AERA guide | Section 4: Risk characterization

In risk characterization, information from each of the risk analysis elements described above (source characterization, pollutant identification, emission estimation, exposure and toxicity assessment) are summarized and integrated into quantitative and qualitative expressions of risk.

USEPA’s Science Policy Council Handbook, Risk Characterization, states that “The goal of risk characterization is to clearly communicate the key findings and their strengths and limitations so its use in decision making can be put into context with the other information critical to evaluating options.” It also suggests that important traits of a high-quality risk characterization are transparency, clarity, consistency and reasonableness.

USEPA’s definition of transparency is “The characterization needs to fully and explicitly disclose the risk assessment methods, default assumptions, logic, rationale, extrapolations, uncertainties, and overall strength of each step in the assessment.”

Such transparency and other traits of a high-quality risk characterization will be achieved by following the AERA guidance. The forms and guidance were designed so that AERA submittals are clear; they are consistent with federal and state requirements and guidelines; and AERA risk results are reasonable. These qualities are obtained through use of the forms, as they provide information on the accuracy of the assumptions and data, the level of refinement and the relevance to current and future land use.

By following the steps described in previous sections of this guidance document, the magnitude of a hypothetical individual’s exposure and pollutant intake by inhaling air toxics, incidentally ingesting pollutants in the soil and ingesting pollutants in foods will have been estimated. The next step in estimating risks (both cancer and non-cancer) is to compare that individual’s exposure and intake levels with benchmark toxicity values for those pollutants.

This comparison results in probabilities that an individual will develop cancer over a lifetime of exposure Major assumptions, scientific judgments, and to the extent possible, estimates of the uncertainties are also presented in the risk characterization.

AERA quantitative risk estimation

In the quantitative risk estimation portion of an AERA, cancer risk and hazard descriptors (“risks”) are intended to convey information about the potential risks to hypothetical individuals potentially impacted by emissions from a facility or project. Quantitative risk estimates from an AERA may include:

  • multi-pathway and inhalation cancer risks from facility emissions
  • multi-pathway and inhalation non-cancer hazard quotients and indices from facility emissions for both chronic and acute exposure durations
  • estimates of blood levels in children associated with exposure to lead from a facility
  • non-cancer hazard quotient from ingestion of mercury in fish tissue
  • where appropriate, potential cumulative risks from nearby sources

The AERA methods for calculating risks differ depending on the tool used, whether an air toxic is a carcinogen, the nature and number of exposure pathways being assessed, and whether cumulative risks are being assessed.

Regardless of which tool is used, the basic equations for calculating risks from individual pollutants are the same and are calculated as follows:

Table 10: Equations for calculating risks from individual pollutants

Risk equation description
 
Risk equation
 
Definition of terms
 
Carcinogens (inhalation)
 
Inhalation cancer risk=IEC x UR
 
IEC = inhalation exposure concentration in air (µg/m3)

UR = Inhalation unit risk (µg/m3)-1
 
Carcinogens (ingestion)
 
Ingestion cancer risk=Intake x SF
 
Intake = daily intake of pollutant

SF = Cancer slope factor (mg/kg-day)-1
 
Total cancer risk
 
Total Cancer Risk = ∑ Ingestion Cancer Risks + ∑ Inhalation Cancer Risks
 
Non-carcinogens (inhalation)
 
Inhalation Hazard Quotient=IEC/RfC
 
IEC = exposure concentration in air (µg/m3)

RfC = reference concentration  (µg/m3)
 
Non-carcinogens (ingestion)
 
Ingestion Hazard Quotient=Intake/RfD
 
Intake = daily intake of pollutant

RfD = Reference dose (mg/kg-day)
 
Total non-cancer
 
Hazard Index = ∑ Ingestion Hazard Quotients + ∑ Inhalation Hazard Quotients
 

 

Total risks estimated in an AERA are calculated by adding the individual risks for each air toxic in each pathway of concern (i.e., inhalation, ingestion), then summing the risk for each receptor-type evaluated (e.g., resident, farmer) for all pathways.

Each AERA tool estimates total risks in a somewhat different way. Tool-specific equations and methods are described below.

RASS and Q/CHI 

The RASS and the Q/CHI both provide a screening level assessment of inhalation and indirect exposure pathways by using Multi-Pathway Screening Factors. These risk screening methods correspond with AERA  screening levels 1 and 2 (MPSFs). The basic calculations performed in the RASS and Q/CHI spreadsheets are shown below.

