Read JEESUVPerfPaper.pdf text version

303

NOTE / NOTE Pulsed UV lamp performance and comparison with UV mercury lamps1

Raymond Schaefer, Michael Grapperhaus, Ian Schaefer, and Karl Linden

Abstract: Pulsed lamps based on electric discharges in xenon are of interest for water treatment because they are free of mercury, have instant-on capability, and may provide enhanced effects due to the high irradiance of pulses or spectral differences. This study provides quantitative comparisons of standard mercury UV lamps with both a commercial flashlamp and a pulsed surface discharge lamp. Unlike mercury lamps, the UV performance of pulsed lamps is a function of operating parameters. In this study the measured UV efficiency of a flashlamp, with a specified practical lifetime, increased as the pulse length decreased, from 4.4% at 796 µs to 9.0% at 71 µs. The surface discharge lamp, which overcomes limitations of flashlamps, had a measured UV efficiency of 17% at 12 µs. In comparison, standard commercial low pressure and medium pressure mercury lamps evaluated in this study had UV efficiencies of 34.6% and 12.2%, respectively. Key words: pulsed UV, water treatment, flashlamp, surface discharge, mercury. Résumé : Les lampes à impulsions basées sur les décharges électriques dans le xénon sont intéressantes pour le traitement de l'eau puisqu'elles ne contiennent pas de mercure; elles possèdent la capacité d'être mises en marche instantanément et elles peuvent avoir des effets accrus en raison de la haute irradiance des impulsions ou des différences spectrales. La présente étude compare quantitativement les lampes standards UV à la vapeur de mercure avec des lampes-éclairs commerciales et une lampe à décharge superficielle à impulsions. Contrairement aux lampes à vapeur de mercure, le rendement UV des lampes à impulsions est fonction des paramètres de fonctionnement. Dans cette étude, l'efficacité mesurée d'une lampe-éclair UV, ayant une durée de vie utile spécifiée, augmentait à mesure que diminuait la longueur de l'impulsion, de 4,4 % à 796 µs à 9,0 % à 71 µs. La lampe à décharge superficielle, qui comble les limites des lampeséclairs, présentait une efficacité UV mesurée de 17 % à 12 µs. En comparaison, les lampes commerciales standards à faible et à moyenne pression de vapeur de mercure évaluées lors de cette étude présentaient respectivement des efficacités UV de 34,6 % et de 12,2 %. Mots-clés : UV à impulsions, traitement de l'eau, lampe-éclair, décharge superficielle, mercure. [Traduit par la Rédaction]

Introduction

Ultraviolet light is a proven means for disinfection and remediation of water, as well as many other applications. UV is used widely in Europe for water disinfection and its use is becoming more widespread in the US with EPA acceptance of UV disinfection of drinking water, as indicated by two 2006 regulations, the Stage 2 Disinfection Byproducts (DBP's) Rule (EPA 2006a) and the Long Term 2 (LT2) Enhanced Surface Water Treatment Rule (EPA 2006b). UV may be preferable over more traditional

chemical disinfectants because of their tendency to produce disinfection by-products of regulatory concern. Furthermore, UV has been shown to be highly effective in the inactivation of difficult to treat protozoan pathogens of health concern, such as Giardia lampda and Cryptosporidium parvum (Linden and Mofidi 2003). Mercury lamps are the standard for most UV commercial applications. However, environmental and health concerns from potential mercury contamination due to lamp failure, lamp dis-

Received 1 September 2005. Revision accepted 20 November 2006. Published on the NRC Research Press Web site at http://jees.nrc.ca/ on 23 May 2007. R. Schaefer,2 M. Grapperhaus, and I. Schaefer. Phoenix Science & Technology, Inc., 27 Industrial Avenue, Chelmsford, MA 01824, USA. K. Linden. Department of Civil and Environmental Engineering, Duke University, Durham, NC 27708, USA. Written discussion of this note is welcomed and will be received by the Editor until 30 September 2007.

1

This note is one of a selection of papers published in this special issue on application of ultraviolet light to air, water and wastewater treatment. 2 Corresponding author (e-mail: [email protected]).

