Janice Roberts at the M&G reports on water issues in South Africa.  The article's headline is 'SA water demand will exceed supply by 2025'.  So, first off, the title is wrong.  Second, the article doesn't tell us what the underlying models are, what's being held constant, what's varying.  We need to know these in order to understand what the author is trying to say, because she evidently isn't saying it well.

So, first off, Supply and Demand are functions, they are not quantities (see the graph adjacent stolen from wikipedia and yes, I understand that I'm adhering to certain weird assumptions using D&S graphs, but I believe they remain a decent heuristic).  If you want to talk about quantities, as in kilolitres or megalitres of water, then you should say 'the quantity of water demanded in the future will exceed the (future? current?) quantities supplied in South Africa'. Again, a function (Demand) cannot 'exceed' another function (Supply).

But, that statement is also incomplete.  Why? It doesn't talk about prices. So what the statement probably means is that the, at current prices, future quantities of water demanded will exceed the quantities supplied.  This is a bit nonsensical because the price will surely change and people will respond accordingly either by consuming less water for luxury uses, e.g., replacing huge water-sucking gardens with water-wise gardens, or having shorter showers, not installing or getting rid of a pool, or they'll find ways to recycle their water, e.g., grey water systems. 

Now, because water is a necessity, firms will probably have manage this with step-wise pricing systems as it is done now, with free water for everyone up to a point (currently six kilolitres per household regardless of size, again this is silly because it should be by household size, but the government has no way to credibly measure household size, and therefore some households are over-provided and some under-provided, sigh, oh, noting which I think this amount should probably be increased), and then water priced higher and higher per kilolitre the more that a household consumes.  My suggestion, as a complete novice of resource economics, would be to price non-essential water use even higher than it currently is priced.  But, this relies on an assumption that the amount of water provided to household actually affects water demand, and specifically that the fraction of the household demand in South Africa comprising those households that demand more than essential uses of water is sufficiently high to allow others to continue to use it. So this takes us to what the article could be saying, that the quantities of domestically-located water supplied will be exceeded by the essential quantities demanded by South Africa's populace by 2025.  That would be sad, but we could import water (pricey, I know), desalinate water (the technology of which will surely improve by 2025, making it cheaper) and so much more. But, of course, none of this makes for a decent headline of a nationally read newspaper. 

Also, please note that the article didn't speak at all about industrial water uses and industrial demand (here I'm talking about the function).  If SA experiences economic growth, and manufacturing growth particularly, then industrial demand (again the function) will increase, thus increasing the quantity demanded by industry as a whole.  This will obviously affected the pricing strategy of South Africa's water authorities.  Why? Well, they want firms to come here for FDI as FDI may (or may not) improve economic growth and increase employment.  But, because of international competition, firms will look for places with the cheapest inputs, e.g. water, electricity, labour, if South African prices increase they may choose not to come here and to go somewhere else.  So we have conflicting things happening here.  Firms that are currently located in SA may stay regardless of the cost increase and just lay-off some people (yes that sucks) to cut costs.  Or they may close (again job losses, which suck).  But with closing firms or firms cutting back, demand may decrease, or at least offset the increase in demand predicted from other sources (increased population).  But, my suspicion is that industrial demand and household demand will both increase, which will lead the price to increase and will lead firms to look for new technologies and people to look for other ways to use their current water before hitting higher prices (imagine richer households installing little alarms once they hit the next tier of water consumption - a loud voice chiming through the household 'Water is now 2 thousand rand per kilolitre!' or something).  I also favour social norm interventions and propaganda (yes, really) about water consumption, stuff like 'Do your duty, don't shower after gym!' (yes, that's a joke, who goes to gym? It's the rich folk who can afford to pay higher prices for water anyway! Duhuh...).

The problems all come from the 'other things equal' that the article implicitly held to, without telling us what was being held equal.  I get that my preferences for information probably exceed that of your average Johan, but come on, even he wants to know enough to make deliberate and informed decisions about his water consumption.