Investigation of Spatial Variation of Air Toxics in Three Houston-Area Neighborhoods Using Passive Sampling Thomas H. Stock, Maria T. Morandi, Masoud Afshar, Valerie Murray, Silvia Maberti University of Texas School of Public Health, Houston, TX Kuenja C. Chung U.S. EPA, Region 6, Dallas, TX

Advantages of Passive Samplers for Ambient Air Monitoring Small, light weight and requires no power Small, light weight and requires no power Easy to handle in the field Easy to handle in the field Relatively unobtrusive Relatively unobtrusive Multiple simultaneous samples Multiple simultaneous samples Locations not accessible with conventional methods Locations not accessible with conventional methods Fewer security concerns Fewer security concerns

Field Evaluation Studies of 3M OVM for Ambient Air Monitoring Phase 1. Feasibility and determination of operational parameters, including limited method comparison Phase 1. Feasibility and determination of operational parameters, including limited method comparison Phase 2. Year-long comparison of central monitoring site vs. nearby centroid Phase 2. Year-long comparison of central monitoring site vs. nearby centroid Phase 3. Intensive spatial variation study Phase 3. Intensive spatial variation study

Phase 3 Study Design Simultaneous monitoring in three areas, defined as a 2-mile radius around each TCEQ monitoring site: Aldine, Clinton Dr. and Deer Park. Simultaneous monitoring in three areas, defined as a 2-mile radius around each TCEQ monitoring site: Aldine, Clinton Dr. and Deer Park. In each area, simultaneous monitoring at TCEQ monitoring site, census tract centroid, and 10 homes of volunteers. In each area, simultaneous monitoring at TCEQ monitoring site, census tract centroid, and 10 homes of volunteers. 72-hour sampling, performed 6 times during May-September. 72-hour sampling, performed 6 times during May-September.

Houston Area Monitoring Sites

Study Areas

Aldine Monitoring Station

Clinton Drive Monitoring Station

Deer Park Monitoring Station

OVM Sample Shelter

OVM Samples and Field Blank

Target Compounds and 72-hr MDLs VOC Lot 1 MDL (ug/m3) Lot 2 MDL (ug/m3) n-Pentane Methylene chloride MTBE n-Hexane Methylcyclopentane Methyl ethyl ketone Chloroform ,3-Dimethylpentane Carbon tetrachloride Benzene Trichloroethylene Toluene Tetrachloroethylene Ethyl benzene

Target Compounds and 72-hr MDLs VOC Lot 1 MDL (ug/m3) Lot 2 MDL (ug/m3) n-Nonane m/p-Xylene o-Xylene Styrene α-Pinene n-Decane ,3,5-Trimethylbenzene ethyl-2-methyl benzene β -Pinene ,2,4-Trimethylbenzene d-Limonene ,2,3-Trimethylbenzene p-Dichlorobenzene Naphthalene

Results of Phase 3 Sampling In both the Aldine and Clinton study areas, concentrations of many VOCs outside some homes frequently exceeded measurements at the central site. In both the Aldine and Clinton study areas, concentrations of many VOCs outside some homes frequently exceeded measurements at the central site. In the Deer Park area, house measurements were in closer agreement with the central site measurements. In the Deer Park area, house measurements were in closer agreement with the central site measurements. Concentrations measured at Aldine House 7 (A7) had the highest median values for 20 compounds in the Aldine area. Concentrations measured at Aldine House 7 (A7) had the highest median values for 20 compounds in the Aldine area. Concentrations measured at Clinton House 10 (C10) had the highest median values for 11 compounds in the Clinton area. Concentrations measured at Clinton House 10 (C10) had the highest median values for 11 compounds in the Clinton area.

Supplementary Sampling In order to differentiate between very localized sources of VOCs outside homes and more area-wide sources, 5 OVMs were placed in different positions around each of two homes with elevated concentrations (House A7 and C10). In order to differentiate between very localized sources of VOCs outside homes and more area-wide sources, 5 OVMs were placed in different positions around each of two homes with elevated concentrations (House A7 and C10). 72-hour sampling performed twice during February-April. 72-hour sampling performed twice during February-April.

Results of Supplementary Sampling Samples from House A7 showed elevated levels of many compounds at only one position, that employed previously for the 6 warm weather samplings. Samples from House A7 showed elevated levels of many compounds at only one position, that employed previously for the 6 warm weather samplings. Interviews indicated that this position (above the garage door) was influenced by nearby resident activities (servicing/refueling of tree-cutting equipment, frequent idling and washing/waxing of vehicles in driveway). Interviews indicated that this position (above the garage door) was influenced by nearby resident activities (servicing/refueling of tree-cutting equipment, frequent idling and washing/waxing of vehicles in driveway). Samples from House C10 were uniformly elevated, indicating the influence of more remote source(s) affecting a wider area, but not characterized by the central site measurements. Samples from House C10 were uniformly elevated, indicating the influence of more remote source(s) affecting a wider area, but not characterized by the central site measurements.

Conclusions Passive sampling is a useful technique for evaluating spatial heterogeneity of ambient air concentrations of VOCs. Passive sampling is a useful technique for evaluating spatial heterogeneity of ambient air concentrations of VOCs. Concentrations of many VOCs in two Houston areas are frequently underestimated by measurements at central monitoring sites. Concentrations of many VOCs in two Houston areas are frequently underestimated by measurements at central monitoring sites. Elevated levels of compounds appear to be persistent, i.e., similar results were observed during warm and cooler weather sampling. Elevated levels of compounds appear to be persistent, i.e., similar results were observed during warm and cooler weather sampling. Spatial mapping of concentrations around homes can be used successfully to differentiate between observed elevated concentrations caused by specific localized resident activities (House A7) and those due to more remote sources affecting a larger area (House C10). Spatial mapping of concentrations around homes can be used successfully to differentiate between observed elevated concentrations caused by specific localized resident activities (House A7) and those due to more remote sources affecting a larger area (House C10).