Optimization of Biological Wastewater Treatment by Modeling Microbial Community Behavior Under Varying Nutrient Loading Rates

EPA Grant Number: FP917786
Title: Optimization of Biological Wastewater Treatment by Modeling Microbial Community Behavior Under Varying Nutrient Loading Rates
Investigators: Rubinstein, Rebecca L
Institution: University of Connecticut
EPA Project Officer: Lee, Sonja
Project Period: September 1, 2015 through August 31, 2018
Project Amount: $132,000
RFA: STAR Graduate Fellowships (2015) RFA Text |  Recipients Lists
Research Category: Academic Fellowships

Objective:

Nutrient removal is a critical component of wastewater treatment with respect to environmental protection. It is dependent upon a complex microbial consortium that is sensitive to influent characteristics, process control strategies, and ambient conditions. This study has three aims: to characterize the variation of the activated sludge microbiome at a very high spatial and temporal resolution; to apply a machine learning approach to model the microbial community response to varying nutrient loads, environmental conditions, and process control techniques; and to optimize operating conditions based upon the model that will result in most efficient treatment.

Approach:

This research will be conducted at the University of Connecticut Water Pollution Control Facility (WPCF). Samples will be collected daily, year-round for approximately 2 years, at 6 locations along the length of the activated sludge carousel. Alkalinity, pH, conductivity, ammonia, nitrite, nitrate, and phosphate will be measured. The microbial consortia will also be sampled and stored for genetic sequencing to determine consortia members and community dynamics. A machine learning approach will then be applied to predict microbial community responses using the collected data, process control notes, and the data collected by the WPCF laboratory. The developed model will then be used to optimize treatment techniques.

Expected Results:

This research will provide insight into the microbial community structure and dynamics of an activated sludge treatment system and how those factors impact treatment efficiency. The data collection timeframe covers a wide range of ambient conditions and influent loading rates. It also incorporates the impact of system upgrades in addition to standard operating conditions. As a result, the collection of an extremely comprehensive dataset will result in a very flexible and robust model. The model developed will be used to predict how the activated sludge microbial community responds to varying influent and ambient conditions, as well as to various applied treatment techniques.

Supplemental Keywords:

wastewater, nutrient removal, microbiome, machine learning

Progress and Final Reports:

  • 2016
  • 2017
  • Final