It is probably safe to say that mining is the most controversial aspect of the entire cryptocurrency industry. Stories and articles condemning the energy consumption of Bitcoin in particular have been surfacing for almost as long as the blockchain itself has been running. The environmental aspect of crypto mining has definitely drawn its fair share of ire over the years. But why is mining necessary? How justified are the criticisms levelled against it?

First things first, it is important to understand just how crucial mining operations are within the context of blockchain infrastructure. Fundamentally, a blockchain is a continuously maintained record of transactions. The revolutionary aspect of the technology lies in how the blockchain reaches a consensus on which transactions are included within it. In order to add transactions to the blockchain, users first broadcast their intentions to the network, where they are collected into something called a mempool, which is a kind of waiting room for all unconfirmed transactions. At this stage, nothing has been finalised, users have merely stated how much crypto they want to send, and to where. To set things in stone, to permanently add those transactions to the blockchain, a measure of authority is needed.

This is where miners step into the game. Mining is the process of solving a very complex mathematical problem and receiving a reward for getting the correct answer. Imagine a problem that cannot be solved by any known mathematical formula, the answer can only be brute forced with a huge amount of computing power. It would be like putting a code into a safe over and over again until it opened, except there are trillions of possible combinations.

Miners throw their computing power at the problem until someone solves it. Concretely, the solution requires the generation of something called a hash, which can then be used by other network participants to check the validity of the operation. Very difficult to calculate; very easy to verify. Whoever generates the correct hash first is responsible for adding all those unconfirmed mempool transactions to the next block, making them permanent and updating the blockchain to its new state. The victorious miner then receives a reward in the form of the cryptocurrency they were mining.

Miners are in essence the stewards of the crypto sphere, they underpin the security and immutability of the blockchain, but are also responsible for introducing new coins into the circulating supply. Adding to the blockchain requires expending a large amount of computing power, and by extension, electricity. These energy costs are typically covered by selling the mined rewards. This economic incentive is the cornerstone of blockchain technology; everything depends on it.

Mining comes down to running a computer program on a piece of hardware, and then running it for as long and as hard as possible. Anyone can download the necessary software, set up a mining operation and start mining Bitcoin for example. Given that this is the case, what is to stop the whole world from getting in on the action and mining as many bitcoin as they want?

This is indeed how it all started. In the prehistoric days of crypto, one person alone mined the first blocks of the Bitcoin blockchain: Satoshi Nakamoto. Later on, more people got on board and contributed their processing power to the network, earning block rewards for doing so. The problem is that Bitcoin has a fixed supply, so the more people mining it, the faster the supply limit would be reached. Once the last bitcoin had been mined, there would be no more block reward, leaving very little incentive for miners to keep adding blocks (barring transaction fees), effectively killing the chain due to an absence of computing power securing it.

This is where a mechanism called the difficulty adjustment comes into play. The Bitcoin blockchain is designed so that a new block is mined roughly every ten minutes. This keeps the network at a relatively constant speed for a more stable user experience. In order to keep the generation of blocks consistent, the difficulty of mining a block is periodically tweaked to compensate for the variations in total mining power. For Bitcoin, the adjustment occurs every 2016 blocks, or around two weeks. The more people mining, the more difficult mining becomes. This is why mining is much harder now compared to in the past. More people are competing for the same share of the pie.

In the early days, a simple CPU sufficed to mine Bitcoin. Although not very efficient, the difficulty was such that users could make a significant amount of bitcoin just by running the software in the background on a rusty old laptop. As the network grew, so did the amount of computing power dedicated to securing it. We have a useful metric for this, called the hashrate. The hashrate tells us how many hashes are being generated to solve a block by the combined might of all currently operational mining hardware. At the time of writing, the global hashrate stands at over 500 Exahashes per second. That’s a five with twenty zeroes behind it.

It only took a few months for people to realise that graphics cards were much more efficient at mining than CPUs were. The year after that would see the first real dedicated hardware solutions in the form of FPGAs, or Field Programmable Gate Arrays. The year after that would witness the introduction of the first ASICs, or Application Specific Integrated Circuits, which remain the de facto mining hardware to this day. Mining technology evolved because of the pressures from increased competition, increased difficulty, and of course soaring bitcoin prices.

Mining is no longer the basement-dwelling pastime it used to be. Economic realities have long since pushed mining deep into industrial territory. The scale of some of the larger mining farms are truly awe-inspiring. The energy requirements are now on the level of entire power plants. It simply isn’t a viable domestic operation anymore.

