Concentrated Solar
Mirrors focus sunlight to melt salt, and the molten salt holds the heat all night. It is solar that works after dark, and photovoltaics got so cheap it may have been overtaken anyway.
The effect compounds within years. Put it in place and it keeps working.
Origins
Archimedes is said to have burned the Roman fleet at Syracuse with mirrors. Almost certainly he did not, but the idea is very old and very obvious: sunlight, gathered and focused, gets hot.
Auguste Mouchout built a solar steam engine in France in the 1860s, alarmed that coal would run out. In 1913, an American engineer named Frank Shuman built a parabolic-trough solar plant outside Cairo that pumped water from the Nile for irrigation. It worked. It was reported as the future. And then the First World War arrived, cheap oil was discovered in the Middle East, and the whole idea was abandoned for seventy years.
The modern revival began in the Mojave in the 1980s and reached its peak of ambition in the 2000s and 2010s: enormous fields of mirrors in the deserts of Spain, Morocco, California and the Gulf, focusing sunlight onto a tower where salt melts and glows.
And then, awkwardly, photovoltaics fell 90% in price and ate its lunch.
Concentrated solar is now in the strange position of being an elegant, powerful technology that has been substantially outcompeted by a simpler one — and yet it retains a single property that nothing else in solar has, and that property may yet save it: it can store its energy as heat, and heat is cheap to store.
What it actually is
Concentrated solar power (CSP) uses mirrors to focus sunlight onto a receiver, heating a fluid — usually molten salt — to several hundred degrees. That heat boils water, drives a steam turbine, and generates electricity. It is a thermal power station whose fuel is sunlight.
The consequence of being thermal is the whole point: you can store the heat.
Molten salt in a well-insulated tank holds its temperature for many hours. A CSP plant can collect heat all day and generate electricity into the night, or on demand, without a battery. This is dispatchable solar, and it is a genuinely different product from photovoltaics.
Thermal storage is also, per unit of energy stored, dramatically cheaper than electrochemical storage. A tank of hot salt is a great deal less complicated than a lithium battery.
The problem is that photovoltaics became so absurdly cheap that PV-plus-battery now beats CSP on cost in most applications, even accounting for storage. CSP is more complex, requires more water for cooling, needs very high direct sunlight (it does not work in haze or diffuse light, where PV still functions), and has a much smaller manufacturing base and therefore a much weaker learning curve.
It survives where its unique property matters most: long-duration heat storage, and industrial process heat, which is a large, hard-to-decarbonise sector that batteries cannot serve.
The numbers
The storage advantage. CSP with molten salt storage can dispatch electricity for many hours after sunset without any battery. Thermal storage is substantially cheaper per unit of stored energy than electrochemical storage, which is CSP’s one durable structural advantage.
The cost problem. CSP costs have fallen — IRENA recorded a sharp decline of around 46% in one recent year — but they started far higher and the manufacturing base is far smaller. Photovoltaics fell about 90% between 2010 and 2024, driven by a manufacturing scale CSP will never match, because CSP is a construction project and PV is a product.
The resource constraint. CSP requires strong, direct sunlight. It performs poorly in haze, humidity or diffuse light, where PV continues to work. This confines it to a relatively narrow band of the world’s deserts.
The water problem. Steam turbines need cooling, and CSP’s best sites are deserts, which is precisely where water is scarcest. Dry cooling is possible and reduces efficiency.
The surviving case: process heat. Industry needs high-temperature heat — for cement, chemicals, steel — and this is one of the hardest sectors to decarbonise. Concentrated solar makes heat directly, without converting to electricity and back. That may end up being its real contribution and it is barely discussed.
Why it matters
There is something magnificent about a solar tower that is worth saying out loud, even if the accountants have moved on.
You stand in a desert and there are thousands of mirrors, each tracking the sun, and they all point at one place, and at that place the air shimmers and the tower glows white, and it is so bright you cannot look at it. It is one of the few pieces of energy infrastructure that produces genuine awe rather than mere acquiescence. Coal plants do not do that. Gas turbines do not do that.
And what it is doing, at that focal point, is melting salt so that a city can have electricity at midnight from sunlight that fell in the afternoon. There is something almost alchemical in the arrangement.
We include it, honestly, partly because it may lose. Photovoltaics is cheaper and simpler and it is winning, and pretending otherwise would be exactly the kind of wishful thinking we criticise elsewhere. That is how technology works and there is no shame in being outcompeted by something better.
But the deserts are still there, and the sun still falls on them at an intensity that could power a continent, and the ability to hold that heat until nightfall is a property that nothing else in solar possesses. Whether it is needed depends on how well batteries do. That is an honest uncertainty, and we would rather sit in it than pretend to know.
What it actually takes
Finding the niche where it actually wins. Competing with PV-plus-battery on ordinary electricity is a losing game and CSP should stop trying. Where it wins is long-duration thermal storage and industrial process heat, which is a huge, hard-to-abate sector that batteries cannot serve at all.
