Continuous flow residence time distribution function characterization

Disinfection contactors are frequently characterized using tracer studies to determine macromixing patterns. For some applications, such as compliance with the surface water treatment rule, this information can be interpreted directly without much analysis. However, when tracer studies are performed to characterize residence time distributions for design purposes, this information is frequently analyzed using a method of moments approach. This paper shows that the method of moments approach, when compared to a nonlinear regression approach, produces a biased estimate of mean residence time and dimensionless variance, and one with a greater mean square error. The amount of bias in moments estimators is significant in a numerical sense, and leads to qualitatively poor estimation of the resulting residence time distribution. Examples are given of characterization of a pilot-scale chlorine contactor and a pilot-scale ozone contactor.