Inside the DBP Playbook: How Utilities Cut THMs by Optimizing Precursor Removal and Disinfection

Regulators have long tied DBP spikes to the collision of chlorine with natural organic matter (NOM, the organic fraction in raw water). The U.S. EPA’s Stage 1 Disinfectants and Disinfection Byproducts Rule (DBPR) even mandated up to 35–50% total organic carbon (TOC) removal via enhanced coagulation (for example, 35% removal for low‑alkalinity water with 2–4 mg/L TOC; 50% removal for 0–60 mg/L alkalinity and >8 mg/L TOC) (nepis.epa.gov).

That mandate codified what operators already knew: remove precursors up front, then pick the right disinfectant and point of application to cap DBP formation potential (THMFP). Bench and pilot jar tests routinely show the gains when systems dial in coagulation and filtration before disinfection (iwaponline.com).

Enhanced coagulation: pH and dose

Enhanced coagulation—deliberately depressing pH and optimizing metal salt dose—targets the NOM fractions most likely to form trihalomethanes (THMs) and haloacetic acids (HAAs). In surface waters high in humic NOM, jar tests under enhanced conditions commonly achieve 40–60% dissolved organic carbon (DOC) removal via coagulation/flocculation (versus roughly 20% without optimization) (iwaponline.com). EPA’s Stage 1 DBPR set explicit TOC removal targets (35–50%) as noted above (nepis.epa.gov).

Mechanistically, lowering coagulation pH to ~5.5–6.5 and increasing alum dose better destabilize aromatic, hydrophobic NOM—the prime THM precursors (iwaponline.com, iwaponline.com). In a Thailand reservoir study, adding a cationic polymer and 10 mg/L powdered activated carbon (PAC) to polyaluminum coagulation boosted DOC removal to ~65–75% and cut THMFP by ~77–79% (iwaponline.com).

In practice, operators guide optimization via SUVA (specific UV absorbance, a NOM character indicator) and jar tests to reach the TOC‑pH “knee” used in DBP guidelines (nepis.epa.gov). After floc formation, the solids load is removed by sedimentation, often in a clarifier, before filtration and disinfection.

Activated carbon filtration performance

Granular activated carbon (GAC) filters—packed beds of adsorptive carbon media—strip residual organics after sedimentation/filtration. Full‑scale systems typically adsorb ~30–40% of influent DOC (iwaponline.com) using media such as activated carbon. Kim & Kang (2008) reported ~33.7% DOC removal by GAC (in “sand‑filter replacement” mode) during the first 6 months of operation (iwaponline.com).

The DBP impact is direct. Little Falls WTP (NJ) installed GAC in one treatment train and saw distribution‑system THMs drop to ~46 μg/L (nepis.epa.gov). GAC beds do require periodic regeneration or replacement: after ~6–12 months of service, adsorption sites break through and efficiency declines (iwaponline.com).

Biologically active carbon (BAC)—GAC seeded with biofilms—can further degrade NOM, achieving 20–40% DOC removal (iwaponline.com). In a Spanish pilot, BAC removed >25% DOC and cut THMFP by ~30% more than GAC alone (iwaponline.com). Powdered activated carbon additions upstream of filtration (e.g., 10–20 mg/L) also improve hydrophilic DOC removal and lower THMFP when paired with coagulation (iwaponline.com).

Advanced precursor removal options

Beyond conventional trains, advanced processes can further strip DBP precursors. Membrane filtration—such as ultrafiltration or nanofiltration—can remove virtually all organics but at high cost. Ion‑exchange resins (e.g., MIEX®), available as ion‑exchange resin systems, target DOC and have lowered THM precursors by ~30–50% in pilot studies (iwaponline.com).

Pre‑oxidation with ozone or permanganate can break down high‑molecular‑weight NOM, making coagulation more effective; studies report combined ozone+coagulation achieving 10–70% DOC removal (iwaponline.com). All advanced steps must be weighed against cost/complexity.

Disinfectant choice and DBP formation

Switching disinfectants can drastically cut regulated DBPs. Chloramination (adding ammonia to chlorine to form monochloramine) produces only ~5–50% of the adsorbable organic halogens (AOUC) compared to the same free chlorine dose (www.ncbi.nlm.nih.gov), reflecting greatly reduced THM/HAA formation. Dozens of utilities have met DBP limits by chloramination (nepis.epa.gov, nepis.epa.gov).

