface after shutting down the gas extraction systems on the older sections (I-III), however, indicated emissions of landfill gas at some locations, mainly connected to the still existing vertical gas wells. Thus, the landfill operator decided to replace the gas wells by biowindows to minimise the methane escape into the atmosphere, and initiated a pilot project to test the passive degasification via methane oxidising biowindows in 2014. At a first step only two former gas wells in the older landfill sections (sections I and II) were removed and replaced by two biowindows. Each biowindow is about 8 x 8 m in size, consists of a gas distribution layer (0.5 m basalt gravel with a particle size between 30 and 60 mm) and an oxidation layer made of 1.4 m bio-compost mixed with wooden chips (70:30 vol%). The scheme of such a biowindow is shown in figure 1.
Figure 1. Scheme of a biowindow which replaces the gas extraction well (GDB = gas distribution layer)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Figure 2. Site plan of the landfill showing the location of the two pilot-biowindows (red squares) still connected via the former horizontal gas collection pipes 2.2 Pre-investigations To find the optimal locations for the two pilot-biowindows FID-screenings of the surface methane releases in sector I and II, and the gas composition at diverse gas wells were measured several times in April 2014. Moreover, sucking tests at selected gas wells took place at the same time. Based on these measurments and with respect to the daily landfill operation measures the two placements for the biowindows have been selected (see figure 2). In addition to the pre-checks in the field diverse materials and mixtures have been tested on their oxidation performance in lab-test. Counter-gradient column test (air supply at the top and landfill gas supply at the bottom) werde conducted at controlled temperature and moisture conditions in a climate chamber. Three different materials (mechanically-biologically pretreated municpial solid waste (mbpmaterial), sewage-sludge compost and bio-waste compost) have been tested in the column experiments. While the two compost materials have been sieved to a particle-size < 25 mm for the tests, the mbp-material was shreddered to a size < 20 mm. To achieve a suitable pore volume, wooden structure material has been added in a ratio of 70 : 30 vol% (compost : wood). In order to test the maximum oxidation capacity the tests were run for 110 days, divided into three phases with different methane supply (phase 1: 100 Nl CH4/m²d; phase 2: 200 Nl CH4/m²d and phase 3: 300 Nl CH4/m²d). All three mixtures showed 100% oxidation rate after an adaptation phase of 4 to 5 days in phase 1, but only the mbp-material and the bio-compost achieved a full (100%) oxidation in phase 2 (200 Nl CH4/m²d), and about 90% oxidation rate in phase 3 (300 Nl/m²d). The mbpmaterial exhibited a clear generation of exopolymeric substances at the end of the lab-scale experiment and performed a quite high respiration activity, indicating that the long-term performance and stability of this material was not guaranted. Thus, the biocompost performed best in the lab-scale tests and was choosen as substrate for the pilot-biowindows. Main characteristics of the compost material are shown in table 1. Measurements of respiration activity was done according to ÖN S2027-1. Parameters as TOC, Ntotal, loss on ignition, pHvalue und conductivity have been done according to ÖN S2023. Ammonium (photometric), nitrate and sulphate (using an ion chromatograph) were analysed in an eluat according to ÖN S2023.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Table 1. Some characteristics of the applied bio-compost mixture (RA: respiration activity; DM: dry matter)
Parameter pH conductivity NH4-N NO3-N SO4-S Phosphorus Organic matter content (loss on ignition) TOC Ntotal RA4 RA7 Water holding capacity
unit mS/cm mg/kg DM mg/kg DM mg/kg DM % DM % DM
Bio-compost with wood admixture (70:30vol%) 7.6 0.7 50 425 89 1.0 41
% DM % DM mg O2/g DM mg O2/g DM % DM
19 1.7 1.8 2.9 115
The maximum methane oxidation capacity of the bio-compost has been defined as about 275 Nl CH4/m² d. Based on these measurements and including a safety margin for sub-optimal conditions in the field, the specific methane load for the field was defined to be not higher than 5.7 l CH4/m²h (= approx. 136 l CH4/m²d in the field).