Water, in the desert. Steam cycles need cooling and the best sites are the driest places on Earth. Dry cooling works and costs efficiency, and this trade-off is unavoidable.
Scale, which it may never reach. PV benefits from a global manufacturing learning curve. CSP is a construction project, and construction projects do not get cheaper with volume in the way products do. This is the same structural disadvantage that afflicts nuclear, and it is the reason both have struggled against solar panels.
Siting, honestly. Deserts look empty and are not. Desert ecosystems, cryptobiotic soil crusts and desert tortoises take centuries to recover from a bulldozer, and the largest CSP projects have had genuine ecological conflicts.
Where it matters most
The Mojave is where modern CSP was born and where its ecological conflicts have been sharpest.
Morocco and the Sahara hold the Noor complex, among the largest CSP installations on Earth, and the best combination of resource and proximity to demand anywhere.
Spain pioneered commercial CSP and still holds much of the world’s installed capacity, in the Iberian interior.
The Gulf, Chile’s Atacama, and northwest China have the strongest direct sunlight in the world — the Atacama in particular is the sunniest place on the planet.
And anywhere with heavy industry. The process heat case is geographically about factories, not deserts, and it is where the technology’s future may actually lie.
How to tell it’s being done well
Is it competing on the right thing? CSP loses to PV plus battery on ordinary electricity. It wins on long-duration heat and industrial process heat. A project competing on the former is a project with a hard road.
What is the cooling? Wet cooling in a desert is a real conflict. Dry cooling costs efficiency and is usually the right choice.
What was the land? Desert is not empty. Cryptobiotic crusts and desert species recover from disturbance over centuries, not years.
Is the storage actually used? A CSP plant without thermal storage has thrown away its only structural advantage over a solar panel.
What you can do
Anyone
- Concentrated solar is the one form of solar that can generate at midnight, without a battery, by storing heat in molten salt.
- It is also being outcompeted by ordinary solar panels, which is how technology works and is nothing to mourn.
Policymakers
- Support CSP for industrial process heat, not for ordinary electricity. Process heat is a huge hard-to-abate sector that batteries cannot serve, and it is where this technology genuinely wins.
- Desert siting requires the same ecological scrutiny as any other habitat. Desert is not empty land.
Business and investors
- Thermal storage is far cheaper per unit of energy than batteries. Where the end use is heat rather than electricity, that advantage is structural and it is not going away.
- Industrial heat decarbonisation is a large, underserved market and concentrated solar is one of very few options for it.
Who is working on this
We are researching which organizations in our directory of 8,493 actively work on this solution, and we only list an organization once we have verified it. That research is ongoing. In the meantime, search the directory yourself:
Questions
How is concentrated solar different from solar panels?
Solar panels convert sunlight directly into electricity. Concentrated solar uses mirrors to focus sunlight and produce heat, usually melting salt to several hundred degrees, then uses that heat to drive a steam turbine. It is a thermal power station whose fuel is sunlight.
Can concentrated solar generate at night?
Yes, and this is its defining advantage. Molten salt in an insulated tank holds its heat for many hours, so a plant can collect heat all day and generate electricity into the night without any battery. Thermal storage is also far cheaper per unit of stored energy than electrochemical storage.
So why isn't it everywhere?
Because photovoltaics got about 90% cheaper between 2010 and 2024 and comprehensively outcompeted it. PV is a manufactured product with a steep learning curve; CSP is a construction project, and construction projects do not get cheaper with volume. PV plus a battery now beats CSP on cost for most applications.
Does it have any future?
Probably, in a niche it is well suited to and largely ignoring: industrial process heat. Cement, chemicals and steel need high-temperature heat, it is one of the hardest sectors to decarbonise, and batteries cannot serve it at all. Concentrated solar makes heat directly, which may turn out to be its real contribution.
What are the drawbacks?
It needs strong direct sunlight and performs poorly in haze or diffuse light where panels still work. It needs water for cooling, and its best sites are deserts, where water is scarcest. And desert siting has real ecological costs; desert soils and species take centuries to recover from disturbance.
Is it beautiful?
Yes, and we are including that in the assessment. A solar tower with thousands of mirrors tracking the sun and a focal point too bright to look at is one of the few pieces of energy infrastructure that produces genuine awe. That is not an argument. But it is not nothing either.
Sources
- Project Drawdown - Deploy Concentrated Solar (Drawdown Explorer) Framework and classification. Cited, not reproduced.
- IRENA (2025) - Renewable Power Generation Costs in 2024
- IEA - Concentrating Solar Power
- IPCC (2022), AR6 Working Group III - Energy Systems
- SolarPACES - international CSP research collaboration
The solution taxonomy follows the framework popularised by Project Drawdown. The analysis above is our own; for their carbon modeling and rankings, visit them directly.