Topeka, KS replaced high pre‑chlorination with chloramination (3.6 mg/L Cl₂ then ammonia) and cut finished‑water TTHMs by 74% and distribution TTHMs by 71% (nepis.epa.gov). Longer contact times are required: Topeka needed ~11–12 hours of chloramine contact for pathogen kill, with no loss of disinfection efficacy (nepis.epa.gov).

Chlorine dioxide (ClO₂) does not halogenate NOM into THMs/HAA5 (xylem.com); San Diego’s Otay WTP installed ClO₂ generation and halved THM levels under Stage 2 DBPR pressures (xylem.com). ClO₂ does generate chlorite and chlorate, which must be monitored. Ozonation likewise oxidizes NOM without forming THMs or HAAs (xylem.com), though bromate formation in bromide‑rich waters is a concern (regulated to ~10 μg/L). UV disinfection produces essentially no DBPs—it neither chlorinates organics nor leaves a chemical residual (xylem.com)—but a residual disinfectant must then be added; utilities often pair UV with systems like ultraviolet.

Application point and process conditions

Adjusting the timing and location of chlorination also matters. Eliminating or minimizing prechlorination has been effective: moving chlorine dosing from raw water to post‑filtration (or removing side‑stream chlorination) cut THM formation in a multi‑utility survey (nepis.epa.gov). Several large cities achieved compliance by adding chlorine only at the end of the treatment train (e.g., right before the clearwell) instead of early pretreatment (nepis.epa.gov).

Lowering the chlorine dose reduces DBPs roughly proportionally, though with potential trade‑offs for pathogen kill. Process pH shapes speciation: high pH tends to favor THM over HAA formation, so some plants adjust pH downward during chlorination (which can increase HAAs) (nepis.epa.gov, iwaponline.com). In distribution, minimizing detention time or adding ammonia at the far end (boost chloramination) prevents excessive DBP accumulation.

Performance outcomes and regulations

In conventional systems without special optimization, coagulation/clarification typically achieves ~40–60% DOC removal (iwaponline.com), which can still leave THMs above standards in TOC‑rich waters. Enhanced coagulation (depressed pH and optimized alum dose) may reach 50–70% NOM removal in favorable cases (iwaponline.com), cutting THMFP accordingly. GAC filtration can then remove another ~30% of DOC (iwaponline.com).

On the disinfection side, switching to chloramine routinely delivers 60–75% lower TTHMs (as in Topeka’s 74% finished‑water and 71% distribution reductions; nepis.epa.gov). Introducing ozone or ClO₂ can eliminate nearly all THM formation (subject to secondary byproducts) (xylem.com).

Regulations commonly set TTHMs at ~0.08–0.10 mg/L (80–100 μg/L) (frontiersin.org, xylem.com). Indonesia’s drinking‑water standard is 100 μg/L TTHM (water.co.id). A utility hitting ~40–50 μg/L—like Little Falls’ ~46 μg/L after GAC—is well within guidelines (nepis.epa.gov).

Multi‑barrier playbook and pilots

Utilities deploy a multi‑barrier approach: optimize coagulation (via jar testing) to maximize NOM removal, supplement with GAC or PAC when needed, and then tailor disinfection mode and location. U.S. experience shows that meeting Stage 2 DBP limits usually requires either advanced precursor removal or switching disinfection mode—or both (nepis.epa.gov, nepis.epa.gov). Raw‑water NOM level matters: a humic lake source might justify enhanced coagulation plus a GAC contactor to remove ~70–80% of organics, whereas a clear mountain‑water system might switch to chloramination with minimal upstream changes.

Whatever the path, data drive decisions. DOC/SUVA monitoring and THMFP tests—paired with pilot trials—quantify expected DBP reductions before full‑scale change (nepis.epa.gov, nepis.epa.gov). Regulatory guidelines and case studies (Indonesia Permenkes, WHO, US EPA Stage 1/2) and peer‑reviewed water‑treatment research were consulted to quantify removal efficiencies and outcomes (water.co.id, iwaponline.com, iwaponline.com, nepis.epa.gov, xylem.com).