People have often lacked energy sources – for heating, cooking, and all sorts of other things. Actually, when you think about it, that is pretty weird, given that the vast quantity of radioactivity in the subsoil of the earth – chiefly in the form of uranium, thorium, and potassium – contributes substantially to the fact that 99% of the soil is hotter than 1000ºC and only 0.1% – the thin crust on which we live – cooler than 100ºC. Added to that is the fact that global energy from solar radiation is about 7000 times greater than our current global consumption of energy.

Nevertheless, throughout the history of humankind, the struggle to obtain the necessary energy to sustain life has often been one of the hardest. However, this was generally due to lack of innovation, as when our ancestors would sometimes die of cold, lying on twigs and branches that they could have turned into a big, warming bonfire – if they had just known how to light it. In other words, there was no lack of resources; merely lack of the technology for using them. Once again, this illustrates that the ultimate resource is always innovation.

Together with the use of animals for transport and the like, burning wood and twigs (not forgetting manure) was people's main source of energy in the Stone, Bronze, and Iron Ages. Approximately 3000 years ago, however, the Chinese began to use coal to extract iron, and directed natural emissions of natural gas through bamboo pipes so they could use it for boiling seawater and extracting salt. And, nearly 2000 years ago, in Dacia in present-day Romania, the Romans extracted oil. However, the large-scale commercialisation of oil extraction only kicked off in the mid-nineteenth century in the United States. Following that, the use of fossil fuels – both coal and gas – rose steadily throughout the world.

Fossil fuels – the culmination is approaching

Given the fact that this has been so important to our civilisation, it is interesting that some analysts believe that the use of fossil fuels will come to an end pretty soon. The following graph from the University of Utah provides an (uncertain) estimate of development in the world's total energy supply.

According to this particular estimate, we are now only around 15 years from ‘peak fossil’, which incidentally includes peak oil and peak gas – unfortunately followed by ‘peak coal’ just a few years later. However, I have to add that there is considerable disagreement among analysts about these forecasts. That said, this forecast's revolutionary development can be ascribed chiefly to three factors: (1) the fact that global population growth is slowing down; (2) we are developing new forms of energy; and (3) we are getting better at utilising the energy we consume. Just look at the graph here, which shows the development in US GDP in relation to the development in Americans’ energy consumption.

Forecast for the development of the global energy supply, 2017–2050

US energy consumption per unit of GDP

The population is growing, and wealth is increasing, while energy consumption per GDP unit is falling – and this development is expected to continue. It is not some pie-in-the-sky fantasy. The development is already well underway.

However, we should not expect that this will also lead to a drop in global energy consumption in the foreseeable future. This is partly due to the following rule:

27. When technological advancements result in more efficient use of resources, prices can fall, which can then lead to increased consumption of that same resource (Jevon's Paradox).

Almost anywhere in our current energy infrastructure, efficiency may continue to improve significantly. An example: our use of traditional light bulbs resulted in the loss of 99% of the original energy. However, as we change to LED, the efficiency improves approximately six-fold, and it continues to improve since networks and LEDs get ever more efficient.

The efficiency of solar panels has also been substantially streamlined and has followed Swanson's Law, which states that the price of solar modules tends to drop 20% for each doubling of cumulative shipped volume. In recent years, this has halved the cost per unit approximately every 10 years. However, neither solar cells nor LED are IT and, within the foreseeable future, they will encounter some physical limitations.

Vast, constantly growing reserves

When a peak in the global consumption of fossil fuel comes, will it be created by lack of supply or by lack of demand? For many years, as previously mentioned, people believed that the adventure would come to an end on account of lack of supply. Now we actually seem to be quite close to this peak fossil energy, but the reason is not that the supply will fail, but that the demand will, just as the Stone Age did not come to an end because we ran out of stone.

This is remarkable. But what makes the story even more amusing is that the end of the oil era is likely to happen while the reserves are almost at their highest level ever – and while the real prices are rarely low.

Let's take a look at the reserves first. This graph shows the developments in oil and gas reserves since 1980, when panic about a future depletion of oil was rife.

Proved oil and gas reserves

As can be seen, these reserves have increased steadily, even though we have also burned huge amounts of oil and gas.

