A method-of-lines formulation for a model of reactive settling in tanks with varying cross-sectional area

TL;DR

This paper develops a one-dimensional reactive settling model with variable cross-sectional area for secondary settling tanks, combining sedimentation, compression, and biokinetic reactions, and introduces a finite difference scheme ensuring physically relevant solutions.

Contribution

It presents a novel method-of-lines formulation for reactive settling in tanks with variable cross-section, incorporating complex PDEs and ensuring invariant-region properties.

Findings

Model accurately simulates denitrification in wastewater tanks.

Finite difference scheme maintains physical relevance of solutions.

The approach is compatible with existing wastewater treatment software.

Abstract

Reactive settling denotes the process of sedimentation of small solid particles dispersed in a viscous fluid with simultaneous reactions between the components that constitute the solid and liquid phases. This process is of particular importance for the simulation and control of secondary settling tanks (SSTs) in water resource recovery facilities (WRRFs), formerly known as wastewater treatment plants. A spatially one-dimensional model of reactive settling in an SST is formulated by combining a mechanistic model of sedimentation with compression with a model of biokinetic reactions. In addition, the cross-sectional area of the tank is allowed to vary as a function of height. The final model is a system of strongly degenerate parabolic, nonlinear partial differential equations (PDEs)that include discontinuous coefficients to describe the feed, underflow and overflow mechanisms, as well as singular source terms that model the feed mechanism. A finite difference scheme for the final model is derived by first deriving a method-of-lines formulation (discrete in space, continuous in time), and then passing to a fully discrete scheme by a time discretization. The advantage of this formulation is its compatibility with common practice in development of software for WRRFs. The main mathematical result is an invariant-region property, which implies that physically relevant numerical solutions are produced. Simulations of denitrification in SSTs in wastewater treatment illustrate the model and its discretization.