This story was originally published by Slate and is reproduced here as part of the Climate Desk collaboration.

Last April, the cryptocurrency world announced its own virtual iteration of the Paris Agreement: the Crypto Climate Accord. The alliance bills itself as “a private sector-led initiative for the entire crypto community focused on decarbonizing the cryptocurrency and blockchain industry in record time.” Its goal is to transition the crypto industry to renewable energy sources in time for the 2025 United Nations climate conference. By 2040, it seeks to “achieve net-zero emissions for the entire crypto industry.”

Why does crypto need its own climate pact? Because it has a massive carbon footprint, one that’s kept growing as interest in cryptocurrencies — not to mention the sheer number of cryptocurrencies — has grown. A 2019 study in the science journal Joule estimated that, at the lowest bounds, Bitcoin’s power consumption emitted about 22 million metric tons of carbon dioxide the previous year. For context, that’s about 10 percent of the global railway sector’s annual emissions — and it’s just for one currency, even if it’s a major one. Such figures are a bad look for the industry’s public image, which is why phrases like “green crypto” and “clean crypto” are suddenly popping up everywhere, fueling efforts like the new climate accord. Crypto’s dirty reputation is an existential problem — so for the sake of both the planet and the industry, it’s worth examining how the many, many “clean” crypto initiatives, currencies, blockchains, and marketplaces for non-fungible tokens, or NFTs, actually stack up.

Here are just a few of them: In early 2021, artist Memo Akten wrote “A Guide to Ecofriendly Cryptoart (NFTs),” pointing not only to more “sustainable” networks but also to methods that could make existing blockchains more energy-efficient. Artist Damien Hirst, back in April, linked 10,000 physical artworks to NFTs verified on the “eco-friendly” ledger Palm. Lists of supposedly environmentally friendly coins and networks often include Tezos, an open-source platform backed by Quincy Jones; Cardano, currently the eighth-largest cryptocurrency in terms of market capitalization; and Ripple, which is in the midst of an SEC brawl as well as a sharp drop in value following January’s crypto crash. That same month, the crypto company EGridd announced a “green” blockchain system powered by magnetic generators.

Grist thanks its sponsors. Become one.

Crypto’s primary energy issue arises from the dominant method of mining coins, known as “proof of work.” I’ve previously explained how proof of work operates on chains like Ethereum: To earn coins, which are heavily encrypted in blocks in order to ensure security and authenticity, miners use their processors to solve complex puzzles for individual blockchain entries. Their systems have to figure out a particular block’s matching “key,” which is a code written by an algorithm, and match their own guesses to the block until they find the correct code line. Standard processing units are usually too inefficient to keep up with such needs, while the processing machines that are more energy-efficient can usually only be used for Bitcoin mining. These more economical processors tend to have a short life span, so when they die, they simply pile up as e-waste. That’s all just for extracting the coins, mind you — sending them to other users for transactions also requires hefty amounts of electricity, as the majority of the computers cued in to the blockchain must verify each transfer in order to ensure its legitimacy.

This applies to any asset minted and transferred on a proof-of-work blockchain, such as NFTs on Ethereum. According to the most recent estimates from the Bitcoin Energy Consumption Index, a data project headed by Dutch economist Alex de Vries, a single transaction of the world’s most popular cryptocurrency uses about 2,157 kilowatt-hours of energy, an amount that could power one household for about 74 days. In 2021, there were roughly 95 million total transactions, adding up to emit an estimated 65 million metric tons of carbon dioxide and exceeding the annual emissions of Papua New Guinea. China and Kazakhstan, formerly Bitcoin-happy nations, have straight-up halted crypto farming within their borders in order to take care of this energy problem. According to a Citigroup report, the energy now used for global Bitcoin operations is about 66 times greater than it was in 2015.

One can’t be blamed for being skeptical of institutional fixes: The supposedly less energy-intensive iteration of the widely used Ethereum network, known as ETH 2.0, was supposed to go live years ago; Ethereum’s founders now promise it will arrive in complete form sometime this year.

The Crypto Climate Accord wants to start fueling crypto with renewables as opposed to fossil fuels, but at the moment, that simply isn’t an option. We don’t have enough renewable energy around the world to meet climate goals even without taking crypto into account; running crypto systems will require that major countries have surplus renewable-produced energy. Already, areas with dedicated green power sources for crypto, like the Nordic states, are running low on the surplus power capacity required for digital mining. Bitcoin’s energy use has shot up over the past year, and Scandinavia’s supply of excess power — about 30 terawatt-hours in an average year — is projected to decline as governments redirect it toward the development of fuels like hydrogen, while also exporting clean power to the rest of Europe.

Grist thanks its sponsors. Become one.

Few countries have majority-renewable-powered grids, and those that do, like Iceland and Denmark, already use much less energy than the countries that tend to be major crypto hubs, like Nigeria and the Philippines. The countries friendliest to crypto have decades to go before they meet the clean energy benchmarks required to power most of their regions and clean up their emissions — much less hold enough excess capacity to regularly mine Bitcoin. (Before China banned crypto mining last year, it provided the largest source of renewable energy for crypto operations.)

