A problem built into our relationship with energy itself. Photo by Ferdinando Scianna/Magnum


Deep warming

Even if we ‘solve’ global warming, we face an older, slower problem. Waste heat could radically alter Earth’s future

by Mark Buchanan + BIO

A problem built into our relationship with energy itself. Photo by Ferdinando Scianna/Magnum

The world will be transformed. By 2050, we will be driving electric cars and flying in aircraft running on synthetic fuels produced through solar and wind energy. New energy-efficient technologies, most likely harnessing artificial intelligence, will dominate nearly all human activities from farming to heavy industry. The fossil fuel industry will be in the final stages of a terminal decline. Nuclear fusion and other new energy sources may have become widespread. Perhaps our planet will even be orbited by massive solar arrays capturing cosmic energy from sunlight and generating seemingly endless energy for all our needs.

That is one possible future for humanity. It’s an optimistic view of how radical changes to energy production might help us slow or avoid the worst outcomes of global warming. In a report from 1965, scientists from the US government warned that our ongoing use of fossil fuels would cause global warming with potentially disastrous consequences for Earth’s climate. The report, one of the first government-produced documents to predict a major crisis caused by humanity’s large-scale activities, noted that the likely consequences would include higher global temperatures, the melting of the ice caps and rising sea levels. ‘Through his worldwide industrial civilisation,’ the report concluded, ‘Man is unwittingly conducting a vast geophysical experiment’ – an experiment with a highly uncertain outcome, but clear and important risks for life on Earth.

Since then, we’ve dithered and doubted and argued about what to do, but still have not managed to take serious action to reduce greenhouse gas emissions, which continue to rise. Governments around the planet have promised to phase out emissions in the coming decades and transition to ‘green energy’. But global temperatures may be rising faster than we expected: some climate scientists worry that rapid rises could create new problems and positive feedback loops that may accelerate climate destabilisation and make parts of the world uninhabitable long before a hoped-for transition is possible.

Despite this bleak vision of the future, there are reasons for optimists to hope due to progress on cleaner sources of renewable energy, especially solar power. Around 2010, solar energy generation accounted for less than 1 per cent of the electricity generated by humanity. But experts believe that, by 2027, due to falling costs, better technology and exponential growth in new installations, solar power will become the largest global energy source for producing electricity. If progress on renewables continues, we might find a way to resolve the warming problem linked to greenhouse gas emissions. By 2050, large-scale societal and ecological changes might have helped us avoid the worst consequences of our extensive use of fossil fuels.

It’s a momentous challenge. And it won’t be easy. But this story of transformation only hints at the true depth of the future problems humanity will confront in managing our energy use and its influence over our climate.

As scientists are gradually learning, even if we solve the immediate warming problem linked to the greenhouse effect, there’s another warming problem steadily growing beneath it. Let’s call it the ‘deep warming’ problem. This deeper problem also raises Earth’s surface temperature but, unlike global warming, it has nothing to do with greenhouse gases and our use of fossil fuels. It stems directly from our use of energy in all forms and our tendency to use more energy over time – a problem created by the inevitable waste heat that is generated whenever we use energy to do something. Yes, the world may well be transformed by 2050. Carbon dioxide levels may stabilise or fall thanks to advanced AI-assisted technologies that run on energy harvested from the sun and wind. And the fossil fuel industry may be taking its last breaths. But we will still face a deeper problem. That’s because ‘deep warming’ is not created by the release of greenhouse gases into the atmosphere. It’s a problem built into our relationship with energy itself.

