Testing Procedure
The testing was performed in two stages. The first stage operated the test chamber with the lamp off. (See table 1) This developed the control data or the base line bacterial levels for the comparison. The second stage operated the test chamber with the lamp on. (See table 2)
Two airflow rates were used to evaluate the lamp effectiveness based on exposure time. Airflow velocities through the ducts of a typical residential heating and cooling system range from 300 to 500 feet per min (fpm). For this study a base air velocity of 500 fpm was used. To decrease the exposure time, a second test was conducted with the airflow in the duct doubled to 1000 fpm. Since the effectiveness of UV lamps is based on the UV light output and exposure time, doubling the airflow reduces the effectiveness of the lamp.
The bacterium was cultured and the cells harvested to provide a suspension of known cell density. This was further diluted to provide gallon quantities of a test suspension containing an estimated 95,000 CFU/ml. This suspension was pumped through the spray nozzle mounted in the center of the duct inlet.
Five air samples were taken for each of the test velocities at short intervals (typically ½ to 2 minutes). This produced a large sample volume of air and reduced the levels of back ground bacteria and molds counts. The plate counts (colony forming units or CFU) for each of the five tests were totaled and divided by the total test volume of air. This produced the comparison value of (269 CFU/FT3 of air) for the 500 FPM airflow and (107.5 CFU/FT3 of air) for the 1000 FPM airflow. Due to apparent efficiency losses in the sampling method at the 1000 FPM velocity, the bacterium count yielded a 60% drop instead of the anticipated 50% reduction due to the velocity change.
Four air samples were taken at 1, 2.5, 3, 5, 6 & 10 minute intervals for each of the test velocities with the lamp on. The longer sample times with the lamp on were needed to obtain plate counts which would provide reliable estimates of the efficiency of disinfection, but with this, more background organisms were found. The plate counts were (18.00 CFU/FT3 of air for the
Page 5
