An Overview of State Approaches to VApOr IntruSIOn

Copyright 2007 Air & Waste Management Association10 em february 2007 awma.org

emfeature

No single government agency has overall responsibility for vapor intrusion (VI) in the United States. Oversight is divided up among various federal, state, and local agencies, depending on the site in question and its status. Outside the United States, multinational corporations are addressing VI at some sites, but the work typically is not regulatory driven.

The U.S. Environmental Protection Agency (EPA) issued draft VI guidance in November 2002,1 but it is not applicable to all sites. It was intended for screening-level evaluations for Resource Conservation and Recovery Act (RCRA), Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), and brownfields sites and it is not applicable at sites regulated under EPA’s Underground Storage Tank (UST) program. Furthermore, as stated in the guidance, “OSHA [the Occupational Safety and Health Administration] generally will take the lead role in address-ing occupational exposures.”

The EPA guidance uses a three-tiered approach to evalu-ate sites, which includes comparing data for one or more media to conservative screening levels. The screening levels were developed using assumed attenuation coefficients (α values), where α is the concentration ratio for a given compound (e.g., indoor air or soil gas).

Only a few states—Massachusetts, Connecticut, New Jersey, and Maine—had VI guidance or regulations prior to EPA’s November 2002 draft guidance. The interest in VI since that time, however, has become more widespread

low groundwater or frozen soils.The authors of this article are members of the ASTM

E50 VI workgroup. As part of the effort to develop an ASTM method for VI, we have reviewed the various state guidance documents and tabulated key information to better under-stand how the various states approach VI issues. The informa-tion is believed to be accurate as of June 2006. We recognize that any such summary is a point-in-time snapshot of the guidance—especially in a fast-moving area such as VI—but we think it a useful exercise.

SUMMARY OF STATE VI GUIDANCEWe completed a list of states with VI guidance and devel-oped a spreadsheet that identified key elements generic to one or more guidance documents. The authors reviewed the guidance for selected states and populated the spread-sheet. The results are summarized below, followed by a brief discussion.

State Documents Included in ReviewTwenty-three states were included in the review. These included all states known to have VI guidance plus six states—Alabama, Florida, Georgia, North Carolina, South Carolina, and Tennessee—that largely defer to the EPA draft guidance.

Screening ValuesAll state guidance documents except one contain screening values based on groundwater, soil, or soil gas concentrations. The lone exception is New York. The types of screening values used by each state are summarized in Table 1.

The assumptions used to calculate screening values vary from state to state. For example, New Jersey uses an assumed building ventilation rate of 0.25 air changes per hour (ACH) for both residential and industrial buildings, whereas Loui-siana uses 0.5 ACH for residential buildings and 0.83 ACH for industrial buildings.

Bart M. Eklund, CIH, is principal scientist with URS Corp., Austin, TX; David J. Folkes, P.E., is president of EnviroGroup Ltd., Centennial, CO; Julie Kabel, MPH, is human health risk assessor with ENSR International,

Westford, MA; and Rachel Farnum is environmental engineer with General Electric, Niskayuna, NY.

E-mail: [email protected].

An Overview of State Approaches to

VApOr IntruSIOn

and numerous state regulatory agencies have issued guidance. As of mid-2006, roughly two-fifths of the 50 states have state-specific guidance. The authors of these state documents were certainly aware of the draft EPA guidance and conscious decisions presumably were made to either model the state guidance on the EPA guidance or to employ an alternative approach. In many cases, efforts were made to address issues not fully covered in the EPA guidance. In addition, the state guidance sometimes addresses variables that might be more relevant to their particular state than to the nation as a whole, such as very deep or very shal-

An Overview of State Approaches to

VApOr IntruSIOn

by Bart Eklund, David Folkes, Julie Kabel, and Rachel Farnumby Bart Eklund, David Folkes, Julie Kabel, and Rachel Farnum

Copyright 2007 Air & Waste Management Associationawma.org february 2007 em 11

Some, but not all, states exclude sites from screening if they are more than a specified distance from the source of contamination (i.e., the source of the vapors). A 100-ft lateral or vertical separation value is used by five states—California, Colorado, New Hampshire, New Jersey, and Pennsylvania. New Hampshire reduces the 100-ft value to 30 ft for petroleum contamination. New Jersey also reduces the separation value to 30 ft for petroleum contamination, but excludes free product (i.e., petroleum spills or releases). A 100-ft lateral separation value is used by Alaska and Ohio, which matches the EPA guidance. Indiana uses 100 ft for chlorinated compounds and 50 ft for benzene, toluene, ethylbenzene, and xylene (BTEX). Two states use a 30-ft distance (Connecticut and Massachusetts). The Massachu-setts 30-ft value applies only to horizontal separation; 15 ft is used for vertical separation.

