I’ve had the energy conversation more times than I can count. Usually it comes up at dinner, or after a talk I’ve given, and it usually comes from someone who is otherwise curious about Bitcoin and is looking for a reason to be sceptical. “But what about the energy?” It’s become a kind of rhetorical shorthand, a way of expressing discomfort with Bitcoin that sounds principled. I’m not dismissing the concern. Energy use is a legitimate topic. But the version of the argument that circulates in most conversations is badly informed, and the actual data tells a more complicated and interesting story than “Bitcoin is boiling the oceans.”
Bitcoin’s energy use is real, measurable, and significant. It is also, when examined carefully, proportionate in ways that the headlines rarely acknowledge, increasingly sourced from renewables, and functionally different in kind from most energy comparisons that get made about it. Here is what the data actually shows, as best as I can render it without pretending the uncertainty doesn’t exist.
The actual consumption numbers
The Bitcoin network consumes somewhere between 100 and 200 terawatt-hours of electricity per year, depending on the methodology used and the time of measurement. The Cambridge Centre for Alternative Finance runs what is probably the most widely cited ongoing estimate of Bitcoin’s energy consumption, the Cambridge Bitcoin Electricity Consumption Index, and its figures have generally sat in this range. To put it in country terms: the annual electricity consumption is comparable to that of Poland, or Argentina, or the Netherlands, mid-sized developed economy equivalents.
That sounds large in isolation. Most things do, at scale. The relevant question is: large compared to what, and for what purpose?
The global financial system uses significantly more energy than Bitcoin. Data centres processing financial transactions, bank branch networks, ATM networks, card payment processing infrastructure, armoured vehicle fleets, physical cash printing and transport, and the gold mining and refining industry combined consume an energy footprint that credible analyses estimate at five to ten times Bitcoin’s current consumption. This isn’t a defence of Bitcoin’s energy use, it’s context. If you’re comfortable with the energy cost of maintaining the existing financial system, the comparison is relevant. If you’re not, that discomfort should apply proportionally to both.
Gold mining alone, for context, consumes roughly 100 to 130 terawatt-hours per year, comparable to Bitcoin, while also requiring vast quantities of water, producing significant chemical waste, and disturbing land at industrial scale. The environmental case against gold mining is considerably stronger than the case against Bitcoin mining in many respects, yet gold mining is rarely the target of the kind of cultural anxiety that Bitcoin mining attracts.
Where the energy comes from
The more interesting question is not how much energy Bitcoin uses, but where that energy comes from. And this is where the picture diverges significantly from the “boiling the oceans” narrative.
Bitcoin mining has a distinctive economic property: it can be located anywhere there is cheap electricity and an internet connection. It doesn’t need to be near consumers. It doesn’t need to be in a city. It doesn’t need to follow the geography of any particular industry. This mobility means that Bitcoin mining gravitates naturally toward the cheapest available electricity, and the cheapest available electricity is frequently stranded renewable energy.
Stranded energy is electricity that is generated but cannot be economically transported or used at the point of generation. Hydroelectric plants in remote mountainous regions. Natural gas that would otherwise be flared at oil production sites. Wind farms that produce surplus electricity during off-peak hours when the grid can’t absorb it all. Solar installations in locations with more generation capacity than local demand. These energy sources are cheap precisely because they can’t easily be sold to normal electricity consumers. Bitcoin miners are willing buyers.
Industry surveys, and I acknowledge these numbers require caution because methodologies vary and self-reporting has obvious limitations, have estimated that somewhere between 50% and 75% of Bitcoin mining uses sustainable or low-carbon energy sources. The Cambridge Centre’s work on this question has produced figures in a similar range. The directional trend has been consistent: the proportion of renewable energy in Bitcoin’s mix has increased over time as miners optimise for cost and as renewable energy has become cheaper relative to fossil fuels.
Bitcoin mining as a grid tool
One of the more counterintuitive aspects of Bitcoin’s energy story is the emerging role of Bitcoin miners as demand-flexibility assets for electrical grids. Miners can ramp their consumption up or down quickly, in some cases within seconds, in response to grid conditions. This makes them useful partners for grid operators managing the intermittency of renewable energy.
When wind or solar generation exceeds demand, grid operators traditionally have to curtail generation, essentially asking producers to turn off, or accept the electricity at very low or negative prices. Bitcoin miners can absorb that excess, providing a revenue source for renewable generators at times when no other buyer exists. When grid demand spikes and electricity is needed for other users, miners can shut down quickly, freeing capacity for residential and industrial demand.
