Solar Energy

Ember (reported by PV Tech, 20 March 2026) counts 647 GW installed in 2025, up 11%, with cumulative capacity near 2.9 TW and likely to pass 3 TW in early 2026. SolarPower Europe (22 June 2026) counts 664 GW, up 12%, and says the global fleet passed 3 TW in early 2026. Solar and wind together added a record 814 GW in 2025 (Ember).

• SolarPower Europe's medium scenario: 612 GW in 2026, down 8%, driven by a 24% fall in China; growth expected to resume in 2027.
• China installed 378-382 GW in 2025 (Ember; SolarPower Europe), with a record 111 GW in May. The China Photovoltaic Industry Association forecasts 180-240 GW for 2026, down from 315 GW in 2025 (as reported by Ember via PV Tech, March 2026).
• SolarPower Europe projects annual installations of about 864 GW by 2030 and a fleet of about 6.6 TW.

SolarPower Europe (22 June 2026) reports India installed 45.7 GW in 2025, up 49%, ahead of the United States. Ember puts India at 49 GWdc, up from 31.9 GW. The EU-27 added 67.2 GW (SolarPower Europe). Solar produced 2,778 TWh in 2025, roughly 9% of global electricity demand.

IRENA's 24/7 renewables analysis (reported by PV Tech, 6 May 2026) puts utility-scale solar and onshore wind near $40/MWh in 2025, less than half the over $100/MWh of new combined-cycle gas. Solar-plus-storage in strong regions ranged from $54 to $82/MWh in 2025, down from over $100/MWh in 2020. Battery system prices reportedly fell about 30% in 2025.

A Section 232 proclamation signed 6 August 2026 takes effect 4 December 2026 (pv magazine India, 10 August 2026). It sets a 15% tariff plus minimum import prices of $21/kg for polysilicon, $100/kg for ingots and wafers, $0.22/W for cells and $0.38/W for modules. Law-firm summaries note combined duties on Chinese-origin products could exceed 65% when stacked with existing tariffs. This is a concrete step toward the trade walls the deck flags.

Ember (via PV Tech, March 2026) estimates solar and wind output since the Middle East conflict began has avoided nearly 330 TWh of gas generation, a potential saving of more than $40 billion.

Not refreshed: the latest spot module price (the $0.10/W figure is kept), the $667/kW install cost and $44/MWh average LCOE (press coverage of IRENA gives conflicting years), the share of Chinese manufacturing, and the best-site $18/MWh cost.

1. 1954Bell Labs: the first practical solar cell
2. 1976$100/W: space tech, priced like jewellery
3. 2010Grid parity: solar matches retail power in sunny spots
4. 2016Auctions tumble: record bids under $30/MWh
5. 2025Cheapest ever: the lowest-cost electricity in history

1. Deploy: subsidies and mandates created early demand.
2. Scale: factories grew; processes standardised.
3. Learn: each doubling of output cut costs ~20%.
4. Undercut: cheaper solar won new markets, then repeat.

1. Photon strikes: sunlight hits a silicon wafer.
2. Electron freed: the photon knocks an electron loose.
3. Current flows: the cell's junction herds electrons one way.
4. Power converts: inverters turn DC into grid-ready AC.

1. Polysilicon: refined silicon, energy-intensive, few players.
2. Cells & wafers: the precision heart, highly concentrated.
3. Modules (panels): assembled, framed, shipped worldwide.
4. Inverters & electronics: the brains converting and managing power.
5. Development & install: site, permits, labour, now most of the cost.

• TOPCon & HJT: next-gen cell designs pushing efficiency past 25%.
• Bifacial: panels harvesting reflected light from both faces.
• Trackers: farms that follow the sun, lifting output ~20%.
• Perovskites: the tandem frontier, lab cells above 33%.
• Agrivoltaics: crops and panels sharing the same land.
• Robot O&M: drones and robots cleaning and inspecting at scale.

Solar's flaw is famous: the sun sets. The fix is arriving on solar's own learning curve, batteries. Storage turns midday abundance into evening power. Its cost curve decides solar's ceiling.

Solved (or soon)

• Evening peak. 4-hour batteries handle it.
• Daily cycling. Charge noon, discharge night.
• Grid services. Batteries stabilise frequency.
• Sunny regions. Firm solar is arriving.

