Department of Health Seal

TGM for the Implementation of the Hawai'i State Contingency Plan
Section 7.8
SAMPLING APPROACHES AND EQUIPMENT

7.8 SAMPLING APPROACHES AND EQUIPMENT

The collection of soil vapor or indoor air samples can be more involved and complex than soil or groundwater sample collection. This is due in part to the need for special sampling equipment and containers to address the reactivity of vapor-phase chemicals and the need to prevent leaks during sample collection.

Based on the type of sampling equipment and containers, soil vapor or indoor air sampling approaches can be grouped into the following categories: (1) whole air sampling; (2) sorbent tube sampling; (3) passive sampling using sorbent materials; and (4) flux chamber sampling. The first two methods are often referred to as active sampling (see Hartman 2002). Whole air sampling involves collecting a volume of gas in a sample container, such as a Summa canister or a Tedlar bag, and analyzing the gas from that container directly. The concentration of targeted compounds is directly reported.

Sorbent tube sampling involves drawing a specified volume of soil vapor or indoor air through a sorbent material using a pump or other vacuum source and analyzing the sorbent material. The concentration of a targeted compound in the media tested is calculated by dividing the mass of the compounds collected on the sorbent material by the volume of vapor or air drawn through the sampler. If reanalysis of a sample might be required, then a collection method that relies on extraction of the sample (e.g., sorbent tubes) rather than purging (e.g., Summa canister) should be considered. The ability to reanalyze a sample collected in a Summa canister is limited due to the fixed volume collected versus the volume required for a specific analytical method. This should be discussed in more detail with the laboratory as needed.

Passive sampling methods rely on the placement of a sorbent material at a sample location for a period of time. The mass of a targeted compound collected on the sampler is then measured. Flux chambers are traditionally used to measure vapor emission rates from point sources such as waste ponds. Their use in soil vapor investigations is more limited but in some instances can be beneficial. A comparison of the key advantages and disadvantages to each of these approaches is provided in Table 7-3.

Similar to soil and groundwater sample collection, a combination of sampling approaches can be used if analyzing for a broad range of chemical compounds. For example, relatively inexpensive passive sampling can be used to initially screen a site for shallow vapor plumes and assist in the identification of areas for active soil vapor sampling. For the investigation of vapors associated with releases of diesel and other middle distillate fuels, a combined used of both Summa canisters and sorbent tubes is recommended (see Section 7.13.1.2). Data from Summa canister samples can be used to determine concentrations of individual VOC and short-range TPH compounds (e.g., <C12 aliphatics). Sorbent tube samples can be collected to evaluate longer range TPH compounds and semi-volatile organic compounds (SVOCs).

Each of these methods is discussed in more detail in the following sections. Consultation with the analytical laboratory to discuss specific standard operating procedures and sampling methods is strongly recommended during the planning phase for each project requiring the collection of soil vapor or indoor air samples.

Equipment that could come in contact with a vapor sample should be cleaned or decontaminated between samples to avoid cross contamination by trace levels of contaminants. Summa canisters and sorbent tubes should be certified clean by the laboratory. Used tubing should be disposed of. Reuse of Swageloks and ferrules between samples is discouraged. This is primarily a concern for the collection of outdoor or indoor air samples, where data are compared to very low action levels.

Table 7-3 Comparison of Soil Vapor & Indoor Air Sampling Approaches

Sampling Approach

Container/ Equipment

Advantages

Disadvantages

Active

Summa Canister

Familiar, widely accepted, rugged, no pump required (vacuum), excellent inert surface, low detection levels (ppbv), up to 30-day hold time from time of sample collection, easily air-shipped

Cost, bulky in field, slower fill time, fixed volume, collection of VOCs over restricted volatility range

Tedlar Bag

Inexpensive, availability, easily transported, rapidly filled

Less rugged, fixed volume, requires additional collection equipment, inner surface may sorb VOCs, medium detection levels (ppmv), short holding time, not recommended for air shipping

Sorbent Tube

Large-volume samples possible, collection of VOCs over a larger volatility range, low detection levels (ppbv), up to 40-day hold time (extraction within seven days of sample collection), easily transported and air shipped

Cost, requires additional collection equipment, saturation of tubes possible, sorbent media varies with respect to VOC and anticipated VOC concentration, tubes require storage at 4°C

Passive

Sorbent

Cost, ease of use, multiple samplers can be combined for analysis, long sampling times

Cannot directly measure vapor concentration

Water

Estimation of vapor concentrations possible, multiple samplers can be combined for analysis, long sampling times

Currently costly to install, cannot directly measure vapor concentration

Flux Chamber

Flux Chamber

Measures VOC flux at surface

Does not measure in-situ concentrations, identification of vapor emission points difficult




Table 7-4 Common Soil Vapor Concentration Unit Conversion Factors

 

Units

 

Convert to:

 

Multiply By:

μg/L

μg/m3

1,000

mg/m3

μg/m3

1,000

ppmv

ppbv

1,000

ppbv

μg/m3

MW/24

μg/m3

ppbv

24/MW

ppbv

ppmv

0.001

Definitions:
MW - molecular weight
mg/m3 - milligrams per cubic meter
μg/m3 - micrograms per cubic meter
μg/L - micrograms per liter
ppbv - parts per billion by volume
ppmv - parts per million by volume




Figure-7-5

Figure 7-5 Summa Canisters (spherical and cylindrical containers) with Flow Controllers (smaller gauges and blue box).




Figure-7-6

Figure 7-6 Summa Canister and Flow Controller Setups (note smaller flow controller on left).




Figure-7-7a
Figure-7-7b

Figure 7-7: Summa Canister and Flow Controller Parts.




Figure-7-8

Figure 7-8: One-liter Tedlar Bag with Disposable Syringe and Three-way Valve for Filling.




Figure-7-9a
Figure-7-9b

Figure 7-9: Sorbent Tubes. Upper photo: Sorbent tubes connected in series with a union fitting. Lower photo: Single sorbent tube connected to a 60ml syringe for collection of vapor sample (see also Figure 7-27).




Figure-7-10

Figure 7-10 Two Examples of Passive Soil Vapor Sample Collectors.




Figure-7-11

Figure 7-11: Two Examples of Indoor Air Passive Sample Collectors.




Figure-7-12

Figure 7-12: Passive Diffusion Sampler (PDS). Schematic of sampler on left and photo of a sampler being installed on the right.




Table 7-5 Comparison of TCE and PCE Results for Passive Diffusion Sampler and Active Soil Vapor Sample

Location

TCE Results

PCE Results

Mean PDS

Mean Active

PDS/Active Percentage

Mean PDS

Mean Active

PDS/Active Percentage

1

4,536

6,500

70%

ND

ND

ND

2

27,584

16,000

172%

384

300

128%

3

56,001

20,000

280%

752

285

264%

4

41,073

25,000

164%

744

385

193%

5

466

190

245%

ND

ND

ND

6

3,283

550

597%

ND

ND

ND

7

5,234

2,000

262%

ND

ND

ND

8

1,970

1,900

104%

ND

ND

ND

9

503

440

114%

ND

ND

ND

10

482

1,500

321%

ND

ND

ND

Average

233%

195%




Figure-7-13

Figure 7-13 Summary of high-density, passive sampler data for PCE vapors beneath the slab of a former dry cleaner: a) True data resolution based on PCE mass reported for each grid cell; (b) Extrapolated isopleth map based on assignment of data to center point of grid cell and use of contouring program.




Figure-7-14

Figure 7-14. Designation of soil vapor DU beneath a building slab for collection of LVP samples; recommended default DU volume of 3,000 liters represents the default, daily vapor entry rate used to develop HDOH (2017a) soil vapor action levels for vapor intrusion risk.




Figure-7-15

Figure 7-15. Example options for designation of purge points for collection of LVP subslab vapor samples: A) High-risk occupancy room within building; B) Subslab utility trench and preferential pathway; C) High-concentration area based on result of small-volume vapor sample data; D) Center of slab.




