Various empirical and probabilistic kinetic inactivation models that c
an be used to assist in the design and analysis of potable water disin
fection systems were reviewed. Models were derived for both disinfecta
nt demand-free and demand conditions. Ozone was used to inactivate het
erotrophic plate count bacteria that were grown in natural water under
low nutrient conditions and enumerated using R2A agar at 20 degrees C
for 7 days. Experiments were conducted at 22 degrees C in 0.05 M (pH
6.9) phosphate buffer in bench-scale, batch 250 mL reactors. This disi
nfection data set, characterized by tailing-off behavior, was used to
assess Chick-Watson, Hem-type, Rational, Hom-Power law, and Selleck mo
del fit to the observed logarithmic survival ratios. It was found that
the Chick-Watson model did not adequately represent the ozone disinfe
ction kinetics. A Hom-type model incorporating a first-order disappear
ance term for ozone residual was found to best describe the observed i
nactivation of heterotrophic plate count bacteria. Named the incomplet
e gamma Hom model, it was found to be a robust kinetic model. The prop
osed incomplete gamma Hom model can be used to generate simple design
charts for a wide range of disinfectant types, organisms, and conditio
ns, as an aid to the design of water disinfection systems.