I. Introduction II. Model Development III. Summary of Model Input Parameters IV. Selected Model Results V. Single-Source and Regional Results VI. Concl usions and Recommenaa tions Acknowledgments Appendix A: Detailed Description of Model Input Parameters Appendix B: Nine-house Study Appendix C: Selected Graphic Output of Preliminary Model Predictions
PREFACEThis report is the first in a series intended to assess residential air-pollution concentrations in the U.S. housing stock. A "macromodel" has been developed to simulate/predict selected residential air-pollution concentration distributions across specified house populations.Combustion pollutants are initially addressed because (1) they are a major class of indoor pollutants with documented adverse health effects at elevated concentrations,the underlying mathematical relationships are somewhat understood, and (3) sufficient input data exist to allow the macromodel to obtain results (i~e.,predictions). This report addresses the macromodel development and preliminary predictions for carbon monoxide, nitrogen dioxide, and respirable suspended particles. Subsequent reports address (1) sensitivity analysis of the macromodel, (2) micro/macro comparisons, and (3) preliminary predictions for organic compounds and sulphur dioxide.i ABSTRACT A simulation model (also call a "macromodel") has been developed to predict residential air pollutant concentration distributions for specified populations. The model inputs include the market penetration of pollution sources, pollution source characteristics (e.g., emission rates, source usage rates), building characteristics (e.g., house volume, air exchange rates), and meteorological parameters (e.g., outside temperature). Four geographically distinct regions of the U.S. have been modeled using Monte Carlo and deterministic simulation techniques. Single-source simulations were also conducted. The highest predicted CO and N02 residential concentrations were associated with the winter-time use of unvented gas and kerosene space heaters. The highest predicted respirable suspended particulate concentrations were associated with indoor cigarette smoking and the winter-time use of non-airtight wood stoves, radiant kerosene heaters, convective unvented gas space heaters, and oil forced-air furnaces. Future field studies in this area should (1) fill information gaps identified in this report, and (2) collect information on the macromodel input parameters to properly interpret the results. It is almost more important to measure the parameters that affect indoor concentration than it is to measure the concentrations themselves.
ii
I -2Approaches to assess total exposures to air pollutants fall into three categories: (1) personal monitoring (i.e., dosimetry); (2) monitoring of concentrations in the various human environments; and (3) modeling of exposures based on a wide array of information, including pollutant sources, building characteristics, pollutant behavior, and human activities. These approaches can be complementary. Moreover, the ...