An actinometric method to characterize performance of reflecting UVC reactors used for water treatment

The applicability of chemical actinometry to characterize the fluence in UV reactors with reflections, non-parallel light, and variable water transmittance is limited due to the unknown effective path length or hydraulic shortcuts within the reactor. In this study, the effects of reflection and transmittance on actinometry were examined and a new, optimized and easy method for determining fluence was developed. KI/KIO₃ and uridine actinometry experiments were carried out under controlled conditions using a collimated beam apparatus and a completely mixed batch reactor with or without diffuse reflection and compared to biodosimetry results. Whereas optically opaque actinometers such as KI/KIO₃ are not directly capable of predicting the fluence of reflecting reactors, the results of uridine actinometry are influenced by reflection and transmission. To precisely predict the fluence rate in UV reactors with uridine, knowledge about the effective optical path length of the light is needed. Here, an existing method to mathematically calculate the optical path length was adopted and optimized for uridine actinometry. Results for average fluence were validated by biodosimetry using MS2 phages under different degrees of reflection and transmission. It could be shown that by modifying the bottom of the reactor with diffusely reflecting polytetrafluoroethylene foil, the fluence rate was increased by a factor of approximately 2.6 and the path length by factor of 2.4. When only half of the bottom was covered with reflective foil, fluence rate increased by a factor of 1.8 and path length by 1.8. Although this new approach cannot replace biodosimetry, to predict the fluence distribution received by microorganisms, it can provide means to characterize more complex reactor designs, validate results of advanced reactor modeling, and quantify fluence for non-parallel irradiation and reflective light, especially for the application of high fluence (e.g., advanced oxidation processes), where biodosimetry may be too sensitive. Further, comparing the fluence obtained with actinometry to the results of biodosimetry might qualitatively indicate hydraulic short cuts or unideal fluence distributions for flow-through reactors.

Inactivation of Alicyclobacillus acidoterrestris spores, single or composite Escherichia coli and native microbiota in isotonic fruit-flavoured sports drinks processed by UV-C light

Pasteurized sports drinks and other fruit-based beverages are susceptible to deterioration due to thermal processing ineffectiveness to inactivate certain spoilage microorganisms, like Alicyclobacillus acidoterrestris. This represents a major challenge for the beverage industry. The goals of this study were to: i) investigate the UV-C inactivation (annular thin film unit, actinometrical delivered fluence: 795-1270 mJ/cm², 10-15 min, 20 °C, 1.8 L/h, Re_h = 391-1067, recirculation mode operation) and the evolution during refrigerated storage of A. acidoterrestris ATCC 49025 spores and single or composite Escherichia coli ATCC 25922 in isotonic sports drinks (ISDs) made from orange (orange-ISD, UVT% = 81) or orange-banana-mango-kiwi-strawberry-lemon juices (multi-fruit-ISD, UVT% = 91), compared to a turbid orange-tangerine juice (OT juice, UVT% = 40); ii) assess the effect of pH, °Brix, A_254nm, turbidity, colour and particle size of the ISDs and juice on microbial inactivation, iii) evaluate the evolution of native microbiota during cold storage, iv) investigate the Coroller, biphasic, Weibull, and Weibull-plus-tail models' ability to describe microbial inactivation and v) measure 5-hydroxymethylfurfural (HMF) formation. The modified biodosimetry method was used to calculate the germicidal UV-C fluences. Heat pasteurization (T-coil, 80 °C/6 min) was evaluated as the control treatment. UV-C was highly effective at inactivating E. coli as 4.1-5.1 and 4.5-5.6 log reductions were determined in the multi-fruit-ISD and orange-ISD, respectively, barely impacted by the background microbiota. No significant differences were recorded for the inactivation of E. coli in the UV-C and T-coil systems. Whereas, a significantly higher inactivation of A. acidoterrestris spores was achieved by UV-C (3.7-4.0 log reductions), compared to the negligible one achieved by the thermal treatment. Even though E. coli inactivation curves were similar in shape, UV-C was less effective when a cocktail of other E. coli strains was present. In comparison to the OT juice, the ISDs' inactivation kinetics were markedly different in shape, with a rapid decrease in population during the first minutes of treatment. The germicidal fluence (H_{d biod}) corresponding to A. acidoterrestris (19.1 mJ/cm²) was selected as it was higher than the one obtained for E. coli (11.0 mJ/cm²). UV-C induced 2.8- or 1.3 and 2.3- or 0.8 log-reductions of total aerobes or moulds and yeasts in the multi-fruit-ISD and orange-ISD, respectively. Compared to the other models, the Coroller and biphasic models showed a better fit and more accurate parameter estimates. UV-C-induced HMF production was not significant in the ISDs. The current study found that the UV-C treatment was more effective than typical heat pasteurization for inactivating A. acidoterrestris spores in isotonic drinks, following a similar trend for E. coli and native microbiota.

A review of the fluence determination methods for UV reactors: Ensuring the reliability of UV disinfection

Ultraviolet (UV) is a green and effective technique that has been widely applied in water disinfection. The reliability of UV disinfection is an important issue, in which the aim is to ensure the delivery of adequate real-time fluence in a UV reactor. Unlike chemical disinfection systems whose disinfection dose can be directly measured with disinfectant residuals, UV is a physical process and the determination of fluence is complicated in practical reactors. To date, several fluence determination methods have been developed, including conventional methods such as biodosimetry and model simulation, as well as emerging methods such as dyed microsphere method and the model-detector method. However, a systematic and comprehensive review of these methods is still needed to discuss the attributes and application scenarios of each method. In this review, we summarized the principal theories, procedures, applications, and pros/cons of these fluence determination methods. Further, the selection and application of appropriate fluence determination methods were discussed based on different purposes (e.g., feedbacks for reactor design, evidence for third-party validation, as well as on-site determination and long-term monitoring of fluence). Overall, this review could provide useful information and new insights regarding the application of current fluence determination methods to ensure the reliability of UV disinfection.