Estimating the total energy that digital consumes is not easy because digital is everywhere and in practically everything. For example, a new car or truck “can have up to 100 networked microprocessors running 150 million lines of code with thousands of supporting active components,” Jeff Darrow writes for Electronic Design. As the digital Web gets further embedded with Internet of Things (IoT), digital will become woven into everything from our clothes to our food.
The energy demand of digital is “growing by 4% per year, in stark contrast to the trend of global GDP’s energy intensity evolution, which is declining by 1.8% per year,” a Shift Project report stated in 2019. “The explosion of video uses (Skype, streaming, etc.) and the increased consumption of frequently renewed digital equipment are the main drivers of this inflation.” In a world of climate change, where much of the world is trying to reduce energy use, digital’s hunger for energy is growing.
There are three major areas where digital consumes energy:
It used to be that the things we made (cars, fridges) used more energy during their lifetime of use than during the period of their manufacture. For example, a car can last from eight to 15 years, according to US Consumer Reports, whereas the lifespan of a computer is typically three to five years, and the lifespan of a smartphone is typically two to three years. In general, the more use you get out of something, the better it is for the environment. “If you make a car last to 200,000 miles rather than 100,000, then the emissions for each mile the car does in its lifetime may drop by as much as 50%, as a result of getting more distance out of the initial manufacturing emissions,” Mike Berners-Lee and Duncan Clark wrote for The Guardian in 2010.
To manufacture something made of steel or plastic requires “between 1 and 10 megajoule[s] of electricity per kilogram of material,” explains Kris De Decker, editor of Low-Tech magazine. However, to manufacture a semiconductor requires between 1,000 and 100,000 megajoules per kilogram of material. In 2011, researchers from Arizona State University found that of the total energy consumed by a computer during its lifetime, 70% of that is consumed during its manufacture. In 2019, Ethical Consumer reported that the manufacturing stage for a typical smartphone can create up to 80% of its total pollution.
It is thus nearly always better to keep using a digital device than to change it for a newer version. A huge amount of pollution was released making the phone in your hand and the computer on your table. Try to get the maximum possible use out of it. It’s greener to use than to buy new. And when you buy, buy for quality, buy things that will last. Cheap costs the Earth. There’s nothing worse for the planet than throwaway digital devices.
Energy during use
Over time, digital products have generally become a lot more energy efficient. A typical data center in 2019 was a lot more efficient than its equivalent in 2010. A study published in Science in 2020 found that while the computing output of data centers globally grew sixfold from 2010 to 2018, their energy consumption only rose by 6%. However, because of the explosion of data and digital devices, the OECD reported that in OECD countries pollution caused by digital had grown by about 450 million tons since 2013, while “overall OECD CO2 pollution actually decreased by 250 million tons.”
According to a report from New York University’s AI Institute, the technology sector (excluding TVs and consumer electronics) would cause 3.6% of global greenhouse gas pollution in 2020, more than double what it caused in 2007. In a worst-case scenario, this figure could rise to 14% by 2040. In 2019, the Financial Times reported that digital “contributes to about 4 per cent of global greenhouse gas pollution… 37% higher than what it was in 2010.”
Digital pollution was the equivalent to that of the fuel used by the aviation industry in 2018, according to an article by Nicola Jones in Nature. She was reporting on research by Anders Andrae which estimated that by 2030, in a best-case scenario, digital would be consuming 8% of global electricity, and in a worst case, it could be 21%. What could accelerate the worst case? “If the computationally intensive cryptocurrency Bitcoin continues to grow,” Jones wrote, “a sharp rise in energy demand could come sooner rather than later.” (More on Bitcoin later.)
An even bigger problem will be if the rest of the world tries to catch up with US pollution habits. In 2018, US energy consumption per person was about four times what an average citizen of the world would consume, according to the US Energy Information Administration. Driven by IoT, the number of connected devices is exploding, and connectivity equals electricity. According to Cisco:
- In the United States, there will be an average of 13.6 networked devices per person by 2022, up from 8.1 in 2017.
- In China, there will be 4.4 networked devices per person by 2022, up from 2.8 in 2017.
- In India, there will be 1.5 networked devices per person by 2022, up from 1.2 in 2017.
- Globally, there will be 3.6 networked devices per person by 2022, up from 2.4 in 2017.
We have an ever-growing hunger for data, particularly “rich” data, such as videos, and we want higher and higher resolutions. Netflix estimates that when you’re streaming in standard definition, you’re using about one GB per hour. However, if you’re using high definition, that jumps to up to three GB per hour. If you double the definition you usually treble the weight, and much of this definition is wholly unnecessary because you can’t see the difference. It’s useless waste.
According to Cisco:
- In the United States, Internet traffic will grow 3.1-fold from 2017 to 2022, an annual growth rate of 25%. Average traffic per person per month will grow from 120.2 GB in 2017 to 304.6 GB in 2022.
