Dismantling the Energy Per Transaction Fallacy
Bitcoin is often criticized for its energy consumption, but critics often miss a few key points. We examine the "per transaction fallacy" and show that it's based on faulty assumptions.
Of the many criticisms levied against Bitcoin, the most prominent is that Bitcoin is detrimental to the environment because it uses too much energy per transaction. The most common approach to the argument looks like this:
- Determine how many transactions Bitcoin does per day.
- Determine how much energy Bitcoin uses per day.
- Divide energy use by transactions to find energy used per transaction.
- Extrapolate how much energy would be required if Bitcoin were used for all global transactions.
- Bitcoin's energy use per transaction looks high. Shockingly high.
- Based on this calculation, Bitcoin would require exponentially more energy than VISA to process the same number of transactions—carry the two, boom—Bitcoin is bad for the environment. It's inefficient, an energy hog, and is going to lead to the destruction of the planet.
Are these critics right? Is Bitcoin inefficient? Is Bitcoin as energy-consumptive as they say? Is Bitcoin that threatening to the environment?
Not even close.
If you’re starting to check out, stay with us for a bit longer—we'll explain everything.
We want to take a purely educational approach—the purpose of this article is to analyze just one part of this Sisyphean debate. We'll go into why the per transaction discourse is a fallacious way of looking at Bitcoin's energy consumption. Specifically, we'll discuss how certain authorities extrapolate data to reach conclusions that are provably and demonstrably false.
What Does this Discourse Look Like?
Over the past few years, there's been no shortage of reports, scientific articles, studies, and news outlets that use this line of thinking. This calculation has led to real headlines like this:
In 2017, Newsweek and the World Economic Forum reported that the Bitcoin network would consume "all of the world's energy by 2020."
Another 2018 study from Nature published by Mora et. al predicted that "Bitcoin emissions alone could push global warming above 2°C."
Both use the "energy per transaction" calculation to reach their conclusions. The first is so glaringly false, as we're now in 2022, and the Bitcoin network uses 0.5% of the world's energy, according to the New York Times. The second is more complex, but its underlying logic is still highly flawed.
Why? Let's dive in.
How Is this Calculation Used?
According to a 2022 Digiconomist study, Bitcoin uses approximately 2090 kilowatt-hours of energy per transaction. VISA, on the other hand, uses only 148 kilowatt-hours of energy per 100,000 transactions.
Again, this data is high. Really high. So high that it's almost impossible not to jump to the conclusion that we should do everything we can to stop Bitcoin.
With this data, it starts to make sense how Bitcoin would use all the world's energy by 2020. It also makes sense how Bitcoin could conceivably raise global temperatures by 2°C
What the Calculation Gets Wrong
We'll put it bluntly—there is no relation between Bitcoin transactions and systemic energy consumption. None at all.
You can see this quite clearly by looking at Bitcoin's transactions over time and comparing it to the energy consumption over time.
From 2017-2022, Bitcoin's daily transactions have been range-bound between roughly 250,000-450,000 transactions per day. Meanwhile, according to Digiconomist, in that same period, Bitcoin's energy consumption has risen from 10 terawatt-hours to a whopping 200 terawatt-hours of energy per year. The energy use has been increasing but, comparatively, the transactions haven't.
It's abundantly clear that the relationship between transactions and energy consumption is not linear, and in fact, the two bear no relationship at all. They don't move with one another in a dependent relationship, which is what the "energy per transaction" calculation would have us believe. These critics employ what's called the False Cause Fallacy, where X and Y are assumed to have a cause-and-effect relationship when, in reality, they don't. What makes things more egregious is that they employ the Linear Projection Fallacy on top of it, where past trends are assumed to continue into the future regardless of any new information. Put simply, we have not only a falsely constructed correlation but also a faulty projection.
It's like saying that the number of dogs in San Francisco is dependent on the number of televisions sold (when we know it's not), and then predicting that there will be a massive increase in the number of dogs in San Francisco—simply because televisions sales have increased. It's a false assumption extrapolated to a ridiculous (and obviously false) conclusion. The argument is simply not rooted in fact.
