"But, But ... Bitcoin Uses Too Much Electricity"

Fact: Bitcoin uses electricity.

Opinion (and a possible discussion starter): Bitcoin uses 'a lot' of electricity.

Opinion (and a conversation non-starter): Bitcoin uses 'too much' electricity.

Opinion (of a luddite): Bitcoin's energy use makes it a non-viable option for value exchange.

Like many of the discussions swirling around Bitcoin, the electricity consumption argument is one that would benefit from some context. To that end, let's talk about Bitcoin, blockchain and energy consumption.

I say 'Byzantine' you say 'Generals', 'Byzantine!' ...

Bitcoin is the cryptocurrency of the Bitcoin blockchain. Blockchain technology is both decentralized and distributed, so this means it must solve "The Byzantine Generals Problem" if it is to be a viable medium of transaction. The Generals problem in a nutshell: everyone has a copy of the Bitcoin ledger, but nobody can be trusted to have the "correct" copy. So, how do you generate consensus among all the ledger copy holders? In Bitcoin's case, they do so via a computationally expensive procedure called "Proof of Work". This procedure is a cryptographic puzzle and the first person to solve the puzzle 'mints' the blockchain 'block' and receives Bitcoin as compensation. In the event that Bitcoin is deemed 'worth something' in the world fiduciary markets, the miner's have created 'money out of thin air' at the expense of electricity powering their computers.

Disclaimer: This same explanatory paragraph exists in many blockchain articles.

Blockchain is not necessarily efficient ...

Bitcoin is transactionally efficient but is inefficient in just about every other way. It is computationally expensive (due to its Proof-of-Work consensus mechanism) and it is data and storage wise expensive (due to the fact that everyone has a copy of the entire ledger). But the ability to transfer value point-to-point is rivaled only by cash. And it is clearly superior to cash when the transactional distance is more than 10 feet. Bitcoin can transact value to the other side of the globe at the rate new blocks are minted (about 1 every 10 minutes).

And by way of clarification, this is fundamentally different than what happens when you use a credit card to purchase something on the other side of the globe. In a credit card purchase the transfer of value is not efficient or immediate in any sense of those words. True, it feels immediate. But it is not and is more akin to a "3 day loan" than it is to an immediate exchange. It is a complex and fragile system replete with intermediaries adding little to no value but taking a piece of the pie for their "trouble".

The Role of Electricity in Bitcoin's Blockchain

As a rule, we tend to like efficiency. Efficiency is directly correlated to many edifying things like sustainability and that zen-like feeling of balance that we get when we feel we are being good stewards of all we have been given. A cursory view of Bitcoin leads one to assume that the person that dreamed-up this 'magic internet money' was wasteful. So it is not surprising that so many people have zeroed in on Bitcoin's energy consumption. It seems wasteful and unnecessary.

Far from being unnecessary, the energy consumption is a defense mechanism. The energy consumption requirements make a Bitcoin blockchain takeover unlikely due to the sheer cost required to mint enough blocks fast enough. And this is where the "Bitcoin uses 'too much' electricity" argument breaks down in my opinion.

That Vault has too much Steel in it ...

You know those banks that have a "vault" visible from the bank lobby for all their customers to see?

Well, "that vault has too much steel in it." said no customer, ever. That would be silly because it is intuitively obvious what the steel is there for. It protects our money. And money is useful. You can use it to buy all sorts of wonderful things ... like friends and love!

In much the same way that the steel protects your money - the effort required to mint new blocks on Bitcoin's blockchain protects the value of Bitcoin. And this bank analogy is convenient because it brings up the obvious counter-argument to the notion that Bitcoin's energy consumption is illegitimate. Look at that picture above. Anyone arguing that Bitcoin's energy consumption is 'too high' has a lot of explaining to do. The amount of time, material and energy applied to the marbled and wood-paneled financial institutions the world over, seems to me to be fairly high and a much better target for indignation than Bitcoin's energy consumption. Especially in light of the modern banking consumer's tendency to avoid going into any of these anachronistic monuments to a bygone era with the same revulsion typically reserved for going to the DMV or to clown college.

How much it would cost to "hack" the Bitcoin Blockchain

My good friend Geoff Duda recently posted this article which was a direct inspiration to this article you are reading. The linked article above attempts to estimate Bitcoin's blockchain energy consumption by the same underlying metric that everyone else appears to use: the blockchain's hashrate. Yes, there are a number of problems with that and the Cambridge School of Alternative Finance (who generated the numbers behind the graph) was quick to point those problems out in their explanation of where the numbers came from. But it made me curious as to what "incredibly expensive" would be in terms of US dollars.

The article calls out that the total estimated annual consumption for the Bitcoin blockchain is around 61.76 terawatt-hours. So let's assume that I decided to take over Bitcoin's blockchain ... I would need at least 51% of the computational power of the blockchain so I would need to procure about about 31 terawatt-hours right? Not so fast. Since I would be inserting myself into the existing blockchain ecosystem I would either have to take over existing compute resources already participating in Bitcoin mining (in which case my target is 31 terawatt-hours) or I would need to bring an additional 62 terawatt-hours if I could not get anyone to cooperate with me. For purposes of this thought experiment let's assume the old Collins charm was able to persuade participating compute resources to join "team Collins". How much power are we talking about?

Keep in mind, the reason I am doing this is because with this much compute power I could "hack" Bitcoin's Proof-of-Work consensus mechanism. This would allow me to theoretically take over the blockchain's hashrate allowing me to generate blocks fast enough to make my ledger the longest - which is pretty much how Bitcoin decides who is "right" (which ledger is the true copy). So back to my scenario here, if I attempt to do this in one day how much would it cost? I asked my good friend Geoff Duda to weigh in because "nobody told me there'd be math" and he has forgotten more about energy than I have ever known. According to Geoff ...

87.7 million kWh per day or the entire output of a 3,653 MW generator. Big for sure; the largest coal fired plant in the US has a nameplate capacity of 3,520 MW. The largest coal plant in the world is Taichung in Taiwan (5,788 MW nameplate).  Source: Wikipedia on both so caveat emptor.

-Geoff Duda - Teleios Commodities

So, how much would this cost? Well, ignoring completely what such an energy demand spike would do to realtime energy prices, the complexity of co-opting computer resources from fully engaged entities, the lunacy of attempting to procure the entire output from one of the largest power assets in the world, I could then takeover the Bitcoin blockchain for about $9 million per day. And here's the kicker, the likely result of such action would be that the Bitcoin blockchain participants would see what was happening and would simply fork the blockchain from a block before I started this activity. So I would have control of a blockchain with nobody else transacting on it. So the cryptocurrency on that blockchain would be worthless.

Conclusion

Bitcoin does use a lot of energy. But the energy consumption is purposeful and is congruent with the digitization of the world financial system. Whether Bitcoin existed or not, we are rapidly creating a world where everything financial is digital. It should not surprise us that energy consumption will increase in such a world.