If, as expected, production peaks in a limited number of years, my guess is that, during the subsequent period of scaling down, we will consume more or less the same as we did on the way up. If so, that means that our consumption of coal, oil, and gas will gradually de-escalate over perhaps 100 years or so, after which we will leave the vast majority of our fossil deposits in the subsoil, and never use them as fuel.

And that brings us to the issue of prices, because leaders in Saudi Arabia, Russia, the United States, and other oil-producing countries are quite clear about what is happening. They now have to live with a feeling that can be compared, say, to 10 restaurants, each of which has baked 100 beefsteaks – in other words, a total of 1000 beefsteaks – for that evening's guests, only to discover that just 500 guests are coming. As I write this, Saudi Arabia has shown an interest to list (and thereby partly sell off) their national oil company (which, by the way, was the most profitable company in the world in 2018), and the Americans are fracking oil at an unbelievable rate – they are actually expected to become net self-sufficient in oil around 2022 and then to start exporting more oil than they import. Sell, sell, sell! And the production costs of fracking are dropping.

So, the current reserves of coal, oil, and gas are vast, and larger than we will probably ever use. But, on top of that, we have methane hydrate, a combustible hydrocarbon, believed to be 2–10 times more widespread than natural gas, which means it could potentially stretch to centuries or millennia of further consumption. The Japanese, for instance, are actually experimenting with it. But it seems unlikely to me ever to become big business – we have plenty of coal, oil, and gas, for which we already have a well-developed extraction and distribution structure.

To understand what will happen as the fossil era gradually winds down, it is not a bad idea to be aware of this rule, which I have taken the liberty of formulating on the basis of research by the physicist Cesare Marchetti:

28. The world's energy supply is undergoing an exponential shift from carbon to hydrogen, which, if it continues, will be completed by around 2150 (Marchetti's Law).

This shift started around 1860 and, if the rule continues to hold true, it will be just about completed by about 2150.

The global shift from coal to hydrogen

In other words, over an approximately 200-year period, humankind will implement an exponential shift from carbon to hydrogen. But do not misunderstand me. It does not necessarily mean that we will be driving around in hydrogen cars with fuel cells. What it does mean is that, unlike hydrogen, the carbon atom is not an energy source, but more akin to the branch on which the useful grapes grow. Carbon in itself does not provide the desired energy gain; during the burning process, it merely lets go of the valuable hydrogen. Carbon then reacts with oxygen to form carbon dioxide, whereas vast amounts of energy is released as hydrogen reacts with oxygen to create water. The reason for Marchetti's Law is that, while there is an immense amount of carbon per hydrogen unit (and energy unit) in wood, there is less in coal, even less in oil, and very little in natural gas. The ratio goes from 0.1 hydrogen atom per carbon atom in wood (i.e. 0.1:1) to 0.5:1 in coal and 2:1 in oil to 4:1 in natural gas. In other words, gas is approximately 40 times less carbon-intensive than wood, and eight times less carbon-intensive than coal.

Human carbon dioxide emissions since 1850

As the illustration shows, our carbon dioxide emissions have increased fairly linearly since World War II due to Jevon's Paradox more than offsetting Marchetti's Law. We reached the halfway mark through the shift from a mainly carbon economy to a mainly hydrogen economy back in 1935. And, if this trend continues, we will have a 90% hydrogen economy by 2100, when our total energy consumption will also be much greater than today, but when our consumption of fossil fuels will probably be lower than now.

According to the International Energy Agency (IEA), an energy research institute created by the 30 OECD countries, as we head towards 2040, natural gas and renewable forms of energy (mainly solar and wind) will be the fastest-growing forms of energy (in absolute figures), but energy-saving technologies will also play an important role. However, there are certain limits to how much wind energy can and, in my opinion, should be scaled. In 2009, the serial entrepreneur Saul Griffith estimated that if, over a period of 25 years, we were to convert 80% of global energy production (which was 16 TW) to new forms of energy, the consequences would be highly significant. So, transferred to more recent figures and something we can relate to, the following applies:

• If 2 out of 11.5 TW is to be covered by photovoltaic solar power (which generates electricity), in 25 years we need to install 80 000 km² of solar panels to generate electricity. That is the equivalent of 1.8 times more than the total land area of Denmark. If a further 2 out of 11.5 TW is to be covered by solar thermal panels (which heats up fluids), in 25 years we need to make just under 40 000 km², which amounts to just under the area of Denmark. If we put the two types of solar cells together (photovoltaic and thermal), so that solar cells constitute 4 out of 11.5 TW, or 25% of a global energy consumption of 16 TW, we will need to build just under 120 000 km² over 25 years. That means covering the whole of Denmark, the Netherlands, Switzerland, and North Korea entirely with solar panels.
• If a further 2 out of 11.5 TW is to be replaced by wind turbines (3 MW turbines), we will need to erect 2.6 million of them over 25 years, or alternatively 800 000 giant wind turbines such as the V-164, which can produce up to 9.5 MW. Given that the average service life of a wind turbine is approximately 25 years, we will then perpetually have to scrap and replace approximately 100 000 regular wind turbines or 25 000 giant wind turbines annually.
• If we were ultimately to convert 2 TW to nuclear power, and if each of them was about 1000–1600 MW, we would have to build between 2000 and 1250 reactors. Given that there are normally two to four reactors per nuclear power plant, this would correspond to about 700 or 800 plants.

Of course, all these figures will change in tandem with technological development, but it is interesting to see that a single nuclear power plant supplies roughly the same amount of energy as just under 3000 wind turbines, 1000 giant wind turbines, or 160 km² of solar panels – and this is without accounting for the complicating fact that sometimes the wind does not blow and the sun does not shine. In this context, however, it must be said that 160 km² of solar panels do not necessarily have to take up space and replace forests and meadows. As we will see later, it is possible to make building-integrated solar panels, such as solar-utilising roof tiles, flat roofs, walls, and even windows, so that a building actually becomes a virtual solar panel without it being immediately visible or taking up more space for that reason. Especially in the wide belt around the equator and a long way above it, and in certain places within the temperate zone, this is already starting to make economic sense.

Nuclear power – a renaissance on the way?

When it comes to nuclear power, as I write this book, there are approximately 450 nuclear reactors in the world, providing about 7% of all energy, including around 17% of all electrical power. Nuclear power is shrouded in an exceptionally apocalyptic aura, making it difficult for people to relate objectively to the technology. This is a result of a number of extremely effective campaigns – complemented by dramatic Hollywood films – at times when safety was significantly inferior to now and the issue of waste unresolved. And there have also been accidents. However, even though these accidents have led to relatively few or no fatalities, and often very minor long-term effects, they have always entailed a significant setback for nuclear power. This is unfair, because, based on the experience of more than half a century, nuclear power – even in its first, primitive versions – has proved to be the safest form of energy in the world by far. This is even though most existing nuclear power plants are based on technology that has long been superseded by something far safer.

A comprehensive study conducted by Markandya and Wilkinson showed all the fatalities – both direct and indirect – per TWh, produced by different forms of energy. Here it is:

• lignite (brown coal): 32.75 fatalities
• coal: 24.62 fatalities
• oil: 18.43 fatalities
• biomass: 4.63 fatalities
• natural gas: 2.82 fatalities
• nuclear power: 0.07 fatalities.

It is striking that the global death rate for a coal-based energy unit has been approximately 400 times greater than that for nuclear power, which was hampered by the ‘safety risk’. According to some estimates, the activism against nuclear has led to something like 10 million extra fatalities.

I would also like to add that the tremendous fear of radioactive radiation in itself is unjustified. Yes, of course you can die from an overdose of radioactivity, just as you can die from falling off a ladder when attempting to install a solar panel – or from lung cancer caused by the smoke from a coal-fired power plant. But then, scientists have been struck by studies which showed that people who lived in an environment with an unusually high daily dose of radioactivity had less cancer than the rest of the population. This led to two interesting meta studies: reviews of the overall research on the effects of radioactivity on health. Both reports concluded that a certain daily dose of radioactivity is actually healthy and prevents cancer, and that the optimum dose was 20–100 times more than we receive on average. It sounds totally off the wall, but is probably due to the fact that radioactivity stimulates chemical error correction in DNA – although it can also cause errors.

