oEPA
EPA/600/R-12/673 | September 2012 www.epa.gov/research
United States
Environmental Protection
Agency
Fluctuation of Indoor Radon
and VOC Concentrations
Due to Seasonal Variations
RESEARCH AND DEVELOPMENT
-------
-------
Fluctuation of Indoor Radon
and VOC Concentrations Due
to Seasonal Variations
Prepared for
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Las Vegas, NV89119
Prepared by
RTI International
3040 Cornwallis Road
Research Triangle Park, NC 27709
and
ARCADIS U.S., Inc.
4915Prospectus Drive, Suite F
Durham, NC 27713
RTI Project Number 0213151.001
Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect
official Agency policy. Mention of trade names and commercial products does not constitute
endorsement or recommendation for use.
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
-------
Table of Contents
1. Executive Summary
1.1 Background...
1.2 Purpose and Objectives ....
1.3 Conclusions
1.3.1 Seasonal Variation and Influence of HVAC
1.3.2 Relationship Between Subsurface and Indoor Air Concentrations ..
1.4.3 Relationship Between Radon and VOCs
Table of Contents
-l
-l
-l
-2
-2
-3
-3
1.3.4 Conclusions: The Use of External Soil Gas Samples as a Surrogate Sampling
Location 1-3
1.3.5 Conclusions: The Duration Over Which Passive Samplers (Solvent Extracted
Radial Style Charcoal) Provided Useful Integration of Indoor Air
Concentrations 1-4
1.3.6 Conclusions: Groundwater vs. Vadose Zone Sources as Controls on Indoor
Concentrations at This Site 1-4
2. Introduction 2-1
2.1 Background 2-2
2.1.1 Variability in Vapor Intrusion Studies 2-3
2.1.2 Vapor Attenuation Factors 2-7
2.1.3 Potential for Use of Radon as a Vapor Intrusion Tracer 2-8
2.1.4 Passive VOC Sampling 2-10
2.2 Objectives 2-14
2.2.1 Time Scale and Measurement of Independent and Dependent Variables 2-15
2.2.2 Data Quality Objectives and Criteria 2-16
3. Methods 3-1
3.1 Site Description 3-1
3.1.1 Area Geology/Hydrogeology 3-1
3.1.2 Area Potential Sources 3-3
3.1.3 Building Description 3-7
3.1.4 Building Occupancy During Sampling 3-9
3.1.5 Initial Site Screening 3-9
3.1.6 Initial Conceptual Site Model 3-13
3.2 Building Renovation 3-14
3.2.1 HVAC Refurbishment and Operations 3-14
3.2.2 Plumbing Refurbishment and Sealing 3-17
3.3 Monitoring Infrastructure Installation (Wells, SGPs, Embedded Temperature Sensors) 3-18
3.4 VOC Sampling and Analysis 3-23
3.4.1 Indoor (Passive, Summa Canister) 3-23
3.4.2 Subslab and Soil Gas (TO-17 and Summa Canister) 3-25
3.4.3 Online Gas Chromatograph 3-25
3.4.4 Groundwater 3-27
3.5 Radon Sampling and Analysis 3-27
3.5.1 Indoor Air Radon Sampling and Analysis 3-27
3.5.2 Subslab and Soil Gas Radon Sampling and Analysis 3-28
3.5.3 Continuous (Real-Time) Indoor Air Radon Sampling and Analysis 3-29
3.6 Physical Parameters Monitoring 3-29
3.6.1 On-Site Weather Station 3-29
3.6.2 Indoor Temperature 3-31
3.6.3 Soil Temperature 3-31
in
-------
Table of Contents
3.6.4 Soil Moisture 3-31
3.6.5 Potentiometric Surface/Water Levels 3-32
3.6.6 Differential Pressure 3-32
3.6.7 Air Exchange Rate 3-32
3.6.8 Crack Monitoring 3-33
4. Results and Discussion: Quality Assurance Checks of Individual Data Sets 4-1
4.1 VOC Sampling—Indoor Air-Passive—Air Toxics Ltd. (ATL) 4-1
4.1.1 Blanks 4-1
4.1.2 Surrogate Recoveries 4-3
4.1.3 Laboratory Control Sample Recoveries 4-3
4.1.4 Duplicates 4-3
4.2 VOC Sampling—Subslab and Soil Gas (TO-17)—U.S. EPA 4-4
4.2.1 Blanks 4-4
4.2.2 Surrogate Recoveries 4-7
4.2.3 Laboratory Control Sample Recoveries 4-7
4.3 VOC Sampling—Subslab and Soil Gas (TO-17)—ATL 4-8
4.3.1 Blanks 4-8
4.3.2 Surrogate Recoveries 4-10
4.3.3 Laboratory Control Sample Recoveries 4-11
4.3.4 Duplicates 4-11
4.4 VOC Sampling—Subslab and Indoor Air (TO-15)—ATL 4-12
4.4.1 Blanks 4-12
4.4.2 Surrogate Recoveries 4-13
4.4.3 Laboratory Control Sample Recoveries 4-13
4.4.4 Duplicates 4-14
4.5 Online GC (Soil Gas and Indoor Air) 4-14
4.5.1 Blanks 4-14
4.5.2 Initial Calibration 4-15
4.5.3 Continuing Calibration 4-15
4.5.4 Calibration Check via Comparison to Fixed Laboratory (TO-15 vs. Online
GC) 4-15
4.5.5 Agreement of Online GC Results with TO-17 Verification Samples 4-16
4.5.6 Agreement of Integrated Online GC Results with Passive Samplers 4-18
4.6 Radon 4-25
4.6.1 Indoor Air: Comparison of Electrets Field, ARCADIS to Charcoal Analyzed
by U.S. EPA Radiation and Indoor Environment (R&IE) National Laboratory 4-25
4.6.2 Comparision of Average of Real Time Alphaguard to Electrets and Charcoal
Canisters 4-28
4.6.3 Quality Assurance Checks of Electrets 4-30
4.7 On-Site Weather Station vs. National Weather Service (NWS) 4-31
4.8 Database 4-33
4.8.1 Checks on Laboratory Reports 4-33
4.8.2 Database Checks 4-34
5. Results and Discussion: VOC Concentration Temporal Trends and Relationship to HVAC 5-1
5.1 VOC Seasonal Trends Based on Weekly, Biweekly, and Monthly Measurements for
52+Weeks 5-1
5.1.1 Indoor Air 5-1
5.1.2 Subslab Soil Gas 5-4
5.1.3 Shallow and Deep Soil Gas 5-8
5.2 Radon Seasonal Trends (based on Weekly Measurements) 5-23
IV
-------
Table of Contents
5.2.1 Indoor Air 5-23
5.2.2 Subslab and Wall Port Soil Gas 5-26
5.2.3 Deep Soil Gas 5-29
5.3 VOC Short-Term Variability (Based on Daily and Hourly VOC Sampling) 5-35
5.3.1 Indoor air 5-35
5.3.2 Subsurface Soil Gas Data 5-41
5.4 Radon Short Term Variability (Based on Daily and More Frequent Measurements) 5-48
5.4.1 Indoor Air 5-48
5.4.2 Subslab, Wall Port, and Deeper Soil Gas Radon Data 5-52
5.5 Outdoor Climate/Weather Data 5-52
6. Results and Discussion: Establishing the Relationship between VOCs and Radon in
Subslab/Subsurface Soil Gas and Indoor Air 6-1
6.1 Correlation between Soil Gas VOC and Radon Concentrations 6-3
6.2 Correlation between the Indoor Air Concentration and Radon 6-15
6.3 Radon and VOC Soil Gas Spatial Distributions 6-16
6.4 Spatial Correlations in Radon and VOCs Analyzed Separately 6-18
6.5 Correlations in Indoor Air VOC and Radon Temporal Trends 6-22
7. Results and Discussion: Attenuation of Soil Gas VOCs and Radon 7-1
7.1 Subslab to Indoor Air Attenuation Factor Temporal Range 7-1
7.2 Subslab Attenuation Factors for Each Side of the Duplex 7-3
7.3 Attenuation Factors Calculated for Each Side of the Duplex Using Subslab and
Shallow Soil Gas Samples 7-6
8. Results and Discussion: Can Near-Building External Samples Be Used as a Surrogate
Sampling Location? 8-1
8.1 Comparison of External Soil Gas to Subslab Soil Gas 8-1
8.1.1 Chloroform 8-1
8.1.2 Tetrachloroethylene (PCE) 8-2
8.1.3 Trichloroethylene (TCE) 8-3
8.1.4 Radon 8-4
8.2 Comparison of Wall Ports to Subslab External Soil Gas 8-5
9. Results and Discussion: Over what durations do solvent extracted passive samplers provide
useful integration of indoor air concentrations? (Is uptake rate constant?) 9-1
9.1 Comparison of Daily to Weekly Samples 9-5
9.2 Comparison of Weekly to Biweekly Samples 9-8
9.3 Comparison of Weekly to Monthly Samples 9-9
9.4 Comparison of Weekly to Quarterly Samples 9-10
9.5 Comparison of Weekly to Semiannual and Annual Samples 9-10
9.5.1 Comparing Radiello Samplers to SKC Samplers 9-11
9.6 Conclusions 9-12
10. Results and Discussion: Determine if observed changes in indoor air concentration of volatile
organics of interest are mechanistically attributable to changes in vapor intrusion 10-1
10.1 Air Exchange Rate Results and Seasonal Variability—Does this Control Indoor Air
Concentration? 10-1
10.2 Direct Differential Pressure Results—Are They Predictive of Indoor Air
Concentrations by Themselves? 10-3
10.3 Inferred Driving Force from Temperature Differentials—Is This Predictive of Indoor
Air Concentrations by Itself? 10-8
10.4 HVAC System Cycles 10-12
-------
Table of Contents
10.5 Trends in Subslab Concentration—Do They Predict Indoor Air Concentration Trends
by Themselves? 10-16
10.6 Do Trends In Shallow and Deep Soil Gas Predict Indoor Air Concentrations? 10-16
10.7 Ambient Concentrations—Are They Significant? 10-17
10.8 Potentiometric Surface/Water Levels—Do They Predict Subslab Concentrations? 10-17
11. Results and Discussion: Do groundwater concentrations control soil gas concentrations at this
site? And thus indoor air concentrations? 11-1
11.1 Potentiometric Surface Changes (and Correlation to Local Surface Water Bodies) 11-1
11.2 Groundwater Concentration Trend 11-5
11.2.2 Is the Groundwater Concentration Trend Correlated to Potentiometric
Surface? 11-8
11.2.3 Is the Groundwater Concentration Trend Correlated to Indoor Air
Concentrations? 11-8
11.3 Soil Moisture Trends 11-8
11.3.1 Correlation with Rainfall Measurements 11-8
11.3.3 Relationship to Observed Stratigraphy 11-9
11.4 Correlation of Groundwater Concentration Changes to Deep Soil Gas 11-11
11.5 Revisions to Conceptual Site Model 11-14
12. Results and Discussion: Special Studies 12-1
12.1 Summary of Temporary vs. Permanent Subslab Sampling Study 12-1
12.2 Summary of Fan Testing 12-1
12.2.1 Fan Test Objectives 12-3
12.2.2 Fan Test Experimental Methods 12-3
12.2.3 Fan Test Results and Discussion 12-4
12.2.4 Fan Test Lessons Learned 12-11
12.3 Testing Utility of Consumer-Grade Radon Device (Safety Siren Pro) 12-11
12.3.1 Consumer-Grade Radon Device Test Objectives 12-12
12.3.2 Consumer-Grade Radon Device Test Methods 12-12
12.3.3 Consumer-Grade Radon Detector Test Results and Discussion 12-12
13. Conclusions and Recommendations 13-2
13.1 Conclusions 13-2
13.1.1 Seasonal Variation and Influence of HVAC 13-2
13.1.2 The Relationship Between Subsurface and Indoor Air Concentrations 13-2
13.1.3 The Relationship Between Radon and VOCs 13-3
13.1.4 The Use of External Soil Gas Samples as a Surrogate Sampling Location 13-3
13.1.5 The Duration Over Which Passive Samplers (Solvent Extracted Radial Style
Charcoal) Provided Useful Integration of Indoor Air Concentrations 13-3
13.1.6 Groundwater vs. Vadose Zone Sources as Controls on Indoor Concentrations
at this Site 13-4
13.2 Practical Implications for Practitioners 13-4
13.2.1 Sampling to Characterize Seasonal Variations 13-4
13.2.2 Using Fan Induced Depressurization in Vapor Intrusion Studies 13-5
13.2.3 Performance of Temporary Subslab Sampling Ports 13-5
13.2.4 Performance of Consumer Grade Radon Detector 13-6
13.3 Recommendations 13-6
13.3.1 Recommendations for Vapor Intrusion Research Generally 13-6
13.3.2 Recommendations Regarding Further Study of this Test Site 13-7
14. References 14-1
VI
-------
Table of Contents
Appendix A: Soil Boring Logs
Appendix B: Temporary vs. Permanent Subslab Port Study
List of Figures
2-1. An overview of important vapor intrusion pathways (U.S. EPA, 2002a) 2-2
2-2. Soil gas and groundwater concentrations below a slab (Schumacher et al., 2010) 2-6
3 -1. Lithological fence diagram showing the maj or soil types beneath the 422/420 house 3-2
3-2. Aerial view of duplex, 420/422 East 28th Street, showing nearby sanitary and storm
sewers 3-3
3 -3. East side of house (on right) and adj oining commercial quadraplex visible (left) 3-4
3-4. Roof of adjacent commercial quadraplex 3-4
3-5. Looking toward southeast corner of adjacent commercial quadraplex 3-5
3-6. Visual evidence of historic dry cleaners in area 3-6
3-7. Front view of house during summer 2011 sampling, with fan testing and weather station 3-7
3-8. Front view of duplex under winter conditions showing designation of sides and FfVAC
setup 3-8
3-9. 422 (left) and 420 East 28th Street in January 2011 3-8
3-10. Test building floor plan showing sampling locations used in preliminary screening 3-11
3-11. Basement supply register in newly installed FfVAC system 3-14
3-12. Common returns from first and second floors in newly installed FfVAC system 3-15
3-13. Gas-fired forced hot air FTVAC system installed in 422 3-15
3-14. Floor cracks in 422 basement, central area, contrast enhanced 3-17
3-15. Weathered cement in walls and floor cracks in 422, contrast enhanced 3-17
3-16. Floor drain, 422, 1st floor laundry area 3-18
3-17. Nested monitoring well 1 and SGP1 are located immediately south of the 422 side front
wall. SGP1-16.5 and MW-1A is by the wall, to the left of the sign. SGP1-9 and 1-13 as
well as MW-1C are by the wall, to the right of the sign (next to the pile of bricks). SGP1-
3.5 and 1-6, and MW-1B are in the installation visible in the center foreground 3-19
3-18. Exterior of test building showing utility corridors, ground surface cover, monitoring wells
(MWs), soil gas points (SGPs), thermocouples (TCs), and moisture sensors (MSs) 3-20
3-19. Interior floor plans of 420 and 422 East 28th Street showing sampling locations 3-21
3-20. Interior SGP9 (top) and SSP-4 (bottom) 3-22
3-21. Wall port 2 3-22
3-22. Passive indoor air sampling rack: 422 first floor 3-23
3-23. Ambient sampler shelters on telephone pole near duplex 3-24
3 -24. Monitoring well MW-3, installed in the basement and completed on the first floor 3-27
3-25. Front view of 420/422 duplex with location of weather station sensors indicated with red
arrow 3-30
3-26. Calibrated crack monitor 3-33
4-1. Correlation between radon measured using the electret and charcoal methods 4-27
4-2. Aerial view of study house, showing potential influences on wind velocity; red arrow
indicates study house 4-32
4-3. Comparison of National Weather Service Indianapolis temperature data to weather
station at 422 East 28th Street 4-32
4-4. Comparison of National Weather Service Indianapolis relative humidity to weather
station at 422 East 28th Street 4-33
4-5. Comparison of National Weather Service wind speed data to weather station at 422 East
28th Street 4-33
5-1. PCE in indoor and ambient air vs. time (7-day Radiello samples) 5-2
vn
-------
Table of Contents
5-2. Chloroform in indoor and ambient air vs. time (7-day Radiello samples) 5-2
5-3. PCE, benzene, and toluene in indoor and ambient air 5-3
5-4. PCE and chloroform in 422 first-floor indoor air; weekly, biweekly, and monthly
duration Radiello samples 5-3
5-5. Plot of subslab chloroform concentrations vs. time 5-4
5-6. Plot of subslab chloroform concentrations vs. time, first intensive sampling period 5-5
5-7. Plot of subslab PCE concentrations vs. time 5-6
5-8. Plot of subslab PCE concentrations vs. time, first intensive sampling period 5-6
5-9. Plot of wall port chloroform concentrations vs. time 5-7
5-10. Plot of wall port PCE concentrations vs. time 5-8
5-11. Chloroform concentrations at each of the SGP1 ports vs. time 5-11
5-12. PCE concentrations at each of the SGP1 ports vs. time 5-11
5-13. Chloroform concentrations at each of the SGP2 ports vs. time 5-12
5-14. PCE concentrations at each of the SGP2 ports vs. time 5-12
5-15. Chloroform concentrations at each of the SGP3 ports vs. time 5-13
5-16. PCE concentrations at each of the SGP3 ports vs. time 5-13
5-17. Chloroform concentrations at each of the SGP4 ports vs. time 5-14
5-18. PCE concentrations at each of the SGP4 ports vs. time 5-14
5-19. Chloroform concentrations at each of the SGP5 ports vs. time 5-15
5-20. PCE concentrations at each of the SGP5 ports vs. time 5-15
5-21. Chloroform concentrations at each of the SGP6 ports vs. time 5-16
5-22. PCE concentrations at each of the SGP6 ports vs. time 5-16
5-23. Chloroform concentrations at each of the SGP7 ports vs. time 5-17
5-24. PCE concentrations at each of the SGP7 ports vs. time 5-17
5-25. Chloroform concentrations at SGP8 and9 ports vs. time 5-18
5-26. PCE concentrations at SGP8 and 9 ports vs. time 5-18
5-27. Chloroform concentrations at each of the SGP10 ports vs. time 5-19
5-28. PCE concentrations at each of the SGP10 ports vs. time 5-19
5-29. Chloroform concentrations at each of the SGP11 ports vs. time 5-20
5-30. PCE concentrations at each of the SGP11 ports vs. time 5-20
5-31. Chloroform concentrations at each of the SGP12 ports vs. time 5-21
5-32. PCE concentrations at each of the SGP12 ports vs. time 5-21
5-33. Subslab PCE concentrations over a 1-week period during the first intensive round 5-22
5-34. Subslab PCE concentrations over a 1-week period during the second intensive round 5-22
5-35. Weekly electret readings for all locations 5-24
5-36. Data for the downstairs continuously recording Alphaguard versus time 5-25
5-37. Data for the upstairs continuously recording Alphaguard versus time 5-26
5-38. This is a kriged radon image taken from subslab, wall, and multidepth soil gas data 5-27
5-39. Subslab Alphaguard data versus time 5-28
5-40. Wall port Alphaguard data versus time 5-28
5-41. Kriged radon image taken from subslab, wall, and multidepth soil gas data 5-30
5-42. Handheld Alphaguard data taken from soil gas ports at location 1 versus time 5-31
5 -43. Handheld Alphaguard data taken from soil gas ports at location 4 versus time 5-31
5-44. Handheld Alphaguard data taken from soil gas ports at location 5 versus time 5-32
5-45. Handheld Alphaguard data taken from soil gas ports at location 7 versus time 5-32
5-46. Handheld Alphaguard data taken from soil gas ports at location 8 versus time 5-33
5-47. Handheld Alphaguard data taken from soil gas ports at location 9 versus time 5-33
5-48. Handheld Alphaguard data taken from soil gas ports at location 10 versus time 5-34
5-49. Handheld Alphaguard data taken from soil gas ports at location 11 versus time 5-34
5-50. Handheld Alphaguard data taken from soil gas ports at location 12 versus time 5-35
5-51. Online GC chloroform indoor air data for 422 first floor 5-36
Vlll
-------
Table of Contents
5-52. Online GC chloroform indoor air data for 422 basement 5-36
5-53. Online GC chloroform indoor air data for 420 first floor 5-38
5-54. Online GC chloroform indoor air data for 420 basement 5-38
5-55. Online GC PCE indoor air data for 422 first floor 5-39
5-56. Online GC PCE indoor air data for 422 basement 5-39
5-57. Online GC PCE indoor air data for 420 first floor 5-40
5-58. Online GC PCE indoor air data for 420 basement 5-40
5-59. Online GC subsurface chloroform soil gas data—Phase 1 and Phase 2 5-42
5-60. Online GC subsurface chloroform soil gas data—Phase 1 5-42
5-61. Online GC subsurface chloroform soil gas data—Phase 2 5-43
5-62. Online GC subsurface PCE soil gas data—Phase 1 and Phase 2 5-44
5-63. Online GC subsurface PCE soil gas data—Phase 1 5-44
5-64. Online GC subsurface PCE soil gas data—Phase 2 5-45
5-65. Method TO-17 data for SSP-4 5-46
5-66. Online GC PCE measurements in SSP-4 5-46
5-67. Comparison of online GC measurements of PCE and chloroform in SGP9 at 6 ft 5-47
5-68. Electret indoor air radon concentrations for the first intensive round 5-49
5-69. Electret indoor air radon concentrations for the second intensive round 5-49
5-70. Radon concentrations from the downstairs stationary Alphaguard during the second
intensive round 5-50
5-71. Radon concentrations from the upstairs stationary Alphaguard during the second
intensive round 5-50
5-72. Electret indoor air radon concentrations for the third intensive round 5-51
5-73. Radon concentrations from the downstairs stationary Alphaguard during the third
intensive round 5-51
5-74. Radon concentrations from the upstairs stationary Alphaguard during the third intensive
round 5-52
5-75. Temperature records from the external temperature monitor and the HOBO devices at
seven indoor locations on the 422 and 420 sides of the house 5-54
5-76. Indoor temperature as recorded inside the 422 second floor office 5-55
5-77. Stacked hydrological graph with depth to water in feet (top—red circles), discharge at
Fall Creek in ft3/s (middle—blue line), and rainfall in inches (bottom—green line) 5-56
5-78. Plot of barometric pressure (inches of Hg) external to the 422/420 house overtime 5-57
5-79. Barometric pressure (Pa) on the 422 side of the house overtime 5-58
6-1. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and
radon(blue)forSGPll at 6 ft bis 6-4
6-2. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP1 at 9 ft bis 6-5
6-3. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP4 at 9 ft bis 6-6
6-4. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP4 at 13 ft bis 6-7
6-5. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP5 at 9 ft bis 6-8
6-6. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP7 at 9 ft bis 6-9
6-7. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP8 at 6 ft bis 6-10
6-8. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP9 at 6 ft bis 6-11
IX
-------
Table of Contents
6-9. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP10 at 6 ft bis 6-12
6-10. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP 11 at 6 ft bis 6-13
6-11. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP12 at 6 ft bis 6-14
6-12. Cross-correlation plots for the time series of log-VOCs and radon by indoor air location 6-15
6-13. Comparison of nearly collocated subslab and shallow internal soil gas ports 6-17
6-14. Concentration distributions (ug/m3 [VOCs] or pCi/L [radon]) and significance tests for
nearly collocated subslab and shallow internal soil gas ports 6-18
6-15. Evaluation of spatial effect north and south basement by VOC and radon for 422 East
28th St.—cumulative distribution plots where the x axis represents concentration (ug/m3
or pCi/L) and the y-axis concentration 6-19
6-16. Evaluation of spatial effect north and south basement by VOC and radon for 420 East
28th Street—cumulative distribution plots where the x axis represents concentration
(ug/m3 or pCi/L) and the y-axis concentration 6-19
6-17. Cross-correlation between north and south basement indoor air by VOCs and radon 6-20
6-18. Comparison of temporal trends at north and south basement sampling locations 6-21
6-19. Autocorrelation function for chloroform, PCE, and radon by location (site and
north/south basement) 6-23
7-1. Subslab (or 9-ft soil gas) to indoor air AFs for individual sample locations, with the
number of calculated AFs in each case indicated by the number directly below each
whisker 7-2
7-2. Subslab to indoor air attenuation factors, calculated for each side of the duplex using only
subslab points 7-4
7-3. Range of weekly chloroform and PCE concentrations in indoor air, subslab, and 9-ft
interior soil gas samples over study period, with the number of calculated AFs in each
case indicated below the whiskers 7-5
7-4. Attenuation factors vs. time: calculated for each side of the duplex using subslab soil gas
ports and the shallowest of the nested interior soil gas ports 7-7
8-1. Box and whisker plots of chloroform distribution in soil gas at varying depths
(concentration is log scale) 8-2
8-2. Box and whisker plots of PCE distribution at various depths 8-3
8-3. Box and whisker plots of TCE distribution at various depths 8-4
8-4. Box and whisker plots of radon concentration at various depths 8-5
9-1. Kernal densities of%Bias for important VOCs 9-4
9-2. The effect of vapor pressure on sorbent performance 9-5
10-1. 422 subslab vs. basement differential pressure (positive values indicate greater
pressurization of the subslab and thus flow toward the basement) 10-4
10-2. Basement vs. upstairs differential pressure (positive values indicate pressurization of the
basement relative to the upstairs) 10-5
10-3. Deep vs. shallow soil gas differential pressure beneath 422 East 28th Street (positive
values indicate a greater pressure in the deep soil gas relative to the shallow soil gas) 10-5
10-4. Basement vs. exterior (above grade) differential pressure at 422 East 28th Street (positive
values indicate that the basement pressure is higher than the pressure in exterior air) 10-6
10-5. Subslab vs. basement differential pressure at 420 East 28th Street (positive values
indicate higher pressure in subslab than in the basement, thus flow toward the basement) 10-7
10-6. Exterior barometric pressure measurements overtime, 420/422 East 28th Street 10-7
10-7. Stack effect driving force for 422 East 28th Street overtime 10-9
10-8. Stack effect driving force 420 (unheated) side overtime 10-9
10-9. Summer and fall interior and ambient temperatures 10-10
-------
Table of Contents
10-10. PCE concentrations in indoor air vs. stack effect driving force (log scale of
concentration) 10-11
10-11. Chloroform concentrations in indoor air vs. stack effect driving force (log scale of
concentration) 10-11
10-12. Radon concentrations (electret measurements) in indoor air vs. stack effect driving force 10-12
10-13. Differential pressure measurements graphed with HVAC system on/off cycles; HVAC
system status at times when basement vs. exterior differential pressure was observed is
annotated 10-14
10-14. Selected period of indoor and ambient temperature data, green arrow marks December 18
to December 23 HVAC system outage 10-14
10-15. Selected period of indoor and ambient temperature data, green arrow shows March 11 to
March 16 period of HVAC outage 10-15
10-16. PCE, online GC data, 422 side, larger data points used to mark periods of heating system
failure 10-15
10-17. Average depth to water at house, discharge of nearby creek (USGS) and rainfall at house
compared (intensive sampling rounds shown with vertical lines) 10-18
10-18. Storm and sanitary sewers near the test duplex at 420/422 East 28th Street 10-19
10-19. Floor drain in first floor laundry room on 422 side of duplex 10-20
11-1. Stacked graph presenting depth to water in feet (top—red circles), discharge at Fall Creek
in ft3/s (middle—blue line), and rainfall in inches (bottom—green line) 11-2
11-2. Monitoring well water data for PCE versus time 11-6
11-3. Flooded SGP water data for PCE versus time 11-7
11-4. Soil moisture: irrometer moisture data in centibars for the interior and exterior of the
422/420 house 11-10
11-5. Lithological fence diagram showing some of the major soil types beneath the 422/420
house 11-10
11-6. Lithological fence diagram showing some of the major soil types beneath the 422/420
house 11-11
11-7. PCE concentrations at each of the SGP1 ports overtime 11-12
11-8. PCE concentrations at each of the SGP2 ports vs. time 11-12
11-9. PCE concentrations at each of the SGP5 ports overtime 11-13
11-10. PCE concentrations at each of the SGP6 ports overtime 11-13
11-11. PCE concentrations at SGP8 and SGP9 ports overtime 11-14
11-12. Temporal plot of log indoor air concentration for VOCs (|ig/m3) and radon (pCi/L) by
sample location over the study period 11-16
12-1. Fan test matrix 12-4
12-2. Fan position in stairwell (note plastic sheet over doorway) 12-4
12-3. Differential pressure before, during, and after fan tests (fan tests denoted by vertical
bars) 12-5
12-4. Radon on second floor before, during and after fan tests (fan tests denoted by vertical
bars) 12-6
12-5. VOC Field instrument data before, during, and after fan tests (fan tests denoted by
vertical bars) 12-7
12-6. Note height of basement window and vent (sealed for this study) 12-8
12-7. Subslab and soil gas VOC data during fan test period 12-9
12-8. VOC data before, during, and after fan testing, Method TO-15 12-10
12-9. Comparison of fan test responses of radon, PCE, and chloroform in 422 basement 12-11
12-10. Comparison of electret and Safety Siren results 12-13
XI
-------
Table of Contents
List of Tables
2-1. VOC Indoor Air Sampling Method Options 2-13
2-2. Factors Causing Temporal Change in Vapor Intrusion and How They Are Observed and
Measured 2-15
2-3. Data Quality Objectives and Criteria 2-17
3-1. Preliminary Indoor Air VOC Screening Results—Fan Off, Basement 3-10
3-2. Preliminary Indoor Air Radon Screening—Fan Off, Basement 3-12
3-3. VOC Results (ug/m3) for Subslab and Soil Gas at 422 E. 28th St., Indianapolis—Fan Off 3-12
3-4. VOCs (ug/m3) in Indoor Air and Subslab Soil Gas, 420 & 422 E. 28th St., Indianapolis-
Fan On 3-12
3-5. Radon (pCi/L) in Indoor Air & Subslab Gas at 420 & 422 E. 28th St. Indianapolis— Fan
Off and On 3-12
3-6. Groundwater Screening Data 3-13
3-7. Soil Analysis from MW-1 Boring at Multiple Depths 3-13
4-1. Indoor Air Passive Field Blank Summary—Radiello 130 4-1
4-2. Indoor Air Passive Trip Blank Summary—Radiello 130 4-2
4-3. Indoor Air Passive Laboratory Blank Summary—Radiello 130 4-2
4-4. Indoor Air Passive Surrogate Summary—Radiello 130 4-3
4-5. Indoor Air Passive LCS Summary—Radiello 130 4-3
4-6. Indoor Air Passive Laboratory Precision (LCS/LCSD) Summary—Radiello 130 4-4
4-7. Subslab and Soil Gas—EPA Field Blank Summary—TO-17 4-5
4-8. Subslab and Soil Gas—EPA Trip Blank Summary—TO-17 4-6
4-9. Subslab and Soil Gas—EPA Laboratory Blank Summary—TO-17 4-6
4-10. Subslab and Soil Gas—EPA Fridge Blank Summary—TO-17 4-7
4-11. EPA TO-17 Surrogate Recovery Summary 4-7
4-12. EPA TO-17 LCS Summary 4-8
4-13. Subslab and Soil Gas—ATL Field Blank Summary—TO-17 4-9
4-14. Subslab and Soil Gas—ATL Trip Blank Summary—TO-17 4-10
4-15. Subslab and Soil Gas—ATL Lab Blank Summary—TO-17 4-10
4-16. ATL TO-17 Surrogate Recovery Summary 4-11
4-17. ATL TO-17 LCS Summary 4-11
4-18. ATL TO-17 Laboratory Precision (LCS/LCSD) Summary 4-12
4-19. Subslab and Indoor Air—ATL Lab Blank Summary—TO-15 4-13
4-20. ATL TO-15 Surrogate Recovery Summary 4-13
4-21. ATL TO-15 LCS Summary 4-14
4-22. ATL TO-15 Laboratory Precision (LCS/LCSD) Summary 4-14
4-23. Interlaboratory Results: Spiked Verification Samples 4-17
4-24. Interlaboratory Statistics: Spiked Verification Samples 4-18
4-25. Comparison of Online GC to Radiello Results 4-19
4-26. Comparison between Electrets and Charcoal Canisters at the 422/420 EPA House from
January 19-26,2011 4-26
4-27. Comparison of Electret and Charcoal Canister Data from April 27, 2011, to May 4, 2011 4-26
4-28. Comparison of Charcoal and Electret Radon December 28, 2011, to January 4, 2012 4-27
4-29. Comparison between 422 Base N Alphaguards and Electrets from March 30, 2011, and
May 18,2011 4-28
4-30. Comparison of Real-Time Alphaguard to Integrated Electret August through October 4-28
4-31. Comparison of Real-time Alphaguards to Integrated Electret Measurements December
28, 2011 to January 4, 2012 4-29
xn
-------
Table of Contents
4-32. Comparison of Real-Time Alphaguard to Integrated Electret Measurements January
through March 4-30
5-1. Frequency ofNondetects (%) by Soil Gas Point or Cluster 5-9
5-2. Frequency ofNondetects by Depth and Compound 5-9
5-3. Summary Meteorological Data for Central Indiana 5-53
6-1. Counts of Records with Flag by VOCs and Flag Type for Soil Gas 6-1
6-2. Counts of Records with Flag by VOCs and Flag Type for Indoor Air 6-2
6-3. Comparison of Mean and Median Concentrations in North and South Sampling
Locations, 422 Side of Duplex 6-22
9-1. Summary Statistics for %Bias by Comparison Period and VOC 9-3
9-2. Summary Statistics Individual Concentration Measurements by VOC and Period 9-6
9-3. Summary Statistics for %Bias Comparing Daily vs. Weekly Period and VOC 9-7
9-4. Evaluation of Daily vs. Weekly Differences 9-8
9-5. Evaluation of Weekly vs. Biweekly Differences (Outlier Included) 9-9
9-6. Evaluation of Weekly vs. Biweekly Differences (Outlier Removed) 9-9
9-7. Evaluation of Weekly vs. Monthly Differences 9-10
9-8. Evaluation of Weekly vs. Quarterly Differences 9-10
9-9. Evaluation of Weekly vs. Semiannual Differences 9-11
9-10. Evaluation of Weekly vs. Annual Differences 9-11
9-11. Average %Bias for Average Weekly Radiello Measurements Compared with Semiannual
and Annual Modified SKC 575 Charcoal Badges 9-12
10-1. April/May 2011 Air Exchange Rate Measurement Results 10-1
10-2. September 2011 Air Exchange Rate Measurement Results 10-2
10-3. National Survey of Air Exchange Rates, Reprinted from the EPA Exposure Factor
Handbook (U.S. EPA, 2011) 10-2
10-4. Drain Sampling Data April 13-April 21, 2011 ((ig/m3) 10-21
10-5. NY Times Database Report on Indianapolis Drinking Water Showing Relative
Concentrations of PCE and Chloroform 10-21
11-1. External Soil Gas Locations and Their Flooded Status during Different Times 11-3
11-2. Internal Soil Gas Locations and Their Flooded Status during Different Times 11-4
11-3. Groundwater Monitoring Well Information 11-5
12-1. Quality Objectives and Criteria for Special Studies 12-2
12-2. Comparison of Safety Siren, Alphaguard, and Electret Data 12-13
Xlll
-------
Notice
The information in this document has been funded wholly by the United States Environmental Protection
Agency under contract number EP-C-11-036 to the Research Triangle Institute. It has been subjected to
the Agency's peer and administrative review and has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement or recommendation for
use.
Acknowledgments
This project was conceived, directed, and managed by Brian Schumacher and John Zimmerman of US
EPA NERL. Robert Truesdale (RTI International) and Chris Lutes (ARCADIS) led the project, with
report input by Brian Cosky (ARCADIS), Breda Munoz and Robert Norberg (RTI), Heidi Hayes (Air
Toxics Ltd.), and Blayne Hartman (Hartman Environmental Geoscience). The authors would also like to
thank the following for their valuable input to the project:
• Leigh Riley Evens, Executive Director, Jackie Nytes, Former Executive Director, and
Nate Lichti, Assistant Director, Mapleton-Fall Creek Development Corporation, who
loaned the house used for this work.
• Dale Greenwell EPA NRMRL and Ron Mosley (retired), radon and equipment support
• Gregory Budd, US EPA Radiation and Indoor Environments National Laboratory, radon
QC sample analysis and instrument support
• Ausha Scott, analytical support, Air Toxics Ltd.
• Alan Williams and Jade Morgan, U.S. EPA NERL, TO-17 analytical support
• Robert Uppencamp ARCADIS - field, data interpretation, and site selection support;
Sara Jonker, ARCADIS, field support
• Rebecca Forbort and Valerie Kull, ARCADIS, Shu-yi Lin, RTI, data management and
analysis
• Scott Forsberg, Harvard School of Public Health, air exchange rate measurement support
• Susan Beck and Sharon Barrell, RTI, document preparation and editing support.
XIV
-------
Section 1—Executive Summary
Table of Contents
1. Executive Summary 1-1
1.1 Background 1-1
1.2 Purpose and Objectives 1-1
1.3 Conclusions 1-2
1.3.1 Seasonal Variation and Influence of HVAC 1-2
1.3.2 Relationship Between Subsurface and Indoor Air Concentrations 1-3
1.4.3 Relationship Between Radon and VOCs 1-3
1.3.4 Conclusions: The Use of External Soil Gas Samples as a Surrogate Sampling
Location 1-3
1.3.5 Conclusions: The Duration Over Which Passive Samplers (Solvent Extracted
Radial Style Charcoal) Provided Useful Integration of Indoor Air
Concentrations 1-4
1.3.6 Conclusions: Groundwater vs. Vadose Zone Sources as Controls on Indoor
Concentrations at This Site 1-4
-------
Section 1—Executive Summary
1. Executive Summary
1.1 Background
Vapor intrusion is the migration of subsurface vapors, including radon and volatile organic compounds
(VOCs), from the subsurface to indoor air. Vapor intrusion has emerged as a priority VOC exposure
pathway at many hazardous waste sites nationwide. Vapor intrusion occurs because of the pressure and
concentration differentials between indoor air and soil gas. Indoor environments are often negatively
pressurized with respect to outdoor air and soil gas, and this pressure difference allows soil gas with
subsurface vapors to flow into indoor air through advection. In addition, concentration differentials may
cause VOCs to migrate from areas of higher to lower concentrations through diffusion, which also causes
vapor intrusion.
The vapor intrusion exposure pathway extends from the contaminant source, which can be free product or
contaminated groundwater, to indoor air exposure points. Contaminated matrices therefore may include
groundwater, soil, soil gas, and indoor air. VOC contaminants of concern typically include halogenated
solvents such as trichloroethene (TCE), tetrachloroethene (PCE), and chloroform, and degradation
products of TCE and PCE including dichloroethenes and vinyl chloride. Petroleum hydrocarbons, such as
the aromatic VOCs benzene, toluene, and xylenes, can also cause vapor intrusion. Radon is a colorless
radioactive gas that is released by radioactive decay of radionuclides in soil, where it migrates into homes
through vapor intrusion in a similar fashion to VOCs. This project will focus on the vapor intrusion of
halogenated VOCs, which are relatively recalcitrant (resistant) to biodegradation in aerobic soils and
groundwater, and radon, which has a radioactive half-life of about 3.8 days.
1.2 Purpose and Objectives
The main purpose of this work assignment is to better characterize the spatial and temporal variability of
vapor intrusion by collecting a full year's dataset of weekly measurements of subslab soil gas, external
soil gas, and indoor air, on a single house that is impacted by vapor intrusion of radon and VOCs. By
examining both short-term and long-term (average annual) concentrations, the project will provide
valuable information on how to best take and evaluate measurements to estimate long-term, chronic risk
for VOCs. We also studied the relationship between radon and VOC vapor intrusion in a house affected
by both. The radon literature could provide valuable lessons for VOC vapor intrusion if there is a
relationship, and radon, being much cheaper to measure than VOCs, could be an important tool in
improving the investigation and mitigation of VOC vapor intrusion. Finally, we investigated the long-
term performance of modified sorbent-based measurement techniques for time-integrated measurements
of indoor air VOCs.
The project investigated distributional changes in VOC and radon concentrations in the indoor air,
subslab, and subsurface soil gas from an underground source (groundwater source and/or vadose zone
source) adjacent to a residence or small commercial building. The time frame of this study is 2 years in
order to evaluate the effects due to seasonal variations on radon and VOC vapor intrusion.
There were four primary objectives for this research effort:
1. Identify any seasonal fluxes in radon and VOC concentrations as they relate to a typical use of
HVAC in the building.
2. Establish relationship between subslab/subsurface soil gas and indoor air concentrations of VOCs
and possibly radon.
3. Determine relationship of radon to VOC concentrations at a given site.
1-1
-------
Section 1—Executive Summary
4. Examine if near-building external samples could be used as a surrogate sampling location.
Five secondary objectives were also addressed by the study:
1. Evaluate the duration over which solvent-extracted passive samplers provide useful integration of
indoor air concentrations (i.e., over what duration is the uptake rate constant?).
2. Characterize the near-building environment sufficiently to explain the observed variation of
VOCs and radon in indoor air.
3. Determine whether the observed changes in indoor air concentration of volatile organics of
interest can be mechanistically attributed to changes in vapor intrusion.
4. Confirm that the two analytical laboratories (Air Toxics and U.S. Environmental Protection
Agency [EPA]) can produce soil gas VOC data with sufficient agreement, that the variance
between laboratories is not significant compared with the variance within laboratories or the
changes in the underlying phenomena being observed.
5. Evaluate the extent to which groundwater concentrations and/or vadose zone sources control soil
gas and indoor air VOC concentrations at this site.
Characteristics of the experimental design and data quality objectives (DQOs) developed to meet these
objectives are provided in Section 2 of this document.
1.3 Conclusions
The conclusions of this study represent the fruit of an intensive study of a single early 20th century duplex
in a particular geological setting—glaciofluvial deposits in Indianapolis, IN. Few other VOC vapor
intrusion studies have collected a dataset of comparable detail, and those have been conducted in
buildings of significantly different age or geological context.1
1.3.1 Seasonal Variation and Influence of HVAC
• Lower VOC concentrations were observed in indoor air in summer. These VOC concentrations in
indoor air are controlled not only by "building envelope-specific" factors, but they are also
significantly influenced by seasonal variations in subsurface concentration distributions,
especially in shallow/subslab soil gas where a weaker seasonal trend was observed.
• In indoor air, peak concentrations were seen in different months of the 2011 winter for PCE
(January) and chloroform (March) on the first floor of this duplex. Temporal trends for
chloroform and PCE differed markedly in fall 201 I/winter 2012 between the heated and unheated
sides of the duplex: the unheated side showed a much steeper decline in spring than the heated
side. Thus, complex data patterns for multiple VOCs in the same structure can be expected even
in the absence of occupant-related sources or activities.
• Stack-effect driving force calculations based on measurements of indoor/outdoor temperature
differential were predictive of indoor air concentrations. These stack effects included not only the
winter stack effect but also solar stack effects observed during summer and early fall. The cooling
effect of window air conditioners appeared to provide some protection against vapor intrusion, at
least for radon, during the summer months.
Johnson, Op. Cit. also numerous case studies compiled in U.S. EPA (2012c).
1-2
-------
Section 1—Executive Summary
• A repeatable seasonal effect of higher concentrations during winter was seen for chloroform and
radon, but not all winters are equal. Winter 2011 and winter 2012 were very different
climactically, and peak PCE concentrations observed in January 2011 were not equaled in 2012.
Inter-year climatic variations are well known even by lay stakeholders, but their role in vapor
intrusion studies may be underappreciated.
1.3.2 Relationship Between Subsurface and Indoor Air Concentrations
• PCE, chloroform, and radon have different spatial patterns in soil gas at this site.
• PCE and chloroform appear to have deep sources.
• Soil gas VOCs at some, but not all, high concentration sampling ports display a similar temporal
pattern to that observed in indoor air, with higher concentrations during winter months.
• Sewer lines and laterals likely play some role in contaminant fate and transport in this system.
Elevated concentrations of PCE and chloroform are present in the headspace of sewer gas. Their
role in lateral transport through the vadose zone and into the subslab of the duplex will be
elucidated through future geophysical studies.
• There is a strong seasonal component to the PCE and chloroform indoor concentrations (see
Section 11). The seasonal component appears to be correlated to the strength of the stack effect,
but it is not the only variable that controls indoor air concentrations.
1.4.3 Relationship Between Radon and VOCs
• Long-term (weekly and greater) radon concentrations in subslab air were more stable than VOC
concentrations, presumably because the shallow soils themselves were the dominant source of
radon and VOCs originate at a greater depth/distance.
• Radon concentrations in indoor air showed approximately an order of magnitude short term (< 1
day) variation—greater short-term variation than was observed for VOCs.
• The 1-week integration time dataset for radon had less seasonal variability than VOCs in indoor
air.
• Statistical cross-correlation testing found that radon and VOCs were positively cross-correlated at
several indoor air sampling locations (5% critical level). In laymen's terms, we are quite
confident that when radon concentrations go up, VOC concentrations will also go up in indoor
air. Some cross-correlations of radon and VOCs were observed at soil gas ports, but these cross-
correlations were less consistent/strong.
• Radon provided a qualitative indication that soil gas was entering this house. Thus, radon would
have been a useful aid to VOC data interpretation if the house had been occupied and had
numerous potential indoor sources. However, long-term radon exposure would not have
completely predicted VOC exposure in this house over all time scales.
1.3.4 Conclusions: The Use of External Soil Gas Samples as a Surrogate Sampling
Location
• High concentrations of VOCs and radon were seen in tight loams directly under building (subslab
ports and 6-ft soil gas ports) but not in external soil gas above the level of the basement floor (3.5
ft bis).
• External soil gas samples collected at 6 ft bis, the depth of the basement floor, had substantial
VOC concentration variability and would have underpredicted subslab concentrations.
• In deep soil gas (13 and 16.5 ft), there was close agreement between the mean chloroform and
radon concentrations at points underneath the building and outside of the building. In deep soil
1-3
-------
Section 1—Executive Summary
gas, PCE concentrations appeared lower on average and more variable for the points outside of
the building than for the points beneath the building.
1.3.5 Conclusions: The Duration Over Which Passive Samplers (Solvent Extracted
Radial Style Charcoal) Provided Useful Integration of Indoor Air Concentrations
• Excellent agreement was observed between numerical averages of successive 7-day exposure
samples with the results of single passive samplers exposed for 14 days (almost always within
+/- 30%) for all compounds, despite dramatic temporal variability. This suggests uniform uptake
rates for these time periods.
• The PCE, benzene, hexane, and toluene passive samplers tested provide good integration over
durations from 7 to 28 days. Chloroform integration was less effective for durations greater than 2
weeks.
• The PCE and toluene passive samplers provide good integration of concentrations over durations
from 7 to 364 days.
• Temporal variability in 1-week duration indoor VOC samples over the course of a year of >20x
were observed. For certain less-volatile compounds, passive samplers allow cost-effective
acquisition of long-term average concentration data.
• Vapor pressure predicted well the relative durations over which different compounds could be
collected with the passive samplers.
1.3.6 Conclusions: Groundwater vs. Vadose Zone Sources as Controls on Indoor
Concentrations at This Site
• The potentiometric surface at this house responds within days to rain events.
• Chloroform concentration trends visually correlate with hydrogeological changes.
• Chloroform concentrations in soil gas peak have their highest concentrations just above the water
table.
• Chloroform is present in highest concentration in deep soil gas. Substantial chloroform has been
historically detected in groundwater on a site 200 ft to the southwest. Chloroform was also
detected in groundwater at this house in preliminary sampling. Further studies are planned to
determine if the lack of detections in recent groundwater samples on site indicate migration
through deep soil gas from off-site sources or losses in the sampling and analysis process.
Chloroform attenuation is substantial between the area just above the water table and the 6-ft-
depth below the structure. Chloroform is also substantially attenuated between subslab air and
indoor air.
• PCE is apparently widely spatially distributed in site groundwater at concentrations well below
the current 5 ng/L MCL.2 The calculated volatilization from these shallow groundwater
concentrations matches observed deep soil gas concentrations. Only a moderate degree of
attenuation occurs in those deep soil concentrations as they are drawn toward the basement of the
structure. Substantial attenuation occurs in the upper 6 ft of the site external soil gas, which is
composed of finer grained materials than the soils. Substantial attenuation also occurs across the
building envelope between subslab and indoor air.
• The relative importance of the potential sources of PCE and chloroform—historic drycleaners,
the adjacent commercial/industrial quadraplex, and storm sewers/drinking water disinfection—is
unclear.
http://water.epa.gov/drink/contaminants/index.cfm
IT
-------
Section 2—Introduction
Table of Contents
2. Introduction 2-1
2.1 Background 2-2
2.1.1 Variability in Vapor Intrusion Studies 2-3
2.1.2 Vapor Attenuation Factors 2-7
2.1.3 Potential for Use of Radon as a Vapor Intrusion Tracer 2-8
2.1.4 Passive VOC Sampling 2-10
2.2 Objectives 2-14
2.2.1 Time Scale and Measurement of Independent and Dependent Variables 2-15
2.2.2 Data Quality Objectives and Criteria 2-16
List of Figures
2-1. An overview of important vapor intrusion pathways (U.S. EPA, 2002a) 2-2
2-2. Soil gas and groundwater concentrations below a slab (Schumacher et al., 2010) 2-6
List of Tables
2-1. VOC Indoor Air Sampling Method Options 2-13
2-2. Factors Causing Temporal Change in Vapor Intrusion and How They Are Observed and
Measured 2-15
2-3. Data Quality Objectives and Criteria 2-17
2-i
-------
Section 2—Introduction
2. Introduction
Vapor intrusion is the migration of subsurface vapors, including radon and volatile organic compounds
(VOCs), from the subsurface to indoor air. Vapor intrusion has emerged as a priority VOC exposure
pathway at many hazardous waste sites nationwide. Vapor intrusion occurs due to the pressure and
concentration differentials between indoor air and soil gas. Indoor environments are often negatively
pressurized with respect to outdoor air and soil gas, and this pressure difference allows soil gas with
subsurface vapors to flow into indoor air through advection. In addition, concentration differentials may
cause VOCs to migrate from areas of higher to lower concentrations through diffusion, which also causes
vapor intrusion.
The vapor intrusion exposure pathway extends from the contaminant source, which can be free product or
contaminated groundwater, to indoor air exposure points. Contaminated matrices, therefore, may include
groundwater, soil, soil gas, and indoor air. VOC contaminants of concern typically include halogenated
solvents such as trichloroethene (TCE), tetrachloroethylene (PCE), and chloroform, and degradation
products of TCE and PCE, including dichloroethylenes and vinyl chloride. Petroleum hydrocarbons, such
as the aromatic VOCs benzene, toluene, and xylenes, can also cause vapor intrusion. These contaminants
can be present in the vapor phase, dissolved phase, as a free phase (nonaqueous phase liquid or NAPL), or
in a sorbed phase on soil or aquifer materials. Vapor intrusion of halogenated VOCs has been identified as
an important exposure pathway at many contaminated sites, including Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA), Resource Conservation and Recovery Act
(RCRA), and Brownfield sites. Vapor intrusion also has occurred at leaking petroleum underground
storage tank (UST) sites but is less prevalent because petroleum hydrocarbons are biodegradable.
Radon is a colorless radioactive gas that is released by radioactive decay of radionuclides in soil. Radon
can migrate into homes through the vapor intrusion pathway in a similar fashion to VOCs. Radon is high
in areas where the radioactive precursors to radon occur at relatively high concentrations in soil (as with
the subject house of this investigation) and affects many more homes across the United States than
halogenated VOCs. Low-cost testing and effective mitigation methods are available for radon, and the
pathway has been studied extensively by EPA and other organizations.
This project focused on halogenated VOCs, which are relatively recalcitrant (resistant) to biodegradation
in aerobic soils and groundwater, and radon, which has a radioactive half-life of about 3.8 days. Of the
predominant two VOCs found in this house (chloroform and PCE), PCE is generally considered quite
recalcitrant, with an aerobic half-life in groundwater of 1 to 2 years (Howard et al., 1991). Studies of
chloroform biodegradation under aerobic conditions are mixed, with some showing recalcitrance (e.g., a
0.2 to 5-year half-life in Howard et al., 1991) and others showing moderate cometabolic biodegradation
with methylene chloride and chloromethane as sequential degradation products (AFCEE, 2004; ATSDR,
1997).
Current practice for evaluating the vapor intrusion pathway involves a combination of mathematical
modeling and direct measurements in groundwater, external soil gas, subslab soil gas, and indoor air. No
single line of evidence is considered definitive, and direct measurements can be costly, especially where
significant spatial and temporal variability require repeated measurements at multiple locations to
accurately assess the chronic risks of long-term VOC exposure. The main focus of this work assignment
is to better characterize this variability by collecting a full year's dataset of weekly measurements of
subslab soil gas, external soil gas, and indoor air, on a single building that is impacted by vapor intrusion
of radon and VOCs. By examining both short-term and long-term (average annual) concentrations, the
project provides valuable information on how to best take and evaluate measurements to estimate long-
term, chronic risk for VOCs. We further elucidate the relationship between radon and VOC vapor
2-1
-------
Section 2—Introduction
intrusion in a house affected by both. The radon literature could provide valuable lessons for VOC vapor
intrusion if there is a relationship, and radon, being much cheaper to measure than VOCs, could be an
important tool in improving the investigation and mitigation of VOC vapor intrusion.
2.1 Background
An overview of the VOC vapor intrusion pathway is shown in Figure 2-1; the building in which exposure
occurs is shown in the center. Three main routes of VOC migration have been defined:
1. Movement of VOC vapors from shallow soil sources through the unsaturated (vadose) zone
2. Transport of VOCs through groundwater, followed by partitioning of VOCs from the most
shallow layer of groundwater into vadose zone soil gas
3. Vapor movement through preferential pathways such as utility corridors
Figure 2-1. An overview of important vapor intrusion pathways (U.S. EPA, 2002a).
In portions of these three routes, advective forces predominate and in others diffusive forces dominate
transport.
The final step of vapor intrusion typically involves soil gas moving from immediately below the building
slab into the indoor air. This subslab space is often significantly more permeable than the bulk vadose
zone soil, either because a gravel drainage layer was intentionally used or the soils have shrunk back from
the slab in places. In those cases, the subslab space is expected to serve as a common plenum allowing the
lateral mixing of VOCs that reach the building through multiple pathways. In other cases, the subslab
space may not be so interconnected, resulting in differing subslab VOC concentrations at different
locations across the slab.
2-2
-------
Section 2—Introduction
Vapor and liquid transport processes and their interactions with various geologic and physical site settings
(including building construction and design), under given meteorological conditions, control migration
through the vapor intrusion pathway. Variations in building design; construction, use, and maintenance;
site-specific stratigraphy; subslab composition; temporal variation in atmospheric pressure; temperature;
precipitation; infiltration; soil moisture; water table elevation; and other factors combine to create a
complex and dynamic system. Important factors controlling vapor intrusion at many sites include (NJ
DEP, 2005):
• Biodegradation of VOCs as they migrate in the vadose zone,
• Site stratigraphy,
• Soil moisture and groundwater recharge,
• Fluctuations in water table elevation, and
• Temporal and inter-building variations in the operation of ventilation systems in
commercial/industrial buildings.
The U.S. Environmental Protection Agency (U.S. EPA) Office of Research and Development (ORD) has
developed a subslab sampling protocol (U.S. EPA, 2005a). The available information to date indicates
that properly placed subslab sampling can significantly increase the reliability of vapor intrusion
estimates and environmental decisions. One important reason is that assessment of subslab VOCs allows
the investigator to forensically determine the contributions to indoor air from the soil gas immediately
underlying the slab. However, one disadvantage of subslab sampling is its intrusiveness to the occupants
of the building. In addition, if a subslab probe is not properly installed, it could provide a preferential
route for contaminant migration. Also, if the space beneath a slab is not interconnected and well mixed,
individual samples may not give an accurate picture of VOC concentrations beneath the slab.
This project explored and further developed several promising cost-effective techniques to evaluate the
vapor intrusion pathway and improve data quality. Two primary tools were investigated: (1) using
modified sorbent-based measurement techniques for time-integrated measurements of indoor air VOCs
and (2) using radon as a tracer for assessing VOC vapor intrusion. The project also investigated
measurements of pressure differentials (subslab vs. indoor), meteorological conditions, crack size, and air
exchange rates in the context of the chemical-specific measurements described above. These physical
measurements are not stand-alone tools nor are they the emphases of the current research program, but are
necessary supporting tools for developing a conceptual understanding of spatial variability, temporal
seasonal effects, and a mass balance around a building subject to vapor intrusion.
2.1.1 Variability in Vapor Intrusion Studies
This project focuses on observing changes in vapor intrusion over a 2-year period. In order to express
quantitatively our goals for this project, it is necessary to understand the causes and typical ranges of
spatial and temporal variation in various matrices studied for vapor intrusion assessment.
Through measurements of radon and VOC vapor intrusion under various conditions, several studies have
provided insight into the complexity of temporal variability in indoor air concentrations attributable to
vapor intrusion—the primary focus of this work. Nazaroff et al. (1987) studied how induced-pressure
variations can influence radon transport from soil into buildings with roughly hourly resolution. In a more
recent study, Mosley (2007) presented the results of experiments, showing that induced building-pressure
variations influence both the temporal and spatial variability of both radon and chlorinated VOCs in
subslab samples and in indoor air (hourly sampling for radon). Schuver and Mosley (2009) have also
reviewed numerous studies of radon indoor concentrations, in which multiple repeated indoor air samples
were collected with hourly, daily, weekly, monthly, 3-month, and annual sample durations for study
2-3
-------
Section 2—Introduction
periods of up to 3 years; however, soil gas datasets with such detailed measurements of both radon and
VOCs are rare.
Several radon studies have demonstrated that barometric pressure fluctuations can affect the transport of
soil gas into buildings (Robinson and Sextro, 1997; Robinson et al., 1997). The impact of barometric
pressure fluctuations on indoor air is influenced by the interaction of the building structures and
conditions, as well as other concurrent factors, such as wind (Luo et al., 2006, 2009). Changes in
atmospheric conditions (e.g., pressure, wind) and building conditions (e.g., open doors and windows) may
temporarily over- or under-pressurize a building. Based on long-term pressure differential datasets
acquired by ARCADIS and EPA's National Risk Management Research Laboratory (NRMRL) at an
Indianapolis study site at which both radon and VOCs are being measured in both subslab and indoor air,
other factors that may cause temporal and spatial variability in soil vapor and indoor air concentrations
include
• fluctuation in building air exchange rates due to resident behavior/HVAC operations,
• fluctuations in outdoor/indoor temperature difference, and
• rainfall events and resultant infiltration and fluctuations in the water table elevation.
The pressure difference between a house-sized building and the surrounding soil is usually most
significant within 1 m to 2 m of the structure, but measurable effects have been reported up to 5 m from
the structure (Nazaroff et al., 1987). Temperature differences or unbalanced mechanical ventilation are
likely to induce a symmetrical pressure distribution in the subsurface, but the wind load on a building
adds an asymmetrical component to the pressure and distribution of contaminants in soil gas.
Folkes et al. (2009) summarize several large groundwater, subslab, and indoor air datasets collected, with
sampling frequencies ranging from quarterly to annually during investigations of vapor intrusion from
chlorinated VOC plumes beneath hundreds of homes in Colorado and New York. They analyzed these
datasets to illustrate the temporal and spatial distributions in the concentration of VOCs. In a study of the
vapor intrusion pathway at the Raymark Superfund site, DiGiulio et al. (2006) showed that measured
concentrations of chlorinated VOCs in subslab exhibited spatial and temporal variability between
neighboring houses and within individual houses. Similar variability in subslab chlorinated VOC
concentrations within and between houses has been observed during vapor intrusion evaluations of
several sites in New York State (Wertz and Festa, 2007).
In scenarios with coarser soils (e.g., sands, gravels), the soil gas permeability is high, and changes in
building pressurization may affect the airflow field and the resultant soil vapor concentration profiles near
buildings. In scenarios with fine-grained soils (e.g., silts, clays), the soil gas permeability is low and soil
gas flow rates (Qs) may be negligible and not affect the subsurface concentration. Nevertheless, in both
soil-type scenarios, over-pressurization of the building may still significantly reduce the indoor air
concentration due to the reversal of soil gas flow direction from the building into the soil (Abreu and
Johnson, 2005, 2006).
A wind-induced, non-uniform pressure distribution on the ground surface on either side of a house may
cause spatial and temporal variability in the subslab soil vapor concentration distribution if the wind is
strong and the soil gas permeability is high (Luo et al., 2006, 2009). In addition, during or after a rainfall
event, the subsurface beneath the building may have a lower moisture content than the adjacent areas due
to water infiltration.
2-4
-------
Section 2—Introduction
Spatial Variability
Spatially, reports of several orders of magnitude variability without apparent patterns between indoor air
and subslab concentrations for adjacent structures in a neighborhood are very common (see, for example,
Dawson, 2008). Six orders of magnitude in subslab concentration variability were reported by Eklund and
Burrows (2009) for one building of 8,290 sq ft. As shown in Figure 2-2, Schumacher and coworkers
observed more than three orders of magnitude concentration variability in shallow soil gas below a slab
over 50 lateral ft (Schumacher et al., 2010), suggesting a strong effect of impervious surfaces both in
limiting soil gas exchange with the atmosphere and in maintaining relatively high concentrations of
VOCs in shallow groundwater. Schumacher and others (2010) also observed two orders of magnitude
concentration variability with a depth change of 10 ft in the unsaturated zone within one borehole. Lee
and others (2010) observed two orders of magnitude variability in subslab concentration within a small
townhouse. Studies by McHugh and others (2007) have generally found markedly less variability in
indoor air concentrations than in subslab concentrations, probably due to the greater degree of mixing in
the indoor environment.
Temporal Variability
Temporal variability has been summarized by ITRC (2007), which states in Section D.4.10:
IBM, Endicott, New York
Recent data from a large site in
Endicott, New York, collected over a
15-month period showed soil gas
concentration variations of less than
a factor of 2 at depths greater than 5
ftbgs.
Variations in soil gas concentrations due to temporal effects
are principally due to temperature changes, precipitation, and
activities within any overlying structure. Variations are
greater in samples taken close to the surface and dampen with
increasing depth. In 2006 there were a number of studies on
temporal variation in soil gas concentrations, and more are
under way or planned in 2007 by USEPA and independent
groups. To date these studies have shown that short-term
variations in soil gas concentrations at depths 4 feet or deeper are less than a factor of 2 and that
seasonal variations in colder climates are less than a factor of 5 (Hartman 2006). Larger variations
may be expected in areas of greater temperature variation and during heavy periods of precipitation,
as described below.
• Temperature. Effects on soil gas concentrations due to actual changes in the vadose zone
temperature are minimal. The bigger effect is due to changes in an overlying heating or
HVAC system and the ventilation of the structure due to open doors and windows. In colder
climates, worse-case scenarios are most likely in the winter season. The radon literature
suggests that temporal variations in soil gas are typically less than a factor of 2 and that
seasonal effects are less than a factor of 5. If soil gas values are more than a factor of 5
below acceptable levels, repeated sampling is likely not necessary regardless of the season.
If the measured values are within a factor of 5 of allowable risk levels, then repeated
sampling may be appropriate.
2-5
-------
Section 2—Introduction
v>EPA
Uniteu Sla
Environmental Protection
Agsncy
Measured Soil Gas
Profile for TCE - Phase 2
S1-1 S'-a ST-2 ST-3 S;-8 M-4
2
4
i—
UJ
4,10
57, DOC |
1
- 6
.
a
10
0
|
SUB
! 6601
| 9,150,
1 SS.OC'd 1
1
|
1 2901
| 6.700 1
tl
23,000'
| 30.QjO|
|52,000i
I
1 331
1 "1
1 n. sod
| 2,600|
1 ?'7r'~l
j//y/i
\
i i
i i
i i
i i
i4" i
193 1
,230 |
,83 ,
|ND |
IZD '
|34 |
!8SO 1
^^NrA%nKA^
|ND ,
INB 1
1 i
,SD |
V
W///////A
,ND
IHB
ND
1
• 50 121 :--! •.< i'l * ' •--• " -. )
10 20 30 40 50 60 70 SO 9(
FEET
Figure 2-2. Soil gas and groundwater concentrations below a slab (Schumacher et al., 2010).
• Precipitation. Infiltration from rainfall can potentially Impact soil gas concentrations by
displacing the soil gas, dissolving VOCs, and by creating a "cap " above the soil gas. In
many settings, infiltration from large storms penetrates into only the uppermost vadose zone.
In general, soil gas samples collected at depths greater than about 3-5 feet bgs or under
foundations or areas with surface cover are unlikely to be significantly affected. Soil gas
samples collected closer to the surface (<3feet) with no surface cover may be affected. If the
moisture has penetrated to the sampling zone, it typically can be recognized by difficulty in
collecting soil gas samples. If high vacuum readings are encountered when collecting a
sample or drops of moisture are evident in the sampling system or sample, measured values
should be considered as minimum values.
• Barometric Pressure. Barometric pressure variations are unlikely to have a significant effect
on soil gas concentrations at depths exceeding 3-5 feet bgs unless a major storm front is
passing by. A recent study in Wyoming (Luo et al. 2006) has shown little to no relationship
between barometric pressure and soil gas oxygen concentrations for a site with a water
table at ~15 feet bgs.
In summary, temporal variations in soil gas concentrations, even for northern climates, are minor
compared with the conservative nature of the risk-based screening levels. If soil gas values are a
factor of 5-10 times below the risk-based screening levels, there likely is no need to do repeated
sampling unless a major change in conditions occurs at the site (e.g., elevated water table, significant
seasonal change in rainfall)
2-6
-------
Section 2—Introduction
And in Section D.8 of the same document ITRC notes:
Short-term temporal variability in subsurface vapor intrusion occurs in response to changes in
weather conditions (temperature, wind, barometric pressure, etc.), and the variability in indoor air
samples generally decreases as the duration of the sample increases because the influences tend to
average out over longer intervals. Published information on temporal variability in indoor air quality
shows concentrations with a range of a factor of 2-5 for 24-hour samples (Kuehster, Folkes, and
Wannamaker 2004; McAlary et al. 2002). If grab samples are used to assess indoor air quality, a
factor of safety (at least a factor of 5) should be used to adjust for short-term fluctuations before
comparing the results to risk-based target concentrations. Long-term integrated average samples (up
to several days) are technically feasible, using a slower flow rate this is the USEPA recommended
approach for radon monitoring). Indoor air sampling during unusual weather conditions should
generally be avoided.
In Section D. 11.8, ITRC goes on to discuss the effect of meteorological changes on vapor intrusion:
A variety of weather conditions can influence soil gas or indoor air concentrations. The radon
literature suggests that temporal variations in the soil gas are typically less than a factor of 2 during
a season and less than a factor of 5 from season to season). . . Forensic approaches were used at the
Redfield Rifles site in Colorado to determine whether the source ofsubslab contaminants was in the
vadose zone or the overlying structure (McHugh, De Blanc, and Pokluda 2006). D-28 Endicott, New
York and Casper, Wyoming are in agreement with the radon results. For soil gas, the importance of
these variables will be greater the closer the samples are to the surface and are unlikely to be
important at depths greater than 3-5 feet below the surface or structure foundation.
Measurement Variability
Beyond spatial and temporal variability, the underlying uncertainty of the measurements used to assess
vapor intrusion must also be considered. Many measurements of vapor intrusion, both in indoor air and
subslab soil gas have traditionally relied on Summa canister samples analyzed by methods TO-14/TO-15
(U.S. EPA, 1996, 1999). Method TO-15 specifies an audit accuracy of 30% and a replicate precision of
25% as performance criteria. But even those figures do not fully convey the interlaboratory variability
observed for these methods when applied to the low concentrations typical of indoor air studies. As Lutes
and coworkers (2010) reported:
• In two recent TO-15 or 8260 interlaboratory comparisons administered by the company ERA for
gas phase samples the acceptance range for tetrachloroethylene results were:
- 4.31-22.3 ppbv (July-Sept 2009 study)
- 31.6-74.1 ^ig/L (October-November 2007 study)
• For comparison in a 2007 TO-14/TO-15 study conducted by Scott Specialty Gasses, the reported
values for toluene reported by 12 labs varied from 3.1 to 18.6 ppb.
2.1.2 Vapor Attenuation Factors
One common way of evaluating the impact of subsurface vapors on indoor air quality is to compute the
ratio of indoor air concentration to subslab soil vapor concentration. EPA has defined the resulting
"attenuation factor" as follows: "The attenuation factor, a, is a proportionality constant relating indoor air
concentrations (Cmdoorair) to the concentrations of vapors in soil gas (CSOiigas) or groundwater (Cgr0undwater)
concentrations." For soil gas to indoor air, the equation is as follows:
^indoor air ™SG ^soil gas •
2-7
-------
Section 2—Introduction
For groundwater, a similar equation is used, except that the dimensionless Henry's Law Constant (H) is
used to convert the dissolved VOC concentration in groundwater to the corresponding equilibrium vapor
concentration:
^indoor air ~~ &GW ^groundwater -H--
A larger a indicates less attenuation and a smaller value indicates more attenuation. The greater the
attenuation factor, the greater the indoor concentration.
Within any one given site, the attenuation factors
• between groundwater and indoor air typically vary 2 to 3 orders of magnitude and
• between external soil gas and indoor air typically vary 2 to 4 orders of magnitude.
Subslab soil gas and indoor air typically vary 2 to 4 orders of magnitude (Dawson and Schuver, 2010).
2.1.3 Potential for Use of Radon as a Vapor Intrusion Tracer
Radon, a naturally occurring radioactive gas, is a potentially useful surrogate for assessing VOC vapor
intrusion because the physics of radon intrusion into indoor air is nearly identical to VOC vapor intrusion.
Radon is ubiquitous in the soil and present at measurable quantities throughout the United States. Indeed,
much of the research in VOC vapor intrusion is an expansion of earlier work on radon intrusion. It is less
expensive to measure radon than VOCs, and the radon measurements could be a useful screening tool to
target buildings for additional vapor intrusion assessment.
Radon provides a nearly unique tracer for vapor intrusion because its presence in the indoor environment
is usually a result of radon in the soil gas immediately surrounding a building. In general, the entry
mechanisms are believed to be the same for VOCs and radon in soil gas. Thus, measured radon entry rates
should be a good predictor of relative entry rates for VOCs. The advantages of using radon as a tracer for
vapor intrusion characterization include:
• Measurements of radon are easier, more accurate and precise, and much less expensive than
canister measurements of VOCs (typically less than 10% of the VOC analysis cost). Passive
indoor sampling for radon costs approximately $5 to $20 per sample. Active radon sampling
(indoor air and subslab) uses some of the same equipment and setup as for VOCs. This minimizes
sampling times and cost.
• High levels of indoor radon identify buildings as vulnerable to soil gas entry.
• Because of the low sampling/analytical costs, it is possible to increase the number of field
measurements. This, in turn, increases confidence in the field evaluation.
• Because mitigation systems are the same for radon and VOCs, and because radon and VOCs
behave similarly in the vicinity of the building, radon measurements before and after installation
of vapor intrusion mitigation systems may be useful for assessing mitigation system performance
for VOCs as well.
In summary, the limited data gathered to date suggest that radon measurement may be an inexpensive,
reliable surrogate for VOC measurement when characterizing vapor intrusion and may significantly
enhance vapor intrusion characterization and decision making, particularly when used in conjunction with
subslab sampling. However, several key aspects and assumptions of this approach need to be verified
before it can be put into widespread use.
2-8
-------
Section 2—Introduction
For radon to be a valuable tracer:
• Radon detection in building interiors should be quantitatively possible across the wide range of
subslab concentrations encountered in the United States. Ideally these measurements can be made
with inexpensive passive methods (i.e., charcoal or electrets).
• Radon route and mechanism of entry should be similar to that of VOCs of interest, once both
species are present in the subslab soil gas. This would imply that the subslab attenuation factors
for radon and VOCs are similar.
• Variance in the natural vadose zone (unsaturated soil) radon concentration across a given building
footprint should be low enough to allow radon to be a useful indicator.
• Concentrations of radon and the VOCs of concern should be well correlated in subslab soil gas.
This would not necessarily be expected based on the fact that radon and VOCs have different
sources. However, they may indeed be approximately correlated if the VOC(s) of interest and
radon are both widely dispersed in deep soil gas. In this case, the concentrations of both radon
and VOCs at various locations in the subslab may be controlled primarily by the ratio of flow
from the deep soil gas to the flow from ambient air (in which both VOC and radon concentrations
would be expected to be low).
• Interior sources of radon should be negligible.
The loss rates to sink effects in the indoor environment should be similar or negligible for radon and
VOCs so that the air exchange rate forms the primary control on indoor air concentration once vapor
intrusion has occurred.
To our knowledge, this concept was first applied in a relatively small study (Cody et al., 2003) at the
Raymark Superfund Site in Connecticut. The study compared the intrusion behavior of radon and
individual VOCs by determining attenuation factors between the subslab and indoor (basement) air in 11
houses. The results indicated that the use of radon measurements in the subslab and basement areas was
promising as a conservative predictor of indoor VOC concentrations when the subslab VOC
concentrations were known. Further work at the Raymark site (U.S. EPA, 2005b) statistically compared
basement and subslab concentration ratios for radon and VOCs associated with vapor intrusion. Of six
test locations, three showed that basement/subslab concentration ratios for radon and VOCs associated
with subsurface contamination were similar. Three had statistically different ratios, suggesting that further
research was needed to evaluate the usefulness of radon in evaluating vapor intrusion. Conservative
VOCs (those believed to be associated only with subsurface contamination) were a better predictor of
other individual volatile compounds associated with vapor intrusion than was radon.
A three-building complex, commercial case study of the radon tracer approach was published by Wisbeck
et al. (2006). Radon and indoor air attenuation factors were calculated for five sampling points and were
generally well correlated. Subslab radon concentrations varied by approximately a factor of 10 across the
five sampling points.
Results of an earlier test program at Orion Park Housing units at Moffett Field have been preliminarily
reported (Mosley, 2007). Results showed:
• Low levels of radon can be measured with sufficient accuracy to be used in analysis of vapor
intrusion problems.
• Radon is a promising, low-cost surrogate for soil gas contaminants; however, as with VOCs
themselves, the complete distribution under the slab must be known in order to properly interpret
its impact on indoor measurements.
2-9
-------
Section 2—Introduction
• Unexpectedly, the subslab areas under each unit were segmented. The four subslab sampling
points installed in one unit were not in good communication with one another. An introduced
tracer, SF6, moved very slowly and not very uniformly under the slab.
• Results showed that for sites where subslab soil conditions lead to interrupted flow and poor
communication beneath the slab, a subslab measurement at a single point is not very reliable for
estimating potential vapor intrusion problems. The average value of subslab measurements at
several locations also may not yield a reliable estimate of indoor concentrations. When subslab
communication is poor, one must identify a connection between subslab contaminants and a
viable entry path.
The potential usefulness of the radon tracer was studied in 2007-2010 by EPA NRMRL at Moffett Field
in California and in the Wheeler building in Indianapolis. These studies are summarized in three draft
peer-reviewed papers that have been submitted for EPA internal review:
• Vapor Intrusion Evaluation Using Radon as a Naturally Occurring Tracer. In this paper we
compile data from five study sites where radon has been used in VOC vapor intrusion
investigations and attenuation factors were calculated. A total of 17 buildings are included in the
dataset, a mix of commercial and residential, in a wide variety of geographical areas within the
United States. Some correlation between radon and VOC attenuation factors was seen, but it was
not perfect.
• Randomized Experiment on Radon Tracer Screening for Vapor Intrusion in a Renovated
Historical Building Complex. This study focused on a renovated former industrial facility now
being reused as residential, public, and office space. Fifty locations within the complex were
originally screened for radon using passive sampling techniques. Two subsets of these sample
locations were selected for passive VOC sampling, one randomly and the other based on the
radon information. The upstairs radon-guided samples were significantly higher in TCE than the
randomly selected locations. The portions of the building complex where the radon guidance
appeared to provide predictive power were understandable in terms of the building design and the
concept of the open basement serving as a common plenum.
