BIOCOVER-REDUCTION OF GREENHOUSE GAS EMISSIONS FROM LANDFILLS BY USE OF ENGINEERED BIO-COVERS

Emission of methane from landfills due to anaerobic decomposition of organic material is one of the most important envi ronmental concerns with regards to solid waste management. This is due to the amount of methane released from landfills globally and the relatively high global warming potential of methane. An approach to reduce emissions is to improve conditions for biological oxidation of methane in the top cover using engineered biocovers. A demonstration project was init iated at the Technical Universi ty of Denmark under the EU Life Env ironment program, where this technology is applied in full scale at section I on Fakse landfill in Denmark. Construction of the full scale biocover at the test site was completed at time of writing. The main project objective was to document the construction and efficiency of the system. The project actions consist of a logical order of tasks performed in able to meet the objectives of the project. At fi rst the landfill was characterized. Expected landfill gas production was estimated based upon the collected data using models. Then, a baseli ne study was performed, consisting of an evaluation of the spatial variabil ity in methane emission at the s ite. The total methane emission from the landfill was measured by use of a tracer technique. M ixtures of locally available soils and organic waste residuals were tested by laboratory batch and column experiments. A cover improvement plan included details on material additions to selected areas of the landfill, maintenance plans of the total landfill cover. A plan for monitoring performance was setup. The emissions after the cover improvement wi ll be compared to the emissions obtained during the baseline study. Scenarios for other landfills will be calculated based on the experiences obtained from the studied landfill.


This new suggested innovative technology is to be demonstrated at Fakse Landfill, Denmark
(see Figure I). The landfill is divided in two parts; section I which was in use from 1 98 1 until I 996 will be the focus of the project activities. This part of the landfill has an area of I 3 ha and received mixed waste. Section I contains 700,000 tons of waste. The oldest part is finally covered with a relatively thick layer of soil having low gas permeability, while the rest of the landfill is temporary cover with a thinner layer of the same soil type.
The idea of the project is to construct a biocover system by establishing permeable regions, so-called windows, containing locally available materials with proven high methane oxidation potential.