Screening values for three selected compounds—ben-zene, trichloroethylene (TCE), and tetrachloroethylene (PCE)—are shown in Table 2. Readers are urged to read the specific state documents for further clarification, as individual states may have more than one set of screening values that apply to residential settings.

As seen in Table 2, the screening values may vary from state to state by three or four orders of magnitude for groundwater and soil gas. There is less variability from state to state for indoor air screening levels, presumably because indoor air is a more direct measure of exposure requir-ing fewer assumptions. Even for indoor air, however, the differences from state to state can be significant given the absolute magnitude of the screening values compared with typical background levels in both indoor air and ambient air. Also, for compounds such as TCE the screening values may approach analytical reporting limits in some cases.

The differences in screening levels from state to state are due, in part, to differences in the risk level that is considered to be acceptable and the attenuation coefficient (α) that is assumed. These topics are addressed below.

Risk LevelsExcept for New York, the various state guidance documents use a specified acceptable risk level for individual carcino-gens of 1E-05 or 1E-06. The risk levels used by each state are summarized in Table 3. Eleven states—Alaska, California, Colorado, Connecticut, Louisiana, Maine, Massachusetts, Minnesota, New Hampshire, Oklahoma, and Oregon—use 1E-06 for screening purposes, while five states—In-diana, Maine, Michigan, Ohio, and Pennsylvania—use 1E-05. New Jersey uses the higher of 1E-06 or the TO-15

Roughly two-fifths of the 50 states have state-specific VI guidance.

Some states preclude the use of groundwater screening values at sites with very shallow groundwater. The minimum distance below foundation to groundwater varies from 2 ft for New Jersey to 15 ft for Connecticut and New Hampshire. Four states—Alaska, Indiana, Ohio, and Pennsylvania—use a 5-ft distance, which matches the value used in the 2002 EPA guidance.

Table 1. Types of screening values used in state VI guidance.

Types of Generic Screening Values

Number Ground- Shallow Deep Indoor Non- ofState water Soil Soil Gas Soil Gas Air Residential VOCs Comments

Alaska Yes No Yes Yes Yes No 100+ California No No Yes Yes Yes Yes 18 Colorado Yes No No No Yes No 50+ Connecticut Yes No Yes No Yes Yes 47 Can adjust for soil type and site-specific dataIndiana Yes Yes Yes No Yes Yes 50+a Can adjust for site-specific dataLouisiana Yes Yes No No Yes Yes 130+ Can adjust for soil type and site-specific dataMaine No No No No Yes No 7 Massachusetts Yes No Yes No Yesb No 50+c Can adjust some values for soil type and site-specific dataMichigan Yes Yes Yes Yes Yes Yes 50+ Minnesota No No Yes Yes Yes No 50+ New Hampshire Yes No Yes Yes Yes Yes 31 Can adjust for site-specific dataNew Jersey Yes No Yes No Yes Yes 50+ Can adjust for soil type and site-specific dataNew York No No No No Yes No 5 Ohio Yes No Yes Yes Yes Yes EPA list Same screening values and adjustments as in EPA guidanceOklahoma Yes Yes Yes Yes Yes No 50+ Oregon Yes Yes No No Yes Yes 32 Pennsylvania Yes Yes No No Yes Yes 100+ Do not use at sites with free liquid

Notes: aIndoor air - four to five for other media. bBackground values. cSoil-gas screening values for seven compounds.


reporting limit. Only five states—Colorado, Louisiana, Massachusetts, New Hampshire, and Pennsylvania— address cumulative risk from multiple compounds. For noncarcinogens, all states use a hazard quotient (HQ) of 1 or a hazard index (HI) of 1, except for New Hampshire and Massachusetts, which use a value of 0.2. HQ values apply to individual compounds and HI values apply to cumulative hazard from multiple compounds.

