- Email: [email protected]
Passive landfill gas emission – Influence of atmospheric pressure and implications for the operation of methane-oxidising biofilters
Waste Management 26 (2006) 245–251 www.elsevier.com/locate/wasman
Passive landfill gas emission – Influence of atmospheric pressure and implications for the operation of methane-oxidising biofilters Julia Gebert *, Alexander Groengroeft University of Hamburg, Institute of Soil Science, Allende-Platz 2, 20146 Hamburg, Germany Accepted 19 January 2005 Available online 23 March 2005
Abstract A passively vented landfill site in Northern Germany was monitored for gas emission dynamics through high resolution measurements of landfill gas pressure, flow rate and composition as well as atmospheric pressure and temperature. Landfill gas emission could be directly related to atmospheric pressure changes on all scales as induced by the autooscillation of air, diurnal variations and the passage of pressure highs and lows. Gas flux reversed every 20 h on average, with 50% of emission phases lasting only 10 h or less. During gas emission phases, methane loads fed to a connected methane oxidising biofiltration unit varied between near zero and 247 g CH4 h1 m3 filter material. Emission dynamics not only influenced the amount of methane fed to the biofilter but also the establishment of gas composition profiles within the biofilter, thus being of high relevance for biofilter operation. The duration of the gas emission phase emerged as most significant variable for the distribution of landfill gas components within the biofilter. 2005 Elsevier Ltd. All rights reserved.
1. Introduction The microbial oxidation of methane in biofilters is considered an alternative to the treatment of landfill gas (LFG) methane emissions that do not meet gas flow rate and methane content requirements for utilization or flaring (Figueroa, 1996; Gebert et al., 2003; Streese and Stegmann, 2003). Methane is aerobically oxidised to carbon dioxide, thereby serving as the sole carbonand energy-source for the methane oxidizing bacteria. Reviews of the taxonomy and physiology of methane oxidising bacteria are given by Hanson and Hanson (1996) and Bowman (2000). Biofiltration as a means for the abatement of anthropogenic methane emissions may be applied at landfills in the initial phase of operation, old landfills or sites containing material of low gas generation rate. The latter field of application will prove *
Corresponding author. Tel.: +49 40 42838 6595; fax: +49 40 42838 2024. E-mail address: [email protected] (J. Gebert). 0956-053X/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2005.01.022
increasingly relevant when the EC landfill directive (1999/31) comes into effect, which stipulates that only material of low biological activity may be deposited. In all the cases mentioned, the gas extraction system will have been shut down or not yet been put into operation, so that landfill gas emission is passive. The passive emission of landfill gas in general follows the pressure gradient between landfill and atmosphere and is thus influenced by atmospheric pressure dynamics. The relevance of the meteorological condition on LFG emission has been observed on site by various authors (Galle et al., 2001; Christophersen et al., 2001; Lewitz, 2001; Czepiel et al., 2003). Also, barometric pressure was used as a key variable in gas flux models developed by Young (1990, 1992), Nastev et al. (2001) and Poulsem et al. (2003). All of the studies mentioned consider LFG emissions across the landfill surface. However, the sensitivity of gas fluxes to barometric pressure may be much higher in the landfill venting system itself. Gas treatment plants, in this case a biofilter, connected to the venting system of passively operated landfills may therefore
246
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
receive highly variable loads of methane and oxygen. This means that the conditions for optimum operation may be highly variable. The study therefore aimed at analysing landfill gas emission behaviour of the passively vented harbour sludge disposal site in Hamburg, Northern Germany, in order to gather information on principal biofilter operational conditions.
2. Material and methods Fig. 1 schematically illustrates the arrangement of gas wells, the emission monitoring unit, and the biofilter, as operated from October 1999 to date. Landfill gas from two gas wells venting different sections of the landfill (1, 2) was combined (3) and then split up again to be fed to the two biofilter chambers (4, 5). The biofilter is an upflow system consisting of five layers (top to base) inside a 15 m3 polyethylene container with an inclined base: A = topsoil (10 cm), covered with grass vegetation, B = sand (1.5 cm), C = gravel (1.5 cm), D = crushed porous clay (67 cm)
and E = gravel for water drainage (10–30 cm). The biofilter container is subdivided into chamber 1 at a size of 6 m3 and chamber 2 at a size of 9 m3, which can be operated independently. Crude landfill gas is distributed via horizontal and parallel slotted gas supply pipes embedded in the top centimeters of the drainage gravel layer (E). The individual gas supply pipes are 20 cm apart. Oxygen is supplied by diffusion from the atmosphere via the biofilter surface. As the biofilter is integrated into the landfill cover without supplementary heating, biofilter temperature follows ambient temperature. Filter material humidity solely relies on precipitation and landfill gas humidity. More details on the filter material are given in Gebert et al. (2003). The cover system of the landfill itself consists of a recultivation layer (1.5 m thick) above a drainage layer (1 m) and a mineral barrier (1.5 m, processed and compacted harbour sludge, saturated hydraulic conductivity = 1.2 · 109 m s1). A detailed description of the cover system is given by Tresselt et al. (1998). The measuring section set up to characterise landfill gas emission behaviour comprised sensors for the
Fig. 1. Gas well and biofiltration unit setup at the harbour sludge disposal site Francop. 1–5, A–E: see text.
Table 1 Range, error, resolution and logging frequency of sensors used to characterise landfill gas emission Parameter
Sensor
Range
Error
Resolution
Logging frequency
LFG pressure
XCX 0.3 DN, SENSPECIAL Co.
20 to +20 hPa
±0.5% of range
0.001 hPa
10 min
LFG and atm. temperature
PT100, DRIESSEN UND KERN Co.
10 to +80 C
±0.3 C
0.01 C
1h
1
1
LFG flowrate
TSI 8475–075
0.05–2.5 m s
±1.6%
0.001 m s
10 min
LFG composition
NDIR-gas analyser BE-4000, BERNT Co.
CH4: 0–100 vol% CO2: 0–100 vol% O2: 0–25 vol%
CH4, CO2: ±2% of range
0.01 vol%
10 min
PTB 101 C, VAISALA Co.
900–1100 hPa
±0.25 hPa
1 hPa
10 min
Atm. pressure
O2:±1% of range
Landfill gas (LFG) pressure, temperature, flow rate and composition were measured in Section 3 of the emission monitoring unit (see Fig. 1). Sensor errors are given as declared by the manufacturers.
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
registration of temperature, pressure against the atmosphere, composition (CH4, CO2, O2) and flow rate of landfill gas. Additionally, atmospheric temperature (1 m above ground in the shade) and pressure (at the biofilter surface) were recorded (see Table 1).