Figure-7-16

Figure 7-16. Floor drain and suspect deep cracks sealed with bentonite slurry to minimize downward leakage of indoor air into purge point during LVP sample collection.




Figure-7-17

Figure 7-17. Simplified schematic of Large Volume Purge sampling train.




Figure-7-18

Figure 7-18. Example design of LVP sample collection system.




Figure-7-19

Figure 7-19. Installation of LVP vapor extraction point used in HDOH (2017c) field study: a) Circular saw used to cut eight-inch hole in concrete for installation of vapor point and protective casing (latter not normally included); b) Completed hole; c) Two-inch PVC vapor point; d) Completed vapor point (interior sealed with cement grout). A smaller diameter hole will normally be adequate for a two-inch soil vapor point.




Figure-7-20

Figure 7-20. Example, completed field LVP sample collection set up (HDOH 2017c; Shop-Vac pump not shown).




Figure-7-21a
Figure-7-21b

Figure 7-21: Schematic Diagram and Photograph of Flux Chamber




7.8.1 Whole Air Sampling

Whole air samples are typically collected in Summa canisters when definitive data with low detection levels are required; however other containers such as Tedlar bags, gas-tight vials, and syringes are also suitable for some applications. Low detection levels can also be obtained using sorbent tubes and TO-17 analytical methods, although the volume of air or vapor drawn through the tubes is limited by the sorptive capacity of the media used. Individual laboratories often publish guidance on the use of various whole-air methods (e.g., Air Toxics 2012). Whole air sampling (or other active soil vapor sampling) is recommended to quantify concentrations of vapor-phase chemicals for an exposure or risk assessment. Contaminant concentrations can be quantified in units of volume of gas per unit volume of air (e.g., parts per million by volume [ppmv] or part per billion by volume [ppbv]) or in units of mass per unit volume (e.g., milligrams per liter [mg/L] or micrograms per cubic meter [µg/m3]). Concentrations should be reported in units of µg/m3 for comparison to HDOH EALs.

It is important to note that unlike aqueous samples, volume units are not equivalent to mass units for gaseous samples. The molecular weight of the analyte and the temperature and pressure of the sample must be used to convert from volume units to mass units. The conversion can be achieved using the Ideal Gas Law equation. The following equation simplifies the Ideal Gas Law equation assuming atmospheric pressure (one atmosphere = 760 millimeters of mercury [mm Hg]) and standard room temperature of 298 Kelvin (K) (25 degrees Celsius [° C]):

µg/m3 = ppbv x MW
  24.45
     
Where:  
  ppbv = Parts per billion by volume
  µg/m3 = Micrograms per cubic meter
  MW = Molecular Weight (chemical-specific)

Table 7-4 presents other common unit conversion factors and HDOH has developed a Vapor Unit Conversion spreadsheet that converts between concentrations and is available for download from the EHE web page (HDOH 2016).

See Table H in Appendix 1 of the HDOH EHE guidance for the molecular weight of common chemicals (HDOH, 2016). In general, target compounds conducive to whole air sampling are chemically stable and have a vapor pressure greater than 0.1 torr at 25 ˚ C and 760 millimeters of mercury (one atmosphere). Effective recovery of chemicals from active soil vapor samples depends on sample humidity, chemical activity of the sample matrix, and the sample container’s degree of inertness.

7.8.1.1 Summa Canisters

A Summa canister is a stainless steel container that is placed under a vacuum and then used to collect a soil vapor or air sample. The canister is cleaned internally using electropolishing and chemical deactivation to produce an interior surface that is nearly chemically inert. This minimizes reactions with the vapor sample and maximizes recovery of volatile compounds from the container. Recovery is generally limited to chemicals with up to ten carbon molecules for aromatic compounds, however, including naphthalene, and up to twelve carbon molecules for aliphatic compounds.

Summa canister volumes range from 400 milliliters (ml) to 6 liters, with one-liter and six-liter canisters being most commonly deployed in the field (Figures 7-5 & 7-6). Larger canisters provide more sample volume for the laboratory and allow for lower detection levels. For indoor or outdoor air sampling, six-liter canisters are typically most appropriate as they support the low detection levels necessary for risk assessment or screening against indoor air standards. They are also better suited to collecting time-integrated samples (e.g., eight to twenty four-hours). Smaller canisters are typically used for soil vapor samples, for which screening levels are typically several orders of magnitude higher than for indoor air. Collection of vapor samples greater than one-liter can be problematic as well as time consuming at sites with relatively tight soils. HDOH recommends a minimum sample size of one-liter for soil vapor samples in order to assist in the collection of data that are representative of the site being investigated. HDOH further recommends the use of Summa canister data and/or sorbent tube data (provided a minimum one-liter sample is drawn) for final decision making purposes.

General procedures when planning a soil vapor investigation using Summa canisters include:

  1. Obtain the field equipment checklist and inventory equipment and materials needed for the soil investigation before proceeding to rent, obtain and stage all equipment
  2. Order clean-certified Summas and related equipment from the laboratory based on the number of samples to be collected plus a minimum of one extra canister and flow controller. Ideally this should be done by two weeks before the scheduled field work in order to give the laboratory adequate time to prepare the canisters and ship them to Hawai‘i.
  3. Check the vacuum of all the canisters as soon as they arrive. Use the pressure gauge the lab sends with the canister. If more than one gauge is sent, use the same gauge to check the canisters before and after sampling in order to ensure consistency. Record the date, time and readings on both the canister tags and field forms. Vacuum readings should be approximately 30 inches of mercury. Consider rejection any canisters that differ from others by three inches of mercury or more.

Summa canisters should be certified clean by the analytical laboratory that supplies them. The laboratory cleans the Summa canisters after each use. The cleaning process is certified by filling a canister with a clean gas (e.g., nitrogen) and then analyzing the gas using method TO-14 or TO-15. Canisters are typically either batch certified or individually certified. For batch certification, a portion of canisters from a cleaning batch is tested (e.g., 10%). For individual or 100% certification, each individual canister is tested.

The Summa canister is prepared by the lab for sampling by evacuating the contents to a vacuum of approximately 30 inches of mercury (in Hg). This ensures that the volume of air drawn into the canister will be approximately equal to the canister volume. The vacuum in each canister should be documented prior to sampling. This can be done by attaching a separate vacuum gauge to the canister and opening the intake valve. The flow controller can also be attached then capped at the intake port and the Summa canister intake valve again opened. Doing both allows testing of the flow controller vacuum gauge for accuracy, a not-uncommon field problem.

Flow controllers (or flow restrictors) are essential equipment for the collection of vapor samples with Summa canisters. The controllers limit the rate at which a sample can be drawn into a canister and assure that the sample flow rate is appropriate for the targeted sample collection time (see Section 7.10.3). Flow controllers also help ensure that an excessive vacuum is not applied to soil. Doing so could potentially strip VOCs from free product or sorbed to soil and bias the vapor sample collected. Older flow controllers tend to be bulky, less reliable and increase the chance for leaks (e.g., see Figure 7-6). Newer flow controllers are more compact and easier to use in the field (See Figure 7-7). In some cases, they may come preset and pre-attached to the Summa canister. A vacuum gauge at the vapor collection point is used to monitor the vacuum pulled during the collection of LVP vapor samples (Section 7.8.5).

The analytical laboratory providing the Summa canisters and flow controllers should be consulted during the planning stages to ensure the appropriate size canisters and appropriately calibrated flow controllers are provided (Figure 7-7). Laboratories typically provide matched canisters and flow controllers. It is advisable to order extra canisters and flow controllers in case insufficient vacuum is present in a canister or the flow controller does not work properly.

