"Urgent" repairs to the Kariba dam on the Zambezi river have been delayed for two years, but the World Bank says there is no danger. In 2015 the Institute of Risk Management South Africa published a report which included the quote “if nothing is done, the dam will collapse in three years”.
This led to an AIM report headlined "Race against time to save Kariba dam". There has been similar South African press coverage in 2014.
Apocalyptic predictions were made of the dam breaking and a flood rushing down the Zambezi river and destroying the Cahora Bassa dam as well.
The Kariba dam is between Zambia and Zimbabwe, upstream from and similar to Cahora Bassa, generating electricity and providing some flood control. The dam holds one of the largest reservoirs in the world, with a capacity of 181 cubic kilometres.
It was built in 1956-59, during the colonial era, with World Bank funding, and design problems were soon recognised. The dam was designed by French engineer Andre Coyne and is similar in design to his Malpasset Dam in France, which collapsed in 1959, destroying two villages and killing 423 people.
Water from the dam goes through hydropower stations on both banks and water flows throughout the year. In addition, there are six giant flood gates which can be opened during the rainy season to dump flood waters. These have two problems.
The first design error became clear quickly. It was assumed that the water could simply pour out onto the river bed below the dam, but by 1962 the plunging water had already cut a hole 30 metres deep in the rock. By 1981, after the independence of Zimbabwe, inspection showed that the hole was 80 metres deep and that water was splashing back toward the dam, eroding the rock at the base of the dam.
In the early 1980s, the joint Zambia-Zimbabwe Zambezi River Authority (ZRA) introduced a new policy that no more than three gates could be opened and that adjoining gates could not both be opened. This stopped the deepening and widening of the plunge pool, and World Bank sources argue that since then the dam has been safe from collapse. But it also meant that with only three of six gates possible to use, Kariba could not dump as large a flood, so the lake water level had to be kept lower, which in turn limited the amount of electricity that could be generated.
The second design problem appeared later. The stone in the concrete should be inert, but aggregates containing reactive silica materials are common, and highly alkaline cement reacts with the silicon dioxide to form a gel that swells and weakens the concrete. The flood gate structures have now swollen and there is a danger that they could jam open (which would slowly drain the late) or closed. This chemical reaction was not well understood at the time of construction. Furthermore the concrete was not built with the kind of reinforcements which are now used in such structures.
The World Bank is now coordinating a $294 million project to cut 300,000 cubic metres of rock from the downstream side of the plunge pool and reshape it to stopped water splashing back toward the dam, and to rebuild the six flood gates.
Funding is complex: $100 mn from the European Development Fund, $75 mn from the World Bank, $75 mn from the African Development Bank (AfDB), $25 mn from Sweden, and $19 mn from the ZRA itself. The World Bank is not allowed to fund a project which benefits Zimbabwe, so it had to argue that the protecting the dam was "urgent" and of regional benefit. Money is being lent only to Zambia, which then passes it on to the ZRA.
Putting the package together caused huge delays. The United States did not support the project in the World Bank (but only abstained, so it could go ahead) and a fight between the AfDB and the European Union caused delays. Technical questions and fiduciary questions also caused delays.
Work should have started last year (2015) but the plunge pool contract will not be issued until at least November this year (2016). Work can only be done in the dry season and will take three years. The contract for the flood gates will not be issued until June 2017. One gate at a time will be repaired and the project will take eight years.
There seems no evidence that the dam is in danger of collapse, but there are serious worries about the flood gates. Clearly the 55-year old dam needs work, and the two year delay shows how hard it is to package even an "urgent" project with multiple donors and that is partly in Zimbabwe.
So has the World Bank itself used scare tactics? In attempting the rectify the mistakes of its own colonial project and put together a difficult funding package, did the World Bank raise the hype and danger warnings? It is interesting that the 24 November 2014 project appraisal document which contains all the project details and risk assessments was not made available on the World Bank website, making it difficult for outsiders to assess the project.
The key appraisal document has only just been posted, after we noted its absence, here>
All the project documents are here>
Low water: One reason that World Bank is more sanguine about Kariba is that after two years of drought, the lake behind Kariba on 28 December was only 14% full, compared with 51% a year earlier, according ZRA. This is partly caused by the inability to keep water from earlier high rainfall years due to need to keep the lake level lower. But this very low level means it is highly unlikely that the flood gates will be needed in the next two years, which could allow the work to progress faster.
Power production has already been halved, and on 7 January the Zambian Energy Minister, Dora Siliya, that a continued absence of rains could force the power plants to shut down altogether.
There is, however, no such problem further downstream at the Cahora Bassa dam. Mozambique's National Water Directorate (DNA) said Cahora Bassa reservoir is 61% full, which is enough to continue normal electricity generation.