Solar power

• 1. Introduction

  ›

  In 2025, solar power generated 9% of the world's electricity. A single solar cell can run a calculator, and a roof full of panels can power a remote home off the grid. The same physics scales up to gigawatt-sized photovoltaic power stations that feed national networks. Three-quarters of all new generation capacity being added is now solar. The cost of panels keeps falling, and grid-connected solar capacity has roughly doubled every three years. What is the trick that turns sunlight straight into an electric current? Who first noticed that light could do this work, and why did the idea sit dormant for decades? How did a technology once reserved for satellites and a handful of homes become the cheapest electricity for new installations in most countries? And what stands in the way of letting the sun carry even more of the load?

• 2. Turning Light Into Current

  ›

  The photovoltaic effect is the heart of the matter: inside a solar cell, light is converted directly into electric current. Charles Fritts built the first solar cell in the 1880s, and the German industrialist Ernst Werner von Siemens was among those who saw how important the discovery was. In 1931, the German engineer Bruno Lange built a photo cell using silver selenide in place of copper oxide. Those early selenium prototypes converted less than 1% of the light striking them into electricity. Following the work of Russell Ohl in the 1940s, researchers Gerald Pearson, Calvin Fuller and Daryl Chapin created the silicon solar cell in 1954. These first silicon cells cost US$286 per watt and reached efficiencies of 4.5-6%. In 1957, Mohamed M. Atalla developed silicon surface passivation by thermal oxidation at Bell Labs, a process that has since been critical to how efficient cells can be. Today over 90% of the market is crystalline silicon. A photovoltaic array produces direct current that rises and falls with the intensity of the sunlight. For most uses that power must be turned into alternating current through inverters. Multiple cells are joined inside panels, panels are wired into arrays, and the arrays feed an inverter that delivers power at the desired voltage and, for AC, the desired frequency. Not every cell is rigid silicon. A thin-film solar cell, a second-generation design, is made by depositing one or more thin layers of photovoltaic material on a substrate such as glass, plastic or metal. Commercial thin-film technologies include cadmium telluride, copper indium gallium diselenide, and amorphous silicon.

• 3. Two Ways To Harvest

  ›

  Concentrated solar power takes a different route to electricity. Instead of converting light directly, it uses lenses or mirrors and tracking systems to focus a large area of sunlight onto a hot spot, then uses that heat to drive a conventional steam turbine. The approach is also called concentrated solar thermal. Its great strength is heat storage. A CSP plant can hold thermal energy and dispatch electricity over as long as a 24-hour period, and because peak demand often arrives around 5 pm, many plants carry 3 to 5 hours of thermal storage. Yet the economics have kept it small. The levelized cost of electricity from CSP is over twice that of PV, and less than 1% of solar power comes from it. Hybrid systems blend solar with other resources. Hydro, wind and batteries are commonly paired with solar so the combined output can track demand or at least smooth out the swings. Adding panels on or around existing hydro reservoirs is especially useful, because hydro is usually more flexible than wind and cheaper at scale than batteries, and the existing power lines can sometimes carry the new generation.

• 4. From Coal Scare To Satellites

  ›

  The early push for solar in the 1860s grew from a fear that coal would soon run short, seen in the experiments of Augustin Mouchot. In 1884, Charles Fritts installed the world's first rooftop photovoltaic array, using 1%-efficient selenium cells, on a New York City roof. Then progress stalled, as cheap and abundant coal and petroleum made solar look unnecessary in the early 20th century. Bell Telephone Laboratories revived the field in the 1950s with silicon wafers coated with boron. Its Bell Solar Battery was described as 6% efficient, and a square yard of panels could generate 50 watts. The first satellite carrying solar panels launched in 1957. By the 1970s panels were still too costly for much beyond spacecraft. In 1974 it was estimated that only six private homes in all of North America were entirely heated or cooled by working solar systems. The 1973 oil embargo and the 1979 energy crisis changed the climate of opinion and brought fresh attention to the technology. Governments built research institutions, including SERI, now NREL, in the United States, NEDO in Japan, and Fraunhofer ISE in Germany. President Jimmy Carter set a target of drawing 20% of U.S. energy from solar by the year 2000, but his successor, Ronald Reagan, removed the funding for renewable research. Falling oil prices in the early 1980s slowed the growth of photovoltaics from 1984 to 1996.

