Bitcoin has the potential to revolutionise financial markets and enjoys substantial global influence. However, the energy demands of bitcoin mining contribute to rising carbon emissions. In a study, Nuri C Onat, Murat Kucukvar and colleagues found that bitcoin mining has a significant carbon footprint embodied in its complex global supply chains. They write that the environmental consequences of digital currencies must be taken into account.
Each bitcoin transaction generates carbon emissions roughly equivalent to driving a gasoline-powered car between 1,600 and 2,600 kilometres. This highlights the high environmental cost of each transaction on the network and underscores the need for solutions to reduce this impact.
Bitcoin mining relies on the proof of work (PoW) system, in which high-performance computers work to verify transactions and secure the blockchain network. This process demands enormous amounts of electricity, estimated to reach around 63 terawatt-hours (TWh) annually, roughly the same as the annual energy consumption of Poland. As fossil fuels remain a primary source of electricity in many mining regions, this energy use translates directly into greenhouse gas emissions, which drive global climate change. The energy consumption of bitcoin mining adds a considerable amount to global carbon emissions and are comparable to the annual emissions of entire nations.
Where are the emissions?
The findings reveal that approximately 46 per cent of global Bitcoin mining emissions originate within the United States, producing around 15.1 million metric tons of CO₂ annually. Despite regulatory restrictions, China remains a major player, both as a significant emitter of bitcoin mining and as a top producer and supplier of bitcoin mining equipment. Kazakhstan and China also contribute significantly, with 20 per cent and 13 per cent of total mining emissions, respectively.
Due to emissions embodied in the global supply chains of bitcoin mining, regional emission reduction efforts alone may not be enough. For example, Norway, despite its clean energy grid, also faces indirect emissions due to bitcoin mining. Around 74 per cent of Norway’s mining-related emissions stem from imported equipment manufactured in regions like China, where carbon-intensive energy sources are prevalent. This underscores the need for mitigation strategies that go beyond national borders, accounting for emissions embedded in global trade and supply chains. Additionally, the findings highlight that coal-powered electricity generation remains a primary source of greenhouse gases, calling for a strategic shift toward cleaner energy alternatives.
Broader implications: beyond bitcoin
Bitcoin’s energy demand is indicative of a broader trend among emerging technologies. Advanced artificial intelligence models, for instance, also require significant computational resources for development and operation. As these technologies expand, so too will their carbon footprint, highlighting the need for sustainable approaches across digital sectors. If left unchecked, the energy requirements of these innovations could pose further challenges to global climate mitigation efforts.
Building a sustainable digital future
Mining operations powered by renewable energy, such as solar or wind, could drastically reduce emissions. Policymakers could incentivise or require mining hubs to rely on clean energy sources, particularly in regions with abundant renewable resources.
Additionally, exploring methods for implementing a carbon tax could help reduce the growing emissions from bitcoin mining. Although the decentralised and largely unregulated nature of cryptocurrencies poses challenges to the regulation of carbon taxes, innovative solutions within the crypto space are possible. For example, blockchain technology could be leveraged to help lower mining-related carbon emissions.
Tackling emissions associated with bitcoin’s entire supply chain is essential. This includes considering the carbon footprint of manufacturing and transporting mining equipment. Transparency in carbon accounting, especially for indirect, so-called Scope 3, emissions could help hold mining operations accountable for both direct and indirect emissions.
To reduce the reliance on energy-intensive proof of work systems, the industry could consider adopting alternative consensus mechanisms like proof of stake (PoS), which has a significantly lower energy requirement, offering a viable path toward sustainable digital currency models. However, this is not a simple solution, as various consensus mechanisms can impact bitcoin’s value as a decentralised and trustless financial asset, presenting a challenge often referred to as the blockchain trilemma.
Although bitcoin’s carbon emissions are concerning, the global financial system’s energy demands in hubs like New York, London and Tokyo are also significant, relying on energy-intensive servers and supercomputers for transaction processing. Therefore, strategies such as improving data centre efficiency, using renewable energy and advancing cloud computing can help reduce emissions from emerging digital technologies. Yet, as financial transactions grow, finding sustainable ways to reduce the sector’s environmental impact remains challenging.
There is an urgent need to address bitcoin mining’s carbon footprint, not only for cryptocurrency but as a blueprint for managing the environmental impact of digital technologies, including AI as well as the financial sector at large. With coordinated action, the cryptocurrency sector can become a model of sustainability, balancing innovation with environmental responsibility. Addressing bitcoin’s emissions, along with those of other emerging technologies, will be crucial as we strive to meet global climate targets and mitigate the risks of climate change.
Sign up for our weekly newsletter here.
- This blog post is based on Carbon footprint of global Bitcoin mining: emissions beyond borders, by Nuri C Onat, Rateb Jabbar, Murat Kucukvar, Tadesse Wakjira, Adeeb A Kutty and Noora Fetais, in Sustainability Science.
- The post represents the views of the author(s), not the position of LSE Business Review or the London School of Economics and Political Science.