J. Environ. Eng. Sci. 6: 303­310 (2007) doi: 10.1139/S06-068 © 2007 NRC Canada

304

J. Environ. Eng. Sci. Vol. 6, 2007 Fig. 1. Light measurement set-up.

posal, and the necessary warm-up time of mercury lamps (where quick response is an issue, such as in Homeland Security applications) has led to interest in pursuing the efficacy of alternative non-mercury based lamp technologies. Pulsed xenon flashlamp technology is mercury free, has instant-on capability, and has received attention for a wide range of disinfection and remediation applications (Liang et al. 2003; Mofidi et al. 2001; Wekhof 2003; Weber and Scheible 2003; MacGregor et al. 1998; Rowan et al. 1999; Krishnamurthy et al. 2004; McDonald et al. 2000; Sharma and Demirci 2003; Hillegas and Demirci 2003). Furthermore, some reports indicate possible increases in inactivation rates of microbes and reduction rates of organic chemicals from the pulsed nature of the light (Wekhof 2003; Grapperhaus et al. 2005; Sharma and Demirci 2003; Hillegas and Demirci 2003). However, quantitative UV spectra, flashlamp efficiency and performance comparisons with mercury UV lamps have generally not been reported (Linden & Mofidi 2003 AWWARF). As a result, evaluations of pulsed lamp performance are often qualitative and without the data needed for quantitative comparisons. The UV efficiency and performance of mercury lamps has been fairly well characterized. The UVC efficiency of medium pressure mercury (MPM) lamps, for instance ranges from 15 to 20% over the range of practical lamp operating conditions (Phillips 1983). In comparison, pulsed lamps have UV efficiencies varying by more than a factor of two while the lifetime varies over several orders of magnitude. For commercial flashlamps, the lifetime is related to its electrical operating point in relation to an "explosion limit" characteristic of the lamp. The UV efficiency depends on pulse length, with flashlamps commonly operating with pulse lengths ranging from a few tens of microseconds to over a millisecond. One purpose of this study is to evaluate one pulsed nonmercury flashlamp and to provide a quantitative demonstration of its variation in UV efficiency with both pulse length for operating conditions with practical lifetime, as well as for high stress conditions in which the UV efficiency is higher, but lifetime is reduced. In addition, this paper provides quantitative information on a second pulsed lamp technology; a new pulsed surface discharge (SD) lamp (US Patent numbers 5,945,790 and 6,724,134) that may operate at stress levels well above the explosion limit of flashlamps while producing greater UV efficiency (Grapperhaus et al. 2005; Schaefer 2003, 2004; Schaefer et al. 2005). The study includes measurements of the UV spectra and efficiency of a standard low pressure mercury (LPM) and a medium pressure mercury (MPM) UV lamp, as well comparisons of these spectra and efficiencies with pulsed UV lamps. Results presented here summarize quantitative measurements for the pulsed and continuous-wave UV lamps and provide a baseline that may be used for future reporting of research using pulsed lamps for water treatment applications.

Lamp

1.83 m

Spectrometer

Experimental methods

Quantitative UV measurements Multi-channel spectral measurement of light output is required for pulsed lamps in order to capture the entire broadband

spectrum of the pulse. The measurements reported here used an Ocean Optics (model USB2000), multi-channel spectrometer with 2048 elements over the wavelength range of 200­800 nm, with a resolution of 0.29 nm. The spectrometer was calibrated using a pair of lamps with NIST traceable calibrated spectral irradiance (Oriel Model 63345 deuterium lamp for 200­400 nm and Oriel Model 63358 quartz tungsten halogen lamp for 400­ 800 nm). Light was collected for a fixed time interval, which produced a spectral irradiance calibration for the spectrometer (J cm-2 nm-1 count-1 ). For pulsed light measurements, calibrated neutral density filters were used to ensure a linear and non-saturated response of the spectrometer. Also, for the pulsed lamps, five pulses were measured and averaged to determine UV efficiency. For lamp output measurements the spectrometer was perpendicular to the lamp at a distance of 1.83 m, where the entire lamp is within the acceptance angle of the spectrometer slit and the irradiance is in the 1/r2 regime as shown in Fig. 1. Lamp spectra measurements were made, with a second background measurement identical but with the lamp off. The measured spectrum with the background subtracted was used for the net count versus wavelength. The counts were then adjusted for the transmission of any neutral density filters used, and converted to irradiance using the measured spectrometer sensitivity with wavelength. The total spectral output of the lamp over one pulse, S, (J nm-1 ) is the integral over 4 steradians of the irradiance at a distance r, S = r 2 4 I d . Thus, the measured irradiance at one position is related to the lamp spectral output by I = S/(f r 2 ), where f is the angular factor between the measurement location and the total lamp output. For an isotropic source (uniform angular distribution) f is 4 , while for a Lambertian source (cosine angular distribution) f is 2 . Using measured angular output, the flashlamp factor is 3.6 , while for the surface discharge (SD) lamp f is 3.8 and falls within the usual range observed in pulsed sources (Marshak 1984).

