Passive Soil Gas Surveys - Mass to Concentration Tie-In Brochure

Proceedings of the AW&MA Vapor Intrusion 2009 Conference San Diego, California January 27-30, 2009 Mass to Concentration Tie-In for Passive Soil Gas Surveys: Improved Technique for Source Area, Spatial Variability and Vapor Intrusion Assessment Joseph E. Odencrantz, Ph.D., P.E.1, Harry O’Neill2 and Paul C. Johnson, Ph.D.3 1 Beacon Environmental Services, Inc. 2121 Yacht Yankee, Newport Beach, CA 92660 2 Beacon Environmental Services, Inc., 323 Williams Street, Bel Air, MD 21014 3 Arizona State University, Brickyard BY640, 699 S. Mill Avenue, Tempe, AZ 85281 ABSTRACT The mass to concentration tie-in [MtoC Tie-In] correlates passive soil gas (PSG) data in mass to active soil gas data in concentration determined by EPA Method TO-17. Passive soil gas surveys consist of rapid deployment of hydrophobic sorbents (dozens to several hundred locations typically installed in one day) to a depth of six-inches to three-feet in a grid pattern with exposure in the field from three days to two weeks to target a wide variety of organic compounds. A power function is used on a compound-to-compound basis to correlate spatially varying mass (nanograms) from selected locations within a passive soil gas survey to concentration (ug/m3) at those same locations. The correlation from selected PSG locations is applied to the remainder of the PSG grid. The MtoC Tie-In correlations provides added value to a PSG survey, with the PSG data then used to estimate risk throughout the limits of the investigation for quantitative assessment. The results from a site in Northern California show the MtoC Tie-In correlations for both Benzene and Total Petroleum Hydrocarbons (TPH). The correlations are applied on a compound-to-compound basis to the remaining locations in the PSG-grid to provide an estimate of concentration that can be used for comparison to risk/screening levels or fate and transport diagnostic tools (partitioning equations, solubility laws, etc.). An example of how the correlations are applied is presented in tabular form. The results from a chlorinated solvent survey show the MtoC Tie-In correlation from a site in Maryland for Tetrachloroethylene (PCE). In this instance, there was a near-perfect relationship between the PSG mass and the active soil gas concentration (R-squared value of 1). The concentration estimated throughout a PSG-grid enables a vast new realm of interpretive power at sites. Several other sites are discussed. INTRODUCTION The MtoC Tie-In correlation relates passive soil gas (PSG) data in mass to active soil gas data in concentration determined by EPA Method TO-17 or TO-15. To obtain quality data, passive soil gas samples must be analyzed following EPA Method 8260B, or equivalent, that includes internal standards and surrogates with all analyses. A power function is used on a compound-to-compound basis to correlate mass (ng) from selected locations within a passive soil gas survey to concentration (ug/m3) at those same locations. The correlation from selected PSG locations is applied to the remainder of the PSG grid. The MtoC Tie-In correlation provides added value to a PSG survey and the PSG data can then be used to estimate risk throughout the limits of the investigation with a concentration array. The MtoC Tie-In brings PSG testing to a level III/IV tool for quantitative assessment as opposed to a screening tool/semi-quantitative method as defined by the EPA (1988). BACKGROUND The interpretation of time-integrated PSG surveys can be augmented whereby select locations at a site are chosen where active soil gas measurements can be used to develop a relationship between passive soil gas mass and active soil gas concentration. This is termed “mass-to-concentration tie-in”. At the completion of a PSG survey, select locations where PSG samples were obtained are developed into a shallow active soil gas vapor ports, e.g. three to five locales for surveys less than 50 PSG locations and 5-10% for surveys over 100 PSG locations. The tie-in locations should be chosen such that there is adequate distribution across the mass ranges encountered on a COC basis from the PSG results and the active soil gas sampling times should be determined in accordance with the detection limit, that is, the mass found at the PSG location and the sampling volume to be used for the active sampling. Upon completion of the active soil gas sampling at the select locations of the PSG survey, a correlation can be established between PSG mass and active soil gas concentrations that can be applied to the remaining PSG locations across the site. For example, for a PSG survey that consisted of 100 locations and the tie-in correlations were establish at five of these locations, the established correlation between mass and concentration can be applied at the remaining 95 locations to provide a concentration estimate on a COC-basis. These concentration estimates can be used to (i) determine regions across the PSG grid where additional investigation is needed, (ii) establish regions across the PSG grid where there are either exceedences in regulatory screening levels or risk-based areas of concern and (iii) assess areas in the PSG grid where vapor intrusion issues are of concern. The benefits of using the tie-in