Power plants drive free chlorine to near-zero to protect metals, ion‑exchange resins, and membranes. Two methods dominate—granular activated carbon (GAC) and sodium bisulfite—and each brings different kinetics, byproducts, footprint, and biofouling risk.
In heat‑recovery steam generator (HRSG) service, free chlorine is the wrong oxidant in the wrong place. It attacks ion‑exchange resins and stainless steels, so operators target non‑detectable residuals—about 0.0–0.1 mg/L—downstream of pretreatment (dechlorinationforro.blogspot.com) (www.ncbi.nlm.nih.gov). That protects downstream ion‑exchange resin units and, where used, Reverse Osmosis (RO) systems from oxidant damage.
Two approaches do the heavy lifting: chemical dechlorination with sodium bisulfite (SBS, NaHSO₃; often fed as sodium metabisulfite, SMBS, Na₂S₂O₅) and granular activated carbon filtration (GAC). Other techniques—aeration, UV, chloramination—are rarely used in large power‑plant raw water.
Chemical dechlorination kinetics and dosing
SBS reacts rapidly and nearly stoichiometrically with chlorine (1 mol Cl₂ + 1 mol bisulfite → products). In theory, 1.34 mg of Na₂S₂O₅ removes 1.00 mg of free Cl₂, but in practice engineers feed about 2–3 mg per mg Cl₂ to ensure complete quench (www.researchgate.net). Neutralizing 0.5–1.0 mg/L typically takes ~1.5–3 mg/L SBS, mixed in‑line; the reaction completes in seconds, so no large contact basin is needed.
Performance is decisive: properly dosed SBS removes >99% of free chlorine—often to <0.01 mg/L—and does so substantially faster than other chemicals in comparative testing (www.researchgate.net). The main byproduct is sulfate (Na₂SO₄), which raises total dissolved solids (TDS) but is inert; for example, a 3 mg/L SBS dose generates roughly 3–5 mg/L Na₂SO₄.
Capital needs are minimal—a skid with an accurate dosing pump—and control is straightforward via ORP (oxidation‑reduction potential) or residual monitors. A 2013 RO pretreatment guide even calls SBS the “chlorine reducing agent of choice” for large systems (docslib.org). Still, over‑dosing or poor mixing can leave residual reducing agent that fosters microbial fouling downstream, and SBS adds dissolved solids that may marginally increase boiler blowdown needs. SBS solutions (pH ~5) require proper storage and handling; dosing is a respiratory irritant hazard and can form SO₂ if misdosed.
For procurement and operations, dechlorination chemicals are commonly supplied as a dechlorination agent, fed continuously in proportion to chlorine load.
GAC filtration mechanism and capacity
GAC (granular activated carbon) removes chlorine by catalytic reduction on the carbon surface, not by simple adsorption. It also pulls down organics, total organic carbon (TOC), pesticides, and trihalomethane (THM) precursors, delivering dual benefits (www.ncbi.nlm.nih.gov) (docslib.org). Fresh beds are nearly 100% effective initially.
But chlorine “breakthrough” is a matter of time, dose, and velocity. With 5 mg/L Cl₂ feed running at 4 gpm/ft² (≈9 m/h), one study saw ~10 days to reach 0.5 mg/L at the outlet (www.ncbi.nlm.nih.gov). At longer contact (~15 minutes), ~25 days passed before 0.5 mg/L appeared in the first 10% of the bed (www.ncbi.nlm.nih.gov). Another trial found 50% breakthrough (~0.25 mg/L) in 25 days at a faster 0.3 minute contact (www.ncbi.nlm.nih.gov). In short, under heavy Cl₂ loading, breakthrough occurs on a days‑to‑weeks timescale (www.ncbi.nlm.nih.gov) (www.ncbi.nlm.nih.gov).
Longer term, GAC capacity is finite: studies suggest ~1–4 grams of Cl₂ can be removed per gram of carbon. With 1–2 mg/L in feed, a bed often exhausts over ≈2–3 years and will be replaced or regenerated (www.ncbi.nlm.nih.gov). In practice, filters are sized for long runs, with weekly backwashes in summer and monthly in winter to manage headloss (ncbi.nlm.nih.gov). At about 1–2 mg/L Cl₂ in raw water, a GAC bed often lasts >1 year on service (www.ncbi.nlm.nih.gov), often delivering 20–50% TOC reduction as an extra benefit.
GAC requires capital: a large pressure vessel or multi‑tank array with thousands of kilograms of media. It is common for small systems (<100 gpm), while large flows need multiple vessels. Plants typically specify activated carbon media in robust housings; for high‑pressure service, steel filter housings are standard.
