🌍 Water Contaminant Risk Assessment

Source Water & Subsurface Contaminant Transport

Drinking-water contamination begins long before a treatment plant intake. Lead enters at the very end of the network through corroding pipes; arsenic is a naturally occurring geogenic contaminant weathering out of bedrock into aquifers; PFAS are synthetic, virtually indestructible industrial chemicals that migrate as slow-moving groundwater plumes. Each has a fundamentally different origin, transport mechanism, and control strategy.

• ~9.2 M: US lead service lines (2023) (EPA national inventory)
• 10 ppb: Arsenic MCL (EPA, effective since 2006)
• >14,000: PFAS compounds identified (per OECD/EPA inventories)
• 2014–19: Flint, MI crisis (switched source, no corrosion control)

Lead — a distribution-network problem, not a source-water problem

Unlike arsenic or PFAS, lead is almost never present in the raw source water itself. It is introduced downstream, inside the distribution system, primarily from lead service lines (LSLs) — the pipes connecting the water main to individual homes, widely installed before lead pipe was banned for new installations by the 1986 Safe Drinking Water Act amendments.

When water is corrosive (low pH, low alkalinity, or lacking a protective scale), it dissolves lead from pipe walls, lead solder joints, and brass fixtures. The 2014–2019 Flint, Michigan crisis is the canonical case: the city switched its source from Lake Huron (via Detroit) to the Flint River without adding a corrosion inhibitor (orthophosphate). The more corrosive, higher-chloride river water stripped the protective scale that had built up inside LSLs over decades, and lead concentrations in some homes exceeded 100 ppb — many multiples of the EPA action level.

As of the EPA's 2023 national inventory, an estimated 9.2 million lead service lines remain in use across the United States, disproportionately concentrated in older Midwest and Northeast cities (Chicago alone accounts for roughly 400,000).

Because lead enters at the tap, not the source, source-water treatment cannot remove it. The only reliable controls are corrosion-control chemistry (orthophosphate dosing to rebuild protective pipe scale) and physically replacing the lead pipe.

Arsenic — geogenic weathering into aquifers

Arsenic is naturally present in many rock-forming minerals, most notably arsenopyrite (FeAsS). Where groundwater flows through arsenic-bearing bedrock or alluvial sediments — parts of the US Southwest, New England, the upper Midwest, and most severely the Bengal Basin of Bangladesh and West Bengal, India — arsenic dissolves into the aquifer at concentrations that can reach hundreds of ppb without any industrial input at all.

Arsenic speciation controls both toxicity and treatability. In groundwater it occurs mainly as:

• Arsenite, As(III) — the dominant form in anoxic/reducing groundwater; more toxic, uncharged at neutral pH, and poorly removed by conventional coagulation or adsorption without pre-oxidation. • Arsenate, As(V) — the dominant form in oxic surface water; negatively charged, binds readily to iron/aluminum coagulant flocs and activated alumina.

This speciation dependence is why treatment trains for groundwater arsenic almost always include an oxidation step (chlorine, permanganate, or ozone) to convert As(III) to As(V) before removal.

PFAS — persistent industrial "forever chemicals"

Per- and polyfluoroalkyl substances (PFAS) are a class of over 14,000 synthetic compounds characterized by carbon–fluorine bonds — among the strongest single bonds in organic chemistry — making them extraordinarily resistant to hydrolysis, photolysis, and biodegradation. This same property that made them useful (as firefighting foams, non-stick coatings, and stain repellents) makes them essentially permanent environmental contaminants.

Primary drinking-water sources of PFAS contamination:

• Aqueous film-forming foam (AFFF): used for decades at military bases and airports for fuel-fire suppression training; PFOS and PFOA leach directly into soil and groundwater at these sites, producing plumes that can extend for miles. • Industrial manufacturing and disposal: historical PFAS production and use by companies including 3M and DuPont (the subject of major PFOA litigation in West Virginia/Ohio) discharged PFAS to rivers and landfills. • Landfill leachate and biosolids application: PFAS-containing consumer products break down in landfills and can migrate to groundwater, or PFAS in wastewater biosolids applied as fertilizer can leach to soil and groundwater.

