In March 2026, the EPA, the Department of Justice, and Washington State announced a $668 million settlement with Boeing, the City of Seattle, King County, and more than 100 other responsible parties to clean up 177 acres of the Lower Duwamish Waterway. Dredging and capping will run for at least a decade. Forty-one hazardous substances — PCBs, arsenic, polycyclic aromatic hydrocarbons, dioxins, furans — are buried in those sediments. The settlement signals a milestone, but it does not signal the end of the project. It signals the start of a thirty-year monitoring liability that, today, nobody has the data to defend.
Two months later, on May 4, the EPA announced Five-Year Reviews at four heartland Superfund sites — a routine reminder that the federal government continues to ask the same question of every capped or dredged site in the country: is the remedy still protective? The answer is supposed to be backed by data. On most contaminated-sediment sites, the data is a handful of cores collected once a year.
That mismatch — a thirty-year liability supported by an annual snapshot — is the next frontier of sediment-quality risk management. It is also where the financial exposure quietly migrates after the dredge barges leave the channel.
The Problem: Post-Remediation Monitoring Is Where Liability Hides
Most of the industry's attention on sediment-quality data focuses on the front end of a project: the remedial investigation, the feasibility study, the design memorandum. That is the stage where consultants get scrutinized and where the dollar figures get nailed down. But the financial life of a $668 million sediment remedy does not end at the ribbon-cutting. It continues through:
- EPA Five-Year Reviews requiring documentation that the remedy remains protective.
- Long-term operation, maintenance, and monitoring (OM&M) obligations written into consent decrees.
- Adaptive management triggers that require additional work when contaminant concentrations exceed thresholds.
- Recontamination liability from external sources — combined sewer overflows, upland runoff, atmospheric deposition — that the original cleanup did not eliminate.
At the Duwamish/Diagonal Combined Sewer Overflow site in Washington State, recontamination attributable to ongoing CSO discharges has already been documented on a previously remediated sediment cap. At other Superfund sites, monitoring has flagged scour from high-velocity flows and ice-jam events capable of redistributing PCBs once every decade. None of those problems were detectable from annual sediment cores. By the time they showed up in the data, they had been quietly accruing for years.
The technical failures driving this liability are not the ones the industry usually talks about. They are different failures, in a different phase of the project, and they require a different conversation.
1. The Cap Performance Blind Spot
A sediment cap is engineered to do one thing: keep buried contaminants from re-entering the water column at concentrations that exceed remedial goals. To verify that, you need to know what is actually crossing the sediment-water interface — continuously, at the concentrations and time scales at which it matters.
What the industry uses instead is surface sediment chemistry sampled once a year. That measurement is a state variable. It tells you what accumulated on top of the cap between visits. It does not tell you what rate contaminants are crossing the interface, when that rate spikes, what drives the spikes, or whether the cap itself is the source.
For a manager defending a Five-Year Review, the gap is enormous. “Concentrations on the cap surface were 12% higher than last year” is not a defensible answer to “is the remedy still performing.” It is a starting point for an argument — and arguments at the Five-Year stage are how operation-and-maintenance budgets blow through their reserves and how consent decrees get reopened.
2. The Recontamination Attribution Problem
When cap-surface concentrations do trend upward, the manager faces a question that conventional monitoring cannot answer: where did the new contamination come from? The candidates are usually some combination of:
- Migration of contaminants upward through the cap from the underlying sediment (cap failure).
- Fresh deposition from external sources — stormwater, CSOs, upland runoff (continuing source loading).
- Water-column processes — particle settling, microbial transformations — that mimic a flux signal without the sediment doing anything.
These three pathways carry profoundly different financial consequences. Cap failure triggers re-design and re-construction. External source loading shifts cost to upstream permittees. Water-column artifacts trigger nothing — but only if you can prove that’s what they are. Without simultaneous, defensible measurement of sediment-only flux and ambient water-column processes, the responsible party often ends up paying for all three possibilities because they cannot rule any of them out. This is the same attribution problem that plagues pre-design nutrient budgets, but at the post-remediation stage, the dollar exposure is concentrated on a much shorter list of parties.
From Snapshot Audits to Continuous Custody of the Remedy
The fix is not a new model. It is a different relationship to data — what we call continuous custody of the remedy. Instead of treating the sediment cap as a static engineered feature that gets inspected annually, the manager treats it as a dynamic interface whose performance is documented continuously, the same way a process plant documents emissions or a wastewater operator documents effluent.
This re-framing matters because Five-Year Reviews, OM&M plans, and adaptive-management decisions are all evidentiary processes. They reward the party that brings defensible data and penalize the party that brings interpretation. Continuous sediment intelligence flips which side of that line the responsible party stands on. Sediment Technologies’ two-tier platform exists to make that shift operational.
Tier 1: NutrientOS — The Decision Engine for the Long Tail
Most environmental consulting teams supporting Superfund OM&M phases inherit a tangled archive: design memos, monitoring well data, surface sediment results, modeling parameters, regulatory letters. NutrientOS ingests that archive — watershed inputs, historical water-quality records, regulatory thresholds, literature values — and builds a unified picture of what the data does and does not support.
For a long-tail monitoring program, the highest-value output is the gap analysis. NutrientOS surfaces the specific quantitative claims a Five-Year Review will require, ranks the financial exposure of each unsupported claim, and tells the team where additional measurement will materially reduce risk. It is the diagnostic step that decides whether continuous monitoring is a luxury or a budget-saving necessity. On a $668 million cleanup, that question deserves more than a literature review.
Tier 2: CAROSEL — The Continuous Custody Instrument
When NutrientOS identifies a defensibility gap — typically around cap performance, recontamination attribution, or compliance-threshold proximity — CAROSEL closes it. The Chamber Array for Observing Sediments Long-term, developed at Florida Atlantic University’s Harbor Branch Oceanographic Institute, is the first commercially available platform for long-duration, autonomous benthic flux monitoring.
For Superfund and contaminated-sediment applications specifically, three CAROSEL capabilities convert the post-remediation monitoring problem into a tractable engineering exercise:
- Sensor-agnostic flux measurement across the analytes that drive Superfund decisions — heavy metals, organic contaminants, ammonium, phosphate, dissolved oxygen — on the same platform.
- True attribution via a sealed-plate reference cell that measures ambient water-column processes simultaneously with the chamber, separating sediment-only flux from water-column noise and from external deposition signals.
- Continuous deployment that captures the episodic events — storm flows, CSO discharges, scour events — that drive most recontamination and that annual coring will never see.
Together, NutrientOS and CAROSEL give the responsible party something the conventional monitoring program cannot: a defensible, time-resolved record of remedy performance that survives a Five-Year Review without requiring the consent decree to be reopened.
The era of treating the post-remediation phase as a custodial afterthought is ending. The next $668 million settlement is already being negotiated somewhere. The data that defends it should already be in the field.
It is time to put sediment intelligence on the long-tail side of the project, not just the design side.
“A sediment cap without continuous flux monitoring is a $668 million bet that nothing changes for thirty years.”
