Gas emissions from biodegradable waste in United Kingdom landfills

Original Article Gas emissions from biodegradable waste in United Kingdom landfills Waste Management & Research 29(1) 69 76! The Author(s) 2011 Reprints and permissions: sagepub.co.uk/journalspermissions.nav DOI: 10.1177/0734242X10389510 wmr.sagepub.com Sally Maree Donovan, Jilang Pan, Thomas Bateson, Jan R Gronow and Nikolaos Voulvoulis Abstract The aim of this research was to predict the effect that the biodegradable municipal waste (BMW) diversion targets in the European Union landfill directive (99/31/EC) would have on landfill gas emissions. This is important for continued mitigation of these emissions. Work was undertaken in three stages using the GasSim model (v1.03) developed by the Environment Agency (England and Wales). The first stage considered the contribution to gas emissions made by each biodegradable component of the waste stream. The second stage considered how gas emissions from a landfill accepting biodegradable wastes with reduced biodegradable content would be affected. The third stage looked at the contribution to gas emissions from real samples of biologically pretreated BMW. For the first two stages, data on the waste components were available in the model. For the third stage samples were obtained from four different biological treatment facilities and the required parameters determined experimentally. The results of stage 1 indicated that in the first 15 years of the landfill the putrescible fraction makes the most significant contribution, after which paper/card becomes the most significant. The second stage found that biodegradability must be reduced by at least 60% to achieve a reduction in overall methane generation. The third stage found that emissions from samples of biologically pretreated BMW would result in a significant reduction in gas emissions over untreated waste, particularly in the early stage of the landfill lifetime; however, low level emissions would continue to occur for the long term. Keywords Mechanical biological treatment, compost-like outputs, biodegradation, landfill gas generation Date received: 25 June 2010; accepted: 12 October 2010 Introduction The landfilling of biodegradable wastes can lead to many environmental problems including fires and explosions, odour nuisance, vegetation effects, local air quality impacts and greenhouse gas emissions. In order to reduce these impacts many countries are aiming to reduce the biodegradable content of landfilled waste. In the EU, the landfill directive (Council of the European Union 1999) contains targets for diverting the biodegradable fraction of municipal solid waste from landfills, which must be achieved by all member states. Biodegradable fractions make up the majority of municipal solid waste (MSW), particularly food, gardening and paper/card wastes. Initiating a separate collection of organic wastes for recycling to land as a soil conditioner goes some way to achieving this diversion, but the residual waste stream still contains a significant proportion of biodegradable waste which must be treated prior to landfilling. Mechanical biological treatment (MBT), in which a combination of sorting the waste using mechanical techniques, and treating the biodegradable fraction using either aerobic or anaerobic methods is becoming one of the major methods for treating the residual MSW. There is a variety of MBT processes which treat the waste to different extents in different ways. It is becoming one of the major technologies for treating residual MSW, as it is cheaper, more flexible and more widely accepted by the general public than thermal treatments. In order to achieve complete diversion, the outputs of the biological processing would be applied to land as a soil conditioner. Unfortunately the outputs of these are Centre for Environmental Policy, Imperial College London, UK. Corresponding author: Nikolaos Voulvoulis, Room 321, Mechanical Engineering Building, Imperial College London, South Kensington, London, SW7 2AZ UK Email: n.voulvoulis@imperial.ac.uk

70 Waste Management & Research 29(1) often unsuitable for application to land due to insufficient degradation or contamination, and so are ending up in landfills. The environmental impacts of a landfill accepting large quantities of partially degraded wastes are largely unknown, giving rise to this research. One of the main concerns that have been expressed by landfill operators is the impact it will have on their ability to capture the methane-rich landfill gas. The concerns in this case are twofold: first, the large amount of gas generated in the early stages of a landfill accepting untreated wastes was one of the main sources of revenue as the volume of the gas was large enough to generate a significant amount of energy; and second, it is a requirement of the landfill directive that gas emissions