Because PFAS molecules are mobile and do not sorb strongly to most soils, plumes migrate largely undiminished with the groundwater flow, reaching drinking-water wells miles from the original source, sometimes decades after the release occurred.

Treatment-Plant Removal Efficiency by Process

No single treatment process removes all three contaminant classes equally well. Effective utilities layer coagulation/flocculation, granular activated carbon (GAC), reverse osmosis (RO), and ion exchange (IX) — selecting and sequencing processes based on which contaminant, and which chemical species, dominates their source water.

• >95%: RO removal — PFAS/arsenic (membrane rejection)
• 60–90%: Coagulation — As(V) removal (with ferric coagulant)
• ~30%: Coagulation — As(III) removal (without pre-oxidation)
• 50–90%: GAC — long-chain PFAS (PFOA/PFOS; declines with use)

Coagulation/flocculation — effective for arsenic, not for lead or PFAS

Coagulation uses iron (ferric chloride/sulfate) or aluminum (alum) salts to form positively charged metal hydroxide flocs that adsorb and co-precipitate dissolved contaminants, which are then removed by sedimentation and filtration.

For arsenic, this is highly effective — but only for the As(V) (arsenate) species, which is anionic and binds strongly to ferric hydroxide floc surfaces, achieving 60–90% removal. As(III) (arsenite) is largely uncharged at neutral pH and passes through coagulation nearly untouched (~30% removal or less) unless first oxidized to As(V) with chlorine or potassium permanganate.

For PFAS, coagulation is essentially ineffective — PFAS molecules are too small and too water-soluble to be captured by metal hydroxide flocs; typical removals are under 10%.

For lead, coagulation at the plant is largely irrelevant because lead is not present in the source water — it is introduced downstream in the distribution system, so plant-scale removal cannot address it at all.

GAC, reverse osmosis, and ion exchange for PFAS and arsenic

Granular activated carbon (GAC) removes organic contaminants by adsorption onto its high-surface-area porous carbon matrix (500–1,500 m²/g). For PFAS, GAC performance is strongly chain-length dependent: long-chain compounds like PFOA and PFOS adsorb well (50–90% removal, improving with fresh media and longer empty-bed contact time), while short-chain PFAS (PFBA, PFBS) break through much faster and are poorly captured. GAC beds require periodic regeneration or replacement as adsorption capacity is exhausted — utilities monitor for "breakthrough" using influent/effluent sampling.

Reverse osmosis (RO) forces water through a semi-permeable membrane under pressure, physically excluding molecules and ions above a size/charge threshold. RO is the most broadly effective single technology across all three contaminant classes: >95% removal of PFAS (including short-chain compounds GAC struggles with), >95% removal of arsenic (both species, once past the membrane's size-exclusion and charge-repulsion mechanisms), and effective for dissolved lead. The tradeoffs are high energy cost, membrane fouling, and a concentrated reject brine stream requiring disposal.

Ion exchange (IX), particularly anion exchange resin (AIX) for PFAS, exchanges contaminant ions for benign ions (typically chloride) on a resin bed. AIX is highly effective for PFAS sulfonates (PFOS-family) and shows good performance for anionic As(V); performance for As(III) is poor without pre-oxidation, mirroring the coagulation limitation.

Because no single process handles all three contaminants and both arsenic oxidation states, real-world compliance often requires an oxidation pre-step (chlorine/permanganate) followed by coagulation for arsenic, plus a parallel GAC or IX train specifically for PFAS — adding significant capital and operating cost for small utilities.

Removal efficiency by process and contaminant

Tap Water Monitoring & Regulatory Limit Comparison

Treatment efficacy means nothing without verification at the tap. Utilities conduct compliance monitoring using ICP-MS for trace metals and LC-MS/MS for PFAS, then compare results against enforceable federal limits — the EPA Lead and Copper Rule action level, the arsenic Maximum Contaminant Level, and the 2024 PFAS National Primary Drinking Water Regulation.

