Bitcoin is recognised as a store of value making it comparable to gold. In 2020, 3,200 tonnes of gold was mined equating to approximately 90,301,440 million tonnes of CO2 emitted.
In comparison Bitcoin is estimated to emit around 37 million tonnes of CO2 over the course of 2021, with China powering 65% of the hash rate in March 2021. China pledges to be net zero by 2050 along with most of the world under the Paris agreement. Under the assumption, this is achieved this would indicate Bitcoin would be powered by renewable energy by 2050 making it a zero-carbon technology. In addition, gold mining is renowned to be one of the most destructive industries responsible for polluting drinking water with cyanide, mercury and other heavy metals, whilst destroying pristine environments and causing damaging health effects. Removing reliance on the need for gold not only has the potential to decarbonise the gold reserve industry but also reduces these negative environmental and health impacts. However, the reliance on countries sticking to the Paris Agreement is needed in order to decarbonise Bitcoin.
Edit: Information on China’s energy policy
China uses the most cost effective renewables and could viably generate 60% of its energy with green energy by 2030. It is estimated the renewable energy implementation could save China around 11% in monetary cost. Feasibly following the laws of economics Bitcoin miners will be more likely to use renewable energy sources if they are cheaper. Therefore, reducing Bitcoin’s carbon footprint.

Notice that rejected energy accounts for around two-thirds of all electricity generation. This energy is produced but ultimately does not go to useful work. The amount wasted annually is around 66.7 quadrillions BTU’s (“quads”) of energy.  For perspective, that is the energy equivalent of wasting 2.3 billion metric tons of coal every year.

The potential for Bitcoin-powered renewables is already evident in China. A 2019 report by Coinshares found that approximately 75% of Bitcoin mining comes from renewable energy sources, much of which stems from newly created hydroelectricity. These new revenue streams have brought power plants online which otherwise would not have been economically viable given existing.

So, if we think of Bitcoin as a battery, what can we do with it?  The key properties of Bitcoin’s battery are: 1) always on and permissionless (no need to find customers, just plug and go) and 2) naturally seeking low-cost electricity: it will always buy when the price is right.

Given those properties, Bitcoin’s battery can assist renewable builds (and electric grids more generally) in a number of ways:

• Interconnection queues: when you develop new energy resources, you must apply to get them connected to the grid. Texas alone has over 100 GW of renewables in its queue. These queues can take years to clear. In the meantime, these assets could be online and earning Bitcoin.
  Project finance: Renewable developers need capital to finance build-outs before they have customers. Bitcoin’s battery is always ready to be the first customer.
• Geographic issues: Sometimes the sunniest, windiest places are not the ones with the most customers, so it’s hard to justify the development of new renewables. Bitcoin’s battery solves this, becoming a “virtual transmission line” of sorts.
• Timing & grid balance: Sometimes when the sun shines and when the wind blows is not when we need the most electricity. Yet, electric grids are marketplaces that must stay in perfect balance between supply and demand. Therefore, grid-connected renewables often have to “curtail” (turn off) if the are producing too much energy at the wrong time. Bitcoin’s battery is ready to buy 24/7/365 when the price is right, and turning up and down as needed, and participating via direct power purchase agreements as well as via demand response programs.
• Underperformance: Related to the timing & balance issues above, often times, renewables produce more energy than is needed on their grid, leading to subpar financial performance. Bitcoin’s battery is ready to buy if no one else will.
• Cleaning the grid: Even outside of renewable generation, Bitcoin’s battery can help improve both emissions and the energy mix. For example, Crusoe Energy attaches efficient turbines and mining equipment to existing gas flaring sites, both improving emissions and converting energy into Bitcoin’s battery. Taking this a step further, you could even then take those profits and reinvest them in on-grid renewables elsewhere, another twist on the idea of Bitcoin as a “virtual transmission line” (aka battery).