RASS calculations

Inhalation risks are estimated in the RASS by first comparing each modeled air toxic concentration with its respective inhalation health benchmark (IHB) concentration. Acute inhalation risks are a comparison of maximum modeled hourly air concentrations with acute IHBs. Chronic inhalation risks (both cancer and non-cancer) are comparisons of maximum modeled annual air concentrations with chronic IHB concentrations. For non-carcinogens, this is the reference concentration (RfC). For carcinogens, the IHB concentration is the concentration in air that could result in an excess lifetime cancer risk of 1 in 100,000 (1 x 10-5). Pollutant-specific risks are then summed to obtain total inhalation risks.

AERA total facility inhalation equation

Ingestion risks are estimated in the RASS by multiplying pollutant-specific inhalation risks by MPSF values. Cancer risks and hazard quotients are summed across all pollutants for the inhalation and ingestion exposure routes. This is done for the farmer, resident and urban gardener. The fish consumption pathway is not evaluated in the RASS.

AERA total facility ingestion risk

Total risks from each exposure route (inhalation and ingestion) are summed and displayed in the RiskCalcs tab of the RASS. A summary of the total risks for each exposure route and exposure scenarios is displayed in the Summary tab of the RASS.

AERA otal facility risk r

Q/CHI calculations

The Q/CHI method is useful when there are multiple sources and more refined spatial and temporal modeling is desired. The Q/CHI spreadsheet is a “RASS-like” spreadsheet that calculates emission rate/chemical health index ratios (Q/CHI). These pollutant-specific ratios are then summed through pollutants, but they remain specific to emission units and exposure scenarios. For ingestion-based exposure scenarios the Q/CHI value is multiplied by a Multi-Pathway Screening Factor prior to Q/CHI value summation. The Q/CHI sum is then entered into AERMOD in the place of pollutant-specific or unitized emission rates. AERMOD modeling using this method results in risk estimates at each modeling receptor. The risk estimates are paired in time and space and may be mapped using a geographic information system. More detailed instructions for the use of the Q/CHI spreadsheet are included in the ReadMe tab within the spreadsheet itself.

If the only exposure pathway of interest is inhalation, total facility risks are obtained by summing pollutant-specific Q/CHI values to obtain total inhalation Q/CHI values, which then may serve as input to AERMOD to obtain total facility inhalation risks. For most projects the ingestion pathway is also of interest, so pollutant-specific ingestion-based Q/CHI values are obtained by multiplying inhalation Q/CHI values by respective MPSFs. Inhalation and ingestion Q/CHI values are summed in the spreadsheet to provide total inhalation and ingestion input to AERMOD. The outcome is total modeled inhalation plus ingestion risks.

Once each risk estimate is calculated by AERMOD these risk estimates are input into the Q/CHI spreadsheet under the tab for the appropriate exposure scenario (e.g. Acute Q_CHI, for the maximum acute inhalation hazard index). This allows the spreadsheet to calculate pollutant-specific hazard quotients and cancer risks.

Note: The pollutant-specific hazard quotients and cancer risks will not sum to the same values as the AERMOD produced risk estimates. This is due to the fact that the pollutant-specific risks calculated in the “RiskDrv” tab are maxima from each location, rather than being paired in time and space. To achieve a more accurate estimation of pollutant-specific risk estimates, one may multiply the percent contribution of that pollutant by the total risk estimate (acute inhalation HI * 97% contribution from nitrogen dioxide as calculated from the RiskDrvr tab). The percent contribution is calculated from the hazard quotients or cancer risks calculated on the “RiskDrv” tab.

Multi-pathway screening factors (MPSFs)

Multi-Pathway Screening Factors (MPSFs) are embedded in the RASS and Q/Chi spreadsheet analysis to estimate ingestion risks from eating vegetables and non-fish foods.

Multi-pathway Screening Factors are defined as ratios of the maximum risk from the indirect (ingestion) exposure routes to the maximum risk from the direct (inhalation) exposure route as shown below.

The MPSFs were developed from ingestion to inhalation ratios modeled using the Minnesota MNRiskS cumulative air pollution tool. This tool is described in Pratt et al. 2012. The model inputs used to calculate the ratios were from point source emissions, and the 98th percentile summary was used once all ratios statewide were calculated. Some near-site ratios were eliminated if point estimations were close enough such that deposition was being estimated by dispersion was still aloft. MPSFs were developed for three scenarios, the adult farmer, the adult resident, and the urban gardener. 

