Modeling for Optimal Management of Agricultural and Domestic Wastewater Loading to Streams

Abstract

Abstract. A simulation/optimization (S/0) model to aid managing multiobjective wastewater loading to streams while maintaining adequate downstream water quality is presented. The conflicting objectives are to maximize the human and dairy cattle populations from which treated wastewater can be discharged to the river system. Nonindustrial municipal (domestic) wastewater undergoes primary and secondary treatment by a sewage treatment plant (STP) before entering as a steady point source. Dairy wastewater is treated by overland flow (OLF) land treatment before entering the stream as a controlled steady diffuse source. Maximum dual-source loading strategies which do not degrade downstream water quality beyond specified limits are developed. For each computed loading strategy, an optimal OLF system design is also determined. The E constraint method is used to obtain sets of noninferior solutions. Sets of noninferior solutions are represented graphically to show the trade-off between human and bovine populations that can be maintained. Each set is computed for a different upstream flow rate to illustrate sensitivity to nondeterministic upstream flow rates. The nonlinear constraints utilized restrict downstream concentrations of 5-day biochemical oxygen demand, dissolved oxygen, nitrogen (organic, ammonia, nitrite, and nitrate), organic and dissolved phosphorus, and chlorophyll a. Concentrations are described via regression equations. The new regression expressions, surrogates for the complex advective-dispersive equation, permit rapid and feasible solutions by this unique S/0 model.

Simulation and optimization technologies for petroleum waste management and remediation process control

“…The regression model was built based on 600 runs of a transport model. Ejaz and Peralta (1995) used regression equations based on 729 runs of an advective-dispersive model to predict downstream concentrations of several constituents from the upstream flow rate and constituent concentration. Cooper et al (1998) developed a simulation/regression/optimization (S/R/O) model to predict, analyze and optimize the LNAPL recovery process.…”

Performance of an Artificial Neural Network model for simulating saltwater intrusion process in coastal aquifers when training with noisy data

“…Many researchers have presented various approximation models to simulate the flow and transport processes in ground water. Ejaz and Peralta (1985) presented regression equations to predict downstream concentration of several constituents from the upstream flow rate and constituent concentration. Alley (1986) presented regression equations to relate the variation in pumping and recharge rate at five decisions wells to the concentration at nine control locations for two dimensional transport processes in an aquifer.…”

Controlling wasteloads from point and nonpoint sources to river system by GIS-aided Epsilon Robust Optimization model

“…For example, Ejaz and Peralta (1995) proposed a simulatione optimization (S/O) model for showing the trade-off between human and bovine populations from which treated wastewater can be discharged into the river system, and solved this multiobjective programming problem by use of the 3-constraint method. Uncertainty of inflow rates at the upstream end of a river was taken into account through obtaining different sets of noninferior solutions with different inflow rates.…”