Environmental Problems

“At present, we are stealing the future, selling it in the present, and calling it gross domestic product.”

– Paul Hawken

The technical inefficiencies of cryptocurrencies are the mark of a technology that is over-extended and not fit for purpose. However, what is even more concerning is the environmental footprint these technologies introduce into the world. Bitcoin and currencies that use proof of work consensus scheme require massive energy consumption to maintain their networks. This feature is central to their operation and is the mechanism that allegedly “builds trust” in the network. No network participant has any privileged status except in the amount of energy they expend to maintain the consistency of the network itself. The amount of energy spent in this global block lottery results in an expected direct return per watt, which is statistically predictable. In a nutshell, the premise of mining is to prove how much power one can waste, and the more power one can waste, the more resources one receives in return. The system is fundamentally inefficient in its design (Gerard 2017; Weaver 2018; Küfeoğlu and Özkuran 2019).

The impacts of new technology have to be considered within a framework of environmental shortcomings since the stakes of our era are staggeringly dire. Climate change is happening now, the global temperature on earth is rising, and the polar ice caps are melting. In Greenland, the glaciers are receding six times faster than expected. Climate change is not an abstract phenomenon happening elsewhere; it is happening everywhere ever day, and technology plays an important part in the future of our planet.

Wasteful Mining

The bitcoin network automatically adjusts the difficulty of mining so that each block takes an average of 10 minutes to mine. If a miner performs a fixed number H of hashes every 10 minutes, each hash has a 1/H chance of mining a block. If a miner performs N hashes every minute, the number of blocks they can expect to find per minute follows a binomial distribution with maximum N and with probability 1/H. We can then compute the expected value of the binomial distribution, calculated by multiplying the number of trials by the probability of success. The expected return is the expected value of computing a block times the block reward amount (currently 6.25 bitcoins). This expected return can be calculated as a single variable function: the number of hashes a miner can compute per second. This unit is dimensionalized in hashes/second and often denoted in megahashes (106 or Mhash/s), gigahashes (109 Ghash/s), or terrahashes (1012 Thash/s). At the time of writing, for a data center to consistently produce an expected return of 1 bitcoin per day would require approximately 140,000 Thash/s. These computations are spread over many racks of computing devices running in parallel(Stoll, Klaaßen, and Gallersdörfer 2019).

Off-the-shelf computers can be used for building cryptocurrency mining rigs. However, the performance of these devices is suboptimal compared to dedicated mining hardware. A top-of-the-line CPU such as the Intel i9 processor can perform 5.2 Mhash/s with an expected return of 0.00000001 bitcoin per week or at the time of writing $0.00012 per week. Mining using standard CPUs found in most computers is generally infeasible. However, a single specialized graphics processing unit (GPU) such as an Nvidia GeForce RTX 2070 uses a model of execution with a great deal more parallelism, can have up to 42 Mhash/s, and has an expected return of 0.0051 bitcoin per month or $60.30 per month. One of these graphics cards retails between $500 and $1000 per device and consumes 214 watts of power.

As many people learned during the California gold rush, sometimes the most profitable activity is not the mining itself but selling picks and shovels to the miners. Globally there is a cottage industry of services and providers selling cryptocurrency mining wares, and this global trade has drastically driven up the price and lowered the supply of high-end graphics processing hardware. Cryptocurrency’s volatility has led to the usual economics during speculative bubbles. Bitcoin has created a veritable arms race of mining equipment that attempts to optimize hashes per watt. This context makes a massive network of data centers worldwide, all clustered around areas of cheap(Dindar and Gül 2021; Benetton, Compiani, and Morse 2021) power where the input capital per watt can yield an optimal return on investment. Regions such as Siberia in Russia, Texas in the United States, and Xinjiang in China have seen upticks in cryptocurrency mining activity due to their geographic proximity to cheap fossil fuel power.

However, the environmental impact is that we are drawing more power from the grid, burning more fossil fuels to maintain this cryptocurrency network, and lining the pockets of cryptocurrency miners. For the bitcoin network, with only 5% of activity corresponding to economic transactions (Vigna 2019), this would result in a truly staggering amount of economic and environmental waste if we compute the volume of carbon emissions required to sustain this entire scheme. Each miner, in their short-term interest, has an incentive to waste more power to have more chance of earning the reward. However in aggregate the amount of power is used for almost no purpose, and the amount of waste is staggering.

