Guest Author – Written by Bart Eklund, The Global Practice Leader for Vapor Intrusion at AECOM
Active and passive sampling are the two primary approaches for collecting a soil-gas sample for subsequent off-site analysis. For an active sample, soil gas typically is withdrawn from the soil using the vacuum within an evacuated, stainless-steel canister or pulled through a sorbent using a pump. For a passive sample, a sorbent typically is placed in the ground where it will encounter soil gas that moves through the soil matrix via molecular diffusion or advection (pressure-driven transport).
Note that both active and passive methods may employ sorbents. Commonly used sorbents for active sampling include Carbotrap (e.g., TO-17), activated charcoal (e.g., radon), and silica gel (e.g., methanol). So, while all passive sampling involves sorbents, not all sorbent sampling is passive.
Relative Advantages and Disadvantages of Canister Sampling
The default approach in the US for soil-gas sampling to evaluate potential vapor intrusion (VI) has long been to collect active samples in canisters for subsequent off-site GC/MS analysis using USEPA Method TO-15 in full-scan mode. There are several reasons for this: 1) the standard TO-15 analyte list includes the primary risk drivers for VI at many sites (in particular benzene, trichloroethylene [TCE], and tetrachloroethylene [PCE]), 2) the approach is capable of addressing a large number of volatile organic compounds (VOCs), 3) the detection limits are sufficient to determine whether screening levels are exceeded or not, 4) the method is applicable over a very wide range of concentrations, 5) multiple analytical methods can be run on a single sample, 6) the infrastructure of testing laboratories, sampling media, cleaning procedures, etc. was already in place for ambient (outdoor) air studies when VI emerged as a topic, 7) background contamination from sampling media generally is not an issue, and 8) the sampling does not require field crews with extensive air sampling experience.
The relative disadvantages of the canister/TO-15 approach include: 1) it is not appropriate or well-suited for certain analytes, such as semi-volatile organics (SVOCs), and 2) drill rigs or extensive hand augering are needed to collect exterior soil-gas samples at desired depths (generally 5 ft. [1.5m] below ground surface [bgs] at a minimum).
Relative Advantages and Disadvantages of Passive Sorbent Sampling
Passive sampling approaches are the best option for some tasks. The primary advantages of passive sorbent sampling are: that the installation is relatively simple and can be performed using a hand drill. This can result in substantial cost savings relative to using a drill rig. Other advantages include: 2) the sampling is extremely simple and requires little experience or technical knowledge (so simple a caveman or a geologist could do it), and 3) the sampling media are very small, which simplifies packing and shipping.
The ability to provide time-integrated data over a one- or two-week period is sometimes cited as an advantage. For most scenarios, however, there is no evidence of short-term temporal variability in soil gas concentrations. So, a grab sample and a one-week sample should yield equivalent results.
With any sorbent method, there is a need to address potential background contributions from the sorbent. This is typically achieved by analyzing multiple method blanks or trip blanks. Also, as with any sorbent method, it is important to collect enough mass to achieve an adequate detection limit but not so much mass as to saturate the sorbent. If a realistic estimate of the soil-gas concentration can be made, the sampling duration can be optimized.
There are several potential limitations specific to passive sorbent sampling of soil gas that should be kept in mind. 1) The sampling depth tends to be relatively shallow (e.g., 18 in. [50 cm] bgs). Therefore, the data are somewhat more prone to the effect of environmental variables such as infiltrating rainwater than deeper samples. 2) There is no simple way to perform a leak check. Therefore, the potential for a poor installation to result in a low bias due to intrusion of atmospheric air cannot be directly assessed. 3) The analysis yields emission rate data, with units of mass/time, for a given VOC. If a number of samples are collected over the same time period, the data set provides the spatial variability of mass. The data can be converted to a concentration (e.g., ppbv) if the uptake rate of the sampler for a particular VOC is known. For tight clayey soils, however, the calculated concentration will be biased low if the transport of vapors through the soil is slower than the uptake rate of the sampler. In other words, the vapors surrounding the sampler are not replenished as fast as the vapors are sorbed. This is usually termed the “starvation effect.” For the above reasons, many regulators do not accept passive soil gas data for input to a human health risk assessment.
Field Applications of Passive Sorbent Sampling
The best approach to use for a given situation will depend on the objectives of the study (i.e., what question is being addressed?). For sub-slab soil-gas sampling, either active or passive methods are good options. The choice may come down to the specific analytes of interest, whether there are any schedule constraints, and the experience of the sampling crew.
Historically, the primary use of passive soil-gas sampling has been exterior soil gas sampling to map out the plume of impacted groundwater (and/or impacted soil) relatively early in the investigative process at a site. For example, one or two passive samplers per residential lot have been used in a neighborhood of about 100 single-family residential buildings to refine the conceptual site model. The data set allows areas of relatively low, medium, high, and very high impacts to be color-coded and mapped in 2-D. This can be used by decision-makers to screen out certain portions of the neighborhood from further consideration and identify locations where follow-up exterior soil-gas sampling or testing inside of buildings is warranted. Where both active and passive soil-gas data have been collected at the same location, the data tend to show good agreement.
Mapping studies often rely upon sampling along utility right-of-ways (e.g., between sidewalks and streets) to minimize the need to get signed access agreements. These locations, however, may yield false positive results near storm/sanitary sewer lines, which often are not leak-tight and may have released small amounts of VOCs in the past. This can be addressed by avoiding sampling in proximity to sewer lines or by collecting multiple samples to define the lateral concentration gradient radiating out from the sewer line.
Bart Eklund is the Global Practice Leader for Vapor Intrusion at AECOM. He has developed both sampling approaches and analytical methods during his 40-year career.