Justin Peter Gardiner Lowe  /  Commentary

What Actually Stops a Project in Africa

Three constraints, in order: power and who is permitted to finance it, the route to the coast, and whether the terms you signed still exist in five years.

Justin Peter Gardiner Lowe  •  1 August 2026

Investment committees discuss Africa as though it presented a single risk, usually labelled political and usually priced with a blunt increase in the discount rate. That habit is expensive in both directions. It kills projects that would have worked and it waves through projects that were never going to. In my experience the continent presents three constraints, they are largely independent of one another, and only one of them is political.

One: power

Roughly 600 million people in sub-Saharan Africa live without electricity, about 80 percent of the global access gap. On current policy the International Energy Agency expects 645 million people worldwide to still be without power in 2030, with 85 percent of them in sub-Saharan Africa. These are usually presented as humanitarian statistics. They are also the single most important industrial fact about the continent.

The United States Energy Secretary, Chris Wright, has made the point more bluntly than most people in development finance are willing to. He calls energy the enabler of everything that we do, and he sets the disparity out in terms that are hard to argue with: the average person in a wealthy country consumes around thirteen barrels of oil equivalent a year, the average African less than one. Whatever one makes of the politics attached to that argument, the arithmetic is not in dispute, and it has a direct consequence for anyone trying to build a mine or a processing plant.

A mine can run on diesel. Diesel is expensive, but it is a line item. A refinery, a smelter or a hydrometallurgical plant cannot run on diesel in any economic sense. It needs firm, cheap power, continuously, for decades. That is the actual reason African ore leaves the continent as concentrate and comes back as product. It is not a failure of ambition or of local policy. You cannot refine what you cannot power, and the countries with the deposits are precisely the countries where the grid is weakest.

This connects directly to the midstream problem in Western supply chains. Every serious proposal to move processing out of one dominant jurisdiction runs into the question of where the cheap firm power will come from. Gas in Mozambique and Tanzania, hydro in the Congo basin and Ethiopia, and geothermal in the Rift Valley are the honest answers. They all require capital deployed years before the plant, and none of them are financed by a mining project alone.

And who is allowed to finance it

The power constraint is not only physical. It is financial, and this is the part of the argument that tends to get left out of polite discussion. For most of the past decade the institutions that fund African infrastructure have been under instruction to withdraw from the cheapest firm power available to the continent.

The World Bank began phasing out support for upstream gas in 2017. At COP26 in 2021, thirty-nine governments and institutions signed the Glasgow Statement, committing to end new direct international public finance for unabated fossil fuel projects abroad. The signatories included the United States, the United Kingdom, Canada, Germany, France and Italy. The International Institute for Sustainable Development estimated the pledge would redirect close to 28 billion dollars a year. Whatever one thinks of the climate logic, the industrial consequence is not ambiguous: a gas-to-power project in Mozambique or Nigeria has spent years competing for capital against a mandate that formally excludes it, while the same governments continued to permit gas generation at home.

That asymmetry is the specific objection African energy ministers have been making for a decade, and it is not a fringe position. It is the mainstream view across the continent's energy ministries, and it has had a measurable effect on where capital went. Money flowed toward generation that is cheap per megawatt-hour installed rather than firm. Solar is genuinely the least expensive electricity to build today, and it should be built. It does not smelt anything at three in the morning.

The consequence closes the circle back to the first constraint. A grid that cannot carry industrial baseload cannot carry processing, so ore continues to leave as concentrate, so the value continues to be captured elsewhere, so the country remains dependent on commodity exports and remains too poor to finance its own grid. Climate finance policy and critical minerals supply chain policy have been working directly against one another for most of a decade, and they are almost never held in the same file by the same officials.

