Mathematical models of the response of populations of microorganisms exposed to ultraviolet germicidal irradiation (UVGI) are developed that include two-stage response curves and shoulder effects. Models are used to develop a C++ computer program that is capable of predicting the performance of UVGI air disinfection systems.
The algorithms are based on models for 1) the intensity field of UVGI lamps, 2) the intensity field due to UVGI reflective enclosures, and 3) the kill rate of microorganisms to UVGI exposure as they pass through the modeled intensity field. The validity of the UVGI lamp model is established by comparison with lamp photosensor data.
The validity of the overall predictive model is established by comparison of predictions with laboratory bioassays for two species of airborne pathogens–Serratia marcescens and Bacillus subtilis. First stage rate constants, second stage rate constants, and the defining shoulder parameters are determined for Aspergillus niger and Rhizopus nigricans based on bioassay data, and it is shown how predictions using only single-stage rate constants can deviate significantly from predictions using the complete survival curve.
A dimensional analysis of UVGI systems identifies nine dimensionless parameters responsible for determining the effectiveness of any rectangular UVGI system. A factorial analysis of the dimensionless parameters based on data output by the program identifies the most critical parameters and the inter-relationships that determine UVGI system effectiveness.
Response surfaces are generated using program output to illustrate the inter-relationships of the dimensionless parameters. The optimum values of the dimensionless parameters have summarized that result in optimized performance.
Economic optimization is demonstrated by a series of examples that calculate life cycle costs, and principles of economic optimization are summarized. Conclusions are presented that will produce more energy-efficient and effective designs and a proposed model for improved UVGI systems is presented.

Download the White Paper