Originally published: August 2021, updated January 2023
Bitcoin’s energy usage has been in the news for years.
It’s often criticized for using too much energy, or not making efficient use of its energy, or in extreme cases, being an outright climate/energy disaster.
For example, back in December 2017, Newsweek ran a piece called, “Bitcoin Mining on Track to Consume All of the World’s Energy By 2020“:
The World Economic Forum ran similar headlines as well, because these various organizations were all referencing the same flawed sources.
Well, here in 2023, I can obviously say that didn’t happen. The University of Cambridge is the most-cited source for estimating Bitcoin’s energy consumption, and they place its estimated annualized rate at under 100 TWh currently. Even at its estimated peak, it was under 150 TWh.
Since the world uses over 176,000 TWh of energy per year, that means that the entire Bitcoin network, at its peak consumption level, uses less than 0.1% of the world’s energy consumption. That’s for a network with 100+ million estimated users worth hundreds of billions of dollars. Other estimates vary slightly but they arrive at similar numbers overall.
Bitcoin’s energy usage is a rounding error as far as global energy usage is concerned. And I mean that literally; when scientists estimate that the world uses a certain amount of energy in a given year, they can easily be off by a couple percentage points in either direction, let alone a couple tenths of a percent. Bitcoin uses less than than one tenth of one percent.
Plus, a large percentage of the energy that Bitcoin uses is from otherwise-wasted energy sources.
In the very long run, if Bitcoin is wildly successful and becomes a systemically important asset and payment system used by over a billion people at 10-20x its current market capitalization, it could reach several tenths of one percent of global energy usage, and maybe upwards of 1% with aggressive modeling. On the other hand if it is unsuccessful, and doesn’t grow much from current levels, its energy usage will stagnate and shrink as the block subsidies continue to diminish. I’ll dive into those numbers later in this piece.
The point is that the Bitcoin network is and forever will be a rounding error as far as global energy consumption is concerned, whether it’s successful or not, and its energy usage won’t exceed its long run utility (however high or low that utility ends up being).
In fact, as someone with a background in electrical engineering, part of why I was drawn to Bitcoin in the first place was by seeing how efficiently its network uses energy. I could have picked any blockchain to invest in, or could have avoided the digital asset space altogether, or could switch my investment to another blockchain at any time. And yet, when I crunch the numbers, I think Bitcoin makes particularly elegant usage of energy, and is getting more energy efficient over time. Plus, there is no substitute for proof-of-work, which requires energy.
If that’s the case and Bitcoin’s energy usage is practically irrelevant on the global scale, how can journalists make such large, sensationalized errors? How can they literally be off by three orders of magnitude, and think that Bitcoin will use 100% of energy while instead it ends up using 0.1%? The answer is that it often comes down to them not understanding the scaling process that the network is going through, or basically understanding how any of it works at all.
And it’s easier to sensationalize things for pageviews or political gain. For example, it’s commonly said that the Bitcoin network uses more energy than some countries. That’s true, but then so does Google, Youtube, Facebook, Amazon, the cruise industry, Christmas lights, household drying machines, private jets, the zinc industry, and basically any other sizable platform or industry. From that list, Bitcoin’s energy usage is the closest to that of the cruise industry’s energy usage, but bitcoins are used by more people, and the network scales far better. If people were 10% more efficient at shutting off their electronic devices when not using them, then that alone would save more energy than the global Bitcoin network uses.
It’s important to understand whether or not Bitcoin is an environmental problem and whether or not it is energy efficient, because these obviously affect its probability of being a good investment and a good payment network.
If Bitcoin did indeed have serious energy scaling problems, it would eventually fail against competitors in the private marketplace by not offering enough utility for its expensive energy consumption.
From an engineering perspective, Bitcoin’s energy usage isn’t a problem when you actually run the numbers, but it takes an understanding of how it works in order to 1) estimate its long-term energy usage reasonably, and 2) what the trade-offs are if you use a different approach than what bitcoin uses.
Let’s take a look at how that works. You can start from the beginning or jump to the chapters of interest to you.
- Understanding Bitcoin’s Purpose
- Why Bitcoin Uses Electricity
- Bitcoin’s Efficient Scaling Pattern
- How Bitcoin Uses Otherwise-Wasted Energy
- Bitcoin’s Proof-of-Work vs Alternative Methods
- Final Thoughts and Summary
Understanding Bitcoin’s Purpose
Before we dive into its energy usage, it helps to summarize what bitcoins are used for, and what problems the network was designed to optimize for. From there, we can then look at the energy usage and decide whether or not it’s achieving its goal.
Bitcoin is not trying to be the cheapest payment network, although in many contexts it ironically can be, when you consider the Lightning network (one of Bitcoin’s second-layer scaling systems).
Instead, Bitcoin is trying to be a decentralized bearer asset and payment/settlement network, and for 14 years has been succeeding.
In terms of the asset, bitcoins are something that users can self-custody with encryption if they want to, and that have a fixed supply. This makes bitcoins resistant against dilution and difficult to confiscate. In terms of the payment network, Bitcoin can be used to send payments that don’t rely on the permission or verification from any centralized entity. This makes the network censorship-resistant and interoperable with many different payment systems internationally.
Almost anything that claims to be more efficient than the Bitcoin network at making payments or storing assets for you, is also more centralized. Centralized things inherently tend to be efficient. Sending payments between parties and tracking peoples’ balances can be as simple as updating a centralized internal database, which is nearly free.
The problem, of course, is that centralized things don’t tend to be resilient, and so there is a trade-off between efficiency and resiliency. When you put all your eggs in one basket, give that basket to your friend, and she drops the basket or refuses to give it back, you’re out of luck. Relying on a centralized entity cedes control to others, which rests on the premise that those others are moral and competent. Plus, as far as payment networks go, they’re tied to fiat currency, which is useful as a medium of exchange but over the long run doesn’t hold its value.
Bitcoin is a public distributed ledger with a series of private and public keys. As long as they have access to the internet (including satellite internet if need be), users can send bitcoins (including fractional bitcoins) to others by using their private keys. They can hold their private keys offline and can receive bitcoin when offline; they just need an internet connection to send them, confirm their balance, and those sorts of actions.
This ranks bitcoins among the most portable and hard-to-confiscate assets in the world. They can be sent internationally between different parties, can be memorized and brought anywhere in the world, and can be stored offline with no counterparty. Considering that they also have a supply cap, it’s not surprising that folks in many emerging markets have turned to them, rather than relying entirely on a local unsound currency that may be inflating at 10%, 20%, 50%, or 100%+ per year in some cases. Bitcoin is global money spread across a decentralized cloud.
Even in developed markets, which have lower rates of inflation, people have to deal with negative inflation-adjusted interest rates. For example, the following chart shows the interest rate of 1-year US Treasuries minus the prevailing 1-year official inflation rate. When it’s below zero, it means that Treasuries (and doubly-so for bank accounts) are not paying enough interest to keep up with the official rate of inflation:
Chart Source: St. Louis Fed
While some countries can partially ban Bitcoin by disallowing banks to send money to crypto exchanges or by banning large-scale bitcoin mining, it’s nearly impossible to stop peer-to-peer transactions, especially in countries with reasonable property rights and freedom of expression. Banning individual interaction with Bitcoin is akin to banning information or speech, since it’s just an open source public ledger and people can memorize numbers to access it, and can use satellite connections to bypass local internet service providers if necessary. It would require a very authoritarian approach with draconian enforcement to truly stamp it out of use.
On the other hand, some countries have embraced it, such as El Salvador that made it legal tender. This is in significant part because El Salvador is highly reliant on remittance payments, and it’s nearly free to send remittance payments via the Lightning network that runs on top of the Bitcoin network, since all of it is open source and keeps getting more developed over time. A number of other countries that have previously banned or were about to ban it have stepped back their bans.
My article on bearer assets included a large section on the practical use-cases of bitcoins, both in developed and developing countries. Aside from being an asset with a fixed supply cap, here are some example use-cases of the censorship-resistant payment and portability aspect:
1) Reuters reported in February 2021 how Russian opposition leader and anti-corruption lawyer used bitcoin as a censorship-resistant payment, as Putin’s authorities block all of the traditional permissioned payment rails:
Russian authorities periodically block the bank accounts of Navalny’s Anti-Corruption Foundation, a separate organisation he founded which conducts investigations into official corruption.
“They are always trying to close down our bank accounts – but we always find some kind of workaround,” said Volkov.
“We use bitcoin because it’s a good legal means of payment. The fact that we have bitcoin payments as an alternative helps to defend us from the Russian authorities. They see if they close down other more traditional channels, we will still have bitcoin. It’s like insurance.”
2) The Guardian reported in July 2021 about how Nigerian merchants and protest groups used bitcoin’s censorship-resistant payment attributes to go around FX currency blocks to carry out their business, and to receive funds even when their bank accounts were suspended.
The clampdown was financial too. Civil society organisations, protest groups and individuals in favour of the demonstrations who were raising funds to free protesters or supply demonstrators with first aid and food had their bank accounts suddenly suspended.