• Case Study: Using Multiple Lines of Evidence to Distinguish Indoor and Vapor Intrusion Sources
in a Historic Building : This paper uses datasets developed at the Southeast Neighborhood
Development Corporation (SEND) Wheeler Arts Building site in Indianapolis, Indiana, to
demonstrate the use of multiple lines of evidence in distinguishing indoor from subsurface
sources in a complex multiuse, multiunit building. The use of radon as a quantitative tracer for
vapor intrusion source discrimination is shown as well as the use of differential pressure data as
an additional line of evidence. Box and whisker plots of the distribution of indoor air pollutants
on multiple floors are used to distinguish pollutants with predominant subslab sources from those
with predominant indoor sources. Those pollutants, which the box and whisker analysis suggests
have indoor sources, are also corroborated from the literature as having very common indoor
sources expected in this building, including arts and crafts activities, human exhalation, consumer
products, and tobacco smoking.
2.1.4 Passive VOC Sampling
Sorbent-based methods are an emerging technology for vapor intrusion assessment. Current standard
practices for indoor air VOC monitoring in the United States include the use of ultra-clean, passivated,
and evacuated (i.e., under a negative pressure) stainless steel canisters for sample collection. Practitioners
frequently use 8- to 48-hour integrated samples with Summa canisters in an attempt to average over an
exposure period. This is the U.S. "gold standard" for indoor air analysis, but it is expensive to implement.
Professional experience, including that gathered under WA 4-46 but not yet published, shows that the
2-10
-------
Section 2—Introduction
flow controllers currently used in commercial practice are subject to substantial flow rate and final
pressure errors when set for integration times in excess of 24 hours (Hayes, 2008).
Active and passive sorbent sampling techniques are already in use in the United States for personal air
monitoring for industrial workers and are outlined in both OSHA Sampling and Analytical Methods
(http://www.osha.gov/dts/sltc/methods/toc.html) and NIOSH Manual of Analytical Methods
(http://www.cdc.gov/niosh/nmam/). Typical sampling scenarios involve the collection of active or passive
samples to monitor a single chemical used in the workplace over a period of up to 10 hours. These
methods are designed to meet OSHA PELs, which are typically in the parts per million (ppm) range and
consequently several orders of magnitude higher than risk-based indoor air screening levels and not
suitable for ambient air measurements without modification.
Active sorbent methods (e.g., TO-17) have also been published by EPA for VOC measurements in
ambient air (U.S. EPA, 1999). However, in those methods, air samples are normally actively collected
over 1 hour, using a sample pump with a sampling rate of 16.7 mL/min to 66.7 mL/min, yielding total
sample volumes between 1 and 4 L. Sampling intervals can be extended beyond 1 hour; however, care
must be taken to ensure breakthrough volumes are not exceeded in order to quantitatively retain the
compounds of interest on the sorbent tube. Given the minimum pump flow rate cited in TO-17 of 10
mL/min, the practical upper limit for chlorinated VOCs using a multi-bed thermal desorption sorbent tube
is on the order of 10 L (Marotta et al., 2012) up to 20 L for select VOCs yielding a corresponding
maximum collection period of 8 to 24 hours.
One way to lower the detection limits and control day-to-day variability is to sample over a longer period
of time. Recent studies have shown that it may be feasible to use passive sorbent samplers to collect a
continuous indoor air sample over several weeks. This approach would provide a lower detection limit, be
cost-effective, and result in a time-integrated composite sample. Laboratory and field evaluations of such
an approach for ambient and indoor air applications have been published and showed promising results
for sampling durations of up to 14 days. Exposure of badge-type charcoal passive samplers to controlled
atmospheres of 10 ppb to 200 ppb benzene, toluene, and m-xylene showed good performance when
deployed for 14 days (Oury et al., 2006). A field study published by Begerow and others (1999) showed
comparability between two charcoal-based passive sampler geometries, badge and tube-style for 4-week
indoor and outdoor air samples. Field evaluations were also conducted using radial charcoal and thermal
desorption Radiello® samplers to determine performance over a 14-day period. Ambient BTEX
measurements using the Radiello samplers compared well to active sorbent sampling results (Cocheo et
al., 2009).
During testing at Orion Park, Moffett Field in California by EPA NRMRL APPCD, EPA Region IX, and
ARCADIS compared measurements of VOCs by Method TO-15 to three different sorbent systems:
1. Radial: Activated Charcoal (with CS2 extraction: GC/MS)
2. Radial: Carbograph 4 (TO-17: Thermal Desorption [TD] GC/MS)
3. Axial: Chromosorb 106 thermal desorption tube (TO-17: TD GC/MS)
Testing was also performed at the Wheeler site in Indianapolis comparing Summa canisters to Radiello
solvent-extracted samplers. Across the two sites, the Radiello solvent extracted showed good agreement
to TO-15 and precision at both sites for chlorinated compounds. Agreement was poor for polar
compounds: ethanol, MEK, MIBK, and acetone. Radiello thermal desorption correlated well with Summa
TO-15 but gave noticeably lower concentrations, suggesting that 2 weeks is too long an integration time
for these samplers. The agreement of the axial (tube) method was inferior (Mosley, 2008; Lutes, 2010).
2-11
-------
Section 2—Introduction
Table 2-1 compares the characteristics of commercially available passive sampler geometries and
available sorbent configurations. The geometry of the sampler (radial, badge, or tube) largely determines
the sampling rate or uptake rate, with the radial design resulting in the highest sampling rate and the tube-
style the lowest sampling rate. The permeation sampler relies on permeation of the vapor-phase
compound through the polydimethylsiloxane (PDMS) membrane and adsorption to the sorbent bed
behind the membrane. The greater the sampling rate, the greater the mass of VOCs adsorbed onto the
sorbent bed. In addition to the passive geometries available, sorbent pairings fall into two main
categories—charcoal based and thermally desorbable. Charcoal-based materials are characterized as very
strong sorbents with a large surface area and a corresponding high adsorption capacity. To efficiently
extract adsorbed compounds for measurement in the laboratory, an aggressive solvent extraction is
required. The thermally desorbable sorbents are generally much weaker than charcoal with a smaller
surface area, allowing for analysis of the adsorbed compounds through thermal extraction. As Table 2-1
shows, when comparing the same passive geometry, the thermally desorbed model provides the lowest
detection limits, while the charcoal-based solvent-extracted system allows for longer sampling times as
well as a greater dynamic range because the high capacity of the charcoal minimizes sorbent saturation
under conditions of high analyte or background matrix.
European agencies have developed standard methods for passive sampling for VOCs that are applicable
to the range of concentrations and durations to be tested in this project:
• Methods for the Determination of Hazardous Substances (MDHS) 88: Volatile Organic
Compounds in Air: Laboratory Method Using Diffusive Samplers, Solvent Desorption and Gas
Chromatography, December 1997. Published by the Health and Safety Executive of the United
Kingdom: http://www.hse.gov.uk/index.htm.
• Methods for the Determination of Hazardous Substances (MDHS) 80: Volatile Organic
Compounds in Air: Laboratory Method Using Diffusive Solid Sorbent Tubes, Thermal Desorption
and Gas Chromatography, August 1995. Published by the Health and Safety Executive of the
United Kingdom: http://www.hse.gov.uk/index.htm
• Ambient air quality—Standard method for measurement of benzene concentrations - Part 4:
Diffusive sampling followed by thermal desorption and gas chromatography, EN 14662-4:2005.
Published by the European Committee for Standardization.
• Ambient air quality—Standard method for measurement of benzene concentrations - Part 5:
Diffusive sampling followed by solvent desorption and gas chromatography, EN 14662-5:2005.
Published by the European Committee for Standardization. (Also published as the British
Standard BS EN 14662-5:2005)
• Indoor air quality—Diffusive samplers for the determination of concentrations of gases and
vapors—Guide for selection, use, and maintenance, EN 14412:2004. Published by the European
Committee for Standardization.
Given the wide range of sampling durations required for this project, several diffusive sampler
configurations are recommended to meet anticipated project objectives for indoor air measurements. For
short-term samples (less than 7 days), the sampler must have sufficient sensitivity to measure the low
VOC concentrations that are expected in the indoor air. Thermally desorbable sorbents paired with a
badge or radial-style geometry can effectively be used for the 24-hour samples and yield low reporting
limits. The badge style is recommended over the radial style given the larger number of chlorinated
compounds for which sampling rates have been generated and validated. For durations of greater than 7
days, stronger sorbents with higher adsorptive capacity are recommended, which require solvent
extraction. Although the solvent extraction is less sensitive than thermal desorption, the high sampling
rate of the radial sampler geometry over durations of 7 to 30 days will result in sample reporting limits
essentially equivalent or lower than those generated using the thermal desorption technique.
2-12
-------
Section 2—Introduction
Table 2-1. VOC Indoor Air Sampling Method Options
Parameter
Collection media
Ease of
deployment
Media and
shipping cost
Method and
analysis
Estimated
analytical reporting
limit
Expected
sampling rate
Recommended
sampling duration
Estimated sample
reporting limits3
Applicable range
of chlorinated
solvents (based
on available
sampling rates)
Whole Air
Summa Canister
(TO-15)
Good
high
TO-15GC/MS
0.05-0.1 ug/m3
0.5-3.5 mL/min
Typically 24
hours
~0.05(SIM)-0.1
ug/m3
TCE/PCE and all
breakdown
products
including vinyl
chloride (VC)
Sorbent-
Active
Multi-bed
AID sorbent
tubes (TO-
17)
Good
medium
TO-17
GC/MS
1-10 ng
10-200
mL/min
8-24 hours
-0.1-1 ug/m3
TCE/PCE
and all
breakdown
products
including VC
Sorbent-Diffusive
Radial:
Charcoal
(Radiello
130)
Excellent
low
Solvent
Extraction
GC/MS or
GC/FID
1 00-200 ng
-60 mL/min
Up to 30
days
-0.1-0.4
ug/m3
TCE, PCE,
111-TCA,
chloroform
Radial: TD
sorbent
(Radiello 145)
Excellent
high
TO-17
GC/MS
1-1 Ong
-25 mL/min
Up to 7 days
for chlorinated
solvents
-0.005-0.05
ug/m3
TCE, PCE,
111-TCA
Badge: Charcoal
type
(SKC 575, 3M
OVM3500)
Excellent
low
Solvent Extraction
GC/MS or
GC/FID
75-200 ng
-10 mL/min SKC
-30 mL/min 3M
Up to 4 weeks
-0.25-2 ug/m3
Validated for a
wide range of
chlorinated
solvents for 8
hours, several for
up to 30 days.
Badge: TD
sorbents selected
by deployment
time: (SKC Ultra I,
II, III)
Excellent
medium
TO-17 GC/MS
1-10 ng
-10 mL/min
1-7 days
-0.01-0.1 ug/m3
TCE, PCE, DCE,
111-TCA,
chloroform, 12-
DCA, cis-12-DCE,
trans-12-DCE, 11-
DCA.
Tube: TD
sorbents (e.g.,
Chromosorb 106)
Excellent
medium
TO-17 GC/MS
1-10 ng
-0.5 mL/min
In general, up to
4 weeks).
-0.2-2 ug/m3
TCE, PCE, 111-
TCA
Permeation:
Charcoal
type
(WMS™)
Excellent
low
Solvent
Extraction
GC/MS
50-200 ng
-0.5- 5
mL/min
Up to 30
days
-1-40 ug/m3
TCE, PCE,
and most
breakdown
products
Permeation:
TD sorbent
(WMS™)
Excellent
low
TO-17
GC/MS
1-1 Ong
-0.5- 5
ml_/min
Up to 30
days
-1-40 ug/m3
TCE, PCE,
and most
breakdown
products
Normalized to a 7-day period for diffusive samplers.
2-13
-------
Section 2—Introduction
Very few studies have evaluated VOC measurements using diffusive samplers beyond 30 days, and
determining if this is possible is one objective of this study. The sorbent selection, the sampler geometry,
and the target chemical's volatility all may have a significant impact on the successful application of
diffusive samplers to extended deployment periods. The few published studies evaluating sampling
intervals greater than 30 days are largely focused on the measurement of BTEX (Bertoni et al., 2001;
Brown and Crump, 1993), and the stability of chlorinated compounds on sorbents in the presence of
humidity and the variability of the sampling rate past 30 days are not well understood for any of the
diffusive samplers under consideration for this study.
Given the previous studies and the existence of standard methods for this application in Europe, the 1-
and 2-week Radiello passive samplers for VOCs are considered sufficiently accurate and precise to be the
primary VOC measurement tool in this project and are used as a basis of comparison for longer duration
samples.
2.2 Objectives
The main goal of this project is to investigate distributional changes in VOC and radon concentrations in
the indoor air, subslab, and subsurface soil gas from an underground source (groundwater source and/or
vadose zone source) adjacent to a residence or small commercial building. The time frame of this study is
over a year (about 14 months) in order to evaluate the effects due to seasonal variations on radon and
VOC vapor intrusion.
There are four primary objectives for this research effort:
1. Identify any seasonal fluxes in radon and VOC concentrations as they relate to a typical use of
HVAC in the building.
2. Establish the relationship between subslab/subsurface soil gas and indoor air concentrations of
VOCs and possibly radon.
3. Determine the relationship of radon to VOC concentrations at a given site.
4. Examine if near-building external samples could be used as a surrogate sampling location.
Five secondary objectives have also been defined:
1. Evaluate the duration over which solvent-extracted passive samplers provide useful integration of
indoor air concentrations (i.e., over what duration is the uptake rate constant?).
2. Characterize the near-building environment sufficiently to explain the observed variation of
VOCs and radon in indoor air.
3. Determine whether the observed changes in indoor air concentration of volatile organics of
interest can be mechanistically attributed to changes in vapor intrusion.
4. Confirm that the two analytical laboratories (Air Toxics and US EPA) can produce soil gas VOC
data with sufficient agreement, that the variance between laboratories is not significant compared
with the variance between laboratories or the changes in the underlying phenomena being
observed.
5. Evaluate the extent to which groundwater concentrations and/or vadose zone sources control soil
gas and indoor air VOC concentrations at this site.
Characteristics of the experimental design and data quality objectives developed to meet these objectives
are described below.
2-14
-------
Section 2—Introduction
2.2.1 Time Scale and Measurement of Independent and Dependent Variables
In our overall study design, we used weekly measurements to observe our dependent variable—indoor air
concentration. We expected the indoor air concentration to be dependent on the flux from vapor intrusion
from soil gas. Our dependent variable is thus controlled by a series of independent variables with different
time cycles that affect the vapor intrusion process, including barometric pressure, soil moisture, soil
temperature, water level, HVAC operation, and air temperature.
In the course of this study, we monitored or measured most of these independent variables or their
surrogates and different frequencies balancing on the general desire for continuous measurements against
logistic considerations. Table 2-2 considers these time-scale issues and the implications they may have
for our test matrix. Figures in Nazaroff and Nero (1988) show examples of how such independent
variables controlled indoor radon concentrations in previous studies.
Table 2-2. Factors Causing Temporal Change in Vapor Intrusion and
How They Are Observed and Measured
Independent
Variables/Causes
of Variability
HVAC system
on/off
Diurnal
temperatu re/wi nd
(night/day)
Barometric
pumping from
weather fronts
Water table
fluctuations
Soil and
groundwater
temperature
change
Vadose zone
moisture change
Stack-effect,
heating vs. cooling
season
Expected Time Cycle
• 10 min-1 hour
• 24 hour
• 2-3 days typical
• Barometric pressure: 2-3
days
• Major rain events:
irregular
• Seasonal climate:
monthly
• Surface events: blasting,
railroad operations, etc.:
minutes, irregular
• Annual/seasonal
• Seasonal major rain
events?
• Daily and seasonal
Indoor VOC & Soil Gas
Measurement Intervals
Available to Observe at
These Time Scales
None: all measurements 24
hours or longer
None: all measurements 24
hours or longer
Weekly, except for daily
samples and continuous
measurements during
intensive periods.
Monthly integrated indoor air
samples
Weekly, biweekly, and
quarterly samples of indoor
air and soil gas
Weekly samples of indoor air
and soil gas
Weekly samples of indoor air
and soil gas
Measurements of
Independent Variables
Available
Measurement with data
logger was planned every
five minutes within heating
season.
Weather station: at least one
data point per hour
Weather station: ambient
pressure logging with at least
one point per hour.
Monthly water-level
measurements
Soil temperature logging with
thermocouples: one or more
points per hour. Groundwater
temperature monthly during
sampling.
Once per hour at 5 depths
Differential pressures, indoor
temperatures: 15-minute
rolling average
2-15
-------
Section 2—Introduction
2.2.2 Data Quality Objectives and Criteria
Table 2-3 summarizes the data quality objectives and criteria for this project. Each objective is expressed
first qualitatively, and then each objective is expressed in quantitative and statistical terms where possible.
The measurements that were used to achieve each objective are also listed. More details on the
measurements to be made are given in Section 3 of this report.
2-16
-------
Section 2—Introduction
Table 2-3. Data Quality Objectives and Criteria
Task Order Objective
Measurements Used
Study Question
Performance or Acceptance Criteria
Primary (Explicit) Objectives
A-1. Determine relationship of
radon to VOC concentrations
in soil gas and in indoor air.
Radon and VOC measurements in
indoor air and soil gas (subslab and
subsurface).
Is there a statistically significant
correlation of radon to VOC
concentrations for each medium?
Replicate measurements: +/-30%
Completeness'. Number of measurements in
medium is adequate to draw quantitative
conclusions.
B-1. Establish attenuation
between subslab and indoor
air concentrations of VOCs
and radon.
Radon and VOC measurements in
subslab soil gas and basement
indoor air.
What is the range and mean of
the subslab to indoor air
attenuation factors for VOCs and
radon? Is there a statistical
relationship?
Replicate measurements: +/-30%
Attenuation factor range: 2-4 orders of magnitude.
Completeness: Adequate measurements for
quantitative conclusions.
C-1. Examine whether
external soil gas samples near
building can be used as
surrogates for subslab, and at
what depth.
Correlated (by time) external soil gas
sampled between 4 ft and 16 ft bgs,
subslab soil gas, and indoor air.
At what depth does the external
soil gas adequately predict
average subslab and indoor air
concentrations, using EPA
attenuation factors? What is the
statistical goodness of fit
between predicted and actual
indoor air values?
Replicate measurements: +/-30%, based on
several orders of magnitude variability in soil gas
and subslab attenuation factors.
Attenuation factor range: 2-4 orders of magnitude.
Completeness: Adequate measurements for
quantitative conclusions.
D-1. Identify seasonal
variations in radon and VOC
vapor intrusion flux (i.e., indoor
air concentrations) and relate
to the use of home HVAC
system.
Weather information, VOC and radon
concentration time series in indoor
air, differential pressure, air exchange
rate.
Are there statistically significant
seasonal trends in radon and
VOC indoor air concentrations? If
so, do they correlate with HVAC
operation and differential
pressures across the slab? Are
there alternative factors?
Replicate measurements: +/-30%, based on 2*-
5x expected seasonal variability in indoor air
concentration.
Completeness: Adequate measurements for
quantitative conclusions.
(continued)
2-17
-------
Section 2—Introduction
Table 2-3. Data Quality Objectives and Criteria (continued)
Task Order Objective
Measurements Used
Study Question
Performance or Acceptance Criteria
Secondary (Implicit) Objectives
lmplicit-1. Determine sample
duration limits for solvent-
extracted passive indoor air
samplers (i.e., is uptake rate
constant overtime?).
Solvent-extracted passive sorbent
samples with varying duration at
same location: weekly, biweekly,
monthly, quarterly, annually. 1-week
samples are the standard for
comparison.
Are the integrated concentrations
measured with 1-week samples
statistically equivalent to those
measured for longer periods?
Replicate measurements: +/-30%, based on
typical risk assessment safety factors of at least
one order of magnitude.
Completeness: Adequate measurements for
quantitative conclusions.
lmplicit-2. Determine if
observed changes in indoor
air concentration of volatile
organics of interest are
mechanistically attributable to
changes in vapor intrusion
Indoor concentration, pressure
differential, concentrations in
immediate subslab soil gas samples,
concentrations in the shallow soil gas
samples immediately adjacent to the
basement walls, concentrations in
ambient air.
Is the observed variance in indoor
air concentration correlated to
changes in differential pressure
across the slab, immediate
subslab concentration, soil gas
concentration immediately
adjacent to the basement walls,
or some combination of these?
Is the observed variance in indoor
air concentration correlated to
changes in ambient air
concentration?
+/-30% accuracy for soil gas and indoor air
concentration. +/-0.5 Pa for differential pressure
(expected range +10 Pa to -10 Pa).
Indoor air, immediate subslab soil, and soil gas
concentrations outside the basement wall were
measured weekly. Differential pressure was
measured more frequently. Thus, an analysis of
which factors contribute to the variance in 52
measurements of the independent variable is
possible.
lmplicit-3. Characterize the
near building environment
sufficiently to allow future 3D
modeling of this site
All measurements in the text matrix
contribute, including soil lithological
logging, utility corridor mapping,
characterization of soil TOC and bulk
density, building air exchange rate.
Since an extensive modeling
exercise is not currently funded, a
formal numerical criterion for
model fit to field data is not being
established at this time.
The planned dataset is as or more extensive than
any known vapor intrusion dataset on a single
building. The accuracy requirements for the
principal measurements defined in the other
objectives are anticipated to be adequate for future
modeling as well.
lmplicit-4. Confirm that the two
analytical laboratories (Air
Toxics and US EPA) can
produce soil gas VOC data
with sufficient agreement, that
the variance between
laboratories is not significant
compared to the changes in
the underlying phenomena
being observed
Collocated or split duplicate soil gas
samples analyzed by both
laboratories. A set of replicate soil
gas samples were acquired by
following normal soil gas sampling
purge procedures for this project.
Then four samples were collected in
rapid sequence. Samples A and C
were submitted to Air Toxics,
samples B and D to EPA for analysis.
Do the Air Toxics and EPA
analyses of duplicate/collocated
soil gas samples agree with each
other to+/-30%?
Is the variability between
duplicates analyzed by two
different laboratories significantly
greater then duplicates analyzed
by any one laboratory?
We plan to acquire at least 5 quartets of split
samples for analyses by the two laboratories.
(continued)
2-18
-------
Section 2—Introduction
Table 2-3. Data Quality Objectives and Criteria (continued)
Task Order Objective
Measurements Used
Study Question
Performance or Acceptance Criteria
lmplicit-5. Evaluate the extent
to which groundwater
concentrations control soil gas
concentrations at this site and
thus indoor air concentrations.
Measurement of VOCs in
groundwater, soil gas at various
depths and indoors.
Is the temporal variability in
immediate subslab soil gas
concentration primarily
attributable to the variability in
deep soil gas concentration?
Is the temporal variability in deep
soil gas concentration primarily
attributable to the temporal
variability in groundwater
concentration?
Is the temporal variability in
indoor air concentration primarily
attributable to the temporal
variability in groundwater
concentration?
Does the three-dimensional
pattern of subslab and external
soil gas concentrations suggest
that the primary source of VOCs
to indoor air is migrating from a
groundwater source (between 16
ft bgs) or a source in the vadose
zone? This can be evaluated by
determining if the highest soil gas
concentrations in subslab soil gas
are matched or exceeded by
deep soil gas or shallow soil gas
external to the house.
Each measurement is expected to be +/- 30%
accuracy or better. Because only two groundwater
well clusters are planned, data analysis for this
objective focuses on those two clusters and the
soil gas sampling points most proximate to them.
Each of the groundwater wells were sampled
monthly.
2-19
-------
Section 3—Methods
Table of Contents
3. Methods 3-1
3.1 Site Description 3-1
3.1.1 Area Geology/Hydrogeology 3-1
3.1.2 Area Potential Sources 3-3
3.1.3 Building Description 3-7
3.1.4 Building Occupancy During Sampling 3-9
3.1.5 Initial Site Screening 3-9
3.1.6 Initial Conceptual Site Model 3-13
3.2 Building Renovation 3-14
3.2.1 HVAC Refurbishment and Operations 3-14
3.2.2 Plumbing Refurbishment and Sealing 3-17
3.3 Monitoring Infrastructure Installation (Wells, SGPs, Embedded Temperature Sensors) 3-18
3.4 VOC Sampling and Analysis 3-23
3.4.1 Indoor (Passive, Summa Canister) 3-23
3.4.2 Subslab and Soil Gas (TO-17 and Summa Canister) 3-25
3.4.3 Online Gas Chromatograph 3-25
3.4.4 Groundwater 3-27
3.5 Radon Sampling and Analysis 3-27
3.5.1 Indoor Air Radon Sampling and Analysis 3-27
3.5.2 Subslab and Soil Gas Radon Sampling and Analysis 3-28
3.5.3 Continuous (Real-Time) Indoor Air Radon Sampling and Analysis 3-29
3.6 Physical Parameters Monitoring 3-29
3.6.1 On-Site Weather Station 3-29
3.6.2 Indoor Temperature 3-31
3.6.3 Soil Temperature 3-31
3.6.4 Soil Moisture 3-31
3.6.5 Potentiometric Surface/Water Levels 3-32
3.6.6 Differential Pressure 3-32
3.6.7 Air Exchange Rate 3-32
3.6.8 Crack Monitoring 3-33
List of Figures
3 -1. Lithological fence diagram showing the maj or soil types beneath the 422/420 house 3-2
3-2. Aerial view of duplex, 420/422 East 28th Street, showing nearby sanitary and storm
sewers 3-3
3 -3. East side of house (on right) and adj oining commercial quadraplex visible (left) 3-4
3-4. Roof of adjacent commercial quadraplex 3-4
3-5. Looking toward southeast corner of adjacent commercial quadraplex 3-5
3-6. Visual evidence of historic dry cleaners in area 3-6
3-7. Front view of house during summer 2011 sampling, with fan testing and weather station 3-7
3 -8. Front view of duplex under winter conditions showing designation of sides and HVAC
setup 3-8
3-9. 422 (left) and 420 East 28th Street in January 2011 3-8
3-10. Test building floor plan showing sampling locations used in preliminary screening 3-11
3-11. Basement supply register in newly installed HVAC system 3-14
3-12. Common returns from first and second floors in newly installed HVAC system 3-15
3-13. Gas-fired forced hot air HVAC system installed in 422 3-15
3-i
-------
Section 3—Methods
3-14. Floor cracks in 422 basement, central area, contrast enhanced 3-17
3-15. Weathered cement in walls and floor cracks in 422, contrast enhanced 3-17
3-16. Floor drain, 422, 1st floor laundry area 3-18
3-17. Nested monitoring well 1 and SGP1 are located immediately south of the 422 side front
wall. SGP1-16.5 and MW-1A is by the wall, to the left of the sign. SGP1-9 and 1-13 as
well as MW-1C are by the wall, to the right of the sign (next to the pile of bricks). SGP1-
3.5 and 1-6, and MW-1B are in the installation visible in the center foreground 3-19
3-18. Exterior of test building showing utility corridors, ground surface cover, monitoring wells
(MWs), soil gas points (SGPs), thermocouples (TCs), and moisture sensors (MSs) 3-20
3-19. Interior floor plans of 420 and 422 East 28th Street showing sampling locations 3-21
3-20. Interior SGP9 (top) and SSP-4 (bottom) 3-22
3-21. Wall port 2 3-22
3-22. Passive indoor air sampling rack: 422 first floor 3-23
3-23. Ambient sampler shelters on telephone pole near duplex 3-24
3 -24. Monitoring well MW-3, installed in the basement and completed on the first floor 3-27
3-25. Front view of 420/422 duplex with location of weather station sensors indicated with red
arrow 3-30
3-26. Calibrated crack monitor 3-33
List of Tables
3-1. Preliminary Indoor Air VOC Screening Results—Fan Off, Basement 3-10
3-2. Preliminary Indoor Air Radon Screening—Fan Off, Basement 3-12
3-3. VOC Results (ug/m3) for Subslab and Soil Gas at 422 E. 28th St., Indianapolis—Fan Off 3-12
3-4. VOCs (ug/m3) in Indoor Air and Subslab Soil Gas, 420 & 422 E. 28th St., Indianapolis-
Fan On 3-12
3-5. Radon (pCi/L) in Indoor Air & Subslab Gas at 420 & 422 E. 28th St. Indianapolis— Fan
Off and On 3-12
3-6. Groundwater Screening Data 3-13
3-7. Soil Analysis from MW-1 Boring at Multiple Depths 3-13
3-ii
-------
Section 3—Methods
3. Methods
3.1 Site Description
We selected a vacant residential duplex at 420/422 East 28th Street in Indianapolis for testing. This house
lies in the Mapleton-Fall Creek neighborhood (IndyGov, 2012). This area of Indianapolis was initially a
farming settlement known as Mapleton founded in the 1840s. The primary residential development in this
area occurred in the late 1800s and early 1900s. Commercial development on the immediate cross street,
Central Avenue, began in the 1920s.
3.1.1 Area Geology/Hydrogeology
Several soil borings were advanced in the area immediately surrounding the house, during monitoring
well (MW) construction and soil gas port (SGP) installation. SGP1A, SGP1B, and SGP1C, as well as
MW-1A and MW-1B, were installed on April 29, 2010. All additional SGPs and MWs on the exterior of
the house were installed between August 30 and September 1, 2010. SGPs and MW-3 located below the
footprint of the house were installed in September 2010. Three-dimensional visualizations of subsurface
lithology are presented in Figure 3-1. Boring logs are included in Appendix A.
In the southern portion of the property, topsoil extends down to about 0.5 to 1 ft. Beneath the topsoil is
sand or silt mixed with cinders, coal fragments, or ash to about 1.5 ft. From 1.5 ft to between 5 and 6 ft is
silt or silty sand with varying amounts of clay. Some trace gravels start at about 7 ft, and underlying that
layer are sands and gravels to between 15 and 16 ft. Beneath the sand and gravel layer is generally sand.
To the east side of the property, at the surface, are soils with a visibly high organic content and a gravel or
a concrete sidewalk. Underlying the surface soil from 1 to 3 ft is sand or clayey sand, with some gravel
and coal fragments in some borings. Beneath that layer down to 7 ft is predominantly clay with some sand
or silt. Underlying that layer is a layer of sand with some clay and gravel down to about 12 to 14 ft. From
14 to 16.25 ft is a layer of sand with gravel to 16.5 ft.
To the north side of the property, the first foot is fill, sand, and gravel. From 1 to 3 ft is brick, with sand
and weathered brick to 3.5 ft. The brick constituent in this location is possibly a remnant of a former
exterior basement stairwell. From 3.4 to 6.25 ft is a silty, sandy clay. From 6.5 to 8 ft is sand, with sand,
gravel, and some clay down to 12 ft. From 12 to 16 ft is all sand.
On the west side of the property, the first half-foot beneath the surface is the concrete sidewalk.
Underlying that to 1.25 ft is fill, cinders, and gravel. From 1.25 to 6.75 ft is a silty, sandy clay with trace
gravel. The layer beneath that to 15.5 ft is sand and gravel with some clay followed by sand to the end of
the boring at 16.5 ft. See Section 6.1 for additional information on site soils.
3-1
-------
Section 3—Methods
Figure 3-1. Lithological fence diagram showing the major soil types beneath the 422/420 house.
In the top figure, the view is toward the north from the street in front of the house. The bottom figure shows a view
toward the south from the backyard. The empty white area at the top of the soil figure represents the house
basement. In the immediate vicinity of the house, silt and clay (brown) are present until 7.5 to 8 ft below land surface
(bis). After that, sand and gravel (burnt orange) alternate with layers of sand (orange).
3-2
-------
Section 3—Methods
3.1.2 Area Potential Sources
The site location, as illustrated in Figure 3-2, is bounded to the south by 28th Street, to the west by North
New Jersey Street, and to the east by Central Avenue. There is a large stream, Fall Creek, approximately
300 ft to the south of the site toward which groundwater generally trends. Across the street south of the
site, there is a parking lot and to the east there is an open field. Across an alley to the west of the site,
there is an open lot with a grassy area and a paved parking lot. Adjacent to the north side of the site there
are backyards of the residential buildings along Central Avenue.
420 E. 28th St, Indianapolis, IN
4
4
Storm_sewers.shp
Sanitary_sewers.shp
Gas Lines
Storm_sewer_structures.shp
Sanitary_sewer_structures.shp
CSO's
Figure 3-2. Aerial view of duplex, 420/422 East 28th Street, showing nearby sanitary
and storm sewers.
Immediately adjacent to the studied duplex (approximately 10 ft east) lies a small commercial/residential
quadraplex (Figures 3-3, 3-4, and 3-5) with a diverse, primarily commercial history dating back to 1930.
The four portions of the building are numbered as 424 East 28th Street, 426 East 28th Street, 2802
Central Avenue, and 2804 Central Avenue.
3-3
-------
Section 3—Methods
Figure 3-3. East side of house (on right) and adjoining commercial quadraplex visible (left).
Figure 3-4. Roof of adjacent commercial quadraplex.
3-4
-------
Section 3—Methods
Figure 3-5. Looking toward southeast corner of adjacent commercial quadraplex.
Among the historic uses of parts of that building were a pharmacy and beauty supply, radio, fur, and
detector companies. Regarding most of the businesses that occupied that space, only their names are
currently known and those names do not match any businesses with a current local or Internet presence.
Thus, chemical uses, though probable, are not documented. The back part of the adjacent building at 2804
Central Avenue has historically been occupied by "Wolf Fur Co." Later in 1954, the same location was
occupied by the "Avideo Detectors Telaveta." In 1930, it was occupied by "Gould & Schildmoler ENEN"
and "Home Radio Co." The records for the adjacent buildings (424 to 428 East 28th Street and 2802 to
2804 Central Avenue) show a number of drug store and beauty shop uses. There are substantial gaps in
the records for these properties; there seems to be little or nothing reported about what was occupying
these locations between 1970 and 2000.
There were 9 to 10 historic laundry cleaners located less than a quarter of a mile to the north of the
422/420 house, and one was a quarter of a mile to the west (Figure 3-6). These laundry cleaners were
listed as hand and steam laundries, pressers, and driers. The most recent laundry was present in 1970
(Environmental Data Resources (EDR) Radius Map, June 15, 2010). In the fall of 2010, we observed
Mapleton-Fall Creek Development Corporation (MFCDC) staff excavating an underground storage tank
that appeared to contain product at a dry cleaner several blocks upgradient from the 422/420 house.
There were three historic gas stations or auto service and repair shops within a quarter of a mile to the
north as well. The most recent auto repair shop was present in 1990 (EDR Radius Map, June 15, 2010).
3-5
-------
Section 3—Methods
Figure 3-6. Visual evidence of historic dry cleaners in area.
The property southwest of the intersection of East 28th Street and Central Avenue was historically mildly
impacted with petroleum hydrocarbons and managed as a Brownfield named "Mapleton-Fall Creek Site"
or "Fall Creek Central Project." This site was closed after tank and soil removal. One round of volatile
organic compound (VOC) groundwater data was acquired at that location that showed detectable
chloroform (8.9 to 22.1 (ig/L in a June 2005 sampling event). These previous studies showed that the
study area has sand and gravel geology from approximately 7 to 25 ft below land surface (bis) and
groundwater at approximately at 16 ft bis. The upper 7 ft of the stratigraphy is heterogonous, variously
described as including fill materials, loam, and silty and moist sandy clay.
Based on the general topography of the area and professional experience in this portion of Indianapolis,
groundwater is thought to flow from the north of the 422/420 house south of the house to Fall Creek.
Thus, many of the historic laundries or auto shops that are potential contaminant sources are generally
upgradient of the studied house.
The 422/420 duplex is located between Central Avenue and its associated alleyway on 28th Street. The
immediate area receives a moderate amount of traffic, but the Central Avenue/Fall Creek Parkway
intersection is very busy throughout most of the day. Traffic could be a contributing factor to petroleum-
based contaminants in surface soils.
3-6
-------
Section 3—Methods
3.1.3 Building Description
3.1.3.1 Building Age, Condition, and HVAC
The tested house located at 422/420 East 28th Street, Indianapolis, IN (Figure 3-7) is an early twentieth
century duplex, dating from before 1915 because it is present on the 1915 Sanborn map of the area. Based
on the mirrored floor plans of the two sides, it is likely that the house was always a duplex. Construction
is wood frame on a brick foundation with a poured concrete basement floor. Interior floor materials
include tile, carpet, and wood flooring.
Figure 3-7. Front view of house during summer 2011 sampling,
with fan testing and weather station.
The duplex at 422/420 was previously vacant and is now owned by Mapleton-Fall Creek Development in
Indianapolis. Before our involvement, the house had been vandalized and stripped of all valuable metals
and fixtures. In all likelihood, the house was never to be restored for use. An operative from the
Indianapolis ARCADIS office acquired the use of the house for the duration of the project. The vandalism
and theft of household items included the following: all copper wiring and tubing, most plumbing
fixtures, and many outlets. Vandals destroyed the previous HVAC unit, probably in an attempt to obtain
any valuable metals. We restored power to the house in September 2010. A gas-fired forced air HVAC
unit was installed on the 422 side in October 2010 by Edward's Electric for use in this project (Figure
3-8). The house had no air conditioning (AC) system, and we chose to install window-mounted units,
which would have been the likely type used by any tenants in this house.
3-7
-------
Section 3—Methods
420 NO
Heated
Heated
Figure 3-8. Front view of duplex under winter conditions showing
designation of sides and HVAC setup.
Figure 3-9. 422 (left) and 420 East 28th Street in January 2011.
There are internal and external visual clues indicating (Figure 3-9) the house has been updated several
times. For example, visual clues suggest that a previous FfVAC unit had been installed that was not native
3-8
-------
Section 3—Methods
to the house's original construction. In the basement, there is evidence of former coal chutes (possibly)
and cisterns on both the 420 and 422 sides. The probable coal chutes and old windows had been blocked
by cinder blocks before ARCADIS occupancy. The cisterns had also been cemented over. Comments
made by electricians in the basement suggest that at one time the house had been heated by an old style
furnace, indicated by cemented-over holes in the walls, but that the furnace had been gone for some time.
3.1.3.2 Building Utilities/Potential Entry Points
The electric lines connect to the house at the northwest corner of the 420 side. Because all original wiring
native to the house had been removed by vandals before the project, we had to have the junction box
rewired to the city electrical line and run new lines within the house to new outlets at designated points.
The gas line connects only to the furnace from an access line in the south wall of the 422 side. Both the
electrical lines and the gas line were emplaced by Edward's Electrical during the furnace installation and
enter the house at the original entry points for each utility.
Sanitary sewer lines run immediately south of the house along East 28th Street. Sanitary and combined
sewer lines run less than one block east and west of the house along Central Avenue and New Jersey
Street (see previous Figure 3-2). There is a sewer line running beneath the basement floor along the
length of the 422 side from north to south that was buried and cemented over sometime after the floor's
original construction. PVC drain lines join this sewer line, running laterally from the plumbing on both
sides of the duplex. The HVAC unit drains condensation into a floor grill leading to the lateral. A
nonfunctional water line enters the house from the south. Large, cinder-blocked portions of the north
interior basement walls of both sides of the duplex along with brick strata in borings have been observed.
We interpret these cinder-blocked walls to be vestigial entranceways to the basement from a time when
the basement was accessed from the back yard, rather than from an interior basement door.
3.1.4 Building Occupancy During Sampling
The initial concept for the 422/420 house was to create an environment free from lifestyle-related indoor
air sources, but operated as though the space were occupied, to simulate a living environment. The
422/420 house was borrowed from MFCDC, which owns the property. It was thought that the house
would eventually be torn down because it had been previously abandoned and vandalized. We thought
that the house would be ideal because it had no occupants, limited use beyond the project, an ideal
location, and vapor intrusion was present.
Because the house was in poor condition and the house had no occupants, we could make any alterations
to the house necessary to set up ports, wells, and sensors for observations. Changes were made without
having to consider the occupants' comfort. For example, the fan testing (to be described in Section 12.2)
would have been inconvenient for a homeowner.
To more closely simulate a living environment, a field scientist worked on-site during most normal work
weeks during the year of intensive sampling, for several months before intensive sampling began, and
during many off times as the need required. The intent was to have an individual who would open doors
and windows, move through the environment, and make temperature adjustments when the seasons
dictated, similar to the way a homeowner would. The constant close proximity of the worker to the work
zone also allowed for quick responses to environmental changes. A second floor bedroom on the 422 side
of the duplex was minimally modified and used as an office for the sampling staff member.
3.1.5 Initial Site Screening
A preliminary indoor air screening evaluation was conducted March 15 to 17, 2010, where basement
indoor air samples were collected and analyzed for VOCs and radon. A second radon sampling was
conducted from March 27 through April 1, 2010. The heat and the fan were off during these sampling
3-9
-------
Section 3—Methods
events. The VOC and radon results are presented in Tables 3-1 and 3-2, respectively. Detected
concentrations of perchloroethane (PCE), chloroform, and radon were at levels of 2.8 |o,g/m3 for PCE,
3.3 |o,g/m3 for chloroform, and 4.98 pCi/L for radon.
Initial sampling for subslab and soil gas VOC evaluation took place May 6-7, 2010. The heat and fan
were off during this sampling event. The results are presented in Table 3-3 and indicate higher PCE
concentrations in subslab and deep soil gas. Therefore, it was unclear if the VOC impact at the site is
from groundwater source, from a deep vadose zone source, or from both.
A confirmatory sampling event took place June 23 through June 25, 2010, for VOCs, and June 23 through
July 14, 2010, for radon. At this sampling event, a fan was turned on during sampling to create or increase
the differential pressure that could enhance the vapor intrusion. Samples were taken from indoor air (first
floor and basement) and from the subslab; no soil gas samples were taken in this event. All sampling
locations used in this initial screening are presented in Figure 3-10. The results for the VOC and radon
screening analysis are presented in Tables 3-4 and 3-5, respectively.
Table 3-1. Preliminary Indoor Air VOC Screening Results—Fan Off, Basement
Duration (min)
Compound
t-1,2-Dichloroethene*
cis-1 ,2-Dichloroethene*
Chloroform*
1,1,1 -Trichloroethane
Benzene
1,2-Dichloroethane*
Trichloroethene
Toluene
Tetrachloroethene
Ethylbenzene
m,p-Xylene
o-Xylene
Styrene
1 ,3,5-Trimethylbenzene*
1 ,2,4-Trimethylbenzene
420 E. 28thSt., Indianapolis
1003392-01 A
File: 2879
ng
1.6
<2.0
16
<5.5
160
4.7
4.6
340
52
51
160
59
16
37
17
TO31909
ug/m3
0.024
<0.030
0.27
<0.096
2.0
0.070
0.059
3.9
0.71
0.69
2.1
0.84
0.21
0.59
0.27
422 E. 28thSt., Indianapolis
1003392-02A
File: 2878
ng
2.0
13
198
7.3
160
5.2
17
342
202
51
160
58
14
33
15
f031910
ug/m3
0.030
0.19
3.3
0.127
2.0
0.078
0.21
4.0
2.8
0.69
2.1
0.81
0.18
0.52
0.23
* Estimated sampling rate
Note: No vinyl chloride was identified.
3-10
-------
Section 3—Methods
m Radon Sample (Electret)
* VOC Sample (Radiello)
• Sub-slab Port
"Gaps in walls in basemenl and firs! floor are open passageways, no evidence thai they have had doors
Screening Sample Locations & Notes
A = March 2010 Indoor Air VOC and Radon Sample (Sample ID: 422 E. 28th St)
B = March 2010 Indoor Air VOC and Radon Sample (Sample ID. 420 E. 28th St)
C = June 2010 Indoor Air VOC and Radon Sample (Sample ID: Flrsl Floor)
D - June 2010 Indoor Air VOC and Radon Sample (Sample ID: Basement North)
E = June 2010 Indoor Air VOC and Raon Sample (Sample ID: Basement South)
F = June 2010 Ambient Air VOC and Radon Sample (Sample ID Ambient)
420 E. 28th St.
_A
Horth
KilchOT)
422 E, 28th St.
Dining Room
Box Fan -
LMfig Room /\
l-'.w h
f.i-.|rti, .m
/\
B.-.LM
Bedroom
BASEMENT
FIRST FLOOR
SECOND FLOOR
Figure 3-10. Test building floor plan showing sampling locations used in preliminary screening.
3-11
-------
Section 3—Methods
Table 3-2. Preliminary Indoor Air Radon Screening—Fan Off, Basement
Sample ID
422 E. 28th St.
420 E. 28th St.
422 E. 28th St.
420 E. 28th St.
Start Test
Start Date/Time
3/26/109:32
3/26/109:38
3/15/1014:34
3/15/1014:42
Finish Test
End Date/Time
4/1/1015:32
4/1/1015:38
3/17/1014:32
3/17/1014:41
Days
6
6
2
2
Radon
pCi/L
3.5
4.2
5.0
3.5
Table 3-3. VOC Results (ug/m ) forSubslab and Soil Gas at 422 E. 28th St., Indianapolis—Fan Off
Location
SSP-1 (subslab center, 422 E. 28th St.)
SSP-2 (subslab northeast, 422 E. 28th St.)
SGP1B(9ft)
SGP1A(13ft)
SGP1C(16.5ft)
Date
5/7/2010
5/7/2010
5/6/2010
5/6/2010
5/6/2010
Carbon
Disulfide
<36
<34
43
<34
43
Chloroform
<56
<53
140
130
120
PCE
170
<73
100
90
130
The data presented in Table 3-4 and Table 3-5 show an significant increase in indoor air and subslab
concentrations when the fan is on, indicating that vapor intrusion was enhanced by building under-
pressurization.
Table 3-4. VOCs (ug/m3) in Indoor Air and Subslab Soil Gas, 420 & 422 E. 28th St., Indianapolis-
Fan On
Location
Indoor air FF (First floor, 422 E. 28th St.)
Basement north (422 E. 28th St.)
Basement south (422 E. 28th St.)
SSP-1 (subslab center, 422 E. 28th St.)
SSP-2 (subslab NE, 422 E. 28th St.)
SSP-3 (subslab, 420 E. 28th St.)
Date
6/23-25/2010
6/23-25/2010
6/23-25/2010
6/23-25/2010
6/23-25/2010
6/23-25/2010
Carbon
Disulfide
Not analyzed
Not analyzed
Not analyzed
<35
<35
<35
Chloroform
6.9
64
27
<55
<55
<55
PCE
3.6
49
24
330
<76
<77
Table 3-5. Radon (pCi/L) in Indoor Air & Subslab Gas at 420 & 422 E. 28th St. Indianapolis-
Fan Off and On
Location
Indoor air (basement 420 E. 28th St.)
Indoor air FF (first floor, 422 E. 28th St.)
Basement north (422 E. 28th St.)
Basement south (422 E. 28th St.)
SS-1 (subslab center, 422 E. 28th St.)
SS-2 (subslab NE, 422 E. 28th St.)
SS-3 (subslab, 420 E. 28th St.)
FanOff(Mar-Apr/2010)
3.5
Not measured
5.0
Not measured
Not measured
Not measured
Not measured
Fan On (Jun-Jul/2010)
Not measured
2.5
7.8
13
530
1,100
220
3-12
-------
Section 3—Methods
Groundwater had detectable but very low PCE and chloroform (Table 3-6), although the data are suspect
because of the qualifiers related to the very low levels of analytes in the sample. Table 3-7 shows that the
soils analyzed were predominantly sandy and alkaline.
Table 3-6. Groundwater Screening Data
Sample ID
MW-1A-1
MW-1A-4
MW-1B-1
MW-1B-2
MW-1B-3
MW-1B-4
Concentration (ng/mL)
Chloroform
1.9
2.2
2.9
2.8
3.0
2.8
BU
BU
U
U
U
U
PCE
0.61
0.60
0.61
0.49
0.53
0.46
BU
BU
BU
BU
BU
BU
B - ng of analyte detected in sample is not greater than 10 times the ng of analyte detected in the method blank
U - ng of analyte detected in the sample is below the lowest calibration curve concentration of 25 ng
Method blank
0.26
0.37
Table 3-7. Soil Analysis from MW-1 Boring at Multiple Depths
Sample ID
TO-977-10ft
TO-9710-12ft
TO-9712-15ft
TO-9715-17ft
Field
Moisture
%
11.58
17.49
14.21
17.55
PH
8.18
8.26
8.33
8.45
Carbon and Nitrogen
Analysis
%
Nitrogen
0.032
0.027
0.029
0.023
% Carbon
(inorganic)
5.156
6.021
6.033
3.726
Pipette Particle Size Analysis
% Sand
83.97
79.39
78.22
96.35
% Silt
14.22
17.59
19.25
2.96
% Clay
1.81
3.01
2.52
0.69
Total %
100
100
100
100
%c
Removed in
Particle Size
Pretreatment
(organic
carbon)
0.3233
0.2255
0.1976
0.1124
3.1.6 Initial Conceptual Site Model
The initial conceptual site model for this structure was that a vapor intrusion source was most likely
present in shallow and subslab soil gas due to historical dry cleaning facilities and adjacent commercial
uses. Radon impacts were suspected because Marion County, Indiana, is in EPA's Zone 1—highest risk
for radon. Detectable concentrations of chlorinated hydrocarbons were detected during initial site
screening and responded to depressurization of the structure by fans.
The source of VOCs observed at this duplex was initially suspected to be transport of contaminants either
• through a groundwater pathway from upgradient dry cleaners or
• released into the shallow vadose zone during the operations of the adjacent commercial
quadraplex.
Later discussions also suggested that an additional potential source is likely disinfection by-products in
city drinking water.
3-13
-------
Section 3—Methods
3.2 Building Renovation
3.2.1 HVAC Refurbishment and Operations
Because project objectives included evaluating the effects of a central heater and air conditioners, we
wished to use the 420 side of the duplex as a "control" without heating, and because the house had been
extensively vandalized, Edward's Electric was hired to install new ductwork with a new gas-fired forced
hot air HVAC unit on the 422 side only. The furnace ran initially from November 19, 2010, until June 22,
2011, and then again from November 19, 2011, until June 1, 2012. The system is illustrated in Figures
3-11 through 3-13.
Figure 3-11. Basement supply register in newly installed HVAC system.
3-14
-------
Section 3—Methods
Figure 3-12. Common returns from first and second floors in newly installed HVAC system.
^j*
Figure 3-13. Gas-fired forced hot air HVAC system installed in 422.
3-15
-------
Section 3—Methods
Window-mounted AC units were initially installed in two upstairs windows on both the 422 and 420 sides
of the house. However, thieves broke into the house, stealing the initial four AC units on July 12 and 15,
2011. After a security system upgrade, two replacement AC units were installed in two of the 422 upstairs
windows. Therefore, during the summer of 2011 there were periods when:
• both sides of the duplex were cooled with air AC (June 29, 2011, until July 12, 2011),
• neither side of the duplex was cooled with AC (before June 29, 2011, and after July 15, 2011,
until August 3, 2011), and
• the 422 side was cooled but the 420 side was not (August 3, 2011, until October 24, 2011).
Part of the intent behind using this formerly vacant house is that we could operate it under nearly normal
residential conditions without having to consider the residents or consumer product-related sources,
because there were none. These conditions allowed us to control the environment (avoiding indoor
sources of contaminants, adjusting environmental conditions at will, adding data collection devices, etc.)
without having to consider occupant permission, inconvenience, or potential tampering. The house
environment was kept as residents might keep it, but an ARCADIS operative was only on-site during the
work week, with occasional weekend work, as opposed to full-time occupancy.
Because of its age, the 422/420 duplex building envelope is particularly leaky. However, every effort was
made at the beginning of the project to further the simulation that this was a normal house. We
endeavored to make any repairs necessary to put the house in a state similar to one in which an actual
resident of limited means would live. Any holes made by vandals in pipes or walls were sealed with foam
(Great Stuff) or medium density flat (MDF) board, with enough time before sampling began to allow for
drying and ventilation. When holes were made to install wiring, SGPs, MWs, gas chromatography (GC)
tubing, or any utility meant for this project, we attempted to seal openings so there was no additional air
communication between normally partitioned areas. New ductwork was installed with the new HVAC on
the 422 side, so there were no unusual air leaks between floors. Some ductwork is partially in place on the
420 side but not connected. A new front door on the 422 side was installed in an attempt to minimize
leakage. But no attempt was made to go beyond what a normal homeowner would normally do. As a
result, the heated/cooled side maintained a moderate temperature in the high 60s/low 70s (Fahrenheit) in
the winter, and high 70s/low 80s (Fahrenheit) in the summer.
Despite these repairs, potential air entry points still exist as in any home, especially of the age of this
duplex. These entry points could include the edges of any of the windows or doors, exposed brick work,
and cracks (Figures 3-14 and 3-15). In the basement, entry could be through bricks in the basement walls,
potentially through cracks in the cement floor, and possibly through the sewer lateral in the floor.
3-16
-------
Section 3—Methods
Figure 3-14. Floor cracks in 422 basement, central area, contrast enhanced.
Figure 3-15. Weathered cement in walls and floor cracks in 422, contrast enhanced.
3.2.2 Plumbing Refurbishment and Sealing
Initially in early to mid-2010 before sampling, the house had a strong odor of sewage. Two causes were
located and addressed well before sampling began:
• A hole was found on the top of a sewer drain line running horizontally along the basement
ceiling. It was taped and later foamed to seal it.
• A vagrant had used an upstairs toilet despite the lack of running water in the house or visible
water in the toilet. This issue was addressed by adding a large quantity of kitty litter (bentonite).
This measure appeared to solve the odor issue.
However, a decision was made that complete repair of the plumbing system of the home was not
necessary for the project purposes and would be costly.
3-17
-------
Section 3—Methods
On approximately March 31, 2011, a sewer gas odor was detected in the residence that had not been
previously noted. The likely source was dried-out traps in the water pipes. Olfactory observations
suggested that the primary source of sewer gas infiltration was a floor drain (Figure 3-16), and the drain
was initially covered with a metal plate as a temporary measure.
There are many circumstances in occupied houses that allow openings for sewer gas infiltration. It seems
that it is relatively common for home inspectors to encounter them during property transactions. Entrance
of VOCs through sewer lines is a widely recognized aspect of the vapor intrusion problem.
At that time, all sewer and water connections were surveyed, and several were sampled with passive
Radiellos from March 14, 2011 to April 21, 2011. The only VOC results of potential significance came
from a drain in the 422 kitchen, likely used as a washer drain. After this sampling, on May 10, 2011, all
open lines were sealed, some with both bentonite and cement plugs.
Figure 3-16. Floor drain, 422,1st floor laundry area.
3.3 Monitoring Infrastructure Installation (Wells, SGPs, Embedded Temperature
Sensors)
The monitoring systems were installed in three main phases. The first phase occurred in April 2010,
during which a few SGPs (SGP1A, SGP1B, SGP1C), SSPs, and MWs (MW-1A, MW-1B) were installed
to check for evidence of vapor intrusion in initial site selection using a hollow stem auger for the MWs
and geoprobe for the SGPs. The second took place August 30 to September 1, 2010, using a geoprobe and
included all additional exterior SGPs and MWs (e.g., Figure 3-17). The last phase began at the same time
as the second phase and included all interior ports and wells and all monitoring sensors, but it took longer
because of difficulties penetrating the soil layers beneath the structure with the equipment that could be
used to work in the fairly confined basement. The interior port installation was started with hand auguring
equipment, but, because of the difficulty involved, rock coring equipment was finally used to complete all
interior SGPs, water, and sensor wells.
3-18
-------
Section 3—Methods
Figure 3-17. Nested monitoring well 1 and SGP1 are located immediately south of the 422 side
front wall. SGP1 -16.5 and MW-1A is by the wall, to the left of the sign. SGP1 -9 and 1 -13 as well as
MW-1C are by the wall, to the right of the sign (next to the pile of bricks). SGP1 -3.5 and 1 -6, and
MW-1B are in the installation visible in the center foreground.
Figure 3-18 shows the exterior monitoring locations, and Figure 3-19 shows the interior monitoring
locations. The installed monitoring network includes the following:
• Seven groundwater monitoring wells. Two three-well clusters (MW-1 and MW-2, at 16 to 21 ft,
21 to 24 ft, and 24 to 26 ft) were installed with a hollow stem auger and 2-inch PVC casings.
MW-3, which was installed in the basement and completed on the first floor, is 1 inch in diameter
and has a screen depth between 13.4 and 18.4 ft bis.
• Seven external soil gas locations, 5 depths each (3.5, 6, 9, 13, 16.5 ft bis), designated SGP1
through SGP7. These locations were installed with a geoprobe with 6-inch stainless steel screens
completed to the surface with 1/4-inch OD Teflon.
• Five internal (basement) soil gas locations, 4 depths each (6, 9, 13, 16.5 ft bis) designated SGP8
through SGP12.
• Seven conventional subslab locations, designated SSP-1 through SSP-7
• Four basement wall ports (WPs), designated WP-1 through WP-4. These ports were constructed
as is typical for SSPs except they were drilled horizontally into the basement wall approximately
2.5 ft above the basement floor.
3-19
-------
Section 3—Methods
erior
7422 East 28th St
TC 16.5
TC13
Possible
Location of
Historic Basement
Stairs
SGP6
TC 6 MS 3.5
TC 9 MS 6
/ MS 9
/ MS 13
tt TC 3.5
~ MS 16.5
Possible Locati
Historic Basenu
Stairs
MW2 /
Electric
Meter
(Working)
Duplex
Side 1
.422^
V)
a.
I
3L
Covered
Porch
Meter '
(Working)
Phone
Line X
SGP1
73
CL>
o
(D
a
00
c
ID
Figure 3-18. Exterior of test building showing utility corridors, ground surface cover, monitoring
wells (MWs), soil gas points (SGPs), thermocouples (TCs), and moisture sensors (MSs).
3-20
-------
Section 3—Methods
Cinde
Seclit
Brick
rblock .-*
,,or Basement L
X* 420 East 28th St IN
SSP 6
0
Open
Walkway
Ii
Wall Port 4
/
/ /"" N SGP 11
I Cistern 1 f\
V^^^^j/ ccr
O SSP •}
SSP 7 uPto
1st Floor
1 1 1 I
Open Open
Walkway Walkway
SSP 5
0
0
SGP 12
r^
Basement
422 East 28th St
o\ Wall Port 1 \
s"V N N
NV^s^ SGP 10
^^t^^ SSP 2
i
Open
Walkway
1— i— I 1 1
Wall Port 2
( ""N ^
SSH 1V_X
T^SRPR TC6'MS6-
7P ^ S r\ M\*,I M5165
IstFloor/T Y ,MW3
/ Open
%%# Walkway
o OSGPQ
SSP 4
Wall Port 3
i\i
L
Cinderblock
Section of
Brick Wall •) '
Figure 3-19. Interior floor plans of 420 and 422 East 28th Street showing sampling locations.
3-21
-------
Section 3—Methods
Example subslab, soil gas, and wall ports are shown in Figures 3-20 and 3-21.
Figure 3-20. Interior SGP9 (top) and SSP-4 (bottom).
Figure 3-21. Wall port 2.
3-22
-------
Section 3—Methods
3.4 VOC Sampling and Analysis
3.4.1 Indoor (Passive, Summa Canister)
The overwhelming majority of the indoor passive sampling was done with Radiello 130s supplied by and
analyzed by Air Toxics Ltd. For comparison two different types of SKC badges were also used that were
specifically adapted to use at very short or long sampling durations. Some indoor air data were taken with
Summa canisters during the fan testing.
For passive sampling, several racks were set up to facilitate arranging groups of samplers in consistent
locations for different durations during the run of the project. These racks were ordinary laundry drying
racks that can be purchased inexpensively at most department stores (Figure 3-22). The racks were ideal
in that they allowed multiple samplers to be placed at the same, or similar, levels within the normal
breathing zone. One rack was placed in each of the following locations on both the 422 and 420 sides (six
total): first floor center room, northern basement room, and southern basement room.
Figure 3-22. Passive indoor air sampling rack: 422 first floor.
At each rack, a specific location was assigned for one of several durations: 7, 14, 28, 91, 182, and 364
days, each approximately 6 inches apart to minimize the potential for starvation effects. Enough spaces on
the rack remained for duplicates of those durations, plus special locations occupied during intensive
rounds. SKC badges were primarily hung on the back portion of the racks, in a similar manner to the
Radiellos.
3-23
-------
Section 3—Methods
In addition to these indoor racks, a special ambient location had to be made to accommodate the samplers.
A hood was purchased to house the samplers and was mounted on a telephone pole by the alley near the
house (Figure 3-23). This hood housed all of the Radiellos and badges for the different day durations.
Figure 3-23. Ambient sampler shelters on telephone pole near duplex.
Sampling of Radiellos consisted of removing the white diffusive body from its backing shield, opening
the glass vial that contained the new screened Radiello 130 and allowing it to slide into the white body,
then the white body was replaced in its backing plate with a new sample number. The old one was then
sealed in a glass vial for shipping. Each week, Radiellos of the appropriate durations were stopped and
replacements were started. For example, when the 7- and 14-day Radiellos were stopped, new ones were
put up in their places. The 7-day samples were taken down the following week, followed by the 14-day
samples the week after. This arrangement allowed us to compare the results of different time durations to
each other (e.g., four weekly samples against the monthly for the same time period). Additionally, during
some of the intensive rounds, daily Radiellos were taken to compare them to the weekly time increments.
3-24
-------
Section 3—Methods
SKC 575 badges with the secondary diffusion cover were used for comparing longest Radiello durations
(the 182- and 364-day time periods). These solvent-extracted charcoal badges have been used in the
literature for durations of 4 weeks and longer. SKC Ultra Badges (thermally desorbed) were used for 24-
hour and 7-day sampling during an intensive round and short-term sampling during a fan test. Both
Radiellos and SKC badges were provided by and returned to Air Toxics Ltd. for analysis.
Summa canisters (6-L, Method TO-15) were used for preliminary site screening, indoor air sampling
before and after the fan testing (Section 12.1), and for a study comparing temporary and permanent
subslab ports (Section 12.1). These canisters were acquired from and returned to Air Toxics Ltd. for
analysis. This project did not include an extensive comparison of Summa canisters to passive samplers
(for example Radiello) because numerous such comparisons have been performed by others (see
discussion in Section 2.1.4).
3.4.2 Subslab and Soil Gas (TO-17 and Summa Canister)
The primary method of subslab and soil gas sampling for VOCs was by TO-17. In this method, athermo-
desorption tube, with a female Swagelok end, was connected to each sampling port in turn. Each port had
its own male union connected to a valve. Before sampling, the port was purged with an SKC Universal
XR pump set to IL/min. To ensure that the soil gas sample represented soil gas and not the air in the
sampling line, five well volumes were purged via an exhaust line that ran away from the operator for
exterior ports or out of a basement window in the case of the interior ports. The fittings were attached
with wrenches, and an airtight syringe was mounted onto the other end of the TO-17 tube. When these
steps were complete, the port's valve was opened, and the syringe was used to draw 200 mL of air
through the TO-17 tube over a period of a minute. Then the port valve was closed, and the TO-17 tube
was removed and sealed for shipping.
Samples were taken from operational ports at no less than three depths each week. Initially, the preferred
depths to sample were 3.5, 9, and 16.5 ft bis exterior and 6, 9, and 16.5 ft bis interior. However, a higher
than expected water table prevented the sampling of the 16.5-ft depths for most of the duration of the
project. Unusually high water tables or perched/infiltrating water occasionally made other SGPs
inoperative. In addition, all wall ports were sampled each week, as well as a subset of the SSPs.
The majority of the TO-17 tubes collected were prepared and analyzed by the EPA National Exposure
Research Laboratory (NERL). For the extensive sampling of the intensive rounds, additional TO-17 tubes
were prepared and analyzed by Air Toxics. An intercomparison study of the two TO-17 laboratories was
conducted (see Section 4.2.4 of this report). During the intensive rounds, all functioning ports (not made
inoperative by water) were sampled at least once each day of the round. For a few days of each round,
several locations were sampled multiple times of the day with the intention of comparing hourly and daily
variability to the normal weekly variability.
Some soil gas samples were acquired with Summa canisters (Method TO-15) during the initial site
screening before the start of the main project. These samples were taken from the earliest of the drilled
subslabs and SGPs (e.g., SSP-1, SGP1). Also, Summa canisters were used to obtain soil gas during the
temporary versus permanent subslab SGP special study (see Section 12.1 for a description of this special
study).
3.4.3 Online Gas Chromatograph
The GC was provided and overseen by Dr. Blayne Hartman of Hartman Environmental Geoscience. They
used an electron capture detector (ECD, EPA Method 8021) and had 16 available sampling channels
controlled sequentially by a multiport stream selection valve. The channels were distributed as follows:
3-25
-------
Section 3—Methods
• One was initially connected to the nitrogen tank but later was connected to a line to outdoor air
(about 4 ft from the house) in order to provide an ambient air comparison.
• One was connected to a trichloroethylene (TCE) standard periodically.
• Two were blanks used to clear the instrument after each run.
• There were 12 sample channels: four indoor air, three subslab probes, one wall port, three house-
interior soil gas probes, and one house-exterior soil gas probe.
All sampling lines were constructed of 1/16-inch OD stainless steel tubing (except the 420 first floor line
that has about a 20-ft section of 1/8-inch OD stainless steel tubing at the sampling end). The different
diameters of tubing were based on available materials and were not expected to have any significant
impact on the operability of the system. The tubing for all lines ran from a multipoint stream
selector valve at the GC along interior walls to the sampling points. At the sample locations, the indoor
air lines hung suspended over passive sampler racks within the breathing zone. For SGPs and SSPs, each
tube was connected to a sampling port by means of appropriate Swagelok male/female fittings. Lines for
sampling were brought to atmospheric pressure before sampling and were sampled for 30 to 60 seconds in
each cycle.
When connected to the GC by the selector valve, the sampling point would be open for the GC to sample
but was closed when the switching valve was connected to another sampling point. The system has a
sample injection valve with an adsorbent trap or 1-cc sample loop, uses computerized data acquisition
with PeakSimple software, and can take approximately nine samples per location per day. Blayne
Hartman had constant access to the GC via a Wi-Fi connection installed at the house for instrument
monitoring. The GC was in operation at the 422/420 house from August 11, 2011, to October 17, 2011,
and again from December 1, 2011, to February 2, 2012. We checked for carryover with port 14 in the
sequence (either a TCE calibration standard or system blank). Port 13 was the last port of a series of high-
concentration PCE and chloroform SGPs. Port 14 (and later ports 15 and 16 as well) were thus used to
monitor for the possibility of carryover before returning the cycle to port 1 -5 for the indoor air analyses.
The tubing from each sample location was connected to the stream selector valve. At any time, one of the
entering tubes was connected to the adsorbent trap or sample loop depending on the position of the stream
selector valve. A low-flow vacuum pump would draw the vapor sample through the tubing at a rate of 25
cc/min to 40 cc/minute for 30 to 90 seconds to purge the sample tubing and ensure the sample in the
sample loop was from the selected sample location. When purging was complete, the sample injection
valve would rotate and inject the sample into the GC for analysis. Cycle time from start of purging to the
end of the analysis was approximately 10 minutes. When the analysis was complete, the stream selector
valve would advance to the next position (next sample location) and the process would repeat itself. This
sequence would continue uninterrupted until stopped by the operator.
In the first phase of the automated program (August 2011 to October 2011), the vapor sample from each
location was concentrated onto an adsorbent trap. Volumes passed over the trap were adjusted depending
on the vapor concentration at each location and ranged from 20 cc to 80 cc. Higher sample volumes were
collected on the trap for lower concentration locations such as indoor air. Lower sample volumes were
used for soil gas.
In the second phase of the program (December 2011 to February 2012), the adsorbent trap was eliminated
and the sample was passed through a 1-cc sample loop for direct injection into the GC. This modification
was made to minimize carry-over between the high-concentration soil gas samples and the low-
concentration indoor air samples and to speed up the analysis.
3-26
-------
Section 3—Methods
3.4.4 Groundwater
Groundwater samples were taken approximately monthly with permeable diffusion bags (PDBs) from
EON Products Inc. The 422/420 duplex has six exterior MWs (two clusters of three) and one single-depth
interior well installed in the basement and completed on the first floor (Figure 3-24). The exterior wells
are arranged in groups of three in the front and the back yards. Each group of three is divided into depths
of 16 to 21 ft, 21 to 24 ft, and 24 to 26 ft bis. The interior well (MW-3) is about 18 ft bis, but the casing
extends up to the first floor for ease of access, so it is about 24 ft deep at its access point. The exterior
wells are 2 inches in diameter, and the internal well is 1 inch in diameter. PDBs for the exterior wells are
12 by 1.75 inches, and the interior is 18 by 0.75 inches. PDBs were deployed for at least 2 weeks, and a
new set of PDBs was cycled through almost monthly. PDBs were filled initially with deionized water
provided by the EPA NERL laboratory. Groundwater samples were shipped to EPA for VOC analysis by
Methods 5030/8260.
Figure 3-24. Monitoring well MW-3, installed in the basement and completed on the first floor.
3.5 Radon Sampling and Analysis
3.5.1 Indoor Air Radon Sampling and Analysis
The primary radon sampling method was electrets ion chambers collecting radon samples passively in
indoor air for the same 7-day intervals as Radiello-collected VOCs. The following secondary methods
were, however, also used for radon in indoor air:
• stationary alphaguards at two locations to provide greater time resolution,
• carbon absorbers for a quality control (QC) comparison, and
• consumer-grade ionization chamber-based detector (Safety Siren Pro Series 3 manufactured by
Family Safety Products Inc.) for comparison.
3-27
-------
Section 3—Methods
Each method is described in detail below.
Rad Elec, E-Perm, ST-type (short-term) electrets were used according to EPA 402-R-92-004 (U.S. EPA,
1992). These electrets were primarily deployed in s-chambers, but h-chambers were used on a few
occasions. To sample, electrets were opened within their chambers at their assigned locations for a week
(or a day during intensive rounds). After a week, the chambers were closed, all electrets were allowed to
equilibrate for an hour to the room temperature where they would be read, and then their voltages were
read on a Rad-Elec electret voltage reader. Start and stop times, as well as voltages, were recorded and the
electrets redeployed. The voltages, configurations (e.g., ST electrets in s-chambers), dates, and times
would then be incorporated into a calculation used to convert voltage to pCi/L, with background gamma
correction.
The electrets reader was calibrated weekly with three standards. In addition, an electret blank test was run
weekly to test for effects of the chamber on the electrets. In this test, an electret not used during the
sampling was inserted into one of the used electret chambers (closed) and then read to determine whether
there had been any voltage drop from the previous week's reading.
Initially, one electret was hung in a mesh bag from each of the passive sampler racks each week (plus one
duplicate at one location). Additionally, an electret was also housed in the ambient sampler hood, but that
electret was stolen. The ambient electret location was then switched to a nearby tree, but that electret was
also stolen. Finally, the ambient electret was kept in a permeable bag and hung from a hook about 2 ft
from the house. This location proved to be ideal. On December 28, 2011, a new electret was added in the
422 second floor office to be used in conjunction with the radon siren testing.
Charcoal canisters from the U.S. EPA Radiation and Indoor Environments (R&IE) National Laboratory
were set out on the sampling racks on three separate occasions to check the accuracy of the electret
readings (U.S. EPA, 1990). They were simply opened for a week (matching an electret sampling period),
closed, and shipped back to EPA for testing. Section 3.5.3 discusses the stationary Alphaguards that were
also used on the project for indoor air radon measurement.
A consumer-grade radon detector (safety siren testing) was a late-stage addition to the project. Six Pro
Series 3 safety siren radon gas detectors were deployed on December 23, 2011, and were in use until the
last electret readings were taken on March 1, 2012. Each was installed at one of six locations: 422 second
floor office, 422 first floor center room, 422 basement south, 422 basement north, 420 first floor center
room, and 420 basement south. The intention of the test was to determine the agreement among the radon
sirens, electrets, stationary alphaguards, and (for 1 week) charcoal canisters. The sirens can be read once
each week, so their readings were taken when the other data types were being acquired and their readings
compared.
3.5.2 Subslab and Soil Gas Radon Sampling and Analysis
Radon readings were collected approximately weekly (and daily during intensive rounds) with a portable
Alphaguard Professional Radon Monitor from Genitron instruments. Operations were based on EPA
guidelines for using continuous radon monitors (U.S. EPA, 1992). More information on the Alphaguard
can be found at www.genitron.de/products/products .html. During normal weekly sampling, this device
was connected to subslab, soil gas, and wall ports with an SKC Universal XR pump set to 1 L/min. Tubes
connected the sample port to the pump (with a moisture filter on the sampling end) and the pump to the
Alphaguard. A purge line led away from the operator for exterior sampling and out of basement windows
for interior sampling locations. The Alphaguard requires a 10-minute cycle of uninterrupted air flow from
the sample location for an accurate reading. Because a certain amount of time was needed for movement
between, one 10-minute cycle was spent relocating and then another to sample at the next location. Thus,
each sample port needed 20 minutes to sample.
3-28
-------
Section 3—Methods
Because radon has a short half-life (3.8 days) and the migration time from substantial depths for soil gas
is estimated to be months to years (Kurtz and Folkes, 2008; Carr et al., 2011), radon sampling focused on
the shallowest depths and, therefore, differed from the VOC sampling strategy. Exterior sampling
consisted of the shallowest soil gas ports available of the wells closest to the house. Usually, these ports
were the 3.5 and 6-ft deep ports of SGP1, 7, 4, and 5. Periodically, these depths would not yield a sample,
presumably due to moisture infiltration. In such cases, the next shallowest depths were chosen. Routine
interior sampling included all wall ports, five of the SSPs, and the shallowest intervals of the nested
interior SGPs. When sampling during the intensive weeks, all locations were sampled multiple times, and
some locations were sampled more than once per day.
For normal weekly sampling, first an ambient reading was taken outdoors and approximately 20 ft away
from the 422/420 house. Then, lines to be sampled would be purged with the SKC pump (five soil gas
point volumes, calculated based on the depth). Finally, the pump would be connected to the Alphaguard
to acquire a full 10-minute sample.
The Alphaguard has a readout screen that details the results of the analysis at the end of each 10-minute
cycle. The data provided are Rn (Bq/m3), relative humidity (%), pressure (mbar), and temperature (° C).
These data were recorded each week in a spreadsheet and the Bq/m3 converted to pCi/L.
3.5.3 Continuous (Real-Time) Indoor Air Radon Sampling and Analysis
The real-time Alphaguards are essentially the same as the hand-held Alphaguard instrument used to
sample from the SGPs, except they are not fitted with the same nozzle type, because they are not
connected to external pumps. Rather, in this application they are operated in a diffusion mode. These
Alphaguards are intended to be placed to give readings in specific rooms. In the case of the 422/420
duplex, one unit was placed in the 422 second floor office, and the other was placed in the 422 North
basement area. These units stayed in their locations, except for brief, periodic data downloadings. These
units were first regularly deployed on March 31, 2011, and were in near-continuous operation for 1 year.
The data are produced by the instrument in the same units as the portable Alphaguard (requiring
conversion to pCi/L) and data points are collected every 10 minutes. However, because these devices
were not moved, all 10-minute cycles are usable. The real-time Alphaguards are used in conjunction with
Data Expert software, also from Genitron Instruments. Once each week, the Alphaguards were connected
to the computer (the one in the basement required briefly moving the instrument to download), and the
software downloaded the readings for the week. These data were then saved as text files for later
conversion to Excel spreadsheet files.
3.6 Physical Parameters Monitoring
3.6.1 On-Site Weather Station
This project used a Davis Vantage Vue Weather Station on-site with Weather Link data logger and
software (Figure 3-25). The components consist of the outdoor monitoring unit, the indoor receiver, and
the computer connection. The outdoor monitoring unit was mounted on an accessible portion of the
422/420 house roof. The unit was mounted on steel pipes, but 5 ft above the highest roof deck (that of the
attic dormer).