© 2007 NRC Canada

Schaefer et al. Table 1. Flashlamp design relationships. Description Capacitor energy Pulse length Dynamic impedance Damping parameter Circuit impedance Explosion limit Lifetime Pulse repetition frequency Life Equation Eo = 1/2CV Tp = LC Ko = 0.38 /d(P )0.2 a = Ko / Vo /Zo = 0.8 Zo = L/C Eex = 94d Tp NL = (Eex /Eo )8.5 f = P /Eo

2

305 Table 2. Flashlamp system parameters. Units Joule µs Ohm-Amp1/2 Dimensionless Ohm Joule Pulses Hertz Hours C (µF) 97 197 370 566 756 Parameter C L d P P Description Capacitance Inductance Arc length Lamp bore inner diameter Xenon fill pressure Continuous power Value Variable Variable 15 12 450 1800 Units µF µH cm mm Torr Watt

Table 3. Pulse length variation test parameters for 3000 h life. Life = NL /f V (kV) 1.9 1.7 1.5 1.5 1.4 Eo (J) 167 281 427 607 677 L (µH) 5.6 13.9 32.4 52.5 85 Ipk (kA) 3.4 2.9 2.4 2.2 1.9 tel (µs) 71 164 344 556 796 UV efficiency (%) 10.5 7.8 6.7 6.1 5.5 PRF (Hz) 9.6 5.7 3.7 2.6 2.4

Flashlamp operation Flashlamps provide pulsed light from pulsed electrical discharges generated between two electrodes inside a quartz envelope filled with inert gas (typically xenon). A pulsed power electrical driver serves both to initiate a plasma electrical discharge through the gas and to provide electrical power to the plasma to generate pulsed light output. Flashlamps are commercially available for a wide range of applications, and detailed descriptions of their characteristic and proper operation are available from manufacturers (Perkin-Elmer 2005). This study employed the manufacturer's design relationships listed in Table 1 to select circuit parameters for a series of pulse lengths from about 100­1000 µs, which is within the range of applicability of the design relationships. For pulse lengths shorter than about 100 µs, flashlamp lifetime is less than predicted by the standard design relationships. Flashlamp lifetime is generally given in terms of number of pulses, which depends on the operating parameters of the lamp relative to an explosion limit (Perkin-Elmer 2005). For mercury lamps, on the other hand, the lifetime is specified in hours. For this study, in order to compare practical flashlamp performance to mercury lamp performance, the flashlamp lifetime is specified at a fixed average power. Specifically, the flashlamp pulse operating conditions, including pulse repetition rate, were chosen so that its average input power per unit length is the same as the commercial MPM lamp used in this study, namely 105 W cm-1 . The flashlamp energy per pulse and the pulse repetition frequency are selected to provide this average power for a specified operational life, which was 3000 h for this study. The flashlamp circuit design relationships are listed in Table 1, and the flashlamp parameters are listed in Table 2. The damping parameter for this study is 0.8, the condition for critical damping of the circuit and the optimal condition for lifetime. The flashlamp is model FXC-1607-6 from EG&G (now Perkin-Elmer) a special order lamp designed for high power operation. Lamp and circuit parameters are listed in Table 2. For evaluating UV performance as a function of pulse length variation, the flashlamp life was fixed at 3000 h, which is the lower limit of expected life for MPM lamps, which ranges upwards of 5000 h. For flashlamp lifetimes longer than 3000 h, the pulse parameters will change in such a way that the UV efficiency will be less than that reported here. The electrical