Incidentally, this observation is part of a broader phenomenon called ‘hormesis’, in which small doses of otherwise hazardous substances or conditions can be healthy. For example, it has been discovered that small doses of pesticides can be healthy because they stimulate or train our chemical defence mechanisms. Similarly, moderate exposure to bacteria is healthy because it stimulates the immune system and prevents autoimmune diseases such as allergy. One modern recommendation, therefore, is to eat 3.5 kg of dirt per year. When I was a boy, many of my friends were instructed to wash their hands before they ate. But my parents did not tell me to, nor did I do so with my own children. Everyone needs a bit of dirt! Moderate exposure to stress can also make us more mentally robust, just as moderate exposure to physical hardships makes us physically stronger. Yes, of course we can sustain harm or die from too much radioactivity, toxins, infections, stress, and physical exhaustion, but, on the other hand, we should not have too little of them all.

Unfortunately, our old friends the fact-resistant hedgehogs have played a significant role in the nuclear power debate. But so far, the most important rational arguments against nuclear power have been related to the waste issue and to the fact that it has been very expensive to build nuclear power plants, since too few were built to make it cheap.

Now, however, we are developing smaller, far safer, and much cheaper nuclear reactors, which make it likely that nuclear power will have a well-deserved renaissance – especially in Asia. For example, the old reactor designs from Chernobyl and Fukushima Daiichi have been replaced by new ones, including Toshiba's 4S (Super Safe, Small, and Simple) or Lawrence Livermore Lab's SSTAR (Small, Sealed, Transportable, Autonomous Reactor), which are also referred to as ‘atomic batteries’. In addition, in the future we are likely to be able to transmute the waste into something harmless.

At the same time, much is happening on the actual fuel front. Deploying already-known technologies, we will have uranium for the following thousands of years of consumption, and if we also start extracting it from the sea, we will have enough for hundreds of thousands of years.

Nuclear fusion – the ultimate energy solution

However, the ultimate technology is nuclear fusion, and governments and private companies are now competing to get there first. In this case, the fuel is hydrogen. Hydrogen, it should be said, is element number 1, in the top left-hand corner of the periodic table, and it is also the most prevalent atom in the universe. It comes in three variants or isotopes:

• 99.98% of it consists of the stable protium (¹H) with one proton and one electron;
• 0.02% consists of the equally stable deuterium (²H) with one proton and one electron, but also with one neutron – this is also referred to as heavy hydrogen, because the addition of the neutron adds to its weight;
• there are quite minimal amounts of tritium (³H), which, as well as one proton and one electron, has two neutrons – this is why it is also called super-heavy hydrogen.

Tritium develops spontaneously in the atmosphere when hydrogen is hit by neutrons whizzing from the universe in the form of so-called cosmic radiation. The reason why it is still extremely rare in nature is that it is unstable and, with an average half-life of just about 12 years, it is transformed to helium, which is number 2 in the periodic table and the second-most prevalent atom in the world.

Today, large amounts of tritium are produced in traditional heavy water reactors. Otherwise, it can be created by bombarding lithium with neutrons. Lithium, it has to be said, has three protons, three electrons, and usually four (sometimes three) neutrons. As is obvious from its name, it is used in lithium batteries: for example, in electric cars or notebook computers like the one I am writing on. If one day we all drive electric cars, we will drown in used lithium batteries, unless we change to another type.

Now comes the interesting part. If you combine deuterium and tritium in a process of so-called nuclear fusion, the result is helium. Helium has two protons, two neutrons, and two electrons, which means that there is one neutron too many in the process, and it is ejected at tremendous speed. Together, the resultant helium and the ejected neutron weigh a mere 99.3% of what the deuterium and tritium, which it was created out of, weighed. The remaining 0.7% mass is converted into energy: cf. Einstein's famous equation E = mc². In fact, a lot of energy!

In other words, here we have a process that involves the most and the second-most prevalent atoms in the universe, which, incidentally, are responsible for well-nigh 100% of all energy generation in the universe: the energy of all the stars.

There is something really ultimate about it. It is, as it were, the mother lode of energy – the energy form above all others. And it is even four times more compact than thorium itself. The energy content of the fuel for nuclear fusion is approximately 10 million times greater per unit weight than coal. In addition, the waste from nuclear fusion will be our harmless helium. It happens to be what we use for blowing up balloons for children's birthday parties, which gives some indication of just how harmless it actually is.