In Texas, which has a large deregulated electricity market and significant wind generation, Bitcoin miners have become significant participants in demand response programmes. ERCOT, the Texas grid operator, has acknowledged miners as useful grid resources. Similar dynamics are emerging in other markets with high renewable penetration.
In South Africa, this argument has particular relevance. Eskom’s capacity constraints and load shedding have been caused by a mismatch between available generation capacity and demand, compounded by the failure to build new generation capacity and the deterioration of existing coal infrastructure. The transition toward renewables, rooftop solar, wind, and independent power producers, creates exactly the kind of intermittency challenge that flexible demand can help manage. Bitcoin mining, structured correctly, is a form of flexible demand. That’s not a solution to Eskom’s problems, but it’s a real consideration for how South Africa manages its energy transition.
The comparison problem
The energy comparisons that circulate in popular media about Bitcoin are almost uniformly misleading, and they’re misleading in ways that make Bitcoin look worse than it is. The most common form: “A single Bitcoin transaction uses as much energy as X households for Y days” or “as much energy as watching Z videos on YouTube.”
These comparisons are methodologically broken. Bitcoin’s energy consumption is largely fixed at the network level, it doesn’t scale linearly with transaction volume. The Bitcoin network consumes roughly the same amount of energy whether it processes 100,000 transactions a day or 500,000. Dividing total energy consumption by number of transactions produces a figure that makes each transaction look absurdly expensive, while ignoring the fact that Bitcoin’s transaction layer is not where its primary settlement function lies. Bitcoin is better understood as a base-layer settlement network, think of it as analogous to the clearing and settlement infrastructure between banks, not as a retail payment system competing with Visa on per-transaction efficiency.
Comparing Bitcoin’s energy use to sending an email or streaming a video is not a serious comparison. It conflates two entirely different functions and two entirely different scales of value being transferred and secured.
The legitimate debate
None of this means Bitcoin’s energy use is without cost or tradeoff. There is a legitimate debate about whether Bitcoin produces sufficient value to justify its energy consumption, and I think that debate should be engaged with honestly rather than dismissed.
The case for “yes” is that Bitcoin provides monetary sovereignty, censorship resistance, and a fixed-supply store of value to anyone in the world with an internet connection, without requiring trust in any institution. The value it provides is difficult to quantify, particularly for people in countries with failing currencies or authoritarian governments where alternative monetary systems are not just convenient but necessary. For a South African investor watching the rand, for a Venezuelan who lived through hyperinflation, for a Nigerian who experienced sudden restrictions on dollar-denominated transactions, Bitcoin’s energy cost looks different than it does to a commentator in a stable currency country with functioning financial infrastructure.
The case for “maybe not” notes that energy consumed for Bitcoin is energy not available for other uses, and that in a world where decarbonisation is urgent and energy is genuinely constrained in many regions, the opportunity cost of directing energy toward mining is real. I don’t think this case is as strong as its proponents claim, most Bitcoin mining consumes energy that would otherwise be wasted or is at the margin of the grid, but it’s not a frivolous argument.
What is not a legitimate argument is the one that stops at “Bitcoin uses a lot of energy” without the context of what that energy is doing, where it comes from, how it compares to alternative systems, and whether the value produced justifies the cost. That’s not an environmental argument. It’s a rhetorical shortcut.
What the data actually shows
Bitcoin’s energy use is significant in absolute terms and proportionate in relative terms. It is increasingly sourced from renewables and stranded energy. It plays a growing role as a grid flexibility tool in markets with high renewable penetration. It compares favourably to the energy footprint of the financial system it is being proposed as an alternative to.
The honest question, whether the value Bitcoin produces justifies its energy cost, is one I think the answer to is yes, but I acknowledge that’s a judgment that involves values and not just data. What I’m less patient with is the version of the energy argument that stops at the headline number and never asks the next question. The next questions are where the real analysis lives, and the real analysis is more interesting than the headlines suggest.
That afternoon in a Cape Town car park when I first understood proof of work, what struck me was that Bitcoin’s energy use isn’t waste. It’s the physical anchor of digital security. Every watt consumed is a watt that would have to be matched to undo what was built. That’s not a rounding error. That’s the design.