The outdoor unit contains all the exterior monitoring equipment (for example, wind speed cups, rain
gauge) and has a solar panel/battery backup for power. The outdoor unit transmits a radio signal to the
indoor receiver, which also records the data every half hour. The indoor unit is human readable and can
also be used to set a variety of parameters. The indoor unit also records the house interior data at its
location, in this case the 422 second floor office. Once each week, the data were downloaded from the
3-29
-------
Section 3—Methods
indoor unit onto the computer containing the Weather Link software. These data were saved as a text file
and later compiled in an Excel spreadsheet file. Many parameters are recorded; the key ones required for
this project are temperature (° F, interior and exterior), relative humidity (percent), wind speed (miles per
hour), and wind direction (16 points [22.5 degrees] on compass rose).
Initially, and at least every 6 months, the results from this on-site system were compared with other
nearby weather stations in Indianapolis using at least 1 day's observations. The National Weather Service
(NWS) Indianapolis International Airport (KIND) is approximately 15 miles southwest from site. The
Indianapolis NWS station at Eagle Creek Airpark (KEYE) is approximately 9 miles west of the site.
There is also a private weather station available online closer to the site in Indianapolis, IN
(KININDIA33).
Figure 3-25. Front view of 420/422 duplex with location of weather station sensors
indicated with red arrow.
3-30
-------
Section 3—Methods
3.6.2 Indoor Temperature
Although the indoor weather station unit can record temperature, it only does so in the 422 second floor
office where it is located. Because temperature readings were required at all sample locations to allow
adjustment of the passive sampler data for uptake rate variation due to temperature, another form of data
collection was necessary. HOBOs data loggers, made by Onset (http://www.onsetcomp.com/), were
placed one at each of the six passive sampler racks in the house. HOBOs record temperature (° F) and
relative humidity (percent) every 30 minutes. Once a week, these data were recorded by taking them to
the computer with the HOBOware reading software and later importing those data to an Excel
spreadsheet file. Special spreadsheets were created to provide this information for the different Radiello
time durations to the passive sampler analytical laboratory.
3.6.3 Soil Temperature
Soil temperature was recorded by thermocouples from Omega (Type T, Hermetically Sealed Tip
Insulated Thermocouples, HSTC-TT-T-24S-120). During the initial house setup, holes were drilled
beneath the basement slab and backyard soils of the duplex to accommodate thermocouple probes with
end points set at different depths. Wires were inserted in approximately 2-inch diameter holes with
weights loosely attached near the ends. The holes were allowed to cave in and backfill naturally. The
thermocouple wires run from their holes to male/female connectors (sealed from the elements in rubber
"boots") and from there to a data acquisition system (PDAQ 56 by lOtech), where the data were recorded
to the software on the computer. A reading was taken approximately every 15 minutes. The
thermocouples wired to the PDAQ roughly corresponded to the depths of the SGPs: inside at 6, 9, 13, and
16.5 ft bis; outside at 1, 3.5, 6, and 13 ft bis. However, there is one thermocouple (outside 16.5 ft) that is
wired into an Omega data logger (OM-EL-USB-TC). The thermocouple data were most typically
collected at 15-minute intervals.
3.6.4 Soil Moisture
Soil moisture was recorded by implanted Watermark moisture sensors. The units of measurement for the
soil moisture sensors are explained by Smaj stria and Harrison (2002):
Water potential is commonly measured in units of bars (and centibars in the English system
of measurement) or kilopascals (in metric units). One bar is approximately equal to one
atmosphere (14.7 lb/in2) of pressure. One centibar is equal to one kilopascal. Because
water is held by capillary forces within unsaturated soil pore spaces, its water potential is
negative, indicating that the water is under tension and that work must be done to extract
water from the soil. A water potential reading ofO indicates that the soil is saturated, and
plant roots may suffer from lack of oxygen. As the soil dries, water becomes less available
and the water potential becomes more negative. The negative sign is usually omitted for
convenience when soil water potentials are measured.
The soil water matrix potential can be converted into volumetric water content using known equations.
Moisture content is often measured in fixed laboratories as gravimetric water content. To convert
gravimetric water content to volumetric water, multiply the gravimetric water content by the bulk specific
gravity of the material.
These sensors were also installed in the holes drilled during the house setup. Before insertion, the sensors
had to be presoaked in water to prepare them. The sensors are pill-shaped devices at the end of a wire.
The wire was run up through a PVC pipe of the appropriate length for the depth and the wire grasped
manually. The sensor could then be placed to the appropriate depth within the hole, the PVC pipe
withdrawn, and the soil backfill allowed to fill in naturally. Wires extend to the Watermark 900M
monitor, which reads and records the data every 30 minutes. Once each week these data were downloaded
3-31
-------
Section 3—Methods
to the Watergraph 3.1 software on the computer. Data were recorded in centibars. The sensors were
installed to approximately correspond to the SGP depths: inside at 6, 13, and 16.5 ft bis and outside at 3.5,
6,9, 13, and 16.5 ft bis.
3.6.5 Potentiometric Surface/Water Levels
Water levels in the seven wells on-site (two outdoor three-well clusters and a single well in the basement)
were taken periodically with a Solinst water-level meter. The water-level results were compared against
U.S. Geological Survey (USGS) stream gauge data for Fall Creek at Millersville, site 03352500 near the
house.
3.6.6 Differential Pressure
Differential pressure readings were monitored by Setra Model 264 low differential pressure transducer.
These units contain a pressure-sensitive diaphragm that measures pressure changes from the exterior
high/low poles. The poles had tubing connected that ran from the areas to be measured. Some Setra poles
were left open as an interior reference at a particular location. The configurations on the 422 side were as
follows: subslab versus basement, basement versus upstairs, deep soil gas versus shallow soil gas, and
basement versus exterior (out of the basement window). Only one unit was located on the 420 side, and it
was used for subslab versus basement. Three lines used SGPs as access points: 422 deep soil gas versus
shallow soil gas used SGP8-6 and SGP8-13; 422 subslab versus basement used SSP-1; and 420 subslab
versus basement used SGP11-9. When these locations had to be sampled for VOCs, the ports would be
closed, disconnected from the Setras, purged, and sampled. Afterward, the ports would be reconnected to
the Setras and opened again.
The four Setras on the 422 side of the house are wired into the Personal Measurement Device, PMD-
1208LS from Measurement Computing. The PMD is connected to the computer and uses TracerDaq
software. Readings are taken every 15 minutes. The one Setra on the 420 side is connected to the PDAQ
device and also takes a reading every 15 minutes (but not necessarily the same 15-minute interval as the
PMD Setras).
In the beginning of the project, the Setras were laid flat on their supporting surfaces. In February 2011,
manufacturer's guidance was found indicating that they should be mounted vertically. The manufacturer
stated that correcting for the different mounting could be done by blocking the poles in the horizontal
position to determine their "zero readings" and then record those same readings in the vertical position to
determine the offset. The offset could then be factored in to change the horizontal position data to
vertical. By March 31, 2011, all were hung in this manner, and the early data corrected.
3.6.7 Air Exchange Rate
To determine the air exchange rate, capillary adsorption tubes (CATs) were used in conjunction with
para-dimethylcyclohexane (PDCH) and para-methylcyclohexane (PMCH) emitters, provided by the
Harvard School of Public Health (HSPH) (EPA Method IP-4). The emitters are small metal shells
containing a fluid (either PDCH or PMCH), and the shells are contained within a foam wrapping. The
fluid releases a tracer gas at a measured constant rate, which is picked up by the CATs when in place. One
stopper end of the CAT is removed when the samplers were deployed for periods of 1 week to allow
sampling of the tracer gas by the adsorbent medium.
On April 22, 2011, in the 422 side of the house, 10 of the PDCH emitters were placed in the basement, 10
PMCH emitters were placed on the first floor, and nine PMCH emitters were placed on the second floor.
Care was taken that emitters be placed far enough from each other and from walls (about 3 to 4 ft). The
placement locations also allowed unrestricted air flow.
3-32
-------
Section 3—Methods
CATs were used for sampling for air exchange rate measurement on two occasions. The first was from
April 27, 2011, to May 4, 2011, and the second was from September 23, 2011, to September 29, 2011. On
the first occasion, CATs were deployed: one on the 422 first floor (center room) and two in the 422
basement (one duplicate). One was also placed in 420 on the first floor (center room) and in the 420
basement (center room). On the second occasion, CATs were only deployed on the 422 side of the house.
One was in the 422 office on the second floor, one on the first floor (center room), and two were placed in
the basement center room (one duplicate). When sampling, CATs were placed on their sides with one cap
removed and slightly tipped at one end so the open end pointed toward the ground. After sampling, the
CATs were sealed and sent to HSPH for analysis.
3.6.8 Crack Monitoring
The basement floors and walls were visually inspected for significant cracks (i.e., ones where vapors
could infiltrate from subsurface soils). For the three most significant cracks, a calibrated crack monitor
(Figure 3-26) was installed across the crack. This device consists of two plates that move independently.
One plate is white with a black millimeter grid; the other is transparent with red crosshairs centered over
the grid. When the monitor is secured with epoxy or screws across a crack, the crosshairs shift vertically
or horizontally on the grid, making crack movement easily visible and trackable. It was installed with a 5-
Minute® Epoxy, a rapid-curing, general-purpose adhesive that bonds rigid, durable substrates such as
metals, glass, ceramics, concrete, and wood in all combinations. The position of the monitor was recorded
monthly and indicated that the monitored cracks did not move during the course of the study.
BEN MEADOWS Calibrated Crack Monitor-
Figure 3-26. Calibrated crack monitor.
3-33
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table of Contents
4. Results and Discussion: Quality Assurance Checks of Individual Data Sets 4-1
4.1 VOC Sampling—Indoor Air-Passive—Air Toxics Ltd. (ATL) 4-1
4.1.1 Blanks 4-1
4.1.2 Surrogate Recoveries 4-3
4.1.3 Laboratory Control Sample Recoveries 4-3
4.1.4 Duplicates 4-3
4.2 VOC Sampling—Subslab and Soil Gas (TO-17)—U.S. EPA 4-4
4.2.1 Blanks 4-4
4.2.2 Surrogate Recoveries 4-7
4.2.3 Laboratory Control Sample Recoveries 4-7
4.3 VOC Sampling—Subslab and Soil Gas (TO-17)—ATL 4-8
4.3.1 Blanks 4-8
4.3.2 Surrogate Recoveries 4-10
4.3.3 Laboratory Control Sample Recoveries 4-11
4.3.4 Duplicates 4-11
4.4 VOC Sampling—Subslab and Indoor Air (TO-15)—ATL 4-12
4.4.1 Blanks 4-12
4.4.2 Surrogate Recoveries 4-13
4.4.3 Laboratory Control Sample Recoveries 4-13
4.4.4 Duplicates 4-14
4.5 Online GC (Soil Gas and Indoor Air) 4-14
4.5.1 Blanks 4-14
4.5.2 Initial Calibration 4-15
4.5.3 Continuing Calibration 4-15
4.5.4 Calibration Check via Comparison to Fixed Laboratory (TO-15 vs. Online
GC) 4-15
4.5.5 Agreement of Online GC Results with TO-17 Verification Samples 4-16
4.5.6 Agreement of Integrated Online GC Results with Passive Samplers 4-18
4.6 Radon 4-25
4.6.1 Indoor Air: Comparison of Electrets Field, ARCADIS to Charcoal Analyzed
by U.S. EPA R&IE National Laboratory 4-25
4.6.2 Comparision of Average of Real Time Alphaguard to Electrets and Charcoal
Canisters 4-28
4.6.3 Quality Assurance Checks of Electrets 4-30
4.7 On-Site Weather Station vs. National Weather Service (NWS) 4-31
4.8 Database 4-33
4.8.1 Checks on Laboratory Reports 4-33
4.8.2 Database Checks 4-34
List of Figures
4-1. Correlation between radon measured using the electret and charcoal methods 4-27
4-2. Aerial view of study house, showing potential influences on wind velocity, red arrow
indicates study house 4-32
4-3. Comparison of National Weather Service Indianapolis temperature data to weather
station at 422 East 28th Street 4-32
4-4. Comparison of National Weather Service Indianapolis relative humidity to weather
station at 422 East 28th Street 4-33
4-i
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4-5. Comparison of National Weather Service wind speed data to weather station at 422 East
28th Street 4-33
List of Tables
4-1. Indoor Air Passive Field Blank Summary—Radiello 130 4-1
4-2. Indoor Air Passive Trip Blank Summary—Radiello 130 4-2
4-3. Indoor Air Passive Laboratory Blank Summary—Radiello 130 4-2
4-4. Indoor Air Passive Surrogate Summary—Radiello 130 4-3
4-5. Indoor Air Passive LCS Summary—Radiello 130 4-3
4-6. Indoor Air Passive Laboratory Precision (LCS/LCSD) Summary—Radiello 130 4-4
4-7. Subslab and Soil Gas—EPA Field Blank Summary—TO-17 4-5
4-8. Subslab and Soil Gas—EPA Trip Blank Summary—TO-17 4-6
4-9. Subslab and Soil Gas—EPA Laboratory Blank Summary—TO-17 4-6
4-10. Subslab and Soil Gas—EPA Fridge Blank Summary—TO-17 4-7
4-11. EPA TO-17 Surrogate Recovery Summary 4-7
4-12. EPA TO-17 LCS Summary 4-8
4-13. Subslab and Soil Gas—ATL Field Blank Summary—TO-17 4-9
4-14. Subslab and Soil Gas—ATL Trip Blank Summary—TO-17 4-10
4-15. Subslab and Soil Gas—ATL Lab Blank Summary—TO-17 4-10
4-16. ATL TO-17 Surrogate Recovery Summary 4-11
4-17. ATL TO-17 LCS Summary 4-11
4-18. ATL TO-17 Laboratory Precision (LCS/LCSD) Summary 4-12
4-19. Subslab and Indoor Air—ATL Lab Blank Summary—TO-15 4-13
4-20. ATL TO-15 Surrogate Recovery Summary 4-13
4-21. ATL TO-15 LCS Summary 4-14
4-22. ATL TO-15 Laboratory Precision (LCS/LCSD) Summary 4-14
4-23. Interlaboratory Results: Spiked Verification Samples 4-17
4-24. Interlaboratory Statistics: Spiked Verification Samples 4-18
4-25. Comparison of Online GC to Radiello Results 4-19
4-26. Comparison between Electrets and Charcoal Canisters at the 422/420 EPA House from
January 19-26, 2011 4-26
4-27. Comparison of Electret and Charcoal Canister Data from April 27, 2011, to May 4, 2011 4-26
4-28. Comparison of Charcoal and Electret Radon December 28, 2011, to January 4, 2012 4-27
4-29. Comparison between 422 Base N Alphaguards and Electrets from March 30, 2011, and
May 18, 2011 4-28
4-30. Comparison of Real-Time Alphaguard to Integrated Electret August through October 4-28
4-31. Comparison of Real-time Alphaguards to Integrated Electret Measurements December
28, 2011 to January 4, 2012 4-29
4-32. Comparison of Real-time Alphaguard to Integrated Electret Measurements January
through March 4-30
4-ii
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4. Results and Discussion: Quality Assurance Checks of Individual
Data Sets
4.1 VOC Sampling—Indoor Air-Passive—Air Toxics Ltd. (ATL)
4.1.1 Blanks
Field blanks, trip blanks, and laboratory blanks were used to evaluate false positives and/or high bias due
to transport, storage, sample handling, and sorbent contamination. Field blanks were collected using a
blank Radiello 130 cartridge from the media sample batch sent to the field from the laboratory. The
cartridge was removed from the sealed storage vial and transferred to the diffusive housing in a similar
manner to sample deployment. The cartridge was then immediately removed from the housing, returned
to the storage vial, and sealed for shipment back to the laboratory with the field samples. In general, a
field blank was collected with each shipment to the laboratory. A total of 47 field blanks were submitted
over the duration of the project.
Blank Radiello cartridges from the media batches were also assigned as trip blanks. The cartridge was not
opened or removed from the storage vial but was sent back to the laboratory along with the field samples.
There were 22 trip blanks submitted for analysis.
In the case of the laboratory blank, a Radiello 130 cartridge was extracted with each analytical batch to
measure background from the sorbent and the extraction process. A total of 72 unique laboratory blanks
were analyzed and reported over the duration of the project.
To assist in data interpretation, all blank samples and all field sample results were evaluated down to the
method detection limit (MDL). The results of the field, trip, and laboratory blanks are summarized in
Tables 4-1, 4-2, and 4-3. The number of blanks with detections above the reporting limit (RL) and MDL
are tabulated. Summary statistics were then calculated on this subset of positive detections.
Table 4-1. Indoor Air Passive Field Blank Summary—Radiello 130
Benzene
Chloroform
cis-1,2-DCE
Hexane
PCE
Toluene
TCE
RL(Mg)
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Number of Field Blanks
Analyzed
47
47
47
47
47
47
47
Cone. >
RL
0
0
0
4
0
1
0
RL> Cone.
>MDL
38
0
0
9
2
21
5
% of Field
Blanks
with
Detections
81%
0%
0%
28%
9%
47%
11%
Mean
Blank
Cone.
(M9)
0.11
NA
NA
0.099
0.032
0.040
0.015
Std Dev
(M9)
0.042
NA
NA
0.091
0.020
0.036
0.0093
Min
(M9)
0.040
NA
NA
0.033
0.0067
0.014
0.0064
Max
(M9)
0.18
NA
NA
0.35
0.049
0.17
0.031
NA = Not applicable
Benzene was detected above the MDL but below the RL in a majority of the field, trip, and lab blanks at
similar background levels. The average of the positive detections was 0.11, 0.10, and 0.12 (ig for the
field, trip, and lab blanks, respectively. The benzene blank levels are largely due to benzene
contamination present in the carbon disulfide extraction solvent. Although the laboratory used high purity
(99.99%) carbon disulfide reagent, benzene is present as a common contaminant in this solvent.
4-1
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-2. Indoor Air Passive Trip Blank Summary—Radiello 130
Benzene
Chloroform
cis-1,2-DCE
Hexane
PCE
Toluene
TCE
RL(ug)
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Number of Trip Blanks
Analyzed
22
22
22
22
22
22
22
Cone. >
RL
0
0
0
0
0
0
0
RL> Cone.
>MDL
20
0
0
10
2
18
4
% of Trip
Blanks
with
Detections
91%
0%
0%
45%
9%
82%
18%
Mean
Blank
Cone.
(M9)
0.10
NA
NA
0.049
0.015
0.020
0.024
Std Dev
(M9)
0.039
NA
NA
0.012
0.009
0.008
0.0159
Min
(M9)
0.042
NA
NA
0.036
0.0087
0.012
0.0094
Max
(M9)
0.16
NA
NA
0.07
0.022
0.041
0.043
NA = Not applicable
Table 4-3. Indoor Air Passive Laboratory Blank Summary—Radiello 130
Benzene
Chloroform
cis-1,2-DCE
Hexane
PCE
Toluene
TCE
RL(ug)
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Number of Lab Blanks
Analyzed
73
73
73
73
73
73
73
Cone. >
RL
0
0
0
0
0
0
0
RL> Cone.
>MDL
67
0
0
18
2
52
4
% of Lab
Blanks
with
Detections
92%
0%
0%
25%
3%
71%
6%
Mean
Blank
Cone.
(ug)
0.12
NA
NA
0.053
0.0081
0.025
0.022
Std Dev
(ug)
0.043
NA
NA
0.019
0.00042
0.014
0.0068
Min
(ug)
0.039
NA
NA
0.034
0.0078
0.012
0.013
Max
(ug)
0.22
NA
NA
0.083
0.0084
0.064
0.027
NA = Not applicable
Although the benzene background levels were below the RL, a positive bias is expected for the daily
Radiello and a large subset of the weekly indoor air samples. Longer duration samples would normally
collect more mass and, thus, would not be significantly affected. The mass of benzene adsorbed by the
sorbent cartridge over 1 day averaged 0.19 (ig, similar to the levels detected in the blanks. For the weekly
samples, the average mass measured on the cartridge was 0.62 (ig; however, approximately half of the
weekly samples contained benzene levels that were less than 5 times the blank levels. Sample deployment
times greater than a week demonstrated less positive bias from the blank because proportionally more
benzene mass was collected by the diffusive sampler from the indoor air environment.
Hexane and toluene were also commonly detected in the field, trip, and lab blanks above the MDL. In the
case of the field blanks, several had concentrations above the RL for hexane and toluene. All detections in
the trip and lab blanks were below the RL but above the MDL. Similar to benzene, a positive bias for
hexane and toluene is anticipated for the daily Radiello samples due to the blank levels. The average mass
collected on the sorbent for the daily passive samples was 0.11 and 0.19 (ig for hexane and toluene,
respectively. A positive bias is expected for hexane for the weekly samples as well with average sample
mass collected of 0.44 (ig. Blank levels of toluene are not significant when evaluating the weekly samples
because the mass collected is generally greater than 10 times blank levels. Longer duration samples would
normally collect more mass and thus would not be significantly affected.
4-2
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
No detections of chloroform or cis-l,2-dichloroethene (cis-l,2-DCE) were measured in any of the blanks.
For a small percentage of the blanks, low concentrations detections above the MDL were measured for
tetrachloroethene (PCE) and trichloroethene (TCE).
4.1.2 Surrogate Recoveries
To monitor extraction efficiency, 5.0 (ig of toluene-d8 was spiked into each field sample and quality
control (QC) sample Radiello 130 cartridge immediately prior to extraction. The recoveries were
evaluated against laboratory limits of 70 to 130%. All surrogate recoveries met the laboratory criterion,
and summary statistics are presented in Table 4-4.
Table 4-4. Indoor Air Passive Surrogate Summary—Radiello 130
Parameter
Number of surrogate recoveries measured
Average recovery (%R)
Standard deviation (%R)
Minimum recovery (%R)
Maximum recovery (%R)
Result
1,255
102.8
5.9
87
122
4.1.3 Laboratory Control Sample Recoveries
Accuracy of the extraction and analysis step for the target compounds was evaluated by analyzing a
laboratory control sample (LCS). An unused Radiello cartridge was spiked with a standard containing
5.0 (ig of each compound of interest. The laboratory acceptance criterion for LCS recovery was 70 to
130%. All LCS recoveries met the control limits of 70 to 130%, and summary statistics are presented in
Table 4-5.
Table 4-5. Indoor Air Passive LCS Summary—Radiello 130
Benzene
Chloroform
cis-1,2-DCE
Hexane
PCE
Toluene
TCE
Number of LCS
Analyzed
73
73
73
73
73
73
73
Mean LCS %
Recovery
93
96
95
101
98
94
97
LCS Std
Dev (%R)
11.0
11.5
8.7
14.6
9.8
9.8
8.6
Min
(%R)
71
70
72
71
73
73
73
Max (%R)
116
122
121
130
125
117
118
4.1.4 Duplicates
Sample precision was evaluated by collecting field duplicates and by analyzing laboratory control sample
duplicates (LCSDs). Field duplicates were collected for approximately every 10 field samples, and an
LCSD was prepared and analyzed with each sample preparation batch. Because the LCSD was a second
cartridge prepared and extracted in the same manner as the LCS, the relative percentage difference
(%RPD) represents the precision of the analytical method from extraction through analysis. The method
precision is summarized in Table 4-6. The laboratory acceptance criterion of %RPD < 25% was met for
all compounds except for benzene in two analytical batches and hexane in five analytical batches.
4-3
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-6. Indoor Air Passive Laboratory Precision (LCS/LCSD) Summary—Radiello 130
Benzene
Chloroform
cis-1,2-DCE
Hexane
PCE
Toluene
TCE
Number of
LCSD
Analyzed
73
73
73
73
73
73
73
Mean
%RPD
9%
10%
5%
11%
5%
5%
5%
Std Dev.
(%RPD)
8%
7%
4%
9%
4%
4%
4%
Min (%RPD)
0%
0%
0%
0%
0%
0%
0%
Max (%RPD)
29%
25%
19%
37%
19%
19%
14%
Number of
Exceedances
2
0
0
5
0
0
0
4.2 VOC Sampling— Subslab and Soil Gas (TO-17)—U.S. EPA
4.2.1 Blanks
Field, trip, refrigerator, and laboratory blanks were used to evaluate false positives and/or high bias due to
transport, storage, sample handling, and sorbent contamination. Field blanks were collected using a blank
Tenax TA TO-17 sorbent tube from the media sample batch sent to the field from the laboratory. The
Swagelok end caps were removed as if to prepare for sample collection; however, no soil vapor was
pulled through the tube. The end caps were immediately replaced, and the tube was sent back to the
laboratory with the field samples. Typically, a field blank was collected with each shipment to the
laboratory. A total of 98 field blanks were submitted over the duration of the project.
Blank Tenax TA TO-17 sorbent tubes from the media batches were also assigned as trip blanks. The tube
remained capped and wrapped in aluminum foil and was sent from the laboratory to the field and back to
the laboratory along with the field samples. There were 85 trip blanks submitted for analysis.
In the case of the laboratory blank, a Tenax TA TO-17 tube was analyzed with each analytical batch to
measure background from the sorbent tubes and instrumentation. A total of 251 lab blanks were analyzed
and reported over the duration of the project.
For a refrigerator (fridge) blank, a Tenax TA TO-17 tube was stored and analyzed with each sample batch
to measure background from the sample storage refrigerator. The tubes were stored in the refrigerator,
capped, and sealed in a zippered bag on top of the jars containing the samples that were received as a
batch. The fridge blanks were placed in the refrigerator with a sample batch and remained in the
refrigerator with the batch until all the samples from that batch had been analyzed. Thus, the fridge blanks
were in the refrigerator longer than some of the samples within a batch. A total of 48 fridge blanks were
analyzed and reported over the duration of the project.
To assist in data interpretation, all blank samples and all field sample results were evaluated down to the
MDL. The results of the field, trip, laboratory, and fridge blanks are summarized in Tables 4-7, 4-8, 4-9,
and 4-10. The number of blanks with detections above the RL and MDL are tabulated. Summary statistics
were then calculated on this subset of positive detections.
Benzene was detected above the MDL in 54%, 42%, 39%, and 46% of the field, trip, laboratory, and
fridge blanks, respectively. The average of the positive detections was 1.4, 1.2, 1.7, and 1.2 nanogram
(ng) for the field, trip, lab, and fridge blanks, respectively. Six laboratory blanks had benzene
concentrations above the RL of 5.0 ng. The benzene blank levels are largely due to background
4-4
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
contribution from the Tenax TA polymer, which can break down during the heating step to generate low
levels of benzene.
The concentrations of benzene in the TO-17 soil vapor samples were similar in magnitude to those
measured in the field blanks. Of the 2,270 TO-17 soil vapor samples analyzed by EPA, 73% of the
samples had a positive detection of benzene. Of the samples that had a positive detection for benzene,
only 2% had a detected concentration above the RL of 5.0 ng. The second most common contaminant in
these blank samples was toluene, which has also been reported as a Tenax breakdown product (MacLeod
and Ames, 1986; Cao and Hewitt, 1994).
Detections of the key compounds that form the focus of this work—PCE, chloroform, and TCE—
occurred in 6% of the samples or less of the hundreds of the field trip and lab blanks analyzed. However,
the percentage of refrigerator blanks with PCE and TCE contamination was considerably higher—19%.
Table 4-7. Subslab and Soil Gas—EPA Field Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
RL
(ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
MDL
(ng)
0.87
0.48
0.76
0.85
0.44
0.60
1.3
1.2
1.2
Number of Field Blanks
Analyzed
98
98
98
98
98
98
98
98
98
Cone.
>RL
0
0
5
0
0
0
1
0
3
RL>
Cone.
> MDL
53
9
0
1
2
4
1
11
0
% of Field
Blanks with
Detections
54
9
5
1
2
4
1
11
3
Mean
Blank
Cone.
(ng)
1.4
3.4
72
1.5
1.6
6.4
2.8
1.3
14
Std. Dev.
(ng)
0.5
1.4
110
NA
1.5
3.8
NA
0.3
1.1
Min
(ng)
0.81
1.7
3.0
1.5
2.2
2.5
2.8
1.1
13
Max
(ng)
3.0
6.4
260
1.5
4.4
11
2.8
1.9
16
NA = Not applicable
4-5
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-8. Subslab and Soil Gas—EPA Trip Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
RL
(ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
MDL
(ng)
0.87
0.48
0.76
0.85
0.44
0.60
1.3
1.2
1.2
Number of Trip Blanks
Analyzed
85
85
85
85
85
85
85
85
85
Cone.
>RL
0
0
4
0
0
0
4
3
2
RL>
Cone. >
MDL
36
9
1
0
2
4
0
13
0
% of Trip
Blanks
with
Detections
42
11
6
0
2
5
5
19
2
Mean
Blank
Cone, (ng)
1.2
2.6
42
0
2.0
2.8
18
3.6
3.7
Std.
Dev.
(ng)
0.5
0.8
45
0
1.5
0.8
11
4.8
2.0
Min
(ng)
0.81
1.6
2.0
0
1.0
2.2
2.3
1.0
2.3
Max
(ng)
2.6
4.0
120
0
3.0
4.0
27
19
5.2
Table 4-9. Subslab and Soil Gas—EPA Laboratory Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
RL
(ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
MDL
(ng)
0.87
0.48
0.76
0.85
0.44
0.60
1.3
1.2
1.2
Number of Lab Blanks
Analyzed
251
251
251
251
251
251
251
251
251
Cone.
>RL
6
4
7
0
0
0
0
2
3
RL>
Cone. >
MDL
92
42
2
0
4
6
1
29
2
% of Lab
Blanks with
Detections
39
18
4
0
2
2
0.4
12
2
Mean
Blank
Cone.
(ng)
1.7
9.6
2.1
0
4.8
3.0
1.8
7.4
7.4
Std.
Dev.
(ng)
1.6
9.2
0.33
0
2.9
1.3
NA
6.0
6.0
Min
(ng)
0.80
0.87
1.3
0
1.8
5.6
1.8
1.0
1.4
Max (ng)
12
52
2.5
0
8.7
2.4
1.8
8.1
16
NA= Not applicable
4-6
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-10. Subslab and Soil Gas—EPA Fridge Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
RL
(ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
MDL
(ng)
0.87
0.48
0.76
0.85
0.44
0.60
1.3
1.2
1.2
Number of Fridge Blanks
Analyzed
48
48
48
48
48
48
48
48
48
Cone. >
RL
0
0
2
0
0
0
6
4
8
RL>Conc
.> MDL
22
2
0
0
3
4
3
8
1
% of Fridge
Blanks with
Detections
46
4
4
0
6
8
19
25
19
Mean
Blank
Cone.
(ng)
1.2
2.3
2.3
0
1.0
6.1
3.7
10
7.4
Std.
Dev.
(ng)
0.40
0.69
0.29
0
0.08
7.5
1.6
23
4.6
Min
(ng)
0.81
1.8
2.1
0
0.88
1.8
2.0
0.96
1.5
Max
(ng)
1.8
2.8
2.5
0
1.0
17
3.5
82
17
4.2.2 Surrogate Recoveries
To monitor analytical efficiency, 5.3 ng of bromochloromethane were loaded onto each QC and field
sample sorbent tube along with the vapor phase internal standard mix during sample analysis. Field
surrogates were not included in the scope of this project. The recoveries were evaluated against laboratory
limits of 70 to 130%. Most surrogate recoveries met the laboratory criterion, and summary statistics are
presented in Table 4-11.
Table 4-11. EPA TO-17 Surrogate Recovery Summary
Parameter
Number of surrogate recoveries measured
Average recovery (%R)
Standard deviation (%R)
Minimum recovery (%R)
Maximum recovery (%R)
Result
3,370
105
14
27
354
4.2.3 Laboratory Control Sample Recoveries
Analytical accuracy was evaluated by analyzing an LCS. Two clean Tenax TA TO-17 sorbent tubes were
spiked with a calibration standard from a source independent from the primary calibration standard and
analyzed after each initial calibration. The spike contained approximately 100 ng of each target
compound. The performance of the EPA TO-17 LCS spikes is summarized in Table 4-12. A total of 10
LCS samples were evaluated, and all met the laboratory RLs with the exceptions of five outliers for
carbon disulfide, four outliers for methylene chloride, and one outlier for cis-l,2-DCE.
4-7
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-12. EPA TO-17 LCS Summary
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
Number
of LCS
Analyzed
10
10
10
10
10
10
10
10
10
Mean LCS
%
Recovery
101
117
96
105
98
111
85
102
100
LCS Std
Dev (%R)
11
64
11
10
11
71
8.1
13
12
Min (%R)
86
24
82
96
72
29
71
80
80
Max (%R)
118
272
122
133
120
291
97
128
120
LCS
Recovery
Limits
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
Number of
Exceedances
0
5
0
1
0
4
0
0
0
4.3 VOC Sampling— Subslab and Soil Gas (TO-17)—A TL
4.3.1 Blanks
Field blanks, trip blanks, and laboratory blanks were used to evaluate false positives and/or high bias due
to transport, storage, sample handling, and sorbent contamination. Field blanks were collected using a
blank Tenax TA TO-17 sorbent tube from the media sample batch sent to the field from the laboratory.
The Swagelok end caps were removed to prepare for sample collection; however, no soil vapor was
pulled through the tube. The end caps were immediately replaced, and the tube was sent back to the
laboratory with the field samples. Typically, a field blank was collected with each shipment to the
laboratory. A total of 18 field blanks were submitted over the duration of the project.
Blank Tenax TA TO-17 sorbent tubes from the media batches were also assigned as trip blanks. The tube
remained capped and wrapped in aluminum foil and was sent back to the laboratory along with the field
samples. There were five trip blanks submitted for analysis.
In the case of the laboratory blank, a Tenax TA TO-17 tube was analyzed with each analytical batch to
measure background from the sorbent tubes and instrumentation. A total of 26 lab blanks were analyzed
and reported over the duration of the project.
To assist in data interpretation, all blank samples and all field sample results were evaluated down to the
MDL. The results of the field, trip, and laboratory blanks are summarized in Tables 4-13, 4-14, and 4-15.
The number of blanks with detections above the RL and MDL are tabulated. Summary statistics were
then calculated on this subset of positive detections.
Benzene was detected above the MDL in all of the field blanks and a majority of the trip and lab blanks.
The average of the positive detections was 3.0, 3.4, and 2.1 ng for the field, trip, and lab blanks,
respectively. Two field blanks had benzene concentrations above the RL of 5.0 ng. The benzene blank
levels are largely due to background contribution from the Tenax TA polymer, which can break down
during the heating step to generate low levels of benzene.
The concentrations of benzene in the TO-17 soil vapor samples were similar in magnitude to those
measured in the field blanks. Of the 382 TO-17 soil vapor samples analyzed by ATL, 93% of the samples
4-8
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
had a positive detection of benzene. The average benzene sample mass measured was 3.6 ng with a
standard deviation of 2.1 ng, a minimum concentration of 1.2 ng, and a maximum concentration of 15 ng.
Methylene chloride was detected at levels above the RL for a set of five field blanks collected on
December 15, 2011. These field blanks had concentrations above the RL of 50 ng with an average
concentration of 90 ng and concentrations ranging from 68 to 130 ng. Similar levels of methylene
chloride were measured in the samples collected over December 11, 2011, to December 15, 2011, and a
positive bias for these sets of samples is expected.
For the chlorinated compounds of concern (chloroform, cis-l,2-DCE, PCE, and TCE), the mean blank
concentrations were typically less than one-half the RL, and sample detections above the RL are not
expected to exhibit a significant positive bias based on the blank levels.
Table 4-13. Subslab and Soil Gas—ATL Field Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene
chloride
PCE
Toluene
TCE
RL (ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
Number of Field Blanks
Analyzed
18
18
18
18
18
18
18
18
18
Cone. >
RL
2
2
0
1
1
5
0
2
0
RL>
Cone. >
MDL
16
7
14
0
6
0*
0
4
11
% of Field
Blanks
with
Detections
100%
50%
78%
6%
39%
28%
0%
33%
61%
Mean
Blank
Cone.
(ng)
3.0
5.5
0.76
2.8
4.3
90
NA
3.7
0.44
Std
Dev
(ng)
1.6
5.0
0.50
NA
4.1
26
NA
2.4
0.090
Min
(ng)
1.5
1.4
0.37
2.8
0.78
68
NA
0.53
0.28
Max
(ng)
6.9
14
1.9
2.8
12
130
NA
6.6
0.60
NA = Not applicable
* Not all blanks were reported to the MDL.
4-9
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-14. Subslab and Soil Gas—ATL Trip Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene
chloride
PCE
Toluene
TCE
RL (ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
Number of Trip Blanks
Analyzed
5
5
5
5
5
5
5
5
5
Cone. >
RL
0
1
0
0
0
0
0
1
0
RL>
Cone. >
MDL
4
3
0
0
2
3*
1
1
2
% of Trip
Blanks
with
Detections
80%
80%
0%
0%
40%
60%
20%
40%
40%
Mean
Blank
Cone.
(ng)
3.4
2.6
NA
NA
1.3
14
0.73
4.2
0.88
Std
Dev
(ng)
0.86
2.5
NA
NA
0.42
6.2
NA
4.9
0.31
Min
(ng)
2.4
1.2
NA
NA
1.0
7.4
0.73
0.72
0.66
Max
(ng)
4.5
6.3
NA
NA
1.6
19
0.73
7.6
1.1
NA = Not applicable
* Not all blanks were reported to the MDL.
Table 4-15. Subslab and Soil Gas—ATL Lab Blank Summary—TO-17
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene
chloride
PCE
Toluene
TCE
RL (ng)
5.0
5.0
2.0
2.0
10
50
2.0
5.0
2.0
Number of Lab Blanks
Analyzed
26
26
26
26
26
26
26
26
26
Cone. >
RL
0
0
0
0
0
0
0
0
0
RL>
Cone. >
MDL
15
12
11
4
2
4*
2
4
7
% of Lab
Blanks
with
Detections
58%
46%
42%
15%
8%
15%
8%
15%
27%
Mean
Blank
Cone.
(ng)
2.1
2.1
0.56
0.66
0.90
3.5
0.34
0.62
0.48
Std
Dev
(ng)
0.46
0.81
0.29
0.50
0.42
4.3
0.20
0.46
0.073
Min
(ng)
1.4
1.0
0.34
0.35
0.60
1.3
0.20
0.32
0.35
Max
(ng)
2.8
3.9
1.4
1.4
1.2
9.9
0.48
1.3
0.57
NA = Not applicable
* Not all blanks were reported to the MDL.
4.3.2 Surrogate Recoveries
To monitor analytical efficiency, 36 ng of bromofluorobenzene were loaded onto each QC and field
sample sorbent tube along with the vapor phase internal standard mix during sample analysis. Field
surrogates were not included in the scope of this project. The recoveries were evaluated against laboratory
limits of 70 to 130%. All surrogate recoveries met the laboratory criterion, and summary statistics are
presented in Table 4-16.
4-10
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-16. ATL TO-17 Surrogate Recovery Summary
Parameter
Number of surrogate recoveries measured
Average recovery (%R)
Standard deviation (%R)
Minimum recovery (%R)
Maximum recovery (%R)
Result
510
100.6
7.5
77
119
4.3.3 Laboratory Control Sample Recoveries
Analytical accuracy was evaluated by analyzing an LCS. A clean Tenax TA TO-17 sorbent tube was
spiked with a calibration standard from a source independent from the primary calibration standard and
analyzed with each analytical batch. The spike typically contained approximately 50 to 150 ng of each
target compound. The performance of the ATL TO-17 LCS spikes is summarized in Table 4-17. A total
of 25 LCS samples were evaluated, and all met the laboratory RLs with the exception of one outlier for
benzene and carbon disulfide and two outliers for hexane.
Table 4-17. ATL TO-17 LCS Summary
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
Number of
LCS
Analyzed
25
25
25
25
25
25
25
25
25
Mean LCS
%
Recovery
82
110
85
91
107
97
81
81
83
LCS Std Dev
(%R)
7.2
22
5.7
6.4
15.4
15.6
8.5
8.8
7.1
Min (%R)
68
72
75
80
77
64
70
70
72
Max (%R)
97
157
96
105
136
127
103
100
96
LCS
Recovery
Limits
70-130%
50-150%
70-130%
70-130%
70-130%
50-150%
70-130%
70-130%
70-130%
Number of
Exceedances
1
1
0
0
2
0
0
0
0
4.3.4 Duplicates
Sample precision was evaluated by collecting field duplicates and by analyzing LCSDs. Field duplicates
were collected for approximately every 10 field samples, and an LCSD was analyzed with each sample
preparation batch. The LCSD was prepared by analyzing the spiked LCS sorbent tube a second time using
the recollection feature of the automated thermal desorption unit. As such, the LCSD provides both
verification of the re-collection step as well as an evaluation of instrument precision. The instrument
precision is summarized in Table 4-18. The laboratory acceptance criterion of %RPD <_20% was met for
all compounds.
4-11
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-18. ATL TO-17 Laboratory Precision (LCS/LCSD) Summary
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene
chloride
PCE
Toluene
TCE
Number of LCSD
Analyzed
25
25
25
25
25
25
25
25
25
Mean (%RPD)
4%
3%
2%
2%
2%
3%
1%
2%
1%
Std Dev. (%RPD)
2%
3%
2%
2%
2%
3%
2%
2%
1%
Min (%RPD)
1%
0%
0%
0%
0%
0%
0%
0%
0%
Max (%RPD)
8%
11%
8%
7%
7%
11%
5%
8%
4%
4.4 VOC Sampling— Subslab and Indoor Air (TO-15)—A TL
A total of 13 subslab and 13 ambient (indoor and outdoor) air samples were collected in Summa canisters
and analyzed by EPA Method TO-15. The subslab samples were analyzed using the laboratory standard
TO-15 method with base RLs of 0.5 to 2.0 ppbv. The ambient air samples were analyzed using a more
sensitive TO-15 instrument configuration (low-level) with base RLs of 0.1 to 0.5 ppbv.
4.4.1 Blanks
Laboratory blanks were used to evaluate false positives and/or high bias due to laboratory handling and
analysis. Lab blanks were prepared by filling a Summa canister with humidified ultra high purity (UHP)
nitrogen or zero air and analyzing in the same manner as the field samples. A total of four unique TO-15
lab blanks were analyzed. One was analyzed on the standard TO-15 unit along with the subslab samples.
Three were analyzed on the low-level TO-15 units along with the ambient samples. All lab blanks and
field sample results were evaluated down to the MDL. The results for the TO-15 lab blanks are
summarized in Table 4-19.
Detections above the MDL but below the RL were reported for carbon disulfide, PCE, and toluene in the
standard TO-15 lab blank. PCE and toluene were reported at less than one-half the RL, and carbon
disulfide was slightly higher than one-half the RL. Similar levels of carbon disulfide were detected in the
associated subslab samples. Sample detections of PCE and toluene above the RL are not expected to
exhibit a positive bias based on the lab blank concentrations.
In the case of the TO-15 low-level analysis, carbon disulfide was detected above the MDL in two of the
lab blanks and methylene chloride was detected above the MDL in one of the lab blanks. Associated
samples had similar concentrations above the MDL but below the RL.
4-12
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-19. Subslab and Indoor Air—ATL Lab Blank Summary—TO-15
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
Tetrachloroethene
Toluene
Trichloroethene
Reporting Limit (|jg/m3)
Standard
1.6
1.6
2.4
2.0
1.8
1.7
3.4
1.9
2.7
Low Level
0.32
1.6
0.49
0.40
0.35
0.69
0.68
0.38
0.54
Laboratory Blank Summary (|jg/m3)
Standard
<1.6
0.97
<2.4
<2.0
<1.8
<1.7
0.83
0.38
<2.7
Low Level
<0.32
<1.6
<0.49
<0.40
<0.35
0.51
<0.68
<0.38
<0.54
<0.32
0.20
<0.49
<0.40
<0.35
<0.69
<0.68
<0.38
<0.54
<0.32
0.23
<0.49
<0.40
<0.35
<0.69
<0.68
<0.38
<0.54
4.4.2 Surrogate Recoveries
To monitor analytical performance, a vapor-phase surrogate mix was loaded onto the TO-15 concentrator
during sample introduction. The three surrogates monitored were l,2-DCE-d4, toluene-d8, and
bromofluorobenzene. The spiking level was 25 ppbv for standard TO-15 and 5.0 ppbv for low-level
TO-15. The recoveries were evaluated against laboratory limits of 70 to 130%. All sample surrogate
recoveries met the laboratory criterion, and summary statistics are presented in Table 4-20.
Table 4-20. ATL TO-15 Surrogate Recovery Summary
Parameter
Number of surrogate
recoveries measured
Average recovery (%R)
Standard deviation (%R)
Minimum recovery (%R)
Maximum recovery (%R)
1,2-DCE-d4
26
97
12
86
124
Toluene-d8
26
98
1.7
94
100
Bromofluorobenzene
26
97
8.2
89
114
4.4.3 Laboratory Control Sample Recoveries
Analytical accuracy was evaluated by analyzing an LCS. An LCS working standard was prepared in a
Summa canister using a National Institute of Standards and Technology (NIST)-traceable vapor standard
independent from the primary calibration standard cylinder. The spiking level was 50 ppbv for the
standard TO-15 method and 10 ppbv for the low-level TO-15 analysis. The performance of the ATL TO-
15 LCSs is summarized in Table 4-21. A total of four unique LCS spikes were evaluated, and all met the
laboratory RLs.
4-13
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-21. ATL TO-15 LCS Summary
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
Number of
LCS
Analyzed
4
4
4
4
4
4
4
4
4
Mean LCS
%
Recovery
97
107
97
93
90
85
95
96
97
LCS Std
Dev (%R)
14
14
11
10
10
12
11
13
10
Min (%R)
78
90
80
80
76
77
79
77
83
Max (%R)
110
124
105
104
98
102
104
108
105
LCS
Recovery
Limits
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
4.4.4 Duplicates
Sample precision was evaluated by collecting field duplicates and by analyzing LCSDs. Field duplicates
were collected for approximately every 10 field samples (air samples do not have duplicate samples; only
collocated samples), and an LCSD was analyzed with each sample preparation batch. The LCSD was
prepared by analyzing the LCS working standard a second time. The instrument precision is summarized
in Table 4-22. The laboratory acceptance criterion of %RPD ^25% was met for all compounds.
Table 4-22. ATL TO-15 Laboratory Precision (LCS/LCSD) Summary
Benzene
Carbon disulfide
Chloroform
cis-1,2-DCE
Hexane
Methylene chloride
PCE
Toluene
TCE
Number of
LCSD
Analyzed
4
4
4
4
4
4
4
4
4
Mean %RPD
2%
2%
1%
1%
1%
3%
3%
2%
2%
Std Dev.
(%RPD)
3%
2%
2%
1%
1%
2%
2%
2%
1%
Min (%RPD)
0%
1%
0%
0%
0%
1%
1%
0%
0%
Max (%RPD)
7%
4%
4%
3%
1%
7%
5%
5%
3%
4.5 Online GC (Soil Gas and Indoor Air)
4.5.1 Blanks
Instrument blanks were analyzed at least once per analysis cycle of the 12 sampling locations. Nitrogen or
outdoor air was analyzed at the beginning of the analysis cycle (stream selector valve port #1). System
blanks (no vapor sample injected) were analyzed twice per analysis cycle at the end of the analysis cycle
(stream selector valve ports #15 and #16).
4-14
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4.5.2 Initial Calibration
For Phase 1 (August 11, 2011, to October 17, 2011), an initial calibration curve for PCE and chloroform
(CHC13) was performed at the start of the monitoring program as follows:
• PCE: Two points at concentrations of 13 ug/m3 and 70 ug/m3
• CHC13: A single point at a concentration 10 ug/m3, with a separate linearity study after the initial
deployment
Additional calibration points were not possible because of problems with the calibration standards
brought to the site during instrument setup. Although these one- and two-point calibrations were less than
desired, and they were corrected in the second round (see below), the data from the two sampling phases
matched up fairly well, indicating that the limited calibration points in the first round still gave
representative data.
For Phase 2 (December 1,2011, to February 16, 2012), initial calibrations were as follows:
• PCE low range: six points at concentrations from 0.7 ug/m3 to 23 ug/m3
• PCE high range: three points at concentrations from 3.5 ug/m3 to 69 ug/m3
• CHC13 low range: four points at concentrations from 3.3 ug/m3 to 55 ug/m3
• CHC13 high range: three points at concentrations from 55 ug/m3 to 270 ug/m3
The MDL for the on-site gas chromatograph (GC) was around 1 ug/m3.
4.5.3 Continuing Calibration
Continuing calibration could not be performed using the compounds of interest because the calibration
standard could contaminate the indoor air values. Instead, a surrogate compound, TCE, was used for
continuing calibration. The TCE was plumbed to stream selector port #14 with the intent it would be
analyzed in every analytical cycle of the 16 ports. However, during both phases of the program, the TCE
calibration standard ran out inexplicably fast. Several attempts were made to discover the source of the
leak and replace the standard, but on each occasion the calibration gas was lost. As an alternative, a
calibration check comparing the performance of the field instrument to a laboratory-based instrument
with site sample was performed as discussed in the next section.
4.5.4 Calibration Check via Comparison to Fixed Laboratory (TO-15 vs. Online GC)
Verification samples were collected and analyzed by H&P Mobile Geochemistry during each phase as
follows.
For Phase 1, an indoor air sample was collected from the 422 1st floor on October 11, 2011, and
compared to the on-site instrument to check on the reported concentration values. The results were as
follows (ug/m3):
On-site GC H&P TO-15
CHC13 1.7 0.8
PCE 3 1.3
In addition, a 24-hour time composite indoor air sample was collected from the 422 first floor and the
basement on September 22, 2011, and compared to the on-site instruments values over the same time
period to check on the reported low concentration values. The results were as follows (ug/m3):
4-15
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
On-siteGC ATLTO-15
422 First floor:
CHC13 1.0 0.24
PCE 1.75 0.40
422 Basement:
CHC13 1.7 0.41
PCE 3.5 0.94
Based on these data and the data summarized in Section 4.5.6, we decided that the online GC chloroform
low values (<5 ug/m3) should be adjusted down by a factor of 2 (conservatively) and the online GC PCE
low values (<5 ug/m3) should be adjusted down by a factor of 3.
For Phase 2, a sample was collected from probe SP8-9 on December 11, 2011, and compared with the on-
site instrument. The results were as follows (ug/m3):
On-site GC H&P TO-15
CHC13 118 100
PCE 140 160
Based on these results, no adjustments in the online GC data were made. In the setup for the second
phase, eliminating the concentration trap and adding additional calibration points resulted in a better data
match than in Phase 1. Given that the primary purpose of the online GC was to look for temporal
variations, rather than making direct concentration comparisons to the other methods, the calibration
results above were determined to be adequate for this research goal.
4.5.5 Agreement of Online GC Results with TO-17 Verification Samples
ATL prepared four 3-L Tedlar bags1 each containing approximately 2 L of vapor labeled A, B, C, and D
and sent them to the Indianapolis field site. Bags A and B were duplicate nitrogen blanks. Bags C and D
were duplicate spikes with chloroform, TCE, and PCE drawn from a common Summa canister. Analyses
were performed of these bags using the online GC and by ARCADIS staff collecting TO-17 samples
directly from the bags and submitting them to the National Exposure Research Laboratory (NERL) for
analysis. ATL also performed analyses before sending the bags to Indianapolis and after their return from
the field. Results of these interlaboratory comparisons are provided in Table 4-23 and statistical
comparison in Table 4-24.
1 Tedlar bags were used because of the need to pull samples with a syringe and the online GC at atmospheric
pressure. Method 0040 and other studies have shown that standards are stable in Tedlar bags for up to 72 hours for
a variety of chlorinated ethanes, ethenes (including TCE and PCE), and carbon tetrachloride.
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-23. Interlaboratory Results: Spiked Verification Samples
Bag
D
A
B
C
D
C
D
B
A
D
B
A
D
C
C
Laboratory
Air Toxics
Hartman
Hartman
Hartmann
Hartman
Air Toxics
Air Toxics
EPA NERL
EPA NERL
EPA NERL
EPA NERL
EPA NERL
EPA NERL
EPA NERL
EPA NERL
Subsample
Date
8/9/2011
8/11/2011
8/11/2011
8/11/2011
8/11/2011
8/12/2011
8/12/2011
8/10/2011
8/10/2011
8/10/2011
8/10/2011
8/10/2011
8/10/2011
8/10/2011
8/10/2011
Analysis Date
8/9/2011
8/11/2011
8/11/2011
8/11/2011
8/11/2011
8/12/2011
8/12/2011
8/14/2011
8/14/2011
8/14/2011
8/14/2011
8/14/2011
8/14/2011
8/14/2011
8/14/2011
PCE
flag
<
<
U
U
U
U
PCE
|jg/m3
8.5
8.5
85
8.5
8.5
80
89
84
PCE
ppbv
21
2
2
20
20
13
12
1.2
1.2
12.3
1.2
1.2
11.6
12.9
12.2
TCE
flag
<
<
U
U
U
U
TCE
|jg/m3
6.7
6.7
110
6.7
6.7
110
110
110
TCE
ppbv
34
2
2
28
23
16
16
1.2
1.2
20.1
1.2
1.2
20.1
20.1
20.1
Chloroform
Flag
<
<
B
U
B
B
Chloroform
|jg/m3
12
6.2
140
12
11
130
140
130
Chloroform
ppbv
42
2
2
40
40
20
21
2.4
1.3
28.2
2.4
2.2
26.2
28.2
26.2
Data quality flags: "<" = less than, "U" = compound analyzed and reported as below the MDL; "B"
detected in the associated method blank.
= Compound concentration is flagged because the compound was
4-17
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-24.1nte(laboratory Statistics: Spiked Verification Samples
Data Summary for Interlab
Data
Chemical
Chloroform
Tetrachloroethene
Trichloroethene
Actual
(TO-15)
42
21
34
Interlab comparison
Standard Samples after Pooling c and d: Interlab Comparison Using
Standard Results
Mean (ppbv)
Air
Toxics
(N=3)
27.7
15.3
22.0
EPA
NERL
(N=4)
27.2
12.3
20.1
Hart man
(N=2)
40.0
20.0
25.5
% Difference (% error)**
Air Toxics
vs. EPA
NERL
1.64
22.24
8.80
Air
Toxics
vs.
Hartman
38.03
47.95
23.46
Air Toxics
vs.
Hartman
36.45
26.42
14.74
4.5.6 Agreement of Integrated Online GC Results with Passive Samplers
Table 4-25 compares the concentrations measured by the 1-week Radiello samples to the concentrations
calculated by averaging the online GC results. For chloroform, agreement is generally remarkably good
for the first 4 weeks of instrument operation. The results for this period are generally within 50 relative
percent difference, which we considered good for this comparison between two different methods, given
that variability in interlaboratory comparisons for split samples of VOCs using one method can be larger.
Expressed as a ratio during this period, the online GC result is always between 0.6 and 1.9 times the
Radiello result.
However, for chloroform, agreement is noticeably worse in succeeding weeks (after September 14, 2011).
Generally, the chloroform values reported from the online GC are 1 to 3 times higher than the values from
the corresponding Radiello sample, although higher ratios up to 6 times higher were occasionally
observed, associated with the lowest concentration Radiello results. During the period when ambient
samples were also collected with the online GC, those results tended to comprise a more significant
fraction of the measured indoor air values than was seen in the Radiello samples. This result suggests the
possible existence of an elevated baseline in the online GC data.
For PCE, the relationship between the online GC and the Radiello samples appears more stable with the
vast majority of the results showing online GC results 1 to 3 times higher than the corresponding Radiello
data.
Despite the substantial differences between the absolute values for either compound measured by the two
methods, when the data are examined in terms of the ratio of concentrations on the first floor to
concentrations measured in the basement, there is reasonably close agreement between the two
instruments.
4-18
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-25. Comparison of Online GC to Radiello Results
Radiello
Location
Code
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
Online
GC
Probe
P4
P4
P2
P5
P5
P3
P4
P4
P2
P5
P5
P3
P4
P4
P2
Radiello
Interval:
Start
8/10/11
15:51
8/10/11
15:43
8/10/11
15:38
8/10/11
15:25
8/10/11
15:31
8/10/11
15:19
8/17/11
18:13
8/17/11
17:55
8/17/11
17:36
8/17/11
17:01
8/17/11
17:17
8/17/11
16:34
8/24/11
16:20
8/24/11
16:22
8/24/11
16:16
Radiello
Interval:
End
8/17/11
18:11
8/17/11
17:53
8/17/11
17:34
8/17/11
16:54
8/17/11
17:13
8/17/11
16:31
8/24/11
16:17
8/24/11
16:21
8/24/11
16:14
8/24/11
15:51
8/24/11
15:58
8/24/11
15:44
8/31/11
15:58
8/31/11
15:51
8/31/11
15:44
Number
of
Matched
GC Runs
63
63
63
62
62
62
70
70
70
71
71
71
65
65
65
Online GC Statistics
Mean
Chloroform
0.179
0.179
0.146
0.240
0.240
0.120
0.201
0.201
0.164
0.226
0.226
0.130
0.249
0.249
0.193
N Chloro-
form
57
57
63
56
56
62
70
70
70
71
71
71
65
65
65
Mean
PCE
0.572
0.572
0.373
0.953
0.953
0.415
0.700
0.700
0.643
0.827
0.827
0.504
0.625
0.625
0.594
N
PCE
51
51
45
58
58
46
69
69
70
71
71
70
65
65
65
Radiello
Measure-
ments
Chloroform
0.23
0.22
0.23
0.2
0.2
0.18
0.12
0.13
0.18
0.13
0.17
0.15
0.24
0.19
0.24
PCE
0.34
0.42
0.32
0.56
0.65
0.23
0.31
0.38
0.3
0.32
0.53
0.21
0.33
0.4
0.33
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
-25%
-21%
-45%
18%
18%
-40%
50%
43%
-9%
54%
28%
-15%
4%
27%
-22%
PCE
51%
31%
15%
52%
38%
57%
77%
59%
73%
88%
44%
82%
62%
44%
57%
Ratio (online
Result/Radiello)
Chloro-
form
0.78
0.81
0.63
1.20
1.20
0.67
1.67
1.54
0.91
1.74
1.33
0.86
1.04
1.31
0.80
PCE
1.68
1.36
1.17
1.70
1.47
1.81
2.26
1.84
2.14
2.58
1.56
2.40
1.89
1.56
1.80
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
0.82
0.50
0.82
0.57
0.77
Tetrachloro-
ethene
0.65
0.44
0.92
0.61
0.95
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
1.02
0.90
1.44
1.00
1.12
Tetrachloro-
ethene
0.84
0.38
0.87
0.49
0.90
4-19
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Radiello
Location
Code
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
Online
GC
Probe
P5
P5
P3
P4
P4
P2
P5
P5
P3
P4
P4
P2
P5
P5
P3
P4
Radiello
Interval:
Start
8/24/11
15:53
8/24/11
16:00
8/24/11
15:47
9/7/11
15:41
9/7/11
15:36
9/7/11
15:29
9/7/11
15:17
9/7/11
15:22
9/7/11
15:12
9/14/11
18:28
9/14/11
18:11
9/14/11
17:49
9/14/11
17:14
9/14/11
17:29
9/14/11
16:59
9/21/11
17:31
Radiello
Interval:
End
8/31/11
3:17
8/31/11
15:24
8/31/11
15:10
9/14/11
18:25
9/14/11
18:09
9/14/11
17:48
9/14/11
17:12
9/14/11
17:27
9/14/11
16:57
9/21/11
17:29
9/21/11
17:23
9/21/11
17:18
9/21/11
16:57
9/21/11
17:02
9/21/11
16:50
9/28/11
16:22
Number
of
Matched
GC Runs
60
64
64
53
53
52
53
53
53
51
51
51
51
51
51
62
Online GC Statistics
Mean
Chloroform
0.165
0.164
0.101
0.294
0.294
0.282
0.711
0.711
0.439
0.683
0.683
0.698
1.096
1.096
0.506
0.460
N Chloro-
form
60
64
64
53
53
52
53
53
53
51
51
51
51
51
51
62
Mean
PCE
0.564
0.557
0.294
0.582
0.582
0.768
1.137
1.137
0.761
0.678
0.678
0.948
1.138
1.138
0.544
0.586
N
PCE
60
64
64
53
53
52
53
53
52
44
44
43
47
47
43
36
Radiello
Measure-
ments
Chloroform
0.088
0.099
0.12
0.24
0.26
0.21
0.42
0.71
0.59
0.35
0.25
0.26
0.44
0.38
0.23
0.092
PCE
0.16
0.2
0.11
0.25
0.36
0.22
0.6
0.89
0.77
0.53
0.63
0.35
0.65
0.94
0.33
0.26
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
61%
50%
-17%
20%
12%
29%
51%
0%
-29%
64%
93%
91%
85%
97%
75%
133%
PCE
112%
94%
91%
80%
47%
111%
62%
24%
-1%
25%
7%
92%
55%
19%
49%
77%
Ratio (online
Result/Radiello)
Chloro-
form
1.88
1.66
0.84
1.23
1.13
1.34
1.69
1.00
0.74
1.95
2.73
2.68
2.49
2.88
2.20
5.00
PCE
3.52
2.78
2.67
2.33
1.62
3.49
1.89
1.28
0.99
1.28
1.08
2.71
1.75
1.21
1.65
2.25
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
0.62
0.96
0.62
1.02
0.46
Tetrachloro-
ethene
0.53
1.32
0.67
1.40
0.48
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
1.28
0.84
1.04
0.87
0.56
Tetrachloro-
ethene
0.61
0.72
1.03
0.60
0.42
4-20
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Radiello
Location
Code
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
Online
GC
Probe
P4
P2
P5
P5
P3
P4
P4
P2
P5
P5
P3
P4
P4
P2
P5
P5
Radiello
Interval:
Start
9/21/11
17:25
9/21/11
17:20
9/21/11
16:59
9/21/11
17:05
9/21/11
16:53
10/6/11
16:53
10/6/11
16:43
10/6/11
16:33
10/6/11
16:09
10/6/11
16:20
10/6/11
15:59
11/30/11
13:33
11/30/11
13:28
11/30/11
13:21
11/30/11
13:01
11/30/11
13:07
Radiello
Interval:
End
9/28/11
16:09
9/28/11
15:58
9/28/11
15:26
9/28/11
15:39
9/28/11
15:08
10/12/11
16:51
10/12/11
16:34
10/12/11
16:16
10/12/11
15:39
10/12/11
15:53
10/12/11
15:24
12/7/11
18:26
12/7/11
18:12
12/7/11
17:49
12/7/11
16:40
12/7/11
17:08
Number
of
Matched
GC Runs
62
62
62
62
62
53
53
53
54
53
53
91
91
91
90
91
Online GC Statistics
Mean
Chloroform
0.460
0.577
0.783
0.783
0.427
0.667
0.667
0.776
0.981
0.984
0.701
0.527
0.527
0.776
0.779
N Chloro-
form
62
62
62
62
62
53
53
53
54
53
53
24
24
0
90
91
Mean
PCE
0.586
0.908
1.158
1.158
0.538
0.845
0.845
1.288
1.282
1.283
0.947
0.436
0.436
0.220
1.300
1.304
N
PCE
36
36
39
39
33
47
47
42
45
44
43
89
89
89
90
91
Radiello
Measure-
ments
Chloroform
0.089
0.094
0.17
0.22
0.14
0.32
0.26
0.23
0.31
0.52
0.53
0.23
0.28
0.23
0.38
0.42
PCE
0.27
0.18
0.54
0.6
0.27
0.6
0.58
0.51
1
1
1.1
0.41
0.4
0.32
0.56
0.75
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
135%
144%
129%
112%
101%
70%
88%
109%
104%
62%
28%
78%
61%
NA
68%
60%
PCE
74%
134%
73%
63%
66%
34%
37%
87%
25%
25%
-15%
6%
9%
-37%
80%
54%
Ratio (online
Result/Radiello)
Chloro-
form
5.17
6.13
4.61
3.56
3.05
2.08
2.56
3.37
3.16
1.89
1.32
2.29
1.88
0.00
2.04
1.86
PCE
2.17
5.04
2.14
1.93
1.99
1.41
1.46
2.53
1.28
1.28
0.86
1.06
1.09
0.69
2.32
1.74
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
1.25
0.55
1.16
0.71
NA
Tetrachloro-
ethene
1.55
0.46
1.52
0.74
0.51
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
1.04
0.72
0.79
1.28
0.90
Tetrachloro-
ethene
0.68
0.47
0.86
1.10
0.79
4-21
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Radiello
Location
Code
422First-A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
Ambient - A
420bASEn-i
420bASEs-i
420flRST-i
422bASEn-i
422bASEs-i
422flRST-i
422flRST-j
AMBIENT -
DOWN
Online
GC
Probe
P3
P4
P4
P2
P5
P5
P3
P1
P4
P4
P2
P5
P5
P3
P3
P1
Radiello
Interval:
Start
11/30/11
12:56
12/7/11
18:27
12/7/11
18:13
12/7/11
17:51
12/7/11
16:43
12/7/11
17:10
12/7/11
16:19
12/7/11
17:33
128/11
13:02
128/11
12:57
128/11
12:41
128/11
12:25
128/11
12:33
128/11
12:12
128/11
12:16
128/11
13:17
Radiello
Interval:
End
12/7/11
16:17
12/14/11
16:44
12/14/11
16:20
12/14/11
16:09
12/14/11
15:33
12/14/11
15:41
12/14/11
15:27
12/14/11
17:00
12/15/11
15:49
12/15/11
15:46
12/15/11
15:43
12/15/11
15:29
12/15/11
15:35
12/15/11
15:08
12/15/11
15:10
12/15/11
16:12
Number
of
Matched
GC Runs
90
100
101
101
101
100
101
101
82
82
82
82
82
82
82
83
Online GC Statistics
Mean
Chloroform
0.515
0.581
0.581
1.092
1.092
0.640
0.512
0.600
0.600
1.094
1.094
0.653
0.653
0.564
N Chloro-
form
42
55
55
0
101
100
98
6
45
45
0
82
82
76
76
8
Mean
PCE
0.581
0.765
0.761
0.560
1.835
1.836
0.995
0.420
0.851
0.851
0.800
1.872
1.872
1.049
1.049
0.555
N
PCE
90
100
101
99
100
99
101
101
82
82
80
81
81
82
82
83
Radiello
Measure-
ments
Chloroform
0.33
0.37
0.3
0.22
0.55
0.86
0.39
0.13
0.27
0.26
0.18
0.57
0.8
0.36
0.36
0.12
PCE
0.45
0.39
0.41
0.31
0.63
0.93
0.48
0.23
0.4
0.34
0.25
0.57
0.9
0.47
0.47
0.21
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
44%
44%
64%
NA
66%
24%
49%
119%
76%
79%
NA
63%
31%
58%
58%
130%
PCE
25%
65%
60%
57%
98%
66%
70%
59%
72%
86%
105%
107%
70%
76%
76%
90%
Ratio (online
Result/Radiello)
Chloro-
form
1.56
1.57
1.94
0.00
1.99
1.27
1.64
3.94
2.22
2.31
0.00
1.92
1.37
1.81
1.81
4.70
PCE
1.29
1.96
1.86
1.81
2.91
1.97
2.07
1.83
2.13
2.50
3.20
3.28
2.08
2.23
2.23
2.64
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
0.66
NA
0.59
NA
0.60
Tetrachloro-
ethene
0.45
0.74
0.54
0.94
0.56
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
0.83
0.66
0.55
0.68
0.53
Tetrachloro-
ethene
0.69
0.78
0.62
0.68
0.64
4-22
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Radiello
Location
Code
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
Ambient - A
420BaseN-A
420BaseS-A
420First-A
422BaseN-A
422BaseS-A
422First-A
Ambient - A
420BaseN-A
420BaseS-A
Online
GC
Probe
P4
P4
P2
P5
P5
P3
P1
P4
P4
P2
P5
P5
P3
P1
P4
P4
Radiello
Interval:
Start
12/14/11
16:47
12/14/11
16:21
12/14/11
16:42
12/14/11
15:35
12/14/11
15:42
12/14/11
15:28
12/14/11
17:01
1/4/12
17:04
1/4/12
17:19
1/4/12
16:55
1/4/12
14:41
1/4/12
14:58
1/4/12
14:24
1/4/12
16:39
1/11/12
15:07
1/11/12
15:01
Radiello
Interval:
End
12/22/11
17:38
12/22/11
17:26
12/22/11
17:12
1222/11
18:04
1222/11
18:16
12/22/11
17:52
12/22/11
16:49
1/11/12
17:35
1/11/12
17:19
1/11/12
16:55
1/11/12
14:41
1/11/12
14:58
1/11/12
14:24
1/11/12
16:39
1/18/12
15:07
1/18/12
15:01
Number
of
Matched
GC Runs
104
104
103
105
105
104
105
98
98
98
98
98
98
98
102
102
Online GC Statistics
Mean
Chloroform
0.696
0.696
0.845
0.845
0.540
0.612
0.342
0.342
0.290
0.959
0.959
0.527
0.284
0.295
0.295
N Chloro-
form
12
12
0
89
89
55
5
45
45
39
98
98
94
36
92
92
Mean
PCE
0.271
0.271
0.268
1.560
1.560
0.644
0.482
0.302
0.302
0.168
1.307
1.307
0.569
0.279
0.297
0.297
N
PCE
98
98
101
102
102
99
94
98
98
98
98
98
98
94
102
102
Radiello
Measure-
ments
Chloroform
0.12
0.12
0.067
0.42
0.61
0.26
0.051
0.14
0.14
0.12
0.42
0.68
0.22
0.1
0.16
0.14
PCE
0.15
0.16
0.09
0.43
0.62
0.32
0.094
0.2
0.2
0.17
0.46
0.71
0.35
0.18
0.3
0.19
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
141%
141%
NA
67%
32%
70%
169%
84%
84%
83%
78%
34%
82%
96%
59%
71%
PCE
57%
51%
99%
114%
86%
67%
135%
41%
41%
-1%
96%
59%
48%
43%
-1%
44%
Ratio (online
Result/Radiello)
Chloro-
form
5.80
5.80
0.00
2.01
1.39
2.08
12.00
2.44
2.44
2.42
2.28
1.41
2.40
2.84
1.84
2.11
PCE
1.80
1.69
2.98
3.63
2.52
2.01
5.13
1.51
1.51
0.99
2.84
1.84
1.63
1.55
0.99
1.57
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
NA
0.64
0.85
0.55
Tetrachloro-
ethene
0.99
0.41
0.56
0.44
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
0.56
0.50
0.86
0.40
Tetrachloro-
ethene
0.58
0.61
0.85
0.60
4-23
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Radiello
Location
Code
420First-A
422BaseN-A
422BaseS-A
422First-A
Ambient - A
GC
Probe
P2
P5
P5
P3
P1
Radiello
Interval:
Start
1/11/12
14:56
1/11/12
14:44
1/11/12
14:50
1/11/12
14:38
1/11/12
15:13
Radiello
Interval:
End
1/18/12
14:56
1/18/12
14:44
1/18/12
14:50
1/18/12
14:38
1/18/12
15:13
Number
of
Matched
GC Runs
102
101
101
101
101
Online GC Statistics
Mean
Chloroform
0.290
1.070
1.070
0.553
0.274
N Chloro-
form
62
100
100
100
45
Mean
PCE
0.249
2.159
2.159
0.924
0.079
N
PCE
102
100
100
100
95
Radiello
Measure-
ments
Chloroform
0.11
0.44
0.73
0.36
0.09
PCE
0.14
0.75
1.1
0.54
0.067
% Difference =
(on line-
Radiello)/average
(Online, Radiellos)
Chloro-
form
90%
83%
38%
42%
101%
PCE
56%
97%
65%
52%
16%
Ratio (online
Result/Radiello)
Chloro-
form
2.64
2.43
1.47
1.54
3.05
PCE
1.78
2.88
1.96
1.71
1.18
Ratio
(First/Basement)
From Online GC
GC
Chloro-
form
0.98
0.52
Tetrachloro-
ethene
0.84
0.43
Ratio (First/Avg
Basement) from
Radiello Data
Radiello
Chloro-
form
0.73
0.62
Tetrachloro-
ethene
0.57
0.58
4-24
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4.6 Radon
4.6.1 Indoor Air: Comparison of Electrets Field, ARCADIS to Charcoal Analyzed by
U.S. EPA Radiation and Indoor Environment (R&IE) National Laboratory
Three comparisons were made between electrets and charcoal canisters. Charcoal canisters were provided
by and analyzed by the U.S. EPA R&IE National Laboratory Center for Indoor Environments in Las
Vegas, Nevada. ARCADIS collected charcoal canister samples and electret samples. Electrets were
obtained from Rad Elec (Frederick, Maryland) and read by ARCADIS on site before and after
deployment. The charcoal canisters were used as a QC check on three separate occasions: January 19,
2011 to January 26, 2011; April 27, 2011 to May 4, 2011; and December 28, 2011 to January 4, 2012.
Charcoal canisters (plus duplicates) were placed at indoor locations and the ambient locations that were
routinely being used for electret monitoring. When the results were received, the sample plus its duplicate
were averaged together to obtain a result for the location. This result was then compared with the electret
result for that location and time period.
For the first occasion, the relative percent difference between the two methods was 20% or less (Table
4-26). The maximum absolute difference was 0.63 pCi. A relative percent difference could not be
calculated for the ambient, which was below the detection limit with the charcoal method (below
detection limits; BDL).
On the second occasion, five of six comparisons showed a relative percent difference of 20% or less and
four of the six comparisons were within 0.5 pCi/L of each other (Table 4-27).
The exceptions were 422 basement north and 420 basement south, which were within 0.9 pCi/L of each
other. The ambient was again BDL by the charcoal method, as would have been predicted from the
electret data.
For the third occasion, December 28, 2011 to January 4, 2012, the absolute difference between the
methods is at or below 0.3 pCi/L and RPD is <6% for all samples (Table 4-28). The ambient charcoal
sample was below the detection limit and that detection limit was equal to the ambient value reported by
the electret method.
4-25
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-26. Comparison between Electrets and Charcoal Canisters at the 422/420 EPA House from
January 19-26, 2011
Sample
Location
422 First floor
422 First floor
422 Basement N
422 Basement N
420 First floor
420 First floor
420 Basement N
420 Basement N
Ambient
Ambient
Electret Rn
(pCi/L)
5.14
8.44
1.68
3.98
0.03
Charcoal Rn
(pCi/L)
4.8
4.6
8
8.8
1.7
1.6
3.3
3.4
<0.5
<0.5
Charcoal
Average
4.7
8.4
1.65
3.35
<0.5
Absolute
Difference (pCi/L)
0.44
0.04
0.03
0.63
RPD (%)
6.84%
5.35%
-1.18%
18.68%
Table 4-27. Comparison of Electret and Charcoal Canister Data from April 27, 2011, to May 4, 2011
Location
Ambient
Ambient duplicate
422 First floor
422 First floor duplicate
422 Basement S
422 Basement S duplicate
422 Basement N
422 Basement N duplicate
420 First floor
420 First floor duplicate
420 Basement S
420 Basement S duplicate
420 Basement N
420 Basement N duplicate
Field blank
Field blank
Elect re
tData
(pCi/L)
0.47
2.72
7.39
7.14
6.77
0.98
4.58
4.48
NA
NA
Charcoal
Canister
Radon Activity
(pCi/L)
<0.5
<0.5
2.8
2.4
7.3
6.7
6.3
5.8
1.3
1.5
3.8
3.7
4.2
3.7
<0.5
<0.5
Charcoal
Canister
Average
Radon Activity
(pCi/L)
2.6
7
6.05
1.4
3.75
3.95
Absolute
Difference
(pCi/L)
0.12
0.39
0.905
-0.42
0.83
0.53
RPD (%)
4.51%
5.42%
13.92%
-35.29%
19.93%
12.57%
4-26
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
f. 4
Electretvs. CharcoalComparision(late April- early May 2011)
Radon
1:1 Line
4 5
Electret (pCi/l)
Figure 4-1. Correlation between radon
Table 4-28. Comparison of Charcoal and
measured using the electret and charcoal methods.