This has begun to move. The World Bank lifted its prohibition on financing nuclear power in June 2025 and has reopened the question of upstream gas, with Ajay Banga arguing that meeting rising power demand in low and middle income countries is among the most urgent development challenges there is. The Bank's Mission 300 programme targets electricity for 300 million people in sub-Saharan Africa by 2030, and that target is not deliverable on intermittent capacity alone. The direction of travel has changed. A decade of deferred projects has not un-deferred itself.

The objection to all of this deserves to be stated properly rather than dismissed. African civil society groups warn that gas infrastructure built now locks in thirty years of emissions and carries real stranded asset risk if export markets price carbon. That risk is genuine. It is also borne by the project sponsor, who is paid to take it. The cost of continued energy poverty is borne by a population that had no part in the decision and receives no premium for bearing it. Those two risks are routinely discussed as though they were comparable. They are not.

Two: the route to the coast

A deposit's value is a function of its distance from a working port, and that function is unforgiving. Trucking concentrate over a thousand kilometres of damaged road can consume the entire margin of an otherwise economic orebody, which is why so many resource statements in central Africa describe assets that are geologically excellent and commercially stranded.

The Lobito Corridor is the most consequential attempt to change this in a generation: 1,300 kilometres from the Angolan port of Lobito to Ndola in Zambia, running through the Congolese copperbelt, backed by the European Union's Global Gateway package and by United States development finance, with the first Congolese copper reaching the United States by that route in August 2024. China responded in November 2025 with a 1.4 billion dollar agreement to modernise the competing TAZARA line to Dar es Salaam.

Two observations follow. First, corridor competition is good news for producers, because a single route is a chokepoint regardless of who owns it, and flood damage that interrupted Lobito services this year demonstrated the point without anyone having to behave badly. Second, no project should be underwritten on the assumption that a corridor will be operating at nameplate by the time it is needed. Rail rehabilitation runs late everywhere, and the discipline is to model the project on the logistics that exist today, then treat the corridor as upside.

Three: whether the contract survives

The political risk that matters is rarely the one that makes the news. Coups are dramatic and occasionally irrelevant to an operating asset. What damages returns is the steady revision of fiscal terms after the capital has been sunk and the commodity price has risen.

The record of the past three years is unambiguous. Mali's 2023 code raised combined state and local ownership from 20 to 35 percent and royalties to as much as 10 percent; the dispute that followed cost Barrick a 430 million dollar settlement and, by one account, 1.9 billion dollars of revenue, while Malian gold production fell 19 percent in 2025. Burkina Faso raised its free carried interest from 10 to 15 percent in 2024 and took 35 percent of the Kiaka mine in April 2026. Niger seized the Somaïr uranium operation from Orano. Ghana introduced a sliding royalty in March 2026 that rises with the gold price. Investment disputes filed at ICSID reached a ten-year high in the first ten months of 2025.

The instinct of most sponsors is to negotiate the lowest possible state take and then defend it. That is precisely the structure that invites revision, because a government watching a foreign operator earn windfall margins at a fixed royalty will eventually act, and its population will support it. The better structure concedes the point in advance: a fiscal ratchet under which the state's share rises automatically with the commodity price, paired with a genuine local processing commitment. It costs something in the good years. It buys the only thing that matters over a twenty-year asset life, which is that nobody has an incentive to reopen the agreement.

What this means in practice

Projects that work in Africa have three things in common, and none of them is a better orebody. They have power secured before the final investment decision rather than assumed. They have a route to a port that exists today and a second one that might exist later. And they have a fiscal structure that gives the host government more money when prices rise, without anyone having to renegotiate anything.

Everything else is a deposit. A deposit is not a project, and the gap between the two is where most of the capital committed to this continent has historically been lost.

Figures cited are drawn from the International Energy Agency's electricity access data, the Glasgow Statement and International Institute for Sustainable Development estimates, World Bank policy announcements, published reporting on African mining fiscal terms, and public statements by the United States Department of Energy. Justin Peter Gardiner Lowe is the founder of Gardiner Investments.

← The midstream problemNext: the uranium contracting cliff →