To reiterate, Bitcoin's total transactions haven't increased considerably in years while the network's energy consumption has exploded. So if Bitcoin's energy consumption is increasing at an outsized rate compared to its transactions, it begs the question: Why? What is this energy used for? And how could this ever be considered a good thing?
What the Energy Actually Goes Toward
Again—there is no relationship between Bitcoin base layer transactions and total energy consumption. At all.
But then what does the energy go toward?
Securing the network and getting bitcoin in return.
Miners don't expend energy to validate transactions. They expend energy to earn bitcoin.
Energy expenditure is directly related to getting that reward, which some call coin issuance. Confirming transactions is just a byproduct of securing the network. Bitcoin miners could confirm 1 transaction or 100,000—it doesn't matter. As long as they win the block, they get their bitcoin.
Why is This a Good Thing?
The Bitcoin network is trustless. It doesn't rely on third parties to exist or to function. This is a system that we can all understand and verify. It's a system that doesn't require any centralized authority.
So why is it so secure? How could something with no central entity possibly be secure?
The Bitcoin network is secured by the fact that it costs a lot of money to attack. If someone wanted to double-spend bitcoin, they would need to expend 51% of Bitcoin’s total energy output—which, as we now know, is a lot. Even if an attacker achieved 51% of total hash power, they would only be profitable for a brief window of time, as they'd have to sustain those high costs continuously. Hash power can be added to the entire network, so this attacker would likely need more than 51% of total hash power to successfully control the network for any extended period.
This security creates a very high barrier to entry for would-be attackers, and that is precisely why Bitcoin is so successful. The more energy that the network requires, the more difficult it is to attack. This is essential for a trustless system that hasn't ever been hacked, successfully stopped, or taken down.
Energy expenditure is security. This is Bitcoin's security model, and it's the most secure network the world has ever seen.
But this begs yet another, more concerning question: Why are we allowing Bitcoin miners to use so much energy to process only 450,000 transactions a day? Isn't that nothing? Who cares about security if the system processes so few transactions? According to VISA, its network processes over 150 million transactions a day. That's over 36,000 times the number of transactions that Bitcoin processes. Seems inefficient, right?
The Base Layer is for Settlement, Not Payments
This element of Bitcoin is hard for some people to understand, as they view Bitcoin in its current form as the final iteration of Bitcoin's evolution. That could not be further from the truth. Bitcoin as a system, the one that miners expend energy on for coin rewards and security, the one that does about 450,000 transactions per day, the one that we've talked about in this article; it's just one thing—a settlement layer.
The point of settlement, in the context of a transaction, is to ensure that all parties involved have the amount of funds that they’re supposed to have. Officially. Settled. Done. In Bitcoin and in any other settlement layer, this ensures that there's no dispute once one party has sent funds to the other party. The base layer is not supposed to be a payment protocol—it's supposed to be a dispute resolution (settlement) protocol.
Financial settlement layers have existed for some time. We use them today, but they're buried underneath several other layers that we use for day-to-day transactions. When you go to buy your coffee with VISA, many things must come first for a transaction to be "settled."
Banks have to talk to one another to settle funds. If you've heard of things like Fedwire, SWIFT, CHIPS, or ACH, these are all examples of financial settlement layers that banks use to transfer money to one another.
Settlement layers have one primary purpose: to move large amounts of value around securely. They don't care about small transactions like buying coffee. They care about big ones, like moving millions of dollars from one bank to another.
And guess how many transactions these settlement layers process in a day? About 800,000. Not much more than Bitcoin.
Historically, we've worked around this by building faster layers on top of these settlement layers to make payments more efficient—layers like VISA.
This is how companies like VISA can process transactions so quickly. But they're not actually processing transactions on the settlement layer. They're just bouncing payments around in advance before they finalize after a few days. Your money isn't settled—that happens only after banks communicate and transfer funds through drawn-out and expensive processes. And this is why banks charge such high fees. It's to cover the costs of these intermediary layers, middlemen, and the inefficiencies of our current financial system.
We get fast transactions, but it's all clunky. And the consumers pay for it.
Bringing it Back to Bitcoin
If VISA can process 150 million transactions per day on top of a settlement layer that handles 800,000 transactions per day, one thing becomes clear—a settlement layer tells us nothing about a system’s total transactions as a whole.