Finding new ways to harness more energy has been a constant theme of human development. The evolution of humanity – from early modes of hunter-gathering to farming and industry – has involved large systematic increases in our per-capita energy use. The British historian and archaeologist Ian Morris estimates, in his book Foragers, Farmers, and Fossil Fuels: How Human Values Evolve (2015), that early human hunter-gatherers, living more than 10,000 years ago, ‘captured’ around 5,000 kcal per person per day by consuming food, burning fuel, making clothing, building shelter, or through other activities. Later, after we turned to farming and enlisted the energies of domesticated animals, we were able to harness as much as 30,000 kcal per day. In the late 17th century, the exploitation of coal and steam power marked another leap: by 1970, the use of fossil fuels allowed humans to consume some 230,000 kcal per person per day. (When we think about humanity writ large as ‘humans’, it’s important to acknowledge that the average person in the wealthiest nations consumes up to 100 times more energy than the average person in the poorest nations.) As the global population has risen and people have invented new energy-dependent technologies, our global energy use has continued to climb.

In many respects, this is great. We can now do more with less effort and achieve things that were unimaginable to the 17th-century inventors of steam engines, let alone to our hominin ancestors. We’ve made powerful mining machines, superfast trains, lasers for use in telecommunications and brain-imaging equipment. But these creations, while helping us, are also subtly heating the planet.

All the energy we humans use – to heat our homes, run our factories, propel our automobiles and aircraft, or to run our electronics – eventually ends up as heat in the environment. In the shorter term, most of the energy we use flows directly into the environment. It gets there through hot exhaust gases, friction between tires and roads, the noises generated by powerful engines, which spread out, dissipate, and eventually end up as heat. However, a small portion of the energy we use gets stored in physical changes, such as in new steel, plastic or concrete. It’s stored in our cities and technologies. In the longer term, as these materials break down, the energy stored inside also finds its way into the environment as heat. This is a direct consequence of the well-tested principles of thermodynamics.

Waste heat will pose a problem that is every bit as serious as global warming from greenhouse gases

In the early decades of the 21st century, this heat created by simply using energy, known as ‘waste heat’, is not so serious. It’s equivalent to roughly 2 per cent of the planetary heating imbalance caused by greenhouse gases – for now. But, with the passing of time, the problem is likely to get much more serious. That’s because humans have a historical tendency to consistently discover and produce things, creating entirely new technologies and industries in the process: domesticated animals for farming; railways and automobiles; global air travel and shipping; personal computers, the internet and mobile phones. The result of such activities is that we end up using more and more energy, despite improved energy efficiency in nearly every area of technology.

During the past two centuries at least (and likely for much longer), our yearly energy use has doubled roughly every 30 to 50 years. Our energy use seems to be growing exponentially, a trend that shows every sign of continuing. We keep finding new things to do and almost everything we invent requires more and more energy: consider the enormous energy demands of cryptocurrency mining or the accelerating energy requirements of AI.

If this historical trend continues, scientists estimate waste heat will pose a problem in roughly 150-200 years that is every bit as serious as the current problem of global warming from greenhouse gases. However, deep heating will be more pernicious as we won’t be able to avoid it by merely shifting from one kind energy to another. A profound problem will loom before us: can we set strict limits on all the energy we use? Can we reign in the seemingly inexorable expansion of our activities to avoid destroying our own environment?

Deep warming is a problem hiding beneath global warming, but one that will become prominent if and when we manage to solve the more pressing issue of greenhouse gases. It remains just out of sight, which might explain why scientists only became concerned about the ‘waste heat’ problem around 15 years ago.

One of the first people to describe the problem is the Harvard astrophysicist Eric Chaisson, who discussed the issue of waste heat in a paper titled ‘Long-Term Global Heating from Energy Usage’ (2008). He concluded that our technological society may be facing a fundamental limit to growth due to ‘unavoidable global heating … dictated solely by the second law of thermodynamics, a biogeophysical effect often ignored when estimating future planetary warming scenarios’. When I emailed Chaisson to learn more, he told me the history of his thinking on the problem:

It was on a night flight, Paris-Boston [circa] 2006, after a UNESCO meeting on the environment when it dawned on me that the IPCC were overlooking something. While others on the plane slept, I crunched some numbers literally on the back of an envelope … and then hoped I was wrong, that is, hoped that I was incorrect in thinking that the very act of using energy heats the air, however slightly now.

