How Solar Panels Work: From Photon to Plug

A solar panel is a slab of glass, silicon and aluminium with no moving parts. It turns roughly a fifth of the sunlight that hits it into electricity. That sounds modest, yet a south-facing 6 kWp array in central Europe delivers about 5,500–7,000 kWh a year, close to what a typical European household uses. Below we follow the energy step by step: how a photon frees an electron, how cells become modules, why you need an inverter, where the losses go, and how to estimate the output of a real system.

Step 1: the photovoltaic effect

Sunlight arrives as photons, small packets of energy. Silicon is the material in more than 95% of panels sold worldwide, according to the Fraunhofer ISE Photovoltaics Report, and it has a band gap of about 1.12 electron-volts. A photon carrying more energy than that can knock an electron out of its bond in the crystal and leave a positively charged “hole” behind. In practice this means light with a wavelength shorter than about 1,100 nm can be used: all visible light plus part of the near infrared.

Left alone, a freed electron would recombine with a hole within microseconds and give its energy back as heat. What makes a solar cell work is the p-n junction. The silicon wafer is doped with a trace of boron in one region (p-type, with spare holes) and a trace of phosphorus in the other (n-type, with spare electrons). A permanent electric field forms where the two meet. Electrons freed near the junction get swept to the n-side and holes to the p-side. Link the two sides with an external wire and the electrons flow through it, doing work on the way. That is the current that charges your phone.

Two physical limits follow:

• Photons with less energy than the band gap pass through or warm the cell. They produce nothing.
• Photons with more energy still free only one electron each, and the surplus turns into heat (“thermalisation”).

Together these set the theoretical ceiling for a single-junction silicon cell at around 29–30% (the Shockley–Queisser analysis, refined for silicon). Our solar panel efficiency guide covers these limits and the technologies that push against them.

Step 2: from cell to module

One silicon cell gives only about 0.6–0.7 volts at open circuit and roughly 0.5–0.6 V at its maximum power point, whatever its size. Current grows with area and light: a modern 182 mm or 210 mm cell delivers around 10–18 A in full sun. To reach a usable voltage, the cells inside a module are wired in series.

Half-cut cells are exactly that: each cell sliced in two. This halves the current in each string of cells, which cuts resistive losses (they scale with the square of current) and makes the module more tolerant of partial shade. Bypass diodes let the current skip a shaded third of the module instead of dragging the whole module down to the output of its weakest cell.

What “Wp” means

Every module is flash-tested under Standard Test Conditions (STC): 1,000 W/m² irradiance, 25 °C cell temperature and a standard solar spectrum (AM1.5). The measured power becomes the nameplate in watts-peak (Wp). Treat it as a laboratory reference, not an everyday output. At a clear summer noon in Europe a well-tilted panel can see 1,000–1,100 W/m², but by then its cells are 30–40 °C hotter than 25 °C. Real output is therefore usually 80–90% of nameplate, even at the best moment of the year.

Step 3: the inverter turns DC into AC

Panels produce direct current at a voltage that shifts with light and temperature. Homes and grids run on alternating current at 230 V/50 Hz (Europe) or 120/240 V/60 Hz (North America). The inverter has two jobs:

1. Maximum power point tracking (MPPT). A PV module’s current-voltage curve has a single “knee” where power (V × I) peaks. That point moves every second as clouds pass and cells warm up, and the inverter keeps adjusting the operating voltage to stay on it.
2. Conversion. Power electronics chop the DC into a synchronised AC sine wave. Modern string inverters reach European weighted efficiencies of 96–98%.

You will meet three common layouts:

• String inverter. Modules wired in series strings (typically 8–20 modules, 300–1,000 V DC) feed one central box. Cheapest per watt, but one shaded module can pull down its whole string.
• Microinverters. A small inverter sits behind each module or pair, so every module runs at its own maximum power point. Better on complex, shaded roofs; more expensive.
• String inverter + DC optimisers. Module-level MPPT with a central inverter, a middle ground.

A hybrid inverter adds a battery port so surplus midday energy can be stored. Use the solar battery calculator to size one.

Step 4: from inverter to plug

The AC output connects to your consumer unit (fuse board). Electricity takes the path of least impedance, so whatever is running in the house uses it first: fridge, heat pump, washing machine. Only the surplus flows out through the meter to the grid. At night, or when demand exceeds solar output, the grid covers the difference. Nothing has to switch; it happens physically and instantly.

