We’re almost 1/4 of the way through the 21st century, and sustainability has become a hot topic. So what will the world be like, from a sustainability perspective, in 2050? Let’s look at this question when it comes to one of the great environmental challenges – the greenhouse gas emissions driving climate change. Now making predictions is kind of fun, you hope you’re right but you’re also pretty sure that you’ll probably be wrong. But what I really love about this exercise, is how imagining the future opens up possibilities.
2050 is the year in which most countries have made commitments to achieve net zero greenhouse gas emissions. Perhaps because it’s a convenient round number – exactly halfway through the century, and also, several years ago, back when those commitments were made, 2050 seemed like a long way away. As we approach 2025, however, 2050 doesn’t seem that far away. Will net zero by 2050 be achieved? Well, at our current rate of progress it looks extremely unlikely. But could that change?
This is part 1 of a 2 part essay on this topic, and has a focus on the technical likelihood and feasibility of net zero.
Part 2 will look at the non-technical aspects, with a focus on making it an easy choice for energy users to decarbonize. These musings about 2050 go well beyond my 2025 predictions!
Technologies – 5 silver bullets.
Five silver bullet solutions that could help us get to net zero by 2050 include renewable energy, batteries, heat pumps, nuclear power, and AGI.
Renewable energy
The stunning drops in the cost of solar PV panels – built with silicon – show no sign of abating. This century solar PV has gone from, in 2000, being a very expensive technology only employed in space craft and in remote areas, to now, on a kWh basis, at utility scale being the second cheapest source of electrical energy (when the sun is shining). Onshore wind is the cheapest (when the wind is blowing). Whilst wind hasn’t seen the same dramatic reductions in price as solar, there have still been phenomenal improvements. Turbines are getting bigger and bigger, and the cost of offshore wind per kWh is competitive with coal (when the wind is blowing). https://www.lazard.com/media/xemfey0k/lazards-lcoeplus-june-2024-_vf.pdf
As discussed later, we can expect the cost of solar to continue to drop, and wind as well, but at a slower rate.
Batteries
Batteries have come from nowhere to provide short term storage of around one hour (by 2020), to now seen as capable of economically providing up to four hours of storage in markets with higher electricity prices. The Rocky Mountains Institute (RMI) sees battery prices continuing to plummet, whilst the energy density increases – that is the kWh stored per kg weight of the battery. They forecast that battery energy density will increase enough to make the use of batteries viable for ships travelling up to 1,500 km by 2030 and eventually for long haul aircraft. https://rmi.org/insight/x-change-batteries/
Batteries can be used to store energy from wind or solar plants, and make it available when there is no wind or sun. They can be used as a buffer with nuclear power plants, which are most efficient at full load, to better match the varying demand for electricity. They are modular and scalable.
Heat pumps
Heat pumps are a century old technology getting a new lease on life for low temperature heat applications, such as domestic hot water or space heating, that have traditionally relied on natural gas. Roughly four times as efficient per unit of useful energy produced than gas heaters, the economics are best in countries with low electricity prices. A lot of natural gas is used for low temperature heating applications that heat pumps are well suited to, and as decarbonization involves electrifying and switching to renewable energy sources, heat pumps reduce the amount of additional electricity generation that will need to be installed to replace natural gas.
Additionally, in the colder climates where natural gas is widely used for building heating, heat pumps reduce the large winter-summer imbalance in building energy consumption. In colder climates, many gas heated buildings easily use at least twice as much total energy in winter months as they do in summer months. This creates a seasonal imbalance in total building energy use, that poses a challenge when it comes to decarbonizing electricity supply with renewables. Solar energy production is highest in summer and lowest in winter, the opposite of the energy demand of buildings in cold climates. Heat pumps don’t eliminate this imbalance, but they reduce it, making it easier to get to net zero by 2050.
Nuclear power
This is a polarizing topic, but it can’t be ignored. Countries which have never had nuclear power plants are now looking more closely at it. The United Arab Emirates has relatively recently installed four nuclear reactors (from South Korea). In Bangladesh, the first of several reactors being built by a Russian company is close to being switched on. Egypt and Turkey are also building reactors. https://world-nuclear.org/information-library/country-profiles/others/emerging-nuclear-energy-countries
However, there is the problem of nuclear waste, the economics aren’t currently that good, they take a long time to build, and their deployment may require legislation where nuclear power bans exist.