Feminist Coalition, a collective of 13 young women founded during the demonstrations, came to national attention as they raised funds for protest groups and supported demonstration efforts. When the women’s accounts were also suspended, the group began taking bitcoin donations, eventually raising $150,000 for its fighting fund through cryptocurrency.
3) Alex Gladstein, with the Human Rights Foundation, discusses the human rights angle of the Bitcoin network in April 2021:
So for Venezuelans to go through this has been nothing short of totally heartbreaking. However, there are a lot of folks who got involved early, like earlier, a lot of young people, a lot of young folks figured it out in 2015, ’16, they were mining at the time. One of these guys I interviewed, he actually helped start Ledn, which is one of the larger industry services now based out of Canada.
But he and his brothers were mining bitcoin for a couple of years there and they ended up having to escape. The government came with like an armed squad and seized all their mining equipment. Thankfully no one got hurt. The government was very perplexed. They saw the mining equipment and they thought that they meant they got the bitcoin. But that’s not how it works. So they were able to use the bitcoin to start a new life in Canada. I thought that was really amazing.
I interviewed another guy who escaped to Argentina. He got involved in some sort of dispute with the government where they claimed he was a criminal, even though he wasn’t, but now he’s able to send money back to his mother who is in Venezuela in bitcoin. She uses it to support herself. There’s just so many stories like this.
And I think for me, one of the most powerful things as someone whose family went through the Holocaust, was this idea of like, you could flee your country and back then you only brought what was on your back, like the clothing on your back. But today you can bring your wealth with you, which is truly remarkable.
And I’ve given advice to people who are, for example, leaving Iran these days, help them out on this. People are selling their homes and they’re converting to bitcoin and they’re getting on a plane, getting the heck out, and they’re bringing it with them in this digital format. People are sending money in and out of Syria to people who are stuck there.
People have escaped countries like Sudan. I interviewed a guy from Sudan, which has a horrible inflation problem that their inflation is in the hundreds. They have an inflation rate of something like 150 or 200%. And he’s living in Europe and he’s sending the hardest money around back to his family in Khartoum. And they’re able to get by through that. We’re early here. Again, I think the estimate based on Coinbase numbers is that maybe 10% of Americans have interacted with Bitcoin or cryptocurrency.
The global number is lower than that, especially in these emerging markets. If the global is two percent and America’s 10%, it’s probably way less than that in a lot of these emerging markets, but hey, in some of them, look at Turkey, man, look at Argentina, look at Nigeria. These are huge countries, 200 million, 100 million, 45 million people. They’ve got the highest per capita usage in those countries. So it’s a changing world.
Perhaps more broadly and concisely, one of Alex’s best quotes about the topic is this:
We’re at the outset of great digital financial transformation, where the money we use on a daily basis is evolving from a bearer asset—one that doesn’t reveal anything about us—into a mechanism of surveillance and control.
This is more urgent for some people in this world, and maybe less urgent for others, depending on the political regime they live under. When I’m looking at this new form of money that’s not controlled by governments or corporations, I’m thinking about the big picture of today’s world, where we have 4.2 billion people living under authoritarianism and 1.2 billion people living under double- or triple-digit inflation. When we talk about the fact that money is broken, this isn’t theoretical, and it isn’t just about one country.
It’s much bigger than that. This is a world where hundreds of millions of people deal with 15 percent, 20 percent, 25 percent inflation, where their time and energy, and the currency that they earn their wages in, is literally disappearing.
At the same time, you have billions of people whose bank accounts can potentially be frozen based on their opinions or ideas.
To summarize, as digital global money with a set of censorship-resistant payment rails, Bitcoin is the application of software towards finance. And not just towards the surface layer like fintech, but towards the root layer of bearer assets and settlement networks, which is part of why it’s so controversial.
The firm Chainalysis, which is used by law enforcement to track public blockchains, has found across multiple years that only 0.50% to 2% of crypto transactions are for illegal activity, such as scams, ransomware, or drug purchases (which happens to be lower than most estimates for the percentage of fiat currency transactions that are for illegal purposes). The vast majority of the usage is estimated to be for investment purposes or legal payments.
As described above, there are many authoritarian regimes where protesting and certain types of speech are considered illegal, and where basic economic interactions can be blocked, and so we could say that bitcoin facilitates those types of “moral yet illegal” behaviors. Oppressed people having access to open source software to aid in basic human freedoms and economic interactions, making it more complex for authoritarians to deal with them, is not what I would consider unethical, and indeed quite the contrary.
Why Bitcoin Uses Electricity
The Bitcoin network is programmed to create a new block on average every ten minutes and add that block to the blockchain, which consists of hundreds of thousands of blocks since inception in 2009.
A new block is produced by a bitcoin miner (a specialized computer) solving a cryptographic puzzle that the previous block created, and the miner can package thousands of bitcoin transactions currently in the queue, into that block. That’s how transactions get settled. The network is programmed to target average block times of ten minutes, meaning on average every ten minutes a block of thousands of transactions is added to the blockchain.
If miners drop off the network and new blocks on average start taking longer than ten minutes to produce, the network is automatically programmed to make the puzzle easier by a quantified amount, so that blocks go back to an every-ten-minute average schedule. Likewise, if a lot of miners join the network and blocks get added to the blockchain faster than every ten minutes on average, the network will make the puzzle harder. This is known as the “difficulty adjustment”, and is one of the key programming challenges that Satoshi Nakamoto solved to make the network work properly.
So, at any given time, there are millions of bitcoin mining machines around the world looking to solve the puzzle and create the next block, and there’s a natural feedback mechanism to ensure that blocks are created on average every ten minutes, regardless of how many miners are on the network.
Back in 2021, China banned crypto mining, and approximately half the network went offline and started moving elsewhere. Bitcoin’s payment network briefly slowed down a bit, but otherwise kept working with 100% uptime. The difficulty adjustment then kicked in, and brought the network back up to its target speed. Imagine if Amazon was told with one week’s notice that it had to move half of its server capacity internationally; it would likely experience uptime issues for its services for the rest of the year as it moved and rebuilt half of the system.
If a miner creates an invalid block, meaning one that doesn’t conform to the consensus rules of the node network, the network discards it. If two miners produce a valid block at around the same time, the winner will be decided by which one gets found by the rest of the network first and has another block produced and added onto it, becoming the longer (and thus official) blockchain.
This process is known as “proof-of-work”. Millions of machines are using electricity to apply processing power to solve cryptographic puzzles left by the most recent block. This may seem like a waste of energy, but it’s what keeps the system decentralized. Work is the arbiter of truth, in this case. There is no central authority that decides what constitutes a valid block or a valid set of transactions; the longest blockchain is verifiable at any given time, and is recognized as truth by the rest of the network based on code. The longest blockchain is the one with the most work put into it, and that also meets the consensus criteria that the node network checks.
The more energy that Bitcoin’s network uses, the more secure it is against most types of attacks. Many of the tiny non-Bitcoin blockchains have been victims of 51% attacks, where a single entity temporarily or permanently gains control of over 51% of the processing power on the network, and uses that majority of processing power to re-organize blocks and perform double-spend transactions (which is essentially theft).
This chart, for example, shows Bitcoin’s network processing power compared to the processing power of some of its hard fork copycats:
Chart Source: BitInfoCharts.com
Both of those other blockchains only have 1% or less of the Bitcoin network’s total processing power, and have been hit by malicious block re-orgs. In fact, if just 1% of bitcoin miners decide to do a 51% attack on either of those two hard forks, they can. The same is not true for the other direction, since it is the Bitcoin network that has a far larger network of miners and energy usage than them, by two orders of magnitude.
That shows the importance of network effects in the blockchain industry, and why Bitcoin’s energy usage has kept it uniquely secure.
When someone asks, “can’t you just copy Bitcoin?”, that’s why the answer is “no”. You can replicate the open source code, but you can’t replicate the fact that tens of thousands of people are running full nodes, you can’t replicate the fact that millions of ASIC miners are securing the Bitcoin network and not your copycat network, and you can’t replicate the fact that thousands of developers are working on making the Bitcoin network better every day rather than working on your copycat network. And Lightning’s number of open channels and liquidity can’t be easily replicated either; it took years to build.
Trying to copy Bitcoin would be like if I copied the content from Wikipedia and hosted it on my website. Technically it could be done, but it wouldn’t do much. It wouldn’t gain the real Wikipedia’s traffic, because it wouldn’t have the hundreds of millions of links pointing to it from other websites. And it wouldn’t be updated like the real Wikipedia, because there’s no way I could convince the majority of those volunteer editors to come work on my version instead. Unless I could somehow succeed in the herculean task of convincing the majority of the network to move over to my version, it would always just be a shadow of the real one with a tiny fraction of the value.
Bitcoin’s Efficient Scaling Pattern
When Bitcoin was created, it was designed so that every ten minutes when a miner produces a new block of transactions, the miner that produced it earns 50 bitcoins. After four years, it was pre-programmed to drop to 25 new bitcoins per block. Four years later, it was 12.5 bitcoins per block. Four years after that, in the current era, it’s down to 6.25 bitcoins per block.