The third attack would be that Bitcoin would move 7 transactions per second and that in a nutshell for such a system even a few kilowatts now would be wasted (curious, then, that Attivissimo remembers this after accepting Bitcoin donations for years, on its blog, of this great problem).
Quickly: this number, unfortunately, arises from a profound misunderstanding of the real decoupling between the technological infrastructure called "blockchain" and the transfer of value in Bitcoin.
It is frequent, but this does not make it any less wrong.
Premise: "It's a feature!", Or rather the size of the Blockchain is deliberately small.
Being able to write on a blockchain must be a luxury, so that it remains decentralized, so that everyone, at any time and with a relatively low effort, can independently verify the correctness of transactions on the network, and not just the large banking institutions.
However, nowadays, at every moment of the day, Bitcoins, or contracts that are related to their value, are brokered and traded on cryptocurrency exchanges, on custodial platforms, by means of pegged tokens, on sidechains, on Lighting Networks, by means of CFDs. , with various levels of enforcement.
And this happens through a number of transactions that are much more than 7 per second, believe me! Claiming the opposite would be like expecting our morning coffee paid for at the bar to be recorded by all the backups of all the nodes of all the Eurozone interbank circuits.
So how many transactions, really?
In general, it is a non-measurable number, and it will be less and less, the estimates will be more and more heuristic, also thanks to privacy-preserving platforms.
We can say, to give a general idea, that every transaction that takes place on an order book of any platform, which represents Bitcoin, can only exist thanks to the fact that, underneath, there is the Proof Of Work which, when necessary, allows settlement of a precise state of a chain of transactions of indefinite length (my token goes to you, you give it to him, you give it to the other, who breaks it in three and gives it to the other, who collects 8 ). Aren't you going to expect all the coins you exchange to end up written down somewhere?
Quite simply, people every day rely on intermediaries to exchange Bitcoin value without using a blockchain directly.
The global concept of decentralization is maintained, relocation is increased with confidence in the last mile (not always! LN!) But the connection element always remains the last, only, true, mandatory, a digital ledger in which the compensation movements.
Comparing, therefore, the phantom 7 transactions (which is a number that was good in 2013, today there are many more even on-chain) to the world's ability to transact Bitcoin, and tie it to the PoW, is a bit like expecting that there are, at all times, a sufficient quantity of vans to move all the gold in all the vaults of the world that intend to exchange gold.

Bitcoin is a settlement system, not a payments aggregator

First things first. What is Bitcoin[1] and what is it not?

Bitcoin is a settlement system like FedWire, it is not a payments aggregator like Visa. I constantly see Bitcoin compared to Visa, MasterCard, or PayPal, and this is the main source of mathematical atrocities whereby Bitcoin’s overall electricity cost is divided by its transactions and then compared to something it’s not. Energy use per settlement transaction is a nonsensical metric by which to judge Bitcoin’s energy use.

Just like the 800,000 or so daily FedWire transactions are not a good measure of the total amount of daily Dollar (USD) transactions, Bitcoin’s 325,000[2] or so daily transactions are not a good measure of the total amount of daily bitcoin (BTC/XBT) transactions. Most bitcoin transactions are not visible. They take place inside the payment aggregation systems of exchanges, on the Lightning network, and yes, even inside of actual aggregators like PayPal, Square, or MasterCard. Only periodically are they settled onto the Bitcoin blockchain as visible transactions.

Solutions like this are referred to as network layering. This is a tried and tested approach to separating casual retail transactions from heavier settlement transactions and it is exactly how we already do things in the fiat monetary and payment systems. In such a system, the base layer, like FedWire (or Bitcoin), only acts as the final arbitrator of settlement transactions, everything else, and that is the vast majority of all transactions, happen in higher payment aggregation layers, which are often entirely different systems.

In other words, Bitcoin is not a competitor to Visa, MasterCard or Paypal. Bitcoin is an independent monetary system that aggregators can make use of.

Presenting Bitcoin’s electricity consumption in terms of its daily number of settlement transactions is a red herring.