Department of Health Seal

TGM for the Implementation of the Hawai'i State Contingency Plan
Section 7.5
COLLECTION OF REPRESENTATIVE SOIL VAPOR SAMPLES

7.5 COLLECTION OF REPRESENTATIVE SOIL VAPOR SAMPLES

The basic concepts of site investigation design and data collection and interpretation discussed in Section 3 and Section 4 of this guidance apply to soil vapor and indoor air as well as of soil and sediment. A systematic approach (Section 3.3) should be used to identify potential environmental issues of concern develop a preliminary, conceptual site model (CSM), designate “Decision Units (DUs)” for sample collection with specific, investigation objectives and decision statements, collect and interpret sample data, refine the CSM and recommend additional investigation and/or remedial action as appropriate. This step-by-step approach to data collection should be documented in the work plan prepared for the site investigation.

The interpretation of soil vapor and indoor air data is discussed in Section 7.14. It is important that data be collected in a manner that is reflective of the investigation questions being asked. Data for traditional, small-volume samples collected from single points can be compared to HDOH subslab vapor action levels for initial identification of large-scale, subsurface vapor plumes that could pose potential vapor intrusion concerns (HDOH 2017b; see also Section 7.14.1). Random, small-scale variability of VOC concentrations within a vapor plume (e.g., at the scale of a one-liter vapor sample) can, however, lead to erroneous estimates of vapor plume boundaries (“false negatives”) and estimation of vapor intrusion risk (Section 13.2; see also Brewer et al. 2014). Small-scale, “hot spot” and “cold spot” VOC patterns based on single samples within a large vapor plume can likewise be artifacts of random variability and very misleading of actual site conditions.

Although useful for initial screening purposes, HDOH soil vapor action levels for vapor intrusion apply to the mean or “true” concentration of a targeted VOC for the total volume of vapor anticipated to intrude a building over several years. Comparison of deep soil vapor data or data for small-volume samples collected immediately beneath a slab can useful for initial screening purposes but is not strictly appropriate for evaluation of vapor intrusion risk (refer to Section 13.2). This is similar to limitations on comparison of small-volume, “discrete” soil sample data to HDOH EALs rather than data for large-mass samples collected from well-thought-out, targeted DUs (Section 4.3; see also Brewer et al. 2017a, b).

Assumptions regarding the representativeness of small-volume vapor sample data should be used with other lines of evidence to assess long-term, vapor intrusion risk. This includes the nature of known releases, soil data (preferably MIS; Section 4), groundwater data and Large Volume Purge (LVP) vapor data collected directly beneath the slab, as discussed below and in Section 7.4.

A more systematic and well-thought-out sample collection approach similar to that used to assess the risk posed by contaminated soil is required to reliably assess vapor intrusion risk (see Section 3.4). The results of the Site Scoping (Section 3.1) and Systematic Planning process (Section 3.2) should be used to designate slab areas within the subject building for subslab vapor sample collection. This could include testing of subslab vapors:

  • Beneath known or hypothetical, vapor entry points;
  • Within known or suspect subslab utility trenches that could serve as preferential pathways for vapor flow;
  • Above suspect soil or groundwater source areas;
  • Beneath areas of the building with high-risk usage (e.g., daycare center) or, in the absence of other information;
  • Beneath the center of the building slab or other potential vapor accumulation areas beneath the slab.

An indoor air study (Section 7.7) can in some cases be used to identify the general location of vapor entry points. In most cases, however, vapor entry points (if present) are rarely known during the initial stages of an investigation. As an alternative, subslab vapor samples for assessment of vapor intrusion risk are typically drawn from hypothetical entry points in worst-case areas of the slab.

A risk-based DU volume of soil vapor should be designated for sample collection, similar to the approach used for characterization of soil. The objective of sample collection is to estimate the true (“mean”) concentration of targeted VOCs within this targeted volume of vapor. A default, DU vapor volume of 3,000 liters is recommended for use in Hawai´i. This represents the volume of subslab vapor assumed to intrude a building through a single gap in a floor over a one-day period, based on a daily vapor entry rate of 2 L/minute. The same vapor entry rate is used to calculate subslab soil vapor action levels presented in the HDOH EAL guidance (HDOH 2017a) and is predicted to be appropriate for tropical climates (Brewer et al. 2014). Larger subslab vapor DU volumes are appropriate for non-tropical climates due to potentially higher, vapor entry rates during periods when a building is being heated. The following vapor entry rates are estimated for different climate zones by Brewer et al. (2014):

Climate Region Average
Cooling Days
per Year
Average
Neutral or
Heating Days
per Year
Estimated
Annual-Average
Vapor Entry Rate
(L/min)
Estimated
Annual-Average
Vapor Entry Rate
(L/day)
Cold 62 303 4.5 6,466
Warm 122 243 4.0 5,756
Mediterranean 199 166 3.4 4,845
Tropical 365 0 2.0 2,880

The daily vapor entry rate estimated for tropical climate zones, rounded to 3,000 liters/day, was referenced for use as a default subslab vapor volume for LVP sample collection. Note that an LVP DU volume of 7,000 liters was used in the 2016 HDOH field study of LVP sample collection (HDOH 2017c), discussed in Section 7.8.5. This is excessively large for use in Hawai´i but might be appropriate for the collection of LVP samples in cold climate zones.

As a default, subslab vapor in soil or fill material within 15-25cm of the building slab should be targeted for sample collection. For example, screened, vapor extraction points might be installed to a depth of 15cm beneath the slab and a series of 3,000-liter, LVP vapor samples collected.

Characterization of targeted vapor DUs could in theory be accomplished by collection of an “adequate” number of small-volume vapor sample points. The number of samples required to obtain a representative concentration of the targeted VOCs in the vapor and use of the resulting data to assess vapor intrusion risk is uncertain, however. Use of individual data points is not strictly appropriate, since the small volume of vapors represented cannot be assumed to represent vapors intruding the building or even the general concentration of VOCs in vapors in the immediately surrounding area (see Section 13.2). Statistical analysis can be used to estimate a mean concentration for a set of small-volume, vapor sample data points, but the total volume of vapor directly represented by the samples will again be very small in comparison to the DU volume of vapors of interest. The field representativeness of a single set of small-volume, vapor points cannot be directly assessed in the absence of multiple, replicate sets of individual data points for comparison. This same problem hampers the reliability of single sets of discrete soil samples as discussed in Section 4.3.

Sample data that represent large, risk-based volumes of vapor (e.g., thousands of liters), similar to the concept of “Multi Increment” soil sample data (Section 4) are required for more reliable characterization of subsurface vapor plumes and vapor intrusion risk. Such “LVP” methods are currently mostly widely used for testing of vapors beneath building slabs where breakthrough to indoor air can be minimized (Section 7.8.5). The collection of LVP samples helps ensure that isolated, subslab vapor “hot-spots” that might be missed by small-volume vapor samples are incorporated into the data used to assess potential vapor intrusion risk and provides a volume-weighted average vapor concentration more applicable to comparison with subslab vapor action levels (HDOH 2017a; Section 7.14.1). Direct correlation of LVP data to identified impacts to indoor air might still not be practical, given the typical lack of knowledge of the exact point of vapor entry into a building, if in fact this is occurring.