Department of Health Seal

TGM for the Implementation of the Hawai'i State Contingency Plan
Section 13.7


In a very limited number cases a traditional Human Health Risk Assessment (HHRA) will be needed to more fully evaluate and document direct exposure concerns identified in the Environmental Hazard Evaluation (HDOH, 2016). An HHRA justifies and employs alternative models and assumptions to develop site-specific screening or final cleanup levels or quantitatively evaluate actual risk posed to human receptors. This type of Advanced Environmental Hazard Evaluation typically follows methodologies, assumptions, and risk assessment models for traditional, detailed HHRAs. Portions of the Tier 1 models still may be retained for some components of the Advanced Environmental Hazard Evaluation.

A HHRA can be described as a scientific process used to estimate the probability of adverse health effects resulting from human exposure to hazardous substances. In 1986, the USEPA established risk assessment guidelines to provide consistency and technical support between the USEPA and other regulatory agencies. The HEER Office recommends that HHRAs be prepared following USEPA risk assessment guidelines. The fundamentals of USEPA’s HHRA methodology are presented in Risk Assessment Guidance for Superfund. Volume I, Human Health Evaluation Manual (Part A) (USEPA, 1989e).

A HHRA applies four evaluation components as the basis for characterizing potential health risks posed to current and/or potential future receptors at a site (USEPA, 1989e), as shown in Table 13-5.

Table 13-5 Evaluation Components for Characterizing Potential Health Risks to Current and/or Potential Future Receptors


Data Usability Evaluation/ Selection of COPCs

  • Collate the site investigation data
  • Evaluate the data
  • Select COPCs


Toxicity Assessment

  • Identify toxicity values for the COPCs


Exposure Assessment

  • Identify exposure scenario
  • Identify exposure pathways
  • Identify exposure factors for receptors
  • Quantify exposure point concentrations
  • Calculate exposure for each COPC/medium/pathway combination


Risk Characterization

  • Calculate the cancer risks and non-cancer hazard indices.
  • Summarize the site risks by COPC and medium for the receptors, exposure scenarios, and exposure pathways
  • Identify key uncertainties and evaluate their potential impacts on the results.

A summary of the basic components of a human heath risk assessment is provided below. Detailed guidance for preparing a HHRA is beyond the scope of this Section. Refer to the references provided at the end of this section for additional guidance and information.

It is vital to ensure that all potential environmental hazards have been considered at a site when conducting a site-specific human health risk assessment. Environmental Hazard Evaluations that only consider risk to human health (e.g., direct exposure to contaminted soil) will not be considered acceptable by the HEER Office.


The components of a traditional HHRA are briefly reviewed below. DATA USABILITY EVALUATION/SELECTION OF COPCS

Most of this task is completed during site investigation and presented in the site investigation report. Elements of data assessment are presented in Section 3.9. A preliminary list of COPCs is developed in the scoping process during site investigation planning (see Section 3.3). The COPCs are refined as the site investigation objectives are developed and the site investigation data are acquired (see Section 3).

To prepare a HHRA, it may be necessary to collate data from several site investigations and assess the collective data. Comparing the collated data to Tier 1 EALs is a good starting point for refining and finalizing the list of COPCs. It is important to remember that COPCs may be uniquely selected for individual decision units (see Section 3.6). In addition, COPCs may be uniquely selected for each environmental medium (e.g., COPCs for soil may be different that the COPCs for groundwater). The identified list of COPCs will be the focus of the HHRA. EXPOSURE ASSESSMENT

Exposure scenarios, exposure pathways, and exposure factors for receptors are identified, exposure point concentrations quantified, and exposures calculated for each COPC/medium/pathway combination in the Exposure Assessment. The statistical evaluation of soil data to estimate representative contaminant concentrations within decision units (including exposure point concentrations for evaluation of direct-exposure hazards) is discussed in Section 4. The average concentration of exposure over the target exposure duration is used for estimation of risk.

