Lamp and in-duct device modeling for UVGI system performance prediction

Ultraviolet germicidal irradiation (UVGI) is one of the most promising technologies for control of airborne indoor bio-contaminants, such as viruses and bacteria, in health care, security, and general air quality applications. UVGI is the use of radiation, mainly from the UVC band, to inactivate microorganisms. For a given level of control, UVGI can significantly reduce requirements for ventilation rate and filtration efficiency, which makes it an attractive alternative or complement to those established methods of ventilation and filtration. However, there is a lack of information and well validated models in the literature for UV lamp performance and UVGI device design in in-duct application. The present study attempts to develop a better basis for selecting, designing and commissioning in-duct UVGI systems.This research investigated the performance of UVC lamps and in-duct UVGI devices under a range of operating conditions encountered in HVAC system applications. Three low pressure mercury, UVC lamp types were selected for experiments. Lamp "Type 1" is a Philips -- TUV 25W standard output hot cathode lamp. Lamp "Type 2" is a Philips -- TUV PLL 60W HO/4P high output hot cathode lamp. Lamp "Type 3" is a Light Sources 782L10 standard output cold cathode lamp. The outputs of the lamps were monitored in the lamp depreciation and life tests. Twenty-one combinations of different current levels, cycling rates, and lamp types were measured over 5500 hours in a burn-in room. The impact of the cycling rate on output was not clearly shown in these lamps. The impact of current was shown in "Type 1" and "Type 2" lamps. Operating standard output lamps (Type 1) at above rated current slightly increases the depreciation rate on lamp output. Operating high output lamps (Type 2) at reduced current decreases the depreciation rate on lamp output.With the results from the ambient respond tests, the impacts of ambient conditions (air velocity and temperature) and flow orientation on UV lamp output were significant. The experimental results showed UVC lamp output variation was very significant (>70%) over the range of conditions typically encountered in in-duct UVGI applications due to convective "wind chill" effects. In addition, analytical heat transfer models were used to develop calibrated heat transfer models which require a relatively small amount of experimental data to predict the lamp output in new conditions with good agreement (<8.5%). It was more accurate and effective than the purely empirical model for an interpolation and a near-range extrapolation. These calibrated heat transfer models could be used as design tools for considering the impacts of operating conditions and selecting the appropriate lamp type and orientation.Previous investigations of UVGI irradiance fields have used inverse square law and view factor methods. Both of these approaches have significant limitations, particularly when geometry is complex. This research applied the Monte Carlo (ray tracing) methods to fluence distribution in a UVGI device considering the reflectivity of surfaces and lamp configuration. After including the anisotropic reflection properties of the duct surfaces and adjusting the thermal impacts on the lamp output, the simple ray-tracing computer model provided a reasonable (+/-17%) prediction of planar irradiance measurements that were recorded in the test section of the UVGI device. Finally, this study evaluated some possible procedures for estimating the spherical irradiance within a UVGI device base on the planar irradiances on a cube.With this study, the output variations of the UVC lamps that commonly used in the in-duct UVGI device were thoroughly investigated for the factors of age and thermal conditions. The impact of these factors in a system were demonstrated in a case study at the end, which found that a potential magnitude of error (20% under-estimation bio-contaminant concentration in a space) may result from failure to account for ambient effects on the output reduction of a typical UVGI device.