A routine event, with numbers attached
Most people treat a main break as a one-off. The literature EPA publishes treats it as ordinary maintenance.
0.27 water main breaks per mile of pipe per year — which the paper states “would equate to 237,600 water main breaks in the United States annually.”
Across a network it puts at 880,000 miles serving roughly 223 million people, of which approximately 48 percent is unlined cast iron and 26 percent is judged to be in fair structural condition.
Roughly 4,400 miles are replaced each year, against 13,200 miles of new pipe installed.
Where the source comes from
The same caveat as the water age page, and it matters as much here.
This is an issue paper prepared by AWWA with Economic and Engineering Services, published by EPA in 2002. EPA states the issue papers “only present available information and do not represent Agency policy”, and that where prepared outside EPA, “EPA does not endorse those papers.”
The underlying figures come from surveys of the early to mid 1990s. They establish scale and mechanism, not today’s counts.
Two of the eight pathways run through the trench
The paper reproduces the recognised pathways for microbial contamination of distribution systems, in order of relative risk:
- Water treatment contaminant breakthrough
- Transitory contamination
- Cross-connections
- Water main breaks and/or repairs
- Uncovered storage reservoirs
- New main installations
- Covered storage reservoirs
- Purposeful contamination
Two are tied directly to construction and repair. And the paper makes a point worth keeping:
Contamination via water main breaks and/or repairs generally carries a higher risk than contamination from new main installations.
A repair is dirtier than an installation. The trench is already wet, the pipe is already in the ground, and the work is unplanned.
It adds that construction and repair “can indirectly provide opportunities for contamination pathways, such as cross-connections.” → Cross-connections and backflow: the route no treatment reaches
What you see is not what is flagged
The paper’s table of water quality issues from construction and repair, with its own health markers.
| Issue | Group | Direct health impact flagged |
|---|---|---|
| Pathogen contamination | Microbiological | Yes |
| Harmful chemical contamination | Chemical | Yes |
| Exposure to excess chlorine | Chemical | Yes |
| Loss of disinfectant residual | Chemical | No |
| Taste and odour | Chemical | No |
| pH stability | Chemical | No |
| Turbidity | Physical | No |
| Colour | Physical | No |
Turbidity and colour — the two you can actually see — carry no marker. Excess chlorine does.
That inverts what most people assume when brown water comes out of the tap after street work. → Turbidity: the line that is really a pathogen check
Why there may be more chlorine, not less
Disinfecting a repaired main is standard practice. The paper summarises it from ANSI/AWWA Standard C651-92, which recommends three methods.
So the taste after a repair can run in either direction: excess chlorine from the disinfection, or loss of disinfectant residual if the section sat stagnant.
One detail from that standard is worth noting. It suggests a temporary flushing and testing connection with a cross-connection control device matched to the hazard — but “backflow protection during main disinfection is optional, unless specifically identified in the purchaser’s specifications.”
The case the paper uses to make the point
Cabool, Missouri, December 1989 to January 1990.
An outbreak of E. coli O157:H7 that “resulted in 240 cases of diarrhea and 4 deaths.” The conclusion was that contaminants entered the distribution system through two major pipe breaks and 43 service meter failures during unusually cold weather.
The detail the paper draws out: the utility “did not practice disinfection following main repairs, relying instead on flushing the repaired main with finished water.” Modelling indicated the contaminant was possibly introduced from sewer overflows happening at the same time.
One event, in one town, thirty-odd years ago, reported at third hand. It is here because it is the case that links repair practice to outcome, not because it describes present-day risk anywhere. → Coliform bacteria: an indicator, not a pathogen
The pipe may not have been clean before it went in
A smaller finding, and an uncomfortable one.
The paper cites a Philadelphia case where a 12-inch pipe “was placed on the curbside of a street for two days with no protection.” Swabs of its internal surface found 75 percent positive for non-coliform bacterial growth.
It notes that ANSI/AWWA C600-93 requires pipe interiors be “kept free from dirt or foreign matter at all times” — while the current NSF/ANSI 61 standard does not provide specific requirements for the sanitary conditions of piping materials intended for drinking water.
NSF/ANSI 61 covers what materials leach. Whether the pipe was kept clean before burial is a different question, and the paper reports the standard as silent on it. → The NSF standards, and what each one covers
The decision framework
1. If a notice was issued, it governs
It reflects what happened, where, and what the utility is doing. Nothing general beats a specific instruction. → Boil water notices, and the two households that need an exception
2. Run the cold tap until it clears
Discoloured water after street work is disturbed sediment. It is the part you can see and the part you can act on.
