Hadas Mamane, Tel Aviv University
Authors: Hadas Mamane, Yifaat Betzalel, Yoram Gerchman
13:20 - 13:40

UV-Light Emitting Diodes (UV-LEDs) are an emerging technology for water disinfection. The projected shift from classical UV lamps to UV LEDs may allow design flexibility but also result in the need to develop novel tools suitable for their testing. In this study, we present the use of microtiter plates to develop a simple and rapid approach to obtain dose response curves and “Petri factor” of collimated beam LED systems. These plates contain multiple "wells" from 6 to 1536 wells with different spatial organization (2x3 for 6 well to 32X48 for 1536 well plate), and offer spatial resolution and the possibility for simultaneously testing multiple repeats and conditions. Together with Aquisense Technologies we designed a wide LED irradiation apparatus that irradiates the area of a microtiter plate, with a petri factor above 0.9 and tested for photon flux, uniformity and dose-response for the different UV LEDs. At first, all rows are covered, and each row is then exposed to irradiation under mixing according to the predetermined and cumulative exposure time. This approach can be useful for rapidly identifying disinfection conditions (e.g. UV wavelengths, doses, environmental conditions) that will be effective for inactivation, and could be applied later in a continuous system. This method may also be implemented for samples that have low volumes or also for luminescent biosensors, microbes dyed with fluorescent dyes, or for determining concurrently absorbance and UV exposures for all samples.

Surapong Rattanakul, The University of Tokyo
Authors: Surapong Rattanakul and Kumiko Oguma
13:40 - 14:00

UV light emitting diodes (UV-LEDs) have gained a lot of interest for water disinfection because of its advantages over conventional low-pressure UV (LPUV) systems. Many of recent studies has investigated UV-LEDs inactivation efficiency on various microbes including UV-resistant adenovirus, but the information on how viruses get inactivated by the UV-LEDs, or by UV radiation in general, is still limited. In this study, UV-LEDs with peak emission at 265, 280 and 300 nm as well as LPUV lamp emitting at 254 nm were introduced to inactivate bacteriophage MS2, a surrogate of pathogenic viruses in water. Subsequently, the infectivity assay, molecular biological assay using reverse-transcription quantitative polymerase chain reaction (RT-qPCR), attachment assay and genome replication assay were employed to quantify the loss of infectivity, the level of genome damage, the loss in attachment ability, and the loss in genome replication ability, respectively. The first-order inactivation rate constants (k) of different wavelengths in each assay were calculated from dose-response curves to indicate the reduction magnitude in each assay and a statistical difference was tested by the analysis of covariance (ANCOVA). It was revealed that kinfectivity_265 was highest followed by kinfectivity_254, kinfectivity_280 and kinfectivity_300, respectively, indicating UV-LEDs at 265 nm was the most effective wavelength per fluence for MS2 inactivation in this study. Although the UV-LEDs at 300 nm was much less effective than 265 and 280 nm, it still posed a potential to be an alternative choice for virus inactivation. Based on these results, it was revealed that UV-LEDs can be an option for water treatment and virus inactivation mechanisms may be different among UV-LEDs at different emission wavelengths. The details of inactivation mechanisms will be intensively discussed in the presentation.

Kumiko Oguma, University of Tokyo
Authors: Kumiko Oguma, Surapong Rattanakul
14:00 - 14:20

Various studies have demonstrated that UV light-emitting diodes (UV-LEDs) are effective in the inactivation of indicator/surrogate microorganisms such as Escherichia coli and bacteriophages, but studies targeting pathogenic microorganisms are still limited. In this study, we have adopted UV-LEDs emitting at 265, 280 and 300 nm to inactivate pathogens including Pseudomonas aeruginosa and Legionella pneumophila in comparison with E. coli, bacteriophage Qβ and Bacillus subtilis spores. Conventional low-pressure mercury UV lamp was also used for comparison. Inactivation kinetics of each microorganism and each wavelength were described with the multi-target and Geeraead’s models considering both lag and tailing phases in the inactivation profiles, and the relevance of the regression models was statistically tested by R-squared and root mean square error (RMSE) values. It was revealed that log-linear inactivation profile fitted well with L. pneumophila and bacteriophage Qβ, while sigmoid curve was appropriate in the inactivation of P. aeruginosa, E. coli and B. subtilis spores. The log-linear curve worked well with multi-target model whereas sigmoid curve fitted better with Geeraead’s model. According to the fluence-based inactivation rate constant, the most effective wavelength in inactivation was 265 nm followed by 254, 280 and 300 nm. Meanwhile, in the efficiency based on the electrical energy consumption per inactivation, 280 nm was the best among the UV-LED options. The result of this study offers new findings to expand UV-LED application to water disinfection.

Kari Sholtes, University of Colorado Boulder
Authors: Kari Sholtes, Oliver Lawal, Jennifer Pagan, Karl Linden
14:20 - 14:40

The development of a protocol was initiated in 2015 to propose standardized deep Ultraviolet Light Emitting Diode (UV-C LED) output testing conditions, methods, and reporting which will enable greater transparency to system designers, regulators, researchers, and end users. This protocol is an IUVA initiative, undertaken by a working group of the IUVA Manufacturers Council, and is designed to not only facilitate standardized laboratory methods but also create and validate a formal protocol which laboratories can use to better describe the output of their UV-LEDs. The protocol was designed to test and compare standardized UV-C LED lamp emission spectra (power output and wavelength) with standardized equipment under different ambient conditions with different operators.

An independent third party at the University of Colorado Boulder developed and validated the proposed testing protocol for the measurement of UV-LED lamp output. The standardized protocol and calibrated equipment was sent to fourteen domestic and international participants (UV LED manufacturers, system designers, and researchers) for round robin testing in 2016. Measurements consisted of experimental conditions, output spectrum, and radiometer readings. Minimal variability resulted from operational differences (i.e., distance from and orientation of the source to the sensor) when comparing the round robin testing equipment. However, significant variability resulted from the in-house testing equipment owned by the various participants. The final protocol, along with strategies for minimizing the variability between each participants’ in-house equipment, have been developed and will be presented and discussed.