STATE OF THE ART
Estimations of the methane production at Danish landfills without LFG utilization were carried out in 1 998 [8]. Based on the original data the potential methane surface emissions 1 Kalmar ECO-TECH KALMAR, SWEDEN, November 26- 28,2007 have been calculated for 2004. Typical potential methane surface emissions are in the order of 1 2-20 g CHa 4 m· 2 d· with most emissions in the 2-1 0 range. For a new landfill with a thickness of 1 5 meter receiving municipal solid waste a typical methane surface emission is 30 g CH4 m· 2 d-.The attenuation of methane in landfill top covers by methane oxidation has been subject to several studies [ I ], [4], [5], [9], [ I I ]. The top covers often consist of a lower anaerobic zone where pore gas consists mainly of LFG, and an upper methanotrophic zone characterized by a pore gas mixture of LFG and atmospheric air diffusing into the top cover from the atmosphere. Studies of landfill covers through simulating column experiments have 1 obtained methane oxidation capacities in the order of I 00-250 g CH4 m· 2 d· . Thus, the oxidation capacity determined is much higher that the observed potential methane emissions for landfills as mentioned above. This indicates that biological methane oxidation in landfill soil covers indeed is a possible technology, which especially for landfills with medium to low methane production may be a very cost-e ffective solution to reduction of green house gas em1ss1ons.
Attempts to enhance the attenuation process have worked with addition of organic material (compost, sludge etc.) to the soil. It proves that manipulation of landfill cover soils to maximize their methane oxidation potential might provide a complementary strategy for controlling emissions. However, at present, landfill caps are not designed with methane oxidation in mind. At many landfills the quality of the cover is in many cases not perfect, and can be of a very heterogeneous nature, and in many cases consisting of clay soils with low gas permeabilities. In such cases, regions with much higher gas permeabilities may exist. Besides, cracks may form during dry periods, or by settlement or erosion, and methane venting through such regions will not reside in the cover long enough to be oxidized. In these cases a significant attenuation bene fit could be achieved with relatively simple design and/or manipulation of the top-soil cover followed up by continuous monitoring of the cover. For example, attenuation can be considerably improved thanks to better soil porosity allowing better through-flow of methane and oxygen and selection of materials with significant methane oxidation potential.
The stimulating e ffects of biocovers on methane oxidation in the field have been studied [6] . They have measured the methane emission from Austrian landfill test cells covered with sewage sludge compost or municipal solid waste compost during a three-year period. For both type of biocovers the results indicated a significant removal of methane . No methane emission could be detected at the surface during the whole measuring period. The research, however, did not include quantitative measurements of LFG fluxes through the cover before and after cover improvement, so it is difficult to estimate a methane oxidation rate from the results. This high performance was obtained when a layer of coarse gravel was installed beneath the compost layer. This layer facilitated a homogeneous gas distribution. The Austrian experiment which was performed in a temperate climate is the only field scale documentation of methane oxidation stimulation in landfill soil covers.
Mixtures of materials (including organic waste materials, coarse sandy and gravel type soils, and landfill soil containing the methanothrophic bacteria) in the soil covers probably constitute the most efficient biocovers for methane oxidation. Studies to optimize the methane oxidation performance by mixing different materials have only been carried out in a few cases and have never been followed up by ful l scale implementation at landfi lls.
Documentation of the efficiency of improved covers at landfi lls for mitigation greenhouse gas emission is not a fully deve loped methodology. A documentation procedure includes measurements of the whole landfila l emission of methane before and after the cover has been improved for better mitigating the methane emissions. Several methods exist for whole landfi l l emission measurement, inca luding mass balance modeling, flux chamber techniques. micrometeorological techniques, and tracer techniques [3 ] . The tracer technique as developed by the Chalmer group [3] seems to be the most reliable method for landfills.
Policy development is needed to secure a sustainable and cost effective management of landfi l l gas at landfil ls also incorporating low technology solutions based on natural methane emission reduction by biocovers. To obtain a reliable technology the reduction in methane emission from the landfi l l by installing a biocover system a guide line needs to be developed. The guideline should describe the full procedure of establishing a biocover system on a landfil ls and methods for documenting the efficiency of the instal led biocover system for mitigating the methane emission from the landfill.

PROJECT OVE RV I EW
The full title of the BIOCOVER project is "Reduction of Greenhouse Gas Emissions from Landfil ls by use of Engineered Bio-covers'·. The project is funded by the LIFE I l l ENVI RONMENT program, the Danish Environmental Protection Agency, and RENOSAM and runs from August 2005 to November 2008. The project is divided into a logical order of activities. The project activities are as fol lows: • Initial characterization of landfill. The landfi l l is characterized according to area. volume, waste characterization, soil thicknesses and types, and vegetation. Expected landfi l l gas production is estimated based upon the col lected data using available landfil l gas generation models.

• Baseline study of methane emissions. On each of the landfi l ls an initial evaluation of the spatial variability in methane emission is carried out, and the total methane emission is measured by use of the tracer technique. • Testing improvement strategies in the laboratory. Mixtures of locally available soils
and organic waste residuals are tested in the laboratory to determine the best suited material mixture to be used for cover improvement. • Improvement of cover layer. On the basis of the baseline study which showed the spatial variability of the gas emission, and the tests on the improvement strategies, a cover improvement plan is developed. The plan will include details on material additions to selected areas of the landfill and maintenance plans of the total landfi ll cover. The covers are improved following the proposed plans.
• Establishing full scale demonstration biocover system. A fter the cover has been improved a monitoring plan is setup including the activities as described under --Baseline study of methane emissions . . . The spatial variability and the whole landfill emissions will be evaluated, and the emissions after the cover improvement will be compared to the emissions obtained during the baseline study. Evaluation of the oxidation efficiency at selected localities within the landfills will be made using the probe/flux chamber approach.
• Analysis of the economic viability of the biocover technology. The standard cost for this reduction technology (in DKK/tons CO 2 G WP removed) will be calculated and compared to other reduction technologies, especially the gas flaring technology. Scenarios for other landfills will be calculated based on the experiences obtained from the two studied landfills. There is no LFG extraction system instal led at Fakse landfill.

