The history of solar power is a slow build with one explosive ending. The physics was figured out in the 1830s. The first practical device came out of Bell Labs in 1954, and by 1956 a watt of it cost $286, useful only because a satellite needed power and nothing else worked. Between 1976 and 2024 module prices fell about 375 times, and global capacity grew from kilowatts to nearly 3 TW. Here's the timeline, with the numbers that mattered at each step. For where all that capacity ended up, see seven real solar applications.
Solar wasn't reinvented to conquer the world. The same silicon junction got about 375 times cheaper between 1976 and 2024.
When Was the Photovoltaic Effect Actually Discovered?
The photovoltaic effect was discovered in 1839 by 19-year-old French physicist Edmond Becquerel. Working in his father's lab, he placed silver chloride in an acidic solution, connected it to platinum electrodes, lit it, and measured a voltage and current. It was the first photovoltaic cell, and it was useless for power. The discovery then sat as a curiosity for decades, since the semiconductor physics to explain it didn't exist yet.
Several intermediate steps mattered for getting from Becquerel to practical cells. Willoughby Smith discovered photoconductivity in selenium in 1873. Charles Fritts coated selenium with a thin layer of gold in 1883 and got about 1% efficiency. Einstein published his photoelectric paper in 1905 and won the 1921 Nobel Prize for it. Jan Czochralski invented his crystal-pulling method in 1916 (published in 1918), the route later used to grow the monocrystalline silicon in most of today's cells. And in 1940, Bell Labs researcher Russell Ohl found a cracked silicon sample that generated current under light, a p-n junction discovered by accident, though it managed only about 1% efficiency.
How Did Bell Labs Build the First Solar Cell?
The modern silicon solar cell came from three Bell Labs researchers working on different problems. Daryl Chapin needed a power source for telephone equipment in the tropics, where dry-cell batteries degraded too fast, and the selenium cells then on the market converted less than 0.5% of sunlight (Perlin). Word of Chapin's problems reached physicist Gerald Pearson, and the group moved to silicon. Chemist Calvin Fuller then made it work. His first cells used lithium, which migrated through the silicon at room temperature and moved the junction away from the surface. The fix was arsenic-doped silicon with a thin layer of boron on top, which put the p-n junction right under the surface where the light is absorbed.
Early cells hit about 4%, and the best ones reached nearly 6% in early 1954, up from selenium's under 0.5%. Bell announced the invention on April 25, 1954, running a toy Ferris wheel and a radio transmitter on sunlight.
The reality was less dramatic. A watt of Bell solar cells cost $286 in 1956, and Chapin worked out that an array big enough for an average house would cost about $1.43 million. Economically useless on the ground. So what saved them? Satellites. As researcher Martin Wolf put it, the Space Age was "the salvation of the solar cell industry" (ASES).
What Made Vanguard 1 Important?
Vanguard 1, launched March 17, 1958, was the first satellite to use solar cells. It carried two transmitters: a 10 mW unit on mercury batteries and a 5 mW beacon fed by six solar cells. The battery transmitter went silent in 1958. The solar beacon kept going for more than six years, with the last signals received in May 1964. Earlier satellites on batteries alone had stopped transmitting within weeks.
Vanguard 1 is still in orbit, the oldest artificial object up there, with an expected lifetime now put at about 240 years. Solar has stayed a space staple since; Juno, in 2016, became the first solar-powered spacecraft at Jupiter's distance, where all eight earlier probes had relied on nuclear power. For how space-rated solar evolved differently from terrestrial PV, see our piece on using solar panels in space.
How Did the 1970s Oil Crisis Change Things?
The 1973 Arab oil embargo sparked the first major government push for terrestrial photovoltaics. Congress passed five energy bills in 1974 and the Solar Energy Research Institute, SERI, the lab now known as NREL, was created in 1977. President Carter installed 32 solar thermal panels on the White House roof in June 1979 to heat water; the Reagan administration removed them in 1986 during a roof resurfacing, judging them not cost-effective.
The 1978 Energy Tax Act created the first US investment tax credit for solar, 10% for business property, with residential credits worth up to $2,500. The Solar Photovoltaic Energy Act the same year committed $1.2 billion over ten years to PV research and demonstration. The credit was expanded to 15% in 1980, and the 10% solar credit was made permanent in 1992 (CRS), while the Institute for Energy Research puts the 1978 homeowner credit at a maximum of $2,500. The residential credit's story has ended differently: under Public Law 119-21, the 25D credit isn't allowed for expenditures made after December 31, 2025. See our 2026 solar tax credits piece for the current landscape.