3. Results and discussion 3.1. Landfill gas emission Fig. 2 exemplarily shows how the course of atmospheric pressure at the biofilter surface, landfill gas pressure, flow rate and composition relate to each other, as measured in Section 3 of the biofilter gas supply system during May 2000 while operating both biofilter chambers. Pressure differences in the gas supply pipe (‘‘LFG pressure against the atmosphere’’) were small at 0.8 to 1.6 hPa during the entire period of operation. They alternated between overpressure and underpressure compared to atmospheric pressure at the biofilter surface. Overpressure corresponded to a decrease of atmospheric pressure and caused the emission of gas from the landfill via the biofilter into the atmosphere (positive values for ‘‘LFG flow rate’’). Conversely, underpressure in the gas supply system corresponded to phases of rising atmospheric pressure, drawing atmospheric air via the biofilter into the landfill (negative values for ‘‘LFG flow rate’’), thus regularly aerating the landfill body. This process was also modelled by Nastev et al. (2001). Overall, landfill gas pressure was very low with a maximum value of 1.6 hPa measured over the entire period of operation. Landfill surface methane emissions were never observed over several years of emission monitoring using static chambers. However, analysis of vertical gas distribution in the recultivation layer showed that gas is in
Fig. 2. Atmospheric pressure, landfill gas pressure, flow rate and composition in May 2000 as measured in Section 3 of the gas supply system (see Fig. 1): LFG = landfill gas; D = difference.
247
principle released into the cover system, at individual sampling spots resulting in CH4-concentrations of near 40 vol% (Gebert and Groengroeft, 2004). Methane oxidation potential of the recultivation layer was 3 g CH4 h1 m2 (at 20 C), as calculated from laboratory activity tests. This is in the range of values given by other authors for the methane oxidation potential of landfill covers (0.3–16 g CH4 h1 m2; Whalen et al., 1990; Jones and Nedwell, 1993; Kightley et al., 1995; Humer and Lechner, 2000). However, it could not be determined if the methane detected in the cover layer originated solely from the top compacted sludge barrier, which supposedly is impermeable to gas ascending from below due to its compaction and water saturation, or if the cover system received substantial amounts of gas from the landfill body which would thus explain the small pressures observed in the venting system. The regular reversal of gas fluxes was clearly reflected by gas composition as measured in the gas supply pipe (Fig. 2, ‘‘LFG composition’’): with decreasing atmospheric pressure, subsequent increasing landfill gas pressure and positive gas fluxes, the methane content increases steadily (maximum concentration in the years of 2000–2002: 56 vol%) while the oxygen content drops to zero. As the direction of flux changes, methane concentration in the supply pipe rapidly falls to zero whereas oxygen reaches atmospheric concentration. Hardly any phases of constant and high positive fluxes and constant methane concentrations were observed. Gas emission usually collapsed after the steady increase in methane content had reached its maximum. This indicated that emission is only driven by a momentary pressure buildup inside the landfill, which is not sustained over longer periods of time. In the period shown in Fig. 2, 14 changes in gas flux direction occurred. Atmospheric pressure at the earthÕs surface may generally vary on different scales due to the following processes: (A) autooscillation of air, (B) warming and cooling of air caused by the daily alternation of solarisation, and (C) passage of atmospheric pressure highs and lows. The associated variations in atmospheric pressure occur on respective pressure- and time-scales and induce corresponding changes in landfill gas pressure (Fig. 3). The autooscillation of air (A), detected at a resolution of 10 min, induces atmospheric pressure variations of less than 1 hPa, which always underly the greater pressure variations. However, there may well exist even smaller variations in pressure on a smaller time scale which would need a higher measurement resolution in order to be detected. Graph B indicates that air temperature and atmospheric pressure vary diurnally. During a stable general weather condition, changes in atmospheric pressure are caused by the daily periodic warming and cooling of the upper surface of the earth. Rising temperature results in a ground level pressure drop due to the rise of the warming air masses whereas falling
248
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
Fig. 3 illustrates that the pressure in the biofilter gas supply pipe reacts to atmospheric pressure changes on all scales. The reaction is independent from the actual gas flow direction and therefore influences both gas emission and air influx scenarios. Graph C in particular indicates that landfill gas pressure (against the atmosphere) is a function of the gradient of atmospheric pressure and not of its absolute value. This interrelation has also been found by Young (1990) who modelled the regulation of gas emission by a fictitious landfill. Simulation results showed that gas emission variations are transient effects as a reaction to rising or falling atmospheric pressure. On site, Christophersen et al. (2001) also found that LFG fluxes correlated best with the pressure gradient (inversely) and only weakly to absolute barometric pressure. This is also illustrated in Fig. 4, depicting the inversely proportional relationship between the change of atmospheric pressure (Datm. pressure) and the change of landfill gas pressure (DLFG pressure). However, the results stand in contrast to the findings of Czepiel et al. (2003). They described that whole-landfill methane emission as measured by an atmospheric tracer method inversely correlated with absolute values of surface atmospheric pressure. An actual reversal of the gas flux as found here was never observed in the field studies mentioned above, but only modelled by Nastev et al. (2001). The extent of advective gas emission sensitivity to barometric pressure changes in general depends on the degree of landfill sealing, including the effectivity and permeability of the gas collection and the connected gas treatment system. Furthermore, it varies with changing water free pore space of the recultivation layer as a result of precipitation, frost or soil aggregation. For example, Christophersen et al. (2001) could demonstrate that the moisture content of a landfill cover soil accounted for the largest part of the gas flux variation. The influence of atmospheric pressure on landfill gas emission behaviour may therefore vary significantly
temperature causes a rise in pressure as a result of sinking of the cooling air. The diurnal pressure changes induced by solarisation may amount to around 4 hPa. A week of unstable weather condition is depicted in graph C. The influence of atmospheric pressure highs and lows can last for several days and may cause pressure variations between 990 and 1030 hPa. As graph C indicates, there is a short lag time between atmospheric and landfill gas pressure. This is due to the fact that when atmospheric pressure stagnates after a decrease and before a new increase landfill gas pressure will not increase further but decrease again.
0.3 LFG pressure [∆hPa h-1]
Fig. 3. Scales of atmospheric pressure variations and corresponding course of lfg pressure as induced by autooscillation of air (A), diurnal variation (B), and atmospheric pressure highs and lows (C): Atm. = atmospheric; LFG = landfill gas; D = difference.