The transformation of energy into heat is among the most ubiquitous processes of physics

Chaisson drafted the idea up as a paper and sent it to an academic journal. Two anonymous reviewers were eager for it to be published. ‘A third tried his damnedest to kill it,’ Chaisson said, the reviewer claiming the findings were ‘irrelevant and distracting’. After it was finally published, the paper got some traction when it was covered by a journalist and ran as a feature story on the front page of The Boston Globe. The numbers Chaisson crunched, predictions of our mounting waste heat, were even run on a supercomputer at the US National Center for Atmospheric Research, by Mark Flanner, a professor of earth system science. Flanner, Chaisson suspected at the time, was likely ‘out to prove it wrong’. But, ‘after his machine crunched for many hours’, he saw the same results that Chaisson had written on the back of an envelope that night in the plane.

Around the same time, also in 2008, two engineers, Nick Cowern and Chihak Ahn, wrote a research paper entirely independent of Chaisson’s work, but with similar conclusions. This was how I first came across the problem. Cowern and Ahn’s study estimated the total amount of waste heat we’re currently releasing to the environment, and found that it is, right now, quite small. But, like Chaisson, they acknowledged that the problem would eventually become serious unless steps were taken to avoid it.

That’s some of the early history of thinking in this area. But these two papers, and a few other analyses since, point to the same unsettling conclusion: what I am calling ‘deep warming’ will be a big problem for humanity at some point in the not-too-distant future. The precise date is far from certain. It might be 150 years, or 400, or 800, but it’s in the relatively near future, not the distant future of, say, thousands or millions of years. This is our future.

The transformation of energy into heat is among the most ubiquitous processes of physics. As cars drive down roads, trains roar along railways, planes cross the skies and industrial plants turn raw materials into refined products, energy gets turned into heat, which is the scientific word for energy stored in the disorganised motions of molecules at the microscopic level. As a plane flies from Paris to Boston, it burns fuel and thrusts hot gases into the air, generates lots of sound and stirs up contrails. These swirls of air give rise to swirls on smaller scales which in turn make smaller ones until the energy ultimately ends up lost in heat – the air is a little warmer than before, the molecules making it up moving about a little more vigorously. A similar process takes place when energy is used by the tiny electrical currents inside the microchips of computers, silently carrying out computations. Energy used always ends up as heat. Decades ago, research by the IBM physicist Rolf Landauer showed that a computation involving even a single computing bit will release a certain minimum amount of heat to the environment.

How this happens is described by the laws of thermodynamics, which were described in the mid-19th century by scientists including Sadi Carnot in France and Rudolf Clausius in Germany. Two key ‘laws’ summarise its main principles.

The first law of thermodynamics simply states that the total quantity of energy never changes but is conserved. Energy, in other words, never disappears, but only changes form. The energy initially stored in an aircraft’s fuel, for example, can be changed into the energetic motion of the plane. Turn on an electric heater, and energy initially held in electric currents gets turned into heat, which spreads into the air, walls and fabric of your house. The total energy remains the same, but it markedly changes form.

We’re generating waste heat all the time with everything we do

The second law of thermodynamics, equally important, is more subtle and states that, in natural processes, the transformation of energy always moves from more organised and useful forms to less organised and less useful forms. For an aircraft, the energy initially concentrated in jet fuel ends up dissipated in stirred-up winds, sounds and heat spread over vast areas of the atmosphere in a largely invisible way. It’s the same with the electric heater: the organised useful energy in the electric currents gets dissipated and spread into the low-grade warmth of the walls, then leaks into the outside air. Although the amount of energy remains the same, it gradually turns into less organised, less usable forms. The end point of the energy process produces waste heat. And we’re generating it all the time with everything we do.

Data on world energy consumption shows that, collectively, all humans on Earth are currently using about 170,000 terawatt-hours (TWh), which is a lot of energy in absolute terms – a terawatt-hour is the total energy consumed in one hour by any process using energy at a rate of 1 trillion watts. This huge number isn’t surprising, as it represents all the energy being used every day by the billions of cars and homes around the world, as well as by industry, farming, construction, air traffic and so on. But, in the early 21st century, the warming from this energy is still much less than the planetary heating due to greenhouse gases.