The economics argument against nuclear– which shows that the levelized cost of energy (LCOE) is much higher than for solar and wind – doesn’t take into account the fact that the sun doesn’t shine all the time, but that electricity is always used, every second of the day, in almost every facility that uses energy. Nor does it take into account wind droughts – something experienced in Australia in recent months. LCOE simply doesn’t provide a fair basis for comparing technologies when considering the 24/7/365 demand for power.
However, on the other hand, the LCOE for nuclear is higher than for coal. To decarbonize generation sources cheaper than coal are needed, and nuclear as its currently built, isn’t one of them.
There is also the assumption that nuclear technology is somehow stuck in a time warp, and that there is no scope for the technology to be improved, for costs to drop and for faster construction times. Now there is some some merit to this argument – as the construction sector more broadly seems have had very few efficiency gains over the last 20, 50 years, in complete contrast to sectors such as semi-conductors. But perhaps AGI will change that!
AGI
If you are a sustainability professional reading this, you, my friend, are probably a master of acronyms. You don’t have to think twice about using acronyms like ESG, CSR, SBT, CBAM, ICE etc, but perhaps you are stumped with Artificial General Intelligence 😊.
AGI is a type of artificial intelligence that equals or betters human capabilities. Who was not shocked when ChatGPT came out – and who hasn’t used it augment their capacity to quickly write something? Futurists look forward to a world where AGI can be employed to help solve problems like climate change, by augmenting human capacity. I believe it will make a great contribution to net zero by 2050.
OpenAI predict that AGI will be achieved by 2027; futurist Ray Kurzweil, by 2029. And after that? AGI will continue to improve.
AGI, coupled with improvements in sensors, motors, materials and the mechanical components of robots, mean that that it may be possible to convert a construction site, from something that is slowly built by human labourers to something that can be built much faster, to a better quality standard, and much cheaper. Perhaps one such construction site could be a nuclear power plant. Possibly AGI could study fission and fusion to an extent unimaginable now, resulting in new power plant designs that have minimal long-life radioactive waste, or no waste – the long-hoped promise of nuclear fusion.
Another construction area where AGI robots might be able to greatly reduce costs is in the construction of electrical transmission lines needed to link up solar and wind farms. As with nuclear not necessarily easy to do. For example, if the current major costs are associated with land purchases for overhead cables, can AGI and materials advances be used to revolutionize underground cabling?
Heating and cooling account for most of the energy used in buildings globally. The reason for this is that many buildings are old and are poorly insulated and sealed. And single pane glass windows, and even double glazing, still transfers a lot of unwanted heat (either in our out of buildings) as do aluminium window frames.
One of the key impediments to the deployment of solar, improved insulation and greatly improved glazing to existing buildings is the high labour cost associated with this. AGI powered robots have the potential to radically change this.
Of course this proposes other questions, like what will be the future of work. And there are also fears that AGI will do away with humans. But these aren’t the topics of this essay.
If AGI is new to you, I strongly suggest that you spend some time learning more about it.
Can these technology silver bullets enable net zero by 2050?
The RMI forecasts for batteries are based on a version of Moore’s law – that computing power would double every year for at least a decade – and proved remarkably precinct back in the 1965 when proposed by Gordon Moore of Intel. In 1975 he modified it to say there would be doubling every 2 years. And he was right, for many more years. Now, whilst the doubling in capacity is now taking longer than 2 years, there are still improvements. Nearly 60 years on, and the principle of Moore’s law – that computing power will continue to increase – still holds!