0.2 0.1 0.0 -0.1 -0.2
-4
Y = 8.6 x 10 - 0.027 X 2 R = 0.47
-0.3 -5 -4 -3 -2 -1 0 1 2 3 4 -1 Atm. Pressure [∆ hPa h ]
5
Fig. 4. Relation between change of atmospheric pressure and landfill gas pressure: Atm. = atmospheric; LFG = landfill gas; D = difference, n = 500.
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
-4 -1.0
-0.5 0.0 0.5 1.0 LFG pressure [∆hPa]
1.5
Fig. 5. Gas flow rate as a function of landfill gas pressure: LFG = landfill gas; D = difference, n = 705.
from site to site. For the investigated landfill it could be observed at one occasion during winter that frost in the biofilter cover layer led to an increase of venting system pressure up to 40 hPa. The effects of individual precipitation events on gas emission through the biofilter were not monitored. A substantial decrease of gas permeability due to water filled pore space, however, can be excluded as the predominant share of filter volume is made up of very coarse and therefore well-draining material (see Fig. 1). Fig. 5 demonstrates how pressure in the gas supply pipe (LFG pressure) relates to the gas flow rate. As can be deduced from the BoyleÕs law, gas flow rate correlates positively with gas pressure: the volume exchanged via gas emission (positive value) or air influx (negative values) increases with increasing overpressure/underpressure in the gas collection system. However, the relation given in Fig. 5 is sigmoid, meaning that at higher (absolute) landfill gas pressures the gas flow rate remains below the values expected for a proportional relationship. Presuming accurate sensor operation, this implies that biofilter permeability confines the gas volume exchangeable per time unit. This is further substantiated by the difference in the mean landfill gas flow rate observed when operating different biofilter volumes, as shown in Table 2. The mean landfill gas flow rate in phase 2 exceeds that in phase 3 by a factor of 2.5. This is exactly equivalent to the relation between the biofilter volumes operated during the two phases (15 m3 in phase 2 and 6 m3 in
1 2 3
Duration 05/00–09/00 10/00–05/01 06/01–05/02
200
80
150
60
100
Frequency Cumul. frequency Cumul. CH4 load
50 0
0
250
20 40 60 80 100 120 140 160 Duration of lfg emission [h]
Mean LFG flow rate (m3 h1)
FRSTD
6+x 15 6
1.08 1.87 0.74
1.46 2.5 1
40 20 0
100
B
200
80
150 Frequency Cumul. frequency
100
0
Biofilter volume operated (m3)
FRSTD = flow rate standardised to mean flowrate in phase 3.
100
A
50
Table 2 Mean LFG flowrate during three phases of biofilter operation Phase
250
60 40 20
0
20 40 60 80 100 120 140 160
Cumulative frequency/CH4 load [%]
-2
Cumulative frequency [%]
0
Frequency
2
phase 3). In phase 1, only chamber 1 (6 m3) was fed with landfill gas; however, due to a leak in the water drainage system, gas could to some extent evade into chamber 2 as well. This is reflected by a higher flow rate than in phase 3, when chamber 1 was operated correctly. From the data for atmospheric pressure and gas flow rate, it was calculated that for the given setup of gas collection and biofiltration unit at the investigated landfill an atmospheric pressure change of 10 hPa induces a gas exchange of around 4 m3 per m3 of biofilter. As Fig. 6 shows, the frequency distribution of emission (plot A) and air influx phase (plot B) duration is very similar. In approximately 50% of all cases, the individual phases last for only 10 h or less. Only 20% of all phases are longer than 30 h. Gas flux reverses on average every 20 h, thus more than once a day. The shortest phase registered was 10 min, both for gas emission and air influx, corresponding to the resolution of measurement. With the reversal of gas flow direction, the biofilter microorganisms are regularly deprived of their energyand carbon-source methane, sometimes for up to a week. Laboratory experiments with the filter material, however, showed that the deprivation of methane on the time scale observed here does not impair methane oxidation activity in the biofilter (Gebert et al., 2003). Depending on the incubation temperature, it took up to five weeks of ‘‘starving’’ conditions to reduce methane
Frequency
Gas flow rate [m3 h-1]
4
249
0
Duration of air influx [h]
Fig. 6. Frequency distribution and cumulative curve of landfill gas emission (A) and atmospheric air influx (B) and cumulative curve of methane load in the period 11.05.2000–07.08.2002: LFG = landfill gas, n(A) = 475, n(B) = 449.
autooscillation of air
∆temperature
∆general weather situation
∆atmospheric pressure + LFG pressure gas emission
- LFG pressure air influx
1400
100
1200
80
1000 60
800 600
frequency cumul. frequency cumul. CH4 load
400 200
[CH4]↑ [O2] ↓
[CH4]↓ [O2]↑
Fig. 7. Conceptual diagram on the regulation on landfill gas emission on the investigated landfill site: D = difference, LFG = landfill gas, [CH4], ½O2 ¼ CH 4 and O2-concentration in landfill gas, › = increasing, fl = decreasing.
oxidation activity and even after half a year without substrate significant methane oxidation activity could still be detected and quickly restored to a high level after addition of methane. Fig. 6A also includes a percentage cumulative curve for the loads of methane emitted during the individual emission phases. These were calculated by partitioning the total mass methane emitted during the period 11.05.2000–07.08.2002 over the individual emission episodes. As the curve indicates, short emission phases of 10 h or less, which constitute 50% of all cases, carry less than 5% of the total methane load emitted by the landfill. However, emission phases of more than 30 h duration, which represent only 20% of all cases and 33% of the time, transport more than 80% of the total methane load. Fig. 7 gives the regulatory mechanism for landfill gas flow rate and composition hypothesized for the investigated biofilter/landfill site, considering the described connections between atmospheric temperature, atmospheric pressure, landfill gas pressure and gas flow rate. As a result of the delicate dependency of landfill gas emission on atmospheric pressure variation, the biofilter connected to the gas collection system receives highly variable loads of methane, as illustrated Fig. 8. For reasons of data handling practicability, not all data collected (10 min resolution), but only every 12th data set was included in the frequency distribution analysis (2 h resolution). The data presented in Fig. 8 is also included in Fig. 6A. During emission phases, the biofilter receives a specific methane load between close to zero and 247 g h1 m3 with an arithmetic mean of 19 g h1 m3 and a median of 9.5 g h1 m3. Of the total methane load, 50% is emitted by pulses of less than 33 g h1 m3 and only 10% by pulses greater than 107 g h1 m3. In combination with the other essential parameters of biofilter operation (temperature, oxygen supply), the frequency distribution of methane input provides a useful data basis for biofilter dimensioning. As the discrepancy
0
0
20 40 60 80 100 120 140 160 -1
-3
specific methane load [g h m ]
40 20 0
cumulative frequency/CH4 load [%]
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
frequency
250
Fig. 8. Frequency distribution and cumulative curve of frequency and specific biofilter methane load in the period 02.05.2000–07.08.2002, n = 4751.
between the cumulative curves of frequency and methane load shows, biofilter layout with respect to desired methane reduction rates depends on the choice of either time or total methane load as reference variable.
