This study is an attempt to assess CH4 and N2O emissions from all the treatment steps of a wastewater treatment plant (WWTP) in Sweden, serving 145 000 persons, and an adjacent biogas production facility. We have used novel mid-IR ground-based remote sensing with a hyperspectral camera to visualize and quantify the emissions on 21 days during a year, with resulting yearly fluxes of 90.4 ± 4.3 tonne CH4/yr and 10.9 ± 1.3 tonne N2O/yr for the entire plant. The most highly emitting CH4 source was found to be sludge storage, which is seldom included in literature as in-situ methods are not suitable for measuring emissions extended over large surfaces, still contributing 90 % to the total CH4 emission in our case. The dominating N2O source was found to be a Stable High rate Ammonia Removal Over Nitrite reactor, contributing 89 % to the total N2O emissions. We also discovered several unexpected CH4 sources. Incomplete flaring of CH4 gave fluxes of at least 30 kg CH4/min, corresponding to plume concentrations of 2.5 %. Such highly episodic fluxes could double the plant-wide yearly emissions if they occur 2 days per year. From a distance of 250 m we found a leak in the biogas production facility, corresponding to 1.1 % of the CH4 produced, and that loading of organic material onto trucks from a biofertilizer storage tank contributed with high emissions during loading events. These results indicate that WWTP emissions globally may have been grossly underestimated and that it is essential to have effective methods that can measure all types of fluxes, and discover new potential sources, in order to make adequate priorities and to take effective actions to mitigate greenhouse gas emissions from WWTPs.
Keywords: Biogas production; Hyperspectral imaging; Methane; Nitrous oxide; Wastewater treatment.
Copyright © 2021 The Authors. Published by Elsevier Inc. All rights reserved.