When to Regenerate Activated Carbon and When to Replace It

The rule that decides whether you regenerate a carbon bed or change it out is not economic. It is mechanistic. If the contaminant was physically adsorbed, the loading is reversible and the carbon is a reusable medium. If it was chemically converted, the carbon is a consumed reagent and no furnace brings it back.

Here is that rule with a number attached. Ammonasorb II, an acid-impregnated ammonia carbon, delivered 19.650 mg NH₃/g when new in NASA Johnson Space Center testing at roughly 20 ppm ammonia. After one regeneration cycle it recovered about 2.4 mg/g — near 12 %. The impregnant had not been fouled; it had been spent, for the reasons set out in ammonia, and why virgin carbon does almost nothing.

Nine loadings, nine mechanisms. Everything in the first two columns determines the third; nothing about the purchase price does.

The short version

• Physisorbed loadings — solvents, VOCs, chlorinated solvents — are reversible. Chemisorbed loadings are not.
• Acid-impregnated ammonia carbon: 19.650 mg NH₃/g new, about 2.4 mg/g after one regeneration cycle, roughly 12 % recovery.
• Published reactivation furnace temperatures span 600–1,010 °C. In-situ steam regeneration runs far cooler — EPA gives 250–350 °F (≈121–177 °C) at 3–4 lb of steam per lb of adsorbed VOC.
• Each reactivation cycle destroys 5–10 % of the carbon, requiring virgin makeup. Separately, EPA puts 3–5 % of adsorbed organics as chemisorbed or very strongly physisorbed and difficult to desorb.
• Spent carbon is hazardous waste if it contains a listed waste or fails a characteristic test — mercury D009 at 0.2 mg/L, benzene D018 at 0.5 mg/L. Above 260 mg/kg total mercury, EPA’s land disposal restrictions require RMERC retorting.

Is “regeneration” the same thing as “reactivation”?

Not in Chemviron’s usage, where reactivation is a furnace pass at roughly 900–950 °C in a low-oxygen environment and steam or hot-nitrogen regeneration rarely gets above 100 °C — but yes in the usage of several other reputable suppliers, who apply the two words interchangeably. It is a real terminology inconsistency, worth knowing before you read a quotation.

Chemviron draws the sharpest line. Reactivation is a thermal process in a multi-hearth furnace or rotary kiln at temperatures it puts at 900–950 °C, in a low-oxygen environment. Regeneration, in the same document, may be done with steam or hot nitrogen that “rarely gets above 100 °C”, leaving carbon that remains partially spent and can still contain some or all of the original adsorbates.

EPA’s Air Pollution Control Cost Manual describes the low-temperature route without calling it anything else: carbon is generally regenerated by heating to 250–350 °F, often with steam, at 3–4 lb of steam per lb of adsorbed VOC (EPA’s worked example uses 3.5). That is in-vessel and on-site — what a solvent-recovery adsorber does every eight to twelve hours. FEECO, a kiln OEM, stages the high-temperature route as drying, then devolatilisation at 650–1,000 °F (343–538 °C), then activation at 1,750–1,850 °F (954–1,010 °C) with steam injection.

Published reactivation temperatures do not converge:

The defensible statement is a band: 600–1,010 °C. Anyone quoting a single universal number is quoting their own furnace. And General Carbon and FEECO both use “regeneration” and “reactivation” interchangeably, while Chemviron does not. The distinction is useful — it tells you whether you are being sold a 150 °C steam cycle or a 950 °C furnace pass — but it is not an industry-wide convention. When a quotation says “regeneration”, ask for the temperature.

What each cycle actually costs you

A reactivation cycle costs you three separate things, routinely conflated: 5–10 % of the carbon destroyed outright, a 3–5 % heel of organics that will not desorb, and a capacity recovery that published sources put anywhere from about 50 % to full restoration depending on which process is meant.

Carbon loss per cycle: 5–10 %. FEECO states losses of 5–10 % are common during reactivation and that virgin carbon must be added to make up the difference; General Carbon gives the same figure and names the mechanisms — breakage, dust and gasification. This is physical destruction of the medium, which is why a reactivation contract is never a pure swap and always carries a makeup line.

The heel: 3–5 %. EPA states that approximately 3 to 5 percent of organics adsorbed on virgin activated carbon is chemisorbed or very strongly physically adsorbed and difficult to desorb during regeneration. That is not carbon you lose; it is pore volume you do not get back.

