721342

research-article2017
WMR0010.1177/0734242X17721342Waste Management & ResearchLucernoni et al.

Original Article

Waste Management & Research

Sampling method for the determination of

2017, Vol. 35(10) 1034­–1044
© The Author(s) 2017
Reprints and permissions:
methane emissions from landfill surfaces sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0734242X17721342
https://doi.org/10.1177/0734242X17721342
journals.sagepub.com/home/wmr

Federico Lucernoni, Matteo Rizzotto, Laura Capelli, Valentina Busini,

Renato Del Rosso and Selena Sironi

Abstract
The first aim of this work is the definition and the study of a suitable sampling method for the measurement of landfill gas (LFG)
emissions from landfill surfaces, since, up to now, there are no codified nor universally accepted sampling methods for this specific
task. The studied sampling method is based on the use of a static hood. The research work involves a preliminary theoretical study
for the hood design, experimental tests for the definition of the optimal sampling procedures, and simulations of the hood fluid-
dynamics for the system validation. The second aim of this study is the investigation of the correlations between LFG emissions and
meteorological conditions, whose identification would be very useful in terms of effective landfill management and pollution control.
This involved a wide literature study for the selection of those parameters that seem to have an influence on LFG emission, and the
collection of a great number of experimental data on a target site, which led to the conclusion that atmospheric pressure and soil
humidity are the parameters that mostly affect LFG emissions.

Keywords
Landfill, sampling, passive area sources, static hood, flux chamber, CFD, surface emissions

Received 2nd March 2017, accepted 24th June 2017 by Associate Editor Arne Ragossnig.

Introduction
Landfills are significant sources of pollution (Kumar et al., 2004). conditions and emissions, such correlations are undemonstrated,
Disposal of waste in such sites leads to the generation of a lea- mostly contrasting, and never quantified. For this reason, it is not
chate, which may pollute the land and the aquifer (Renou et al., known how to account for the effect of changing meteorological
2008), and a biogas, a mixture of carbon dioxide (CO2) and meth- conditions on the LFG emissions.
ane (CH4), greenhouse gases, and pollutants (Young and Parker, The first aim of this work is the definition of a sampling
1983). Landfills are typically also an important source of odour method that allows for reproducible and repeatable measure-
pollution (Palmiotto et al., 2014), because of the presence of ments. As already mentioned, to date the best method to conduct
traces of compounds in landfill gas (LFG), characterized by very LFG sampling on landfill surfaces is still debated within the sci-
low odour-detection thresholds (Capelli et al., 2008; Davoli entific community.
et al., 2003), given that LFG emissions can be related to odour On a regulatory level, there are two different approaches, both
emissions from a landfill (Lucernoni et al., 2016b). based on so-called “hood methods”, involving the use of a spe-
Even though modern landfills are always equipped with a gas cific sampling hood that isolates a portion of the surface to be
capture system, a portion of the gas escapes and is emitted into sampled. The oldest – and more consolidated – is the EPA
the atmosphere through the surface. Thus, the possibility of quan- method proposed by the U.S. Environmental Protection Agency
tifying the LFG surface emissions and monitoring them over (Klenbusch, 1986), which entails the use of a flux chamber (FC)
time may represent an important aspect for landfill operation and flushed with a neutral gas flow (N2 or air) (Di Trapani et al., 2013;
management (Mosher et al., 1999). The quantification of LFG Gebert et al., 2011; Park and Shin, 2001; Reinhart et al., 1992).
emissions, and their possible correlation to odour emissions,
requires the periodical execution of specific measurement cam-
paigns, which should be carried out by means of a suitable sam- Politecnico di Milano, Department of Chemistry, Materials and
Chemical Engineering “Giulio Natta”, Milan, Italy
pling method that should be repeatable, reproducible, and
accurate. This is a complicated task; to date, there are no codified Corresponding author:
nor universally accepted sampling methods. Laura Capelli, Politecnico di Milano, Department of Chemistry,
Materials and Chemical Engineering “Giulio Natta”, Piazza L. da Vinci
Whilst there have been several studies that hypothesize the 32, 20133 Milan, Italy.
existence of some sort of correlation between meteorological Email: laura.capelli@polimi.it
Lucernoni et al. 1035

