Green by geography

The Lake Turkana Wind Farm in northern Kenya is an impressive sight. Hooked up to the country’s grid in July 2019, the farm consists of 365 individual wind turbines which can supply more than 17 per cent of Kenya’s electricity at peak times. By far the largest wind farm in Africa, it plays a key role in Kenya’s ambition to produce all of its electricity from renewable sources by 2030 while also ensuring that the country’s population has access to cheap, reliable power. Currently, around three-quarters of homes have access to electricity, a number that is far higher than its neighbours, such as Ethiopia (56 per cent) and Tanzania (52 per cent).

Besides wind, Kenya’s other main sources of energy are currently geothermal (almost 40 per cent) and hydropower (20-30 per cent). Fossil fuels account for just 10 per cent of electricity production, and are used primarily as a backup or balancing load during peak hours or dry spells. Green by geography, this impressively sustainable energy mix offers Kenya an opportunity to use its capacity to produce cheap, clean power as an industrial policy weapon, the home of data centres and green manufacturing, rather than just a worthy boast and talking point at climate conferences.

A report in March suggested that Kenya's data centre market is poised for significant growth, projected to increase from 266 million US dollars in 2025 to 805 million US dollars by 2031. That, however, is highly dependent on Kenya’s ability to produce more electricity. A massive data centre planned by Microsoft and the UAE’s G42, first announced in 2024, is currently on hold, reportedly due to Microsoft's unfulfilled demands for a guaranteed annual minimum purchase of server capacities. Kenya's President William Ruto has also pointed to logistical hurdles, saying that the project's energy demand exceeds available resources. “We would need to switch off half the country for the data center to be powered,” he said earlier this year. This should not be read as an indictment of Kenya’s (or Africa’s) ability to cope with the huge demand for power of data centres: countries across the world are similarly struggling. The International Energy Agency (IEA) believes that from 2024 to 2030, data centre electricity consumption will grow by around 15 per cent per year, more than four times faster than the growth of total electricity consumption from all other sectors.

Indeed, the stalled Microsoft project perhaps says less about Kenya's available resources or ambitions than about how it has chosen to sell them. The country offers something genuinely rare, namely cheap, local, green power. The Microsoft deal is not misguided or necessarily wrong, but it arguably fails to make the most of what Kenya has to offer. A first phase of 100 megawatts (just a tenth of the eventual target), was never going to be supplied by an electricity grid whose peak demand has already touched 2,444 megawatts against 3,000 megawatts of capacity, and where the evenings are already sometimes dark.

The cheapest new capacity Kenya can find is the power it has already demonstrated it can generate quickly, at scale, and at the right cost.

A possible, alternative solution, and certainly a model worth scaling, can already be seen in operation in the geothermal-rich lands of Olkaria, northwest of Kenya’s capital Nairobi, where EcoCloud's Project Eagle, a geothermal-powered campus that the country' s prime cabinet secretary, Musalia Mudavadi, launched earlier in 2026, sits inside state electricity producer KenGen's Green Energy Park and draws its 60 megawatts directly from geothermal wells, not from the existing grid that provides light to Kenyan homes. Kenya would do well to add more anchor tenants of 60 to 100 megawatts, their electricity supply ring-fenced at source, rather than gigawatt projects that might end up putting the needs of data servers in direct competition with those of the country’s households. Peter Njenga, KenGen's boss, has grasped the logic, adding a steel plant to the same park in March on the argument that the same wells can serve homes and industry simultaneously.

KenGen currently uses only a tenth (or so) of the Rift Valley's estimated 10,000 megawatts of geothermal potential, so the scope is certainly there to do so. Its pipeline of 326 megawatts of new Olkaria units, a 63-megawatt revamp of Olkaria I and a 200-megawatt wind farm at Marsabit all point the right way, but arguably do so far too slowly. The political capital spent on naming KenGen operator of a future nuclear plant would almost certainly repay Kenya faster if it was spent on drilling steam wells. It can often take decades for new nuclear power plants to come online, while geothermal wells can be powering homes (and data centres, or anything else) in a much shorter timeframe. The cheapest new capacity Kenya can find is the power it has already demonstrated it can generate quickly, at scale, and at the right cost.

Back in 2007, Kenya taught the world how to move money by phone, launching M-Pesa, which turned the most basic of mobile phones into a digital wallet for tens of millions of unbanked people (more than 60 million people now use its services across Africa). Two decades on, its next reinvention could be selling reliable green computing capacity to a power-hungry planet, and few countries are better placed to do that, because so few countries are so green by geography. Focusing less on the power it currently lacks and selling the power it already has, tucked away in the ground, would appear to be the way forward.