Engineering analysis

Destroying PFAS in Saline RO Reject: Why Whole-Brine SCWO and BDD Fail and Concentrate-Then-HALT Wins

None of whole-brine SCWO, raw-brine boron-doped diamond (BDD), or foam fractionation alone hits PFAS destruction targets on saline RO reject — each is disqualified by the salt matrix. The viable architecture is concentrate-then-destroy: use foam plus PFAS-selective ion exchange (with a hardness-exclusion front-end) to pull ~7.6 kg/day of PFAS out of ~53 t/day of salt, then feed the small concentrate to a subcritical Hydrothermal Alkaline Treatment (HALT) reactor that tolerates dissolved NaCl and fixes fluorine as soluble NaF. Before spending capital, retire the unknowns with ~$150K of testing over one quarter: confirm the bulk-brine disposal path, get a merchant off-site destruction quote for the ~2.8 t/yr of PFAS, and run a HALT treatability test to check that Ca/sulfate hardness does not scale the reactor.

Key takeaways

  • Whole-brine SCWO fails physically: crossing the critical point drops NaCl solubility ~1000×, leaving a 3.5 wt% feed 100–350× supersaturated, and liberated fluoride forms HF that corrodes the vessel at ~600 °C.
  • Raw-brine BDD fails chemically: 15,000 mg/L chloride becomes a chlorate (~6.3 mol/m²/day) and perchlorate factory against 2–6 µg/L limits, and it cannot defluorinate short-chain/GenX PFAS.
  • The real fork is salt management: tolerate the NaCl (subcritical HALT at ~350 °C, ~16.5 MPa keeps dielectric ~15–20 so NaCl stays dissolved >300 g/L), but exclude the inverse-solubility Ca/sulfate hardness that can still scale the reactor.
  • The governing figure of merit is $/kg PFAS destroyed, not $/1,000 gal — a whole-brine reactor spends ≥99.9% of its energy and oxidant on salt and ~0.02% on the ~7.6 kg/day of PFAS.
  • Concentration is the master lever: it moves the matrix:target ratio from ~30,000:1 molar (whole brine) toward <100:1, a 3+ order-of-magnitude gain no reactor optimization can match.
  • Total de-risking spend is ~$150K over one quarter to protect a $10–30M capital decision, with a 4-month gate: proceed to HALT only if concentration factor ≥50× and hardness does not scale the reactor.

PFAS destruction architectures for saline RO reject: performance is nearly tied, salt tolerance is the divider

CriterionA1: Separate then SCWOA2: Separate then BDDB: Concentrate then HALT
Long + short-chain >99% destructionYes (non-selective)Long yes; short-chain riskYes (non-selective)
Perchlorate/oxychlorine riskNoneOnly if chloride fully removed firstNone
Salt-management burdenHigh — needs low-salt and low-volume feedHighest — must also strip chlorideLow — tolerates residual NaCl
Closest real data pointPeterson SFB SCWO (low-salt)Foam + electro-oxidation pilotAquagga/3M HALT on IX regen brine
Reactor materialsSiC/TaC/YF3 (exotic)ElectrodesNi-alloy if caustic-SCC survives
Modeled CAPEX~$15–30MHigher than A1 and B~$8–15M
Frontier positionViable hedgeDominatedPrimary candidate

Frequently asked

Why can't supercritical water oxidation (SCWO) treat whole saline brine directly?

Above water's critical point the dielectric constant collapses (~80 to 2–6), dropping NaCl solubility ~1000× so a 3.5 wt% feed becomes 100–350× supersaturated and plugs the reactor. Liberated fluoride also forms corrosive HF at ~600 °C, and over 99.9% of the energy is spent destroying salt rather than PFAS.

What is HALT and why does it tolerate salt?

Hydrothermal Alkaline Treatment (HALT) is a subcritical process running at roughly 350 °C and 16.5 MPa with excess NaOH. Because it stays below the critical point, water's dielectric constant remains ~15–20, so NaCl stays fully dissolved above 300 g/L and fluorine exits as soluble NaF instead of forming corrosive HF.

Is chloride or hardness the bigger problem for a subcritical PFAS reactor?

Once you drop the supercritical premise, dissolved NaCl stops mattering to the reactor. The species to fear is the Ca/sulfate hardness, the inverse-solubility component that can still scale even a subcritical reactor — which is why a hardness-exclusion front-end and a treatability test are the decisive pivot.

Should PFAS be destroyed on-site or shipped off-site?

The PFAS mass is only ~2.8 t/yr, so a merchant off-site destruction or regeneration quote in $/kg-PFAS may beat any on-site build. Get that quote before committing capital; a one-week phone call can obsolete the entire on-site train.

Does the reactor need to meet the sub-4 ppt PFAS limit?

No. The <4 ppt target is a train-level product-water spec met by RO plus a GAC/IX guard bed, not a reactor spec. The reactor's job is >99% mass destruction of the concentrate; chasing the last two 9s inside the reactor over-sizes it by orders of magnitude.

Why does raw-brine BDD create a new regulatory problem?

The 15,000 mg/L chloride in RO reject is oxidized into chlorate (~6.3 mol/m²/day) and accumulating perchlorate, both regulated against 2–6 µg/L limits, which BDD cannot self-remediate. It also cannot reliably defluorinate short-chain and ether PFAS such as PFBA, PFBS, and GenX.

This answer is distilled from a full Sparlo analysis — the root-cause trade-off, cross-domain paths, and vendor-ready options.

Read the full analysis →
Analyzed with Sparlo — an AI engine that stress-tests hard engineering decisions against their constraints.

This site uses cookies to improve your experience.