The other one, adopted by the UK Environment Agency (UK EA, concentration (PIC) data over multiple, non-overlapping
2010), entails the use of a static hood (SH), in which the increase beam paths (Babilotte et al., 2010).
of methane concentration is measured over time. The static cham- •• The differential absorption LiDAR – or Light Detection And
ber measurement approach, for methane and LFG emission deter- Ranging – (DiAL) technique is a laser-based remote monitor-
mination from landfill surfaces, is the most often used in Italy and ing, enabling range-resolved concentration measures of a
it is also commonly applied in many cases outside of the UK large variety of atmospheric chemical compounds both in the
because it is relatively inexpensive, simple, and highly sensitive at infra-red (e.g. CH4, C2H6, etc.) and ultra-violet wavelength
detecting even small fluxes (Abichou et al., 2006; Bogner et al., spectrum (e.g. NOx, SOx, etc.) with a ppm sensitivity at ranges
1995; Cardellini et al., 2003; Schroth et al., 2012); thus, new and higher than 500 m. The system consists of an accessorized
optimized designs for static chambers have been recently pro- self-powered truck. In the DiAL, the laser is operated alter-
posed (Lucernoni et al., 2016a; Rachor et al., 2013). nately at two adjacent wavelengths. The on-resonant wave-
Moreover, hood methods, and especially fluxed hoods, are by length is chosen to be at a wavelength that is absorbed by the
far the preferred method for the assessment of odour emissions target species. The off-resonant wavelength is chosen to be at
from passive area sources (Capelli et al., 2013; Hudson and a wavelength that is not absorbed by the target species in order
Ayoko, 2008; Parker et al., 2013), and also from landfill surfaces to avoid interferences. Emission fluxes are measured scanning
(Frechen et al., 2004; Romain et al., 2008; Sarkar and Hobbs, the DiAL measurement beam in a vertical plane downwind of
2002; Sironi et al., 2005). the target sources and determining the total concentration of
However, as far as the assessment of CH4 and LFG emissions CH4 above the background in that plane. Vertical planes are
from landfills are concerned, several alternatives to hood sam- typically 600 m × 600 m with a range resolution of 25 m verti-
pling methodologies exist, which are worthy of mentioning: cally and 5 m horizontally. To determine the emission flux of
CH4 due to the landfill site itself, the background CH4 needs to
•• The main alternative is the tracer gas (TG) method, which be subtracted from the concentration profiles before the flux is
entails the controlled release on the emission surface of a calculated (Babilotte et al., 2010).
given traceable gas (e.g., SF6), used to simulate the landfill •• The inverse modelling (IM) entails the use of a concentration
gaseous emissions (Börjesson et al., 2000; Spokas et al., analyser that provides real-time CH4 concentration measures.
2003). There are different options depending on the TG used Concentration measures are performed at discrete receptors
and on the kind of measurement performed. The main two downwind of the landfill. CH4 concentrations and associated
possibilities are mobile or static measurement. The mobile Global Positioning System (GPS) coordinates are then used
plume measurement (MPM) consists of driving with a tune- in an atmospheric dispersion modelling software (e.g.
able diode laser spectrometer (TDL) along a downwind tran- ADMS3, CALPUFF, etc.). The software performs an IM
sect perpendicular to the wind direction around 200 m from analysis according to the geo-referenced CH4 concentrations.
the site to measure TG concentrations in the plume’s cross The output is a CH4 emission factor for each landfill cell
section. An inlet tube is located at the front of the van above (Babilotte et al., 2010).
the cabin to let the outside air come into the TDL system; this
avoids additional mixing. A TG (e.g. N2O) is released with a As previously mentioned, there are several literature works
known constant flow rate from the source and used as a ref- describing the possibility to use hood methods for measuring
erence compound to calibrate the model. The released gas both LFG (Abichou et al., 2006; Bogner et al., 1995; Cardellini
flow is controlled; before and after the experiments the tracer et al., 2003; Di Trapani et al., 2013; Gebert et al., 2011; Liu et al.,
bottles are weighted to know the exact volume of TG lost 2015; Park and Shin, 2001; Reinhart et al., 1992; Schroth et al.,
within the release period. The emission is calculated from 2012) and odour (Frechen et al., 2004; Romain et al., 2008;
the measured/modelled concentration levels above the back- Sarkar and Hobbs, 2002; Sironi et al., 2005) emissions from
ground (Babilotte et al., 2010). The static plume measure- landfills. As demonstrated in other previous works, odour emis-
ment (SPM) entails using vacuum gas bottles installed at a sions from landfill surfaces can be determined either by means of
road downwind of the source. The bottles are evacuated direct odour sampling, or indirectly; that is, by measuring the
before the measurements with a vacuum pump and will fill LFG emissions and then multiplying the LFG flux by the LFG
themselves to approximately 0.5 bar, requiring a defined odour concentration (Lucernoni et al., 2016b, 2017). In this sec-
time. After, the bottles are closed and analysed with a TDL ond case, the odour concentration of the LFG emitted through the
system. One of the bottles is used for the assessment of the landfill surface needs to be estimated by relating the CH4 and the
background concentration of CH4 and TG; this one needs to odour concentration of samples collected over the landfill surface
be located upwind of the landfill. The emission from the (Lucernoni et al., 2016b, 2017). Even though odour measurement
source is calculated from the measured/modelled concentra- is not the primary objective of this work, because of the above-
tion levels above the background (Babilotte et al., 2010; mentioned possibility to relate CH4 and odour concentration data,
Simpson et al., 1995). hood methods were preferred as the investigated method for the
•• The radial plume mapping (RPM) entails the use of optical development of a sampling methodology that allows both the
remote sensing (ORS) for the collection of path-integrated measurement of LFG fluxes and of odour emissions.
1036 Waste Management & Research 35(10)

Figure 1.  The sampling hood designed at the Politecnico di Milano operated as a flux chamber (FC; left) and as a static hood
(SH; right).