circuit parameters and pulse repetition rates for the different test pulse lengths are summarized in Table 3. The capacitance values correspond to capacitors available for this experiment, the inductance was chosen to vary the pulse length, and the charging voltage chosen to achieve the 3000 h life. In addition, testing included operation of the flashlamp at higher stress conditions that have higher UV efficiency but shorter lifetime. The UV efficiency increases as the pulse length decreases. Thus, to provide an indication of possible high UV efficiency flashlamp operating points, additional tests were made starting with the shortest pulse length listed in Table 3, 71 µs, and increasing the charge voltage, with the pulse energy approaching the explosion limit. Flashlamp set-up The experimental set-up for the flashlamp is shown schematically and pictured in Fig. 2. The flashlamp is driven with a pulse discharge circuit, consisting of a capacitor bank, an air core inductor, and a triggered spark gap switch. In addition, a trigger wire is wound around the length of the discharge tube, as recommended by the manufacturer. To pulse the lamp, the capacitor bank is charged to a high voltage with a Kaiser Systems, Inc power supply (LS-1500). When the capacitor bank is fully charged, a trigger pulse is sent to the high voltage pulse generator for the initiation wire, which pre-ionizes the discharge gas. The trigger pulse also fires the pulse generator on the trigger pin for the spark gap switch. When the spark gap switch closes, the capacitor bank discharges into the lamp. The capacitance, inductance, and charging voltage are selected to produce a critically damped LCR circuit waveform (i.e., the time dependence of the current is of form te-t ). The total inductance consists of the air core inductor and the inherent circuit inductance, which was determined using a short circuit test. In the short circuit test, a copper bar replaced the flashlamp and the air core inductor was removed.

© 2007 NRC Canada

306 Fig. 2. Flashlamp experimental set-up (a) picture, (b) diagram.

J. Environ. Eng. Sci. Vol. 6, 2007

Spark Inductor gap switch

(a)

Flashlamp

Capacitor bank (b)

Triggered spark gap

High voltage pulse generator Trigger controller

L Flashlamp dc -dc Converter

C

High voltage pulse generator

Lcircuit

R circuit

The short circuit was operated using a voltage that produces a peak current comparable to the flashlamp discharge. The oscillation of the current was matched to an under-damped LCR circuit to determine the inherent circuit inductance and resistance. The measurements showed the circuit resistance was less than 40 mOhm and the circuit inductance was 1.2 µH. Surface discharge lamp description and set-up The surface discharge (SD) lamp is a new pulsed lamp that, in comparison to flashlamps, can operate at much higher energy per pulse (for the same pulse length) and has correspondingly higher UV efficiency. The SD lamp is used for both water treatment and paint stripping (Grapperhaus et al. 2005; Schaefer 2001, 2003, 2004; Schaefer et al. 2005; Schaefer and Grapperhaus 2006a, 2006b). In a SD lamp, a high power electrical pulse discharges along the outer surface of a dielectric substrate. Figure 3a shows a SD lamp with a fused silica tube as the substrate, inside a much larger tube (envelope) that contains xenon gas. The electrical discharge produces a plasma along the surface of the substrate, as shown is the open shutter photograph in Fig. 3b. The outer envelope serves only to contain the xenon, and plays no role in the plasma formation and evolution, unlike flashlamps. Because of its relatively large diameter envelope

(76.2 mm in Fig. 3a), the SD lamp lifetime can be much longer than flashlamp lifetimes. The experimental set-up for the SD lamp is shown schematically in Fig. 3c. The SD has a two-component pulse discharge circuit, one to initiate the discharge (Csp and Cpk ) and the other (Cmain ) to drive the discharge. To pulse the lamp, the capacitors are charged to specified voltages. When the capacitors are fully charged, a trigger pulse initiates spark gap switch 1, and, after a predetermined delay, a second trigger pulse initiates spark gap switch 2. When each spark gap switch closes, the capacitor discharges into the lamp. The capacitance and charging voltage are varied to provide different operating conditions. Mercury lamps The mercury lamps used in this study were commercial lamps supplied by Trojan Technologies, Inc. The UV output from the LPM lamp is primarily at 254 nm, and was measured to have a UV efficiency of 34.6%. In water treatment applications, the MPM lamp is placed inside a 76.2 mm diameter envelope, the same size as the SD lamp. The UV output from the MPM lamp from 200 to 300 nm was measured through the envelope and was 325 W, while the electrical power used was 2660 W, giving a UV efficiency of 12.2%. Although the efficiency of LMP lamps is