Only those pollutants with a ratio of one (rounded values) or higher were assigned MPSFs in the RASS. For many volatile compounds, the inhalation risks are higher than ingestion risks, so the calculated MPSFs are less than one. Pollutants that accumulate in the food chain (PBTs) have higher ingestion risks, and therefore have MPSFs greater than one. MPSFs were rounded to whole numbers to better reflect the appropriate level of certainty.

The general development of the MPSFs prior to the MNRiskS update is described in detail in Pratt and Dymond, 2009.

HHRAP-based analyses (refined Multi-pathway AERA, Tier 3)

The USEPA developed an approach for conducting multi-pathway, site-specific human health risk assessments on hazardous waste combustors, Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities (HHRAP). The guidance was developed to describe the evaluation of inhalation risks and to provide procedures for estimating risks from indirect exposure pathways. Equations for estimating potential cancer risks and non-cancer hazards are provided in HHRAP Appendix C.

The HHRAP procedures provide default assumptions that typically reflect national averages. MPCA staff reviewed the default values provided in the HHRAP guidance and have performed some sensitivity analyses. MPCA-recommended values are discussed below and provided in the Microsoft Office document icon Refined HHRAP-Based Analysis Form (AERA-26 Form). Many of the alternative values in the AERA-26 form are Minnesota-specific (e.g., average annual precipitation) and are intended to reflect Minnesota conditions more accurately while producing a health-protective analysis. The AERA-26 form provides additional recommendations reflecting the most recent, scientifically defensible data and approaches (e.g., using the most current regulatory air dispersion model and the most current toxicity values).

In addition to the exposure scenarios evaluated using the RASS and Q/CHI tools, HHRAP-based tools are equipped to evaluate the fish consumption exposure pathway.

Toxicity values in HHRAP Analyses

Inhalation toxicity values from the RASS need to be used in HHRAP-based tools (e.g., IRAP). A hierarchy similar to that used to compile inhalation values was been followed by MPCA staff to tabulate ingestion toxicity values for use in HHRAP-based analyses. These values are available upon request.

Acute analyses are performed using the RASS. If a rare circumstance arises where it may be deemed appropriate, after discussion with MPCA risk assessment staff, to perform the acute analysis using a HHRAP-based tool, the acute toxicity values from the RASS needs to be entered in the tool. Most HHRAP-based software acute values are emergency levels issued by DOE as part of their Temporary Emergency Exposure Limits or are from the USEPA Acute Inhalation Exposure Guideline Levels - Level 1 (AEGL 1s) Database. These sources are not part of the AERA hierarchy of toxicity information sources.

Mercury analysis using MMREM

In general, facilities that emit more than 1 pound of mercury a year (actual facility emissions, potential controlled emissions may differ) and/or are located near water body(s) may be asked to estimate potential human health risks from mercury. Fish consumption is the primary pathway of concern for mercury, and needs to be assessed using the “MPCA Mercury Risk Estimating Method” (MMREM). This approach relies on measured mercury concentrations in Minnesota fish and measured ambient background mercury deposition. The Office spreadsheet icon MMREM spreadsheetis used to estimate an incremental hazard index from eating the methyl mercury in fish from the water body(s) assessed. Inputs to the spreadsheet include: measured mercury concentrations in Minnesota fish tissue from the water body(s) being assessed (or from representative water bodies), the area of the water body, the area of the terrestrial portion of the watershed, averaged modeled concentrations of speciated mercury over the water body, and the modeled average concentration over the terrestrial watershed area. A detailed description of this process is provided in a PDF icon MMREM guidance document. Additional details and the scoping process for performing mercury-based analyses are described in the Microsoft Office document icon MMREM protocol form.

The RASS may be used to estimate inhalation risks from mercury exposure.

Lead analysis using the IEUBK model

A non-cancer health benchmark is not currently available in the MPCA/MDH toxicity value hierarchy because there is no known threshold for potential non-cancer health effects associated with lead exposure. There is, therefore, no direct methodology for calculating non-cancer risk estimates for air lead emissions. The Center for Disease Control (CDC) and MDH, however, provide consistent recommendations for a blood-lead reference level for triggering medical and prevention actions. This creates concern for lead and a desire to screen lead for potential non-cancer health effects, but requires consideration beyond a calculation of an inhalation hazard quotient from lead emissions.