Learning curves and scale
Variations of Moore’s law can be applied to other sectors, which demonstrate similar characteristics to semi-conductors. A variation of Moore’s law would be to state that “anything that can be manufactured in every-increasing quantities will drop in costs at a rate in proportion to the growth in annual production.” Let me illustrate with three examples:
- Transistors, on which Moore’s law is based. A transistor is basically an electronic switch. Well, we can’t really label them transistors any more, “chip” might be a better word that describes a semiconductor device containing many, many transistors – in the tens to hundreds of billions in a computer chip. Potential market size just continues to grow as more and more of the tools and equipment we use in daily life incorporates electronics, and is now probably on track to reach the trillions (X,000,0000,0000,0000) per capita. Have a look at https://en.wikipedia.org/wiki/Transistor_count. If you are reading this on a phone – remember your phone has more computing power that the space craft that first took people to the moon. No one doubts that performance will continue to improve whilst costs drop.
- Solar panels. The current potential market is roughly in the order of dozens per capita, and there is still scope for large growth in annual production. Costs are a tiny fraction of what they were in the 1990s, back when I first used them, and efficiency is considerably higher. The market is much smaller than the silicon chip market, but is still a huge, untapped market (say 50 solar panels per capita x 8 billion people = half a trillion solar panels). It’s inevitable that performance will continue to improve, whilst costs drop -including the costs of inverters – which use silicon chips! In fact there is a variation of Moore’s law, called Swanson’s law just for solar PV – which states that the price of PV modules drops 20% for every doubling of cumulative shipped volume.
- Cars. Potential market is perhaps around 0.5 per capita, so perhaps 2 billion globally. A much smaller potential market that solar panels, and perhaps even less so when full self driving cars exist and few people own their own car even more. It’s not really a growth market as a whole, although EVs are. As a result car’s have tended not to drop too much in price, but have increased in safety and luxury. This is consistent with my variation of Moore’s law – as there is little to no growth in annual production there hasn’t been the dropping of costs we have seen in other sectors.
Applying learning curves and scale to the silver bullets
So if we take these principles associated with learning curves and scale, and the silver bullets discussed earlier, and look at the world’s greenhouse gas emissions, will net zero by 2050 be achieved, with fossil fuels no longer used to provide energy?
Let’s look at renewable energy and batteries. It is widely recognized that wind and solar are now the cheapest forms of energy – when the wind is blowing and the sun is shining. And prices will continue to drop whilst performance improves. Can batteries make net zero possible when it is not possible to use energy from the sun or the wind directly? Where nuclear is deployed could batteries help enable net zero by serving as a load buffer, and eliminating the need for natural gas peaking plants?
The potential market for batteries – or more specifically let’s say an individual battery cell (of which typically there are currently hundreds to thousands in an electric car) is in the order of hundreds to thousands per capita. This is larger than the potential market for solar panels, probably by one or two orders of magnitude. There is HUGE scope for market growth, and thus a lowering of costs and improvement in performance. If, like me, you have been amazed at how cheap solar panels have become so quickly, be prepared to be amazed at what battery prices will do. RMIs predictions that batteries will be used to power ships and aircraft may seem fanciful, but once we consider experience curves, they don’t look out of place. And that could mean that batteries are moved around just like fuel tankers currently move diesel and gasoline around. Batteries will play a pivotal role in net zero by 2050.
Can AGI break the learning curve and scale paradigm?
With all due respect to my friends in favour of nuclear, as a custom-built power plant with little opportunity to benefit from experience curves due to very low volumes, I think that nuclear will struggle to compete – unless AGI can be used alongside human ingenuity to greatly improve the case for nuclear, by lowering costs, speeding up deployment times, slashing or eliminating the waste problem and safety concerns. That a very big ask, but, in comparison to solar and wind, nuclear can be used as a “drop in” replacement for fossil power plants as they retire, unlike wind and solar farms which can require large transmission investments to connect to the grid.
If human ingenuity can harness AGI to cause a lowering of costs in products that aren’t manufactured in every increasing volumes and at large scale, whilst also greatly improving them, this opens up many possibilities that will reduce the cost of decarbonization and make net zero by 2050 easier to achieve.
All these advances give me hope that net zero by 2050 is technologically achievable. But will our collective choices as humans, large and small, in how we use energy, get us there?
That is the subject of part two of Net Zero by 2050.
All images in this article have been AI generated by DALL:E
