P&T vs. PRB: Which Remedy Wins?
Pump-and-treat is the default. A permeable reactive barrier costs more upfront but works passively. With early intervention, three extraction wells contain the plume in every Monte Carlo realization and P&T beats PRB at every horizon.
How Does Pump-and-Treat Work?
The concept is simple: extract contaminated groundwater, run it through granular activated carbon (GAC) filters to strip the PFAS, and discharge the clean water back to the aquifer or a surface outfall. Pump-and-treat is the most common PFAS remedy because it's well-understood, regulators trust it, and the capital costs are manageable.
Our configuration places three extraction wells across the plume centerline at x = 600 m — between the source and the municipal well field at x = 800 m. The pumps run continuously, pulling contaminated water through GAC for the life of the project. The 50-realization Monte Carlo samples Kd, hydraulic conductivity, gradient, and porosity from published ranges to test whether the well field is reliably protected.
Does P&T Protect the Well?
The extraction wells at x = 600 m intercept the plume before it reaches the monitoring well at x = 800 m. With early intervention, the cleanup-time question is moot: the well is never contaminated to begin with. The 50-realization ensemble samples the full published range of Kd, K, gradient, and porosity. Containment success rate: 100%.
Modeling caveat: these results assume P&T begins at the contamination event (t = 0). In practice, contamination is detected after years of migration; delaying P&T 10–30 years would change this picture. The model demonstrates that early intervention with P&T is robust at this site geometry — it does not model late-discovery scenarios where the plume has already passed the extraction wells.
What Does It Actually Cost?
Capex is $30M at t = 0 (well field, treatment plant, discharge infrastructure). Across the 50-realization Monte Carlo ensemble, total NPV converges to $31.9M at every horizon — capex dominates; discounted opex is small because containment is reliable, and the 3% real discount rate compresses the long tail.
P50 NPV equals P95 NPV at $31.9M. There is no tail to budget against in this configuration.
For early-intervention P&T at this site, NPV is essentially fixed at ~$32M. Budget risk lives elsewhere — in late detection, in compliance scope changes, in well-field expansion. Not in the cleanup duration distribution.
Is a PRB Better?
A permeable reactive barrier is a trench filled with reactive media (typically GAC or ion exchange resin) installed perpendicular to groundwater flow. The plume flows through it passively — no pumps, no energy, no operators. Capex is higher ($50M vs. $30M for P&T), and the media must be replaced every 15 years at roughly $20M each cycle.
In principle the PRB has no operating tail to amortize. In this configuration, that advantage is overwhelmed by capex plus replacement: 30-year NPV is $80.9M, 100-year NPV is $99.1M. P&T runs continuously but its discounted opex tops out near $32M.
At the 30-year horizon, P&T NPV (P50 = P95 = $31.9M) is about 40% of PRB ($80.9M). Stretch the horizon to 100 years and PRB grows toward $100M while P&T stays flat. The PRB only beats P&T if pumping has to run far longer than the plume actually lasts — which is exactly the scenario this site geometry rules out.
The decision is robust here, not contested. P&T started at t = 0 wins deterministically and stochastically. The interesting fidelity question shifts: when does a PRB look better? Answer: when extraction wells can’t be sited downgradient of the source, when continuous opex carries political risk, or when discovery is late and the plume has already passed the candidate well field. None of those apply at this site.
How We Modeled This
Remediation costs from EPA (2021) treatment technology report and ITRC (2023) PFAS guidance. All costs in net present value at a 3% real discount rate. P&T: $30M CAPEX (well field, treatment plant, discharge infrastructure) plus $2M/year O&M (energy, GAC replacement, monitoring, labor). PRB: $50M CAPEX (excavation, reactive media, backfill) plus $20M media replacement every 15 years.
Cleanup duration driven by MODFLOW 6 transport model with Kd sampled from Anderson et al. (2019) range (0.5–20 L/kg, log-normal). 50 realizations × 150-year horizon. Each realization runs P&T until all monitoring wells read below 4 ppt for 4 consecutive quarters.
Sources: EPA (2021) EPA/600/R-21/164. ITRC (2023) PFAS Technical and Regulatory Guidance. NPV at 3% real discount rate. P&T: $30M CAPEX + $2M/yr. PRB: $50M CAPEX + $20M replacement every 15 yr.
Why Turning Off the Foam Doesn't Help
AFFF use stopped at most bases by 2000. You might expect concentrations to decline after the source is removed. We tested this: 30 years of active source, then complete removal. At 80 years, the difference in well concentration is less than 1%.
The reason: the aquifer itself has become the source. Decades of PFAS sorbed to soil particles throughout the plume body slowly re-release into groundwater, sustaining contamination long after the original foam stopped flowing. With a retardation factor of 9.5, only about 10% of the PFAS mass is in the water at any given time — the other 90% is stuck to soil, slowly desorbing over decades.
This is why walking away doesn't work. Even after AFFF stops, the aquifer itself is the source: decades of PFAS sorbed to soil slowly re-release into groundwater. With retardation factor 9.5, only ~10% of the mass is in water at any moment; the other 90% is on soil, desorbing for decades. Q3 above shows P&T started at t=0 contains the plume cleanly. This section shows the alternative — doing nothing — doesn't decay away on its own.
MODFLOW 6 GWF+GWT. Source: 100 ppb constant concentration for 30 stress periods (years), then CNC boundary removed. Compared to infinite source at 80 years. Difference: <1% in well concentration, 26% reduction in plume area (138,550 m² vs 187,100 m²).