- Featured image provided by Shutterstock
- When you leave a comment, you’re agreeing to our Comment Policy.
Print Friendly
“Bitcoin has the potential to revolutionise financial markets“
Bitcoin has had a long time to do this but so far has only created new forms of financial speculation and facilitated the explosive growth of Ransomware.
I addressed the possibility of sustainable or green Bitcoin in 2021.
https://blogs.lse.ac.uk/businessreview/2021/05/25/how-green-is-my-bitcoin/
The per transaction method of evaluating bitcoin energy use cited in this article, and upon which the conclusions of this article rest, has been refuted by four subsequent authors in academic literature including Sai & Vranken (2023). Of the most recent 12 peer reviewed publications on bitcoin and energy, 11 of them have stopped using this metric, as it leads to incorrect conclusions about bitcoin environmental impact.
There are other issues in this post also, which directly impact the conclusions reached. For example, the data used is now more than 3 years old. This matters materially because high-fossil fuel using countries such as Kazakhstan now have very little bitcoin mining occurring (River survey, 2023)
Finally there is a large body of peer reviewed research that suggests positive environmental externalities from bitcoin mining. These have been documented by the Digital Assets Research Institute (da-ri.org). Any objective analysis of bitcoin mining impact should consider both positive and negative externalities.
Thank you for your comment, Daniel. The emissions are not based on a per-transaction method; rather, it’s an alternative way of presenting information about total annual emissions. Our estimates are calculated per hash rate, though this isn’t a term widely recognized by the general public.
Additionally, emissions estimates for Bitcoin mining are constantly changing, making precise calculations or forecasts difficult. However, these estimates can still underscore the significance of this growing issue.
The energy sources used by miners remain dynamic. For this reason, I believe that more systematic and sustainable solutions are necessary, considering fluctuating nature of emissions, hash rates, energy demand, etc.
Please see the article for more details: https://link.springer.com/article/10.1007/s11625-024-01576-5
You say emissions are not on a per transaction method, but that is exactly what the article opens with:
“Each bitcoin transaction generates carbon emissions roughly equivalent to driving a gasoline-powered car between 1,600 and 2,600 kilometres.”
Also, here is a supporting info for those who wonder what functional unit is and what transaction means
Transaction is used a functional unit of Bitcoin system in the analysis. A functional unit (FU) is a measurable representation of a product’s function, serving as the standard reference point for all impact assessment calculations. In addition, other functional units such as “per hashrate” and “total emission estimations” are estimated and on average transaction emissions are calculated derived from hashrate-based estimation methodology. Results from Transaction-based, ownership-based, and hashrate-based approaches can vary depending on the definition of “transaction” as well as analysed time period.
A Bitcoin transaction is a digitally signed message that transfers ownership of bitcoins from one address to another within the Bitcoin network. Each transaction comprises inputs—references to previously unspent transaction outputs (UTXOs)—and outputs, which define new UTXOs assigned to recipients’ addresses. This structure ensures that bitcoins are spent only once, preventing double-spending. All transactions are publicly recorded on the blockchain, a decentralized ledger maintained by a network of nodes, providing transparency and security.
Tony,
We’ve calculated emissions for both metrics. Total emissions are estimated based on the hashrate, and emissions per transaction are derived by dividing the total emissions by the total number of transactions—providing an average emissions figure per transaction. This method remains accurate, though it’s not a bottom-up approach that calculates emissions by estimating those embedded in each transaction. Instead, it’s a top-down approach.
That said, the key issue is still that emissions from Bitcoin mining are on the rise, and we should focus on finding solutions rather than questioning the study’s estimations based on a single statement. It’s essential to understand the calculation method before drawing conclusions. Apologies for any confusion caused.
Furthermore, here is a text from the study for your reference about the functional unit (please refer to life cycle assessment methodology to understand what functional unit is)
“We point out that the studies of McCook (2018) and De Vries (2018) use a single hash rate value despite calculating
the total hashes mined annually while quantifying the total energy consumption stemming from Bitcoin mining. In these studies, along with the study by Mora et al. (2018), we follow the argument that Bitcoin mining depends on the use of per transaction as an activity variable when calculating the Bitcoin energy consumption. The authors follow the
argument based on the fact that Bitcoin mining depends on the number of blocks that require “proof-of-work” (PoW) to be solved along with the present level of difficulty of the network and not on the total number of transactions alone. Further, the average time required per block to be resolved remains constant at 6 blocks per hour (Masanet et al. 2019). At every 2016 block, the difficulty is adjusted to maintain a buffer of 10 min before the introduction of each block (Narayanan et al. 2016; Stoll et al. 2019). Thus, with an increase in the number of transactions, the number of blocks will not increase, rather it will lead these blocks to accommodate more transactions, where the block transaction limit may be scaled over time. The “SegWit” soft fork change in the transaction batching and the proposed increase in the block size for SegWit2x are some examples to account for this raised proposition. It is thus evident that there exists no relationship between the transactions in a block and the effort to mine the block. A similar argument was raised by Masanet et al. (2019) claiming the use of transactions as a design flaw when calculating the associated energy consumption and CO2 emissions associated with Bitcoin mining. However, translating Bitcoin energy consumption into CO2 emission is arguably inaccurate and can cause further inconsistency in the results.