Land use is a critical element in developing exposure scenarios because it determines potential receptor populations. Land use assumptions are based on a factual understanding of site-specific conditions and reasonably anticipated future use. Discussions with land owners and local land use planning offices are helpful in determining future land use. Typical land uses to consider include residential and commercial/industrial.

An exposure pathway describes the course that a chemical takes from a source to a receptor. Potential exposures are evaluated by considering the following four factors:

  • A source of potentially hazardous substances
  • A contaminated media, such as soil
  • An exposure or contact point with the contaminated medium
  • An exposure route for chemical intake by a receptor

An exposure pathway is considered complete when it has all four factors. Designation of an exposure pathway as complete indicates that human exposure is possible, but does not necessarily mean that exposure will occur nor that exposure will occur at the levels estimated in the HHRA. When any one of the factors is missing in the pathway, it is considered incomplete. Incomplete exposure pathways do not pose a health hazard and are not evaluated. The key step in analyzing exposure pathways is to determine whether there are any plausible routes of human exposure to chemicals detected at the site.

The hypothetical receptor that is typically evaluated in each exposure scenario is assumed to have a reasonable maximum exposure (RME) by any potential exposure route. The RME, as defined by the USEPA (USEPA, 1989e), is the "highest exposure that is reasonably expected to occur" and is intended to best represent "a conservative exposure estimate that is within the range of possible exposures." The assumption of exposure represents a conservative approach. This approach is recommended by regulatory risk assessment guidance in order to make the health risk assessment sufficiently protective of potential receptors.

Exposure factors, including exposure point concentrations for receptors are necessary to quantify exposures for each COPC/medium/pathway combination. Exposure factors are available in Exposure Factors Handbook (USEPA, 1997e), Soil Screening Guidance: Technical Background Document (USEPA 1996b), and Supplemental Guidance for Developing Soil Screening Levels for Superfund Sites (USEPA, 2002e), among other sources. An exposure point concentration is a reasonable estimate of the concentration likely to be contacted over time by potential receptors (USEPA, 1989e). A discussion of approaches to determine average contaminant concentrations for exposure areas or decision units is presented in Sections 3 and Section 4 .

The final step in the exposure assessment is quantifying the Average Daily Intake of COPCs for the identified potential receptors in the exposure scenarios. The daily intake equations are available in Risk Assessment Guidance for Superfund. Volume I, Human Health Evaluation Manual (Part A) (USEPA, 1989e). TOXICITY ASSESSMENT

The purpose of the Toxicity Assessment is to weigh available evidence regarding the potential for the COPCs to cause adverse effects in exposed individuals and to provide, where possible, an estimate of the relationship between the extent of exposure to a COPC and the increased likelihood and/or severity of adverse effects (USEPA, 1989e). Because the USEPA has established values for the toxicity of most typically encountered COPCs, the toxicity assessment generally consists of locating and collating toxicity information and combining it with the exposure assessment information to calculate human health risks.

A cancer slope factor is a numerical estimate of potency of a chemical that is multiplied by the Lifetime Average Daily Intake to give a probability of an individual developing cancer over a lifetime. A reference dose is defined as an estimate (with uncertainty spanning perhaps an order of magnitude or greater) of a daily exposure level for the human population, including sensitive subgroups, that is likely to be without an appreciable risk of deleterious effects during a portion of a lifetime (USEPA, 1989e).

Refer to Appendix 1 of the HDOH EHE guidance for a summary of toxicity factors selected for development of environmental action levels. Except as noted, the toxicity factors for the EALs reflect those used for USEPA Regional Screening Levels guidance (USEPA, 2012). RISK CHARACTERIZATION

For complete pathways, risk characterization combines the Exposure Assessment and Toxicity Assessment to produce quantitative estimates of potential health risks associated with the COPCs.