The money produced real hardware. A 1980 house in Carlisle got the first building-integrated PV system, thin-film cells passed 10% efficiency that year, and by 1985 a 6 MW plant at Carissa Plains was feeding Southern California Edison (EIA). Still, commercial deployment moved slowly. Module prices fell from about $100 per watt in 1976 to $12 in 1990 (constant 2025 dollars, Our World in Data), real progress, but far above grid parity.
How Did Solar Get Cheap?
Our World in Data's series, built from Nemet, Farmer and Lafond, and IRENA's module price index, shows prices falling from $99.60 per watt in 1976 to $4.75 in 2008 and $0.27 in 2024, all in constant 2025 dollars. The last 16 years alone were an 18-fold drop. Three forces compounded. Scale. Cumulative capacity passed 1 TW in 2022, a milestone that took more than 40 years, and then climbed to nearly 3 TW in less than three more (IEA PVPS). Efficiency. SunPower's E20 panel reached 20.1% in 2012, the best silicon cell now stands at 27.81%, and Longi's founder claims each point of cell efficiency cuts system cost by over 5%. Chinese manufacturing. An IEA report in 2022 found China holding over 80% of all manufacturing stages critical to solar panels, including 79% of polysilicon capacity, 97% of wafers, and 85% of cells in 2021. Overcapacity then pushed Chinese module prices down by more than 60% from early 2023 levels, and Western makers felt it: SunPower itself went bankrupt in August 2024.
Cheap modules don't mean cheap rooftops. Berkeley Lab's 2025 data put the median US residential system installed in 2024 at $3.5/W for cash buyers and $4.7/W for loan-financed ones. A $0.27/W module is roughly a tenth of that bill or less; the rest is permitting, interconnection, sales overhead, and labor, which is why IEA PVPS says deployment bottlenecks have shifted from module cost toward permitting and grid connection.
When Did Solar Become a Real Power Source?
In 2003, solar historian John Perlin noted that about a billion watts of PV modules were in use worldwide. Twenty-two years later the number is 2,973 GW. IEA PVPS counts 2,276 GW at the end of 2024, a record 698 GW added in 2025, and 39 countries that each installed at least 1 GW. China holds 1,464 GW, almost half the world total; the EU has 406 GW, the US 268 GW, India 180 GW, and Germany 117 GW. Ember's data show solar generated 2,778 TWh in 2025, up 30% from 2,143 TWh in 2024, enough to meet 75% of the rise in global electricity demand.
| Year | Milestone | Price/scale marker |
|---|---|---|
| 1839 | Becquerel discovers the photovoltaic effect | N/A |
| 1954 | Bell Labs builds first practical silicon cell | About 4-6% efficiency; $286/W by 1956 |
| 1958 | Vanguard 1, first solar-powered satellite | Six cells; 5 mW beacon ran 6+ years |
| 1976 | Early terrestrial era | About $100/W (2025 dollars) |
| 2008 | Pre-Chinese-scale module pricing | About $4.75/W (2025 dollars) |
| 2022 | Global capacity passes 1 TW | More than 40 years to reach |
| 2024 | Module price and capacity | About $0.27/W; 2,276 GW installed |
| 2025 | Record year | 698 GW added; 2,973 GW cumulative |
Policy shaped the road. Germany's Renewable Energy Sources Act of 2000 lifted the feed-in tariff to about EUR 0.51 per kWh, and the PV market there boomed from 2004 on. By 2020 the IEA could say that, for projects with cheap financing and good sunlight, solar was the cheapest electricity in history. The shift from policy-driven deployment (feed-in tariffs, net metering, mandates) to economic deployment happened gradually through the late 2010s. For which countries lead by capacity, see our piece on where solar energy is used.
Summary
Solar started as an 1839 observation by Edmond Becquerel and took 115 years to become a practical device. Bell Labs built the first working silicon cell in 1954 at roughly 4-6% efficiency, and satellites used it before homes did. The 1970s oil crisis funded the first major government push, including the 1978 Energy Tax Act. The price collapse came from scale, better cells, and Chinese manufacturing: about $100/W in 1976, $4.75/W in 2008, and $0.27/W in 2024 in today's dollars. Global installed capacity reached 2,973 GW by the end of 2025. For good sites with cheap financing, the IEA called solar the cheapest electricity in history back in 2020, and the limits now are storage, grid capacity, and policy rather than physics, and the storage half of that equation is what home batteries like the Tesla Powerwall 3 address at the household level.