Bitcoin Electricity Usage: Is It Worth It?
Imagine there was a type of money or digital asset that was minted over a century or so, but could be used not only during the minting period but for many centuries to come. The energy in creating it might seem large at the time, but be a very worthwhile long term investment. That asset is Bitcoin.
A year ago I wrote about the electricity requirements of Bitcoin mining and told people that we should calm down. I noted that gold mining consumes much more power than Bitcoin mining. This is not surprising since the above-ground gold supply is worth, at $8 trillion, about 40 times as much as the present Bitcoin market cap.
Bitcoin’s Power Consumption
There are many alt coins that use other types of consensus algorithms rather than the Proof of Work used in Bitcoin mining, but these have much lower security and much less value as a result. A relatively small number of coins are mined using Proof of Work, with Ethereum, Bitcoin Cash, and Litecoin being the most important after Bitcoin. None of these has a computational hashing rate of even 3% that of Bitcoin, however.
As of the end of May 2019, we put together a list of the top mining pools and estimated that the top 8 mined coins amounted to $24 million of economic production per day or $8.6 billion per year, and 2/3 of that was accounted for by Bitcoin. The overall value and Bitcoin share have both risen since then.
There have been a number of estimates of Bitcoin power consumption over the last few years, and several of these are summarized in a recent Congressional Research Service report. The table below shows the findings in Megawatts by year, and I have taken the geometric mean where there were multiple estimates from various sources within a given year.
While there are often concerns expressed about the Proof of Work electricity consumption amounting to that of mid-sized countries such as Ireland or Colombia, it remains well under 1% of global usage, by an order of magnitude.
However, Bitcoin is growing value rapidly over the long term and the mining hashrate continues to reach new records; it is currently around 70 Exahashes per second. As a result, as we see in the table, the electrical power consumption has grown rapidly as well, more than doubling each year on average.
Since the current global generating capacity is around 7,365 Gigawatts, the current Bitcoin mining power requirement is just over 0.1% of global capacity. Total global generating capacity grows around 2% to 3% per annum.
The point of the electricity usage is to generate hashrate, the number of Proof of Work cryptographic calculations per second. The higher the hashrate, the greater the chance of winning the reward for a given block. Also, more hashrate from the world’s mining pools means greater security and thus more value for Bitcoin.
During the current decade, total mining hashrate has increased by a factor of over a trillion as the number of miners has grown and as the miners have deployed specialized hardware. During the same interval, the Bitcoin price increased by a factor of 10 million from its very first price discovery at 1/10 of a cent.
Very roughly it seems as if the market capitalization of Bitcoin rises in proportion to the square root of the aggregate hashrate, but with much variance. Hashrate tends to follow, not lead, Bitcoin’s price increases.
I have run a regression on the log base 10 values of Bitcoin power requirements in Table 1 and find a slope of 0.38, corresponding to an average factor of 2.4 increase in power consumed per year.
In Table 2 above we forecast the future electric power requirement assuming the same rate of increase observed during the past 5 years. We compare that to the current global generation capacity of 7365 GigaWatts and also assume global generating capacity increases by 2% per year.
Unusual Supply-Demand Relationship
So what does this possible continued rapid increase in electricity employed for Bitcoin mining and crypto mining in general suggest?
First one must realize that Bitcoin economics are nothing like typical supply and demand economics. An increasing amount of mining and electricity utilized does not impact the supply rate of Bitcoin. The supply rate for new coins is coded into the Nakamoto consensus and determined by the pre-determined block reward Halvings that occur roughly every four years. (These actually occur at every 210,000 blocks and that is closer to three years and nine months between Halvings).
If the mining hash rate leaps, there is a difficulty adjustment automatically made inside the Proof of Work algorithm that keeps the ledger block interval on a nearly 10-minute schedule, so that the aggregate block reward each day is close to uniform, until the next Halving.
There have been two Halving events to date, in 2012 and 2016, that cut the number of Bitcoins awarded per 10 minute block from 50 to 25 and then again from 25 to 12.5, the present block reward. So there are currently 1800 new Bitcoins generated per day.
Next year, in May of 2020, the block reward will be halved once again to 6.25 Bitcoins per block. Then the number of new Bitcoins generated will drop to 900 per day.
This supply rate is fully determined by the Bitcoin code and will not change whether the electricity usage doubles or triples or stays the same.
Here is how Bitcoin economics works. Supply tightening with Halvings, as well as with the continued growth of the existing stockpile relative to annual supply, cause the price to rise. Rising prices are an incentive for more mining, causing the hash rate to grow. The increased hash rate enhances security, which also adds to Bitcoin’s value.
Electrical power consumption has been growing as the 2/3 power of the hashrate. The Bitcoin market cap has been growing more slowly than electrical consumption, by about 20% per year, which seems as if it could become untenable in the intermediate-term.