There are also crypto advocates who put forth dubious cases for digital currencies they claim are actually paving the path for clean power. Jack Dorsey’s company Block, back when it was still known as Square, released a white paper claiming Bitcoin mining is necessary to incentivize the scaling of renewable energy, an argument that doesn’t quite hold up to scrutiny or play out in practice. Many green-blockchain advocates tout their purchasing and trading of carbon offsets, but these so-called offsets often only add to carbon emissions; others advertise themselves as “carbon-neutral,” promoting a shaky concept that’s mostly allowed energy firms aiming for “net-zero” emissions to not substantively reduce their carbon footprints.

So there are a lot of “green crypto” initiatives that are easy to dismiss as pure hype. At the same time, there are many digital traders, artists, engineers, and true believers who have been working for years, out of genuine concern, to try to build and scale solutions to crypto’s environmental problem.

One of their biggest tools at hand involves alternative mining programs. In 2012, two developers introduced a new practice known as “proof of stake” to help make ledgers’ transaction fees cheaper and scale crypto mining to a level Ethereum couldn’t yet reach. Proof of stake is less energy-intensive than proof of work, since it doesn’t require every miner to be online in order to verify transactions. Instead, certain participants — those who’ve been in a given network longer than others and have the most currency invested into a collective “staking pool” — are randomly selected as “validators,” who can verify transactions and update the blockchain. If they execute their high-knowledge duties sufficiently, they will gain tokens; if they go offline or validate bad transactions, they can lose their privileges and their coins. An offshoot of proof of stake known as delegated proof of stake allows users to pour their currencies into staking pools that can each be attached to a particular “delegate,” who’s directly elected by network denizens by a vote. These delegates are assigned block by block in the system; no one leader may hold the reins for too long.

The primary benefit of proof of stake is the fact that not as many processing units are needed to run mining operations; Ethereum estimates this can slash energy requirements by up to 99 percent. There’s a similar benefit to proof of authority, the method invented by one of Ethereum’s co-founders and championed in Damien Hirst’s massive NFT drop. The difference between proof of authority and proof of stake is that validators are preapproved to validate transactions based on their reputation in the market rather than by just their value holdings, and these validators are not allowed to verify consecutive blocks.

That sounds promising from the outside. But those drawn to crypto for the security and “decentralization” may not care for proof of stake or proof of authority as much. It’s far easier for a bad actor to dominate a proof of stake space than a proof of work one, since for the former you’d only need to hold more than half the currencies on hand, while for the latter you’d need to control more than half the hardware infrastructure. And it’s easier for validators and delegates in proof of stake systems to keep gaining coins than it is for other users; plus, the technical skills required for such roles have a high barrier to entry. Some users think this ruins the whole point of crypto, which is ostensibly to democratize a financial system that runs parallel to fiat currency — per a recent Guardian report, “a majority of bitcoiners remain adamant that [proof of stake] is not a worthy replacement for [proof of work].” That may explain why, 10 years after its introduction, proof of stake’s most popular networks and currencies, like Tezos and Solana, are still nowhere near as dominant in the market as proof of work–based ledgers are.

There’s another alternative called proof of space, which allows mining to be powered by smaller and more efficient hard discs rather than massive processors. Just last year, a proof of space currency named Chia gained popularity in China so quickly that it led to shortages of hard discs along the supply chain. But then China banned crypto mining, and proof of space tokens haven’t recovered from that setback.

What about climate-conscious miners who favor Ethereum and proof of work’s security but not its power consumption? The “Guide to Ecofriendly Cryptoart” mentions a few potential fixes within Ethereum itself. Lazy minting is the practice of delaying the actual creation of an NFT on the relevant ledger until it is purchased; though it’s more cost-effective for NFT creators, it’s not necessarily much more energy-efficient. “Bridges” allow users to transfer their data from one blockchain ledger to another if they wish, which certainly provides a useful offramp but doesn’t fix the issues within ETH. So-called sidechains and Layer 2 scaling solutions allow Ethereum to be “bridged” to parallel blockchains that may use proof of stake, giving such chains a core security while not necessitating the same high-cost and high-energy transactions for mining and trading tokens. Yet all this still lets ETH off the hook for its energy-intensive nature.

At this point, it’s difficult to see an immediate fix to crypto’s energy woes, more than 12 years after it debuted. To be fair, many crypto evangelists make the accurate point that video gaming, streaming, and other data-heavy internet habits are also extremely energy-intensive; a 2019 report from the California Energy Commission found that video gaming within the state alone consumed about as much electricity as the entirety of Sri Lanka.

Still, each of the aforementioned activities is a unique source of entertainment as well as of cultural and informational value. Crypto, on the other hand, is a speculative asset that requires mass faith and collective buy-in over regulation and universally established commerce. If you believe in crypto’s mission but worry for our natural environment, it’s certainly more advisable to work in a proof of stake system than a proof of work one. But then you’re working in a system that negates some of the very benefits of security and decentralization crypto was intended to offer — and that still has a carbon footprint all its own, even if it doesn’t equate to Bitcoin’s. Just because one system isn’t as environmentally ruinous as another doesn’t mean that the system is itself “green.” Add to all this the scams and fluctuating values that rock crypto on a regular basis, and it becomes increasingly difficult to see just what it is we’re doing here.