Natalie Hull, University of Colorado Boulder
Authors: Natalie Hull and Karl Linden
14:40 - 15:00

A UV LED flow-through reactor was validated for disinfection performance using MS2 coliphage as a surrogate, over a range of UV transmittance (UVT) levels and flow rates and then piloted in the field to study sustainability, robustness, and water disinfection performance over time.  The reactor pilot site was at a local small drinking water treatment plant downstream of slow sand filtration.  UVT in the UV reactor influent was measured daily, along with temperature, pH, and turbidity.  ATP, total coliform, E. coli, and total organic carbon (TOC) were measured bi-weekly in the treatment plant influent, slow sand filter effluent, UV LED effluent, and compared to the chlorine effluent from a parallel chlorination train.  Disinfection performance in the field was measured by MS2 challenge testing immediately after installation, followed by quarterly measurements to monitor performance over time and during each season.  Data on electrical and maintenance requirements were collected for life cycle sustainability analyses.  Additionally, biofilm development in pipes was compared upstream and downstream of UV and chlorine disinfection processes.  DNA was extracted and analyzed from biofilm that developed on 1-inch pipe sections to determine the effect of each disinfection process on attached microbial communities.  The field challenge testing with MS2 in the small system filter effluent aligned with modeled results from the validation bench testing, indicating no detrimental effects from the water matrix in the field upon installation.  Continued sampling is underway to monitor longitudinal performance.  Longitudinal evaluation of the flow-through UV LED system will provide data necessary for design improvements and scale-up of the reactor, allowing faster adoption in the future. Testing will be complete by January 2018.  This is the first longitudinal pilot investigation of the first commercially available flow-through UV LED disinfection reactor at a drinking water treatment plant.

Majid Keshavarzfathy, University of British Columbia
Authors: Majid Keshavarzfathy, Fariborz Taghipour
15:40 - 16:00

Ultraviolet light-emitting diode (UV-LED) is a new UV source, which emits UV in a small bandwidth with a predesigned peak wavelength. UV-LED has small footprint, does not contain mercury and requires DC power supply. These features have made UV-LED exceptional for small UV-disinfection units. To design an effective UV-LED reactor, the integration of the three phenomena hydrodynamics, kinetics, and radiation is crucial. The amount of UV received by microorganisms depends on the flow and radiation distributions in the reactor. In this study, a comprehensive radiation model was developed and integrated with a CFD-based model to simulate the UV-LED reactor in an Eulerian framework to simulate the reactor performance. This research presents a radiation model based on the actual UV-LED radiation pattern. In our approach, the radiation transport equation (RTE) was adapted for water treatment applications and solved to assess the effect of wall reflectivity on the radiation distribution of a cylindrical reactor. The radiation model includes specular and diffuse reflections from the walls. The radiation distribution was evaluated by optical method and the reactor performance was assessed using biodosimetry technique at different operating conditions. The optical measurements indicated the precision of the radiation model. The good agreement between the simulation and experimental data showed the comprehensive radiation model in an Eulerian framework can be successfully used for simulation and optimization of UV-LED reactors.

Oliver Lawal, AquiSense Technologies
Authors: Gary Hunter, Saketh Thanneeru, Oliver Lawal
16:00 - 16:20

It’s been almost three years since the launch of first commercially available UV-C LED water disinfection system. The dialogue of the UV-C LED system disinfection is moving from possibilities into real world applications. Due to the technology being in its infancy, most of the systems operate at low flows and high UV-Transmittance values, but the aggressive growth nature of LEDs, enables these systems towards more challenging operations. The inherent benefits of the LEDs bring about a major shift in UV industry.

In the wastewater industry, energy and operation costs are major factors in selection of a disinfection technology. UV-C LED systems are based on a solid-state technology, which generally tends to operate with low electrical consumption and less relative space to conventional technology. The pilot study was aimed to study the benefits of a UV-C LED system at Mill Creek Wastewater Treatment Plant located in Cincinnati, Ohio, USA.

The Mill Creek plant is an activated sludge wastewater treatment plant that is highly impacted by wet weather flows that currently uses hypochlorite for disinfection of the treated effluent.  The UV-C LED system would acts as a disinfection step after secondary treatment replacing the hypochlorite system. The main focus of this study is to understand the performance, limitations, and challenges of LED UV in the wastewater environment with aspirations of a future large-scale system. The factors that will be looked at are disinfection efficacy, energy use, operating cost, and life-cycle cost. The results from the pilot will feed into eventually a municipal system design using UV-C LEDs, decreasing not only the costs and maintenance of the systems but also harmful chemical usage almost negligible compared to current consumption.

Junichi Shiga, METAWATER Co., Ltd.
Authors: Junichi Shiga,Takahito Sugimoto,Tsukasa Kusano,Masahiro Ito, Shinya Watanabe,Eiichi Tsuga
16:20 - 16:40

RED (Reduction Equivalent UV Dose) of mercury-free DUV-LED reactor was researched by the study. The reactor was designed for full-scale water treatment, which capacity is 2,000 m3 per day. RED was measured by biological dosimeter. MS2 bacteriophage was used for RED measurement as challenge organism. Two different tests were conducted to measure DUV-LED strength in various conditions.

In the light output adjustment testing, RED was measured at six different level of DUV-LED output between 50 and 100% with constant flow rate of 2,000m3 per day.

The test result showed RED increased linearly with DUV-LED. RED was 26.2 mJ/cm2 at output level of 100%.

In the water flow adjustment testing, RED was measured at five different flow rate between 1,000 and 2,300m3 per day under constant DUV-LED output level. The output level of DUV-LED was set to 70% considering both its life-span and the performance of the DUV-LED . At DUV-LED output of 70% and under flow rate of 2000m3/day, the test resulted in 16.1 mJ/cm2 RED.

These results present that enough UV dose for disinfection was obtained by the full scale DUV-LED reactor.

Further long term testing is planned to obtain characteristic of DUV-LED and performance of UV monitoring device to acquire JWRC（Japan Water Research Center） Certification of Ultraviolet Disinfection.

Sara Beck, Eawag
Authors: Sara E. Beck; Hong Nguyen; Poonyanooch Suwan; Thusitha Rathneyeke; Thammarat Koottatep
16:40 - 17:00

This research investigated the application of ultraviolet-C light emitting diodes (UV-C LEDs) for wastewater reuse for agriculture in low- to middle-income countries. Research conducted at the Asian Institute of Technology (AIT) north of Bangkok, Thailand evaluated options for pretreating domestic wastewater with subsequent disinfection by UV-C LED irradiation.

An initial study evaluated UV disinfection of two sources of wastewater following microfiltration. Domestic wastewater from a primary sedimentation tank and effluent from a conventional septic tank were pumped through a submerged woven-fiber microfiltration bioreactor to decrease the suspended solids and turbidity for effective UV disinfection. The permeates were pumped through flow-through UV reactors manufactured in-house, including a 4 x 16W UV-C LED reactor emitting at 280 nm as well as a 20W low pressure (LP) UV reactor for comparison. For permeate with a minimum UVT of 40%, the UV-C LED flow-through reactor operating at 10 mL/min inactivated MS2 by 3.5-log, compared to over 7-log reduction by the LP UV reactor at 1.5 L/min.