For carcinogens, cancer risks are calculated according to the following equation (USEPA, 1989e):

Incremental Lifetime Cancer Risk = Lifetime Average Daily Intake x Cancer Slope Factor

For non-carcinogens, hazard quotients are calculated according to the following equation (USEPA, 1989e):

Hazard Quotient = Average Daily Intake / Reference Dose

Incremental lifetime cancer risk probabilities may be compared to the USEPA acceptable risk levels. The USEPA has established a potentially acceptable range of 10-4 to 10-6 for lifetime cancer risk (USEPA, 1989e, 1991b). Remediation or risk management is almost always warranted at sites where the estimated cancer risk exceeds 10-4. For sites where the estimated risk is between 10-4 and 10-6, the need for active remediation or risk management is evaluated on a site-specific basis (i.e., risks within this range are "potentially acceptable", depending on site-specific considerations) (USEPA, 1991b). It should be noted that the calculated risk values are upper-bound estimates of excess cancer risk potentially arising from lifetime exposure to the chemical in question. A number of assumptions have been made in the derivation of the values, many of which are likely to overestimate exposure and toxicity. The actual incidence of cancer is likely to be lower than these estimates and may be zero.

The non-cancer hazard index is based on a comparison of the estimated site-related dose to the EPA acceptable dose. A hazard index of less than or equal to one indicates no potential for non-cancer health hazard (USEPA, 2001c).


An exposure point concentration is a reasonable estimate of the concentration likely to be contacted over time by potential receptors (USEPA, 1989e). An exposure point concentration must be determined for a given site and used for comparison to the HDOH Tier 1 EALs. If a given site has more than one decision unit, exposure point concentrations must be developed for each decision unit. Statistical approaches are generally necessary to determine the exposure point concentrations. The method for deriving exposure point concentrations will depend upon the sampling strategy used at a given site. Approaches for sites with Multi-increment sampling and judgmental/biased sampling strategies are presented below. Soil sampling approaches are summarized in Section 4 and briefly reviewed below. MULTI-INCREMENT SAMPLING STRATEGY

For sites where a multi-increment sampling strategy is used to collect samples, replicate samples from the same decision unit allow for statistical calculation of several important quantities, including the standard deviation, the relative standard deviation (RSD), and the 95 percent (%) upper confidence level (UCL) on the mean. The RSD is initially evaluated to ensure this measure meets site data quality objectives. The mean and standard deviation are used to calculate the 95% UCL, which can be used as the exposure point concentration. The 95% UCL represents a value that we are 95% confident does not exceed the true mean of the population (e.g., in an individual decision unit).

To calculate the 95% UCL for multi-increment sample (MIS) data, three or more MIS replicates for a single DU are required see Section 4. Triplicates are typically recommended for MIS data, either for a single decision unit (DU), or one DU sampled in triplicate representing a "batch" of up to 2-10 similar DUs (e.g. DUs with similar soil type, contaminant sources and analytes, history of use, etc. such as former agricultural fields for a particular crop/area). If the decision units are heterogeneous in key physical characteristics, contaminant sources, or history of use, triplicates must be collected from each decision unit to calculate a 95% UCL for each decision unit.

When the DU replicate data meet an initial data quality measure for relative standard deviation (e.g. typically <35% RSD), the data are estimated to approximate a statistically "normal distribution", and in these cases the "Student’s- t" UCL method is recommended if the 95% UCL is calculated (see Section 4). Where one triplicate sample is collected to evaluate sample variation for multiple (or a batch of) similar DUs, the replicate DU data would be used to estimate the 95% UCL for the selected replicate DU as well as the other DUs in the batch (see Section Others have suggested that the Chebyshev UCL method, in addition to the Student’s- t UCL method, may be appropriate for analysis of MIS data in site-specific situations (ITRC, 2012).