A 10% certification of Summa canisters (i.e., 10% of canisters) is recommended for standard TO-14 or TO-15 analysis for soil vapor or other applications where very low detection levels are not required. This is appropriate for routine ambient air applications and the collection of high-concentration soil vapor and landfill gas samples where parts-per-million or parts-per-billion reporting levels are required. A 100% certification is recommended when “Low Level” or selected ion monitoring (SIM) analyses will be conducted and parts-per-trillion reporting levels are required (e.g. for indoor air). If desired, certification can usually be provided at an additional cost for specific pairs of flow controllers and canisters that are labeled as such by the laboratory.

The Summa canister valve should be closed and sample collection ceased once a residual vacuum of 3-5 inches of mercury is reached and the final vacuum recorded. This will help notify the lab of potential container leakage and compromised samples during storage and shipment. A holding time of 30 days is recommended once the sample has been collected (USEPA 1999b). Most laboratories recommend that canisters be returned within 14 days of receipt in order to help ensure the integrity of the canister and ensure that hold times are not exceeded prior to analysis.

7.8.1.2 Tedlar Bags

HDOH recommends the use of Summa canisters or sorbent tubes for sample collection if the data are to be used for final, decision making purposes. Tedlar bags are flexible, plastic bags that can be used for the collection of air or vapor samples with a syringe or a lung box. Tedlar bags can offer an inexpensive, screening tool for initial site investigations or monitoring, however. A Tedlar bag is made from two layers of Tedlar film sealed at the edges and containing a valve allowing for soil vapor or indoor air sample collection using a syringe or lung box . Tedlar is a trade name for a polyvinyl fluoride film that exhibits a low permeability to gases, chemical inertness, weathering resistance, and low off-gassing. The manufacturer of Tedlar, DuPont, announced in 2009 that they would phase out support for Tedlar film in the sample bag market. As a result, true Tedlar bags are becoming more difficult to obtain and are being replaced by bags made of alternative materials.

Similar to canisters, Tedlar bags range in volume up to five liters, and can be used to collect high-concentration grab samples or ambient air samples. The Tedlar bags can be used for projects involving analysis of low concentrations of compounds, in the ppbv range; however, they are less robust and more prone to leaking and diffusion than Summa canisters. Shipment by air is generally not recommended, since a decrease in atmospheric pressure as the plane ascends can cause a Tedlar bag sample to expand and leak. If the use of Tedlar bags and shipment by air is not avoidable, then the bags should only be filled to 50% or less of their total capacity.

As described in Section 7.10.4, soil vapor or indoor air samples are collected in Tedlar bags using a lung box or a glass syringe equipped with a 3-way valve (Figure 7-8). Note that a plastic syringe is depicted in the figure. The use of glass syringes for sample collection is recommended due to the potential sorption of VOCs to plastic. Tedlar bags are normally disposed of following use to collect one sample since VOCs can absorb to the bag interior. This excludes bags used to screen for fixed gases like O2, CO2 and CH4, which are not significantly sorptive.

7.8.1.3 Whole Air Sample Handling

In order to preserve the integrity of whole air samples, the following guidelines should be observed.

  • Do not chill samples during storage as is common with soil and water samples.
  • Sample containers should not be left in the direct sunlight.
  • The maximum holding time in Summa canisters is typically 30 days from the time the canisters are initially cleaned, depending on the lab (labs often request return of canisters within 14 days of receipt).
  • The maximum holding time for Tedlar bags is 24 to 72 hours after the sample has been collected, depending upon the compound (CalEPA 2012, SDC 2011).

Petroleum compounds and biogenic gases (e.g., methane, carbon dioxide) are less stable and should be analyzed within 24 hours for samples collected in Tedlar bags, while chlorinated compounds are more stable and bags can typically be held for up to 72 hours. However, as noted above, Tedlar bags are being phased out and replaced by bags made of alternative materials. Appropriate holding times vary depending on the material. Therefore, the supplier of sample bags not made of Tedlar, typically the laboratory, should be consulted regarding the appropriate holding time for the site contaminants of concern.

7.8.2 Sorbent Tube Sampling

Sorbent tube sampling involves drawing a known volume of soil vapor or air through a sorbent material using a pump or other vacuum source and analyzing the sorbent material. Concentration is calculated by dividing the mass of a targeted compound by the volume of vapor or air drawn through the sampler. Photographs of sorbent tubes and sorbent tube sampling trains are provided in Figure 7-9. Individual laboratories publish guidance on the use of various sorbent methods (e.g., Air Toxics 2012b). Method TO-17 is the most common setup and analysis used for sorbent tubes (see Section 7.13).

Sorbent tubes are an optional method for short-chain VOCs (e.g., BTEX, PCE, TCE, etc., including naphthalene) and the most appropriate method for longer-chain, SVOCs that cannot be recovered from a Summa canister. The latter includes acenaphthene, methylnaphthalenes and other PAHs with molecular weights up to 200 but including pyrene, with a molecular weight of 202 (see HDOH, 2011b). As discussed below, sorbent tubes in combination with Summa canisters are recommended for testing of vapors associated with diesel and other middle distillate fuels if a significant (e.g., >10%) amount of longer-chain hydrocarbons could be present (e.g., >C10 aromatics or C12 aliphatics; see Section 7.13.1.2).

A variety of sorbent cartridges and pumping systems are provided by commercial vendors or laboratories. It is important to discuss the anticipated types and concentrations of target VOCs and SVOCs with the laboratory in order to optimize the type and amount of sorbent used to prepare the tubes. Sorbent tubes are typically shipped and stored chilled to 4°C but should be brought to ambient temperature prior to use in the field.

A low-flow pump or syringe is used to draw soil vapor or air through the sorbent over a pre-established time period. A maximum flow rate of 200 ml/minute is recommended in order to minimize the risk of leaks around the probe annulus as well as minimize the vacuum imposed on the soil and stripping of VOCs from the soil or free product (see Section 7.10.3). Pumps are typically used for the collection of larger volume, indoor or outdoor air samples. If a pump is used then the volume of soil vapor drawn through the tube is calculated by multiplying the average flow rate by the draw time. This will require recording and averaging the flow rate several times if it varies over collection of the sample.

Calibrated syringes that can be easily read in the field provide a more accurate estimation of the volume drawn through a sorbent tube for small-volume samples. A syringe draw time of no less than 15 seconds, for example, is recommended for a 50ml soil vapor sample. This is the maximum draw volume typically allowed by laboratories for collection of high-concentration soil vapor samples associated with petroleum in order to avoid saturation of the sorbent material in the tube and multiple dilutions at the laboratory. Note that the syringes should not be re-used between sample points to avoid potential contamination of sorbent tube media due to a high concentration breakthrough in a previously drawn sample.

The presence of very high concentrations of volatile compounds at some sites can significantly limit the volume of soil vapor that can be drawn through a sorbent tube without saturation of the sorbent material. Unlike canister samples, sorbent tubes have maximum reportable concentrations for VOCs, based on the sorptive capacity of the material used. Once this capacity is reached, breakthough will occur and true concentration of the chemical present cannot be determined. This can be addressed in part by using larger sorbent tubes, adjusting the sorptive material used and/or connecting two or more sorptive tubes in series and adding the masses of targeted VOCs captured in each tube.

Note that PIDs primarily target aromatic compounds and are not good indicators of total TPH levels in soil vapors without inclusion of a correction factor, since vapors are likely to be dominated by aliphatic compounds. This is especially important to consider for testing of aromatic-poor vapors from diesel fuel or other middle distillate fuels (refer to HEER Office petroleum vapor study; HDOH, 2012). PID readings for similar vapor concentrations from gasoline versus diesel can be significantly lower for diesel. An FID can be used to minimize this problem but they are not widely used in Hawai‘i. An FID responds to both methane and petroleum-related compounds, but does not respond strongly to chlorinated solvents, and perhaps most important for rural or inter-island work in Hawai‘i requires a ready supply of hydrogen. High humidity or low-oxygen environments can also extinguish the FID flame. This can be an issue for screening of subsurface vapor associated with degrading petroleum releases.

