Sensitivity analysis of the USEPA WINSPM2.5 separator

Factors affecting the performance of the USEPA WINS PM2.5 separator have be en systematically evaluated. In conjunction with the separator's laboratory calibrated penetration curve, analysis of the governing equation that desc ribes conventional impactor performance was used to predict changes in cutp oint as a function of impactor dimensions, flow rate, uncertainties in ambi ent temperature and pressure measurement, and the temperature and pressure of the sampled air volume. By integrating the resulting performance curves with three idealized ambient aerosol size distributions, the effect of thes e parameters on measured PM2.5 concentration was predicted. Results showed that allowable variations in impactor jet width, flow rate, diffusion oil v olumes, and ambient temperature and pressure measurement result in relative ly minimal PM2.5 mass concentration measurement biases. Loading of the WINS well with previously collected particles slightly reduces the separator's cutpoint and thus slightly reduces expected PM2.5 mass concentrations. Vari ations in ambient pressure produce negligible changes in the performance of the WINS. While not causing a true measurement bias as defined by the regu lations, low ambient temperatures naturally affect the airstream's properti es and inherently shifts the WINS' cutpoint to slightly lower values. Labor atory-induced crystallization of the DOW 704 diffusion oil produced no appr eciable changes in either the position or shape of the WINS separation curv e.