The development of the sampling methodology involved the (2013). The hood has a squared base, 50 cm × 50 cm, with a
following: a preliminary theoretical study for the hood design; height of 10 cm and it is connected to the outside by means of a
experimental tests for the definition of optimal sampling proce- 3-m-long Teflon tube that keeps the internal pressure equal to the
dures and operative conditions; and a fluid-dynamic study atmospheric pressure in order to avoid over-pressures that may
exploiting computational fluid dynamics (CFD) simulations for affect the emission. The device is made of steel. A 10-cm-long
the system validation. tube on the top of the hood allows for the CH4 concentration
The second goal of the study is the investigation of the corre- measurement by attaching a Flame Ionization Detector (FID),
lations between LFG emissions and meteorological parameters. which has a suction flow of 1.06 l min–1, to the outlet tube for 2
This involved a wide literature study for the selection of those min. The CH4 concentration analyser used is a Crowcon GasTec
parameters that seem to have an influence on the LFG emissions. portable FID, with a detection range of 0–10,000 ppm and an
The identification of such a correlation would be useful in terms accuracy of 1 ppm.
of effective landfill management and pollution control. One particular feature of this newly designed hood is that it
can be operated both as a FC or as a SH; that is, with or without
Materials and methods the introduction of a neutral sweep air flow, respectively. This
gives the advantage of allowing both the direct and indirect
Experimental campaigns measurement of odour emissions from landfill area sources.
The site is a landfill in Northern Italy, operative since 1993; it has Direct odour measurement needs to be performed using the hood
an extension of 250,000 m2, subdivided into six allotments, of as a FC, since the withdrawal of the olfactometric sample vol-
which only one is still operational. The landfill waste storage ume would perturb the internal SH too much; this is less true for
capacity amounts to roughly 6,200,000 m3. The LFG collection a FC whose internal volume is continuously flushed by a neutral
system sucks 2200 m3 h–1 of gas, which is burnt in four co-gener- gas flow. On the other hand, indirect odour emission measure-
ators for the production of electric energy. The campaigns were ments can be carried out with both hoods, since this method
planned with a frequency of two times per week on average, for a relies primarily on the evaluation of the LFG flux through the
total of 40, from December 2014 to November 2015 in allotments measurement of the CH4 concentration over the landfill surface,
1 and 2, both closed and covered with a clay layer, without water- and the subsequent estimation of the odour emission rate by
proofing seal. Four different points were identified for LFG sam- multiplication with the LFG odour concentration (Lucernoni
pling from December 2014 to March 2015; since two of these et al., 2016b, 2017).
points resulted scarcely emissive (not measurable), these points In order to evaluate the possibility to use the designed hood in
were discarded and replaced with three new points for the meas- both modes (fluxed or static), in the period from December to
urements from March 2015 to November 2015. Measures were March, the hood was operated both as a FC and a SH to compare
always carried out at the same time, around 11 a.m., in order to the two sampling methods.
prevent additional variability. The procedure defined for the FC mode (Figure 1, left) pro-
vides that the hood is fluxed with a neutral air flow of 200 l h–1
for a period of 12 min by connecting an air bottle to the tube on
Materials the top of the hood.
The device developed at the Politecnico di Milano for CH4 sam- The SH mode procedure provides that the hood is positioned
pling over landfill surfaces is a hood (Figure 1), which was on the landfill surface for a period of 10 min before measuring
designed based on the one described in the UK EA (UK EA, the CH4 concentration. From the CH4 concentration, it is possible
2010) and by the modified version proposed by Rachor et al. to calculate the specific emissive LFG flow, in l m–2 h–1. The
Lucernoni et al. 1037

presence of the lateral tube connecting the interior of the hood sampling hood is fundamental for the correct interpretation of the
with the external ambient assuring isobaric conditions during sampled data (Prata Jr. et al., 2016). CFD simulations with the
sampling also guarantees that a higher CH4 concentration inside software ANSYS Fluent were performed with the aim of verify-
the hood is avoided without continuously sweeping the air out of ing the assumptions of linear growth of the CH4 concentration
the hood (Lucernoni et al., 2016a; Rachor et al., 2013). inside the hood and of equivalence of measured CH4 concentra-
For the fluxed mode, it is possible to write the CH4 mass bal- tion and average CH4 concentration inside the hood, on which the
ance as: method (equation (4)) is based. This fluid-dynamic study is fun-
damental to verify the appropriateness of the developed sampling
method, thereby involving the verification of the pertinence of
Q out ⋅ cCH 4
, mis = Q in ⋅ cCH 4
, in + Q LFG ⋅ cCH 4
, LFG (1) the choice of the sampling point and of the sampling time, thus
proving the representativeness of the adopted procedure for the
In equation (1), (Q out ) is the total flow [ l h −1 ] equal to the inlet determination of the emitted LFG flow rate. The advantage of
air flow ( Q in = 200l h −1 ) plus the emitted LFG flow ( Q LFG ), this approach is to avoid the perturbation in the concentration
(cCH , mis ) is the measured CH4 concentration [µmolmol−1 ],
4
inside the chamber induced by the flow of the FID, which would
(cCH ,in ) is the CH4 concentration in the neutral air equal to zero
4
make it very difficult experimentally to evaluate the concentra-
[µmolmol−1 ], and (cCH , LFG ) is the concentration in the pure
4
tion in a precise point and impossible during a single test.
LFG equal to 500,000 [µmol mol−1 ] (i.e. 50% molar fraction). In order to apply CFD, it was first necessary to generate a suit-
The CH4 concentration of 50% in the pure LFG is a datum able mesh that provides the software with a discrete representa-
obtained from the operational information regarding the LFG tion of the hood geometry. The adopted mesh is non-structured
collection and combustion system of the landfill. Since the emit- with a refinement of 14 layers of structured mesh at the inlet
ted LFG flow rate is much lower than the fluxed neutral air, it is boundary (as shown in Figure 4) and a total number of 3,000,000
( )
possible to assume that Q out and Q in are equal. The specific ( ) cells. Then the simulation settings have to be defined: the source
LFG flow Q ( 
)
 −2 −1 
LFG  lm h  is obtained by dividing by the base term was set equal to 0.25 lm −2 h −1  , which is the mean specific
area of the hood ( Sh ), which is 0.25 m2: LFG flow emitted from the landfill surface, derived from the
experimental campaigns on site. This value is not the final datum,
 Qin ⋅ cCH , mis
Q LFG = but it is a value obtained during the first campaigns, in order to
4

