(548e) Electrochemical Systems and Machine Learning for Cost-Effective Wastewater Treatment to Maximize Renewable Energy

Clean water is an essential and limited resource. Growing food demand and fertilizer use has increased the need for wastewater treatment systems. In Iowa, wastewater nitrogen levels range from 2.0 to 360 mg/L, while allowable nitrate levels in bottled water are 10 mg/L. Wastewater treatment facilities can lower nitrogen levels by more than 66%, but they require energy input and are expensive to operate. This study investigates utilizing renewable energy and electrochemical systems to lower wastewater treatment costs.

Renewable energy is abundant and inexpensive but intermittent. Because of this, we developed machine learning (ML) predictive models to operate the system during excess energy availability or high-water nutrient loading times. To further reduce costs, we investigate the sale of hydroxycinnamic acid (HCA) as a high-value by-product from the electrochemical treatment of wastewater streams.

For this project, a deep neural network (DNN) ML model was trained using US geological survey (USGS) data on nitrate levels for rivers across the US. The DNN predicts hourly nitrate levels within a 2% accuracy. An hourly timeframe allows the electrochemical system to switch between solar, battery, or grid electricity depending on availability and cost. At nitrate levels below 60 mg/L, the electrochemical system stores excess solar electricity using a battery unit. Above 60 mg/L, the system treats the wastewater stream to produce HCA. The HCA and sale of excess electricity allows the system to be profitable.

There is a required balance between the benefits of renewable energy options and the costs. This presentation will discuss the potential of using electrochemical systems and chemicals to reduce wastewater treatment costs from point sources by optimizing renewable energy use. We will also discuss how ML models can help address the critical technical parameters and challenges of electrochemical wastewater systems. This intersection of technology and chemistry could solve many of the problems posed by wastewater treatment, possibly even making it a profitable effort in order to drive interest in the process.