A phenomenological model of biocide action against microbial biofilms
was derived. Processes incorporated in the model include bulk flow in
and out of a well-mixed reactor, transport of dissolved species into t
he biofilm, substrate consumption by bacterial metabolism, bacterial g
rowth, advection of cell mass within the biofilm, cell detachment from
the biofilm, cell death, and biocide concentration-dependent disinfec
tion. Simulations were performed to analyze the general behavior of th
e model and to perform preliminary sensitivity analysis to identify ke
y input parameters. The model captured several general features of ant
imicrobial agent action against biofilms that have been observed widel
y by experimenters and practitioners. These included (1) rapid disinfe
ction followed by biofilm regrowth, (2) slower detachment than disinfe
ction, and (3) reduced susceptibility of microorganisms in biofilms. T
he results support the plausibility of a mechanism of biofilm resistan
ce in which the biocide is neutralized by reaction with biofilm consti
tuents, leading to a reduction in the bulk biocide concentration and,
more significantly, biocide concentration gradients within the biofilm
. Sensitivity experiments and analyses identified which input paramete
rs influence key response variables. Each of three response variables
was sensitive to each of the five input parameters, but they were most
sensitive to the initial biofilm thickness and next most sensitive to
the biocide disinfection rate coefficient. Statistical regression mod
eling produced simple equations for approximating the response variabl
es for situations within the range of conditions covered by the sensit
ivity experiment. The model should be useful as a tool for studying al
ternative biocide control strategies. For example, the simulations sug
gested that a good; interval between pulses of biocide is the time to
minimum thickness. (C) 1996 John Wiley & Sons, Inc.