Chloramine Residual Collapse in Summer Dead Ends
A 6.2 MGD community system has used chloramination for three years to control THMs. Free chlorine is applied ahead of a four-hour clearwell and ammonia is dosed as the water leaves the plant, targeting 3.0 mg/L total chlorine and free ammonia below 0.1 mg/L at the plant tap. Finished pH is 8.1, alkalinity 120 mg/L. The system has 85 miles of main, much of it 1950s unlined cast iron, plus several 2-inch dead-end stubs serving 20-40 homes each. Over the last six weeks (July-August, water temperature 79 degrees F), total chlorine at the three farthest dead-end sites fell from 2.4 to 0.4 mg/L. Free ammonia at those sites is 0.6-1.1 mg/L, nitrite 0.3 mg/L, and HPC at one site rose to 480 CFU/mL. Plant effluent has not changed: 3.0 mg/L total chlorine, 0.05 mg/L free ammonia, no coliform hits. Two dead-end customers have called about a swampy taste.
1. What is your diagnosis, and what sampling would confirm it or rule out something else?
Read the worked answer
The pattern is nitrification in the distribution system, not a plant problem. The decisive evidence is the split: plant effluent is stable at 3.0 mg/L total chlorine with almost no free ammonia, but at the dead ends total chlorine has collapsed while free ammonia has risen to 0.6-1.1 mg/L and nitrite has appeared at 0.3 mg/L. Free ammonia rising in the system can only come from chloramine decay, and nitrite is the fingerprint of ammonia-oxidizing bacteria. HPC of 480 CFU/mL is consistent with regrowth in a pipe that has lost its residual.
Confirm by sampling along the trunk line from the plant outward, recording total chlorine, free ammonia, nitrite, and temperature at each point, so you can see where the residual breaks. Check the ammonia feed calibration and the chlorine-to-ammonia ratio, because a plant-side overfeed of ammonia would show up as free ammonia at the plant tap, and it does not. Rule out a recent main break, a new high-demand industrial customer, and a cross-connection pulling in ammonia or organics. Also check whether the affected stubs are truly dead ends with no blowoff.
2. Give your corrective actions in the order you would take them, with the risk attached to each.
Read the worked answer
Start with the cheapest and fastest: unidirectional flushing of the affected dead ends to pull the ammonia-laden, nitrified water out, then establish a routine blowoff schedule for those stubs. Risk: discolored water complaints, so notify customers first.
Second, tighten the plant chemistry. Increase the chlorine-to-ammonia ratio so free ammonia leaving the plant stays near zero and total chlorine holds above 2.0 mg/L in the distribution; verify the ammonia feed pump is calibrated and not drifting. Third, consider raising finished pH toward 8.3-8.5, which favors monochloramine over dichloramine and slows decay. Risk: check that pH change against your corrosion-control plan and lead/copper sampling, because moving pH on a system with lead service lines can mobilize lead.
If those do not hold residual within a few weeks, plan a temporary free-chlorine conversion with the ammonia feed shut off, breakpoint chlorination to establish a free residual, and 3-4 weeks of operation before reverting. Risk: taste and odor complaints, higher THM and HAA formation, and the same lead-mobilization concern, plus you must tell Ohio EPA before changing treatment.
3. What monitoring do you put in place for the rest of the season, and what would push you into regulatory notification?
Read the worked answer
Treat the dead ends as your early-warning network: weekly total chlorine, free ammonia, and nitrite at the three worst sites, plus HPC while nitrite is present. Keep the routine coliform and chlorine-residual monitoring intact, and add sites where the trunk line shows the steepest residual drop so you are not only watching the worst case. Track temperature, because the problem will ease as water cools but will return next July.
Regulatory triggers are coliform or E. coli positives in the routine or repeat samples, exceeding the chloramine maximum residual disinfectant level of 4.0 mg/L, or failing to maintain the required minimum residual in the distribution. Any of those means notification under the rules. HPC below 500 CFU/mL is a process indicator, not a violation, so do not overreact to it; use it to decide whether flushing and ratio changes are working. Log the nitrite and ammonia data with the residual data on one sheet so the trend, not a single sample, drives the decision.