This pattern will continue every four years until new bitcoin generation asymptotically approaches zero, and the hard cap of 21 million bitcoins is reached sometime after the year 2100. Miners will earn a vanishingly small number of fractional bitcoins for producing new blocks within a few decades. Out of the 21 million, 19.2 million bitcoins have already been created.
So, Bitcoin was highly inflationary in the beginning, but it has an increasingly disinflationary monetary policy until it approaches outright zero inflation, and its miner revenue scales similarly.
However, miners also earn transaction fees. Senders pay transaction fees, denominated in fractional bitcoin, to ensure their transaction gets into the blockchain in a timely manner.
In the early days, blocks were often not full, so transaction fees were minimal. However, as Bitcoin became more widespread, blocks reliably became full, and transaction fees became a small but more meaningful part of miner fees.
Here is a table of the Bitcoin network’s average market capitalization, annual miner revenue (including block subsidies and transaction fees), and the percentage of the market capitalization spent on miner revenue each year:
Bitcoin miner revenue has grown at a historically strong rate, but the network always spent a smaller percentage of its market capitalization on miner revenue than the year prior. This isn’t a decision by any centralized party; it’s a combination of the algorithm, the value of the network, and individual miner decisions whether to mine or not based on the economics of the situation.
That’s what journalists and other people who don’t understand the algorithm often miss: the declining block subsidy. This results in Bitcoin’s inflation rate going down, along with miner revenue as a percentage of Bitcoin’s total market capitalization until it reaches more of a steady state based on transaction fees alone.
Ironically, some analysts and critics of the network are concerned that Bitcoin won’t use enough energy to remain secure in the future when it relies mostly on transaction fees. I don’t view that as a primary concern either, but it has a higher likelihood of being a problem than the reverse situation of the network using too much energy (which by design is almost impossible; it can only use energy if people are getting a lot of utility out of it).
Since miners spend most of their costs on electricity and need to remain solvent, that miner revenue represents the high-end for how much energy the Bitcoin network is using in dollar terms. In reality it is less than that, due to miners having considerable other expenses and sometimes making a profit.
Next, we can look at just the portion of the miner revenue that comes from transaction fees, which is a subset of the previous chart:
We can see that transaction fees are a tiny portion of Bitcoin’s market capitalization each year. The highest year in percent terms was 2017 during the bubble peak. Efficiency improvements such as Segwit and transaction batching have been made since 2017, so even in the heart of the early 2021 bull run, the network didn’t reach those levels again.
In response to high fees in 2017 and 2021, exchanges and users made more efficient usage of block space:
In addition, from mid-2018 through mid-2020, Bitcoin’s OP_RETURN feature was heavily used, which drove up fees somewhat even in a bear market:
Chart Source: Transaction Fee Info
OP_RETURN allows for the insertion of arbitrary data into the blockchain, which can be useful for some purposes but can also be spammy in low-fee environments. Other blockchains can use it to secure themselves by embedding what are basically “checkpoints” into the Bitcoin blockchain, and there was a two-year period where that was very popular. In other words, a considerable percentage of the fees in 2018, 2019, and 2020 were not associated with moving bitcoins around, but rather were for other non-bitcoin purposes that eventually migrated to other blockchains or just went away.
The Bitcoin network is now down to less than 2% of its market cap being spent on miner revenue each year, including a fraction of one percent on fees. In 2024 there will be another block subsidy halving, which will probably bring miner revenue down closer to 1% of market capitalization.
In 2028 there will be yet another block subsidy halving, and another in 2032. After that point, the block subsidy will be so tiny that a large portion of miner revenue will be made up of transaction fees, and miner revenue will likely be less than 1% of market capitalization, approaching some steady-state situation based on fees that is hard to model (since it depends on overall network adoption, utility, and velocity).
Since we can’t know for sure what the steady state will be due to variable market-driven transaction fees, here’s a table of potential long-term Bitcoin market capitalizations (vertical axis) and annual miner revenue as a percentage of market capitalizations (horizontal axis) in the future:
If Bitcoin fails to grow for one reason or another, and becomes a stagnant project or permanently remains around its current market capitalization of less than $1 trillion, its miner revenue will significantly decrease from current levels as block subsidies diminish. By the 2030s, bitcoin miner revenue will probably be around 0.50% of market capitalization or less, and so the network will be stuck at 2018-2020 energy spending levels or less.
If Bitcoin becomes systemically important, let’s say $5-$10 trillion (representing a per-coin price of $250k to $500k) with hundreds of millions of users, then at 0.50% annual miner revenue relative to market capitalization, that would be $25-$50 billion. This could represent perhaps 0.3% of global energy usage.
If we say it reaches a very high price of one million dollars per coin, for a critically important market capitalization of $20 trillion or more, with billions of users, then at 0.25% to 0.50% annual miner revenue relative to market capitalization, that would be $50-$100 billion. This could represent perhaps 0.6% of global energy usage, which seems appropriate for a network used by billions of people for multiple purposes, as it would need to be at that point in order to reach such a high value.
By that point, it would be big enough that it’s likely replacing energy used by parts of the global banking system. There are tens of millions of people working in banks and fintech companies around the world, with office buildings, office equipment, payment servers, and more. The application of software to money at the root layer, just like other industries, brings efficiencies and reduces the need for employment and equipment and real estate in certain parts of legacy infrastructure, freeing up those human resources and corresponding energy usage for other productive purposes.
We can also re-run these figures for annual transaction volumes rather than market capitalization, which is likely a better way of looking at it. Volumes are more closely tied to the fee market than market capitalization is, but the true volume is difficult to measure unlike the market capitalization. So, showing both market capitalization and volume can be helpful, and we can see where they might diverge. Historically, bitcoin’s monetary velocity (annual transaction volume, divided by market capitalization) is rather volatile:
The above volume figure excludes obvious change sent back to the sender’s address, but still includes a lot of noise. There is a more conservative calculation of volume by Coin Metrics that excludes various short-term transaction hops (therefore factoring out a lot of exchange cold storage shuffling and the usage of mixers). The velocity using this adjusted volume estimate has averaged 5x to 10x, and has been less volatile:
Here is a table of the Bitcoin network’s adjusted transaction volume, annual miner revenue (including block subsidies and transaction fees), and the percentage of the annual adjusted volume spent on miner revenue each year:
Here’s the same table, but only including the transaction fee portion of miner revenue:
We can see in that chart that fees were pretty miniscule until 2017. That was when blocks started to become full on a regular basis, and a meaningful fee market developed.
Now that a lot of block space efficiency gains have been realized and transaction spam has been reduced, I expect that future increases in transaction volume will create more persistent upward fee pressure. This will likely lead to a higher structural average transaction fee, but these fees would still be a very low percentage of the average transaction size.
Here’s a table of potential long-term Bitcoin annual adjusted transaction volumes (vertical axis) and annual miner fees as a percentage of annual adjusted transaction volumes (horizontal axis) in the future:
It’s hard to say what the Bitcoin network’s on-chain velocity will be in a couple decades, and what its annual on-chain volume will be. Right now, bitcoins are primarily used for investment/savings purposes, and so on-chain velocity is pretty low. If we reach a pivot point where Bitcoin is more integrated into the global payments network, the exchange rate becomes less volatile, and more people use bitcoins for payments, velocity could increase by quite a bit, resulting in some of the higher-end energy usage estimates coming to pass (perhaps upwards of 1% of global energy).
At the current time, the Bitcoin network’s electricity consumption is estimated to emit less CO2 than random things we don’t think about, like tumble driers or zinc production:
Source: Nic Carter, Demystifying Bitcoin
If Bitcoin becomes wildly successful with trillions of dollars of utility for users, we could potentially see it consume an amount of energy per year that is comparable to aluminum production. In other words, despite reaching a massive scale and serving numerous purposes, it would still be comparable to various other random industries.
The network would need to become truly massive to surpass aluminum production and surpass 1% of global energy usage, and with the uniquely flexible way that bitcoin miners interact with the electrical grid and stranded resources, even that high-energy scenario wouldn’t be a bad thing.
Scaling By Layers and the “Cost Per Transaction” Fallacy
The Bitcoin network can do a maximum of a few hundred thousand base layer transactions per day. That’s about five transactions per second. This number has increased slightly over time due to occasional upgrades that improve transaction density.
This transaction limit is often unfavorably compared to a network like Visa, which can process tens of thousands of transactions per second.
Due to that, critics often point out that Bitcoin’s energy usage per transaction is very high, and thus the network is inefficient and should be avoided for ESG reasons. There are two problems with that reasoning, however.
The first problem with that reasoning is the fact that Bitcoin historically uses energy whether or not transactions are occurring, due to the block subsidy. The way to think about it is that a large portion of that energy is used simply for securing the network against transaction censoring or deep block reorgs, and keeping it attractive as a settlement network and store of value. One block might have 1,200 transactions. The next block might have 2,500 transactions. The block after that might have 1,800 transactions. Meanwhile, the same number of miners are hooked up to the network between those subsequent blocks, verifying the blocks, and paying for electricity. Whether blocks are full or not, they’re using roughly the same amount of energy.