Non-detect data are much less commonly encountered when using a multi-increment sampling approach compared to a discrete sampling approach. In addition, if any non-detects are far below relevant site Environmental Action Levels they are unlikely to affect site decision-making. Where non-detect data are encountered in MIS investigations, including calculations for contaminant totals that are comprised of any individual congeners that may be non-detect, the HEER Office recommends simple substitution of ½ the method detection limit (MDL). DISCRETE SAMPLING STRATEGY

The HEER Office strongly recommends the use of Multi-increment sampling strategies over discrete sampling strategies to establish representative soil contaminant concentrations for decision units. When using a discrete sampling strategy, sample locations should be randomly selected within the decision unit; and effort should be made to maximize the number of discrete samples collected within each decision unit (30 or more discrete samples from each decision unit would be best, or as close to this number as feasible).

A statistical derivation of the exposure point concentrations is performed with the discrete data collected, typically the 95% UCL on the mean. Use statistical methods to estimate site-specific exposure point concentrations and evaluate environmental hazards. If sample data are limited, maximum-detected concentrations of the COPCs should be compared to the HDOH Interim Final EALs to evaluate potential environmental concerns. A judgmental (or biased) sampling method is not used for determining exposure point concentrations as data cannot be inferred (calculated statistically) beyond the specific sample locations collected.

Sample data collected outside of impacted areas generally should not be included in estimation of exposure point concentrations. Instead, use decision unit strategies to divide the site and group data points prior to calculating exposure point concentrations (see Sections 3 and Section 4 for more information regarding decision units). For residential land use scenarios, sample data should not be averaged over areas greater than the size of a typical yard or backyard [e.g., 100 to 500 meter2 (CalEPA, 1996b) or 1,000 to 5,000 feet2 (HDOH, 2016)].

The 95% UCL on the arithmetic mean was traditionally used as the exposure point concentration for discrete data (USEPA, 1989e), but has under some cases been supplanted by alternative approaches that support greater accuracy. The United States Environmental Protection Agency (USEPA) has published several guidance documents that focus on calculating UCLs for discrete data to be used as exposure point concentrations (USEPA, 1992b, 2002d) and released software to aid in these calculations. The California Environmental Protection Agency (CalEPA) provides additional guidance for estimating exposure point concentrations using discrete data in its Preliminary Endangerment Assessment Guidance Manual (CalEPA, 1994a) and Supplemental Guidance for Human Health Multimedia Risk Assessments of Hazardous Waste Sites and Permitted Facilities (CalEPA, 1996b), among other sources.

It is common for environmental data sets that have not been collected with a Multi-increment sampling approach to contain non-detect data that must be addressed during calculation of exposure point concentrations. For handling data analysis for large discrete data sets, the HEER Office recommends the following approaches for assessing non-detect data (USEPA, 2007b) and (Helsel, 2005):

  • Kaplan-Meier – a non-parametric method with no distribution assumptions. This method is usable where non-detects are <50% of the data set
  • MLE – ( not Cohen’s MLE) a lognormal distribution is assumed in this method that is usable when the number of non-detect values is >50% and a lognormal distribution has been established
  • Robust regression on order statistics – this method is usable in all other cases.


Special considerations must be made to conduct a human health risk assessment for sites at which lead is a COPC. For lead risk assessments, the EPA currently recommends two models to assess exposure, depending on the age of the receptor population. For children, exposure assessments should be performed using the Integrated Exposure Uptake Biokinetic Model for Lead in Children (IEUBK) (USEPA, 1994b). For adults, EPA’s 1996 adult interim guidance should be used – the "Adult Lead Model" (USEPA, 1996e). These models are available for download free-of-charge from the USEPA web site.

Both models take into account intake and uptake components of lead exposure, allow the user to input site-specific data (exposure frequency, sources of lead, as well as others), and predict potential upper-bound blood lead concentrations. Predicted blood lead values provide one indication of the associated lead exposure for both current and potential future populations (USEPA, 2002, USEPA, 2002i).

