Bitcoin in a nutshell is a decentralized monetary ledger that converts electrical energy into peer to peer money. This essay explores the energy consumption of the Bitcoin network at different price points. A very basic familiarity with nuts and bolts of Bitcoin mining is assumed. We start with estimating the energy consumption of the network at present, then extrapolate it for 100K & 1M USD Bitcoin. The reason why we can do this extrapolation is the fact that mining difficulty and network hashing power ultimately catches up to the price, keeping the return on investment for mining hardware stable.This enables us to roughly estimate power consumption at higher and higher price points. Before we dive into the estimation, I want to make it very clear that this is a back of the envelope type calculation. Also website like digiconomist already tracks the energy consumption of the bitcoin network in detail, but I want the reader to understand the variables behind this estimation so as to get a better handle on extrapolating the power consumption in future.
Lets estimate current energy consumption
According to blockchain.com the current network hashing power is around 50 Million TH/s (trillion hashes/second). The most efficient mining hardware currently available is Antminer S9 which delivers 14 TH/s for 1.35 KW of power, so roughly 10 TH/s for 1 KW. Now if assume that the entire Bitcoin mining network is running on this hardware, we get a lower bound for the power consumption. I am aware of the fact that most of the mining is done at industrial scale with access to super cheap energy (sub 5 cents/KWh), this makes even the far less efficient hardware still profitable. But lets just get back to estimating the lower bound. 50 Million /10, we get the current power consumption as 5 Million KW. This is the lower bound, to estimate a more realistic power consumption of the current network, we need to compensate for less efficient hardware and add the power required to cool the mining rigs. Lets assume that the average hashing power per KW is around 9 TH/s instead of 10 TH/s. This will give us 5.55 Million KW (50 Million/ 9). Now we need to add the power consumed for cooling the rigs. According to Bitfury (institutional miner and manufacturer of mining ASICs) CEO , around 30% to 40% of total electricity consumption goes into cooling the rigs. For simplicity we just add 30% on top of 5.55 Million KW. So a total of 7.15 Million KW, lets round down for simplicity to 7 Million KW or 7 Giga (Billion)Watt. Global power consumption is between 15 to 18 Tera (Trillion)Watt according to different sources. This means current Bitcoin network consumes less than 0.05% (0.043% to be more accurate) of global power consumption. But what happens when the price of Bitcoin goes to a million dollar ?
Extrapolating power consumption for a million dollar Bitcoin
Before I delve into the analysis, I want to make it very clear that this extrapolation depends on Bitcoin network using Proof of Work as the consensus algorithm in the future and not switching to Proof of Stake or any other algorithm. It is highly unlikely Bitcoin will ever switch to any other consensus mechanism.
Lets imagine that the price of Bitcoin jumps to 1 Million dollar (US) tomorrow, what happens to mining profits ? We assume mining difficulty remains nearly the same for another year. The basis of our calculation here is Antminer S9 ASIC. The cost of this hardware currently is about 650 USD
At current price of Bitcoin and access to cheap electricity it takes about 11 months to recover the cost of mining hardware. If the price hits 1 Million dollar it will take just a little over a day to recover the cost. You can basically make 200K on an investment of 650, this is insane. The point I am trying to make here is that it becomes extremely lucrative and hence it can’t stay like that for very long. Orders of magnitude more mining rigs will be deployed eventually, pushing the return on investment for the hardware back to its historical average. This is obviously an extremely unrealistic scenario but even if takes Bitcoin 5,7 or 10 years to get to a million dollar, mining difficulty will ultimately catch up. This means the electricity consumption will also scale proportionally. Sure, we will have more efficient hardware in the future meaning more hashes per second for same KW of power but lot more rigs (same level of power consumption) will be deployed as long as price keeps climbing. Another factor to consider here is reward halving. Lets imagine a scenario when the price of Bitcoin hits 1 Million dollar by 2025. The block reward will only be a little over 3 Bitcoins, its currently 12.5 Bitcoins. So when the price of Bitcoin goes from nearly 6.2K at the moment to 1 Milllion in next 7 years, a factor of 160 the mining revenue will grow only by a factor of 40,because reward has been reduced by a factor of 4. The power consumption will also rise from 7 GW to 7 x 40 = 280 GW. Assuming an annual increase of 5% in global power demand, the world will need 24 TW of power in 7 years. Around 1.2 % of that would be required to run the Bitcoin network.
Cost of running the network when Bitcoin hits 1 Million dollar
Seven years from now, the number of Bitcoin mined will be around 19.5 million. Assuming a reasonable 1.5 Million to be lost forever, there should be around 18 million Bitcoins in circulation by then. If the price goes to a million dollar, the network will be valued at 18 Trillion dollar. Now we need to convert 280 GW of power into KWh of energy. 280 GW = 280 x 10⁶ KW
280 x 10⁶ x 24 x 365 = 2.45 Trillion KWh annually. Assuming the cheapest electricity is available at 5 cents per KWh by then, that would mean an annual cost of staggering 122 Billion dollars (lower bound) to maintain the network. Or is it really that staggering ?
Securing a network worth 18 Trillion dollar at a cost of 122 Billion (less than 1%) is actually quite cheap. I am yet to estimate dollar and energy cost of gold mining and global banking system but my guess is its not going to be significantly cheaper, could be even more expensive. After all what should be the cost of human liberty and freedom ? If that amount buys 7 Billion people liberty, then its worth it. Getting every human on the Bitcoin ecosystem is one of the most audacious goals and when completed, it would be one of the greatest achievement.