from landfills are captured and utilized where possible. Where emission rates are too low for energy generation they must be flared; however, at very low concentrations the ability to capture the gas and flare it becomes more difficult. Thus if low levels of emissions are occurring over a long time they could be a huge and expensive challenge, after having already lost their main source of revenue. The overall aim of this research was therefore to investigate the way gas emissions from landfills will be altered as the biodegradable municipal waste diversion targets come into force. This was achieved through three objectives: assessing the contribution that each biodegradable fraction makes to methane generation; assessing the effect that reducing the biodegradability of the total incoming waste stream will have on methane generation; and finally assessing the contribution biologically pretreated wastes will make to methane emissions. The GasSim model developed by the Environment Agency (England and Wales) formed the basis for studying all three objectives. Methodology The GasSim model The GasSim model developed by the Environment Agency (England and Wales) takes a first-order kinetic approach, meaning the main limiting factor of the landfill gas production rate is the amount of substrate available. Rather than calculating the degradation rate of every component of the incoming waste stream, which would become extremely complicated, the components are classified as either slow, medium or fast. This simplification has been shown to correlate well with experimental data from landfill sites. The determination of degradation rate is based on cellulose and hemicellulose content. The moisture content of a landfill has been shown to have a strong influence on methane production as high moisture content limits transport of oxygen, an inhibitor to methanogenesis, from the atmosphere; facilitates transport of microbes and nutrients between microenvironments; and dilutes the concentration of inhibitors (Christensen et al. 1996). The GasSim model allows two methods for detailed input of moisture content, depending on available data. The first allows input as dry, average or wet, which is taken as less than 30%, between 30 and 60% and greater than 60%, respectively. Where the moisture content exceeds 70% it is taken to act as an inhibitor to methane generation and should be input as dry. The second method calculates the moisture content based on infiltration, waste density, effective porosity, leachate head, hydraulic conductivity, adsorptive capacity and leachate recirculation. All the inputs in the model can be entered as a probability distribution function, reflecting the largely unpredictable nature of landfills. A Monte Carlo simulation is performed, based on a user-defined number of iterations. The main interest for this research was the variation in gas emissions with changes in the incoming waste stream; thus all other parameters were constant for each scenario, and taken from example landfill sites contained within the model. Contribution to methane from different waste types The first stage of the research used the GasSim model to predict the contribution to methane generation from each biodegradable fraction of the waste stream. Site-specific parameters were taken from the predefined Bucks landfill, which represents typical UK landfill sites and contains very detailed site-specific annual waste disposal data for a reasonably long operational time, from 1978 to 1985. It accepted 2 000 000 t of domestic waste over the 8 years, and covered an area of 70 000 m 2 (200 m north to south, 350 m east to west). The 1980 2010 waste stream, predefined by the model, was used as the base case of this section, as it is representative of 1995, the base year of the biodegradable waste diversion targets in the directive. Simulations were conducted over a period of 100 years. The first part of the simulation assumed that the whole landfill received only one type of waste and the iteration was repeated for each individual waste type. For the second part of this stage, a landfill with a typical incoming waste composition, as defined by the Environment Agency, for 1980 2010 was modelled, followed by a landfill accepting the same waste stream with the paper/ card fraction set to 0 (the percentage was evenly distributed over other waste types). The model was run again with paper/card returned to its normal percentage and putrescibles set to 0, and so on for all the different biodegradable components of the waste stream. The difference in total methane projected to be generated over 100 years between the landfill accepting all wastes, and landfills accepting waste with one type excluded was then calcuated. This modeling method was then repeated with future waste compositions for 2010, 2015 and 2020, as predicted by the Environment Agency.

