Validation of UV disinfection systems for recalcitrant microorganisms using Lagrangian actinometry

Po-Shun Chan, Purdue University

Abstract

Lagrangian actinometry (LA) has shown advantages as a tool for validation of UV reactors, such as ability to measure dose distribution. However, several problems involving dyed microspheres (DMS) used in LA have been identified. First, ambient particles present in some water sources could have similar sizes and fluorescence intensity (FI) with the DMS, and could interfere in flow cytometry analysis. Second, the wavelength-dependent dose responses of DMS need to be examined for validation of polychromatic UV reactors. Third, LA applications of high dose range (more than 200 mJ/cm2) for inactivation of UV-resistant microorganisms have not been studied. The primary goal of this research is to develop DMS method for validation of “high dose” reactors. A collimated medium-pressure (MP) UV source and bandpass optical filters were used to conduct wavelength-dependent dose-response experiments for three different types of DMS (nominally 6, 10 and 15 µm) over the entire germicidal UV spectrum for a UV dose range of 0–450 mJ/cm2. Internal (microsphere) standards were implemented and applied for wavelength-dependent dose-response samples to calibrate the measurements in flow cytometry. In applying this approach, variations of the instrument optical responses were reduced. Action spectra of three different types of DMS were established. The action spectra of 6 and 10 µm DMS were similar to several microorganisms within the most effective range of wavelengths for a MP UV disinfection system, which supports the ability of prediction of microbial disinfection performance using LA. A 3-LPHO-lamp reactor and an 18-LPHO-lamp reactor were validated by LA and biodosimetry using three different sizes of DMS, 6, 10 and 15 µm and environmental isolates of Bacillus pumilus spores. Four flow rates were applied for validation of each reactor. There were some significant variations between replicate samples at lower flow rates for the 3-lamp reactor, possibly due to instabilities of the reactor at low flow rates. The ability to capture these variations is also beneficial to understand the limitation of the reactor. LA and biodosimetry results did not agree with each other for the 18-lamp-reactor test, possibly due to extrapolation of DMS dose responses, uncertainty in spore dose-response relationship, and sample contamination.

Degree

Ph.D.

Advisors

Blatchley, Purdue University.

Subject Area

Environmental engineering

COinS