Optimising water supply in Ethiopia

Gain insights to the field of addressing critical water supply challenges in sometimes equally challenging surroundings. Through the use of district metering areas and network pressure management, water loss and non-revenue levels are remarkably reduced.

By Gertjan Van Der Ende, Management Trainee at Evides Drinkwater.

A story of a young, happy, concerned water specialist in his, fortunately not entirely lonely, adventure to make the world a bit better. For about 5 months I worked with the drinking water company of Addis Ababa, the capital of Ethiopia.

There you are. Mentally exhausted, because you just mentioned a number of tourist destinations to visit with a tight face. It was already too late to arrange a work visa, so the tourist visa was actually the only realistic option. Fortunately, the customs man is not too critical around this hour, and he decides that my story is credible. It is 6 o'clock in the morning, and the drive from the airport to the hotel is quiet. Addis Ababa still sleeps. This city with 4 million inhabitants is, possibly, not a 24/7 city. Furthermore, contrary to what most of my friends and family thought, it is not a warm, dusty one. Actually, it is a pretty organised city, with a temperature that dances around 23 degrees all year round. So ideal.

Well now, contrary to what they now thought in customs, I did not come for the elegant churches in Lalibela, the lava lake of Erta Ale, or the very nice Ethiopian food. I came to help the local water company with one of their main problems: Non-revenue water, or in short, NRW.


NRW stands for Non-Revenue Water. Actually, it is a very simple sum. You take all the water that is produced, and subtract all the sold water cubes. What is left is non-revenue water. That part of your water that is never paid for. In practice, this includes losses because not all clients receive an invoice, because there are holes in the pipes, and losses because meters are broken.


DMA’s, or district metering areas, are more or less the answer to NRW. This is a strategy that consists of dividing your pipeline network into all small areas (200 to 5000 connections). Here it is precisely determined how much water goes into it (using a water meter) and determines how much is sold. Because these are small manageable areas, the reasons for the loss can be traced and resolved.

NRW (see above for explanation) is a phenomenon that is not limited to the developing world. This is sometimes a major problem in Western countries as well. We in the Netherlands also suffer from NRW, but luckily we have a better hand than the Ethiopians. In the Netherlands we are at 5 % NRW. The English often go to 25 %, but in Addis they were talking about 40 %. This is not only a waste of lost income. First of all it is a real pity, because the limited availability of water then means no water to some parts of the city.
The goal of my project was to make a firm impression of how high that NRW percentage actually was, and then to halve this percentage. This will all happen with the help of DMA’s.

'On the ground' in Ethiopia
Freshly arrived, I obviously had a plan ready. After extensive acquaintance with all employees, we were able to adjust this plan to the local taste and approve it. Very smooth. My boss, however, was very ‘ambitious’. He said he would be satisfied if I could realise just half of my plan. On one hand, a very reassuring statement (the bar is, apparently, not that high), but on the other hand it is also very frightening, because how many setbacks are normal here?
Anyway: I had a plan, I had 4 smart local engineers and about 10 mechanics - let's see where the ship is stranding.

It quickly became apparent that my added value was not so much in the technical knowledge that I brought. DMA’s are simply not that complicated. My added value was my ability, list drive, my access to (higher) authority and my direct link to project funds.

No pressure ...
In the first DMA, the water entered the area through a single pipe. There were no pipes that left the area. In itself a very simple setup. The water meter was installed quite quickly. Now it was important to test whether the area actually had only one incoming and no outgoing pipe. For this we do a so-called 'zero pressure test'. In this process, the incoming pipe is shut off, so that there is no water supply, and therefore no pressure in the area. This is all closely monitored by pressure gauges that register, both the pressure within the DMA and the pressure outside the DMA. When there is pressure in the area around the DMA, and the pressure is zero in the DMA itself, we can conclude that there are no other connections that we knew nothing about. To minimise the burden on our customers, we decided to do this in the night. The next morning, however, it appeared that some pressure gauges in the DMA still showed a major pressure. The suspicion was that this was hydrostatic pressure, by water that was left in the pipe. The only way to know this for sure was to ensure that the pipes would be completely empty. The plan was to use a low-lying fire hydrant and drain the water on the street. As far as I am concerned, not an issue, but my colleagues found this, not entirely unjustified, a waste of water. They did not work for nothing in the non-revenue water team. The solution ended up being a water truck. It was arranged that a water truck would pass, into which we would pump all the remaining water from the pipeline. This was already the second attempt of the zero pressure test, and therefore also second day that this district would be closed off for water.

Sounded like a good idea. Sadly, when the car had well arrived at the fire hydrant, the connection of the fire hydrant did not fit on the water truck ... It was already in the afternoon of my free Saturday, and I was done with it in the meantime. So after a short discussion, we agreed that we would discharge the water on the street. The lost water was a pity, but this was an investment in the future. When we arrived back at the fire hydrant, the tap was opened full of enthusiasm. And .... nothing happened. The local residents were still able to tell us that a valve was built underneath the pavement, which unfortunately is now unreachable. Another (free Saturday) day thrown away, attempt three approaching the following Monday. Who then also assumes that the only fire hydrant in the district is actually functioning ...

Surface boxes
To make valves accessible, we use surface boxes in the Netherlands. A tube with a lid - not all that exciting. The one we needed had to be very strong, given that it would be placed in the middle of the main road. Sadly, these kinds of products are not found everywhere on the shelves in Ethiopia. Import and export does not go so smoothly here. What I understand from the situation is, that the shortage of hard currency (dollars, euros) is so acute that it actually affects the entire economy. There are no dollars left for anyone. Construction projects are at a standstill, and imported items are poorly available.
And, not only large projects felt this, we too. In all my inventiveness and creativity I had found a steel company that could make a cast iron surface box for me. We had designed it together and there was even a mould made of sheet steel. I was quite proud of this. Tailor-made materials, more than strong enough, theft-proof because of the absurd weight, and also a great support to the local economy. My academic education of a combination of development economics and civil engineering got it all hot. The mould that was made was approved, and we were able to proceed to the next step. This step consisted of manufacturing a mould from sand. Here the liquid cast iron was poured, after which it would cool and harden. Anyway, we were of course talking about currency issues. As it turns out, because of the complex shape of the design, a certain chemical substance was needed. This stuff kept the sand together, but was unfortunately not available. Why not? Import has not been possible for some time ...
In the meantime I got a surface box at our own warehouse. A beautiful model of the prestigious AVK. Fully plastic (cast-iron manhole covers are often stolen in Ethiopia), and yet strong enough for the purpose. I take it with me in my suitcase, and if I like it, we fly in a few pallets.

I would be able to add some of these stories, but it seems like a lot was difficult and slow in Ethiopia. That would not do justice to my time there. There is also a lot of success there. Four DMA’s have been set up, nine water meters have been installed and three manholes have been installed. All measurements for determining the current NRW percentage have been done, and various training courses in the field of NRW and GIS have been provided. The NRW percentages of sometimes more than 50 % painfully exposed how much income is lost. At the same time this also gives hope for the future, because there is still so much low-hanging fruit to pick.

We can conclude that the estimates of about 40 % NRW were relatively accurate. In the established DMA’s the figures varied between 36 % and 52 %. We are now ready for phase two of the project in which the problems have to be solved. I'll fly back next month. With an AVK surface box in my suitcase. Let's see if I can speak that right to customs.

AVK solutions

Learn more about the many ways we can assist in efficient water supply projects.

Go to water supply

Gain Knowledge

Enter our knowledge area, where current conversation and recent technologies (incl. NRW and DMA) are in focus.

gain knowledge