The massive growth in hash rate between 2017 and 2021 was largely fuelled by the development of gargantuan mining farms in China. Entire warehouses full of dedicated mining rigs were directly hooked up to hydroelectric dams in the western reaches of China. Tucked away in the remote mountainous regions of Sichuan, Xinjiang and Inner Mongolia, vast operations were busy at work, taking full advantage of the excess electricity during the rainy seasons. During this time, it is estimated that up to 80% of the world’s hashrate was concentrated in China.

Such a geographical concentration of mining power gave rise to understandable concerns about the uncensorable nature of blockchain technology. Is it really a global currency if one nation has a controlling share of the mechanism that secures it? Debatable, but such concerns would dissipate before they could manifest, when the Chinese government subsequently enacted brutal measures to shut down all mining operations within its borders. China has famously “banned crypto” many times over the past few years, only to walk back on such measures further down the line. In the case of mining however, the ban was final. By mid-2021, the vast majority of Chinese miners had packed up and moved their operations abroad, never to return. The balance in hash power shifted significantly following the crackdown, firstly across the border to Kazakhstan, but more permanently to the United States, which now boasts about 40% of the global hashrate, higher than any other nation. Within the United States, Texas is in the process of firmly establishing itself as the central hub of crypto mining and is now home to the largest mining farms in the world.

Moving on to the energy consumption aspect of mining, this is where the data becomes a bit more difficult to pin down. We know exactly what the hashrate is, thanks to the way proof-of-work blockchains operate and the way difficulty adjustments work, but this doesn’t perfectly correlate with energy consumption for the following reasons.

The first is that we don’t know which type of miner is being used to do the work. Mining equipment has evolved dramatically in the last ten years. Current ASICs are one hundred times more efficient compared to the first models, going from roughly 2000 Joules per TeraHash in 2013 to under 20 J/TH today. We have no idea which miners are being used.

An elaborate balancing act comes into play here. A mining operation that has been around for a while will no doubt have a bunch of old miners lying around collecting dust. It won’t typically be worth plugging them in because the electricity costs of running them will outweigh the profit gained by selling the crypto they generate. The more efficient miners will increase the hashrate to the point that it simply is not profitable to connect the old miners to the grid. The profitability of a miner depends on many things: electricity cost, mining difficulty, miner energy efficiency, but also the price of the crypto being mined. What if the price were to suddenly shoot up? Suddenly the calculations change and it might be worth bringing less efficient models back online, if only before the difficulty increases again.

The second point is related to where the electricity comes from, which ties back to the geographic aspect of mining. In China, mining operations have been notoriously difficult to track down because a lot of the time, and particularly since the 2021 crackdown, such operations have been intentionally obfuscated. Some mining farms were often not even connected to the main power grid, but hooked straight into local power stations. In the case of hydroelectric dams, the fact is that during periods of heavy rain they can remain full for weeks, often draining excess water, bypassing the turbines entirely because there is no energy demand on the grid. In this situation, the electricity is essentially free, which massively changes the viability of older mining equipment. Redundancy is never a factor for such setups; a mining farm is not a data centre. A broken mining rig is unceremoniously thrown away, immediately switched out for a functional replacement. The mining rigs are run hard until they break.

Despite the difficulty in establishing exact figures on global energy consumption, we can still establish reasonably accurate estimates. Claims of the Bitcoin network consuming the same amount of electricity as entire countries are indeed accurate. In 2024, the annualised electricity consumption of Bitcoin mining activities is estimated to be around 150 TWh, which puts it on par with the likes of Poland, Malaysia or Argentina, or roughly 0.5% of global demand. A new bull run in the crypto markets would swiftly change the economic balance of mining and probably result in a huge increase in this figure.

The significance of the above figure really boils down to a question of opinion. So what if crypto mining uses a lot of energy? We waste about a tenth of all generated electricity just in transmission and distribution losses. In the grand scheme of things, how big of an environmental impact does crypto mining represent?

The massive energy expenditure of proof-of-work blockchains is not a design flaw, it is entirely intentional. The energy is what secures the chain. If a nefarious actor wanted to attack the blockchain, reorganising the chain of transactions for personal profit, they would need a controlling share of the power to do so. This is what is known as a 51% attack. To put it bluntly, Bitcoin is secure because it would take the energy demand of entire nations to subvert it. Ethereum used to have comparable energy requirements, until it changed its consensus mechanism from proof-of-work to proof-of-stake, which reduced its electricity demand a thousandfold.

If there is one convincing argument against cryptocurrency mining, this is probably it. If the Ethereum blockchain, and other PoS blockchains, can convince enough people that they are just as secure as their PoW counterparts of a similar size, then it does become harder to defend the massive energy requirements of PoW blockchains. Perhaps there is another way. Time will tell.