[TOC]    3.  Model Implementation [Prev. Page]   [Next Page]

This chapter discusses the development of the numerical model. The model simulates two-dimensional saturated steady-state flow with advection and dispersion of multiple reactive solutes. It includes the processes of Langmuir and Freundlich isotherms, as well as first order and higher reactions. It uses single, double, and competitive Monod kinetics to represent biological processes. The model is also capable of representing other processes such as intermediate toxicity and cometabolic transformation by combinations of these reactions.

The fundamental equation that represents the sum of these processes is:

(3.1)
Where: d C/d t = time rate of change in concentration (M/L3·T)
  C = concentration (M/L3)
  D = dispersion coefficient (L2/T)
  V = pore water velocity (L/T)
  R = net rate of reaction (sink/source term) (M/L3·T)

There are several options available when solving equation (3.1). Finite difference and finite element methods are very popular. Operator-splitting methods break down equation (3.1) into sub-problems and solve each sub-problem sequentially or simultaneously using the most appropriate method. A literature review was conducted to determine which methods have been employed by other researchers. Tables 3.1.1 through 3.1.4 summarize the results of a literature search of numerical models involving biodegradation.

The tables below illustrate that no single model includes all of the desired processes outlined above. Construction of a new model is necessary. Construction is chosen over modification of an existing program for multiple reasons. Modifying an existing model would require time and effort to learn how the existing code works. It would also be easy to introduce programming errors due to unfamiliarity with the existing code. In addition, most existing programs are written in FORTRAN. Inclusion of multiple different types of reactions is far easier to implement in C++.

Table 3.1.1 Comparison of Numerical Methods, Dimensions and Flow Type

Author(s) Model Dimen-sions Flow Type & Kinetics Advection & Dispersion Number of Components
Chen & McTernan (1992) MMGTM Three Steady-State Saturated Transient Two, Substrate and Oxygen
Chen et al. (1992)   One Steady-State Saturated Transient Coupled Five
Dhawn (1993) -- One Steady-State Saturated Diffusion Only Two, Substrate and Oxygen
Kindred (1989) -- One Steady-State Saturated Transient, Coupled Three
Lindstrom (1992) -- One Steady-State Saturated Transient, Decoupled Two
Mills (1991) COMETº Two Steady-State Saturated Transient, Coupled One
Odencrantz (1992) -- Two Steady-State Saturated Transient, Decoupled Two, Electron Donor & Acceptor
Rifai (1988) Bioplume II Two Steady-State Saturated Transient Decoupled Two, Hydrocarbon and Oxygen
Semprini & McCarty (1991b) -- One Steady-State Saturated Transient, Coupled Two, Electron Donor & Acceptor
Sleep, Sykes (1993) -- Three Variably Saturated, Transient Transient, Coupled Arbitrary
Srinivasan, Mercer (1988) Bio1D One Steady-State Saturated Transient Coupled 1º and 2º Substrates and Oxygen
Taylor & Jaffé (1990b) -- One Transient, Saturated Transient, Coupled Two, Substrate & Biomass
Tim & Mostaghmi (1991) VIROTRANSæ One Variably Saturated, Transient Transient, Coupled One, Virus
Zheng (1993) -- Three Steady-State Saturated Transient, Decoupled One
What is Desired   At Least Two Steady-State or Transient   Four or greater

This study was for biological stimulation only, not a biodegradation study.
Macropores
ºColloids-Metal Transport Model incorporates EPA’s CML model, movement of contaminants on moving colloids.
æVirus Transport Model