Electret Radon December 28, 2011, to January 4, 2012
Canister ID
877138
877113
877137
877115
877133
877107
877139
877136
877128
877111
877108
877140
877110
877131
877130
Radon
Activity
(pCi/l)
3.1
3.2
2.8
2.7
1.1
1.0
10.0
9.9
9.6
9.4
4.8
4.7
5.0
5.3
<0.5
Charcoal
Average
(pCi/l)
3.2
2.8
1.1
10.0
9.5
4.8
5.2
Location
420 Basement N
420 Basement N Dup
420 Basement S
420 Basement S Dup
420 First floor
420 First floor Dup
422 Basement N
422 Basement N Dup
422 Basement S
422 Basement S Dup
422 First floor
422 First floor Dup
422 Office
422 Office Dup
Ambient
Electrets
(pCi/L)
3.34
2.72
1.09
10.22
10.35
9.57
4.86
4.92
0.5
Absolute
Difference
(pCi/L)
-0.2
0.0
0.0
-0.3
-0.1
-0.1
0.2
NA
RPD (%)
-5.86%
1.10%
-3.74%
-2.67%
-0.73%
-2.29%
4.57%
NA
4-27
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4.6.2 Comparision of Average of Real Time Alphaguard to Electrets and Charcoal
Canisters
Stationary Alphaguard units provided by the U.S. EPA were used for real-time monitoring of indoor air
radon at two locations (422 Basement North and 422 Office (2nd floor)). Several comparisons were made
between the stationary Alphaguard data and electrets located nearby (initially at 422 basement north and
later at both 422 basement north and 422 office).
The first comparison took place over several weeks between March 30, 2011 and May 18, 2011 (Table
4-29). The absolute difference ranged from -0.04 pCi/L to 1.44 pCi/L. The relative percent difference
ranged from -0.50% to 26.04%.
Table 4-29. Comparison between 422 Base N Alphaguards and Electrets
from March 30, 2011, and May 18, 2011
Date Range
03/30-04/07
04/07-04/13
04/13-04/20
04/20-04/27
04/27-05/04
05/04-05/1 1
05/11-05/18
Alphaguard
Reading
(pCi/L)
6.18
5.90
8.41
6.25
6.92
4.66
6.15
Elect ret
(pCi/L)
6.30
4.94
6.97
4.04
7.14
2.93
5.81
Electret
Dup(pCi/L)
4.98
5.87
7.83
5.58
6.77
4.50
6.01
Electret Ave
(pCi/L)
5.64
5.41
7.40
4.81
6.96
3.72
5.91
Absolute
Difference
(pCi/L)
0.54
0.50
1.01
1.44
-0.04
0.95
0.24
Relative
Percent
Difference
9.14%
8.76%
12.78%
26.04%
-0.50%
22.57%
3.98%
For the second comparison, which occurred from August 3, 2011 to October 6, 2011,in the 422BaseN
location, the absolute difference ranged from -1.11 pCi/L to 2.42 pCi/L. The relative percent difference
ranged from -40.18% to 30.76% (Table 4-30).
Table 4-30. Comparison of Real-Time Alphaguard to Integrated Electret August through October
End Date/
Time
8/3/201 1
8/10/2011
8/17/2011
8/24/201 1
8/31/2011
9/7/201 1
9/14/2011
9/21/2011
9/28/201 1
10/6/2011
Rn (pCi/L) A
Guard
(averaged
over a week)
6.85
7.24
8.38
3.84
2.21
4.34
6.09
8.69
12.51
10.33
Rn (pCi/L)
Electrets 422
Base N
6.85
7.25
7.53
3.48
2.17
4.52
5.68
8.03
11.67
7.83
Rn (pCi/L)
Electrets 422
Base N Dup
5.14
6.79
7.20
3.00
4.46
1.84
5.44
7.84
11.44
7.99
Average of
Duplicate
Electrets
(pCi/L)
6.00
7.02
7.37
3.24
3.32
3.18
5.56
7.94
11.56
7.91
Absolute
Difference
(pCi/L)
0.85
0.22
1.02
0.60
-1.11
1.16
0.53
0.75
0.96
2.42
Relative
Percent
Difference
13.26%
3.09%
12.91%
16.93%
-40.18%
30.76%
9.16%
9.05%
7.97%
26.53%
4-28
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
During the third comparison, electrets, the Alphaguard, and the charcoal canisters were compared from
December 28, 2011 to January 4, 2012. Only the 422 office and 422 basement north were compared by all
three methods during this time. The absolute difference between the canisters and Alphaguard ranged
from -0.05 pCi/L to 0.15pCi/L, and the absolute difference between the electrets and Alphaguard ranged
from -0.08pCi/L to 0.29pCi/L. The relative percent difference between canisters and Alphaguard ranged
from -0.50% to 2.96%, and the relative percent difference between electrets and Alphaguard ranged from
-1.61% to 2.81% (Table 4-31).
Table 4-31. Comparison of Real-time Alphaguards to Integrated Electret Measurements December
28,2011 to January 4, 2012
o
ro
o
422BaseN
422 Office
.&
o
c
o
•o
IV
u_
'c
(7
0
9.95
5.15
_
—
,S
•ffi
1
LJJ
10.22
4.92
i
o
Q-
3
Q
•ffi
1
LLJ
10.35
o
Q.
o
>
<
•ffi
1
LJJ
10.29
ro ^f
3 ."^
D)O
If
^ .2
Z "ro
« •>;
g |
9 <
10.00
5.00
c
S «
"ffi §
-Q O)
p -^
c Q-
£ <
.*= c
Q ro
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Table 4-32. Comparison of Real-Time Alphaguard to Integrated Electret Measurements
January through March
C
"fo
Q
7-01/04/12
01/04/12-
01/11/12
01/11/12-
01/18/12
01/18/12-
01/25/12
01/25/12-
02/01/12
02/01/12-
02/08/12
02/08/12-
02/15/12
02/15/12-
02/22/12
02/22/12-
03/01/12
422 Base N Alphaguard
Reading (pCi/L)
10
8.78
9.73
8.52
7.71
8.68
8.44
7.74
8.48
Office Alphaguard Reading
(pCi/L)
5
4.69
5.09
4.79
4.46
4.78
4.80
4.3
4.74
422 Base N Electret (pCi/L)
10.22
9.05
9.34
7.83
8.24
8.60
8.28
6.08
9.00
422 Base N Dup Electret
(pCi/L)
10.35
9.11
9.73
7.98
8.03
8.62
7.47
5.82
9.00
422 Base N Electret Average
(pCi/L)
10.29
9.08
9.54
7.91
8.14
8.61
7.88
5.95
9.00
Office Electret (pCi/L)
4.92
4.56
4.88
4.74
4.15
4.58
4.41
3.68
3.97
Absolute Difference between
422 Base N Alphaguards and
Electrets (pCi/L)
-0.29
0.30
0.19
0.61
-0.43
0.06
0.56
1.79
-0.52
Absolute Difference between
Office Alphaguards and
Electrets (pCi/L)
0.08
0.13
0.21
0.05
0.31
0.20
0.39
0.62
0.77
422 Base N Relative Percent
Difference
-2.81%
-3.36%
2.02%
7.49%
-5.36%
0.81%
6.93%
26.15%
-5.95%
Office Relative Percent
Difference
1.61%
2.81%
4.21%
1.05%
7.20%
4.27%
8.47%
15.54%
17.68%
4.6.3 Quality Assurance Checks of Electrets
QC was performed on the electret reader and on the chambers holding the electrets. The QC check on the
reader was performed by placing reference electrets within the reader each week to measure any deviation
from the standard. The standard reference electrets were of 0 V, 245 V, and 250 V. Over the duration of
the project, the readings on the 0 V electret fluctuated but stayed within 4 V of its nominal value. The 245
V electret, with only two exceptions, stayed within 20 V of its stated value. It steadily declined over the
duration of the project, hitting a low before slowly rising toward the end of the project. The 250 V electret
stayed within 6 V of its nominal value, showing a slight decline toward the end of the project.
To check for drift within the electret chambers, a normal electret was placed in a closed electret chamber
each week and then read on the voltage meter to measure any change in the voltage from the previous
week's readings. This method would indicate any deviation caused by the chambers. Near the beginning
of the project, this electret dropped an average of 5 V/4 weeks or 1.25 V per week. The rate was even
lower in the second half of the project to a drop of 5 V/30 weeks or 0.16 V per week. These rates of drift
are insignificant because the actual observed voltage change at the indoor sampling locations was
typically 25V per week or more.
4-30
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
4.7 On-Site Weather Station vs. National Weather Service (NWS)
A VantageVue weather station from Davis Instruments was installed at the 422/420 house. Because it was
not safe to mount the station directly on the peak of the roof, it was mounted on vertical rods raised to the
approximate peak elevation from the edge of the second story roof. The trees near the house, especially to
the north, are quite tall, equal to or higher than the weather station. Branches extend close to the house on
the northwest corner. The house is much taller than the neighboring building to the east. There is also a
neighboring two-story residential structure to the northeast, approximately 30 to 40 ft away. A seven-
story commercial structure is approximately 150 ft southwest of the studied duplex. Essentially, the only
side completely free from all air current obstructions is the southern side, which borders 28th Street
(Figure 4-2).
A 2-month comparison between the house weather station data and NWS data was made from
September 17, 2011 to November 17, 2011, as a QC check. Three parameters were compared:
temperature, relative humidity, and wind speed. For temperature, the data from the two weather stations
match very well, only differing by an average of 0.5 ° F (Figure 4-3). Relative humidity at both weather
stations differed by an average of approximately 4% (Figure 4-4). House wind speed and that of the
NWS differed by an average of approximately 6 mph; the airport weather station was generally higher.
This difference is likely due to the local NWS station being at the Indianapolis International Airport. The
Indianapolis International Airport (KIND) weather station is located in the middle of the runways at the
Indianapolis Airport approximately 500 meters from the nearest building. Thus, the readings obtained at
the house are probably a better representation of the wind speeds that directly impinge on the house
(Figure 4-5).
4-31
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Figure 4-2. Aerial view of study house, showing potential influences on wind velocity;
red arrow indicates study house.
Temperature Comparison
% %
\
Date
Figure 4-3. Comparison of National Weather Service Indianapolis
temperature data to weather station at 422 East 28th Street.
4-32
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
Relative Humidity Comparison
JS
O 40
•»••»• « ~» *
.» *•. » ••»
^ «»».»»»»• • •,
» •_*_ JLJL •.
r^
*• *i»
>422WS RH(%)
• NWS Relative Humidity(%)
\ \ \
Date
\
Figure 4-4. Comparison of National Weather Service Indianapolis
relative humidity to weather station at 422 East 28th Street.
Wind Speed Comparison
Q.
»422 WS Wind Speed (mph)
• NWS Wind Speed (mph)
***
» «
,, \ \ \ \ \ \
Date
Figure 4-5. Comparison of National Weather Service wind speed data
to weather station at 422 East 28th Street.
4.8 Database
4.8.1 Checks on Laboratory Reports
Throughout the project, the ARCADIS project manager briefly reviewed laboratory reports as they were
received from the VOC analytical laboratories. The primary focus of these checks was on blanks and
4-33
-------
Section 4—Results and Discussion: QA Checks of Individual Data Sets
ambient samples as a sampling performance indicator as well as the general consistency and
reasonableness of the trends in reported concentrations for key analytes: PCE and chloroform.
The ARCADIS project manager also performed a manual review of the electrets radon computations in
the spreadsheet used for those calculations. He also reviewed that data set regularly and interacted with
the field scientist collecting these data when any anomalous results were observed.
The lead analyst (from Hartman-Environmental Geosciences), an ARCADIS principal scientist, and an
RTI scientist were all involved in reviewing the online GC calculations. For suspect values, QC checks
performed included calibration checks and chromatogram reviews.
4.8.2 Database Checks
A Microsoft Access database was developed and used to compile results for volatile organic compounds
(VOCs) (TO-17, TO-15, and passive indoor air) and radon in indoor air and soil gas (electret and portable
Alphaguard).
The following QC checks were performed on this database:
• The ARCADIS field scientist responsible for the majority of the field sampling performed a fairly
intense check of the reports received from laboratories against his own records. He checked for
the following: approximate number of each sample type (to determine what reports were still
pending) and a line-by-line check of the sample times, dates, and sample numbers of each sample
type. The assignment of sample locations was also reviewed. Notes of any discrepancies and
corrections were sent to the ARCADIS database manager.
• During the initial portions of the project, the ATL technical director manually prepared an Excel
spreadsheet from laboratory reports comparing the results of passive samplers exposed at the
same location for multiple durations and calculating percentage bias. The ARCADIS project
manager then used that spreadsheet to spot check the calculations of percentage bias performed in
the database. After correcting for slight differences in the percentage bias formula used, excellent
agreement was found. This agreement indicates that, at least for the calculations spot checked,
both the calculation and the importation of the underlying concentration data from electronic
deliverable files into the database are being performed correctly.
• During the initial portions of the project, the ATL technical director manually prepared an Excel
spreadsheet of indoor air VOC results from laboratory reports. That Excel spreadsheet was used
to prepare temporal trend plots of indoor concentrations for key analytes for the first 18 weeks of
the project before the Access database was fully implemented. The ARCADIS project manager
then confirmed that the essential features of these temporal trend plots (such as range of
concentrations and overall temporal trends) were consistent between these plots and similar plots
generated from the Access database. This result indicates for this period that the importation of
the underlying concentration data from electronic deliverable files into the database is being
performed correctly.
• The ARCADIS project manager provided to the database manager a design document for the
reports to be generated, including definitions of key formulas and variables. The design document
was prepared based on the project objectives in the Quality Assurance Project Plan (QAPP). As
database reports were prepared, the ARCADIS project manager reviewed their format and
content and requested changes as necessary.
• The ARCADIS project manager and database manager both spot checked the transfer of the
NERL results for groundwater into the database.
4-34
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Table of Contents
5. Results and Discussion: VOC Concentration Temporal Trends and Relationship to HVAC 5-1
5.1 VOC Seasonal Trends Based on Weekly, Biweekly, and Monthly Measurements for
52+Weeks 5-1
5.1.1 Indoor Air 5-1
5.1.2 Subslab Soil Gas 5-4
5.1.3 Shallow and Deep Soil Gas 5-8
5.2 Radon Seasonal Trends (based on Weekly Measurements) 5-23
5.2.1 Indoor Air 5-23
5.2.2 Subslab and Wall Port Soil Gas 5-26
5.2.3 Deep Soil Gas 5-29
5.3 VOC Short-Term Variability (Based on Daily and Hourly VOC Sampling) 5-35
5.3.1 Indoor air 5-35
5.3.2 Subsurface Soil Gas Data 5-41
5.4 Radon Short Term Variability (Based on Daily and More Frequent Measurements) 5-48
5.4.1 Indoor Air 5-48
5.4.2 Subslab, Wall Port, and Deeper Soil Gas Radon Data 5-52
5.5 Outdoor Climate/Weather Data 5-52
List of Figures
5-1. PCE in indoor and ambient air vs. time (7-day Radiello samples) 5-2
5-2. Chloroform in indoor and ambient air vs. time (7-day Radiello samples) 5-2
5-3. PCE, benzene, and toluene in indoor and ambient air 5-3
5-4. PCE and chloroform in 422 first-floor indoor air; weekly, biweekly, and monthly
duration Radiello samples 5-3
5-5. Plot of subslab chloroform concentrations vs. time 5-4
5-6. Plot of subslab chloroform concentrations vs. time, first intensive sampling period 5-5
5-7. Plot of subslab PCE concentrations vs. time 5-6
5-8. Plot of subslab PCE concentrations vs. time, first intensive sampling period 5-6
5-9. Plot of wall port chloroform concentrations vs. time 5-7
5-10. Plot of wall port PCE concentrations vs. time 5-8
5-11. Chloroform concentrations at each of the SGP1 ports vs. time 5-11
5-12. PCE concentrations at each of the SGP1 ports vs. time 5-11
5-13. Chloroform concentrations at each of the SGP2 ports vs. time 5-12
5-14. PCE concentrations at each of the SGP2 ports vs. time 5-12
5-15. Chloroform concentrations at each of the SGP3 ports vs. time 5-13
5-16. PCE concentrations at each of the SGP3 ports vs. time 5-13
5-17. Chloroform concentrations at each of the SGP4 ports vs. time 5-14
5-18. PCE concentrations at each of the SGP4 ports vs. time 5-14
5-19. Chloroform concentrations at each of the SGP5 ports vs. time 5-15
5-20. PCE concentrations at each of the SGP5 ports vs. time 5-15
5-21. Chloroform concentrations at each of the SGP6 ports vs. time 5-16
5-22. PCE concentrations at each of the SGP6 ports vs. time 5-16
5-23. Chloroform concentrations at each of the SGP7 ports vs. time 5-17
5-24. PCE concentrations at each of the SGP7 ports vs. time 5-17
5-25. Chloroform concentrations at SGP8 and9 ports vs. time 5-18
5-26. PCE concentrations at SGP8 and 9 ports vs. time 5-18
5-27. Chloroform concentrations at each of the SGP10 ports vs. time 5-19
5-i
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5-28. PCE concentrations at each of the SGP10 ports vs. time 5-19
5-29. Chloroform concentrations at each of the SGP11 ports vs. time 5-20
5-30. PCE concentrations at each of the SGP11 ports vs. time 5-20
5-31. Chloroform concentrations at each of the SGP12 ports vs. time 5-21
5-32. PCE concentrations at each of the SGP12 ports vs. time 5-21
5-33. Subslab PCE concentrations over a 1-week period during the first intensive round 5-22
5-34. Subslab PCE concentrations over a 1-week period during the second intensive round 5-22
5-35. Weekly electret readings for all locations 5-24
5-36. Data for the downstairs continuously recording Alphaguard versus time 5-25
5-37. Data for the upstairs continuously recording Alphaguard versus time 5-26
5-38. This is a kriged radon image taken from subslab, wall, and multidepth soil gas data 5-27
5-39. Subslab Alphaguard data versus time 5-28
5-40. Wall port Alphaguard data versus time 5-28
5-41. Kriged radon image taken from subslab, wall, and multidepth soil gas data 5-30
5 -42. Handheld Alphaguard data taken from soil gas ports at location 1 versus time 5-31
5 -43. Handheld Alphaguard data taken from soil gas ports at location 4 versus time 5-31
5-44. Handheld Alphaguard data taken from soil gas ports at location 5 versus time 5-32
5-45. Handheld Alphaguard data taken from soil gas ports at location 7 versus time 5-32
5 -46. Handheld Alphaguard data taken from soil gas ports at location 8 versus time 5-33
5 -47. Handheld Alphaguard data taken from soil gas ports at location 9 versus time 5-33
5-48. Handheld Alphaguard data taken from soil gas ports at location 10 versus time 5-34
5-49. Handheld Alphaguard data taken from soil gas ports at location 11 versus time 5-34
5-50. Handheld Alphaguard data taken from soil gas ports at location 12 versus time 5-35
5-51. Online GC chloroform indoor air data for 422 first floor 5-36
5-52. Online GC chloroform indoor air data for 422 basement 5-36
5-53. Online GC chloroform indoor air data for 420 first floor 5-38
5-54. Online GC chloroform indoor air data for 420 basement 5-38
5-55. Online GC PCE indoor air data for 422 first floor 5-39
5-56. Online GC PCE indoor air data for 422 basement 5-39
5-57. Online GC PCE indoor air data for 420 first floor 5-40
5-58. Online GC PCE indoor air data for 420 basement 5-40
5-59. Online GC subsurface chloroform soil gas data—Phase 1 and Phase 2 5-42
5-60. Online GC subsurface chloroform soil gas data—Phase 1 5-42
5-61. Online GC subsurface chloroform soil gas data—Phase 2 5-43
5-62. Online GC subsurface PCE soil gas data—Phase 1 and Phase 2 5-44
5-63. Online GC subsurface PCE soil gas data—Phase 1 5-44
5-64. Online GC subsurface PCE soil gas data—Phase 2 5-45
5-65. Method TO-17 data for SSP-4 5-46
5-66. Online GC PCE measurements in SSP-4 5-46
5-67. Comparison of online GC measurements of PCE and chloroform in SGP9 at 6 ft 5-47
5-68. Electret indoor air radon concentrations for the first intensive round 5-49
5-69. Electret indoor air radon concentrations for the second intensive round 5-49
5-70. Radon concentrations from the downstairs stationary Alphaguard during the second
intensive round 5-50
5-71. Radon concentrations from the upstairs stationary Alphaguard during the second
intensive round 5-50
5-72. Electret indoor air radon concentrations for the third intensive round 5-51
5-73. Radon concentrations from the downstairs stationary Alphaguard during the third
intensive round 5-51
5-74. Radon concentrations from the upstairs stationary Alphaguard during the third intensive
round 5-52
5-n
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5-75. Temperature records from the external temperature monitor and the HOBO devices at
seven indoor locations on the 422 and 420 sides of the house 5-54
5-76. Indoor temperature as recorded inside the 422 second floor office 5-55
5-77. Stacked hydrological graph with depth to water in feet (top—red circles), discharge at
Fall Creek in ft3/s (middle—blue line), and rainfall in inches (bottom—green line) 5-56
5-78. Plot of barometric pressure (inches of Hg) external to the 422/420 house overtime 5-57
5-79. Barometric pressure (Pa) on the 422 side of the house overtime 5-58
List of Tables
5-1. Frequency of Nondetects (%) by Soil Gas Point or Cluster 5-9
5-2. Frequency of Nondetects by Depth and Compound 5-9
5-3. Summary Meteorological Data for Central Indiana 5-53
5-in
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5. Results and Discussion: VOC Concentration Temporal Trends
and Relationship to HVAC
5.1 VOC Seasonal Trends Based on Weekly, Biweekly, and Monthly
Measurements for 52+ Weeks
5.1.1 Indoor Air
Figures 5-1 and 5-2 show PCE and chloroform versus time, respectively, at all seven air monitoring
locations (including ambient). PCE levels at all six locations follow the same general trend of starting
higher at the beginning of the project, dropping to a low in spring, and rising slightly and leveling out
through the end of the project (see Figure 5-1). However, the timing of the spring minimum differed
substantially for the unheated side of the duplex (when it occurred in late March) from the heated side of
the duplex (where the minimum was reached in July). The highest readings were generally found at 422
basement south except during brief periods when first floor concentrations were higher (mostly periods of
fan operations, see section 12.2).
Chloroform's behavior can be summarized in four main trends (see Figure 5-2):
1. Broadly, the six indoor locations show a general concentration decline from a localized maximum
at the beginning of the sampling interval in January 2011. The minimum is reached at the end of
spring on the 422 side of the house (early July), as with PCE. Also similar to PCE's behavior, the
chloroform minimum on the 420 side of the house occurs much earlier in the year (March).
2. However, the levels at the 422 first floor sampling location rise abruptly to a maximum in March
2011 immediately after the first brief drop in January (see Figure 5-2). During this maximum, the
first floor concentrations exceed those of even the basement stations. The 422 basement sampling
stations show a less dramatic rise in this period.
3. Chloroform concentrations reached a minimum in July 2011 and began steadily increasing
thereafter, forming a generally U-shaped curve. The winter 2012 levels more closely approach
their original highs than do the corresponding PCE results.
4. The second maximum for chloroform occurs in October 2011 for the 420 (unheated) locations
and is followed by a considerable decline through the winter months. The second peak occurs
later (December 2012) on the 422 (heated) side of the duplex and concentrations stay near that
maximum until February 2012.
With the exception of the elevated chloroform from late February to late March 2011, the highest
chloroform levels were found at 422 basement south, the same station that was generally highest for PCE
(Figures 5-1 and 5-2).
Figure 5-3 shows PCE, benzene, and toluene at 422 basement south versus time, along with ambient
levels of benzene and toluene. Although both benzene and toluene are above their action levels (benzene
= 0.31(ig/m3; toluene = 0.0052(ig/m3; EPA Regional Screening Level (RSL) Summary Table, Nov.,
2011), each tends to trend similarly to its respective ambient; this is not the case with PCE, which is
almost always considerable higher than its ambient. This suggests that benzene and toluene at this
location are likely dominated by regional ambient sources, not soil gas vapor intrusion.
Figure 5-4 shows concentrations for 7-, 14-, and 28-day durations over the extent of the project at the 422
first floor sampling station versus time for both PCE and chloroform. Generally, sampling at all three
durations shows the same trends, the only exception being the brief fan tests that influence the weekly
5-1
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCEin Indoor and Ambient Air vs. Time
E
5
:.-;
9
5
2
Figure 5-1. PCE in indoor and ambient air vs. time (7-day Radiello samples).
Chloroform in Indoor and Ambient Air vs. Time
Q
«
•
3
Figure 5-2. Chloroform in indoor and ambient air vs. time (7-day Radiello samples).
5-2
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCE, Benzene and Toluene in Indoor and Ambient Air
m
I
. o o
cRfeo
0 O O
'0 ouo
o
o
o
:
* rf>-CL
\ \ \
Date
Figure 5-3. PCE, benzene, and toluene in indoor and ambient air.
Weekly, Biweekly, and Monthly Radiellosat 422 First Floor vs. Time
$
(M»
Figure 5-4. PCE and chloroform in 422 first-floor indoor air; weekly, biweekly,
and monthly duration Radiello samples.
5-3
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
samples more dramatically than the longer duration samples (see Section 12.2). The comparison of
absolute concentrations measured by samplers of different durations is discussed in Section 9.
5.1.2 Subslab Soil Gas
Data are presented in this section from seven subslab ports (SSPs), numbered 1 through 7, and four wall
ports (WPs), numbered 1 through 4. On the 422 side of the house are SSP1, 2, and 4 and WP1 through 3.
Given its low initial concentration and nearness to soil gas port (SGP) 10-6, SSP2 was sampled relatively
infrequently. On the 420 side of the house are SSPS, and 5 through 7, and WP4. The basements of both
sides of the duplex are each divided into thirds in the interior. There is generally one subslab port per
basement division, with one section on the 420 side having two. The wall ports are located on the exterior
walls of the duplex. WP1 and 3 are each located in the centers of the north and south ends of the 422
basement, and WP-2 is in the center of the east side of the 422 basement. WP-4 is located in the center of
the west wall of the 420 basement.
Figures 5-5 and 5-6, and 5-7 and 5-8 are chloroform and PCE concentrations versus time, respectively,
with the Figures 5-6 and 5-8 representing intensive sampling periods.1 For chloroform as shown in
Figure 5-5, most of the ports on the unheated 420 side (the various crosses and the square) are generally
stable for most of the duration of the project. Notable exceptions are the vertical alignment of data points
on the plot (indicating concentration variability over a short time period) That occur during intensive
periods of sampling.
Concentration ug/m3
i-1
£ ° S 8 §
7 M M o o o o
Subslab Chloroform Concentrations
^flp* x* \\^t.
^ X
*F ttA
+*+ + ti.
m •>•- d* ™ S xP
i • _I_
K J^
t x «
X
DOOh-»OOOl-»O
^LnOOi-»NicnuDNi4ii
JOh-1l-1NiOl-1NiO
r ^ S § Date S S § > S
DOOOh^Ph^PK
hloroform SSP-2 Chloroform ASSP-3Chloroform XSSP-4 Chloroform XSSP-5 Chloroform •SSP-6Chloroform SSP-7 Chloroform
Figure 5-5. Plot of subslab chloroform concentrations vs. time.
During the normal times, the subslab samples were collected during regular daytime working hours, while the
intensive periods involved two shifts of personnel, allowing up to three samples to be collected, generally early
morning, midday, and evening.
5-4
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration u.g/m3
» ^6
t/i ,_. h^ O O
7 ° ^ S 8 8 8
1-1
Subslab Chloroform Concentrations First Intensive Round
t x N * x x x
r « • t * » »
I
iji j_
J M -
f
DOOOOOOO
hloroform SSP-2 Chloroform ASSP-3Chloroform XSSP-4Chloroform XSSP-5 Chloroform •SSP-G Chloroform SSP-7 Chloroform
Figure 5-6. Plot of subslab chloroform concentrations vs. time, first intensive sampling period.
This may indicate that there is a diurnal pattern in the subslab sampling that is only perceptible during the
intensive periods (Figures 5-6 and 5-8).2 Another notable observation on the 420 side occurred from July
14, 2011, to August 3, 2011, between the time when thieves stole the house air conditioners (ACs) (both
sides) and when they were replaced (422 side). Chloroform approached its highest levels on the 420 side
during this time. Chloroform on the 422 side (shown in Figure 5-5 as the circles, diamonds, and
triangles) traces a rough sinusoidal trend over months, although the different ports are somewhat out of
phase. These trends generally show lows during the warmer months (SSP-1 and SSP-4 seem to both reach
a minimum in August/September 2011 and highs during cooler months). It is also notable that the
concentration increases abruptly two orders of magnitude between August 27 and September 8, 2011, a
period of time during which a series of fan tests (coded B and F) intended to simulate the stack effect
expected under winter conditions were conducted (as discussed in Section 12.2). Another smaller rise
occurs from September 30 to October 14. Fan test "I" was conducted from October 6 to October 14.
In both cases, the subslab ports on the 422 side have higher concentrations of PCE and chloroform than
those on the 420 (unheated) side of the structure. In Figure 5-7, SSP PCE concentrations versus time
more prominantly display a simple pattern of high and low concentration changes during warmer and
cooler months across the range of ports. Most of the ports on the 420 side of the house and SSP-4 on the
422 side show highs during the warmer months and lows during the cooler months. A notable exception
is SSP-1, which shows the opposite trend to all the others.
During normal times, the subslab samples were collected during regular daytime working hours, while the
intensive periods involved two shifts of personnel, allowing up to three samples to be collected, generally early
morning, midday, and evening.
5-5
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Subslab PCE Concentrations
§
O
O
°
O
O
on ug
Concen
O
O
M
O
o
P
M
%>
v
^^PW(^«^PK
+
X « * X
SSP-1PCE SSP-2PCE ASSP-3PCE XSSP-4PCE XSSP-5 PCE »SSP-6PCE SSP-7 PCE
Figure 5-7. Plot of subslab PCE concentrations vs. time.
10000
1000
m
^
bn i
i 100
c
o
£
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Neither compound when graphed for the wall ports (Figure 5-9 nor Figure 5-10) shows the same clear
patterns of highs and lows found during the changing seasons in Figures 5-5 and 5-7. Figure 5-9 plots
chloroform concentrations at the four WPs versus time. Most ports are generally stable throughout the
project time period. As with the subslab samples, vertical alignments of data points on the plot occur
during intensive rounds. These chloroform levels do not show the same kind of spike during the period
when the ACs were stolen as for SSP chloroform. Highs for WP-3 in January through February and
September through October 2011 seem to suggest influence of the snow and ice and fan testing,
respectively. The greater temporal flucations of the wall ports as compared with the subslab ports may be
attributable to their more shallow depths (approximately 1.5 ft bis).
Figure 5-10 plots PCE concentrations at the four WPs versus time. Most are generally stable over time.
Vertical alignments of data points are seen during the intensive rounds of sampling. The high
concentrations of PCE in WP-3 at the beginning of the project could be due to the snow and ice capping
event during the severe winter of January and February 2011. Highs in September and October might be
attributable to the fan testing during that time.
Wall Port Chloroform Concentrations
10000
1000
00
o
01
u
o
u
100
10
0.1
AXA
A
•A*
A A
XAXXM XX XXX *X*AX X XXXXXXX X
X
t? Date t?
WP-l Chloroform
WP-2 Chloroform A WP-3 Chloroform XWP-4 Chloroform
Figure 5-9. Plot of wall port chloroform concentrations vs. time.
5-7
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration ng/m3
o S 8
P h^ O O O
h-» h-1 O O O O
nT /OT /TT
Wall Port PCE Concentrations
X
A
A
» X
'
AA A A
X A A A*
X XX XxSxXXXX V.xmK**XX XtULXM. XXXXXXX)tilD§;XXXXXXXXX
» x
^OOOOOOh^h^OO
D B t? B B Date ^ fj ^ ^ ^ ^
WP-1PCE WP-2PCE AWP-3PCE XWP-4PCE
Figure 5-10. Plot of wall port PCE concentrations MS. time.
5.1.3 Shallow and Deep Soil Gas
A series of 12 nested SGPs surround the 422/420 house or originate in the basements of either side of the
duplex. The five depths at each of the external nested locations are as follows: 3.5 ft bgs, 6 ft bgs, 9 ft bgs,
13 ft bgs, and 16.5 ft bgs. At the internal nested locations there are only four depths; the 3.5-ft depth is
omitted because the basement floor is at ~5 ft bgs. External to the house, there are seven nested locations,
notated SGP1 through 1'. Internal to the house are the nested locations notated SGPS through 12. Each
individual port is notated based on its location and its depth (e.g., SGP1-3.5 for the 3.5-ft depth at the
SGP1 location; see Figures 5-11 through 5-32). Groundwater levels varied throughout the project but
remained high enough most of the time to render the 16.5-ft probe depths inaccessible for soil gas
sampling for much of the project.
Concentrations are generally highest in the deepest ports of each cluster and decrease at shallower depths.
This pattern is consistent with expectations for attenuation of vapor intrusion of VOCs originating from a
deep source (whether in the vadose zone or groundwater). This attenuation pattern appears to be more
pronounced for chloroform (frequently two to three orders of magnitude) than for PCE (generally one
order of magnitude).
An analysis of the frequencies of nondetects was performed for each compound by borehole and depth.
Of the boreholes outside the house footprint, only SGP1 (just south of the 422 part of the duplex) had less
than a 20% frequency of nondetects for both PCE and chloroform (Table 5-1).
5-8
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Table 5-1. Frequency of Nondetects (%) by Soil Gas Point or Cluster
Location Benzene Chloroform Hexane PCE
Toluene TCE
SGP1
10
69
66
58
67
63
74
68
71
63
61
69
70
74
54
80
50
61
68
65
62
66
68
76
68
77
79
87
84
85
67
71
81
82
85
52
100
90
58
86
87
66
82
83
70
All of the wall ports have more than 20% nondetects for all compounds except benzene (Table 5-2).
Table 5-2. Frequency of Nondetects by Depth and Compound
Benzene Chloroform Hexane Tetrachloroethene Toluene Trichloroethene
Wall Port
Sub-Slab
3.5
^B 6
9
13
16.5
Nondetects are infrequent (<20%) in almost all the SSPs for PCE but more frequent for chloroform and
the 420 side of the duplex. Interestingly many subslab ports are consistently detectable (>80%) for
benzene as well. Benzene is also consistently detectable at the 6 and 16.5-ft depth intervals. For the trend
in nondetects by depth, we see about what we would expect for a deep vapor intrusion source; there are
fewer nondetects at lower depths. As mentioned before, PCE is under 20% nondetects at the subslab level
(depth = 5 ft) and from 6 ft down in the deeper soil gas samples. Chloroform had fewer than 20%
5-9
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
nondetects only at depth deeper than 13 ft. Benzene was under 20% nondetects only at the deepest depth
of 16 ft. No other compounds were consistently detectable.
Thus, the shallowest depths (3.5 ft and 6 ft) were generally the most stable, with little fluctuation because
most results were below the detection limit. The 9-ft depths had periods of stability as well (see Figures
5-15, 5-16, 5-21,5-27,5-29, and 5-31). Notable exceptions to the shallow stability can be found at SGP1
and, to differing degrees, all of the indoor ports, SGP8 through!2 where the shallow concentrations were
higher and thus less affected by nondetects. At each of those ports, shallow concentrations seem to
partially track the seasonal variations of the deeper ports (see Figures 5-11, 5-12, and 5-25 - 5-32). At
SGP3 and 4, the deeper ports are often low or stable (see Figures 5-15 through 5-18).
Many of the deeper ports at each location (9 ft through 13 ft, sometimes 16.5 ft) show what appears to be
a rough cycle responding to seasonal changes (see Figures 5-19, 5-20, 5-23, 5-24, 5-25, and 5-29), that is,
concentrations are higher in the cooler months and lower in the warmer. SGP3 and 4 are too diffuse to
show much of atrend (see Figures 5-15 through 5-18). SGP1 and 2 show the opposite trend, especially
for chloroform at SGP1-13 and PCE at SGP2-9 (see Figures 5-11 and 5-14). To a degree, a negative
trend can also be seen for PCE at SGP8-6, and chloroform at SGPs 11-13 and 12-13 (see Figures 5-26
and 5-30).
Some prominent points among the figures might be attributed to natural or project-related phenomena.
Although samples were taken multiple times per week, and in some cases per day, during the intensive
rounds (yielding as many as >12 successive samples at some locations during a week), there were no
discernable or notable trends in the data. This suggests that there is probably not a strong diurnal variance
in subslab soil gas concentrations at this duplex and that the frequency of sampling (and thus the artificial
volumetric flow in the subsurface induced by frequent sampling) does not appear to be significant (for
example, see Figures 5-33 and 5-34). High concentrations found at the beginning of the project but
tapering off toward the spring could be due to the period of heavy snow and ice in the very cold winter of
January and February 2011 (for example, see Figure 5-14). In all of the PCE figures (even numbered
Figures 5-12 through 5-32), there is a cluster of points offset above the long term from mid-May to early
July visible. This is also visible in the wall port PCE plot presented in the previous section. It is possible
that this clustering resulted from drastic temperature fluctuations that occurred during that time. For
example, the low on May 16, 2011, was 43.8 degrees F, and the high roughly 2.5 weeks later was 95.5
degrees F on June 4,2011.
5-10
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
m
00
i 100
o
Concentrat
p ^
h^ h^ O
>/Tn
SGP1 Chloroform Concentrations
X
A ^ A* A^A ^^^A\
JPA A A A A v A^^
•"A 4 ^
• A A
.
• • ^B
DOOh-»OOOl-»O
^ ^ > > Date S S > i£ S
0000 "«"«- ^ ,_, ^ ^ M
»SGPl-3.5Chloroform SGP1-6 Chloroform ASGPl-9Chloroform XSGP1-13 Chloroform XSGPl-16.5Chloroform
Figure 5-11. Chloroform concentrations at each of the SGP1 ports vs. time.
10000
1000
m
£
i 100
o
Concentrat'
p i-
h^ h^ O
SGP1 PCE Concentrations
X
X v ^^^^^
X ^ ^X^ A vA^'^Mu AAA >^v< X
A* f*X^^^^XCA nJ^/^^AA
iXBl^^ • A B BB^^
B B X
X
DOOh-»OOOl-»O
S o o o uale P P P P G
»SGP1-3.5PCE BSGP1-6PCE ASGP1-9PCE XSGP1-13PCE XSGP1-16.5 PCE
Figure 5-12. PCE concentrations at each of the SGP1 ports vs. time.
5-11
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
SGP2 Chloroform Concentrations
m
E
i 100
c
o
+J
TO
L.
+J
u
c
o
u
x
X X X X., Xx X ., X,,
xxx xx xxxxx
A
A A. A *! AA
AjA A^T1^^ AA ^ ^A^A A * ^^
AX A A
A
A *
* ** * **M*** ***** ***** **** *•• *•** «K* ** * f *«** * *
• SGP2-3. 5 Chloroform SGP2-6 Chloroform ASGP2-9 Chloroform XSGP2-13 Chloroform XSGP2-16.5 Chloroform
Figure 5-13. Chloroform concentrations at each of the SGP2 ports vs. time.
Concentration ng/m3
S
p S 8 8
h^ h^ O O O O
nT /OT /TT
SGP2 PCE Concentrations
AA
AMA
*" X X A XAA iW ^ ^^ A
X X 4 '
A
*
************** ******* * ** *•»*******<* ****^ ***** *
«
•*OOOOOOl-»l-»OO
j|_i[\jjiCn-vJUDi-»NiNi4ii
^ONi|-1ONi|-1ONi|-1O
0-vJCTl-vJCTlCTl^UJUJI— »l— »
^ h-1 h-1 h-1 h-1 DStG h-1 h-1 h-1 h-1 h-1 h-1
Dl-1!-1!-1!-1!-1!-1!-1!-1^!^!
SGP2-3.5 PCE SGP2-6 PCE ASGP2-9 PCE X SGP2-13 PCE X SGP2-16.5 PCE
Figure 5-14. PCE concentrations at each of the SGP2 ports vs. time.
5-12
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration ng/m3
o 6 8
° ^ S 8 8 8
nT /OT /TT
SGP3 Chloroform Concentrations
* X
A x
X X
xxx
^ 4 X •
«: Hi* *********** A** •* *** A» A* A ^
4
A/
A A A
»*•»»« MftggA A AAA A A
^OOOOOOh^h^OO
SGP3-3. 5 Chloroform SGP3-6 Chloroform ASGP3-9 Chloroform XSGP3-13 Chloroform XSGP3-16.5 Chloroform
Figure 5-15. Chloroform concentrations at each
Concentration ng/m3
p S 8 8
h-» h-» O O O O
of the SGP3 ports vs. time.
SGP3 PCE Concentrations
x
A
A
X
ft Hat K«i*Jk*ft* ft* AAA »* AA* A«* iSi it A
X
A
X A A
AAAAAXMXAAAA A AAA A A
•*OOOOOOh-»l-»OO
? § § § § Date § § § § G G
»SGP3-3.5PCE BSGP3-6PCE ASGP3-9PCE XSGP3-13PCE XSGP3-16.5 PCE
Figure 5-16. PCE concentrations at each of the SGP3 ports vs. time.
5-13
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
m
E
hiQ
IJ* 1QQ
c
o
+J
£5
+J
U
£
O
1
h
,0
E
SGP4 Chloroform Concentrations
A
A
XX XA AXA X XX
* ti* A/^
AA* • A*
1
•
•*OOOOOOl-»l-»OO
j|_i[\jjiCn-vJUDi-»NiNi4ii
i'ONi|-1ONi|-1ONi|-1O
0-vJCTl-vJCTlCTl^UJUJI— »l— »
^ h-1 h-1 h-1 h-1 DStG I-1 I-1 I-1 I-1 I-1 I-1
3 ,_i ,_i ,_i ,_i "-"Mfc*. |_i |_i |_i |_i |sj |sj
SGP4-3. 5 Chloroform SGP4-9 Chloroform ASGP4-13 Chloroform XSGP4-16. 5 Chloroform
Figure 5-17. Chloroform concentrations at each of the SGP4 ports vs. time.
Concentration ng/m3
S
p S 8 8
l-» l-» O O O O
SGP4 PCE Concentrations
A A
A
• -* • -A*^^%A:-A. ^A- .
i"*X ^A AAAAAAO BB — . A
«• • •• ••• HA •••••• • *
•
•*OOOOOOh-»l-»OO
i|-i[\J4^Cri-vJUDi-»NiNi4ii
^ONi|-1ONi|-1ONi|-1O
o-vjcn-vjcncrijiUJUJi-1!-1
^ h-1 h-1 h-1 h-1 DSt6 h-1 h-1 h-1 h-1 h-1 h-1
-, ,_» ,_» ,_» ,_» I^tlLt ,_» ,_» ,_» ,_» ^j ^j
SGP4-3.5 PCE SGP4-9 PCE ASGP4-13 PCE XSGP4-16.5 PCE
Figure 5-18. PCE concentrations at each of the SGP4 ports vs. time.
5-14
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
SGP5 Chloroform Concentrations
1000
m
£
cuo
c
O
+J
k.
+J
u
c
o
u
h
XY %
J^C x^_ .fX v x ^XXXX^X
X x^ v>0 ^ ^^ X ^^ X ^v^X X
X ^ X v
^ ^< ^K T'^'; ™ ?* X ^
A A . X X AAA AA A
*A
A 4 A
A A*
AA A A A
"
•*OOOOOOl-»l-»OO
•tt-tt-tl-^-Sl-^-^-tt-tl-tl-^
*SGP5-3. 5 Chloroform SGP5-6 Chloroform A SGP5-9 Chloroform XSGP5-13 Chloroform XSGP5-16.5 Chloroform
Figure 5-19. Chloroform concentrations at each of the SGP5 ports MS. time.
Concentration ng/m3
S
p S 8 8
h-» h-» O O O O
nT /OT /TT
SGP5 PCE Concentrations
x
x
•
A XA±A J^AA^** AA
A AA AA+4AAAAA AA*A A A A A A A A4> A+AA A 4* A* 4* 4*4444* ft A AAA A A
** •
•*OOOOOOl-»l-»OO
j|_i[\J4^CTi-vJUDi-»NiNi4ii
^'ON)|-1ON)|-1ON)|-1O
0-vJCTl-vJCTlCTl^UJUJI— »l— »
j |_i |_i |_i |_i Date h-1 h-1 h-1 h-1 h-1 h-1
Dh- '1— '1— '1— '1— '1— '1— '1— » N) N)
SGP5-3.5PCE SGP5-6PCE ASGP5-9PCE XSGP5-13PCE XSGP5-16.5 PCE
Figure 5-20. PCE concentrations at each of the SGP5 ports MS. time.
5-15
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
SGP6 Chloroform Concentrations
m
£
oo
i 100
o
4-J
E
u
c
o
u
h
A
v* v ** X X *>K ^ ^^XX
xA xx x ' xxx^x x
^ \s X
A
A A A ^ A A
* -•
AA I
• A **» ^ AA A ABIH^ • •»»»••«*•««***** • •!• A •
A
^OOOOOOh^h^OO
h^^SJJiOI^JJ.O^SJSJJi
00 ^J 01 ~J 01 01 ^ LO LO ^ ^
Oh- 'h- 'h- 'h- 'h- 'h- 'h- 'h- 'N1N1
»SGP6-3.5Chloroform SGP6-6 Chloroform ASGP6-9 Chloroform XSGP6-13 Chlorform XSGP6-16. 5 Chloroform
Figure 5-21. Chloroform concentrations at each of the SGP6 ports vs. time.
SGP6 PCE Concentrations
O
i
O
h-»
8
O
Concentration ng
S
O O
p
h-»
• •••••••••«
OOOOOh-»
[\J4^Cn-vJUDi-»
Ni|-1ONi|-1O
cn-vjcncnj^uJ
> > > Date >>>
1— » 1— » 1— » 1— » 1— » 1— »
SGP6-3.5PCE SGP6-6PCE ASGP6-9PCE XSGP6-13PCE XSGP6-16.5 PCE
Figure 5-22. PCE concentrations at each of the SGP6 ports MS. time.
5-16
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
m
OO
i 100
g
1
£ 10
u
c
o
u
1
0.1
i-
^
.a
i-
c
»SGP7
SGP7 Chloroform Concentrations
* M^ ~*^
*** jf* ">K
x ' x* * ^xxxxx x *XA A>
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration ng/m3
S
p S 8 8
h-» h-» O O O O
SGP8 and SGP9 Chloroform Concentrations
*•**•xixxx xt^x^ •'%o«** t x*+ x
*• *** •"* ? XxX
x •
•«•••• •» K
^OOOOOOh^h^OO
^ h-1 h-1 h-1 h-1 0316 h-1 h-1 h-1 h-1 h-1 h-1
»SGP8-6 Chloroform SGP8-9Chloroform ASGP8-13 Chloroform XSGP8-16.5 Chloroform
XSGP9-6 Chloroform •SGP9-9Chloroform SGP9-13 Chloroform SGP9-16.5 Chloroform
Figure 5-25. Chloroform concentrations at SGP8 and9 ports vs. time.
SGP8 and SGP9 Chloroform Concentrations
10000
1000
100
O
"+J
(G
53 10
u
c
o
u
0.1
•*••
x x
x ^ . +
"A'li ^
»*+ «
^ Date >
SGP8-6 Chloroform SGP8-9Chloroform ASGP8-13Chloroform XSGP8-16.5 Chloroform
XSGP9-6 Chloroform »SGP9-9 Chloroform +SGP9-13 Chloroform SGP9-16.5 Chloroform
Figure 5-26. PCE concentrations at SGP8 and 9 ports vs. time.
5-18
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration ng/m3
o 6 8
° ^ S 8 8 8
SGP10 Chloroform Concentrations
•
X*XX- A ^ A*
. ^ : - 4 .-"«"•" ' H.' rA -
•*%• ''""'..»..
x
+ • +•
• • ^••^ ••* • «»••» «• •• • *• •••••^^•^ •
•
-»OOOOOOh-»l-»OO
-i|_i|\j.CiCn-vJUDl-»NlNl.ei
^ON1|-1ON1|-1ON1|-1O
o-vjcn-vjcncnjiUJUJi-1!-1
^ h-1 h-1 h-1 h-1 Date I-1 I-1 I-1 I-1 I-1 I-1
3 l_i l_i l_i l_i —«-•••*" |_i |_i |_i |_i |sj |sj
*SGP10-6 Chloroform SGP10-9 Chloroform ASGP10-13 Chloroform XSGP10-16. 5 Chloroform
Figure 5-27. Chloroform concentrations at each of the SGP10 ports vs. time.
Concentration ng/m3
h^ O
P S 8 8
h^ h^ O O O O
DT/QT/TT
SGP10 PCE Concentrations
*A AAAAAAAAAAAAA A
«-o:;;- /^^^•rt*.
•> * * P" ••• B i
• AH •
»OOOOOOl-»l-»OO
i|_i[\jjiCn-vJUDi-»NiNi4ii
»ONi|-1ONi|-1ONi|-1O
3 -vj cn -vj cn |^a*o en 4^ ui ui i-» i-»
ll_ll_ll_ll_ll_ll_ll_ll_l[\JI\J
4SGP10-6 PCE SGP10-9 PCE ASGP10-13 PCE XSGP10-16.5 PCE
Figure 5-28. PCE concentrations at each of the SGP10 ports vs. time.
5-19
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
SGP11 Chloroform Concentrations
m
£
1 —
00
i 100
c
o
4J
E
+J
S 10
U
c
o
u
1
\-
\-
X
x AA A
X ^ AAA A A jA
• -.I*"' """•**• , » """" '"'•••.
A^m * * H^
• 4^ + * w
* • • A •B»l^4 • •••••
•*OOOOOOh-»l-»OO
1— » O N) 1— * O N) 1— * O N) 1— 'O
CO -vj CT1 -vj CT1 r^Q+A O*i •& UJ UJ 1— » 1— »
SPPPPtitititiGG
SGPll-6 Chloroform SGP11-9 Chloroform ASGP11-13 Chloroform XSGP11-16. 5 Chloroform
Figure 5-29. Chloroform concentrations at each of the SGP11 ports MS. time.
10000
1000
m
£
00
i 100
o
Concentrat
p ^
h^ h^ O
SGP11 PCE Concentrations
X ^ » ^d__^M A
****!**** " •
X
•*OOOOOOh-»l-»OO
CO *~J C71 *~J C71 r^*)+A C71 ^ UJ UJ h- ' h- '
Oh- 'h- 'h- 'h- 'h- 'h- 'h- 'h- 'N1N1
4SGP11-6PCE BSGP11-9PCE ASGP11-13PCE XSGP11-16.5 PCE
Figure 5-30. PCE concentrations at each of the SGP11 ports MS. time.
5-20
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
m
Concentration jigy
S
o o
1
0.1
I-
•t
1
SGP12 Chloroform Concentrations
x
v AAAAH A |AAA A
^ XX x i*A AA A AA AAAAAAA
. - . fP/ " "B*^-.- B/ % VA
4 ******* ******* * **.
AAAAAAAA
•
* * **
t^^^^^^^
^ o NI h-1 °rv|VJ ^ °
•~ • — . • — . • — . • — . uaie • — . • — . • — .
Oh-»h-»h-»h-»h-»h-»h-»
Nl h-1 O
h-1 Nl Nl
»SGP12-6 Chloroform SGP12-9 Chloroform ASGP12-13 Chloroform XSGP12-16.5Chloroform
Figure 5-31. Chloroform concentrations at each of the SGP12 ports vs. time.
10000
1000
m
£
i 100
c
E
+J
S 10
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Subslab PCE Concentrations First
on
E
"So >
3. 100
c
0
4-1
re
4-1
Intensive Round
> • • t •
< x X g x
•
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5.2 Radon Seasonal Trends (based on Weekly Measurements)
5.2.1 Indoor Air
Radon for indoor air was recorded on electrets from E-Perm in Fredrick, MD, or on Genitron Alphaguard
units. The Alphaguard units were kept in stationary locations, one in the 422 second floor office and one
in the 422 north basement. The electrets from all locations were read once per week on the same day,
except during intensive rounds, when they were read once per day (only the weekly readings are included
here; see Figure 5-35). The stationary Alphaguards were set to read continuously every 10 minutes and
their data downloaded once per week (see Figures 5-36 and 5-37).
The electret readings are fairly stable over a 1-year period, beginning and ending at a similar
concentration. The general pattern for high concentrations is ~10 pCi/L in the cooler months and ~5 pCi/L
in the warmer months. The high for the winter months was 12.22 pCi/L at 422 basement south in week 8,
and the low for the summer months was 0.15 pCi/L at 422 first in week 33 (see Figure 5-35).
The Alphaguard units were not brought online until the end of March 2011, so the electret figure and the
Alphaguard figures cannot be aligned directly; the Alphaguard figures begin with week 13 on the electret
figure (see Figures 5-35 through Figure 5-37). Both Alphaguard figures show considerably more
fluctuation than the electret figure because of their constant taking of readings. The Alphaguard figures
also show a rough downward trend toward the warmer months and a rise toward the cooler months, with
some fluctuation possibly due to weather changes (see Figures 5-35 and 5-37).
The intensive rounds, the main fan test period, and the period when the AC units were gone are marked
on the figures (see Figures 5-35 through 5-37). The intensive rounds were included more to give an idea
of the conditions taking place during each of the rounds rather than to suggest that intense sampling
changed the normal patterns of readings.
On Figure 5-35, the general trend is to decrease toward the warmer months (weeks 20-36), but some
areas do not fit the pattern. The AC units were installed on both sides of the duplex on June 29, 2011
(week 26). After this, readings hit a low in week 27. All ACs were stolen by July 15, 2011, and missing
until replaced on the 422 side on August 3, 2011 (weeks 28 through 30). It is during this time when the
ACs are missing that radon levels reached their highest for the summer months (Figure 5-35). A possible
explanation for this is from the solar stack effect—the hot sun heats up the air in the higher stories of the
house, producing an effect similar to what is achieved with the heater in winter. Something similar can be
seen from weeks 23 through 26 just before the ACs were initially installed (see Figure 5-35). The
readings start to rise and then fall sharply after the ACs were installed. Some other prominent points
during the summer occur during the fan test period, when the basement was depressurized, increasing
radon infiltration (see Figure 5-35).
J^adon, in Figures 5-36 and 5-37, appears to reach much lower concentrations in warmer months than
cooler ones; however, radon also appears to fluctuate less wildly in warmer months. Both figures show a
reaction to the loss of the AC units, but the upstairs office Alphaguard shows a more immediate and
definite response (see Figures 5-36 and 5-37). Both Figures 5-36 and 5-37 show the effects of the fan
tests on the stationary Alphaguard readings, but, again, the upstairs office Alphaguard shows a better
defined response to the fan testing. The prominent peaks are not the same for both, suggesting fan tests
had variable effects on different regions of the house (see Figures 5-36 and 5-37).
5-23
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Figure 5-35. Weekly electret readings for all locations.
Arrows indicate intensive-round weeks. Black bars indicate the period when the ACs were missing. Red bars indicate
the main fan test period. One reading extended beyond the current range of the figure. It reached 31.9 pCi/L and is
indicated by the arrow and text. The original range was truncated to better view the data.
5-24
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Downstairs Alphaguard Data
AC Fan
Loss Testing
Date/Time vs Rn (pCi/L)
Date
Figure 5-36. Data for the downstairs continuously recording Alphaguard
versus time.
Data were recorded every 10 minutes. Note that readings did not begin until the end of March 2011. The two
intensive rounds within this data range are indicated by horizontal bars. Black vertical bars mark the period of AC
loss. Red bars indicate the fan test period.
5-25
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Upstairs Alphaguard Data
AC Fan
Loss Testing
^
Date/Time vs Rn (pCi/L)
Date
Figure 5-37. Data for the upstairs continuously recording Alphaguard
versus time.
Data were recorded every 10 minutes. Note that readings did not begin until the end of March 2011. The two
intensive rounds within this data range are indicated by horizontal bars. Black vertical bars mark the period of AC
loss. Red bars indicate the fan test period.
5.2.2 Subslab and Wall Port Soil Gas
There were seven SSPs and fourWPs at the 422/420 house. SSPs-1, -2, and -4, and WPs-1 through -3
were on the 422 side, and SSP-3, -5, -6, and -7, and WP-4 were on the 420 side. Radon readings at most
ports were taken each week with a handheld Alphaguard unit.
Figure 5-38 shows radon concentrations as they were distributed around the 422/420 house
(superimposed within the kriged image) during the first week of data collection. The general pattern is
that radon is at its lowest concentration closer to the surface and increased to 1,000 pCi/L through 1,200
pCi/L at greater depths (see Figure 5-38). However, there is a zone of much higher concentrations from
about 6 ft through 9 ft (see Figure 5-38). This zone of higher concentrations appears to be greater toward
the southwest of the house (lower right of the figure).
Figures 5-39 and 5-40 show the radon concentration at SSPs and WPs for the duration of the project. The
radon concentration for the SSPs in Figure 5-39 is fairly stable among the ports with higher
concentrations (at or above 1,000 pCi/L): SSP-1, SSP-4, SSP-5, and SSP-6). Figure 5-39 shows that SSP-
5-26
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
4 and -6 have the highest concentrations. In contrast, Figure 5-40 shows a more diffuse concentration
pattern among all the WPs. Additionally, among WP-2 through -4, there appears to be a change in the
distribution pattern that is higher in the autumn and lower at all other times (see Figure 5-40). As with the
VOCs in Section 5.1.2., it is possible to see a vertical lines of measurements in Figures 5-39 and 5-40
that correspond to the intensive sampling rounds conducted in early March, August, and December. These
bars indicate a diurnal pattern of data variability only revealed by more frequent sampling, similar to what
has been observed in other radon studies (see Section 2 for references)..
There is a rough agreement between the general pattern for SSPs and WPs and the kriged radon of Figure
5-38. The WPs tend to have fairly low concentrations compared with subslab concentrations, and the port
of highest concentration is located in the southwest region of the house (see Figures 5-38 through 5-40).
However, Figure 5-38 shows the zone of higher concentrations as being deeper than just beneath the slab.
Also, SSP-6 has fairly high concentrations, but that is in an area of the northwest section of the house,
which should be low according to the kriged map (see Figures 5-38 and 5-40). As described in section 2,
unlike chlorinated VOCs, radon has a short half life (3.8 days) and therefore its subsurface concentration
is very influenced by the geologic materials immediately surrounding a sample point. Therefore these
anomalies could simply represent small-scale heterogeneities in subsurface materials with respect to their
radon generation potential.
3D View of Kriged Radon Data From
Subslab, Wall and Multi-depth Soil Gas
cncfrntrdbon Cut Btoika • W**k On*
Figure 5-38. This is a kriged radon image taken from subslab, wall, and multidepth soil gas data.
To the left is a key showing the color code for different radon concentrations. The image follows that color code. On
the image, a map of the 422/420 house is superimposed at depth. The bottom of the map faces south. Note that this
is just the first week's data.
5-27
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Subslab Alphaguard
o
ro
L.
+j
o
u
h-1 h-1 h-1 h-1 h-1 DSlG h-1 h-1 h-1 h-1 h-1
O h-» h-» h-» h-» *"•**•*- ,_» ,_» ,_» ,_» |xj
SSP-lAlphaguard SSP-2 Alphaguard ASSP-3 Alphaguard XSSP-4 Alphaguard XSSP-5Alphaguard «SSP-6 Alphaguard SSP-7 Alphaguard
Figure 5-39. Subslab Alphaguard data versus time.
Concentration (pCi/L)
M O
o MOO
P M 0 0 0
I-1 M 0 0 0 0
Wall Port Alphaguard
•
A A*A** *^t* ** X ** lt *«****•**»
A At • /• . XM ^lA5^x * •^•"KAXA^* A
A ^A*A ix^X *^* ^X . X" A «• AA A^
>A XA ^*AAX " X X%XX •M^JX™ A
• x •
x x
*OOOOOOl-»l-»OO
i|_i[\jjiCn-vJUDi-»NiNi4ii
iONi|-1ONi|-1ONi|-1O
D-vjcn-vjcncnjiUJUJi-1!-1
i |_i |_i |_i |_i Datp h-1 h-1 h-1 h-1 h-1 h-1
3 h-1 h-1 h-1 h-1 **» *-^ |_» |_» |_» |_» |SJ |SJ
WP-1 Alphaguard WP-2 Alphaguard A WP-3 Alphaguard XWP-4 Alphaguard
Figure 5-40. Wall port Alphaguard data versus time.
5-28
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5.2.3 Deep Soil Gas
There were 12 nested SGPs at the 422/420 house. At each of the seven ports exterior to the house, there
were five depths: 3.5 ft, 6 ft, 9 ft, 13 ft, and 16.5 ft. At the five interior ports, four of the five depths were
used (6 ft, 9 ft, 13 ft, and 16.5 ft), the shallowest (3.5 ft) being eliminated because of the depth of the
basement slab (~4 ft) beneath the ground surface. Radon data were taken each week with a Genitron
Alphaguard. The sampling strategy each week was to obtain a reading at the shallowest two ports at each
of the closest locations to the house exterior (SGP1-6 and -9; SGP4-9 and -13; SGP5-3.5 and -9; SGP7-
3.5 and -9), the four WPs, the SSPs (SSP-1, -4, -5, -6, -7), and the shallowest ports at each of the
basement SGPs (SGP8-6, SGP9-6, SGP10-6, SGP11-6, and SGP12-6).
Figure 5-41 is a kriged map of radon concentrations as they were distributed around the 422/420 house
during the first week of data collection. As previously discussed, radon is at its lowest concentration
closer to the surface and increased to about 1,000 pCi/L through 1,200 pCi/L at greater depths (see
Figure 5-41). However, there is a zone of much higher concentrations from about 6 ft to 9 ft (see Figure
5-41) toward the southwest of the house (lower right of the figure).
Figures 5-42 through 5-45 present the data from the ports external to the house, and Figures 5-46
through 5-50 are internal to the house. Most are generally stable throughout the duration of the project,
with little fluctuations. However, Figures 5-44 and 5-45 (SGP5 and 7) show what might be seasonal
(winter) lows for the 3.5-ft ports.
The data from Figures 5-42 through 5-50 agree fairly well with the kriged distribution of radon shown in
Figure 5-41. The locations with some of the highest concentrations are found toward the southwest of the
house (SGP1-6 and SGP9-6; Figure 5-42 and 5-47). However, SGP7-9 also has some higher
concentrations (Figure 5-45), which would not be expected from looking at the radon distribution
mapped in Figure 5-41.
The deeper ports at the soil gas locations were taken less frequently. As a result, only general
characteristics of the deeper soil gas activity can be inferred. Deeper soil gas was stable for the duration
of the project, except the 13-ft interval decreased overtime at SGP1 (see Figure 5-42), and the 13-ft
interval increased overtime at SGP5 and 9 (see Figures 5-44 and 5-47).
5-29
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
3D View of Kriged Radon Data From
Subslab, Wall and Multi-depth Soil Gas
Figure 5-41. Kriged radon image taken from subslab, wall, and multidepth soil gas data.
To the left is a key showing the color code for different radon concentrations. The image follows that color code. On
the image, a map of the 422/420 house is superimposed at depth. The bottom of the map faces south. Note that this
is just the first week's data.
5-30
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
SGP1 Alphaguard
1000
—I
Ll
c
O
4-1
TO
4-»
C
U
c
O
u
H
1-
. r.. ...._, T^-^s-:i^v
»K* * FTT ^A\ ^ *^ ^™A ~ * ^. A^AA^A ^ * A ^
\ > * A *
A^ •
A*
^
iOOOOOOMMOO
w°m^°m£SK£°
O H1 H1 H1 H1 L/dlC h-1 H1 H1 H1 N) N)
*SGPl-3.5 Alphaguard SGP1-6 Alphaguard A SGP1-9 Alphaguard XSGP1-13 Alphaguard XSGP1-16. 5 Alphaguard
Figure 5-42. Handheld Alphaguard data taken from soil gas ports
at location 1 versus time.
10000
1000
1
u
^ 100
c
o
Concentrat
9 ^ g
TT
SGP4 Alphaguard
^ .. vmm^m _„...
>C* A A A A^AAA A A ^ AAAAA ^AAAAA|* A*AAA
X
iOOOOOOMMOO
S 3 S 5 g£ K S K P g
o P P P £ uate p p p £ £ g
»SGP4-3.5Alphaguard SGP4-9 Alphaguard ASGP4-13 Alphaguard XSGP4-16. 5 Alphaguard
Figure 5-43. Handheld Alphaguard data taken from soil gas ports
at location 4 versus time.
5-31
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
1000
U
ntration (
; i
0
u
i-
o
c
SGP5 Alphaguard
^ •
•
*
• * ******
iOOOOOOMMOO
5 £ P S P Date P P S P K K
SGP5-3.5Alphaguard BSGP5-6Alphaguard ASGP5-9 Alphaguard XSGP5-13 Alphaguard XSGP5-16. 5 Alphaguard
Figure 5-44. Handheld Alphaguard data taken from soil gas ports
at location 5 versus time.
Concentration (pCi/L)
1-^ 0
(-1 h^ O O
9 i-^ooo
l-^l-^OOOO
SGP7 Alphaguard
AAAJP + A AAA A A A^ AiA AA A ^ AA^AA /A V* f AA^AA
X^ %/^l** **
* « •
* t* * * •* *
* * t » * «
. * * » *
» *
" *
* »
JOOOOOOMMOO
JMNJ-P>Cn--JlX>l-1NJNJ-P>
JONJMONJMONJMO
a-Jcn-Jcncn-piUiUii-1!-1
5 M M M M Datfi M M M M NJ NJ
SGP7-3.5 Alphaguard SGP7-6 Alphaguard A SGP7-9 Alphaguard XSGP7-13 Alphaguard XCSGP7-16. 5 Alphaguard
Figure 5-45. Handheld Alphaguard data taken from soil gas ports
at location 7 versus time.
5-32
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
u
0
2
flj 10
u
O
U
1
0.1
H
SGP8 Alphaguard
» *** ****** *^ ******* ^ ^*»**^**»»» *^*»^.»**4. »
^" ^
.
^OOOOOOMMOO
MONJMONJMONJMO
O M M M M L/aTc [_i [_i [_i [_i |\j |\j
»SGP8-6Alphaguard SGP8-9 Alphaguard ASGP8-13 Alphaguard XSGP8-16.5 Alphaguard
Figure 5-46. Handheld Alphaguard data taken from soil gas ports
at location 8 versus time.
10000
1000
U
_Q; 100
C
O
2
g 10
U
0
u
1
0.1
H
SGP9 Alphaguard
»**»» *•»* *** **** ****** *** ** 4* *$*******%* *<***** *« *
• f f
*& I
JOOOOOOMMOO
I P § P § Date § § B £ K K
»SGP9-6Alphaguard SGP9-9 Alphaguard ASGP9-13Alphaguard XSGP9-16.5 Alphaguard
Figure 5-47. Handheld Alphaguard data taken from soil gas ports
at location 9 versus time.
5-33
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
1
u
Q- 100
c
O
ns
4-1
c
QJ 10
u
c
O
u
0.1
H
SGP10 Alphaguard
t»<* * • ^
•iOOOOOOl-il-iOO
oo^jai^jaiai^LuLuH1!-1
H1 H1 H1 H1 H1 Date ^ ^ ^ ^ ^ ^
»SGP10-6Alphaguard BSGP10-9 Alphaguard ASGP10-13Alphaguard XSGP10-16.5 Alphaguard
Figure 5-48. Handheld Alphaguard data taken from soil gas ports
at location 10 versus time.
Concentration (pCi/L)
h^ O
p ^588
I-* I-* O O O O
n-r /n-r/TT
SGPll Alphaguard
***** *rtX ***• **».«* \** * ** *»**%»/»»* *S%/»* *
"* :*
* *
•^OOOOOOl-il-iOO
-*!-*N)-&CJl"vJlDl-1N)N)-&
^ON)l-iON)l-iON)l-iO
O-«JOl-«JOlOl-tiUJUJl-i|-i
^ HI HI HI HI Platp HI i-i i-i i-i i-i i-i
3 i_i i_i i_i i_i L^aic: |_i |_i |_i |_i nj nj
SGP11-6 Alphaguard SGP11-9 Alphaguard ASGP11-13 Alphaguard XSGP11-16.5 Alphaguard
Figure 5-49. Handheld Alphaguard data taken from soil gas ports
at location 11 versus time.
5-34
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
10000
1000
J~
u
0
2
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
CHCI3 Field GC Indoor Air Data - 422
7/26/2011
4/1/2012
Figure 5-51. Online GC chloroform indoor air data for 422 first floor.
CHCI3 Field GC Indoor Air Data - 422
I~liii MUM I
INI I I I III 11 I I
EH DD [Mm
EMU ED ID
-n-n—n-
7/26/2011
9/14/2011
11/3/2011 12/23/2011 2/11/2012
Date & Time
O Basement422
4/1/2012
Figure 5-52. Online GC chloroform indoor air data for 422 basement.
5-36
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
phase, but slightly lower than on the 422 side during the second phase and showed less scatter. Similar to
the 422 side, there was an increase in values starting in September and continuing into October (Figure 5-
53). Other than these step changes, short-term temporal variations were generally less than a factor of 2.
Measured values for the 420 basement ranged from ~0.3 ug/m3 to ~1.0 ug/m3 (Figure 5-54). A less
distinct step change is seen at this port in late September. Aside from that step change, short-term
temporal variations were generally less than a factor of two. Values were slightly lower than values
measured in the 422 basement especially during the second phase.
5.3.1.2 Tetrachloroethylene (PCE)
Measured values for the 422 first floor ranged from 0.2 ug/m3 to -2.2 ug/m3, although the vast majority
of values ranged from 0.5 ug/m3 to 1.0 ug/m3 (Figure 5-55). Generally, the levels were similar for both
sampling phases, although there were periods of higher values in the second (winter) phase. Temporal
variation during the first phase was generally less than a factor of two, but short-term temporal variations
in the second phase were up to a factor of four. Measured values for the 422 basement ranged from -0.3
ug/m3 to -3.2 ug/m3. Temporal variations were less than a factor of two during the first (summer-fall)
phase, but short-term temporal variations in the second (winter) phase were up to a factor of four, similar
to the variations seen on the first floor (Figure 5-56). This could be related to cooler temperatures or
greater temperature swings during the colder months.
Measured values for the 420 first floor (the non-heated part of the house) ranged from detection level
(-0.1 ug/m3) to -2.2 ug/m3 (Figure 5-57). Generally the values were higher with little temporal variation
in the summer-fall phase and lower with much greater short-term variation during the winter phase.
Temporal variation during the first phase was generally less than a factor of two, but short-term temporal
variations in the second phase were up to a factor of 10. Similarly, measured values for 420 basement
ranged from detection level (-0.1 ug/m3) to -2.2 ug/m3, with similar patterns to those seen on the first
floor: little temporal variation during the summer-fall (<2x) and higher short-term variations during the
winter phase of a factor of 10 (Figure 5-58). Although this is the same general pattern observed for the
422 heated side, the unheated nature of the 420 side of the building seems to have intensified the effect.
5.3.1.3 Comparison Between the Two Sampling Phases
Chloroform
The concentrations measured in both the basement and on the first floor of both units remained relatively
consistent over both sampling phases from August 2011 to February 2012. The computed percent
standard deviation (%RSD) for each floor was as follows:
422 first floor: 22%
422 basement 17%
420 first floor: 30%
420 basement 10%
The lack of a change in concentrations in the 422 side of the duplex is surprising because of the large
increase in subslab concentrations observed under this unit as described in Section 5.3.2.
Tetrachloroethylene
The concentrations measured on the basement and first floor of unit 422 remained relatively consistent
across both sampling phases, from August 2011 to February 2012.
422 first floor: 12%
422 basement 26%
5-37
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
3.0 -
2.5 -
_ 2.0 -
m
1
Concentration
M
|
5 1.0 -
0.5 -
o.o -
7/261
CHCI3 Field GC Indoor Air Data - 420
*
0 0
5E
'^yit ***&*
2011 9/14/2011 11/3/2011 12/23/2011 2/11/2012 4/1/2012
Date & Time
•> 1st 420
Figure 5-53. Online GC chloroform indoor air data for 420 first floor.
3.0 -
2.5 •
_2.0 •
m
E
S
|l.5-
s
8
|
5 1.0 -
0.5 •
0.0 •
7/261
CHCI3 Field GC Indoor Air Data - 420
D
D
g
. o|n^SoB DJo ^ °o °
w fi B 6 "^^4^
^g^^
2011 9/14/2011 11/3/2011 12/23/2011 2/11/2012 4/1/2012
Date & Time
n Basement 420
Figure 5-54. Online GC chloroform indoor air data for 420 basement.
5-38
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCE Field GC Indoor Air Data - 422
7/26/2011
9/14/2011
11/3/2011 12/23/2011
Date & Time
2/11/2012
4/1/2012
Figure 5-55. Online GC PCE indoor air data for 422 first floor.
PCE Field GC Indoor Air Data - 422
9/14/2011
11/3/2011 12/23/2011
Date & Time
n Basement 422
2/11/2012
4/1/2012
Figure 5-56. Online GC PCE indoor air data for 422 basement.
5-39
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCE Field GC Indoor Air Data - 420
-I-
7/26/2011 9/14/2011 11/3/2011 12/23/2011
Date & Time
2/11/2012
4/1/2012
Figure 5-57. Online GC PCE indoor air data for 420 first floor.
PCE Field GC Indoor Air Data - 420
1
I
s
7/26/2011
9/14/2011
11/3/2011 12/23/2011
Date & Time
° Basement 420
2/11/2012
4/1/2012
Figure 5-58. Online GC PCE indoor air data for 420 basement.
5-40
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
However, greater variation was observed on both floors of the unheated 420 unit:
420 first floor: 88%
420 basement 50%
The larger variations are due to the short-term temporal variations of up to a factor of 10, as shown in
Figures 5-57 and 5-58. As previously discussed this may be due to greater temperature swings in the
unheated part of the house.
5.3.2 Subsurface Soil Gas Data
Subsurface concentrations were monitored at eight locations with the automated GC:
• three subslab locations: SSP-2, SSP-4, and SSP-7
• four soil gas locations: SGP2-9 ft, SGP8-9 ft, SGP9-6 ft and SGP11-13 ft
• one location in the wall on the side of the basement (WP-3).
Approximately 600 measurements per location were collected in Phase 1 and approximately 900
measurements per location were collected in Phase 2 at each of these eight locations.
5.3.2.1 Chloroform
The chloroform data from the automated GC for all locations for both sampling phases are summarized in
Figure 5-59 and for the separate phases in Figures 5-60 and 5-61.
In the first phase of the program, chloroform values were relatively constant until approximately
September 13. At that time, the instrument inexplicably stopped and was not restarted until 2 days later on
September 15. Upon restart, there is an abrupt increase in all the chloroform values but not the
tetrachloroethylene values. This shift occurred because of a change in the chloroform baseline definition
by the integration software and was not due to changes in the actual chloroform concentrations.
The following behaviors were observed in the first phase (Figure 5-60):
• Temporal variation is generally less than a factor of 2 at all the sample locations during this phase
except for location WP-3.
• At probe WP-3, the concentrations show repeated high and low variations of a factor of 3 to 5
times occurring over time scales of several days. WP-3 was the only location to exhibit this
behavior.
In the second phase of the program (Figure 5-61), the following behaviors were observed:
• Probe WP-3 continued showing the same oscillations as in the first phase.
• Probes SGP9-6 ft and SSP-4 showed a continual rise in concentrations throughout the sampling
period, increasing by approximately 2 to 2.5 times above the starting concentration of the second
phase. This same increase at SGP9-6 ft was also observable in the extractive samples (method
TO-17 data set) as a trend running from late August to December. This pattern was not seen in
the first phase of the program. Despite the large concentration increase of chloroform during this
second phase, there was no concurrent increase in the indoor air concentrations of chloroform
measured by the online GC in either the basement or first floor of unit 422.
• Chloroform variations in all the other subsurface probes were less than 50%.