correlations across the PSG grid instead of collecting only active soil gas measurements are that a large number of sample locations can be simultaneously and rapidly measured to provide concentration estimates over a large area with both minimum disruption/disturbance at a site and in a cost effective, efficient manner. MtoC Relationship for Chlorinated Compounds at a Single PSG Location An example of the relationship between PSG mass and active soil gas concentration is shown by the following plot. Figure 1 shows the correlation between PSG mass and active soil gas concentration for four compounds at one location. In this instance the most impacted PSG locale, from an array of 22 samplers at the site, was chosen for follow-up sampling using EPA Method TO-15. The location was Indiana and the investigation took place in November 2003 at a former dry cleaning facility inside the building. The PSG samplers were exposed for six days in the shallow vadose zone and were deployed to a depth of 14-16 inches below grade. An active soil gas sample was collected through the concrete slab with a vapor probe set to a depth of 18 inches. The active sample was collected over a period of 24 hours in a six-liter canister. y = 1.3368x1.1896R2 = 0.99771001,00010,000100,0001,000,000100 1,000 10,000 100,000PSG (ng)TO-15 (ug/m3) Figure 1. Mass to concentration relationship for four chlorinated compounds at a site in Indiana The relationship between the passive and active data across the four compounds is very strong with nearly a perfect correlation factor. The time-integrated passive soil gas mass collected over a six-day period is related to a shorter-term active soil gas sample concentration through the correlation above. The utility of PSG as an investigatory tool is greatly enhanced by relating active soil gas concentration to the array of deployed samplers. MtoC Relationship for Chlorinated Compounds at Multiple PSG Locations An example of the relationship between PSG mass and active soil gas concentrations at three different locations at site in the eastern United States demonstrates the spatial strength of the correlations. Thirty-seven PSG samplers were installed for a period of seven days in the subsurface at this site. Three PSG locales were chosen for follow-up sampling using active sampling via EPA Method TO-17 with 30 min soil gas samples taken at a flowrate of 20 mL/min. The site location was in northern Maryland and the investigation took place in September 2007 at a former dry cleaning facility with samples collected inside the building through the concrete slab. Figure 2 shows tetrachloroethylene (PCE) results for both passive and active samples at three PSG locations. y = 3.5778x0.8223R2 = 0.99161001,00010,000100 1,000 10,000PSG (ng)TO-17 (ug/m3) Figure 2. Mass to concentration relationship for PCE at three locations at a site in Maryland. PCE TCE cis-DCE trans-DCE The relationship between the time-integrated PSG and active soil gas data is nearly perfect. The spatial characteristics between active and passive soil gas sampling is impressive at this site. The three tie-in points are approximately twenty feet apart in a triangular configuration. The correlation was applied at other parts of the site where the PSG survey revealed other significant quantities of PCE. The active soil gas probes were installed to a depth of eighteen inches and contained a sand pack. More on the typical soil gas probe installation is provided below. At each tie-in location the field technician drilled three holes through the slab. Each hole was drilled using a hammer drill with a 1-½”-diameter drill bit to a depth of 18 inches. A ¼”-diameter length of Teflon® tubing was lowered down the hole until it was approximately one-inch from the bottom of the hole. The tubing had a screen implant inserted in the lower end of the tubing to prevent the uptake of particle matter. The end of the tubing extending above slab had a stainless steel valve that was swaged to it prior to sampling. After the tubing was lowered into the hole, dry, clean silica was slowly poured into the hole until reaching the lower level of the concrete slab. A layer of granular bentonite, approximately 1-inch thick, was then poured on top of the silica and moistened by adding distilled water. This process was repeated until the hydrated bentonite filled the annular space of the hole to grade. The hydrated bentonite was then compacted around the tubing and finished once again to grade. Using a 60 mL syringe, the sampling hole and train was slowly purged of three soil gas pore volumes and the valve was left in the closed position. MASS TO CONCENTRATION TIE-IN: PETROLEUM HYDROCARBON SITE The following results from a site in Northern California show the MtoC Tie-In correlations for both Benzene and Total Petroleum Hydrocarbons (TPH) from C5 to C10. Thirty-six time-integrated PSG samplers were exposed soil gas at shallow depths for seven days and three of those locations, inside a mini-mart, were chosen to conduct the follow-up active soil gas testing as tie-in points, with 20-30 min samples at a flowrate of 20 mL/min. The three tie-in points are approximately twenty feet apart in a triangular configuration. Figure 3 shows the MtoC correlations for benzene and TPH C5-C10. Figure 3. Benzene and