Concentrations of greenhouse gases such as CO2 and methane are quite small, and only make a fractional difference to how much of the Sun’s energy gets trapped in the atmosphere, rather than making it back out to space. Even so, this fractional difference has a huge effect because the stream of energy arriving from the Sun to Earth is so large. Current estimates of this greenhouse energy imbalance come to around 0.87 W per square meter, which translates into a total energy figure about 50 times larger than our waste heat. That’s reassuring. But as Cowern and Ahn wrote in their 2008 paper, things aren’t likely to stay this way over time because our energy usage keeps rising. Unless, that is, we can find some radical way to break the trend of using ever more energy.

One common objection to the idea of the deep warming is to claim that the problem won’t really arise. ‘Don’t worry,’ someone might say, ‘with efficient technology, we’re going to find ways to stop using more energy; though we’ll end up doing more things in the future, we’ll use less energy.’ This may sound plausible at first, because we are indeed getting more efficient at using energy in most areas of technology. Our cars, appliances and laptops are all doing more with less energy. If efficiency keeps improving, perhaps we can learn to run these things with almost no energy at all? Not likely, because there are limits to energy efficiency.

Over the past few decades, the efficiency of heating in homes – including oil and gas furnaces, and boilers used to heat water – has increased from less than 50 per cent to well above 90 per cent of what is theoretically possible. That’s good news, but there’s not much more efficiency to be realised in basic heating. The efficiency of lighting has also vastly improved, with modern LED lighting turning something like 70 per cent of the applied electrical energy into light. We will gain some efficiencies as older lighting gets completely replaced by LEDs, but there’s not a lot of room left for future efficiency improvements. Similar efficiency limits arise in the growing or cooking of food; in the manufacturing of cars, bikes and electronic devices; in transportation, as we’re taken from place to place; in the running of search engines, translation software, GPT-4 or other large-language models.

Even if we made significant improvements in the efficiencies of these technologies, we will only have bought a little time. These changes won’t delay by much the date when deep warming becomes a problem we must reckon with.

Optimising efficiencies is just a temporary reprieve, not a radical change in our human future

As a thought experiment, suppose we could immediately improve the energy efficiency of everything we do by a factor of 10 – a fantastically optimistic proposal. That is, imagine the energy output of humans on Earth has been reduced 10 times, from 170,000 TWh to 17,000 TWh. If our energy use keeps expanding, doubling every 30-50 years or so (as it has for centuries), then a 10-fold increase in waste heat will happen in just over three doubling times, which is about 130 years: 17,000 TWh doubles to 34,000 TWh, which doubles to 68,000 TWh, which doubles to 136,000 TWh, and so on. All those improvements in energy efficiency would quickly evaporate. The date when deep warming hits would recede by 130 years or so, but not much more. Optimising efficiencies is just a temporary reprieve, not a radical change in our human future.

Improvements in energy efficiency can also have an inverse effect on our overall energy use. It’s easy to think that if we make a technology more efficient, we’ll then use less energy through the technology. But economists are deeply aware of a paradoxical effect known as ‘rebound’, whereby improved energy efficiency, by making the use of a technology cheaper, actually leads to more widespread use of that technology – and more energy use too. The classic example, as noted by the British economist William Stanley Jevons in his book The Coal Question (1865), is the invention of the steam engine. This new technology could extract energy from burning coal more efficiently, but it also made possible so many new applications that the use of coal increased. A recent study by economists suggests that, across the economy, such rebound effects might easily swallow at least 50 per cent of any efficiency gains in energy use. Something similar has already happened with LED lights, for which people have found thousands of new uses.