In effort to screen for potential non-cancer impacts from lead, the NAAQS ambient air quality standard was included in the RASS as a surrogate for an RfC. If a screening RASS results in a risk estimate for lead above 10% of the ambient air quality standard, or if there is a modeled lead NAAQS exceedance, then the MPCA recommends that blood lead concentrations are estimated and compared to recommended reference levels for triggering medical and prevention actions.

The USEPA has developed the Integrated Exposure Uptake Biokinetic (IEUBK) model to estimate blood-lead levels in children associated with multiple lead exposure pathways (air, soil, dust, diet, drinking water, and maternal lead). The IEUBK model integrates several assumptions about the complex exposure patterns and physiological handling of lead by the body, and predicts blood lead levels and distributions for children 0 to 7 years of age. The IEUBK exposure module includes default media concentrations and media intake rates, including ingestion rates for air, drinking water, soil/dust, diet, and other sources. The default intake parameters selected for use in the IEUBK model are from the central observations of the ranges of values.

The model defaults may be used for all parameters except air and soil lead concentrations. The highest modeled project-related annual average lead air concentration and the highest lead soil concentration from the HHRAP-based modeling exercise needs to be used. Indoor dust concentrations can be assumed to be the same as outdoor. Modeled blood-lead levels can then be compared with the MDH and CDC recommended reference level.

In special circumstances, the Adult Lead Methodology model may be requested.

Persistent, bioaccumulative and toxic pollutants (PBTs)

Organic pollutants that might be considered PBTs were identified using the USEPA PBT profiler model. Potential inorganic PBTs were identified using a more comprehensive list adopted by the European Union. The PBT pollutants identified by the MPCA using these resources with IHB values are indicated on the RiskCalcs page of the RASS.

PBT pollutants without multi-pathway screening factors

Some pollutants may be considered PBTs and have toxicity information available with which to assess the ingestion pathways. Some of these PBT pollutants do not have inhalation toxicity information and a MPSF cannot therefore be calculated. A list of air toxics considered to be PBTs is provided in the RASS in the “RiskCalcs” tabs. These pollutants need to be discussed in the qualitative section of the AERA in Microsoft Office document icon AERA form-02.

Quantitative risk estimation of special mixtures

Petroleum hydrocarbons—Aliphatic (C7 – C11)

Air toxics emissions of mixtures that include primarily aliphatic hydrocarbons, in the C7 – C11 range, and with less than 1.5 % aromatics, need to be entered in the RASS as “petroleum hydrocarbons, aliphatic (C7 – C11)”. Consistent with the general approach described above for the treatment of mixtures, the mass of all aromatics with available IHBs (e.g., benzene) needs to be subtracted from the mixture to be assessed separately. If a petroleum hydrocarbon mixture contains a higher fraction of aromatics than 1.5% and subtracting the mass of aromatics with IHBs reduces the fraction to less than 1.5%, the remaining mixture needs to be assessed using the recommended inhalation health benchmark derived from aromatized petroleum stream inhalation studies Information documenting the composition of the mixtures assessed with this IHB needs to be provided by the project proposer.

Dioxins/Furans

Specific dioxin and furan congener emissions need to be entered into the RASS if the information is available. If specific congener emissions are not available, but total emissions of the congener group is available, the total congener group emissions are entered into that row in the RASS Emissions tab. For example, if emissions are not available for each specific penta-chlorinated dioxins (PeCDD), but the total of this congener group is available, the group emissions are entered into the RASS Emissions tab on the row for “pentachlorodibenzodioxins, all isomers”. If a project proposer converts individual dioxin/furan congener emission rates to 2,3,7,8-TCDD equivalents, these emissions can be summed and entered into the RASS Emissions page on the line for 2,3,7,8-TCDD equivalents (Air Pollutant Identification Number 00-09-1). If refined multi-pathway modeling (e.g. HHRAP-based software) is being conducted, the preferred method is to enter specific dioxin and furan congener emissions rather than 2,3,7,8 TCDD equivalent emissions.

Air pollutant identification numbers

If specific air toxics emissions information is available (and the Chemical Abstract Service [CAS] number of the emitted air toxic matches a CAS number on the spreadsheet), enter it directly. In some cases, there is no available CAS number. MPCA air pollutant identification numbers were developed to facilitate the tracking and analysis of air toxics in the RASS.

