Prediction and quantifying parameter importance in simultaneous anaerobic sulfide and nitrate removal process using...

Area of Science:

• Environmental Engineering
• Biotechnology
• Wastewater Treatment

Background:

• Simultaneous removal of sulfide and nitrate is crucial for wastewater treatment.
• Upflow anaerobic sludge bed (UASB) reactors are effective for anaerobic processes.
• Predictive modeling can optimize wastewater treatment efficiency.

Purpose of the Study:

• To develop and validate an artificial neural network (ANN) model for predicting simultaneous anaerobic sulfide and nitrate removal.
• To evaluate the influence of key operational parameters on reactor performance.
• To compare different methods for quantifying parameter importance.

Main Methods:

• An artificial neural network (ANN) model was developed using influent sulfide/nitrate concentrations, S/N ratio, pH, and HRT as inputs.
• Model performance was assessed using RMSE, MAE, MARE, and R².
• Parameter importance was determined using connection weights, Garson's algorithm, and partial derivatives (PaD).

Main Results:

• The ANN model demonstrated good predictive performance for the simultaneous sulfide and nitrate removal process.
• Hydraulic retention time (HRT) was identified as the most significant parameter affecting reactor performance.
• Sulfide and nitrate removal efficiencies, along with sulfate and nitrogen production, were accurately predicted.

Conclusions:

• ANN modeling is a viable tool for predicting the performance of UASB reactors in simultaneous sulfide and nitrate removal.
• Optimizing HRT is critical for enhancing the efficiency of this combined wastewater treatment process.
• The study provides insights into the complex interactions governing anaerobic sulfide and nitrate removal.