© 2007 NRC Canada

Schaefer et al. Fig. 3. Surface discharge (SD) lamp (a) photograph, (b) open shutter photograph of the discharge, (c) circuit diagram.

307

(a)

Electrodes

Envelope

Substrate (b)

(c) High voltage pulse generator Triggered spark gap High voltage pulse generator Triggered spark gap Trigger controller

Capacitor charger Csp

Capacitor charger

Cpk

Lamp

Cmain

Energy spectrum (J nm-1)

much higher than MPM lamps, the power per unit lamp length is much smaller. Consequently, LPM lamps tend to be used in low power applications, and MPM lamps used in high power applications. Low pressure mercury and MPM represent the range of UV lamp capabilities.

Fig. 4. Flashlamp light spectra for 3000 h predicted life.

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 Increasing pulse length Pulse length (ms) 796 556 344 164 71

Results and discussion

Flashlamp spectral and UV performance Flashlamp operating conditions that match the input power density (105 W cm-2 ) and lifetime (3000 h) of the MPM lamp are considered first. The optical spectra (200­800 nm) for a single pulse of the flashlamp are exhibited in Fig. 4, for the pulse lengths in Table 3. The total light output for a single pulse decreases as the pulse length decreases, because the energy per pulse decreases to maintain the specified lifetime of 3000 h. For a more pertinent comparison, the lamp output spectrum for a given pulse is normalized by the input energy, giving a spectral efficiency. Figure 5 shows that the spectral UV efficiency increases as the pulse length decreases. For the longest pulse length the spectral UV efficiency decreases from 300 to 200 nm.As the pulse lengths are shortened, the UV efficiency increases, because shorter pulses produce a higher power density

Wavelength (nm)

inside the flashlamp, which heats the plasma to higher temperatures, shifting the maximum in the spectrum from the visible for the longest pulse length to the UV for the shortest pulse length tested. For this study the total UV efficiency is defined as the ratio of the total light energy in 200­300 nm to the input electrical

© 2007 NRC Canada

308 Fig. 5. Normalized flashlamp UV spectra for 3000 h predicted life.

J. Environ. Eng. Sci. Vol. 6, 2007 Fig. 7. Flashlamp light spectra dependence on charging voltage.

Spectral efficiency (J nm-1 J-1 electrical )

0.004

3.0

Pulse length (ms) 71 164 344 556 796

Voltage

Energy spectrum (J nm-1)

2.5 Increasing voltage 2.0 1.5 1.0 0.5 0

0.003

0.002

5.0 kV 4.5 kV 4.0 kV 3.5 kV 3.0 kV 2.5 kV

0.001 Decreasing pulse length 220 240 260 280 300

0 200

200

300

400

500

600

700

800

Wavelength (nm)

Fig. 6. Pulse length dependence of flashlamp UV efficiency.

10 9 8 7 6 5 4 3 2 1 0 0 100 200 300 400 500 600 700 800 900

Wavelength (nm)

Fig. 8. Dependence of flashlamp UV efficiency and lifetime on charging voltage.

10000 1000 100 14 12 10 8 6 4 2 0 5.0

Efficiency (%)

10 1 0.1 0.01 0.001

0.0001 1.5

2.0

2.5

3.0

3.5

4.0

4.5

Voltage (kV)