4. Conclusions Driven by atmospheric pressure variation the direction of advective gas flow reverses on average more than once a day at the investigated passively vented landfill site. The reaction of gas emission to atmospheric pressure changes even below 1 hPa is immediate and highly sensitive, resulting in a very high variability of flow rate (in both directions) and methane concentration of landfill gas. Correspondingly, the biofilter connected to the landfill venting system receives very variable loads of methane. Gradients in atmospheric pressure as induced by the autooscillation of air are usually too small to result in a reversion of flux direction, also due to the lag volume given by the biofilter. Diurnal variations causing atmospheric pressure changes of up to 4 hPa may determine gas flow direction but are often superimposed by longer pressure fluctations associated with atmospheric highs and lows. Although, the porosity of the filter material is very high with a gas permeable volume in the porous clay layer of 78%, the biofilter was shown to confine gas exchange between the landfill and the atmosphere. Gas flow rate and emission phase duration not only govern methane supply to the methanotrophic bacteria but also determine the degree of counterdirective diffusive oxygen supply via the filter surface. Emission dynamics therefore strongly influence vertical gas distribution in the biofilter and thus conditions for successful operation. A significant share of the total methane load is emitted in phases of longer duration. These, however, correspond with higher methane concentrations and lower oxygen concentrations across the filter profile,
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
thus causing suboptimal conditions for methane oxidation. On the other hand, the regular reversion of fluxes leads to an aeration of the biofilter profile and may even be indispensable for filter operation as observed here. For adequate dimensioning of connected biofilters without further regulation of input flow, the range of methane input variability as well as the specific gas flow rate determining diffusive oxygen supply must therefore be preestimated. Aeration of the landfill body as a result of regular flux reversal may enable substantial amounts of methane to be oxidised prior to emission. Using the O2-content of the landfill gas in relation to its N2-content, it was roughly estimated that atmospheric air regularly introduced into the landfill body on average may enable oxidation of 46% of the amount of methane that would otherwise have been emitted via the venting system (data not shown). The extent of this welcome effect of passive landfill ventilation in relation to the overall amount of gas produced in general depends on the effectiveness of the gas collection system, i.e., the degree of gas retrieval. As methanogens are strictly anaerobic organisms, substantial input of oxygen may on the other hand impact the extent of gas generation inside the landfill. However, as the deposited harbour sludge layers are 1.5 m thick, almost water-saturated and compacted, it is doubtful whether the oxygen introduced by gas flow reversal will expand substantially into the methane-forming sites. The issue as such was not addressed in the investigations. Acknowledgements The authors wish to thank the German Federal Ministry for Education and Research (BMBF) for funding and the Hamburg Department of Port and River Engineering for facilitating the work presented in this paper.
References Bowman, J., 2000. The Methanotrophs – The families Methylococcaceae and Methylocystaceae. In: The Prokaryotes: An Evolving electronic resource for the microbiological community, third ed. Release 3.1, 1/20/2000. Springer, New York. Available from:
251
emissions and methane oxidation. Waste Management and Research 19, 595–612. Czepiel, P.M., Shorter, J.H., Mosher, B., Allwine, E., McManus, J.B., Harriss, R.C., Kolb, C.E., Lamb, B.K., 2003. The influence of atmospheric pressure on landfill methane emissions. Waste Management 23 (7), 593–598. Figueroa, R.A., 1996. Landfill gas treatmant by biofilters. In: Christensen, T.H., Cossu, R., Stegmann, R. (Eds.), Landfilling of Waste: Biogas. E & FN Spon, pp. 535–559. Galle, B., Samuelsson, J., Svensson, B.H., Bo¨rjesson, G., 2001. Measurements of methane emissions from landfills using a time correlation tracer method based on FTIR absorption spectroscopy. Environmental Science and Technology 35, 21– 25. Gebert, J., Groengroeft, A., Miehlich, G., 2003. Kinetics of microbial landfill methane oxidation in biofilters. Waste Management 23, 609–619. Gebert, J., Groengroeft, A., 2004. Gas formation and management in Francop. Technical report to the Department of Port and River Engineering of the City of Hamburg (in German). Hanson, R.S., Hanson, T.E., 1996. Methanotrophic bacteria. FEMS Microbiology Reviews 60, 439–471. Humer, M., Lechner, P., 2000. Methane oxidation in optimised landfill cover layers under different seasonal conditions. In: Proceedings 1st Intercontinental Landfill Research Symposia, Lulea. Jones, H.A., Nedwell, D.B., 1993. Methane emission and methane oxidation in landfill cover soil. FEMS Microbiological Ecology 102, 185–195. Kightley, D., Nedwell, D.B., Cooper, M., 1995. Capacity for methane oxidation in landfill cover soils measured in laboratory-scale soil microcosms. Applied and Environmental Microbiology 61, 592– 601. Lewitz, H., 2001. Beobachtungen und Ru¨ckschlu¨sse im Zusammenhang mit dem Gashaushalt der Deponie Georgswerder. TerraTech 6, 47–51 (in German). Nastev, M., Therrien, R., Lefebvre, R., Ge´linas, P., 2001. Gas production and migration in landfills and geological materials. Journal of Contaminant Hydrology 1095, 187–211. Poulsem, T.G., Christophersen, M., Moldrup, P., Kjeldsen, P., 2003. Relating landfill gas emissions to atmospheric pressure using numerical modelling and state-space analysis. Waste Management and Research 21, 356–366. Streese, J., Stegmann, R., 2003. Microbial oxidation of methane from old landfills in biofilters. Waste Management 23, 573–580. Tresselt, K., Miehlich, G., Groengroeft, A., Melchior, S., Berger, K., Harms, C., 1998. Harbour sludge as barrier material in landfill cover systems. Water Science and Technology 37, 307– 313. Whalen, S.C., Reeburgh, W.S., Sandbeck, K.A., 1990. Rapid methane oxidation in a landfill cover soil. Applied and Environmental Microbiology 56, 3405–3411. Young, A., 1990. Volumetric changes in landfill gas flux in response to variations in atmospheric pressure. Waste Management and Research 8, 379–385. Young, A., 1992. The effect of fluctuations in atmospheric pressure on landfill gas migration and composition. Water, Air and Soil Pollution 64, 604–616.
