This article is an edited version of one that was published on The Bitcoin Adoption Forecast. It is reproduced in partnership.
It is rapidly becoming clear that Bitcoin mining can now legitimately be called “greentech”. An overwhelming majority of recent peer-reviewed academic papers on Bitcoin mining and energy have now established its significant environmental benefits.
But as the argument that “Bitcoin uses too much energy” has over time been substantively disproven, another myth remains a vector by which critics attack Bitcoin’s environmental impact: e-waste, or the idea that retired ASIC mining equipment has a polluting effect.
However, this argument is just as flawed as the “too much energy” one; it’s just that it hasn’t yet been formally debunked. Until now.
This week, the Cambridge Centre for Alternative Finance published a wide-ranging report on the digital mining industry that suggested that only a small proportion of mining ASICs are retired every year, and of those, nearly 90% are recycled, sold or repurposed.
Recent research by the team at DARI have led to similar conclusions – that Bitcoin mining generates only a very small amount of e-waste – as the ASICs remain profitable for much longer than previously thought, and that there are are very strong incentives for mining companies to recycle or reuse their equipment. These findings, which are corroborated by the independent conclusions of the Cambridge Centre for Alternative Finance report, provide very strong evidence that Bitcoin contributes very little to global e-waste.
Origins of the Myth
Looking at the real picture of Bitcoin and e-waste, it’s crucial to see where the myth originated.
Back in May 2018, Dutch central bank employee Alex de Vries published a five-page commentary called “Bitcoin’s Growing Energy Problem”. While this piece was not on e-waste, the commentary marked the invention of a technique for measuring Bitcoin resource use “per transaction”.
De Vries later applied this same logic to e-waste in a 2021 paper titled – you guessed it – “Bitcoin’s growing eWaste problem”.
And thus, the e-waste myth was born.
It’s worth noting, that de Vries’s article was a “commentary” article, not research. A commentary is similar to an academic opinion piece. It is used to forward the author’s perspective without the need for new empirical data. Commentaries are also restricted in the number of references they can use. Meaning that often these kind of publications can make claims that are unverifiable due to the lack of traceability of sources.
The crux of de Vries assertions in this commentary was: “We can measure Bitcoin’s use of energy per transaction to show that the Bitcoin network cannot scale without vastly increasing emissions”.
His commentary gained significant traction. In 2017 there was one article on Bitcoin’s energy usage. In 2018, the year of de Vries’ commentary, there were more than 400. Each of these articles either draws directly from de Vries’ flawed paper, or indirectly does so by referencing one that does.
The problem is that deVries’ perspective was inaccurate because Bitcoin’s energy use does not come from its transactions. Bitcoin could handle a thousand times more transactions, yet use no more energy. Indeed, a study by researchers at Cambridge University’s Centre for Alternative Finance describes the per-transaction measure as “not a meaningful metric”.
But the media and other authors kept citing it anyway. Then the regulators, central banks, and NGOs joined them.
The per-transaction metric was refuted a further four times, this time in academic journals (Masenet et al. 2018, Dittmar and Praktiknjo 2019, Sedlmeir et al. 2020, and Sai and Vranken 2024). But the truth did not stand in the way of a good story. The metric not only kept on spreading, it mutated.
In total, de Vries’ energy-per-transaction metric evolved to include three other variants:
- Emissions per transaction
- Water footprint per transaction
- e-waste per transaction
The Reality
Now, what really happens to old mining rigs at the end of their useful lifespan?
To be sure, e-waste is a real ecological issue. It’s not just the space it takes in landfills, it’s the fact that it often contains toxic materials such as lithium and cadmium. UNICEF has shown, for example, that smartphone e-waste, a major contributor to global e-waste, can negatively impact the health of children in the developing world.
E-waste is an emotive issue. As suchit’s important that we conduct a nuanced analysis when looking at claims concerning Bitcoin mining and e-waste.
Unlike cell phones, it turns out that Bitcoin miners (ASICs) 1) do not contain heavy metals, 2) can be easily and profitably recycled without landfill waste, and 3) are used nearly three times as long as the average smartphone.
Like the per-transaction methodology flaw and other errors in previous studies, de Vries’ paper had a basic data flaw around the claimed age at which mining units reached end-of-life.
For one, de Vries claimed that Bitcoin mining rigs were used for 1.12-2.15 years. This range was given without empirical justification, however. The life of a latest-generation Bitcoin mining rig (the S19 series) has actually been over 5 years. De Vries’ research assumptions were off by at least 132%-346%.
But there’s an even more significant omission. It turns out that for the most part, Bitcoin e-waste isn’t actually wasted. Yes, you heard that right. Bitcoin mining e-waste is an oxymoron.
It makes very little sense for a Bitcoin mining company to dispose of ASIC miners into a landfill. This is forthe simple reason that unlike so much other e-waste, it’s profitable to recycle them. It can even generate carbon credits to do so. For example, Hut8, a publicly traded Bitcoin mining company, generated 7,500 carbon credits through its most recent recycling of their e-waste.
Similarly, ERS, a global electronics reuse and recycling company that has contracts with five of the top eleven Bitcoin mining companies in the world, confirmed that end-of-life ASIC miners are actively being recycled by Bitcoin mining companies. They also reported that most of the large operators are “receptive to having a sustainable strategy in place to handle mining rigs that have been deemed obsolete”.