(2)
cCH , LFG ⋅ Sh 4
perform the CFD study assessing the appropriateness of the
When operating the hood as a SH (Figure 1, right), it is possible adopted sampling method in the early stages of the project. The
to write the mass balance for CH4 as: model also requires setting the values of pressure, temperature,
and diffusivity coefficients. The simulations were run consider-
d cCH
V⋅ 4
= Q LFG ⋅ cCH , LFG
(3) ing a total time of 12 min: 10 min of SH positioning + 2 min for
dt 4

the FID analysis.

In equation (3), (V ) is the volume of the hood equal to 25 [ l],
 d cCH  is the average CH concentration variation inside the Correlation between meteorological data
 
4
4
 dt  and LFG emissions
hood over time, (Q LFG ) is the emitted LFG flow [ lh −1 ], and
(cCH , LFG ) is the concentration in the pure LFG equal to 500,000 In order to investigate the existence of a correlation between LFG
4

[µmolmol−1 ]. The specific LFG flow may be obtained by making emissions from the landfill surface and meteorological condi-
two assumptions: tions, a great number of emission and meteorological data are
required. Before starting the analysis of the experimental data
1. the CH4 concentration inside the hood has a linear growth acquired from the measurement campaigns on site, a thorough
over time; review of the scientific literature on the subject was performed to
2. the CH4 concentration that is measured by means of the FID is identify the meteorological parameters that other authors had
equal to the average CH4 concentration inside the hood cCH . investigated as possibly affecting the LFG emissions. As a first
4

result of this initial literature review step, Table 1 presents a bib-

The specific LFG flow can be computed using equation (4): liographic overview of what has been written by other research-
cCH ⋅V ers on this matter by summarizing their statements and
 , mis
Q LFG = (4) observations about the influence of atmospheric parameters on
4

cCH , LFG ⋅ ∆t ⋅ Sh
4
LFG emissions.
Finally, the meteorological data considered for the study were
rainfall, temperature, atmospheric pressure, wind speed, relative
Computational Fluid Dynamics (CFD)
humidity, and solar radiation. In addition, the soil humidity was
simulations also considered, which was determined experimentally by col-
CFD was applied to study the static sampling procedure, since lecting soil samples and weighting them before and after drying,
the understanding of the fluid-dynamic behaviour of the thus assessing the humidity as the removed water fraction.
1038 Waste Management & Research 35(10)

Table 1.  Bibliographic overview of the statements and observations made by other researchers about the influence of
atmospheric parameters on LFG emissions.

Atmospheric parameter Source(s) Correlation type Author’s hypothesis

Atmospheric pressure (P) McBain et al., 2005; Reinhart Negative correlation (–): Landfill “respiration”
et al., 1992 P increases and LFG flow
decreases
Temperature (T) Park and Shin, 2001 Positive correlation (+): T Waste decomposition process is
increases and LFG flow enhanced
increases
Temperature (T) Rachor et al., 2013 Negative correlation (–): Temperature governs microbial
T increases and LFG flow CH4 oxidation because chemical
decreases processes in the cells are faster
at greater energy supplies (i.e.
during the warm season)
Soil humidity (U) Rachor et al., 2013 Positive correlation (+): U CH4 production is enhanced in
increases and LFG flow a moist environment and CH4
increases oxidation is disadvantaged by lack
of O2 (pores occupied by H2O)
Soil humidity (U) Rachor et al., 2013 Negative correlation (–): CH4 oxidation is favoured, more
U increases and LFG flow favourable environment for
decreases methanotroph bacteria
Wind velocity (v) McBain et al., 2005 Positive correlation (+): v Diffusive resistance is lowered,
increases and LFG flow the boundary layer thickness
increases decreases
Wind velocity (v) Rachor et al., 2013; Reinhart Negative correlation (–): Wind velocity influences the
et al., 1992 v increases and LFG flow pressure value on soil surface, P
decreases increases
Rainfalls (p) Rachor et al., 2013 Negative correlation (–): Cork effect, water prevents LFG
p increases and LFG flow from going into the atmosphere
decreases