How to report quantitative risk estimates

Risk results are estimated differently depending on the tool used. Risk results are displayed in various ways in the RASS and Q/CHI spreadsheets, both for the total facility and by pollutant. Mercury fish consumption hazard quotients are presented in the MMREM spreadsheet.

The following HHRAP results need to be submitted in a spreadsheet:

  • Indirect, and inhalation cancer and non-cancer risk results by exposure scenario for all receptors
  • Indirect, inhalation cancer and non-cancer risk results by exposure scenario for the risk driver pollutants (only for the maximum receptors).

Rounding and significant figures

Due to the uncertainties and variability of data included in a final AERA risk estimate, it is important to discuss rounding and significant figures. The MPCA intends for AERA risk estimates to:

  • Reflect uncertainty and variability
  • Contain transparent calculations
  • Be protective of human health

The MPCA realizes that no general guidance about significant figures will completely fulfill all three of the objectives mentioned above because of uncertainty in emission estimates, toxicity information, and modeling.

Standard rules for rounding apply which will commonly lead to an answer of one significant figure in both risk and hazard estimates. Hazard quotients, hazard indices, and cancer risk estimates are usually reported as one significant figure for presentation and summary purposes. The MPCA recommends rounding only the final reported results, not the intermediate calculations.

In some circumstances there may be enough information to report single pollutant risk estimates to more than 1 significant figure. This may be the case when there is a toxicity value with low uncertainty, facility specific toxicity value, and nearby meteorological data. However, this is case specific and the MPCA will need to review the work.

Tables showing risk calculations may also require more than 1 significant figure in order to represent the calculations transparently. More than 1 significant figure needs to be used in these circumstances.

Displaying risk contours using maps

If conducting a refined risk assessment using the Emission Rate/Chemical Health Index (Q/CHI) process or a receptor grid-based HHRAP type multi-pathway analysis, a project proposer can use AERMOD to generate risk isopleth maps. Locations of all receptors need to be noted on the map. Submit only isopleth maps for risk results above 0.1 (0.1 in 100,000 for cancer estimates).

Total facility

The use of the RASS in any form results in the summation of maximum modeled risks from all facility sources regardless of where they are located or the time in which they were modeled. This typically results in overly conservative risk estimates. For example, a maximum concentration for pollutant 1 may be modeled in year 2 of the meteorological data, and a maximum concentration for pollutant 2 may be modeled in year 3 of the meteorological data. Furthermore, the maximum modeled concentrations for pollutant 1 and pollutant 2 may not occur in the same location. Since the RASS sums the maximum modeled concentrations from all pollutants from all sources, it is therefore usually the case that the RASS results in facility-specific screening risk estimates that exceed facility risk guidelines. Further investigation is then required using more facility specific information, more refined dispersion modeling, or closer investigation of possible exposure pathways, as discussed in previous sections of this guidance.

Infrequently, however, screening results for the entire facility may be below human health benchmarks or facility risk guidelines. In these cases there is no further need for refining the AERA with more facility-specific information. The RASS used for this screening exercise must be submitted to the MPCA for review and approval.

Ethanol facilities

Some proposed ethanol facilities meeting certain criteria will not be asked to complete an AERA. A project proposer is asked to fill out Form Microsoft Office document icon AERA-13 Determination Checklist for Proposed Ethanol Facilities to determine if the facility meets the criteria. This form will be reviewed by MPCA staff. Further information is included in a more detailed guidance document entitled PDF icon Planning and Constructing an Ethanol Plant in Minnesota: A Guidance Document.

Screening out pollutants and sources using the RASS

Screening analyses may be used as part of the scoping process to eliminate sources, pollutants or exposure pathways that are unlikely to result in human health risks or the risk are negligible. The RASS can be used with default, conservative assumptions to provide initial AERA screening results in order to complete this type of screening exercise.

Risk-driver levels

The AERA risk management criteria for relatively low total facility risks (risk-driver levels) are:

  • pollutant-specific hazard quotients of less than 0.1 (without rounding and including the sum of hazard quotients for that pollutant across all exposure routes)
  • pollutant-specific cancer risks less than 10-6 (without rounding and including the sum of risks for that pollutant across all exposure routes)
  • cancer risks less than 10-6 and hazard indices of less than 0.1 specific to a single emission unit.

