There are three proven ways to remove PFAS from water: granular activated carbon, anion exchange resin, and reverse osmosis. All three work. They differ in which PFAS they hold, what they cost to run, and what happens to the PFAS afterward.
For most utilities and industrial sites the answer is GAC, alone or in front of ion exchange. RO is the right tool at the tap and in a few specialized plants. Here is how we reach that conclusion.
This is the comparison article in the PFAS section of our Activated Carbon for Water Treatment hub. We sell carbon, and we say so up front. The comparison below is the one we give buyers anyway, because recommending carbon where resin or RO belongs produces a failed system and a lost customer.
Option 1: Granular Activated Carbon
GAC removes PFAS by adsorption. The fluorinated tail of the molecule prefers the carbon surface to water and sticks there. Long-chain compounds, PFOS and PFOA above all, are held above 99 percent for 20,000 to 50,000 bed volumes at adequate contact time. Short-chain compounds like PFBA break through in the low thousands.
Carbon's advantages are cost, simplicity, and the end of life. It is the cheapest media per gallon treated, the equipment is a pressure vessel with no chemicals or power beyond pumping, and spent carbon can be thermally reactivated, which destroys the PFAS.
Its limit is short-chain capacity. The compound-by-compound data is in Does Activated Carbon Remove PFAS? and the system design in GAC for PFAS Removal.
Option 2: Anion Exchange Resin
PFAS carry a negatively charged head group. A PFAS-selective anion exchange resin holds positively charged sites that bind that head, while the resin's hydrophobic backbone grabs the tail. Two mechanisms at once means resin has higher capacity than carbon for every chain length, and for short-chain compounds the gap is large: PFBA bed volumes in the tens of thousands versus carbon's hundreds to low thousands.
The costs are real. Resin is several times the price of carbon per unit volume, and PFAS-selective resin is single-use: once exhausted it is incinerated, not regenerated, so the PFAS goes to a furnace either way.
Resin also needs cleaner feed water than carbon, because suspended solids and high TOC foul it, and it can release nitrate or sulfate it has exchanged away. Most PFAS resin installations have a GAC bed in front of them for that reason.
Option 3: Reverse Osmosis
RO rejects PFAS physically. The molecule is too large to pass the membrane, so removal is above 99 percent for every chain length, short and long alike, with no adsorption capacity to exhaust. RO also removes dissolved salts, nitrate, and nearly everything else.
That completeness is the cost. RO needs high pressure and therefore energy, wastes 15 to 25 percent of feed water as concentrate, and that concentrate contains all the PFAS at four to five times inlet concentration with nowhere to go.
Disposal of PFAS-laden concentrate is an unsolved problem at plant scale. At the kitchen tap, where the concentrate goes down the drain and volumes are tiny, RO is an excellent PFAS answer. For a municipal plant, it is rarely the first choice.
"Every PFAS technology moves the molecule somewhere. Only carbon moves it somewhere it can be destroyed on the way back."
Carbeva Technical Notes
| GAC virgin | GAC reactivated | PFAS-selective resin | RO membrane |
| 1x | 0.5x | 4–8x | n/a, energy |
Carbon vs. RO vs. Ion Exchange at a Glance
The table we put in front of buyers before any product discussion.
| Parameter | Granular activated carbon | Anion exchange resin | Reverse osmosis |
|---|---|---|---|
| Mechanism | Adsorption | Charge binding plus adsorption | Membrane rejection |
| Long-chain PFAS | Excellent | Excellent | Excellent |
| Short-chain PFAS | Limited | Very good | Excellent |
| Removes TDS, nitrate | No | Some nitrate, sulfate | Yes |
| Media cost | Lowest | High | Moderate, plus energy |
| Energy | Pumping only | Pumping only | High pressure |
| Water loss | Backwash only | Minimal | 15–25% concentrate |
| Feed water sensitivity | Tolerant | Needs low TOC and solids | Needs extensive pretreatment |
| Spent media fate | Thermal reactivation, PFAS destroyed | Incineration, single use | Concentrate with no destruction |
| Best fit | Municipal, industrial, lead position | Short-chain polish behind GAC | Point of use, brackish water |
How to Choose
Start from the PFAS profile, not the technology.
If the water is dominated by long-chain compounds, PFOS, PFOA, PFNA, PFHxS, GAC alone at 20 minutes total contact in lead-lag configuration is the complete answer and the lowest-cost one. Most groundwater contaminated by firefighting foam or legacy manufacturing falls here.
If short-chain compounds or GenX are present at levels that matter against your limit, the proven arrangement is GAC in lead taking the long-chain load and protecting the resin, followed by a PFAS-selective anion exchange bed for the short-chain fraction. The carbon extends resin life several times over and keeps the expensive media doing only the job carbon cannot.
If the water also needs TDS, nitrate, or hardness reduction, or the application is a single tap, RO does all of it in one step and the concentrate problem is manageable at that scale. Put a carbon prefilter ahead of it regardless: RO membranes need chlorine removed, and carbon does that while taking the PFAS load off the membrane.
Common Mistakes When Choosing a PFAS Technology
- Choosing on removal percentage alone. All three hit 99 percent on day one. Capacity, cost per gallon, and spent media are the real differences.
- Putting resin on raw water. TOC and solids foul it within months. GAC belongs in front.
- Specifying plant-scale RO without a concentrate plan. The PFAS has to go somewhere, and that somewhere is regulated.
- Treating total PFAS as the design number. Chain distribution decides whether carbon alone is enough.
- Skipping the pilot. Every vendor figure is from someone else's water.
Working Through Your PFAS Options With Us
Send us the PFAS results by compound, flow, TOC, and any other treatment targets. We will tell you whether carbon alone is enough, where a resin polish belongs, and when RO is the better fit. Where carbon is the answer, we will supply the grade with a pilot quantity and its batch certificate. Where it is not, we will say so.
For the complete water treatment map, return to the pillar guide.
Frequently Asked Questions
What is the best way to remove PFAS from water?
For municipal and industrial supplies, granular activated carbon in lead-lag configuration, with a PFAS-selective ion exchange polish added when short-chain compounds are significant. For a single tap, reverse osmosis with a carbon prefilter is the simplest complete answer.
Is reverse osmosis or activated carbon better for PFAS?
RO removes all PFAS chain lengths above 99 percent but wastes 15 to 25 percent of water as PFAS-concentrated reject and needs high pressure. Carbon is cheaper, simpler, and its spent media can be reactivated to destroy PFAS, but it is limited on short-chain compounds. RO suits point of use; carbon suits plant scale.
Does ion exchange remove PFAS better than carbon?
Per unit volume, yes, especially for short-chain PFAS. But PFAS-selective resin costs several times more, is single-use, and fouls on organics and solids, so it is almost always installed behind a GAC bed rather than instead of one.
Can boiling or a standard pitcher filter remove PFAS?
Boiling does not remove PFAS and concentrates it slightly. Pitcher filters vary: those certified to NSF/ANSI 53 or 58 for PFOS and PFOA reduce them, but most basic taste-and-odor pitchers are not designed or certified for PFAS. Check for the specific certification.