Granular activated carbon (GAC) is the most field-validated sorption media for PFAS removal in drinking water, capable of 100% removal of long-chain PFOA and PFOS during early bed life, with effectiveness governed by carbon type, bed depth, and flow rate [S1][S3].
US EPA's first enforceable national drinking-water standard sets 4 ng/L (parts per trillion) limits for PFOA and PFOS, a regulation projected to protect roughly 100 million people and now mandating GAC, ion exchange, or high-pressure membrane systems at affected utilities [S2].
What activated carbon does in PFAS service
GAC removes PFAS by adsorption onto a highly porous carbon matrix derived from bituminous coal, lignite, coconut shell, or wood, materials selected for surface area and pore-size distribution that match the target molecule [S1]. Longer-chain perfluorinated compounds such as PFOA (C8) and PFOS (C8) adsorb efficiently; shorter-chain analogues like PFBS (C4) and PFBA (C4) slip through the same bed unless carbon selection and contact time are tuned specifically for them [S1]. Powdered activated carbon (PAC), the same base material milled finer, is dosed into the clarification stage rather than used in a flow-through bed; EPA researchers note PAC achieves only modest percent removals at high doses and creates a PFAS-laden sludge handling problem that utilities must close-loop [S1].
2026 supply shift: domestic capacity and pricing exposure
The defining 2026 supply development is a wave of new US domestic production modules designed to cut reliance on overseas bituminous and coconut-shell carbon. BioEnergy Development Inc. has deployed a 3,000 tons/year production module in Montana using US wood waste as feedstock, with a four-unit system projected to scale output to 12,000 tons annually [S2]. A separate Water Research Foundation project is studying PAC for PFAS removal from wastewater effluent, with $185,000 in funding and a 2028 completion target, signaling that PAC-grade material demand will be formally benchmarked within this decade [S6].
From a sourcing perspective, this matters because activated carbon pricing tracks both raw feedstock (coal, coconut shell, wood) and the energy intensity of activation. Where reactivation services are used, a spent GAC stream can be thermally reactivated to destroy more than 99.99% of adsorbed PFAS, allowing the carbon to be redeployed in service and reducing virgin-material demand over a multi-cycle operating life [S5].
Comparison of the main PFAS treatment options
Activated carbon sits in a sorption-based treatment family alongside ion exchange (IX) resins and high-pressure membranes (nanofiltration, reverse osmosis); all three are designated Best Available Technology under current US PFAS frameworks, but they trade off on different operating axes [S2][S4]. GAC is typically the lowest-cost option at scale for long-chain PFAS in surface water, IX resins achieve high removal across a broader PFAS chain-length range including short chains, and nanofiltration/reverse osmosis provide a physical barrier but generate a concentrated PFAS reject stream that must be handled downstream [S4][S8]. A 2025 life-cycle review reported activated carbon PFAS treatment emissions in the range of 1 to 80 kg CO2-equivalent per gram of PFAS removed, reflecting wide variation driven by carbon source, regeneration energy, and target compound [S7]. For utilities trialing options in late 2026, the decision is rarely a single-technology buy: most full-scale designs pair GAC lead-lag beds with polishing IX or membrane stages to cover short-chain residuals [S4][S8].
Selection criteria engineers should lock down first
Three specifications drive the GAC selection call. First, the target PFAS chain-length profile: influent dominated by PFOA/PFOS can run on standard bituminous GAC; short-chain PFBS/PFBA loads require finer-tuned pore structure or blending with IX [S1]. Second, empty bed contact time (EBCT), which US EPA guidance links directly to breakthrough behavior, and which in practice is sized in the 10 to 20 minute range for PFAS service. Third, the spent-carbon disposition pathway, because thermal reactivation with verified 99.99% PFAS destruction is the most established closed-loop route, while landfill or incineration without reactivation locks the operator into recurring virgin-carbon purchase and disposal cost [S5].
Where GAC is the wrong primary tool
GAC alone is a poor fit for very short-chain PFAS-dominated streams, for high-total-dissolved-solids brines where competitive adsorption knocks out capacity, and for applications where the downstream sludge or spent-carbon handling chain cannot guarantee destruction of the captured PFAS [S1][S4]. In those cases, the ITRC PFAS treatment technology guidance recommends pairing GAC with IX, nanofiltration, or reverse osmosis, or routing the concentrate to a high-temperature thermal destruction unit rather than a regenerable sorption loop [S4].
Standards, reactivation, and circular-economics reality
Designers specifying reactivation services should treat the 99.99% PFAS destruction figure as a process-performance claim tied to furnace residence time and temperature, not a default for any reactivation run, and should require lot-traceable destruction verification from the reactivation vendor [S5]. Norit's GAC, PAC, and reactivation portfolio and Calgon Carbon's reactivation-enabled product lines are both positioning for the same utility demand wave that EPA's 4 ng/L limit and Europe's 500 ng/L total-PFAS cap are creating [S2]. Operators in the 2026 procurement cycle should be tracking two signals: the actual commissioning of the Montana 12,000-tons/year scale-up, and any tightening of European compound-specific PFAS thresholds that would push demand toward IX-augmented or membrane-polished GAC trains rather than GAC alone [S2].
The underlying component specifications are covered under ballast water treatment, carbon fiber, and carbon steel.
Background reading: Alumina 85% vs 99.5%: Picking the Right Purity Grade for the Job.