Elimination of pollutants from further evaluation does not imply the complete elimination of risk. It means the contribution from these air toxics to the total facility risk is insignificant relative to the potential risks posed by the pollutants retained for further analysis. The emission source and pollutant elimination process will work most effectively by using default dispersion factors first, so that pollutants and sources can be identified appropriately as insignificant contributors to a facility’s air emission risk. Once the scope of the AERA has been narrowed using the lookup table dispersion factors, the user may choose to run AERMOD to generate more site-specific dispersion factors. If the RASS is used as a screening or scoping tool, documentation must be provided to demonstrate the rationale for excluding pollutants or emissions sources from further analysis.

Qualitative information is important to decision makers and interested community members because it provides context for the quantitative risk estimates. The contextual information is used in all of the AERA steps but is particularly important in developing a protocol; and in understanding and communicating the key AERA findings. The qualitative information used in an AERA is documented by the project proposer in the AERA-02 form.

Qualitative information resources

The following resources can be useful in finding contextual information  to support  decision-makers in making fully informed decisions.

Table 11: Quick reference table for qualitative information

AERA 02 form information
 
What to include
 
Resources
 
Receptors and sensitive populations
 
Schools, daycares, recreation centers/playgrounds, nursing homes, hospitals, and residence locations
 
Aerial photos from sites referenced above or local records, databases.
 
General neighborhood information
 
Nearest residents if not addressed under Receptors and Sensitive Populations.
 
U.S. Census Bureau:  http://www.census.gov/

Minnesota Census Quick Facts: http://quickfacts.census.gov/qfd/maps/minnesota_map.html

and http://www.census.gov/census2000/states/mn.html
 
Nearby facilities
 
Map and/or list of permitted facilities with air emissions;  not limited to facilities with air permits
 
Minnesota Environmental Data Access:

What’s In My Neighborhood?: http://www.pca.state.mn.us/backyard/neighborhood.html
 
Zoning
 
Description of zoning within a 10 km radius where available
 
Zoning maps are searchable on the internet for most counties in Minnesota – use your preferred search engine to find “MN zoning maps”
 
Land use
 
Provide map showing land use within a 10 km radius including farming, forests, residential and industrial areas. It is recommended to verify information with a site visit.
 
Minnesota County Land Use Maps: http://www.mnplan.state.mn.us/maps/LandUse/

Minnesota Land Use and Cover: http://www.mngeo.state.mn.us/landuse/
 
Risk receptor information and isopleths
 
Maps can be generated using AERMOD when using the Q/CHI methodology. Maps can be produced for each exposure time and scenario, e.g., acute inhalation, by overlaying the risk isopleths with an aerial photograph of the area.
 
AERMOD software

Aerial photographs obtained from either the Agency or other GIS-based source.
 
Fishable water bodies
 
Provide map with labels of fishable water bodies. Information on accessibility to water body should be provided when available.
 
Lake Finder: http://www.dnr.state.mn.us/lakefind/index.html
 
Farming locations
 
Provide map showing farming locations surrounding facility. Additional information regarding crop types, animals raised, number of animals, farm size, and other qualitative information about the farm may be provided.
 

Minnesota County Land Use Maps: http://www.mnplan.state.mn.us/maps/LandUse/
 

 

 

 

Environmental Justice Policy: Implementation in the AERA

It is the policy of the MPCA to incorporate the concepts of environmental justice into agency work, including AERAs.   The MPCA policy incorporates the USEPA definition:

“The U.S. EPA defines Environmental Justice as the fair treatment and meaningful involvement of all people regardless of race, color, national origin, or income with respect to the development, implementation, and enforcement of environmental laws, regulations and policies.”

The AERA provides a tool by which quantitative risk results may be considered within the context of potential disproportionate impacts on low income and otherwise historically disadvantaged communities. Once an AERA is received, MPCA staff will collect the information listed below. This information will be incorporated into the AERA summary completed by MPCA staff and presented to Risk Managers.

  • Results from the USEPA EJScreen tool
  • The median household income by census block group in the area surrounding a facility fenceline in comparison to the county average household median income
  • The percentage of the population that is non-white, by census block group surrounding a facility fenceline in comparison to the county average non-white population percentage.

The comparison area may be modified depending on the facility location. In some circumstances a comparison to city-wide averages may be more appropriate than county wide averages.