Results and discussion – and, thus, also in terms of related odour – emission fluxes, it
was decided to give preference to the SH method, since it is less
Definition of the sampling method demanding, especially as far as logistics are concerned, as it does
Figure 2 shows the LFG specific flow data obtained in the differ- not involve the need to provide a neutral sweep air flow (i.e. no
ent sampling points for all experimental campaigns. In order to need to transport and consume air bottles, or to use a rotameter).
define the most suitable sampling method, emission data obtained The mean specific LFG flow rate derived from the experimen-
with the SH have been compared with those obtained with the FC tal measurements turned out to be equal to 0.39 l m–2 h–1. This
until March: the comparison showed no dramatic differences value is similar to that found in literature in a study by Palmiotto
between the two methods (Figure 3). et al. (2014), in a municipal solid waste (MSW) landfill also
The average flux highlighted in Figure 2 is obtained by com- located in Northern Italy, which is likely to have both landfill
puting the arithmetic mean, in accordance with the criteria in the waste quality and landfill management operations similar to
guideline of the UK EA (UK EA, 2010). those of the landfill considered for this study. In that case, the
The mean specific LFG flux value turned out to be 0.39 obtained surface LFG flux is equal to 0.2 l m–2 h–1 (Palmiotto
l m–2 h–1. There is a significant variability of the experimental et al., 2014). Parallel to this, a CFD simulation of the sampling
data (Figure 2), which is the case in environmental measurement with the hood was performed to see if the sampling system
campaigns, since the measurement is highly affected by different adopted was suitable for the study.
factors. In Figure 2, the majority of the flux values are rather low,
while the high ones represent a smaller share. The method
CFD simulation of the sampling
adopted to assess the LFG flux has been used in a recent publica-
procedure with SH
tion to assess odour emissions from landfill surfaces (Lucernoni
et al., 2016b). In order to verify the assumptions of linear growth of the CH4
By plotting the LFG flux data (Figure 3) with the SH data on concentration inside the SH and of equivalence of measured CH4
the x-axis and the FC data on the y-axis, it is possible to see how concentration and average CH4 concentration inside the hood,
the resulting points are very close to the line, representing perfect which the adopted sampling method relies on, it was decided to
equivalence between the two methods (i.e. x = y). Once it was evaluate the CH4 concentration at some specific points inside the
verified that both methods provide similar results in terms of CH4 hood with a CFD simulation. The LFG specific flux imposed as
 Lucernoni et al.

Figure 2.  Results of the experimental campaigns (specific LFG flow rates in l m–2 h–1).
1039
1040 Waste Management & Research 35(10)

the source term for the CFD simulations is not 0.39 l m–2 h–1, hood. This linear growth for short horizon times is observed for
which was the result of the experimental campaigns, but 0.25 all the points on the central axis of the hood.
l m–2 h–1. The reason for this discrepancy is because it was neces- The CFD simulation also allowed us to verify the second
sary to run the preliminary simulations by means of CFD to assumption – that the CH4 concentration read by the FID is rep-
describe the fluid-dynamic behaviour of the hood before com- resentative of the average CH4 concentration inside the hood – by
pleting the experimental campaigns in the field to validate the analysing the CH4 concentration contours inside the hood over
sampling system. For this reason, the specific LFG flux of 0.25 time; that is, during the first 10 min, in which the SH is posi-
l m–2 h–1 used for the simulations is a partial result that was tioned over the landfill surface and during the following 2 min of
obtained after the first campaigns. The fact that the “true” value the FID measurement duration.
resulting after completion of the experimental data collection The LFG flow rate considered as the source term (i.e. 0.25
turned out to be 0.39 l m–2 h–1, does not affect the significance l m–2 h–1) is rather low, so no mixing is foreseen inside the hood;
and validity of the CFD simulations, since the two values do not this was confirmed by the simulation showing that the CH4 con-
differ substantially. In order to confirm this assumption, other centration distribution inside the SH after 10 min is not homoge-
simulations were subsequently run, changing the LFG specific neous, but stratified (Figure 5(a)). The simulation results presented
flux in a range from 0.05 l m–2 h–1 to 1 l m–2 h–1, as described in in Figure 5(a) also show that the stratification is bell-shaped – the
the paper by Lucernoni et al. (2016a). typical concentration distribution in the case of plug flow with
Figure 4 shows the results of the simulation relevant to the axial diffusion. This is corroborated by the order of magnitude of
point at the outlet orifice on top of the hood, where the shorter the diffusive velocity, 10–4 m s–1, higher than the convective
tube is attached and where the FID is inserted for the concentra- velocity inside the hood, and the order of magnitude of 10–7 m s–1.
tion measurement. The values at each minute from 0 to 10 in Thus, after the established “deposition” time of 10 min, the CH4
kmol m–3 and the resulting plot over time are reported. It is pos- concentration at the sampling point (the outlet) is equal to 196
sible to observe that the concentration points are linearly interpo- μmol mol–1, a value very close to the average CH4 concentration
lated, with a correlation index R 2 of 0.9991, confirming the first inside the hood after 10 min (this contour is highlighted with the
hypothesis of linear increase of the CH4 concentration inside the dashed line in Figure 5(a)), equal to 208 μmol mol–1, which can
also be calculated as shown in equation (5):


Q * ∆t * Sh (5)
LFG * cCH , LFG
cCH = 4

4
V
After the FID operation time of 2 min (Figure 5(b)), the system is
perturbed with respect to the static deposition period (Figure
5(a)), giving a fairly mixed system. The simulation results show
that the CH4 concentration at the sampling point after this time
(10 + 2 min) is 260 μmol mol–1, compared to an average CH4
concentration in the chamber of 248 μmol mol–1. These results
confirm both the appropriateness of the choice of the sampling
point and, consequently, the legitimacy of the assumption of con-
Figure 3.  Comparison between LFG emission data obtained sidering the CH4 concentration value measured with the FID at
with flux chamber (FC) and static hood (SH) (December the sampling point as representative of the average CH4 concen-
2014–March 2015). tration inside the hood. A more detailed study of the trend of