New York uses a risk level of 1E-06 when evaluating TCE data, selecting an acceptable indoor air concentration (i.e., 5 µg/m3) that falls within their acceptable risk range of 1E-04 to 1E-06. New York considers only noncarcinogen effects when evaluating PCE data. Massachusetts uses typical background concentrations in place of risk-based screening values for indoor air, if the background concentration is higher.

States typically use multiple sources of toxicological data, with an emphasis on EPA sources such as the Inte-grated Risk Information System (IRIS), National Center for Environmental Assessment (NCEA), Health Effects Assessment Summary Tables (HEAST), and the Region III Risk Based Concentration (RBC) screening levels. Other sources of toxicological values include ATSDR’s Minimal Risk Levels (MRLs) and values developed by state agencies. The source that is used can be a significant decision for compounds such as TCE, where there may be two orders of magnitude difference in concentration for a given risk level, given differences in the underlying cancer slope factor.2

Attenuation CoefficientsThe existing EPA guidance in Tier 2a uses default attenu-ation coefficients (α) of 0.1 for shallow soil gas or subslab

soil gas, 0.01 for deep soil gas, and 0.001 for groundwater (applied to the equilibrium soil gas concentration based on the Henry’s Law coefficient for the compound). EPA is expected to revise these values and use an α of 0.02 for sub-slab soil gas and possibly shallow, exterior soil gas as well. (The applicability of shallow exterior soil gas is currently a matter of debate within the community of regulators.) Shallow soil gas is typically defined as <5 ft below the base of the foundation. Deep soil gas in this context implies data from >5 ft below the base of the foundation.

The attenuation coefficients used by each state are summarized in Table 4. In general, the state values are similar to the EPA values. Six states (California, Connecticut, Michigan, New Hampshire, New Jersey, and Pennsylvania) use less conservative values for shallow soil gas than the 2002 EPA value. One state (Indiana) gives separate α values for subslab soil gas and shallow soil gas and two states (Colorado and Oklahoma) indicate that their shallow soil gas value is based on subslab soil gas. California has different default values for different building types.

Only three states (New Jersey, Massachusetts, and New Hampshire) apply an adjustment to account for aerobic degradation of aromatic hydrocarbons, such as BTEX. In each case, the α values are adjusted by a factor of 10 to make them less conservative. New Hampshire applies the correction only to groundwater data.

The 2002 EPA guidance assumes that there is no attenuation between a crawl space and the overlying build-ing (i.e., α = 1). Six states (California, Colorado, Indiana, New Hampshire, New Jersey, and Oklahoma) explicitly address crawl spaces and each uses α = 1.

Table 2. Residential screening levels for selected VOCs.

Benzene TCE PCE

State Groundwater Soil Gas Indoor Air Groundwater Soil Gas Indoor Air Groundwater Soil Gas Indoor Air

Alaska 5 3.1 0.31 5 0.22 0.022 5 8.1 0.81California NA 36.2 0.084 NA 528 1.22 NA 180 0.41Colorado 15 NA 0.23 5 NA 0.016 5 NA 0.31Connecticut 130 2490 3.3 27 752 1 340 3798 5Indiana 95-850 250-1400; 2.5 4.6-700 120-2000; 1.2-4.1 7.4-1100 320-5200; 3.2-10 25-140a 2-200a 32–520a

Louisiana 2,900 NA 12 10,000 NA 59 15,000 NA 110Maine NA NA 10b NA NA NA NA NA NAMassachusetts 2000 NA 0.3 30 NA 1.37 50 NA 0.04Michigan 5600 150 2.9 15,000 700 14 25,000 2100 42Minnesota NA 1.3-4.5 1.3-4.5 NA NA NA NA NA 20New Hampshire 2000 95 1.9 50 54 1.1 80 68 1.4New Jersey 15 16 2 1 27 3 1 34 3New York NA NA NA NA NA 5 NA NA 100Ohio 14 31 3.1 -- 122 12.2 11 81 8.1Oklahoma 5 3.1 0.27 5 0.17 0.017 5 0.33 0.33Oregon 160 NA 0.27 6.6 NA 0.018 78 NA 0.34Pennsylvania 3500 NA 2.7 14,000 NA 12 42,000 NA 36

Notes: Units are µg/L for groundwater and µg/m3 for soil gas and indoor air. See individual state guidance documents for additional information, including limitations and exceptions. Trigger or action levels for mitigation based on indoor air concentrations may be higher than the screening levels shown. aSecond range of values shown is for subslab soil gas. bChronic exposure value.

