Fixed Sulfur Smell, Now Chlorine Taste: Second-Fault Masking Decision Tree

Why this matters

A sulfur-removal install or repair, often a chlorination injection plus retention plus carbon polish stage, that successfully eliminates the rotten-egg smell sometimes uncovers a chlorine taste at the tap a few weeks later. The customer reads this as a new fault introduced by the visit when the actual chemistry is straightforward: the sulfur was masking the chlorine, the carbon polish stage is no longer capturing all of the residual, or the chlorine injection dose has drifted upward. This tree separates root causes and gives the corrective action without making the visit feel like a callback.

The decision flow at a glance:

  Sulfur fixed, now chlorine taste - second fault?
  |
  +-- 1. Carbon exhausted / short EBCT? --> REPLACE /
  |                                         RESIZE
  |                                         CARBON
  |
  +-- 2. Chlorine dose drifted high? -----> REBALANCE
  |                                         DOSE TO
  |                                         CURRENT
  |
  +-- 3. Sulfide load dropped? -----------> REDUCE DOSE
  |                                         (LESS TO
  |                                         OXIDIZE)
  |
  +-- 4. Retention time reduced? ---------> FIX TANK /
  |                                         CONTACT TIME
  |
  +-- 5. Carbon not backwashed? ----------> BACKWASH
  |                                         SCHEDULE +
  |                                         REBED

Symptom presentation

Customer call several days to several weeks after a sulfur removal service: rotten-egg smell is gone, but cold treated water now tastes or smells of chlorine, bleach, or pool. Treated water tests show free chlorine present at the cold tap (typically 0.5 to 3 ppm), no sulfide detected, hardness and iron in spec. Customer may report the new taste appeared gradually or all at once after a specific event (a backwash cycle, a media change, a power-cycle of the injection pump).

Quick checks at the tap and the equipment

• Test free chlorine and total chlorine at the cold tap. Free above 0.2 ppm at a downstream tap means the carbon polish is not removing residual.
• Test free chlorine at the injection point or the tap immediately downstream of the retention tank to confirm dose. A dose target for sulfide oxidation is typically 1 to 2 ppm above the sulfide ppm being oxidized; a dose drifting toward 5 ppm or higher is too high.
• Inspect the chlorine injection pump: stroke setting, suction line condition, chemical tank concentration, and date of last refill.
• Inspect the retention tank: contact time at typical household flow, internal baffling intact, any short-circuit path.
• Inspect the carbon polish stage: media age, bed depth, backwash flow correct, no channeling.
• Confirm the system layout has the carbon stage AFTER the retention tank, not before. A misordered install puts the carbon in front of the chlorination and the customer gets chlorine straight to the tap.

Isolation tree

Branch A: carbon stage exhausted or short on contact time Activated carbon adsorbs chlorine effectively at appropriate empty-bed contact time (EBCT), typically a minimum of 1 to 3 minutes depending on bed type and chlorine concentration. A bed that has reached its capacity for chlorine adsorption breaks through, and the customer tastes chlorine that was previously captured. A bed with adequate volume but undersized at peak household flow lets chlorine through during high-draw periods (showering, laundry running concurrent).

Branch B: chlorine dose drifted high A peristaltic injection pump that has had its stroke setting bumped, a chemical tank refilled with a stronger concentration than spec, or a dose that was set high during initial commissioning to ensure sulfide oxidation and never reduced once steady-state was reached. The carbon was sized for a target dose; an above-target dose breaks through faster.

Branch C: sulfide load decreased Wells with seasonal sulfide variation can present at higher concentration in some months than others. The injection dose set during peak sulfide presents now as excess chlorine because there is less sulfide to oxidize. Free chlorine residual after the retention tank rises, and the carbon faces a higher chlorine load than the original sizing assumed.

Branch D: retention tank contact time issue The retention tank may have lost effective volume through internal sediment accumulation, a failed baffle, or being undersized at the household's peak flow. A reduced retention time means less of the injected chlorine is consumed by sulfide oxidation, and more passes downstream as free residual to the carbon, accelerating carbon exhaustion.

Branch E: carbon backwash insufficient A carbon bed that is not being backwashed at adequate frequency or flow rate accumulates fines, develops channels, and bypasses water around its capture zone. The bed depth looks correct on inspection but the effective bed is much smaller. This commonly appears 3 to 6 months after install when the customer's flow demand has stabilized but the backwash schedule was set for a different load.