5-41
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
CHCI3 Field GC Subsurface Air Data
4/1/2012
1SGP9-6 »SGP8-9 BSGP2-9 DSGP11-13 • SSP-7 D SSP-4
Figure 5-59. Online GC subsurface chloroform soil gas data—Phase 1 and Phase 2.
CHCI3 Field GC Subsurface Air Data
8/5/2011 8/15/2011 8/25/2011 9/4/2011 9/14/2011 9/24/2011 10/4/2011 10/14/2011 10/24/2011
Date & Time
SSP-2 BWP-3 "SGP9-6 «SGP8-9 BSGP2-9 QSGP11-13 • SSP-7 Q SSP-4
Figure 5-60. Online GC subsurface chloroform soil gas data—Phase 1.
5-42
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
CHCI3 Field GC Soil Gas Data
11^23/2011 12/3/2011 12/13/2011 12/23/2011 1/2/2012
1/12/2012
Date & Time
1/22/2012 2/1/2012 2/11/2012 2/21/2012
Figure 5-61. Online GC subsurface chloroform soil gas data—Phase 2.
5.3.2.2 Tetrachloroethylene (PCE)
The tetrachloroethylene data from the automated GC for all locations for both sampling phases are
summarized in Figure 5-62 and for the separate phases in Figures 5-63 and 5-64.
In the first phase of the program (Figure 5-63), it appears as if there is a lot of fluctuation in the
subsurface values. However, inspection of the individual locations shows the following:
• Probes SGP2-9 ft, SGP8-9 ft, and SGP9-6 ft show only slight temporal variations of 20% to 50%
over the sampling period.
• There are two probes that field records suggest may have been inadvertently closed for a period
of time:
- SGP11-13 ft 8/29/11 @ 15:16 closed; 9/9/11 between 14:00 and 15:00 opened
- SSP-7 ft 8/29/11 @ 15:36 closed; 9/9/11 between 14:00 and 15:00 opened
If those periods of inadvertent closure are discounted, then variation during this phase was less than a
factor of 3 at these two ports.
• Probes SSP-4 and SSP-2 also show less than a factor of 2 temporal variation over most of the
sampling period. However, both of these probes contain a group of analyses when the values
dropped rapidly by large amounts and then increased rapidly back to the prior values (Figure
5-61). The cause for this behavior is not clear. The drop in SSP-2 data occurred at times
5-43
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCE Field GC Soil Gas Data
2/11/2012
4/1/2012
SGP2-9 DSGP11-13
1 SSP-7 D SSP-4
Figure 5-62. Online GC subsurface PCE soil gas data—Phase 1 and Phase 2.
PCE Field GC Soil Gas Data
8/5/2011 8/15/2011 8/25/2011 9/4/2011 9/14/2011 9/24/2011 10/4/2011 10/14/2011 10/24/2011
Date & Time
1SGP9-61 »SGP8-9 "SGP2-9 QSGP11-13 • SSP-7 Q SSP-4
Figure 5-63. Online GC subsurface PCE soil gas data—Phase 1.
5-44
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
PCE Field GC Subsurface Air Data
Figure 5-64. Online GC subsurface PCE soil gas data—Phase 2.
that may suggest an effect of the fan tests (discussed in Section 12.2). The SSP-4 drop offs
happen more frequently and do not appear to be caused by the fan tests. The TO-17 data for SSP-
4 PCE over the whole year also did show considerable variability (Figure 5-65). The pattern of
this subslab probe's plot is reminiscent of Johnson's observation of data from another house:
"There are long periods of relative VI activity with sporadic VI inactivity" (Johnson et al., 2012).
• Probe WP-3 concentrations show repeated high and low variations of a factor of 3 to 5 times
occurring over weekly time scales. These fluctuations are similar to the chloroform variations
seen in this same probe.
In the second phase of the program (Figure 5-64), the following behaviors were observed:
• Probes SGP2-9 ft, SGP8-9 ft, SGP9-6 ft, SGP11-13 ft, and SSP-7 show slight temporal variations
of 20% to 50% over the sampling period.
• Probe SSP-4 is constant within 25% for most of this phase of observation but shows two periods
of a rapid drop in values down to near-zero values and then a quick rebound to the predrop values
(Figure 5-66). This probe is located very close both spatially and within 18 inches vertically to
probe SGP9-6 ft. SGP9-6 ft had similar PCE concentrations and did not show the same rapid
variations. However, the drop in values is also seen in the method TO-17 samples of location
SSP-4 at other times. This suggests that the behavior at SSP-4 was due to air leakage in the thin
void zone that often exists under concrete slabs (DePersio and Fitzgerald, 1995) and thus had less
influence on the SGP9-6 ft probe, which had a wider screened interval.
5-45
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Concentration (ug/m3)
ooooooooo
VOC Data for SSP-4 (Method TO-17)
•
• A
A
• A
*** A •
:« A •
A A A
• A A
^ * APCESSP-4
^J* * ^ • ChlorotormSSP-4
• A A A A
A A A
•
\ \ \ \ \ \ \ \
Date
Figure 5-65. Method TO-17 data for SSP-4.
PCE-SSP-4(Portl3)
Figure 5-66. Online GC PCE measurements in SSP-4.
Probe SSP-2 was mostly constant but showed three periods of concentration variations of
approximately a factor of 2 lasting over several days each.
5-46
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
• Probes SGP11-13 ft and SSP-7 did not show the rapid drop in values seen during the first phase,
suggesting that the behavior in the first phase might indeed be due to valve closure, not actual
variations in the soil gas concentrations as discussed above.
• Probe WP-3 continued to show the same oscillations as in the first phase with slightly greater
variations of a factor of 5 to 8 times occurring overtime scales of several days. These fluctuations
are similar to the chloroform variations seen in this same probe.
• The PCE concentrations at locations SGP9-6 ft and SSP-4 decreased slightly over the sampling
period in contrast to the CHC13 concentrations, which showed large increases in these two probes
over the same time period (Figure 5-67 shows data from SGP9-6). This trend was also observed
in the TO-17 sampling of this port during the same time period. This is indicative of different
sources for the chloroform and tetrachloroethylene.
In summary, except for probe WP-3, the regular short time scale (< 14 day) temporal variations in PCE
seen in all the subsurface probes are typically less than a factor of 2 and generally less than 50%. Probe
WP-3 is located closest to the ground surface (~3 ft bgs) so the variations detected might be due to
surface influences. Neither rain events, snow events, nor any other changing meteorological conditions
seemed to have much effect on the SSPs or on SGPs 6 ft bgs or deeper. SSP-4 showed long periods at
relatively steady elevated concentrations punctuated by short intervals of dramatically lower
concentrations.
Soil gas concentration variations that were observed at WP-3, and to a lesser extent at SSP-2, occurred
over a period of days, indicating that there is little advantage to collecting 24-hour composite samples
versus instantaneous grab samples.
PCE&CHCI3-SGP9-6
Figure 5-67. Comparison of online GC measurements of PCE and chloroform in SGPS at 6 ft.
5-47
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
5.4 Radon Short Term Variability (Based on Daily and More Frequent
Measurements)
5.4.1 Indoor Air
Indoor air radon was measured during the intensive rounds using E-Perm electrets and Genitron
Alphaguard monitors. Electrets were located at all six indoor locations on the 422/420 sides of the house
and the ambient location. One stationary Alphaguard was located in the 422 north basement, and the other
was located in the 422 upstairs office. At the time of the first intensive round, the stationary Alphaguards
were not collecting data, but they were in regular use throughout the rest of the project.
Electret radon data for the first intensive round are shown in Figure 5-68. All measurements were higher
than ambient, and all of the 422 locations were higher than the 420 locations. This could be explained by
the 422 side being the heated side during a fairly cold winter (i.e., a greater stack effect for 422). Also, the
south side of the basement usually has higher concentrations on both the 422 and 420 sides of the duplex.
For the second intensive round (across 2 project weeks), the electret radon concentrations showed the
opposite pattern to the first intensive round (see Figures 5-68 and 5-69), with higher radon concentrations
on the 420 side. This could be due to the ACs running on the 422 but not on the 420 side.
The stationary Alphaguard data during the second intensive round are similar downstairs and upstairs (see
Figures 5-70 and 5-71). The downstairs Alphaguard (Figure 5-70) showed a regularly repeated pattern of
daily peaks and troughs, with the peaks occurring during the early morning of each day. The upstairs
Alphaguard (Figure 5-71) showed a similar, but more diffuse, pattern than the downstairs graph. The
early-morning peaks may have occurred when the sun shone on 422 basement north and heated that
portion of the basement. The more diffuse pattern for the upstairs radon may be from additional mixing of
ambient air in the upstairs portion of the house.
The third intensive round electret radon data (Figure 5-72) looked very similar to the pattern for the first
(Figure 5-68). Both occurred during the colder period of the year when the heater was in use on the 422
side of the house. Again, the south side of the house (422 and 420 sides) generally showed greater radon
concentrations than other locations.
The stationary Alphaguards for the third intensive show similar patterns for downstairs (Figure 5-73) and
upstairs (Figure 5-74) radon levels, although there were lower concentrations of radon upstairs.
5-48
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
First Intensive Round
Location
Figure 5-68. Electret indoor air radon concentrations for the first intensive round.
Second Intensive Round
• 1 nd ]« Rourd IstiVitt
• 2 m I nt Raund I id Wsst
Location
Figure 5-69. Electret indoor air radon concentrations for the second intensive round.
Note that the round lasted for 7 consecutive days across 2 weeks.
5-49
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Downstairs Alpha guard Second Intensive Round
Figure 5-70. Radon concentrations from the downstairs stationary Alphaguard
during the second intensive round.
2nd Intensive Upstairs Alphaguard
-2nd intoH we upstei rs
Figure 5-71. Radon concentrations from the upstairs stationary Alphaguard
during the second intensive round.
5-50
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Third Intensive Round
1D.OO -
9.00
• 3rd mi Round ] LI Wee*
• 3rd mi Round 2nd W..-L-k
100
0.00
Location
Figure 5-72. Electret indoor air radon concentrations for the third intensive round.
Note that the round lasted for 7 consecutive days across 2 weeks.
Third Intensive Downstairs Alphaguard
stairs Alp hagyard
Figure 5-73. Radon concentrations from the downstairs stationary Alphaguard
during the third intensive round.
5-51
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
3rd Intensive Upstairs Alphaguard
Figure 5-74. Radon concentrations from the upstairs stationary Alphaguard
during the third intensive round.
5.4.2 Subslab, Wall Port, and Deeper Soil Gas Radon Data
Subslab, wall, and SGPs were sampled with the Genitron Alphaguard for both the weekly sampling
routines and the intensive rounds. The only change for the intensive rounds was the frequency of the
sampling, sometimes daily or multiple times per day. These radon data are not included in this report but
showed similar trends as in the weekly data discussed in Section 5.2.
5.5 Outdoor Climate/Weather Data
External and internal weather parameters were measured at the 422/420 house on a Vantage Vue weather
monitor. Internal temperatures were recorded by HOBO data loggers. Barometric pressure readings were
taken about every 15 minutes by Setra pressure sensors. Data were downloaded from these sources
approximately once per week. Well water levels were measured approximately once per month.
Table 5-3 presents data from monthly weather summaries for 2011 and 2012 published by the Indiana
State Climate Office (Scheeringa, 2011-2012). The 2011-2012 project year can be summarized as an
eventful period for Indiana weather. At the beginning of the project, central Indiana received more snow
than usual and temperatures in the region started lower than usual. As the weather warmed, central
Indiana experienced almost 50 tornadoes with over 60 for the state in 2011 (the usual for the state is -22
per year [Scheeringa, 2011-2012]). Additionally, in the period of January through March 2012, Indiana
experienced 11 tornadoes, about half of the usual yearly allotment. April was the wettest April on Indiana
record, and the summer was hot and windy in central Indiana, with an 8-day heat wave in late July. The
winter of 2011-2012 had very little snow, and March was the warmest on Indiana record.
Figure 5-75 shows the temperature record from the external temperature monitor and HOBO devices
placed at seven indoor locations on the 422 and 420 sides of the house. Figure 5-76 shows indoor
5-52
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Table 5-3. Summary Meteorological Data for Central Indiana
Month/Year
January 20 11
February 20 11
March 2011
April 2011
May 2011
June 2011
July 2011
August 20 11
September
2011
October 2011
November
2011
December
2011
January 201 2
February 201 2
March 201 2
State
Average T
(deg F)
23.1,2.9V
30.7
41 .4, 0.7 A
53.3, 1.9 A
62.5, 0.5 A
72.6, 1.6 A
79.2, 4.6 A
72.4, 0.8 A
63.6, 2 v
54, 0.1 A
46.6, 4.2 A
36.7, 5.6 A
32.3, 6.3 A
35.1, 4.5 A
54.4, 13.7 A
Central IN
Average T
(deg F)
22.3, 3v
30.5, 0.8 A
41.3, 1.3 A
53.1, 2.3 A
62.4, 0.7 A
72, 1.3 A
79, 4.7 A
72.9, 0.7 A
63.4,1.9V
53.7, 0.3 A
46.6, 4.7 A
36.2, 5.5 A
31 .8, 6.4 A
34.9, 5.1 A
54.2, 14.1 A
State
Average
Precipitatio
n(")
1.51,0.92V
4.17,2.28A
3.5, 0.1 A
9.69
6.47, 2.06 A
5.34, 1.14 A
2.79,1.31V
2.59,1.2V
5.39, 2.3 A
3.32, 0.42 A
6.25, 2.66 A
4.55, 1.49 A
3.39, 0.96 A
1.56,0.73V
2.74, 0.66 v
Central
Average
Precipitatio
n(")
1.63,0.71V
5.1,2.98A
3.7, 0.4 A
9.1, -5.5 A
6.03, 1.63 A
5.4, 1.3 A
1.6,2.65V
2.81,0.95V
3.56, 2.58 A
3.38, 0.56 A
6.02, 2.38 A
5.08, 2.10 A
3.51, 1.17 A
1.34,0.94V
3.38, 0.1 A
Special
Notes on
Central IN
5 -12" snow
4- 14" snow
Wettest IN
April, 27
tornadoes
21 tornadoes
8-day heat
wave
Windy
Windy
light snow
light snow
1-2" snow
1-2" snow
Warmest IN
March
Week 1 T
Average
(deg F)
1v
2v
1A
4A
5v
8A
2A
6A
2v
1A
normal
normal
7A
13A
4A
Notes
Feb 1 and 2,
3 - 8" snow
2.2" rain
1 .7" rain
1 .8" rain
-0.5" rain
0.6" rain
0.5" rain
0.9" rain
0.7" rain
1 .5" rain
0.1 "rain
0.3" rain
1 .2" rain
Week 2 T
Average
(degF)
3v
4v
1A
4A
8A
2v
3A
1v
4v
5A
7A
4A
7A
normal
20 A
Notes
Feb 10, state
aveT17v
0.5" rain
1.2" rain
0.6" rain
1 .4" rain
0.1 "rain
1 .5" rain
0.3" rain
0.4" rain
0.5" rain
1 .3" rain
0.5" rain
0.2" rain
0.2" rain
Week 3 T
Average
(degF)
6v
10A
9A
3v
6v
normal
9A
normal
normal
4v
4A
11A
2A
2A
26 A
Notes
on 01/21/11,
19 degF
below
normal
0.7" rain
3.4" rain
0.9" rain
2.1 "rain
0.5" rain
0.1 "rain
1 .7" rain
2.5" rain
1 .3" rain
1.1 "rain
1.2" rain
0.3" rain
1 .3" rain
Week 4 T
Average
(deg F)
2v
2v
5v
2A
4A
1v
7A
1 v
3v
2v
5A
~9A
18A
3A
9A
Notes
0.25" rain
3.0" rain
2.6" rain
0.7" rain
0.6" rain
0.6" rain
2.4" rain
0.4" rain
3.8" rain
0.75" rain
1 .6" rain
0.5" rain
0.25" rain
Note that the symbols "A" and "v" mean "above" and "below" normal, respectively, and that the weekly values show how the weekly averages differ from normal (from Scheeringa and Hudson, 2011,
2012).
5-53
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
HOBO and Ambient Temperatures
Figure 5-75. Temperature records from the external temperature monitor and the HOBO devices at
seven indoor locations on the 422 and 420 sides of the house.
Dashed lines indicate the periods of AC use, and the colored solid lines indicate the fan test times.
5-54
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Indoor Office Temperature
110
Figure 5-76. Indoor temperature as recorded inside
the 422 second floor office.
temperature data recorded in the office set up for this work on the 422 side of the house. The same
general trend can be seen in both figures, cycling from the winter lows to the summer highs. The lowest
temperature occurred on the unheated 420 first floor at -26 degrees Fahrenheit on February 10, 2011, and
the highest temperature occurred at the 422 office at 102 degrees Fahrenheit on July 21,2011.
As stated in Section 3.2.1., the gas-fired furnace was run from November 19, 2010 until June 22, 2011,
and then from November 7, 2011, until June 1, 2012, on the 422 side only, with no heating unit on the
420 side. Initially, window-mounted ACs ran on both sides of the duplex from June 29, 2011 until July
12, 2011. When the ACs were replaced, they were replaced on the 422 side only and ran from March 3,
2011, until October 24, 2011. Figures 5-75 and 5-76 show some of the highest temperatures occurring
during the period between the AC theft and when they were replaced on the 422 side, along with higher
temperatures on the 420 side where the AC units were not replaced. The higher temperatures between AC
periods could be a result of the solar stack effect, which may have been driving the higher radon and VOC
concentrations observed during that time (see Section 5.2.1).
Temperature lows seen in Figures 5-75 and 5-76 track fairly well with what is represented in Table 5-3,
as the external temperature line (yellow line of Figure 5-75) and the internal HOBOs on the unheated 420
side of the house (light blue, dark blue, tan lines of Figure 5-75) show. Highs for the summer heat wave
also can be seen on both figures.
The most obvious features of the stacked hydrological graph of Figure 5-77 are the prominent highs in
rainfall and stream discharge, coupled with the high water levels measured during gauging. These highs
align well with the period of heavy snowfall and rain experienced in central Indiana (see Table 5-3). Dips
5-55
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
Comparison between Water Level, Discharge, and Rainfall
•£
o
5
0
6500
6000
5500
5000
4500
4000
3500
3000
2500
2000
1500
1000
1st Intensive Round
2nd Intensive Round n 3rd Intensive
^~~-~^_ Round
ii ---^^^
II
ii
II
ii
II
M
II
II
II
II
a
10
15
20
a
a
"
o
3-
\ \
\,
ff
Date
%
VA
^
Figure 5-77. Stacked hydrological graph with depth to water in feet (top—red circles), discharge at
Fall Creek in ft3/s (middle—blue line), and rainfall in inches (bottom—green line).
All are over time for the duration of the project. Intensive sampling rounds are marked by dashed and solid lines.
5-56
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
in stream discharge and the lower depths during well gauging match well with the much hotter drier
summer period.
Transition weather can be quite turbulent, and much of the 2011-2012 year seemed like transition
weather, with its tornadoes, record highs, and rainy periods (see Table 5-3). Figures 5-78 and 5-79 show
pressure readings taken outside the 422/420 house (Figure 5-78) and inside (Figure 5-79). For the time
period represented by this report (January 2011 through March 2012), the figures are fairly similar, with
prominent highs and lows during the cooler seasons and transitional weather times and more stable
periods during the warmer months.
PressureExternalto the House
20S
I:JHR
?*H $ i f.c LS- i 5 ?ft*
-==- -=,
s
u
I
Figure 5-78. Plot of barometric pressure (inches of Hg) external to the 422/420 house over time.
5-57
-------
Section 5—Results and Discussion: VOC Concentration
Temporal Trends and Relationship to HVAC
422 Pressure Data
ra
£
CL
K^ IkN
• Date vs 422 s\t> (Pa)
Date vs 422 bvu (Pa)
Date vs 422 dsgssg (Pa)
« Date vs 422 bve (Pa)
Date
Figure 5-79. Barometric pressure (Pa) on the 422 side of the house overtime.
5-58
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Table of Contents
6. Results and Discussion: Establishing the Relationship between VOCs and Radon in
Subslab/Subsurface Soil Gas and Indoor Air 6-1
6.1 Correlation between Soil Gas VOC and Radon Concentrations 6-3
6.2 Correlation between the Indoor Air Concentration and Radon 6-15
6.3 Radon and VOC Soil Gas Spatial Distributions 6-16
6.4 Spatial Correlations in Radon and VOCs Analyzed Separately 6-18
6.5 Correlations in Indoor Air VOC and Radon Temporal Trends 6-22
List of Figures
6-1. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and
radon(blue)forSGPll at 6 ft bis 6-4
6-2. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP1 at 9 ft bis 6-5
6-3. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP4 at 9 ft bis 6-6
6-4. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP4 at 13 ft bis 6-7
6-5. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP5 at 9 ft bis 6-8
6-6. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP7 at 9 ft bis 6-9
6-7. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP8 at 6 ft bis 6-10
6-8. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP9 at 6 ft bis 6-11
6-9. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP10 at 6 ft bis 6-12
6-10. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP 11 at 6 ft bis 6-13
6-11. Temporal trends and cross-correlograms of chloroform (red), PCE (green), and radon
(blue) for SGP12 at 6 ft bis 6-14
6-12. Cross-correlation plots for the time series of log-VOCs and radon by indoor air location 6-15
6-13. Comparison of nearly collocated subslab and shallow internal soil gas ports 6-17
6-14. Concentration distributions (ug/m3 [VOCs] or pCi/L [radon]) and significance tests for
nearly collocated subslab and shallow internal soil gas ports 6-18
6-15. Evaluation of spatial effect north and south basement by VOC and radon for 422 East
28th St.—cumulative distribution plots where the x axis represents concentration (ug/m3
or pCi/L) and the y-axis concentration 6-19
6-16. Evaluation of spatial effect north and south basement by VOC and radon for 420 East
28th Street—cumulative distribution plots where the x axis represents concentration
(ug/m3 or pCi/L) and the y-axis concentration 6-19
6-17. Cross-correlation between north and south basement indoor air by VOCs and radon 6-20
6-18. Comparison of temporal trends at north and south basement sampling locations 6-21
6-19. Autocorrelation function for chloroform, PCE, and radon by location (site and
north/south basement) 6-23
6-i
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
List of Tables
6-1. Counts of Records with Flag by VOCs and Flag Type for Soil Gas 6-1
6-2. Counts of Records with Flag by VOCs and Flag Type for Indoor Air 6-2
6-3. Comparison of Mean and Median Concentrations in North and South Sampling
Locations, 422 Side of Duplex 6-22
6-ii
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
6. Results and Discussion: Establishing the Relationship between
VOCs and Radon in Subslab/Subsurface Soil Gas and Indoor Air
In this chapter we explore whether radon, a parameter that can be inexpensively and rapidly analyzed in
the field, correlates with VOCs for which field analysis is more difficult and costly. In the statistical
analysis of environmental data, the number of nondetects affects the precision of estimates. As the level
of censoring due to nondetects or data quality issues increases, most of the correlation methods result in
highly biased correlation estimates (Newton and Rudela, 2007). To begin to assess the significance of
detection-limit limitations on the data from this study, we looked at flags related to data quality in the
study database. Tables 6-1 and 6-2 display the number and types of flags in the data for radon and for
each VOC measured in soil gas and indoor air, respectively. For soil gas, the proportion of nondetects
ranged from 23% (benzene) to 93% (hexane). A large proportion of nondetects also were reported for
trichloroethene1 (TCE, 89%) and toluene (78%). For indoor air, only cis-l,2-dichloroethene2 (cis-
1,2_DCE, 84%) shows nondetects. Because the proportion of nondetects in soil gas was reasonable (no
more than 50%, see Helsel [2005] for more detail), and because these compounds appear to be from
Table 6-1. Counts of Records with Flag by VOCs and Flag Type for Soil Gas
Flag Type
U
J
E
B
Q
I
S
C
ND
% Nondetect
Radon
0
0
0
0
0
0
0
0
0
0
PCE
576
126
0
14
0
477
52
174
46
24%
TCE
2,098
127
0
0
0
312
51
1
192
89%
Chloroform
853
226
0
28
0
499
52
117
50
35%
Benzene
594
284
0
965
0
492
49
9
11
23%
Hexane
2,147
131
0
0
0
323
50
17
243
93%
Toluene
1,908
235
0
122
0
337
49
1
104
78%
U Compound not detected and reported as MDL
J Compound concentration is estimated because detection was between the lowest calibration standard
concentration and the MDL
E Compound concentration is estimated because the concentration was above the highest calibration standard
concentration
B Compound concentration is flagged because the compound was detected in the associated method blank
Q Value failed project QC criteria
I Associated internal standard failed project QC criteria
S Associated surrogate standard failed project QC criteria
C Associated calibration verification standard failed project QC criteria
ND Nondetect
1 Includes flags U and nondetect
Also known as trichloroethylene
' Also known as cis-l,2-dichlorethylene
6-1
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Table 6-2. Counts of Records with Flag by VOCs and Flag Type for Indoor Air
Flag Type
U
J
E
B
Q
1
S
C
ND
% Nondetect
Radon
0
0
0
0
0
0
0
0
0
0%
Benzene
0
18
0
0
0
0
0
0
0
0%
Chloroform
0
164
0
0
0
0
0
0
12
2.9 %
cis-1,2-DCE
0
53
0
0
0
0
0
360
350
84%
Hexane
0
0
0
0
0
0
0
0
0
0%
PCE
0
91
0
0
0
0
0
0
0
0%
Toluene
0
0
0
0
0
0
0
0
0
0%
U Compound not detected and reported as below MDL
J Compound concentration is estimated because detection was between the lowest calibration standard
concentration and the MDL
E Compound concentration is estimated because the concentration was above the highest calibration standard
concentration
B Compound concentration is flagged because the compound was detected in the associated method blank
Q Value failed project QC criteria
I Associated internal standard failed project QC criteria
S Associated surrogate standard failed project QC criteria
C Associated calibration verification standard failed project QC criteria
ND Nondetect
1 Includes flags U and nondetect
subsurface sources based on other lines of evidence (see Section 11), we decided to focus the VOC and
radon analysis on tetrachloroethene3 (PCE, 24% nondetect) and chloroform (35% nondetects).
In environmental science it is not uncommon to observe shifts in time in the correlation between two time
series. For example, one series may have a delayed response to the other series or perhaps a delayed
response to a common stimulus affecting both series. The simple correlation coefficient between two
series properly aligned in time is inadequate to characterize the relationship in such situations. An
alternative is the cross-correlation function, which takes into account the possible lagged correlation
between the two time series. The cross-correlation function can assume values between -1 and 1, with a
high correlation indicating a periodicity in the signal of the corresponding time duration. Lag k cross-
correlation coefficient explores the correlation between week t from series 1 with week t+k in series 2.
The cross-correlation at lag 0 has a similar interpretation as the Pearson correlation coefficient.
To assess the correlation between the radon time series and each of the VOC time series, cross-correlation
coefficients were calculated at several lags, measured in weeks. A positive cross-correlation at lagl
coefficient suggests that observing increments of radon in 1 week is associated with an increasing trend in
VOCs 1 week later. Similarly, a negative cross-correlation of lag k suggests that an increasing trend in
random is correlated with a decreasing trend in the VOCs observed k weeks later.
1 Also known as tetrachloroethylene
6-2
-------
Section 6 — Results and Discussion: Establishing the Relationship between VOCs and Radon
r
Critical bounds were calculated at the 5% significant level, given by /v" to determine the significance
+ y
of the cross-correlation coefficients. If any of the cross-correlations exceeds the cutoff /^" then the
cross-correlation coefficient is deemed statistically significant from zero. The significance of the cross-
correlation coefficient is interpreted as a sign of positive or negative correction between radon and the
VOCs at k weeks apart. . Significant lags on both sides of 0 suggest that the relationship between the time
series goes two ways (one does not drive the other), which we would expect to see because changes in
radon do not directly cause VOC concentration rises, but there are a common collection of factors that
affect them both. Multiple significant lags means that the relationship is "smeared" in time — the
concentration of radon today is related to the concentration of PCE today, tomorrow, and the next day for
example. This also makes physical sense because the average residence time of VOCs in soil gas is higher
than that for radon at depth because of the relatively short half-life of radon. Significant correlations at
large lags may suggest some type of noise in the data that may result from autocorrelation of any of the
time series.
6.1 Correlation between Soil Gas VOC and Radon Concentrations
The correlation between the soil gas concentration for two VOCs (PCE and chloroform) and radon was
investigated using the cross-correlation function for the combinations of soil gas probe sites and depths
with a sample size large enough to allow analysis. Figure 6-1 shows temporal trends from an example
time series for the two VOCs and radon in soil gas at one soil gas probe site and depth (6 ft bis in
SGP1 1). The second panel in Figure 6-1 shows the correlograms for evaluating the correlation between
the two VOCs and radon at 6 ft bis in the same soil gas probe (SGP1). Blue lines denote the confidence
bands; spikes exceeding these confidence bands represent cross-correlations that are statistically
significant suggesting that the VOCs and radon are correlated at that lag time. For a 6 ft bis, only
chloroform is negatively correlated with radon in 2-week lag. Significant spikes in the correlograms in the
diagonal suggest autocorrelation for both VOCs and radon. Auto-correlation is to be expected in these
data sets, since soil gas concentrations change slowly at this site with respect to the weekly frequency of
the measurements performed.
6-3
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP1-6
O
Ac-2311
Week
Autocorrelation and Cross-Correlation
1.0-
o.o-
Chloroform, Chloroform
•
•^ Hi ' ^ ^-~
Chloroform, Tetrachloroethene | Chloroform, Radon
— ^ H^ _ ^^___H K __^
I . • .
12 0
Significant at the
0.95 level
i.o-
0.5-1
o.o-
-n^J
Radon, Chloroform
_^ H ^_^R^«_
-^ ^
Radon, Tetrachloroethene
—
Radon, Radon
••
-12 -9
-3 0 -12
-9 -6 -3
Lag (Weeks)
12
Figure 6-1. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon(blue) for SGP11 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-4
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP1-9
1000
Autocorrelation and Cross-Correlation
Chloroform. Tetrachloroethene
Chloroform. Radon
12 0
12 0
12
Tetrachloroethene. Chloroform T
etrachloroethene. Tetrachloroethei
Tetrachloroethene. Radon
Significant at the
0.95 level
0 0
12 0
12
'
False
True
1.0-
0.5-
o.o-
Radon,
Chloroform
-— — — — ™^ — — — — ™ — *-
Radon. Tetrachloroethene
-— ^^ M™ — • — ^ —
Radon, Radon
I
-• — — — — •••" — ^^—
-12
0 -12
0 0
12
Lag (Weeks)
Figure 6-2. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP1 at 9 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-5
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
13333
Jan 2011
Apr 2011
SGP4-9
Jul2011
Oct2011
Jan2D12
Compound
— Chloroform
TetrachlDroethene
- Radon
Week
Autocorrelation and Cross-Correlation
0.8-
0.4-
o.o-
0.8-
0.4-
o.o-
Chloroform. Chloroform
1
•• -• —
Chloroform Tetrachloroethene
__ ^m ^—^m
•
0 4 8 12 0 4
Tetrachloroethene, Chloroform T
•B — • ••B^
trachloroethene,
Chloroform, Radon
••
8 12 0 4 8 12
Tetrachloroethei
I
^•••••~— —
Tetrachloroethene, Radon
^m H^H
i i i i . i i . i i i i
Significant at the
0.95 level
• False
True
Radon, Chloroform
Radon Tetrachloroethene
:
Radon, Radon
1
•
-12 -8 -4 0-12-8 4 00 4
Lag (Weeks)
8 12
Figure 6-3. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP4 at 9 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-6
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
1000:
r
J
Apr 2011
SGP4-13
JUI2011 Oct2011
Week
Jan 2012
Autocorrelation and Cross-Correlation
Chloroform, Tetrachloroethene
ft
Chloroform, Radon
12 0
12 0
0
to.o
o
12
Tetrachloroethene.
^H ^Mm
Chloroform T
trachloroethene
1
•
Tetrachloroether
Tetrachloroethene, Radon
•— — •- *
Significant at the
0.95 level
-12 -9
-3
12 0 3
12
•
False
True
0.8
04
00
Radon, Chloroform
Radon, Tetrachloroethene
_
~— •— ™— "•"
Radon, Radon
1... .
-12 -9
-3 0 -12 -9 -6-3 00
Lag (Weeks)
Figure 6-4. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP4 at 13 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-7
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP5-9
1000
100:
Jan 2011
A(x2011
Jul2011
Jan 2012
Week
Autocorrelation and Cross-Correlation
Chloroform. Tetrachloroethene
Chloroform Radon
12
12 0
12
o
to
o
O
1.0-
0.5-
n n-
n c -
Tetrachloroethene
ChloroformT
— — — •— • — — •••— ^
-12 -8
trachloroethene.
Tetrachloroether
•
"•• — — — — — ~ — __
-4 0 0
4
8
Tetrachloroethene,
Radon
B— • ™ — ~ ^~~
12 0
4
8
12
Significant at the
0.95 level
'
False
True
1 0
0 5
0.0
-0.5
Radon. Chloroform
Radon. Tetrachloroethene
Radon. Radon
-12
0 -12
0 0
12
Lag (Weeks)
Figure 6-5. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP5 at 9 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-8
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP7-9
Apr 2011
Week
Autocorrelation and Cross-Correlation
1.0-
0.5-
o.o-
-0.5-
i;
~ 1.0-
o> 0.5-
2 o.o-
0-0.5-
O
1.0-
0.5-
-0 fi-
Chloroform, Chloroform
|._.
••••—•—• _
Chloroform. Tetrachloroethene
— _
0 4 8 12 0 4
Tetrachloroethene. ChloroformT
Chloroform, Radon
™~" — — —
8 12 0 4
trachloroethene, Tetrachloroethe
!••••••••—_
-12-8-4 00 4
Radon, Chloroform
8 12
Tetrachloroethene, Radon
— — ~™^ — ^ — "M^-™™""
3 12 0 4
Radon, Tetrachloroethene
,,__,_
Radon,
J
•
8 12
Radon
Significant at the
0.95 level
• False
True
-12
-4 0-12-8 -4 00 4
Lag (Weeks)
8 12
Figure 6-6. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP7 at 9 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-9
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
o
Jan 2011
Apr 2011
SGP8-6
JUI2011
Oct 201 1
Jan 2012
Week
Autocorrelation and Cross-Correlation
Chloroform. Chloroform Chloroform. Tetrachloroethene Chloroform. Radon
0
12 0
8 12 0
8 12
o
O
Tetrachloroethene. Chloroform Tbtrachloroethene, Tetrachloroethei
Tetrachloroethene, Radon
-12 -8
0 0
12 0
8 12
0.8
0.4
00
Radon, Chloroform
Radon. Tetrachloroethene
Radon, Radon
-12 -8
0-12-8 -4 00 4
Lag (Weeks)
12
Significant at the
0.95 level
False
True
I
Figure 6-7. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP8 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L
6-10
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Jen 2011
SGP9-6
Apf2011
JUI2011
octant
Week
Jan 2012
Autocorrelation and Cross-Correlation
Chloroform, Chloroform Chloroform. Tetrachloroethene Chloroform. Radon
Tetrachloroethene. ChloroformT
strachloroethene. Tetrachloroethe
1.0-
0.5-
o.o-
_n s-
Radon, Chloroform
•^
Radon, Tetrachloroethene
™— " ^M™ ™ — —
Radon, Radon
•
*•• ' _ ^r-
-12
4 0-12-8 -4 00
Lag (Weeks)
8 12
Significant at the
Figure 6-8. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP9 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-11
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP10-6
Jan 2011
Ac-2011
r 2012
Week
Autocorrelation and Cross-Correlation
1.0-
0.5-
Chloroform. Chloroform
|
•
Chloroform, Tetrachloroethene
Chloroform, Radon
™ •
12 0
12 0
12
I 1.0-
ro 0 5-
£00-
o 0.5
Tetrachloroethene. Chloroform T*
-
jtrachloroethene. Tetrachloroethe
-12 -84 00
?
1
Tetrachloroethene, Radon
•^ — . — . ^
20 8 12
Radon, Chloroform
Radon, Tetrachloroethene
Radon, Radon
-12
0-12
-8 -4
Lag (Weeks)
00
12
Significant at the
0.95 level
I
False
True
Figure 6-9. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP10 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-12
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
E
3
100:
SGP11-6
Jan 2011
Ap«2011
JUI2011
Cot 20 11
Jan 2012
Week
Autocorrelation and Cross-Correlation
Chloroform. Chloroform
Chloroform, Tetrachloroethene
Chloroform. Radon
8 12
Significant at the
Tetrachloroethene. Chloroform!
etrachloroethene. Tetrachloroethen
Tetrachloroethene. Radon
8 12 0
-12 -8 -4 0-12-8 -4 00
Figure 6-10. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP11 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L.
6-13
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
SGP12-6
7
Jan 2011
Ac- "J11
JUI2011
Oct2011
Jan 2012
Week
Autocorrelation and Cross-Correlation
LO-
OS-
u
I 1.0-
™ 0.5-
S o.o-
o ° J
LO-
CI 5-
oo-
Chloroform Chloroform
1,
_n
0
Chloroform,
Tetrachloroethene
••BB»»— •••—
4 8 12 0 4
Tetrachloroethene, ChloroformT
_• _•• • •••
-12
Chloroform
. Radon
^1 ^l^'
8 12 0 4
trachloroethene, Tetrachloroethe
• _
^•^•••— —
-8-4 00 4
Radon, Chloroform
-12
8
12
Tetrachloroethene, Radon
^•^H ^•^•^^^•^•^H ^H ^H
^P
8 12 0 4
Radon, Tetrachloroethene
-• — ^ — ^ ^^^^^
-8 4 0 -12 -8
Lag
8
12
Radon, Radon
I
M — H —
400 4
(Weeks)
8
12
Significant at the
0.95 level
I
False
True
Figure 6-11. Temporal trends and cross-correlograms of chloroform (red),
PCE (green), and radon (blue) for SGP12 at 6 ft bis.
Note that radon concentrations are plotted in pCi/L.
Cross-correlograms for chloroform and PCE with radon at SGP4 at 9 and 13 ft bis are shown in Figures
6-3 and 6-4. Only chloroform is positively correlated with radon at 9 ft at lags 4 and 5. At 13 ft, positive
correlation was observed at lag 9 between PCE and radon; this particular cross-correlation might be the
result of some noise or induced by the autocorrelation observed in both VOCs and radon. There is no
obvious physical mechanism that could lead to a lag this long.
Both VOCs were positively and negatively correlated with radon at lag 6 and lag 0 with chloroform and
PCE at SGP5 at 9 ft bis, respectively (Figure 6-5). Both VOCs and radon show autocorrelation at 0 and
other lags; presence of autocorrelations might affect the cross-correlation with radon.
Radon was correlated with PCE at lag 1 and chloroform at lag 2 at SGP7 at 9 ft bis (Figure 6-6).
However, the presence of autocorrelation observed in both VOCs and radon suggests that more in-depth
analysis is needed to try to eliminate the autocorrelation in order to better assess the cross-correlation.
6-14
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
No correlation with radon was observed at SGP8 at 6 ft bis (Figure 6-7). Autocorrelation was observed in
both VOCs and radon. Only PCE was negatively correlated with radon at SGP9 at 6 ft bis at lags 2 and 7
(Figure 6-8).
For SGP10 at 6 ft bis, both VOCs have negative correlation with radon at lags 1 (PCE) and 8
(chloroform) (Figure 6-9). For SGP11 at 6 ft bis (Figure 6-10), both VOCs showed negative correlation
with radon at lags 2, 8 (chloroform), and 10 (PCE). For SGP-12, both VOCs showed negative correlation
with radon at 6 ft bis at lags larger than 6 (Figure 6-11).
To summarize, from a practitioners perspective, the lack of a consistent, positive correlation between
radon and VOCs in soil gas at a consistent lag time suggests that monitoring radon in soil gas would not
be a practical tool for predicting variations of VOCs in soil gas.
6.2 Correlation between the Indoor Air Concentration and Radon
Figure 6-12 shows cross-correlation plots for the time series of log-VOCs and radon by indoor air
location. Cross-correlations at different lags exceed the confidence bands, suggesting that VOCs are
positively correlated with radon for all of the locations and that chloroform appears to have a stronger
positive correlation (larger cross-correlation coefficients) with radon than PCE in the study house. Further
analyses are needed to empirically model the relationship between the VOCs and radon, but these results
do indicate that there is a statistically significant relationship between radon and the VOC concentrations.
Developing such a model will allow us to describe what proportion of the temporal variability in VOCs
can be predicted using radon, which is desirable because the cost of measuring radon is lower than the
cost for measuring VOCs. From a practitioners perspective these results suggest that monitoring radon in
indoor air could provide a helpful indication of the direction in which VOC concentrations in indoor air
are moving.
Chloroform and Radon - Air samples, Location =422BaseS
Tetrachioroethene and Radon -Air samples, Location =422BaseS
Lag
Chloroform and Radon - Air samples, Location =422First
Lag
Tetrachioroethene and Radon -Air samples, Location =422First
Lag
Lag
Figure 6-12. Cross-correlation plots for the time series of log-VOCs
and radon by indoor air location.
6-15
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
6.3 Radon and VOC Soil Gas Spatial Distributions
The study design for this project in many cases nearly collocated conventionally installed subslab ports
(in which the drill bit passes through the slab and 3 inches or less into the soil) with the upper intervals of
multidepth interior soil gas monitoring points, which were installed as 6-inch stainless steel mesh screens
starting immediately below the floor. Conceptually, we expect that a shrinkage crack or other gap directly
beneath the slab will have more influence on the subslab ports than the wider screened soil gas ports.
In some instances, there appeared to be significant differences in the concentrations observed at these two
very similar depth intervals and the shape of the temporal trends, as shown in Figure 6-13. These graphs
present the concentrations over the course of the year of data with locally weighted scatterplot smoothing
(LOWESS) applied (Cleveland, 1981; Cleveland and Devlin, 1988) and 95% confidence intervals shown
for both locations. LOWESS makes the lines look neater while overlooking the issue of the different
sample size available at different times. Interpretation of these plots should be made with caution given
the different precision achieved at each lag. The confidence bands around the smoothed line are a
function of the amount of data and the variability in the data. Smaller sample sizes and larger variability
result in wider confidence intervals. Except for radon, overlap of the confidence bands suggests no
difference between SGP and SSP locations. Formal testing should be used to statistically confirm whether
the distribution of the analytes in the subslab ports (SSPs) is the same as the corresponding distribution of
the shallow internal soil gas ports (SGPs).
The Kolmogorov-Smirnov (K-S) goodness of fit test was used to compare the significance of the
distributions of the nearly colocated SSP and shallow internal SGP data. K-S is a nonparametric test for
assessing whether two groups of data come from the same distribution. Figure 6-14 shows the
distributions and the results of the K-S tests. Results for TCE are not very robust given the high number
of nondetects for that analyte. For all comparisons between SSP and SGP, the K-S tests resulted in
significant p-values for radon suggesting that the SSP and SGP data come from different distributions.
The distributions of the VOCs and radon data collected from SSP4 and SGP9 showed a difference that
will be investigated in future work.
6-16
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
How Similar are SSPs and SGPs?
100;
Soil Gas Port
— Sub-Slab Port
Date
Figure 6-13. Comparison of nearly collocated subslab and shallow internal soil gas ports.
6-17
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Distributions and K-S Test P-Values
0.9-
0.6-
-J North
-''South
0.3-
0.0-
10
0.1 1 10
Concentration (p.g /m )
o 1
10
Figure 6-14. Concentration distributions (ug/m [VOCs] or pCi/L [radon]) and significance tests for
nearly collocated subslab and shallow internal soil gas ports.
6.4 Spatial Correlations in Radon and VOCs Analyzed Separately
Although we expected the basements on each side of the duplex to comprise a single HVAC zone after
visual inspection, they may have different entry points for soil gas and fresh air. They also may not be
well mixed between compartments. Therefore, we sampled at two locations within each basement and
used the K-S test to compare the distribution of measured concentrations between sites located at the
north and south ends of the basements. If the data provide evidence against the null hypothesis that the
two distributions were not different, we could then merge the north and south sampling point
concentrations for future analysis. Significant differences in the two distributions were found between the
north and south for PCE and radon on the 422 side of the house (Figures 6-15 and 6-16) for all
compounds, suggesting that in this side of the house other factors may be affecting the observed outcomes
of the anlytes.
The cross-correlations between the north and south sampling locations within each basement were also
examined (Figure 6-17). For chloroform and PCE, a positive correlation exists between the two sides of
the house. In contrast, the distribution of the north and south sides at different sides of the house have
positive and negative correlations, suggesting that other external factors result in high concentrations in
6-18
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
one side when the other side is low, and vice versa. Further analysis should be performed to determine if
these effects that influence the distribution of the radon can be determined and quantified.
Ham
Side
Figure 6-15. Evaluation of spatial effect north and south basement by VOC and radon for 422 East
28th St.—cumulative distribution plots where the x axis represents concentration (ug/m3 or pCi/L)
and the y-axis concentration.
Nortl
Side
KSIest 006
Figure 6-16. Evaluation of spatial effect north and south basement by VOC and radon for 420 East
28th Street—cumulative distribution plots where the x axis represents concentration (ug/m3 or
pCi/L) and the y-axis concentration.
6-19
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
North vs South for Chloroformat site 420 Side
North vs South for Chloroformat site 422 Side
F 3-
0
13
f>
E ™ _
o o
0
I/I
& CN
0 CD ~
-1
Mil
I
5 -10
I
I I I
i
C
Li
LL
III
i
5
i I I I I I I • I 1
II
l i
E 10 1
E
o fr
"S d
p
O ~,
o ^H -
p
O d ~
i -1
i i
| 1
I
E -10
I
, I
I
I I I
_
i
(
L,
S
I.
1
II i .
1 1 1
I
E 10 1'
North vs South for Tetrachloroetheneat site 420 Sid North vs South for Tetrachloroetheneat site 422 Sic
1 «* -
i "
8CN
0
lf>
1 „
o d ~
-1
, M I
I
5 -10
I , I
r
III
=
c
Li
I I
i if
[[[LL,
I I
I E 10 1
s
1
•w ^
Q
O ^ _
o d ~
;
I
1
5 -10
l
c
(
L,
,,..
I
t E 10 1
g
North vs South for Radon at site 420 Side
North vs South for Radonat site 422 Side
o
'•5
5
,ll
i
-10
l ,
_
:
I
I ' I ' ' " I
i i i
0 E 10 1!
Lac
g
'•5
ro
d
p-i
d
I
E
I '
i
-10
I
I
C
— 1_
II
f
c
I
I
» 10 1!
Lag
Figure 6-17. Cross-correlation between north and south basement indoor air by VOCs and radon.
In Figure 6-18, the temporal trends on the north and south sides of the basement are visualized. These
graphs present the concentrations over the course of the year of data with a smoother curve depicting the
temporal trend and 95% confidence bands for both locations and radon and VOCs. As mentioned earlier,
the smoother makes the lines look neater, and it takes into account the issue of different sample size at
different times and locations. The confidence bands around the smoothed line are affected by the sample
size and the variability in the data.
6-20
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
How similar are the North and South sides of each basement?
North
South
0.1 =
Jan2011 Apr2011 Jul2011 Oct2011 Jan 2012 Jan 2011 Apr2011 Jul2011 Oct2011 Jan 2012
Date
Figure 6-18. Comparison of temporal trends at north and south basement sampling locations.
Although it appears that the south side of the 422 basement produces consistently higher concentrations
for the vapor intrusion-related constituents (PCE, chloroform, and radon), the overlapping of the
confidence bands suggests that the temporal trends have similar distributions at each location (Figure
6-18). Formal testing (K-S test) is needed to determine quantitatively whether the data from the north and
south have the same distribution. Thus, the indoor air concentration distribution somewhat reflects the
subslab distribution in that the VOC concentrations at the northern ports SGP10-6 and SSP-2 are typically
an order of magnitude or more lower than the central and southern ports beneath the 422 duplex (see
Section 5.1 for VOC data). Interestingly, the difference between northern, central, and southern soil gas
ports is much less marked for radon (see Section 5.2). However, when expressed as year-long mean or
median concentrations (Table 6-3), the differences between the northern and southern locations, although
statistically significant, are unlikely to be large enough to lead to differing management decisions under
these site conditions.
6-21
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Table 6-3. Comparison of Mean and Median Concentrations in North
and South Sampling Locations, 422 Side of Duplex
Compound
Radon
Radon
Benzene
Benzene
Chloroform
Chloroform
Hexane
Hexane
Tetrachloroethene
Tetrachloroethene
Toluene
Toluene
Trichloroethene
Trichloroethene
Location
North
South
North
South
North
South
North
South
North
South
North
South
North
South
Mean (pCi/L
or ug/m3)
5.60
6.47
0.78
0.78
0.29
0.34
0.64
0.67
0.94
1.30
1.71
1.70
0.11
0.14
Median (pCi/L
or ug/m3)
5.19
6.22
0.70
0.72
0.20
0.19
0.52
0.55
0.38
0.43
1.40
1.40
0.05
0.05
6.5 Correlations in Indoor Air VOC and Radon Temporal Trends
The autocorrelation function (ACF) is a collection of correlation coefficients between the series and lags
of itself over time. If the ACF plot is contained within the blue dashed lines, there is no temporal
correlation and the observations can be considered independent. Figure 6-19 displays the ACF for each
VOC and radon. All ACF plots show spikes exceeding the confidence bands suggesting a temporal
correlation and that the measurements are not independent. The nonindependence of the data suggests that
standard tests cannot be applied to the data without previously removing the temporal trend. Applying
standard tests to correlated data may result in under-estimated standard errors and larger p-values, which
will increase the likelihood of not rejecting a hypothesis of difference.
6-22
-------
Section 6—Results and Discussion: Establishing the Relationship between VOCs and Radon
Temporal correlation for Temporal correlation for
Chloroform at site 420 Base I Chloroform at site 420Base!
Temporal correlation for Temporal correlation for
Chloroform at site 422Basel Chloroform at site 422Base
O
"I D
o
8 °
£
0 "
0
IU
I 1
III 1
1 1 1 1 1 • 1 1 1 II
1 1 1 1
0 5 10 15
Lag
O
"S °
O
S3 =
£
(J (VJ
f
I,
MIL
ntti,,
i i i
5 10 15
Lag
|
~ CO
-^ o
o
tf) O
£
0 ™.
o
7
0
i
5
"I
ll..
Ml||
1 !
10 15
Lag
|
is *•&
1 °
O
w o
0 OJ
O ~ -
T
0
-\ i
ll,,
'' 1}
i i i
5 10 15
Lag
Tetra ch I oroeth e ne
at site 420BaseH
Tetrachloroethene
at site420BaseS
Tetrachloroethene
at site 422BaseH
Tetrachloroethene
at site 422BaseS
1
| s:
o
o fN
D
J,
Itn,,
i i i i
0 5 10 IE-
Lag
Temporal correlation for
Radonat site 420BaseN
-
s -
-
CN
• _
1
{
1
UN
Illl
i
5
i i
10 15
Lag
E "~
1 ^ -
t °
o
yj
o «
0
I
)
III,,..
, , 1
1 1 1
5 10 15
Lag
E ^
E
o
(0 O
(J (N
O
hi,,,
I I I
) 5 10 15
Lag
Temporal correlation for
Radon at site 420BaseS
Temporal correlation for
Radon at site 422BaseN
Temporal correlation for
Radon at site 422Base S
.1
.s
I
Figure 6-19. Autocorrelation function for chloroform, PCE, and radon by location
(site and north/south basement).
6-23
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
Table of Contents
7. Results and Discussion: Attenuation of Soil Gas VOCs and Radon 7-1
7.1 Subslab to Indoor Air Attenuation Factor Temporal Range 7-1
7.2 Subslab Attenuation Factors for Each Side of the Duplex 7-3
7.3 Attenuation Factors Calculated for Each Side of the Duplex Using Subslab and
Shallow Soil Gas Samples 7-6
List of Figures
7-1. Subslab (or 9-ft soil gas) to indoor air AFs for individual sample locations, with the
number of calculated AFs in each case indicated by the number directly below each
whisker 7-2
7-2. Subslab to indoor air attenuation factors, calculated for each side of the duplex using
only subslab points 7-4
7-3. Range of weekly chloroform and PCE concentrations in indoor air, subslab, and 9-ft
interior soil gas samples over study period, with the number of calculated AFs in
each case indicated below the whiskers 7-5
7-4. Attenuation factors vs. time: calculated for each side of the duplex using subslab soil
gas ports and the shallowest of the nested interior soil gas ports 7-7
7-i
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
7. Results and Discussion: Attenuation of Soil Gas VOCs and
Radon
In this section we explore the relationship between radon and VOC levels in indoor air and concentrations
measured in subslab and deeper soil gas, which is usually portrayed through the vapor intrusion
attenuation factor (AF). As described in Section 2.1.2, the vapor intrusion AF is the indoor air
concentration divided by a subsurface soil gas concentration at the same time and location. For example,
if subslab soil gas concentrations are 100 times the indoor air concentration measured at the same time,
the subslab AF would be 0.01. This section focuses on a general comparison of subslab and deeper soil
gas AFs for the entire project. Additional analysis of seasonal, weekly, and daily trends and relationships
will be accomplished in the next phase of this project.
7.1 Subslab to Indoor Air Attenuation Factor Temporal Range
After confirming that the number of nondetectable results in the indoor air and soil gas datasets to be used
was very small and, thus, not an issue, individual weekly AFs were calculated for PCE, radon, and
chloroform1 at each sampling location by sorting the data by location and week and then averaging all
observations from the same week and location. For soil gas concentration (the AF denominator), the
sample points for this analysis included all subslab soil gas points (SSP1 through SSP7) as well as the 9-ft
deep soil gas probes installed within the building (SGP8 through SGP12). For indoor air, weekly
measurements for the north and south basement of each side of the duplex were averaged to generate a
single weekly indoor air concentration (the AF numerator) for each side of the duplex that was used to
calculate AFs for all subslab or soil gas sample points on that side.
Results are presented as box and whisker plots in Figure 7-1. In these plots:
• Concentration (ug/m3 for VOCs and pCi/L for radon) is plotted on a logarithmic axis.
• The median of the data is represented by a dark black horizontal line across the box.
• The 25% to 75% range of the distribution is represented by the box.
• The whiskers go to the last point before the outlier cutoff, which is +/- 1.5 times the interquartile
range (75th percentile-25th percentile) (R Development Core Team, 2012; Wickham, 2009).
• Individual outlying data points above or below the whiskers are plotted as dots.
• 420 AFs are plotted in red and 422 AFs are plotted in green.
• The number of AFs are included below each box and whiskers plot.
Because each box and whiskers plot provides the distributional statistics for a single subslab or deeper
soil gas sampling point, the span of the plot represents the temporal variability of the AF at that point.
Notable observations that can be made from Figure 7-1 include:
• The inter-quartile range is generally quite narrow compared with other AF distributions that have
been published. This can be attributed to the fact that each box and whiskers plot is fixed in space
and, therefore, represents temporal variability in attenuation only. The high number of
measurements made here enable us to conclude that for most of the year, the AF at a particular
sample point is fairly stable. However, the full distribution is as wide as two orders of magnitude
for PCE and chloroform, suggesting that any single sampling event could yield an AF far enough
from the mean to markedly affect site management decisions.
1 The number of nondetects was significant for other analytes (e.g., TCE, benzene, hexane, 1,1-DCE) measured in
soil gas and indoor air. See Section 5.
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
IE
0.1
0.01 =
0.001-
O 0.1 E
"CD
ZJ
0)
0.01 E
0.001-
0.1 =
0.01:
o.ooiH
Radon
26 8 8 17 8 176 20 6 91 74 75 91
Chloroform
T
H
45 79 58 60 59 164 5 10 62 70 71 47
Tetrachloroethene
45 79 58 60 59 164 5 10 62 70 71 47
Side
420 Side
422 Side
Location
Figure 7-1. Subslab (or 9-ft soil gas) to indoor air AFs for individual sample locations, with the
number of calculated AFs in each case indicated by the number directly below each whisker.
7-2
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
• There is considerably less variability in the measured radon AFs than for VOC AFs. The reason
for this is not yet known, but the VOC and radon figures presented in Sections 5.1 and 5.2 (e.g.,
compare Figures 5-5 [chloroform] and 5-7 [PCE] with Figure 5-37 [radon] for subslab) suggest
that there is much more stability in radon concentrations in the subslab and shallow soil gas than
for VOCs. One reasonable explanation of this result is that the radon concentrations are expected
to be controlled by the emanation rate characteristic of the soil immediately surrounding a probe
and the rate of barometric (or stack effect induced) pumping (see Section 2). VOC concentrations
are subject to all of those causes of variability as well as processes occurring in the deep
subsurface along the longer VOC migration path from the source (e.g., water table fluctuations,
temperature dependent processes). The short half-life of radon (3.8 days vs. a year or more for
PCE) prevents subsurface migration being a significant influence on radon vapor intrusion as it is
with VOCs. Because the radon source is primarily from the soils immediately surrounding a
building, radon AFs may provide useful information about building envelope-specific processes.
• There appears to be reasonable agreement between the chloroform and PCE AFs on the 422 side
of the duplex. The measures of central tendency of the chloroform AF distributions for the 420
side of the duplex tend to be somewhat higher than those for the 422 side.
• The variability in AFs for different sampling points is greater on the 422 side, with the 420 side
having fairly consistent values from point to point for both subslab and the 9-ft deep soil gas
probes. It is not known whether the greater variability on the 422 side is due to the influence of
the heating system installed on that side or to differences in subsurface characteristics.
Geophysical tests just completed at the site for the follow-on project may provide insights on the
latter hypothesis.
• Although the subslab samples would be expected to have higher AFs (i.e., lower attenuation) than
the deeper (9-ft) soil gas, differences in AFs for the subslab vs. 9-ft soil gas are not markedly
apparent from visual observations.
These conclusions are based on observation of Figure 7-1, other lines of evidence from this study, and
general knowledge of vapor intrusion processes. Additional statistical analyses will be required to
determine which observations are statistically significant and which are not.
7.2 Subslab Attenuation Factors for Each Side of the Duplex
In Figure 7-2, we plot the average attenuation within each basement (420 and 422) over the course of the
study, calculated as follows:
420AvgAttenuation = Avg(420BaseS, 420BaseN)/Avg(SSP-3, SSP-5, SSP-6, SSP-7)
422AvgAttenuation = Avg(422BaseS, 422BaseN) /Avg(SSP-l, SSP-2, SSP-4).
In Figure 7-2, AFs appear to be much more variable on the unheated 420 side of the duplex. The heated
422 side of the duplex shows only two major "bumps" in the trend, one of which appears to correspond to
the fan tests conducted in September and October 2011 (discussed in Section 12.2), and the other effects
only chloroform at the very end of the study period.
7-3
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
D.1-
0.01-
Zl
So. 001
01
CD
I
0.01-
422 Side
Compound
Radon
~ Chloroform
Tetrachloroethene
— Trichloroethene
0.001-
Week
Figure 7-2. Subslab to indoor air attenuation factors, calculated for
each side of the duplex using only subslab points.
It is also notable that for chloroform and PCE, the AF temporal series plots bear little resemblance to the
indoor concentration plots presented in Section 5.1 (see Figures 5-1 and 5-2). This dissimilarity suggests
that the variance of concentration in indoor air is driven more by variance in subslab or deeper soil gas
concentrations (Figure 7-3) and less by changes in the building-specific AF across the slab. Figure 7-2
also does not show an obvious correspondence between AFs for radon and the VOCs, suggesting that
7-4
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
different processes are controlling these contaminants On the 422 side the radon AF generally agrees well
with the PCE and chloroform AFs. That is not the case on the 420 side, where the apparent attenuation of
VOCs is generally less attenuation than that for radon.
1000E
10 =
-ig
Chloroform
E3 V
h
45 79 58 60 59 164 5 10 62 70 71 47 137 112
$
45 79 58 60 59 164 5 10 62 70 71 47 137 112
Side
^420 Side
tit 422 Side
Location
Figure 7-3. Range of weekly chloroform and PCE concentrations in indoor air, subslab, and 9-ft
interior soil gas samples over study period, with the number of calculated AFs in each case
indicated below the whiskers.
7-5
-------
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
7.3 Attenuation Factors Calculated for Each Side of the Duplex Using Subslab
and Shallow Soil Gas Samples
The AFs for this analysis were calculated for each side of the duplex using all available shallow soil gas,
subslab soil gas, and indoor air measurements. The two different AFs were calculated as follows:
AF420SGBF = Avg(420BaseN, 420BaseS) / Avg(SSP~3, SSP-5, SSP-6, SSP-77, SGPU-6, SGPU-6)
AF422SGBF = Avg(422BaseN,422BaseS)/Avg(SSP-\, SSP-2, SSP-4, SGP&-6, SGP9-6, SGPW-6)
The 422 side VOC AFs calculated in this way (Figure 7-4) show significant variance through the year,
generally decreasing gradually from the start of the testing during the severe winter of January 2011
through July 2011. They then rise only modestly into the milder winter of 2012.
Several points in early September and early October stand out from the gradual temporal trends as
anomalously high AFs. These show the influence of fan testing, in which a box fan was placed at the head
of the stairs of the 422 side of the duplex withdrawing air from the basement. This would be the expected
result of depressurization of the basement space. The concentration and differential pressure results of the
fan tests are discussed in Section 12.2.
On the 422 side of the duplex, there appears to be fairly close agreement between the AFs for PCE and
radon except for a brief period in the winter of 2011, and the chloroform AFs seem to be somewhat lower
than the PCE or radon AFs. On the 420 side of the duplex, the chloroform AFs are generally higher (less
attenuation) than those for PCE, which is on the high side of the radon AF distribution. On both sides of
the duplex, the radon AFs show less temporal variability than the VOC AFs (Figure 7-4).
7-6
-------
10
0.1
0.01
0.001
0.0001
O
1-^
0
Section 7—Results and Discussion: Attenuation of Soil Gas VOCs and Radon
Fan Test
^yr-
^
•i
Chloroform
TCE
O Radon
o o
CT)
CT)
en
*^.
NJ
O
422 (Heated) Subslab to Basement Attenuation Factors
Figure 7-4. Attenuation factors vs. time: calculated for each side of the duplex using subslab soil
gas ports and the shallowest of the nested interior soil gas ports.
7-7
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
Table of Contents
Results and Discussion: Can Near-Building External Samples Be Used as a Surrogate
Sampling Location? 8-1
8.1 Comparison of External Soil Gas to Subslab Soil Gas 8-1
8.1.1 Chloroform 8-1
8.1.2 Tetrachloroethylene (PCE) 8-2
8.1.3 TCE 8-3
8.1.4 Radon 8-4
8.2 Comparison of Wall Ports to Subslab External Soil Gas 8-5
List of Figures
8-1. Box and whisker plots of chloroform distribution in soil gas at varying depths
(concentration is log scale) 8-2
8-2. Box and whisker plots of PCE distribution at various depths 8-3
8-3. Box and whisker plots of TCE distribution at various depths 8-4
8-4. Box and whisker plots of radon concentration at various depths 8-5
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
8. Results and Discussion: Can Near-Building External Samples Be
Used as a Surrogate Sampling Location?
The conventional assumption in the vapor intrusion field and in regulatory guidance documents is that
subslab concentrations represent the best descriptor of a subsurface vapor intrusion source of the
commonly collected lines of evidence. Although that assumption has been recently challenged by those
who emphasize the potential for buildings to also contribute soil gas into the subslab space because of
building overpressurization (e.g., McHugh et al., 2006), it will be taken as a given for the purposes of this
chapter.
Because subslab samples are generally considered by vapor intrusion investigators to be more intrusive on
the lives of residents than exterior samples, some practitioners and responsible parties strongly desire to
make maximum use of exterior soil gas as a line of evidence before resorting to subslab sampling inside
the house. Exterior soil gas points are often installed within 10 horizontal ft of the foundation edge, as
was done in this study. Some regulatory agencies (e.g., some California agencies) have suggested
multidepth exterior soil gas as a useful line of evidence and prefer deep to shallow soil gas as a
conservative/definitive estimate of concentrations under the building slab. The construction of multidepth
soil gas ports extending vertically below the interior of the building is, however, rare outside of research
studies. Thus, we compare the trends of concentration versus depth for the multidepth soil gas points
installed in the exterior and interior of the house and compare those with the subslab concentrations.
In order to examine these trends, the soil gas data for all time points sampled were used to prepare a series
of box and whisker plots for subslab and interior, and exterior soil gas samples at different depths. In
these plots:
• Concentration (ug/m3) is plotted on a logarithmic axis.
• The median of the data is represented by a dark black line.
• The 25% to 75% range of the distribution is represented by the box.
• The whiskers go to the last point before the outlier cutoff, which is +/- 1.5 times the interquartile
range (75th percentile-25th percentile) (R Development Core Team, 2012; Wickham, 2009).
• Individual outlying data points are plotted as dots.
In these plots, the exterior soil gas points at each depth are grouped as a population (orange boxes), and
the interior soil gas points at each depth are grouped as another population (blue boxes), dataset The range
of results for these many samples (over 60 points sampled weekly for over 50 weeks) illustrates the
potential variability that could result with the much fewer data points that are taken during a more typical
vapor intrusion investigation. Because it is so extensive, this dataset can be used to model decisions that
would result if few samples are taken, inside or outside the study building or at different times during the
year.
In reviewing these box plots, the reader should keep in mind that the subslab samples in this study duplex
are beneath the basement and approximately 5 ft bis., The percentages of nondetects for the data that
went into these plots are tabulated by compound and depth in Tables 5-1 and 5-2 in Section 5.1.3.
8.1 Comparison of External Soil Gas to Subslab Soil Gas
8.1.1 Chloroform
Chloroform mean concentrations increase with depth (Figure 8-1). The highest median concentrations are
associated with the samples collected just above the water table: 13 or 16.5 ft (note that the 16.5-ft depth
could not be sampled for soil gas at many times because the water table rose above that depth). This
3-1
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
suggests a groundwater source or transport pathway for chloroform in the vicinity of the house. The
median concentration decreases nearly two orders of magnitude with depth, which suggests considerable
natural attenuation. Although the mechanisms for this attenuation are not known and can include
dilution/barometric pumping, chloroform is also subject to biodegradation both by anaerobic organisms
and cometabolically under aerobic conditions (AFCEE, 2004). This attenuation is so dramatic that by the
3.5-ft depth most of the observations are nondetects. Thus, the 3.5-ft exterior soil gas would not have
predicted the subslab concentration well. The interquartile range for chloroform in subslab
(approximately 5 ft) is quite similar to the interquartile range at 6 ft for exterior and interior soil gas. The
subslab concentration median is, however, considerably higher. The median chloroform concentration in
subslab soil gas lies between the median concentrations in exterior soil gas at 9 to 13 ft.
Chloroform
bub-blab
T C
J.D
:t:
"~" — ^
£ o
"o. 9
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
distribution at the external 9 ft, which has also been observed in visualizations of datasets from specific
time periods, is somewhat anomalous and has not been fully explained. The distributions in exterior soil
gas consistently show much more interquartile variability than the distributions in interior multidepth soil
gas. The interquartile variability increases with decreasing depth for the exterior clusters.
Tetrachloroethene
Sub-Slab-
3.5-
6-
CL
O)
Q
9-
13-
16.5-
Wall Port-
f
I
"
• **** •••!•• *+ **
I
F-l Exterior
F-l Interior
1
10
100
1000
Concentration(M.Q /m3)
Figure 8-2. Box and whisker plots of PCE distribution at various depths.
8.1.3 Trichloroethylene (TCE)
At many sites, PCE and TCE have similar distributions because TCE is the first biodegradation product of
PCE, and releases of mixed PCE/TCE sources are also common. At this site, however, TCE is much less
frequently detected than PCE (see Table 5-1 in Section 5.1.3). The most frequent depth of detection was
subslab and 6 ft. The distributions in subslab and at 6-ft exterior soil gas are similar (Figure 8-3).
Comparison of indoor and outdoor air concentrations over time suggests that ambient air is a primary
source of TCE when concentrations are low but that subsurface sources are more important when VOC
concentrations are higher (see Section 7).
8-3
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
Trichloroethene
Q_
4
N
44
•
HI
• 4
** *
***
*• **
** *
*
«•
* *
MH
*
•••
«
4
* *
F-;-| Exterior
Interior
10
100
Concentration^ /m3)
Figure 8-3. Box and whisker plots of TCE distribution at various depths.
8.1.4 Radon
The distribution of radon with depth (Figure 8-4) shows the highest concentrations at 6 ft and in subslab.
Concentrations are relatively uniform with depth except at 3.5 ft. The interquartile distributions are
notably narrower than those observed for the VOCs. This is consistent with a conceptual understanding in
which radon is generated from a wide variety of geological materials surrounding the sampling points but
has a relatively short half-life (days versus a year or more for PCE). The notably lower concentrations and
wider interquartile range at 3.5 ft and for the wall port samples are consistent with a greater degree of
barometric pumping expected in shallow soils.
8-4
-------
Section 8—Results and Discussion: Can Near-Building External Samples
Be Used as a Surrogate Sampling Location?
Radon
Sub-Slab-
3.5-
Q.
0)
Q
13-
16.5-
WallPort-
Exterior
Interior
10
100
1000
10000
pCi/L
Figure 8-4. Box and whisker plots of radon concentration at various depths.
8.2 Comparison of Wall Ports to Subslab External Soil Gas
Concentrations in wall ports (see plots above and in Section 5.1.2) are generally quite low. This is
consistent with their shallower depths and the notably lower concentrations in the 3.5-ft exterior soil gas
samples.
It does not though necessarily follow that the flux through the basement walls is an insignificant
contribution to vapor intrusion. It should be noted that the wall surface area is much higher than the
basement floor surface area and that the relative volumetric gas flows through the walls and floor are not
known. Tracer studies of this duplex in a follow-on project should provide greater insight on this
question.
8-5
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table of Contents
9. Results and Discussion: Over what durations do solvent extracted passive samplers provide
useful integration of indoor air concentrations? (Is uptake rate constant?) 9-1
9.1 Comparison of Daily to Weekly Samples 9-5
9.2 Comparison of Weekly to Biweekly Samples 9-8
9.3 Comparison of Weekly to Monthly Samples 9-9
9.4 Comparison of Weekly to Quarterly Samples 9-10
9.5 Comparison of Weekly to Semiannual and Annual Samples 9-10
9.5.1 Comparing Radiello Samplers to SKC Samplers 9-11
9.6 Conclusions 9-12
List of Figures
9-1. Kernal densities of%Bias for important VOCs 9-4
9-2. The effect of vapor pressure on sorbent performance 9-5
List of Tables
9-1. Summary Statistics for %Bias by Comparison Period and VOC 9-3
9-2. Summary Statistics Individual Concentration Measurements by VOC and Period 9-6
9-3. Summary Statistics for %Bias Comparing Daily vs. Weekly Period and VOC 9-7
9-4. Evaluation of Daily vs. Weekly Differences 9-8
9-5. Evaluation of Weekly vs. Biweekly Differences (Outlier Included) 9-9
9-6. Evaluation of Weekly vs. Biweekly Differences (Outlier Removed) 9-9
9-7. Evaluation of Weekly vs. Monthly Differences 9-10
9-8. Evaluation of Weekly vs. Quarterly Differences 9-10
9-9. Evaluation of Weekly vs. Semiannual Differences 9-11
9-10. Evaluation of Weekly vs. Annual Differences 9-11
9-11. Average %Bias for Average Weekly Radiello Measurements Compared with Semiannual
and Annual Modified SKC 575 Charcoal Badges 9-12
9-i
-------
Section 9 — Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
9. Results and Discussion: Over what durations do solvent
extracted passive samplers provide useful integration of indoor
air concentrations? (Is uptake rate constant?)
The reliability of passive samplers in measuring VOC concentrations largely depends on whether the
uptake rate is constant given the environmental conditions and the sample duration. Prolonged exposure
of passive samplers can result in reduced net uptake rates due to back-diffusion or loss of sorptive
capacity. Loss of adsorbed chemicals occurs when the concentration at the adsorbing surface is
sufficiently high that the uptake rate decreases. To evaluate the performance of charcoal solvent-extracted
passive samplers over periods ranging from 1 day to 1 year, the VOC concentrations measured for
extended time intervals were compared with the average concentrations measured concurrently over
shorter time segments. With the exception of a daily sample deployment over 7 days in the spring and
again in the winter, the shortest interval used in our study was a 1-week duration. Weekly samples were
collected concurrently with biweekly (2-week), monthly (4-week), quarterly (13-week), semiannual (26-
week), and annual (52-week) samples.
For each sampling interval, Radiello charcoal passive samplers were deployed. The high sampling rates of
the radial style sampler provided good sensitivity for indoor air measurements for the weekly samples.
Additionally, the charcoal sorbent cartridge was selected over the thermally desorbable cartridge because
of its stronger retention characteristics for the target VOCs and its higher VOC loading capacity, both
beneficial attributes for long-term sample exposure. The uptake rates used to generate sample
concentrations were published by the Radiello manufacturer, Fondazione Salvatore Maugeri, Padova,
Italy, based on measurements in a standard atmosphere chamber (Sigma- Aldrich, 2012). The rates were
corrected for the average temperature recorded over the sampling duration using the equation:
K
where K is the measured temperature in Kelvin, QK is the uptake rate at temperature K, and Q298 is the
published reference rate at 298K.
In addition to the Radiello sampler, the SKC 575 badge packed with charcoal and equipped with a
secondary diffusive barrier was deployed for the two longest sampling periods, the semiannual and annual
intervals. The badge paired with the secondary barrier has an uptake rate approximately 100 times lower
than the Radiello sampler. With the exception of hexane, the modified badge uptake rates were provided
by the manufacturer SKC Inc., Eighty Four, PA (Coyne, 2010). The uptake rate for hexane was estimated
by dividing the standard published SKC 575 badge uptake rates by a factor of 28.5 based on the
corresponding reduction in the diffusive surface area when using the secondary diffusive cover. This
alternative sampler was deployed over 6 months and a year to determine if lower uptakes rates were more
stable over the prolonged exposures and less subject to back-diffusion effects and possible interference
from water adsorption. Discussion of the SKC badge performance as compared with the Radiello
performance is presented in Section 9.6.
Evaluation of the passive sampler over the exposure period was determined by comparing the numerical
average of the shorter time segments (e.g., 2 weeks, 4 weeks) to the concurrent integrated measurement
(e.g., biweekly, monthly). For each interval evaluated, the relative percent difference (%Bias) was
calculated using the equation:
%Bias = Cf '~ Cl x 100%
(Lp, + LI/ z.)
9-1
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
CA = Average concentration of the shorter duration sample
Ci = Measured concentration of integrated sample over same period
A positive %Bias value indicates the average concentration of the shorter duration measurements was
higher as compared with the longer integrated sample concentration. Similarly, a negative %Bias
indicates that the shorter measurement technique will underestimate the actual vapor concentration.
Several explanations are possible for a positive bias. A positive %Bias is expected when the actual uptake
rate is lower than the published rate such as in the case of back-diffusion. Also, a positive %Bias can be
observed when the shorter duration has a high bias due to artifacts from the sorbent material or the
extraction process. The acceptance criterion to demonstrate equivalency is +30%, which was established
as equivalent to the data quality objective set for replicate samples for this work, based on what is defined
as acceptable reproducibility in vapor intrusion field studies. If the reported concentration was a
nondetect, the %Bias calculation was performed using half of the reporting limit for the corresponding
concentration. Table 9-1 displays summary statistics of the %Bias for each target VOC by type of
comparison period (weekly vs. biweekly, weekly vs. monthly, weekly vs. semiannual, or weekly vs.
annual). With the exemption of PCE (biweekly and monthly) and toluene (biweekly and semiannual), the
average of the %Bias favored the shorter measurements for all comparison periods (i.e., the longer period
sample consistently underestimated the actual vapor concentration) and all VOCs. Chloroform and
hexane showed the two larger standard deviations with respect to the average of %Bias across all
comparison periods, while toluene and tetrachloroethene had smaller standard deviations. A combination
of a smaller average concentration and larger standard deviation results in a high rate of %Bias not
meeting the acceptance criterion as shown in Figures 9-1.
