EPA Four-Step Risk Assessment Paradigm

Quick Answer: EPA risk assessment is four steps: (1) hazard identification, (2) dose-response, (3) exposure assessment, (4) risk characterization. Know what each step produces and how they connect to cleanup goals and permits.

Risk assessment (~6% of FE Environmental) quantifies whether chemical exposures pose unacceptable health risk. Environmental engineers use it at Superfund sites, brownfields, air toxics reviews, and drinking water standards development.

Step 1 — Hazard Identification

Ask: Can the agent cause harm? Review epidemiology, animal studies, and mechanistic data.

Hazard identification does not quantify risk — it establishes plausibility of harm. A chemical may be hazardous but pose negligible risk if exposure is tiny.

Exam trap: Listing a chemical as a HAP or CERCLA hazardous substance signals regulatory concern but is not the same as completing a site-specific risk assessment.

Step 2 — Dose-Response Assessment

Links dose (mg/kg-day) or air concentration to probability or severity of effect.

Non-cancer chronic toxicity:

• Reference Dose (RfD) — oral daily dose unlikely to cause appreciable harm over lifetime (with uncertainty factors).
• Reference Concentration (RfC) — inhalation analog.

Cancer:

• Slope factor (SF) — upper-bound potency (risk per mg/kg-day) for genotoxic carcinogens.
• Unit risk — excess lifetime cancer risk per µg/m³ in air.

Uncertainty factors divide NOAEL/LOAEL from animal studies to account for interspecies and intraspecies variability.

Worked example: NOAEL = 100 mg/kg-day; total uncertainty factor = 1,000 → RfD = 0.1 mg/kg-day.

Step 3 — Exposure Assessment

Estimates average daily dose by route:

[ \text{ADD} = \frac{C \times IR \times EF \times ED}{BW \times AT} ]

Pathways: groundwater ingestion, soil ingestion (children), inhalation of volatiles and particulates, dermal contact, food chain (bioaccumulation).

Exposure point concentration (EPC) may use 95% UCL of soil samples — conservative for risk, not the arithmetic mean.

Step 4 — Risk Characterization

Combines prior steps:

Non-cancer hazard quotient (HQ): [ HQ = \frac{\text{ADD}}{RfD} ] Sum HQs across pathways/chemicals for hazard index (HI). HI > 1 suggests potential non-cancer concern (not a legal bright line everywhere).

Cancer risk: [ \text{Risk} = ADD \times SF \quad \text{(oral)} ] or concentration × unit risk (inhalation).

EPA often uses 1×10⁻⁶ to 1×10⁻⁴ as a managerial range for site cleanup; 1×10⁻⁶ is a common point of departure in Superfund (policy context).

Uncertainty analysis documents data gaps — which assumptions drive results.

Ecological Risk Assessment (Brief)

Parallel framework for receptors (fish, birds, plants) with LOAEL/NOAEL in field or lab studies. Sediment quality criteria and tissue residues may apply.

Risk Assessment vs. Risk Management

Engineers conduct assessment; agencies and owners choose action levels.

Common FE Scenarios

• Identify correct step from a narrative ("toxicologist reviews rat cancer study" → hazard identification).
• Pick pathway needing highest exposure (child ingesting contaminated soil vs. adult worker inhalation).
• Know that RfD is for non-cancer chronic oral reference.

Quality and Data Issues

• Detection limits — nondetects handled with half DL or statistical models (conceptual).
• Bioavailability adjustments reduce effective dose for some metals in soil.
• Monte Carlo replaces single-point estimates with distributions (advanced; conceptual on FE).

Exam trap: HQ > 1 indicates potential exceedance of the reference dose — it is not a clinical diagnosis of disease in an individual.

The four-step paradigm structures every human-health risk problem on the exam. Memorize the sequence and what each step outputs before tackling dose and HQ math in the next section.

Uncertainty and Data Quality

Tier 1 screening uses conservative assumptions (high intake, 95% UCL concentrations). Tier 2 refines with site-specific exposure frequency or bioavailability. Monte Carlo simulation replaces single-point estimates with distributions — useful when variance is large but beyond most FE calculations.

Data quality objectives (DQOs) plan how many soil samples and what spacing before excavation decisions — conceptual link between statistics chapter and risk.

Ecological Risk Assessment

Ecological risk assessment (ERA) evaluates receptors such as fish, birds, and benthic organisms. Sediment quality guidelines and tissue residue criteria complement human-health risk. A remedial action level protective of humans may not protect ecological receptors — dual pathways matter at wetlands and fisheries.

FE Study Drill

Memorize the four steps in order and one output of each: hazard identification (yes/no harm potential), dose-response (RfD or SF), exposure assessment (ADD by pathway), risk characterization (HQ, cancer risk, uncertainty narrative). Exam items frequently ask which step a described activity belongs to.

Four Steps Expanded

Hazard identification — does agent cause effect? Dose-response — CSF, RfD, unit risk. Exposure assessment — intake pathways. Risk characterization — HQ, cancer risk, uncertainty.

Default vs Site-Specific

Screening uses conservative defaults (body weight 70 kg, 70 yr lifetime). Site-specific replaces with measured concentrations and receptor populations.

Uncertainty Factors

RfD = NOAEL / (UF × MF). UF accounts for interspecies, intraspecies, LOAEL→NOAEL, subchronic→chronic.

Ecological Risk

Parallel paradigm for receptors (fish, birds) — not identical to human health risk on FE but similar structure conceptually.