Adhering to these caveats, we ground our calculations for the energy consumption based on per total hashes required to solve a block mined annually and not per transaction alone.
The energy consumed by the Bitcoin network was calculated and the results were consistent with recent studies that used similar assumptions while quantifying the energy consumption (CoinShares Research2022). This consistency provides confidence that our methodology in this aspect is robust and dependable.
Please see Table 2 Recent studies on the carbon footprints of Bitcoin mining Estimation Publication year Title Authors
35.17–71.36 Mt CO2e 2023 Carbon Footprint Comparison of Bitcoin and Conventional Currencies
in a Life Cycle Analysis Perspective Pagone et al. (2023)
85.89 Mt CO2e 2023 The environmental footprint of bitcoin mining across the globe: Call
for urgent action, Chamanara et al. (2023)
64.18 Mt CO2e 2023 An analysis of energy consumption and carbon footprints of cryptocurrencies and possible solutions Kohli et al. (2023)
90.6 Mt CO2e 2024 Energy, Water, and Carbon Footprints Landscape of Cryptocurrencies
and Sustainable Solutions 2024 Laimon and Almadadha (2024)
Furthermore, total estimations we calculated might be even “underestimated”. For example, in recent studies:
Laimon and Almadadha (2024) 90.6 Mt CO2e,
Kohli et al. (2023) 64.18 Mt CO2e,
Chamanara et al. (2023) 85.89 Mt CO2e,
Pagone et al. (2023) 35.17–71.36 Mt CO2e.
In our study, we found that annual emissions are 32.74 MtCOe (only).
Pagone E, Hart A, Salonitis K (2023) Carbon footprint comparison of bitcoin and conventional currencies in a life cycle analysis perspective. Proc CIRP 116:468–473
Chamanara S, Ghaffarizadeh SA, Madani K (2023) The environmental footprint of Bitcoin mining across the globe: call for urgent action. Earth’s Future 11(10):e2023EF003871
Kohli V, Chakravarty S, Chamola V, Sangwan KS, Zeadally S (2023) An analysis of energy consumption and carbon footprints of cryptocurrencies and possible solutions. Digit Commun Netw 9(1):79–89
Laimon, M and Almadadha R (2024) Energy, water, and carbon footprints landscape of cryptocurrencies and sustainable solutions. SSRN 1-15 (preprint)
It is rightly said ‘all that glitters is not gold’. This applies to bitcoin mining too. Amidst a global craze and craving for bitcoins, a path-breaking study by Professors Nuri C Onat, Murat Kucukvar and their associates draws attention to an alarming level of carbon footprint linked to complex global supply chains of bitcoin mining. The authors cogently argue that detrimental environmental impact of digital currencies ought to be taken seriously. It is essential to tackle bitcoin mining’s carbon footprint both for cryptocurrency and as a prototype for managing the environmental impact of digital technologies like artificial intelligence as well as the financial sector at large. By adopting a series of joint actions, the cryptocurrency sector can become sustainable that would reconcile innovation with environmental responsibility. In order to meet global climate targets and mitigate the risks of climate change, it is crucial to handle bitcoin’s emissions and those of other new technologies. The necessary measures include actions such as improving data centre efficiency, using renewable energy and advancing cloud computing. technologies.
Respectully, this is not a robust rebuttal. You say “The emissions are not based on a per-transaction method” however you are still using the methodology, which has been discredited in 4 separate academic journals. It is inappropriate to mention this method, other than to refute it which your paper does not do. It is therefore misleading, and serves to perpetuate the misunderstanding that Bitcoin emissions can be measured per transaction.
Furthermore, the data is this paper is now several years out of date (Kazakhstan now represents under 1% of global hashrate according to contemporary datasets (River).
The calculations also do not take into consideration Cambridge’s notes on their own website
1. that the mining map is now almost 3 years old and “likely overestimates emissions by 25%” based on this one factor along
2. Does not include factors that can “be reasonably expected to decrease emissions” such as methane mitigation
3. Does not include factors that can “be reasonably expected to decrease emissions” such as offgrid mining
The use of the discredited and widely disputed work of Mora et al, de Vries, even if you did not use their calculations, together with the exclusion of the majority of papers on Bitcoin and energy most of which demonstrate positive environmental externalities, points to significant gaps in the prior literature search process.
Respectfully, I do not find the study to be path-breaking at all. Please state in what way you find it to be path-breaking specifically? In fact, I would counter-argue that it has more in common with the first iteration of scholarship on Bitcoin mining between 2018-2022, much of which has now been refuted, while ignoring the large body of work demonstrating the counter-intuitive but very real and measurable positive environmental externalities such as the decarbonizing impact of Demand response, frequency regulation, methane mitigation, heat recycling and more. Any objective measure of environmental impact must examine both positive and negative externalities. While I respect the authors, I find this particular expression of their work to be lacking by that objectivity benchmark.