Lets Boil Some Oceans
My friend recently sent me this gem of a post written by Carl Beekhuizen of The Ethereum Foundation Blog.
Upon reading, I began to feel some déjà vu coming along, and decided to try my hand at busting some FUD (fear, uncertainty, doubt)! Some of the ideas I tackle here are in no particular order with respect to the chronology of the blog post, and may not even be in the blog post explicitly. However, I’d like to cast a wider net here so do forgive me if I joust with some windmills. In any case, let's begin
The key comment that set me off has to do with energy accounting, specifically the use of the energy per transaction metric:
“If energy consumption per-transaction is more your speed, that’s ~35Wh/tx (avg ~60K gas/tx) or about 20 minutes of TV. By contrast, Ethereum PoW uses the equivalent energy of a house for 2.8 days per transaction and Bitcoin consumes 38 house-days worth.”
OK….
The energy use of proof of work is not directly proportional to the number of transactions in a block. Thinking this is the case is a fundamental misunderstanding of how the validation and consensus mechanism works (maybe even an intentional deception). Miners validate blocks primarily for the block reward (currently 6.25 BTC/block) and so the amount of hash power that comes online is a response to the rise/fall in the price of BTC, which will make mining operations more/less profitable to run. The number of transactions is a total red herring here. Using the E/txn metric leads to the faulty assumption that there is some fundamental relation between how many transactions take place on the network and how much energy miners use. One might give those using the metric the benefit of the doubt and assume that they are making some subtler point about network scalability (can’t think of one but I’d love to hear it), but then we are venturing into territory that has been well-tread (Lightning Network, Liquid, etc.). I will forego more discussion on the transaction figure of merit as it has been more eloquently covered elsewhere
Now suppose we can agree that the energy accounting for BTC mining, and is accurate. We then have to consider what kind of energy is being used to validate blocks. It turns out that the economics of BTC mining is such that miners are incentivized to use the absolute cheapest energy sources available. The cheapest energy that would end up being used is waste energy (eg. natural gas flaring, stranded hydroelectric, geothermal, etc.). The rub here is that most of this energy has — using economic jargon — “zero marginal utility” to humanity because it is too costly to transport that energy to a place where people live and can use it. So in any situation in which there is no possibility of profitably selling energy to the grid, you could potentially store that energy in monetary form. If you follow the thread of this reasoning further you realize that BTC mining actually can serve as an “energy buyer of last resort”, creating a price floor where there was none before. This even has implications in incentivizing renewable energy operations by alleviating demand mismatch problems, allowing energy providers to cost-effectively diversify their energy portfolio. There are so many other amazing things we could do in a more decentralized energy market that I may write another one of these just to talk about it.
Now from here on out my comments may start to get very sardonic, pithy, or downright rude, but when viewed in the context of a sociopathic financial system, and an endless clown world of recycled FUD, it’s an insignificant personal indulgence all things considered.
Anyone can easily spin up factoids about how INSERT TECHNOLOGY THAT I DON’T LIKE HERE uses as much energy as a small country, so the people who do it are bad and are LITERALLY BOILING OCEANS. It’s an easy rhetorical hat trick, so I’ll try my hand at it:
GLOBAL REFRIGERATOR USE uses as much energy as a small country! If you want high-shelf-life foods there are plenty of alternatives like canned foods (love me some Bush’s baked beans) and packaged foods (mmmm Doritos). Anyone stupid and irresponsible enough to use refrigerators is LITERALLY BOILING OCEANS. what’s that? You feel like it’s ok for you to be able to enjoy some ice cream at the end of your double-shift, or maybe store some tomatoes from the farmer’s market for your delicious caesar salad from scratch? SCREW YOU THINK OF THE PLANET YOU SWINE.
oh, man… I think I got a bit carried away there…
I’m really sorry about that. I’ll calm down and talk about something else for a bit. Do you know what I’ve realized lately? It’s hard to find some silence these days. Some time to really sit down and reflect. Be at peace. Find my zen. Know what I mean? I’ve actually started washing my clothes by hand. Yeah! I got the clothesline, washing board pins the whole shebang. It’s soothing to just get into the flow of working with my hands. I get into this meditative state and thoughts just pour out of me. I'll share one with you that I thought was particularly profound:
WASHING MACHINES IN AGGREGATE use as much energy as a small country! These IDIOTS who don’t have the time to wash their clothes by hand are LITERALLY BOILING OCEANS
The bit might be getting a little stale at this point, but I just want to do one more. It's super neat I swear.