Pulse width (ms)

energy, and the measured variation is plotted in Fig. 6 as a function of pulse length. The UV efficiency ranges from about 4.8% at 796 µs to about 9.0% at 71 µs, showing the increase in efficiency with shorter pulse length seen in the spectral UV efficiency in Fig. 5. Operating with short pulse length at higher pulse energy allows the flashlamp simultaneously to have high peak power and high UV efficiency, but significantly reduces the lifetime of the lamp. The spectra for a high stress test series are shown in Fig. 7, where the starting point is the 3000 h lifetime, 71 µs pulse length case (2.5 kV charging voltage), which has the highest UV efficiency of the 3000 h lifetime cases (see Fig. 6). The higher stress operating points correspond to increasing charge voltage (i.e., increasing energy per pulse). The spectral shape for this test series is similar for all operating points, with increasing output for increased pulse energy, but the shorter wavelengths increase more quickly with pulse energy. The shift in output to shorter wavelengths results in higher UV efficiency and is caused by the higher plasma temperature during the discharge, which is a result of the higher discharge power density.

Unfortunately, the increase in UV efficiency at higher stress operating conditions results in a significantly shorter lifetime. Figure 8 shows the corresponding UV efficiency and calculated lifetime for the high stress operating condition as a function of the charge voltage. The lifetime falls off quickly as the voltage increases, while the UV efficiency increases from about 9.0% for a 3000 h life up to over 12% efficiency near the explosion limit. Across all the operating conditions examined in this study, the flashlamp UV efficiency varied by more than a factor of 2.5, as shown in Table 3 and Fig. 8, with vastly shorter predicted lifetimes for the high UV efficiency conditions. This shows the importance of specifying flashlamp operating conditions and measuring the UV output when using a flashlamp UV source for water treatment. Operating a flashlamp under high stress conditions could result in a short-term performance comparable to the MPM lamp, but would be impractical for a commercial system because of short lamp life. Surface discharge UV performance A more complete characterization of surface discharge (SD) lamp performance is found elsewhere (Schaefer 2001, 2003,

© 2007 NRC Canada

UV efficiency (%)

Life (Hours)

Schaefer et al. Table 4. Irradiance estimates at the surface of a 3 in. diameter envelope (for retrofit). Lamp SD1 SD2 Flashlamp MPM LPM C (µF) 4 118 97 NA NA V (kV) 18 5.07 1.9 NA NA Ave. center irradiance UV (W cm-2 ) 19 000 10 400 572 0.686 0.014 Ave. center irradiance light (W cm-2 ) 78 400 34 400 2 330 2.28 0.016 tpulse (µs) 12.4 37.2 71.0 Continuous Continuous E200 /Eelelc (%) 17.2 12.2 9.0 12.2 34.6

309

Fig. 9. Comparison of flashlamp and SD lamp (4 µF, 18 kV; and 118 µF, 5 kV) UV light spectra.

Fig. 10. Comparison of SD, LPM, and MPM lamp UV light spectra.

J electrical -1 )

3.5 × 10-3 3.0 × 10-3 2.5 × 10

-3

Normalized spectral power (W nm-1 cm-1)

4.0 × 10-3

0.7 0.6 0.5 0.4 0.3

MPM lamp

-1

Spectral efficiency (J nm

2.0 × 10-3 1.5 × 10-3 1.0 × 10-3 0.5 × 10-3 0.0 200

SD: 4 uF 18 kV

SD: 118 uF 5 kV

SD lamp

0.2 0.1

Flashlamp 220 240 260 280 300

LPM lamp

0 200 220 240 260 280 300

Wavelength (nm)

2004; Schaefer et al. 2005). Surface discharge lamp data in this study exhibits a range of pulse parameters that provide a variation of pulse length and irradiance, at conditions where the lifetime is expected to be much greater than 3000 h, although life testing is not complete for this new lamp technology. Two operating points for the SD lamps listed in Table 4 represent the shortest and longest pulse lengths and may represent a range appropriate for UV water treatment have UV efficiencies of 12.2% and 17.2%. The two corresponding SD spectra and the best flashlamp UV efficiency (9.0%) operating point for 3000 h life are superimposed in Fig. 9. The spectral shapes are similar, indicative of discharges in xenon, with the SD spectra being more filled in, with higher broadband levels and higher UV efficiency. UV lamp comparative performance The comparative spectral performance of the LMP, MPM, and SD (17.2% example) lamps are shown in Fig. 10. The LPM lamp spectrum is centered approximately on 254 nm. The spectra from the SD and MPM lamps have different shapes, the SD with more output below 250 nm. The SD and MPM lamps have comparable levels between 250 and 300 nm, with the MPM having strong peaks while the SD spectrum is comparatively flat. Table 4 provides a quantitative comparison of the UV efficiency, pulse length, and average irradiance during the pulse. The SD UV efficiencies range from the same as, to about 40% higher than, the MPM lamp, with the best practical flashlamp