Uncertainty and variability in AERAs

The risk assessment process is subject to uncertainty and variability from a variety of sources. These are inherent in human health risk assessment and are not unique to the MPCA AERA. The USEPA defines these terms as follows:

Uncertainty

Uncertainty refers to our inability to know for sure - it is often due to incomplete data. For example, when assessing the potential for risks to people, toxicology studies generally involve dosing of sexually mature test animals such as rats as a surrogate for humans. Since we don't really know how differently humans and rats respond, USEPA often employs the use of an uncertainty factor to account for possible differences. Additional consideration may also be made if there is some reason to believe that the very young are more susceptible than adults, or if key toxicology studies are not available.

Variability

This refers to the range of toxic response or exposure - for example, the dose that might cause a toxic response can vary from one person to the next depending on factors such as genetic differences, and preexisting medical conditions. Exposure may vary from one person to the next depending on factors such as where one works, time spent indoors or out, where one lives, and how much people eat or drink.

Conservatism (assumptions made to produce health-protective risk estimates)

The main difference between uncertainty and variability is that variability can only be better characterized, but not necessarily reduced. In addressing uncertainty and variability, the AERA and USEPA risk assessments, include assumptions that may increase risk estimates. This is called, “conservatism”, and is incorporated into human health risk assessments to ensure that they are as health protective as is reasonable.

Reporting uncertainty and variability in an AERA

Table 12 provides an example of several AERA parameters and possible uncertainties that may be associated with each. An informed qualitative judgment needs to be provided as to the effect of each parameter on the risk estimate.

The following table is a template with examples filled in, however it is not comprehensive. The information in this table may not always represent an accurate scenario on a case-by-case basis.
 

Table 12: Example AERA uncertainty and variability table

AERA component
 
Description
 
Effect on risk estimate
 
Approximate magnitude
 
Provide the risk analysis parameter being discussed (examples are provided below)
 
Describe the information source and the specific way the information was used, or what assumption was made.
 
Report whether this likely results in an overestimate, underestimate, or if it may over- or under-estimate overall risk estimates
 
High (would change overall risk estimate by more than 10X),

Moderate (would change overall risk estimate by approximately 2X),

Low (would change risk estimate by less than 2X),

 or Unknown
 
Facility characterization
 
Assumed 24 hour operation, actual operation limited to 6am to 9pm
 
Over-estimate
 
Moderate
 
Chemical identification
 
Literature review of similar facilities but facility studies may not have considered all possible pollutants
 
May underestimate
 
High to unknown
 
Emissions estimates
 
Emission factor source was AP-42: controls on facility emission units differ from AP-42 scenario
 
May over-or under-estimate
 
Unknown
 
Emission factor source AP-42: AP-42 factor is based on average of stack test results
 
Under-estimate
 
Moderate
 
Stack tests with non-detects: Non-detects assumed to be at detection limit to estimate emissions
 
Over-estimate
 
Moderate
 
The worst-case fuel is used to estimate each air toxic emission rate from a boiler allowed to operate with multiple fuels
 
Over-estimate
 
Unknown
 
Stack tests with non-detects: Non-detects assumed to be zero in emissions estimates
 
Under-estimate
 
 
Air dispersion modeling
 
RASS lookup tables: modeled maximum concentrations from several stacks were summed, stacks are distant, which in reality results in maximum concentrations at very different receptor locations
 
Overestimate
 
High
 
RASS lookup tables: Only one short stack  and nearby receptors
 
May over- or under-estimate
 
Unknown
 
AERMOD used to calculate dispersion coefficients entered into RASS: pollutants are not paired in time and space
 
Overestimate
 
Low
 
Exposure scenario
 
Estimated farmer scenario on fenceline, but closest potential farmer is 1km from fenceline.
 
Over-estimate
 
High
 
Multi-pathway assessment
 
MPSFs: unknown how well default assumptions in model represent site conditions
 
Over-or under-estimate
 
Unknown
 
Level 3 multi-pathway analysis: unknown how well default assumptions in model represent site conditions
 
Over-or under-estimate
 
Unknown
 
Toxicity assessment
 
MPCA/MDH hierarchy: a California REL value an order of magnitude lower was published last month. Pollutant was only one of 3 risk-drivers and overall risks will not change much
 
Under-estimate
 
Low
 
Quantitative cancer risk estimates were not adjusted using early life-stage exposure adjustment factors for linear carcinogens.
 
May under-estimate
 
Low
 
Summed hazard index regardless of health endpoints
 
Over-estimate
 
Moderate