Copyright 2007 Air & Waste Management Associationawma.org february 2007 em 13

Trigger LevelsSeveral states have trigger levels that call for immediate action or mitigation if exceeded. These trigger levels typically are based on indoor air concentrations, but also can be based on soil gas concentrations or a combination of indoor air and soil gas. Among the states that include trigger levels in their guidance are Colorado, Indiana, Maine, Minnesota, New Jersey, and Ohio. The decision matrices used by New York also can be thought of as trigger levels. The trigger levels may be based on acute (short-term) risk, but Ohio uses 1E-05 to trigger immedi-ate action and Colorado uses 1E-05 to trigger mitigation for most compounds (1E-04 is used for TCE and PCE due to background considerations).

DISCUSSIONMost states that have issued VI guidance have patterned their guidance on the 2002 EPA guidance to the extent that they have used a tiered evaluation process with the initial tiers employing conservative screening values for multiple types of media that are risk-based concentrations. Evalu-ation of VI begins—and often ends—with data collected outside the building. A line-of-evidence approach may be used with soil gas monitoring as one key source of informa-tion. The use of modeling with site-specific data appears to have been superseded to a large extent by screening values based on conservative assumptions. This trend may be reversed when the revised EPA spreadsheets are released, which incorporate constraints to ensure reasonably con-servative results.

The exception among the states is New York, whose guidance is unique. New York employs a risk manage-ment approach that limits incremental exposure above assumed background levels rather than estimating the indoor air concentration due to VI and calculating the risk for that fraction. New York also is unique in mov-ing immediately to measurements inside the building if any volatile organic compounds (VOCs) are detected in soil gas at a site. At this point, however, the New York guidance only addresses five compounds or classes of compounds and the decision matrices only cover three VOCs.

There is a great deal of variation from state to state regarding the level of detail included in the VI guidance. New Jersey arguably has the most comprehensive guid-ance in terms of addressing specific issues or scenarios of potential interest. The various state guidance documents generally indicate that alternative approaches will be considered. In actual practice, regulatory staff members may be reluctant to deviate from the written guidance with the degree of reluctance roughly inversely propor-tional to their level of experience with VI sites.

While noting the differences from state to state, we offer no opinion as to which state guidance is “best” or which is “right” in terms of technical issues. Each state is free to choose what risk levels it finds acceptable, so inevitably there will be differences from state to state. Ultimately, there can be a trade-off between conserva-tiveness and effectiveness. The public may tend to judge more conservative screening values as being better, but

Table 3. Risk values used for screening purposes.

Risk Levels

Cancer Cumulative Non-Cancer State Risk Cancer Risk Risk Source of Toxicological Data Comments

Alaska 1.E-06 Not addressed HI = 1 EPA Nov 2002 VI Guidance California 1.E-06 Not addressed HQ = 1 OEHHA (Cal/EPA), IRIS, HEAST OEHHA is the primary sourceColorado 1.E-06 1.E-05 HQ = 1 EPA Region 3 1,1-Dce from CDPHEConnecticut 1.E-06 Not addressed HQ = 1 IRIS, HEAST, ATSDR MRL, CalEPA Indiana 1.0E-05 Not addressed HQ = 1 IRIS, HEAST, EPA Provisional values, internal IDEM values Louisiana 1.E-06 Risk Range = HI = 1 IRIS, NCEA, HEAST, etc. 1E-04 to 1E-06 Maine 1.E-05 Not addressed IRIS, ATSDR MRL Massachusetts 1.E-06 1.E-05 HI = 0.2 IRIS, HEAST, EPA Provisional values, MADEP Michigan 1.E-05 Not addressed HQ = 1 IRIS, HEAST, MDEQ internal values, etc. Minnesota 1.E-06 Not Addressed HQ = 1 MDH, EPA-IRIS RFCs New Hampshire 1.E-06 1E-06 risk HQ = 0.2 EPA Hierarchy and HQ=0.02New Jersey 1.E-06 Not addressed HQ = 1 EPA Region 3 Higher of 1E-06 or TO-15 RLNew York 1E-04 to 1E-06 NA NA NA Ohio 1E-05 Per EPA HI = 1 Same as EPA 2002 except uses Cal EPA for TCE Oklahoma 1.E-06 Not addressed HQ = 1 EPA Region 6 Oregon 1.E-06 Not addressed HQ = 1 IRIS, NCEA, OEHHAPennsylvania 1.E-05 1.E-05 HQ = 1 Not specified

