Bitcoin's Energy Use: What the Data Actually Shows

Bitcoin's energy use is the feature, not the flaw. The electricity a miner burns is precisely what makes a confirmed transaction impossible to reverse, and the network would be worthless without it. That is the part the headlines skip. I have spent years watching South Africans hesitate over a number they have never once seen placed next to the systems Bitcoin actually competes with.

Why Bitcoin uses energy at all

Every ten minutes the network adds a block to its ledger, and to win the right to add it a miner has to feed data through a hash function billions of times until an answer fits. There is no clever shortcut. It is brute computational effort, and the effort costs electricity by design. That cost is not a bug someone forgot to optimise away. It is the whole point.

Proof-of-work anchors digital truth in physical cost. To rewrite a single old transaction an attacker would have to redo the work for that block and every block stacked on top of it, faster than the entire rest of the network combined. In practice that is not merely hard. It is economically suicidal, because the hardware and the power bill would cost more than anything the attack could steal, and the reward for playing honestly quietly evaporates the moment you try to cheat.

So the electricity buys something specific. It buys finality that no government, no company and no court can undo. Traditional settlement leans on banks and central counterparties and lawyers to enforce that finality. Bitcoin swaps all of them for physics.

That swap is the whole trade.

The total computing power pointed at the network is called the hash rate, and it is a direct reading of how costly an attack would be. Higher hash rate, higher wall. In September 2025 that wall crossed one zettahash, a thousand exahashes a second, a level that would have sounded absurd a few years earlier. Energy and security do not merely correlate here. They are the same quantity measured two ways. Each new block layered onto the chain makes revising history a little more expensive again, so the network does not become harder to attack because it grows more complicated. It becomes harder to attack because it grows more expensive to lie to. I set out the full mechanism in proof of work, explained.

The comparison most critics skip

A single number, stripped of anything to compare it against, tells you almost nothing. Bitcoin is often cited at between 100 and 200 terawatt-hours a year, roughly a Poland or a Netherlands. The sentence stops there as if the case were made. It has not. The real question is never how much energy Bitcoin uses. It is what that energy secures and what the alternatives cost.

Take gold, the asset Bitcoin most resembles as a store of value. Gold mining draws electricity in the same broad range, roughly 100 to 130 terawatt-hours a year, and then adds the parts nobody puts on the poster: diesel for the generators and the earthmovers, the water and chemicals that leach the ore, the open pits that never grow back. Bitcoin's footprint is almost entirely the electricity. There is no hole in the ground.

Now hold it against the system Bitcoin is actually trying to replace. The legacy financial machine runs on bank data centres, branch networks, card rails, the SWIFT clearing infrastructure that moves money between institutions, and layers of redundancy few people ever see. There is a further cost most estimates never dare to price at all: the military protection of the global oil flows that underwrite the dollar itself. Set that beside a settlement layer whose security you can verify from an ordinary laptop and the double standard becomes hard to miss.

The awkward part is that the comparison is almost never run. Estimates put the financial system's own consumption at several times Bitcoin's. The branches, the card networks and the cash-in-transit fleets get treated as wallpaper rather than an energy cost, so they never reach the same page as the Bitcoin figure. One number gets a headline. The other gets ignored because we grew up inside it.

Bitcoin settles irreversible transactions every day of the year, with no weekend and no holiday. Judging its power bill without once weighing it against the machinery it competes with is not analysis. It is a press release.

Where the electricity actually comes from

Miners are not sentimental. Electricity is their single largest cost and they run on margins thin enough to make a supermarket blush, which forces them to hunt down the cheapest power on earth, and the cheapest power is almost always power nobody else can use.

Stranded energy is generation with no line to carry it to a city: remote hydro, a wind farm producing more than the grid can absorb, natural gas flared off at a wellhead and burned into the sky for nothing. Curtailed energy is power a grid makes but cannot distribute at that moment, usually a midday glut of solar or a gusty night of wind. Both are waste unless something shows up to consume them on the spot. A mining rig can, because it can sit anywhere and switch on or off in seconds.

That flexibility turns a supposed villain into a grid asset. In Texas, where a large slice of the network's mining lives, operators routinely power down during peak demand and cold snaps, releasing electricity back to households exactly when the grid is stretched thinnest. They are often the first load to curtail, not the last. A factory cannot do that. A smelter cannot do that. A Bitcoin miner does it before breakfast.

Picture that logic dropped onto our own map. A wind farm off the Cape coast on a howling night, producing more than the lines can carry, is producing waste unless a buyer exists who can appear beside it and leave the instant the grid needs the power back. That buyer is almost impossible to find in heavy industry, which needs to run continuously to pay for itself. A container of mining rigs is exactly that buyer, and it is the rare industrial load a fragile grid can switch off without breaking anything. I am not arguing South Africa should race to mine. I am pointing out that the technology critics fear is, in the right hands, a tool for turning stranded generation into something rather than nothing.

None of this makes Bitcoin "green," and I would not insult you by claiming it. The renewable and wasted share is real and it is rising, but it rises for a hard economic reason rather than a marketing one: free and stranded power is the cheapest power, and the cheapest power is what a rational miner chases. The geography keeps shifting too. A few years ago something like three quarters of the network's mining sat in China; a ban there scattered it across the world, and the United States now holds the largest single share.

What this means for South African investors

Here is where the local version of the argument usually arrives, and it deserves a straight answer. Our grid is fragile. Eskom's failures are not a talking point, they are a Tuesday, and a decade of load-shedding has taught every South African to treat spare electricity as sacred. So when someone hears "Bitcoin uses as much power as a country" the instinct to bristle is entirely reasonable in a place that has queued for its own.

But that instinct is aimed at the wrong target. Mining and owning are two different activities that happen to share a name.

When you buy Bitcoin through a South African broker you are acquiring coins that already exist. You are not plugging in a rig. You are not drawing a single extra watt from Eskom. The mining happens on the other side of the planet, in Texas and Ethiopia and Paraguay, and it happens whether you hold one satoshi or none. The energy debate is a real and interesting question about the network's global security. It is simply not an argument against Bitcoin as a place to keep your savings, any more than the diesel burned in a Free State gold mine is an argument against wearing a wedding ring.

The reason a rand-earning saver should care about that distant energy is precisely that it works. The miners burning power in a dozen jurisdictions are what make the settlement guarantees credible for someone in George or Sandton who wants a store of value the Reserve Bank cannot quietly print more of. That is the same case I make in Bitcoin as sound money, and it is why the properties matter more than the power bill. If you are still forming a view on what the asset even is, I would start with what Bitcoin actually is before worrying about its electricity.

There is also a fairness point worth naming plainly, because South Africans hear it least often. Gold has been a respectable savings asset in this country for generations, dug from some of the deepest and most energy-hungry shafts on earth, and nobody has ever told a client to feel guilty about a Krugerrand. The selective outrage says more about unfamiliarity than about physics, a pattern I trace in why investors stopped respecting gold and commodities.

My honest position, then. The energy question is a fair thing to ask and a poor reason to sit out. Ask it, get a real answer, then decide on the asset's merits rather than on a number someone handed you without a ruler. In fifteen years of these conversations I have watched the energy objection do its real work not as a reason but as an excuse, a comfortable place to stop for someone who was never going to move anyway. The genuinely curious ask the question, hear the comparison, and get on with the actual decision of whether a fixed-supply asset belongs in their savings. If you would like that answer applied to your own situation, book a call and we can look at your position properly rather than in the abstract.