Rainwater Harvesting

A torrential tropical rain started 30 minutes ago – I love it!

Much middle-class and almost all higher-class housing is made from concrete blocks constituted from beach sand and cement. Given the small area of our coastline and the fact that the beaches are even more limited, given we are a precipitous volcanic island, the extraction of sand has had, in some places, a disastrous environmental effect with beaches destroyed and communities at risk of falling into the sea.

Wood, even our superb tropical hardwoods, are generally viewed as a lower-class construction material – except by a few, of whom I am one! Obviously, the exploitation of tropical hardwood is also a great problem but most of our wood is coming from current and abandoned plantations rather than primary forest, where trees (if the will exists), can be replanted.

Our house has ended up about 75% wood and 25% concrete. The wood is mostly local African Teak (a different species from the most-valued teak, but still very, very good) and Jackfruit (stairs and bathroom counter-tops), another very good white hardwood.

Most of the urban population depends on public taps supplied with mountain water from a colonial supply system probably 50 years old, just patched up over the years Rural populations collect, often contaminated, river water. Our nearest clean freshwater supply is a kilometre or more away … once upon a time there had been buried underground piping from the source to the nearest public tap but someone stole it! The local rural population relies on rivers and springs.

So I decided on rainwater collection.

Everyone here thinks rainwater is unclean, even given the occasional cholera outbreaks resulting from contaminated river water.

The rich will construct a cuboid semi-subterranean tank from the aforementioned concrete blocks and mortar with water supplied from the public water supply network. The problem here is threefold – “cylindrical or spherical shapes optimize the use of materials and the wall strength”, the joints between concrete blocks sealed with mortar are points of weakness as are the joints between vertical and horizontal.

We constructed our first hemispherical ferrocement tanks of approximately 40 m3.

The inspiration for the design and construction technique came from the drawing below with some modifications.

Firstly, the size – I halved it as the tank in the drawing was for a rural primary school in Kenya! The modifications were as follows:

a) the draw-off pipe is not at the bottom of the tank for various reasons. We send water from the tanks to a 2000 litre tank on small tower with an electric pump which we connect when necessary. The water falls by gravity to the house taps. Water just below the surface will be cleaner than at the bottom of the tank where detritus will accumulate. So the draw-off pipe is a hose tied to a buoy that rides up-and-down a vertical pipe wedged in-between the roof and floor of the tank.

b) as we do not have a bottom draw-off pipe, we did not construct the 5 cubic metre run-off reservoir.

c) we have had a problem with our very large overflow pipes (4 inch – necessary because of the intensity of the rain) - of giant African snails crawling up and drowning at the bottom of the tank – solution has been a small piece of mosquito netting and inner-tube rubber to secure the netting to the PVC tubing.

d) filtering – from the guttering around the house and bloody great metal boxes on each corner of the house (to try and capture the sheer quantity of water that occurs during a tropical storm), the water passes through a coarse filter of stone chips and then a finer filter of mosquito-sized netting

The three years, four starts of rainy seasons, have been a learning experience … A torrential downpour on 30 September 2004, a late beginning of the rainy season, the day we moved in … so a planned purchase of water was not necessary.

During our first rainy season, supply with one tank was more than enough, but come our dry season from May through September, there is no rain and the one tank would get lower and lower … Nanda insists on washing clothes by hand in a wash-tank practically everyday in spite of a washing-machine! They don’t trust washing-machines, thinking they can do a better job by hand and not understanding they are beating the shit out of my clothes! Perhaps Nanda would feel differently if she had to trek to the river! Anyway, I had wildly underestimated our consumption.

Over the first year it soon became apparent that 40 m3 would not be enough for an entire year so we constructed another tank of c. 30 m3.

In 2005 and 2006 the rains also started late. This year, with obviously much less domestic consumption, the rains have already started.

So what is ferrocement?

Basically we made a net of chickenwire and barbed wire and applied a cement-powdered stone (not wanting to use beach-sand which as being ecologically would have leached out salt) mixture in layers over this. As it is subterranean the underlying earth will provide support.

Reference: John Gould and Erik Nissen-Petersen. 1999. Rainwater Catchment Systems for Domestic Supply. Intermediate Technology Publications.