Developing an artificial intelligence-based WRRF nitrous oxide mitigation road map: The Eindhoven N2O mitigation case study

An N2O risk roadmap was developed to apply an artificial intelligence-based N2O risk model and diagnose risk of producing nitrous oxide in water resource recovery facilities (WRRFs), based on expert knowledge, identify opportunities to mitigate N2O production, and implement mitigation strategies in full-scale. The N2O risk roadmap was followed for the Eindhoven WRRF in The Netherlands. The four-step process resulted in first identifying that significant risk of producing N2O existed, and that risk is due to both Low DO risk and High DO risk, which implicates both N2O pathways from ammonia oxidizing bacteria. Assessing risk results along with process data then helped to determine that DO could potentially be The risk results and stabilized to mitigate N2O risk and eliminate peaks of ammonia. approach were verified with diagnostic N2O measurements, which also helped in identifying and implementing a specific mitigation strategy in a full-scale control test. The control test results verified that N2O production could be mitigated based on the risk results, as well as that peaks of ammonia and periods of over-aeration can be eliminated. This would represent approximately a 40 percent reduction in the total N2O and WRRF electricity related greenhouse gas emissions. Therefore, the AI-based N2O risk roadmap has proven to be an effective and practical approach for using expert knowledge in significantly reducing WRRF GHG emissions.