When possible, screening level Photoionization Detector (PID) data should be provided to the laboratory prior to sample collection in order to assist the lab in optimization of sorbent tube preparation. For heavily contaminated, petroleum-release sites in particular, the amount of soil vapor drawn through a sorbent tube might still be limited to volumes as small as 50ml. Smaller volumes are not recommended under any circumstances as they are unlikely to be representative of site conditions. This should be discussed with the laboratory prior to preparation of the sorbent tubes for sample collection. If necessary, a series of connected sorbent tubes can be used to collect larger-volume samples (see first photo in Figure 7-9).

If sorbent tubes are to be used in a high-concentration, soil vapor environment (e.g., to evaluate TPH in vapors associated with diesel-contaminated soil or groundwater) and the volume of vapors to be drawn is less than one liter then the concurrent collection of a one-liter or larger Summa canister sample is also recommended (see Section 7.13). The Summa canister sample should be collected first to help ensure that the vapor point is adequately purged and to improve the representativeness of the sorbent tube sample. The well point should then be closed using a valve or pinched shut (similar to the sampling train leak test), using a small length of flexible tubing to prevent the backflow of ambient air into the tubing and soil. Allow adequate time for the vacuum on the soil to dissipate with the vapor sampling point remaining closed. This could take several minutes for tight soils. The sorbent tube sampling train should then be connected, the vapor point re-opened, and the sample collected.

After the sample is drawn, the sorbent tube should be chilled to 4°C and sent to the laboratory for analysis. The concentration of a targeted chemical in the original vapor is calculated as the mass of the chemical sorbed divided by the volume of vapor drawn through the sorbent.

The storage and holding time for sorbent tubes vary depending on the sorbent material used and targeted VOCs but are typically up to 30 days after the tubes are prepared. Removal and testing of the sorbent material may be required by the laboratory within 14 days of sample collection for some methods.

Concurrent PID data should be provided to the laboratory in order to determine if dilution of sample is necessary prior to analysis and help reduce analytical time and costs. Note that petroleum vapors are dominated by aliphatic compounds. PIDs primarily target aromatic compounds and are not good indicators of total TPH levels in soil vapors without inclusion of a correction factor. This is especially important to remember for aromatic-poor vapors from diesel fuel or other middle distillate fuels (refer to HEER Office petroleum vapor study; HDOH, 2012). PID readings for similar vapor concentrations from gasoline versus diesel can be significantly lower for the latter. A Flame Ionization Detector (FID) can be used to minimize this problem but they are not widely used in Hawai‘i. High humidity or low-oxygen environments can extinguish the FID flame. An FID also responds to methane as well as petroleum-related compounds, does not respond strongly to chlorinated solvents, and perhaps most important for rural or inter-island work in Hawai‘i requires a ready supply of hydrogen.

When used in combination at a petroleum site, the Summa canister sample should be tested for TPH as the sum of C5-C12 compounds (Section 7.13) as well as targeted, individual compounds (e.g., BTEX and naphthalene) using TO-15 or an equivalent method. The sorbent tube sample should be tested for TPH as the sum of C5-C18 compounds using TO-17 or an equivalent method. Although not directly comparable due to different lab methods, the difference in the two, reported concentrations of TPH in the vapor samples will give some idea of the proportion of compounds greater than C12. As an alternative, the lab can be asked to quantify TPH in the sorbent tube sample as the both the sum of C5 to C18 compounds and C12 and higher compounds. The data can then be used to evaluate the most appropriate sample collection method for characterization of the site. For example, if less than 10% of the total TPH is estimated to be composed of C12 and higher compounds then Summa canisters can be used to collect additional samples (see also Section 7.13.2)

7.8.3 Passive Sampling

Passive sampling involves using adsorbent materials to collect vapor phase chemicals without the use of a pump or Summa canister. The vapor is not induced to flow over the adsorbent; instead the chemicals in the vapor passively contact the adsorbent and adsorb to it. Both VOCs and SVOCs are captured by the adsorbent material and can be characterized, although extremely volatile chemicals (e.g., vinyl chloride) may not concentrate sufficiently on the adsorbent and the less-volatile SVOCs may not have sufficient vapor pressure to be detectable.

Passive sampling approaches requires less equipment and is more straightforward in the field than active sampling. Data for samplers can be used to identify vapor-phase chemicals for additional site characterizations and vapor intrusion studies (e.g., USEPA 2009). These methods give a time-integrated measurement and capture temporal variations in VOC concentrations that could be missed with short-duration, active samples. Passive sampling methods have also been used to estimate VOC concentrations in soil vapor.

Calculation of the vapor concentration from passive sampling results is, however, sometimes attempted by estimating the volume of vapor that passes by the buried adsorbent during the burial time period and a vapor diffusion model. Another method is to determine an uptake rate for the passive collector. At present, comparative studies between actively measured soil vapor concentrations and those estimated from passive sampling are at best within an order of magnitude (e.g., USEPA 2009). Due to this uncertainty, passive soil vapor data are considered to give qualitative or semi-quantitative vapor concentration results and generally are considered a screening tool for soil vapor investigations. An emerging exception could be the use of water-based passive samples, discussed in a following section.

Implementation of passive sampling approaches requires less equipment and is more straightforward in the field than active sampling. Therefore, passive sampling can be beneficial for the initial stages of site characterization or vapor intrusion studies.

Advantages of passive sampling include:

  • Quick and relatively inexpensive method to investigate large areas and to map plumes;
  • Able to detect any contaminant that has an appreciable vapor pressure and can be adsorbed in sufficient quantity to determine relative presence or absence, including lighter-end SVOCs like naphthalene, even if present in very low concentrations;
  • Results can be used to more cost-effectively design and optimize follow-on active sampling;
  • Individual samples can be combined for extraction and testing to increase coverage for targeted area (check with lab prior to collection);
  • Useful in situations where active methods may not be applicable, (e.g., areas of extremely low permeability and high moisture content, high-traffic/limited access areas, etc.);
  • Can be used to find preferential pathways into a structure or around a structure, such as utility corridors;

Disadvantages of passive sampling include:

  • Not useful for generating vertical vapor profiles unless sampling intervals can be effectively isolated;
  • Actual vapor concentrations can only be estimated, limiting use of the method to screening, plume mapping, and other semi-quantitative applications.

An additional drawback is that passive sampling requires two visits to the field, one to deploy the adsorbents and a second trip to retrieve them, and does not allow for the acquisition of real-time data. Intentional or accidental disturbance and even vandalism of passive samplers can also be a problem

Passive sampling can be applied to either soil vapor or indoor air. Although the principle is the same in application to these media, the sampling equipment is different, as described in the two sections below.

7.8.3.1 Passive Soil Vapor Sample Collectors

Qualitative passive soil vapor sample collection involves placing an adsorbent into the subsurface for a pre-specified exposure period to allow the adsorption of soil vapor chemicals onto the adsorbent material. Compound uptake rates are not required to be known for this approach. The sample exposure duration can be from days to weeks, with a typical exposure period of one week to identify source areas and two weeks when tracking contamination in groundwater or when heavier molecular weight compounds are of concern (e.g., naphthalene and 2-methylnaphthalene). Longer exposure durations are recommended for comparison of data to lower action levels. Request that the technology vendor provide specific guidance on sample exposure periods based on sampling objectives and site conditions.

Passive soil vapor sample collection involves placing an adsorbent into the subsurface for a known exposure period (e.g., one to two weeks) to allow the adsorption of soil vapor chemicals onto the adsorbent material. Sample collection procedures are described in Section 7.12. The absorbent is typically placed in the upper end of an inverted container having an open bottom, or in a fine wire mesh or polymeric material, to facilitate contact with the soil vapor but not the soil. Photographs of two vendor-supplied sample collectors are provided in Figure 7-10.

Although the results are qualitative, passive soil vapor sampling can provide useful information when investigating subsurface vapor plumes or preferential pathways for vapor intrusion studies.