A follow-up study evaluated less expensive pre-treatment options as well as operation and maintenance of a commercially-available UV-C LED reactor. Domestic wastewater was continuously flowed through a tube settler and sand filter before exposure to a UV-C LED flow-through reactor emitting at 280 nm. For pre-treated wastewater at a UVT of 70%, a 30 mL/min flow rate was sufficient for 3.7-log reduction of MS2. Fouling of the reactor by organic calcium, magnesium, phosphorus, and iron complexes decreased disinfection efficacy by approximately 1-log reduction of MS2 after 8 days. Fouling layers were completely removed with four hours of citric acid exposure.

This work also determined the inactivation rate of Ascaris lumbricoides eggs by UV-C LEDs at 280 nm compared to LPUV. Ascaris eggs are responsible for roundworm infections in 1.4 billion people worldwide, predominantly in low- to middle-income countries.

Oliver Lawal, AquiSense Technologies
Authors: Dr. Jennifer Pagan & Mitchel Hansen
17:00 - 17:20

New technologies often begin life at very high cost and low accessibility to some markets. There are often limits in place that prevent fully utilizing newer technology; but, over time costs come down, availability increases and new possibilities open up. Ultraviolet Light Emitting Diode (LED)technology can be seen to follow the same path. This technology has been used for applications such as sensors and low-flow water treatment products but not much more. It is the premise that advancements in output power and efficiency of UV-C LEDs will see it grow into new markets – including city level water treatment. The question is when.

Mercury-vapor technology solutions still have some inherent engineering limitations on/off cycle restrictions, footprint, warm-up time, heat dissipation, etc. On the other hand, UV-C LEDs function very differently and show huge potential for addressing real operational needs.

Expanding the topic into two key different applications we can further address how LEDs can treat municipal scale flows – wastewater and potable water. Also discussed will be the difference between medium-pressure lamps and low-pressure lamps and which technology is more of a risk from disruption. This paper will seek to balance the various factors (some obvious and some not obvious) that will determine the timing of the UV-LED technology entry into municipal sized treatment applications.

Karl Linden, University of Colorado-Boulder
08:40 - 09:00

While conventional UV technology has traditionally utilized low pressure (LP) mercury vapor lamps emitting at 253.7 nm due to familiarity and simplicity, the use of polychromatic UV sources such as medium pressure (MP) UV lamps is now equally likely.  This talk will draw on the characteristics of LP and MP systems to investigate differences in fundamental action of each source for mechanisms of inactivation, photoreactivation, production of disinfection byproducts and toxicity.  Attributes of each for UV-based advanced oxidation processes will also be detailed.  Finally, the emergence of UV-LEDs will be compared for their potential in revolutionizing the way UV processes are implemented in water treatment.  This work will draw on the findings from the Linden Research Group over the past 10 years to paint a comprehensive picture of how MP UV may or may not be beneficial in various situations.

Ian Mayor-Smith, Hanovia Ltd
Authors: Ian Mayor-Smith, Benjamin Woods, Jutta Eggers, Karl-Heinz Schon
09:00 - 09:20

The use of mercury based Low Pressure (LP) and Medium Pressure (MP) lamps are widely used Ultraviolet (UV) sources for the application of water disinfection. As a critical component within a water treatment reactor, the design and material component selection of a lamp can significantly impact a lamps output and continued performance over the life of its use. The ability to accurately measure the output of a lamp is critical not only to the design and modelling of system performance but also to ensure the consistency of manufacturing over time and verification of any replacement components during manufacturing or the need for complete lamp replacement, for example due to product obsolescence.

The replacement or change of such critical component following the validation of a UV system has the potential to change reactor performance impacting the validity of testing carried out. The USEPA UV Disinfection Guidance Manual (UVDGM) provides guidance for when a reactor should be considered needing to be “re-validated” which in respect of lamp assembly changes include; lamp arc length, lamp envelope diameter, lamp envelope UV transmittance (185-400nm), mercury content of the lamp and argon content of the lamp. All of these aspects should effect the lamp output and performance and in respect of MP lamps could affect spectral output and spectral isotropy. In addition critical component temperatures can provide early indicative measures to the likelihood of lamp life performance. The ability to accurately and repeatability measure total UV output, spectral output and temperature profiles of a lamp is key to the assessment of such a critical component.

This presentation describes the lamp measurement comparison between three partners within the ECO-UV consortium from lamp round robins. Additionally a novel customised lamp measurement apparatus is presented enabling the spectral and thermal characterisation of UV lamps both linearly and rotationally along the axis of a lamp. Spectral and thermal measurement comparisons are presented of lamp variants with nominally the same design criteria for the potential use in the same reactor design.

Karl-Heinz Schön, TZW: DVGW-Technologiezentrum Wasser
Authors: Karl-Heinz Schön, Jutta Eggers
09:20 - 09:40

Low Pressure (LP) Ultraviolet (UV) lamps are commonly used UV sources for water treatment. LP lamps can be sensitive to a changing water temperature, which can vary significantly with key factors being due to location and season. This can influence the operating and maintenance costs for the water suppliers due to a quicker lamp replacement and increasing power usage. Varying methods of performance stabilisation are used by lamp manufactures to address this challenge. Common approaches used are; custom designs for an expected narrow region of temperature variation, changing the amalgam filling (additional elements selected based on temperature requirements) to enable a broader range of temperature operation or the use of an appendix providing active thermal control with an electronic controller.

The harmonised DVGW-DIN/ÖNORM test protocol for LP UV-disinfection devices includes an in depth characterization of lamps, ballasts and other reactor components, partly with a varying water temperature requiring a standardised test reactor to be constructed.

First measurements in the test reactors showed variation in lamp output by up to + 20% for full powered amalgam lamps with a rising power consumption if the water temperatures changes from 7°C to 25°C. Other lamps, optimised for lower water temperatures, had a decreasing UV output of about 10 – 20 % within the same temperature range.

This presentation shows measurements of different LP lamp types - both vertically and horizontally installed - under controlled conditions. It compares the electric and optical characteristics of lamps both at full power and when dimmed to lower powers and is looking at the warmup times and step responses by different water temperatures.

