Assessing N2O production and short-cut nitrogen removal in Post Aerobic Digestion through enhanced monitoring and advanced process modeling

Nitrous Oxide (N2O) is a greenhouse gas emitted from resource recovery facilities (WRRFs) utilizing biological nutrient removal (BNR) at a rate between 0.1-1.8 percent of the influent total kjeldahl nitrogen (TKN) loading. In addition to having a global warming potential more than 300-times that of Carbon Dioxide, N2O is a precursor to ground-level ozone formation. With the Environmental Protection Agency (EPA)'s reclassification of the Denver Metro/North Front Range ozone nonattainment area from Moderate to Serious nonattainment in December 2019, the City of Boulder (city) has become increasingly interested in understanding the impact that the Boulder WRRF (BWRRF) has on GHG emissions. One such concern is the continued use of the BWRRF's Post Aerobic Digester (PAD), which utilizes operating conditions similar to those identified to increase N¬2O production in secondary processes (e.g. low dissolved oxygen, transient anoxia, etc.). Enhanced sampling around PAD and process modeling on with improved nitrogen biokinetics was completed to better predict nitrogen removal pathways occurring within PAD and the potential for N2O release. Initial results indicate minimal N2O generation through steady state loading conditions (<0.05 percent of PAD loading). There are, however, likely intermittent excursions of N2O and possibly nitric oxide (NO) that may require further evaluation. It is expected that N2O emissions from PAD represent a small portion of the overall GHG footprint of the WRRF. In addition, N2O emissions from PAD through this initial monitoring period appear to be significantly less than those reported for alternative sidestream nitrogen removal processes (e.g. anammox). The study was completed between January and February of 2020 at the BWRRF. The study included the installation of an online N2O sensor (Unisense, Aarhus, Denmark) to monitor aqueous N2O and nitrogen sampling as noted in Table 1. Gaseous emission rates of N¬2O and NH3 were estimated utilizing the transfer coefficient (kLa). Estimates based on lab collected data were compared to the values predicted within the Sumo© process model (data not shown). Process modeling using enhanced nitrogen biokinetics was completed to provide further insights into the possible nitrogen pathways occurring within the PAD process. A process flow diagram of the model can be seen in Figure 1. The model base is denoted as Sumo4N and includes two primary ammonia oxidizing bacteria (AOB) pathways for N2O/NO generation described by the 2-P model (Pocquet et al. 2016) along with heterotrophic reactions related to NO and N2O. The 2-P model includes the nitrifier denitrification pathway (ND) and incomplete oxidation of hydroxylamine pathway (NN). The matrix adds seven nitrogen related state variables, including intermediates in ammonia oxidation (NH2OH and NO) and twelve reactions related to nitrogen transformations. Figure 2 shows a comparison between conventional 2-step nitrification/denitrification (Henze et al. 2000) and the Sum4N approach.