Since the statistics involved in the proof of work system are readily quantifiable, it is possible to estimate the energy required to sustain the bitcoin network. Alex de Vries and others estimate that the bitcoin network consumes 87.1 TWh (terawatt hours) of electrical power annually as of September 30, 2019. (Vries 2020)

This amount of wasted energy on bitcoin is comparable to the energy consumption of the entire nation of Argentina, a country of 50 million people. In his paper, de Vries also estimates that the network has doubled its electricity consumption between 2018 and 2019. This figure represents 43% of the current global data center electricity across the entire IT industry, as estimated by the International Energy Agency. (“Key World Energy Statistics 2019” 2019)

Bitcoin consumes nation-state levels of energy to process a minuscule amount of transactions considerably slower than any other payment method.

For a comparison with the traditional IT and banking sectors, the Visa network processed 111.2 billion transactions in 2016. An internal audit of the company reported its data centers 674,922 gigajoules, which during the course of the year amounts to 21.4 megawatts (106 watts). All of Google’s data centers globally used an annual 5.7 terawatts (1012 watts). This includes operations which provide anyone on the internet with Google search and video streaming of YouTube.

The bitcoin network consumes more power than all the Amazon, Microsoft, Facebook, Netflix, Google, and Microsoft’s data centers combined.

Environmental Horrors

The per transaction costs of bitcoin is an even more alarming statistic. The transaction statistics from aggregated chain data indicate the bitcoin network is performing 326,140 transactions per day or 119,041,100 transactions per year. The per transaction energy cost is 2077.54 kWh, or the equivalent to the power consumption of an average US household over 75.67 days. Comparatively, the Visa network can perform 100,000 transactions for 151 kWh, and a single transaction takes 0.002 kWh.1

The yearly carbon output of this energy consumption is quantifiable as a percentage of power derived from fossil fuel emissions, and the bitcoin network is estimated to emit 51.9 megatons of carbon dioxide annually. A single bitcoin transaction alone produces 270 kg of CO2. (Goodkind, Jones, and Berrens 2020; Vries 2020)

On top of the carbon footprint and energy waste, every single data center running computer hardware to perform useless proof of work computations emits a steady stream of e-waste in discarded graphics cards, ASICs, and servers. Each of these chips contains an abundance of heavy metals and carcinogens such as lead, cadmium, mercury, and chromium that are dumped straight into the landfill after being discarded. The annual e-waste of the bitcoin network amounts to all the cell phones, laptops, tablets, and computers of every person in the Netherlands put together. A single transaction on the bitcoin network amounts to the amortized destruction of 2.29 iPhones.

The bitcoin network requires constant hardware replacement and produces a continuous stream of waste from broken and exhausted components. (Vries and Stoll 2021) A substantial new change to the software protocol of a cryptocurrency network may invalidate the previous purchases and require a complete overhaul and repurchasing of all global mining hardware, specially for dedicated ASIC miners. The network produces 11,000,000 kg of electronic waste annually or 96 grams per transaction. This annual e-waste(Jana et al. 2021) is equivalent to several small countries and 482,456 people living at the German standard of 22.8 kg of e-waste per person per year. Moreover, approximately 98% of bitcoin mining equipment will become obsolete before returning any value (Peplow 2019).

Bitcoin is a single network among hundreds that use similarly wasteful proof-of-work models. It is challenging computation to estimate the global energy cost and CO2 emissions across the entire cryptocurrency sector, but the total sum of hundreds of proof-of-work currencies could conservatively be 50% on top of bitcoin energy requirements. Gallersdörfer et al. estimate that “bitcoin accounts for 2/3 of the total energy consumption, and understudied cryptocurrencies represent the remaining 1/3. Therefore, understudied currencies add nearly 50% on top of bitcoin’s energy hunger.” The entire cost in terms of health and climate damages caused by the continued operation of these services is an alarming number and is deserving of further study and estimation. (Gallersdörfer et al. 2020; Vries 2019)

Whether bitcoin has a legitimate claim on any of society’s resources is a question that does not have a scientific answer, it is fundamentally an ethical question. There are many activities where humans burn massive amounts of fossil fuels for entertainment activities or activites that do not serve any productive purpose. For example, Americans burn 6.6 TWh annually for holiday lighting. The software industry must ask whether we should sustain a perpetually wasteful activity in perpetuity.