However, once a Bitcoin is mined, it is forever, and its value can also continue to increase in the future. Thus the market cap of the Bitcoin network should continue to grow indefinitely even as new supply issuance becomes smaller and smaller.
More Government Regulation
What increases in electricity utilized will do is bring further government regulation, since electricity is one of the most regulated industries around the world. Bitcoin itself cannot be banned, but mining and exchanges can be banned or regulated.
The competition will also drive miners toward more efficient ASIC-based mining computers and the most favorable locations for power.
The good news is that Bitcoin mining operations have been seeking out the lowest priced electricity environments and best cooling environments. These are often based on renewable power sources, especially hydropower located in cool climates. One recent estimate was that 74% of the consumed electricity worldwide is from renewables.
Since Bitcoin mining can happen at any time of day, it is suitable for off-peak electricity usage. Mining pools are increasingly multinational and can move mining workload around the globe by the time of day as appropriate. Bitcoin mining could actually encourage renewable electricity development since it can provide economic returns from off-peak production that would otherwise be underutilized.
Some of the most important geographies for Bitcoin mining are in China, including Sichuan province, and Canada, especially Quebec, and in the U.S., Washington State, and upstate New York. In Europe, Scandinavia is a popular location. Most of these locations have substantial hydropower sources.
There is very cheap energy available from natural gas in locations where the gas is flared off. For example, in the massive Bakken formation found in the northern U.S. and in Canada, almost one-quarter of the natural gas is wasted, flared off because there is no economically viable way to get it to market. One can colocate Bitcoin mining rigs in shipping containers at the source of the natural gas, and convert the waste product to value. This is happening in Canada today.
Various governments, national, regional, and municipal, have often stepped in to restrict or regulate mining operations, although the province of Quebec has actually provided incentives to miners.
What we would expect is that as power consumed grows to a few percents of the global supply that regulations and limitations will become even tighter. Perhaps mining will be allowed only at certain times of the day and night, or higher electricity prices that typically come into force during peak hours will cause miners to go offline each day.
All of these effects will tend to lock in the most efficient mining pool operations as part of an increasingly utility-like industry.
Now one can expect that individuals will continue to mine, offering their equipment as part of a pool on-line or even as gray market unlicensed “wildcat” miners. But it will continue to be harder for them to compete if not based in a low-cost location.
The same number of Bitcoins will be produced per day regardless of any these developments. If the total electricity supplied to mining operations is restricted in one location, then there will just be different winners and losers in the race for block rewards. If some miners shut-in or restrict operations, voluntarily or due to government restrictions, the profit margin for others will be improved.
Those who are in place and build out their operations most efficiently and aggressively in the most favorable locations now will be in the best position going forward to capture Bitcoin production market share.
The next three years will be very telling, as Bitcoin’s electricity consumption reaches 1% of global electricity level. How successful will mining pools and miners be in accessing low-cost electricity? Will they be able to maintain a high renewable resource fraction in their consumption?
Over the next couple of decades as there are continued Halving events roughly every four years, even with continued price increases for Bitcoin, miners will need to increase their revenue from transaction fees. Eventually, miners will become fee-based transaction validators, first and foremost. If the price of Bitcoin continues to rise this will be a very viable business model.
Bitcoin does not really need to compete with Visa and MasterCard for small transactions, although second-layer payment channels with Lightning make this possible. Rather it competes with large value transfers including wire transfers on the SWIFT network, or transfers of gold bullion.
Imagine it is the year 2039. The supply of Bitcoin is now well over 20 million and barely growing, increasing by only 7 Bitcoin per day. A single Bitcoin is worth $2 million. Bitcoin “miners” or block validators are compensated mainly from transaction fees, which are substantial, given the enormous $40 trillion capitalizations of the Bitcoin network.
A Satoshi, the Bitcoin sub-unit, is worth 2 cents. Micro-transactions and small transactions are easily handled off-chain by a Lightning network or other second-layer solutions, with batched transactions committed back to the main chain once a month, or once a quarter.
Central banks have added Bitcoin to their reserves, alongside their gold and foreign exchange holdings. Fiat currency is better backed than during the wild days of unbridled central bank quantitative easing in the 2008–2028 period.
Will the electricity expended in creating the supply of funds for this multi-trillion dollar network have been worth it? I believe so.
http://orionx.net/2018/07/will-bitcoin-consume-all-electricity/ Will Bitcoin consume all electricity?
http://orionx.net/2019/06/top-50-of-crypto-mining-june-2019/ Top 50 of Crypto Mining, June 2019
“Bitcoin, Blockchain, and the Energy Sector” Congressional Research Service R45863, https://crsreports.congress.gov
https://bitcoinist.com/bitcoin-mining-waste-oil-industry/ story on capturing natural gas in situ