Passive landfill gas emission – Influence of atmospheric pressure and implications for the operation of methane-oxidising biofilters Julia Gebert *, Alexander Groengroeft University of Hamburg, Institute of Soil Science, Allende-Platz 2, 20146 Hamburg, Germany Accepted 19 January 2005 Available online 23 March 2005
Abstract A passively vented landfill site in Northern Germany was monitored for gas emission dynamics through high resolution measurements of landfill gas pressure, flow rate and composition as well as atmospheric pressure and temperature. Landfill gas emission could be directly related to atmospheric pressure changes on all scales as induced by the autooscillation of air, diurnal variations and the passage of pressure highs and lows. Gas flux reversed every 20 h on average, with 50% of emission phases lasting only 10 h or less. During gas emission phases, methane loads fed to a connected methane oxidising biofiltration unit varied between near zero and 247 g CH4 h1 m3 filter material. Emission dynamics not only influenced the amount of methane fed to the biofilter but also the establishment of gas composition profiles within the biofilter, thus being of high relevance for biofilter operation. The duration of the gas emission phase emerged as most significant variable for the distribution of landfill gas components within the biofilter. 2005 Elsevier Ltd. All rights reserved.
1. Introduction The microbial oxidation of methane in biofilters is considered an alternative to the treatment of landfill gas (LFG) methane emissions that do not meet gas flow rate and methane content requirements for utilization or flaring (Figueroa, 1996; Gebert et al., 2003; Streese and Stegmann, 2003). Methane is aerobically oxidised to carbon dioxide, thereby serving as the sole carbonand energy-source for the methane oxidizing bacteria. Reviews of the taxonomy and physiology of methane oxidising bacteria are given by Hanson and Hanson (1996) and Bowman (2000). Biofiltration as a means for the abatement of anthropogenic methane emissions may be applied at landfills in the initial phase of operation, old landfills or sites containing material of low gas generation rate. The latter field of application will prove *
Corresponding author. Tel.: +49 40 42838 6595; fax: +49 40 42838 2024. E-mail address: [email protected] (J. Gebert). 0956-053X/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2005.01.022
increasingly relevant when the EC landfill directive (1999/31) comes into effect, which stipulates that only material of low biological activity may be deposited. In all the cases mentioned, the gas extraction system will have been shut down or not yet been put into operation, so that landfill gas emission is passive. The passive emission of landfill gas in general follows the pressure gradient between landfill and atmosphere and is thus influenced by atmospheric pressure dynamics. The relevance of the meteorological condition on LFG emission has been observed on site by various authors (Galle et al., 2001; Christophersen et al., 2001; Lewitz, 2001; Czepiel et al., 2003). Also, barometric pressure was used as a key variable in gas flux models developed by Young (1990, 1992), Nastev et al. (2001) and Poulsem et al. (2003). All of the studies mentioned consider LFG emissions across the landfill surface. However, the sensitivity of gas fluxes to barometric pressure may be much higher in the landfill venting system itself. Gas treatment plants, in this case a biofilter, connected to the venting system of passively operated landfills may therefore
246
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
receive highly variable loads of methane and oxygen. This means that the conditions for optimum operation may be highly variable. The study therefore aimed at analysing landfill gas emission behaviour of the passively vented harbour sludge disposal site in Hamburg, Northern Germany, in order to gather information on principal biofilter operational conditions.
2. Material and methods Fig. 1 schematically illustrates the arrangement of gas wells, the emission monitoring unit, and the biofilter, as operated from October 1999 to date. Landfill gas from two gas wells venting different sections of the landfill (1, 2) was combined (3) and then split up again to be fed to the two biofilter chambers (4, 5). The biofilter is an upflow system consisting of five layers (top to base) inside a 15 m3 polyethylene container with an inclined base: A = topsoil (10 cm), covered with grass vegetation, B = sand (1.5 cm), C = gravel (1.5 cm), D = crushed porous clay (67 cm)
and E = gravel for water drainage (10–30 cm). The biofilter container is subdivided into chamber 1 at a size of 6 m3 and chamber 2 at a size of 9 m3, which can be operated independently. Crude landfill gas is distributed via horizontal and parallel slotted gas supply pipes embedded in the top centimeters of the drainage gravel layer (E). The individual gas supply pipes are 20 cm apart. Oxygen is supplied by diffusion from the atmosphere via the biofilter surface. As the biofilter is integrated into the landfill cover without supplementary heating, biofilter temperature follows ambient temperature. Filter material humidity solely relies on precipitation and landfill gas humidity. More details on the filter material are given in Gebert et al. (2003). The cover system of the landfill itself consists of a recultivation layer (1.5 m thick) above a drainage layer (1 m) and a mineral barrier (1.5 m, processed and compacted harbour sludge, saturated hydraulic conductivity = 1.2 · 109 m s1). A detailed description of the cover system is given by Tresselt et al. (1998). The measuring section set up to characterise landfill gas emission behaviour comprised sensors for the
Fig. 1. Gas well and biofiltration unit setup at the harbour sludge disposal site Francop. 1–5, A–E: see text.
Table 1 Range, error, resolution and logging frequency of sensors used to characterise landfill gas emission Parameter
Sensor
Range
Error
Resolution
Logging frequency
LFG pressure
XCX 0.3 DN, SENSPECIAL Co.
20 to +20 hPa
±0.5% of range
0.001 hPa
10 min
LFG and atm. temperature
PT100, DRIESSEN UND KERN Co.
10 to +80 C
±0.3 C
0.01 C
1h
1
1
LFG flowrate
TSI 8475–075
0.05–2.5 m s
±1.6%
0.001 m s
10 min
LFG composition
NDIR-gas analyser BE-4000, BERNT Co.
CH4: 0–100 vol% CO2: 0–100 vol% O2: 0–25 vol%
CH4, CO2: ±2% of range
0.01 vol%
10 min
PTB 101 C, VAISALA Co.
900–1100 hPa
±0.25 hPa
1 hPa
10 min
Atm. pressure
O2:±1% of range
Landfill gas (LFG) pressure, temperature, flow rate and composition were measured in Section 3 of the emission monitoring unit (see Fig. 1). Sensor errors are given as declared by the manufacturers.
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
registration of temperature, pressure against the atmosphere, composition (CH4, CO2, O2) and flow rate of landfill gas. Additionally, atmospheric temperature (1 m above ground in the shade) and pressure (at the biofilter surface) were recorded (see Table 1).
