Cross-Validation of Liquid-Phase and Offgas N2O Emission Measurements at Full-Scale Wastewater Treatment Plants

Danish utilities will soon face regulation based on nitrous oxide (N[sub]2[/sub]O) emissions from biological wastewater treatment. The Danish EPA-funded N[sub]2[/sub]O Mobile project demonstrated a comprehensive, multi-sensor approach for measuring and validating N[sub]2[/sub]O emissions across multiple wastewater treatment plants. Using liquid N[sub]2[/sub]O and offgas sensors operating simultaneously at up to eight positions-either within one or across several bioprocess tanks-the project developed a mobile setup for complete detection of N[sub]2[/sub]O emissions. Through extensive measurement campaigns, the project enabled detailed emission profiling at both tank and plant levels. The results form the basis for a standardized method to extrapolate N[sub]2[/sub]O emissions from single-point measurements to total plant emissions, supporting future regulatory implementation. This is important for other utilities as it provides a practical blueprint for measuring N[sub]2[/sub]O emissions reliably [b]Results and discussion[/b]: Two full-scale WWTPs were monitored; Aalborg Vest (3 treatment trains, 6 tanks, 265.000 PE) and Lynetten (10 treatment trains, 20 tanks, 1.000.000 PE). The WWTPs have fully mixed aeration tanks and are full BNR bioprocesses with fine bubble diffused air and intermitted aeration. N[sub]2[/sub]O emissions were monitored at single tanks and across entire WWTPs using 8 offgas hoods (UpWater, 8 min intervals) and 8 liquid phase N[sub]2[/sub]O sensors (Unisense Environment, 1 min intervals), installed in pairs at 8 process locations. Each pair, placed in proximity, collected data that was pre-processed and transferred to a cloud server for centralized processing in Hubgrade (Veolia). Emission calculations and time series matching were conducted with Python. At both WWTPs, comprehensive 5-month cross-validation study was undertaken between liquid-phase N[sub]2[/sub]O sensors and offgas-based emission measurements with all-in-one tank ([b]Figure 1[/b]). In both WWTPs all monitoring positions showed comparable N[sub]2[/sub]O liquid (mgN/L) and offgas (ppm) concentrations, demonstrating fully mixed raceways. Comparison of each liquid N[sub]2[/sub]O/offgas pair, utilizing hood airflow, demonstrated strong correlations for both Aalborg slope 0.95, R[sup]2[/sup] = 0.98) and Lynetten (slope 1.10, R[sup]2[/sup] = 0.99), as established through published monitoring and mass transfer models (Foley et al., Schulthess, R. & Gujer) and advanced hood systems (Gruber et al.) ([b]Figure 2[/b]). Additionally, when tank airflow was normalized per m[sup]2[/sup] of aerated surface for the same pairs, similarly linear correlations were observed at both wastewater treatment plants (slopes 0.94 & 1.12, R[sup]2[/sup]=0.98 & 0.97). Due to low emissions at Aalborg Vest, comparisons were over a high ammonium load period. Comparing emissions calculated from hood measured airflow (1m2) to those based on tank-wide airflow reveals much weaker correlations, as shown in [b]Figure 2[/b] for both WWTPs (lower left and upper right quadrants), mainly due to differences in airflow across the tank surface. For both liquid N[sub]2[/sub]O and gas phase monitoring technologies used, depending solely on airflow measurements at one or a few hoods can lead to considerable uncertainty in total N[sub]2[/sub]O emission estimates, even within a single zone or bioprocess tank. By calculating or installing distributed airflow meters at tank or zone level, overall emission reporting accuracy can be greatly improved, eliminating errors caused by uneven or localized airflow. Results from multi-tank monitoring in 8 of 20 process tanks at Lynetten show that true mean emissions were 31.54 kgN/day (SD = 6.06 kgN/day, mean of all liquid N[sub]2[/sub]O sensor measurements, [b]Figure 3 & Table 1[/b]). To statistically test monitoring requirement for a ~15% accuracy, progressively increasing the number of measured tanks 1 to 8 tanks (17–100% of the total), the relative uncertainty decreased from originally 22% demonstrating that approx. 30% coverage is required to obtain estimates within 10–15% of the true plant mean. This finding aligns with earlier multi-tank assessments covering 100% at Aalborg Vest, where daily emissions rates ranged from 0.45 to 0.93 kgN/day and an average emission rate of 0.71±0.19 kgN/day ([b]Table 2[/b]). Highlighting this variability, statistically testing the number of tanks needed to get within 10-15% of the Aalborg Vest mean value shows that at least 50% of the bioprocess volume should be monitored. [b]Conclusions[/b]:This report summarizes a multi-site assessment of two N[sub]2[/sub]O measurement methods for emission quantification. Long-term validation showed that liquid-phase N[sub]2[/sub]O sensor and offgas emission measurements produce comparable results, confirming the reliability of mass-transfer models for liquid-based calculations. Accurate bioprocess emission estimates require airflow data from all tanks or zones, as relying on local readings may lead to significant errors; local tank or zone level airflow improve reporting accuracy. Measurements across multiple treatment trains reveal that monitoring at least 30-50% of bioprocesses helps reduce uncertainty caused by uneven loading. The study identifies best practices for extrapolating single-point measurements to whole-plant emissions, offering validated guidelines to support accurate total emissions reporting in wastewater treatment plants