Figure 4.  (a) Methane concentration values over time; (b) concentration trend over time at the hood outlet; and (c) mesh
representation for the device highlighting the outlet.
Lucernoni et al. 1041

Figure 5. CH4 concentration distribution after 10 min (a) and after 10 + 2 min (b).

concentration inside the hood and the validation of the CFD Moreover, in those laboratory studies, significant variations of
results with the experimental data has been presented elsewhere CH4 emissions were observed only for considerable tempera-
(Lucernoni et al., 2016a). ture differences by controlling other parameters. For this study,
samples were always collected in the morning, given that the
temperature differences were significant only when comparing
LFG emissions and meteorological
measurements carried out in different seasons; and in such a
parameters long time frame and with other parameters affecting the CH4
The first step of this part of the work was the selection of the emissions more so than temperature, it is unlikely that such
meteorological parameters deemed as the most influential on the effects of temperature on the potential of methane oxidation are
LFG emissions. The wind velocity and the wind direction were observed.
excluded since the sampling method adopted is a hood isolating Air humidity does not seem to have a significant influence on
the sampling area, making the action of the wind uninfluential, LFG emissions, which was hypothesized a priori since there is no
especially for the present study where the source is located in a theoretical justification for that and since no mention of a possi-
region characterized by weak winds. Air temperature proved to ble influence of air humidity on the emissions was found in the
have a small influence on emissions, probably because the waste literature. Up to now, there is no conclusive evidence that solar
decomposition process is in an advanced stage and external air radiation directly affects the emissions, even if several undergo-
temperature has a minimal influence on the phenomenon. There ing studies are trying to study a possible cross-correlation
could also be an alternative explanation: considering that the air between radiation and atmospheric pressure, since both parame-
temperature only has an influence on soil temperature up to a ters are indicative of the weather conditions. Rainfall was not
certain depth (1–2 m below the surface), the temperature in the directly investigated as a parameter; the consideration of soil
landfill body is typically constant. As most of the waste is located humidity was preferred. Therefore, the variables considered for
deeper than that, it is possible that the air temperature will not the study were atmospheric pressure and soil humidity. As
affect LFG generation. explained in the scientific literature, it is possible to ascribe the
On the other hand, there are some studies in literature inves- negative influence of pressure increase on LFG emissions to a
tigating the effects of temperature on microbial methane oxida- phenomenon called landfill “respiration”: in some cases, a weak
tion in landfill cover soils, which might, therefore, have negative correlation was found between the two variables.
measurable consequences on methane emissions from the land- According to this explanation, a pressure increase “pushes” the
fill surface to the atmosphere (e.g. Börjesson et al., 2004; Einola biogas into the soil, obstructing the emission. In order to investi-
et al., 2007; Spokas and Bogner, 2011). However, most of these gate the effectiveness of this hypothesis, the daily and monthly
studies have been performed on a laboratory scale or by isolat- pressure trends were analysed as well as the pressure gradients
ing a portion of the cover soil to be tested, and in general, up to during the 3 and 6 hours preceding each campaign. The only cor-
now, there have been very few field studies that have attempted relations that were identified are those between emissions and
to investigate the relative contribution of the effective temporal instantaneous pressure at the time of the sampling, and the aver-
dynamics and the relative contribution of such environmental age pressure in the preceding 24 hours and 48 hours. The prelimi-
parameters directly on landfill surfaces (Scheutz et al., 2009). nary analysis of the data shows a correlation that – in contrast to
For this reason, it is very difficult to make accurate and quanti- the results of some other studies – is positive; even though weak,
tative considerations about the possibility that temperature is the correlation seems to indicate an increase of the LFG emission
negatively correlated to CH4 emissions from landfill surfaces. with atmospheric pressure. This observed correlation might be
1042 Waste Management & Research 35(10)

Figure 6.  LFG emission vs. environmental parameters. (a) LFG emission vs. atmospheric pressure at the time of sampling; (b)
LFG emission vs. average atmospheric pressure in the 24 h preceding the sampling; and (c) LFG emission vs. soil humidity.