Wavelength (nm)

point having approximately three-quarters the efficiency of the MPM lamp. Light irradiances are included in Table 4 to provide an indication of the inherent difference between continuous and pulsed light, which may play a role in possible differences in water treatment rates with pulsed lamps. Except for LPM, the irradiances are for light leaving a 76.2 mm diameter envelope (or water jacket) for each lamp, a common configuration for water treatment. Note that the highest SD UV irradiance, at 19 000 Wcm-2 , is more than a factor of 30 higher than the flashlamp, and the flashlamp UV irradiance is about three orders of magnitude higher than for the MPM, with an irradiance less than 1 W cm-2 . Table 4 also lists the total irradiance, which includes light from 200 to 700 nm.

Conclusions

The UV efficiency, lifetime, and output intensity of pulsed xenon lamps varies strongly with the lamp and electrical pulse operating conditions. For the flashlamp tests reported here, the UV efficiency varied by more than a factor of 2.5, with the highest efficiency corresponding to the shortest lifetime tested. For a 3000 h predicted life, based on the manufacturer relationships, the highest flashlamp UV efficiency was 9.0% compared to 12.2% for MPM lamps. The UV efficiencies of a new pulsed SD lamp are12.2% and 17.2% for the operating points reported

© 2007 NRC Canada

310

J. Environ. Eng. Sci. Vol. 6, 2007 McDonald, K.F., Curry, R.D., Clevenger, T.E., Unklesbay, K., Eisenstart,A., Golden, J., and Morgan, R.D. 2000.A comparison of pulsed and continuous ultraviolet light sources for the decontamination of surfaces. IEEE Trans. Plasma Sci. 28(5): 1581­1587. Mofidi, A.A., Baribeau, H., Rochelle, P.A., De Leon, R., Coffey, B.M., and Green, J.F. 2001. Disinfection of Cryptosporidium parvum with polychromatic UV light. J. Am. Water Works Assoc. 93(6): 95­109. Perkin-Elmer. 2005. Citing online sources: high performance flash and arc lamps [online]. Available from http://optoelectronics.perkinelmer.com/content/RelatedLinks/ flashcatalog.pdf [cited 26 August 2005]. Phillips, R. 1983. Sources and applications of ultraviolet radiation, chapter 8: the medium pressure mercury lamp. Academic Press, New York, N.Y. Rowan, N.J., MacGregor, S.J., Anderson, J.G., Fouracre, R.A., McIlvaney, L., and Farish, O. 1999. Pulsed-light inactivation of foodrelated microorganisms. Appl. Environ. Microbiol. 65(3): 1312­ 1315. Schaefer, R.B. 2001. Innovative ultraviolet light source for disinfection of drinking water. Final Report for EPA, Phase I SBIR, Contract No. 68-D-00-257 EPA 2001; EPA SBIR Phase II, Innovative ultraviolet light source for disinfection of drinking water, Contract no. 68-D01-057. Schaefer, R.B. 2003. Ultraviolet (UV) water remediation with surface discharge UV lamps. SBIR Phase I Contract No. DMI-0109181, Final Report to NSF, December 2001; NSF SBIR Phase II, Ultraviolet (UV) water remediation with surface discharge lamp, Contract No. DMI-0237472. Schaefer, R.B. 2004. Surface discharge high intensity ultraviolet lamp systems. Final report to NIST Advanced Technology Program, Contract No. 70NANB1H3053. Schaefer, R., and Grapperhaus, M. 2006a. Innovative lead paint removal using high intensity light pulses. HUD Healthy Homes and Technical Study Project Final Report, Contract No. MALTS009002. Schaefer, R.B., and Grapperhaus, M. 2006b. Paint removal from architectural surfaces with an innovative pulsed light source. EPA SBIR Phase II Final Report, Contract No. 68-D-03-046. Schaefer, R.B., Grapperhaus, M., and Linden, K. 2005. New surface discharge pulsed UV light source. CDROM Proceedings of the 3rd International Congress on Ultraviolet Technologies, Whistler, BC, May 2005, International Ultraviolet Association, PO Box 1110, Ayr, ON, Canada N0B 1E0. Sharma, R.R., and Demirci, A. 2003. Inactivation of Escherichia coli O157:H7 on inoculated alfalfa seeds with pulsed ultraviolet light and response surface modeling. J. Food Sci. 68(4): 1448­1453. Weber, E., and Scheible, K. 2003. Pulsed-UV unit may inactivate biological agents. J. Am. Water Works Assoc. 95(6): 34­46. Wekhof, A. 2003. Sterilization of packaged pharmaceutical solutions, packaging and surgical tools with pulsed UV light. CDROM Proceedings of the 2nd International Congress on Ultraviolet Technologies, Vienna, Austria, July 2003, International Ultraviolet Association, PO Box 1110, Ayr, ON, Canada N0B 1E0.

