Science, geopolitics and economics of solar photovoltaic energy

In this article I show a brief mix of science (obtaining silicon), geopolitics (China’s hegemony) and economics (evolution of solar photovoltaic energy costs).

Cycle of obtaining a photovoltaic module from its raw material: silicon. Here I focus on obtaining silicon

The solar photovoltaic energy production industry owes its success, among other reasons, to obtaining high-purity silicon and manufacturing solar panels at extraordinarily competitive prices. Here I detail it.

  1. Science: obtaining silicon

The raw material of the photovoltaic industry is quartz-rich sand, a crystalline form of silicon oxide. Silicon factories heat it to 1,900°C in electric arc furnaces with some carbon, in the form of coke. The oxygen in the sand reacts with the carbon to create carbon oxide, leaving molten “polysilicon”. This step requires a lot of energy to complete

SiO2 + C ? Si + CO2

It is then cooled, ground and reacted with hydrochloric acid to produce a volatile liquid called trichlorosilane, which is repeatedly distilled to remove all traces of impurities.

Si + 3HCl ? SiHCl3 + H2

SiHCl3 + H2 ? Si + 3HCl

The most advanced factories purify this so-called “metallurgical grade” silicon to a purity of “10 nines”, meaning that the polysilicon they obtain is 99.99999999%. This silicon is called “electronic purity”.

Flowchart of the production process for electronic-grade silicon

Until the beginning of this century, electronic-grade silicon dominated the market entirely, being purchased by chip-making industries. The solar cell industry lived off the “leftovers.” But in the mid-2000s, government subsidies in many countries (including our own) to encourage renewable energy caused the demand for photovoltaic energy to increase the demand for purified silicon beyond what the chip industry could do without. As the price of silicon rose, a number of companies in Asia began to invest heavily to build silicon-processing and purification factories to supply the photovoltaic industry.

  1. Geopolitics: China, the world’s leading producer of photovoltaic silicon

China quickly took the lead, and kept it. By 2023, Chinese companies will make 93% of all the world’s silicon for solar cells. Some of these industries are end-to-end manufacturers and also make the solar cells. Others sell the suitably purified silicon to other companies that cut it into wafers and polish it to make the solar cells.

Silicon manufacturers benefit from the fact that they are key players in China’s industrial strategy. There have been some bankruptcies, but the Chinese government has provided loans to many over-indebted companies. China’s two largest silicon producers, GCL-Poly and Tongwei, each had production capacity of 370,000 tons in 2023, more than enough to meet global demand. Tongwei is investing some $4 billion in a new plant that will eventually be able to produce 400,000 tons a year. China has planned facilities capable of producing 7 million tons a year, enough to make 3.5 TW of solar panels a year.

In terms of the amount of silicon such factories would require, these amounts are enormous. But it is worth noting that, compared with the material needs of other energy technologies, they are tiny. For example, coal production requires about 8 billion tons a year; add oil and gas to that and the figure doubles.

The Chinese industry is also fiercely competitive in the sense that only companies that make more or less the same thing can be competitive, but that is still a drawback. This is because solar cells are standardized products that are made using very similar steps from one manufacturer to the next. So manufacturers compete fiercely by trying to cut costs, either by making cells that produce fractionally more electricity for a given level of illumination or that cost less. But the “less,” in this case, is tiny.

Solar PV not only creates relentless competition on the supply side. It also generates an incredibly diverse demand thanks to perhaps the greatest advantage of the technology: its modularity. The revolutionary thing about solar energy is that it targets all types of customers. From the citizen who buys a €2 charger for his mobile phone to the company that develops 1 GW photovoltaic plants, everyone who uses solar energy basically buys the same product. There is no other energy generation technology in which one unit or a million units of the same product are installed, depending on the application. It is unimaginable for a citizen to install a nuclear power plant on the roof of his house for his own consumption of electricity.

  1. Economy: the astonishing price reduction of solar panels

The key to how this demand grows must be found in the “learning curve” of the sector. From the mid-1970s to the beginning of the 2020s, the cumulative installed capacity of photovoltaic energy multiplied by a million. Over that time frame, prices have fallen 500-fold (from ~€100/W in 1970 to €0.2/W in 2020). That represents a 21% decrease in costs for every doubling of installed capacity, which means a halving of costs for every 360% increase in installed capacity.

Learning curve for photovoltaic technology. As cumulative installed capacity increases, the price of solar energy declines exponentially. For more than four decades, the price of solar panels has fallen by 20% with each doubling of global cumulative capacity

Not only have solar panels become cheaper faster than the other big renewable technology, wind, they have also gone relatively unnoticed. In the case of wind, a more efficient generator means bigger turbines, on bigger towers, which means a big visual impact. In the case of photovoltaic installations, and even though covering tracts of land with them bothers some people in some places, in general solar panels are popular, they enjoy more “social license” than any other form of energy generation, whether renewable, fossil or nuclear. And this is only just beginning.