TPH MtoC relationships from a service station site in northern California. TPH C5-C10 y = 0.001x1.4821R2 = 0.937511,0001,000,0001,000 10,000 100,000 1,000,000PSG (ng)Method TO-17 (ug/m3)Benzene y = 0.0904x1.3919R2 = 0.92551101001,00010,00010 100 1,000 10,000PSG (ng)Method TO-17 (ug/m3) The correlations are applied on a compound-to-compound basis to the remaining locations in the PSG-grid to provide an estimate of concentration that can be used for comparison to risk/screening levels or fate and transport diagnostic tools (partitioning equations, solubility laws, etc.). An example of how the correlations are applied are illustrated in the following table whereby Benzene and TPH concentrations are tabulated from the results of the PSG survey and the correlations are applied throughout the remainder of the PSG-grid. The bolded values indicate values that are above published regulatory screening levels. Table 1. Application of Tie-In Correlation to Portion of PSG Grid A second site in northern California with similar conditions yielded the following correlations and power functions. For the sake of brevity, the graphs are not shown. An example of how to apply the developed correlations is provided below. Benzene Active = 0.8682 x PSG (ng)1.0889 R-squared 0.8801 [PSG range 17-14,698 ng and TO-17 range 15.9-62,535. ug/m3] Example at a PSG value of 3,000 ng results in 5,307. ug/m3 [0.8682 x PSG (3,000.)1.0889] TPH C5-C10 Active = 0.0007 x PSG (ng)1.5946 R-squared 0.9409 [PSG range 2,500.-612,980 ng and TO-17 range 151.-804,083. ug/m3] Example at a PSG value of 30,000 ng results in 9,645. ug/m3 [0.0007 x PSG (30,000)1.5946] MASS TO CONCENTRATION TIE-IN: MIXED PETROLEUM HYDROCARBON AND CHLORINATED SOLVENT SITE A passive soil gas survey was conducted at a site in northern California with an active remediation system that consists of dual-phased extraction for soil vapor and groundwater. The purpose of conducting a time-integrated PSG survey was to characterize the soil gas regime under the slab of an operating business and to screen areas for risk evaluation/further investigation. Upon completion of the PGS survey, active soil gas sampling at five locations was performed to develop an MtoC correlation (four samples at ten min. and one sample for thirty min. at a flowrate of 20 mL/min). The MtoC correlation was used to quantify vapor intrusion risk Units SV-14 SV-15 SV-16 SV-17ng 259 28 2,160 2,425ug/m3 207 9 3,957 4,649ng 58,033 37,313 325,988 460,879ug/m3 11,488 5,970 148,291 247,743BenzeneTPH C5-C10 and to identify locations of future remediation and investigation efforts. Figure 4 shows the five locations of the tie-in points along with a site map containing the locations of soil borings, groundwater wells and the location of the treatment plant. The PSG samplers remained in the subsurface for eleven days. Figure 5 shows the spatial distribution from the PSG survey for BTEX and PCE side by side. Figure 4. Tie-in points for MtoC correlations at an active site in northern California. BTEX PCE Figure 5. PSG survey spatial variability summary for BTEX and PCE. Units of ng. The time-integrated PSG survey for BTEX revealed there are sources of petroleum that are beyond the influence of the current treatment system. In other words, the previous discrete sampling did not reveal the complete picture of the source areas. As an example, there is a total BTEX mass of 198,552 nanograms at PV-8. Figure 6 shows the MtoC correlations for both BTEX and PCE from the five tie-in points at the site. BTEXy = 1.6682x1.2367R2 = 0.89751001,00010,000100,0001,000,00010,000,000100 1,000 10,000 100,000 1,000,000PSG (ng)Method TO-17 (ug/m3)PCEy = 1.4027x0.9637R2 = 0.75341001,00010,000100,000100 1,000 10,000PSG (ng)Method TO-17 (ug/m3) Figure 6. MtoC correlations for PCE and TPH at site in northern California. As an example of how to use the MtoC correlations, the active soil gas concentration at PV-1 for BTEX is 8,727. ug/m3. All the other non-tie-in PSG locations could also have concentrations developed through the correlations. The discovery of a PCE soil gas plume at the site sourced at PV-9 is a significant finding. The PCE MtoC correlation can be applied at PV-9 and the active soil gas estimate there is 7,487. ug/m3. Once again , the other non-tie-in PSG locations concentration can be estimated through application of the correlation. The MtoC correlations, developed for seven compounds at this same site, revealed there is clearly a trend between the range of active soil gas concentrations encountered from the tie-in points and the strength of the correlations. Figure 7 is a plot of the range of active soil gas concentrations encountered for each compound versus the strength of the MtoC correlation. y = 0.2435Ln(x) + 0.5899R2 = 0.830700.20.40.60.810 1 2 3 4 5Order of Magnitude TO-17 VariabilityMtoC Correlation Coefficient Figure 7. Relationship between MtoC correlation and soil gas concentration range for the tie-in points. PCE TCA TCE Benzene BTEX TMB 1,2,4 MTBE The logarithmic line of best fit reveals two orders of magnitude variability of concentration of any compound results in an R-squared of 0.759 which is very strong and significant. Similarly, at one order of magnitude active soil gas concentration variability results in a correlation coefficient of 0.590 