2.3 Monitoring measures After the pre-tests in the laboratory (material selection and characterisation), the check of the landfill site (mapping of gas composition and emissions) and the construction of the biowindows a continuous monitoring of the biowindows has been conducted between summer 2014 and the end of 2016. The monitoring program included quarterly emission flux measurements (using both a closed chamber and a dynamic open chamber method), FID-mappings, gas concentration and temperature profiling of the biowindows, and a yearly FID-check of the entire landfill surface. The FID-screenings were done using a Thermo Scientific FID (TVA-100; accuracy ±2.5 ppm in the range of 1.0 to 10,000 ppm) in a minimum grid of 1,5 x 1,5 m on each biowindow, visible disturbances (lack in vegetation etc.) on the surface were specifically measured. In each biowindow four measuring points for the gas profils were installed; each measuring point (nest) included 4 steel probes reaching into different depths (20, 50, 100 and 140 cm). The concentrations of oxygen, methane and carbon dioxide were directly measured at the probes with a portable gas analyzer (Multigas analyser LMSx). Temperature were measured in parallel in the probes using resistor-type thermometers (PT-100 sensors, class B; Fa. Testo). Moreover, in summer 2016 an additional sample taking campaign of the compost material from the biowindows (for e.g. water content, respiration activity according to the measurement methods defined above) as well as in-situ field tests (permeability, soil physical parameters, etc.) have been conducted. Due to difficulties with the air permeameter equipment, mainly soil hydraulic parameters describing water retention and storage have been determined and evaluated. On both pilotwindows spots with closed grass cover and “hotspots” were chosen and the biocompost layer was surveyed in three depths – at the surface, in 40 cm and in 80 cm – comprising following in-
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
situ experiments: single-ring infiltrometer (Reynolds, Elrick and Youngs, 2002), constant head well permeameter (Guelph permeameter according to Reynolds and Elrick, 2002), tension infiltrometer (Soil Measurement Systems, 2008). Undisturbed soil samples of 200 cm3 volume gained with the core method were analysed in the laboratory for dry bulk density and the only one status quantity, volumetric water content θ. For comparison θ had been measured in-situ with a two-rod sensor based on uncalibrated time domain reflectometry (Trase System, Soil Moisture Company). The underlying gas distribution layer was not investigated.
3 RESULTS AND DISCUSSION The previous data show that the landfill gas supply to the biowindows is very heterogeneous, showing a high temporal and spatial variability. During the first months of the pilot phase diverse technical adaptions have been implemented to optimise the gas distribution and to increase the oxidation performance of the windows, since already in the first weeks after the initial installation of the biowindows a very heterogeneous spatial gas distribution was detectable with the FID at the surface. Particularly in the edges, where the gas collection pipes entered the biowindows, hotspots occured and were visible due to a lack of vegetation (see figure 3). Moreover, due to the fact that the landfill operator wanted to built just two bio-windows (with a total biofiltration area of about 128 m²) in the pilot phase, and under the theoretical assumption (worst case) that all of the generated gas may penetrate through these two permeable windows (expecting a homogeneous gas distribution within the biowindows) and not through the more cohesive temporary cover material, a theoretical methane load of about 590 Nl CH4/m²d (at a filter area of 128 m²) could maximally occur, which would strongly exceed the maximal oxidation capacity of the filter material investigated in the lab-tests.
Figure 3. Hotspot occuring at the edge of the biowindow where the old gas collection pipe enters the gas distribution layer of the biowindow (gas probe nest M1 close to the visible hotspot)
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
3.1 Gas concentration profiles and gas variability over time Gas concentration and temperature profils have been measured at least quarterly. In figure 4 a concentration profile from a probe nest located at the hotspot of biowindow 1 (left side) and one located on an area with adequate methane load (right side) are shown. The profiles on the right side show typical gas curves indicating methane oxidation (shift in CH4 to CO2 ratio, depletion (consumtion) of oxygen, methane no longer detectable in the upper part of the cover); while the profiles measured on the hotspot clearly indicate the overload with landfill gas (methane) at this spot. Oxygen does not penetrate deeply into the cover (affected by the high landfill gas flow from below and due to an intensive consumption in the upper 20 cm), there is no clear shift in the ratio of CH4 to CO2. The curves indicate dilution of the landfill gas in the upper part due to air infiltration, and only limited oxidation activity can be expected in the upper 20 cm. methane concentration (vol%)
depth (cm)
methane concentration (vol%) and temperature (°C)
depth (cm)
Figure 4. Exemplary gas concentration profiles measured on/close to a visible hotspot (left) and on an “undisturbed” spot (right) at biowindow 1
In addition to the strong spatial variability within the biowindows, also temporal variability in the methane load of the biowindows have been observed, indicated by the changing methane concentrations in the depth of 140 cm in the gas distribution layer made of coarse gravel (see figure 5 exemplary for biowindow 1). The methane concentrations measured within the landfill body (prior to the installation of the pilot windows and sporadically checked in still existing gas wells; data not shown) usually ranged between 15 and 35 %vol. The methane concentrations in 140 cm depth were quite similar in all probes at the first measurement dates in 2014, meaning the first weeks after biowindow installation (April (not shown) and June). Afterwards, the methane concentrations drifted apart. In probe nests M1 and M3 (closer to the inlet of the old gas collection pipe) the methane contents were higher compared to M2 and M4. The temporal variability was on the one hand obviously influenced by seasonal (climatic) changes (most probably impact on gas generation), and on the other hand due to operational measures, like the opening and closing of still existing gas wells and gates, which were all still connected to the main gas collection pipes, and thus consequently impacting the gas flow to the biowindows. Some specific test periods regarding the controlling of the gas flows (and loads to the
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
biowindows) were conducted during 2016. The whole corresponding data set and final results are still under evaluation, but the impact on the variability is indicated in figure 5 (monitoring period 2016).