One evolving approach is to subdivide a site into targeted Decision Units (DU) for screening characterization. Active soil vapor sampling will be targeted for the DU with the highest, relative concentration of VOCs identified in passive samples. Rather than deploy a single passive sampler in each DU, multiple samplers are deployed to provide better coverage and then combined at the laboratory for a single extraction and analysis. For example, five to ten passive samplers can be installed within each targeted area of a site. After collection, the laboratory can be instructed to combine and carry out a single extraction for groups of samplers from targeted DUs. This increases the accuracy and quality of field data without increasing lab cost.

USEPA conducted a verification study of the major vendor-supplied passive diffusion sample collectors in the late 1990’s (USEPA 1998, USEPA 1998d). As part of this study, the results of passive diffusion samplers were compared to active soil vapor measurements. These studies showed that:

  • The passive soil vapor sampling systems detected the same compounds in each sample as the active method, as well as several VOCs that the active method did not detect. This performance characteristic suggests that the passive soil vapor sampling systems may detect VOCs that are at lower concentrations in the subsurface than the active soil vapor sampling method can detect and/or that the passive samples were able to better capture temporal changes in vapor concentrations due to the longer exposure period.
  • The results also indicated a general, relative correlation between passive soil vapor sampling results and active method data (e.g., high or low). However, at high contaminant levels, the ratio between the passive and active results decreased, suggesting that sorbent saturation might have occurred. This decreases the resolution capability of the passive samplers in heavily contaminated areas.
  • Because the passive soil vapor sampling systems and the active method use different techniques to collect soil vapor samples, it is not expected that the two methods will provide the same response or that the data will be directly comparable.

7.8.3.2 Passive Sampling of Indoor Air

Passive indoor air sample collection involves hanging an adsorbent-containing sample collector at a location in a building where indoor air chemical concentration data are desired. A typical sample collector consists of a glass vial with the adsorbent material inside – the vial is capped with a gas-permeable membrane, which allows vapors to enter the vial but excludes any particulate matter in the air.

As discussed earlier, another alternative is to deploy multiple samplers within targeted rooms, floors, etc., to provide better coverage and then have the laboratory combine groups of samplers for a single extraction and analysis. This can help reduce concerns about air flow and the inclusion of stagnant areas of the building in the indoor air evaluation.

Passive sampling of indoor air is a technique that has two primary applications. First, as a low-cost screening tool, the technique can be used to provide wide coverage of a building or set of buildings with a minimal amount of field equipment. Second, because the adsorbent equilibrates with the indoor air over a longer period of time than is typically practical with whole air sampling, the result of passive sampling reflects a longer-term average concentration that can be useful as another line of evidence in risk assessment.

Photographs of two vendor-supplied sample collectors are provided in Figure 7-11.

7.8.3.3 Emerging Technologies

The field of soil vapor and indoor air passive sampling is rapidly developing. There are a number of technologies that can be applicable to specific site characterization needs. One emerging technology, the use of passive diffusion samplers, is described below.

High-Density Passive Sampler Deployment

The use of high-densities of passive samplers to characterize the in situ nature of vapor plumes beneath building slabs was evaluated in an HDOH (2017) investigation of a PCE vapor plume beneath the slab of a former dry cleaner. The approach involves the installation of multiple, rather than single, passive samplers beneath targeted areas of the slab and subsequent collection and combination of the samplers for testing under a single analysis. Testing of multiple individual samplers installed within a single 300 ft2 grid cell indicated significant variability between closely located points. This has significant implications regarding the reliability of small-volume vapor sample data to accurately delineate plume boundaries and variability within larger-scale plumes (Section 7.5; see also Section 13.2 and Brewer et al. 2014).

Four samplers were installed in each grid cell for the HDOH (2017) study and combined for analysis. Replicate sets of samplers (triplicates) were installed in three grid cells in order to assess the precisions of the data. The replicate data indicted very good precision. The number of samplers required to obtain consistent, reproducible results for a targeted area has not been studied in detail and at this point is necessarily site specific.

Isopleth maps of the vapor plume identified beneath the dry cleaner in the HDOH field study are presented in Figure 7-13. The first map reflects the true resolution of the passive sampler data at the scale of an individual grid cell. The second map was generated by assigning the concentration of tetrachloroethylene (PCE) reported for each cell to the center point of the cell and then using a contouring program to generate corresponding isopleths. While clearly superior to typical, small-volume vapor sample investigations, the practicality of installing large numbers of passive samples beneath a building slab will necessarily be site specific.

The combination of multiple passive samplers for testing should be discussed with the laboratory prior to the commencement of field work. This approach helps to capture and represent small-scale, random heterogeneity of VOC concentrations within a targeted area and provide more representative data for site characterization purposes. Although less well-defined in terms of sampling theory, this is similar in concept to the collection of a “Multi Increment” soil sample from multiple, rather than a single point within a targeted area for improved data resolution and reliability (Section 4.2). Refer to the HDOH 2017c study for additional information.

Passive Diffusion Sampler

A passive diffusion sampler (PDS) has been developed by the USEPA Office of Research and Development for soil vapor characterization (Paul 2009). This sampler uses water as the media into which contaminants partition rather than the solid adsorbent approach described above. This sampling technology is in the developmental phase, but it has advantages for characterization of contaminants that might not be well adsorbed on solid media, including more polar compounds. More reliable estimates of VOC concentrations in soil vapor may be possible.

The PDS is constructed using a 40 ml VOA vial filled with de-ionized water and with the Teflon septa replaced with a vapor-permeable membrane. The PDS is inserted into a custom-made messenger (hollowed-out plastic cylinder) and deployed in two-inch diameter, monitoring wells with a screened interval placed at the desired soil vapor depth interval. Figure 7-12 shows a schematic of the PDS and a photo of the sampler being deployed in the field. An O-ring on the messenger seals the targeted depth interval from ambient air. Further installation details and a Standard Operating Procedure (SOP) are available from the developer of this technology (Paul 2009).

Once the sampler is installed, the permeable membrane in the PDS is exposed to the screened interval and contaminants diffuse through the membrane into the water-filled PDS until the water reaches equilibrium with the surrounding soil vapor. The PDS is recovered from the well after an appropriate equilibration period, typically one month for most VOCs. The addition of preservatives to the PDS sample after collection should be considered in the similar manner as done for groundwater samples (see Section 6). The water is then analyzed for targeted VOCs, with results expressed in units of mass per volume of water (e.g., ug/L). The concentration of the VOC in water is then multiplied by the Henry’s Law constant for that chemical to estimate the average, equilibrium concentration of the VOC in the surrounding soil vapor.

Several comparative field studies of this technology in application to petroleum and chlorinated solvents have been carried out (e.g., USEPA 2009). Table 7-5 lists the results of one study wherein PDS sampler results were converted to vapor concentrations using Henry’s Law and then compared to a collocated active (e.g., Summa) soil vapor measurement. In this study, the PDS estimated concentrations of vapor-phase VOCs were consistently higher than those reported for the active samples for both PCE and TCE (USEPA 2009). Among other possibilities, this suggests either: 1) A consistent error in conversion of dissolved-phase VOCs to equivalent vapor-phase VOCs and/or 2) The existence of subsurface spatial and/or temporal vapor “hot spots” that were captured by the PDS sampler due to their longer exposure time but missed by the short-duration active samples.

Another potential advantage of PDS samples is the ability to collect and then combine multiple samples within individual, targeted Decision Unit areas for comparison to adjacent areas or for estimation of time-averaged VOC concentrations for use in vapor intrusion studies. Whether or not PDS samples can indeed be used to obtain a more representative picture of long-term subsurface soil vapor conditions at a site is still under investigation.