partially explained based on some considerations found in the since during rainfall a sort of “cork” effect is observed due to the
manual “Solid Waste Engineering” by Sirini et al. (2010), which obstruction of the pores and reduced diffusivity, and measured
discusses the migration of trace gases across the landfill surface emissions from the landfill surface are 0. In order to provide an
(equation (6)) in Chapter 15: explanation for this experimental evidence, some considerations
about the operation of the studied landfill should be made. In the
Di * α 4 / 3 * ( Ci , atm − Ci , s * Wi ) (6) examined landfill, the common practice of “leachate recircula-
Ji =
L tion” inside the waste mass is not performed; the leachate formed
during waste fermentation is compelled by gravity to move
Sirini et al. indicate that the emission flux of the specific gas (Ji) downwards, given that the upper levels of waste will be typically
is a function of several parameters, such as molecular diffusivity dry and thus generate less biogas due to slower fermentation
(Di), soil porosity (α), atmospheric gas concentration (Ci,atm), kinetics. Moreover, given that the closed allotments of the stud-
saturation gas concentration (Ci,s), scaling factor (Wi), and land- ied landfill are not fully waterproofed, during rainfall, water
fill covering thickness (L). It is possible to assume that all the trickles across the soil, wetting those wastes stored in the upper
parameters contained in the equation are not affected by atmos- levels which are normally dry; this might speed up the LFG pro-
pheric pressure, except the diffusivity, which is related to the soil duction kinetics and cause higher emissions. However, there may
porosity. Therefore, it is possible that, since high pressure indi- be an alternative reason as well: the hotspots can be assumed to
cates a “no-rain condition”, the soil porosity will be higher and contain wider pores than the soil around these hotspots. Therefore,
thus emissions will be higher as well. if it starts raining the pores in the soil may be obstructed more
As an example, the correlation between specific LFG flow and earlier than the pores in the hotspot. This may drive a larger
rate and atmospheric pressure (instantaneous and average of the portion of the emission through the hotspot without affecting the
preceding 24 hours) relevant to sampling point #3 is shown in overall landfill emission.
Figure 6(a) and (b). The correlation observed in Figure 6(c) looks more exponen-
Concerning the correlation between LFG emissions and soil tial than linear. This observation would be coherent with the
humidity, a positive correlation was found and the LFG emission influence of the humidity found in most expressions for the CH4
seems to increase with the soil humidity (e.g. Figure 6(c) shows production kinetics via waste fermentation. As an example, the
the correlation between LFG emission and soil humidity relevant equation used by the US EPA software LandGEM (Alexander
to sampling point #3). This is only true in the case of no rain, et al., 2005) describes an exponential trend whereby the exponent
Lucernoni et al. 1043