Donovan et al. 71 Theoretical assessment of methane generation from biologically pretreated wastes The second stage of the investigation used the GasSim model to predict how wastes that have been pretreated through an MBT process would behave in a landfill. MBT processes and their outputs vary greatly and thus the extent of reduction in biodegradability potential also varies. This will affect landfill gas generation as well as settlement and leachate contamination. Values have been reported in the literature that range from 35 to 60% depending on the waste composition, and the specific MBT process. The theoretical prediction of the gas generation rate of MBT outputs accounted for this variation by modelling organic content reduction rates of 50, 60, 70, 80 and 90%. Methane generation from biologically pretreated wastes For the third stage of the investigation, samples were obtained from four different biological waste treatment plants. The first, called F1, was taken from a composting plant accepting source-segregated kitchen and garden waste from households in a ratio of approximately 30 garden to 5 food. It was composted in tunnels with forced aeration for 2 to 3 weeks, and matured for 5 weeks. The second, F2, was an anaerobic digestion plant accepting only kitchen waste from households and grass cuttings from civic amenity sites. It operated a continuous stirred tank reactor system at approximately 40 C. The third, F3, was a composting facility taking in coarsely screened mixed waste. The composting process lasted for 6 weeks. The fourth, F4, was an MBT taking in mixed MSW in black bags. The biological stage was an in-vessel composting process lasting 2 weeks. Each of these samples was analysed for moisture, cellulose and hemicellulose content. The details of the experiment and the results are presented in Donovan et al. (2010a). The results of this analysis were then used to predict gas generation from these kinds of waste in a landfill, using the GasSim model. The landfill parameters were from the Berks landfill, contained within the model. This was 240 100 m 2 (490 m 490 m) and accepted 1.6 Mt of MSW from 1978 to 2001. Similar to the previous sections, the only variation in the model input parameters was the waste composition. Nine scenarios were modelled: a landfill accepting untreated municipal solid waste, based on the Waste Strategy for England (Defra 2007) called the base case scenario ; four landfills accepting only biologically pretreated waste from each of the four facilities described above, called the CLOs only scenarios; and four landfills accepting half untreated waste and half biologically pretreated waste called the half half scenario. In each of the scenarios the total amount of waste accepted at the landfill remained the same with only the proportion of each component of the waste stream altered. It was difficult to determine a realistic combination of raw and treated waste, as issues such as recycling rates, alternative uses for CLOs and future waste arising were beyond the scope of this investigation. For this stage of the modelling the period was increased from 100 to 150 years, as it had been suggested that 100 years was too brief a period. Results and discussion The results of stage 1, in which the methane generation of each waste fraction under the same landfill conditions was modelled, are shown in Table 1. The highest methane generation rates were found in the paper/card categories especially other paper, which is defined as papers excluding newspapers and magazines. All cards are generally easy to degrade and result in significant gas generation. Therefore paper/ card is of primary importance. The second and third highest generation rates were found in the miscellaneous and putrescible fractions, making them of secondary and tertiary importance, in terms of total gas generation over 100 years. Figure 1 shows that for putrescibles, a sharp increase in gas production over a short period of time was seen, followed by a dramatic decrease after it reached the peak. In contrast, most fractions in the paper/card category showed a slower increase and declined more gradually after reaching the production peak. Therefore if a shorter period were assessed, the level of importance in terms of total methane generation would change. Most notably, as the time period decreased, the importance of the putrescible fraction increased, and became the primary waste stream when the time period was less than about 14 years. Table 1. Average annual methane generation from individual waste fractions as calculated by GasSim Waste fractions Sub-categories Average annual methane generation (m 3 h 1 year 1 ) Level of importance Paper/card Newspapers 132.2 Primary Magazines 156.3 Other paper 616.6 Liquid cartons 282.7 Card 282.9 Other Card 282.7 Textile Textile 140.6 Quaternary Miscellaneous Disposable 187.9 Secondary Other 187.8 miscellaneous Putrescibles Garden waste 78.5 Tertiary Other 155.8 Fines 10 mm 140.3 Quaternary

