This second, and concluding, part of this study evaluated changes in sampling efficiency of respirable size-selective samplers due to air pulsations generated by the selected personal sampling pumps characterized in Part I (Lee E, Lee L, Möhlmann C et al. Evaluation of pump pulsation in respirable size-selective sampling: Part I. Pulsation measurements. Ann Occup Hyg 2013). Nine particle sizes of monodisperse ammonium fluorescein (from 1 to 9 µm mass median aerodynamic diameter) were generated individually by a vibrating orifice aerosol generator from dilute solutions of fluorescein in aqueous ammonia and then injected into an environmental chamber. To collect these particles, 10-mm nylon cyclones, also known as Dorr-Oliver (DO) cyclones, were used with five medium volumetric flow rate pumps. Those were the Apex IS, HFS513, GilAir5, Elite5, and Basic5 pumps, which were found in Part I to generate pulsations of 5% (the lowest), 25%, 30%, 56%, and 70% (the highest), respectively. GK2.69 cyclones were used with the Legacy [pump pulsation (PP) = 15%] and Elite12 (PP = 41%) pumps for collection at high flows. The DO cyclone was also used to evaluate changes in sampling efficiency due to pulse shape. The HFS513 pump, which generates a more complex pulse shape, was compared to a single sine wave fluctuation generated by a piston. The luminescent intensity of the fluorescein extracted from each sample was measured with a luminescence spectrometer. Sampling efficiencies were obtained by dividing the intensity of the fluorescein extracted from the filter placed in a cyclone with the intensity obtained from the filter used with a sharp-edged reference sampler. Then, sampling efficiency curves were generated using a sigmoid function with three parameters and each sampling efficiency curve was compared to that of the reference cyclone by constructing bias maps. In general, no change in sampling efficiency (bias under +/-10%) was observed until pulsations exceeded 25% for the DO cyclone. However, for three models of pumps producing 30%, 56%, and 70% pulsations, substantial changes were confirmed. The GK2.69 cyclone showed a similar pattern to that of the DO cyclone, i.e. no change in sampling efficiency for the Legacy producing 15% pulsation and a substantial change for the Elite12 producing 41% pulsation. Pulse shape did not cause any change in sampling efficiency when compared to the single sine wave. The findings suggest that 25% pulsation at the inlet of the cyclone as measured by this test can be acceptable for the respirable particle collection. If this test is used in place of that currently in European standards (EN 1232-1997 and EN 12919-1999) or is used in any International Organization for Standardization standard, then a 25% pulsation criterion could be adopted. This work suggests that a 10% criterion as currently specified in the European standards for testing may be overly restrictive and not able to be met by many pumps on the market. Further work is recommended to determine which criterion would be applicable to this test if it is to be retained in its current form.
Eun Gyung Lee, National Institute for Occupational Safety and Health, Health Effects Laboratory Division, Exposure Assessment Branch, 1095 Willowdale Road, Morgantown, WV 26505