Property – Construction & Engineering · Editorial
By Moakanyi Magazine · China-in-Africa · June 2026
Uganda switched on 600 MW of new generation in 2024 for a grid whose peak demand sits closer to 1,000 MW. On paper that looks like a surplus built ahead of need. That is precisely the point: Karuma is not a response to today's demand but a bet on tomorrow's industry.
China's foreign ministry lists Karuma among the FOCAC cooperation outcomes, according to Beijing. The independent record fills in the figures – and the strain – behind the headline capacity, and it is the strain that decides whether the bet pays.
600 MW, commissioned after twelve years
The Karuma Hydroelectric Power Station was commissioned on 26 September 2024, adding 600 MW to the national grid through six vertical Francis turbines of 100 MW each. China's Sinohydro Corporation, a subsidiary of state-owned Power China, executed the build over roughly twelve years – far longer than the schedule a quick capacity boost implies.
The delay is part of the story. A plant conceived to close a power gap instead took more than a decade to deliver, and over those years the gap itself moved as Uganda added other generation. The long schedule carried its own costs in interest and escalation, and it pushed the day of reckoning – when the country must actually use the power – out to a future that has now arrived.
Twelve years is a long time to wait for power a country has not yet learned to use.
US$1.4 billion borrowed, US$1.688 billion spent
China Exim Bank financed the project with a US$1.4 billion loan, with the Ugandan government covering the balance; the total cost is reported at about US$1.688 billion. Reporting also points to accrued interest on delayed payments – the predictable tail of a project that overran its timeline, where slippage on one side feeds penalty costs on the other.
Karuma lifts Uganda's installed capacity to around 2,045 MW, against a current peak demand of between 900 and 1,000 MW. The gap between capacity and demand is the wager made explicit: build the power first, and bet that industry follows it. Roughly half the new plant's output has, for now, no buyer – a stark figure when the financing behind it is US$1.4 billion in foreign-currency debt.
The loan bought megawatts; whether it bought a customer for them is the open account.
Power waiting for a load
Idle capacity is not free. Generation that runs below its potential still has to service its debt, and a 600 MW plant earning on a fraction of its output is a financing risk as much as an engineering achievement. The shortfall lands on tariffs or the treasury – either consumers pay more per unit to cover the fixed cost, or the government absorbs the gap. Uganda's case for Karuma rests on industrialisation, mineral processing and transmission build-out arriving to absorb the surplus, including the lines needed to move power from the Nile to where factories might rise.
That makes Karuma a clean test of the build-ahead model. If factories and an extended grid materialise, the early surplus looks like foresight and the unit cost of power falls as demand catches up. If they lag, Uganda is left servicing US$1.4 billion of debt on power it cannot yet sell – and the longer that lasts, the more the plant looks like capacity stranded ahead of its market.
Power built ahead of demand is foresight or stranded cost – and only the next decade decides which.
The continental version of the same bet
Karuma is one instance of a pattern that runs across the continent: large, Chinese-built hydropower commissioned ahead of the demand it assumes, on the theory that reliable, cheap power will pull industry into being. The theory is not unreasonable – power shortages have throttled African manufacturing for decades. But it loads the risk onto the host's debt rather than the contractor's, and it makes the demand-creation step the part most likely to disappoint.
For governments weighing the next dam, Uganda's experience is the cautionary detail: the engineering is the solved problem, the twelve-year schedule and the absent load are the hard ones. A turbine spins on command; an industrial base does not.
Building the megawatts is the easy half of the bet – summoning the industry to use them is the rest.
