When does solar become the largest source of electricity?
The viral chart showing solar overtaking coal in 2033 comes from WEO 2024. The IEA's 2025 outlook adds a scenario in which coal stays number one to 2035.
The question of when solar becomes the largest source of electricity has a widely shared answer — 2033 — and that answer has a specific origin. It is the Stated Policies Scenario of the IEA's World Energy Outlook 2024, charted by Carbon Brief on 18 October 2024. The IEA's next edition, published 13 months later, added a scenario in which it does not happen within the projection period at all.
What each outlook actually said
| Claim | Detail | Источник |
|---|---|---|
| Solar quadruples by 2030 | From 2023 levels; more than nine-fold by 2050 | WEO 2024 (STEPS), via Carbon Brief |
| Crossover dates | Passes nuclear, hydro and wind in 2026; gas in 2031; coal in 2033 | WEO 2024 (STEPS), via Carbon Brief |
| Coal decline | Down 13% by 2030 and 34% by 2035 | WEO 2024 (STEPS), via Carbon Brief |
| Solar capacity in 2050 | Above 16,000 GW, 30% higher than the 2023 edition expected | WEO 2024, via Carbon Brief |
| New scenario added | Current Policies Scenario, based only on enacted laws and measures | WEO 2025, published 12 Nov 2025 |
| Solar additions in CPS | Stall at around 540 GW a year to 2035, roughly 2024 levels | IEA WEO 2025 |
| Coal in CPS | Remains the largest single source of global power generation over the next ten years | IEA WEO 2025, via pv magazine |
| Renewables share in CPS | From one third of global generation in 2024 to just under half by 2035 | IEA WEO 2025, via pv magazine |
| Scenario dropped | The Announced Pledges Scenario is not included in WEO 2025 | IEA WEO 2025 |
| Warming outcome | Almost 3 °C in 2100 under CPS; 2.5 °C under STEPS | IEA WEO 2025 |
Both editions are internally consistent. They differ because they answer different questions. STEPS reads today's policy direction of travel, including measures formally proposed but not yet adopted. CPS counts only what is already enacted and takes a cautious view of how fast new technology gets integrated. The 2033 crossover lives in the first and not in the second.
The variable is the grid, not the panel
The gap between the two scenarios is mostly an integration story. The IEA reports that investment in electricity generation has risen almost 70% since 2015 to around USD 1 trillion a year, while annual grid spending grew at less than half that pace to USD 400 billion. The consequences it lists are congestion, delayed connections, rising curtailment of wind and solar output, more frequent negative wholesale prices and slow permitting for grid projects.
That is the same bottleneck visible in the actual data. Ember found EU solar installations fell 0.7% in 2025 to 65.1 GW, the first decline in a decade, as middays saturated. Battery additions are the partial answer — the IEA counted more than 75 GW added in 2024 — but it cautions that batteries cannot cover seasonal flexibility needs.
One under-reported number from WEO 2025 cuts the other way. The IEA estimates a 2% absolute improvement in average panel efficiency yields over 10% more electricity from the same rated capacity, and that an accelerated repowering schedule could mean up to 15% higher global solar output by 2035 without a single extra hectare. Efficiency and stock turnover are a lever independent of how many gigawatts get installed.
Why the Gulf sits on the favourable side of this fork
One WEO 2025 finding matters more here than the crossover date: by 2035, 80% of growth in energy consumption occurs in regions with high-quality solar irradiation, against roughly half in medium- to low-solar regions over the past decade. The IEA's own conclusion is that this should translate into higher average capacity factors and potentially lower average generation costs than currently registered.
- For a villa or warehouse owner, none of this changes your payback. That is set by consumption, roof area and the tariff you displace, not by a global crossover year. The honest use of these scenarios is directional: nothing in either outlook suggests module costs rise, and both have renewables meeting the largest share of demand growth. Our page on what installed capacity actually costs in Dubai is the number that governs your decision.
- For investors, the scenario spread is the risk disclosure. A model built on the STEPS trajectory and a model built on CPS produce materially different merchant revenue and curtailment assumptions. In the GCC that exposure is currently muted, because capacity is contracted through long-dated PPAs and availability-payment storage agreements rather than sold merchant. Read the contract structure before the scenario.
- For EPC contractors, grid connection is the scarce good. If the binding constraint globally is integration rather than modules, the durable business is in interconnection, protection and storage commissioning, not in bolting down panels faster.
- Treat the crossover date as trivia. Solar reached 10% of global electricity in H1 2026, which is a measured fact rather than a projection. Whether the coal crossover lands in 2033, 2038 or outside the window changes very little about whether a Gulf rooftop is worth building this year.
The IEA is blunt about the status of all of this: "none of the scenarios should be viewed as a forecast", and the scenario analysis is designed to inform decisions rather than predict them. A chart with a crossing point invites the opposite reading, which is why it travels so well.
Источники: Carbon Brief — Solar surge will send coal power tumbling by 2030, IEA data reveals (18 October 2024), IEA — World Energy Outlook 2025, executive summary (12 November 2025), pv magazine — IEA warns global solar growth may plateau at 540 GW annually to 2035 (12 November 2025), Ember — Batteries have unlocked the era of anytime solar (12 August 2026).