contains the kinetic constant (k), which depends mainly on the Davoli E, Gangai ML, Morselli L, et al. (2003) Characterization of odor-
ants emissions from landfills by SPME and GC/MS. Chemosphere 51:
soil humidity:
357–368.
n 1 Di Trapani D, Di Bella G and Viviani G (2013) Uncontrolled methane emis-
 Mi 
QCH =4 ∑∑kL  10  e
i =1 j = 0 ,1
0
− ktij
(7) sions from a MSW landfill surface: Influence of landfill features and side
slopes. Waste Management 33: 2108–2115.
Einola JKM, Kettunen RH and Rintala JA (2007) Responses of methane oxi-
dation to temperature and water content in cover soil of a boreal landfill.
Conclusions Soil Biology & Biochemistry 39: 1156–1164.
Frechen FB, Frey M, Wett M, et al. (2004) Aerodynamic performance
The first aim of the work was the definition of a reliable and of a low-speed wind tunnel. Water Science and Technology 50:
57–64.
reproducible sampling procedure. The optimal procedure defined
Gebert J, Rachor I, Gröngröft A, et al. (2011) Temporal variability of soil
is based on the use of a SH, with a static sampling time of 10 min gas composition in landfill covers. Waste Management 31: 935–945.
followed by 2 min required by the FID for the CH4 concentration Hudson N and Ayoko GA (2008) Odour sampling 2. Comparison of physi-
cal and aerodynamic characteristics of sampling devices: A review.
measurement. The CFD simulations confirmed the basic assump-
Bioresource Technology 99: 3993–4007.
tions for the definition of this sampling method: linear CH4 con- Klenbusch MR (1986) Measurement of gaseous emission rates from land
centration growth during the established sampling time and surfaces using an emission isolation flux chamber (User’s guide. No.
equivalence between measured CH4 concentration and average PB-86-223161/XAB). Austin, TX: Radian Corporation.
Kumar S, Gaikwad SA, Shekdar AV, et al. (2004) Estimation method for
CH4 concentration inside the hood, as well as the appropriateness national methane emission from solid waste landfills. Atmospheric
of the chosen sampling point. Environment 38: 3481–3487.
The second aim was a preliminary study of the correlation Liu Y, Lu W, Li D, et al. (2015) Estimation of volatile compounds emission
rates from the working face of a large anaerobic landfill in China using a
between LFG emissions and meteorological parameters; the most
wind tunnel system. Atmospheric Environment 111: 213–221.
influential parameters were identified, these being atmospheric Lucernoni F, Capelli L and Sironi S (2017) Comparison of different
pressure and soil humidity. Some preliminary positive correla- approaches for the estimation of odour emissions from landfill surfaces.
Waste Management 63: 345–353.
tions were observed between those parameters and the LFG
Lucernoni F, Rizzotto M, Tapparo F, et al. (2016a) Use of CFD for static
emissions from the landfill surface. sampling hood design: An example for methane flux assessment on land-
fill surfaces. Chemosphere 163: 259–269.
Declaration of conflicting interests Lucernoni F, Tapparo F, Capelli L, et al. (2016b) Evaluation of an odour
emission factor (OEF) to estimate odour emissions from landfill surfaces.
The authors declared no potential conflicts of interest with respect to Atmospheric Environment 144: 87–99.
the research, authorship, and/or publication of this article. McBain M, Warland J, McBride R, et al. (2005) Micrometeorological meas-
urements of N2O and CH4 emissions from a municipal solid waste land-
Funding fill. Waste Management & Research 23: 409–419.
Mosher B, Czepiel P and Harriss R (1999) Methane emissions at nine land-
The authors received no financial support for the research, author- fill sites in the northeastern United States. Environmental Science &
ship, and/or publication of this article. Technology 33: 2088–2094.
Palmiotto M, Fattore E, Paiano V, et al. (2014) Influence of a municipal solid
waste landfill in the surrounding environment: Toxicological risk and
References
odor nuisance effects. Environment International 68: 16–24.
Abichou T, Chanton J, Powelson D, et al. (2006) Methane flux and oxida- Park JW and Shin HC (2001) Surface emissions of landfill gas from solid
tion at two types of intermediate landfill covers. Waste Management 26: waste landfill. Atmospheric Environment 35: 3445–3451.
1305–1312. Parker D, Ham J, Woodbury B, et al. (2013) Standardization of flux cham-
Alexander A, Burklin C and Singleton A (2005) Landfill Gas Emissions ber and wind tunnel flux measurements for quantifying volatile organic
Model (LandGEM) Version 3.02 User’s Guide, US-EPA-600/R-05/047. compound and ammonia emissions from area sources at animal feeding
Babilotte A, Lagier T, Fiani E, et al. (2010) Fugitive methane emissions from operations. Atmospheric Environment 66: 72–83.
landfills: Field comparison of five methods on a French landfill. Journal Prata Jr AA, Santos JM, Beghi SP, et al. (2016) Dynamic flux chamber meas-
of Environmental Management 136: 777–784. urements of hydrogen sulphide emission rate from a quiescent surface – A
Bogner J, Spokas K, Burton E, et al. (1995) Landfills as atmospheric methane computational evaluation. Chemosphere 146: 426–434.
sources and sinks. Chemosphere 9: 4119–4130. Rachor IM, Gebert J, Grongroft A, et al. (2013) Variability of methane emis-
Börjesson G, Danielsson Ä and Svensson BH (2000) Methane fluxes from a sions from an old landfill over different time-scales. European Journal of
Swedish landfill determined by geostatistical treatment of static chamber Soil Science 64: 16–26.
measurements. Environmental Science & Technology 34: 4044–4050. Reinhart DR, Cooper DC and Walker BL (1992) Flux chamber design and
Börjesson G, Sundh I and Svensson BH (2004) Microbial oxidation of CH4 operation for the measurement of municipal solid waste landfill gas
at different temperatures in landfill cover soils. FEMS Microbiology emission rates. Journal of the Air & Waste Management Association 42:
Ecology 48: 305–312. 1067–1070.
Capelli L, Sironi S, Del Rosso R, et al. (2008) A comparative and critical Renou S, Givaudan JG, Poulain S, et al. (2008) Landfill leachate treat-
evaluation of odour assessment methods on a landfill site. Atmospheric ment: Review and opportunity. Journal of Hazardous Materials 150:
Environment 42: 7050–7058. 468–493.
Capelli L, Sironi S and Del Rosso R (2013) Odor sampling: Techniques and Romain AC, Delva J and Nicolas J (2008) Complementary approaches to
strategies for the estimation of odor emission rates from different source measure environmental odours emitted by landfill areas. Sensors and
types. Sensors 13: 938–955. Actuators B: Chemical 131: 18–23.
Cardellini C, Chiodini G, Frondini F, et al. (2003) Accumulation chamber Sarkar U and Hobbs SH (2002) Odour from municipal waste (MSW) land-
measurements of methane fluxes: Application to volcanic-geothermal fills: A study of the analysis of perception. Environment International
areas and landfills. Applied Geochemistry 18: 45–54. 27: 655–662.
1044 Waste Management & Research 35(10)

Scheutz C, Kjeldsen P, Bogner JE, et al. (2009) Microbial methane oxidation Sironi S, Capelli L, Centola P, et al. (2005) Odour emission factors for assess-
processes and technologies for mitigation of landfill gas emissions. Waste ment and prediction of Italian MSW landfills odour impact. Atmospheric
Management & Research 27: 409–455. Environment 39: 5387–5394.
Schroth MH, Eugster W, Gómez KE, et al. (2012) Above- and below-ground Spokas KA and Bogner JE (2011) Limits and dynamics of methane oxidation
methane fluxes and methanotrophic activity in a landfill-cover soil. Waste in landfill cover soils. Waste Management 31: 823–832.
Management 32: 879–889. Spokas K, Graff C, Morcet M, et al. (2003) Implications of the spatial variability of
Simpson IJ, Thurtell GW, Kidd GE, et al. (1995) Tunable diode laser landfill emission rates on geospatial analyses. Waste Management 23: 599–607.
measurements of methane fluxes from an irrigated rice paddy UK Environment Agency (UK EA) (2010) Guidance on monitoring landfill
field in the Philippines. Journal of Geophysical Research 100: gas surface emissions. Almondsbury, Bristol, UK: UK EA.
7283–7290. Young PJ and Parker A (1983) The identification and possible environmental
Sirini P, Tchobanoglous G and Noto La Diega RC (2010) Ingegneria dei impact of trace gases and vapours in landfill gas. Waste Management and
Rifiuti Solidi. The McGraw-Hill Companies: Milan, chapter 15. Research 1: 213–226.