- Globally, average traffic per person per month will grow from 16.2 GB in 2017 to 49.8 GB in 2022.
- In China, average traffic per person per month will grow from 15.6 GB in 2017 to 66.1 GB in 2022.
- In India, average traffic per person per month will grow from 2.4 GB in 2017 to 13.9 GB in 2022.
Let’s say someone is using 140 GB of data a month. What sort of pollution is that causing? That’s not an easy question to answer. If you search about data energy usage, then you’re likely to find studies that state that transmitting and storing one gigabyte of data consumes 7 kWh, or 3.9 kWh or 0.06 kWh: a huge variance. You will find that the 7 kWh figure is quoted quite a bit, but if you dig deeper you will find that the study this figure is based on was published in 2009, and that networks have become much more energy efficient since then.
Here lies a data integrity problem. Not much gets archived or deleted on the Web. Wrong and out-of-date information is everywhere. Older content often comes up earlier in search results than fresher, more recently published content. As Norbert Wiener stated in 1950, “The idea that information can be stored in a changing world without an overwhelming depreciation in its value is false.” As the Web gets older, more and more of it becomes “fake” because the information it seeks to communicate is no longer accurate. You have to do your homework if you want to get close to the facts.
Turns out that the 0.06 kWh is a figure published in 2015 that was based on a review of multiple other studies and seems to be the more accurate one. According to this analysis, the amount of energy required to transmit and store data has been decreasing by half every two years since 2000. Extrapolating that trend outwards to 2019, we get a figure of 0.015 kWh per GB, and that is the figure I will use for this book.
Thus, a monthly data usage of 140 GB per US citizen comes to 2.1 kWh or about 0.59 kg of pollution. The population of the US is 327 million. So, every year US citizens’ data production and consumption habits emit about 2.4 billion kg of CO2. You’d need to plant 230 million trees to deal with that pollution. By 2022, Cisco estimates it will be 304 GB per person per month. We’d need to plant over 500 million trees to deal with that.
Digital continues to develop technologies that have massive energy needs. Blockchain is a technology that uses the network to allow secure, transparent and anonymous transactions to occur. It is an alternative to the traditional banking system, as value, transactions and communications are authenticated, not by one central authority, but by all the members of the network. It is a truly interesting idea with tremendous potential, but its energy needs are huge.
Bitcoin is an example of a cryptocurrency using blockchain. Bitcoin is a “decentralized digital currency without a central bank or single administrator that can be sent from user to user on the peer-to-peer bitcoin network without the need for intermediaries,” according to Wikipedia. Creating bitcoins, which is called “mining,” creates huge amounts of pollution. Nature magazine has estimated that the mining of Bitcoin and similar cryptocurrencies consumes more energy than most other minerals of an equivalent market value. Basically, mining a dollar’s worth of Bitcoin is more expensive than mining a dollar’s worth of gold. As well, a cryptocurrency transaction costs as much in energy as hundreds of thousands of Visa-type transactions. Blockchain and cryptocurrencies need to become much more energy conscious and environmentally friendly.
Many tech companies are aware of the environmental issues and are seeking to make digital devices and data centers more energy efficient. A lot of the more advanced data centers use substantial quantities of renewable energy, and their energy efficiency improved dramatically between 2010 and 2020. This is good but it’s not addressing the core issue, which is overconsumption. We consume too much digital, way, way too much, and it’s exploding, exploding, exploding. The business case of so many in digital is overconsumption. The more we consume of digital, the more money they make. Let’s say we eat too much ice cream. Well, the tech industry is telling us: “Hey, look, we’ve found more energy-efficient ways to make ice cream!”
This idea that because it’s renewable we can use as much as we want is a dangerous one, and unfortunately feeds the myth of digital, which is all about unlimited power, space and energy. In the psychology of the Internet, cyberspace has no boundaries, has no weight, and the information superhighway goes on and on without end, without limit. There is an end, though, and there is a weight.
Every single thing that happens in digital has an energy cost that is charged to the Earth. Every byte of data takes the Earth’s energy to create and takes the Earth’s energy to transit and store. The windmills and solar panels that make our renewable energy all need to be manufactured, transported, installed, maintained and retired. That all creates pollution. The renewable energy they create needs to be transmitted across creaking electrical infrastructure and will feed billions of hungry digital machines that were themselves manufactured and will ultimately become waste; some of the worst and most toxic waste ever to have been created.
“It’s the most powerful greenhouse gas known to humanity, and emissions have risen rapidly in recent years,” the BBC reported in 2019. “Sulphur hexafluoride, or SF6, is widely used in the electrical industry to prevent short circuits and accidents. But leaks of the little-known gas in the UK and the rest of the EU in 2017 were the equivalent of putting an extra 1.3 million cars on the road.” Wind turbines use SF6.