Bruno Ferran, SUEZ
Authors: Bruno Ferran, Wei Yang
09:40 - 10:00

A batch of 18 low-pressure high-output germicidal UV lamps was continuously aged between July 2012 and December 2015. Aging was conducted in accordance with the “Lamp Age Factor Testing” section inside the 2nd edition of the NWRI/AwwaRF “UV Disinfection Guidelines for Drinking and Water Reuse”. Lamps were measured both in air and water to evaluate the progressive change in UVC output over time. Lamp testing in air was conducted in accordance with the first protocol for testing the UV efficiency of low-pressure and low-pressure high-output UV lamps by the Manufacturers’ Council of the International Ultraviolet Association (IUVA) published in 2008 (IUVA News Vol. 10 No. 1). Lamp testing in water was conducted using a single-lamp stainless steel UV reactor continuously fed with high transmittance temperature controlled water. Local UVC intensities were measured through water using an International Light IL1700 radiometer at five locations alongside the lamp axis covering the entire arc length of each test lamp. Intensity measurements in air and water were conducted at regular intervals throughout the aging and witnessed by a third party. The lamp aging factor (EOLL) was developed by comparison to reference (i.e. unaged) lamps to minimize experimental errors. This paper will present in details the protocols followed during this aging and the data obtained but most importantly this paper will demonstrate that lamp intensity measurements in air and water resulted in the same lamp aging factor at end of life.

Alois W. Schmalwieser, University of Veterinary Medicine
Authors: Alois W. Schmalwieser, Georg Hrischmann, Regina Sommer
10:40 - 11:00

The only practicable way to control the disinfection capability of an UV disinfection plant for drinking water all the time is to use a UV radiometer. According to the Austrian Standard M5873, this plant radiometer is a standardized part of each plant. The standard defines a so-called reference radiometer (RRM) as well. This is necessary because a plant radiometer has to be controlled periodically. A RRM is a hand-held device which has to fulfil high quality criteria and must be almost insensitive to environmental conditions. In this paper the principles of the concept behind the RRM are explained together with the requirements of such a device. Further on, the test methods are presented as well as a summary of test results from all RRMs developed during the past 15 years. It is shown that the radiation monitoring concept of the Austrian Standard was successfully practicable and that the international acceptance of the Austrian Standard is justified.

Gord Knight, Trojan Technologies
11:00 - 11:20

Quartz sleeves for ultraviolet (UV) lamps are widely used in water and wastewater disinfection applications. In most UV systems each UV lamp is accommodated in a lamp sleeve. The purpose of the sleeve is to provide a watertight seal around the lamp, to allow transmission of UV radiation into water, and to provide optimal thermal conditions for lamp operation. Sleeve UV transmittance (UVT) can be measured with a spectrophotometer, where scattering light is considered to be negligible. Sleeve fouling is a common phenomenon in UV disinfection. When sleeve fouls the UVT measurement using a spectrophotometer would underestimate the UVT value, i.e., one would obtain a lower UVT than the true value for a fouled sleeve. The UVT underestimation from the spectrophotometer method is primarily due to light scattering from fouling, and due to the difficulty for a spectrophotometer to collect scattered light. The underestimation of the fouled sleeve UVT measurement is not often discussed in the UV disinfection industry, probably due to the fact that reliable UVT measurement requires advanced spectroradiometer instrument setup and using a spectrophotometer with integrating sphere is not practical. The far-field (FF) radial irradiance measurement is discussed, which could accurately measure the true UVT value of a fouled sleeve.  A comparison between spectrophotometer method and FF irradiance method demonstrates the significant effect of sleeve fouling on UVT measurement.

Niklas Papathanasiou, sglux GmbH
Authors: Dr. Niklas Papathanasiou*, Gabriel Hopfenmüller, Michael Matalla and Dr. Tilman Weiss
11:20 - 11:40

In UV water purification applications UV sensors are monitoring the dosage of UV irradiation as according to ÖNORM and DVGW standards. sglux is manufacturing such sensors employing opaque synthetic quartz glass diffusors as entrance windows. This paper investigates the influence of high intensity UV irradiation on the transmission behavior of these diffusors. Different types of microporous quartz glass were used with varying number of bubbles per unit volume. The sensors were continuously monitored while irradiated by a 1kW medium pressure Hg lamp with an total UV irradiance of 1000mW/cm2 for 1000 hours. Before and after the aging period the total transmissions of the diffusors were measured.

A 15% signal decrease of sensors employing porous quartz glass could be observed within the first 200h of irradiation. Afterwards no further degradation was detectable with the current set-up. The total transmission and reflection of the sensor windows were measured before and after the aging experiment. The total transmission between 350nm and 500nm is reduced by 7-10% due to the 1000h stress while below 350nm the degradation becomes more severe. Up to 25% loss in transmission at 260nm is observed. These correlates well with the observed decrease of the sensor signal by 15%, which integrates the irradiance between 220nm and 390nm. Samples with clear quartz glass windows show nearly no loss in transmission between 350nm and 500nm, while about 3% loss in the UVC region is detected. The origin of the reduced transmission is not clear and was not in the scope of this study. However, the results prove, that it is necessary to perform an initial UV treatment of porous quartz glass to achieve stable signal reading for UV sensor employing such diffusors.

Stephanie Gora, University of Toronto
Authors: Stephanie Gora and Susan Andrews
11:40 - 12:00

Adsorption is a well established water treatment process used to remove organic contaminants, including disinfection byproduct (DBP) precursors. Existing adsorbents are effective but difficult to regenerate. Although most research on titanium dioxide (TiO2) has focused on the use of TiO2 in advanced oxidation processes (AOPs), it has both adsorptive and oxidative abilities. As with other AOPs, the degradation of complex organic molecules via TiO2 photocatalysis is a multistep process and full mineralization may not be achieved within an acceptable treatment time frame. This risk can be avoided by instead removing contaminants via adsorption to TiO2, separating the used TiO2 from the treated water, regenerating the materials via photocatalysis, then recycling the regenerated TiO2 back into the main treatment tank.  This two-stage process would rely exclusively on adsorption for contaminant removal from the water, thus avoiding the formation of undesirable byproducts during the treatment step. The key to making this process work is the development of a TiO2 material that can be removed from the treated water quickly.

Two engineered TiO2 nanomaterials were synthesized via a simple hydrothermal method. The nanomaterials had similar dimensions but differed in terms of surface characteristics, surface area, and crystal structure. Unlike the industry standard TiO2 nanoparticles, both engineered nanomaterials were easily removed from the treated water via settling or filtration. Experiments were conducted to evaluate the materials’ ability to remove DBP precursors from natural river water samples gathered from two Canadian water treatment plants via adsorption. The engineered nanomaterials differed in terms of their ability to remove DBP precursors. The adsorption of DOC, UV254, THM precursors, and HAA precursors by the engineered TiO2 nanomaterials fit a modified Freundlich adsorption isotherm model. Both materials were repeatedly regenerated via exposure to UVA light.