Table 3.1.2 Comparison of Numerical Methods, Biological Reactions

Author(s) Model Biological Kinetics Cometab-olism Competitive Inhibition Contaminant Availability
Chen & McTernan (1992) MMGTM Single, Double Monod & First-Order No No Soluble and Sorbed
Chen et al. (1992)   No Yes No Soluble Only
Dhawn (1993) -- Double Monod -- No Soluble Only
Kindred (1989) -- Double Monod Yes No Soluble Only
Lindstrom (1992) -- Double Monod No No Soluble Only
Mills (1991) COMET -- -- -- --
Odencrantz (1992) -- Single, Double Monod & Biofilm -- -- Soluble Only
Rifai (1988) Bioplume II Instantaneous No No Soluble Only
Semprini & McCarty (1991b) -- Double Monod & Biofilm -- -- Soluble Only
Sleep, Sykes (1993) -- -- -- -- --
Srinivasan, Mercer (1988) Bio1D Double Monod No No Soluble Only
Taylor & Jaffé (1990b) -- Biofilm, (effectiveness factor) -- -- Soluble Only
Tim & Mostaghmi (1991) VIROTRANS First-Order Decay -- -- --
Zheng (1993) -- -- -- -- --
What is Desired   Monod and First Order Yes Yes Soluble and Sorbed

Double Monod Kinetics were modified by including a term that accounts for the minimum substrate value below which nothing happens.

Table 3.1.3 Comparison of Numerical Methods, Biomass Transport, and Sorption Kinetics

Author(s) Model Biomass Transport Biomass Growth Sorption Partitioning Sorption Kinetics
Chen & McTernan (1992) MMGTM No Yes, Decoupled Linear, Langmuir & Freundlich Equilibrium & Non-Equilibrium
Chen et al. (1992) -- No Yes Linear Equilibrium
Dhawn (1993) -- No Yes, Decoupled Linear Equilibrium
Kindred (1989) -- No Yes Linear Equilibrium
Lindstrom (1992) -- No Yes Linear Equilibrium
Mills (1991) COMET Colloid Transport -- Linear for Soil and Colloids Equilibrium
Odencrantz (1992) -- No Yes, Decoupled Linear Equilibrium
Rifai (1988) Bioplume II No No Linear Equilibrium (inferred)
Semprini & McCarty (1991b) -- No Yes, Decoupled Linear Equilibrium & Non-Equilibrium
Sleep, Sykes (1993) -- -- -- none none
Srinivasan, Mercer (1988) Bio1D No No Linear, Langmuir & Freundlich Equilibrium
Taylor & Jaffé (1990b) -- Yes Yes -- --
Tim & Mostaghmi (1991) VIROTRANS Yes No Linear for Virus Equilibrium
Zheng (1993) -- -- -- Linear Equilibrium
What is Desired   Desired for Completeness Yes Linear, Langmuir & Freundlich Equilibrium and Non-Equilibrium

Decoupled from the fluid flow equation, does not effect fluid flow.

Table 3.1.4 Comparison of Numerical Methods, Numerical Techniques

Author(s) Model Hard Coded for Acceptor Numerical Technique Abiotic Reactions
Chen & McTernan (1992) MMGTM Yes Crank-Nicholson Finite-Difference, Newton-Raphson No
Chen et al. (1992)   Yes Galerkin F.E. Picard Iteration No
Dhawn (1993) -- Yes ISML No
Kindred (1989) ` Yes Optimal Test Function No
Lindstrom (1992) -- Yes Eulerian-Legrangian No
Mills (1991) COMET -- -- --
Odencrantz (1992) -- No Operator-splitting No
Rifai (1988) Bioplume II Yes Method of Characteristics No
Semprini & McCarty (1991b) -- No Finite Difference, Runge-Kutta No
Sleep, Sykes (1993) -- -- Finite Difference, IMPESC No
Srinivasan, Mercer (1988) Bio1D Yes Finite-Difference Crank Nicholson Newton-Raphson No
Taylor & Jaffé (1990b) -- -- Galerkin Finite Element, Weighted Finite-Difference No
Tim & Mostaghmi (1991) VIROTRANS -- Galerkin Finite Element, Newton-Raphson, Picard Iteration No
Zheng (1993) -- -- MOC block-Centered, Finite-Difference, Foreword Particles No
What is Desired -- No -- Desired for Completeness

The coupled set of ODE’s were integrated using the International Mathematical & Scientific Library subroutine LGEAR and DPDES.


[Home]   [Table of Contents] [Prev. Page]   [Next Page]

A Two Dimensional Numerical Model for Simulating the Movement and Biodegradation of Contaminants in a Saturated Aquifer
© Copyright 1996, Jason E. Fabritz. All Rights Reserved.