The exponential growth of solar energy: a new era of cheap and unlimited energy
Recently, the prestigious British magazine The Economist has published an article dedicated to highlighting the future of solar energy in the global energy mix. Given its interest and because it can only be read in English and under subscription, I offer here the main contents of the article, with some personal contributions that I think will be of interest, especially the graphic content, which is all personal, as well as all the links to the various sources that appear in the text. I have also modified the wording of some paragraphs to make them more understandable, and incorporated new data that reinforce the theses of the aforementioned article.

  1. When did this start?

It has been 70 years since AT&T’s Bell Labs presented a new technology to convert sunlight into energy. On April 25, 1954, Calvin Fuller, Gerald Pearson and Daryl Chapin announced the invention of the first operational silicon solar cell. The telephone company hoped to replace the batteries that powered equipment in places where electricity was not available. They made this known at a press event where one of these solar cells illuminated by a lamp turned a toy Ferris wheel non-stop.

The exponential growth of solar energy: a new era of cheap and unlimited energy

Above: Bell Labs’ “photovoltaic Ferris wheel.” Below: the first application of solar cells: remote telephone stations

  1. The present and the foreseeable future

Today, solar energy has far surpassed the toy phase. The panels already occupy 10,000 km2 (the area of ??Asturias) and this year they will supply the world with nearly 6% of its electricity, that is, almost three times all the electricity consumed by the United States in 1954. However, this historic growth is only the second most striking aspect of the solar energy boom. The first and most remarkable is that it has only just begun.

The exponential growth of solar power: a new era of cheap and unlimited energy

GW of cumulative installed solar power (GW) worldwide from 2019 to 2023 and forecast to 2028

Calling the rise of solar power exponential is not hyperbole, but a fact. Installed solar capacity doubles roughly every three years and increases tenfold every decade. Rarely does one see such sustained growth in anything that matters. Ten years ago, when it was a tenth of its current size, solar power was considered marginal even by experts who knew how fast it had grown. The next tenfold increase will be equivalent to increasing the world’s entire fleet of nuclear reactors eightfold in less time than it normally takes to build just one.

Solar cells will likely be the largest source of electrical power on the planet by the mid-2030s. By the 2040s, they could be the largest source not just of electricity, but of all energy. If current trends continue, the total cost of the electricity they produce promises to be less than half of the cheapest available today. This will not stop climate change, but it could slow it down much more quickly. Much of the world – including Africa, where 600 million people still cannot light their homes – will be able to access abundant, cheap energy. It will be a new and transformative sensation for humanity.

To understand that this is not a delusional dream of environmentalists, consider the “learning curve” of photovoltaic technology. As the cumulative production of a manufactured product increases, costs fall. As costs fall, demand increases. As demand increases, production increases, costs fall again, and so on. That is exactly what has been happening to solar photovoltaics for the past 50 years. The graph illustrates this:

The exponential growth of solar energy: a new era of cheap and unlimited energy

Learning curve of the photovoltaic industry, for the period 1976-2023

Obviously, this cannot last forever: production, demand, or both are always limited. In previous energy transitions – from wood to coal, from coal to oil, or from oil to gas – the efficiency of extraction increased, but was eventually offset by the cost of finding more and more fuel.

Solar energy faces no such limitation, as there is no problem finding more “fuel.” Add to that a few unique factors: the resources needed to produce solar cells and install them in solar farms are silicon-rich sand, sunny locations, and human ingenuity, all of which are abundant. Making the cells also requires energy, but solar energy is making it abundant as well. As for demand, it is huge and elastic: make electricity cheaper, and people will find new uses for it. The result is that, unlike previous energy sources, solar energy has consistently become cheaper and will continue to do so.

  1. Limitations and bottlenecks

Given humans’ propensity to live outside daylight hours, solar energy needs to be supplemented by storage and other technologies. These problems may be solved as batteries and fuels created by electrolysis become cheaper. On the other hand, heavy industry, aviation, and freight transport are difficult to electrify.

Another, no less important problem is that the vast majority of the world’s solar panels, and almost all of the purified silicon they are made from, come from China. Its solar industry is highly competitive, heavily subsidized, and outpacing current demand—quite an achievement when you consider all the solar capacity China is installing within its own borders. This means that Chinese capacity is large enough to sustain expansion for years, even if some of the companies involved go under and some investments dry up.

The Exponential Growth of Solar Power: A New Era of Cheap, Unlimited Energy

The 10 Largest PV Module Manufacturers in the World. They’re All Chinese (Despite the name, Canadian Solar is also a Chinese manufacturer)

In the long run, a world in which more power is generated without oil and gas coming from unstable or politically hostile countries will be more reliable. Still, the fact that a vital industry resides almost entirely in a single country is worrying, a concern that the United States and Europe feel keenly, which is why they have put tariffs on Chinese solar modules.

If anything, cheap and abundant energy will make new uses and activities possible. People who could never afford it will start lighting their homes. Cheap energy can purify water and even desalinate it, it can power the machinery of artificial intelligence, it can make billions of homes and offices more bearable in summers that will become increasingly hot over the next few decades.

The Sun will light up a world in the coming decades where no one needs to go without electricity and where access to energy will empower all those it reaches.

Ignacio Mártil
Professor of Electronics at the Complutense University of Madrid and member of the Royal Spanish Society of Physics