72 Waste Management & Research 29(1) Methane generated (m 3 h 1 ) 4500 4000 3500 3000 2500 2000 1500 newspapers magazines Other paper Liquid cartons Card Other card Textile Disposable Other miscellaneous Garden waste Others putrescible Fines Sewage sludge Composted Lncination ash 1000 500 0 1975 1980 1985 1990 1995 2000 2005 Year Figure 1. Annual contribution to mean methane generation from individual components of the waste stream. Table 2. Predicted biodegradable content of UK waste stream going to landfill (Environment Agency) Waste content 1980 2010 waste stream (%) 2010 waste stream (%) 2015 waste stream (%) 2020 waste stream (%) Paper/card 33.5 25.2 16.7 11.7 Textile 2.4 1.8 1.2 0.8 Miscellaneous 8.0 6.0 4.0 2.8 Putrescible 20.8 15.6 10.4 7.2 Fines 7.1 5.3 3.6 2.5 Within Table 1, it can be seen that the range of methane production rates vary from less than 100 to more than 600 m 3 h 1, indicating the methane generation rate of different materials varies significantly, therefore removing certain fractions of the waste stream, especially those that result in greater methane generation rates, would seem a more appropriate strategy, than diverting a proportion of each fraction. Table 2 shows the composition of the biodegradable components of the incoming waste stream as defined by the Environment Agency for 1980 2010, as well as their predicted future waste compositions for 2010, 2015 and 2020. The non-biodegradable fractions are not included as they are not relevant to the current work. Table 3 shows the change in methane generation between a landfill accepting a normal waste stream, compared with a landfill accepting a waste stream with one biodegradable waste type excluded. During modelling, the proportion of certain waste fractions was taken away from the total and spread among the rest of the waste fractions accordingly. As expected the effect of removing each one differs significantly; for example, removing the paper/card fraction resulted in a 52% reduction in total methane production, whereas the reduction after removing the textiles fraction was insignificant. According to these results the reduction in methane production achieved beyond compliance with the landfill directive targets is highly significant for the removal of paper/card and quite significant for the reduction in putrescibles. Noticeably, although the proportions of biodegradable waste fractions decreased in the waste streams, the percentage of reduction in methane production achieved by removal of that fraction increased significantly. This relationship indicates that to achieve significant reduction in landfill gas generation, the reduction/sorting of certain targeted fractions, especially paper, card, green waste and other putrescibles should be prioritized.

Donovan et al. 73 Table 3. Reduction in methane generation with the removal of biodegradable fractions of the waste stream Waste without 1980 2010 CH 4 reduction (%) 2010 CH 4 reduction (%) 2015 CH 4 reduction (%) 2020 CH 4 reduction (%) Paper/card 46.2 52.1 56.7 59.3 Textile 0.3 a 0.2 0.9 1.1 Miscellaneous 1.3 3.3 5.3 6.5 Putrescible 2.6 a 3.5 9.2 12.1 Fines 1.0 a 0.7 2.6 3.7 a Negative numbers indicate the small proportion and significant influence of fractions, which degrade easily and produce methane. The Environment Agency s role is to ensure that the most appropriate waste disposal options are applied in the UK without causing harm to health or the environment. As the landfill directive indicates, however, sorting does not necessarily need to be applied to all types of waste. Those that contribute most to methane generation could be targeted and treated before landfill. A study conducted by AEA technology for the Greater London Authority (Poll 2004) showed that 52% of residual household collected waste is potentially recyclable and a further 34% is potentially compostable. However, this does not take into account the suitability of the material for recycling; for example, not all paper and card can be targeted for recycling. Moreover, the maximum amount of recyclable and compostable material that could actually be collected will be lower because of contamination issues. As the first phase of the study indicated, separate collection of paper, card, organic kitchen and garden waste for either recycling or composting is a useful and efficient first step to divert biodegradable waste from landfills. In areas in which the separate collection of organic waste is already established or is not feasible, MBT is a way to divert BMW in order to achieve the desired reduction in gas emissions from a landfill. MBT of residual MSW is becoming common practice in many European member states, especially the Netherlands, Austria and Germany, thus much useful data have been generated from these countries. MBT outputs differ significantly from untreated MSW in that they are more homogeneous, have a smaller average particle size, lower biological activity and more soil-like properties. Researchers at the University of Hanover (Kuehle-Weidemeier & Doedens, 2003), have been working on the use of MBT