7.8.4 Large Volume Purge Sampling

“Large Volume Purge (LVP)” soil vapor collection methods, referred to as “High Purge Volume” samples by McAlary et al. (2010), have been used sporadically to assess vapor intrusion risk since the early 2000s but only casually mentioned in USEPA or state agency guidance (e.g., CalEPA 2015). Under this approach an active or passive vapor sample is continuously collected from a stream of vapor being purged from a point installed into the bottom-floor slab of a building. Problems hindering routine use of the approach included: 1) Lack of awareness of the limitations of traditional, soil vapor data (see Section 13.2), 2) Lack of a systematic approach to soil vapor investigations and designation of risk-based, “Decision Units (DUs)” of vapor for sample collection and characterization, 3) Limited information on the engineering design of LVP sampling collection system; and 4) Misplaced concerns regarding the need to identify the exact, subslab source of vapors purged during sample collection.

These issues were evaluated in an HDOH field study of the collection of LVP vapor samples carried out in 2016 (HDOH 2017c). The use of high-density, passive sampler installation to characterize the in situ nature of vapor plumes beneath building slabs and assist in designation of LVP sample collection locations was also included in the HODH study (see Section 7.8.3.3). A brief overview of the design of the sampling system is provided below. Refer to Section 7.5, Section 13.2 and the HDOH (2017c) study for additional background information. The sample collection design presented is intended as an example only and is similar in nature to a standard, soil vapor extraction pilot test. It is anticipated that more efficient LVP sample collection methods will be developed in the future.

7.8.4.1 Investigation Objectives and LVP DU Designation

Large Volume Purge vapor data are used to more directly assess potential vapor intrusion risk at existing buildings, rather than characterization of in situ VOC concentrations in vapors beneath a slab (refer to Section 7.5). This should be clearly stated in the project workplan. The latter might be necessary if LVP data indicate a potentially significant risk, or might be carried out beforehand in order to assist in in designation of an LVP sample collection point.

A default, subslab vapor DU volume of 3,000 liters is recommended (Figure 7-14). This reflects a default vapor entry rate of 2 L/minute estimated by Brewer et al. (2014) for buildings in tropical climate zones (see Section 7.5). Deviations from the default volume should be discussed in the workplan. A consecutive series of five LVP purges is recommended. This is intended to reflect potential vapor intrusion through the designated LVP sample collection point over a five-day period and better capture large-scale variability of VOC concentrations within a vapor plume underlying a building slab. Collecting separate samples over a series of LVP purges also reduces the risk of biasing the full data set if leakage into the system is identified during later stages of sample collection.

7.8.4.2 LVP Sample Point Designation and Slab Preparation

State the rationale for designation of the targeted LVP sample collection point or points in the investigation workplan. Potential LVP sample collection points include:

  • At or near suspect or known vapor entry points;
  • Within or adjacent to known or suspect subslab utility trenches that could serve as preferential pathways for vapor flow;
  • Directly above suspect, subsurface soil or groundwater source areas; and
  • Sensitive-use areas of the building, or in the center of the slab (Figure 7-15).

Existing small-volume vapor sample data might also be used to designate an LVP collection point, if available, although the collection in advance of small-volume vapor samples is not necessary unless a source area above groundwater is specifically suspected or if significantly heterogeneity of VOC concentrations within an underlying groundwater plume is suspected.

Obtain as-built designs for the targeted slab or consult a structural engineer prior to designation of points for LVP sample collection. Geophysical toning and a review of as-built building plans should be carried out to identify subsurface utilities and the presence of rebar in the slab. Slabs under commercial buildings are typically between four and 20 inches thick and may or may not be reinforced with steel reinforcing bars (rebar) or other material. Slabs constructed for commercial or industrial buildings are not typically uniform. A slab will typically be thicker (supported by an underlying foundation) in areas anticipated to bear significant weight from machinery or walls. These structures could in theory compartmentalize and isolate individual pockets of vapor beneath a slab and should be taken into consideration for designation of sample collection points if as-built diagrams of the building slab and foundation are available.

Seal visible and accessible utility penetrations, floor drains, cracks suspected to penetrate the slab and other potential routes for downward leakage of indoor air to the extent practical during the collection of LVP samples (Figure 7-16). Methods to seal cracks and gaps in floors include bentonite slurry and heavy-duty tape. Avoid the use of compounds with volatile compounds that could affect sampling results. Methods to assess leakage through the slab during the collection of LVP samples are described in Section 7.8.4.6.

The distance from the sample collection point that the floor should be sealed can be estimated as the worst-case, vapor draw area with respect to the targeted, total DU purge volume. Assume, for example, that a one-meter wide by one-meter deep utility trench is present beneath or nearby the selected sample point and could serve as a preferential pathway for vapor flow. Based on a total LVP purge volume of 15,000 liters (i.e., five, 3,000 liter purges) and an effective, air-filled porosity of 28% (default value used in HDOH EAL calculations), the vapors will be drawn from an approximate 50 m3 volume of soil beneath the slab. This suggests that the slab within 25 m of the LVP purge point should be sealed, assuming an equal length of influence in both directions along the utility trench. The source area of the vapors cannot be determined from the purge data and will be influenced by preferential pathways in the soil. Exact knowledge of the source area is not important for assessment of potential vapor intrusion risk but could be useful for identification of subslab source areas and/or utility corridors and other features that could serve as preferential pathways. The source of purged vapors can be approximated by installing additional probe points through the slab and monitoring of the vacuum drawn at in different locations around the LVP point. This might be desirable if significant levels of VOCs are reported in LVP samples and a better understanding of vapor source areas is needed, as discussed below.

These distances are for example only. In practice, all potential, accessible gaps within the slab should be sealed prior to sample collection as a conservative measure and in order to minimize the potential for leakage of indoor air into the sample and call into question the reliability of the data. Leak checks should then be included as part of LVP sample collection (Section 7.8.4.4).

The exact vapor draw area does not need to be determined as part of an LVP investigation since, aside from an increased flow rate, sample collection is intended to directly mimic upward vapor flow through the designated point. This includes potential leakage of outdoor air under the edges of the slab during sample LVP purges. As discussed in Section 13.2, subslab vapors that intrude and impact indoor air in most cases originate as outdoor air that has been drawn in under the edges of the slab and is contaminated by volatile chemicals slowly diffusing out of a soil or groundwater source as the air flows toward the vapor entry point.

Knowledge of the approximate location of the vapor source area might, however, be beneficial as part of a follow-up characterization to identify subslab source areas and design remedial action plans if LVP data indicate potentially adverse impacts to indoor air. Be aware, however, that subslab vapor plumes are often not co-located with the soil or groundwater source area. An additional soil and/or groundwater investigation will typically be required to identify vapor source areas. Strategies for the investigation of soil and groundwater contamination by volatile chemicals are discussed in Section 4 and Section 6, respectively.

7.8.4.3 LVP Sample Train Design and Test

A detailed description of the system used in the HDOH (2017c) LVP field test is provided in the report for that study. The system was modeled largely after an approach published by McAlary et al. (2010) and is similar to designs used for a soil vapor extraction pilot test. It is anticipated that the design utilized in the field study can be scaled down for routine use.

A schematic of the LVP design is provided in Figure 7-17. The design used in the HDOH (2017c) field study is depicted in Figure 7-18. The basic configuration consists of a two-inch polyvinyl chloride (PVC) pipe connected to a vapor sampling point installed in the center of the slab. A Shop-Vac® was used in the field study to produce a vacuum on the sample point and purge the targeted volumes of vapors. Rotron or similar types of fans or blowers can also be utilized. Multiple sample ports should be installed into the PVC piping to allow the vacuum on the well point and vapor flow rate to be monitored and ensure a continuous draw of a vapors from the purge stream into or through the selected collection apparatus.

In practice the volume of vapors purged and the mass of VOCs in the purge stream will be relatively small. The system should be exhausted to outdoor air in order to avoid adverse impacts to indoor air and downwind of nearby receptors.