Whether you choose to make a transaction or not does not materially change how much energy the Bitcoin network is using at that time. Bitcoin’s energy usage comes from miners earning the block subsidy and average transaction fees, and is denominated in bitcoin and thus based on the value per bitcoin, which mainly comes from people holding bitcoin as a store of value, not spending it. Transaction volumes only affect the transaction fee portion, and only the longer-run average transaction fee matters.
Think of it like running your dishwasher each night. Whether it’s 50% or 90% full when you run it, it still uses about the same amount of resources per run. The marginal extra dish or utensil doesn’t materially affect the dishwasher’s energy usage.
Another analogy would be keeping your computer on all day, and either sending 20 emails or 100 emails. The marginal amount of energy “per email” that you send isn’t relevant, because regardless of how many emails you send that particular day, your computer is on and using approximately its baseline resource level.
The second problem with that reasoning is the idea that this limit of about 5 transaction per second is the true limit, which it is not. In reality, the Bitcoin network has multiple layers, just like the current financial system. It’s a settlement network.
Visa is merely a layer on top of a deeper payment network. In the United States, for example, we have the Fedwire system which settles approximately one quadrillion dollars of gross volume per year. That’s the gross settlement layer that banks use to perform large transactions with each other. This system performs approximately the same number of transactions as Bitcoin per year and has scaled up slowly as needed, but those Fedwire transaction amounts are very large, representing millions of dollars each. On top of that layer, there are things like Visa, PayPal, Venmo, people writing physical checks to each other, and so forth.
Source: FRB Services
If you send me a credit card payment, for example, that seems pretty instantaneous to both of us, but in reality it’s not. When the transaction seems finished to us, in reality our two banks just conversed and made an IOU between themselves. Sometime later, they will batch it with many other consumer transactions and settle their books with a big settlement transaction. There’s no limit to how many surface-layer transactions can occur, because there is no limit to the size of those massive settlement transactions. Each settlement transaction represents thousands of smaller payments.
Similarly, the Bitcoin network has additional layers: Lightning, Liquid, RSK, custodial ecosystems, and more. However, unlike the banking system that depends on long settlement times and IOUs, many of Bitcoin’s layers are designed to minimize trust, via software.
Method 1) Lightning (Second Layer)
The Lightning network is a trustless smart contract layer that runs on top of the Bitcoin network and is suitable for smaller transactions.
Like the Bitcoin network itself, nobody “owns” the Lightning network; it’s a set of open source standards that multiple companies make open source implementations of. Those Lightning transactions are instantaneous and nearly free, and have no upper limit in how many can occur per second as the network continues to grow. It went active in early 2018, and by early 2021, it reached enough channel liquidity, with major apps and exchanges starting to use it, that it became truly functional and reasonably mature. It has been enjoying steady growing in the two years since then.
Chart Source: Lightning Explorer
With Lightning, two people can open a channel with each other using a base layer transaction, and then send any number of instant transactions between each other. Days, weeks, months, or years later (whenever they want to), they can close that channel with a second base layer transaction. That means dozens, hundreds, or thousands of mini-transactions can be combined into two base layer transactions. It’s like keeping a bar tab open and settling at the end of the night, or the end of the month, except it doesn’t rely on trust but instead relies on programmed smart contracts that ensure the bar tab is settled.
In addition, you can send a Lightning transaction through the network to someone else. If Alice has a channel open with Bob, and Bob has a channel open with Cody, then Alice can send a payment to Cody through Bob as the intermediary, even though she doesn’t have a channel with Cody. And it’s all trustless; all based on programmed smart contracts using the Bitcoin base layer as the settlement assurance.
If someone has a couple Lightning channels open with well-connected counterparties, she could send and receive many transactions to/from any number of people through the network (including all the vast numbers of people on the network that she doesn’t have any channels open with), and settle them in a small number of base layer transactions when opening or closing those channels.
To give an idea of how efficient Lightning is, there were some gaming demos at the 2021 Bitcoin Conference in Miami involving “sat streaming”, and THNDR Games processed 13,571 Lightning transactions during the conference for an average fee of 1.4 sats per transaction, or roughly $0.0005 USD per transaction.
And as previously mentioned, layer three developments allow the Lightning network to be used for secure, decentralized, and censorship-resistant data transmission as well.
Method 2) Federations (Side Chains)
Liquid is an open-source federated side chain of Bitcoin, meant for large entities like exchanges to settle bitcoins with each other in a faster and less expensive way than with base-layer transactions. Beyond that, it has many other uses as well.
With Liquid, bitcoins get locked up into L-BTC tokens with a “peg-in” transaction, and those L-BTC tokens can run on this side chain with faster settlement times until an entity decides to unlock the bitcoins from that chain and bring them back to the base layer with a “peg-out” transaction.
A federation of nodes runs the Liquid network. The Liquid network can’t fully replicate the rock-solid security and decentralization level of the bitcoin base layer, but it has much faster transaction throughput, and has a reasonable degree of security and decentralization.
In other words, an entity can make a known trade-off to lock up some bitcoin, and get L-BTC tokens in return, and those tokens use bitcoin as their foundation of value, but provide additional speed and features. Between each peg-in and peg-out transaction, L-BTC will typically be used for many transactions.
Similarly, any smart contract platform can lock up bitcoin in a similar way that Liquid does. The largest one, Wrapped Bitcoin or “WBTC” consists of bitcoins that are wrapped in an Ethereum token, and can therefore trade in Ethereum-based DeFi ecosystems. There are also DeFi projects that run on Bitcoin, using wrapped bitcoins in their ecosystems as well.
There is no limitation to how many federated sidechains can be built, allowing bitcoins to move faster and cheaper, but with some counterparty trade-offs.
Method 3) Exchanges and Neo-Banks (Custodial Relationships)
Custodial crypto exchanges and bitcoin platforms are also scaling methods. When millions of people trade on Coinbase, for example, those are not millions of base layer transactions. Those are transactions within Coinbase’s centralized database. Only when people deposit or withdraw bitcoins to/from Coinbase, or Coinbase sends bitcoins to or from cold storage or to another exchange in a big batch, would there be an on-chain transaction.
As another example, millions of people use Cash App to buy and hold bitcoins. They can send bitcoins to another Cash App account, for free. Those are not base layer transactions; those are merely transactions within Cash App’s database. It only becomes a base layer transaction if they deposit or withdraw their bitcoins to/from Cash App with an external source, or if Cash App moves its users’ bitcoin around internally, to or from cold storage in large batches.
Unlike Lightning and Liquid, however, those types of custodial accounts like Coinbase and Cash App require that you trust the centralized platform, similar to how you trust your bank.
There has been interesting development lately in open source protocols like Fedimint that allow for the creation of small, private, community-based custodian arrangements. With that technology, small ecosystems can set themselves up anywhere in the world, as like modern community banks.
If Bitcoin continues to be successful and grows in users and market capitalization, base layer transactions will become increasingly used for settlement transactions, not day-to-day transactions. The fee market is what modulates its usage.
If transaction fees are averaging $1 for the base layer because block space isn’t been heavily used, for example, then $100 amounts are fine to do base layer transactions with. However, if transaction fees are averaging $50 for the base layer, those small transactions would be better with Lightning. There are many transactions worth hundreds of thousands or millions of dollars on the base layer, and they don’t mind paying a $50 fee or higher.
Ultimately, Bitcoin is comparable to Fedwire, as a massive settlement network that can do about 5 transactions per second, with no limit to transaction size, and thus no limit to how much value it can settle per day. On top of that, multiple layers ranging from trustless Lightning, to semi-trusted Liquid, to fully-trusted custodial platforms, can account for hundreds of thousands of transactions per second, with no upper limit. As a result, the cost and energy per transaction can become negligible over time.
The Tech Stock Comparison
When an investor is picking through early-stage unprofitable tech companies to invest in, they need to model out the future to ensure that, if successful, its revenue will scale faster than expenses.
Consider a new social network, for example. It will typically be unprofitable or minimally-profitable in the beginning because it requires high base expenses to pay a team to run the service. Even if it has zero users in the beginning, it still has those base expenses to get it started and running.
However, if it’s successful and well-managed, it becomes cheaper to add each user. Expenses still do grow over time (more employees, more facilities, more servers, etc) but if the company is well-designed, those expenses should grow more slowly than users and revenue, which leads the company towards profitability and robust profit margins.
In other words:
Bitcoin is not a company, but it is an efficient network. By design, its expenses scale more slowly than its utility, due to its declining block subsidy that eventually results in a security and usage model based only on transaction fees.
We can’t know for sure how much energy Bitcoin will ultimately use, since we don’t know how many people will use it, how they will use it, or what the fee market will look like a decade or more in the future. If it’s successful it will consume more energy than if it’s not successful, but due to its declining block subsidy, its utility will greatly outpace its energy usage in either scenario. That’s the key takeaway.