This is why self-consumption drives the economics. A kWh you use directly replaces a retail kWh (often €0.20–0.40 in Europe), while an exported kWh may earn only a feed-in tariff or net-metering credit. The net metering calculator shows how export rules change what a system is worth.

Grid-tied inverters must also disconnect within fractions of a second if the grid fails (anti-islanding protection), so line workers are not put at risk. That is why an ordinary solar system goes dark in a power cut unless it has a battery with a backup function.

Where the energy goes: losses from sunlight to socket

Take 1,000 W of sunlight striking one square metre of panel at STC irradiance. A typical rooftop system delivers less than a fifth of that kilowatt to the socket:

The losses after the module are rolled into the performance ratio (PR), the share of the module’s nameplate energy that actually reaches the AC meter. A well-designed modern system reaches a PR of 75–88%. Below 70% usually points to shading, faults or poor design.

Estimating annual output

The quickest reliable estimate needs two inputs: the annual irradiation on the panel plane at your site and the performance ratio.

Worked example: 6 kWp in three cities

Assume a 6 kWp array tilted close to its optimum, with a PR of 0.80. Irradiation values for optimal tilt are approximate, from PVGIS (JRC) and NREL’s PVWatts:

For Bucharest: 6 kWp × 1,550 kWh/m² × 0.80 ≈ 7,440 kWh per year. The same panels produce roughly 75% more in Seville than in Hamburg. Location outweighs every other design choice. For an hour-by-hour simulation with real satellite weather at your address, use our solar panel calculator. To work back from your annual bill to a module count, try how many solar panels do I need. You can also click any spot on the Energy Explorer to see its solar resource.

Orientation, tilt and shading

Output follows irradiance on the panel plane, so geometry matters:

• Azimuth. Facing the equator (south in the northern hemisphere) gives the most energy over a year. East or west orientations typically lose 10–20% at mid-latitudes, but they move production into the mornings or evenings.
• Tilt. At mid-latitudes the annual optimum is usually 30–40°, a little less than the latitude. Being off by ±10–15° costs only a few per cent. Our guide to the best angle for solar panels and the solar tilt angle calculator go into detail.
• Shading. This loss gets underestimated more than any other. Cells are in series, so shading a single row of them can cut a module’s output by a third or more before the bypass diode kicks in. A chimney shadow over two modules in winter can cost more than a 10° tilt error.

What happens over 25+ years

Silicon modules degrade slowly. The causes: light-induced degradation in the first hours of sun (1–3% for some older p-type cells, far less for modern n-type), UV yellowing of encapsulants, micro-cracks from thermal cycling and, in hot humid climates, potential-induced degradation. Jordan and Kurtz (NREL, 2013) analysed nearly 2,000 published degradation rates and reported a median of about 0.5% per year. Premium n-type modules are now warranted at 0.3–0.4% per year.

The inverter is the part most likely to need replacing. Budget for one replacement over the life of a system: string inverters typically last 10–15 years, while microinverters often carry 20–25-year warranties.

Beyond silicon: other cell types

• Thin film (CdTe, CIGS). A few micrometres of absorber on glass. CdTe is common in US utility plants, with module efficiencies of about 19–20%.
• Tandem perovskite-on-silicon. A perovskite top cell absorbs blue light efficiently and silicon handles the red. Laboratory cells have passed 33% on NREL’s Best Research-Cell Efficiency Chart, but commercial durability is still unproven.
• Bifacial modules. Glass-glass modules that also collect light reflected onto their rear side. They add roughly 5–15% over bright ground (white roofs, sand, snow) and much less on dark roofs.

Key takeaways

• The p-n junction in a cell separates light-freed electrons and produces DC. Cells in series make a module, and an inverter converts the DC into grid-quality AC.
• Nameplate (Wp) is a laboratory figure. Real energy depends on the irradiation at your site and a performance ratio of about 0.75–0.88.
• Location sets the ceiling, from under 1,000 kWh/kWp in northern Europe to over 1,700 in deserts. Compare countries on the world energy map or in our country profiles.
• Heat, shading and inverter losses are what you can control. Good design gains more than any upgrade to premium panels.

Figures in this guide are typical ranges drawn from the sources above; your own numbers depend on location, prices and equipment. Use the linked calculators for site-specific estimates.