If gains in efficiency won’t buy us lots of time, how about other factors, such as a reduction of the global population? Scientists generally believe that the current human population of more than 8 billion people is well beyond the limits of our finite planet, especially if a large fraction of this population aspires to the resource-intensive lifestyles of wealthy nations. Some estimates suggest that a more sustainable population might be more like 2 billion, which could reduce energy use significantly, potentially by a factor of three or four. However, this isn’t a real solution: again, as with the example of improved energy efficiency, a one-time reduction of our energy consumption by a factor of three will quickly be swallowed up by an inexorable rise in energy use. If Earth’s population were suddenly reduced to 2 billion – about a quarter of the current population – our energy gains would initially be enormous. But those gains would be erased in two doubling times, or roughly 60-100 years, as our energy demands would grow fourfold.

So, why aren’t more people talking about this? The deep warming problem is starting to get more attention. It was recently mentioned on Twitter by the German climate scientist Stefan Rahmstorf, who cautioned that nuclear fusion, despite excitement over recent advances, won’t arrive in time to save us from our waste heat, and might make the problem worse. By providing another cheap source of energy, fusion energy could accelerate both the growth of our energy use and the reckoning of deep warming. A student of Rahmstorf’s, Peter Steiglechner, wrote his master’s thesis on the problem in 2018. Recognition of deep warming and its long-term implications for humanity is spreading. But what can we do about the problem?

Avoiding or delaying deep warming will involve slowing the rise of our waste heat, which means restricting the amount of energy we use and also choosing energy sources that exacerbate the problem as little as possible. Unlike the energy from fossil fuels or nuclear power, which add to our waste energy burden, renewable energy sources intercept energy that is already on its way to Earth, rather than producing additional waste heat. In this sense, the deep warming problem is another reason to pursue renewable energy sources such as solar or wind rather than alternatives such as nuclear fusion, fission or even geothermal power. If we derive energy from any of these sources, we’re unleashing new flows of energy into the Earth system without making a compensating reduction. As a result, all such sources will add to the waste heat problem. However, if renewable sources of energy are deployed correctly, they need not add to our deposition of waste heat in the environment. By using this energy, we produce no more waste heat than would have been created by sunlight in the first place.

Take the example of wind energy. Sunlight first stirs winds into motion by heating parts of the planet unequally, causing vast cells of convection. As wind churns through the atmosphere, blows through trees and over mountains and waves, most of its energy gets turned into heat, ending up in the microscopic motions of molecules. If we harvest some of this wind energy through turbines, it will also be turned into heat in the form of stored energy. But, crucially, no more heat is generated than if there had been no turbines to capture the wind.

The same can hold true for solar energy. In an array of solar cells, if each cell only collects the sunlight falling on it – which would ordinarily have been absorbed by Earth’s surface – then the cells don’t alter how much waste heat gets produced as they generate energy. The light that would have warmed Earth’s surface instead goes into the solar cells, gets used by people for some purpose, and then later ends up as heat. In this way we reduce the amount of heat being absorbed by Earth by precisely the same amount as the energy we are extracting for human use. We are not adding to overall planetary heating. This keeps the waste energy burden unchanged, at least in the relatively near future, even if we go on extracting and using ever larger amounts of energy.

Covering deserts in dark panels would absorb a lot more energy than the desert floor

Chaisson summarised the problem quite clearly in 2008:

I’m now of the opinion … that any energy that’s dug up on Earth – including all fossil fuels of course, but also nuclear and ground-sourced geothermal – will inevitably produce waste heat as a byproduct of humankind’s use of energy. The only exception to that is energy arriving from beyond Earth, this is energy here and now and not dug up, namely the many solar energies (plural) caused by the Sun’s rays landing here daily … The need to avoid waste heat is indeed the single, strongest, scientific argument to embrace solar energies of all types.

But not just any method of gathering solar energy will avoid the deep warming problem. Doing so requires careful engineering. For example, covering deserts with solar panels would add to planetary heating because deserts reflect a lot of incident light back out to space, so it is never absorbed by Earth (and therefore doesn’t produce waste heat). Covering deserts in dark panels would absorb a lot more energy than the desert floor and would heat the planet further.