Figure 5. Temporal variation in methane concentrations measured in 140 cm depth (in gas distribution layer) in the four different measuring points installed in biowindow 1 3.2 FID-screenings and spatial gas distribution The intensity of diffuse and preferential methane escapes at the surface of the biowindows were checked with the FID at least quaterly. The data indicate that particularly in the edges, where the old (slotted) gas collection pipes enter the biowindows, methane preferentially escaped. This was even visible with the naked eye (see figure 3). In figure 6 an exemplary FID-mapping of biowindow 1 is presented, showing the high spatial variability of the methane releases (methane concentrations up to 2,000 ppm at the visible hot spot and zero emissions in most of the remaining (grass covered) area of the biowindow.
Figure 6. Spatial variation in methane concentration at the surface of biowindow 1 indicating the visible hotspot
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
Due to the convective gas escape the hotspot dried off more and more over time (less moisture content compared to the other material of methane oxidation layer were measured; data not presented). Thus, the gas permeability at the hotspot increased as well (lower water content => more air filled pores), which again lead to an enhanced gas flow and an “overdose” of the methanotrophic microorganisms.
3.3 Gas flux measurements Based on the results of the FID-measurements the location for the flux-measurements using two different types of chambers (open wind tunnel and closed dynamic chamber) was selected. The bars shown in figure 7 present the mean value of at least 3 measurements at the same spot and date (range shown as well). The orange bars indicate the measurements directly on a visible hotspot, the grey bars show emission rates measured on grass covered spots (with small and without pre-detected methane concentrations using the FID).
Figure 7. Temporal variation in methane fluxes (g/m²d) on hotspots (orange bars) and on placements in the middle (no visible hot-spot; grey bars) of biowindow 1 (G19) and 2 (G23)
The emission rates of the (non-hotspots) placements of the windows ranged between 0 – 11.0 g CH4/m²d (0 – 4.0 kg CH4/m²a) which is below the current limit value for gaseous emissions from temporary landfill covers according to the Austrian landfill directive (DVO 2008, appendix 3 lit. C (BGBl. II 39/2008)) which is set to 5 kg CH4/m²a. On the visible hot-spots the maximum emissions ranged between 14.9 up to extreme values of 562.4 g CH4/m²d (5.4 to 205 kg CH4/m²a). This means the current limit value for maximum hot-spot emissions on temporary covers for municipal solid waste landfills (containing a high amount of biodegradable waste) set in the Austrian landfill directive (< 10 kg CH4/m²a) was strongly exceeded on some measurement dates at the visible hot-spot.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
3.4 In-situ tests on water and gas permeability In additon to the monitoring of the gas flows and profiles, in summer 2016 field test have been performed in order to analyse soil physical parameters. The best information about soil properties governing vertical fluid movement in the soil matrix and in the macropores is given by ring infiltrometer methods. Field saturated hydraulic conductivity for tap water, corrected for 10 °C was kfs = 2.2×10−5 m·s−1 and 4.3×10−5 m·s−1 at the surface (measured on two dates), 5.1×10−6 m·s−1 at 40 cm depth and 1.3×10−5 m·s−1 at 80 cm depth at plot 1 (biowindow 1, grass covered spot without detected methane release at the surface). The values for plot 2 (biowindow 1, hotspot) at the same depths are 1.1×10−4 m·s−1, 2.0×10−7 m·s−1 and 4.0×10−5 m·s−1. These results reveal no clear distribution over profile depth or distinct properties related to one of the plots. The values were confirmed by similar results determined with guelph permeameter experiments, especially the small value at 40 cm depth on plot 1. This is not a matter of course considering a probable small-scale heterogeneity and anisotropy and different dependency on the latter by the two methods. Soil was moderately wet on both field days. Dry bulk density deduced from cylinder samples was very small and ranged for both plots from ρd = 0.4 g·cm−3 to 0.6 g·cm−3. Under not to wet conditions it can be stated that vertical gas convection will not be restricted reasonably in the biocompost layer. Close to water saturation at the bottom of the layer a barrier effect may be speculated, leading to a horizontal redistribution of the gas in the GDB, an upward breakthrough in few macropores or preferential flowpaths and reaching the surface, e.g. at the hot spots visible in figure 3. Validation would require knowledge of the water retention behaviour of both layers at the transition and either simulation or monitoring of soil water movement up to a full parameterization for a 3D two-phase flow simulation.