here, with an intensity four orders of magnitude larger than the MPM lamp. Surface discharge lamps are pulsed, instant-on, mercury free lamps with high UV efficiency. The comparative lamp parameters show the importance of providing quantitative measurements of pulsed lamps for UV applications.

Acknowledgements

Development of the SD lamp was supported by the Advanced Technology Program (ATP) under a NIST Cooperative Agreement, the flashlamp characterization under NSF Grant No. DMI-0237472. Technical assistance by John Gallagher and Erin O'Brien of Phoenix Science & Technology is gratefully acknowledged.

References

Bohrerova, Z., Mamane-Gravetz, H., Linden, K., Grapperhaus, M., and Schaefer, R.B. 2005. Comparative inactivation of Bacillus subitilis spores and MS-2 bacteriophage using standard mercury lamps and a new, pulsed, surface discharge lamp. CDROM Proceedings of the 3rd International Congress on Ultraviolet Technologies, Whistler, BC, May 2005, International Ultraviolet Association, PO Box 1110, Ayr, ON, Canada N0B 1E0. EPA. 2006a. National primary drinking water regulations: stage 2 disinfectants and disinfection byproducts rule. Fed. Regist. 71(2): 388. EPA. 2006b. National primary drinking water regulations: long term 2 enhanced surface water treatment rule. Fed. Regist. 71(3): 654. Grapperhaus, M., Schaefer, R.B., and Linden, K. 2005. Modeling of a new UV test cell for evaluation of lamp intensity effects on water treatment. CDROM Proceedings of the 3rd International Congress on Ultraviolet Technologies, Whistler, BC, May 2005, International Ultraviolet Association, PO Box 1110, Ayr, ON, Canada N0B 1E0. Hillegas, S., and Demirci, A. 2003. Inactivation of Clostridium sporogens in clover honey by pulsed UV-light treatment. ASAE Meeting, Las Vegas, Nev., 27­30 July 2003, Paper No. 037017. Krishnamurthy, K., Demirci, A., and Irudayaraj, J. 2004. Inactivation of Staphylococcus aureus by pulsed UV-light sterilization. J. Food Prot. 67(5): 1027­1030. Liang, S., Min, J.H., Davis, M.K., Green, J.F., and Remer, D.S. 2003. Use of pulsed-UV processes to destroy NDMA. J. Am. Water Works Assoc. 95(9): 121­131. Linden, K.G., and Mofidi, A.A. 2003. Disinfection efficiency and dose measurement of polychromatic UV light: high-quality water -- monitoring and treatment. AwwaRF Report 90931F, IWA Publishing, London, UK. MacGregor, S.J., Rowan, N.J., McIlvaney, L., Anderson, J.G., Fouracre, R.A., and Farish, O. 1998. Light inactivation of foodrelated pathogenic bacteria using a pulsed power source. Lett. Appl. Microbiol. 27: 67­70. Marshak, I.S. 1984. Pulsed light sources. Chapter 5. Radiation characteristics of flashlamps. Consultants Bureau, New York, N.Y.

© 2007 NRC Canada

Information

i:\

8 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

1191777