• 15 ppb: Lead action level (EPA Lead and Copper Rule)
• 10 ppb: Arsenic MCL (EPA, since 2006 (was 50 ppb))
• 4.0 ppt: PFOA / PFOS MCL (EPA 2024 final rule, each)
• 10 ppt: GenX (HFPO-DA) MCL (2024 rule; PFHxS/PFNA also 10 ppt)

Analytical methods — ICP-MS for metals, LC-MS/MS for PFAS

Inductively Coupled Plasma Mass Spectrometry (ICP-MS) is the standard method (EPA Method 200.8) for quantifying lead and arsenic in drinking water. Samples are nebulized into an argon plasma at ~10,000 K, ionizing essentially every element; ions are separated by mass-to-charge ratio in a quadrupole or magnetic-sector analyzer. ICP-MS routinely achieves detection limits below 0.1 ppb — well under the regulatory thresholds — and can quantify dozens of elements simultaneously from a single injection.

Liquid Chromatography–Tandem Mass Spectrometry (LC-MS/MS, EPA Method 537.1 / 533) is required for PFAS because these compounds are non-volatile and thermally labile, ruling out gas-chromatography approaches. Water samples are extracted by solid-phase extraction (SPE) to concentrate trace-level PFAS, separated by liquid chromatography, then detected by tandem mass spectrometry using multiple-reaction monitoring, which fragments each PFAS ion into characteristic daughter ions for unambiguous identification even at part-per-trillion concentrations — roughly a thousandfold more sensitive than typical ICP-MS metal detection limits, reflecting PFAS toxicity at vastly lower concentrations.

Lead and Copper Rule — an action level, not a health-based MCL

Lead is regulated differently from arsenic and PFAS. Because lead enters primarily from plumbing rather than source water, EPA cannot set a treatment-based Maximum Contaminant Level at the plant. Instead, the Lead and Copper Rule (LCR, 1991; substantially revised by the Lead and Copper Rule Improvements, finalized 2024) sets a 15 ppb action level based on 90th-percentile results from tap sampling at high-risk homes (typically those with lead service lines).

Exceeding the action level does not itself constitute a violation — it triggers required actions: enhanced corrosion control treatment, public education, and accelerated lead service line replacement. The 2024 LCR Improvements require most utilities to replace all remaining lead service lines within 10 years (by 2037 for most systems) and lowered the action level trigger for replacement programs.

The arsenic MCL of 10 ppb, in force since 2006 (tightened from the prior 50 ppb standard set in 1942), is a true enforceable Maximum Contaminant Level: exceeding it is a direct violation requiring immediate corrective treatment.

The 2024 EPA PFAS National Primary Drinking Water Regulation was the first-ever federally enforceable limit for PFAS in the 50-year history of the Safe Drinking Water Act, setting individual MCLs of 4.0 ppt for PFOA and PFOS, 10 ppt for PFHxS, PFNA, and GenX (HFPO-DA), plus a Hazard Index limit for mixtures of these compounds with PFBS.

Why distribution-network sampling location matters

Because lead contamination originates inside individual service lines rather than uniformly across a system, a single sample from a treatment-plant outlet is meaningless for lead compliance. Utilities must sample at the tap in homes most likely to have lead — those built before the 1986 lead-pipe ban, especially those with confirmed or suspected lead service lines — and the same street can show compliant water at one home and multiples of the action level two doors down, depending on which service line, solder, and fixture materials are present and how long water has stood stagnant in the pipe (first-draw vs. flushed samples can differ several-fold).

Arsenic and PFAS, by contrast, are far more uniform across a distribution network because they originate in the source water itself rather than the plumbing, so a representative sample near the plant outlet is generally predictive of system-wide levels — though PFAS blending from multiple wells with different contamination levels can still create localized variation.

Dose–Response Modeling — From Concentration to Individual Risk

A measured concentration only becomes a health risk once translated into an estimated dose an individual actually receives, then compared against toxicological benchmarks. Non-cancer risk is screened using the Hazard Quotient; cancer risk from genotoxic carcinogens like inorganic arsenic is estimated using a cancer slope factor. Lead is treated as a special case with no identified safe threshold.

