Evaluating The Inactivation Efficiency Of An Ultraviolet Light Emitting Diodes Point-of-Use (UV-LEDs POU) Device For Drinking Water Disinfection

INTRODUCTION
The use of ultraviolet (UV) lamps in water treatment is a promising alternative to conventional water disinfection techniques (e.g., chlorination and ozonation). UV disinfection inactivates a broad range of microorganisms without chemicals, thereby eliminating the problem of disinfection byproducts. However, UV disinfection technology also have some disadvantages. The use of mercury lamps creates heat, and UV systems are often complex due to the necessary contact time for disinfection. Water turbidity must be taken into consideration when measuring contact time for UV disinfection. Moreover, the lamps need to be cleaned or replaced periodically and, most importantly, could pose a risk of mercury contamination into the water if a lamp breaks. UV light emitting diodes (UV-LEDs) are an emerging UV treatment technology that eliminates several of these disadvantages, offering many benefits over mercury containing UV lamps. They are smaller, lighter, and mercury-free while providing effective water disinfection performance in a simple and environmentally friendly manner employing a wider UV spectrum selection. UV-LEDs can be incorporated into point-of-use (POU) treatment devices for water disinfection. POU water disinfection systems can improve public health in rural communities that often lack access to clean water. Therefore, it is imperative to evaluate the disinfection efficacy of these residential POU UV-LEDs devices. In this study, we investigated the inactivation efficiency of a POU UV-LEDs device against heterotrophic bacteria and E. coli under various operating conditions, including UV wavelength, flow rate, and UV radiative power output.
MATERIALS AND METHODS
The UV-LEDs POU device consists of two UV reactors, a working volume of 59 mL per reactor, two UV-LEDs plates with six chips per plate, and a controller. Each UV reactor is assembled with one of the UV-LEDs plates at the end of the reactor. Three different UV-LEDs chip arrays, which emit at 265 nm, 278 nm, or combination of two (i.e., three chips of each wavelength in a UV-LEDs plate), are on the UV-LEDs plate, enabling distinctive combinations within the UV wavelengths. For each test, 20 L of dechlorinated hot tap water from a simulated home plumbing system at U.S. EPA in Cincinnati, Ohio, was used as feed water. The feed water was cooled to room temperature before use. The flow rate varied from 1 L/min to 4 L/min. The UV radiative power output was adjusted by the controller that selectively turns on and off the individual UV-LEDs chips on the plate. UV-LEDs exposure of heterotrophic bacteria naturally occurring in the feed water was measured during bench-scale experiments. Experiments also included E. coli that was spiked into the feed water with an initial concentration of approximately 10[sup]6[/sup] CFU/mL. The inactivation efficiency of heterotrophic bacteria and E. coli was determined by analyzing the concentration of a target microorganism in influent (before UV treatment) and effluent (after UV treatment) using the spread plate culture method onto R2A agar and nutrient agar, respectively (BD biosciences). RESULTS AND DISCUSSION Heterotrophic bacteria were more susceptible to treatment at 265 nm–265 nm than 278 nm–278 nm under consecutive exposure condition (Fig. 1-A). Moreover, a combined wavelength of either 265 nm–278 nm (sequential exposure) or 265 nm/278 nm–265 nm/278 nm (simultaneous and consecutive exposure) achieved improved inactivation efficiency, implying a synergistic effect when UV wavelengths are combined. Increasing flow rate resulted in a decreased inactivation efficacy in heterotrophic bacteria due to the lowered UV exposure time (the mean hydraulic residence time is only 0.9 seconds per each reactor at 4 L/min). E. coli was much more susceptible to all the UV exposure combinations tested than heterotrophic bacteria. While no synergistic bactericidal effect of dual wavelength against E. coli was observed (Fig. 1-B), the inactivation rate of E. coli was greater than 5-log, which was beyond the detection limit, when lower flow rates were used (3.5 seconds per each reactor at 1 L/min). It is also important to note that the inactivation efficiency of E. coli was achieved with only one-sixth of total energy consumption that was used for heterotrophic bacteria. A single UV-LEDs chip per reactor (approximately 2.1–2.2 W depending on the wavelength) exhibited greater than 2-log inactivation even at 4 L/min. Based on the current experimental results, the specific energy consumption for 4-log inactivation of E. coli is estimated. As seen in Figure 2, the fast flow rate consumes less energy, which allows for a larger amount of treated water for the same period of time. However, faster flow rates (e.g., 2 and 4 L/min) can result in a reduced overall UV exposure time for the feed water, possibly decreasing water disinfection performance.
CONCLUSION
The results from this UV-LEDs POU device demonstrated a potential synergistic effect of the combined UV-LEDs wavelengths on heterotrophic bacteria inactivation. Moreover, the device effectively inactivated E. coli while consuming much lower energy than that was required for heterotrophic bacteria. These findings suggest further research on the UV-LEDs POU devices are required to elucidate the mechanisms behind the results as well as optimizing the operating conditions that will make UV-LEDs POU devices viable.