Predictive Emissions Monitoring Systems (PEMS)

A Predictive Emissions Monitoring System (PEMS) is a software-based approach that estimates and monitors emissions in real time using operational wastewater treatment plant (WWTP) data and advanced analytics. Instead of relying solely on physical monitoring equipment, PEMS uses machine learning and statistical models to predict emissions from process parameters such as temperature, airflow, loading conditions and -concentrations. This enables continuous insight into emissions performance while reducing the need for extensive monitoring hardware.

PEMS provides a scalable and cost-effective solution for monitoring nitrous oxide (N₂O) emissions. Measurements from selected aeration lanes are used to train predictive models that estimate emissions across the entire WWTP. These models are continuously improved as new monitoring data becomes available, allowing WWTPs to maintain reliable emissions tracking even when continuous measurements are limited.

Model performance is validated using the Relative Accuracy (RA) method defined by the U.S. Environmental Protection Agency (EPA) for PEMS. Maintaining RA values below 20% demonstrates that model predictions closely match measured emissions, ensuring confidence for environmental reporting, regulatory alignment, and operational decision-making.