Your bathtub can probably hold about 300 l of water, and the hydrogen in that water probably contains all the deuterium you will need for your entire life's personal energy consumption, if supplied with nuclear fusion. In addition, we must then obtain an equivalent amount of tritium, which can be made from the amount of lithium found in two small batteries from a normal notebook computer. These amounts are so small that scientists estimate we can provide enough of it to supply the world with clean energy for somewhere between 30 million and countless billions of years – solely through nuclear fusion. In this context, note that the lowest estimate – 30 million years – is about 100 000 times longer than the 300 or so years the fossil period is likely to last. Also note the extra twist to this story. If we actually introduce nuclear fusion later this century, it will be the solution that will propel Marchetti's Law towards a pure hydrogen economy.

But how will we accomplish this? The major technological challenge is to achieve a so-called triple product, which is critical density multiplied by temperature multiplied by time. We have this within stars, including the sun, where the pressure is 100 billion atmospheres, the temperature 15 million degrees, and the process fortunately permanent. Since the 1960s, work has been done to raise the triple product, which has actually doubled on average every 1.8 years:

29. Since the 1960s, nuclear fusion experiments have doubled triple product – the combination of density, temperature, and confinement time – every 1.8 years.

That means it has increased approximately 100 000 times since the start of experiments and, at the time of writing, we are only missing a factor of five before we reach the level required for the major engineering breakthrough in the form of a functioning reactor. In other words, we are almost – but not quite – there. That’s the good news. The bad news is that it has been a struggle to achieve the necessary density.

Now there are about 20 fusion reactors up and running and a dozen in the pipeline or under construction. The majority fall into two categories: magnetic fusion and inertial fusion. The former uses a circular or twisted-circular fusion chamber, into which the deuterium and tritium are blown. They are held around the centre with a huge magnetic field while being heated to a point at which the fusion process (hopefully) occurs. Here it is primarily the input of heat, rather than pressure, that is needed to trigger the process.

The leading project of this kind in the world has arguably been the Joint European Torus (JET) project in the United Kingdom, which has set the ‘world triple product record’. However, if it has not already been surpassed by the Chinese or others, it probably will be by ITER in France, which is funded by the European Union, China, India, Japan, Korea, Russia, and the United States. ITER is expected to start in 2025, and there are reports that the process should work by around 2027 – give or take a few years, I assume. However, neutral nuclear power experts I have spoken to are extremely sceptical vis-à-vis magnetic fusion technology and think it may very well be a dead end.

The major alternative is inertial fusion, in which the experts I have talked to have greater faith. Here pressure rather than heat is the key element that triggers the process. One principle involves shooting ‘cartridges’ of deuterium and tritium into a combustion chamber and then hitting them in the air with a number of extremely powerful laser beams, which then ignite them. It is a bit like automatic clay pigeon shooting. The pioneer in this field is the National Ignition Facility (NIF) in the United States, which has already achieved short-term partial ignitions, and in France they are in the process of creating a laser mégajoule (LMJ), which should be capable of achieving the same.

In addition to such public experimental reactors, there are a number of others, many of which are private. In this context, one of the alternative concepts is to shoot fuel components into each other with particle accelerators. Another, which is being developed by General Fusion in Canada, uses a circular space, in which molten lead and lithium rotates. This eddy current creates a vortex, into which one pumps deuterium and tritium. This is then ignited by targeting the container with rhythmic piston strokes from the outside. Another intriguing project is Sparc, which is a collaboration between Massachusetts Institute of Technology (MIT) and Commonwealth Fusion Systems. They expect to have a commercially usable reactor up and running by around 2030.

In addition to getting triple product up to the point where they have stable nuclear fusion, the goal is to increase the amount of energy that comes out in relation to what needs to be brought in. In the industry, this ratio is referred to as ‘Q’, and a Q below 1 obviously makes no commercial sense. Most experiments are now aiming at a Q of at least 10.

