Collecting Rainwater

Collecting rainwater in Tanzania

I have some experience with rainwater collection in the tropics. Both for household use and for increasing the survival of planted trees in dry areas.

A rainwater tank that worked well

Gathering rainwater from the roof of the house where it will be used can be very practical in many areas. When we lived in the northwestern corner of Tanzania in 1988-89, we built a rainwater tank with a couple of improvements that worked well.

  1. Dividing the tank to form a smaller tank near the top (to maintain the water’s (height) potential energy – which gives a higher water pressure {and then better water flow} even when the tank contained a small amount of water) and a lower part to save that overflowing from the upper part, as long-term reserve
  2. Use of the roof of the tank itself as part of the rain collection. Water tanks are usually covered due to many good reasons (avoid algal growth, mosquito larvae, prevent evaporation, prevent direct contamination). But this leads to the rain that hits the tank itself to be lead off, to no avail, while the rain that hits the house next door is carefully led into the tank. The house had a surface area of ​​42 m2 , so using the top of the tank (about 2.5 m2) increased the effective area by 6% (Actually more because we were unable to have rain gutters around the entire roof.) All that was needed was a small vertical concrete wall along the edge of the tank top.
tank where rain on it flows and water until it goes straight to the bottom = no pressure tank with upper part to be filled first
To the left is a typical rain tank. The rain that hits the tank runs off. The water entering the tank goes straight to the bottom. When there is not much water in the tank, there is no pressure that can send water to a faucet at a useful height. To the right is a tank of the type we built and used. It also takes in the water that lands on the roof of the tank. The water stays at the higher level until the upper tank is filled. This gives pressure enough to deliver water even to a shower (which is not too high) even if there is only a little water in the tank.

Planting holes for trees in the dry areas

In the woodland plantation project in Karatu (near Ngorongoro, Tanzania) in 1993, a number of experimental activities were also included. One question concerned how the planting hole design could affect the survival of trees. (The number of surviving trees is significantly more important than the number of planted trees.) For example, a local Catholic priest had a very good survival, but each tree had a 1 meter wide and deep hole, partially filled with compost and manure. It required a very large amount of work. Common planting holes (about 30 * 30cm) had significantly poorer survival, but were much easier to make.

Planting hole has two ditches that collect surface water toward it. I saw an interesting idea about ”microcatchments” in a book from VITA. ”Microcatchments” were simply small dikes that caught and led water to the planting hole.

However, there were no proposals for measures for these. I did a spreadsheet calculation that used input data: the slope of the ground, the largest rainfall in a short time, the size of the hole itself, the width and depth of ditches, how quickly the ground absorbed water poured into a sample hole. The calculation was expressed in suggestions of how high the ”wall” below the hole would be and how long the two dikes would be. Microcatchment planting holes worked best. At the moment, I can’t find that file, but if anyone is interested, I can find or try to recreate it.

Experience from ferro-cement water tanks 

Tanzania 1988/89

I made a ferrocement water tank at the school where I worked in 1988/9, Iluhya Sec. School in Northwestern Tanzania. There were a number of discarded leaky galvanized water tanks hiding in the grass around the school. By simply making one lengthwise cut in the one that was in best shape, then resecuring with some self-tapping screws and metal strips, it could be used as a form for making ferrocement tanks that were longer (taller) than itself. (I think we stretched it outward some, to make it easier to remove.) Ferrocement is sand-cement mortar reinforced with chicken wire and can make strong, though thin-walled, structures (even boats).

The old tank to be used as form is on the old tank foundation ready to be wrapped with chicken wire and cement mortar. Two students that helped.
The old tank to be used as form is on the old tank foundation ready to be wrapped with chicken wire and cement mortar. Two students that helped.
Me and the finished tank
Me and the finished tank

The tank-form was placed on the old water tank foundation. It was lightly oiled with spill-oil then covered with one layer of chicken wire. The chicken wire extended a bit above the form to help form a joint with the extension upward to be made later. One strand of soft 3mm iron wire was tightened into every other corrugation.

Then it was ready to be covered with just enough (as DRY as workable for max final density and strength) sand-cement mortar to fill the corrugations plus about 1cm. (The walls were about 5-6 cm thick when finished.)

ferrocement tank covered with plastic to prevent drying out while hardening
ferrocement tank covered with plastic to prevent drying out while hardening

Very important, after applying the mortar, was to keep the mortar as WET as possible, at least to prevent it from drying as the cement hardens while binding water. In the picture below can be seen the tin and plastic used to protect the cement from drying. But it still had to be splashed several times a day.

We made a low conical roof that we formed on a sand pile beside the tank and lifted up  and fastened to the tank walls.

Low conical roof form
Low conical roof form

Rwanda 2002

I built a ferrocement rainwater tank for my in-laws in Rwanda in 2002. It has worked well, so I have gathered my notes and spreadsheets about how it was built, material availability, problems and possible improvements. These are in English to be more internationally useful. Here’s a link to the draft of the notes: and to the spreadsheet