Gas prod uction modeling
Four available gas production models were applied for estimating the production of Landfill gas at Fakse Landfill. These were :

Figure 4. Total landfill gas production of currently deposited waste material at Fakse landfill estimated using four gas production models [7].
LFG production simulations were performed for both Section I, which is the main focus of the project, and for Section I I . Furthermore, the production of LFG in the capped and uncapped part of Section I has been evaluated separately as has each of the 7 disposal subunits in Section I.
The resulting LFG production for 2005 in Section I of Fakse Landfi l l was estimated in the 1 range from 0.6 -1 .3 million m 3 LFG/yr 5 corresponding to 5 5 5 to I 1 65 kg CH 4 d-(see Figure 4). Seventy-five percent of this gas production was estimated to take place in the uncapped part of Section I comprised by Units 4-7. The reported range includes mode l estimates obtained using the models I PCC, GasSim and Multi Phase Model. LandGE M results were assessed t o overestimate the gas production at Fakse Landfi ll, since the model intended for U.S. landfills with a high amount of municipal solid waste.

Conceptual model of LFG em ission
A conceptual model of LFG emission from Fakse landfi l l was setup based on initial methane concentration screenings, cover soil characterization and considering the technical design of the landfi ll. Clayey soi l has been used for both temporary and final covering. This leads to low permeability, which in turn leads to gas emitting in more concentrated parts of the landfill, where permeabi lity is higher rather than more uniformly upwards through the top soi l over the entire area.
Main pathways of emission were considered to be leachate col lection wells, which are part of the leachate drainage system and high emission areas (hot spots) on the soi l cover. These areas were believed to be found on slopes and parts of the temporary cover, where the soi l cover i s thin, and thereby more permeable.

Measu rem ent of methane em ission through leachate collection system
To measure methane emission rates from the leachate collection system, a continuous tracer release method was used on desired locations. This was done to evaluate the necessity of modifying the leachate system to reduce gas emissions through this pathway, which would constitute a bypass of the biocover system. The principle of the method used was to continuously release a gaseous tracer ( carbon monoxide) at a constant known rate near the source of the methane emission (bottom of leachate wells), and subsequently compare measurements of concentrations of tracer and methane in the pl ume downwind after background concentrations of tracer and methane have been measured. Concentrations of carbon dioxide were also measured.
The basic equation for calculating emission rates through these measurements of concentration was derived from the assumption that the ratio between flow rates of methane and tracer is equal to the ratio between observed concentrations measured downwind. 1 The sum of methane emitting through the wells was 35 1 kg C H 4 d-, which suggested that as much as 50% of the methane produced in the waste emitted through the leachate collection system, by comparing this rate to the results of the LFG production models [2].

