Africa Doesn't Have an Energy Problem, It Has a Location Problem

While the world waits for Africa's grids to catch up, one Nigerian founder is betting that compute should move to the power — not the other way around.

Here's the sentence that should make every African policymaker uncomfortable: the continent is not short of energy. It is short of demand in the right places.

For decades, the story of African power has been told as a scarcity story — not enough megawatts, not enough grid, not enough investment to light up homes and run factories. It's a true story. But it's an incomplete one. Because while hundreds of millions of Africans live without reliable electricity, the continent is simultaneously flaring away enough gas every year to power entire economies. Nigeria alone flared roughly 203 billion standard cubic feet of gas in 2025, a volume that places it among the world's top flaring nations alongside Iraq, Russia, and Iran — not because the gas isn't there, but because the pipelines, processing plants, and transmission lines to move it to where people need it simply don't exist.

That's not an energy problem. That's a location problem. And in 2026, a small but fast-growing group of infrastructure builders has figured out the workaround: if you can't move the power to the demand, move the demand to the power.

The idea nobody in Lagos was pitching two years ago

This is the argument Olumayowa Ogunnusi, founder and CEO of TerraHex, made in a recent feature for the AR Initiative's Energy issue — that Africa's digital future won't be unlocked by waiting for legacy grid infrastructure to arrive. It will be built by rethinking where compute happens in the first place.

It's a deceptively simple pivot. For as long as digital infrastructure has existed, the assumption has been that compute follows power lines — you build a data center where the grid is strong, the fibre is dense, and the population is large. Lagos, Nairobi, Johannesburg. That's where Africa's data centers have historically clustered, and it's why the continent's compute capacity remains a rounding error globally.

The real opportunity isn't in the cities with the strongest grids. It's in the oil fields, gas flares, and hydro sites the grid was never built to reach.

TerraHex, which Ogunnusi founded in 2025 after rebranding from an earlier Bitcoin-mining venture, builds high-performance compute infrastructure directly on top of stranded and underutilised energy — gas that would otherwise be flared at oil and gas production sites in frontier markets. Instead of waiting for that gas to be piped somewhere useful, the company brings modular, containerised data centers to the wellhead itself, running on power that was previously being burned into the atmosphere for nothing.

How "waste gas" becomes a balance sheet

The mechanics are worth understanding, because they explain why this model is spreading faster than most African infrastructure plays. TerraHex has agreed preliminary terms with Shoreline Power for a joint venture at Afisiere in Nigeria's Delta State, where the secured power capacity for an initial proof-of-concept deployment has reached 18 megawatts, with room to scale significantly if the first phase succeeds. The commercial structure behind it, which TerraHex calls "power-for-equity," has energy partners contributing stranded power assets in exchange for equity in the project, while TerraHex supplies the capital, hardware, and operational expertise to turn that power into revenue — first through Bitcoin mining, which monetises the electricity almost immediately, and increasingly through AI inference and fine-tuning workloads as demand for sovereign compute grows.

It's a model with precedent. In the US, companies like Crusoe Energy pioneered "digital flare mitigation," converting gas that oil producers would otherwise burn off into on-site power for compute. Argentina has already adopted versions of it. And the underlying economics are getting more favourable by the year — industry estimates suggest flare-gas power can cut CO₂-equivalent emissions by more than 60% compared to open flaring, simply because engines combust methane more completely than a flare stack does.

Why this matters more in Africa than anywhere else

Africa's numbers make the case louder than any pitch deck could. The continent sits on more than 600 trillion cubic feet of proven natural gas reserves, yet consumes only a sliver of it domestically — most of it has historically been earmarked for export. Meanwhile, global data center electricity demand is projected to nearly double by 2030 as AI training and inference workloads scale, and industry analysts increasingly agree that the binding constraint on AI growth isn't chips or capital. It's power.

That combination — abundant stranded energy on one side, an insatiable and growing global appetite for compute on the other — is exactly the mismatch TerraHex and companies like it are built to close. And the timing lines up with a broader continental push: African Energy Week's 2026 programming has dedicated an entire track to AI and data centers, with African Energy Chamber executive chairman NJ Ayuk framing the gas-to-compute convergence not as a niche energy play but as an economic strategy — one that could anchor jobs, industrial capacity, and a genuine seat at the table in the global AI economy, rather than leaving Africa to simply export the raw energy inputs that power everyone else's data centers.

Today, Africa exports energy while facing chronic power shortages at home. Gas-to-compute infrastructure is one of the few models that closes that gap instead of widening it.

The obvious skepticism — and why it doesn't fully hold

None of this is without friction. Deploying enterprise-grade compute hardware to a remote gas flare in the Niger Delta is a very different proposition from renting rack space in a Lagos data center — equipment has to survive harsh conditions with minimal staffing, gas composition varies site to site, and a single rack of modern GPU servers can run upwards of half a million dollars, meaning a bad deployment doesn't just fail, it strands capital in the very location the model was supposed to unlock. Critics of the "stranded energy" thesis more broadly also point out that some of these arrangements risk crowding out other productive uses of underutilised power rather than genuinely adding new value.

But the regulatory direction of travel is not neutral on this. Nigeria has moved toward progressive penalties for routine gas flaring alongside incentives for gas utilisation, which increasingly makes hosting compute at flare sites a compliance strategy for oil and gas operators, not just an opportunistic side bet. When the alternative to monetising stranded gas is a regulatory fine, the calculus for producers changes fast.

What this means if you're building in Africa's digital economy right now

For founders, investors, and policymakers watching Africa's digital infrastructure conversation, the TerraHex thesis is a useful reframe: the continent doesn't need to win the grid-expansion race before it can compete in AI and compute. It can build compute capacity at the energy source, today, using resources that are currently being wasted. That's a fundamentally different — and faster — path to digital relevance than waiting for transmission infrastructure that may take another decade to arrive.

The bigger question is whether this becomes a genuine industrial category for the continent or stays a handful of proof-of-concept sites in the Niger Delta. Ogunnusi's bet, and the bet of everyone circling this space, is that it's the former — that stranded energy, treated correctly, isn't a liability on Africa's balance sheet. It's an asset nobody has finished pricing yet.

Africa was never short on energy. It was short on the imagination to put compute where the energy already is — and that's finally starting to change.

.