Manganese: Below the Guideline, Above the Complaint Threshold

Discoloured, high manganese water compared to clear water.

Manganese is the parameter where 'compliant' sits furthest from 'no complaints', and most utility monitoring programs have not adjusted for that gap.

Discolouration complaints often follow predictable patterns. They cluster after weekends of low demand, planned flushes, or any event that disturbs settled flow conditions in the distribution network. When the monitoring record is pulled, the results are typically within guideline values. One parameter which may be responsible is manganese, and the reason it keeps appearing in post-event investigations is that most utility monitoring programs are designed to track health risk, not customer experience. Both are legitimate objectives. They just require different thresholds.

Two values, two different problems:

The Australian Drinking Water Guidelines set two separate values for manganese. The health-based guideline is 0.1 mg/L, derived from long-term neurological risk at elevated exposure. The aesthetic guideline is 0.05 mg/L, the level above which taste, staining and discolouration become customer-visible. ADWG Chapter 6 is explicit that aesthetic exceedances are expected to generate complaints even where no health risk exists.

The difference between those two numbers is where most programs face issues. Licence conditions and reporting frameworks typically reference the health-based value as the compliance threshold, which is appropriate for regulatory purposes. However, when internal monitoring triggers are also set near 0.1 mg/L, the program is misaligned with the customer experience. The guideline gap is only the visible half of the problem. Manganese does not have to exceed the aesthetic guideline to cause trouble: sustained levels at concentrations below 0.05 mg/L can still accumulates in the network, so a monitoring record that reads as compliant can be steadily building a deposit that releases later. The aesthetic guideline marks the point of immediate visibility at the customer’s tap, not a safe ceiling for what enters distribution. The ADWG recommends an operational limit of 0.02 mg/L leaving the plant.

Source water and treatment drivers:

Elevated manganese in treated water usually originates at the source, in the treatment process, or both acting in sequence.

In surface water catchments, the primary driver is thermal stratification. During summer, the hypolimnion becomes anoxic, and dissolved manganese mobilises from bed sediments into the water column. Autumn turnover, or drawdown to deeper offtake levels during drought, introduces that manganese-enriched water into the treatment train. Bushfire-affected catchments present a compounding risk: post-fire runoff carries elevated dissolved manganese, and the effect can persist across multiple wet seasons.

On the treatment side, the critical variable is oxidation chemistry. Dissolved manganese must be oxidised to insoluble MnO2 before filtration can remove it. Chlorination alone is a poor oxidant for manganese at typical drinking water pH. Permanganate and ozone perform better, but both are sensitive to pH and contact time. Reliable oxidation generally requires pH above 7.5 and sufficient reaction time before the filter. A plant operating within its licence can still pass dissolved manganese into distribution if oxidation conditions are not actively managed.

The distribution delay:

The more difficult problem for monitoring programs is what happens after the treatment plant gate.

Manganese entering the network at concentrations below the aesthetic guideline does not travel transparently to the customer tap. Over time it accumulates in pipe wall deposits alongside iron and biofilm. Under steady flow conditions those deposits are stable. Under changed conditions, including flow reversals from network switching, velocity increases during demand spikes, and flushing operations, they mobilise. Physical mobilisation is the biggest culprit, but changes in water chemistry can also destabilise those deposits. A drop in chlorine residual shifts the local redox balance and can reduce fixed manganese dioxide back to a soluble form, while a movement in pH alters both its solubility and how tightly it is held to the pipe wall.

The result is a pulse of discoloured, high-manganese water at the customer tap that reflects months or years of accumulated deposit, not current source or treatment performance. This is why post-event sampling so often finds nothing actionable. By the time the investigation is underway, conditions have restabilised and the monitoring record shows compliant results throughout. The event is a deferred consequence of sustained low-level ingress, not an acute treatment failure.

Setting a trigger that prevents complaints rather than confirming them:

A defensible operational trigger for manganese needs to be set with enough margin below the aesthetic guideline to try and reduce build up to a level that is sustainable and is manageable. What that margin looks like will depend on network size, maintenance programs, and deadends and stagnation, but the principle is consistent: by the time a result at the treatment outlet approaches the aesthetic guideline value, the problem is already accumulating downstream. The trigger needs to sit meaningfully below it, not approach it.

The trigger also needs a paired flushing protocol. Reactive flushing in response to complaints is often necessary but can mobilise deposits more widely if velocity and sequencing are not controlled. Programmatic low-velocity flushing in known accumulation zones, scheduled ahead of seasonal turnover and high-demand periods, reduces the deposit inventory before it becomes a complaint source.

Conclusion

If your manganese monitoring trigger is aligned to the health-based guideline, your program is designed to confirm regulatory compliance, not to prevent customer complaints. For most networks those are different objectives, and the gap between them shows up in complaint data rather than monitoring records.

The practical review is straightforward: check where your internal action levels sit relative to the ADWG aesthetic guideline of 0.05 mg/L, and whether your monitoring frequency at the plant will pick up changes in treated water before it's seeded around the network. For utilities with older infrastructure, stratifying storages, or catchments with recent fire history, the exposure window is narrower and the case for tighter triggers is stronger.

Manganese rarely produces a health exceedance. It is, however, very effective at generating sustained customer dissatisfaction while the monitoring record remains clean. Closing the gap between the operational trigger and the aesthetic guideline is the best way to reduce dirty water complaints over the long term and increase customer satisfaction.

Will your organisation look for long-term customer satisfaction and set an operational limit of <0.02 mg/L at the plant, which is less than the aesthetic guideline or opt for a higher limit, still compliant but potentially with some future you headaches?

References

• Australian Drinking Water Guidelines (ADWG), National Health and Medical Research Council, 2011 (updated 2022). Chapter 6, Chemical parameters: manganese.