Capacity recovered: two reputable sources, one clean contradiction. EPA states regenerated carbon typically recovers about 50 % of virgin capacity. Carbotecnia, a reactivation vendor, states reactivated carbon usually recovers its full original capacity. We are not going to arbitrate that, because the two describe different processes: EPA’s 50 % sits in a chapter on in-situ steam regeneration at 121–177 °C, while Carbotecnia’s is about a furnace at 800 °C that pyrolyses the residue and reopens the pore structure. Read each number as attached to its process, not to the word.

Cycle life: not sourced. We could not find a published figure for how many reactivation cycles a carbon survives. If a supplier quotes one, ask what data it rests on.

Which loadings come back, and which do not

The mechanism column drives everything to its right.

Desotec states the physics plainly: in physisorption molecules attach weakly through intermolecular forces and the bond is reversible; in chemisorption they form stronger and often permanent bonds — the route used deliberately for ammonia and hydrogen sulfide. When you buy an impregnated carbon you are buying that second mechanism on purpose. Irreversibility is the product feature.

H₂S is the middle case. On catalytic carbon the gas is converted to elemental sulfur or sulfates — not volatile, but partly water-soluble, and at least one vendor states its catalytic carbon can be regenerated in place by washing with water. Thermally, Coppola and Papurello tested nitrogen, air and CO₂ on H₂S-spent carbon: the best result was CO₂ at 300 °C for 75 minutes, a regeneration ratio of 30 %. Nitrogen at 200 °C did worse; air at 300 °C caused unwanted partial combustion. Thirty per cent is real, but it is not a restored bed. On caustic-impregnated carbon the answer is no — the alkali has become a salt, and the circulated claim that such media regenerate ten times is unverified, as our H₂S capacity article notes.

Siloxanes are the cleanest “no”. Cyclic volatile methylsiloxanes polymerise on the carbon surface into heavier, hardly desorbable polydimethylsiloxanes, while oxidative regeneration converts them to SiO₂ instead; both products block the pores permanently, so the siloxane bed is a throwaway. The mechanism, the evidence and the proof-by-exception are set out in siloxanes in biogas.

Mercury is a hard no with a temperature to prove it. Sulfur impregnation converts Hg⁰ to mercuric sulfide, analogous to cinnabar — a stable solid with a sublimation temperature around 584 °C. General Carbon states outright that thermal reactivation is not suited to carbons loaded with heavy metals or inorganics, naming mercury and lead. SURSORB GG-Hg publishes a mercury loading of min 20 % on that basis: a capacity figure for a consumable, not a duty cycle.

Radioiodine on TEDA carbon is chemically final. The tertiary amine nitrogen attacks methyl iodide by nucleophilic SN2 substitution, forming a non-volatile quaternary ammonium iodide salt — which is why TEDA carbon holds up where physisorption and isotopic exchange fail, and why it does not come back. SURSORB NGC does not disclose its impregnant; ask for the chemistry before planning an end-of-life route.

The palladium case: regenerating the catalyst, not the bed

Palladium is not consumed by CO oxidation; it deactivates structurally, by ionic palladium transforming into metallic clusters, and Peterson and co-workers report in Nature Communications that activity is restored by oxidation at 700 °C in air. That is catalyst regeneration rather than bed regeneration, and 700 °C in air is precisely why the fielded catalyst sits on alumina rather than on an oxidisable carbon support — the chemistry and the patent record are in the impregnation matrix. Treat SURSORB PAL as a consumable.

What happens to spent carbon under US law

Spent activated carbon is a solid waste; whether it is a hazardous waste is a separate question. EPA’s RCRA guidance treats spent carbon as a spent material, unless it results from pollution control — in which case it is a sludge. It becomes hazardous when it contains a listed hazardous waste or exhibits a hazardous waste characteristic. That guidance also notes that while the regeneration facility is exempt from Part 264 standards, storage requires a permit — which is why a reactivator’s yard is permitted and your loading dock is not.

Two routes make spent carbon hazardous. The listed route: if the carbon treated a listed waste, the mixture and derived-from rules carry the listing forward. Calgon’s acceptance FAQ walks through this with the F, K, P and U codes. It is not a judgement about how contaminated the carbon looks; it is a question about what the influent was. The characteristic route: 40 CFR 261.24 Table 1 sets the TCLP thresholds:

Mercury gets its own regime on top. EPA’s land disposal restrictions split mercury wastes at 260 mg/kg total mercury: below that, numerical treatment standards apply; at or above it the waste must be treated by RMERC — the standard at 40 CFR 268.42, defined as retorting or roasting in a thermal processing unit capable of volatilising mercury and subsequently condensing it for recovery. EPA reports mercury recovery ratios of 97.7–98.4 % for that route and explicitly names spent activated carbon among the mercury-bearing wastes retorted on site. Applicable waste codes include D009, K071, K106, P065, P092 and U151.