How Bitcoin Uses Otherwise-Wasted Energy
Beyond a simple calculation of how much energy Bitcoin uses, we should also consider the details for how it uses energy and the types of energy it uses.
People often imagine bitcoin miners competing with other industries for electricity, as though bitcoin mining must push out some other use of electricity. However, because bitcoin miners inherently require extremely cheap electricity sources, they can’t normally compete with normal users of electricity. As a result, bitcoin miners seek out inefficiencies around the world where electricity is being underutilized and wasted.
The vast majority of energy consumers can’t go to where the energy is; the energy has to be brought to them. Humans organize themselves based on geography, mainly around shipping channels. We live in coastal or riverside cities, and in the suburbs of those areas, and around rural areas of fertile land. Not around energy.
We don’t move to where the oil and gas and uranium deposits are; we send folks out to go get the oil and gas and uranium and bring it back to us for consumption in our homes and at gas stations and nearby nuclear stations.
Bitcoin miners are unusual energy consumers in the sense that they can go to wherever the energy source is, as long as they can get some sort of basic internet connection, including a satellite connection if needed. That means they use energy in quite efficient and unusual ways.
Fidelity’s first digital asset analyst and later founding partner of Castle Island Ventures, Nic Carter, described Bitcoin’s energy usage in an insightful way back in 2018:
An interesting externality of PoW coins – they are always-willing energy buyers at 3-5 cents/kWH. And some of the best energy assets are off the grid. This global energy net liberates stranded assets and makes new ones viable.
Imagine a 3D topographic map of the world with cheap energy hotspots being lower and expensive energy being higher. I imagine Bitcoin mining being akin to a glass of water poured over the surface, settling in the nooks and crannies, and smoothing it out.
Although some miners use cheap forms of traditional energy, here is a sampling of some of the novel ways that bitcoin miners use otherwise stranded or unwanted energy to the benefit of themselves and their counterparties.
1) Stranded Hydroelectric Power in China
For a long time, China has been the largest bitcoin mining jurisdiction. At one point, Chinese miners were estimated to account for over 70% of the network, but by spring of 2021 it was estimated to have gradually dipped to under 50% as more competition arose elsewhere. Then, a 2021 ban on Chinese bitcoin mining, likely in order to enforce their capital controls, has sharply reduced Chinese bitcoin mining exposure, and those miners have gone elsewhere.
However, for many years, China was an interesting example of bitcoin mobility. The province of Sichuan has a ton of overbuilt hydroelectric capacity. During the wet season, they produce more clean electricity than they can possibly use, and so it would otherwise be curtailed, wasted. It is stranded energy.
Since bitcoin miners can go to where the energy source is, they used to flock to Sichuan during the wet season to make use of that otherwise wasted energy. Not because they are altruistic environmentalists, but simply because it is cheap and nobody else is making use of it. Electricity that would otherwise be wasted and generate no revenue for the operator, can be sold for extremely cheap levels to someone who can find a use for it.
Here is an estimation by Cambridge of how much each Chinese province contributed to China’s overall bitcoin hash rate throughout the seasons. Sichuan is in yellow at the top:
Chart Source: University of Cambridge, Annotated by Lyn Alden
With the Chinese bitcoin mining ban that came shortly after the end-date of this chart, that hash rate and billions of dollars worth of annual revenue is moving to North America and other countries. But for many years, this was a great example of bitcoin miners mopping up stranded and wasted energy.
2) Flared Gas Bitcoin Mining
Many types of petroleum deposits come with associated natural gas.
If there is a sufficient quantity of this gas, it can be collected and sent via pipeline or other transport networks to be used as a primary energy source, since of course natural gas is extremely useful for electricity and heating. However, if it’s a small amount, then it’s not economic enough to build a pipeline or otherwise collect that gas.
So what happens? It gets vented or flared into the atmosphere, and therefore wasted. Venting means just letting it out into the atmosphere, mainly as methane (which is a stronger greenhouse gas than carbon dioxide). Flaring means it is burned, and thus converted into carbon dioxide and emitted into the atmosphere. A complete waste, either way, and yet still contributing to global greenhouse gases.
In terms of scale, the US Energy Information Administration estimated in its 2020 natural gas annual report that 1.48 billion cubic feet of natural gas was vented or flared per day on average in the United States throughout 2019. That’s about 150 TWh of energy for the year, which is higher than the estimated total peak level of Bitcoin’s annualized energy usage in 2021, according to the University of Cambridge. The University of Cambridge has also estimated that global flare gas recovery potential is 8x larger than the bitcoin network’s energy usage in 2021.
In other words, virtually the entire Bitcoin network in its peak 2021 form could hypothetically be run off of stranded natural gas in the US, let alone the rest of the world.
There are several private bitcoin mining companies that specialize in hooking up trailers of bitcoin miners to oil producers with stranded gas, to make use of that otherwise-wasted energy.
It’s a win/win scenario for producers and the bitcoin miners. The producers get to sell their gas rather than waste it, while earning higher ESG scores and meeting state flaring limits. Bitcoin miners get a super cheap source of energy in the process. Alternatively, some of these bitcoin miners are also willing to sell mining systems to oil and gas producers and set the hardware up for them, so that the producer can directly capture any potential upside in bitcoin’s price.
North Dakota in particular is a major area for stranded gas mining. According to the EIA, nearly 20% of produced natural gas is flared in North Dakota, rather than collected. That wasted gas alone is equivalent to tens of TWh of electricity generation per year. Back in 2014, North Dakota’s wasted amount was more like 35%, and the state implemented rules to try to get that number down. Bitcoin miners can soak a lot of this otherwise wasted energy up, and bitcoin mining is indeed growing in that state.
I reached out to Marty Bent, the Director of Business Development for Great American Mining (editor’s note: at the time of this article’s initial publication), to understand their operations. Great American Mining is a (editor’s note: now-acquired) private company that works with oil and gas producers to deploy mobile bitcoin mining rigs onto the site of oil and gas production, to use their wasted gas to mine bitcoin.
Source: Great American Mining
Marty gave a useful clarification on why pretty much only bitcoin miners can make use of this stranded gas, rather than similar industries like datacenter server farms:
Since the Bitcoin network is a distributed peer-to-peer network that doesn’t depend on any one miner to facilitate transactions, bitcoin miners are better positioned to take advantage of the flare gas opportunity compared to other energy intensive compute processes like server farms because they can stomach disruptions in the field without affecting the uptime of the network materially. Whereas a server farm would not be able to because uptime disruption could seriously affect critical business operations. Beyond this, the amount of data that miners send to mining pools is very small and doesn’t require much bandwidth, so they can operate in very remote areas using cellular data much more trivially than other energy intensive data processes.
-Marty Bent, July 2021, via email
I asked him which types of producers are better suited than others. His answers are the ones in chillier climates (current generation bitcoin miners need good air cooling, which uses energy), and/or in jurisdictions that have more stringent flaring requirements. And due to various limitations, on-shore locations tend to be better-suited than offshore locations:
The opportunity is particularly large in jurisdictions with strict flaring regulations because producers are highly incentivized to reduce their flaring. If they flare too much they are forced to shut in their wells for a period of time.
Yes, shale producers certainly have an advantage over offshore drillers as the electrical requirements and permitting necessary to operate offshore are much more restrictive than onshore drilling. On top of this, the amount of surface area to drop containers and generators on offshore operations is extremely small compared to onshore well pads, so scaling could be an issue. Beyond this, producers in states with relatively cool climates have an advantage, at least in the short to medium term until immersion setups mature. Also, producers who own the natural gas minerals and the production are better positioned because they wouldn’t have to deal with the headache of paying out royalties to mineral rights owners.
-Marty Bent, July 2021, via email
In my view, bitcoin miners using stranded gas (and stranded hydro power) are about the cleanest miners out there. It’s better than solar panels, wind energy, or similar sources, because the energy they are using is literally wasted and sent into the atmosphere otherwise, by companies that are extracting petroleum for other purposes.
3) Landfill Gas Bitcoin Mining
As a prior section showed, associated methane gas found with oil deposits is often wasted, either by venting it into the atmosphere or burning it into carbon dioxide without. It contributes to the accumulation of atmospheric greenhouse gases and represents wasted, stranded energy.
Another big way that methane leaks into the atmosphere is through landfills. An enormous amount of methane leaks out of landfills from decomposing organic matter. Some of the largest landfills capture this methane and use it to generate useful energy, but many smaller landfills around the world, or even large landfills that don’t have economic usage for it, just let it leak into the atmosphere.
A company called Vespene Energy raised $4.3 million in mid-2022 to deploy bitcoin miners at landfill sites to make use of this methane. By capturing methane and burning into carbon dioxide, it creates usable energy and reduces the overall greenhouse effect. And the only way to make the incentive work, is if there is some energy buyer that is flexible enough to set up at small and mid-sized landfills. Enter, bitcoin mining.