It's just a base layer. A foundation.
Put simply, the number of transactions that Bitcoin can theoretically process is not constrained by the number of transactions that it currently processes. We know we're repeating ourselves a lot here, but it needs to be said—the Bitcoin base layer is just a settlement layer. And a really good settlement layer at that.
If we compare settlement systems, Bitcoin settles any amount of funds, from 20 dollars to 20 million dollars, in approximately 10 minutes. Just by virtue of its quick settlement times alone, Bitcoin is a much more useful settlement layer than anything else we currently have. Fedwire, ACH, CHIPS, and SWIFT can take days to settle transactions. And they're not even remotely as secure as Bitcoin
Lightning Network, Bitcoin's VISA
We've talked a lot about Lightning here at NOAH, and for good reason. Lightning Network is a second layer that runs on top of Bitcoin. It's similar to VISA in the sense that it allows for small, frequent transactions. But it's different in that it doesn't require a central authority to run it. It's an incredible piece of technology that allows Bitcoin to process a large number of small transactions quickly, efficiently, and without increasing the load on the Bitcoin base layer.
With the Lightning Network, an individual, group of individuals, or a company can set up their own payment network on top of Bitcoin. This network can be used to process a million transactions per-second without ever touching the Bitcoin base layer. At NOAH, we could send millions of transactions between our friends at ZEBEDEE, for as long as we want, and eventually settle on the Bitcoin base layer when we're ready. And that settlement, after an unlimited number of transactions on Lightning, would only count as one transaction on the actual Bitcoin blockchain. One.
Think about this for a second. And then think about what this means for Bitcoin's energy consumption. Think about how ludicrous it is to try to calculate an energy per transaction cost when Lightning can process in one second as many transactions as the base layer can in an entire day. Think about how this changes the entire conversation.
Comparing Bitcoin's energy consumption to VISA's without considering Lightning is not just a simple mistake. It's nonsensical. It's not based in reality.
The Bitcoin base layer will continue to process a few hundred-thousand transactions per day, much like Fedwire, but its additional layers will process millions, and even billions, of transactions per day.
Comparing Apples to Anacondas
We're not here to say we can ignore Bitcoin's carbon footprint or emancipate it from criticisms of its energy consumption. We can and should be critical. We should question where this energy is coming from, who is using it, and whether it could be used more efficiently.
We're just here to say that when we talk about Bitcoin's energy consumption, we need to be clear about what we're talking about. And we need to be clear that the energy per transaction calculation is a meaningless way to have the energy conversation.
Using energy per transaction, like the headlines we've mentioned, seeks to delegitimize Bitcoin by conflating a useful base layer with the entire system. It seeks to ignore that Bitcoin is just the base layer of a growing, maturing ecosystem of technologies. Making these comparisons is like trying to measure the number of miles per gallon a car can get by measuring the size of its headlights. It just doesn't make any sense. We're comparing a settlement layer to a well-oiled payment infrastructure. It's not comparing apples to oranges. It's comparing apples to anacondas
Wrapping it Up
The dominant narratives are flawed. Fundamentally, they're built upon a house of cards of implausible projections, false equivalencies, and bad logic. They're reductionist ways of thinking about Bitcoin's energy consumption, and if we want to have meaningful conversations rooted in facts, we have to veer away from the logical fallacies that plague these discussions.
There's a deep misconception about how Bitcoin functions, and it all starts with understanding that the energy per transaction metric isn’t how we should Bitcoin's carbon footprint. Data shows us that there is no strong correlation between transactions and energy use. As we've discussed, as second-layer solutions like the Lightning Network gain mass adoption, they will become the dominant method of Bitcoin payments. The majority of these transactions will rarely—if ever—be settled on the Bitcoin base layer, and they won't ever make their way into the flawed energy per transaction metric.
Yes, Bitcoin uses a lot of energy. But what's important is not the amount of energy per transaction. What's important is asking ourselves how important security is. What’s important is discussing its mix of renewables, the efficiency of miners, where that energy is coming from, and whether it's being used to displace more carbon-intensive energy sources.
We're going to have those discussions. Stay tuned for more on Bitcoin and the Environment and how we're going to be tackling these topics.