Copyright 2007 Air & Waste Management Association


Table 4. Attenuation values used in state VI guidance.

Attenuation Coefficients

State Groundwater Shallow Soil Gas Deep Soil Gas BTEX Crawl Spaces

Alaska 0.001 0.1 0.01 NA NACalifornia NA 0.01 - 0.002 same as shallow NA 0.002Colorado NA 0.1 (subslab) NA NA 1Connecticut 0.001 0.001 NA NA NAIndiana NA subslab = 0.1 soil gas = 0.01 0.01 NA 1Louisiana NA NA NA NA NAMaine NA NA NA NA NAMassachusetts Based on J&E model NA NA Adjusted by 10x NAMichigan Based on J&E model 0.02 0.002 NA NAMinnesota NA NA NA NA NANew Hampshire Based on J&E model 0.02 0.02 Groundwater values adjusted by 10x 1New Jersey Based on J&E model 0.02 NA 0.002 1New York NA NA NA NA NAOhio 0.001 0.1 0.01 NA NAOklahoma 0.1 (subslab) 0.1 (8-10 ft) NA 1Oregon 0.002 NA NA NA NAPennsylvania NA 0.01 NA NA NA

this is not necessarily true. A screen is not effective if almost all sites fail. The optimal screening values for groundwater or soil gas are those that minimize the number

of false positive signals while still avoiding most false nega-tive signals. For indoor air, the choice of target values is complicated by the relatively high background levels of some VOCs of interest in both ambient and indoor air. The more conservative the target values, the harder it may be to resolve VI from background sources.

Most states that have VI guidance follow a tiered evaluation approach similar to that given in the 2002 EPA guidance. There are, however, significant differences from state to state in the degree of conservativeness built into the VI guidance, including the screening and action levels for various compounds. There also are significant differences from state to state with regard to the specific data needed or suit-able to evaluate VI. Therefore, responsible parties have difficulty applying a consistent approach across multiple sites if they are located in multiple states. From the perspective of most responsible parties and their contractors, it would be preferable to have well thought-out, detailed national guidance rather than operate under a patchwork of various state and local guidance. Guidance cur-rently being developed by the Interstate Technology and Regulatory Council (ITRC) and practices for evaluating VI during real estate transactions cur-rently being developed by ASTM may provide bases for moving toward more consistent practices across the country. em

Twenty-Sixth Annual InternationalConference on Incineration & Thermal

Treatment Technologies (IT3)

The IT3 Conference is an annual symposium on thermaltreatment technologies for the management of special wastestreams, including radioactive, hazardous chemical, mixed,medical/infectious, explosives, and chemical munitions. Theconference will address a variety of topics with perspectivesfrom regulators, designers, operators, program managers,and research scientists. Presentations will discuss monitoringtechniques, risk assessment and management technologies,institutional management, and regulatory programs andtheir impacts.

For more information, visit www.it3.umd.edu.

May 14-18, 2007 • Phoenix, Arizona REFERENCES1. Draft Guidance for Evaluating

the Vapor Intrusion to Indoor Air Pathway from Groundwater and Soils (Subsurface Vapor Intrusion Guidance); U.S. Environmental Protection Agency; Fed. Regist. 2002, 67, (230), 71169-71172.

2. Trichloroethylene Inhalation Toxic-ity Values and Corresponding Risk Based Indoor Air Concentrations; U.S. Environmental Protection Agency Region 8 Technical Publication, January 26, 2005; available at www.envirogroup.com/TCE_Toxic-ity_VI_Tech%20Pub_1_26_05.pdf.

Copyright 2007 Air & Waste Management AssociationCopyright 2007 Air & Waste Management Association






Copyright 2007 Air & Waste Management Association