BERKELEY, Calif., Aug. 9, 2022 /PRNewswire/ — Vespene Energy, a methane mitigation company, today announced the close of a $4.3M financing round led by Polychain Capital, and joined by a number of climate-focused funds. Vespene installs highly efficient micro-turbines on municipal landfills to convert waste methane into electricity to power a variety of on-site uses, the first of which will be Bitcoin mining data-centers. Vespene’s immediately deployable, and highly scalable technology, enables municipal landfill operators to monetize an otherwise stranded asset while reducing harmful greenhouse gas emissions.
As Vespene describes on their homepage:
We leverage landfill methane to create an on-site energy source that can support broad EV fleet electrification and other variable facility loads. By pairing energy generation with interruptible data processing, we ensure that methane is fully destroyed and the energy is always put to beneficial use.
“Interruptible data processing” in this context is a description of bitcoin mining, and likely phrased in that way to avoid triggering people who would otherwise be interested in their services but have incorrect negative connotations about Bitcoin. Bitcoin mining is just a form of computation and data processing as it relates to transaction ordering, and unlike other uses of data centers, individual bitcoin miners are highly interruptible and flexible.
Some people ask, “can’t we capture this methane without bitcoin miners?” and to that I say, “coulda, shoulda, woulda… but didn’t”. Without proper economic incentives, it just doesn’t happen. The only way it happens for some of the smaller landfill sites is for highly flexible profitable consumers of electricity to come in and make use of it on site, and for that purpose bitcoin is the purest option.
4) Bitcoin Mining as a Grid Battery
Electrical grids have to compensate for two things: changing supply levels and changing demand levels.
Some electrical sources are very consistent, like baseload nuclear power, which can run 24/7. Other sources, like wind and solar and to some extent hydro, are more variable based on what Mother Nature feels like providing in terms of wind, sun, and rain during a given timeframe. Due to this partial variability, electrical supply needs to be overbuilt, so that even on a particularly “low” day of supply generation, it’s still sufficient to provide power to the community.
Moving to the demand side, there are certain days that require more electricity than others. Looking at my gas and electric bill, for example, I use a lot more gas in winter than in summer, because in the summer it’s only used for cooking while in the winter it’s used for cooking and heating. Meanwhile, I use way more electricity in the summer, since I’m using it for air conditioning in that season, in addition to using it for lights and electronics consistently throughout the year. Plus there are peak days, such as the most dangerously hot day of a given year, where just about every single household has the air conditioning system on full blast. Days like that need to be designed for.
So, due to variability on both the supply side and the demand side, electrical grids need to be overbuilt, and to have a lot more power generation capacity than is used on an average day. Some of that capacity could be variable, like natural gas peaking plants that can be rapidly turned on or off as needed. Other types might be ones they can’t control, like solar panels and wind turbines. If you overbuild solar capacity and wind capacity and aren’t using the excess or selling it to another grid, you just waste it.
One of the problems with solar and wind power is that the cost of storage is very high. Despite all of our human ingenuity, we still can’t make very cost-effective batteries at a utility scale. It’s an extraordinary hard physics problem. We can of course make storage batteries for certain ideal conditions, but it’s not cost effective to use them very broadly. And they require a lot of battery metals.
Bitcoin mining makes it profitable to overbuild renewable sources of energy production, since it allows that surplus supply to be monetized. Every community that wants reliable power needs overbuilt electric capacity anyway, and for wind and solar and hydro that’s even more important because they are variable. However, overbuilding is usually not very cost effective, unless you can use it for something profitable and useful when it’s not otherwise needed.
Bitcoin miners are a unique solution to that problem, can make overbuilding profitable, and thus play the indirect role of an energy storage solution.
During the vast majority of the time when there is more supply than demand, bitcoin miners as one of the electricity consumers in the community can power their machines, earn revenue, and pay their electricity costs. If there is a surge in electricity demand or a reduction in supply that would otherwise cause brown-outs in the region, those bitcoin miners can temporarily shut off.
A well-structed commercial rates contract can make this work smoothly. The utility could offer the miner the lowest possible rate in the area, in exchange for them having a higher tolerance for variability and other points of contract flexibility.
Harry Sudock, VP of Strategy at a bitcoin mining company called Griid, explained this to Peter McCormack on his podcast in June 2021:
Curtailing is not the position you ever want to be in as the energy generator. So, let’s use a wind turbine as a really easy example. The turbine goes around once, generates electrons. The market price in some regions is negative, so what they’ll choose to do is just not to send the energy anywhere. It dissipates.
So, if they’re able to strike a deal with another bidder on that energy who can tolerate some intermittent consumption, can use it some of the time, not the other part of the time, that’s a really valuable customer to be able to bring to a market that isn’t necessarily able to support the energy generation on a broader basis.
So, I think bitcoin miners are special and are a huge technological upgrade from the traditional consumers of electricity. We have two, I think of as “energy superpowers”: the first one is that energy is 80% or 90% of our monthly costs; the second is that we can consume on an intermittent basis without harming our business model particularly. So, if someone tells me I need you to shut off your miners 100 hours a year, or 500 hours a year, we don’t say no, we just say, “We need to reflect that in the energy price we pay”.
So, when I’m looking to negotiate a power contract, the way that I frame this is, “I need you to get me the lowest possible cost that you know how to offer. I’m willing to negotiate on every other part of the profile of the load. How big are we going to build the mine; how often do you need that power back; do you need us to serve any other creative purpose within your energy mix or system; do you need us to split our facility into two and to go locate at two different points within place? Great. Do we need to be able to contribute to the security budget of these other pieces of the operation?”
Our job is to drive that energy price as low and competitive as possible and work with producers on every other variable.
For clarity, I would add that their third superpower is their ability to co-locate with the source of electricity generation, and thus cut down on transmission losses to help keep their electricity cheap. Bitcoin miners are unique in that 1) almost their entire operating expense is electricity, 2) they can tolerate intermittent consumption, and 3) they are flexible with their location. As a result, they can sacrifice variables that most other companies cannot, in exchange for rock bottom electricity prices when electricity is abundant.
Due to their ability to go to the source of power, bitcoin miners can also fill in unexpected holes in demand, or other special situations. In that podcast, for example, Sudock described this situation:
This is an anecdote that we’re in the midst of working through right now. A community is zoned to have a new hospital built in their area. There are 17,000 energy customers, house to house, in that utility’s jurisdiction. They are going to bring a hospital that is going to double the amount of energy that this region pulls down. We can all agree that a hospital is a very worthy use of electricity, no argument there.
They rebuild the transmission lines, they build a new substation that’s bigger, that can handle the additional load, and after they do all that, the hospital project falls apart.
So they’re left having invested millions of dollars in this area to attract this large customer. They now have to pass that cost back to those 17,000 households, unless they can find another use for that energy. So what did they do? They called our VP of Energy Management and said, “We’ve got an overbuilt supply here. If we don’t bring in a large scale energy customer, these costs are going to get passed to these households that don’t have the budget to support rising energy prices.”
So, we have this beautiful opportunity to come in, backstop this utility, provide a customer to come in on the back of this deal falling apart, and provide the backbone to this community and to stabilize their energy prices for a decade to come. And so these are the stories of bitcoin mining that don’t get to rise to the surface. It also happens that this energy source is over 60% carbon free.
This is why Jack Dorsey, CEO of Block Inc (SQ), has made the controversial statement that the Bitcoin network incentivizes renewable power. Square (a subsidiary of Block Inc) put out a white paper on the subject in spring 2021. One of the summary points from the paper was this:
Bitcoin miners are unique energy buyers in that they offer highly flexible and easily interruptible load, provide payout in a globally liquid cryptocurrency, and are completely location agnostic, requiring only an internet connection. These combined qualities constitute an extraordinary asset, an energy buyer of last resort that can be turned on or off at a moment’s notice anywhere in the world.
Bitcoin has historically been too small and niche for grid engineers to incorporate it into their plans, but if bitcoin miners become more regular and visible, they can be incorporated into grid designs and rate markets more thoroughly.
If the Bitcoin network gets big enough, every “always on” source of electricity generation can have bitcoin mining equipment co-located with it, to profitably soak up the variable difference between the supply of that power and the surrounding grid demand for that power, so that it doesn’t go to waste. That lowers the cost for the producer and the consumer of that power, and can make renewable energy sources more cost effective.
Author update, December 2021: In the Bitcoin Mining Council’s Q3 2021 Briefing, the CEO of the largest Bitcoin miner, Marathon Digital Holdings, stated that he believes co-location is the future for the industry and that his company has been in talks with several of the largest North American power providers about the topic of co-locating bitcoin miners with power generation, and that those companies have research teams in place for the subject. In November 2021, a startup company called Lancium raised $150 million to build bitcoin mining facilities with the specific design of stabilizing Texas’ energy grid, which received positive attention from one of the state’s senators.
Author update, January 2023: The full year of 2022 showed multiple instances of large publicly-traded bitcoin miners in the United States performing significant curtailment during peak load days, in both hot and cold periods. In December 2022, it was reported that the Tokyo Electric Power Grid would begin using bitcoin mining as a way to monetize surplus energy. The thesis of bitcoin miners as flexible loads continues to be proven and expanded upon over time.