Still hard

• Winter. Weeks of low sun need seasonal storage.
• High latitudes. Less sun, more heating demand.
• Industry heat. Some processes need more than power.
• Land & wires. Storage doesn't site or connect itself.

A two-axis map of representative positions, with cost on the horizontal axis (cheaper to the right) and firmness, meaning always available, on the vertical axis. Nuclear, gas and coal sit in the costly, firm corner. Hydro and solar-plus-storage sit in the cheap, firm 'prize' corner. Wind and solar sit in the cheap, variable corner, with solar cheapest.

• China: installs more than the rest of the world combined, industrial policy at scale
• United States: utility-scale boom in the sunbelt; interconnection queues years long
• India: among the world's cheapest solar, the developing world's template
• Europe: rooftops and policy-driven demand; less sun, high prices make it pay
• The Gulf: record-low auction prices; petrostates hedging with photons
• Africa: off-grid and mini-grid solar leapfrogging the wire entirely

1. Devices (watts)calculators, sensors
2. Rooftopshomes, 5-20 kW
3. Commercialwarehouses, MW-scale
4. Utilityhundreds of MW
5. Gigafarms (gigawatts)desert-scale, GW+

• Midday glut. Solar floods the grid; prices sag toward zero, sometimes below.
• Evening cliff. The sun sets just as demand peaks, the system's hardest hours.
• The response. Batteries charge the glut, discharge the cliff; demand learns to shift.
• The prize. Markets that master this shape get the cheapest power ever offered.

Solar is the first energy source whose price is set in a factory, not a market, and factories only know one direction: down. Oil obeys OPEC; gas obeys winters; solar obeys the learning curve.

• Patient industrial policy. A decade of subsidies, cheap credit and provincial competition.
• Scale begets scale. The biggest factories rode the learning curve fastest.
• Vertical integration. Polysilicon to panel inside one ecosystem.
• Brutal margins. Domestic overcapacity keeps global prices, and rivals' profits, near zero.

• Grid connection: interconnection queues now run 4-7 years in the West.
• Curtailment: midday gluts already force farms to switch off unpaid.
• Land & permits: gigawatt farms need square miles and local consent.
• Negative prices: too much midday solar erodes its own revenue.
• Trade walls: tariffs raise costs and fragment the learning curve.
• Minerals: silver, polysilicon and glass scale, but not for free.

Solar's paradox: the cheaper it gets, the more it cannibalises its own price, because sunshine arrives everywhere at once. Storage, flexible demand and better markets are the answers. All are being built; none is finished.

• Storage everywhere. Batteries co-located as default, not exception.
• Flexible demand. EVs, heating and industry shifting to soak up midday power.
• Bigger, smarter grids. Moving sun from where it shines to where it's needed.
• New market design. Paying for reliability and flexibility, not just energy.

1. Does storage keep collapsing? Sub-$100/kWh batteries change every market.
2. Do grid queues shorten? Permitting reform is the real accelerant.
3. Does 24/7 firm solar scale? Beyond pilots, into standard contracts.
4. How high do trade walls go? Tariffs decide who rides the learning curve.
5. Does demand learn to follow the sun? Flexible load is the quiet revolution.

Fifty years ago solar cost a hundred dollars a watt and powered satellites. Today it costs ten cents and is the cheapest electricity our species has ever produced. The energy transition isn't a hope, it's a price list.

1. The curve already won: solar is the cheapest electricity in history.
2. Factories beat mines: manufactured energy keeps getting cheaper; fuels don't.
3. Storage sets the ceiling: batteries decide how much of the day solar owns.
4. China owns the supply chain: cheap and concentrated, both true at once.
5. The grid is the bottleneck: wires and permits, not panels, gate the future.

• IRENA: Renewable Power Generation Costs 2025 (installs, LCOE, batteries).
• Ember: global electricity and 2025 capacity-addition data.
• IEA: Renewables 2025; solar supply-chain analyses.
• Our World in Data: long-run solar module price series.
• REN21: Global Status Report 2025.
• Industry trackers: module and battery price indices (2024-26).
• Figures marked 'estimate' or 'representative' are compilations by the research team.