The components of the setup depicted in Figure 7-18 include (upstream to downstream):

  • Two-inch Schedule 40 PVC;
  • 1/4-inch wedge valve near the intake "T", with tubing connected to a Dwyer Magnehelic Gauge (0-100 in-H2O) pressure/vacuum gauge;
  • Summa sample valve (1/4-inch wedge valve with Teflon tubing);
  • PID meter and O2/CO2 meter port (1/4-inch wedge valve equipped with Teflon tubing to the PID meter) – opposite side of Summa Port (peristaltic pump used to overcome vacuum imposed on purge stream and draw influent to the PID and O2/CO2 meter port);
  • Pitot Tube port (3/8-inch ID threaded pipe, ½-inch length);
  • Flow meter port (3/8-inch ID threaded pipe, one-inch length).

A summa canister is used to collect a sample from the LVP purge stream in Figure 7-17 and Figure 7-18 (Section 7.8.1). The system could also be designed to allow a continuous stream of vapors to pass through a sorbent collection tube (Section 7.8.2). No experience with the use of sorbent tubes in LVP sample collection systems was available at the time this guidance was prepared. Discussions with consultants familiar with the approach suggest that the use of sorbent tubes to collect LVP samples would be limited due to the vacuum imposed on the purge train.

The PID sample port should be placed immediately opposite the summa sample port and used to monitor oxygen, carbon dioxide and total VOC concentrations periodically using a Tedlar bag and vacuum chamber (“lung box”) during each purge event (see Figure 7-18). This ensures that PID readings will be representative of the collected samples.

Installation of an Averaging Pitot Tube (“Pitot tube”) upstream of the flow meter port is recommended in order to confirm flow rates based on a thermal, anemometer flow meter. A minimum ten-pipe-diameter upstream separation distance and five-pipe-diameter downstream separation distance between any fittings or ports and the Pitot tube and flow meter should be maintained, in order to minimize turbulence that could affect the accuracy of the instruments (see HDOH 2017c).

A limited amount of low-volatile, non-chlorinated, PVC cement can be used at joints in the piping, if necessary to ensure secure fittings. If used, then the vapor purge system should be allowed to aerate for several days prior to use in the field and tested with a PID to confirm that no VOCs are present in the piping and fittings.

Carry out a shut-in test on the sampling train to ensure that no leakage is occurring around joints and fittings. This will be similar to the shut-in test described in Section 7.10.5.1 for soil vapor sampling trains in general. Test the vacuum of the pump (e.g., Shop-Vac) to be used for sample collection by attaching it directly to a vacuum gauge and measuring the vacuum drawn for at least 15 seconds. Connect the pump to the LVP sampling train, close the valve to the extraction point, and re-measure the vacuum drawn for at least 15 seconds. Compare this value to the vacuum previously measured. A difference of greater that 10% indicates a significant leak somewhere in the sampling train. Correct any problems identified and redo the shut-in test. The results of the shut-in test and bench test should be documented and described in the report for the LVP investigation.

Once the system is deemed to be tight, carry out a bench test to optimize the design of the system and evaluate the purge rate under different vacuums imposed on the vapor entry point. Ensure that the flow meter(s) are functioning properly. In the HDOH (2017c) field study, purges were directed into a spherical, latex-rubber weather balloon in order to verify the accuracy of flow meters attached to the purge stream. Verifying the precision of flow rate measurements is critical, since the DU volume of vapors purged per sample is a critical part of sample collection and data evaluation.

A vacuum of between 30 and 40 in-H2O is typical of field conditions when a 6.5 HP Shop-Vac is used to purge a well point (see also McAlary et al. 2010). Note that this is well below the maximum recommended vacuum to be applied to a vapor sample point of 100 in-H2O or seven inches of mercury (in-Hg), intended to avoid stripping of vapors from free product entrained in soil (refer to Section 7.10.3.2; see also CalEPA 2015). A flow rate of 300 to 3,000 liters per minute (10 to 100 standard cubic feet per minute) is typical, depending on the permeability of the material below the floor slab. This corresponds to an estimated purge duration of approximately 1 to 10 minutes to collect an LVP sample from a default, 3,000 liter DU volume of subslab vapor. Smaller diameter piping will require a longer purge time for any given sub-slab permeability because of frictional losses. As discussed in Section 7.8.4.5, correspondence of the purge time with the flow rate of the Summa canister is critical when a canister is used for the collection of an LVP sample.

7.8.4.4 Extraction Point Installation

An example extraction point design is depicted in Figure 7-19. (Note that the building had been removed prior to LVP sample collection.) The LVP extraction point should be installed in a manner that allows access to the targeted depth interval beneath the slab and prevents downward leakage of indoor air during purges of subslab vapor. In the HDOH (2017) LVP field study, the extraction point was constructed as a two-inch PVC well, set within an eight-inch diameter steel casing installed to from the surface to the base of the slab. The latter was installed due to the vulnerability of the sampling point in the field to surface traffic (building previously removed). A narrower diameter installation and even narrower diameter piping will likely be adequate for most investigations.

The LVP extraction well in the HDOH (2017c) study was constructed with 10-slot, two-inch diameter screened PVC with a solid end-cap. Smaller- or larger-diameter well points might also be practical. The well screen should be extended from the base of the concrete pad to the depth of targeted, subslab vapor DU (e.g., 12 inches). Include a solid endcap on the well point in order to help focus the draw area of influence to the targeted layer of soil. Install a sand pack to a height of several inches above the top of the well screen. Add a minimum, two-inch layer of hydrated bentonite above the sand pack. Seal the gap between the extraction point and the slab with Portland cement to further prevent downward leakage. A water dam can be used to confirm the absence of leakage in this seal (see Section 7.10.5.2). Build a dam around the seal and add water. The water should not disappear during the LVP purge. The top of the extraction point can be installed either below or above the top of the slab, depending on the needs of the investigation. Fit the top of the tube with a solid PVC screw cap in order to seal and secure the top of the well.

Alternative extraction point installation designs are possible, provided that the objectives of LVP sample collection are met. Alternative designs should ensure that collection of vapors from the targeted subslab DU interval is optimized, that leakage around the extraction point is minimized, that the resulting flow rate is compatible with the sample collection method and that the resulting samples will be sufficiently representative.

7.8.4.5 LVP Sample Collection

Figure 7-20 depicts a completed, LVP sample collection system set up taken from the HDOH (2017) field study. A field pilot test should be carried out to estimate the flow rate of the LVP under sample collection conditions. The test should be kept as short as possible in order to limit disturbance of subslab vapors. A duration of less than 60 seconds is anticipated to be adequate.

Record the flow rate and estimate the time required to complete the targeted, DU purge volume. Ensure that the minimum draw time is greater than the time required to complete the LVP purge if a Summa canister is to be used to collect the LVP sample (minimum draw rate typically 200 ml/minute). Six liter canisters are recommended in order to ensure that the canisters do not fill prior to completion of the time required to achieve the targeted purge.

The collection of LVP samples can begin immediately after the purge test. Carry out a final shut-in test to ensure the tightness of the sampling train. Connect the sample collection apparatus (e.g., Summa canister) to the sampling train. Carry out a shut-in test on the sampling train itself to ensure tightness. Attach additional sample collection equipment (e.g., Summa canisters or sorbent tubes) to the sampling train as needed to collect replicate samples in accordance with the investigation work plan.

Turn on the pump attached to the LVP system while simultaneously opening the Summa canister valve or the connection to a sorbent tube collection system. This will allow a continuous portion of the purge stream to enter or pass through the sample collection device. Collection of a concurrent, indoor air sample(s) in the immediate vicinity of the LVP well point is also recommended (Section 7.7). The sample should be tested for targeted, subslab VOCs as well as oxygen, carbon dioxide, substances used for leak detection (e.g., isopropyl alcohol) and other gases that might prove useful for evaluation of potential downward leakage of indoor air through the slab and into the LVP sampling point during purges (Section 7.8.4.6).

Include tests for leaks at the well point and at connections in the LVP sampling train upstream of the sample collection ports (Section 7.10.5). This could include placement of rags soaked in isopropyl alcohol (standard rubbing alcohol) around the wellhead extraction point connection, fittings upstream of the vacuum gauge, and fittings immediately downstream of the vacuum gauge and prior to the sample collection ports (see Figure 7-20). Accidental contamination of sampling containers and subsequent contamination of laboratory equipment and bias of test results can be difficult to avoid, however. A simple shut-in test is considered adequate by many field experts.