• 5. The Long Climb To Cheap

  ›

  In the early 2000s, feed-in tariffs reshaped the industry. The policy gave renewables priority on the grid and fixed a price for the electricity they generated, which created investment security and a surge of PV deployment across Europe. For several years European demand drove worldwide growth, then the center of gravity moved to Asia, especially China and Japan. Chinese manufacturers grew to be the largest makers of solar equipment. The price collapse that followed was dramatic. The cost of utility-scale solar PV fell by 85% between 2010 and 2020, while CSP costs fell 68% over the same period. In 2022 the world's solar generation capacity passed 1 terawatt for the first time. Even as the 2021-2022 global energy crisis drove up the price of materials like polysilicon, utility-scale solar remained the least expensive energy source in many countries, because rival fuels such as natural gas grew more costly. Africa is now the world's fastest-growing solar market, aided mostly by China. China today holds about half the world's solar power, and almost half of all solar installed in 2022 sat on rooftops. Solar is forecast to become the largest source of renewable power before the end of the 2020s, surpassing hydropower.

• 6. What Solar Costs

  ›

  Photovoltaic systems burn no fuel, and their modules typically last 25 to 40 years. That changes the shape of the bill: upfront capital and financing make up 80% to 90% of the cost of solar power. This reliance on long-term contracts is a problem in countries where agreements may not be honored, such as some African nations. The fall in module prices has been steep. In 1982 a kilowatt cost roughly 27,000 American dollars, and by 2006 that had dropped to about 4,000 dollars per kilowatt. A full PV system cost around 16,000 dollars per kilowatt in 1992 and about 6,000 dollars per kilowatt in 2008. In 2025 in the US, residential solar runs around 2.50 dollars per watt, though solar shingles cost much more, while utility solar costs are around 25 US cents per watt. Where solar pays best depends on geography. Productivity follows solar irradiance, which shifts through the day and year and depends on latitude and climate. The richest sunlight lies in the arid tropics and subtropics, where low-latitude deserts have few clouds and receive more than ten hours of sun a day. These deserts form a Global Sun Belt circling the planet, stretching across Northern Africa, Southern Africa, Southwest Asia, the Middle East, and Australia, along with smaller deserts in the Americas. Modelling by Exeter University suggests that by 2030 solar will be the least expensive option everywhere except some Nordic countries.

• 7. Catching The Surplus

  ›

  Solar energy is not available at night, so storing it is the central problem for continuous power, especially off-grid and in any future 100% renewable scenario. Solar is intermittent because of the day-night cycle and changing weather, though it can be forecast somewhat by time of day, location and season. In places with hot summers and mild winters, solar matches daytime cooling demand well. Concentrated solar plants store heat directly, often in high-temperature molten salts. Those salts work well because they are low-cost, hold a great deal of heat per unit, and deliver it at temperatures conventional power systems can use. Photovoltaic systems lean more on chemistry and on the grid itself. In stand-alone systems batteries hold the excess, while grid-connected systems can send surplus electricity into the network and earn a credit through net metering or feed-in tariffs. As battery prices fall, PV systems increasingly store a daytime surplus for use at night. Most home batteries are lithium-ion, mainly lithium iron phosphate since around 2021, with some nickel manganese cobalt cells. Falling prices are expected from large plants such as the Tesla Gigafactory 1. The batteries of plug-in electric cars could become storage too, since most vehicles sit parked an average of 95% of the time, letting power flow from the car to the lines and back. The stakes of storage are rising fast. In Europe, too little storage and transmission could waste about 40 terawatt-hours of solar electricity through curtailment in 2026, enough to power Greater London for a year, a 25% increase over 2025.

• 8. Clean Power And Its Politics

  ›

  The life-cycle greenhouse-gas emissions of solar farms are less than 50 grams per kilowatt-hour, and with battery storage could reach up to 150 grams. By comparison, a combined-cycle gas plant without carbon capture emits around 500 grams per kilowatt-hour, and a coal plant about 1,000. Solar carries an upfront environmental cost in manufacturing, with a carbon payback of several years, then delivers clean energy across a roughly 30-year lifetime. The land question is real. The lifecycle surface power density of solar averages about 7 watts per square metre, against about 240 for nuclear and 480 for gas. A 2021 study found that drawing 25% to 80% of electricity from domestic solar farms by 2050 would mean covering 0.5% to 2.8% of the European Union, 0.3% to 1.4% of India, and 1.2% to 5.2% of Japan and South Korea. Materials raise their own concerns. A 2021 International Energy Agency study projects copper demand will double by 2040, and more tellurium and indium may be needed. Then there is the geopolitics of supply. Over 40% of global polysilicon manufacturing capacity sits in Xinjiang in China, which raises concerns about human rights violations. The International Solar Energy Society argues China's dominance is not a problem, both because it estimates solar manufacturing cannot grow beyond 400 billion USD per year, and because other countries would have years to build their own industry if Chinese supply were cut off. There is a deeper political argument in solar's favor. Once installed, solar generation cannot be switched off by geopolitics the way oil and gas can, which is why it strengthens energy security and why some libertarians value it for reducing reliance on government and on fragile electricity grids.