outputs as a means of improving control over their subsequent degradation in landfills. The biological pretreatment in particular aims to reduce contaminant concentration in leachates, gas generation rates and settlement in a landfill. Although easily degradable fractions of the MSW are degraded during biological treatment, significant biodegradable components remain and will thus continue to degrade under landfill conditions. The remaining biodegradability is dependent on the particular MBT process and the incoming waste stream. Mu ller & Bulson (2004) calculated that less than 10% of the biodegradable content would remain after effective organic treatment; however, other studies show reductions of 35 60%. The GasSim modelling results indicate a minimum reduction of 60% of the fast-degrading components; that is, those which would degrade significantly in a composting process (especially putrescibles), are required to attain a meaningful reduction in methane emissions, as shown in Figure 2. The figure also indicates that significant reductions in emissions were seen when putrescible content is reduced by more than 70%; at a reduction of 90% a reduction in methane production of more than 74% is predicted. While MBT processes can be controlled to maximize reduction in biodegradability, the outputs are still also dependent on the composition of the incoming waste stream, which is much more unpredictable. Therefore only general estimates can be made on physical properties of the MBT waste, and outputs will need to be tested frequently in order to understand their behaviour in a landfill. The third stage of the investigation predicted emissions from landfills accepting compost-like outputs from four currently operating biological pretreatment facilities. It was found that a significant reduction in overall methane generation was achieved, compared with untreated waste. A consequent reduction in emissions will have a positive impact on the environment if no attempt is being made to mitigate gas emissions; however, where the mitigation of gas emissions is being practised, which is a legal requirement in many countries, the impact that landfilling CLOs will have on emissions over time is of concern. Figure 3 indicates that gas generation will continue to occur over long time periods, similar to those for untreated waste, reflecting the results of Knox & Robinson (2007) that indicated even though there was significant reductions in peak emissions, the long-term emissions tail at landfills would be similar to that of untreated waste. The low flow rates will make it difficult and more expensive to capture the gas and also limit the ability to utilize it. The Environment Agency s Guidance on Landfill Gas Flaring states that a practical threshold for combustion is 100 m 3 h 1 (Environment Agency 2002). Many models also assume that combustion in flares will fail at this rate and in practice, it has been found that with poor gas collection

74 Waste Management & Research 29(1) 1600 1400 90% reduction Methane generated (m 3 h 1 ) 1200 1000 800 600 400 80% reduction 70% reduction 60% reduction 50% reduction 1980 2010 200 0 1960 1980 2000 2020 2040 2060 2080 2100 Year Figure 2. Methane generation from landfill with decreasing biodegradable content. Methane generated (m 3 h 1 ) 400 350 300 250 200 150 100 Base case F1: 30 parts garden:5 parts kitchen; tunnels 2 3 weeks; maturation 5 weeks F2: 98% kitchen waste with 2% grass cuttings; Ad stirred tank 40 C F3: Coarsely screened mixed waste; 6 weeks composting F4: Black bag rmsw; 2 weeks in-vessel composting 50 0 0 50 100 Simulation year 150 Figure 3. Methane generation from landfilled MBT outputs. infrastructure, combustion may actually fail at generation rates of 180 m 3 h 1. Table 4 indicates the percentage of the 150-year lifetime that the landfills were likely to generate gas below this threshold. In the base case scenario it was 64.7% of the time. In the half half scenario it was increased slightly but stayed within 6% of this value ranging from 66.0% for F1 to 70.7% for F2. In the CLOs only scenario F2 was increased to 80%, however, it is unlikely that a CLOs only

Donovan et al. 75 Table 4. Methane reduction compared to the Defra waste strategy 2007 composition data and lifetime landfill gas passively venting Scenario CLO source Lifetime methane reduction (%) Percentage of lifetime landfill gas generation < 100 m 3 h 1 Percentage of lifetime landfill gas generation < 10 m 3 h 1 Base case 64.7 < 1 Half-half F1 11.7 66.0 14.0 F2 32.9 70.7 14.0 F3 21.7 67.3 13.3 F4 13.3 66.7 14.7 CLOs only F1 23.8 68.0 35.3 F2 65.0 80.0 46.0 F3 42.5 72.0 38.7 F4 28.5 69.3 36.0 landfill would be created. For example, paper and card that generate small amounts of methane over long periods of time are likely to be present, which will increase these generation rates in the later years, of the landfill. However, it is expected that these scenarios will still result in a significant reduction in the amount of revenue generated through selling the gas as a