Intelligent Utility Network Community

The mission of this network is to bring together utility professionals in the power industry who are in the thick of the digital utility transformation. This network incorporates the Mobile Utility special interest group.

61,814 Subscribers

Article Post

The energy toll of cryptocurrencies is overstated, and here’s why…

You might be wondering what an article on cryptocurrencies is doing on a site dedicated to providing news about the energy sector. Just about every person and their grandma has heard about Bitcoin by now, and the influx of altcoins (in addition to the surging popularity and widespread adoption of cryptocurrency in general) has piqued the attention of many people who are concerned about the massive amounts of power that is dedicated to ‘mining’ the digital currency.

For those that you don’t know, Bitcoin—and many other cryptocurrencies, for that matter—require computers to solve incredibly complex mathematical equations in order to successfully mine a block of the currency. Initially, it was viable to mine bitcoin from your regular home PC or laptop, but the last few years have required miners to buy or rent dedicated mining computers that are optimized for mining bitcoin. One of the major reasons that non-specialized computers lost their viability was that they were consuming more energy than they were receiving in bitcoin.

Despite the fact that cryptocurrencies have given millions of people living in developing countries the opportunity to transact with minimal overhead (as bank fees would often be cost-prohibitive to small businesses located in some developing nations, where countries like Somalia only got their first ATM in 2014), there has been a growing concern about the energy toll that the mining of digital currency takes on our world’s already-strained energy supply.

You might think that a peer-to-peer currency such as bitcoin would share the energy load efficiently; however, unfortunately, this isn’t the whole story. While units of cryptocurrency can be traded with negligible effect on your power bill, the mining of cryptocurrency is another story. While I appreciate that many of the readers on here have a good sense of what a kilowatt hour is, some of you may not. As such, I think the best way to truly get an idea of the energy cost pertaining to cryptocurrency—specifically as it pertains to Bitcoin—is to look at the energy usage as compared to other mediums of exchange. This is where things get interesting.

According to figures obtained from Digiconomist’s Bitcoin Energy Consumption Index, it takes about 50,000 US households for Visa to process 350 million transactions, while it takes Bitcoin the equivalent of 2.8 million US households to run 350,000 transactions when things are going well. In other words, Visa’s handling of transactions is 58 times faster than Bitcoin’s. This sounds bad—and, admittedly, it’s far from ideal—but it’s also far superior to more common forms of currency production.

Bitcoin takes 0.8KWh to 4.4KWh per year to run, while minting coins and paper bills in America alone uses about three times as much energy—and Bitcoin is accepted worldwide (barring the half a dozen relatively insignificant countries that have currently banned trading the currency). Although, to be fair, the playing field is somewhat levelled when you realise that the total cash and coin supply of US dollars is about $1.2 trillion, which is about five times Bitcoin’s market cap (which is currently sitting at around $257 billion). Nevertheless, Bitcoin, as a currency, is immutable and cannot be destroyed or worn out like paper money and coins; moreover, Bitcoin cannot be counterfeited like paper money can be. Then there is the recycling of gold (AKA gold smelting), a practice which goes back centuries. Even using the most modern techniques of goldsmithing, the recycling of gold uses, on average, 138KWh per year.

Bitcoin alone consumes a staggering 37 terawatts per year (which is about the same amount of energy that the country of Qatar consumes in a year), and that number will surely continue to grow as the market cap does; this equates to more than 18,000 kilotons of CO2 each year. Therefore, regardless of whether cryptocurrency (and the blockchain technology that underpins it) is the future of digital currency (and currency as a whole), developers who maintain and invent new cryptocurrencies should carefully consider the effect that it can have on the power grid. 


Chris Cosgrove's picture
Chris Cosgrove

There is something amiss in the math presented.  According to what is written, Visa can process 7,000 transactions per household (350,000,000/50,000).  Bitcoin can only process 0.125 transactions per household (350,000/2,800,000).  Is there a typo somewhere?  If that math is correct, then Visa is 56,000 times more efficient at transactions than Bitcoin. 

"According to figures obtained from Digiconomist’s Bitcoin Energy Consumption Index, it takes about 50,000 US households for Visa to process 350 million transactions, while it takes Bitcoin the equivalent of 2.8 million US households to run 350,000 transactions when things are going well."

Phillip Kopp's picture
Phillip Kopp

Chris, good point. The author should clarify the discrepancy between 350,000 and 350M transactions.  And also that this figure is not representative of time, but rather energy value.  Since bank transactions are processed on specialized mainframes, I don't see how they could be 56,000 times more efficient than the ASIC computers that process Bitcoin transactions.  If we assumed that the processing efficiency is roughly equivalent, then it means that it takes 56,000 times longer to process a bitcoin transaction than a Visa transaction?  Lets say a Visa transaction takes about 20 seconds to process in total, that would be 311 hours to process a Bitcoin transaction.  I have seen it done in just a few minutes, so I think the numbers here are definitely off.  

It's probably a typo and should be on the order of 350,000,000, not 350,000 transactions.  Even then, 2.8M households would be a lot of energy, roughly 48kWh/tx (not 4.4)!

We better get a handle on the math :)

Ben Schultz's picture
Ben Schultz

Just thought I should let you know I responded to your comment!

Ben Schultz's picture
Ben Schultz

Thanks for your responses, gentlemen. Actually, I believe the figures are, surprisingly, correct (broadly speaking). One of the big criticisms of Bitcoin has been its inefficiency when compared to more established payment methods. But the security that is provided by the nature of cryptocurrency is invaluable to many.

You can see by these numbers that it does indeed take the equivalent of about eleven houses per transaction (which is actually three more than the previous estimate).

You also have to factor in the energy that it takes to mine each coin, not just the transaction itself. The sustainability of Bitcoin as a currency will be highly dependent on streamlining the payment process. If you have any more questions, please feel free to ask!