Secondary disinfection byproducts (NDMA, bromate)

WTR
0health goal (bromate)

EPA's health-based goal for bromate is zero. The legal limit is 10 ppb.

Settled science

Secondary disinfection byproducts are trace chemicals your water utility creates while disinfecting, beyond the trihalomethanes. The mix depends on the disinfectant: ozone leaves bromate, chlorine leaves chlorate, chlorine dioxide leaves chlorite, and chloramine leaves nitrosamines like NDMA. Reverse osmosis clears the inorganic anions. NDMA is the exception: it resists both carbon and RO, and UV is the control.

What it is, and where it comes from

These are the disinfection byproducts that sit outside the regulated trihalomethanes and haloacetic acids. Disinfection itself creates them; they are not in the source water. Which one a household sees depends on what disinfectant its utility runs. Ozone oxidizes natural bromide into bromate. Chlorine and aging hypochlorite solutions leave chlorate. Chlorine dioxide leaves chlorite. Chloramination, a dominant disinfectant for many US systems, can form nitrosamines such as NDMA. Every US community water system publishes an annual Consumer Confidence Report, so you can run the same compound-by-compound read for any address.

Why it matters

The concern divides by compound. EPA sets bromate's health goal at zero and lists its long-term effect as increased cancer risk, the same no-safe-threshold posture as the genotoxic THMs. NDMA is the standout: ATSDR reports the liver as its primary target organ, and it is classified as a probable or reasonably anticipated human carcinogen by EPA, IARC, and DHHS. It is potent at trace doses, which is why its advisory level is measured in parts per trillion, three orders of magnitude below the part-per-billion byproducts. Chlorite carries a non-cancer endpoint instead: anemia, and nervous-system effects in infants and young children. Chlorate's advisory level rests on thyroid-related effects.

What we grade it against

Where the health research draws the line, versus the legal limit.
ContaminantHealth-based levelLegal limitSource (health-based)
Bromate (ppb)0MCLG zero; increased cancer risk10federal MCLUS EPA NPDWR
Chlorite (ppm)0.8MCLG; anemia, infant nervous-system effects1.0federal MCLUS EPA NPDWR
NDMA (ppt)Probable carcinogenpotent at trace dosesnone setno federal MCL; carried as a CA notification-level chemicalATSDR ToxFAQs; CA SWRCB notification-level program
Chlorate (ppb)Thyroid-effect basisnone setno federal MCL; carried as a CA notification-level chemicalCA SWRCB notification-level program

Health-based levels come from peer-reviewed research and government risk scientists working without cost constraints. Legal limits are enforceable compromises. Your report grades to the health column.

What helps

Direct fixes

  • Reverse osmosis (NSF/ANSI 58) at the tap

    Rejects charged inorganic anions, so it is an effective point-of-use control where bromate, chlorate, or chlorite is the concern.

    Does not reliably reject NDMA or the other nitrosamines, which are small uncharged molecules.

  • Activated carbon filter

    Not a reliable barrier for this group of byproducts on its own.

    Does not adsorb NDMA well, and does not remove the inorganic anions bromate, chlorate, or chlorite; those need RO.

Bigger retrofits

  • UV photolysis (utility-scale or specialized)

    UV light destroys NDMA, the control that matches its chemistry. This parallels 1,4-dioxane: both are small molecules that defeat the carbon and RO toolkit and require destruction rather than capture.

    Not a typical under-sink cartridge. Air stripping is also useless here, because NDMA does not evaporate.

Free and behavioral

  • Read your utility's annual CCR compound by compound

    Tells you which of these byproducts your address actually faces, since the mix shifts with the local disinfectant. The same read runs for any US community water system.

Like the trihalomethanes, these byproducts are the cost of a public-health intervention that did its job: disinfection made water safe to drink, and we manage the ultra-trace residue at the point of use. RO handles the anions but not NDMA, and we will not represent a carbon or RO unit as solving it. Where nitrosamines are the concern, the answer points to UV, not a cartridge.

Common questions

Are these the same as trihalomethanes?

No. They are a separate layer of disinfection byproducts that sit outside the regulated trihalomethanes and haloacetic acids. Different disinfectants leave different residues: ozone leaves bromate, chlorine leaves chlorate, chlorine dioxide leaves chlorite, and chloramine can leave nitrosamines like NDMA.

Will my reverse osmosis filter remove all of them?

Not all of them. RO rejects charged ions efficiently, so it is an effective control for bromate, chlorate, and chlorite. It does not reliably remove NDMA, a small uncharged molecule that passes through both carbon and RO.

So what removes NDMA?

UV light. ATSDR notes that sunlight breaks NDMA down in water, and UV photolysis destroys the molecule. It is a utility-scale or specialized process, not a typical under-sink cartridge. NDMA shares the bind with 1,4-dioxane: capture fails, destruction works.

Which of these should I worry about?

It depends on what disinfectant your utility runs, which you can read in its annual Consumer Confidence Report. In chloramine systems, the nitrosamines matter most. In ozone-treated systems, bromate is the one to watch.

Why is NDMA measured in parts per trillion instead of parts per billion?

Because it is potent at far lower doses. Its advisory level sits about three orders of magnitude below the part-per-billion byproducts like bromate, which is why detecting it at all takes part-per-trillion sensitivity.