BPJ Code

The water problem

With Cape Town's drinking water crisis looming, I worry about whether I will be able to get water from my taps all the way to July/August, which is when the regional dams usually start filling again. Last I heard we have 102 days' useable supply left, and it's mid-March, which takes us to June. June is not July, let alone August, so we're heading for a major problem.

I can't unilaterally solve all of Cape Town's water problems, but I thought I'd try my hand at securing at least my own supply. While I have a wellpoint on my property, I don't know how safe the water is to drink. Maybe I won't die from a month of drinking groundwater, but I wouldn't want to take my chances of years of this practice: cities' groundwater tends to be polluted with all kinds of nasty chemicals. (I'm not too concerned with pathogens - those can be killed. I'm more worried about the kind of nasties that seep down when people change their cars' oil and let the old oil simply drain into the ground.)

My focus isn't exclusive of solving the broader problem of Cape Town simply no longer being secure in having adequate potable water supplies. Technologies that can address this larger problem do interest me, especially if I think I can turn my brains into money.

Possible water sources and technologies for making it potable

  • Groundwater: boiling kills pathogens, but other pollutants remain. Requires distillation or osmosis. I have trivial access.
  • Seawater: similar problems as groundwater, except with a much higher concentration of solutes. It would take me two hours to reach the sea, but there's no practical limit to how much water is available.
  • Rainwater: probably no pathogens if collected on a clean surface. No chemical pollutants worth worrying about - acid rain is for marble statues, not a real concern for humans drinking it. Unfortunately there's very little rain in summer, and I have very little storage capacity, and what I have is empty or contaminated.

First prototype - low-tech solar still

My first prototype was just an old ice cream container with a glass lid, and some old plastic irrigation pipes for drawing off the moist air to a makeshift external condenser (a random plate of aluminium, under shade). Chunks of charcoal in the water serve to collect incident radiation, and as a side benefit provide a larger surface area from which water can evaporate.

While this design did produce a distillate, performance was very low. Over a whole day the condenser collected only about a milliliter of liquid water. A naive physics calculation shows that enough solar energy enters the ice cream container to evaporate 300ml of water. 0.3% efficiency just won't do!

Something I noticed with my first design is how much water was collecting as drops on the glass lid:

This was Mother Nature giving me a hint that drawing off the moist air to be condensed externally was not the way to go. Taking her inspiration, I made a second ice cream container still, this one with a slit irrigation pipe as a gutter to collect condensed water. (This iteration is the one shown above, with the glass lid covered with drops of water ready to coalesce and run down into the gutter.)

Performance with this glass lid condenser was much better: on a good day (hot, windstill, sunny) I collected 28ml of distillate. Still not what's physically available, but 10% is respectable.

Question for readers: where is the energy going, if not to evaporating water which condenses and collects in a jar outside the still?

Going viral, and stepping up the tech ladder

Next thing I knew, my mad inventor friend Jochen Demnitz from Fan Engineering Works had accidentally inspired his brother Jörg to build a larger steel version of my still. I now have custody of it and I think my task is to press it to its best possible performance.

A welded steel still body

Its runoff angle is much steeper than I could achieve with my upcycling tech. Overall its volume is much larger, so I speculate that convective currents can more easily form, and I hope this improves performance. Naive physics suggests that up to 3 liters could be produced on an ideal day.

Lo and behold, on even a largely overcast but windstill day, I drew off over 215ml from this larger still. 7% of the maximum possible is not bad going for an overcast day:

(The next day, sunny but windy, I managed to draw off far less, maybe half as much.) While my scavenged lid/condenser works, the collected water tastes horrible. Warm plastic in contact with water can't be the best for water quality, and in any case this still is intended to have a glass lid.


The FEW brothers are talking to plastics people about molds for producing larger numbers of solar stills. Before that happens, we need to get closer to the 3l/day that's available from the sun's energy. I'm still working on how to improve performance. It probably involves insulating the still body, and using a glass lid might help too.

Scaling up

Yes, I want to make moar clean water. The obvious approach would be to just produce a million or ten stills sized to supply a single person with drinkable water. But I'm thinking of something more sane. With larger devices should come higher efficiencies, and if we're going to produce millions of liters of water daily, we might not want to draw the feedstock from our aquifers. This spells: Concrete barges floating in False Bay. Square kilometers of them, connected to the city's water system by pipeline.

Other topics

  • Vacuum distillation: I suspect much of the inefficiency of my stills has been due to radiation energy going towards increasing the temperature of water (increasing losses to ambient), rather than evaporating it. Distilling under vacuum gives the water nothing else to do but to evaporate.
  • Refrigeration cycle: Way more high-tech (but nothing outlandish), but perhaps offering a higher production intensity. The idea is to run a refrigeration cycle merely as a heat pump, to transport heat from the condenser to the evaporator, rather than with the goal of producing temperatures far off ambient.
  • Osmosis: interesting, but I don't care about it, because it's technologically inaccessible to me.

Thanks for reading!

If you have any questions or comments, please send me an email: bernd@bpj-code.co.za