Maintenance to control biofouling
There’s a trade‑off: once chlorine is removed on carbon, the high surface area and adsorbed organics enable microbial growth. A treatment bulletin warns carbon filters are “notorious for breeding bacteria,” with viable counts doubling in 24 hours unless sanitized (docslib.org). In beverage applications, GAC “was much maligned because of its inability to prevent bacterial growths,” even though chlorine removal per se was effective (www.ncbi.nlm.nih.gov).
Operators respond with frequent backwashing and periodic chemical or thermal sanitization, media replacement, or reactivation. Drinking‑water plants backwash GAC weekly in summer and monthly in winter (ncbi.nlm.nih.gov). Without such maintenance, beds quickly develop high bacteria counts that foul downstream systems. Many HRSG facilities schedule downtime months in advance or lean more heavily on SBS to reduce reliance on a single GAC unit.
Comparative performance and trade‑offs
Removal efficiency: both methods can push free chlorine to near zero. SBS can quench >99.9% immediately—often to <0.01 mg/L—when the dose is right (www.researchgate.net). GAC initially removes essentially all chlorine but will break through as its finite capacity is consumed (www.ncbi.nlm.nih.gov) (www.ncbi.nlm.nih.gov).
Speed and control: SBS acts in seconds and allows tight cascade control with ORP or residual meters. A GAC bed needs designed contact time (seconds to minutes) and careful monitoring for breakthrough; there is no online sensor that “doses a bed,” so operators manage bed life and backwash schedules.
Footprint and cost: chemical dechlorination carries the smallest capital footprint—essentially a mixing loop and pump—while GAC requires large vessels and media. Operating cost flips that: SBS consumes reagent (roughly tens of kilograms per Mgal (≈3.8 m³) at 1 ppm Cl₂), whereas GAC uses little chemical but demands media replacement or reactivation—thousands of kilograms per year at scale.
Water quality impacts: GAC improves overall water quality by removing organics and particulates—often a 20–50% TOC reduction (docslib.org)—and it converts chlorine to chloride without changing TDS. SBS adds sulfate; for every mole of Cl₂ removed, a mole of SO₄²⁻ is produced, raising conductivity and potentially increasing blowdown modestly, though sulfate is typically benign at the low mg/L levels used in HRSG makeup.
Safety and byproducts: SBS introduces a handling hazard (respiratory irritant; potential SO₂ formation); its byproducts are sulfite/sulfate. GAC’s “byproducts” are adsorbed organics on the media, often biodegraded in situ. Operational reliability differs too: a well‑maintained SBS system rarely fails unless pumping or mixing fails—alarmable events—whereas GAC can channel or break through if not maintained; large plants often install tandem vessels to keep one online while another is serviced.
Bottom line: many designs prefer SBS where rapid, small‑footprint dechlorination is required, while GAC is chosen when added organics removal is valuable or for redundancy. An RO pretreatment Q&A even argues there’s “no need for GAC” provided mixing is good (“Forget the GAC”) (thewaternetwork.com). Others run dual trains (SBS first, then GAC polishing) for safety.
Quantitative comparison snapshot
Engineers often remember two figures: SBS eliminates ~100% of Cl₂ in <10 seconds using ~3× stoichiometric dose (www.researchgate.net), while a GAC bed under moderate flow may run ~25 days before any Cl₂ appears, after which it must be reactivated or replaced (www.ncbi.nlm.nih.gov) (www.ncbi.nlm.nih.gov).
Where RO membranes are part of the train, dechlorination protects the polyamide layer as much as it shields ion exchange. Plants that integrate brackish‑water RO often place SBS dosing ahead of membranes and GAC downstream for polishing, keeping both organics and oxidants in check.
Sources: Authoritative water treatment literature and industry reports. Modeled GAC breakthrough data and the “brewing industry” note come from a U.S. National Research Council review (www.ncbi.nlm.nih.gov) (www.ncbi.nlm.nih.gov). Stoichiometric details of sodium bisulfite dosing are documented in water supply studies (www.researchgate.net). Operational practices for GAC (backwash schedules, biogrowth) are noted in recent treatment reviews (ncbi.nlm.nih.gov) (docslib.org).
Specifying equipment and media is application‑dependent, but the two anchors rarely change: dose the oxidant away with a controlled chemical feed and SBS‑type dechlorination agent, or design and maintain a stable bed of activated carbon in an appropriate pressure housing—and monitor for breakthrough to protect both resins and membranes.