USEPA guidance for lead-contaminated soil calls for the comparison of lead concentrations in the <250 micron soil fraction to action levels (USEPA, 2000e). The fine soil fraction is considered to be the particle size fraction most likely to stick to hands and, thus, potentially be incidentally ingested. This guidance also call for the use of the %lt;250 micron soil fraction in bioaccessibility tests (USEPA, 2000e). This also applies for bioaccessibility tests carried out on arsenic-contaminated soils. Concurrent data for the <2mm soil fraction can also be very useful in determining the distribution of lead (and arsenic) in the soil.


In terms of human health risk assessment, parks and similar public recreational areas should be treated like residential backyards and indeed they serve this function in many densely populated areas of Hawai`i (i.e., assumed residential exposure, screening level target excess cancer risk of 10-6 and target hazard index of 1.0). A purely toxicity-based, recreational-use exposure scenario could suggest that substantially higher concentrations of contaminants could be left in place at the site and not pose a threat to human health. This is because of the reduced exposure frequency and duration (e.g., 100 days per year for ten years) assumed in the models. Cleanup levels based on such scenarios can be higher (less stringent) than levels that would be allowed for commercial/industrial properties. This is counterintuitive to the intention of setting aside land for public use and puts an inherent public use restriction on the property (i.e., visitation and use limited to the assumed exposure frequency and duration).

The use of public parks should be unrestricted. Placing restrictions on the use of public parks due to contamination concerns would quite likely not be acceptable to the general public, one of the tenants required for consideration of final site remedial actions. Public parks are also frequented by children, young mothers, elderly people and other groups of people with potentially elevated sensitivities to environmental contaminants. Long-term, future uses of such properties are also difficult to predict.

In some cases, remediation of proposed parklands to unrestricted land-use standards may not be technically or economically practicable. This should be evaluated on a site-specific basis and receive approval from the overseeing regulatory agency as well as private and public stakeholders. In such cases, the appropriateness of allowing unrestricted access to the area should be carefully evaluated. This could include the need to impose access restrictions on the property (i.e., based on the exposure assumptions used in the risk assessment) and/or cap impacted soils with a minimal amount of clean fill. It may also be prudent to post signs at the property entrance that warn of potential health hazards (refer also to HDOH, 2016).


Potentially useful reference documents for preparation of HHRAs include the following:

  • Superfund Exposure Assessment Manual (USEPA, 1988c)
  • Risk Assessment Guidance for Superfund. Volume I, Human Health Evaluation Manual (Part A) (USEPA, 1989e)
  • Technical Support Document: Parameters and Equations Used in the Integrated Exposure Uptake Biokinetic Model for Lead in Children (USEPA, 1994b)
  • Preliminary Endangerment Assessment Guidance Manual (CalEPA, 1994)
  • Standard Provisional Guide for Risk-Based Corrective Action (ASTM, 2004b)
  • Superfund Soil Screening Guidance: Technical Background Document (USEPA, 1996b)
  • Supplemental Guidance For Human Health Multimedia Risk Assessments of Hazardous Waste Sites and Permitted Facilities (CalEPA, 1996b)
  • Exposure Factors Handbook (USEPA, 1997e)
  • Health Effects Summary Tables (USEPA, 1997)
  • Assessing the Significance of Subsurface Contaminant Vapor Migration to Enclosed Spaces (Johnson et. al, 1998)
  • Supplemental Guidance for Developing Soil Screening Levels for Superfund Sites (USEPA, 2002e)
  • USEPA Regional Screening Levels: (USEPA, 2012).

The 2012 USEPA Regional Screening Levels (RSLs) replace Preliminary Remediation Goals (PRGs) previously published by individual USEPA regions (e.g., USEPA, 2012). The USEPA RSLs have been incorporated into HDOH Environmental Action Levels for screening of potential direct-exposure hazards (HDOH, 2016).

The above list of references is not intended to be comprehensive. Additional HHRA guidance should be referred to as needed.