• 2 L/day: Default ingestion rate (EPA adult drinking-water default)
• 1.5: Arsenic oral CSF ((mg/kg-day)⁻¹, EPA IRIS)
• 10⁻⁶–10⁻⁴: Target ELCR range (EPA acceptable risk window)
• none: Lead reference dose (no safe threshold per CDC/EPA)

Hazard Quotient — screening non-cancer risk

For non-carcinogenic effects, EPA risk assessment uses the Hazard Quotient:

HQ = Average Daily Dose (ADD) / Reference Dose (RfD)

where ADD (mg/kg-day) = (Concentration × Ingestion Rate) / Body Weight, using EPA default exposure assumptions of 2 L/day ingestion and 70 kg (adult) or age-adjusted values for children, who receive a higher dose per kilogram of body weight from the same tap-water concentration.

The Reference Dose (RfD) is an estimate of daily lifetime exposure a population, including sensitive subgroups, can sustain without appreciable risk of adverse non-cancer effects — derived from animal or epidemiological studies, divided by uncertainty factors (typically 10× for interspecies extrapolation, 10× for sensitive individuals).

HQ < 1: exposure is below the level of concern. HQ ≥ 1: exposure exceeds the reference dose; potential for adverse effects, warranting risk management action.

For mixtures of multiple PFAS compounds acting through a common mechanism, EPA's 2024 rule uses a Hazard Index (HI) — the sum of Hazard Quotients across PFHxS, PFNA, GenX, and PFBS — with an HI limit of 1.0, since combined exposure to several PFAS congeners below their individual limits can still pose cumulative risk.

Cancer slope factor and excess lifetime cancer risk

Inorganic arsenic is classified by EPA as a Group A human carcinogen (sufficient evidence in humans) based on epidemiological studies of populations with high natural arsenic exposure, notably in Taiwan and Bangladesh, linking arsenic ingestion to bladder, lung, and skin cancers.

For carcinogens without an identified threshold, EPA IRIS assigns a Cancer Slope Factor (CSF) — the increase in cancer risk per unit of lifetime average daily dose — derived from the linearized low-dose extrapolation of dose-response data. For inorganic arsenic, the oral CSF is approximately 1.5 (mg/kg-day)⁻¹.

Excess Lifetime Cancer Risk (ELCR) is then:

ELCR = Lifetime Average Daily Dose × CSF

EPA's generally acceptable risk range for regulatory decision-making is 10⁻⁶ to 10⁻⁴ — meaning no more than 1 additional cancer case per 10,000 to 1,000,000 people exposed over a lifetime, beyond background cancer rates. Drinking water at the 10 ppb arsenic MCL over a lifetime corresponds to an ELCR on the order of 10⁻⁴ to 10⁻³ for some exposure scenarios — near or above the top of the "acceptable" risk band, reflecting that the arsenic MCL is a technology-and-cost-feasibility standard as much as a purely health-based one.

EPA and CDC state explicitly that there is no known safe level of lead exposure — blood lead levels as low as 3.5 µg/dL are associated with measurable cognitive deficits in children. Lead is therefore managed by minimizing exposure toward zero (the LCR "action level" concept) rather than by a threshold reference dose.

PFAS toxicology — immune, developmental, and cancer endpoints

PFOA and PFOS toxicology spans multiple endpoints beyond cancer: immune suppression (reduced vaccine antibody response in children — among the most sensitive endpoints identified), developmental effects (low birth weight), thyroid hormone disruption, elevated cholesterol, and kidney/testicular cancer associations (PFOA specifically, per epidemiological studies around the DuPont Washington Works facility, West Virginia — the C8 Science Panel studies).

Because PFOA and PFOS bioaccumulate with serum half-lives of roughly 2–5 years in humans (far longer than most environmental contaminants, which clear in days to weeks), even low chronic drinking-water exposure produces substantial body-burden accumulation over years, which is a major reason EPA set MCLs at the parts-per-trillion level — near the current limit of reliable analytical quantification — rather than the parts-per-billion levels typical of metals regulation.

Population Risk Characterization & Remediation Strategy

Individual dose-response estimates must be scaled to the exposed population to guide public-health action and resource allocation — and that population is rarely exposed uniformly. Contaminant exposure, particularly from lead service lines, tracks closely with aging infrastructure concentrated in older, historically under-invested, and often lower-income and minority communities, making water contaminant risk an environmental justice issue as much as a technical one.

