The National Institute for Occupational Safety and Health (NIOSH), recognizing the difficulties inherent in using old military data to define modern industrial respirator fit test panels, recently completed a study to develop an anthropometric database of the measurements of heads and faces of civilian respirator users. Based on the data collected, NIOSH researchers developed two new panels for fit testing half-facepiece and full-facepiece respirators. One of the new panels (NIOSH bivariate panel) uses face length and face width. The other panel is based on principal component analysis (PCA) to identify the linear combination of facial dimensions that best explains facial variations. The objective of this study was to investigate the correlation between respirator fit and the new NIOSH respirator fit test panel cells for various respirator sizes. This study was carried out on 30 subjects that were selected in part using the new NIOSH bivariate panel. Fit tests were conducted on the test subjects using a PORTACOUNT device and three exercises. Each subject was tested with three replications of four models of P-100 half-facepiece respirators in three sizes. This study found that respirator size significantly influenced fit within a given panel cell. Face size categories also matched the respirator sizing reasonably well, in that the small, medium, and large face size categories achieved the highest geometric mean fit factors in the small, medium, and large respirator sizes, respectively. The same pattern holds for fit test passing rate. Therefore, a correlation was found between respirator fit and the new NIOSH bivariate fit test panel cells for various respirator sizes. Face sizes classified by the PCA panel also followed a similar pattern with respirator fit although not quite as consistently. For the LANL panel, however, both small and medium faces achieved best fit in small size respirators, and large faces achieved best fit in medium respirators. These findings support the selection of the facial dimensions for developing the new NIOSH bivariate respirator fit test panel.

No abstract

Three-dimensional computational fluid dynamics (CFD) simulations were used to predict the flow field and resulting worker exposures when toxic airborne contaminants were released into the wake region of a mannequin that had its back to the airflow while holding the source of airborne contaminants. The effects of ventilation velocity, free-stream turbulence, and various thermal conditions on fluid flow and exposure levels were evaluated. The results showed good agreement between predicted and experimental concentrations at the mouth at a broad range of airflow velocities when the mannequin was both heated and unheated. When the mannequin was unheated, the exposure level decreased as the ventilation velocity increased. The expectation that buoyancy provided by the heat from the mannequin would be most important at very low velocities and decreasingly important at high velocities was proved true for both the predicted and observed exposures. The result was that when the mannequin was heated to normal human body temperatures, exposure levels had an inverted V relationship with velocity. These findings are important, since they call into question the common practice of modeling human exposures with mannequins at ambient temperatures. In addition, free-stream turbulence could be used to reduce worker exposure to airborne pollutants as suggested by the simulations. CFD enabled a detailed investigation of the effect of particular factors for exposure predictions in a cost-effective way.

No abstract

Objectives National Institute for Occupational Safety and Health–approved P100 filtering facepiece respirators (FFRs) have a higher filter efficiency compared to the N95 filters. However, the former typically produce higher flow resistance (Rf). Consequently, when faceseal leakage is present, the proportion of leakage airflow for P100 FFRs may exceed that of N95s, resulting in a higher total inward leakage (TIL) of the P100. Methods In this manikin-based study, the performance of two pairs of N95 and P100 FFRs (N95-A versus P100-A; N95-B versus P100-B) were compared under five sealing conditions (fully sealed and partially sealed with one, two, or three leaks of 0.8-mm, and one 2-mm leak). Sodium chloride particles (CMD ~45 nm) were used as the challenge aerosol. Respirators were tested under three constant flows (15, 50, and 85 L/min) and three cyclic flows (mean inspiratory flow = 15, 50, and 85 L/min). Both filter penetration (Pfilter) and TIL were determined. The Rf under constant flows was recorded. Based on Pfilter, TIL, and Rf, the quality factor (qf) was calculated to compare the overall performance of N95 and P100 FFRs. Results For a fully sealed condition, the Pfilter was much lower for the P100 FFRs than for the N95 FFRs. When small leaks were inserted (0.8-mm and 2 × 0.8-mm), the TIL was higher for the P100 FFRs than for the N95 FFRs under the lowest tested flow (15 L/min), while for greater leaks (3 × 0.8-mm and 2-mm), the TIL of the P100 FFRs was always higher regardless of the flow. The Rf of P100 FFRs was measured twice as high as the N95. The qf values were also found higher for the N95 FFRs than for the P100 FFRs regardless of leak size and breathing flow. Conclusions With the presence of artificial leakage, a P100 FFR with high-flow-resistance may not be as protective as a low-flow-resistance N95 FFR. This finding suggests that future efforts should be directed to reducing the breathing resistance when designing P100 FFRs.