This is why I wouldn’t consider it a bad thing even if Bitcoin did exceed 1% of global energy usage; by that point it would be highly optimized into the energy system and likely be a net good for the energy system, rather than a net bad.
Bitcoin mining is a highly commoditized industry. The only way for a bitcoin miner to remain solvent over the long run is to use the cheapest sources of electricity, and the cheapest sources are the ones that are otherwise stranded or wasted. During a temporary bitcoin bull run, bitcoin miners can get away with using just about any source of electricity and remain profitable, but over multiple bull/bear cycles, the trend is clear: only the cheapest sources of electricity are viable for bitcoin miners that wish to remain solvent cycle after cycle. They’ll need to make agreements with grids to help balance the load, and/or be co-located with the electricity producer to monetize surplus electricity, and/or they’ll need to make use of various geographically stranded energy resources, such as associated natural gas or landfill emissions.
5) Advancing New Energy Technologies
There is a form of renewable baseload power called Ocean Thermal Energy Conversion or “OTEC” for short. It was first successfully demonstrated a century ago, but hasn’t taken off.
As large amounts of solar energy strike the world’s oceans, it heats the surface significantly compared to the colder depths. This column of warmth represents a way to generate electricity, and also represents a natural energy storage medium. Using a giant tube and a specific energy conversion process, large platforms or ships can tap into this difference between warm surface water and cold deep water to generate baseload power. They can use this to generate electricity directly, or use the electricity to produce liquid fuels.
The problem, so far, has been about scaling. Small OTEC plants have been demonstrated for test purposes, but are uneconomical. Medium-sized OTEC plants don’t make much sense either, because they require a lot of infrastructure to get that power back to land. Large OTEC plants do make economic sense, but nobody will build a large one before medium-sized ones are properly tested.
However, medium-sized OTEC plants are expected to be economic if they don’t need to send their power back to land. If they can be free-roaming offshore and go to the warmest parts of the ocean near the equator, then even at that middle size they can potentially be economical. The problem is that there isn’t any use of electricity out in the middle of the ocean… or at least there didn’t use to be. If they load up the OTEC ship with bitcoin miners, then they can monetize the electricity out in the ocean, prove the concept, and hopefully get funding for the large-sized platforms that can power coastal populations.
A company called OceanBit is attempting to do just that. With bitcoin mining, they believe they can rejuvenate OTEC as an active area of research and development. One of their co-founders also developed a method to incorporate bitcoin mining into the OTEC process, by using some of the cold water that they are pumping up to keep the miners cool, and joining the hot waste heat from the miners into the warm water that they’re already using. I spoke with the co-founders Nate Harmon and Michael Bennett in depth about their work, and find the technology fascinating and, at least over the long run, very promising.
As another example, a company called PRTI converts wasted tires into hydrocarbon commodities. The world produces over a billion wasted tires per year, made out of hydrocarbons, and the majority of them are just burned or buried when they wear out.
PRTI developed a unique sealed boiler process to take those wasted tires and boil them down into their various hydrocarbon commodities, and sell those commodities. However, they also produce some natural gas in this process, which they can use to generate electricity. Since their locations tend to be wherever the tires are sent to rather than in dense population centers, their local electrical grid doesn’t generally have very much use for that electricity. And it’s not enough spare gas to build a pipeline or otherwise use it for many purposes.
So PRTI takes that extra natural gas that they generate, and mines bitcoins on-site with it. As DCD reports:
Product Recovery Technology International (PRTI), is processing discarded tires at a site in Franklinton, north of Raleigh, to create oil, syngas, carbon char, and steel. It is then using the gas to generate electricity which it uses in bitcoin miners in its on-site office. The process is scalable, the energy could be used for other purposes, and the company is planning to roll it out in other sites including Europe.
The funny thing is that even that article gets it wrong in the end. They went and stuck a random anti-bitcoin section later in the piece:
Of course, making Bitcoin isn’t an environmental benefit in itself, as cryptocurrencies simply burn energy to produce abstract value. Crypto mining uses more energy than a country the size of Argentina, and has a very large carbon footprint because not all the energy they use comes from renewable sources. Even when it is done with renewable energy, it diverts that energy from other uses, thereby increasing the carbon footprint of the human race.
If that author had researched the subject deeper, he would know why PRTI mines bitcoin with their excess energy rather than send it back to the grid. As PRTI co-founder and former CEO Jason Williams explained on the Investor’s Podcast Network, the local grid providers offer an extremely low price per kWh to buy energy from PRTI, since they don’t need that energy, because these plants are located near waste sites, not in population centers. Market prices dictate lack of demand like that. This stranded energy that PRTI creates from recycling tires that would otherwise be burned or buried is not “diverting” energy from other uses.
Instead, PRTI merely keeps the relatively small amount of energy that they created, which would otherwise be wasted or sent non-economically into the grid, and mines bitcoin. This helps keep their business profitable, so that they can continue grow to do the good work of recycling tires, and cutting down on that massive global source of pollutants and litter. Bitcoin mining just happens to be the most economic use of the bit of energy they produce from their innovative process.
6) Bootstrapping Developing Country Electricity
Many low-income countries have a lot of energy resources, including hydroelectric capacity. However, they often have a chicken-and-egg problem. It’s often too expensive for them to build the electrical transmission and distribution infrastructure to move that power from where it is generated to where it would be consumed. And they can’t get a lot of capital, because there is not a lot of productive capacity that is sitting there ready to use electricity.
Bitcoin presents an interesting opportunity for some of these developing areas to build out their electrical capacity and generate revenue. If an energy source is developed, bitcoin miners can come in and give that site immediate profits as a guaranteed anchor tenant buyer, until that electricity is put to better use. As Alex Gladstein of the Human Rights Foundation described:
Billions of people in developing nations face the stranded power problem. In order for their economies to grow, they have to expand their electrical infrastructure, a capital-intensive and complex undertaking. But when they, with the help of foreign aid or investment, build power plants to try and capture renewable energy in remote places, that power often has nowhere to go.
In many countries across Africa, for example, there are vast solar, wind and hydro resources. These forces could drive economic activity, but local communities and governments usually lack the resources to invest in the infrastructure to kickstart the process.
Foreign donors and investors are not keen to support projects that do not have a pathway to sustainability or profits. Without strong transmission lines to deliver energy from harvest points to population centers, power plant builders could wait years before they can run without foreign subsidy.
Here is where Bitcoin could be an incentives game-changer. New power plants, no matter how remote, can generate immediate revenue, even with no transmission lines, by directing their energy to the Bitcoin network and turning sunlight, water or wind into money.
As local authorities or customers gradually link up to the power plant, and are willing to pay more for the energy than what miners can afford, the Bitcoin load is lowered, and communities can grow. In this way, economic activity and renewable grids can be bootstrapped by Bitcoin mining. And international aid could provide the spark.
-Alex Gladstein, May 2021
In the 2020 Shareholder Letter for Stone Ridge, the founder and CEO Ross Stevens presented an interesting perspective on bitcoin’s energy usage. He described how bitcoin mining is the first profitable use of energy in human history that doesn’t need to be situated near human settlements. Instead, it can be situated near sources of untapped energy, which allows for the development of infrastructure and provides an incentive for people to come and settle near the untapped energy source.
As he put it:
In doing so, Bitcoin can fundamentally change the economics of energy by introducing a highly profitable use of electricity that’s location independent. The world has never had a profitable use of energy that’s location independent. Now it does. And since fossil fuels are already too expensive to be a profitable source of Bitcoin mining energy, I believe the only long-term, profitable Bitcoin mining will be powered by hydro.
Imagine a future with Bitcoin mining firms, unsubsidized, in extraordinarily isolated locations – visualize a waterfall in a largely population-free part of an African country suffering from abject poverty – easily connected to the Bitcoin network, building serious energy infrastructure to monetize the local clean energy source for mining. However, once the industrial-strength, profitable infrastructure is in place, let’s extend it. Let’s build roads. And housing. And schools. And hospitals. Ultimately leading to human settlement.
The net result can be people locating around new, Bitcoin-driven hydroelectric energy infrastructure, with more and more of humanity clustering around cheap, clean energy sources. Historically, our energy challenge has been to move the power to the people. With Bitcoin, we can move the people to the power.
-Ross Stevens, December 2020
I described this concept of energy bootstrapping as an “interesting idea” in my original 2021 version of this article. During 2022, this vision started to become a reality in Africa. A company called Gridless began using bitcoin mining to incentivize and sustain the deployment of small river hydropower in eastern portions of Africa.
As they describe it on their homepage:
There is immense demand for reliable, clean, and affordable energy across Africa, yet mini-grid energy generators struggle for sustainability. Gridless works with renewable, rural, mini-grid energy generators to monetize the full capacity of their output as a buyer of last resort, as well as serving as an anchor tenant for new energy generation creation.