THE U.S ACTIVE AIRCRAFT CARRIER FLEET uses as much energy as a small country! Furthermore, it's powered by two A4W fission-powered pressurized water reactors. So it does in fact ACTUALLY LITERALLY BOIL OCEANS
Well, maybe that got a little too real. That being said, extraordinary claims require extraordinary evidence, so I’ll pull out the back of my handy-dandy envelope, sit in my undergrad physics degree armchair, put my internet-sleuth hat on, and make an attempt. Before I begin, let me make the disclaimer that I’m just some random person who has put themselves into the awkward position of speculating on secretive naval maneuvers so apply your grains of salt accordingly.
A single Nimitz class aircraft carrier can sustain a maximum speed of 30 knots (yes I totally know how fast that is) With a maximum power usage of 194 MW. Let’s suppose that none of them are in any rush, and they don’t move all that much around the places they are stationed. I am going to be very generous to our compatriots at the U.S Navy and guess that each carrier uses 10% of its maximum power (19.4 MW) over a year of operation and sustains this energy usage for three hours every week (that's about 150 hours over that same year). There are currently 10 active Nimitz class ships in the US fleet, some with over 20 years in operation. So in one year, we have an estimated energy usage of 19.4 MW times 150 operational hours times 10 carriers for a grand total of 29.1 GWh/yr. Now if we look at a list of countries by electricity consumption, we can see that the small island nation of Kiribati had an energy usage of 27.75 GWh/yr as its 2018 estimate. From this point, the reader may verify that 29 is greater than 27. QED.
Now that you’ve let me have my fun I’ll get back to my main point. At any stage of my analysis some of you may have done one or more of the following:
- Balked at a series of numbers that you think I effectively pulled out of my ass
- Remarked that it is easy to do a surface-level 30 minute Wikipedia scrub and that this is not actual research
- Thought that I am wholly ignorant of the wider systems at play
- Stated that my calculations were overly simplistic and that my model fails in its approximation of the complex reality we actually inhabit
If you have done any one of these things you should congratulate yourself, you have pretty much beat me to the punchline. The point here is that it is easy to take some cross-section of human energy use, do some numerical hand-waving, and claim some sort of victory because you have managed to hypnotize your readership into an algebra-induced stupor. Gee whiz isn’t math super hard and cool! However, what this species of FUD does is obfuscate the real discussion we should be having. The real questions we should be asking. Here are a few of these questions:
What are the fundamental advantages and downstream benefits of using hash power (and the collective energy-harvesting resources if humanity as a result) to secure a global monetary base layer now worth $1 trillion (and soon to be worth much more by my reckoning)?
How efficiently is this energy used in providing said security when compared to the energy cost of securing the value stored in the fiat system?
When moving to an alternative consensus algorithm such as Proof of Stake, does the reduction of energy use come with new security risks? What are these risks? Have any concerns associated with these risks been robustly shown to be unfounded based on testing by the open market? (note how network security is not even mentioned in the aforementioned blogpost)
Of course, there are many resources one could seek out to begin to arrive at an answer to these questions, but here are a few that come to my mind:
- Alex Gladstein’s article on the petrodollar system
- Lyn Alden’s newsletter post, also on the petrodollar system
- A section of Michael Saylor’s interview with Kitco News where he talks about Bitcoins energy usage
- An analysis of Ethereum and its prospects also by Lyn Alden
If you have managed to get this far, I thank you for being generous enough to spend your most valuable resource reading my first piece of public writing. It's been almost a year since I started my plummet down the rabbit hole, and I’m grateful for all of the lessons it has already taught me. If you enjoyed this and have the additional generosity in your budget to give me some positive incentive to write more, kindly send some sats my way.
