Safety Through Uncertainty in Wastewater Treatment: Monte Carlo Simulation-Based First Order Kinetic Design of Waste Stabilization Ponds

Abstract

Waste stabilization pond (WSP) systems are a simple and low-cost form of wastewater treatment. Forecasting the treatment efficiency of these systems can be challenging given the complexity of their treatment processes and their dependence on changing climatic conditions. In WSP design, these treatment forecasting uncertainties are compounded by inaccuracies in population statistics. In this thesis, two studies were conducted to develop uncertainty-based methods of design for WSPs. In the first study, a simple uncertainty-based method of design for a facultative-maturation pond system was proposed based on the removal of biochemical oxygen demand and Escherichia coli. Removal of contaminants was modeled using first kinetics assuming a dispersed flow hydraulic regime with uncertainties propagated using Monte Carlo simulations. A sensitivity analysis method was also described for factor prioritization and factor fixing. The method’s use in practice was presented using a case study for the design of a pond system treating residential wastewater. The second study explores two approaches for the integration of a steady state temperature model (SSTM) and a transient bulk temperature model (TBTM) into an uncertainty-based WSP design method. The design method was based on the removal of biochemical oxygen demand using a first order kinetic model assuming a plug flow hydraulic regime. Uncertainties were propagated using Monte Carlo simulations. A case study with two design scenarios was conducted for the sizing of a standalone facultative pond treating residential wastewater. The first design scenario employed the uncertainty-based method with the SSTM approach while the second made use of the TBTM approach. The uncertainty-based method using the SSTM approach produced area requirements 76% larger than the TBTM approach. The significant difference in area requirement estimates of both approaches highlights a need for the accurate modelling of pond temperature and its uncertainty in uncertainty-based design.