Figure 9-1 displays the %Bias density plots for each VOC and each interval comparison. The dotted line
represents a %Bias of 0, and the solid lines bracket the acceptance criterion of +30%. Density plots are
approximations to the probability distribution of the data and are affected by the sample size, so caution
must be used when interpreting density plots based on small samples (n<30). As shown in the legend of
Figure 9-1, the number of available comparisons was less than 30 for the 3-month and longer durations.
Figure 9-1 and Table 9-1 show that as the period of measurements increase (e.g., quarterly, semiannual,
annual) the distribution of %Bias moves away from zero, suggesting that weekly measurements are
increasingly greater than the concentrations determined by integrated measurement.
Table 9-1 and Figure 9-1 show the proportion of %Bias satisfying the acceptance criteria. The %Bias for
benzene, chloroform, and hexane shifts in the positive direction as the measurement time period increases
from monthly to annually, under the concentration ranges seen in our test house. Based on these plots,
this particular passive sampler performs as follows for the chemicals tested:
• Chloroform: performs well up to a 14-day integration period but degrades by 28 days
• Benzene and TCE: performs well up to a 28-day integration period but degrades by 91 days
• Hexane: performs well up to a 91-day integration period but degrades by 182 days
• PCE and toluene: performs well up to 364 days
Extending the sampling duration to quarterly shows that only benzene, hexane, PCE, and toluene have
maintained a relatively stable uptake rate as defined by the %Bias criterion of+30%. The average %Bias
for TCE and the density plot shift to the right, indicating that the uptake rate is showing a drop over the
quarterly interval.
Figure 9-2 shows that as the period of measurements increases from biweekly to annually, the number of
%Bias satisfying the equivalency rate decreases for all VOCs except toluene.
9-2
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-1. Summary Statistics for %B/as by Comparison Period and VOC
Weekly vs. Biweekly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Weekly vs. Monthly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Weekly vs. Quarterly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Weekly vs. Semiannual
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Weekly vs. Annual
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
N
175
175
175
175
175
175
84
84
84
84
84
84
28
28
28
28
28
28
14
14
14
14
14
14
7
7
7
7
7
7
Minimum
^6.20
-72.00
-67.40
-57.90
-66.70
-51.20
-20.80
-28.60
-44.90
-26.10
-25.20
-19.90
8.10
18.20
-13.40
-16.00
-16.10
6.80
25.40
89.30
7.00
-15.00
-19.40
19.50
51.90
162.60
41.60
14.60
-10.30
58.50
25%
-3.35
0.00
-7.90
-7.50
-5.60
1.60
0.53
12.60
-1.15
-7.18
-6.55
7.55
21.48
57.40
12.37
-0.95
-6.13
17.38
33.15
125.30
19.05
-8.70
-14.65
35.60
55.20
165.60
53.10
17.20
-3.25
65.80
Median
9.30
11.30
3.50
0.00
0.50
9.80
13.90
28.35
12.45
-1.65
-0.15
20.35
31.00
85.80
22.00
5.80
-0.05
37.15
37.20
155.90
33.45
5.05
-6.40
41.65
56.70
171.20
77.10
20.30
0.90
69.60
Mean
7.74
11.15
2.75
-0.44
-0.02
10.98
13.39
31.39
11.33
-0.47
0.03
19.09
33.05
88.29
23.05
5.88
2.37
34.17
39.48
147.00
31.61
2.93
-4.84
41.57
60.81
172.30
67.19
22.37
0.80
69.30
75%
20.05
23.80
14.85
6.50
6.70
19.60
23.83
47.58
26.63
7.40
5.58
27.35
43.05
114.10
31.22
10.80
10.90
46.65
46.40
177.20
43.88
10.65
1.33
46.68
67.90
178.00
80.95
27.25
4.65
72.60
Maximum
50.50
59.20
59.60
29.80
33.60
67.40
59.30
94.30
50.90
23.10
35.20
81.30
61.70
172.30
68.20
25.00
25.00
59.10
62.80
187.80
58.20
27.10
18.80
64.80
70.90
185.40
83.50
32.80
12.20
80.20
SD
16.98
18.55
18.85
11.78
12.13
17.23
16.80
26.89
20.00
10.89
10.68
17.65
14.83
43.58
19.29
10.82
12.03
15.94
10.89
36.58
16.71
12.44
11.41
11.91
7.91
8.45
18.10
7.54
111
7.25
9-3
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Bias of Passive, Solvent Extracted Samplers Over Time
Sample Duration
Two Weeks
N=175
One Month
Three Months
/ Six Months
N = 14
~7 One Vear
0 50 -50 -25 0 25 -60 -40 -20 0 20
% Bias (Positive bias indicates the longer sample under-reported concentration)
Figure 9-1. Kernal densities of %B/as for important VOCs.
9-4
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
100 -
90 -
80 -
g 70 -
CO
V
In"
5 60^
#_
| 50
-J5
to
2 40
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-2. Summary Statistics Individual Concentration Measurements by VOC and Period
Period
N
Min.
1stQu.
Median
Mean
3rd Qu.
Max.
SD
Daily
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
86
86
86
86
86
86
0.81
0.23
0.52
0.09
0.44
0.04
1.00
0.26
0.75
0.21
0.86
0.06
1.30
0.50
1.05
0.45
1.10
0.06
1.74
0.82
1.31
0.50
1.89
0.09
2.00
0.80
1.58
0.66
2.05
0.10
7.70
6.60
4.20
1.50
8.00
0.41
1.12
1.10
0.82
0.35
1.94
0.06
Weekly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
371
371
371
371
371
371
0.36
0.06
0.23
0.08
0.50
0.01
0.57
0.12
0.42
0.22
0.96
0.04
0.75
0.20
0.56
0.37
1.40
0.05
0.80
0.33
0.68
1.06
1.76
0.12
0.94
0.38
0.80
0.67
2.35
0.08
2.30
4.00
2.60
22.00
6.00
2.70
0.30
0.42
0.38
2.36
1.08
0.26
Biweekly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
191
191
191
191
191
191
0.38
0.07
0.25
0.10
0.52
0.02
0.58
0.11
0.47
0.25
1.20
0.04
0.67
0.20
0.59
0.43
1.70
0.05
0.72
0.35
0.66
1.22
1.79
0.13
0.85
0.42
0.74
0.79
2.15
0.09
1.60
3.70
2.00
12.00
5.30
1.40
0.23
0.42
0.31
2.27
0.94
0.22
Monthly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
99
99
99
99
99
99
0.43
0.04
0.24
0.14
0.68
0.02
0.59
0.09
0.48
0.21
1.10
0.03
0.68
0.14
0.58
0.32
1.60
0.04
0.69
0.27
0.62
1.08
1.75
0.11
0.78
0.27
0.74
0.68
2.20
0.09
1.40
2.30
1.80
13.00
4.20
1.20
0.16
0.37
0.23
2.22
0.71
0.21
Quarter
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
28
28
28
28
28
28
0.40
0.02
0.34
0.15
0.88
0.02
0.49
0.05
0.43
0.27
1.48
0.03
0.60
0.09
0.53
0.33
1.80
0.05
0.58
0.17
0.55
1.01
1.74
0.10
0.65
0.15
0.67
1.05
2.00
0.11
0.71
1.00
0.80
5.90
3.00
0.48
0.09
0.23
0.14
1.35
0.50
0.11
Semi-annual
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
14
14
14
14
14
14
0.41
0.01
0.34
0.22
1.20
0.04
0.48
0.02
0.42
0.42
1.53
0.05
0.55
0.04
0.47
0.84
1.95
0.07
0.53
0.05
0.51
1.05
1.93
0.08
0.58
0.08
0.58
1.40
2.35
0.11
0.65
0.14
0.75
3.00
2.70
0.18
0.07
0.04
0.13
0.79
0.46
0.04
Annual
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
7
7
7
7
7
7
0.38
0.01
0.29
0.56
1.60
0.05
0.40
0.02
0.29
0.68
1.75
0.06
0.43
0.03
0.33
0.83
1.80
0.06
0.44
0.03
0.37
0.90
1.89
0.06
0.46
0.04
0.39
0.92
2.00
0.07
0.54
0.06
0.60
1.70
2.30
0.09
0.06
0.02
0.11
0.38
0.26
0.01
9-6
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Because of the short exposure time of the daily samples and the low concentrations at the site, the mass
collected for each target compound was typically below the reporting limit, or the mass measured was
biased high because of background levels in the sampling/analysis method. Even with the lowered
reporting limit and reporting results down to the MDL, many site samples showed nondetects for the
chlorinated solvents (cis-l,2-DCE, TCE, and chloroform) making comparisons with the weekly integrated
measurement less useful. Of 86 daily samples collected, the percentage nondetects were benzene 92%,
chloroform 78%, hexane 59%, PCE 8%, toluene 44%, and PCE 77%.
Because PCE was typically at or slightly above the reporting limit in the daily samples (typically 0.12
lig/m3), the average of the concentrations correlated better with the weekly measurements. PCE
concentrations from the daily sample collection ranged from 0.09 to 1.50 vig/m3 (Table 9-2). Chloroform
performed well when comparing the average 24-hour exposure times with the weekly measurements
when concentrations were above 0.5 jig/m
The daily samples did not provide accurate concentrations for benzene or hexane because the mass
measured on each sampler was similar to the mass detected in the associated laboratory blanks . The
average benzene and hexane mass measured on the daily samples was 0.19 and 0.12 \ig, respectively. In
terms of concentration, this mass translates to approximately 1.74 vig/m3 benzene and 1.31 vig/m3 hexane.
Benzene was detected in each laboratory blank with an average blank concentration of 0.086 \ig
(approximately 0.75 vig/m3 for a 24-hour period). Hexane was detected in one-half of the lab blanks with
an average concentration of 0.053 \ig (approximately 0.56 jig/m3). As a result, both benzene and hexane
showed a positive bias when comparing the daily average with the weekly measurement because of the
high bias from background artifacts. Despite background detections of toluene in all of the associated
laboratory blanks averaging 0.018 \ig, the mass measured in each daily sample was typically 10 times
higher, resulting in minimal bias for the average daily calculated concentration (Table 9-3).
Table 9-3. Summary Statistics for %B/as Comparing Daily vs. Weekly Period and VOC
Comparison Period
Daily vs. weekly
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Minimum
18.5
-7.6
-47.1
-19.1
-39.2
-30.4
25%
34.93
16.8
-45.22
-3.325
-0.225
-20.52
Median
48.05
43.15
-33.3
39.55
4.40
-2.55
Mean
48.99
55.10
-27.2
39.93
3.736
-3.893
75%
55.7
96.35
-29.08
84.85
8.825
13.92
Maximum
102
121.5
60.4
110.2
44.10
21.2
SD
20.74
43.94
28.94
45.60
17.21
18.39
Table 9-4 displays the difference between daily and weekly measurements. Strong statistical significance
(p-value<0.001) was detected between the durations for benzene, chloroform, and hexane.
In summary these results suggest that the solvent-extracted Radiello sampler is better used for durations
longer than 1 day, if concentrations are at or below the indoor air concentrations we measured. Other
passive samplers with higher uptake rates are available that are more suitable for short durations at such
low concentrations.
9-7
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-4. Evaluation of Daily vs. Weekly Differences
voc
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (% bias)
37.0
29.7
13.6
-6.2
-14.5
-8.9
Standard Error of the
Estimated Differences
(% Bias)
6.1
12.9
13.4
5.1
5.4
14.4
P-Value
0.000***
0.000***
0.006**
0.431
0.443
0.164
** Denotes statistical significance at 0.01
***Denotes statistical significance at 0.001
9.2 Comparison of Weekly to Biweekly Samples
For this comparison, weekly sample averages were subtracted from the corresponding biweekly samples.
The normality assumption was then evaluated using the Shapiro Wilks test implemented in the software
package Nortest in the R statistical language and environment. The Shapiro Wilks test has more power to
detect normality compared with other normality tests such as the Kolmogorov Smirnov test (Razali and
Wah, 2011). For large sample sizes (n>30), the central limit theorem was used to justify normality of the
average measures. To decide whether the data can be considered a random sample or independent and
identically distributed, a random test or simple sign test was used. If an observed value in the sequence is
influenced by its position in the sequence or by the observations that precede it, the sequence of data
points is not truly a random sample. Paired t-tests were used to compare the weekly to the biweekly
sample averages when both the normality and independence assumptions held. When the independence
assumption failed, a t-test accounting for the correlation in the data was used.
Comparisons were performed with (Table 9-5) and without (Table 9-6) the presence of an outlier sample
that was identified as such but could not be explained.1 In both tables, strong statistical significance (p-
value<0.001) was detected for benzene, chloroform, and TCE. Borderline significance (p-value = 0.055)
and significance (p-value<0.05) were detected for hexane with and without the outlier, respectively. The
differences between the two sequential 7-day samples and the corresponding 14-day samples were thus
small but consistent. Because the study included 175 such comparisons for each compound, the small
differences achieved statistical significance.
The Radiello sampler performed well over the biweekly period with average %Bias for each target VOC
well within the 30% acceptance criterion. For the more volatile compounds tested, the measured
concentration was consistently slightly lower in the 14-day samples. Practitioners may choose, however,
to accept that slight bias in order to gain more cost-effective long-duration observations to account for
temporal variability.
1 The sample identified as an outlier had concentrations of benzene = 1.7, chloroform = 64, hexane = 0.84, PCE =
49, toluene = 3.5, and TCE = 2.4 (ig/m3. We removed this from the analysis in Table 9-4 because, although we
could find no issues with the sample or its analysis, the measured concentrations did not appear reasonable based
on the other measurements.
9-8
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-5. Evaluation of Weekly vs. Biweekly Differences (Outlier Included)
voc
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias)
7.7
11.1
2.8
-0.4
0.0
11.0
Standard Error of the
Estimated Differences
(%Bias)
1.84
1.94
1.42
0.89
0.92
1.89
P-Value
0.000***
0.000***
0.055
0.619
0.980
0.000***
*Denotes statistical significance at 0.001
Table 9-6. Evaluation of Weekly vs. Biweekly Differences (Outlier Removed)
VOC
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias s)
8.0
11.6
3.1
-0.1
0.4
11.3
Standard Error of the
Estimated Differences
(%Bias)
1.92
2.01
1.38
0.84
0.84
1.98
P-Value
0.000***
0.000***
0.025*
0.864
0.668
0.000
* Denotes statistical significance at 0.05
***Denotes statistical significance at 0.001
9.3 Comparison of Weekly to Monthly Samples
A paired t-test was used to evaluate the significance of the difference between weekly and monthly
samples (Table 9-7). The p-value statistic shows that the %Bias of weekly measurements vs. monthly is
significantly different from zero for benzene, chloroform, hexane, and TCE, suggesting that weekly
concentrations are larger than monthly concentrations and that monthly measurements would tend to
underestimate VOC concentrations in indoor air. Although the amount of underestimation, from 11% to
31%, is generally within the target accuracy range, it is consistently biased low. Practitioners may choose
to accept or correct for that slight bias in order to gain more cost-effective long-duration observations to
account for temporal variability.
9-9
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-7. Evaluation of Weekly vs. Monthly Differences
voc
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias)
13.4
31.4
11.3
-0.5
0.0
19.1
Standard Error of the
Estimated Differences
(%Bias)
2.98
6.30
2.18
1.19
1.17
3.35
P-Value
0.001***
0.000**
0.000***
0.696
0.977
0.000***
***Denotes statistical significance at 0.001
9.4 Comparison of Weekly to Quarterly Samples
The difference between weekly and quarterly measurements (Table 9-8) is statistically significantly
different from zero for benzene, chloroform, hexane, tetrachloroethene, and TCE, using a t-test. The
direction is consistent—weekly concentrations are larger than quarterly concentrations. Although the
average bias of PCE and hexane are below the 30% criterion, with 28 comparison datasets for each
compound we were able to detect these modest variations with statistical confidence. However, for all of
the tested compounds except chloroform a practitioner might choose to accept or correct for the negative
bias in the concentration estimate for the quarterly sample in order to benefit from the dramatic cost
savings from using one passive sampler rather than 13 successive 7-day passive samples.
Table 9-8. Evaluation of Weekly vs. Quarterly Differences
VOC
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias)
33.1
88.3
23.1
5.9
2.4
34.2
Standard Error of the
Estimated Differences
(%Bias)
5.52
23.00
3.65
2.04
2.27
6.73
P-Value
0.009**
0.031*
0.000***
0.008**
0.307
0.015*
*Denotes statistical significance at 0.05
**Denotes statistical significance at 0.01
***Denotes statistical significance at 0.001
9.5 Comparison of Weekly to Semiannual and Annual Samples
The difference between weekly and semiannual measurements (Table 9-9) is statistically significantly
different (t-test) from zero for hexane, suggesting that weekly concentrations are larger than semiannual
concentrations. The estimated differences (Table 9-9) and mean bias (Table 9-1) for benzene, TCE,
chloroform were substantially greater than the 30% criteria. The estimated difference and mean bias for
hexane only slightly exceeded the criteria. This suggests larger weekly measurements compared with
9-10
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
corresponding semiannual values and a greater underestimation of indoor air concentrations with longer
term samples.
By inspection, it seems unintuitive that the large estimated difference for chloroform as shown in this
table and in Table 9-1 is associated with such a high p-value. However, we have rechecked this result,
and it is mathematically correct because of the structure of the dataset and the small number of
comparisons (N=14).
Table 9-9. Evaluation of Weekly vs. Semiannual Differences
voc
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias)
39.5
147.0
31.6
2.9
-4.8
41.6
Standard Error of the
Estimated Differences
(%Bias)
5.48
31.46
4.46
3.32
3.05
8.39
P-Value
0.088
0.134
0.000***
0.394
0.137
0.127
***Denotes statistical significance at 0.001
The difference between weekly measurements and annual (Table 9-10) is statistically significantly
different from zero for hexane and tetrachloroethene, suggesting that weekly concentrations are
consistently larger than annual concentrations. Caution must be used when interpreting significance given
the small sample size, which was not sufficient to compare weekly and annual values for benzene,
chloroform, and TCE. The estimated differences (Table 9-10) and mean %Bias (Table 9-1) for both
toluene and PCE were both lower than the 30% criteria even over the full-year duration.
Table 9-10. Evaluation of Weekly vs. Annual Differences
VOC
Benzene
Chloroform
Hexane
Tetrachloroethene
Toluene
Trichloroethene
Estimated Differences
Between
Durations (%Bias)
Not enough data
Not enough data
66.7
21.4
0.8
Not enough data
Standard Error of the
Estimated Differences
(%Bias)
Not enough data
Not enough data
2.38
0.77
0.05
Not enough data
P-Value
Not enough data
Not enough data
0.02*
0.02*
0.58
Not enough data
*Denotes statistical significance at 0.05
9.5.1 Comparing Radiello Samplers to SKC Samplers
In addition to the Radiello samplers, SKC charcoal badge samplers were also deployed over the 26-week
and 52-week periods. The badges were equipped with a secondary diffusive cover to lower the uptake rate
by about 28.5 times as compared with the standard SKC 575 badge configuration. This modified badge
uptake rate was approximately 100 times lower than the Radiello charcoal sampler. The %Bias
calculations for the SKC badge semiannual and annual sample measurements are listed in Table 9-11.
9-11
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
Table 9-11. Average %S/as for Average Weekly Radiello Measurements
Compared with Semiannual and Annual Modified SKC 575 Charcoal Badges
Compound
Chloroform
Hexane
Benzene
Trichloroethene
Toluene
Tetrachloroethene
# Observations per compound
26-Week Average vs.
SKC 575 Semi-
annual Measurement
-54
-90
-74
-58
-32
1.0
14
52-Week Average vs.
SKC 575 Annual
Measurement
-35
-90
-39
-23
-33
13
6*
Despite the prolonged exposures, the low badge sampling rates of approximately 0.5 mL/min resulted in
relatively high VOC reporting limits in the 0.3 to 1.0 (jg/m3 range. For the chlorinated solvents, most
results were below or just above the detection limit. In the case of benzene, hexane, and to a lesser degree
TCE, the associated blank levels contributed to a high bias of the measurement. Because the average of
the weekly Radiello measurement was lower than the extended badge measurement, the %Bias for all
compounds except PCE was negative.
The %Bias became less negative for benzene and TCE over the longer period because the background
mass from the blank was proportionally less than the sample mass adsorbed by the badge sampler. The
average mass of benzene measured on the badge blanks was 0.14 \ig, and the average sample mass
measured on the field samplers was 0.34 \ig and 0.57 \ig for the semiannual and annual periods,
respectively. Blank levels of hexane averaged 0.12 \ig on the badge samplers with average sample mass
concentrations of 0.24 and 0.52 \ig for the semiannual and annual samples. The laboratory also reported
TCE in several of the blanks above the detection limit but below the reporting limit. The average TCE
mass in the SKC blanks was 0.011 \ig, the average mass measured for the semiannual samples was 0.029
jig, and the average mass measured for the annual samples was 0.047 \ig.
Overall, the SKC badge concentrations were higher than the average weekly Radiello measurements even
for those VOCs without significant blank contributions such as chloroform and toluene. As was the case
with the Radiello charcoal sampler, the semiannual and annual PCE concentrations compared well with
the corresponding weekly Radiello measurements. The %Bias data also suggest that uptake rates may be
more uniform for the more volatile VOCs using the modified badge than with the Radiello sampler for
extended periods. For example, TCE appeared to have a more stable uptake rate using the badge sampler
when comparing the %Bias of the annual badge sampler with the annual Radiello sampler. Although the
badge %Bias shifts in the positive direction when extending the duration from 26 weeks to 52 weeks, the
shift appears to be more a function of the badge blank levels resulting in a higher concentration at 6
months than due to a drop in the sampling rate over the year-long period.
9.6 Conclusions
Overall, the radial-style charcoal passive sampler performance over periods from 1 day to 1 year was
dependent on the target compound. For the shortest duration of 1 day, the background contribution from
the sorbent and the extraction procedure did not allow for accurate quantitation of benzene and hexane at
concentrations of <2.0 jig/m3. Additionally, the analytical sensitivity of the solvent-extraction technique
was not sufficient to measure indoor air concentrations at the site without enhancements to the detector
sensitivity. All target compounds showed excellent agreement between the numerical averages of the
9-12
-------
Section 9—Results and Discussion: Over what durations do solvent extracted passive samplers
provide useful integration of indoor air concentrations? (Is uptake rate constant?)
weekly exposures and the 2-week integrated measurement, suggesting uniform uptake rates over this
period within a tolerance appropriate to environmental chemistry. Benzene, hexane, toluene, and PCE
exhibited stable uptake rates at 4 weeks, as when evaluating the dataset with concentrations above the
detection limit. The average % Bias for chloroform was just outside the criterion at the 4-week period.
Hexane performed well for durations up to 13 weeks and was slightly above the bias criteria at 26 weeks.
Toluene and PCE continued to demonstrate stable uptake rates at the quarterly, semiannual, and annual
intervals within our 30% tolerance criteria.
In general, the %Bias data suggest that the stability of the uptake rate is a function of the compound's
volatility, as measured by vapor pressure. As shown in Figures 9-2 and 9-3, the VOCs with higher vapor
pressures shifted toward a positive bias at shorter exposure intervals than VOCs with lower vapor
pressures. The maximum sampling duration defined by meeting the average %Bias criterion follows the
expected order from shortest to longest based on the compound's volatility. The most volatile VOC,
chloroform, was the first VOC to exceed the average +/-30% criterion during the study, and sampling
intervals could not be extended beyond 4 weeks. Benzene and TCE were similar in their volatility and
essentially showed comparable performance in their %Bias data, exceeding the criterion in the quarterly
interval measurement. Hexane was a slight exception to the volatility order performing better than would
have been expected at the 13-week duration. The two least volatile compounds, toluene and PCE,
demonstrated a uniform uptake rate over the course of a year.
Given the VOC concentrations at the site, the charcoal sorbent cartridge had sufficient capacity for a 52-
week duration with mass loadings onto the cartridges well under the manufacturer's recommended limit
of 80 mg. The sum of the target VOC masses collected on the sampler for the year-long samples was less
than 0.2 mg for all of the samples, and the total mass on the samplers was estimated to be generally less
than 1 mg. Additionally, water adsorption did not appear to be interference in the sampler performance
and did not result in any negative effects during sample extraction or analysis. If VOC concentrations are
significantly higher than what was measured at this site, additional consideration should be made
regarding extending the sampling duration to ensure sampler capacity is not exceeded.
The concept of reducing the uptake rate on a charcoal passive sampler to extend the sampling interval is
promising based on the data generated using the modified SKC 575 badge. This small and very limited
dataset suggested that uptake rates may be more stable for more volatile VOCs over extended durations of
up to a year. However, blank levels can make accurate measurements challenging when concentrations
and uptake rates are very low.
9-13
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Table of Contents
10. Results and Discussion: Determine if observed changes in indoor air concentration of volatile
organics of interest are mechanistically attributable to changes in vapor intrusion 10-1
10.1 Air Exchange Rate Results and Seasonal Variability—Does this Control Indoor Air
Concentration? 10-1
10.2 Direct Differential Pressure Results—Are They Predictive of Indoor Air
Concentrations by Themselves? 10-3
10.3 Inferred Driving Force from Temperature Differentials—Is This Predictive of Indoor
Air Concentrations by Itself? 10-8
10.4 HVAC System Cycles 10-12
10.5 Trends in Subslab Concentration—Do They Predict Indoor Air Concentration Trends
by Themselves? 10-16
10.6 Do Trends In Shallow and Deep Soil Gas Predict Indoor Air Concentrations? 10-16
10.7 Ambient Concentrations—Are They Significant? 10-17
10.8 Potentiometric Surface/Water Levels—Do They Predict Subslab Concentrations? 10-17
List of Figures
10-1. 422 subslab vs. basement differential pressure (positive values indicate greater
pressurization of the subslab and thus flow toward the basement) 10-4
10-2. Basement vs. upstairs differential pressure (positive values indicate pressurization of the
basement relative to the upstairs) 10-5
10-3. Deep vs. shallow soil gas differential pressure beneath 422 East 28th Street (positive
values indicate a greater pressure in the deep soil gas relative to the shallow soil gas) 10-5
10-4. Basement vs. exterior (above grade) differential pressure at 422 East 28th Street (positive
values indicate that the basement pressure is higher than the pressure in exterior air) 10-6
10-5. Subslab vs. basement differential pressure at 420 East 28th Street (positive values
indicate higher pressure in subslab than in the basement, thus flow toward the basement) 10-7
10-6. Exterior barometric pressure measurements overtime, 420/422 East 28th Street 10-7
10-7. Stack effect driving force for 422 East 28th Street over time 10-9
10-8. Stack effect driving force 420 (unheated) side overtime 10-9
10-9. Summer and fall interior and ambient temperatures 10-10
10-10. PCE concentrations in indoor air vs. stack effect driving force (log scale of
concentration) 10-11
10-11. Chloroform concentrations in indoor air vs. stack effect driving force (log scale of
concentration) 10-11
10-12. Radon concentrations (electret measurements) in indoor air vs. stack effect driving force 10-12
10-13. Differential pressure measurements graphed with HVAC system on/off cycles; HVAC
system status at times when basement vs. exterior differential pressure was observed is
annotated 10-14
10-14. Selected period of indoor and ambient temperature data, green arrow marks December 18
to December 23 HVAC system outage 10-14
10-15. Selected period of indoor and ambient temperature data, green arrow shows March 11 to
March 16 period of HVAC outage 10-15
10-16. PCE, online GC data, 422 side, larger data points used to mark periods of heating system
failure 10-15
10-17. Average depth to water at house, discharge of nearby creek (USGS) and rainfall at house
compared (intensive sampling rounds shown with vertical lines) 10-18
10-18. Storm and sanitary sewers near the test duplex at 420/422 East 28th Street 10-19
10-19. Floor drain in first floor laundry room on 422 side of duplex 10-20
10-i
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
List of Tables
10-1. April/May 2011 Air Exchange Rate Measurement Results 10-1
10-2. September 2011 Air Exchange Rate Measurement Results 10-2
10-3. National Survey of Air Exchange Rates, Reprinted from the EPA Exposure Factor
Handbook (U.S. EPA, 2011) 10-2
10-4. Drain Sampling Data April 13-April 21, 2011 (ng/m3) 10-21
10-5. NY Times Database Report on Indianapolis Drinking Water Showing Relative
Concentrations of PCE and Chloroform 10-21
10-ii
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
10. Results and Discussion: Determine if observed changes in
indoor air concentration of volatile organics of interest are
mechanistically attributable to changes in vapor intrusion
10.1 Air Exchange Rate Results and Seasonal Variability—Does this Control
Indoor Air Concentration?
Air exchange rate measurements were performed using EPA Method IP-4A, which uses passive emitters
and passive samplers known as capillary adsorption tube samplers (CATS) from
• April 27 to May 11, 2011, and
• September 23 to September 29, 2011.'
The emitters were evenly spaced across their respective floors of the 422 side of the duplex:
• 10 perfluorodimethylcyclohexane (PDCH) emitters in the basement
• 10 perfluoromethylcyclohexane (PMCH) emitters on the first floor
• 9 PMCH emitters on the second floor
No emitters were placed on the 420 side of the duplex, but CATS measurements were made there in the
April/May round to estimate the amount of airflow between sides of the duplex. The emitters were
deployed on April 22, 2011, to allow the building to come to equilibrium before sampling and were
essentially left in place throughout the measurement periods.
As shown in Table 10-1, the April/May 422 basement air exchange rates showed excellent agreement for
the duplicates (both 0.74/hour). As shown in Table 10-2, the September measurements for the basement
(0.64/hour and 0.82/hour) are more variable but bracket the April/May measurements. The first floor
measurements were lower in both measurement periods (0.56 in April/May and 0.48 in September). The
September measurements show a pattern of decreasing air exchange rates up through the building
(basement through second floor office).
Table 10-1. April/May 2011 Air Exchange Rate Measurement Results
Date
Deployed
4/27/2011
4/27/2011
4/27/2011
4/27/2011
4/27/2011
4/27/2011
Date
Collected
5/4/201 1
5/4/201 1
5/4/201 1
5/4/201 1
5/4/201 1
5/4/201 1
CAT ID
11015
8441
779
9167
5273
6963
PMCH
Amount
(Pi)
30.74
28.96
301.47
0
0
0.75
PDCH
Amount
(Pi)
127.51
126.67
25.03
0
0
0
Location
422 basement
422 basement
dup
422 first
420 basement
420 first
Travel blank
Primary
Tracer
Deployed
PDCH
PDCH
PMCH
None
None
None
Temperature
(F)
61.29
61.29
67.82
58.17
61.19
68
Calculated
AER1/Hr
0.74
0.74
0.56
NA
NA
NA
Volume
Ft3
4547
4547
9002
4547
9002
0
Duration
of Test
Minutes
10368
10367
10364
10354
10352
0
Fan testing had ended on September 14 and resumed on October 6.
10-1
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Table 10-2. September 2011 Air Exchange Rate Measurement Results
Date
Deployed
9/23/201 1
9/23/201 1
9/23/201 1
9/23/201 1
Date
Collected
9/29/201 1
9/29/201 1
9/29/201 1
9/29/201 1
CAT ID
12621
18744
18185
9024
PMCH
Amount
(Pi)
406.42
253.51
5.94
4.48
PDCH
Amount
(Pi)
28.96
38.35
108.79
121.27
Location
422 office
422 first
422 basement
422 basement
dup
Primary
Tracer
Deployed
PMCH
PMCH
PDCH
PDCH
Temperature
(F)
72.416
72.416
67.77
67.77
AER1/Hr
0.30
0.48
0.82
0.64
Volume
Ft3
9002
9002
4547
4547
Duration
of Test
Minutes
8594
8594
8591
8591
Measurements performed in April/May 2012 did not show any detectable crossover of either tracer into
the 420 side of the duplex. The detection limit of the method is approximately 1 pi per sample and the
lowest amount of tracer collected in one of the rooms with the emitters for that tracer present was 126 pi.
So less than 1% of the tracer concentration detected in the 422 zones where it was released was present on
the 420 side of the duplex.
The concentration of the tracer released in the basement (PDCH) was about 20% of the basement
concentration on the first floor. The concentration of the tracer released on the first and second floors
(PMCH) was detected at about 2% of the first floor concentration in the basement. These percentages
suggest that during that measurement period more flow was up from the basement to the first floor,
although some flow did come from the first floor down into the basement.
All of the measurements of air exchange rate are near the center of the range of Midwestern values
compiled in EPA's Exposure Factor Handbook (U.S. EPA, 2011; Table 10-3).
Table 10-3. National Survey of Air Exchange Rates, Reprinted from
the EPA Exposure Factor Handbook (U.S. EPA, 2011)
Summary Statistics for Residential Air Exchange Rates (in ACHa), by Region
Arithmetic mean
Arithmetic standard deviation
Geometric mean
Geometric standard deviation
10th percentile
50th percentile
90th percentile
Maximum
West
Region
0.66
0.87
0.47
2.11
0.20
0.43
1.25
23.32
Midwest
Region
0.57
0.63
0.39
2.36
0.16
0.35
1.49
4.52
Northeast
Region
0.71
0.60
0.54
2.14
0.23
0.49
1.33
5.49
South
Region
0.61
0.51
0.46
2.28
0.16
0.49
1.21
3.44
All Regions
0.63
0.65
0.46
2.25
0.18
0.45
1.26
23.32
a ACH = Air exchanges per hour.
Source: Koontz and Rector, 1995, as cited in U.S. EPA (2011), Table 19-24.
10-2
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
10.2 Direct Differential Pressure Results—Are They Predictive of Indoor Air
Concentrations by Themselves?
If:
a. the concentration of VOCs or radon in subslab soil gas was constant,
b. the size of the openings such as cracks into the basement was constant,
c. indoor sources and ambient sources of VOCs were negligible, and
d. the air exchange rate of the basement was constant,
then the flux of VOCs into the basement and the concentration of VOCs in the basement air should be
directly related to the differential pressure. Compressible flow through an orifice (narrow opening, which
likely describes a floor crack) is proportional to the square root of differential pressure as described by
Bernoulli's Law (Lau, 2008). Of these conditions b. and c. are likely satisfied in this case, whereas a. and
d. are likely to not be constant overtime.
Differential pressures were collected for five pairs of locations as described in Section 3.6.6. Data for
these locations are presented as Figures 10-1 through 10-5. The ambient barometric pressure measured at
the house is shown as Figure 10-6. Most of the observed differential pressures show the greatest degree
of variability and highest amplitudes (positive and negative) during the winter months (Figures 10-1,
10-3, 10-4, and 10-5). This variation is consistent with the pattern of the observed barometric pressures,
which are moderate and stable during the summer and show greater fluctuation in other seasons (Figure
10-6).
Generally, the differential pressure from subslab to indoor air (Figure 10-1) on the 422 side of the duplex
shows the existence of a driving force for vapor intrusion through most of the year. The maximum
sustained magnitude of this differential pressure near 5 Pa is in close agreement with that observed in
many residential structures observed for radon vapor intrusion (EPA, 1993b). The observed pattern where
this driving force is lowest in July and August is broadly consistent with the pattern of VOC and radon
indoor air concentrations (see Sections 5.1 and 5.2). The pressure differential on the unheated 420 side
shows a qualitatively similar annual pattern (Figure 10-5) but with less magnitude in the fluctuations.
There is little differential pressure between the basement and upstairs (generally less than 1 Pa) (Figure
10-2). This low differential pressure is consistent with the relatively open plan of the structure and the
absence of airtight vapor barriers between the floors. Thus, there is little resistance to flow vertically
through the structure. The data points with differential pressures greater than +/- 1 Pa are mostly
attributable to depressurization of the basement during fan tests (as discussed in Section 12.2).
The differential pressure between deep and shallow soil gas is often among the highest differential
pressures measured, relatively frequently reaching the upper and lower ranges of the sensor used
(+/- 15 Pa) (Figure 10-3). This pressure difference indicates that a driving force for advective gas flow
often exists between the 6- and 13-ft depths beneath the structure. This observation is at the outer edge of
the predicted advective "zone of influence" of the structure that is part of a widely accepted
conceptualization of the vapor intrusion process. This current understanding is well summarized in the
Users' Guide to the Johnson and Ettinger Model (Environmental Quality Management, 2004):
.... scenario where the source of contamination is incorporated in soil and buried some
distance below the enclosed space floor. At the top boundary of contamination, molecular
diffusion moves the volatilized contaminant toward the soil surface until it reaches the
zone of influence of the building. Here convective air movement within the soil column
transports the vapors through cracks between the foundation and the basement slab
10-3
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
floor. This convective sweep effect is induced by a negative pressure within the structure
caused by a combination of wind effects and stack effects due to building heating and
mechanical ventilation. It is also important to recognize that the advective zone of
influence for soil gas flow is limited to soil immediately adjacent to the building
foundation These results indicate that the advective zone of influence will likely be
limited to a zone within 1 to 2m of the building foundation.
However, it should be noted that:
• During some periods of the year, the 13-ft depth may have been below the water table, which
would certainly bias the differential pressure measurement.
• Others have observed differential soil gas pressures in clusters of soil gas ports at substantial
depths. For example, 2 cm of water differential pressure (196 Pa) is indicated at some times of
day between the 6-ft and 74- and 84-ft ports in data presented by Forbes and coworkers. The
southwestern site in question has a 6-in thick concrete slab (not necessarily a building) and a 90-ft
thick vadose zone of basin fill alluvium. The authors interpret their data as showing barometric
pumping down to the deepest depths with the deepest depths being attenuated and out of phase
with the surface pressure cycle. Thus, they observe a reversal in flow direction between deep and
shallow soil gas (Forbes et al., 1993).
The basement-to-exterior pressure differential shows a regular and substantial fluctuation (Figure 10-4)
with some of the most extreme values in the winter seasons. This differential would be expected to be
influenced by barometric pressure fluctuations in the atmosphere as well as the stack effect.
422 Subslab vs. Basement Differential Pressure
ro
CL
to
11
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Hov Dec Jan Feb Mar Apr
Date
Figure 10-1. 422 subslab vs. basement differential pressure (positive values indicate greater
pressurization of the subslab and thus flow toward the basement).
10-4
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Basement vs. Upstairs Differential Pressure at 422 East 28th Street
Date
Figure 10-2. Basement vs. upstairs differential pressure (positive values indicate
pressurization of the basement relative to the upstairs).
422 Deep Soil Gas vs, Shallow Soil Gas Differential Pressure
20
OJ
en
•*
'if IP
> 4 fit
f, |fj
'f! *'$'
Date
Figure 10-3. Deep vs. shallow soil gas differential pressure beneath
422 East 28th Street (positive values indicate a greater
pressure in the deep soil gas relative to the shallow soil gas).
10-5
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
422 Basement vs. Exterior Differentia I Pressure 422
20
to
Q.
tU
i_
CL-
0 -
-10 -
-20
Date
Figure 10-4. Basement vs. exterior (above grade) differential pressure at 422 East 28th Street
(positive values indicate that the basement pressure is higher than the pressure in exterior air).
10-6
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
420 Subslab vs. Basement Differential Pressure
03
CL
CO
i/5
O)
r- T- r- T- -rr- *- tr- TT- r- r- r- T- O4 CM CM CN
RRRRRRRRRR RRRRR R
Date
Figure 10-5. Subslab vs. basement differential pressure at 420 East 28th Street (positive values
indicate higher pressure in subslab than in the basement, thus flow toward the basement).
Exterior Barometric Pressure at 420/422 East 28th Street
Date
Figure 10-6. Exterior barometric pressure measurements overtime, 420/422 East 28th Street.
10-7
-------
Section 10 — Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
10.3 Inferred Driving Force from Temperature Differentials — Is This Predictive of
Indoor Air Concentrations by Itself?
To examine the stack effect as a potential driving force, we computed the predicted strength of the stack
effect, based on an equation relating flow to the indoor/outdoor temperature differential from Dr. Sam
C.M. Hui (1993):
3.3 Flow caused by thermal forces
If the building's internal resistance is not significant, the flow caused by stack effect may be estimated by:
<7)
0
where Q = ai r flow rate (m3/s)
K = discharge coefficient for the opening (usually assumed to be 0.65)
A = free area of inlet openings (m2)
A/7 = height from lower opening (mid-point) to neutral pressure level (m)
T", = indoor air temperature (K)
TO = outdoor air temperature (K)
We simplified this equation for purposes of plotting by taking only the variable portions:
Q a (Ti -To/Ti)1/2 or Q a (To -Ti/To)1/2
This quantity was then calculated for each 30-minute interval and then averaged over the 1-week period
of operation of the passive samplers.
When we compare the calculated strength of the stack effect, we see that it is stronger and more variable
on the heated side (Figure 10-7) than the unheated side (Figure 10-8) of the duplex. Although the stack
effect is primarily associated in the VOC vapor intrusion field with winter conditions, the existence of a
"solar stack effect" under summer conditions is well known and should not be ignored (University of
Minnesota, 2008). As shown in Figure 10-9, there is a substantial cooling of ambient air at night in
Indianapolis in the summer, but the building tends to hold heat because the windows were not being
opened at night. Thus, it is not unusual to have a condition where the interior is 15° F higher than the
exterior, allowing for a relatively strong stack effect. In this case, the windows could not be opened at
night because the house was not staffed overnight; similar conditions are common in urban residences:
10-8
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Stack Effect Driving Force vs. Time for the 422 (Heated) Side
04C
• j;: :-.:
-a «o
; «
i «
s
a
1
«
-;~
I
g
1
0
A
§
1
\
§
u
^
Dale
Figure 10-7. Stack effect driving force for 422 East 28th Street over time.
Stack Effect Driving Force vs. Time for the 420 Side
5
I
I
I
Figure 10-8. Stack effect driving force 420 (unheated) side overtime.
10-9
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Summer/Fall Temperatures
100
t
\\ \ \ % % \
j &f *"&+ *Vf Vj &i "^5
•^ ^ ^ ^ *- '« •>
-Oon
Dof
: r-
-Iflff
-PW
-Fart
Gen
I Off
MB"
Date
Figure 10-9. Summer and fall interior and ambient temperatures.
Brick and mortar buildings, asphalt streets, and tar roofs absorb daytime heat and slowly
release it at night. Consequently, temperatures in urban areas can be warmer than rural
areas by several degrees both day and night ... Socioeconomic factors also place urban
residents under extra risk. Some people in cities do not have air conditioning, while
people in high crime areas may be afraid to open their windows. (National Weather
Service, 2012)
We then plotted the average strength of the stack effect over one week against indoor air concentrations
of key contaminants. This analysis suggests that the strength of the stack effect explains some but not all
of the variability in indoor concentrations we observed (Figures 10-10 through 10-12). PCE and
chloroform indoor concentrations increase nearly exponentially with the computed stack effect driving
force and appear linear on the semilog plots presented (Figures 10-10 and 10-11). Radon concentrations,
in contrast, apparently increase linearly with computed stack effect driving force (Figure 10-12). Note on
the radon plot that the first floor sampling locations on both sides of the duplex (red bordered data points
in Figure 10-12) appear to overlay each other and lie below the trend of the basement data, as would be
expected due to the greater dilution by outside air on the first floor.
PCE and chloroform also seem to show a greater concentration variability (scatter) than radon at higher
stack effect driving forces. One possible interpretation of this higher scatter follows. The radon
concentration in a horizontal layer under the slab is relatively uniform. The VOC concentrations under the
slab within a
10-10
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
PC E Stack Effect at A II Locations
"s
I
3
o <5> °
-O.D5ODD OJXJDQC o Bsccc c :ici
Stack Effect Driving Force Index from Temperature Data
Figure 10-10. PCE concentrations in indoor air vs. stack effect driving force
(log scale of concentration).
::
m" ,
s
J
1
oon
-OJ3?
Chi on
J $ o
ft 8^^
8 88^?^1
» RDOBro
Stack Effect Drivi
Chloroform Stack Effect at A It Locations
0 0
•
* a
o <»
£
•
o
* -:::i :;::-: r
III:-" IJIOCCO D 2SCOC' CJCOCC
Figure 10-11. Chloroform concentrations in indoor air vs. stack effect driving force
(log scale of concentration).
10-11
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Electret Radon Data vs. the Stack Effect on the 4ZZ Side
(Y Axis Truncated for Clarity, Removing one Data point)
2QXIQ
+422 1st Elect-in
•420 1st ElKt-eB
-5CO
O..QQQQQ O.QSOOQ a..lflQQQ 0.15000 GJOMQ 0-25000
Stack Effect Driving Force Index from Temperature Data (1st vs. outside)
Figure 10-12. Radon concentrations (electret measurements) in indoor air
vs. stack effect driving force.
horizontal layer are very uneven and the spatial pattern changes with time. The more uniform radon
concentration is an expected result because most geological materials can generate at least some radon.
Thus, the short half-life radon concentration is continually renewed. On the other hand, the more variable
chlorinated VOCs have a relatively long half-life under aerobic conditions and can be shifted around in
their spatial position between the slab due to wind effects (U.S. EPA, 2012a). The cracks and other points
of entry to the foundation are probably unevenly distributed horizontally across the foundation. So the
scatter at high stack effect flows could reflect that in some cases the VOCs are "in position" beneath the
key cracks and sometimes they are "out of position" because of wind direction changes.
10.4 HVAC System Cycles
Recalling our objective D-l "Identify any seasonal variations in VI fluxes in radon and VOCs as they
relate to the use of HVAC in the home." We can examine this effect using a number of timescales and
data sets:
1. In Section 10.3, we calculated average stack effect driving force for 1-week sampling periods
from calculations of the driving force made on half-hour intervals. Thus, we observed the effect
of HVAC systems indirectly, because the HVAC maintained a temperature inside the 422 side of
the duplex that was different from that experienced by an unconditioned structure (warmer in
winter and cooler in parts of the summer).
10-12
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
2. We also directly measured the HVAC on/off cycles with 1-minute time resolution. This
measurement was experimentally difficult to make, as well as to graph once the data were
acquired. In Figure 10-13, we present one small time interval of the data, showing the following:
a. On this day, the thermostatically controlled HVAC system was going on and off rapidly—
typically on for 6 or 7 minutes and then off for 6 to 8 minutes, repeatedly.
b. To avoid overloading the data system, the differential pressure sensor was set up to measure
only once every 15 minutes as an instantaneous measurement.
c. Subslab vs. basement and basement vs. upstairs differential pressures were essentially flat
over this 4-hour time interval.
d. The basement vs. exterior differential pressure had quite a bit of variability. However, that
variability does not appear to be connected to the off or on status of the HVAC system.
e. This result is expected because the HVAC system, like most residential systems, is primarily
recirculating air. Thus, the effects across the building envelope are probably controlled by
temperature differentials and wind loads that operate on different timescales/cycles from the
HVAC on/off Therefore, we deprioritized further review of the HVAC system data on the
scale of minutes.
f. We can also examine the effect on the HVAC system on a timescale of hours and days using
a few inadvertent experiments. In Sections 5.2 and 5.4, we show the effects from the thefts of
the window unit air conditioners (which occurred on July 13 and July 15, abruptly taking
them out of service until replacement on August 2 and August 3).
We also observed sharp dips in temperature upstairs in the 422 side of the duplex from March 10 through
March 16, 2011 and from December 18 through December 23, 2011 (Figures 10-14 and 10-15)
associated with furnace breakdowns and repairs. Unfortunately the online gas chromatograph (GC) also
went down on December 22, coming back into service on December 29. The decrease in the indoor-
outdoor temperature differential (and thus stack effect driving force) that occurred when the heat was off
from December 17 to December 23 appears to be associated with a decline in some of the VOC
concentrations inside the house (Figure 10-16).
10-13
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Pressure and HVAC over 4,5 Hours
*
"•?
I
1
ft
K
4
—
^K. .
*
•
_ -—
.
*
•
•
•
on
1
*
m -jr.
oft
m
*
'
- '
b
nil i
•
.h. .
.
XI
•
1
•
n
l 1
g
l
ulrne
Q
•
1
_ -
. _
!
a
«
;
«
V
* on
H
--
* «
nil
• *JKC(ir««llKI*^
* *
!
- ICfM. ff.
s ? I = s £ £ Dtte *
Figure 10-13. Differential pressure measurements graphed with HVAC system on/off
cycles; HVAC system status at times when basement vs. exterior differential
pressure was observed is annotated.
Fall/Winter of 2011 -12 Temperatures
' r',
^*5 "^i ^^L ^
•%, *^ ^ -Sj^
% %
\ '%
V %
Date
Figure 10-1. Selected period of indoor and ambient temperature data, green arrow marks
December 18 to December 23 HVAC system outage.
10-14
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Indoor and AmbientTemperatures
/^V^^/A;^^
k^t-X^W
I
-Li : i-
Date
Figure 10-2. Selected period of indoor and ambient temperature data, green arrow
shows March 11 to March 16 period of HVAC outage.
PCE Fi el d GC Indoor Air Data - 422
J
€
B
3
a. '*•*
B
B D
a D
-O-
g
MjB
LJ
OTT
D a c LJ -*::•:-; :: 2 =;'-'.'
a a a
a o a
ID D
a a OB
B em
a a ODD D
axi aa c
D n m a
—^l^—a
_ 0 D O D <
Q EX] D ni mmc
Figure 10-3. PCE, online GC data, 422 side, larger data points used to mark periods
of heating system failure.
10-15
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
10.5 Trends in Subslab Concentration—Do They Predict Indoor Air
Concentration Trends by Themselves?
There is substantial agreement in the patterns between the subslab data for certain ports and the indoor air
concentrations. The port that shows the highest concentration of PCE and nearly the highest
concentrations of chloroform observed (SSP-1) has the same pattern of high concentrations during the
severe winter of 2010, declining to a minimum in the summer of 2010 and rebounding somewhat in the
fall and winter of 2011 (compare figures in Sections 5.1.1 and 5.1.2). WP-3, also on the 422 (heated) side
of the building, shows a similar trend. It is notable that this correlation occurs between the highest
concentration subslab port and the highest concentration portion of the building interior. It is not known
whether this observation is indeed predictive or merely coincidental because there is a lack of a similar
correlation on the 420 side of the duplex. However, the magnitude of the changes is suggestive; for
example, the decline in PCE between January and July 2011 is approximately 40x in both SSP-1 and 422
basement indoor air. Similarly, the decline between March and August is about 20x in both SSP-1 and
422 basement indoor air. The subslab concentrations seem to increase proportionally more between July
2011 and January 2012 than the approximate 2x increase of PCE seen in the 422 side indoor air.
Similarly, the rise in indoor air of chloroform is about 8x from July to December 2011, whereas the rise in
subslab chloroform was substantially greater (approximately 20x).
10.6 Do Trends In Shallow and Deep Soil Gas Predict Indoor Air
Concentrations?
As noted in Section 5.1.3, a seasonal pattern with higher concentrations in winter and lower
concentrations in summer was also noted in certain soil gas sampling locations further away from the
slab. For chloroform, this seasonal pattern can be discerned in deep external soil gas SGP5-13 (Figure
5-19), SGP6-9 (Figure 5-21), SGP7-13 (Figure 5-23) as well as deep soil gas beneath the slab such as
SGP8-9 (Figure 5-25), SGP9-9 (Figure 5-25), SGP10-13 (Figure 5-27), and SGP11-13 (Figure 5-29).
This pattern clearly cannot be due to temperature, because temperatures at these depths would be
expected to be relatively stable, and the concentration of VOCs in the vapor phase would be expected to
increase with increasing temperature. However, this trend could reflect the following:
• lower groundwater levels in summer (and thus a greater distance between the presumed source of
VOCs and the sampling point allowing greater attenuation), and/or
• the reduced driving force from the stack effect in summer (see Figures 10-7 and 10-8 in Section
10.3) inducing less flow upward, and thus allowing greater attenuation/dilution.
These seasonal soil gas concentration changes in chloroform are about 1 order of magnitude and, thus,
slightly smaller in magnitude than the changes observed in indoor air (see Section 5.1.1). However,
because the magnitude of the driving force for soil gas entry is also decreasing, the effect of lower soil gas
concentrations would be expected to be compounded by reductions in volumetric flux into the structure
(Section 10.3). In order to prove that these soil gas concentration changes are indeed causing the observed
changes in indoor air, it would be necessary to know whether the portions of the subslab soil gas that
exhibit this seasonal trend account for a significant proportion of the soil gas VOC flux into the building.
In later studies we will attempt to determine the impacts of seasonal trends in soil gas through tracer
testing and pumping tests.
A similar seasonal pattern is seen occasionally in PCE: SGP1-16 (Figure 5-12), SGP6-9 (Figure 5-22),
and SGP9-9 (Figure 5-26) (see Section 5.1.3). The physical causes of these trends could be similar to
those for chloroform. However, because these seasonal patterns in soil gas for PCE are less frequent and
clear, the case for them being causative of the observed trends in indoor air PCE is weaker.
10-16
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
10.7 Ambient Concentrations—Are They Significant?
The data presented in Sections 5.1.1 through 5.1.3 show that ambient concentrations are consistently
among the lowest measured for chloroform and PCE. Ambient concentrations of PCE and chloroform
observed at this site are relatively uniform with time, whereas indoor concentrations display a strong
seasonality. A concentration comparison suggests that the only situation in which the ambient
concentration appears to dominate indoor air for PCE and chloroform is for the 420 first floor during
summer.
10.8 Potentiometric Surface/Water Levels—Do They Predict Subslab
Concentrations?
As shown in Figure 10-17, there is a close correlation between the observations of rainfall at the duplex,
rises in the nearby creek, and increases in the groundwater table beneath the duplex. It is not known
whether this rise in the water table reflects the influence of leakage from the nearby combined sewers
(Figure 10-18) or an indirect potentiometric effect from the influx of creek water into the shallow aquifer.
Other than the early September and early October weeks that were affected by the fan tests (discussed in
Section 12.2), the peak indoor air concentrations of chloroform on the 422 side of the duplex peaked in
March and mid-December 2011 (see Section 5.1.1). Although these months were periods of relatively
high stack effect driving force, they do not correspond closely to the maxima of the stack effect driving
force calculation that occurred on approximately January 15, 2011, and January 15, 2012 (see Section
10.3). These chloroform maxima do exactly correlate; however, to significant storm events recorded in
the U.S. Geological Survey (USGS) discharge records of the nearby Fall Creek as well as in the rainfall
and depth to groundwater data sets. Coincidentally, these maxima also corresponded to our first and third
intensive rounds, which were scheduled well in advance, on the basis of anticipated temperatures (not
rainfall). Although it is not as crisp a correlation, there is some evidence of chloroform reaching maxima
in the highest concentration subslab ports (SSP-1 and SSP-4 beneath the 422 side) in March and
December 2011 as well (see Section 5.1.2).
In contrast, the maximum PCE concentrations in 422 indoor air occurred in mid-January 2011 and mid-
January 2012 (see Section 5.1.1). These maxima in PCE do appear to coincide with the stack effect
driving force estimate derived from temperature data (Section 10.3). These two January maxima are also
visible in the plot of the highest concentration subslab PCE port beneath the 422 side (SSP-1, Section
5.1.2). The PCE behavior on the 420 side appears significantly different—it declines more rapidly in
January 2011 and reaches its next maxima earlier—in late November or early December (Section 5.1.1).
The chloroform behavior on the 420 side is also different from the behavior on the 422 side. Chloroform
also declines more rapidly on the 420 side than the 422 side. Chloroform shows three maxima that are not
related to the September/October fan testing—January 15, March 31, and December 15,2011. The latter
two maxima come at the end of significant storm events recorded in the nearby Fall Creek.
10-17
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Comparison between Water Level. Discharge, and Rainfalt
o
o
— 6500
<£ eooo
5 5500
5000
4500
4000
3500
MOO
2500
2MQ
1500
1000
II
II
II
II
II
II
It
j^ JL
II
II
II
II
II
II
II
1st Intensive Round
I'nd Intensive Round l( 3rd lnl«ns(vo
^^^-^, Round
ll
10
u
I
s
a
a
20
Figure 10-17. Average depth to water at house, discharge of nearby creek (USGS)
and rainfall at house compared (intensive sampling rounds shown with vertical lines).
10-18
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
420 E. 28th St, Indianapolis, IN
4 Storm_sewers.shp
4 Sanilary_sewers.shp
Gas Lines
• Storm sewer structures,.slip —»-_ j
• Sanitary _sewerstructure5.slip
• CSO's
S
Figure 10-18. Storm and sanitary sewers near the test duplex at 420/422 East 28th Street.
In summary, it appears that
• chloroform trends in indoor air visually correlate with hydrogeology/storm events.
• PCE trends in indoor air more closely correlate with the stack effect driving force, and
• soil gas data suggest that both chloroform and PCE concentrations peak just above the water
table.
Although our data analysis is ongoing, we have several potential explanations for the temporal trends
observed in this data set:
• A sewer gas odor was noted on the first floor of the duplex on approximately April 1, 2011, and
was traced to a floor drain in the first floor laundry room on the 422 side (Figure 10-19). All
drains and fixtures except those in the basement necessary for the continued operation of the
furnace were sealed as of May 10, 2011. This sewer gas pathway is not considered to be a major
source to indoor air, because neither the appearance of this odor nor the sealing of the drains
coincided with a significant inflection point in the overall indoor air time series (Section 5.1.1).
However, it is probable that this entry route contributed to the higher indoor air concentrations of
chloroform observed on the first floor than in the basement in March 2011.
10-19
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Figure 10-19. Floor drain in first floor laundry room on 422 side of duplex.
Direct samples of sewer gas with Radiello samplers were collected in late April and are
summarized in Table 10-4. The observed ratios of PCE/chloroform and PCE/TCE also support
the conclusion that the first floor drain route is not a controlling factor on concentrations other
than chloroform on the 422 first floor.
Only one sample had concentrations high enough to be credible as a significant source of VOCs
for a large area—the 422 laundry room drain that was also the location of our "sewer gas" odor
observations. Surprisingly, the concentrations in that drain were almost exactly the same for
chloroform and PCE—around 300 (ig/m3, which is not the expected result of a nonspecific
reaction between bleach and natural organic matter, which is a primary source of chloroform in
most sewers (Odabasi, 2008; Whitmore and Corsi, 1994; Hass and Hermann, 1998). Nor is it
what one would expect from Indianapolis city drinking water passing through use and on into the
sewers or leaking from water mains down into underlying sewer pipes. PCE was found in only 1
of 74 samples of Indianapolis drinking water at a concentration of 0.53 ppb. Chloroform was
found in 67 of 73 Indianapolis drinking water samples at an average concentration of 18.9 ppb
(see Table 10-5). The drinking water shows a set of products typical of disinfection by-products,
including trihalomethanes and haleoacetic acids.
10-20
-------
Section 10—Results and Discussion: Determine if observed changes in indoor air concentration
of volatile organics of interest are mechanistically attributable to changes in vapor intrusion
Table 10-4. Drain Sampling Data April 13-April 21, 2011 (|jg/m3)
Location
Trip blank
Field blank
Ambient
422 basement S
422 basement N
422 first floor
420 basement S
420 basement N
420 first floor
422 laundry drain in floor
422 bathroom — in sink
422 basement floor drain (near
furnace)
420 bathroom — in sink
420 laundry room drain in floor
Chloroform
ND
0.064
0.17
0.2
0.18
320
1.2
0.65
1.5
1.6
TCE
ND
ND
0.039
0.097
0.061
0.061
0.043
0.038
0.028
5.6
0.089
0.091
0.054
0.087
PCE
ND
0.011
0.14
1.4
0.78
0.7
0.36
0.31
0.21
310
1.4
1.7
0.96
2.6
Ratio
PCE/Chloroform
-
-
2.19
2.06
1.77
1.06
2.12
1.55
1.17
0.97
1.17
2.62
0.64
1.63
Ratio
PCE/TCE
-
14.43
12.79
11.48
8.37
8.16
7.50
55.36
15.73
18.68
17.78
29.89
ND = no data
Table 10-5. NY Times Database Report on Indianapolis Drinking Water Showing Relative
Concentrations of PCE and Chloroform
Contaminant
Average
Result
Maximum
Result
Health
Limit
Legal
Limit
Number of Tests
Total
Positive
Results
Above
Health
Limit
Above
Legal
Limit
Contaminants above legal limits
Total haloacetic acids
Total trihalomethanes
36.58 ppb
33.56 ppb
107.8
70
—
60
80
44
111
44
105
2
0
3
1
Contaminant below legal limits, but above health guidelines
Tetrachloroethylene
0.00 ppb
0.53
0.06
5
74
1
1
0
Contaminants found within health guidelines and legal limits
Chloroform
Chloromethane
18.90 ppb
0.01 ppb
70
30
80
73
73
67
1
0
0
0
0
Source: New York Times, 2012
• These data suggest that the gas in the floor drain on the first floor of 422 was probably influenced
by a location-specific source of PCE. This could reflect migration of PCE in soil or groundwater
through the sewer's acting as a preferential pathway for vapor intrusion. Alternately, PCE could
be present as a free product in low points of the sewer system from previous residential,
commercial, or industrial uses. Similar pathways have been observed at other locations with
vapor intrusion issues (Distler and Mazierski, 2010).
• Referring back to the indoor air trends discussed in Section 5.1.1, this sewer gas with an even
ratio of chloroform and PCE cannot be the primary driver of the indoor concentrations because
PCE is highest in week 1 and declines steadily through July, whereas chloroform has a distinct
peak in weeks 7 through 10 on the 422 first floor.
10-21
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Table of Contents
11. Results and Discussion: Do groundwater concentrations control soil gas concentrations at this
site? And thus indoor air concentrations? 11-1
11.1 Potentiometric Surface Changes (and Correlation to Local Surface Water Bodies) 11-1
11.2 Groundwater Concentration Trend 11-5
11.2.2 Is the Groundwater Concentration Trend Correlated to Potentiometric
Surface? 11-8
11.2.3 Is the Groundwater Concentration Trend Correlated to Indoor Air
Concentrations? 11-8
11.3 Soil Moisture Trends 11-8
11.3.1 Correlation with Rainfall Measurements 11-8
11.3.3 Relationship to Observed Stratigraphy 11-9
11.4 Correlation of Groundwater Concentration Changes to Deep Soil Gas 11-11
11.5 Revisions to Conceptual Site Model 11-14
List of Figures
11-1. Stacked graph presenting depth to water in feet (top—red circles), discharge at Fall Creek
in ft3/s (middle—blue line), and rainfall in inches (bottom—green line) 11-2
11-2. Monitoring well water data for PCE versus time 11-6
11-3. Flooded SGP water data for PCE versus time 11-7
11-4. Soil moisture: irrometer moisture data in centibars for the interior and exterior of the
422/420 house 11-10
11-5. Lithological fence diagram showing some of the major soil types beneath the 422/420
house 11-10
11-6. Lithological fence diagram showing some of the major soil types beneath the 422/420
house 11-11
11-7. PCE concentrations at each of the SGP1 ports overtime 11-12
11-8. PCE concentrations at each of the SGP2 ports vs. time 11-12
11-9. PCE concentrations at each of the SGP5 ports overtime 11-13
11-10. PCE concentrations at each of the SGP6 ports overtime 11-13
11-11. PCE concentrations at SGP8 and SGP9 ports over time 11-14
11-12. Temporal plot of log indoor air concentration for VOCs (|ig/m3) and radon (pCi/L) by
sample location over the study period 11-16
List of Tables
11-1. External Soil Gas Locations and Their Flooded Status during Different Times 11-3
11-2. Internal Soil Gas Locations and Their Flooded Status during Different Times 11-4
11-3. Groundwater Monitoring Well Information 11-5
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
11. Results and Discussion: Do groundwater concentrations control
soil gas concentrations at this site? And thus indoor air
concentrations?
11.1 Potentiometric Surface Changes (and Correlation to Local Surface Water
Bodies)
Figure 11-1 presents in a stacked representation the relationship between depth to water readings taken at
the 422/420 house, Fall Creek discharge, and the rainfall taken at the house for the duration of the project.
The most noticeable feature is that the highest degree of rainfall occurred in early March 2011. This
feature was immediately reflected in the discharge at Fall Creek and in the decreasing depths to water at
that time. Note that multiple depth readings were taken in early March because of the coincidence of the
heavy rain with the first intensive sampling round. The hotter months appear to be the drier, with the
spring and fall being some of the wetter months.
The water table had been expected to be at about 17 ft, just under the deepest of the soil gas ports
(16.5 ft). However, 2011 appears to have been a wetter than average year, resulting in the water table at
about 15 ft bis for most of the project. There was a particularly dry period in the area from August until
October 2011, which allowed many of the 16.5-ft depths to be sampled again for a brief period (see
Tables 11-1 and 11-2). The intensive rounds can be seen on Figure 11-1, indicated by the vertical dashed
and solid lines. The first intensive round happened to occur during the wettest period of the whole project.
The second intensive took place during one of the drier periods of the project. Stream discharge had
increased toward December, and by this time, the 16.5-ft depths were nonfunctional (see Figure 11-1).
For the most part, all three sections of the stacked graph match well indicating a rapid connection between
precipitation, surface water levels, and groundwater levels. The first large peak has the closest match
between these three parameters. A rainfall peak in March seems to have a time delayed relationship to
similar peaks in the discharge and water depth graphs. All three parts of the stacked graph reflect the drier
period well. However, a series of peaks on the stream discharge figure in December correspond to the
groundwater depth graph, but with only a small peak on the rainfall graph (Figure 11-1). These
December observations suggest a regional rainfall event in the catchment of Fall Creek not recorded as
substantially at the house. In the Indiana weather data, Central Indiana was 2.10 inches above normal
rainfall levels for the month of December (see Table 5-3 in Section 5.5).
Tables 11-1 and 11-2 show recorded dates or date ranges when soil gas ports external to the house (Table
11-1) and internal (Table 11-2) would not pump (WP) presumably because of saturated soils or contained
visible water in their lines (WIL). Red lettering indicates dates during the first intensive round (the wettest
time during the project as indicated in Figure 11-1). Nearly every external soil gas port was at some time
unable to pump or had water in the line, but some had blockages more frequently and a few never
operated again after a blockage occurred. Almost all of the recorded blockages occurred between
January 25, 2011, and May 25, 2011, with only a few occurring afterward (see Tables 11-1 and 11-2).
Many of the dates when blockages occurred coincided with the high water levels and rains during the first
intensive round, but this could also have been a result of checking the ports multiple times per day or
week. We interpret these observations of water in lines or points that would not pump as indicating
infiltrating precipitation.
Ports SGP5-6, SGP6-3.5, and SGP7-6 were unusable from the beginning of the March intensive round
through late May 2011. All 13-ft depths were unusable during the intensive round but became usable and
were used in place of the original plan of frequent sampling from the 16.5-ft depths shortly after (see
Table 11-1).
11-1
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Comparison between Water Level, Discharge, and Rainfall
£
Q
|
u
6000
5500
5000
4500
4000
3500
3000
2500
2000
1500
1000
1st Intensive Round
2nd Intensive Round u 3rd Intensive
"~—-^ Round
ii •~—^_
II
ii
II
ii
II
ii
M
n
M
II
L
\ \
v- •>
0
o
5 1
o
10 *
E
15 =
ff
^
Date
Figure 11-1. Stacked graph presenting depth to water in feet (top — red circles), discharge at Fall
Creek in ft3/s (middle — blue line), and rainfall in inches (bottom — green line).
All are overtime for the duration of the project. Superimposed on these graphs are the durations
of the intensive round, marked by dashed and solid lines.
11-2
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Table 11-1. External Soil Gas Locations and Their Flooded Status during Different Times
Location
SGP1-3.5
SGP1-6
SGP1-9
SGP1-13
SGP1-16.5
SGP2-3.5
SGP2-6
SGP2-9
SGP2-13
SGP2-16.5
SGP3-6
SGP3-9
SGP3-13
SGP3-16.5
SGP4-3.5
SGP4-6
SGP4-9
SGP4-13
SGP4-16.5
SGP5-3.5
SGP5-6
SGP5-9
SGP5-13
SGP5-16.5
SGP6-3.5
SGP6-6
SGP6-9
Date and Status (WP=will not pump; WIL=water in the line; red lettering=1st Intensive Round)
02/15/11 WIL
02/03/11 WIL
03/03/11 WIL
03/02/11 WIL
02/21/11 WP
03/05/11 WIL
03/05/1 1 WP
03/03/11 WIL
03/02/11 WIL
03/16/11 WP
03/02/1 1 WP
12/15/11 WIL
03/02/1 1 WP
02/10/11 WP
02/22/1 1 WP
02/03/1 1 WP
02/22/11 WIL
03/02/1 1 WP
03/02/1 1 WP
03/02/11 WIL
02/18/11 WP
02/17/11 WIL
03/02/1 1 WP
03/02/1 1 WP
01/25/11 WP
05/18/11 WP
01/18/12 WP
Until
04/26/1 1 WP
03/04/1 1 WP
03/06/11 WIL
03/16/11 WIL
03/06/1 1 WP
03/03/1 1 WP
03/06/1 1 WP
03/02/1 1 WP
03/02/1 1 WP
03/02/1 1 WP
03/02/1 1 WP
03/06/1 1 WP
03/16/11 WP
03/03/11 WIL
03/02/1 1 WP
03/08/11 WIL
03/06/1 1 WP
03/16/11 WP
03/02/1 1 WP
03/29/11 WIL
01/18/12 WP
03/22/11 WIL
03/07/1 1 WP
03/07/1 1 WP
03/04/1 1 WP
03/07/1 1 WP
03/16/11 WIL
until
until
03/03/11 WIL
03/07/1 1 WP
03/04/11 WIL
until
03/07/1 1 WP
until
This port was never used again after it flooded. Water would never stay out of it.
03/17/11 WIL
01/18/12 WP
03/05/1 1 WP
12/15/11 WIL
05/25/1 1 WP
05/25/1 1 WP
03/06/1 1 WP
05/25/1 1 WP
05/25/1 1 WP
03/22/11 WIL
03/06/1 1 WP
01/18/12 WP
03/07/1 1 WP
03/08/1 1 WP
This port would never pump again
This port would never pump again
03/07/11 WIL
12/01/11 WP
03/08/11 WIL
03/09/11 WIL
03/09/1 1 WP
03/16/11 WIL
05/03/1 1 WP
(continued)
11-3
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Table 11-1. External Soil Gas Locations and Their Flooded Status during Different Times (continued)
Location
SGP6-13
SGP6-16.5
SGP7-3.5
SGP7-6
SGP7-9
SGP7-13
SGP7-16.5
Date and Status (WP=will not pump; WIL=water in the line; red lettering=1st Intensive Round)
03/06/1 1 WP
03/16/11 WP
03/05/1 1 WP
02/03/1 1 WP
02/23/11 WIL
03/06/1 1 WP
03/16/11 WP
03/07/1 1 WP
04/19/11 WP
03/02/1 1 WP
03/07/1 1 WP
04/20/1 1 WP
until
04/26/1 1 WP
05/25/1 1 WP
05/03/1 1 WP
12/15/11 WP
Note: where multiple successive dates are omitted, the word "until" is used.
The 13-ft depths are flooded by 03/01/11, but they were sampled instead of the 16.5-ft depths as of 03/17/11 for most of the rest of the project.
The 16.5-ft depths became possible to sample for soil gas again for most of the locations as of 08/22/11, except for SGP9-16.5.
However, by 10/22/11 only the 16.5-ft depth at SGP6 could still be used. It was no longer usable a few weeks after that.
Table 11-2. Internal Soil Gas Locations and Their Flooded Status during Different Times
Location
SGP8-16.5
SGP9-16.5
SGP10-16.5
SGP11-16.5
SGP12-16.5
WP-2
SSP-2
Date and Status (WP=won't pump; WIL=water in the line; red lettering=1st Intensive Round)
02/19/11 WIL
02/10/11 WIL
02/19/11 WIL
03/15/11 WIL
02/22/1 1 WP
03/03/11 WIL
03/09/11 WIL
03/16/11 WIL
03/16/11 WP
03/16/11 WIL
03/15/11 WIL
08/22/1 1 WP
08/29/1 1 WP
The 13-ft depths are flooded by 03/01/11, but they were sampled instead of the 16.5-ft depths as of 03/17/11 for most of the rest of the project.
The 16.5-ft depths became possible to sample for soil gas again for most locations as of 08/22/11, except for SGP9-16.5.
However, by 10/22/11 only the 16.5-ft depth at SGP6 could still be sampled. It was no longer usable a few weeks after that.
11-4
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
As stated before, the 16.5-ft depths were flooded and could not be used for soil gas sampling during the
first intensive round. Additionally, SGP1-3.5, SGP4-3.5, and SGP4-6 became unusable as well either
because of water in the line or they were unable to pump (see Table 11-1). SGP4 likely had difficulties
due to its location in a breezeway between the 422/420 house and the building on the adjacent property.
During freezing weather, slush and ice would pond and melt and refreeze in the path along this
breezeway. In warmer weather, rainwater would gather in the walkway, flooding the path. It is possible
that this constant influx of water kept some of the shallow SGP4 ports flooded.
Observations of water in lines or points that would not pump were less frequent at depths above the water
table beneath the house, suggesting that the house created a considerable moisture shadow and that
drainage away from the house was probably good.
11.2 Groundwater Concentration Trend
Groundwater samples were taken approximately once each month from all monitoring wells, generally
with permeable diffusion bags. Sample locations included a nest of three wells at the south side of the
house designated MW-1 A - C, a nest of three wells at the north side of the house called MW-2 A - C, and
a single well called MW-3 inside the house. The lettering of these wells used during sampling is in some
cases different than the lettering used at the time of construction. 1 The lettering of the intervals used
during all groundwater sampling is shown in Table 11-3.
Table 11-3. Groundwater Monitoring Well Information.
Well Name
MW1A
MW1B
MW1C
MW2A
MW2B
MW2C
MW3
Screened Interval
Depth (ft)
24-26
21-24
16-21
24-26
21-24
16-21
-19.5-24.5
PDB Tether Length (ft)
25
22'3"
18'3"
24
22'5"
18'6"
23'2"
Measured Total Depth of Well
(ft on 08/29/1 2)a
>25
23'9"
20'4"
>25
24
21 '2"
24'8"
3 This closely agrees with another set of total depth measurements taken 1/6/11.
The PCE data from these monitoring wells are plotted on Figure 11-2. Chloroform was not detectable in
any of these samples. Note, however, that chloroform had been detected by the EPA laboratory in
samples collected for preliminary site screening (Table 3-6). Chloroform had also been detected in
groundwater in previous work at the Mapleton-Fall Creek site across the intersection of East 28th St and
Central Avenue from our duplex.
Note that the MW1 nest was constructed over two widely sampled mobilizations with the middle interval
constructed last. This is important to consider when interpreting the appended boring logs.
11-5
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Monitoring Well Groundwater PCE Data
2.5
^3_
C
o
01
o
0.5
MDL
or
T3T
DMW1APCE
^MWIBPCE
• MW1C PCE
DMW2APCE
• MW 2C PCE
MW 3 PCE
Date
Figure 11-2. Monitoring well water data for PCE versus time.
Note: the samples within the green bars are below the method detection limit (MDL)
and are reported only at the MDL.
In addition groundwater samples were collected from soil gas points when they were temporarily flooded
using a peristaltic pump. These results are shown in Figure 11-3. Again chloroform was not detected in
any of these samples.
11-6
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Flooded SoilgasPort Water Sample
PCE Data
2.5
O
1"
4-J
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
of the duplex—MW-1C. MW-2C, north of 422 East 28th St, which in February 2011 had one of the
highest concentrations, had all nondetects in several fall rounds, with the exception of one detection in
November.
Our confidence in this groundwater data requiring professional judgment at the lower end of the
instrumental range is greatly strengthened because a Henry's law conversion shows good agreement
between these PCE concentrations and the PCE concentrations observed in gas phase samples from
nearby deep soil gas ports. This suggests that the low groundwater concentrations are at quasi-equilibrium
with deep soil gas concentrations.