“Electronic waste or e-waste describes discarded electrical or electronic devices,” according to Wikipedia. As of 2019, 50 million tons of e-waste were being produced each year, with smartphones accounting for as much waste as desktops, laptops and displays combined. All the commercial aircraft ever built come to about 50 million tons of material. If one year’s e-waste was laid out together, it would cover an area the size of Manhattan. That’s one year’s waste. It’s estimated that by 2050 we will be producing 120 million tons of e-waste a year.
E-waste is more destructive to the environment than many other forms of waste because:
- Short lifespans (typically less than five years) mean that tech churns through raw materials at a tremendous pace.
- High manufacturing energy costs mean that the tech products cause most of their pollution during the manufacturing process. Combine that with their short lifespans and you get a perfect storm of pollution.
- The culture of tech is the culture of change, newness, innovation. The business model depends on super-short lifespans for hardware and software. For the tech industry, growth comes through waste. In tech, nothing is built to last. Everything is built to waste because the driving idea is that there is a new innovation around the corner that will change everything.
- Many digital products are deliberately designed not to be repaired. Smartphones are often designed with special glues, custom screws and batteries that can’t be easily replaced.
- Much e-waste includes toxic materials that are harmful to the environment.
- E-waste tends to involve a highly complex mix of materials, thus making it difficult and costly to recycle.
In 2020, the European Parliament voted overwhelmingly in favour of introducing a single universal charging method for mobile phones. The Parliament estimated that obsolete cables generate more than 51,000 tonnes of e-waste per year. What was Apple’s response? “Regulations that would drive conformity across the type of connector built into all smartphones freeze innovation rather than encourage it,” an official statement from the company read. “Such proposals are bad for the environment and unnecessarily disruptive for customers.” In 2020, we surely live in a world of Orwellian doublespeak. In some tech circles, innovation has become a god, a god that sacrifices the Earth in pursuit of newness and change.
In 2016, according to the US Environmental Protection Agency, only about 12% of e-waste was recycled in the US. The UN estimated that in 2016 about 20% of e-waste was recycled globally. These figures are low but then recycling has always been a bit of a PR mirage, a way industry can market itself as being environmentally conscious. Recycling does not address the core problem: overproduction and overconsumption. Recycling can be like going to mass on a Sunday, thinking it gives a license to be an asshole for the rest of the week.
Digital facilitates a culture of waste because it is a key driver in the world of convenience. Digital is always saying to us: “Don’t spend your physical energy. Spend my electrical energy.” Any digital appliance that uses a standby mode constantly wastes energy. Instant-on televisions, computers in sleep mode, cable set-top boxes that are never turned off, and game consoles, all silently suck energy.
In the US, it’s estimated that this “vampire power” can account for up to 20% of a typical electricity bill. According to a 2015 study by the NRDC, energy use by inactive devices costs “$19 billion a year—about $165 per US household on average—and approximately 50 large (500-megawatt) power plants’ worth of electricity.” The study found that the average US house has 65—yes, 65—devices permanently plugged into electrical outlets. How on Earth are you going to manage 65 devices? That’s the thing. You’re not. It’s overload. Too many devices. Too much stuff. So much that you don’t even think about it anymore.
Spending the wrong energy
“Even now, I still go to the hospital to keep receiving treatments for my arm, and there’s still pain. The pain gets especially bad when it rains, but I grit my teeth and work through that.” If you think that this sounds like someone with repetitive stress injury from spending too long using a computer, you’d be right. This is the voice of Lee “Flash” Young Ho, regarded as one of the legends of electronic gaming, also known as e-sports.
In 2019, there were more than 400 million fans of e-sports, and that is expected to rise to 1.1 billion by 2021. The PC and game consoles that Lee “Flash” Young Ho plays are super-hungry consumers of energy. “Globally, PC gamers use about 75 billion kWh of electricity a year, equivalent to the output of 25 electric power plants,” Bryan Schatz wrote for Grist. You’d need to plant 7.2 billion trees to deal with that sort of pollution.
A study by the British Journal of Medicine of e-sports athletes found that they spent up to 10 hours a day practicing before competitions. 40% did not do any physical exercise, with 15% reporting that they sat for more than three hours without standing. Sitting “raises your risk of heart disease, diabetes, stroke, high blood pressure, and high cholesterol,” according to WebMD.
E-sports embodies the modern digital world: the human body is spending less and less energy, the Earth is spending more and more. Digital is the babysitter. In interacting with digital, the human body is not working in any way like it was designed to work over millions of years. The body is largely passive, a shell for the mind, except for repetitive movements of the arms and eyes. The body’s surplus energy is increasingly being stored as fat.
