Date of Award

12-2009

Degree Name

MS in Civil and Environmental Engineering

Department

Civil and Environmental Engineering

Advisor

Tracy Thatcher

Abstract

Black carbon, a proxy for woodsmoke was measured indoors and outdoors for an occupied residence in Cambria, CA during the winter months of 2009. The purpose was to investigate the infiltration parameters: air exchange rate, deposition rate, and penetration factor. The second part of this study investigated the light absorption properties of organic carbon from residential wood combustion, the dominant fraction of woodsmoke.

To assess woodsmoke variation, a study conducted parallel to the one presented in this thesis (Ward, 2009), a grid array of personal emission monitors (PEMS) and aethalometers were placed in a small area, approximately one square kilometer, within a community in Cambria, California between the months of November 2008 and March 2009. In this study, PEMS were used to collect particles on filters, which were analyzed for tracers for woodsmoke, including levoglucosan, elemental carbon, and organic carbon. Aethalometers measured black carbon, an indicator of carbon combustion. Additional PEMS and aethalometers were placed inside one residential home to better understand infiltration of woodsmoke.

To model the infiltration of woodsmoke, the Lawrence Berkeley National Laboratory Air Infiltration Model was used. The home of interest was chosen such that indoor sources of particulate matter (PM) were minimal. This insures that all PM measured indoors was from outdoor sources, namely household chimneys. While indoor sources such as indoor fires and resuspension of particles were of concern, homes were chosen to minimize these sources.

To investigate the infiltration parameters, four different solution techniques were used. Two of the solution techniques used SOLVER, a Microsoft Excel program, to minimize the sum of squared differences between calculated indoor concentrations and measured indoor concentrations, with all three parameters (air exchange rate, penetration, and deposition) as independent variables. The other two solution techniques used the Air Exchange Rate (AER) model from Lawrence Berkeley National Laboratory (LBNL) (Sherman & Grimsrud, 1980) and then used SOLVER to calculate deposition rate and penetration factor.

Solution techniques 1 and 3, which used SOLVER to find all three parameters, had average penetration factors of 0.94 and 0.97 respectively, while solution techniques 2 and 4, which used the LBNL AER model had average penetration factors of 0.85 and 0.78 respectively. The deposition rates for solution techniques 1,2,3, and 4 were 0.10, 0.07, 0.08, and 0.04 hr-1 respectively. The air exchange rates varied throughout the study and ranged from 0.1 to 0.7 hr-1. The average indoor/outdoor ratio was also found to be 0.75.

The aerosols derived from the study samples were found to have light absorption properties that were heavily spectrally dependent, which is consistent with expectations for wood combustion aerosols. Conversely, traffic derived aerosols are not found to be heavily spectrally dependent and follow the power law relationship of λ-1 whereas our samples followed λ-1.7 across all wavelengths and λ-2.25 for wavelengths less than 600 nm. The reason for the difference in spectral dependence is the presence of light absorbing organic carbon in wood smoke that is not found in diesel aerosols. The optical absorbances were also calculated for our samples and average values were found to be 3 and 1 m2/g for 370 and 450 nm wavelengths respectively.