11.2.2 Is the Groundwater Concentration Trend Correlated to Potentiometric Surface?
Because chloroform was always below the groundwater detection limit at the EPA laboratory,
conclusions cannot be drawn directly, although in other sections, we have argued that the chloroform
trend in deep soil gas is temporally correlated to the potentiometric surface.
Figure 11-1 shows some broad similarities to Figure 11-2. The fraction of monitored intervals in which
PCE is detectable according to professional judgment is highest in the winter and spring months of 2011
when the highest rainfall and streamflow were noted. However, there is not enough complexity and
resolution in these trends to judge whether this is real or merely coincidental.
11.2.3 Is the Groundwater Concentration Trend Correlated to Indoor Air
Concentrations?
Because all samples are nondetects, no conclusions can be drawn for chloroform in groundwater. For
PCE, there is a rough correspondence between Figure 11-2 showing groundwater concentrations and the
trend of PCE in indoor air discussed in Section 5. PCE concentrations in indoor air are highest in the
winter of 2011. They reach a low in July 2011 and only modestly recover in the winter of 2012. Similarly
PCE was detectable according to professional judgment in all but the deepest intervals of groundwater
through late June 2011. According to professional judgment, the number of intervals in which PCE is
detected markedly declines in the groundwater samples taken after July 2011.
11.3 Soil Moisture Trends
11.3.1 Correlation with Rainfall Measurements
Figure 11-4 presents dated from implanted soil moisture probes (i.e., irrometers). There is a rough
correlation between Figures 11-4 and 11-1: the soil is drier in the hotter months and wetter in the cooler
months. In Figure 11-4, all sensors read more saturated soil conditions until late May 2011, when they
began to dry out. This soil moisture trend corresponds reasonably well with the gradual tapering off of the
period of high rainfall in mid-April (see Figure 11-1) and with the observations that many soil gas
sampling points would not pump or showed visible water in March 2011 (Tables 11-1 and 11-2 in Section
11.1). The weather began to get wetter again in October, which corresponds with the increase in Fall
Creek discharge rates (Figure 11-1). These year-long trends in rainfall and water level also correspond
with the period when the 16.5-ft depths became usable again and when they finally stopped pumping.
In Figure 11-4, the sensor at the 6-ft depth beneath the house shows that it was under more saturated
conditions at the beginning of the project, but dried out toward May and through the summer. However,
this sensor continued a slow, steady progression toward drier levels. Note that the sensors themselves
were conditioned by soaking before installation; although that procedure would only have a short term
effect in most soils, it is possible that it led to a longer bias in tighter soils (which are generally found at
shallower depths at this site, see Section 3.1.1). The 13-ftand 16.5-ft depths beneath the building stay
11-8
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
more saturated throughout the project. This is reasonable for beneath the house despite any possible
moisture shadow given the depth to groundwater (assuming also the existence of a capillary fringe).
The outside moisture sensors for the 3.5-ft and 6-ft depths agree with the seasonal rainfall changes (see
Figures 11-1 and 11-4). The moisture readings at the 13-ft outside depth do not correlate with the other
datasets. This probe began the project with readings toward the more saturated end of the scale and then
dried out in May and into the summer. However, it stayed very dry in fall 2011 and winter 2012, even
when other datasets suggest that soil moisture conditions should have gotten wetter. The outdoor sensors
for the 9-ft and 16.5-ft intervals (not shown) yielded readings that, according to the manufacturer, could
mean the sensors are recording extremely dry conditions, they have a broken connection, or the soil dried
out and shrunk away from the sensor and never reconnected. For the 13-ft depth outdoor sensor, based on
the pattern it followed before May 2011, it seems likely that the soil dried away from the sensor and never
regained connection. For the 9-ft and 16.5-ft sensors, it seems more likely that the sensors have a broken
connection. Given that these sensors are designed to be permanently implanted, they cannot easily be
removed for servicing.
11.3.3 Relationship to Observed Stratigraphy
Figures 11-5 and 11-6 show a generalized cross section of the soil types present in the immediate area of
the 422/420 house. The layer beneath the topsoil consists mostly of silt and clay. This layer would be
expected to retain the high amounts of rainfall during the wetter periods, which corresponds to the
shallower depths registering moister on Figure 11-4 for the beginning of the project, and it could also
correspond to the difficulty pumping shallower soil gas ports during this time. The flooding of the deeper
soil gas ports within the sand and sand and gravel layers and the moister readings for the deeper moisture
sensors could be in response to the changing water levels at Fall Creek to the south of the 422/420 house,
because there appears to be a rough correlation between the rise and fall of Fall Creek and the moistening
of some of the sensors (see Figures 11-1 and 11-4).
11-9
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Soil Moisture
dryer
» Outside3.5' bis
outside6r bis
• OutsidelS'bls
• beneath building6r bis
beneath buildinglS' bis
• beneath building 16.5ft bis
more saturated
Figure 11-4. Soil moisture: irrometer moisture data in centibars
for the interior and exterior of the 422/420 house.
Note that lower readings are more saturated.
Figure 11-5. Lithological fence diagram showing some of the major
soil types beneath the 422/420 house.
The view is toward the north from the street in front of the house. The empty white area at the top of the soil figure
represents the house basement. This figure shows that in the immediate vicinity of the house, silt and clay are
present until 7.5-8 ft. After that, sand and gravel alternate with layers of sand.
11-10
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
|SOP3]
„ Depth (ft bgs)
0.0
Lithology Fences
Figure 11-6. Lithological fence diagram showing some of the major
soil types beneath the 422/420 house.
The view is toward the south from the backyard of the house. The empty white area at the top of the soil figure
represents the house basement. This figure shows that in the immediate vicinity of the house, silt and clay are
present until 7.5-8 ft. After that, sand and gravel alternate with layers of sand.
11.4 Correlation of Groundwater Concentration Changes to Deep Soil Gas
MWl and 2 are both a series of three clustered wells, labeled A, B, and C. Each well in the cluster
extends to the following screen depths: 24 ft to 26 ft, 21 ft to 24 ft, and 16 ft to 21 ft, respectively (see
Tablell-3). Because sampling occurred within these intervals, comparing the PCE behavior of the
deepest nearby soil gas ports (the 13-ft and 16.5-ft depths) might provide insight into the linkage between
PCE in the groundwater and the soil gas.
The nearest soil gas ports to the water monitoring wells are
• SGP1 and SGP2 by the MW-1 cluster,
• SGP5 and SGP6 by the MW-2 cluster, and
• SGP8 and SGP9 near MW-3.
Data from each of those SGPs are represented in Figures 11-7 through 11-11.
When soil gas data were available for the deepest soil gas ports (at 13 and 16.5 ft), these SGPs show a
similar year-long pattern to what the groundwater data show (Figures 11-2 and 11-3): the elevated PCE
concentrations in the cooler weather at the beginning of the project gradually decreased toward the
warmer summer months and then rose again as the weather began to cool off. The 16.5-ft depths also
show elevated PCE concentrations at a comparable time to the elevated PCE concentrations in the
groundwater data. Deep soil gas intervals plotted in Figures 11-7 through 11-9 also show a similar
pattern over a short time period for the September/November period: a short rise in September and then
falling off in PCE concentrations after lower summer concentrations (compare with Figure 11-2).
11-11
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Concentration pig/m3
h-i
Hi 0
Hi 0 0
l-i O O O
OI-iOOOO
SGP1 PCE Concentrations
O
x^^fcfr^
• • X
x
J N> h-» "-^ N> Ol h-» "-^ I—1
J "--- O 1— * CT1 "--- -Pi N) "---
- o ^ 5J Date ^ £ ^ ii G
3 O I— * I— * I—1 I—1
O I-1
SGP1-3.5 PCE SGP1-6 PCE A SGP1-9 PCE X SGP1-13 PCE QSGP1-16.5 PCE
Figure 11-7. PCE concentrations at each of the SGP1 ports overtime.
SGP2 PCE Concentrations
1 nnnn
1 nnn
£
i 100
c
0
-t-»
2
-(— '
£ 10
u
c
0
u
01
AA
A
A4A A
TA AAA AA^A^AAA AX jx x>^<>^?
O h^
»SGP2-3.5PCE SGP2-6 PCE ASGP2-9PCE XSGP2-13PCE OSGP2-16.5 PCE
Figure 11-8. PCE concentrations at each of the SGP2 ports vs. time.
11-12
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
Concentration |-ig/m3
i-i
1 — * O
r-, l-i O O
P l-i O O O
l-i l-i O O O O
DT/9T/TT
SGP5 PCE Concentrations
o
X
:-M ^^^^^
ft A* **>*AA.AA* ftftKA KAAAAAA« ** A A 4fc 4* *»*•«»»«*** A ***** *
4 *
? M M M "Date M M M ^ K M
5 1-1
*SGP5-3.5PCE SGP5-6PCE ASGP5-9PCE XSGP5-13PCE OSGP5-16.5 PCE
Figure 11-9. PCE concentrations at each of the SGP5 ports overtime.
10000
1000
m
j:
i 100
c
O
4-1
4-J
V 10
u
c
0
U
1
0.1
1-
J
1-
C
\-
C
SGP6 PCE Concentrations
0
A {p°oOCPxXX XxxXv
0 0 ° • xxXOO><><»A A
A V ^A *A A
A » » A* A
A A AA ^ A^ ^ A^ A A*AAAA A ^A A A
4fA A
• A **» V ^••^B • •»«»*•**•!««»<»«*•»* • •••• •
•
^,H l H l H l H l H l H l H l ^^, h-1 N)
_1 |_1|_1 |_l |_l |_l |_l |SJ
D I-1
SGP6-3.5 PCE • SGP6-6 PCE A SGP6-9 PCE X SGP6-13 PCE O SGP6-16.5 PCE
Figure 11-10. PCE concentrations at each of the SGP6 ports overtime.
11-13
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
SGP8 and SGP9 PCE Concentrations
10000
1000
^
W)
c
_o
4-J
O3
4-1
c
OJ
u
o
U
100
10
0.1
01 -J
CD K>
^ Date 51
SGP8-6PCE SGP8-9 PCE ASGP8-13PCE XSGP8-16.5 PCE XSGP9-6PCE •SGPS-S PCE OSGP9-13 PCE i SGP9-16.5 PCE
Figure 11-11. PCE concentrations at SGP8 and SGP9 ports overtime.
11.5 Revisions to Conceptual Site Model
In Sections 2 and 2.1, we review the generally accepted, general model of vapor intrusion processes. In
Section 3.1.6, we present an initial conceptual site model for this specific duplex on this specific site. A
conceptual site model is always subject to revisions and refinement based on new data. In this section, we
present revisions to the site-specific conceptual site model.
First, it should be noted that this study situation is unusual in that we have gathered an extremely detailed
dataset about one particular duplex known to be in proximity to a group of known, but poorly defined,
potential sources:
• 10 historic drycleaners within 1A mile, believed to be upgradient
• an immediately adjacent building with a complex history of multiple commercial/industrial uses
• aged combined sewers in a community with known chlorinated VOC disinfection by-products in
drinking water.
However, none of these potential sources has been thoroughly investigated or delineated—a situation that
is likely common in urban areas (Dickson et al., 2010). In contrast, vapor intrusion practitioners more
typically work on sites where there is a known anthropogenic source term and some delineation of a
groundwater plume on a scale of hundreds of feet/blocks. But practitioners are frequently dealing with
individual residences about which very little is known, typically only the results of a brief
survey/homeowner interview and a very small number of measurements in indoor or subslab air.
11-14
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
The following revisions to our conceptual site model of this duplex and the vapor intrusion sources
supplying it can be made:
• The groundwater level beneath the house is subject to rapid swings of up to 5 ft over the course of
a few days during seasonal flooding in the adjacent Fall Creek and potentially to the influence of
combined sewers.
• The stack effect caused by indoor/outdoor temperature differentials operates not only during the
heating season, but also during the summer, because of the "solar stack effect" and the storage of
heat in the building during cool late summer/fall nights. Differential pressure measurements
indicate that changes in building differential pressure are reflected in a measureable advective
driving force between the 13-ft depth near the water table and the 6-ft depth directly beneath the
basement. Therefore, in this case, advection may be the primary cause of VOC migration through
the deeper portions of the vadose zone.
• The heterogeneity of the subslab concentrations beneath the duplex suggests the absence of an
engineered gravel layer. Therefore, the subslab region of the house does not behave here as a
well-mixed plenum.
• PCE is apparently widely spatially distributed in site groundwater at concentrations well below
the current 5 (ig/L MCL (U.S. EPA, 2012a). These shallow groundwater concentrations
apparently control deep soil gas concentrations. Only a moderate degree of attenuation occurs in
those deep soil concentrations as they are drawn toward the basement of the structure. Substantial
attenuation occurs in the upper 6 ft of the site external soil gas, which is finer grained materials
than the sandy deeper materials. It is currently unclear whether this is due to gas permeability
contrasts, sorption processes, or most likely barometric pumping dilution. Substantial attenuation
also occurs across the building envelope between subslab and indoor air.
• Chloroform is present in highest concentration in deep soil gas. Substantial chloroform has
historically been detected in groundwater on a site 200 ft to the southwest. Chloroform was also
detected in groundwater at this house in preliminary sampling. Further studies are planned to
determine if the lack of detections in recent groundwater samples on site indicate migration
through deep soil gas from offsite sources or losses in the sampling and analysis process.
Chloroform attenuation is substantial between the area just above the water table and the 6 ft
depth below the structure. Chloroform is further attenuated between subslab air and indoor air.
• The relative importance of the potential sources of PCE and chloroform—historic drycleaners,
the adjacent commercial/industrial quadraplex, and storm sewers/drinking water disinfection—is
unclear.
• Sewer lines and laterals likely play some role in contaminant fate and transport in this system.
Elevated concentrations of PCE and chloroform are present in the headspace of sewer gas. Their
role as a direct entry pathway can be minimized through plumbing trap and vent maintenance.
Their role in lateral transport through the vadose zone and into the subslab of the duplex will be
elucidated through future geophysical studies.
• There is a seasonal component to the PCE and chloroform indoor concentrations (Figure 11-12).
The seasonal component is partially but not completely correlated to the strength of the stack
effect.
• Concentrations of benzene, hexane, and toluene in indoor air are quite similar to ambient levels
and appear to change in lockstep with ambient air, although there are some traces of benzene in
soil gas (Figure 11-12). TCE in indoor air also tracks ambient concentrations when TCE is low,
but follows a trend that very similar to PCE when TCE concentrations were high at the beginning
of the study, suggesting a contribution of subsurface sources to TCE indoor air concentrations.
11-15
-------
Section 11—Results and Discussion: Do groundwater concentrations control soil gas
concentrations at this site? And thus indoor air concentrations?
100-
1.0-
0.1-
E
CD
1.0
0.1
10.0
1.0
0.1
• 1 00
0.10
0.01
o
o
1-
1-
Trichloroethene
Radon
Chloroform
Tetrachloroethene
Trichloroethene
Benzene
Toluene
Hexane
Location
420BaseN
- 420BaseS
— 420First
— 422BaseN
— 422BaseS
— 422First
AMBIENT
Date
Figure 11-12. Temporal plot of log indoor air concentration for VOCs (ug/m3) and radon
(pCi/L) by sample location over the study period.
11-16
-------
Section 12—Results and Discussion: Special Studies
Table of Contents
12. Results and Discussion: Special Studies 12-1
12.1 Summary of Temporary vs. Permanent Subslab Sampling Study 12-1
12.2 Summary of Fan Testing 12-1
12.2.1 Fan Test Objectives 12-3
12.2.2 Fan Test Experimental Methods 12-3
12.2.3 Fan Test Results and Discussion 12-4
12.2 A Fan Test Lessons Learned 12-11
12.3 Testing Utility of Consumer Grade Radon Device (Safety Siren Pro) 12-11
12.3.1 Consumer-Grade Radon Device Test Objectives 12-12
12.3.2 Consumer-Grade Radon Device Test Methods 12-12
12.3.3 Consumer Grade Radon Detector Test Results and Discussion 12-12
List of Figures
12-1. Fan test matrix 12-4
12-2. Fan position in stairwell (note plastic sheet over doorway) 12-4
12-3. Differential pressure before, during, and after fan tests (fan tests denoted by vertical
bars) 12-5
12-4. Radon on second floor before, during and after fan tests (fan tests denoted by vertical
bars) 12-6
12-5. VOC Field instrument data before, during and after fan tests (fan tests denoted by vertical
bars) 12-7
12-6. Note height of basement window and vent (sealed for this study) 12-8
12-7. Subslab and soil gas VOC data during fan test period 12-9
12-8. VOC data before, during, and after fan testing, Method TO-15 12-10
12-9. Comparison of fan test responses of radon, PCE, and chloroform in 422 basement 12-11
12-10. Comparison of electret and Safety Siren results 12-13
List of Tables
12-1. Quality Objectives and Criteria for Special Studies 12-2
12-2. Comparison of Safety Siren, Alphaguard, and Electret Data 12-13
12-i
-------
Section 12—Results and Discussion: Special Studies
12. Results and Discussion: Special Studies
We conducted three special studies in the same facility that go beyond the objectives articulated in
Section 2 around which the overall study was designed. These studies addressed specific research needs
identified after the main study was initiated. In this section, we present methods, results, and discussion
for a series of these studies, which can be perhaps best understood as discrete "mini-projects." These
studies examined the
• VOC measurement efficiency of temporary subslab ports as compared with permanent subslab
port constructions (Section 12.1),
• use of box fans to induce flow into the structure in an attempt to create "worst case" conditions
for vapor intrusion (Section 12.2), and
• use of a consumer grade radon detector as an indicator of vapor intrusion (Section 12.3).
Table 12-1 provides data quality objectives and criteria for these studies and parallels Table 2-3.
12.1 Summary of Temporary vs. Permanent Subslab Sampling Study
Please refer to the report in Appendix B for a full discussion of methods and results of the study
conducted before the main study that compared nearly collocated temporary and permanent subslab ports.
The temporary ports were sealed during the main study.
Under the conditions studied here, the VOC and radon concentrations measured simultaneously in soil
gas using nearly collocated temporary and permanent ports appeared to be independent of the type of
port. The variability between nearly collocated temporary and permanent ports was much less than the
spatial variability between different locations within the same residential duplex.
12.2 Summary of Fan Testing
Vapor intrusion guidance in many jurisdictions has suggested that worst-case conditions occur under
specific meteorological conditions that provide maximum driving forces (i.e., a maximum differential
pressure driving flow into the building). Although many guidance documents (e.g., Department of
Defense guidance, state guidance from New Jersey and Massachusetts) suggest that these conditions
occur in winter, not all datasets show the same seasonality. Also, it is logistically difficult to schedule
extractive sampling events to occur during specific meteorological conditions. Some documents (e.g.,
handbooks by the Electric Power Research Institute and the U.S. Navy) have suggested the possibility of
using fans to quickly depressurize a building and enhance vapor intrusion by creating a pressure gradient
from the subsurface into a building, as was done in a single test using a window fan at a Hill Air Force
Base residence in March 2006 (GSI Environmental, 2008). In this case, TCE increased by two times in
indoor air, and although radon did not, an increase in both the TCE and radon attenuation factors was
observed. A similar method could provide a quick, inexpensive, and easy-to-implement method to
determine the potential for vapor intrusion into a building.
Conversely, it has also been suggested that fan-induced over-pressurization of a building could help
distinguish indoor volatile compound sources from vapor intrusion. Tests were conducted in 2008 in
which positive and negative pressurization fans were each applied to three matched townhouse units at
Moffett Field in California (Mosley et al., 2008; Lutes et al., 2009). Results were not always in the
predicted direction. In another study, a detailed mathematical approach to quantify the relative proportions
of a contaminant attributable to subslab and indoor sources, based on measurements of radon and VOCs
under two baseline and fan perturbed conditions, was developed and tested in a former industrial building
in Indianapolis (Mosley et al., 2010).
12-1
-------
Section 12—Results and Discussion: Special Studies
Table 12-1. Quality Objectives and Criteria for Special Studies
Study Question:
Qualitatively Stated
(from SOW
objectives when
applicable)
Study Questions:
Quantitatively/Statistically
Stated
Measurement Used To
Support Study Question
Measurement Performance or Acceptance Criteria (for this
question)
Temporary vs. Permanent Subslab Sampling Study
Identify functional
differences between
permanent and
temporary subslab
probes.
• Is there a statistically
significant difference
between analyte
concentrations in gas
samples collected from
permanent and temporary
subslab probes?
• Is there evidence of a
significant amount of
leakage of indoor air into
either type of probe during
sample collection?
• Radon and VOC
measurements in subslab
soil gas samples
• Tracer gas (helium)
measurements in subslab
soil gas samples
• Agreement of subslab concentrations within +/-30% was defined
as adequate given the variable nature of subslab soil gas
distribution. Helium concentrations indicative of significant leakage
are addressed in the QAPP. For each comparison 5 pairs of
measurements are available.
Fan Testing
Evaluate the
effectiveness of fan
negatively pressurizing
the structure as a quick,
inexpensive method to
determine the potential
for vapor intrusion into a
building.
Can the use of a simple
fan system create worst
case vapor intrusion
conditions at a time
when they otherwise
would not be occurring?
• Do concentrations of VOCs
and radon and indoor air rise
significantly in response to
installation of a fan
negatively pressurizing the
structure?
• Does the magnitude of the
fan effect exceed the
difference between the 25th
and 75th percentiles of the
distribution of concentrations
measured over the year?
• Radon and VOC
concentrations, both
instrumental and extractive
measurements during a
time period of at least 6
weeks before, during, and
after the fan tests.
• Differential pressure during
the same time period.
• Radon and VOC
concentrations, both
instrumental and extractive
measurements.
• This dataset can be analyzed as an interrupted time series with the
intervention (the fan operation) applied and removed at various
known times. Clearly, however, the intervention was not the only
control on the indoor concentration, because other independent
variables, such as weather phenomena, were measured but not
controlled. The intervention was expected to be abrupt and
permanent (a step function). Can use time series analysis to
evaluate the size of the effect attributable to the fan and whether it
is statistically significant.
(http://www.oreqoneval.orq/ANALYSIS%20OF%20INTERRUPTED
%20TIME%20SERIES%20FINAL.pdf
http://wps.ablonqman.com/wps/media/obiects/2829/2897573/ch18.
fidf)
• The number of data points depends on the duration of the
intervention. We expect, however, to perform at least some fan
tests with a 48-hour or greater duration, which will yield
approximately 20 field GC VOC samples at each of four indoor
locations and 45 or more Alphaguard radon measurements at each
of two indoor locations during the application of the intervention.
Differential pressure data are being acquired every 15 minutes,
which should provide approximately 190 measurements on each of
five channels during the intervention (fan operation). Longer
periods of data with the intervention withdrawn are also expected
before and after the period of fan testing, although not necessarily
between individual fan tests.
• We achieved at least 50 week-long extractive measurements
obtained during time periods when the fan was not used that can
be used to define an annual distribution for both VOC and radon to
which the magnitude of the fan effect can be compared.
• A second distribution can be defined using what is expected to be
at least 400 measurements each for the online Alphaguard radon
and online GC. This dataset, although larger, is expected to be less
representative of the full year's seasonal variance. However, it is
expected to allow evaluation of the diurnal variance in at least two
different seasons.
Consumer Grade Radon Device Testing
Evaluate the ability of a
widely available low
cost consumer grade
radon detector based
on an ionization
chamber to provide a
continuous indication of
soil gas entry into the
structure. (Safety Siren
Pro Series 3
manufactured by
Family Safety Products
Inc.).
• Does the measurement of
radon concentration using
this consumer-based
analyzer agree within +/-
30% to the readings from
the electret and Alphaguard
methods >90% of the time?
• Does the consumer-grade
radon detector provide a
useful indication of the
weekly average infiltration of
VOC containing soil gas?
• The Safety Siren has two
displays — "short term"
(average over previous 7
days) and "long term"
(average from time of last
reset, up to 5 years).
Readings available after a
minimum of 48 hours of
operation. Recorded short-
term reading at each of six
indoor stations weekly.
• Comparison with ongoing
electret and Alphaguard
measurements.
• Month-long correlation test
between consumer-grade
radon detector and other
radon detectors.
• Year-long dataset on
radon/VOC correlation in
this house.
• Six stations were sampled: 422 basement S, 422 basement N
(downstairs Alphaguard), 420 basement S, 422 first floor, 422
second floor office (upstairs Alphaguard), and 420 first floor.
Impractical to use the safety siren for ambient measurement
because of temperature and power issues. Added electrets
measurement during the safety siren test period at 422 second
floor office for an additional comparison.
• At seven stations the Safety Siren can be compared to the electrets
being routinely operated with a one-week duration. Collected 8
weeks of comparative data.
• At two stations the Safety siren can be compared with the
Alphaguards taking hourly data, averaged over the week, providing
at least 8 more pairs of data points.
• Data analysis approaches to be used for comparing methods are
presented in Sections 3.4.4 and 8.3.3 of the main QAPP.
• We will judge the answer to this study question to be "yes" if the
Safety Siren is shown to adequately correlate with the Electrets
(see above) AND radon is shown to be correlated to VOCs in the
main study dataset (see objective A-2 in the main QAPP Table
3-2).
12-2
-------
Section 12—Results and Discussion: Special Studies
12.2.1 Fan Test Objectives
Previous studies of the fan testing method by others were generally designed to provide a line of evidence
to confirm that indoor air contaminants were from vapor intrusion, as opposed to an indoor source. The
fan testing described in this section (performed at the 420/422 East 28th St. site in Indianapolis) has a
somewhat different objective. Using the long-term, detailed temporal variability dataset available for this
site, we evaluated whether a simple fan-induced depressurization could mimic the worst case vapor
intrusion conditions observed at the site. Such a "fan-induced worst case," if available on demand, could
allow the number of sampling rounds, and, thus, the disturbance of residents to be minimized. It could
also expedite the vapor intrusion evaluation process and, if necessary, mitigation, which many
stakeholders would find desirable. We formally stated the objective as follows:
Evaluate the effectiveness of fan negatively pressurizing the structure as a quick,
inexpensive method to determine the potential for vapor intrusion into a building.
Specifically, we seek to determine if the use of a simple fan system can create worst case
vapor intrusion conditions at a time when they otherwise would not be occurring.
12.2.2 Fan Test Experimental Methods
The fan testing study design entailed concurrently evaluating the transient response of indoor air
concentrations, subslab soil gas concentrations, subslab pressure gradients, and air-exchange rates arising
from a change in fan setting (i.e., air-flow rate). The test was conducted primarily on the 422 side of the
duplex (fans will be operated on that side only). Multispeed fans were used as in previous screening tests,
but testing rapidly settled on the highest fan speeds. In principle, the target flow rate of the fan can be
established by monitoring the subslab pressure gradient. We evaluated the initial results rapidly; chose a
fan operating speed and position that yield a maximum (plateau, absolute) value of under-pressurization;
and used these results to begin monitoring subslab soil gas, air-exchange rate, and indoor air quality with
laboratory methods. Laboratory VOC methods included 24-hour passive SKC Ultra samplers and 24-hour
Summa canisters. An additional indoor sampling location on the second floor of the 422 side was
established and used just for this test.
Our plan was that the fan would be operated at a high enough speed to ensure that the pressure differential
across the slab is always negative, but, if possible, not so high of a speed that it induces outside air to flow
down around the outside of the foundation into the subslab area. We monitored the subslab vapor
concentration to ensure that outside air flow did not predominate. If concentrations had decreased over
time to near zero values, we would have interpreted this as outside air flowing in around the foundation,
but this did not occur. If concentrations in the subslab remained steady, we then assumed that the system
was at quasi-equilibrium. If concentrations in the subslab area increased with time, we would have
assumed that, prior to the test, they had been diluted from outflow from the house and were now being
increased by capture of additional soil gas. After a sufficient period of fan operation (intended to
maximize the flux of subsurface vapors into the building), indoor air quality was tested with off-site
methods to determine VOC concentrations.
We initially tested the use of box fans in second floor windows (Test A shown in Figure 12-1); however,
we found that this did not create a strong depressurization effect. We next tested a fan position at the head
of the stairs leading from the basement to the first floor (Tests B through I). Two commercial box fans
were collocated with a measured total flow of 1,224 cfrn. This position provided a stronger effect,
especially when coupled with a plastic curtain to limit the localized flow of air back down the stairs
(Figure 12-1, Tests D through I, see also photo in Figure 12-2). Under these conditions, whether the
second story windows were open or closed had little apparent effect, perhaps because the envelope of this
pre-1920 home is not tight (comparing Test G to other tests in the D through I series).
12-3
-------
Section 12—Results and Discussion: Special Studies
Column!
A.C.s on
A.C.S off
Birdcage fans in windows
Box fans in windows
Basement door open
2nd floor windows open
2nd floor windows closed
Box fans at basement
doorway
Plastic sheet over
basement doorway
Birdcage fans In
basement vents
Cooler temperatures
Basement HVAC vent
closed
Time
I Test A TestB I Test C Test D Test E I Test F TestG I Test H Test!
X
X
X
-9hrs -20min -S.Shrs ~15hrs ~11hrs ~36hrs -92hrs -48hrs -Sdays
Figure 12-1. Fan test matrix.
Figure 12-2. Fan position in stairwell (note plastic sheet over doorway).
Based on the main study information and the initial fan test data, the test matrix (Figure 12-1) was
iteratively refined. Test durations were lengthened as we gathered more confidence in the iterative design.
12.2.3 Fan Test Results and Discussion
As shown in Figures 12-3 and 12-4, the differential pressure and second-floor indoor radon concentration
changed very rapidly (in less than 1 hour) after initiating fan testing. Two collocated box fans virtually
instantaneously gave us a 1 to 3 Pascals deflection in the expected directions, such that a driving force
12-4
-------
Section 12—Results and Discussion: Special Studies
was created between deep and shallow soil gas and between subslab and basement air. Both the pressure
and radon changes gave the appearance of a step function, terminating rapidly after the fans were shut off.
This differential pressure change was in the range of typical seasonal variation for this and other houses
(U.S. EPA, 1993a) but was not as large as the worst case observed at this site (compare results for the full
year presented in Section 10.2). Note in Figure 12-3, for example, that large variations in subslab versus
indoor differential pressure occurred during a 2-week period in late September when no fan tests were
performed. However, the radon values attained during fan testing (e.g., the second-floor levels shown on
Figure 12-4) were similar to the highest radon concentrations observed during any season in this
structure. The basement radon concentrations (not shown) also were elevated to some extent during the
fan tests but did not show as clear of a response to the fan on and off cycles as those on the second floor.
This likely reflects that although operating the fans at the head of the basement stairs enhances the flow of
soil gas containing radon into the basement, it also is discharging radon-laden air from the basement into
the upstairs of the house.
Setra Differential Pressure Sensors 422 Side
-6
Date
Figure 12-3. Differential pressure before, during, and after fan tests
(fan tests denoted by vertical bars).
12-5
-------
Section 12—Results and Discussion: Special Studies
Upstairs (2nd floor) Alphaguard
-E i'
-6 Z s*
-Den
-Dot
-Con
-Cot
-r sn
-FolT
-Gen
-- ;r
-_--
-Hoi
-I0n
-lOff
Figure 12-4. Radon on second floor before, during and after fan tests
(fan tests denoted by vertical bars).
We expect that the flow into the basement will partially comprise soil gas and substantially comprise
exterior air entering the basement through the exposed portion of the basement walls (Figure 12-5)
because
• the basement in this duplex is only partially below ground (Figure 12-6),
• substantial pressure differentials have been measured between the basement and exterior that at
some times indicate flow into the basement,
• only small pressure differentials exist between the upstairs and basement,
• tracer studies indicate that flow across the first floor/basement ceiling is generally upward
(Section 10.1), and
• the basement concentrations of radon are much lower than the radon concentrations in the
subslab.
12-6
-------
Section 12—Results and Discussion: Special Studies
PCE Field GC Indoor Air Data During Fan Testing
='!:" goTt
- 3— -^^~SQTI
— left ifflff
• =;:;-;•:-:: • ;;::'
Oaff
Golf
Chtorofem. F«d GCIndoof Air Dtta During AnTcsting
Figure 12-5. VOC Field instrument data before, during, and after fan tests
(fan tests denoted by vertical bars).
12-7
-------
Section 12—Results and Discussion: Special Studies
Figure 12-6. Note height of basement window and vent (sealed for this study).
In contrast, although the VOC data during the fan test (e.g., Figure 12-5) suggest that measureable
increases in indoor concentrations can be induced with fan testing, this effect was less rapid and
predictable. Subslab and soil gas VOC concentrations (Figure 12-7) did not generally markedly change
during these short-term fan tests. Exceptions to this trend occurred at WP-3 for PCE and SGP2-9 as well
as WP-3 for chloroform, each channel of which does show more than one discontinuity corresponding
with the start or end of a fan test. Analysis of data, both from the on-site GC (Figure 10-5, Section 5.3)
and time-integrated samples (collected with Method TO-15, Figure 12-8), suggests that although the fan
operation increased VOC concentrations moderately over baseline, it did not increase concentrations to
the highest levels observed under natural conditions. For example, the VOC levels in December and
January were considerable higher than those observed during the fan testing (compare results in this
section to Sections 5.1.1 and 5.3.1). Figure 12-9 shows that the radon, PCE, and chloroform
concentrations did not always move together during the fan testing.
The radon, VOC, and differential pressure datasets suggest that although the fans were operated to
directly draw air out of the basement space, their effect may have been greater in increasing the flow of
basement air upstairs, rather than increasing vapor intrusion and the basement indoor air concentrations.
12-8
-------
Section 12—Results and Discussion: Special Studies
PCE field GC Subsurface Air Data During Fan Testing Period
CHCL3 Field GC Subsurf aceAir Data
= Itf-
SoC
-Aft
Fa-
Figure 12-7. Subslab and soil gas VOC data during fan test period.
12-9
-------
Section 12—Results and Discussion: Special Studies
Chloroform Fan Testing Comparision(TO-15)
I !_• -I~11'.
Date
PCE Fan Testing Comparision:TO-15 Data
Date
Figure 12-8. VOC data before, during, and after fan testing, Method TO-15.
12-10
-------
Section 12—Results and Discussion: Special Studies
Comparison of Summer Basement Rn (Alphaguard) and VOCs (GC)
2.5
-Aon
-Aoff
-Bon
-6/Coff
-Don
-Doff
-Eon
-E=ff
-Fon
-Foff
Grr
Goff
Han
Hoff
=1P
-loff
B/10/2011
B/3Q'2Oli
3/1S/2011 9/29/2011 lQ'9/2011 10/1&2O11
Date
CHC3
42 2 EBierrert FCE
Figure 12-9. Comparison of fan test responses of radon, PCE, and
chloroform in 422 basement.
12.2.4 Fan Test Lessons Learned
Achieving depressurization of a building is very sensitive to fan placement within the structure.
Temporary plastic barriers can be effectively used to control flow pathways. Differential pressure
monitoring at several locations during fan testing is valuable because results may not be fully predictable
a priori and probably depend on the specific air sealing of the house envelope. Short-term (several days)
induction of a differential pressure equivalent to worst case natural conditions may not provide worst case
indoor air concentration for VOCs if there is significant seasonal variability in soil gas concentrations. It
can, however, provide some confirmation that vapor intrusion is a significant pathway by increasing
indoor concentrations. Results suggest that, in the house tested, the effects of fans on both pressure and
concentration may not be as powerful as the natural forces that influence the variability of the intrusion of
subsurface VOCs into indoor air and that VOC concentrations are not as greatly affected by fans as radon.
12.3 Testing Utility of Consumer-Grade Radon Device (Safety Siren Pro)
Schuver and Siegel (2011) have:
• highlighted the role of radon as a potential "general tracer of soil-gas entry";
• pointed out that there are multiple benefits from minimizing soil gas entry (including reductions
of problems attributable to moisture/mold, radon, and methane as well as reduction in VOCs);
and
12-11
-------
Section 12—Results and Discussion: Special Studies
• advocated the active involvement of homeowners in observing the building-specific aspects of
vapor intrusion at VOC sites, both as an educational tool (to help homeowners understand
temporal and spatial variability) and a way to an efficient solution to which all stakeholders agree.
Thus, as an additional task in this project, we evaluated the ability of a widely available low-cost ($129)
consumer-grade radon detector based on an ionization chamber to provide a continuous indication of soil
gas entry into the structure (Safety Siren Pro Series 3 manufactured by Family Safety Products Inc.).
12.3.1 Consumer-Grade Radon Device Test Objectives
The objective was stated as "evaluate the ability of a widely available low-cost ($129) consumer grade
radon detector based on an ionization chamber to provide a continuous indication of soil gas entry into the
structure" (Safety Siren Pro Series 31 manufactured by Family Safety Products Inc.).
12.3.2 Consumer-Grade Radon Device Test Methods
The Safety Siren Pro Series 3 is a consumer-grade radon detection instrument that provides continuous
real-time measurement based on an ionization chamber and requires little operator labor. An operating
manual for this instrument is provided as Appendix B of the QAPP, Addendum 2. In this test, we sought
to compare the performance of the safety siren to a well-accepted method (electrets). Secondarily, we
were able to compare the Safety Siren with the online Alphaguard data and the charcoal sampling with
off-site analysis.
The following six stations were used for testing: the 422 basement south, the 422 basement north
(downstairs Alphaguard), the 420 basement south, the 422 first floor, the 422 second floor office (upstairs
Alphaguard), and the 420 first floor. The detector may be placed face up on a tabletop, countertop, or any
flat surface where the ventilation slots will not be blocked. The detector must be kept dust free. A proper
airflow must be maintained through the detector to obtain an air sampling representative of the local
environment. It is impractical to use the Safety Siren for ambient measurement because of temperature
and power issues. The manual restricts the operating environment to 0°C (32°F) to 40°C (104°F). We
added an electret measurement location during the Safety Siren test period in the 422 second floor office
for an additional comparison.
The Safety Siren has two displays—the "short term" is an average over the previous 7 days, and the "long
term" is the average from time of last reset (up to 5 years). The numeric LED display shows the level of
radon gas in pico Curies per liter (pCi/L). The display range is 0.0 to 999.9. Readings are available after a
minimum of 48 hours of operation. We manually recorded the short-term reading at each of six indoor
stations weekly. Data were assembled in spreadsheet form for comparison to electret and Alphaguard
results. The audible alarm was muted.
Every 24 hours, the detector does a self-test. If there is a failure in this self-test, an error message will
appear in the display window.
12.3.3 Consumer-Grade Radon Detector Test Results and Discussion
As shown in Figure 12-10 and Table 12-2, the Safety Siren consumer-grade detector shows reasonably
good agreement with an accepted professional method (electrets) over a range (1 to 5 pCi/L) useful for
determining compliance with EPA's recommend radon action level (4 pCi/L). Above 5 pCi/L, the Safety
Siren tended to dramatically overestimate the radon concentration. Thus, this device would provide an
indication of soil gas entry at low concentrations useful for radon management. In the higher range, the
Safety Siren might overestimate the risk. Thus, the Safety Siren would be useful in showing a homeowner
1 http://www.radonzone.com/radon-detector.html
12-12
-------
Section 12—Results and Discussion: Special Studies
when radon was being effectively excluded, but it might create a somewhat exaggerated impression of
radon vapor intrusion variability if high concentration peaks occurred. Given the range of concentrations
of this house and the accuracy range of the Safety Siren, it will be interesting to observe how the meters
respond when a mitigation system is installed and begins operation in the test house as is planned in a
later study.
Radon Comparision- Consumer Grade Detector to Electrets
12.00
4.00
Radon
-1:1 Line
0.00
4.00 8.00 12.00 16.00
Siren consumer grade Radon detector one week duration (pCi/l)
20.00
Figure 12-10. Comparison of electret and Safety Siren results.
Table 12-2. Comparison of Safety Siren, Alphaguard, and Electret Data
Location
Time
Date
Safety Siren
(pCi/L)
Alphaguard
(pCi/L)
Electrets (pCi/L)
Electret
Duplicates
(pCi/L)
1st Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
15:55
16:03
16:10
16:08
16:13
16:13
1/4/2012
1/4/2012
1/4/2012
1/4/2012
1/4/2012
1/4/2012
6.6
5.4
14.4
14.6
1.4
3.7
~5
-10
4.92
4.86
10.22
9.57
1.09
2.72
10.35
(continued)
12-13
-------
Section 12—Results and Discussion: Special Studies
Table 22-2. Comparison of Safety Siren, Alphaguard, and Electret Data (continued)
Location
Time
Date
Safety Siren
(pCi/L)
Alphaguard
(pCi/L)
Electrets
(pCi/L)
Electret
Duplicates
(pCi/L)
2nd Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
13:59
14:12
14:18
14:19
14:21
14:25
1/11/2012
1/11/2012
1/11/2012
1/11/2012
1/11/2012
1/11/2012
5.7
5.8
12.6
18.6
1.6
3.7
4.69
8.78
4.56
4.37
9.05
8.70
1.18
3.50
9.11
3rd Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
11:25
11:26
11:27
11:28
11:40
11:42
1/18/2012
1/18/2012
1/18/2012
1/18/2012
1/18/2012
1/18/2012
6.9
6.4
13.7
18.8
1.9
3.0
5.09
9.73
4.88
4.46
9.34
8.89
0.98
2.84
9.73
4th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
15:17
15:18
15:20
15:21
15:25
15:26
1/25/2012
1/25/2012
1/25/2012
1/25/2012
1/25/2012
1/25/2012
5.7
5.9
12.2
18.8
1.9
3.8
4.79
8.52
4.74
3.81
7.83
8.12
1.74
3.60
7.98
5th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
14:41
14:40
14:39
14:39
14:38
14:36
2/1/2012
2/1/2012
2/1/2012
2/1/2012
2/1/2012
2/1/2012
5.7
5.5
12.6
18.9
1.0
1.8
4.46
7.71
4.15
3.42
8.24
7.26
0.25
1.27
8.03
6th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
14:03
14:04
14:15
14:15
14:10
14:11
2/8/2012
2/8/2012
2/8/2012
2/8/2012
2/8/2012
2/8/2012
5.2
5.3
13.3
18.9
2.3
5.4
4.78
8.68
4.58
4.48
8.60
9.56
1.09
2.40
8.62
(continued)
12-14
-------
Section 12—Results and Discussion: Special Studies
Table 32-2. Comparison of Safety Siren, Alphaguard, and Electret Data (continued)
Location
Time
Date
Safety Siren
(pCi/L)
Alphaguard
(pCi/L)
Electrets (pCi/L)
Electret
Duplicates
(pCi/L)
7th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
12:19
12:20
12:23
12:25
12:28
12:30
2/15/2012
2/15/2012
2/15/2012
2/15/2012
2/15/2012
2/15/2012
5.6
6.0
13.3
19.1
1.4
3.0
4.80
8.44
4.41
4.15
8.28
8.34
0.36
1.94
7.47
8th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
14:28
14:29
14:30
14:31
14:26
14:25
2/22/2012
2/22/2012
2/22/2012
2/22/2012
2/22/2012
2/22/2012
4.8
5.2
12.0
18.1
1.4
3.7
4.30
7.74
3.68
3.82
6.08
6.56
0.42
2.08
5.82
9th Week Radon Comparison
422 Second floor office
422 First floor
422 Basement N
422 Basement S
420 First floor
420 Basement S
15:40
15:40
15:41
15:42
15:46
15:47
3/1/2012
3/1/2012
3/1/2012
3/1/2012
3/1/2012
3/1/2012
6.1
6.2
12.7
19.6
1.4
2.1
4.74
8.48
3.97
3.88
9.00
10.43
0.45
2.56
9.00
12-15
-------
Section 13—Conclusions and Recommendations
Table of Contents
13. Conclusions and Recommendations 13-2
13.1 Conclusions 13-2
13.1.1 Conclusions: Seasonal Variation and Influence of HVAC 13-2
13.1.2 Conclusions: The Relationship Between Subsurface and Indoor Air
Concentrations 13-2
13.1.3 Conclusions: The Relationship Between Radon and VOCs 13-2
13.1.4 Conclusions: The Use of External Soil Gas Samples as a Surrogate Sampling
Location 13-3
13.1.5 Conclusions: The Duration Over Which Passive Samplers (Solvent Extracted
Radial Style Charcoal) Provided Useful Integration of Indoor Air
Concentrations 13-3
13.1.6 Conclusions: Groundwater vs. Vadose Zone Sources as Controls on Indoor
Concentrations at This Site 13-3
13.2 Practical Implications for Practitioners 13-4
13.2.1 Sampling to Characterize Seasonal Variations 13-4
13.2.2 Using Fan Induced Depressurization in Vapor Intrusion Studies 13-5
13.2.3 Performance of Temporary Sub-slab Sampling Ports 13-5
13.2.4 Performance of Consumer Grade Radon Detector 13-6
13.3 Recommendations 13-6
13.3.1 Recommendations For Vapor Intrusion Research Generally 13-6
13.3.2 Recommendations Regarding Further Data Collection at This Test Site 13-7
13-1
-------
Section 13—Conclusions and Recommendations
13. Conclusions and Recommendations
13.1 Conclusions
The conclusions of this study represent the fruit of an intensive study of a single early 20th century duplex
in a particular geological setting—glaciofluvial deposits in Indianapolis, IN. Few other VOC vapor
intrusion studies have collected a dataset of comparable detail, and those have been conducted in
buildings of significantly different age or geological context.:
13.1.1 Seasonal Variation and Influence of HVAC
• Lower VOC concentrations were observed in indoor air in summer. These VOC concentrations in
indoor air are controlled not only by "building envelope-specific" factors, but they are also
significantly influenced by seasonal variations in subsurface concentration distributions,
especially in shallow/subslab soil gas where a weaker seasonal trend was observed.
• In indoor air, peak concentrations were seen in different months of the 2011 winter for PCE
(January) and chloroform (March) on the first floor of this duplex. Temporal trends for
chloroform and PCE differed markedly in fall 201 I/winter 2012 between the heated and unheated
sides of the duplex: the unheated side showed a much steeper decline in spring than the heated
side. Thus, complex data patterns for multiple VOCs in the same structure can be expected even
in the absence of occupant-related sources or activities.
• Stack-effect driving force calculations based on measurements of indoor/outdoor temperature
differential were predictive of indoor air concentrations. These stack effects included not only the
winter stack effect but also solar stack effects observed during summer and early fall. The cooling
effect of window air conditioners appeared to provide some protection against vapor intrusion, at
least for radon, during the summer months.
• A repeatable seasonal effect of higher concentrations during winter was seen for chloroform and
radon, but not all winters are equal. Winter 2011 and winter 2012 were very different
climactically, and peak PCE concentrations observed in January 2011 were not equaled in 2012.
Inter-year climatic variations are well known even by lay stakeholders, but their role in vapor
intrusion studies may be underappreciated.
13.1.2 The Relationship Between Subsurface and Indoor Air Concentrations
• PCE, chloroform, and radon have different spatial patterns in soil gas at this site.
• PCE and chloroform appear to have deep sources.
• Soil gas VOCs at some, but not all, high concentration sampling ports display a similar temporal
pattern to that observed in indoor air, with higher concentrations during winter months.
• Sewer lines and laterals likely play some role in contaminant fate and transport in this system.
Elevated concentrations of PCE and chloroform are present in the headspace of sewer gas. Their
role in lateral transport through the vadose zone and into the subslab of the duplex will be
elucidated through future geophysical studies.
• There is a strong seasonal component to the PCE and chloroform indoor concentrations (see
Section 11). The seasonal component appears to be correlated to the strength of the stack effect,
but it is not the only variable that controls indoor air concentrations.
Johnson, Op. Cit. also numerous case studies compiled in U.S. EPA (2012c).
13-2
-------
Section 13—Conclusions and Recommendations
13.1.3 The Relationship Between Radon and VOCs
• Long-term (weekly and greater) radon concentrations in subslab air were more stable than VOC
concentrations, presumably because the shallow soils themselves were the dominant source of
radon and VOCs originate at a greater depth/distance.
• Radon concentrations in indoor air showed approximately an order of magnitude short term (< 1
day) variation—greater short-term variation than was observed for VOCs.
• The 1-week integration time dataset for radon had less seasonal variability than VOCs in indoor
air.
• Statistical cross-correlation testing found that radon and VOCs were positively cross-correlated at
several indoor air sampling locations (5% critical level). In laymen's terms, we are quite
confident that when radon concentrations go up, VOC concentrations will also go up in indoor
air. Some cross-correlations of radon and VOCs were observed at soil gas ports, but these cross-
correlations were less consistent/strong.
• Radon provided a qualitative indication that soil gas was entering this house. Thus, radon would
have been a useful aid to VOC data interpretation if the house had been occupied and had
numerous potential indoor sources. However, long-term radon exposure would not have
completely predicted VOC exposure in this house over all time scales.
13.1.4 The Use of External Soil Gas Samples as a Surrogate Sampling Location
• High concentrations of VOCs and radon were seen in tight loams directly under building (subslab
ports and 6-ft soil gas ports) but not in external soil gas above the level of the basement floor (3.5
ft bis).
• External soil gas samples collected at 6 ft bis, the depth of the basement floor, had substantial
VOC concentration variability and would have underpredicted subslab concentrations.
• In deep soil gas (13 and 16.5 ft), there was close agreement between the mean chloroform and
radon concentrations at points underneath the building and outside of the building. In deep soil
gas, PCE concentrations appeared lower on average and more variable for the points outside of
the building than for the points beneath the building.
13.1.5 The Duration Over Which Passive Samplers (Solvent Extracted Radial Style
Charcoal) Provided Useful Integration of Indoor Air Concentrations
• Excellent agreement was observed between numerical averages of successive 7-day exposure
samples with the results of single passive samplers exposed for 14 days (almost always within
+/- 30%) for all compounds, despite dramatic temporal variability. This suggests uniform uptake
rates for these time periods.
• The PCE, benzene, hexane, and toluene passive samplers tested provide good integration over
durations from 7 to 28 days. Chloroform integration was less effective for durations greater than 2
weeks.
• The PCE and toluene passive samplers provide good integration of concentrations over durations
from 7 to 364 days.
• Temporal variability in 1-week duration indoor VOC samples over the course of a year of >20x
were observed. For certain less-volatile compounds, passive samplers allow cost-effective
acquisition of long-term average concentration data.
• Vapor pressure predicted well the relative durations over which different compounds could be
collected with the passive samplers.
13-3
-------
Section 13—Conclusions and Recommendations
13.1.6 Groundwater vs. Vadose Zone Sources as Controls on Indoor Concentrations at
this Site
• The potentiometric surface at this house responds within days to rain events.
• Chloroform concentration trends visually correlate with hydrogeological changes.
• Chloroform concentrations in soil gas peak have their highest concentrations just above the water
table.
• Chloroform is present in highest concentration in deep soil gas. Substantial chloroform has been
historically detected in groundwater on a site 200 ft to the southwest. Chloroform was also
detected in groundwater at this house in preliminary sampling. Further studies are planned to
determine if the lack of detections in recent groundwater samples on site indicate migration
through deep soil gas from off-site sources or losses in the sampling and analysis process.
Chloroform attenuation is substantial between the area just above the water table and the 6-ft-
depth below the structure. Chloroform is also substantially attenuated between subslab air and
indoor air.
• PCE is apparently widely spatially distributed in site groundwater at concentrations well below
the current 5 ng/L MCL.2 The calculated volatilization from these shallow groundwater
concentrations matches observed deep soil gas concentrations. Only a moderate degree of
attenuation occurs in those deep soil concentrations as they are drawn toward the basement of the
structure. Substantial attenuation occurs in the upper 6 ft of the site external soil gas, which is
composed of finer grained materials than the soils. Substantial attenuation also occurs across the
building envelope between subslab and indoor air.
• The relative importance of the potential sources of PCE and chloroform—historic drycleaners,
the adjacent commercial/industrial quadraplex, and storm sewers/drinking water disinfection—is
unclear.
13.2 Practical Implications for Practitioners
In this section we present specific conclusions and observations that directly address common questions
that arise in the investigation of vapor intrusion sites.
13.2.1 Sampling to Characterize Seasonal Variations
13.2.1.1 Indoor Air Sampling for Seasonal Variations
• Current guidance in NJ DEP (2012) calls for sampling in : "Heating season is from November 1
to March 31 (Winter)" NY (2006)3 also ties sampling to "heating season"
• Tying the sampling location to the heating season may not adequately represent seasonal worst
case if soil gas concentrations increase gradually during the winter or are effected by spring high
water.
• CA DTSC (2011)4 guidance formulation may be better on this point: "At a minimum, sampling
data should be obtained over two seasons; late summer/early autumn and late winter/early
spring"
http://water.epa.gov/drink/contaminants/index.cfm
3 FINAL: Guidance for Evaluating Soil Vapor Intrusion in the State of New York
October 2006, Prepared by:NEW YORK STATE DEPARTMENT OF HEALTH
Final Guidance for the Evaluation and Mitigation of Subsurface Vapor Intrusion to Indoor Air (Vapor Intrusion
Guidance). Department of Toxic Substances Control, California Environmental Protection Agency. October 2011
-------
Section 13—Conclusions and Recommendations
• Caution: high water levels at some sites may bring VOCs closer to the occupied space, but at
other sites may represent the temporary occurrence of a fresh water lens.
• Short term variability of 2-5 x was observed in indoor air (semihourly observations over a period
of one week or less). The variability in this case appears to be less severe than the three orders of
magnitude observed by Johnson et al. (2012) in another house under different building and
geological conditions.
13.2.1.2 Soil Gas Sampling for Seasonal Variation
• Short term variability (semihourly observations over a period of a week) was quite low (<2x) in
subslab and shallow soil gas ports. Basement wall ports are the exception and show more short
term variability.
• Except for the wall ports, our results suggest changes in soil gas concentrations occur gradually
over months. Therefore, 15-minute soil gas samples may be adequate, and 24-hour integration
unnecessary.
• A significant, steady rise in soil gas concentrations over the course of a winter was observed;
therefore, sampling at the beginning and end of winter may give substantially different results.
13.2.2 Using Fan Induced Depressurization in Vapor Intrusion Studies
• Achieving depressurization of a building is very sensitive to fan placement within the structure.
Temporary plastic barriers can be effectively used to control flow pathways.
• Differential pressure monitoring at several locations during fan testing is valuable since results
may not be fully predictable a priori.
• Short term (several days) induction of a differential pressure equivalent to worst case natural
conditions may not provide worst case indoor air concentration for VOCs if there is significant
seasonal variability in soil gas concentrations.
• In the house tested, the effects of fans on both pressure and concentration were not as powerful as
the natural forces that influence the variability of the intrusion of subsurface VOCs into indoor
air, and thus VOC concentrations were not as greatly affected by fans as radon concentrations.
13.2.3 Performance of Temporary Subslab Sampling Ports
• Under the conditions studied here, VOC and radon concentrations measured simultaneously in
soil gas using nearly collocated temporary and permanent ports appeared to be independent of the
type of port.
• The variability between nearly collocated temporary and permanent ports was much less than the
spatial variability between different locations within the same residential duplex.
• The agreement of concentrations was achieved even though the clay portion of the seal of the
temporary ports visibly desiccated and cracked. Post sampling leak test results suggested that this
desiccation and cracking was not as detrimental to port seal performance as would have been
expected and suggests that the Teflon tape portion of the seals was serving an important function.
• Post sampling leak tests are advisable (in addition to presampling leak tests) when temporary
ports are used to collect a time integrated sample over a period of several hours.
• These results suggest that temporary subslab sampling ports can provide data equivalent to that
collected from a permanent subslab sampling port at the same time. However, we caution that: (1)
we tested only one type of seal material in one location, (2) the seals were installed by experts
and rigorous quality control, and; thus, (3) these results may not apply to all types of temporary
seals and all building foundations.
13-5
-------
Section 13—Conclusions and Recommendations
13.2.4 Performance of Consumer Grade Radon Detector
• The Safety Siren consumer grade detector shows reasonably good agreement with an accepted
professional method (electrets) over a range of radon concentrations (1-5 pCi/L) useful for
determining compliance with EPA's recommend Radon Action Level of 4 pCi/L. Above 5 pCi/L
the Safety Siren detector tends to dramatically overestimate the radon concentration. Thus, this
device would provide an indication of soil gas entry at low concentrations useful for radon
management. In the higher range it might overestimate the risk.
13.3 Recommendations
13.3.1 Recommendations for Vapor Intrusion Research Generally
• A standardized quantitative format to describe the degree of variability in vapor intrusion studies
would help advance scientific understanding and aid practitioners. Such a format could
- allow the relative importance of temporal and spatial variability to be compared for a given site;
- allow various intensely studied sites to be intercompared to determine the magnitude of
variability to be expected in typical vapor intrusion applications; and
- guide practitioners, regulators, and stakeholders in assessing the relative value of additional
sampling rounds vs. additional sampling points.
• The use of stack-effect driving force calculations based on indoor-outdoor temperature
differentials should be further explored as a practical tool in for monitoring temporal variability.
These measurements can be made at extremely low cost with commercially available equipment.
• Because long exposure periods (up to one year) of VOC passive samplers appear promising in
this study for certain less volatile compounds, we suggest that
attempts be made to replicate these results in other building types, geologic settings, and
other contamination situations;
computational approaches be developed in which uptake rates used in data reporting are
corrected not only for exposure temperature (as they currently are) but also for sampling
duration (based on empirical results and/or vapor pressure). This would address one likely
objection—that passive samplers have predictably consistent negative biases over long
exposure periods; and
EPA-accepted methods (TO series) should be written for passive samplers in indoor air,
potentially based on current UK or EU methods.
• Researchers should begin to compare the cost-effectiveness of multiple tools and strategies that
could potentially be used to monitor or estimate long-term exposure in numerous structures
subject to vapor intrusion. Passive VOC samplers, fan testing, and surrogates such as radon or
indoor/outdoor temperature differential are all approaches that should be further developed and
compared.
• Because the number of we 11-published studies of VOC vapor intrusion under controlled
conditions is very small, additional studies should be undertaken that include frequent sampling
of groundwater, soil gas, and indoor air over long durations. Additional structure types such as
crawl space, slab on grade, mobile home, multifamily, and commercial should be included.
Additional climatic conditions, such as tropical or coastal, should be included, along with
different geologic settings.
13-6
-------
Section 13—Conclusions and Recommendations
13.3.2 Recommendations Regarding Further Study of this Test Site
• Further studies to better elucidate the exact routes of VOC migration through this test duplex
would be valuable. Tracer studies could address this as well as providing insight into the rate at
which VOCs and radon move through this system.
• Additional sampling at this site would help better establish the role of various potential sources in
what is presumed to be a low concentration impacted groundwater plume. This has implications
not only for local residents, but also for the management of potential vapor intrusion issues in
other historic urban neighborhoods.
• Further studies to better establish the roles of sewer mains and laterals in this case would be
valuable.
• Studies of this site that further elucidated deep soil gas attenuation processes, such as the greater
attenuation of chloroform as compared with PCE would contribute to a fuller understanding of
this site.
• 3-D numerical modeling of this site could help evaluate the utility of current state-of-the-art
models of vapor intrusion processes.
• The duplex structure and existing dataset at this site provide opportunities for comparative studies
of vapor intrusion investigation and mitigation techniques.
13-7
-------
Section 14—References
14. References
Abreu, L.D., and P.C. Johnson. 2005. Effect of vapor source —building separation and building
construction on soil vapor intrusion as studied with a three-dimensional numerical model.
Environmental Science and Technology 39(12):4550-4561.
Abreu, L.D., and P.C. Johnson. 2006. Simulating the effect of aerobic biodegradation on soil vapor
intrusion into buildings: influence of degradation rate, source concentrations. Environmental
Science and Technology 40(1):2304-2315.
AFCEE (Air Force Center for Environmental Excellence). 2004. Principles and Practices of Enhanced
Anaerobic Bioremediation of Chlorinated Solvents. Prepared for AFCEE, Brooks City-Base, TX.
Available at
http://costperformance.org/remediation/pdf/principles_and_practices_bioremediation.pdf
Arvela, H. 2008. Radonsafe Foundation, Moisture Prevention and Air Exchange in a Healthy Building.
Finnish Research Programme on Environmental Health. Available at
http://www.ktl.fi/sytty/abstracts/arvela.htm.
ATSDR (Agency for Toxic Substances and Disease Registry). 1997. Toxic Substances Portal -
Toxicological Profile for Chloroform. Available at
http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=53&tid=16.
Batterman, S., T. Metts, and P. Kalliokoski. 2002a. Diffusive uptake in passive and active absorbent
sampling using thermal desorption tubes. Journal of Environmental Monitoring 4:870.
Batterman, S., T. Metts, and P. Kalliokoski. 2002b. Low-flow active and passive sampling of VOCs using
thermal desorption tubes: Theory and application at an offset printing facility. Journal of
Environmental Monitoring 4:361-370.
Begerow, J., E. Jermann, T. Keles, and L. Dunemann. 1999. Performance of two different types of
passive samplers for the GC/ECD-FID determination of environmental VOC levels in air.
Fresenius Journal of Analytical Chemistry 363:399-403.
Bertoni, G., R. Tappa, and I. Allegrini. 2001. Internal consistence of the new "Analyst" Diffusive
Sampler: A long term field test. The Diffusive Monitor 12:2-5.
Bozkurt, O., K.G. Pennell, and E.M. Suuberg. 2007a. Evaluation of Characterization Techniques for
Vapor Intrusion Scenarios using a Three-Dimensional Computational Fluid Dynamics (CFD)
Model. Air and Waste Management Association, Providence, RI, September.
Bozkurt, O., K.G. Pennell, and E.M. Suuberg. 2007b. Characterizing Vapor Intrusion Scenarios Using a
Computational Fluid Dynamics Model. ACS National Meeting. Boston, MA. August.
Brenner, D. 2007. Effect of Meteorological and Site Conditions on, and Relationship of Diurnal Variation
to, Vapor Intrusion at NASA Ames Research Center. Conference Proceeding from Vapor Intrusion:
Learning from the Challenges-Air and Waste Management Association (AWMA), Providence, RI.
Brown, V., and D. Crump. 1993. Appropriate sampling strategies to characterize VOCs in indoor air
using diffusive samplers. Pp. 241-249 m Proceedings of International Conference on VOCs,
London, October 27-28.
Bruno, P., M. Caputi, M. Caselli, G. de Gennaro, and M. de Rienzo. 2005. Reliability of a BTEX radial
diffusive sampler for thermal desorption: Field measurements. Atmospheric Environment 39:1347-
1355.
14-1
-------
Section 14—References
Cao, X.L., and C.N. Hewitt. 1994. Study of the degradtion by ozone of adsorbents and hydrocarbons
adsorbed during the passive sampling of air. Environmental Science and Technology 28:757-725.
Carr, D.B., P.G., Laurent, C. Levy, A.H. Horneman. 2011. Stylistic Modeling ofVadose Zone Transport
Insight into Vapor Intrusion Processes. Presented at the AEHS Workshop: Addressing Regulatory
Challenges in Vapor Intrusion. A State of the Science Update. San Diego CA, March 15.
Case, J., and K. Gorder. 2006. The Investigation of Vapor Intrusion atHillAFB. Conference Proceeding
from Vapor Intrusion: The Next Great Environmental Challenge—An Update. Air and Waste
Management Association (AWMA), Los Angeles, CA.
Cleveland, W.S. 1981. LOWESS: A program for smoothing scatterplots by robust locally weighted
regression. The American Statistician 35 (1): 54.
Cleveland, W.S, and S.J. Devlin. 1988. Locally-weighted regression: an approach to regression analysis
by local fitting. Journal of the American Statistical Association 83 (403): 596-610.
Cocheo, C., C. Boaretto, D. Pagani, F. Quaglio, P. Sacco, L. Zaratin, and D. Cottica. 2009. Field
evaluation of thermal and chemical desorption BTEX radial diffusive sampler Radiello® compared
with active (pumped) samplers for ambient air measurements. Journal of Environmental
Monitoring 77:297-306.
Cody, R.J., A. Lee, and R. Wiley. 2003. Use of Radon Measurements as a Surrogate for Relative Entry
Rates of Volatile Organic Compounds in the Soil Gas. Presented at the AWMA Conference on
Indoor Air Quality Problems and Engineering Solutions. Research Triangle Park, NC, July 21-23.
Colby College. 2009. Regression: Patterns of Variation. Colby College Biology Department. Available at
http://www.colbv.edu/biology/BI17x/regression.html.
Coyne, L. 2010. Personal communication. SKC, Inc. August 17.
Dawson, H. 2008. US EPA 's Vapor Intrusion Database - Preliminary Evaluation of Attenuation Factors
for Chlorinated VOCs in Residences. U.S. EPA Office of Superfund Remediation and Technology
Innovation (OSRTI).
Dawson, H., and H. Schuver. 2010. US EPA's Vapor Intrusion Database: Preliminary Evaluation of
Attenuation Factors for Chlorinated VOCs in Residences. Presented at the 20th Annual
International Conference on Soils, Sediments, Water and Energy, March 16.
DePersio, T., and J. Fitzgerald. 1995. Guidelines forthe Design, Installation, and Operation of Sub-Slab
Depressurization Systems. Massachusetts Department of Environmental Protection, Northeast
Regional Office. Available at http://www.mass.gov/dep/cleanup/laws/ssdle.pdf
Dickson, J., A. Lonegran, and R. Stetson. 2010. Characterization of multiple chlorinated solvent plumes
due to the impact of TCE screening level reduction. International Journal of Soil, Sediment and
Water 3(2):Section 11.
Dietz, R.N., and E.A.Cote. 1982. Air infiltration measurements in a home using convenient
perfluorocarbon tracer technique. Environment International 5(1-6):419-433.
Dietz, R.N., et al. 1986. Detailed description and performance of a passive perflurocarbon tracer system
for building ventilation and air exchange measurements. Pp. 203-254 in Measured Air Leakage of
Buildings: A Symposium. Heinz R. Trechsel, Peter L. Lagus, ASTM Committee E-6 on
Performance of Building Constructions Contributor Heinz R. Trechsel, Peter L. Lagus. Published
by ASTM International.
DiGiulio, D., C. Paul, R. Cody, R. Willey, S.Clifford, P. Kahn, R. Mosley, A. Lee, and K. Christensen.
2006. Assessment of Vapor Intrusion in Homes near the Raymark Superfund Site Using Basement
14-2
-------
Section 14—References
andSubslab Air Samples. EPA/600/R-05/147. U.S. EPA Office of Research and Development,
Ada, OK. March.
Distler, M., and P. Mazierski "Soil Vapor Migration Through Subsurface Utilities", Proceedings 2010
AWMA Vapor Intrusion Specialty Conference.
Eklund, B., and D. Borrows. 2009. Prediction of indoor air quality form soil-gas data at industrial
buildings. Groundwater Monitoring and Remediation 29( 1):118-125.
Environmental Quality Management. 2004. Users Guide for Evaluating Subsurface Vapor Intrusion into
Buildings. Prepared for US EPA under EPA Contract Number: 68-W-02-33, WA No. 004, PN
030224.0002. Available at
http://www.epa.gov/oswer/riskassessment/airmodel/pdf/2004_0222_3phase_users_guide.pdf
Folkes, D., W. Wertz, J. Kurtz, and T. Kuehster. 2009. Observed spatial and temporal distributions of
CVOCs at Colorado and New York vapor intrusion sites. Ground Water Monitoring &
Remediation 29(1):70-80.
Forbes, J., J. Havlena, M. Burkhard, and K. Myers. 1993. Monitoring of VOCs in the deep vadose zone
using multi-port soil gas wells and multi-port soil gas/ground-water wells. Pp. 557-571 in
Proceedings of the Seventh National Symposium on Aquifer Restoration, Ground-water
Monitoring, and Geophysical Methods, National Ground Water Association, Dublin, OH.
Garbesi, K., and R.G. Sextro. 1989. Modeling and field evidence of pressure-driven entry of soil gas into
a house through permeable below-grade walls. Environmental Science and Technology
23(12): 1481-1487.
Greiner, M.A., and R.M. Franza.2003. Barriers and bridges for successful environmental technology
transfer. The Journal of Technology Transfer 28(2): 167-177(11).
GSI Environmental. 2008. Final Report: Detailed Field Investigation of Vapor Intrusion Processes.
ESTCP Project ER-0423, September.
Hall, S.K., J. Chakraborty, and R.J. Ruch. 1997. Chemical Exposure and Toxic Responses. Boca Raton,
FL: CRC Press.
Hass, B.S., and R. Herrmann. 1998. Tracing volatile organic compounds in sewers. Water Science and
Technology 37(1):295-301.
Hayes, H. 2008. Air Toxics Ltd., personal communication to Chris Lutes, ARCADIS, August.
Helsel, D.R. 2005. Nondetects and Data Analysis: Statistics for Censored Environmental Data. Hoboken,
NJ: Wiley.
Howard, P.H., R.S. Boethling, W.F. Jarvis, W.M. Meylan, and E.M. Michalenko. 1991. Handbook of
Environmental Degradation Rates. Chelsea, MI: Lewis Publishers, Inc.
Hui, S.C. 2003. Lecture: Air Movement and Natural Ventilation. Department of Mechanical Engineering,
The University of Hong Kong. Used by Permission from Dr. Hui. Available at
http://www.arch.hku.hk/teaching/lectures/airvent/sect03.htm.
IndyGov. 2012. My Neighborhood. Available at http://maps.indy.gov/myneighborhood/.
ITRC (Interstate Technology and Regulatory Council). 2007. Vapor Intrusion Pathway: A Practical
Guideline. Prepared by The Interstate Technology & Regulatory Council Vapor Intrusion Team.
Available at http: //www. itrcweb. org/documents/VI-1 .pdf.
Johnson, P.C., and R.A. Ettinger. 1991. Heuristic Model for Predicting the Intrusion Rate of Contaminant
Vapors Into Buildings. Environmental Science and Technology 25:1445-1452.
14-3
-------
Section 14—References
Johnson, Paul C., et al. 2012. Vapor Intrusion Above a Dilute CHC Plume: Lessons-Learned from Two
Years of Monitoring. Presented at AEHS Conference San Diego.
Jury, W.A., D. Russo, G. Streile, and H.E. Abd. 1990. Evaluation of volatilization by organic chemicals
residing below the soil surface. Water Resources Research 26(1): 13-20.
Koontz, M.D., and H.E. Rector. 1995. Estimation of Distributions for Residential Air Exchange
Rates. EPA Contract No. 68-D9-0166, Work Assignment No. 3-19. U.S. EPA Office of
Pollution Prevention and Toxics. Washington, DC.
Krewski, D., R. Mallick, J.M. Zielinski, and E.G. Letourneau. 2005. Modeling seasonal variation in
indoor radon concentrations. Journal of Exposure Analysis & Environmental Epidemiology
7J(3):234-243.
Kuehster, T.E., D.J. Folkes, and E.J. Wannamaker. 2004. Seasonal Variation of Observed Indoor Air
Concentrations due to Vapor Intrusion. Presented at the Midwestern States Risk Assessment
Symposium, Indianapolis, IN.
Kurtz, J.P., and D.J. Folkes. 2008. Empirical Data on Lag Time for Vapor Intrusion from a Chlorinated
VOC Groundwater Plume. Presented at AEHS San Diego Conference, March 12.
Lau, P. 2008. Calculation of flow rate from differential pressure devices - orifice plates. Presented to
the European Metrology Association for Thermal Energy Measurement, August 2-4.
Lee, A., K. Baylor, P. Reddy, and M. Plate. 2010. EPA Region 9 's RARE opportunity to Improve Vapor
Intrusion Indoor Air Investigations. Presented at the 20the Annual International Conference on
Soils, Sediments, Water and Energy, March 16.
Linn, W., L. Appel, R. DeZeeuw, P. Doom, J. Doyon, T. Evanson, J. Farrell, D. Hanson, R. Jurgens, J.
So, C. Speigel, D. Switek, and S. Yankey. 2010. Conducting Contamination Assessment Work at
Drycleaning Sites. State Coalition for Remediation of Dry cleaners. Available at
http://www.drvcleancoalition.org/download/assessment.pdf.
Loureiro, C.O., and L.M. Abriola. 1990. Three-dimensional simulation of radon transport into houses
with basements under constant negative pressure. Environmental Science and Technology
24(9): 1338-1348.
Luo, H, P. Dahlen, P. Johnson, T. Creamer, T. Peargin, P. Lundegard, B. Hartman, L. Abreau, and T.
McAlary. 2006. Spatial and temporal variability in hydrocarbon and oxygen concentrations
beneath a building above a shallow NAPL source. Presented at the Battelle Conference on
Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA, May.
Luo, H., P. Dahlen, P.C. Johnson, T. Peargin, and T. Creamer. 2009. Spatial variability of soil-gas
concentrations near and beneath a building overlying shallow petroleum hydrocarbon-impacted
soils. Ground Water Monitoring & Remediation 29(1):81-91.
Lutes, C.C., R. Uppencamp, C. Singer, L. Abreu, R. Mosley, sand D. Greenwell. 2009. Radon Tracer as
a Multipurpose Tool to Enhance Vapor Intrusion Assessment and Mitigation. Poster presented at
the Partners in Environmental Technology Technical Symposium and Workshop, Washington DC,
December.
Lutes, C.C., R. Uppencamp, H. Hayes, D. Greenwell, and R. Mosley. 2010. Long-term Integrated
Samplers for Indoor Air and Sub-slab Soil Gas at VI Sites. Presented at the 2010 Battelle
Conference, Monterey, CA.
MacLeod, G., and J.M. Ames. 1986. Comparative assessment of the artifact background on thermal
desorption of Tenax GC and Tenax TA. Journal of Chromatography 355:393-398.
14-4
-------
Section 14—References
MADEP. 2002. Indoor Air Sampling and Evaluation Guide. Boston, MA: Commonwealth of
Massachusetts Executive Office of Environmental Affairs, Office of Research and Standards,
Department of Environmental Protection.
Marotta, L., M. Snow, and S. Varisco. 2012. Extending the Hydrocarbon Range above Naphthalene for
Soil Vapor and Air Samples Using Automated Thermal Desorption/Gas Chromatography/Mass
Spectrometry (ATD/GC/MS). Presented at National Environmental Monitoring Conference,
Washington DC, August 6-10.
McAlary, T.A., P. Dollar, P. de Haven, R. Moss, G. Wilkinson, J. Llewellyn, and D. Crump. 2002.
Assessment of subsurface vapour transport through Triassic sandstone and quarry fill into indoor
air in Weston Village, Runcorn. Presented at Indoor Air 2002, Monterey, CA.
McHugh, T. E., and Nickels, T. N. 2007. Evaluation of Spatial and Temporal Variability in VOC
Concentrations at Vapor Intrusion Investigation Sites. Conference Proceeding from Vapor
Intrusion: Learning from the Challenges-Air and Waste Management Association (AWMA),
Providence, RI.
McHugh, T., P. De Blanc, and R. Pokluda. 2006. Indoor air as a source of VOC contamination in shallow
soils below buildings. Soil and Sediment Contamination 15:103-22.
McHugh, T.E., T.N. Nichols, and S. Brock. 2007. Evaluation of spatial and temporal variability in VOC
concentrations at vapor intrusion investigation sites. Pp. 129-142 m Proceeding of Air & Waste
Management Association's Vapor Intrusion: Learning from the Challenges, Providence, RI,
September 26-28.
Mosley, R.B. 2007. Use of Radon to Establish a Building-specific Sub-slab Attenuation Factor for
comparison with Similar Quantities measured for Other Vapor Intrusion Contaminants. Presented
at the National Environmental Monitoring Conference, Cambridge, MA, August 19-25.
Mosley, R.B., D. Greenwell, A. Lee, K. Baylor, M. Plate, and C. Lutes. 2008. Use of Integrated Indoor
Radon and Volatile Organic Compounds (VOCs) to Distinguish Soil Sources from Above-Ground
Sources. Extended abstract and oral presentation, AWMA Symposium on Air Quality
Measurement and Technology, Chapel Hill NC, November 6.
Mosley, R.B., D. Greenwell, and C.C. Lutes. 2010. Use of Integrated Indoor Concentrations of Tracer
Gases and Volatile Organic Compounds (VOCs) to Distinguish Soil Sources from Above-Ground
Sources. Poster presented at the Seventh International Remediation of Chlorinated and Recalcitrant
Compounds Conference, Monterey, CA, May 24-27.