3. Results and discussion 3.1. Landfill gas emission Fig. 2 exemplarily shows how the course of atmospheric pressure at the biofilter surface, landfill gas pressure, flow rate and composition relate to each other, as measured in Section 3 of the biofilter gas supply system during May 2000 while operating both biofilter chambers. Pressure differences in the gas supply pipe (‘‘LFG pressure against the atmosphere’’) were small at 0.8 to 1.6 hPa during the entire period of operation. They alternated between overpressure and underpressure compared to atmospheric pressure at the biofilter surface. Overpressure corresponded to a decrease of atmospheric pressure and caused the emission of gas from the landfill via the biofilter into the atmosphere (positive values for ‘‘LFG flow rate’’). Conversely, underpressure in the gas supply system corresponded to phases of rising atmospheric pressure, drawing atmospheric air via the biofilter into the landfill (negative values for ‘‘LFG flow rate’’), thus regularly aerating the landfill body. This process was also modelled by Nastev et al. (2001). Overall, landfill gas pressure was very low with a maximum value of 1.6 hPa measured over the entire period of operation. Landfill surface methane emissions were never observed over several years of emission monitoring using static chambers. However, analysis of vertical gas distribution in the recultivation layer showed that gas is in
Fig. 2. Atmospheric pressure, landfill gas pressure, flow rate and composition in May 2000 as measured in Section 3 of the gas supply system (see Fig. 1): LFG = landfill gas; D = difference.
247
principle released into the cover system, at individual sampling spots resulting in CH4-concentrations of near 40 vol% (Gebert and Groengroeft, 2004). Methane oxidation potential of the recultivation layer was 3 g CH4 h1 m2 (at 20 C), as calculated from laboratory activity tests. This is in the range of values given by other authors for the methane oxidation potential of landfill covers (0.3–16 g CH4 h1 m2; Whalen et al., 1990; Jones and Nedwell, 1993; Kightley et al., 1995; Humer and Lechner, 2000). However, it could not be determined if the methane detected in the cover layer originated solely from the top compacted sludge barrier, which supposedly is impermeable to gas ascending from below due to its compaction and water saturation, or if the cover system received substantial amounts of gas from the landfill body which would thus explain the small pressures observed in the venting system. The regular reversal of gas fluxes was clearly reflected by gas composition as measured in the gas supply pipe (Fig. 2, ‘‘LFG composition’’): with decreasing atmospheric pressure, subsequent increasing landfill gas pressure and positive gas fluxes, the methane content increases steadily (maximum concentration in the years of 2000–2002: 56 vol%) while the oxygen content drops to zero. As the direction of flux changes, methane concentration in the supply pipe rapidly falls to zero whereas oxygen reaches atmospheric concentration. Hardly any phases of constant and high positive fluxes and constant methane concentrations were observed. Gas emission usually collapsed after the steady increase in methane content had reached its maximum. This indicated that emission is only driven by a momentary pressure buildup inside the landfill, which is not sustained over longer periods of time. In the period shown in Fig. 2, 14 changes in gas flux direction occurred. Atmospheric pressure at the earthÕs surface may generally vary on different scales due to the following processes: (A) autooscillation of air, (B) warming and cooling of air caused by the daily alternation of solarisation, and (C) passage of atmospheric pressure highs and lows. The associated variations in atmospheric pressure occur on respective pressure- and time-scales and induce corresponding changes in landfill gas pressure (Fig. 3). The autooscillation of air (A), detected at a resolution of 10 min, induces atmospheric pressure variations of less than 1 hPa, which always underly the greater pressure variations. However, there may well exist even smaller variations in pressure on a smaller time scale which would need a higher measurement resolution in order to be detected. Graph B indicates that air temperature and atmospheric pressure vary diurnally. During a stable general weather condition, changes in atmospheric pressure are caused by the daily periodic warming and cooling of the upper surface of the earth. Rising temperature results in a ground level pressure drop due to the rise of the warming air masses whereas falling
248
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
Fig. 3 illustrates that the pressure in the biofilter gas supply pipe reacts to atmospheric pressure changes on all scales. The reaction is independent from the actual gas flow direction and therefore influences both gas emission and air influx scenarios. Graph C in particular indicates that landfill gas pressure (against the atmosphere) is a function of the gradient of atmospheric pressure and not of its absolute value. This interrelation has also been found by Young (1990) who modelled the regulation of gas emission by a fictitious landfill. Simulation results showed that gas emission variations are transient effects as a reaction to rising or falling atmospheric pressure. On site, Christophersen et al. (2001) also found that LFG fluxes correlated best with the pressure gradient (inversely) and only weakly to absolute barometric pressure. This is also illustrated in Fig. 4, depicting the inversely proportional relationship between the change of atmospheric pressure (Datm. pressure) and the change of landfill gas pressure (DLFG pressure). However, the results stand in contrast to the findings of Czepiel et al. (2003). They described that whole-landfill methane emission as measured by an atmospheric tracer method inversely correlated with absolute values of surface atmospheric pressure. An actual reversal of the gas flux as found here was never observed in the field studies mentioned above, but only modelled by Nastev et al. (2001). The extent of advective gas emission sensitivity to barometric pressure changes in general depends on the degree of landfill sealing, including the effectivity and permeability of the gas collection and the connected gas treatment system. Furthermore, it varies with changing water free pore space of the recultivation layer as a result of precipitation, frost or soil aggregation. For example, Christophersen et al. (2001) could demonstrate that the moisture content of a landfill cover soil accounted for the largest part of the gas flux variation. The influence of atmospheric pressure on landfill gas emission behaviour may therefore vary significantly
temperature causes a rise in pressure as a result of sinking of the cooling air. The diurnal pressure changes induced by solarisation may amount to around 4 hPa. A week of unstable weather condition is depicted in graph C. The influence of atmospheric pressure highs and lows can last for several days and may cause pressure variations between 990 and 1030 hPa. As graph C indicates, there is a short lag time between atmospheric and landfill gas pressure. This is due to the fact that when atmospheric pressure stagnates after a decrease and before a new increase landfill gas pressure will not increase further but decrease again.
0.3 LFG pressure [∆hPa h-1]
Fig. 3. Scales of atmospheric pressure variations and corresponding course of lfg pressure as induced by autooscillation of air (A), diurnal variation (B), and atmospheric pressure highs and lows (C): Atm. = atmospheric; LFG = landfill gas; D = difference.
0.2 0.1 0.0 -0.1 -0.2
-4
Y = 8.6 x 10 - 0.027 X 2 R = 0.47
-0.3 -5 -4 -3 -2 -1 0 1 2 3 4 -1 Atm. Pressure [∆ hPa h ]
5
Fig. 4. Relation between change of atmospheric pressure and landfill gas pressure: Atm. = atmospheric; LFG = landfill gas; D = difference, n = 500.
