USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5106

Back to Table of Contents

Revisions to the HBSL Methodology

HBSLs for unregulated contaminants are calculated using standard USEPA OW equations for establishing drinking-water guideline values (Lifetime Health Advisory (Lifetime HA) and Cancer Risk Concentration values) for the protection of human health; this general approach has not changed since the original HBSL methodology was published (Toccalino and others, 2003).

The revisions to the HBSL methodology presented in this report reflect updates to USEPA cancer classifications, relationships between the equations used to calculate HBSLs and USEPA cancer classifications, changes to the hierarchy of toxicity information sources used to calculate HBSLs, and a departure from defaulting to Lifetime HA values as HBSLs for contaminants with certain cancer classifications. Additionally, this report outlines conditions under which exceptions to the HBSL methodology are made.

Updates to USEPA Cancer Classifications

Cancer classifications from the 1986, 1996, 1999, and 2005 USEPA guidelines for cancer risk assessment are considered in the development of HBSLs (U.S. Environmental Protection Agency, 1986; U.S. Environmental Protection Agency, 1996, 1999, and 2005). The finalization of the USEPA cancer guidelines in 2005 (U.S. Environmental Protection Agency, 2005) does not change the HBSL methodology, but the HBSL methodology has been updated in this report to include the final USEPA cancer classifications.

Early USEPA cancer classifications used alpha-numeric cancer group designations that reflect a qualitative Weight-of-Evidence (WOE) judgment as to the likelihood that a contaminant may be a carcinogen for humans (U.S. Environmental Protection Agency, 1986). Subsequent USEPA draft guidelines (1996 and 1999) and the final 2005 USEPA guidelines for carcinogen risk assessment use descriptive terms for carcinogenicity that replace the earlier alpha-numeric cancer group designations (U.S. Environmental Protection Agency, 2005). The descriptive WOE judgments reflect the likelihood that a contaminant is a human carcinogen and the conditions under which the carcinogenic effects may be expressed. Some contaminants have not yet been evaluated under the final 2005 guidelines, so there are examples of contaminants with cancer classifications explained in the 1986, 1996, 1999, or 2005 guidelines.

Equations Used to Calculate HBSLs

Three USEPA OW equations are used to calculate HBSLs for unregulated contaminants as determined by the USEPA cancer classification for each chemical. These three equations have not changed since the publication of the original HBSL methodology (Toccalino and others, 2003), although the definitions for some terms used in the equations have been clarified. The USEPA’s assumptions for establishing drinking-water guidelines—lifetime ingestion of 2 liters (L) of water per day by a 70-kilogram adult (sidebar 1)—are adopted in each equation.

For carcinogens, the OW equation for calculating Cancer Risk Concentration values is used to calculate an HBSL range. The HBSL range represents a contaminant concentration range in drinking water corresponding to an excess estimated lifetime cancer risk of 1 chance in 1 million (10^-6) to 1 chance in ten thousand (10^-4) (eq. 1). HBSL concentration ranges for carcinogens were developed to be consistent with USEPA procedures and to acknowledge the uncertainty of the estimates. For carcinogens in drinking water, the USEPA considers risk levels of 10^-6 (and for some compounds, risk levels as high as 10^-4) to be protective of human health, provided these levels also are protective of noncancer adverse effects (U.S. Environmental Protection Agency, 1988). The USEPA accepts cancer risk policies from states in the range of 10^-6 to 10^-4 (U.S. Environmental Protection Agency, 1992 and 1995). Cancer Risk Concentration values were called Risk Specific Dose values in earlier HBSL documents (Toccalino and others, 2003; Toccalino and others, 2004; Toccalino and others, 2005).

     (1)

Where

For possible (Group C) carcinogens or contaminants with suggestive evidence of carcinogenic potential, HBSLs are calculated using the OW equation for calculating Lifetime HA values for Group C carcinogens (eq. 2). Equation 2 is the same as that presented in the original HBSL methodology report (Toccalino and others, 2003) except to clarify that the Relative Source Contribution (RSC) and the Risk Management Factor (RMF) values may differ from their default values. For noncarcinogens, HBSLs are calculated using the OW equation for calculating Lifetime HA values (eq. 3).

     (2)

     (3)

Where

The relationships between USEPA WOE cancer classifications and the corresponding OW equations used to calculate HBSLs are shown in Tables 1 and 2.

Because HBSLs are calculated using USEPA cancer classifications, USEPA toxicity data, and standard OW equations for establishing drinking-water guideline values (eqs. 1-3), HBSLs are equivalent to existing USEPA Cancer Risk Concentration and Lifetime HA values (when they exist), except for compounds for which more recent toxicity information has become available. HBSLs are rounded to one significant figure, which is consistent with USEPA OW policy.

