Comparison of the efficacies of troubleshooting methodologies for ventilation systems - a field study.
Seattle, WA: University of Washington, 1996 Jun; :1-78
For a ventilation system to control deleterious worker exposures most efficiently, the system ductwork must distribute the airflow in the correct proportions to all the branches serving the hoods. Even if good distribution is established when the system is fust installed, the distribution may become increasingly unsatisfactory due to particle settling, alteration from the original design, wear, deformation of the ductwork, and other causes. This means that some hoods may receive excess airflow, while others receive a flow that is inadequate to properly protect workers using the hood. Visual inspection often fails to discover changes to ducts because of their opacity and poor accessibility. Thus, "troubleshooting" must rely on measurements of pressures and flows in the ducts to detect and locate alterations that can affect airflow distribution. This field study compares the efficacy of SLX methods of troubleshooting ventilation system branches. Static pressures and airflows were measured on two different systems over a three month period. Repeat measurements were made on each system. The system was then inspected for obstructions or other alterations, cleaned out, and remeasured. Sensitivity and specificity were then calculated for a full range of decision variable thresholds. Methods were compared using receiver operating characteristic curves. The log transformed static pressure ratio and power loss coefficient (X-value) methods performed much better than the use of hood static pressures alone or the method described in Industrial Ventilation Manual (ACGIH, 1995). At a given sensitivity, both methods produce low numbers of cosdy searches for non-existent alterations. The log transformed static pressure ratio method does not require a time consuming velocity traverse, and thus may be the method of choice. The common hood static pressure method and the idealized IVM method both performed poorly. The results of this study provide guidance to industrial hygienists and ventilation professionals as to what troubleshooting methodology is most effective. Equipped with troubleshooting methods that produce few false positives, practitioners may be encouraged to monitor systems more closely and intervene before hood performance has deteriorated to unsatisfactory levels.
Ventilation; Industrial-ventilation; Ventilation-systems; Exhaust-ventilation; Air-flow; Air-contamination; Airborne-particles; Equipment-design; Models
Douglas H. Moody, University of Washington, School of Public Health and Community Medicine, Department of Environmental Health, Room F226D, Box 357234, Seattle, WA 98195-7234
Empirical determination of the error in the ACGIH method of predicting airflow distribution in two ventilation systems
University of Washington, Department of Environmental Health, Seattle, WA