Murry, D.M., and D.E. Burmaster. 1995. Residential air exchange rates in the United States: Empirical
and estimated parametric distributions by season and climatic region. Risk Analysis 7J(4):459-65.
National Weather Service. 2012. Excessive Heat - An Underrated Problem. National Weather Service
Regional Office, Central Regional Headquarters. Available at
http://www.crh.noaa.gov/Image/lsx/wcm/Heat/ExcessiveHeat.pdf
Nazaroff, W.W. 1988. Predicting the rate of 222Rn entry from soil into the basement of a dwelling due to
pressure-driven air flow. Radiation Protection Dosimetry 22(1/4): 199-202.
Nazaroff, W.W., and A.V. Nero (Eds.). 1988. Radon and Its Decay Products in Indoor Air. New York:
John Wiley and Sons.
Nazaroff, W.W., S.R Lewis, S.M. Doyle, B.A. Moed, and A.V. Nero, 1987. Experiments on pollutant
transport from soil into residential basements by pressure-driven airflow. Environmental Science &
Technology 21:459-466.
14-5
-------
Section 14—References
Neptune and Company. 2007. Report for 2007 Annual Vapor Intrusion Sampling for Buildings N210 and
19 and Additional Baseline Sampling for Building 16. Prepared for ISSi/SAIC NASA Ames
Research Center Moffett Field, California 94035-1100 Neptune Project No 09501.
New York Times. 2012. Toxic Waters: Indianapolis Water. May 26. Available at
http://projects.nvtimes.com/toxic-waters/contaminants/in/marion/in5249004-indianapolis-water.
Newton, E., and R. Rudela. 2007. Estimating correlation with multiply censored data arising from the
adjustment of singly censored data. Environmental Science and Technology 47(l):221-228.
NJ DEP (New Jersey Department of Environmental Protection). 2012. Draft Vapor Intrusion Guidance.
January.
Odabasi, M. 2008. Halogenated volatile organic compounds from the use of chlorine-bleach-containing
household products. Environmental Science and Technology 42:1445-1451.
Oury, B., F. Lhuillier, J. Protois, and Y. Morele. 2006. Behavior of the GABIE, 3M3500, PerkinElmer
Tenax TA, and RADIELLO 145 diffusive samplers exposed over a long time to a low
concentration of VOCs. Journal of Occupational and Environmental Hygiene: 547-557.
Parker, G.B. 1985. PNL-SA-13507. Presented at Conference on Conservation in Buildings: Northwest
Perspective, Butte MT, May 20. Abstract available at
http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=5567459.
Pennequin-Cardinala, A. et al. 2005. Performances of the Radiellos diffusive sampler for BTEX
measurements: Influence of environmental conditions and determination of modeled sampling
rates. Atmospheric Environment 39:2535-2544.
R Development Core Team. 2012. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. Available at http://www.R-project.org.
Razali, N., and Y.B. Wah. 2011. Power comparisons of Shapiro-Wilk, Kolmogorov-Smirnov, Lilliefors
and Anderson-Darling tests. Journal of Statistical Modeling and Analytics 2(l):21-33. Available
at http://instatmv.org.my/downloads/e-iurnal%202/3.pdf. Retrieved 5 June 2012.
Revzan K.L., W.J. Fisk, and A.J. Gadgil. 1991. Modeling Radon Entry into Houses with Basements:
Model Description and Verification. Indoor Air 2:173-189.
Robinson A.L., and R.G. Sextro. 1997. Radon entry into buildings driven by atmospheric pressure
fluctuations. Environmental Science and Technology 31:1742-1748.
Robinson, A.L., R.G. Sextro, and W.J. Riley. 1997. Soil-gas entry into houses driven by atmospheric
pressure fluctuations - the influence of soil properties. Atmospheric Environment 31:1487-1495.
Ryan, J.V, P.M. Lemieux, and W.T Preston. 1998. Near-real-time measurement of trace volatile organic
compounds from combustion processes using an on-line gas chromatograph. Waste Management
75(6):403-410.
Scheeringa, K., and K. Hudson. 2011-2012. Monthly Climate Summaries. Indiana State Climate Office.
Available at http://iclimate.org/summary.asp.
Schumacher, B., J. Zimmerman, G. Swanson, J. Elliot, and B. Hartman. 2010. Field Observations on
Ground Covers/Buildings. Presented at the 20the Annual International Conference on Soils,
Sediments, Water and Energy, March 16.
Schuver, H., and L. Siegel. 2011. Vapor Intrusion: Involved-Stakeholder Awareness of the Uncertainty
(and Multiple Benefits of Controls). Community Involvement Training Conference, Crystal City
VA, July. Available at
14-6
-------
Section 14—References
http://epa.gov/ciconference/download/presentationsAVed_OpenTime%20Session_Salon6_Schuver
Siegel.pdf.
Schuver, H.J., and R.B. Mosley. 2009. Investigating vapor intrusion with confidence and efficiency (some
observations from indoor air-based radon intrusion studies). AWMA Vapor Intrusion 2009, San
Diego, CA.
Sigma-Aldrich. 2012. Radiello manual, Volatile organic compounds (VOCs) chemically desorbedwith
CS2, Supelco edition. Available at http://www.sigmaaldrich.com/content/dam/sigma-
aldrich/docs/Supelco/Application Notes/radiello dl d6.pdf. accessed August 2012.
Smajstrla, A.G., and D.S. Harrison. 1998. Tensiometers for Soil Moisture Measurement and Irrigation
Scheduling. CIR487. Agricultural and Biological Engineering Department, Florida Cooperative
Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Available at
http://edis.ifas.ufl.edu/ael46.
The Polis Center. 2012. Study Neighborhoods from the Project on Religion and Urban Culture:
Mapleton-Fall Creek. Footnote 2. The Polis Center, Indianapolis, IN. Available at
http://www.polis.iupui.edu/RUC/Neighborhoods/MapletonFallCreek/MF CNarrative.htm#_ftn2.
Truesdale, R., H. Dawson, and I. Hers. 2005. Vapor Intrusion Database Status and Updates. Presented at
The EPA/AEHS Conference, San Diego, CA, March 14. Available at
http://iavi.rti.org/attachments/Resources/1045-
_Truesdale_Vapor_Intrusion_Database_Status_and_Updates.pdf
U.S. EPA (Environmental Protection Agency). 1990. NAREL standard operating procedures for Radon-
222 measurement using diffusion barrier charcoal canisters. EPA 520/5-90/032 Stock Number
PB91-179002. U.S. EPA National Air and Radiation Environmental Lab., Montgomery AL.
U.S. EPA (Environmental Protection Agency). 1992. Indoor Radon and Radon Decay Product
Measurement, Device Protocols. EPA 402-R-92-004, Office of Radiation Programs, Washington,
DC, July (revised).
U.S. EPA (Environmental Protection Agency). 1993a. Protocols for Radon and Radon Decay Product
Measurements in Homes. EPA-402-R-92-003, May 1993a.
U.S. EPA (Environmental Protection Agency). 1993b. Radon Reduction Techniques for Existing
Detached Houses, Technical Guidance (third edition) for Active Soil Depressurization Systems.
EPA/625/R-93/011.
U.S. EPA (Environmental Protection Agency). 1997. Compendium of Methods for the Determination of
Toxic Organic Compounds in Ambient Air, Second Edition, Compendium Method TO-17:
Determination of Volatile Organic Compounds in Ambient Air Using Active Sampling Onto
Sorbent Tubes. EPA/625/R-96/010b. January.
U.S. EPA (Environmental Protection Agency). 1999. Compendium of Methods for the Determination of
Toxic Organic Compounds in Ambient Air, Second Edition, Compendium Method TO-14Ab:
Determination of Volatile Organic Compounds (VOCs) in Ambient Air Using Specially Prepared
Canisters With Subsequent Analysis By Gas Chromatography. EPA/625/R-96/010b. Available at
http: //www. epa.gov/ttnamti 1 /file s/ambient/airtox/to-14ar.pdf
U.S. EPA (Environmental Protection Agency). 1999. Compendium of Methods for the Determination of
Toxic Organic Compounds in Ambient Air Second Edition Compendium Method TO-15
Determination Of Volatile Organic Compounds (VOCs) In Air Collected In Specially-Prepared
Canisters And Analyzed By Gas Chromatography Mass Spectrometry (GC/MS)EPA/625/R-
96/01 Ob. Available at http://www.epa.gov/ttnamtil/files/ambient/airtox/to-15r.pdf.
14-7
-------
Section 14—References
U.S. EPA (Environmental Protection Agency). 2002a. OSWER Draft Guidance for Evaluating the Vapor
Intrusion to Indoor Air Pathway from Groundwater and Soils (Subsurface Vapor Intrusion
Guidance). EPA530-D-02-004. Available at http://www.epa.gov/correctiveaction/eis/vapor.htm
U.S. EPA (Environmental Protection Agency). 2002b. Guidance on Environmental Data Verification and
Validation. EPA QA/G-8 EPA/240/R-02/004. November. Available at
http://www.epa.gov/quality/qs-docs/g8-final.pdf.
U.S. EPA (Environmental Protection Agency). 2003. A Standardized EPA Protocol for Characterization
of Indoor Air Quality in Large Office Buildings. U.S. EPA Indoor Air Division and Atmospheric
Research and Exposure Assessment Laboratory. Available at http://www.epa.gov/iaq/
base/pdfs/2003_base_protocol.pdf.
U.S. EPA (Environmental Protection Agency). 2005a. DRAFT Standard Operating Procedure (SOP) for
Installation of Sub-Slab Vapor Probes and Sampling Using EPA Method TO-15 to Support Vapor
Intrusion Investigations. Available at http://www.epa.gov/region8/r8risk/pdf/epa_sub-
slabvapor.pdf.
U.S. EPA (Environmental Protection Agency). 2005b. DRAFT Assessment of Vapor Intrusion in Homes
Near the Former Raymark Superfund Site - Recommendations for Testing at Other Sites.
U.S. EPA (Environmental Protection Agency). 2011. Exposure Factors Handbook 2011 Edition (Final).
EPA/600/R-09/052F. 2011. Office of Research and Development, Washington, DC.
U.S. EPA (Environmental Protection Agency). 2012a. Conceptual Model Scenarios for the Vapor
Intrusion Pathway. Office of Solid Waste and Emergency Response, Washington, DC. Available at
http://www.epa.gov/oswer/vaporintrusion/documents/vi-cms-vllfinal-2-24-2012.pdf
U.S. EPA (Environmental Protection Agency). 2012b. EPA On-line Tools for Site Assessment
Calculation: Johnson andEttinger Attenuation Factor. U.S. EPA Ecosystems Research, Athens,
GA. Available at http://www.epa.gov/athens/learn2model/part-two/onsite/jne alpha.htm.
U.S. EPA (Environmental Protection Agency). 2012c. EPA 's Vapor Intrusion Database: Evaluation and
Characterization of Attenuation Factors for Chlorinated Volatile Organic Compounds and
Residential Buildings. EPA 530-R-10-002. Office of Solid Waste and Emergency Response,
Washington, DC. March 16. Available at
http://www.epa.gov/oswer/vaporintrusion/documents/OSWER_2010_Database_Report_03-16-
2012_Final_witherratum_5 08 .pdf.
U.S. EPA (Environmental Protection Agency). 2012c. Drinking Water Contaminants. Available at
http://water.epa.gov/drink/contaminants/index.cfm.
University of Minnesota. 2008. Re-Arch: The Initiative for Renewable Energy in Architecture. Fact Sheet.
University of Minnesota Initiative for Renewable Energy in Architecture (re-ARCH). Available at
http: //www .rearch .umn. edu/factsheets/VentilationFactSheet .pdf.
Wang, F., and I.C. Ward. 2000. The development of a radon entry model for a house with a cellar.
Building and Environment 35:615-631.
Wertz, W., and T. Festa. 2007. The patchy fog model of vapor intrusion. Pp. 28-36 in Proceedings of
AWMA Conference on Vapor Intrusion: Learning from the Challenges, Providence, RI, September
26-28. Pittsburgh, PA: Air and Waste Management Association.
Whitmore, A., and R.L. Corsi. 1994. Measurement of gas-liquid mass transfer coefficients for volatile
organic compounds in sewers. Environmental Progress May: 114-123.
Wickham, H. 2009. ggplot2: elegant graphics for data analysis. New York: Springer.
14-8
-------
Section 14—References
Wilson, L.G., L.G. Everett, and S.J. Cullen (Eds.). 1995. Handbook of Vadose Zone Characterization and
Monitoring. Boca Raton, FL: Lewis Publishers.
Winberry, W. T. et al. 1990. Compendium of Methods for the Determination of Air Pollutants in Indoor
Air. EPA/600/4-90/010. April.
Wisbeck D., C. Sharpe C., A. Frizzell, C. Lutes, and N. Weinberg. 2006. Using naturally occurring radon
as a tracer for vapor intrusion: A case study. ARCADIS, presented at the 2006 Society of Risk
Analysis (SRA) Annual Meeting, Baltimore, MD.
Zhao, Y., and C. Frey. 2006. Uncertainty for data with non-detects: air toxic emissions from
combustion. Human and Ecological Risk Assessment 72:1171-1191.
14-9
-------
Appendix A
Soil Boring Logs
-------
Interior Soil Gas Port (SGP) Borings
SGP8
0 - 2.5 ft below the concrete slab- sandy silt, brown, moist
2.5 - 3.5 ft- tan/brown, slightly moist sand (med-fine); large gravel toward 3.5
3.5 - 6 ft- brown/tan, dry, sand and gravel. Some large gravel
SGP 9
0 - 2.5 ft below the concrete slab- dark brown, clayey, silty sand
2.5 - 3.5 ft - dramatic color change to tan sand with little to no cohesion
3.5 - 4 ft ft - same as the previous type, but the gravel size increases with depth from pea-sized
to stones approximately 1 - 1.5 in long. A few are as big as 2 - 4 in. At about 4 ft, we hit the
gravel layer that hinders the augur, making hand-drilling impossible.
Outdoor Soil Gas Port (SGP) Borings
The following are from the outdoor boring logs taken by Randy Woodruff from August 30
through September 1 of 2010. The depths for SGPs 8 and 9 correspond roughly to depths 5 -
11 ft and 5 - 9 ft respectively. For brevity, I have included only the logs for depths between 5
and 11 ft.
SGP 2
1.3- 6.0 ft- Brown - Dk Brown, silty sandy clay, dry, stiff-firm, friable: si. Plastic @ 3 ft;
decreasing hardness with depth.
6.0 - 6.5 ft- Brown, silty sand with some clays, si. moist, si. plastic, cohesive
6.5 - 7.0 ft- Dk brown-grey, clayey sand with trace gravels, si plastic, si moist
7.0 - 11.0- Lt grey - It brown, sands and gravels, fine - coarse, loose - si cohesive, si moist;
oxidation staining from 8.25 - 8.75 ft; slightly clayey from 9.0 - 9.5 ft.
SGP 3
3.0 - 7.0- Dk brown, silty clay, dry, si plastic, si moist at 6.5 ft
7.0 - 12.5 ft- Lt brown, sands with some clay, si moist, si plastic/cohesive; oxidation staining at
10.0 ft. Large gravels at 10.5 - 12.0 ft.
SGP 4
1.0- 6.75 ft- Dk brown, sandy clay, moist, medium stiff; dry at 4.0 ft, hard
6.75 - 14.0 ft- Lt brown, sands and gravels with some clays, si cohesive - loose, si moist.
SGPS
4.0 - 6.25 ft- brown, silty sandy clay, firm - hard, dry, si plastic
6.25 - 8.0 ft- Lt brown, fine - med sands, si moist, si cohesive - loose
8.0 - 12.0 ft- Lt brown, sands and gravels with some clay, si moist, fine - coarse, loose - si
cohesive.
SGP 7
1.25 - 6.75- Brown, silty sandy clay, dry, si plastic - friable, trace-some gravels; ash and wood
debris at 2.0 ft, stiff; dec gravel content with depth.
6.75 - 15.5 ft- Lt brown, sands and gravels, with some clay, non-plastic, si cohesive, si moist;
fine - coarse sands; inc gravel size at 11.0 ft. Oxidation staining at 10.0 ft.
It looks like the dominant material for the first two feet or so beneath the slab is a dark brown
silty sandy clay, dry to moist in some areas.
-------
Well Construction Log
(Unconsolidated)
/
/
/
'
/
/
/
/
/
^~
_
^
^
—
-
-
-
-
; •
/
/
/
<
/
/
i
/
i
/
\
^
~^ft Project Name and No. TO-97
•^ LAND SURFACE
Well MW-1A Town/City Indianapolis, Indiana
County Marion State IN
6.25 inch diameter Permit No. NA
drilled hole
Land-Surface Elevation and Datum:
feet I""] Surveyed
\
>— Well casing, | | Estimated
2 inch diameter, Installation Date(s) 29-Apr-10
PVC
Drilling Method Geoprobe 6600 - Hollow Stem Auger
'|_| Backfill
fxlGrout Benseal Drilling Contractor WDC
Drilling Fluid NA
1 ft*
Development Technique(s) and Date(s)
Bentonite Qslurry
14.5 ft* [xjpellets Surge Block and Pumping 4/30/2010
Fluid Loss During Drilling NA gallons
16 ft*
Water Removed During Development 80 gallons
v Static Depth to Water 16 feet below M.P.**
N— Well Screen.
2 inch diameter, Pumping Depth to Water feet below M.P.**
PVC
Pumping Duration hours
0.01 slot
Yield gpm Date
/fx|Filter Pack Specific Capacity gpm/ft
|X|Formation Collaspse Well Purpose Monitoring
21 ft*
Remarks
22 ft*
* Depth Below Land Surface
"Measuring Point is Top of Well Casing Unless Otherwise Noted.
Prepared by
Z:\Projects\0213151-EPA_STREAMS_ll\0213151.001-Vapor_lntrusion\Deliverables\Final_Report\Appendices\Appendix_A_bonng_Logs\Appendix_A\april '2UW Logs(^).xls- MW-1A
-------
Well Construction Log
(Unconsolidated)
/
/
/
'
/
/
/
/
/
^~
_
^
^
—
-
-
-
-
; •
/
/
/
<
/
/
i
/
i
/
\
^
~^ft Project Name and No. TO-97
•^ LAND SURFACE
Well MW-1B Town/City Indianapolis, Indiana
County Marion State IN
12.25 inch diameter Permit No. NA
drilled hole
Land-Surface Elevation and Datum:
feet I""] Surveyed
\
>— Well casing, | | Estimated
2 inch diameter, Installation Date(s) 29-Apr-10
PVC
Drilling Method Geoprobe 6600 - Hollow Stem Auger
'|_| Backfill
fxlGrout Benseal Drilling Contractor WDC
Drilling Fluid NA
1 ft*
Development Technique(s) and Date(s)
Bentonite Qslurry
22.4 ft* [xjpellets Surge Block and Pumping 4/30/2010
Fluid Loss During Drilling NA gallons
24.4 ft*
Water Removed During Development 80 gallons
v Static Depth to Water 16 feet below M.P.**
N— Well Screen.
2 inch diameter, Pumping Depth to Water feet below M.P.**
PVC
Pumping Duration hours
0.01 slot
Yield gpm Date
/fx|Filter Pack Specific Capacity gpm/ft
|X|Formation Collaspse Well Purpose Monitoring
26.4 ft*
Remarks
28 ft*
* Depth Below Land Surface
"Measuring Point is Top of Well Casing Unless Otherwise Noted.
Prepared by
Z:\Projects\0213151-EPA_STREAMS_ll\0213151.001-Vapor_lntrusion\Deliverables\Final_Report\Appendices\Appendix_A_bonng_Logs\Appendix_A\april '2UW Logs(^).xls- MW-lb
-------
If*
SOIL CORE / SAMPLING LOG
Boring/Well: MW-1 Project/No.:
Site
Location: 420 E. 28th Street, Indianapolis, IN
Drilling
Contractor: WDC
Drilling Fluid Used: None
Length and Diameter
of Coring Device: L: 5.0' D: 2.0"
Land-Surface Elev.: -- feet
Total Depth Drilled: 27.0 Feet
Prepared
By: RU
TO-97
Drilling
Started: 4/29/2010
Driller: Ron
Drilling Method:
Sampling Interval:
Q Surveyed Water: NA
Hole Diameter: 2.0" Coring Device:
Hammer
Weight:
Page: 1 of 1
Drilling
Completed: 4/29/2010
Helper:
GeoProbe6610DT-HAS
5.0 feet
GeoProbe6610DT
Hammer
Drop: — ins.
Sampling Data:
Depth
7-10'
10-12'
12-15'
15-17'
Grab/Composite
Grab
Grab
Grab
Grab
Time
10:12
10:15
10:17
10:22
QA/QC Collected
-
-
-
-
Laboratory Analysis
pH, TOC, bulk density, moisture
pH, TOC, bulk density, moisture
pH, TOC, bulk density, moisture
pH, TOC, bulk density, moisture
Soil Characterization:
Sample Interval
(Feet bgs)
From To
0
0.5
1.5
5
7
16
18
25
0.5
1.5
5
7
16
18
25
27
Core
Recovery
(Percent)
80
80
80
100
100
PID/FID
Reading
(ppm)
—
—
—
—
—
—
—
—
—
Blow
Counts
(per 6 in.)
—
—
—
—
—
—
—
—
—
Sample
Depth
(Feet bgs)
—
—
—
—
—
—
—
—
—
Sample/Core Description
Topsoil
Black cinders & silt
Med. Brown, moist, clayey silt, low plasticity
Light brown, moist, sandy clayey silt
Silty sand & gravel, moist, poorly sorted, medium sand &
gravel with some large gravel
Brown, fine to medium sand
Brown/tan, Sand & gravel, poorly sorted, water at 18.5'
Gray/brown, wet, fine-medium sand
End of Boring - 27'
-------
ARCAD1S
SOIL GAS PORT CONSTRUCTION DIAGRAM
Valve & Tube Fitting
LAND SURFACE
0.5 ft*
• Flush Mount
Protective Cover
1 Concrete Surface Seal
El granular
- Bentonite Q slurry
Q pellets
Tubing
8.0 ft*
9.0 ft*
1 Screen
£3 granular
- Bentonite Q slurry
Q pellets
12.0 ft*
- Sand Pack
13.0 ft*
-Screen
Depth Below Land Surface
Project: TO-97
City: Indianapolis,
County: Marion
Port: SGP-1 (A&B)
State: IN
GPS Coordinates:
Latitude: NA
Longitude: NA
Land-Surface Elevation and Datum:
NA feet
Q Surveyed
fj Estimated
Installation Date: 4/29/2010
Weather Conditions at Installation: Sunny, Warm
Drilling Contractor: WDC
Driller: Ron
Drilling Method: Geobrobe 6600 Maro-Core
Screen:
Construction: Stainless Steel Mesh
Length: 6 - inches
Tubing:
Construction: Teflon
Diameter: 1/4 - Inch OD
End Valve:
Type/Construction:
End Connection: SS Swagelok Tube Fitting
S D
Volume of Air in Tubing/Screen: mLs
Volume of Air in Sandpack:
Volume of Air Purged at Installation:
Remarks:
mLs
mLs
Prepared by: RU
-------
jJWwv
SOIL GAS PORT CONSTRUCTION DIAGRAM
, — Valve & Tube Fitting
/ LAND SURFACE
! \ ^ Flush Mount
\ Protective Cover
\— Concrete Surface Seal
2_
3_
4 • — Tubing
5_
SI granular
6
Bentonite LJ slurry
D pellets
7_
8
1".
11
12
13
14
1=.
15.5 ft*
16 • Sand Pack
' . 16.5 ft*
\
\_Screen
Project: TO-97 Port:
City: Indianapolis
County: Marion State: IN
GPS Coordinates:
Latitude: NA
Longitude: NA
Land-Surface Elevation and Datum:
Q Surveyed
NA feet
Q Estimatec
Installation Date: 4/29/2010
Weather Conditions at Installation: Sunny, warm
Drilling Contractor: WDC
Driller: Andy
Drilling Method: Geobrobe 6600 Maro-Core
Screen:
Construction: Stainless Steel Mesh
Length: 6 - inches
Tubing:
Construction: Teflon
Diameter: 1/4 - Inch OD
End Valve:
Type/Construction:
End Connection: SS Swagelok Tube Fitting
S D
Volume of Air in Tubing/Screen: m
Volume of Air in bandpack: m
Volume of Air Purged at Installation: m
Remarks:
Prepared by: RU
SGP-1C
Ls
Ls
Ls
Depth Below Land Surface
-------
Soring/Wei
Silt
Drilling
Contractor
Drilling Fli*
^engtiiandl
of Coring Dt
Land-Surfac
Total Depth
Prepared
By
Boring Tern
Sampling P
Di
Soil Chara
Sample/Col
(Fe
a
2.
^I L DnHer/U^e*
Used ' Drilling Method <^Sk^S«2£:
ameter
^M . - Samplire Interval feel
Bev. ^ f
Jrilled /d>'f 1
nation Depth
th
set 1 ISmveyefl [^
eel Hole Diameter
110,$'
Grab/Composite
Ttoe
|Estimated Dannn
Coring Device
Hammer Hammtr
Weight Drop ins.
Water OPS Coordinates
BorinE/WcB Location Sketch 1
OAAJC Saffiples I
terization:
Depth
bis}
To
^>
a.
4
R-
/o
/2-
^y:
/^
/6^
Core
ecovery
(%)
^O
P
rj>$
I1*
•
iff
SOQH
J0& \
/•<*»
{ruction Detail!:
•Dtccllon
ih
^
^
-,, _,.
-.
— —
^
^•— _^
-UJ,
Blow
Counts
per 6 Inches
— —
— ~,
. ~
.
f .
Desc
Diameter
Lengih
Mat era!
riptionlDepth
From - To
3,O ' »^"
[Jt £r *"" j! iff
fjf *" Jl*"^.
fjf, f.f*
&{' K>A
^boratory Analysis • ^K A/
' i
f '
— - - " '-
Sample/Core Description
TtAcsaC, o&fstj^x&'f &V- £<&&• MS%fe S^T^1* Ji
>K (9io«^ ^£*ii. fi^*( *AY ^s5*4^ '
l^fcw; ^/fii 2^s A&Or^-avF li^rkjertt Si
ftunsL. • ^
^&~ &*&*$.*: fM&z&ii-r tu&r&f»+\ SUA<>
-------
ARCADIS Boring Log: GP-7
GERAGHTY& MILLER _
Prefect Name: Maoteton - Fall Creek Data started: 6-21-05 Looaar TY
Project Number: IN000763.Q001. 00003 flat* r™^^. Frii1nr ^
II
5
10-
15-
20-
25-
30-
"f ='
fi
3 4 T 13
1941
434S
22810
B 819 24
1311011
12 11 14 IT
58812
rt 812
•.'•«-)
77t17
5 j
20
:
12
14
12
3
12
20
20
20
f gf
3 't
1 d.6 ~
J 0-0
I 0>0
n 0.0
j °°
J ao
T 0,0
jt 0.0
1 ao
T 0.0
n o.o
J °°
u
i
%
1
*
m
&
«*!*
%
^
;.ZT-
P
1
•w
"*'f*
m
*•'
Sfc
Sal Class
VC
so
OP-
SP
U»X(>
GP
DP-«P
Oascriptoo
Brown ( 10YR 4V3) tuty sandy day. sflfltWy pluBc. slIgMlf cohesivD. non- '
ttleKy, iRghtfy nxil*l, tr»o (pav«l
Dark raddirt brawn (SYR 3/4) doy«y sand, nonpteflc, alofiUy coheslw
nornOcky, motet
Vary pole brawn (10YR 8/4) medium to OOWM sand with traco oraval, dry
nc¥vtv mtrtavt *--*- -*--*
Dark yaitawvsn arw,n (1QYR 4V8) medium to coarw sand and gravei stighttv
wet, poorly graded, sub-rounded: w« Q tfi.O'
Oartt y«iuwtoh (Mown (TOYR 4«) gmval. wet. poorty gmd«d, «u( dry ""
poorty gradaci. aub-toundad
IXI Composite Sample to Lab
Drilling Co.: Earth Exploration
\ Grab Sample to Lab Q]0 Sample Not Analyzed Page 1 of 1
Dffler B. Scott PhHflps
Drilling Method: HSA_
Drilling Fluid: Nona
Sampling Method:
Sampling Interval:
Water Level Start
Water Level Finish:
Split Spoon
2.0-
16.0'
-------
ARCAOIS
Well Constajction Log:
GP-7
Project Name: _
Project Number
Mapteton - Fall Creek
Date:
6-21-05
Well Construction Diagram
Wefl Constructkin
Details
10
15-
20-
25-
30_
DrilBngCa: Earth Exploration
Driller B. Scott PhHIIps
Drilling Method: HSA
Drilling Fluid: None
PROTECTIVE COVER TYPE:
Stick-up
SURFACE SEAL
Bentonite
Total Depth:
ii.rj-
Type:
Top Depth:
Bottom Depth:
ANNULAR SEAL
Inside Diameter
Length:
Material:
CASING
1.0-
no-
PVC
Type:
Top Depth:
Bottom Depth:
SEAL
Bentonite
o.y
n.o1
Type
Inside Diameter:
Slot Size:
Top Depth:
Bottom Depth:
SCREEN
PVC
i.cr
0.010"
13.0-
23.0-
Material:
Top Depth:
Bottom Depth:
FILTER PACK
sand
no-
23.0"
Filter Placement Method:
Seal Placement Method:
Total Depth (TOC):
Water Level Finish:
Gravity Pour
Gravity Pour
2S.V
-------
ARCADIS Boring Log: GP-8 !
GERAGHTY&MILLER
Prefect Number
II
5-
10-
15-
20-
25-
30-
II
ma
"'"
i ma
an IBZS
« 11 « IT
15131827
M30tt»
14 10 14 23
aatia
•4131831
14 « 21 at
Marieton- Fall Creek Date Started: 6-21-05 Loooer TY
IN000763.0001. 00003 Date Cnmpiatarf- Frfltrr TY
!l
II
:
3
3
3
T2 '
12 "
12
3 "
20 "
20 '
20 '
20 '
M
1 00
\ 0.0
1 ° °
£ 0.0
1 0.0
I 0.0
1 00
1
J 0.0
\ 0.0
1 0.0
jl °'°
J ° °
n o.o
^
I*
;%v
%
%
m
P
it
si**
^
s
3*
fi
^
1
MOM,
FILL
vc
UW
^__
BP-»P
Owcrtptton
Brown (10YR 4V3) •% undy day, plsatlc. eofttsiva. itt*y. moltt. tnice
PMH
Very pale brown (lOYRa/4( medium to ooanwcand and ar«v«l dry eoorfy
graded, sub-roundad
Dark yeKowrtsri brown (10YR 4VBJ medHim to coane tand and araval,«wL
poorty graded, tub-rounded
I3 Composite Sample to Lab
DrilimgCo.: Earth Expkxation
I Grab Sample to Lab OJ Sample Not Analyzed Pose 1 of 1
Drilter: B. Scott Phillips
Drfllng Method: HSA
Drilling Fluid: None
Sampling Method:
Sampling Interval: _
Water Level Start _
Water Level Finish:
Split Sooon
2.01
18.0*
-------
ARCADIS
I Ptf**<* Name: Maoleton - Fall Creek
Well Construction Log:
GP-8
Data: ft-?1-0f> Innnnr- TV
j Project Number IN000763.0001.00003 ^itnf. ^
••*•*•••••
0
5
10
15-
20-
25-
30_
DfiHIngCc
Driller: B
Drilling M
Is
I1
•MMH^
—
).: Ea
Scott
Whort
Well Construction Diagram
c
:-
•
»^M«WMW.^HMnM^
rth Exploration
Phillips
HSA
•3
3
5
a-
-c
>
\
\
"•
..
Drilling Fluid: None
F
S
Wen Construction
Details
PROTECTIVE COVER TYPE:
Stick-UP
SURFACE SEAL
TyP* BentonKe
Total Depth: 130-
ANNULAR SEAL
Type:
Top Depth;
Bottom Depth:
CASING
Inside Diameter i.o»
Lengtm 15i0.
Material: pvc
SEAL
Typo: Bentonlta
Top Depth: o O1
Bottom Depth: no-
SCREEN
Type pvc
Slot Size: onio-
Top Depth: 150»
RLTER PACK
Material: s-,^
Top Depth: 13,0.
Bottom Depth: js o*
Iter Placement Method: GravttyPour
eal Placement Method: GravftyP«,r
Total Depth (TOC): 25.0-
Water Level Finish:
-------
ARCADIS Boring Log: GP-9
GERAGHTY&MILLER
Project Name: Marieton - FaB Creek Data Started: 6-22-05 Loaaer TY
Project Number IN000763.0001. 00003 Date Completed- Editor TY
II
0
5
10-
15-
20-
m-
30-
ll
--S1
1 «4.J
ills
9»56
4144
18 18 • 14
18 14 13 IS
|l
fh
M <
& *
12
3
12
20
12
12
12
3
12
20
20
20
t
I
h
I 8I
J 0.0
j ao
J 00
T 0.0
J 00
T 0.0
1 ::
J 0,0
J
if
s1
NfSj
I
I
ij*
1
®
ft
:-'!
i
i
i M
•P
*p
0^^
Con«rt»
Brown (10 VR 4/3) siity sandy clay, nonpiasbc. non-cohesjv«, non-stciy.
Daft yebrw(Ti(10YR4/Q) firw to medium sand, poorty sorted sut>-
round»d , inobi
Oartt jnltoMtoh brown <1GYR 4W) nwrti«i to roars* »and and gimti. dm"
poortyscrtad, sub-rounoad
ftrt ydfcwtoh temm |«3¥R 4») fbw to nwliyffl »«nd. pwxiy «Bttd~i^b-
DiMfc yellow tsh brawn frwdltra to OOtrs* cwxl and g ravel . wat, powly sorted
sub-round ed
^ Composite Sample to Lab
Drilling Co.: Earth Exploration
I Grab Sample to Lab QU Sample Not Analyzed Page 1 of 1
Driller B. Scott Phillips
Dritflng Mettxjd: HSA
DrilHng Fluid: None
. Sampling Method:
SampBnfl Intervat _
Water Level Start _
Water Level Finish:
Split Spoon
2.0*
17.0-
-------
ARCADIS
1 Project Name: Mapleton - FaH Creek
Well Construction Log:
GP-9
Data: R-??-Of5 I rwmr TV
j Project Number: IN000763.0001. 00003 p^,^ jy "~
£?
3!
o"
5-
10-
-
20-
25-
30 J
Drilling Cc
Driller B.
DrilSng Me.
f*l
§" 3 \ Well Construction Diagram
I
|
_
).: Ea
Scotl
-------
ARCADIS
GERAGHTY&MILLER
'reject Name: _
Boring Log: GP-10
M y
Maptefan- Fafl Creek
Project Number
IN000763.0001.00003
Data Started:
Date Completed:
6-22-05
Logger
Editor
TY
TY
•«« J WR 40} s% mdlr day,
rtckx. ^jWJy mow, iraca gr«v«
Vwy pata trawn (10VH WJ «M to nwlium wnd, dry, poo% Mrtad. lub-
'
«y p** bOM (10YR WJ iMdhMI «J OOWM W«i and pMl, «y,
-
poorty •oittd, ttJMwmtfM
Composite Sample to Lab
Drilling Co.: Earth Exploration
1 Grab Sample to Lab JED Sample Not Analyzed Page 1 of 1
Driller B. Scott Phillips
Drilling Mettwd: HSA
Drilling Fluid: None
Sampling Method:
Sampling Interval: _
Water Level Start _
Water Level Finish:
HSA
2.01
17.0-
-------
ARCADIS
Well Construction Log:
GP-10
Q *
0
5
10
15-
20-
25-
30.
!
•HHMBB
WellCc
c
i
*«
fc-
3
=
E
;•
Drilling Co.: Eartfi Exploration
Driller B. Scott Phillips
Drilling Method: HSA
Drilling Fluid: None
Well Construction
Details
PROTECTIVE COVER TYPE:
Sttek-uD
Type:
Total Depth:
SURFACE SEAL
Benionite
Type:
Top Depth:
Bottom Depth:
ANNULAR SEAL
CASINO
Inside Diameter: 1.0"
Length:
Material:
14.o.
pyc
Type:
Top Depth:
Bottom Depth:
SEAL
Bentonfte
off
Type
SCREEN
Inside Diameter
1 rr
0.010"
Top Depth:
Bottom Depth:
14.r/
240*
Material:
Top Depth: __
Bottom Depth:
RLTER PACK
12 O1
, Filter Placement Method:
Seal Placement Method:
Total Depth (TOC):
Water Level Finish:
24 rr
Gravity Pour
Qravfty Pair
24.01
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
BORING LOG
BORING NO
PROJECT NAME. Fall Br-k «.
PROJECT LOCATION- MlmaBa,^ r»
DRHJ.IN8 MTHD:
I TIME STARTED: oe.lB
CLIENT.
DRILLING CONTRACTOR.
N0= INOOOS20.QQP1
TIME FINISHED: 09,30
MTHDi
«-.rfrffT
B/Z3/OD
6EOLOSI8T: B. ri.h»-
SURFftCE ELEVATION: Nft
LITHOLOGIC
DESCRIPTION
BROW FIU, IIOICT
ROMISH BflOUN SWOT UWt mi8T, FTOfflX
.<3 8WIE 6RAT 8TAIHDC rtT ii FEET
LI^T ntoM COWSE aw m BRAVE, HOIST,
ZO.O
25.0
-30.0
BOTTWI OF rear BORINB: is.oo
Scnpl* co)I»ctBd rrom 10-12
SPT • 8TAMMM) PDimwnON TEST
REC" Stmt RECOVERT
W • HWHJCTECTrtBU
FID • FL«K IDKIZATI£» DETECTOR
no' PHDTo-icNramDN DTTECTCB?
J
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
BORING LOB
BORING NO=Bp-g
N0= INOOOS2Q noni,
PROJECT NAME- r.H
PROJECT LOCATION
DRILUN6
DRILLINB CONTRACTOR
SAMPLE MTHD= 8i».vff
TDC STARTED: 09,30
TIME FINISHED: 10
6EOL06IST: 8. ri«h«-
SURFACE ELEVATION: NA
LITHOL06IC
DESCRIPTION
K«BW SANDY LtWl NOMT, FRWBU
REBHSH BfWIN SILT U»l raiBT, tUCSJVE,
s««i LOME; NCN-COEBTVE;
BOTTWI CF TEST BtTONB: 16.00
8PT - BTMOMO POCTRATION TEST
REC • 8«fLE RECOVERT
«J • MTOHZTECTrtBLE
FID • FUW lOKUtiTttM DETECTOR
PUS • PHOTO-ICNGATICM DETECTOR
-------
I
I
I
I
I
I
I
I
1
1
1
I
I
PROJECT NAME
PROJECT
DRILLINS MTHD: Ggopr-ob.
TIME STARTED. 10.^3
Indlonopoll,. in
BORING LOG
BORIN6 NO=BP-3
PROJECT NO--
Mopl«»*0n-Fol I Cr-mmk
TDK FXNI8HED. u-eo
DRILLINS CONTRACTOR- Pcr^urrt Environ--^.. p—.m
SAMPLE MTHD: ei^v. _____
DATE: B/ZS/OO
armnsiBT: ,
SURFACE ELEVATION: NA
LITHOL06IC
DESCRIPTION
^SMKBRO*! SILTT CLAY UN* WJICT,
ryap. onnoo, unts owm
0 «gg CffleaWE, FUMBLE
|jg LlSff BfKBM FINE SWfflY LWt raiST,
IWTTLIN6
rt, ^ DARKER BROW AT 13 FEET
\-f.A\.
-30.0-
BOTTW1 CF TEST BOfONB: 16.00'
SPT • STANDMO PENETRATION TEST
REC- SAffU RECOVERT
M> • NOHXTECTABLE
FID • FLAft lOOantW KTECTCW
PB3 • PHDTP-IONIZATIPM DETECTOR
REMARKS
Irrtervoi
Fro« 1Z -
PASE: I
-------
BORING LOG
BORING
PROJECT NO
PROJECT NMIE. Foi
PROJECT LOCATION
0RILUN6 HTH0
DRILUN6 CONTACTOR
Envlrorman*n( gyvlc
TIME 8TrtRTED= 14.30
TD1E (TNIBHED: 12, .a
DATE; B/23/00
SURFftCE ELEVftTTON: Nrt
LITHOLOGIC
DESCRIPTION
mea"1*
SNOT L1MH, TOST, TRD«I,
WTTLES AT 7.5 FEET
Boapl. ool |«ot«d Trow W - 16
nrervDl
BOTTOn OT TEST BOKDB: 16.00
SPT • BTOMMfiO POCTWtnW TEST
REC • »WU RECOVERY
NON-OtTECTftSLE
FID » njVK JOaZATBW OOECTOR
PIP » PHOTO-IONIMTIDN DETECTW
-------
BORING LOG
BORING NQ=BP-5
N0= INOOOS20.QOOJ
Orffffi"
PROJECT NAME:, r.n
PROJECT LOCATION. i^lnnnpQ,.. TM
DR3LUN6 HTHO
DRILLINS CONTRACTOR
SAMPUHTHO
8/23/00
TBC STARTED: 12.13
TTHE FINISHED: 13.00
6EOL08I8T: e. ri.h^-
suRr«^CE ELEVftncy<= Nfl
LITHOLOGIC
DESCRIPTION
oollnrttd fron 2 -
6WY SANDY UWt tffllBT, COC8IVE,
ool |«rt«J Tro. 11 - 16 r«.t
BOTTOn OF TEST BOONS: 16.00
8PT - STANMflO PDOWnON TEST
IZC • 3AWIE RECOVERY
HD
FID • FLATt ItKtZATJDH DETECTOR
no • PHonncKrz*T DETECTOR
-------
BORING LOG
BORING NQ.-BP-e
: INOOOSZQ.DOqf
PROJECT NAME. Fel
PROJECT LOCATION
DRB.UN8 HTHD:
DRXLUNB CONTRACTOR
SWPLEMTHD:
B/23/DD O3JL08I8T: B ri»
TIOE FINISfCD; 13. ag
SURFACE ELEVATION; Nft
LITHOLOGIC
DESCRIPTION
BUCK HLT LMfl (FILL), HOIST, FR»aE
Does con AT e
aworLiwi OMNEEnomts.
LMHT BROW BAND HTTH TWCE
wo ewir fnniiN HOIST,
BOTTOn OF TEST BOR1N8: 16.00
SPT - STfNDMD PDCTIWniW TEST
REC« WlfPIX RECOVERY
MJ • (WfHlTECTftRX
no • rim. TmoKna* DETECTOR
pro» PHjTO-iONr KTCCTDR
-------
Appendix B
Temporary vs. Permanent Subslab Port Study
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Initial Draft
Interim Report for Subtask 2E
Fluctuation of Indoor Radon and
VOC Concentrations Due to
Seasonal Variations: Temporary vs. Permanent Subslab Port Study
Draft
Category III
U.S. EPA Contract Number EP-C-05-060
Project No. RN007160.0002
Task Order (TO) 97
RTI International
Research Triangle Park, North Carolina
U.S. Environmental Protection Agency
National Engineering Research Laboratory
Las Vegas, Nevada
Prepared By:
ARCADIS U.S., Inc.
4915 Prospectus Drive
Suite F
Durham, NC 27713
December 2010
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Title and Approval Sheet
Initial Draft
Interim Report for Subtask 2E
Fluctuation of Indoor Radon and VOC Concentrations
Due to Seasonal Variations (QA Category III)
Brian Schumacher Date
EPA Task Order Project Officer
Robert Truesdale, RTI Task Order Manager Date
Chris Lutes Date
ARCADIS Task Order Manager
Laura Beach Nessley Date
ARCADIS Quality Assurance Officer
Concurrence
Ed Heithmar, ECB QA Coordinator Date
George M. Brills Date
EPA ESD QA Manager
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table of Contents
Title and Approval Sheet i
List of Tables iii
List of Figures iii
List of Appendices iii
1. Project Background, Definition, and Objectives 4
1.1 Sub-Slab Port Installation and Use 4
1.2 Objectives 5
2. Methods 6
2.1 Probe Installation and Leak Checks 9
2.2 Sub-Slab Sampling Using Summa Type Canisters for VOCs 9
3. Results 12
3.1 Seal Integrity 12
3.2 VOC Sampling Results 13
3.3 Radon Results 16
3.4 Study Limitations 16
4. Conclusions 18
5. References 18
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
List of Tables
Table 1. Test Matrix: Sample Number, Frequency and Location 7
Table 2. Test Matrix: Analytical Methods, Analytes, Laboratory and Turnaround Times 8
Table 3. Summa Canister Sampling Times and Pressures 11
Table 4. Leak Test Results for Temporary Probes 13
Table 5. Radon Data Comparing Temporary and Permanent Probes 18
List of Figures
Figure 1. Interiors of test buildings, showing soil gas points (SGP) and paired single-depth temporary and
permanent sub-slab ports (SSP). 10
Figure 2. PCE concentrations in soil gas from temporary and permanent subslab probes. 14
Figure 3. Chloroform concentrations in soil gas from temporary and permanent subslab probes. 15
Figure 4. TCE concentration in soil gas from temporary and permanent subslab probes. 16
List of Appendices
Appendix A: Photographs of Ports Sampling
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
1. Project Background, Definition, and Objectives
This interim report presents the implementation and results of Subtask 2E, comparing the performance of
permanent and temporary sub-slab sampling ports for measuring volatile organic compounds (VOCs) in soil
gas beneath a basement building slab. This work is described in more detail in Addendum 1 to "Fluctuation
of Indoor Radon and VOC Concentrations Due to Seasonal Variations; Quality Assurance Project Plan"
(ARCADIS, November, 2010) [QAPP Addendum 1].
1.1 Sub-Slab Port Installation and Use
An introduction to vapor intrusion (VI) and the role of sub-slab sampling in investigating VI is provided in the
QAPP. In the original TO-97 study, sub-slab soil gas sampling ports were installed in the basement floor of
the test building at 422 East 28th St., Indianapolis, IN. These ports are considered "permanent" and were
installed in accordance with an SOP provided in the QAPP. Only the single-depth permanent sub-slab ports
installed beneath the basement floor slab were used in this study, not the multi-depth soil gas points.
To ensure comparability in this study, the methods used to install the TO-97 permanent sub-slab ports were
compared with procedures for permanent ports published in two guidance documents:
• Vapor Intrusion Guidance, New Jersey Department of Environmental Protection (NJDEP), dated
October 2005; at http://www.ni.gov/dep/srp/guidance/vaporintrusion/vig.htm ) [VIG]
• Response Engineering and Analytical Contract SOP 2082: Construction and Installation of
Permanent Sub-Slab Soil Gas Wells, dated March 2007
The comparison, detailed in QAPP Addendum 1, showed that the TO-97 installation methods were
functionally equivalent to either guidance document for installation of permanent sub-slab sampling ports.
The following small differences between methods were identified:
• the sequence of drilling the 3/8" and 1" holes
• whether the depth of the 1" diameter hole (that serves to hide the fitting below the floor) is fixed at
a depth of 1 3/8", or adjusted to the depth necessary to sink the fitting
• whether or not clay is used to help support the cement before it dries.
The New Jersey VIG also allows for the installation of temporary ports and the permanent and temporary
port types have different construction methods, materials, and surface seals. The NJ VIG permanent port
consists of an assembly of stainless steel tubing and Swagelok fittings which are cemented into a hole
drilled into the slab, allowing the sampler to repeatedly access the sample point. In contrast the NJ VIG
temporary port procedure allows the use of flexible tubing rather than stainless steel, sealed into a hole in
the slab with "modeling clay, beeswax or other non-volatile emitting and non-shrinking materials...". EPA
Region 2 staff has observed (see correspondence in Appendix A of QAPP Addendum 1) that the temporary
ports are often used and they are commonly sealed with clay or bentonite. EPA Region 2 requested testing
of permanent ports versus temporary points sealed with bentonite to determine whether the seals are
adequate to prevent indoor air from infiltrating into the subsurface during sampling and the methods achieve
comparable results.
Atypical application of a bentonite seal may take one of two approaches:
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
• Method 1. With the tubing in the drill hole, dry bentonite (granular or powdered) is poured into the
annular space, sprayed with water, allowed to hydrate, and then smoothed at the surface, with
edges feathered to make a seal with the floor and tubing. Complete hydration and full coverage of
bentonite in the annular space cannot be ensured using this method. A Teflon tape wrap on the
tubing, required by the NJ VIG, may be effective in preventing any liquids not taken up by the
bentonite from falling into the sub-slab sample space, but this cannot be ensured. Another potential
issue with this method is the degree of hydration and continuity of the bentonite beneath the
surface.
• Method 2. Bentonite is mixed in a container, starting with water, then adding bentonite until no free
water is present and the mixture has the consistency of gel or modeling clay. The mixture is applied
to the top 1-2 inches of tubing (above the Teflon tape barrier required by the NJ VIG) as it is twisted
into the drill hole, and as needed at the surface to fill the remaining annular space and make a seal
with the floor and tubing. This method would appear to be more reliable in preventing liquids from
passing the Teflon barrier, but may also fail to provide an even seal beneath the surface.
A seal installed using either of these methods is potentially subject to air leakage because the flexible tubing
is likely to move during sampling, possibly opening a pathway for air entry. The seal also depends on the
properties of the bentonite, which are likely to change as the bentonite dries. The second bentonite mixing
method described above was used for Task 2E.
1.2 Objectives
The primary objective of TO-97 is to investigate distributional changes in VOC and radon concentrations in
the indoor air, sub-slab, and subsurface from an underground source (groundwater source and/or vadose
zone source) adjacent to a residence or small commercial building. Addendum 1 to the QAPP added the
following goal, which is addressed in this report:
• Compare the quality of sub-slab vapor samples collected from permanent and temporary sub-
slab ports when the seal for the temporary port is constructed of bentonite, and the temporary
tubing is an allowable flexible material consistent with the New Jersey VIG.
The major elements of this task were:
• Installation of Temporary Sub-Slab Ports. On the interior of the building, five new temporary sub-
slab ports were installed. Each temporary port was paired with and installed within 30 cm of an
existing permanent port.
• Sub-Slab Soil Gas Sampling. Soil gas samples from the temporary sub-slab ports were collected
simultaneously with samples from the permanent ports using Method TO-15.
• Other Monitoring. Tracer gas leak testing of sub-slab ports was performed using helium and
handheld air testing instruments.
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
2. Methods
The quality objectives and criteria for Task 2E are described below. The test matrix for Task 2E is indicated
in Tables 1 and 2.
Study Question: Qualitatively Stated (from SOW Objectives when applicable).
Identify differences in functions and results between permanent and temporary sub-slab probes.
Study Question: Quantitatively/Statistically Stated.
(1) Is there a statistically significant difference between analyte concentrations in gas samples collected
from permanent and temporary sub-slab probes?
(2) Is there a measureable amount of leakage of indoor air into either type of probe during sample
collection?
Measurement Used To Support Study Question.
(1) Radon and VOC measurements in sub-slab soil gas samples
(2) Tracer gas (helium) measurements in sub-slab soil gas samples
Measurement Performance or Acceptance Criteria for this question/it of data points anticipated.
(1) Agreement of sub-slab concentrations within +/-30 percent is expected to be adequate given the
variable nature of sub-slab soil gas distribution.
Helium concentrations indicative of significant leakage are in the QAPP.
For each comparison 5 pairs of measurements were available.
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table 1. Test Matrix: Sample Number, Frequency and Location
Matrix
Interior sub-slab
soil gas
Interior sub-slab
soil gas
Sample/Measure-
ment Type
VOCs, TO-15
Summa Canister
Radon
Sample
Integration Time
or Frequency
24 hour
integrated
Limit sample flow
rate to 200 ml/min
or less.
Integration time
depends on
sampler
Estimated
Number of
Primary Sample
Measurements
6 = One time,
simultaneously
with the single
depth permanent
sub-slab ports,
sample the five
paired temporary
sub-slab ports.
Numbers represent
only the samples
from the temporary
ports and
associated QA
6 = One time,
simultaneously
with the single
depth permanent
sub-slab ports,
sample the five
paired temporary
sub-slab ports.
Numbers represent
only the samples
from the temporary
ports and
associated QA
Number of QA
Samples/Measurements
Duplicate
1
1
Equip Blank
0
0
Field Blank
0
0
Ambient
0
0
Total Number
of Samples/
Measurements
6
6
Locations
Interior: 5 single-
depth temporary
ports to be
installed next to
five single depth
permanent ports.
Interior: 5 single-
depth temporary
ports to be
installed next to
five single depth
permanent ports.
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table 2. Test Matrix: Analytical Methods, Analytes, Laboratory and Turnaround Times
Matrix
Interior sub-
slab soil gas
Interior sub-
slab soil gas
Sample/Measure-
ment Type
VOCs, Summa
Canister
Radon
Analytical Method
TO-15
Alphaguard according to Protocol for
Using Continuous Radon Monitors
(CR)to Measure Indoor Radon
Concentrations
http://epa.gov/radon/pubs/devprot3.ht
ml#2.1 and EPA 2-56 MOP:
Alphaguard: Operation of the
Alphaguard Portable R
Analytes
Project VOC target list
Radon
Laboratory
Air Toxics Ltd.
Field
Special Turnaround
Time or Interim Data
Analysis Requirements
None
None
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
2.1 Probe Installation and Leak Checks
Single-depth sub-slab ports of both the temporary and permanent type were constructed to test the two
types side-by-side. Locations of the paired ports are shown in Figure 1, with each "SSP" indicating a paired
temporary and permanent sub-slab port. Each temporary sub-slab port was installed within 30 cm of a
single-depth, permanent port.
Temporary single-depth sub-slab probes were constructed in accordance with the SOP in Appendix B of
QAPP Addendum 1, from the NJ VIG. The ports were constructed with flexible tubing (Teflon), and were
sealed with hydrated bentonite using the following method:
• Mix bentonite in a container, starting with water, and then add bentonite until no free water is
present and the mixture has the consistency of gel or modeling clay. Apply the mixture to the top 1-
2 inches of tubing (above the Teflon tape barrier required by the SOP and NJ VIG) as it is twisted
into the drill hole, and as needed at the surface to fill the remaining annular space and make a seal
with the floor and tubing.
For this application, granular bentonite with particle sizes in the medium to fine sand range were used for
fast hydration and easy mixing (Benseal uniform granular Wyoming sodium bentonite (grouting bentonite) -
Halliburton).
During installation, the adherence of the hydrated bentonite to the slab material and to the tubing was
qualitatively noted, as was the apparent continuity of the subsurface portion of the seal as it was installed.
During leak checks and sampling, a reasonable effort was made not to move the sample tubing at the
ground surface, but visible shrinkage and cracking of the seal was observed and documented as incidental
movement of the tubing occurred and as the seal aged and dried out. Photographs of a typical seal as
constructed and of each individual seal after sampling are provided in Appendix A to show the effect of
aging.
Leak checks were performed on each permanent and temporary port using the tracer gas/shroud method
discussed in section 5.3 of the QAPP. Helium gas was used as the tracer. Leak checks were performed
before sampling at each permanent port and before and after sampling at each temporary port.
2.2 Sub-Slab Sampling Using Summa Type Canisters for VOCs
Sample collection methods for both temporary and permanent sub-slab ports were as described in QAPP
Addendum 1. One round of paired samples was collected from each temporary/permanent pair, for VOCs
and radon.
Sub-slab air samples were collected in evacuated, 6-L Summa-type polished canisters. For sub-slab air
sample collection, a brass or stainless NPT to Swagelok union fitting was used to connect vapor probes to a
"T" fitting made of a stainless steel flexible line and an in-line valve. A portable vacuum pump was used to
purge vapor probes and sampling lines for one minute at a flow rate of 0.1 to 0.2 liter per minute (LPM).
Immediately after the in-line valve on the pump end of the "T" fitting was closed, the Summa canister valve
was opened to collect a grab sample at a maximum rate of 0.1 to 0.2 LPM. The larger sizes of Summa
canisters are equipped with an adjustable critical orifice with back pressure regulator that is calibrated at the
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
laboratory for a target fill time of 24 hours. The sampling start and end times are reported in Table 3. The
SOP called for sampling to cease when canister pressure decreased to within 2 to 7 in Hg. Two samples,
temporary port 3 and permanent port SSP 1 were observed to be insufficiently filled after 24 hours and were
continued for 45-48 hours. One ambient background sample was also collected for comparison to the soil
gas samples. Samples from SSP 5 and ambient had gone to 0 in Hg as observed in the field but were
observed to have some small vacuum with a more sensitive gauge upon receipt in the lab. Given that these
two were in relatively low temperature locations and that many of the other canisters were at 3.5 in Hg or
less we expect but cannot prove that these two canisters filled only slightly more rapidly than the canisters
that had small but observable vacuums in the field after 24 hours.
Basement
420 East 28th St
Basement
422 East 28th St
o
Open
Walkway
-4 1-
SSP6
Wall Port 4
f SSP 3
SSP 7 UP*°
1st Floor
H 1 1 1—
Open Open
Walkway Walkway
SSP 5
O
O
SGP 12
r
\
x"V
s
Up to
Wai I Port 1
x SGP 10
*-s no
V . SSP 2
\ ^ ^
i
1st Floor
TC13'
TC 16.5
X
X
Wai
/
Pa
Open
Walkway
Wall Port 2
( Cistern j '
SSP 1V__X
tjSGPS Tcs;MS6-
O o MS 13'
' y ^ MW 3
Open
Walkway
0 OSGP9
SSP 4
>rt 3
Figure 1. Interiors of test buildings, showing soil gas points (SGP) and paired single-depth temporary
and permanent sub-slab ports (SSP).
10
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table 3. Summa Canister Sampling Times and Pressures
Port*
SSP5
Temp port 5
SSP3
Temp port 3
Ambient
SSP 4 Temp
SSP 4 Temp DUP
SSP 1 Temp
SSP1
SSP 2
SSP 2 Temp
SSP 4
SSP 4 DUP
Can*
5619
12011
5766
14008
4338
12687
4181
12669
35245
31442
5738
13345
12940
Flow Controller*
40376
40597
40259
40487
6010
40281
40085
40324
40145
40522
40658
40701
40060
Initial
Vacuum
Recorded In
Field ("Hg)
29
29
29
28.5
28
30
30
29.5
30
30
30
30
30
Final
Vacuum
("Hg)
Recorded
In Field
0
1.5
2.5
2
0
3
1.5
2
3
1.5
2.5
5.5
3.5
Final
Vaccum
Measured at
Lab
0.6 psi
1.5"Hg
2.5 "Hg
1.0 "Hg
0.4 psi
3.5 "Hg
2.0 "Hg
2.5 "Hg
3.0 "Hg
1.5"Hg
2.5 "Hg
5.0 "Hg
3.5 "Hg
Start date
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
11/30/2010
Start time
19:09
19:23
19:34
16:30
19:57
19:48
19:48
19:27
19:27
19:15
19:15
19:48
19:48
End date
12/1/2010
12/1/2010
12/1/2010
12/2/2010
12/1/2010
12/1/2010
12/1/2010
12/1/2010
12/2/2010
12/1/2010
12/1/2010
12/1/2010
12/1/2010
End time
18:53
19:23
18:58
16:34
19:02
19:59
18:48
19:27
16:34
18:40
19:32
21:27
18:48
Sampling
Duration
(hh:mn)
23.44
24:00
23:32
48:02
23:55
24:11
23:00
24:00
45:07
23:25
24:17
25:39
23:00
Date
Received
at Lab
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
12/4/2010
11
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
3. Results
3.1 Seal Integrity
According to the NJ guidance:
"Another method employs a shroud or plastic sheeting placed around the sample probe. An inert tracer gas
(such as helium) is released under the sheeting. The initial soil gas samples (after purging) can be
monitored using field-screening instruments for elevated concentrations (>5%) of the tracer gas (based on
the original tracer gas concentration in the shroud)."
All of the port seals easily passed this leak test criteria when initially installed. However the leak check
integrity of the temporary ports declined with time as measured with the helium shroud test (Table 4). As
shown in Appendix A, all of the bentonite seals had visible dessication cracks when photographed on
December 3rd, several days al
hour sampling period as well.
December 3rd, several days after installation. This cracking was beginning to be visible at the end of the 24
We suspect given the visible cracks in the bentonite, that the ability of the seals to pass the post test leak
test is primarily attributable to the careful use of Teflon tape around the tube as part of the sealing process,
as required by the NJ guidance. It should also be noted that the effectiveness of the seal is not expected to
be solely dependent on the construction methods for the seal. From first principles we would expect that the
seal effectiveness would also be dependent on:
• The air permeability of the subslab soil and the flow rate of sampling
• The degree to which the field staff can hold the tube immobile during the attachment of sampling
equipment and sampling.
• The humidity of the air around the clay portion of the seal and thus the rate of dessication, if several
hours or days are expected to pass between the creation of the seal and the completion of
sampling.
12
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table 4. Leak Test Results for Temporary Probes
(expressed as percentage of detected helium concentration in port of the measured helium
concentration in the shroud)
Location
Temporary SSP 1
Temporary SSP 2
Temporary SSP 3
Temporary SSP 4
Temporary SSP 5
Leak (%)
11/30/2010
(before
sampling,
nondetects
calculated as
0.070
0.126
0.000
0.000
0.000
11/30/2010
(before
sampling,
nondetects
calculated at
0.070
0.126
0.003
0.003
0.003
12/6/2010 (after
sampling)
0.365
0.288
0.174
0.636
0.850
3.2 VOC Sampling Results
The VOCs detected in soil gas at concentrations markedly different than the ambient sample were PCE,
chloroform, and TCE (as shown in Figures 2-4; and in a data table in Appendix B). Results for PCE,
chloroform and TCE were very similar for the paired permanent and temporary sampling ports. The data
set shows considerable spatial variability around the subslab (subbasement) area of the pair of duplexes
studied, demonstrating the utility of collecting multiple subslab samples in even relatively small structures.
The highest concentrations appear in the central and southern portions of the 422 East 28th St. side of the
duplex.
If:
• only the data set from only the temporary ports were used, OR
• the data set from only the permanent ports were used,
it is highly likely that a practitioner would have reached the same site management decision using either
data set.
{Note: We Are looking at statistical comparisons between the paired samples to show that there was no
significant difference.)
13
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
180
120
100
40
20
SSP1
Results for PCE Temporary vs Permanent Probes
I Permanent
Temporary
SSP2
SSP 3 SSP 4
Sampling Locations
SSP4DUP
SSP 5
Figure 2. PCE concentrations in soil gas from temporary and permanent subslab probes.
14
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
10
Results for Chloroform Temporary vs Permanent Probes
• Permanent
Temporary
SSP1
SSP2
SSP 3 SSP 4
Sampling Locations
SSP4DUP
SSP 5
Figure 3. Chloroform concentrations in soil gas from temporary and permanent subslab probes.
15
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Results for TCE Temporary vs Permanent Probes
• Permanent
Temporary
SSPl SSP2 SSP3 SSP4 SSP4DUP SSP5
Sampling Locations
Figure 4. TCE concentration in soil gas from temporary and permanent subslab probes.
3.3 Radon Results
As shown in Table 5, there was also relatively good agreement between short term field radon
measurements in soil gas made after the VOC sampling in both the temporary and permanent ports, except
in SSP3. The radon concentration in temporary port SSP3 was substantially higher than in the permanent
port at that location. Variability between short term field radon measurements in soil gas made before and
after the VOC sampling was also greatest in location SSP3. This may suggest that port SSP3 is located at
an area with a sharp gradient in radon concentrations over a small area.
3.4 Study Limitations
We would like to caution about several limitations of our conclusions:
• This study was performed at only one site. Based on first principles, the demands on seal
performance are likely to be greatest in subslab sample ports in structures constructed directly on
low permeability soils.
16
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
The temporary ports installed in this study were under the direct field supervision of a highly
experienced staff member who has installed numerous ports over several years. We did not
make any attempt to evaluate the variability in seal quality of ports installed by workers of varying
levels of experience.
This study did not examine whether the repetitive sampling over many months, which is a primary
purpose of permanent sample port installation, would have yielded a different result than one
time use of either permanent or temporary ports.
This study evaluated only one seal material, bentonite clay. The NJ guidance quoted allows a
number of different materials "modeling clay, beeswax or other non-volatile emitting and non-
shrinking materials" to be used with Teflon tape "to create a snug fit when the tubing is twisted
into the hole". The NY state guidance (2006) includes a somewhat different list of permissible
seal materials "the implant should be sealed to the surface with non-VOC-containing and
nonshrinking products for temporary installations (e.g., permagum grout, melted beeswax, putty,
etc.) or cement for permanent installations." We are not aware of any studies that have
compared the seal quality that can be achieved with these different materials to one another.
We caution that the term "modeling clay" used in the NJ guidance to describe an acceptable seal
material is commonly used for a very broad range of product formulations used for a common
artistic and educational purpose. The term can be used for at least four different types of
materials 1) products composed primarily of natural mined clay minerals; 2) products produced by
combining oils, waxes and clay minerals; 3) those made entirely of organic polymers and 4) those
produced dough of flour, cornstarch, oil, water and cream of tartar1. The organic polymer clays
include those primarily composed of polyvinylchloride for example2. These have been reported to
potentially contain residual vinyl chloride (Stopford, W. 2000).
1 Three websites accessed December 18, 2010 http://en.wikipedia.org/wiki/Modellinq clay
http://www.wiseqeek.com/what-are-the-different-types-of-modelinq-clav.htm
http://www.clavsculptinq.org/modelinq-clav/
2http://cdn.dickblick.com/msds/DBH_33901XXXX.pdf
17
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table 5. Radon Data Comparing Temporary and Permanent Probes
Location
SSP 1
SSP2
SSP 3
SSP 4
SSP 5
Radon 11/30/10
Before VOC
Sampling
Permanent Port
pCi/L
1068
1203
219
1865
1089
Radon 12/6/2010, After VOC
Sampling
Temporary Port
pCi/L
735
1108
1151
1659
1214
Permanent Port
pCi/L
719
1338
543
1708
1089
4. Conclusions
Under the conditions studied here VOC and radon concentrations measured simultaneously in soil gas
using nearly collocated temporary and permanent ports appeared to be independent of the type of port.
The variability between nearly collocated temporary and permanent ports was much less than the spatial
variability between different locations within the same residential duplex. The agreement of concentrations
was achieved even though the clay portion of the seal of the temporary ports visibly desiccated and
cracked. Post sampling leak test results suggested that this desiccation and cracking was not as
detrimental to port seal performance as would have been expected, suggesting that the Teflon tape portion
of the seals was serving an important function. Post sampling leak tests are advisable (in addition to
presampling leak tests) when temporary ports are used to collect a time integrated sample over a period of
several hours.
These results suggest that temporary subslab sampling ports can provide data equivalent to that collected
from a permanent subslab sampling port at the same time. However we caution that (1) we tested only one
type of seal material in one location, (2) the seals were installed by experts and rigorous quality control, and
thus (3) these results may not apply to all types of temporary seals and all building foundations.
5. References
New Jersey Department of Environmental Protection (NJDEP). 2005. Vapor Intrusion Guidance.
http://www.ni.qov/dep/srp/quidance/vaporintrusion/viq.htm)
New York State Department of Health (NYSDOH). 2006. Guidance for Evaluating Soil Vapor Intrusionin the
State of New York. Center for Environmental Health, Bureau of Environmental Exposure
Investigation. Albany, NY. October.
US EPA (Environmental Protection Agency). 2007. Construction and Installation of Permanent Sub-Slab
Soil Gas Wells. SOP 2082. Response Engineering and Analytical Contract.
18
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
U.S. EPA (Environmental Protection Agency). 1999. Compendium of Methods for the Determination of
Toxic Organic Compounds in Ambient Air Second Edition Compendium Method TO-15
Determination Of Volatile Organic Compounds (VOCs) in Air Collected in Specially-Prepared
Canisters and Analyzed by Gas Chromatography Mass Spectrometry (GC/MS). EPA/625/R-
96/01 Ob. At http://www.epa.gov/ttnamti1/files/ambient/airtox/to-15r.pdf.
Woodard, S. 2010. Hazard Risk Assessment from the Use of Polymer Clays. Division of Occupational and
Environmental Medicine, Duke University Medical Center; Durham, NC 27710;
http://www.polvmerclavcentral.com/cvclopedia/polymerclav safety.htm May 2000. Accessed
December 18, 2010.
19
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Appendix A: Photographs of Seals
Typical Temporary Port Construction Before Sampling
20
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Typical Port Construction Before Sampling, One Temporary port shown at lower end of picture and one
Permanent port shown in the middle of the frame
21
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 1
22
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 1 close up
23
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Temp Port 2 (129 KB)
24
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Temp Port 2 close up
25
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 3
26
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Temp Port 3 close up
27
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Temp Port 4
28
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 4 close up
29
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 5
30
-------
Initial Draft- 12/21/10 - Do Not Cite or Quote
Temp Port 5 close up
31
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Port 5 additional close up
32
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Effect of tubing movement on bentonite seal after drying has occurred
33
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
34
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Appendix B: Data Summary Table
Temporary vs. Permanent Port Study
Summa Canister VOCs Results (showing only VOCs detected or identified at least
once)
COMPOUND
NAME
Sampling Location
RESULTS (ug/m3)
Qualifier
REPLMT (ug/m3)
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2Temp
SSP-3
SSP-3Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
0.74
0.47
0.46
1.9
2.0
0.89
1.0
1.3
1.6
1.3
1.3
1.4
1.4
1.4
1.3
1.7
J
ND
ND
ND
J
ND
J
ND
ND
J
J
ND
J
J
J
J
J
J
J
J
J
J
J
2.1
2.4
2.3
2.2
2.3
2.3
2.2
2.6
2.4
2.3
2.4
2.1
2.2
2.0
2.3
2.3
2.2
2.3
2.3
2.2
2.5
2.4
2.2
35
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Methylene Chloride
Methylene Chloride
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2Temp
SSP-3
SSP-3Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
1.6
1.1
1.6
66
69
4.2
4.8
13
12
19
18
0.55
1.0
2.2
1.9
1.3
2.1
J
J
J
ND
ND
ND
J
ND
ND
ND
ND
ND
J
ND
ND
ND
ND
J
J
ND
J
J
ND
2.4
2.0
2.2
3.2
3.6
3.6
3.4
3.6
3.6
3.4
3.9
3.7
3.5
3.7
3.1
3.4
2.3
2.6
2.6
2.5
2.6
2.6
2.4
2.8
2.7
2.5
2.7
2.3
2.5
2.2
2.6
36
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Tetrachloroethene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
SSP-1 Temp
SSP-2
SSP-2Temp
SSP-3
SSP-3Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
2.1
1.2
2.7
0.50
150
150
3.7
4.8
16
23
140
140
160
160
27
23
0.91
0.84
1.1
1.8
J
J
ND
ND
ND
ND
ND
ND
ND
J
ND
J
J
J
ND
J
ND
J
ND
J
2.5
2.4
2.5
2.5
2.4
2.8
2.6
2.5
2.6
2.2
2.4
4.4
5.0
5.0
4.8
5.0
5.0
4.7
5.5
5.2
4.9
5.2
4.4
4.8
2.4
2.8
2.8
2.6
2.8
2.8
2.6
37
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
Trichloroethene
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2Temp
SSP-3
SSP-3Temp
SSP-4
SSP-4 Dup
SSP4 Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
5.9
5.2
2.0
7.0
7.3
1.2
ND
ND
ND
J
ND
ND
ND
ND
ND
ND
ND
J
ND
ND
ND
3.0
2.9
2.7
2.9
2.4
2.6
3.5
4.0
3.9
3.8
3.9
3.9
3.7
4.3
4.1
3.9
4.1
3.5
3.8
38
-------
Initial Draft - 12/21/10 - Do Not Cite or Quote
Table . Temporary vs Permanent Port Study
Summa Canister VOCs Results (showing only VOCs detected or identified at least once)
RESULTS REPLMT
COMPOUND NAME Sampling Location (ug/m3) Qualifier (ug/m3)
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Carbon Disulfide
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Chloroform
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Hexane
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Methylene Chloride
Tetrachloroethene
Tfitranhlnrnethfine
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
SSP-1 Temp
SSP-2
SSP-2 Temp
SSP-3
SSP-3 Temp
SSP-4
SSP-4 Dup
SSP4Temp Dup
SSP-4 Temporary
SSP-5
SSP-5 Temporary
Ambient Outdoor 422 Back Porch
SSP-1
0.74
0.47
0.46
1.9
2.0
0.89
1.0
1.3
1.6
1.3
1.3
1.4
1.4
1.4
1.3
1.7
1.6
1.1
1.6
66
69
4.2
4.8
13
12
19
18
0.55
1.0
2.2
1.9
1.3
2.1
2.1
1.2
2.7
0.50
150
J
ND
ND
ND
J
ND
J
ND
ND
J
J
ND
J
J
J
J
J
J
J
J
J
J
J
J
J
J
ND
ND
ND
J
ND
ND
ND
ND
ND
J
ND
ND
ND
ND
J
J
ND
J
J
ND
J
J
ND
ND
ND
ND
ND
ND
ND
J
ND
2.1
2.4
2.3
2.2
2.3
2.3
2.2
2.6
2.4
2.3
2.4
2.1
2.2
2.0
2.3
2.3
2.2
2.3
2.3
2.2
2.5
2.4
2.2
2.4
2.0
2.2
3.2
3.6
3.6
3.4
3.6
3.6
3.4
3.9
3.7
3.5
3.7
3.1
3.4
2.3
2.6
2.6
2.5
2.6
2.6
2.4
2.8
2.7
2.5
2.7
2.3
2.5
2.2
2.6
2.5
2.4
2.5
2.5
2.4
2.8
2.6
2.5
2.6
2.2
2.4
4.4
50
39
-------
Appendix C
Additional Correlograms for Ambient and Indoor Air
-------
Appendix C—Additional Correlograms for Ambient and Indoor Air
Chloroform and Radon - Air samples, Location =AMBIENT Tetrachloroethene and Radon -Air samples, Location =AMBIENT
o
Lag
0
Lag
Chloroform and Radon - Air samples,
Location =420BaseN
Tetrachloroethene and Radon -Air samples,
Location =420BaseH
13 d
. •
n ti id
Mill Illliill II
-20 -10 0
Lag
10 20
E
1 S -
~m o
8 0
O ^
d
I L
II 1 i ill
-20 -10
|
0
Lag
||
||
III
III
10
In
1
20
Chloroform and Radon - Air samples,
Location =420BaseS
Tetrachloroethene and Radon -Air samples,
Location =420BaseS
f
|
1
1
1 1
1 i 1 i , 1
r-i
E d "
c
a.
en
p
9
-15 -10
0
Lag
10 15
-15 -10 -5
0
Lag
10 15
C-2
-------
Appendix C—Additional Correlograms for Ambient and Indoor Air
Chloroform and Radon - Air samples,
Location =422BaseH
o
e o
8 °
m
e **
o o
-i-u -i-L .' - -l-i.
rhi i iiii
>i II II 1 INI
-
Tetrachloroethene and Radon -Air samples,
Location =422BaseN
-
JJJL
i i
i i
-15 -10 -E 0 E 10
Lag
Chloroform and Radon - Air samples,
Location =422BaseS
E
-------
Appendix C—Additional Correlograms for Ambient and Indoor Air
Chloroform and Radon -Air sampletrachloroethene and Radon -Air sar
Location =AMBIENT Location =AMBIENT
OJ
o
E
c
o
ra
£ °
o o
to
2
O
OJ
o
-15
-5 0
10
CO
o
OJ
o
E
c
o
_
0;
o
o
-------
-------
&EPA
United States
Environmental Protection
Agency
Office of Research
and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
EPA/600/R-12/673
September 2012
www.epa.gov
Please make all necessary changes on the below label, detach
or copy and return to the address in the upper left hand corner.
If you do not wish to receive these reports CHECK HERE I I:
detach, or copy this cover, and return to the address in the
upper left hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA PERMIT No. G-35
Recy cl ed /Recy cl able
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |