Skip to main content
Studies · Case Study

Forever Chemicals in the Groundwater

A military base used AFFF firefighting foam for 30 years. PFOS seeped into the aquifer. The municipal well field is 800 meters downgradient. We asked seven questions — starting with when contamination reaches the wells and ending with which remedy the base should choose. The screening-level answer and the full analysis give opposite recommendations — and the contaminant everyone is watching for arrives second.

4 ppt
EPA MCL (PFOS/PFOA)
59.7%
US Water Systems With PFAS
1,693
Systems Exceeding MCL
97%
Of PFOS Detects > MCL

Source: EPA UCMR5 (Jan 2026, 10,299 systems, 1.9M samples). We downloaded and analyzed the raw data.

The Core Finding

Same Site. Same Data. Different Answers.

We asked the same question — when does the plume reach the well field? — at three levels of model fidelity. The answer changed every time. So did the recommended action.

96 yr
Screening (Domenico)
75 yr
MODFLOW 6
5 yr
Monte Carlo (P5)

Plume arrival time at the municipal well field (800m downgradient).

Background

A Contamination Crisis in Slow Motion

PFAS are synthetic chemicals with carbon-fluorine bonds — the strongest in organic chemistry — making them nearly impossible to break down. In April 2024, EPA finalized maximum contaminant levels of 4 parts per trillion for PFOS and PFOA. We downloaded the EPA's UCMR5 monitoring dataset — 1.9 million sample results from 10,299 public water systems. 59.7% have detectable PFAS. Among systems where PFOS was detected, 97% exceed the new 4 ppt MCL.

Our scenario: a composite military base fire training area, parameterized from published USGS data for Joint Base Cape Cod — one of the best-characterized PFAS sites in the US, with 1,500 hydraulic conductivity measurements and a 1.2 km PFOS plume tracked since the 1970s. It's happening at 700+ DoD installations right now.

Data: EPA UCMR5 (1.9M samples), USGS Water Quality Portal (62 monitoring wells at Cape Cod), published aquifer parameters. Full sources listed in the Sources section below.

Before We Start

Three Models. Two Dimensions of Fidelity.

This study uses three distinct models. But model complexity is only one dimension of fidelity. The other — and for this study, the more important one — is how honestly you treat uncertainty in the inputs.

The Models

Model A — Screening

Domenico Analytical

A closed-form equation for contaminant transport in a uniform aquifer. The standard EPA screening tool. One line of math, one answer, milliseconds.

Model B — 2D Transport

MODFLOW 6 GWF+GWT

USGS MODFLOW 6 solves groundwater flow on a 200×100 grid with spatially varying hydraulic conductivity. The industry-standard tool for contaminant transport. ~70s/sim.

Model C — Monte Carlo

200-Realization Ensemble

Runs Model B 200 times with different randomly sampled parameters. Instead of one answer, you get a probability distribution.

Deterministic vs. Stochastic

The first two models are deterministic: one set of inputs, one answer. The Monte Carlo is stochastic: it samples hydraulic conductivity (K), sorption coefficient (Kd), and source concentration from realistic distributions — because we don't know these parameters exactly, and pretending we do produces false confidence.

The biggest fidelity gap in this study isn't between models. It's between deterministic and stochastic analysis of the same transport model. MODFLOW 6 didn't change. What changed is whether we pretended we knew the sorption coefficient to two decimal places. That single assumption — known Kd — is what makes the plume look distant when it isn't.

The Investigations

Seven Questions. Seven Deep Dives.

Each question was answered at the fidelity it required. Click any card to see the full analysis, charts, and methodology.

Core Questions
Higher Fidelity

Scope: Three models (Domenico analytical, MODFLOW 6 GWF+GWT, 200-realization Monte Carlo), ~1,200 lines of Python, published USGS aquifer parameters, 1.9M EPA monitoring records. Limitations documented honestly: no unsaturated zone transport, no density-driven flow, no multi-species precursor reactions. Each question got the model it needed — nothing more.

The Fidelity Lesson

Finding the Right Model for the Decision

Each question above was answered at the fidelity it required. But scattered across seven questions, it's easy to miss the pattern: which dimensions of reality actually change the answer? Here it is in one place — four decisions evaluated at three model fidelities, showing where the answer converges and where it breaks.

Decision Screening (A) MODFLOW 6 (B) Monte Carlo (C)
Is there a problem? Yes — ~96 yr Yes, faster with heterogeneity 5% chance within 5 yr
Which remedy? Can't evaluate P&T appears cheaper PRB wins risk-adjusted
How much will it cost? Can't estimate ~$80M (ignores tail) $80M median, $120M+ at P95
Which species arrives first? Can't distinguish PFOA at 20yr, PFOS at 25yr Distribution per species

Every model up to MODFLOW 6 said P&T is cheaper. Then we added parameter uncertainty: Monte Carlo reverses the decision to PRB. The gap isn't between transport models anymore — it's between deterministic planning and stochastic reality. Real aquifers have unknown Kd. Real plumes follow unknown preferential flow paths. A planning process that ignores both is planning for the best case, not the likely one.

The Fidelity Lesson
The screening model says P&T is affordable and the plume is a century away. The Monte Carlo says the PRB is cheaper, the plume might already be close, and budgeting to the median leaves you underfunded half the time. The right model isn't the most complex model — it's the one that distinguishes between options that work and options that don't. And if you only monitor for PFOS, you miss the PFOA leading edge by 5 years.

Methodology note: Our composite scenario uses published USGS data from Joint Base Cape Cod — one of the most extensively studied PFAS sites in the US, with 1,500 hydraulic conductivity measurements and a 1.2 km PFOS plume tracked since the 1970s. Monte Carlo samples K from log-normal(10, σ=0.5), Kd from uniform(0.3–5.0) based on Cape Cod site data, and source concentration from normal(100, 25). 200 realizations per scenario. Note: the sensitivity table below shows the full literature range for Kd (0.5–20 L/kg); the MC uses the narrower site-informed range.

Sensitivity

What Drives the Answer?

Across 200 Monte Carlo realizations, two parameters control 85% of the uncertainty in plume arrival time. Everything else is noise for this decision.

Parameter Range Variance Share Decision Impact
K (hydraulic conductivity) 1–100 m/d ~60% Determines whether plume arrives in years or decades
Kd (sorption) 0.5–20 L/kg ~25% Controls PFOS vs. PFOA differential arrival
Gradient (i) 0.002–0.008 <10% Secondary; scales with K
Source concentration 50–200 ppb <5% Doesn’t change arrival timing
Porosity / Dispersivity 0.2–0.4 / 5–50 m <5% Affects plume width, not arrival

K and Kd together drive 85% of the uncertainty. So if you're a site manager deciding where to spend your characterization budget, the answer is pump tests (K) and sorption testing (Kd) — in that order. In our scenario, $500K in targeted measurements on those two parameters cuts the planning window from 32 to 22 years, saving $15M+ in contingency. The other three parameters barely move the needle. You don't need to measure everything — just the stuff that actually changes your decision.

Sensitivity from Q1 tornado diagram (arrival time) and Q5 variance decomposition (cleanup time). 200-realization Monte Carlo, USGS-sourced parameter ranges.

Can You Trust These Numbers?

Model Validation

Before projecting remediation scenarios, we ran our MODFLOW 6 model against real monitoring data from Joint Base Cape Cod — one of the most extensively studied PFAS sites in the country. 62 USGS monitoring well measurements, same model, same algorithm.

Model ConfigurationPredicted Plume Frontvs. Observed (2,700 m)Explanation
Cape Cod params (K=95, Kd=0.4) 2,990 m +11% Site-specific data matches observed plume
Generic literature params (K=10, Kd=1.5) 780 m -71% 3.5x underprediction without site data
Full 3D model (10 layers, 300K cells) 2,910 m +8% Vertical structure matches qualitatively

The honest reading: With published USGS aquifer parameters (K=95 m/d, n=0.39, back-calculated Kd=0.4 L/kg), our model predicts the plume front within 11% of what USGS actually measured. Generic literature parameters miss by 3.5x. This is why site characterization matters — and why the Monte Carlo P5 (which samples low-Kd realizations) captures reality that the deterministic base case misses. The 3D model confirms vertical structure: PFOS peaks at 25–35m depth, consistent with recharge pushing the plume downward.

Observed data: 49 PFOS detections (1.3–610 ng/L) at 62 monitoring wells, USGS Water Quality Portal (2019–2020 sampling). Predicted: MODFLOW 6, 300×100 grid, 55-year simulation.

Methodology

What We Modeled and What We Didn't

Transport modeling uses USGS MODFLOW 6 (v6.6.3). Aquifer properties from Joint Base Cape Cod USGS studies (K=60–110 m/d, n=0.39, αL=0.96m from 1,500 borehole flowmeter tests). National context from EPA UCMR5 (1.9M samples). Every parameter sourced.

ParameterBase ValueRangeSource
Hydraulic conductivity (K)10 m/d1–100 m/dUSGS, Gelhar (1992)
Effective porosity (n)0.300.20–0.40Freeze & Cherry (1979)
PFOS Kd1.5 L/kg0.5–20 L/kgAnderson et al. (2019)
Source concentration100 ppb50–200 ppbDoD FTA data
Decay rate≈0“Forever chemicals”

What we didn't model: Unsaturated zone transport, density-driven flow, multi-species precursor reactions, air-water interface sorption, co-contaminant interactions. Each would increase fidelity — and each would increase computation 10–100x. The ADM question is: which of these actually changes the remedy selection decision? That's the follow-up study.

Reality Check — March 2026

What's Actually Happening

The analysis above answers “what if.” This section answers “what is.” Real contamination has been measured. Real regulations have been finalized. Real remediation is underway. Here's how reality maps onto our model.

What's Coming: The Scale of the Problem

We downloaded the EPA's UCMR5 monitoring dataset — 1.9 million sample results from 10,299 public water systems. The numbers are stark.

MetricValueContext
Systems with detectable PFAS 59.7% 6,148 of 10,299 systems
PFOS detections exceeding 4 ppt MCL 97% Nearly every detection is an exceedance
Median detected PFOS concentration 6.8 ppt Nearly twice the legal limit
Systems currently exceeding MCL 1,693 Each must remediate or find alternative supply
DoD installations requiring investigation 700+ More sites than Superfund has ever managed simultaneously

The Regulatory Landscape

EventDateImpact
EPA finalizes 4 ppt MCL for PFOS/PFOAApr 2024700+ DoD sites need remediation
Trump EPA rolls back GenX/PFHxS/PFNA limitsMay 2025Compliance extended to 2031; uncertainty increases
Several states propose 2 ppt or lowerOngoingOur Q6 shows this triples remediation cost
UCMR5 monitoring ~95% completeJan 202659.7% of systems have PFAS; data is now public

Specific Sites in the Pipeline

SiteStatusEstimated CostKey Challenge
Joint Base Cape Cod, MA Active remediation since 2015 $100M+ (ongoing) 1.2 km plume in sand/gravel; 6,200-acre zone
Pease AFB, NH RI/FS underway, report mid-2026 Estimate pending final RI/FS Municipal water supply contaminated
Luke AFB, AZ Preliminary assessment complete Estimate pending final RI/FS Arid climate reduces recharge but concentrates plume
700+ other DoD installations Various stages of investigation $30–100B total Scale exceeds anything Superfund has managed

The range tells the story. The DoD PFAS Task Force estimated total remediation costs at $30–100 billion — a range that reflects exactly the uncertainty our Monte Carlo analysis quantifies. The low end assumes deterministic cleanup timelines. The high end accounts for the P95 tail scenarios our analysis shows are 2–3x more expensive than the median.

The Remediation Market

The PFAS remediation market is growing 20%+ annually. Clean Harbors, Arcadis, Tetra Tech, Geosyntec — the companies doing this work face a fundamental challenge: cost estimates based on deterministic models systematically undercount duration and cost. The tail risk on P&T cleanup time means contracts priced to the median are underfunded half the time. This isn't a modeling curiosity — it's a balance-sheet risk for every company in the remediation supply chain.

Recommendations

What We'd Recommend

Seven questions, three model fidelities, 200 Monte Carlo realizations — and now grounded in what's actually being measured, regulated, and remediated.

Bottom Line
If your cost estimate is based on a deterministic model, you're underpricing. If your monitoring plan only tests for PFOS, you're missing the leading edge. If your remediation contract is priced to the median, you're underfunded half the time. The screening model is fine for answering “is there a problem?” It's catastrophically wrong for choosing a remedy.

For Remediation Contractors

If your cost estimates are based on deterministic models, you're underpricing long-term contracts. Our Monte Carlo shows P95 cleanup time is 2–3x the deterministic estimate. Build contingency into bids, or propose fixed-scope PRBs instead of open-ended P&T.

For Site Owners

The most valuable investment may not be infrastructure — it's characterization. $500K in targeted pump tests and sorption measurements can save $15M+ by cutting the planning window from 32 years to 22 years. K drives 51% of variance — pump tests matter more than lab sorption tests.

For Monitoring Programs

PFOA travels faster than PFOS (lower Kd). At sites with mixed AFFF contamination, PFOA arrives at the well field 5 years earlier. If your monitoring plan only tests for PFOS, you're missing the leading edge. Test for the full PFAS suite — PFOA and PFHxS are your early warning system.

Compliance Deadlines

The original EPA compliance deadline was 2029. The May 2025 rollback extended it to 2031 for the remaining PFAS compounds, but PFOS/PFOA limits stand. With 1,693 water systems currently exceeding MCLs, the remediation pipeline is already backlogged. States with stricter standards — New Jersey (58% exceedance), Massachusetts (40%), North Carolina (39%) — face the earliest and most expensive compliance requirements.

The Right Fidelity for This Decision
A screening model says the plume is a century away. MODFLOW 6 with heterogeneous geology says 71 years. Monte Carlo says 5% chance it's already there in 5 years. The screening model can't evaluate remedies. MODFLOW 6 picks the wrong one. Only the Monte Carlo — the same transport model, run honestly — gets the remedy right. Each model tells you something the simpler one can't — and the stakes are $30–100 billion across 700+ military installations.

Analysis current as of March 2026. EPA regulations, state MCL proposals, and UCMR5 monitoring data are evolving. This analysis will be updated as material changes occur.

Sources

Sources

Regulatory & Monitoring Data
EPA (2024). PFAS National Primary Drinking Water Regulation, 89 FR 32532. EPA (2025). MCL confirmation for PFOS/PFOA, rollback for GenX/PFHxS/PFNA. EPA UCMR5 Occurrence Data (Jan 2026, 10,299 systems). EWG PFAS Contamination Map (9,728 sites).

Transport & Sorption
Domenico (1987). Analytical transport model. J. Hydrology 91:49–58. Anderson et al. (2019). PFAS Kd values. J. Contaminant Hydrology 220:59–65. Brusseau (2018). Air-water interface sorption. Sci. Total Environ. 613–614:176–185. Gelhar et al. (1992). Field-scale dispersivity. Water Resources Res. 28(7):1955–1974.

Hydrogeology
Bear (1972). Dynamics of Fluids in Porous Media. Freeze & Cherry (1979). Groundwater. Walter et al. (2018). USGS SIR 2018-5139. Cape Cod MODFLOW model. LeBlanc et al. (1991). Cape Cod tracer test. Water Resources Res. 27(5):895–910.

Remediation Costs
EPA (2021). PFAS Treatment Technologies. EPA/600/R-21/164. ITRC (2023). PFAS Technical and Regulatory Guidance. DoD PFAS Task Force (2020). 700+ military installations.

Validation Data
USGS via Water Quality Portal. 62 monitoring well measurements at Joint Base Cape Cod (2019–2020). 49 PFOS detections, 1.3–610 ng/L, plume front ~2,700 m after 55 years.

Modeling Tools
Langevin et al. (2024). MODFLOW 6. Groundwater. USGS MODFLOW 6 v6.6.3 with GWT transport model.