Wenjun Sun, Tsinghua University
Authors: Wenjun Sun, Wenjun Liu, Xudong Liang, Shuai Wu
13:20 - 13:40

III-Nitride material based UV-C LEDs (light emitting diodes) technology has been widely studied recent years. This paper will discuss the high efficient and high power AlGaN UV-C LEDs grown on sapphire substrate. The maximum output power over 176mW at 1A high current for a single UV-C LED device has been achieved as shown in figure 1, which by far is the highest output power reported from a single 4040 mil device, and the voltage is only 5.8V, corresponding to a very low power consumption of less than 1 Watt, which is obviously less than mercury lamp power consumption. This high performance of UV-C LEDs light source is combined with high quality epitaxial material with thicker AlN template and optimized device fabrication and packaging technique for better light extraction. Besides these high current performance of the UV-C LEDs, the 100mW device from a single 6868 package at only 300mA was also tested. The reliability of similar devices at different current were tested, and over 10000 hrs long life time was achieved for 100mA stress and over 5000 hrs lifetime for 300mA current stress.  Such good reliability and high power performance demonstrated the great potential for UV-C LEDs to replace conventional mercury lamp. This paper will also present the water disinfection effects of this device.

Naoki Shibata, Nikkiso
Authors: Dr. Naoki Shibata, Tetsuhiko Inazu, Shinya Fukahori, and Dr. Cyril Pernot
13:40 - 14:00

III-Nitride material based UV-C LEDs (light emitting diodes) technology has been widely studied recent years. This paper will discuss the high efficient and high power AlGaN UV-C LEDs grown on sapphire substrate. The maximum output power over 176mW at 1A high current for a single UV-C LED device has been achieved as shown in figure 1, which by far is the highest output power reported from a single 4040 mil device, and the voltage is only 5.8V, corresponding to a very low power consumption of less than 1 Watt, which is obviously less than mercury lamp power consumption. This high performance of UV-C LEDs light source is combined with high quality epitaxial material with thicker AlN template and optimized device fabrication and packaging technique for better light extraction. Besides these high current performance of the UV-C LEDs, the 100mW device from a single 6868 package at only 300mA was also tested. The reliability of similar devices at different current were tested, and over 10000 hrs long life time was achieved for 100mA stress and over 5000 hrs lifetime for 300mA current stress.  Such good reliability and high power performance demonstrated the great potential for UV-C LEDs to replace conventional mercury lamp. This paper will also present the water disinfection effects of this device.

In comparison to traditional UV light sources, AlGaN based deep ultraviolet (UV) light-emitting diodes (LEDs) have many advantages, such as, tunable wavelength (from 365 to sub 250 nm) and low environmental impact. Deep UV LEDs are promising to replace gas-discharge mercury based lamps and are able to address practical applications such as surface, air and water disinfection, UV coating and printing, medical phototherapy, and so on.

As a result of recent progress of high quality AlGaN template on sapphire substrate and optimized device structure, production grade UV LEDs have been realized with both high output power (> 50mW) and high reliability (L70 > 10,000h) under 300 nm deep UV region.

In this presentation, we will describe the challenges for high efficiency UV LEDs and introduce innovative approaches for manufacturing high output power and high reliability deep UV LEDs.  In addition, future challenges and proposed solutions will be discussed.

Richard Simons, AquiSense Technologies
Authors: Dr. Richard Simons, Dr. Jennifer Pagan, Oliver Lawal
14:00 - 14:20

Continued developments of UV-C LEDs have opened opportunities for integration into previous no-go areas for more traditional technologies. Since the early 1900’s the only way to effectively generate germicidal UV has been through mercury vapour or excimer lamps; the electrical, mechanical, and thermal requirements of these devices have often hampered their implementation. The distinct differences in UV-LED design present an opportunity for a paradigm shift, taking UV disinfection where it could never have gone before.

One such environment is space exploration, where mercury-based methods had been previously rejected, but which has now seen the application of UV-C LEDs in surface disinfection aboard the ISS. Plans are also underway for future air and water disinfection systems through several different projects. UV-C LEDs offer a disinfection solution which is chemical-free, durable, lightweight, and has a small footprint; all of which carry significant importance for future launch scenarios.

AquiSense Technologies was invited to collaborate in a pan-European project to deliver a future system for the safe storage, rapid quality monitoring, and on-demand disinfection of drinking water for future space exploration missions. We will provide an update on the current state of development and lessons learned through this collaborative process.

Exploration of new environments leads to unexpected challenges which must be overcome, these challenges breed innovation and can be used as a springboard to advance product design well beyond the initial project scope.

Giuseppe Spataro, University of Siena
Authors: G. Spataro, G. Cevenini, L. Mariani, N. Nante, D. Rosadini, M. Tani, G. Messina
14:20 - 14:40

Introduction

After 201 years from its invention, the stethoscope remains still the most used medical device and symbol of medicine. Surprising its disinfection, after each use, is often neglected and it has been proved that the microbial contamination on its membrane is similar to the one of the hands: one of the major cause of healthcare associated infections.

Aims are: i) to verify the easiness of use of a wearable device which automatically disinfects the stethoscope membrane; ii) to test its effectiveness in hospital.

Methods

A pilot study, with a pre-post intervention design, was conducted in a private clinic from November 2016 to May 2017. Several wearable medical devices, which attach the head of the stethoscope and automatically disinfect its membrane with UV-C LED (wavelength 275nm), were provided to health professionals. Colony Forming Unit (CFU), on petri dishes, measured microbial contamination with/without the use of the devices.

The average percentage of CFU reduction and 95% confidence intervals (CI), was computed through the bias corrected and accelerated bootstrap method, for both groups of no-treated and UV-treated stethoscope membranes. CFU were also compared between groups, using the non parametric Mann-Whitney test for independent samples. Statistical analysis were performed with SPSS.

Results

274 samples were taken from stethoscope membrane: 156 were recorded without the use of the UVC device, while 116 had the treatment. The average percentage of CFU reduction was 94% [CI 91,46-97,45%]. About 10 times more contaminated membranes were associated in the ward of “permanent vegetative state”. Highly significant statistical differences of CFU were found between no-treated and treated membranes (p<0.001).

Conclusion

The use of the automatic and wearable UV-C device resulted practical and easy for the health professionals. UV-C irradiation on stethoscope membranes showed a marked reduction of bacterial load, confirming its high disinfection capacity.

