Bitcoin and pollution — the definitive answer
The problem
Bitcoins are created out of thin air from a computing process called “mining”, a term inspired by how precious metals are extracted from Earth’s crust. In reality, this “mining” is a computationally intensive task with an artificial level of difficulty, required to reach consensus on which “block” in the blockchain gets validated next (on average every 10 minutes). It’s also called Proof-of-Work.
The same computing process also validates ~4 transactions per second, or ~300,000 per day, or ~110M transactions per year (source).
This network of Bitcoin nodes currently consumes ~77 TWh per year (source). This leads to many (mostly negative) comments about Bitcoin’s environmental footprint, such as these:
Bitcoin consumes annually the electricity output of a small country. (source)
Most bitcoin mining is using energy at the source that was uneconomical to use for other purposes, because of the loss experienced in transporting the energy to economic centers. (source)
(On the news of Tesla’s purchase of US$1.5B worth of bitcoins) Legitimizing Bitcoin more than offsets the environmental benefits of electric vehicles. (source)
Bitcoin uses as much electricity as Chile and a single transaction has the same carbon footprint as ~700,000 Visa transactions. (source)
So, what’s going on?
Does Bitcoin pollute?
Yes, Bitcoin does pollute. The Bitcoin network consumes as much electric energy as Chile (but represents only a fraction of other pollutants, so it does NOT “pollute as much as Chile”).
China ranks #1 in the world at 7,225 TWh/year.
The United States consumes ~3,989 TWh of electricity per year.
My home country, Italy, is at 297 TWh/year.
Chile? 75 TWh/year, slightly less than the Bitcoin network. (source)
Yes but… Is that mostly renewable energy?
Hard to say, but… Not really. Here’s why.
Most Bitcoin mining is done in China, with equipment that runs 24/7 (you want to amortize the high capital costs of buying ASICs to mine bitcoins).
The only renewable energy that could also run 24/7 is hydroelectric — and that’s not even always true. Otherwise, your next best option for cheap energy in China is… Coal, which is particularly bad for the environment.
Indeed there are regions in China where there’s abundance of hydroelectric power, and I’d assume that most Bitcoin mining operations try to be there; but in essence only 39% of global Bitcoin mining uses renewables (source), and therefore, 61% of Bitcoin mining does NOT use renewable energy.
Is the VISA network way more efficient than Bitcoin?
One Bitcoin transaction consumes as much energy as 700,000 transactions on the VISA or Mastercard network. That’s quite correct.
But it would be much more fair to compare the amount of money settled (source), in which case:
- VISA handles US$30B/day
- Mastercard handles US$11B/day
- Bitcoin handles US$10B/day
So, no, VISA is not more efficient than Bitcoin. VISA is simply a completely different thing. In fact…
Bitcoin is still mostly a store of value
Bitcoin is not used to buy a cappuccino for the equivalent of $4. It is NOT used to buy stuff. At all. Bitcoin is mostly used as a store of value (and I’d add: mostly for speculative reasons, at least until now).
We know the energy required to run Bitcoin’s network: 77 TWh/year. The average cost paid by miners is US$0.046/kWh, which means that the Bitcoin network costs US$3.5B per year on energy alone (source). 61% of it is powered by non-renewable energy, equal to 47 TWh/year, or ~US$2B.
Currently, 328,500 BTC are mined per year, with an existing total supply of ~18,623,000 (source). At the current price (Feb 10th, 2021) of $46,000 per bitcoin, ~US$15B worth of bitcoin are mined per year.
It is much better, in my view, to compare Bitcoin to another store of value that the world is already using: Gold.
Bitcoin vs Gold in terms of pollution
AISC (median all-in sustaining costs) mining gold cost is at US$923 per ounce. For us in civilized, modern, 21st century countries where the SI is in use, that means US$923 per 28.35 grams, or US$32,557 per kilogram. You can sell a kilogram of gold for ~US$58,000. Most of the gold mined is used as a store of value. A small fraction of it has real use (jewelry, electronics, etc).
There are ~190,000 tons of gold stored somewhere in the world. Each year, ~3,000 tons of gold are mined (source).
The cost of mining gold each year is therefore US$97.6B.
The cost of storing gold is much harder to calculate; I’m going to use a simple approximation here, and say that it equates to 10% of its mining costs.
It’s pretty clear by now that, looking just at costs (without getting into the detail of the actual type of pollution), mining gold is ~50 times more expensive than mining bitcoins AND running the bitcoin network. Bitcoin pollutes far, far less than gold.
mining gold is ~50 times more expensive than mining bitcoins
It’s even harder to compare the type of pollution created by mining gold versus mining bitcoin. What’s easy to do is to simply state that mining gold is MUCH WORSE than mining bitcoins. Mining gold is an environmental disaster. Poisoned waters, solid waste, endangered communities, and so on.
Producing gold for one wedding ring alone generates 20 tons of waste.
Bitcoin pollutes far, far less than gold.
If we consider the hypothesis that Bitcoin (current market cap: ~US$800B) might one day replace Gold (current market cap: ~US$10,600B — source) as store of value, the environmental effect of this switch would be a significant reduction in total pollution.
The end of pollution: From Proof-of-Work (PoW) to Proof-of-Stake (PoS)
Last, but not least, it is reasonable to expect that the Bitcoin network will manage to transition to a Proof-of-Stake consensus mechanism, within at most a few years.
For the non-nerds reading this, it means that instead of requiring complicated calculations to prove that a certain block in the blockchain is confirmed (and therefore, consuming a lot of energy), you need to “stake” your existing assets, therefore consuming much less energy compared to the current “Proof-of-Work” consensus mechanism.
Conclusions
Next time you read of hear someone complaining about Bitcoin’s pollution, point them to this blog post. If they have a wedding ring, tell them what that means for the environment.
But wait, there’s more.
I hope that in a few years Bitcoin will have fully replaced gold as the global reserve. I hope that Gold will fall in value, so much that mining gold becomes anti-economical. The result will be a cleaner, greener world, and a much more flexible store of value.
Hope, unfortunately, is almost always never enough. This innocent, naïve dream of Bitcoin replacing Gold needs to face the reality of our complex financial system, our complex global politics, and so on.
This is still a remote hypothesis, for many reasons. I would give it a <10% chance of happening by the end of year 2030 (~10 years from now).
And let’s not pretend that Bitcoin HODLers are tree huggers; as I wrote earlier in this post, the vast majority of Bitcoin transactions are simply speculation.
We are driven by greed, not by green.
But still, we can be hopeful: our planet would be really lucky if, for once, green is an unintended consequence of our true nature.
Final note: I explicitly decided NOT to talk about the environmental impact of having a global currency like the US dollar. It’s a huge topic and I don’t feel prepared to dissect it properly.