• ~100,000: Flint, MI population exposed (2014–2019 crisis)
• ~23,000: Newark, NJ LSL replaced (accelerated program, 2019–2021)
• 2022: Jackson, MS crisis (system-wide boil-water/collapse)
• 2037: LCR Improvements deadline (nationwide LSL replacement target)

Characterizing population-level risk

Population risk characterization combines the individual-level dose-response outputs (HQ, ELCR) with the size and demographic distribution of the exposed population to estimate the number of excess cases and to prioritize remediation. Key inputs include:

• Population served: from a small rural well system (hundreds of people) to a major metropolitan utility (millions) • Exposure heterogeneity: not all residents receive the same dose — proximity to a PFAS plume, presence of a lead service line, and household water-use patterns all vary • Susceptible subpopulations: children (higher dose per body weight, greater neurodevelopmental sensitivity to lead), pregnant individuals (developmental toxicity), and immunocompromised individuals (PFAS immune effects) drive disproportionate risk within the same nominal exposure

A population-weighted excess case estimate multiplies per-capita ELCR or HQ-exceedance probability by the number of people in each exposure stratum, producing an expected number of attributable adverse health outcomes across the service area — the metric regulators and utilities use to prioritize capital investment.

Environmental justice and disparate exposure

Multiple documented water crises illustrate how contaminant exposure concentrates in disadvantaged communities:

• Flint, Michigan (2014–2019): a majority-Black city under state-appointed emergency management switched water sources to cut costs without adding required corrosion control, exposing roughly 100,000 residents, including thousands of children, to elevated lead. • Newark, New Jersey (2016–2021): elevated lead in a majority-Black and Hispanic city led to a court-ordered, ultimately successful accelerated program that replaced roughly 23,000 lead service lines in about two years — demonstrating that rapid, well-funded replacement is achievable when prioritized. • Jackson, Mississippi (2022): decades of deferred infrastructure investment in a majority-Black city culminated in a near-total system collapse after flooding, leaving most of the city without safe water for weeks.

These cases share common structural drivers: aging pre-1986 infrastructure concentrated in older urban cores, historical disinvestment, and — in some cases — governance failures that delayed corrosion-control or capital-improvement decisions despite known risk. National mapping of lead service line locations (required under the 2024 LCR Improvements) is intended to make this disparity visible and drive equity-weighted prioritization of replacement funding.

The EPA's 2024 Lead and Copper Rule Improvements requires most water systems to achieve 100% lead service line replacement within 10 years and mandates publicly accessible LSL inventories — shifting the burden of proof from residents discovering contamination to utilities proactively identifying and replacing risk.

Remediation strategy — matching the fix to the failure point

Because lead, arsenic, and PFAS fail at different points in the source-to-tap pathway, effective remediation is layered:

• Full lead service line replacement: the only permanent fix for lead — removing the pipe eliminates the exposure pathway entirely, unlike corrosion control, which manages but does not eliminate risk. The 2024 LCR Improvements mandate this nationwide, with an emphasis on replacing full lines (not partial replacements, which can temporarily worsen lead release by disturbing scale). • Point-of-use (POU) and point-of-entry (POE) filters: NSF/ANSI Standard 53-certified activated-carbon or reverse-osmosis filters at the tap or home entry point provide an immediate, lower-cost interim measure while service line replacement or centralized treatment upgrades are completed — critical for protecting residents during the multi-year replacement window. • Wellhead and centralized treatment upgrades: for arsenic and PFAS, installing or upgrading GAC, ion-exchange, or RO treatment at the wellhead or treatment plant addresses the contaminant at its source rather than relying on individual household action, providing protection for the entire service population simultaneously. • Corrosion control optimization: orthophosphate dosing calibrated to source-water chemistry (pH, alkalinity) remains an essential complementary measure everywhere lead service lines are still present, reducing — though not eliminating — lead release while replacement proceeds.

The most effective public-health strategy combines all four: immediate POU protection, accelerated permanent LSL replacement, source-level treatment for geogenic/industrial contaminants, and ongoing compliance monitoring to verify that population-level risk is actually declining toward the EPA's target ELCR range.