In its first year of operation, Gridless has entered five different project contract pilots in rural Kenya alongside HydroBox, an African hydroelectric energy company. Three of these pilots are now operational. Gridless finances the construction and managers the operation of the data centers in these rural communities. The company has now set its sights on expansion into other areas in East Africa.
Obi Nwosu described the concept of a bitcoin-based frontier town, comparing and contrasting it to the concept of a gold-based frontier town. People can go to where stranded energy is (such as small waterfalls), monetize it in an environmentally-friendly way to mine bitcoin initially, and then begin using that energy to build out a sustainable community.
Proof-of-Work vs Proof-of-Stake
So far, we established that bitcoin uses only a sub-0.1% fraction of global energy, and a significant portion of the energy it does use is otherwise stranded or wasted, or renewable. It can also help bring new types of energy to market thanks to its flexibility.
But can we make a blockchain that is even more energy efficient? That’s what proof-of-stake purports to do, compared to Bitcoin’s proof-of-work model. Many newer cryptocurrencies use proof-of-stake as their consensus and security model.
As previously described, proof-of-work is a system where miners compete with electricity and processing power to build the longest blockchain, which becomes the accepted blockchain.
In contrast to this, proof-of-stake is a system where holders of the cryptocurrency lock up or “stake” their coins, and use them to vote on the valid blockchain, and get rewarded with more coins for successfully creating new blocks. Instead of committing electricity and processing power to create new blocks on the blockchain, they’re committing their stake of coins to do so.
Proof-of-work is simple, because there is no need to punish bad miners that try to validate the wrong chain or make invalid blocks that don’t fit the rules of the node network. Their punishment is simply that they spent electricity on blocks that weren’t valid or weren’t included in the longest eventual chain, and therefore they lost money. They self-inflict their own wound, and thus it rarely happens on purpose. There is a tangible connection between the blockchain and real-world resources.
Proof-of-stake is more complex, because there is no connection to real-world resources and so the system needs a way to punish stakers that improperly vote on the “wrong” chain. In addition, they need a way to make sure stakers aren’t voting on all possible chains (which can’t be done with proof-of-work, because it takes real-world resources). So, proof-of-stake is a much more complex system that will take away stakers’ coins if they vote improperly, and has ways of checking to see if they are voting on multiple chains.
A more blunt way to phrase this is that proof-of-stake is based on circular logic, since it lacks unforgeable costliness. In a proof-of-stake network, the state of the ledger determines who the coinholders currently are, and the coinholders update the ongoing state of the ledger. This only works while the network is continually running, since anyone can make an alternative history of the ledger nearly for free. If the network goes offline for any reason, there is no trustless and decentralized source of the ledger’s history to restart the network from- large trusted validators or centralized checkpoints must serve as the authority regarding which history of the ledger is the correct one.
Proof-of-work networks, on the other hand, have unforgeable costliness built into the history of the ledger. Even if the whole network goes offline, individual nodes can re-organize themselves and determine the historical state of the ledger from the heaviest proof-of-work blockchain, since there is no way to fake it.
Basically, if you remove the energy input from a blockchain, then instead you have to inject more governance in its place. The whole purpose of the Bitcoin network and its proof-of-work consensus mechanism is to minimize the need for governance and to maximize its ability to run automatically, and for people to coordinate around it globally without trusting any central source of truth.
Besides a larger amount of complexity, trust, and attack surfaces, another main issue with proof-of-stake is that it can be prone to centralization. The more coins you have, the more voting power you have in terms of transaction ordering, and those with the coins are also the ones earning the new coins from staking. Since they don’t need to expend resources to stake, they can simply increase their overall staking amount as they earn ongoing coins from staking rewards, and exponentially grow their influence on the network over time, which can potentially lead to transaction censorship.
I don’t consider the proof-of-stake model bad for other cryptocurrencies to use, if they are more like a corporation. In fact, proof-of-stake can increase the cost of attacking the protocol, since an attack or group of attackers would need to acquire more than half of the coins (unless they find and exploit a bug, due to the greater attack surface). There are certain DeFi projects or platforms, for example, that can operate like a company and use proof-of-stake to be efficient and costly to attack, if all goes well.
Instead, I just consider proof-of-stake to be unsuitable for a democratized, decentralized, censorship-resistant global money, especially as this article shows how negligible the negative impact of Bitcoin’s energy usage is for the world, along with bringing a number of positives.
Or as Adam Back, co-founder and CEO of Blockstream described:
You see that with other commodity money, like physical gold. It’s a system that works because money has a cost. I think money that doesn’t have a cost ultimately ends up being political in nature. So people closer to the money, the so-called Cantillon Effect, are going to be advantaged.
In a proof-of-work system like the Bitcoin network, power is more distributed between miners, developers, and individual nodes. Your ability to be a miner is based on your ability to put forth capital and find low-cost electricity. Rather than the entrenched miners having an advantage and increasing their advantage over time, newer miners actually have the advantage over existing miners because they buy the newer machines with more processing power per watt, thanks to Moore’s law. Mining businesses, old and new, are all constantly refreshing themselves, making use of new cheap or stranded energy resources, and spending nearly all of their revenue on expenses. Nobody gets an entrenched, exponentially-compounding advantage.
Plus, the Bitcoin network’s designers went to great lengths to make it easy and cheap to run a full node (unlike almost any other cryptocurrency). In the Bitcoin network, the real power rests with the nodes, rather than the miners. If miners try to collude and mine blocks that are invalid, the node network won’t accept them. The only thing that miners can do to attack the network is temporarily censor network transactions and perform surface-level block reorganizations.
Many other blockchains that have come into existence since the Bitcoin network, make multiple trade-offs including making the nodes require immense processing power, bandwidth, and storage, so that only industrial-scale entities can run them, which centralizes the network into a handful of major providers.
Bitcoin’s proof-of-work and small block design keeps a lot of power with the individual users. Anyone running a full node can audit the entire blockchain, verify their individual transactions, and participate in the network effect that ensures consensus.
Final Thoughts and Summary
There is a split in the Bitcoin community about how they should respond to environmental concerns.
Some of the larger companies in the industry have formed the Bitcoin Mining Council to collect volunteer data from miners on the sources of energy used. Their data collection over the past couple years shows that bitcoin mining uses a higher ratio of sustainable energy than just about any country’s typical electrical grid:
Chart Source: Bitcoin Mining Council, Q3 2022 Global Review
The general view of the Bitcoin Mining Council members and their supporters is that information is useful to have, so collecting volunteer data and presenting it to capital allocators and policymakers is inherently helpful. And as they found out from their surveys, bitcoin mining has a high ratio of sustainable energy, which makes sense given the fact that bitcoin inherently seeks out cheap and stranded energy. This gives capital allocators and policymakers a useful set of information to work with, rather than the hysteria that is often presented in media.
Others, such as some smaller miners and some individual users, have pushed back against the very framing of the discussion. In their view, the network shouldn’t have to bend over backwards to meet environmental concerns to a greater degree than other industries, and they’re concerned that regulators might try to push for certain types of energy to be prioritized in the network, without those regulators understanding the details of how the network consumes energy. For example, using stranded natural gas that would otherwise be burned away without use, is about as clean as you can get, despite the fact that it’s a fossil fuel.
Both sides make valid points in my view. As an analyst I find it helpful to have the facts about the industry, but I also see why many in the industry would be cautious about the framing of the debate and any attempts made by policymakers (often based on misconceptions) to change or centralize their activities.
Overall, the main takeaways from this long report are as follows:
-Bitcoin provides a service that people can use to store and transfer value. So far, the market of millions of participants has decided that this network has value, and like anything of value, it consumes energy.
-Bitcoin mining uses less than 0.1% of global energy, and by design cannot use more energy than the utility it is providing to users.
-A sizable chunk of the energy that is used by Bitcoin, is otherwise stranded and wasted energy. This is because bitcoin miners have the unique capability to go to remote locations and deal with inconsistent power that other consumers can’t make use of, as long as it’s cheap.
-The network continues to be more energy efficient each year due to pre-programmed declining block subsidies (structural disinflation). Plus, additional layers like the Lightning Network dramatically expand its per-transaction energy efficiency even further as they are built-out and become increasingly operable. Like any functional financial system, Bitcoin uses a layered scaling approach.
-Blockchains that use other consensus models with lower energy requirements, like proof-of-stake, make trade-offs to do so. There is no free lunch, and these other forms of consensus are not “better” than Bitcoin’s proof-of-work model, since they have more attack surfaces, a lack of unforgeable costliness, and greater risks of centralization.
For those reasons, whether Bitcoin continues to be successful or fails in terms of broader adoption, there’s no risk of the network using too much energy in the grand scheme of things. By any metric, it’s a rounding error as far as global consumption energy is concerned, with a sizable chunk of its energy usage consisting of sustainable or otherwise wasted energy.
People focusing heavily on the environmental “E” side of ESG as it relates to Bitcoin often overlook the “SG” component- social and governance. At the end of the day, I consider Bitcoin to be one of the most ESG assets around, just not in the corporate-sanitized conception that the term ESG is often used in.