Measurement of methane emissions from confined sources using the inverse dispersion method

Bühler, Marcel (2022). Measurement of methane emissions from confined sources using the inverse dispersion method (Unpublished). (Dissertation, University of Bern, Oeschger Centre for Climate Change Research & Institute of Geography)

Text (PhD Thesis)
PhD_Thesis_Marcel_B_hler_published.pdf - Other
Available under License BORIS Standard License.

Download (16MB) | Preview

Greenhouse gas (GHG) emissions are reported in annual national inventories. Globally, the main anthropogenic sources of methane (CH4) are fossil fuel burning, agriculture, landfills, and waste management. The main source of CH4 from agriculture is enteric fermentation in the digestive tract of ruminants and a minor source are emissions from manure management. In 2019, the Swiss Federal council decided that Switzerland must reduce its GHG emission to net-zero until 2050. To reach this goal, the agricultural sector is obliged to contribute to the emission reduction. However, emissions from the agriculture and waste sector imply large uncertainties as, among other reasons, the availability of data based on real-world studies is limited. Several investigations showed that measurements from laboratory- or pilot-scale experiments do often not comply with real-world conditions. For studies under real-world conditions, different measurement approaches are available. One of the most promising methods is the inverse dispersion method (IDM) that was applied in this thesis to measure CH 4 emissions from livestock production and the waste management sector in order to evaluate the method for complex source configurations and to specify emission factors of these sources.

For the IDM, a backward Lagrangian stochastic (bLS) model in combination with concentration measurements up- and downwind of the source using open-path tunable diode laser spectrometers (GasFinders) were employed. GasFinders are simple to use, flexible in their application, and relatively cost-effective measuring devices. However, several challenges were faced and overcome throughout the thesis. The precision of the employed GasFinder model was about 10 times lower than the manufacturer stated, which necessitated adaptation in the measurement setup. Additionally, an intercomparison before or after each measurement campaign was necessary to correct the offset and
span between the employed GasFinders.

In the first two studies presented in this thesis, experiments were conducted to evaluate the IDM. In a third study, experiments were conducted to assess the handling of complex source configurations with the IDM. The emissions determined in the third study were used as a basis for emission factors of Swiss biogas plants (BGPs) and wastewater treatment plants (WWTPs).

In the first study, a known and predefined amount of CH 4 was released by an artificial source in a barn that mimics a dairy housing. For concentration measurements, GasFinders with a path length ii Thesis summary of 110 m were placed in downwind direction of the barn at a distance of 50 m, 100 m, 150 m, and 200 m. At the first three distances, an ultrasonic anemometer was placed in the middle of the GasFinder path length for executing turbulence measurements. Upwind of the barn, an additional GasFinder and an ultrasonic anemometer were installed. The main objective was to test the ideal measurement fetch for the IDM. The results of this experiment are included in the method section, where the conditions and the setup of an IDM measurement campaign are outlined. A release rate of 140 norm litres min -1 was chosen to achieve sufficient concentration enhancement at the GasFinder locations. The mean recovery rates of the experiment were between 0.55 – 0.76.

In the second study, CH4 emission measurements from a naturally ventilated dairy housing were conducted in two measurement campaigns. During part of the campaign duration, emissions were also measured inside the housing with the inhouse tracer ratio method (iTRM). This allowed comparing the IDM with the iTRM, which was considered as a reference method for naturally ventilated livestock housings. For simultaneous emission intervals, the average IDM emissions were lower by 1 % and 8 % compared to the iTRM measurements, which was within the uncertainty of either of the two methods. Additionally, an uncertainty analysis for the IDM showed that measurement campaigns of at least 10 consecutive days are necessary to acquire reliable emission data.

The third study addressed the handling of complex source configurations with the IDM. Emissions from four agricultural BGPs and two WWTPs in Switzerland were measured. The average BGP CH 4 emission varied between 0.39 kg h -1 and 2.22 kg h-1, which was less than 5 % of the plant’s CH4 production. The average CH4 emissions for the two WWTPs were 166 g population-equivalent-1 y -1 and 381 g population-equivalent-1 y -1, respectively. The BGPs often had livestock housings nearby that needed to be discriminated from the plant emission. It was demonstrated how the plant emission can be corrected for the nearby CH4 sources, which confounded the GasFinder measurements. Further, it was demonstrated how to combine multiple GasFinder measurements to a single line concentration for the bLS modelling. WWTPs are complex sources as they consist of multiple sub-sources with different emission strengths spread over a large area. Three different calculation approaches with different degree of details are presented for the combination of the individual sources in the bLS modelling: (i) A polygon over the entire WWTP area as a single source. (ii) All potential sources within the WWTP have a uniform emission density. (iii) Based on literature data, relative weighting of the individual sources is carried out. The maximum difference in emission between the most complex approach (iii) and the simplest approach (i) was 42 %. It could be shown that for large source areas (> 10,000 m2), approach (iii) is the preferred option, whereas for the measured BGPs the simple polygon approach (i) was sufficient.

The recovery rate of the IDM from the release experiment (study 1) was below 1 and somewhat lower than previous studies with a similar experimental setting have shown. I was not able to conclusively identify the reasons for this result, which contrasts with the outcome of study 2 at the naturally ventilated dairy housing with a high consistency between the IDM and the iTRM used as a reference. Therefore, I suggest repeating the release experiment with an adapted setting and additionally roughly mapping the emission plume by a drone or a high-precision handheld sensor to monitor the dispersion of the plume. The field measurements at the WWTPs and the BGPs based on iii the chosen approach of source combination yielded data that are in the expected range according to
current state of knowledge.

The presented PhD thesis supports the aptitude of the IDM to measure emissions from complex sources like farms, BGPs, or WWTPs. Such measurements contribute to increasing the accuracy of national GHG inventories. Nevertheless, I suggest further investigations to better assess the accuracy of the IDM under complex conditions.

Item Type:

Thesis (Dissertation)


08 Faculty of Science > Institute of Geography > Physical Geography > Unit Climatology
10 Strategic Research Centers > Oeschger Centre for Climate Change Research (OCCR)
08 Faculty of Science > Institute of Geography
08 Faculty of Science > Institute of Geography > Physical Geography

UniBE Contributor:

Bühler, Marcel


500 Science > 530 Physics
500 Science > 550 Earth sciences & geology
500 Science > 570 Life sciences; biology
600 Technology > 620 Engineering
600 Technology > 630 Agriculture
900 History > 910 Geography & travel




Marcel Bühler

Date Deposited:

21 Apr 2022 12:05

Last Modified:

21 Apr 2022 12:05




Actions (login required)

Edit item Edit item
Provide Feedback