We are spending electrical energy instead of physical energy and are making the Earth sick and ourselves sick. A US study published in 2019 found that, by the age of eight, children were becoming much more sedentary, and that there was an even bigger drop off in exercising by age eleven. Globally, over 80% of school-going adolescents are not getting enough regular exercise, a 2019 World Health Organization study found.
Digital feeds laziness. Electric scooters that are appearing in many cities are a perfect example of devices that encourage laziness and convenience, litter sidewalks and streets, causing accidents, while damaging the environment. Researchers at North Carolina State University found that nearly half of e-scooter riders would have biked or walked had the scooters not been available. Only 34% had chosen the e-scooter instead of using their car. When everything was taken into account, the e-scooters were producing more greenhouse-gas pollution per passenger mile than a standard diesel bus with high ridership.
Digital feeds congestion and pollution. According to a 2020 report by the Union of Concerned Scientists, ride-hailing has resulted in declining mass transit ridership and increased congestion. “Ride-hailing trips today result in an estimated 69 percent more climate pollution on average than the trips they displace,” the report stated. In the false economy of digital, the costs to the environment are rarely calculated. Instead, digital has become our go-to source for energy. Why go to the bother of repairing something when it’s cheaper and more convenient to replace it with a more powerful version? Why get up and turn the light off when digital can do it for you? Digital feeds our addictions. Not surprisingly, you are likely to find a digital addict with one hand on the mouse and one hand clutching a sugar-saturated drink. Sugar, without doubt the greatest single addiction of the human race, tells us a harsh story of what happens when our addictions control our lives.
Use less electrical energy. Use more of your own physical energy.
World Wide Waste
- Free, weekly email
- Published since 1996
- Read some examples
- Subscribe with your email below
- The United Nations Global E-waste Monitor 2017: Quantities, Flows, and Resources, United Nations, 2017
- The world’s e-waste is a huge problem. It’s also a golden opportunity, Guy Ryder, Houlin Zhao Houlin, World Economic Forum, 2019
- American trash: How an e-waste sting uncovered a shocking betrayal, Colin Lecher, The Verge, 2019
- What’s the carbon footprint of… a new car? Mike Berners-Lee, Duncan Clark, The Guardian, 2010
- Climate change: Electrical industry’s ‘dirty secret’ boosts warming, Matt McGrath, BBC, 2019
- AI Now 2019 Report: Technology CO2 emissions estimate, AI Now Institute, 2019
- Quantum computing could change the way the world uses energy, Vern Brownell, QUARTZ, 2019
- The Energy and Carbon Footprint of the Global ICT and E&M Sectors 2010–2015, Jens Malmodin, Dag Lundén, ResearchGate, 2018
- Total Consumer Power Consumption Forecast, Anders S.G. Andrae, ResearchGate, 2017
- Cloud Computing Is Not the Energy Hog That Had Been Feared, Steve Lohr, The New York Times
- Recalibrating global data center energy-use estimates, E. Masanet, A. Shehabi, N. Lei1, S. Smith, J. Koomey, Science, 2020
- Assessing ICT global emissions footprint: Trends to 2040 & recommendations, Lotfi Belkhir, Ahmed Elmeligi, Science Direct, 2018
- Just because it’s digital doesn’t mean it’s green, FT Alphaville, 2019
- Lean ICT: Towards digital sobriety, The Shift Project, 2019
- What is the United States’ share of world energy consumption? US Energy Information Administration, 2019
- Cisco digital forecasts 2017 – 2022, Cisco
- Why We Need a Speed Limit for the Internet, Kris De Decker, Low Tech Magazine, 2015
- The monster footprint of digital technology, Kris De Decker, Low Tech Magazine, 2009
- Ride-Hailing’s Climate Risks, Union of Concerned Scientists, 2020
- How energy vampires could be sucking your electricity dry and costing you money, Allconnect, 2019
- Video games consume more electricity than 25 power plants can produce, Bryan Schatz, Bryan Schatz, Grist, 2018
- In this guide we investigate, score and rank the ethical and environmental record of 14 desktop computer brands, Tom Bryson, Ethical Consumer, 2019
- The Micromobility Mirage, Frederic Filloux, Medium, 2019
- Computer factories eat way more energy than running the devices they build, Michael Cooney, Network World, 2011
- Managing the health of the eSport athlete: an integrated health management model, J. DiFrancisco-Donoghue, J. Balentine, G. Schmidt, H. Zwibel, BMJ, 2019
- Why Sitting Too Much Is Bad for Your Health, WebMD
- Children’s Sport Participation and Physical Activity Study, Sports Ireland, 2019
- One in three children ‘not active enough’, finds sport survey, BBC, 2018
- Children around the world not getting enough exercise, report says, Tauren Dyson, UPI, 2018
- Young Children are Getting Less Exercise than We Thought, Cleveland Clinic, 2019