Cape Town, South Africa is experiencing an extremely serious drought, with estimates putting April 21, 2018 as the day that the city will be forced to shut off all municipal water supplies and begin water rationing, except to essential services such as hospitals.

One possible solution is to delay this date by turning the supply off and on for scheduled periods of time over the next few months. However, experts have said that this is not a solution that can be implemented, because the pipes aren't supposed to deal with this variation of pressure and leaks and other infrastructure damage will occur.

Is there a way that a city could implement a water rationing system where water is shut on and off periodically, while minimizing leaks and damage to infrastructure?

  • $\begingroup$ So if the pressure change is likely to cause damage then, no at least on the system they have. Replace the system with something better then yes - but will it be worth the cost? $\endgroup$
    – Solar Mike
    Commented Jan 22, 2018 at 7:49
  • $\begingroup$ Cost is a consideration, but timeline is probably the bigger limiting factor. In the long run desalinisation and other costly infrastructure will be used to solve the water crisis, but between now and April is the time frame that any solution needs to be rolled out in, if one exists. $\endgroup$ Commented Jan 22, 2018 at 17:26
  • $\begingroup$ The problem is usually with the valves themselves. I know several cities have ceased exercising isolation valves because the failure rate became too high. New York was one of these. They are building a new aqueduct, at least. So the problem probably has as much to do with the capability to control the water distribution. They may get some water shut off and not be able to turn it on again. $\endgroup$
    – Phil Sweet
    Commented Jan 23, 2018 at 15:21

2 Answers 2


It is difficult to provide a useful answer to this question without additional information. For example, what is the system pressurization - operating psi/bars? Having installed ductile iron water mains, and having pressure tested new installations to destruction (resulting from hairline cracks in new pipes), I can confirm that the effects can potentially be devastating to infrastructure - especially roads and possibly other foundations/buried structures. In addition to the obvious negatives associated with traffic disruption and reinstatement, there will also be substantial water losses should this happen.

Road collapse due to pipe failure.

Where HDPE pipe is used, one of the predictors of service life is the cyclic nature of the pressurization. Cyclic pressurization may not lead to catastrophic failure, as it could for ductile iron, but it will certainly shorten the lifespan of the network.

What is the nature of the network design in terms of control points? Can the network even be controlled? The valves, tees and joint connections will likely be the most fragile part of the system.

Nonetheless, in my view the best chance of implementing such a programme successfully would be to partially shut down the network on a rotational schedule in such a way that essentially the entire network would have a chance to stay pressurized - or at least ensure that the portion of the network that is exposed to risk would be minimized. This would require endpoint valves however. It could be possible, depending upon the nature of the network, to install these at strategic locations.

Whether such an approach is economic, or would provide any benefits at all in terms of a reduction of water consumption, is another question. My guess is it may not, but then again a rationing programme may not be enforceable either.

  • 1
    $\begingroup$ Thanks - I actually don't have the specifics of the system, I'm just a (physics) grad student/member of the general public. I think it sounds like this approach can't work, especially given your statement that it does not necessarily guarantee a substantial reduction in water usage. I'll wait another day or two to see if other answers appear before accepting. $\endgroup$ Commented Jan 22, 2018 at 17:29
  • $\begingroup$ @JonRayner A rotational schedule could work provided that it was properly done. I don't think this is possible within the time frame you indicated - by April - but given a couple of years planning and some funds, it could be organised for future purposes. Yours is an interesting question because it is an intelligent, more forward looking enquiry. Most often city network design is simply an ad-hoc, piecemeal process. Ideally one should establish the network elements, then isolate and test them. $\endgroup$
    – AsymLabs
    Commented Jan 22, 2018 at 20:01

The problem with older and under-maintained (this is South Africa after all) systems is that the pipes get brittle, rusted and joints fragile and the network won't be able to handle the transient forces of constantly opening and closing valves, even if it's done slowly. The networks are in general not designed to withstand such large constant cyclic loading in any way as the water supply is "usually" constant.

Another solution might be to close the valves partway to slow down the flow to a trickle so people consume the water more slowly. Other advantages are that leaks in the system and faucets will lose much less water and might possibly stop due to the pressure reduction.

Problems with this, however, are that some people might not have water at all due to head losses in the network and cavitation might happen at the half closed valves causing damage to it.

Just an update:

In a radio interview this morning, 24 Jan, the Western Cape's premier noted that they are actually drastically reducing the water pressure of suburbs that refuse to reduce their water consumption.

Edit based on Johnathan's comment:

I know there are regulations on the minimum and maximum water pressure to serve the system. The pressure could still be regulated to stay within these limits, but as a temporary solution to extend water availability, having water pressure lower than the norm is a bit better than having people walk or drive to collect their minimum daily ration of 25l of water; which is what will happen if nothing changes by 21 April 2018.

  • $\begingroup$ Also, it may not cause people to reduce their consumption as they will just wait longer until they get the same amount given the lower flow rate... $\endgroup$
    – Solar Mike
    Commented Jan 22, 2018 at 8:42
  • $\begingroup$ @SolarMike True, but it would be quite a pain to shower for half an hour in a trickle of water, plus there's less wastage if you wash your hands or leave the tap open while brushing teeth $\endgroup$
    – ChP
    Commented Jan 22, 2018 at 8:46
  • 1
    $\begingroup$ Getting people to change habits like having the tap open to clean teeth is a really good idea : But having seen the poor take-up with students it is not likely to change very rapidly... $\endgroup$
    – Solar Mike
    Commented Jan 22, 2018 at 8:51
  • 1
    $\begingroup$ @Jonathan, yes, there are indeed municipal regulations Tshwane's minimum is 2.5 bar when I last checked (9 bar max), other cities will probably be similar. This should be adequate for 2 story buildings. Blocks of flats any higher would hopefully have booster pumps and overhead storage tanks by default, most I've designed have. Anyway, with this drought on the verge of being classified as a natural disaster, I doubt anyone can complain about water pressure. $\endgroup$
    – ChP
    Commented Jan 23, 2018 at 12:42
  • 1
    $\begingroup$ By the way Charl, I found your answer helpful - am just slightly leaning towards accepting AsymLabs' answer (ideally I wouldn't have to choose between the two - they both tell useful parts of the same story) $\endgroup$ Commented Jan 23, 2018 at 18:28

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.