Branch F: pH-shifted chlorine speciation At higher pH (above 8.0), free chlorine exists more as hypochlorite ion than hypochlorous acid. Carbon adsorbs both but at different efficiencies, and the perceived taste threshold shifts with pH. A retention or pH-adjustment stage that has shifted treated water pH upward can present as a chlorine taste complaint even at the same total residual.

Decision thresholds

Free chlorine at downstream tap under 0.1 ppm: carbon is working. Taste complaint is not chlorine; investigate other taste sources per the in-spec-but-tastes-off decision tree.

Free chlorine 0.1 to 0.5 ppm: carbon is partially through. Check carbon age, EBCT, and backwash settings. Replace media if past rated volume.

Free chlorine above 0.5 ppm: significant breakthrough. Investigate dose, retention contact time, and carbon condition together.

Injected chlorine dose more than 2 ppm above the current sulfide reading: dose is too high. Reduce stroke or chemical concentration.

Retention tank contact time under 20 minutes at peak flow: tank is undersized for current flow. Consider larger tank or staged retention.

Carbon bed past rated bed volume in throughput, or bed visibly channeled, or fines present in significant volume: replace media.

Confirming diagnosis

Sequence:

1. Test free and total chlorine at: raw water, immediately downstream of injection, immediately downstream of retention tank, downstream of carbon stage, and cold tap nearest the equipment.
2. Calculate the chlorine consumed across each stage. Injection should match the pump stroke and chemical concentration; retention should show a drop equal to or greater than the sulfide load oxidized; carbon should drop residual to near zero.
3. Measure household peak flow and calculate EBCT through the carbon bed at peak flow. EBCT under 1 minute at peak is generally insufficient for chlorine removal regardless of bed condition.
4. Pull a carbon sample and inspect. Healthy granular carbon is uniform, free-flowing, and free of fines. Spent or channeled carbon shows fines, color shift, or visible bed compaction.
5. Verify retention tank serial volume: drain and inspect for sediment accumulation if access permits.

Remediation

Adjust dose first if dose is high: reduce injection stroke to the target above current sulfide. Allow the system to stabilize for 24 to 48 hours, then re-test.

Branch A: replace the carbon at the media manufacturer's rated throughput, and size the bed so EBCT stays inside the media spec at the measured household peak flow, not at average flow. Where the existing vessel cannot hold enough media to make that EBCT, the answer is a larger vessel or a second carbon stage in series, quoted as equipment. Acceptance: free chlorine under 0.1 ppm at the cold tap with a peak draw running.

Branch C: re-set the injection dose to the sulfide reading you measured today, not the one from commissioning, and put the seasonal swing in the customer's file with a re-check scheduled for the opposite season. A dose left at the peak-sulfide setting year-round guarantees this call comes back every time the well drops. Tell the customer the swing is the well, not the equipment.

Branch D: clean the accumulated sediment out of the retention tank and repair or replace a failed baffle, then re-verify contact time at peak flow. Where the tank is simply too small for the household's current peak, no cleaning fixes it; a larger tank or staged retention is the remedy and it is an equipment quote, not a service adjustment.

Branch E: correct the backwash frequency and flow rate to the media manufacturer's spec for the measured service flow. If the bed is already channeled or carrying fines, replace the media first and then run the corrected schedule, because a new schedule on a ruined bed just backwashes a ruined bed.

Branch F: bring treated pH back into the design range by correcting whatever shifted it (neutralizer media exhausted, feed pump setting, blend valve), then re-taste and re-test residual. Where the water genuinely runs high pH by nature, size the carbon for dechlorination at that pH per the media manufacturer's data rather than assuming the original sizing still holds. Do not chase this one by cutting the chlorine dose below what the sulfide load needs; that trades a taste complaint for the smell you were hired to remove.

References

• NSF/ANSI 53, Drinking Water Treatment Units, Health Effects (carbon chlorine reduction).
• NSF/ANSI 42, Drinking Water Treatment Units, Aesthetic Effects (taste, odor, chlorine reduction claims).
• EPA Safe Drinking Water Act, 40 CFR Part 141, maximum residual disinfectant level for chlorine.
• WQA Technical Application Bulletin on hydrogen sulfide removal and chlorination plus carbon design.
• AWWA Manual M20, Water Chlorination and Chloramination Practices and Principles.