The reason for this is the same as why Bitcoin mining companies generally use sustainable energy sources: It is simply more profitable to “do the right thing”. Mining companies earn a profit share on money earned by selling material for recycling.
It’s a fast process too. In November 2023, Hut8’s recycling of 369 metric tonnes of e-waste occurred in just over three weeks. These were mostly obsolete ASIC miners, but also hash boards, cables, and other components. The materials were either sold to third parties for reuse or recycled. None of the materials were sent to landfills.
The reason that Bitcoin mining equipment is easier to recycle than other e-waste is that most mining rigs are cased with recyclable aluminum or steel, while the hardware contains precious metals and resources that can be extracted and reused.
Even the processors on the hashboards are recyclable due to the nature of what they are composed of. For instance, they can contain silver, gold, steel, aluminum, or nickel.
Meanwhile, Bitcoin mining rigs do not contain the toxic metals commonly found in consumer electronics. The most toxic material is the plastic that the fans are made of.
As far as the ASIC miners go, the hash boards are further processed for the metals. The gold and other precious metals are extracted, and what’s left over is pulverized into almost a fine powder, for further processing. The plastics or fiberglass (or a mix of both) that are left on the board then get sent downstream to certified smelters and refiners. And as the industry slowly shifts to immersion cooling for Bitcoin mining, even the need for plastic fans will be eliminated in the future.
Moreover, new AI nanotechnology can now identify various materials at the molecular level for cleaner sorting in industrial recycling as well as electronics (the main benefit is in sorting plastics, but the technique can be applied in many other areas).
Where is the End of Life for Mining Rigs?
So while Bitcoin miners are highly recyclable, they also are productive for far longer than de Vries suggests.
Unlike other electronics, Bitcoin mining rigs are often utilised for secondary purposes after being removed from the racks in data centers.
Even after five or six years, mining companies will often continue to use mining rigs, before sending them to get recycled.
In 2021, Nic Carter observed that as mining companies bought the then-new S19s to mine “around the clock”, they often redeployed old S9 miners to sites powered by intermittent energy sources such as solar and wind.
The logic behind this is that this form of electricity is often cheaper, free or even negatively-priced because it is surplus energy that otherwise would have been wasted, so it makes sense to mine using old S9s that would not be profitable around the clock due to their lower hashing rate (meanwhile, it also does not make sense to confine a new machine to a job for only 10 hours per day).
Intermittent renewable energy is the best candidate for older mining machines that have already recouped their original capital cost and can now afford to mine intermittently on cheap energy such as solar, wind, or even surplus hydro-power in a country with a large seasonal rainfall asymmetry.
Today, mining firms are looking to redeploy old S17s and even S19s to do the job done by S9s in the last cycle. The S9s are finally getting recycled, in some cases a full 8 years after their mid-2016 release. The units themselves, or their hashboards after recycling the casing are often used as self-subsidising household or office heaters.
But there’s yet another alternative use for old mining rigs. Old Bitcoin miners can also be used as “load banks”. Load banks are probably the only piece of equipment ever designed for the sole purpose of wasting electricity.
Sometimes, renewable generators need to curtail energy suddenly, but powering down can create wear and tear on their plant. So, many renewable operators prefer to send some of that power to a load bank, which does nothing other than chew up surplus electricity.
Load bank on the left. Bitcoin mining rig on the right (not to scale)
A 500 kW load bank will set you back $28,490 and is a more challenging e-waste conundrum than any Bitcoin miner. When renewable operators instead use obsolete Bitcoin miners as de facto load banks, it is not only cheaper for them and better for the environment, but they can also earn Bitcoin.
In Summary
Trying to slap the e-waste complaint on Bitcoin mining rigs is a form of gaslighting. Rather, the rigs are probably the only piece of electronics that do not present humanity with an e-waste challenge.
Unlike some consumer electronic e-waste, they:
- last much longer
- don’t have toxic components
- can obviate other wasteful pieces of equipment, even in “retirement”
- can be recycled easily and profitably, with no residual landfill waste.
Bitcoin mining is a story where the economic incentives just seem to continually align with “doing the right thing”.
Because 80% of Bitcoin mining companies’ operational budget is electricity, and the cheapest electricity these days is stranded energy and renewable energy, Bitcoin miners seek out stranded renewable energy.
Because older mining rigs can still profitably mine Bitcoin when electricity prices go very low or negative, they help load-balancing grids struggling with power oversupply while encouraging mining companies to keep old miners operational for as long as possible.
Because Bitcoin mining rigs are easy and profitable to recycle, mining companies are incentivized to take them to equipment recyclers rather than throw them in a landfill at end-of-life.
The difference in viewpoint about the environmental sustainability of Bitcoin can be put down squarely to the amount of research someone is willing to do.
Those unwilling to research how Bitcoin mining and its built-in economic incentives work will continue to rely on junk science commentaries and media reports that picked up those commentaries without critical evaluation.
But those who are willing to investigate the economics of Bitcoin mining will find a plethora of organic incentives playing out to encourage sustainable behavior. None of this needs to be enforced or promoted by any authority.