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
-4 -1.0
-0.5 0.0 0.5 1.0 LFG pressure [∆hPa]
1.5
Fig. 5. Gas flow rate as a function of landfill gas pressure: LFG = landfill gas; D = difference, n = 705.
from site to site. For the investigated landfill it could be observed at one occasion during winter that frost in the biofilter cover layer led to an increase of venting system pressure up to 40 hPa. The effects of individual precipitation events on gas emission through the biofilter were not monitored. A substantial decrease of gas permeability due to water filled pore space, however, can be excluded as the predominant share of filter volume is made up of very coarse and therefore well-draining material (see Fig. 1). Fig. 5 demonstrates how pressure in the gas supply pipe (LFG pressure) relates to the gas flow rate. As can be deduced from the BoyleÕs law, gas flow rate correlates positively with gas pressure: the volume exchanged via gas emission (positive value) or air influx (negative values) increases with increasing overpressure/underpressure in the gas collection system. However, the relation given in Fig. 5 is sigmoid, meaning that at higher (absolute) landfill gas pressures the gas flow rate remains below the values expected for a proportional relationship. Presuming accurate sensor operation, this implies that biofilter permeability confines the gas volume exchangeable per time unit. This is further substantiated by the difference in the mean landfill gas flow rate observed when operating different biofilter volumes, as shown in Table 2. The mean landfill gas flow rate in phase 2 exceeds that in phase 3 by a factor of 2.5. This is exactly equivalent to the relation between the biofilter volumes operated during the two phases (15 m3 in phase 2 and 6 m3 in
1 2 3
Duration 05/00–09/00 10/00–05/01 06/01–05/02
200
80
150
60
100
Frequency Cumul. frequency Cumul. CH4 load
50 0
0
250
20 40 60 80 100 120 140 160 Duration of lfg emission [h]
Mean LFG flow rate (m3 h1)
FRSTD
6+x 15 6
1.08 1.87 0.74
1.46 2.5 1
40 20 0
100
B
200
80
150 Frequency Cumul. frequency
100
0
Biofilter volume operated (m3)
FRSTD = flow rate standardised to mean flowrate in phase 3.
100
A
50
Table 2 Mean LFG flowrate during three phases of biofilter operation Phase
250
60 40 20
0
20 40 60 80 100 120 140 160
Cumulative frequency/CH4 load [%]
-2
Cumulative frequency [%]
0
Frequency
2
phase 3). In phase 1, only chamber 1 (6 m3) was fed with landfill gas; however, due to a leak in the water drainage system, gas could to some extent evade into chamber 2 as well. This is reflected by a higher flow rate than in phase 3, when chamber 1 was operated correctly. From the data for atmospheric pressure and gas flow rate, it was calculated that for the given setup of gas collection and biofiltration unit at the investigated landfill an atmospheric pressure change of 10 hPa induces a gas exchange of around 4 m3 per m3 of biofilter. As Fig. 6 shows, the frequency distribution of emission (plot A) and air influx phase (plot B) duration is very similar. In approximately 50% of all cases, the individual phases last for only 10 h or less. Only 20% of all phases are longer than 30 h. Gas flux reverses on average every 20 h, thus more than once a day. The shortest phase registered was 10 min, both for gas emission and air influx, corresponding to the resolution of measurement. With the reversal of gas flow direction, the biofilter microorganisms are regularly deprived of their energyand carbon-source methane, sometimes for up to a week. Laboratory experiments with the filter material, however, showed that the deprivation of methane on the time scale observed here does not impair methane oxidation activity in the biofilter (Gebert et al., 2003). Depending on the incubation temperature, it took up to five weeks of ‘‘starving’’ conditions to reduce methane
Frequency
Gas flow rate [m3 h-1]
4
249
0
Duration of air influx [h]
Fig. 6. Frequency distribution and cumulative curve of landfill gas emission (A) and atmospheric air influx (B) and cumulative curve of methane load in the period 11.05.2000–07.08.2002: LFG = landfill gas, n(A) = 475, n(B) = 449.
autooscillation of air
∆temperature
∆general weather situation
∆atmospheric pressure + LFG pressure gas emission
- LFG pressure air influx
1400
100
1200
80
1000 60
800 600
frequency cumul. frequency cumul. CH4 load
400 200
[CH4]↑ [O2] ↓
[CH4]↓ [O2]↑
Fig. 7. Conceptual diagram on the regulation on landfill gas emission on the investigated landfill site: D = difference, LFG = landfill gas, [CH4], ½O2 ¼ CH 4 and O2-concentration in landfill gas, › = increasing, fl = decreasing.
oxidation activity and even after half a year without substrate significant methane oxidation activity could still be detected and quickly restored to a high level after addition of methane. Fig. 6A also includes a percentage cumulative curve for the loads of methane emitted during the individual emission phases. These were calculated by partitioning the total mass methane emitted during the period 11.05.2000–07.08.2002 over the individual emission episodes. As the curve indicates, short emission phases of 10 h or less, which constitute 50% of all cases, carry less than 5% of the total methane load emitted by the landfill. However, emission phases of more than 30 h duration, which represent only 20% of all cases and 33% of the time, transport more than 80% of the total methane load. Fig. 7 gives the regulatory mechanism for landfill gas flow rate and composition hypothesized for the investigated biofilter/landfill site, considering the described connections between atmospheric temperature, atmospheric pressure, landfill gas pressure and gas flow rate. As a result of the delicate dependency of landfill gas emission on atmospheric pressure variation, the biofilter connected to the gas collection system receives highly variable loads of methane, as illustrated Fig. 8. For reasons of data handling practicability, not all data collected (10 min resolution), but only every 12th data set was included in the frequency distribution analysis (2 h resolution). The data presented in Fig. 8 is also included in Fig. 6A. During emission phases, the biofilter receives a specific methane load between close to zero and 247 g h1 m3 with an arithmetic mean of 19 g h1 m3 and a median of 9.5 g h1 m3. Of the total methane load, 50% is emitted by pulses of less than 33 g h1 m3 and only 10% by pulses greater than 107 g h1 m3. In combination with the other essential parameters of biofilter operation (temperature, oxygen supply), the frequency distribution of methane input provides a useful data basis for biofilter dimensioning. As the discrepancy
0
0
20 40 60 80 100 120 140 160 -1
-3
specific methane load [g h m ]
40 20 0
cumulative frequency/CH4 load [%]
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
frequency
250
Fig. 8. Frequency distribution and cumulative curve of frequency and specific biofilter methane load in the period 02.05.2000–07.08.2002, n = 4751.
between the cumulative curves of frequency and methane load shows, biofilter layout with respect to desired methane reduction rates depends on the choice of either time or total methane load as reference variable.