One cost that rarely appears in a payback calculation: a US reactivator will charge you to find out whether it can take your carbon at all. Calgon publishes acceptance testing for RCRA-hazardous determination at $1,000 for vapour phase and $1,200 for liquid phase, 2–3 week turnaround.

Why reactivation is a local business — and where Suracsh sits

Suracsh Adsorbents LLC does not operate reactivation services in the United States. We do not accept spent carbon and we do not run a furnace here.

That is structural rather than a hole in the product line. Reactivation means freighting spent carbon — heavy, often wet, sometimes regulated — to a furnace and freighting it back, so the economics are set by haul distance rather than by the thermal process. We looked for a published ratio of freight cost to carbon value and for a defensible haul radius. Neither could be verified, so we are not printing one.

What is verifiable points the same way. Calgon Carbon publicly implemented a US energy surcharge citing ongoing fuel and logistics cost pressures — a thermal processor telling the market its cost base moves with fuel and freight. Service providers describe the offering in logistics terms: transporting, exchanging and reinstalling GAC on a schedule. What is sold is a route as much as a furnace.

So if your duty is genuinely regenerable — solvent recovery, VOC control, chlorinated solvents — use a US reactivator with a furnace near your site. That is a better answer than anything we can offer. Suracsh’s advantage is at the other end: supplying virgin media with the specification published before you buy. Where the duty consumes the carbon, there is no service to buy from anyone and the only question left is disposal.

The decision checklist

1. What is on the carbon, and by what mechanism? If an impregnant did the work, assume it is consumed until someone shows regeneration data on that chemistry.
2. Is it a listed or characteristic waste? Check what the influent was — the mixture and derived-from rules travel with it — then TCLP against 40 CFR 261.24, and the 260 mg/kg threshold separately for mercury.
3. Can the site accept an exchange service? Vessel design, lifting access, permitted storage for the spent bed, and whether your operating permit allows it.
4. How far is the nearest furnace? Get round-trip freight quoted before the reactivation price.
5. Is the carbon a reusable medium or a consumable reagent here? If it is a reagent, the regeneration conversation is moot and the real variables are delivered cost per kg, working capacity and changeout labour.

Every grade named here publishes its full specification in HTML at suracsh.us, and every datasheet downloads without a form or an NDA. The selection guide maps contaminant to grade to specification, and the mechanism table above is the part to work through before a reactivation contract is priced.

Sources

• Chemviron — Reactivation of PFAS-loaded carbon
• US EPA — Air Pollution Control Cost Manual, Chapter 1: Carbon Adsorbers (7th ed.)
• FEECO International — Spent carbon reactivation
• General Carbon — Reactivated carbon
• Carbotecnia — Exhausted activated carbon reactivation
• Desotec — Activated carbon adsorption for the removal of industrial odours
• Pelton Environmental — Carbon reactivation and the lifecycle of GAC
• NASA NTRS 20150008167 — Ammonia sorbent evaluation
• Coppola & Papurello, Clean Technologies 1(1):40–57 (2019) — Activated carbon regeneration for H₂S removal
• General Carbon — Guide to hydrogen sulfide removal
• Gaj, Energies 13(10):2605 — Adsorptive biogas purification from siloxanes, a critical review
• Cabrera-Codony et al., Environ. Sci. Technol. — Optimal activated carbon properties for siloxane removal
• Idaho National Laboratory — Mercury adsorption on sulfur-impregnated carbon
• Peterson et al., Nature Communications 5:4885 (2014) — Low-temperature CO oxidation on atomically dispersed Pd
• Nucleophilic mechanism of TEDA-impregnated carbon for radioiodine
• US EPA RCRA Online RO 11143 (2 April 1986) — Regulatory status of spent activated carbon
• Calgon Carbon — Carbon acceptance frequently asked questions
• 40 CFR 261.24 — Toxicity characteristic (TCLP) Table 1
• US EPA — Waste specific evaluation of the RMERC treatment standard
• Calgon Carbon — US energy surcharge announcement (surcharge effective 6 April 2026)

Questions