4 CONCLUSIONS As it is well-known from former studies and international literature the selection of suitable materials for methane oxidation systems and the technical design, particularly with respect to the homogeneous distribution and supply of methane, are the most important factors for a proper and effective performance of methane oxidation. Especially in the case of switching existing gas wells (point sources) of an (formerly operated) active gas extraction system into passively supplied methane oxidation windows (bio-windows), the distribution and homogenous supply of methane to each biowindow seems to be the biggest challenge - as indicated by ongoing investigations on an older Austrian municipal solid waste landfill with already strongly decreasing gas generation. So far the landfill owner has installed only two pilot-windows. It was expected from the very beginning, that the biofiltration area of two windows would be too small to catch and oxidise all of the still produced (calculated) methane, and this was confirmed during the first monitoring period. However, the pilot phase indicated first important information regarding the distribution of methane in the biowindows, and the behaviour of the landfill gas flows within the still existing gas collection pipes system. It is planned to remove further vertical gas wells and to install about 10 – 12 additional biowindows, with improved technical design and as far as possible controlled methane supply according to the lessons learned from the pilot phase. This project is currently still under evaluation by the local authorities, however, it is expected that it can be implemented in 2018 at the latest.
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
REFERENCES Bogner, J., Abdelrafie Ahmed, M., Diaz, C., Faaij, A., Gao, Q., Hashimoto, S., Mareckova, K., Pipatti, R., Zhang, T. (2007). Waste Management. In: Metz, B., Davidson, O.R., Bosch, P.R., Dave, R., Meyer, L.A. (Eds.), Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA Fischedick, M., Roy, J., Abdel-Aziz, A., Acquaye, A., Allwood, J. M., Ceron, J.-P., Geng, Y., Kheshgi, H., Lanza, A., Perczyk, D., Price, L., Santalla, E., Sheinbaum, C., Tanaka, K. (2014). Industry. In: Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y. Sokona, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J. Savolainen, S. Schlömer, C. von Stechow, T. Zwickel and J.C. Minx (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.. Oonk, H., Boom, T. (1995). Landfill Gas Formation, Recovery and Emissions, TNO-report 95130, TNO, Apeldoorn, the Netherlands. Reynolds, W. D., Elrick, D. E. (2002). Constant Head Well Permeameter (Vadose Zone). In: Dane, J. H., Topp, C. G., (Co-Editors). (2002). Methods of Soil Analysis Part 4: Physical Methods. Madison, Wisconsin: Soil Science Society of America. 821-826 Reynolds, W. D., Elrick, D. E., Youngs, E. G. (2002). Single-Ring and Double- or ConcentricRing Infiltrometers. In: Dane, J. H., Topp, C. G., (Co-Editors). (2002). Methods of Soil Analysis Part 4: Physical Methods. Madison, Wisconsin: Soil Science Society of America. 821-826 Soil Measurement Systems. (2008). Tension Infiltrometer: Users Manual. Tucson, Arizona: Soil Measurement Systems, LLC Spokas, K., Bogner, J., Chanton, J., Morcet, M., Aran, C., Graff, C., Moreau-le-Golvan, Y., Bureau, N., Hebe, I. (2006). Methane mass balance at three landfill sites: what is the efficiency of capture by gas collection systems? Waste Management, 26, 516- 525.