The answers to this fundamental question from outside of the tech industry have raised some alarming extrapolations from current trends. In an environmental study, Mora et al. estimate that Bitcoin emissions alone could push global warming above 2°C (Mora et al. 2018), and Goodkind et al. suggest that (Goodkind, Jones, and Berrens 2020):

Each $1 of bitcoin value created was responsible for $0.49 in health and climate damages in the US and $0.37 in China

Climate change is a runaway phenomenon that may pose an existential threat to human civilization(Howson 2020). Today massive CO2 emissions are a debit on the quality of life for future generations. The problem of cryptocurrency mining needs to be addressed within a framework that considers the quality of life for future generations and in terms of a cost-benefit analysis of running a network that is consuming nation-state levels of power and whose purpose is primarily speculative gambling.

Benetton, Matteo, Giovanni Compiani, and Adair Morse. 2021. “When Cryptomining Comes to Town: High Electricity-Use Spillovers to the Local Economy.” SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3779720.

Dindar, B., and Ö. Gül. 2021. “The Detection of Illicit Cryptocurrency Mining Farms with Innovative Approaches for the Prevention of Electricity Theft.” Energy & Environment, no. April: 0958305X211045066. https://doi.org/10.1177/0958305x211045066.

Gallersdörfer, Ulrich, Lena Klaaßen, Christian Stoll, Ulrich Gallersdo, Lena Klaaßen, Christian Stoll, and Ulrich Gallersdo. 2020. “Energy Consumption of Cryptocurrencies Beyond Bitcoin.” Joule 4 (2018): 2018–21. https://doi.org/10.1016/j.joule.2020.07.013.

Gerard, David. 2017. Attack of the 50 Foot Blockchain: Bitcoin, Blockchain, Ethereum & Smart Contracts. David Gerard.

Goodkind, Andrew L., Benjamin A. Jones, and Robert P. Berrens. 2020. “Cryptodamages: Monetary Value Estimates of the Air Pollution and Human Health Impacts of Cryptocurrency Mining.” Energy Research and Social Science 59 (March 2019): 101281. https://doi.org/10.1016/j.erss.2019.101281.

Howson, Peter. 2020. “Climate Crises and Crypto-Colonialism: Conjuring Value on the Blockchain Frontiers of the Global South.” Frontiers in Blockchain 3 (May). https://doi.org/10.3389/fbloc.2020.00022.

Jana, Rabin K., Indranil Ghosh, Debojyoti Das, and Anupam Dutta. 2021. “Determinants of Electronic Waste Generation in Bitcoin Network: Evidence from the Machine Learning Approach.” Technological Forecasting and Social Change 173. https://doi.org/10.1016/j.techfore.2021.121101.

“Key World Energy Statistics 2019.” 2019. In. International Energy Agency.

Küfeoğlu, Sinan, and Mahmut Özkuran. 2019. “Bitcoin Mining: A Global Review of Energy and Power Demand.” Energy Research and Social Science 58: 101273. https://doi.org/10.1016/j.erss.2019.101273.

Mora, Camilo, Randi L Rollins, Katie Taladay, Michael B Kantar, Mason K Chock, Mio Shimada, and Erik C Franklin. 2018. “Bitcoin Emissions Alone Could Push Global Warming Above 2 c.” Nature Climate Change 8 (11): 931–33.

Peplow, Mark. 2019. “Bitcoin Poses Major Electronic-Waste Problem.” American Chemical Society.

Stoll, Christian, Lena Klaaßen, and Ulrich Gallersdörfer. 2019. “The Carbon Footprint of Bitcoin.” Joule 3 (7): 1647–61. https://doi.org/10.1016/j.joule.2019.05.012.

Vigna, Paul. 2019. “Most Bitcoin Trading Faked by Unregulated Exchanges, Study Finds.” Wall Street Journal.

Vries, Alex de. 2019. “Renewable Energy Will Not Solve Bitcoin’s Sustainability Problem.” Joule 3 (4): 893–98.

———. 2020. “Bitcoin’s Energy Consumption Is Underestimated: A Market Dynamics Approach.” Energy Research & Social Science 70: 101721.

Vries, Alex de, and Christian Stoll. 2021. “Bitcoin’s Growing E-Waste Problem.” Resources, Conservation and Recycling 175 (September): 105901. https://doi.org/10.1016/j.resconrec.2021.105901.

Weaver, Nicholas. 2018. Blockchains and Cryptocurrencies: Burn It with Fire. Berkeley School of Information. https://www.youtube.com/watch?v=xCHab0dNnj4.

  1. A time of writing the bitcoin network was performing 359,405 transactions per day.↩︎