renewable energy source. It will also be necessary for landfill operators to invest in low volume landfill gas flares with re-igniters in order to achieve the requisite gas controls. These can deal with gas volumes down to approximately 10 m 3 h 1 (there is continuous development of flares that can deal with volumes lower than this and so this threshold is likely to reduce with time). To investigate the impacts of this a threshold of 10 m 3 h 1 was assumed and is also presented in Table 4 as a percentage of landfill lifetime generation rates. The base case generated gas at rates below this threshold for less than 1% of the 150 years. The CLOs only scenario showed a significant increase in the time scales of this generation rate of up to 46% for F2. This is of concern as it is likely this may be emitted untreated to the atmosphere. In reality it is unlikely that a landfill would contain CLOs only, and the half half scenario is probably closer to what is likely to occur. In this case it increases to 14%, which amounts to approximately 1055 tonnes of CO 2 equivalents per year. Conclusions The findings of this study could have significant implications for the way landfill management should change as the biodegradable municipal waste diversion targets come into force. First, it was shown that over the lifetime of the landfill the slowly degrading components, especially paper and card waste, make the most significant overall contribution to methane emissions. Thus focusing diversion efforts on these highly degradable fractions of the waste stream will be more effective than applying a general reduction to all waste materials, in terms of reducing gas emissions. The second stage began the investigation into waste materials that had been biologically treated prior to landfilling to reduce their gas generation potential. It was found that a reduction of at least 60% was required to see any reduction in gas generation, and at least 70% to see a significant reduction. The final stage of the investigation indicated that biologically pretreated waste materials will continue to generate gas for at least 150 years after deposition, at very low levels. In order to comply with the landfill directive, these gases must be captured and where possible utilized or else flared to prevent release of methane to the atmosphere. The results of this stage of the investigation indicated that the ability to utilize landfill gas for energy generation will be significantly reduced by pretreating the waste, but the time over which gas is generated will remain long term. Thus the potential for landfill operators to benefit financially from selling the gas for energy generation likely will decline, and greater expenses will be incurred to capture very low level emissions for flaring. References Christensen TH, Kjeldsen P and Lindhardt B (1996) Gas-generating processes in landfills. In: Christensen TH, Cossu R and Stegmann R (eds.) Landfilling of Waste: Biogas. London, UK: E&FN Spon. Council of the European Union (1999) Directive on the landfill of waste (99/31/EC). Official Journal of the European Union L 182: 1 19. Defra (Department for the Environment, Food and Rural Affairs) (2007) Waste Strategy for England. Norwich, UK: Her Majesty s Stationery Office. Donovan SM, Bateson T, Gronow JR and Voulvoulis N (2010a) Characterisation of compost-like-outputs from mechanical biological treatment of municipal solid waste. Journal of the Air and Waste Management 60: 694 701. Donovan SM, Bateson T, Gronow JR and Voulvoulis N (2010b) Modelling the behavior of mechanical biological treatment outputs in landfills using the GasSim model. Science of the Total Environment 408: 1979 1984. Environment Agency (2002) Landfill Gas Flaring Guidance. Bristol, UK: Environment Agency.

76 Waste Management & Research 29(1) Knox K and Robinson H (2007) MBT and thermal treatment of MSW residues: a comparative study of energy balance and long term pollution potential of leachates. In: Cossu, R and Stegmann, R (eds). Proceedings of the Eleventh Waste Management and Landfill Symposium, 1 5 October 2007, p. 470. CISA, Cagliari, Italy. Kuehle-Weidemeier M and Doedens H (2003) Landfilling and properties of mechanical-biological treated municipal waste. Christensen, CH, Cossu, R and Stegmann, R (eds). Proceedings of the Ninth International Waste Management and Landfill Symposium, 5 9 October, 2003, p. 375. CISA, Cagliari, Italy. Mu ller W and Bulson H (2004) Significance of bio-mechanical waste treatment in Europe. In Proc. of the Integrated Waste Management and Pollution Control: Policy and Practice, Research and Solutions, Stratford-upon-Avon, 28 30 September, pp. 22 31. Golder Associates, Nottingham, UK. Pan J and Voulvoulis N (2007) The role of mechanical biological treatment in reducing methane emissions from landfill disposal of municipal solid waste in the United Kingdom. Journal of the Air and Waste Management 57: 155 163. Poll J (2004) Greater London Authority Waste Composition Scoping Study. Harwell, Oxfordshire, UK: AEA Technology.