The OW equations used to calculate HBSLs (eqs. 1-3) are comparable to those used by a variety of state agencies to develop state drinking-water guidelines, although states may use different default exposure assumptions, sources of toxicity data, or modifications to OW’s equations. For example, the NJDEP calculates ground-water-quality criteria (GWQC) for ground waters that have been designated for potable water use (New Jersey Department of Environmental Protection, 2004). The equations used to calculate GWQC values are the same as those used to calculate HBSLs, although for some Group C carcinogens, the NJDEP uses the equation for calculating Cancer Risk Concentration values (eq. 1) instead of the equation for Lifetime HA values for Group C carcinogens (eq. 2) (New Jersey Department of Environmental Protection, 2004). In California, non-regulatory Public Health Goals (PHGs) are developed by the Office of Environmental Health Hazard Assessment (OEHHA) to help maintain the quality of California’s drinking-water supplies (California Environmental Protection Agency, 2003). PHGs are calculated using similar equations and default exposure assumptions as those used to calculate HBSLs, although for noncarcinogens, OEHHA may use different toxicity values (for example, no-observed-adverse-effect-level (NOAEL) values and associated uncertainty factors instead of USEPA reference dose (RfD) values) (California Environmental Protection Agency, 2003). Additionally, PHGs for VOCs account for inhalation and dermal exposure through showering and other household uses of tap water, which decreases PHG values compared to considering ingestion exposure only. For carcinogens, GWQC and PHG values typically are established at the 10^-6 cancer risk level, whereas HBSL values are calculated as concentration ranges that correspond to a cancer risk range of 10^-6 to 10^-4 (eq. 1); both approaches are consistent with USEPA procedures.

Changes to Hierarchy of Toxicity Information Sources

In the original HBSL methodology (Toccalino and others, 2003), a hierarchy of USEPA sources of cancer classifications and toxicity data (cancer slope factors and reference doses, sidebar 2) was used in the development of HBSLs. Cancer classifications and toxicity data from the USEPA Integrated Risk Information System (IRIS) database had the highest priority, followed by the most recent information from the USEPA OW and the Office of Pesticide Programs (OPP). Cancer classifications and toxicity data from the USEPA Health Effects Assessment Summary Tables (HEAST) were assigned a lower priority and were used only in the absence of information from IRIS, OW, and OPP.

The USEPA and USGS participants in the August-September 2005 discussions agreed that the hierarchy of toxicity information used in HBSL development should change because the availability of toxicity information published by the USEPA has changed since the original HBSL methodology was published. Specifically, as part of the pesticide registration review program, the OPP periodically reevaluates toxicity information for pesticides. Updates to toxicity information for pesticides historically were published in the IRIS database and in OPP documents, but are now published in separate OPP risk assessment documents and reregistration eligibility decision documents (and typically not in the IRIS database) (A. Mills, IRIS Program Director, written commun., December 27, 2005). The discontinuation of pesticide toxicity information updates to the IRIS database affects the hierarchy of sources of toxicity information used in the development of HBSLs.

The hierarchy of toxicity information used in HBSL development also was changed to represent the current “state of the science” in the United States. The USEPA and USGS participants in the August-September 2005 discussions determined that “acceptable” toxicity information for developing HBSLs should meet four criteria: (1) United States information (limited to USEPA data); (2) internally (e.g., USEPA) or externally peer reviewed; (3) publicly available (so anyone has access to the information); and (4) most recently available.

The hierarchy of sources of cancer classifications and toxicity data to use in the development of HBSLs was changed to the most recent of five USEPA sources (table 3). Use of this hierarchy has the advantage of providing a mechanism for the timely incorporation of updated toxicity information in the interpretation of water-quality data. HEAST values are no longer used to derive HBSLs because these values are older, not necessarily peer reviewed, and are not readily available to the public.

Departure from Defaulting to Lifetime Health Advisories

In the original HBSL methodology, HBSLs for possible (Group C) carcinogens, contaminants with suggestive evidence of carcinogenic potential, and noncarcinogens defaulted to USEPA Lifetime HA values, when available (Toccalino and others, 2003). The revised HBSL methodology will no longer default to using Lifetime HA values. Instead, HBSLs are now derived using Equation 2 for possible carcinogens and contaminants with suggestive evidence of carcinogenicity, Equation 3 for noncarcinogens, and the most recent toxicity information as noted in Table 3. This revision to the HBSL methodology ensures that HBSLs are consistently based on the most recently available toxicity information.

Exceptions to the HBSL Methodology

When an HBSL differs from an existing USEPA Lifetime HA value and the reason for the difference is not apparent (for example, the same OW toxicity value and equation are used for the HBSL and Lifetime HA), the USGS will consult with the USEPA OW to identify the reason for the discrepancy. the USEPA OW may have technical or policy reasons for modifying the equations for calculating Lifetime HA values for some compounds (J. Donohue, OW, Office of Science and Technology, written commun., January 20, 2006) such as:

• Noncarcinogens for which the RSC is different from the USEPA default value of 20 percent.
• Possible (Group C) carcinogens or contaminants with suggestive evidence of carcinogenic potential for which the RMF is different from the USEPA default value of 10. USEPA policy allows for RMF values to range from 1 to 10.
• Compounds for which different rounding procedures were used. OW has calculated Lifetime HA values using RfD values that were rounded to one significant figure for some compounds and not rounded to one significant figure for other compounds. HBSLs are calculated using the most recent, USEPA peer-reviewed, publicly available toxicity values, as published, regardless of the number of significant figures.

For those compounds with differing HBSL and Lifetime HA values, the HBSL will default to the Lifetime HA value when:

1. The HBSL is based on OW toxicity data (that is, OW has the most recently published toxicity information).
2. OW has established an RSC or RMF for a compound that differs from the default values of 20% and 10, respectively, regardless of whether the HBSL is derived using OW, OPP, or IRIS toxicity data. OW is the only USEPA office with policies on the development and use of RSC and RMF values.

The rationale for any exceptions to the HBSL methodology will be captured on the HBSL website (Toccalino and others, 2006a).

Back to Table of Contents