Record flow data and the vacuum at the extraction point for each purge at an interval adequate to document sampling conditions. A series of readings at the beginning of a purge until conditions stabilize (e.g., every 30 to 60 seconds) followed by a reading at the mid-point and end of the purge is recommended. Use a PID and landfill gas meter with a Tedlar Bag and Lung Box to periodically (or continuously) monitor oxygen, carbon dioxide and if feasible total VOCs during the purge. Record the time required to achieve the target DU purge volume. Cease sample collection if leakage of indoor air into the LVP train is suspected.

Record the starting and final vacuum of the summa canister. Use this to estimate volume of the vapor sample collected in the canister. Discuss minimum sample volume necessary to meet testing requirements with the laboratory prior to sample collection (typically 1-2 liters).

Turn off the LVP sampling train pump when the target purge volume has been reached. Immediately close the Summa canister valve (or port to sorbent tube) as well as the valve to the vapor extraction point. Disconnect the sample collection apparatus (e.g., Summa canister or sorbent tube). Connect the apparatus for collection of the next LVP series sample. Repeat the same steps noted above until the full series of LVP samples targeted for the subject purge point have been collected.

Submit the LVP samples to the laboratory for analysis. Ensure that the samples are tested for oxygen, carbon dioxide and any other gases used to assess potential leakage in addition to VOCs targeted as part of the vapor intrusion investigation.

7.8.4.6 Data Quality Control

Field quality control should include (Section 7.10.5): 1) Shut-in test of LVP sampling train prior to and after connection to vapor extraction; 2) Leak testing of sampling train using isopropyl alcohol or comparable method throughout each purge event; 3) Collection of a background indoor air sample(s); 4) Collection of O2, CO2 and other potential tracer gas data for preliminary subslab vapors prior to sample collection and as part of all LVP and background indoor air sample analyses; and 5) Collection of triplicate LVP sample(s) for the first purge of a sampling event if a non-continuous draw method is used to collect an LVP sample.

Test for and record oxygen and carbon dioxide levels in subslab vapors at the well point prior to the collection of LVP samples. These data will be important for assessment of potential leakage of indoor air into the system during LVP sample collection. All LVP samples should likewise be tested for O2, CO2 and other potential gases that could prove useful in leakage tests (e.g., indoor air contaminants not anticipated to be present in subslab vapors).

Evaluation of the overall integrity of the sampling train during LVP sample collection should be used in conjunction with preliminary subslab sample data, field data recorded during LVP sample collection and LVP sample and indoor air sample data to assess the magnitude of indoor air leakage into the sampling train during purge events. Oxygen levels in subslab vapors are often depleted in comparison with indoor (and outdoor) air. This is accompanied by a typical increase in carbon dioxide levels in subslab vapors.

These observations and data can be used to assess the relative magnitude of leakage into the sampling train during purge events. A leakage rate of <10% is considered insignificant in terms of data quality and use of the data to assess potential vapor intrusion risk (i.e., >90% of sample volume represented by subslab vapors). An absence of significant isopropyl alcohol in the samples implies minimal leakage at these points. Consistent depletion of carbon dioxide in LVP samples in comparison to indoor air is a particularly useful indicator of minimal leakage. Comparison of other tracer gasses found in indoor air but absent or at significantly depleted levels in pre-sample collection, subslab vapors might also prove very useful (e.g., TPH, BTEX, non-targeted solvents, etc.)

The collection of concurrent, replicate samples during an LVP purge to test data representativeness and reproducibility is not necessary for continuously collected samples, since vapor from 100% of the purge stream is included in the resulting data (i.e., replicate samples not normally needed). At least one set of replicate samples (triplicates) per LVP collection event is, however, recommended for sample collection methods that involve only periodic testing of vapors from the purge stream. For example, a small “increment” of vapor might be allowed to enter the sample collection system (or field testing equipment) every minute or some fraction of a minute. In this case the resulting data represents the mean of the vapor increments collected and the representativeness of the complete purge stream cannot be directly assured. The collection of concurrent triplicate samples will allow the precision of a single LVP sample data point to be tested in a manner similar to that applied to the collection of replicate Multi Increment soil and sediment samples (Section 4.2.7). This assumes, of course, that the samples were collected in a scientifically valid manner to begin with.

7.8.4.7 LVP Investigation Report

Information to be provided in the LVP investigation report includes:

  • Site background and summary of existing data;
  • Rationale for targeted DU volume of subslab vapors to be characterized and selection of LVP sample collection point;
  • Summary of sample collection methods;
  • Summary of data quality control measures, including leak detection;
  • Summary of data for targeted VOCs;
  • Investigation conclusions, including evaluation of potential vapor intrusion risks and any limitations on data reliability;
  • Field photographs;
  • Laboratory reports;
  • Field data sheets.

Summary information for an LVP investigation can be included as part of a larger investigation provided that all necessary information is provided.

7.8.5 Flux Chamber Sampling

Flux chambers are enclosures that are placed directly above on the surface (e.g., ground, floor) for a period of time and the resulting contaminant concentration in the enclosure is measured (Kienbusch 1986, Eklund 1992, Hartman 2003, ITRC 2007; Figure 7-21). Flux chambers were originally designed to estimate vapor emissions from open waste pits in terms of mass per unit area per unit time (e.g., mg/m2-hour). This method offers advantages in some cases because it yields the actual flux of the contaminant out of the ground, which eliminates some of the assumptions required when using other types of subsurface data in vapor intrusion models. Unlike soil vapor or indoor air samples, flux chamber data can be used to definitively identify and document the emission of vapors from subsurface sources to the atmosphere or to the interior of buildings. The method has long been used by regulatory agencies at hazardous waste sites and it is widely used for measuring trace emissions from natural soils.

HDOH considers its quantitative value for soil vapor and vapor intrusion assessments to be limited, and HDOH should be contacted prior to the use of flux chambers in site investigations or vapor intrusion studies. Flux chambers are primarily useful as a qualitative tool to locate surface fluxes of VOC contamination and entry points into structures. This is due in part to the small area tested and difficulty in capturing the heterogeneity of subsurface vapors, as well as short term temporal variations in downward versus upward vapor flux (e.g., due to changes in barometric pressure). Use in open areas also does not mimic vapor flux into buildings. The presence of small-scale, preferential pathways in soils (e.g., desiccation cracks, root structures, soil heterogeneity, etc.) to optimize placement of the chambers is also difficult to identify in the field.

The testing is typically conducted in one of two modes: static or dynamic. In dynamic systems, a sweep gas is introduced into the chamber to maintain a large concentration gradient across the emitting surface. The effluent air from the chamber is collected using canisters and analyzed for chemicals of concern. The method is best suited for situations where large fluxes are anticipated. In static systems, a chamber is placed on the ground or floor and the contaminant concentration build-up is measured over time. This method is best suited for situations where lower fluxes are anticipated.

Flux chambers are not well-suited for structures with covered floor surfaces such as single family residences, because the primary entry points of soil vapor into the structure (cracks, holes, sumps, etc.) are often concealed by floor coverings, walls, stairs, etc. For structures, the method has more application to larger industrial and commercial buildings with slab-on-grade construction where the slab is mostly uncovered. A building survey using a real-time analyzer or on-site GC can be used to attempt to identify the primary locations of vapor intrusion.

Regardless of the method used, enough chamber measurements should be collected to get a representative value under the footprint of the building (analogous to placing enough borings on a typical site), and ensure that they are located near edges where the slab meets the footing, over any zones with cracks or conduits, and over the center of the contamination if known. In all cases, it is recommended that chambers should be deployed for long enough periods to enable temporal variations to be assessed, similar to indoor air measurements (8 to 24 hours depending upon the conditions; 24 hours if large temperature differences exist between day & night) (SDC 2011).