Pierrick Boulay, Yole Développement
Authors: Pars MUKISH, Pierrick BOULAY
14:40 - 15:00

UVA LEDs continue to progress in the UV curing space. Continuous improvement of device performance coupled with price reduction has allowed the technology to be increasingly adopted in UV curing applications. Penetration of UV LEDs is increasing but we observe differences in adoption rates depending on application. Small size and low speed applications like spot adhesive and digital inkjets have the highest adoption rate, and most new developments use UV LEDs. This is due to the small module size and low irradiance level needed that limits the extra cost of integrating UV LEDs compared to the total price of systems like inkjet printers. On the other hand, applications that need high speed processes and/or high levels of irradiance such as screen printing or coating applications have lower adoption rates. This is because UV LED performance is not yet good enough to fully replace traditional mercury lamps. To improve performance of UV LED curing systems, new UV LED devices with higher irradiance and longer lifetimes must be developed. However, the industry is also focusing now on other components and functions of the system such as optics, thermal management and drivers. Now that irradiance of UV LED is good enough for most UV curing applications, manufacturers work on other contributors to system performance to develop the next generation of UV curing systems. This trend is also emphasized by the development of customized systems as each application has its own requirements, such as level of irradiance, working distance, and integration in the process. This shift in product development has attracted many new players who want to benefits from the UV LED curing boom. They come from different directions, including the printing equipment industry, traditional curing industry, the visible LED lighting industry, the optic industry, and the photonic industry. Some companies involved have even been specifically created to capitalize on this opportunity, including start-ups and spin-offs.

On the other hand, the UVC LED industry is still small but strong growth is expected in the next 18 months due to dramatic price reductions. In 2016 prices are 1/8-1/10 of what they were in 2015. This has been triggered by the industry’s development, its transition to mass production and improved device performance. With most of the industry believing that $1-$4/mW is the price that would trigger mass market adoption we are getting close to a UVC LED market boom. Another positive sign is that most UVC LED manufacturers are now focusing on developing cost-effective solutions rather than improving device power output. In parallel, the UVC LED industry continues to work on increasing lifetime and developing lower wavelength devices, below 280nm.

In this presentation, we are going to highlight status of the UV LED industry and recent trends but also discuss recent technology and market highlights in both UVA LED and UVC LED fields.

Margaret M. Busse, Purdue University
Authors: Margaret M. Busse, John Maiyo, Emmanuel C. Kipkorir and Ernest R. Blatchley III
16:20 - 16:40

Solar UV disinfection systems have been demonstrated to be effective for treatment of drinking water; however, the implementation of such systems in areas lacking access to safe water provides non-technical challenges that influence their adoption.  To explore these issues, we partnered with a local business, Maji Safi (Eldoret, Kenya), to evaluate the potential for introduction of UV disinfection systems into their business model.  Maji Safi currently provides water treatment options to communities, schools, and households in western Kenya by drilling wells, installing slow sand filters (SSF) with chlorination, selling household ceramic filters, and educating on safe water practices.  This study aimed to provide data on public perception and viability of UV disinfection with commercial reactors based on pre- and post-installation surveys, microbial inactivation experiments, and operational observations.  The study was initiated by examining the perception of drinking water and water treatment among three local schools that each had improved, but not safe, water sources.  The schools each received a SSF system followed by a commercially-available, household-scale UV disinfection system powered by a low-cost solar panel/battery setup.  To further contextualize the perception of these systems, one school also received a chlorination system with two drinking water taps, one from each disinfection method, which were used to directly compare usage.  Experiments were also conducted to examine the potential for the use of solar UV directly in continuous-flow reactors based on compound parabolic collectors (CPCs).  Measurements included comparisons of UVB reflectance among materials that could be used to line the CPC, as well as the potential benefits of solar tracking and the feasibility of local construction.  The comparisons among commercial UV reactors, CPC systems, and chlorination will inform design of disinfection systems for communities based on consumer preferences.

Jantinus Bruins, WLN B.V.
16:40 - 17:00

Low pressure UV disinfection (dose amounts 40 mJ.cm-2) is applied at drinking water facility De Punt (Groningen, The Netherlands). Until 2011, UV-disinfection was the final treatment step, i.e., just before the clean water reservoirs. Monitoring the Biofilm Formation Rate (BFR) of the feed and filtrate of the UV reactors, showed that LP-UV increases the BFR from approximately 1.4 to circa 12 pg. ATP.cm-2.d-1 (results presented at the 5th IUVA World Congress, 2009). These results imply that even low pressure UV substantially elevates the regrowth potential during drinking water treatment under local conditions. This is unwanted since regrowth might lead to higher concentrations of pathogens, e.g. Legionella and other unwanted (micro)organisms in the distribution networks. To limit the regrowth potential of the LP UV disinfection step, following measures were taken: changing the reactor configuration, from a ‘single 40 mJ.cm-2 dose reactor’, into a ‘double, sequential, 20 mJ.cm-2 dose reactor system’; preventing the UV disinfection from being the final treatment step, i.e., LP-UV was placed before Slow Sand Filtration (SSF) as final treatment step. Monitoring the BFR of feed and filtrate of the UV reactors, after the configuration change, showed only a limited increase of the BFR from approximately 3.1 to circa 4.2 pg. ATP cm- 2 d1. This is below the value of 10 pg. ATP.cm-2.d-1, which is found to be an acceptable value regarding the production of biological stable drinking water, without regrowth potential concern. Finally SSF lowered the BFR to a value of approximately 1.3 pg. ATP cm- 2 d1. From this research it can be concluded that changing the reactor configuration is an effective measure to lower the BFR and therefore an excellent move to limit the regrowth potential of the distributed drinking water.

Harold Wright, Carollo Engineers
Authors: Traci Brooks, Mark Heath, Linda Hills, Jeff Adams
08:40 - 09:00

The UVDGM states that an action spectra correction factor (ASCF) should be applied to validation conducted on MP UV reactors to account for differences in the wavelength response of the validation microbes and target pathogens. While the UVDGM provides an approach for determining the ASCF, the approach is conservative for many UV reactors, increasing UV disinfection costs. To address this issue, the WRF Project 4376 developed tables of ASCF values that could be broadly applied to MP UV reactors and guidance for using CFD-based UV dose models to calculate UV reactor specific ASCF values. However, because commercial MP reactors use UV sensors that do not monitor below 240 nm, the approaches used to determine ASCFs does not take credit for pathogen inactivation below 240 nm.

As an alternative, MP reactors can use validation equations that use low and high wavelength UV sensors to monitor the contribution to UV dose delivery by wavelengths below and above 240 nm, respectively. With this approach, microbe log inactivation is predicted as the sum of low (240 nm) log inactivations. The high wavelength component is predicted as a function of UVT at 254 nm and a high wavelength combined variable (SH/S0H)/(QxDLxASCFH) where SH/S0H is the lamp output defined by a high wavelength UV sensor and ASCFH is a high wavelength ASCF. Similarly, the low wavelength component is predicted as a function of a low wavelength UVT and a low wavelength combined variable defined as (SL/S0L)/(QxDLxASCFL) where SL/S0L is the lamp output defined by a low wavelength UV sensor and ASCFL is a low wavelength ASCF.

This approach has the advantage of accounting for target pathogen inactivation at wavelengths below 240 nm, which can significantly reduce the costs of UV disinfection, in particular for adenovirus inactivation credit by up to 3-fold.

