Performance Stability of the Harvard Ambient Particle Concentrator

Abstract

The Harvard Ambient Particle Concentrator (HAPC) has been used routinely for exposure testing for a period of approximately three years. The stability of concentrator performance has been investigated as a function of local meteorological conditions, ambient particle concentrations, composition, and size distribution. Concentrator performance is characterized by the concentration enrichment factor (CEF), a ratio of concentrated particle mass (or sulfate) concentration to the ambient concentration. Over three years of normal operation, the mass and sulfate CEFs averaged 27.9 and 28.6, respectively. The majority of variability in the CEF was found to be related to the mass median aerodynamic diameter (MMAD) of ambient particles and to the HAPC's total operating pressure drop.

The 50% cutpoint of the HAPC is 0.15 μ m. Between 0.15 and 2.5 μ m, the HAPC concentrates ambient particle mass by approximately a factor of 28. It is logical that changes in particle CEF should be influenced by the amount of ambient particle mass within the concentrator's effective size range.

The concentrator operates optimally at a pressure drop of 2.5 in. of water per stage. Total pressure drop ranges from 7.5–25 in. of water, depending on the alignment of the slits of the virtual impactors and other parameters. Increases in the pressure drop result in decrease of the CEF, mainly due to excessive particle loss by impaction on the edges of the collection slit. In contrast, the minor operating pressure drop was found to be consistently lower on days where neither the ambient particle mass, temperature, or relative humidity were high. This subset of days was found to have a higher CEF than days where any of the three variables were high. The HAPC minor flow pressure drops are thus monitored carefully, since they are the most controllable indicator of concentration enrichment efficiency.

Acknowledgments

The authors would like to acknowledge Jessica Sekula, Jim Sullivan, and Denise Belliveau for invaluable assistance in operation and maintenance of the particle concentrator as well as collection and analysis of samples. In addition, we would like to thank Yiming Ding for insightful discussion of the model results. This work was supported by the Health Effects Institute, under grant #5 PO1 ES08129.

Notes

*Collection of these samples began at a later date, only limited data available.

*Included as inverse (1/Minor 3 Pressure) to improve linearity of relationship.

*Included as inverse (1/Minor 3 Pressure) to improve linearity of relationship.