4. Conclusions Driven by atmospheric pressure variation the direction of advective gas flow reverses on average more than once a day at the investigated passively vented landfill site. The reaction of gas emission to atmospheric pressure changes even below 1 hPa is immediate and highly sensitive, resulting in a very high variability of flow rate (in both directions) and methane concentration of landfill gas. Correspondingly, the biofilter connected to the landfill venting system receives very variable loads of methane. Gradients in atmospheric pressure as induced by the autooscillation of air are usually too small to result in a reversion of flux direction, also due to the lag volume given by the biofilter. Diurnal variations causing atmospheric pressure changes of up to 4 hPa may determine gas flow direction but are often superimposed by longer pressure fluctations associated with atmospheric highs and lows. Although, the porosity of the filter material is very high with a gas permeable volume in the porous clay layer of 78%, the biofilter was shown to confine gas exchange between the landfill and the atmosphere. Gas flow rate and emission phase duration not only govern methane supply to the methanotrophic bacteria but also determine the degree of counterdirective diffusive oxygen supply via the filter surface. Emission dynamics therefore strongly influence vertical gas distribution in the biofilter and thus conditions for successful operation. A significant share of the total methane load is emitted in phases of longer duration. These, however, correspond with higher methane concentrations and lower oxygen concentrations across the filter profile,
J. Gebert, A. Groengroeft / Waste Management 26 (2006) 245–251
thus causing suboptimal conditions for methane oxidation. On the other hand, the regular reversion of fluxes leads to an aeration of the biofilter profile and may even be indispensable for filter operation as observed here. For adequate dimensioning of connected biofilters without further regulation of input flow, the range of methane input variability as well as the specific gas flow rate determining diffusive oxygen supply must therefore be preestimated. Aeration of the landfill body as a result of regular flux reversal may enable substantial amounts of methane to be oxidised prior to emission. Using the O2-content of the landfill gas in relation to its N2-content, it was roughly estimated that atmospheric air regularly introduced into the landfill body on average may enable oxidation of 46% of the amount of methane that would otherwise have been emitted via the venting system (data not shown). The extent of this welcome effect of passive landfill ventilation in relation to the overall amount of gas produced in general depends on the effectiveness of the gas collection system, i.e., the degree of gas retrieval. As methanogens are strictly anaerobic organisms, substantial input of oxygen may on the other hand impact the extent of gas generation inside the landfill. However, as the deposited harbour sludge layers are 1.5 m thick, almost water-saturated and compacted, it is doubtful whether the oxygen introduced by gas flow reversal will expand substantially into the methane-forming sites. The issue as such was not addressed in the investigations. Acknowledgements The authors wish to thank the German Federal Ministry for Education and Research (BMBF) for funding and the Hamburg Department of Port and River Engineering for facilitating the work presented in this paper.
References Bowman, J., 2000. The Methanotrophs – The families Methylococcaceae and Methylocystaceae. In: The Prokaryotes: An Evolving electronic resource for the microbiological community, third ed. Release 3.1, 1/20/2000. Springer, New York. Available from:
251
emissions and methane oxidation. Waste Management and Research 19, 595–612. Czepiel, P.M., Shorter, J.H., Mosher, B., Allwine, E., McManus, J.B., Harriss, R.C., Kolb, C.E., Lamb, B.K., 2003. The influence of atmospheric pressure on landfill methane emissions. Waste Management 23 (7), 593–598. Figueroa, R.A., 1996. Landfill gas treatmant by biofilters. In: Christensen, T.H., Cossu, R., Stegmann, R. (Eds.), Landfilling of Waste: Biogas. E & FN Spon, pp. 535–559. Galle, B., Samuelsson, J., Svensson, B.H., Bo¨rjesson, G., 2001. Measurements of methane emissions from landfills using a time correlation tracer method based on FTIR absorption spectroscopy. Environmental Science and Technology 35, 21– 25. Gebert, J., Groengroeft, A., Miehlich, G., 2003. Kinetics of microbial landfill methane oxidation in biofilters. Waste Management 23, 609–619. Gebert, J., Groengroeft, A., 2004. Gas formation and management in Francop. Technical report to the Department of Port and River Engineering of the City of Hamburg (in German). Hanson, R.S., Hanson, T.E., 1996. Methanotrophic bacteria. FEMS Microbiology Reviews 60, 439–471. Humer, M., Lechner, P., 2000. Methane oxidation in optimised landfill cover layers under different seasonal conditions. In: Proceedings 1st Intercontinental Landfill Research Symposia, Lulea. Jones, H.A., Nedwell, D.B., 1993. Methane emission and methane oxidation in landfill cover soil. FEMS Microbiological Ecology 102, 185–195. Kightley, D., Nedwell, D.B., Cooper, M., 1995. Capacity for methane oxidation in landfill cover soils measured in laboratory-scale soil microcosms. Applied and Environmental Microbiology 61, 592– 601. Lewitz, H., 2001. Beobachtungen und Ru¨ckschlu¨sse im Zusammenhang mit dem Gashaushalt der Deponie Georgswerder. TerraTech 6, 47–51 (in German). Nastev, M., Therrien, R., Lefebvre, R., Ge´linas, P., 2001. Gas production and migration in landfills and geological materials. Journal of Contaminant Hydrology 1095, 187–211. Poulsem, T.G., Christophersen, M., Moldrup, P., Kjeldsen, P., 2003. Relating landfill gas emissions to atmospheric pressure using numerical modelling and state-space analysis. Waste Management and Research 21, 356–366. Streese, J., Stegmann, R., 2003. Microbial oxidation of methane from old landfills in biofilters. Waste Management 23, 573–580. Tresselt, K., Miehlich, G., Groengroeft, A., Melchior, S., Berger, K., Harms, C., 1998. Harbour sludge as barrier material in landfill cover systems. Water Science and Technology 37, 307– 313. Whalen, S.C., Reeburgh, W.S., Sandbeck, K.A., 1990. Rapid methane oxidation in a landfill cover soil. Applied and Environmental Microbiology 56, 3405–3411. Young, A., 1990. Volumetric changes in landfill gas flux in response to variations in atmospheric pressure. Waste Management and Research 8, 379–385. Young, A., 1992. The effect of fluctuations in atmospheric pressure on landfill gas migration and composition. Water, Air and Soil Pollution 64, 604–616.