Monitoring Key Water Quality Parameters in Water Resources Systems Using Bioactive Electrodes

Existing water quality sensors in surface, environmental, and drinking water systems are not well suited for long-term, scalable use as they require calibration, replacement of reagents, and are subject to biofouling which degrades measurement accuracy. Microbial Potentiometric Sensors (MPSs) offer an alternative approach to water quality monitoring by monitoring a biofilm-mediated potentiometric response to diverse water quality parameters. MPS biofilms grow naturally on graphite electrodes in diverse aqueous systems, are regenerative, and their potentiometric response correlates with numerous water quality parameters. As such, the overarching hypothesis of this dissertation is that MPS signal can be used to assess water quality trends and that its signal is driven by biofilm vitality. To test this hypothesis, machine learning, statistical regression, and the use of more complex, impedimetric measurement techniques were explored to characterize water quality trends in diverse water systems. This was accomplished by completing three dissertation objectives: 1.) Assess whether Machine Learning/Artificial Intelligence (ML/AI) tools can be used to disaggregate various surface water quality parameter values from Open Circuit Potential (OCP) signals produced by MPSs; 2.) Determine whether residual free chlorine concentration in drinking water could be determined by monitoring MPSs; and 3.) Determine whether OCP and/or Electrochemical Impedance Spectroscopy (EIS)-derived impedance data from an MPS can be used to determine water quality trends while confirming its biological origins. The findings confirm the hypothesis by demonstrating that ML/AI can be used to disaggregate MPS signal and determine numerous water quality parameters, offering unique opportunities for real-time monitoring of aqueous environments. Additionally, MPSs are particularly useful in measuring free chlorine concentrations in drinking water distribution systems which offers opportunities for scalable, in-situ, continuous monitoring of chlorine throughout a distribution network. Finally, the findings demonstrate that coupling MPSs’ OCP signal with more advanced measurement techniques such as EIS can improve understanding of drinking water quality trends, however current open source, affordable technologies capable of conducting EIS are prone to high measurement noise and are not currently accurate enough to be used in drinking water systems.