Mark Heath, Carollo Engineers
Authors: M. Heath, H. Wright, T. Brooks, J. Eggers, T. Schwarzenberger, K. Schön, I. Mayor-Smith, P. Ropic
09:00 - 09:20

This presentation is the second in a two-part series discussing recent cooperative efforts between Carollo's UV Validation Facility in Portland, OR, USA and the DVGW-TZW testing facility in Germany. It builds on the presentation “UV Validation-Bridging the Atlantic Divide Part I: DVGW comes to the United States”.

To navigate the complex global regulatory environment of municipal UV disinfection, equipment manufacturers must conduct extensive and costly validation testing under various testing protocols accepted by different regulatory jurisdictions. With validation costs ranging from $50,000-$100,000 and more for a single validation, these costs add up quickly. To gain return on investment, expanding from the relatively narrow municipal disinfection market into the broader industrial and commercial markets is critical. Validation techniques for municipal applications form the basis for acceptance in many industrial/commercial applications as well. For example, the Model Aquatic Health Code (MAHC) for protection of swimming pool and water activities in the United States has adopted the UVDGM approach for systems using UV disinfection. Providing cost effective validation options for UV manufacturers has been the focus of the cooperative effort between Carollo and TZW.

This presentation will focus on UVDGM testing conducted by Carollo and TZW staff at the DVGW testing facility in October 2016 and March 2017. This testing represents the first time UVDGM testing has been performed in Europe. Testing was conducted on a Hanovia Pureline H+ MP reactor and two Berson Amaline LPHO reactors. Potentially catastrophic complications were encountered with the UV dose response of the MS2 phage used at the beginning of testing. The combination of the test water and the LSA used to adjust the UVT resulted in an  unstable MS2 dose response giving as much as two times greater log inactivation than expected for a given UV dose. The issue was resolved through a detailed set of experiments to evaluate combinations of test water and the UV absorbers used to modify UV transmittance. The solution was to change the UV absorber from LSA to SuperHume. This observation has significant implications for validation testing. This presentation will discuss these complications and their resolution, as well as summarize the efforts needed to establish validation testing under new protocols at an existing validation facility.

Mark Heath, Carollo Engineers
Authors: M. Heath, H. Wright, T. Brooks, J Eggers, T Schwarzenberger, C. Bokermann, R, Bemus
09:20 - 09:40

Ultraviolet disinfection for drinking water treatment has gained worldwide acceptance over the past two decades with its routine use in Europe extending decades earlier. Regulatory approval is governed by different organizations in different parts of the globe. In Europe, regulatory acceptance is primarily governed by the German Technical and Scientific Association for Gas and Water (DVGW) and the Austrian Standards Institute (ONORM), while in the United States the US Environmental Protection Agency (EPA) published the 2006 UV Disinfection Guidance Manual (UVDGM) for compliance with the standards of the Long-Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR).

To demonstrate performance of UV disinfection equipment for regulatory acceptance, full-scale validation testing is required. Validation techniques differ between the various regulatory jurisdictions, and to gain global regulatory acceptance, UV equipment vendors must perform costly validation testing to meet the requirements of each. This often means vendors must ship UV equipment to more than one of the handful of validation test facilities worldwide, incurring substantial costs associated with installation, functional testing, biodosimetry testing and data analysis procedures stipulated by each regulatory jurisdiction. With the cost of validation ranging from $50,000-$100,000 and in some cases substantially more for a single validation, there is a significant barrier for UV vendors achieving global regulatory acceptance of their equipment. In addition, for large capacity UV equipment, there are few options for validation facilities capable of the required flow rates for testing.

Recent cooperative efforts between Carollo's UV Validation Facility in Portland, OR, USA and the DVGW TZW testing facility in Germany allowed a major UV vendor to test one of their higher capacity UV units under both US EPA UVDGM and DVGW testing protocols. The benefits of this cooperative effort allowed the UV vendor to achieve dual validation at a maximum flow capacity not achievable by the German test facility, at the same time realizing substantial cost savings by testing only at one facility. By the end of 2016, further cooperative testing will be conducted on three UV reactors form two UV vendors at the DVGW TZW test facility under the US EPA UVDGM protocols. This will allow the two European vendors to conduct testing locally, thus reducing the costs associated with supporting validation testing in the United States, while achieving certification under US EPA UVDGM protocols.

This presentation will discuss the primary differences between various UV validation protocols and UV dose monitoring philosophies as well as summarize the cooperative testing conducted to date and that proposed moving forward that will provide significant cost and time saving benefits to the UV industry

Yuri Lawryshyn, University of Toronto
Authors: Colin Powell and Yuri Lawryshyn
09:40 - 10:00

Flow induced mixing in UV wastewater reactors has been identified as an important factor in enhancing reactor performance. Current methods for determining mixing are either qualitative or are based on ideal mixing states, and cannot be correlated directly with reactor performance. In this work, the Shannon entropy concept is used to quantify mixing in UV reactors and then modified with a weighting based on the distance a particle in a UV reactor travels from its initial injection position. Using a Monte Carlo simulation, it is shown that the Shannon entropy-based mixedness can increase or decrease when the number of regions by which the system is divided are increased, contrary to current literature. As well, it is proven numerically and theoretically that the mixedness will indeed converge to a finite value with a sufficiently high number of regions. A strong correlation between mixedness and UV reactor performance was shown for two idealized single-lamp UV reactors. This means that the mixedness values can be used as a proxy for reactor performance, allowing modelers to significantly reduce the computational cost required when comparing the performance of different reactor configurations, especially in the case where large eddy simulation (LES) computational fluid dynamics (CFD) models are utilized.

Ernest R. Blatchley III, Purdue University
Authors: Yousra M.H. Ahmed, Xing Li, Angela P. Ortiz, Chengyue Shen, O. Karl Scheible, E.R. Blatchley III
10:40 - 11:00

Most contemporary methods for design and validation of UV disinfection systems do not adequately account for the physical phenomena that govern the performance of these systems.  One consequence of this approach is that UV systems designed and operated based on these methods tend to be over-sized relative to their respective treatment requirements.  Issues that are often ignored or greatly simplified in the design of UV disinfection systems, as well as other applications of UV radiation water (or air) treatment include accounting for the UV dose distribution delivered by the reactor, and for variability in process performance.

To address these issues, a stochastic method has been developed for application of numerical models based on integrated application of computational fluid dynamics and irradiance field (fluence rate field) modeling, or so-called CFD-I models.  CFD-I models provide a method whereby the dose distribution can be accounted for; however, in most applications of CFD-I to UV systems to date, the solution developed by these approaches has been deterministic.  In this research, predictions of variability in process performance were developed through the use of a stochastic process, whereby variability in input parameters was accounted for in the simulation.