3. Do not read colour as danger, or clarity as safety
The paper marks pathogens, chemical contamination and excess chlorine. It does not mark turbidity or colour.
4. Expect taste to move in either direction
Excess chlorine from disinfection, or flat water from a stagnant section. Both are consequences of the same work. → Water age: why the same supply differs street to street
5. Do not buy anything in response
There is no certified claim for “a main break”. The products that exist address the contaminants, not the event.
The decision path
| Your situation | What decides it | Where that lands you |
|---|---|---|
| Notice issued after a break | Specific to the event | Follow it exactly |
| Brown water, no notice | Disturbed sediment | Run the cold tap until it clears |
| Strong chlorine taste after work | Main disinfection is standard | Expected; ask if it persists |
| Flat or stale taste after work | Residual lost in a stagnant section | Usually resolves with flow |
| Repeated breaks in your street | 48 percent of pipe is unlined cast iron | A question for the utility |
| New main being installed | Lower risk than a repair, per the paper | Routine |
| Worried about the pipe’s cleanliness | No sanitary requirement in NSF/ANSI 61 | A standards gap, not a product gap |
| Looking for it on your report | Not a reported parameter | It will not be there |
Choose to act if
- A notice was issued. The action is following it.
- Discolouration persists after running the tap, or returns repeatedly, which is worth reporting to the utility.
Choose to wait if
- The water cleared after running the tap and no notice was issued. That is the expected course of a routine repair.
Choose to ask the utility if
- Breaks recur in your street, or you want to know whether repaired mains are disinfected or only flushed. That distinction is the one the Cabool case turns on.
Mistakes that cost the most money
Buying a filter for a main break. No certified claim addresses an event. The contaminants have claims; the event does not.
Reading brown water as a health emergency. Colour and turbidity are listed without the health marker in this paper’s own table.
Reading clear water as an all-clear. Pathogens and chemical contamination are what carry the marker, and neither is visible.
Ignoring a notice because the water looks fine. The notice responds to the event, not the appearance.
Drinking heavily chlorinated water rather than asking. Exposure to excess chlorine is one of the three flagged issues.
Assuming new pipe is sterile. Seventy-five percent of swabs from one unprotected curbside pipe grew bacteria.
Three situations, and what changes
A break two streets away, brown water for an hour, no notice. The ordinary case. Sediment disturbed by a pressure and flow change, which running the cold tap clears. Nothing in the paper’s table marks colour as a health issue, and no purchase follows.
A notice after a break, in cold weather, with several failures at once. This is closer to the shape of the Cabool event — multiple breaks and meter failures during a freeze. The notice exists because that combination is the one with a documented outcome behind it. Follow it.
A street with repeated breaks. Worth raising with the utility, given that 48 percent of the network in the surveyed period was unlined cast iron and each repair is, by the paper’s own ranking, riskier than a new installation.
The objections worth answering
“Breaks are rare.” At 0.27 per mile per year the paper puts them at roughly 237,600 annually. Rare at your address, routine across the network.
“Treated water cannot be contaminated afterwards.” Two of the eight recognised pathways run directly through construction and repair, which happen entirely downstream of treatment.
“Brown water is obviously the dangerous part.” It is the visible part. The paper marks pathogens, chemical contamination and excess chlorine, and leaves colour and turbidity unmarked.
“The utility disinfects after every repair.” Standard practice says so. Cabool is in this paper specifically because that utility flushed instead, and whether any given utility disinfects is a question you can ask.
“This is EPA’s assessment.” It is not. EPA publishes the paper while stating it does not represent Agency policy and that it does not endorse externally prepared papers.
Who this is for
Households who have seen street work, lost pressure, or had discoloured water, and want to know what actually happens to the water and what follows from it.
It is also for anyone weighing whether a main break is a reason to buy treatment. On the evidence here it is a reason to run the tap and read the notice.
It is not an assessment of any system, and during an active incident the notice governs.
What to settle first
Was a notice issued, and has the water cleared after running the tap?
Those two answers separate a routine repair from an event the utility has decided you need to act on.
→ Cross-connections and backflow: the route no treatment reaches → Boil water notices, and the two households that need an exception → Water age: why the same supply differs street to street → Coliform bacteria: an indicator, not a pathogen → City water treatment, and the three gaps a utility cannot close