We’ll also face serious problems in the long run if our energy appetite keeps increasing. Futurists dream of technologies deployed in space where huge panels would absorb sunlight that would otherwise have passed by Earth and never entered our atmosphere. Ultimately, they believe, this energy could be beamed down to Earth. Like nuclear energy, such technologies would add an additional energy source to the planet without any compensating removal of heating from the sunlight currently striking our planet’s surface. Any effort to produce more energy than is normally available from sunlight at Earth’s surface will only make our heating problems worse.

Deep warming is simply a consequence of the laws of physics and our inquisitive nature. It seems to be in our nature to constantly learn and develop new things, changing our environment in the process. For thousands of years, we have harvested and exploited ever greater quantities of energy in this pursuit, and we appear poised to continue along this path with the rapidly expanding use of renewable energy sources – and perhaps even more novel sources such as nuclear fusion. But this path cannot proceed indefinitely without consequences.

The logic that more energy equals more warming sets up a profound dilemma for our future. The laws of physics and the habits ingrained in us from our long evolutionary history are steering us toward trouble. We may have a technological fix for greenhouse gas warming – just shift from fossil fuels to cleaner energy sources – but there is no technical trick to get us out of the deep warming problem. That won’t stop some scientists from trying.

Perhaps, believing that humanity is incapable of reducing its energy usage, we’ll adopt a fantastic scheme to cool the planet, such as planetary-scale refrigeration or using artificially engineered tornadoes to transport heat from Earth’s surface to the upper atmosphere where it can be radiated away to space. As far-fetched as such approaches sound, scientists have given some serious thought to these and other equally bizarre ideas, which seem wholly in the realm of science fiction. They’re schemes that will likely make the problem worse not better.

We will need to transform the human story. It must become a story of doing less, not more

I see several possibilities for how we might ultimately respond. As with greenhouse gas warming, there will probably be an initial period of disbelief, denial and inaction, as we continue with unconstrained technological advance and growing energy use. Our planet will continue warming. Sooner or later, however, such warming will lead to serious disruptions of the Earth environment and its ecosystems. We won’t be able to ignore this for long, and it may provide a natural counterbalance to our energy use, as our technical and social capacity to generate and use ever more energy will be eroded. We may eventually come to some uncomfortable balance in which we just scrabble out a life on a hot, compromised planet because we lack the moral and organisational ability to restrict our energy use enough to maintain a sound environment.

An alternative would require a radical break with our past: using less energy. Finding a way to use less energy would represent a truly fundamental rupture with all of human history, something entirely novel. A rupture of this magnitude won’t come easily. However, if we could learn to view restrictions on our energy use as a non-negotiable element of life on Earth, we may still be able to do many of the things that make us essentially human: learning, discovering, inventing, creating. In this scenario, any helpful new technology that comes into use and begins using lots of energy would require a balancing reduction in energy use elsewhere. In such a way, we might go on with the future being perpetually new, and possibly better.

None of this is easily achieved and will likely mirror our current struggles to come to agreements on greenhouse gas heating. There will be vicious squabbles, arguments and profound polarisation, quite possibly major wars. Humanity will never have faced a challenge of this magnitude, and we won’t face up to it quickly or easily, I expect. But we must. Planetary heating is in our future – the very near future and further out as well. Many people will find this conclusion surprisingly hard to swallow, perhaps because it implies fundamental restrictions on our future here on Earth: we can’t go on forever using more and more energy, and, at the same time, expecting the planet’s climate to remain stable.

The world will likely be transformed by 2050. And, sometime after that, we will need to transform the human story. The narrative arc of humanity must become a tale of continuing innovation and learning, but also one of careful management. It must become a story, in energy terms, of doing less, not more. There’s no technology for entirely escaping waste heat, only techniques.

This is important to remember as we face up to the extremely urgent challenge of heating linked to fossil-fuel use and greenhouse gases. Global warming is just the beginning of our problems. It’s a testing ground to see if we can manage an intelligent and coordinated response. If we can handle this challenge, we might be better prepared, more capable and resilient as a species to tackle an even harder one.