Long-term monitoring of methane emission rates from a biogas plant in Heidelberg, Germany: quantification of method uncertainties with controlled release experiments

The increasing number of biogas plants in Germany and Europe requires an appropriate strategy to quantify potential methane losses from biogas plants to ensure the sustainability of this renewable energy production. In addition to the large uncertainties in these emission factors, there is little information on the temporal variations of CH4 emissions from biogas plants. A long-term study of CH4 emission rates from a biogas plant in Heidelberg, Germany,  was performed over a five year period from August 2018 to December 2023. For a total of 27 measurement days the CH4 emission rates were determined for this biogas plant. These measurements were performed at different distances, weather conditions and two measurement devices. By averaging the individual emission rates of each measurement day at the biogas plant, an emission rate of 6.7 ± 0.8 kg CH4 h-1 was determined. For the calculation of CH4 emission rates, mobile measurements were performed nearby the biogas plant and analysed by applying a basic Gaussian plume dispersion model. Several controlled CH4 release experiments were carried out to analyse the accuracy and uncertainty of the CH4 emission rates from biogas plants and to improve the method. This included mobile measurements of methane concentration, meteorological measurements and the application of a Gaussian plume model. For this purpose, methane was released from gas cylinders in Heidelberg/Mannheim (Germany) in a controlled manner. Release rates between 0.1 and 0.7 kg CH4 h-1 were set and distances of 5 to 260 m between release point and inlet were investigated. The results obtained by applying a Gaussian plume model were compared with the actual release rates and thus the uncertainty of the method was analyzed in detail. The parameterization of the dispersion coefficients, the stability classes and thus the meteorological conditions during the measurements and other parameters like number of transects driven and averaging methods were assessed.