The results of these simulations were predictions of process performance, including variability, for UV disinfection systems over a range of operating conditions that were representative of their respective operating envelopes.  This method, which is most relevant for large UV systems, provides a protocol for optimizing UV disinfection system design by developing a link between operating conditions and performance, including variability.  As such, this method offers the potential to reduce design and operating costs, while improving process reliability.

Harold Wright, Carollo Engineers
Authors: Mark Heath, Traci Brooks, Linda Hills, and Jeff Adams
11:00 - 11:20

The 2006 UVDGM describes two approaches for UV dose monitoring, namely the UV intensity setpoint approach and the calculated dose approach. While the UV intensity setpoint and its validation were well defined when the UVDGM was published, there was less information and experience on how to implement the calculated dose approach. This knowledge gap has since been addressed through projects funded through the Water Research Foundation and extensive experience with validation testing.

CFD-based UV dose models and validation data show that microbe log inactivation can be predicted as a function of UVT and a combined variable defined as (S/S0)/(Q DL) where S/S0 is the relative lamp output indicated by UV sensors, Q is the flow, and DL is the UV dose per log inactivation of the microbe whose log inactivation is predicted. In theory, UV dose monitoring algorithms using the combined variable can be calibrated using UV validation conducted with MS2 phage and then used to directly predict the log inactivation of the regulated pathogens, namely Cryptosporidium, Giardia, and viruses, by using the DL values specific to those pathogens. In practice, the equation should be calibrated using validation data collected using MS2 and T1UV phage. The validation report can then provide analysis which shows that the equation calibrated using MS2 predicts T1UV log inactivation and vice versa. This analysis provides confidence that the calculated dose approach using the combined variable can directly predict the log inactivation of the targeted or regulated pathogen effectively.

This approach has a number of benefits for PWSs and state regulators. It eliminates the need to apply a correction factor, referred to as the RED bias factor in the UVDGM, when determining the validation factor. It can also reduce the cost of UV validation testing and provide more cost effective selection and implementation of UV technologies.

Yuri Lawryshyn, University of Toronto
Authors: Yuri Lawryshyn, Ying Zhang, Jinpeng Wang, Monika Patel and Ron Hofmann
11:20 - 11:40

The use of UV disinfection for achieving 4-log virus (adenovirus) credit is gaining interest in drinking water applications. As specified in the United States’ Ultraviolet Disinfection Guidance Manual (UVDGM), the UV dose for 4-log virus credit is 186 mJ cm-2, however, there are practical issues associated with validating UV systems to such high doses. A recent paper by Lawryshyn and Hofmann (2015) utilized theory to show that the UV reduction equivalent dose (RED) of two UV reactors in series can be considered to be additive under certain conditions and, specifically, additivity is assured, even under worst case mixing conditions between the two reactors, if the test organism’s UV sensitivity is twice that of the target organism.

The objective of this study is to experimentally demonstrate the validity of the UV dose additivity theory by comparing the RED based on bioassay testing using MS2 bacteriophage for a single reactor to that of the RED of two reactors in series based on bioassay testing using bacillus pumilus (B. pumilus). Since the UV sensitivity of MS2 is more than twice that of B. pumilus, dose addidivity should be assured. Two configurations are tested; one with essentially complete mixing between the reactors and the other where dose paths through the first reactor are positively correlated with those of the second reactor. Numerical CFD models are utilized in addition to the experiments to predict the dose correlations and to compare the bioassay results with the numerical models.

This presentation will be of interest to UV manufacturers, regulators, and professionals responsible for UV dose validation, as it will provide the scientific basis for achieving high UV doses by using reactors in series.

Yuri Lawryshyn, University of Toronto
Authors: Yuri Lawryshyn
11:40 - 12:00

The combined variable, defined as the ratio of the power output percent (based on sensor readings) divided by the flow rate and organism sensitivity, has been shown to provide a strong linear relationship between it and the log inactivation of a given UV reactor. The combined variable approach (CVA) stems from the fact that the log inactivation (LogI) for a single microbial path will be directly proportional to the UV lamp intensity and inversely proportional to the flow rate and organism sensitivity. A key assumption of the CVA is that the UV dose distribution scales linearly with lamp output power and inversely with flow rate. While the first assumption is generally accepted, the second is further based on the assumption that particle trajectories through the UV reactor essentially follow the same paths at varying flow rates. Work in fluid mechanics has shown that obstructions in flow generally create vortices and the frequency and magnitude of the vortices change with flow rate. Thus, the assumption of scaling the dose distribution with flow rate may not be valid under certain circumstances. This work examines the merits and limitations of the CVA by utilizing both theory and numerical experiments. The theory is developed by conceptualizing a simple UV reactor with multiple particle trajectories whose effective water layers, defined as the average distance between the particle and the lamp sleeve, are assumed to be random variables and the impact of the variance of the effective water layers on the CVA is examined. Numerical experiments are developed using both a conceptual theoretical reactor and a realistic computational fluid dynamics (CFD) based reactor. Aspects associated with the assumption of the dose distribution scaling inversely with the flow rate are examined.

Mengkai Li, Chinese Academy of Sciences
Authors: Mengkai Li, Zhimin Qiang, Ernest R. Blatchley III
12:00 - 12:20

Ultraviolet (UV) photoreactors based on medium-pressure (MP) mercury lamps have been widely used in water and wastewater treatment. The performance of these photoreactors is strongly influenced by the spatial photon fluence rate (PFR) distributions which are wavelength-dependent across the most effective range of 200-300 nm; to date, this aspect has not been well understood. To address this issue, PFR distributions in an MP lamp photoreactor with various UV transmittance (UVT) conditions were measured using a 360-degree response micro-fluorescent silica detector (MFSD). To accurately express the optical behavior in an MP photoreactor, PFR, MFSD response PFR (PFRMFSD), and effective germicidal PFR (PFRGER) were defined and compared. The measured axial and radial PFRMFSD values agreed well with the corresponding results from a simulation model (UVCalc®). The PFR and PFRGER were obtained from the measured PFRMFSD by using correction factors calculated by the simulation model. Under identical UVT254 conditions (75% and 85%), the weighted average PFRGER values were 13.318.7% lower than the corresponding PFR values in two test media (copper sulfate and potassium acid phthalate solutions), indicating that PFRGER, rather than PFR should be used in MP photoreactor design to meet the disinfection standards. Based on the measured lamp output, medium absorption spectrum, MFSD response, and microbial deoxyribonucleic acid response spectrum, the detailed relationships between the PFR, PFRMFSD, and PFRGER were elucidated. This work proposes a new method for the accurate description of wavelength-dependent PFR distributions in an MP photoreactor, thus providing an important tool for the optimal design of polychromatic UV photoreactors.