So, when will we have sustainable nuclear fusion with a meaningful Q? Let us just peruse some published timelines for the six nuclear fusion projects:

Here, the light grey indicates that the projects are under development, the dark grey indicates attempts to raise Q to above 1, while the black indicates attempts to raise Q to above 10, at which point it becomes commercially interesting. In other words, by 2030, according to their own published plans, ITER, NIF, and LMJ should be well underway with projects with Q ≥ 10, added to which breakthroughs may occur in various private projects.

It is totally unclear whether these six timelines will turn out to be viable, nor do we know whether other projects will be the ones that first reach the goal. But what we do know is that from the moment a test reactor delivers stable Q ≥ 10, it will take some time before a plant can run profitably and stably. One challenge relates to the thick walls of the fusion chambers of some of the designs. They are dimensioned to absorb the constant bombardment with neutrons from the fusion chamber, which cannot be held by a magnetic field, since they lack electromagnetic charge. This bombardment will be so powerful that the walls with their current materials will only last one to two years. Then the reactor will have to be switched off, and the wall – which in some designs is big, very heavy, and slightly radioactive – replaced. With regard to the radioactivity of the walls, this will be phased out over about 10 years, but, as a professor from MIT once explained, the problem could easily be fixed by immersion in a water basin for 10 years, following which you could probably stand next to it.

Another challenge lies in making the plants smaller and cheaper. ITER, for instance, takes up an area equivalent to 60 football pitches and costs €20 billion. All the private initiatives are aimed at making something more compact and cheaper. For example, the company General Fusion's reactor is located in what they humorously refer to as a (large) garage, and Sparc is 1/65 the size of ITER: partly because they use super-magnetic adhesive tape to create the magnetic field, which is extremely compact. In addition, Sparc seeks to reduce the problem of the core wall by making it much thinner, so that a large number of the neutrons pass through. So, behind this, liquid salt containing lithium will then circulate, turning, as a result of the neutron bombardment, into tritium. That's pretty damned smart! Similarly, as already mentioned, since General Fusion will also mix lithium into their rotating molten metal, they can extract tritium from that. Again, brilliant!

So, what are we to think about all this? I think it doesn't sound impossible that by 2025–2030 we will have at least one project that demonstrates continuous nuclear fusion which has a Q above 10. And it will probably be based on inertial fusion. It is no coincidence that Helion Energy, the NIF, General Fusion, and a number of other serious operators in the field are pursuing various varieties of inertial fusion. But I cannot be more certain than that. No matter which projects eventually reach a high enough Q to make it interesting, the people behind them will be aware when many of their approaches have already become obsolete. Then I guess the activity in the field of fusion, which is currently worth only about $4 billion a year, or around 0.005% of global GDP, will explode into a kind of other-worldly gold rush.

Regarding what is happening with different energy technologies, it is also important to point out that, contrary to how it is often portrayed in the media, the world is not up and running in terms of an impending complete conversion to solar and wind. According to the IEA's 2018 World New Policies Scenario – which is very different from the one from the University of Utah that I described earlier – by 2040, these energy forms will account for just 4.1% of the world's energy supply. And that is IEA's most optimistic scenario, which assumes that every nation's targets for renewable energy will be reached. The same statement assumes that the rest of the world's heavily increasing energy needs will be met with increasing consumption of fossil fuels. In this context, we should remember that solar and wind are not digital technologies, and there is no way they will experience sustained exponential increase in efficiency.

Yes, this illustrates that there are big differences in forecasts, but historically speaking, time and time again, it has taken humankind 60–70 years to make a significant shift in terms of the dominant type of energy, and if we are to repeat that in this context, it is going to require a technological ‘black swan’ – the introduction of one or more radical new technologies. Thorium is one idea and nuclear fusion an even better one – if it works. And other ideas may come up.

Maybe things will work out as follows. In or around 2040, maybe (maybe!) the first commercial nuclear fusion plants will be connected to the power network. Around 2060–2070, nuclear fusion will be one of mankind's dominant forms of energy. And by 2100 it will have become the biggest and actually predominant – maybe with solar energy as number two, followed by wind and natural gas. Of course, I'm only guessing. There are so many uncertainties.

I would also like to add that, ultimately, it may not be deuterium–tritium fusion that ends up taking the lead in fusion. Instead, it may turn out to be fusion between proton and proton or between deuterium and lithium, neither of which produce excess neutrons. Either would be jolly good, in my opinion.