Measurement of methane e mission through hot spots in soil cover
The methane emission through the soil cover at Fakse landfill was considered to occur mostly through so-called hot spots rather than more uniformly through the entire soil cover. Localization of the hot spots and measurement of the methane flux through these were done, to evaluate the need for cover improvement in connection with installation of the biocover. Another objective was to test the presumptions of the conceptual model with regards to the spatial variability of the emission through the soil cover.
To systematically evaluate the spatial variability of the emission. a measurement grid was marked on each of the seven disposal units on the site with grid spacings ranging from 22 to 28 meters. After marking the grid of point locations for measurement, each location was measured for methane concentration just above ground level in four screening campaigns. Concentrations are measured on the grid points and hot spots using a Photovac portable flame ionization detector (FID) (detection limit: 0.5 ppm CH4) connected via I m tube to a 20cm diameter funnel.
Location of hot spots was done by systematically measuring near surface methane concentrations using the F I D and funnel walking slowly along the grid between all grid points in both north-south, and east-west directions. The screening between grid points was supplemented with a screening on slopes and other features were emission through the soil cover was potentially high. On most grid points concentrations were near background level (<2 ppm). On located hot spots, concentrations of several thousand ppm were measured. Hot spots were found almost exclusively on slopes at the site. The observat ions made during concentration screenings confirmed the conceptual model of methane emission at the site.
Through approximately 200 measurements of flux rates primarily at the hot spots, average 1 emission from each of the identified hot spot areas ranged from 5 to 4000 g CH4 111· 2 d-, where the h ighest emissions were measured at the temporarily covered part of the landfill . ln most cases the area of the hot spots were very clearly defined and these were evaluated using the F I D and funnel screening near surface concentrations of methane. Emission through identified hot spots covering a combined area equal to 0.4% of the site was measured using a mobile type flux chamber. Measurements of the concentration inside the flux chamber were done using the FID. Time of each flux measurement was approx imately five minutes, and the concentrations of methane in the flux chamber were measured 6-7 times during each measurement.
Through the flux measurements, a total emission through identified hot spots was measured to 1 be 1 82 kg CH4 d· , of which 97% emitted through the temporarily covered part of section I .
The sum of the emission measurements o f the local sources (leachate collection system and 1 hot spots in soil cover) was 533 kg CH 4 d' .

TOTA L M E T HA N E EMISSION M EASUREMENTS
The total methane emission from the disposal site was measured using a tracer technique, combining controlled tracer gas release from gas bottles placed on the landfill with time resolved concentration measurements downwind the landfill using FTIR absorption spectroscopy.
Initially to the tracer release measurement, a general leak search at the landfill was conducted with the main purpose to identi fy high emission areas for placement of the tracer release bottles. In all two field campaigns were performed; during October 1 1 a -1 2, 2006 and February I 9-20. 2007. At both field campaigns an overall leak search showed that the methane em ission from the old landfill section was localized to the leachate collection wells and some slope areas with cohered with prev ious observations. During the first campaign the methane emission from section I was estimated to be 1 748.8± 1 63 . 2 kg CH 4 h' • During the second campaign the methane emission from section I 1 was estimated to be 732.0±52. 8

kg CH 4 h-[ I O] .
1 The difference between results measurement o f local emissions ( 5 3 3 kg CH 4 h" ) and total 1 emission (average: 740.4 kg CH4 h" ) can be explained by more diffuse emission sources not being quantified in the first case. By comparing the values, it was concluded that the most important sources of methane emission were identified while assessing spatial variability in emissions.

ONGOIN G AND FUTURE WORK
At time of writing, construction of a full scale biocover system to reduce methane emiss ion from Fakse landfill was completed. The biocover material chosen was five year old composted garden waste, produced at the site. Based on results from column experiments investigating methane potentials of available biocover materials (not described in this paper). and measured total methane emission, a necessary total area of the biocover windows was set to 5000 m 2 . The windows were constructed by removing top soil, placing a 1 0-1 5 cm gravel layer and placing I meter of composted garden waste over the gravel layer. After construction of the biocover system, performance will be assessed using whole site methane emission measurement, of which results will be compared to baseline values. Studies on methane oxidation in the biocover windows will also be studies at a more local level. The performance will be studied over a full calendar year, to observe effia ciency under different weather conditions. Finally, a viability analysis of using biocover technology to reduce greenhouse gas emissions from landfills will be done, involving desk studies on five other European landfills using experiences gained from constructing the system at Fakse landfill.

ACKNOW LEDGEM ENTS
This project was funded through the EU Life Env ironment program . We thank Bente Munk and the staff at Fakse landfill for their cooperation.