which is suitable for estimations within order of magnitude variability. LIMITATIONS OF USING MODELS AND EQUATIONS FROM ESTIMATING ACTIVE SOIL GAS CONCENTRATIONS FROM PASSIVE SOIL GAS DATA Passive soil gas surveys have historically been labeled a “site screening tool” for finding sources of volatile organic compounds. The MtoC tie-in concept enables a vast new realm of interpretive power by applying passive soil gas survey results with the developed correlations for risk evaluation, vapor intrusion assessment, source identification and identification of regions that require additional investigation. One of the important assets of using the correlation is that the active soil gas concentrations are measured in the field as opposed to the application of an equation or model with several independent variables. The authors emphasize this point for there are several limitations to the use of equations and models that are discussed in some detail below. The application of models or equations to convert passive soil gas data to active soil gas concentrations will always be questioned because the determination of independent variables is problematic as are the underlying assumptions imbedded with the equations/models. Models provide predictive capability, but their reliability is difficult to defend with limited site-specific data (ITRC, 2007). The use of models and equations may be useful at the very preliminary stages of an investigation; however, their predictions will be superseded by the collection of site-specific data. The use of extreme and non-representative assumptions is the most common weakness of predictive modeling (ITRC, 2007). When models and equations are used for the determination of active soil gas concentration estimates, an additional layer of regulatory hurdles will be necessary, that are, but are not limited to: a sensitivity analysis, parameter estimation difficulties, overcoming skepticism and conceptual model issues. The use of equations or models to determine active soil gas concentrations from passive soil gas survey results presents challenges and direct field measurements are the preferred option. Therefore, applying a time-integrated passive soil gas survey with MtoC Tie-In points allows one to provide concentration data while increasing the number of locations that can be rapidly sampled without the need for regulators or pumps and avoiding the challenges associated with collecting active soil gas data. CONCLUSIONS The MtoC Tie-In correlations can be implemented at sites at various stages of investigation and/or remediation to estimate concentrations throughout a PSG-grid that enables a vast new realm of interpretive power by applying passive soil gas surveys. The following summarize the conclusions from this paper. • The MtoC correlations provide a useful and time-integrated tool to assess site risks and to determine regions of vapor intrusion concerns. • The MtoC Tie-In results have been used successfully to estimate concentrations at PSG locations for the express purpose of comparison to regulatory screening levels and to assess potential areas for active remediation under regulatory oversight. • The MtoC Tie-In correlations developed at selected points within a PSG-grid are strong on a compound-to-compound basis for organic compounds typically encountered at petroleum hydrocarbon and chlorinated solvent impacted sites. • The MtoC Tie-In results can be used for fate and transport diagnostic tools (partitioning equations, solubility laws, etc.) in order to determine areas for additional investigation, zones of biodegradation and to assess natural attenuation factors. • The MtoC correlations are established using a power function to fit PSG mass and TO-17 active soil gas concentration. • There is a strong relationship between the variability of measured soil gas concentration from the MtoC tie-in points and the R-squared (goodness of fit). The applications of the fitted logarithmic equation reveals two orders of magnitude variability of concentrations of a compound results in an r-squared of 0.759. • The use of equations or models to determine active soil gas concentrations from passive soil gas survey results presents challenges and direct field measurements are the preferred option. REFERENCES 1. EPA. Field Screening Methods Catalogue-User`s Guide. Developed by the United States Environmental Protection Agency, Office of Emergency Response and Remedial Actions, EPA/540/2-88/005. 1988. 128 pp. 2. EPA. 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, Center for Environmental Research Information Office of Research and Development U.S. Environmental Protection Agency, Cincinnati, OH 45268. 1999. 3. Interstate Technology and Regulatory Council. Vapor Intrusion Pathway: A Practical Guideline. 2007. http://www.itrcweb.org/gd_VI.asp 4. U.S. EPA Draft Guidance for Evaluating the Vapor Intrusion to Indoor Air Pathway from Groundwater and Soil. 2002. http://www.epa.gov/correctiveaction/eis/vapor/complete.pdf KEY WORDS Passive soil gas sampling; vapor intrusion evaluation; risk assessment; mass to concentration tie-in correlation; sorbent tube; time-integrated sampling; U.S EPA; TO-17; Beacon Environmental Services, Inc.; source area assessment: spatial variability determination: site-specific correlation