Under an answer to How is the quality of tap water in Thailand? I commented:

Usually water in pipes is pressurized so that it comes out of the tap when you open it; is "dirty water making its way into cracks in the clean water pipes" really a known thing? Does it really happen, can you cite an authoritative source for this phenomenon? Sure during a major weather event water treatment plants can be overwhelmed, and breakage of pipes resulting in complete loss of pressure can allow inflow, but I don't think heavy rainfall itself can push dirty water into the pipes against normal levels of pressurization. I'm certainly happy to be proven wrong if you find something.

and someone replied

water coming out with pressure is due to gravity because the tank is on the roof. Usually tanks are filled with water from municipal sources and then that tank water flows through pipes into the building's houses.

to which I replied:

Sure that's quite common, but at what point is there negative pressure such that the dirty water can leak into the pipes? I know this is a DIY SE or Physics SE rather than a travel question, I'm just really curious if this is really a thing, or just something we imagine.

Now those pumps that "pull" water up to the tank on the roof are worrying me; can they in fact be so aggressive that they drop the pressure in a small diameter rural municipal water pipe so much that dirty water could leak in?

Are there other things that can happen, short of catastrophic pressure loss that can lead to this happening?

  • 3
    An aspirator vacuum pump uses flowing water to create a vacuum; I don't see it as being impossible for the same situation to occur in a water supply pipe, but I don't know for sure. Commented Sep 25, 2022 at 10:17
  • 3
    If it can't flow away quickly because it's buried, wouldn't the pressure outside a crack/hole of a buried line equalize with the pressure inside? And if it's equalized and the pressure drops from usage inside the pipe, then the outer pressure would push it back in.
    – rtaft
    Commented Sep 25, 2022 at 19:19
  • 2
    A lot of times the source of those contaminations is never determined. Just recently, Baltimore, MD (USA) had a situation where e-coli was detected in drinking water. Areas of the city were on a boiled water, or do not use alert for a week or two. The emergency has now been lifted, but the source of the e-coli contamination has not been determined.
    – SteveSh
    Commented Sep 25, 2022 at 21:07
  • 1
    @SteveSh That's comforting! (not)
    – Michael
    Commented Sep 25, 2022 at 22:31
  • 1
    Just as a bit of an third-hand anecdote I was told by a firefighter in the US. He said pumper truck can pull enough of a vacuum when attached to the water mains that they can collapse/deform the plastic mains with an audible/tactile thump. But I might consider that a catastrophic loss of pressure.
    – RomaH
    Commented Sep 27, 2022 at 17:47

5 Answers 5


When there is no flow in an underground water supply pipe under pressure, a small crack or a loose joint in the pipe can only leak water. Contaminants can't get in.

However, as soon as there is flow, there is the potential for contaminants to enter the pipe due to the venturi effect. This occurs when there is a restriction, reduction or change in direction of the pipe, such as at an elbow or tee. And as @fraxinus points out, pipe corrosion can easily create both a restriction and an opening at the same place.

The U.S. Environmental Protection Agency publishes a guide (PDF) that discusses contamination of drinking water in distribution systems, and it includes this paragraph:

If a pipe with cracks or leaking joints is exposed to a wet environment, negative pressure can cause water to be drawn in (or to intrude into) the distribution system through backsiphonage (Kirmeyer et al., 2001). A separate issue paper addresses risks from intrusion due to pressure transients.

I was told by a foreman at a sewer construction site in New York City that buried water supply lines that run in the same trench as sewer lines must always be above the sewer line, never below. It's because of the potential for contamination from the venturi effect.

You'll understand how difficult it may be to comply with that rule in densely populated areas considering what it looks like under the streets. (Photo credit: davidgalbraith.org)


  • 4
    The effect is basically you opening your car windows and stuff gets sucked outside. That's probably the most readily experienced example.
    – Nelson
    Commented Sep 27, 2022 at 4:03

Backsiphonage and backpressure

The situation you ask about in your question is called backsiphonage, where pressure demand (suction) in a line that isn't adequately pressurized pulls in contaminants. The converse of that is a backpressure situation, where a pollutant source that's at higher pressure than the city water gets connected to it without any sort of non-return mechanism, causing water and pollutants to flow backwards from the pressure source into the city mains. (Note that in North America, it's very common to get a harmless backpressure situation when water main pressures are low due to thermal expansion in hot water heater tanks -- this is less of an issue in countries that don't use tanked hot water heaters, though.)

Since both phenomena pose a serious risk to public health, backflow prevention and cross-connection control (since you can't have harmful pollutants getting into the supply without a source for them) are major tasks for water utilities the world over. Generally speaking, sources of pollution (such as the wine bottler in DDS' answer, or more mundane sources such as hose bibbs, lawn sprinklers, and the like) have backflow preventers installed in their water supplies to prevent this from happening. (Lawn sprinklers, specialty sinks, and hose bibbs/outdoor taps use vacuum breaker devices, while more industrial applications will use double check valve or reduced pressure type devices.)

The US EPA has a handy manual that discusses this topic in more detail for those who wish to go further.


Here happened that a factory (wine bottling facility) pushed pollutant (wine in our case) in the city water mains. Authorities said the pressure used by the facility was higher than the acquaduct's and a failure of a facility valve.

  • 17
    Ah, turning water into wine. Practically Biblical....
    – Ecnerwal
    Commented Sep 25, 2022 at 14:01
  • 2
    Better than what happened in the town I grew up in... Similar story, back feeding into the municipal water supply under pressure... difference was it was the waste from a chicken processing plant rather than a winery.
    – James T
    Commented Sep 26, 2022 at 8:51
  • 2
    @JamesT And this is why water->sewer connections require air-gaps! you think poo backing up into the basement is bad, but it's nothing compared to poo backing up into the drinking water. Commented Sep 26, 2022 at 18:19

It would depend on the pump and the municipal water line. It's certainly not inconceivable for a pump to pull suction on a low-pressure supply line.

Also, in many cases the roof tank is also because the municipal water line is only pressurized SOME of the time, and at other times has no pressure.

In other cases the municipal water line does not exist, but there's a water truck that drives around and delivers.


Besides cracks and pressure as explained by other answers, also it is common during pipe repair works and in case of negligence in the purifying stations.

Or just too polluted water that the purifying equipment can't cope with. Unfortunately that happens around here especially in small villages.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.