China's solar capacity overtakes coal at 1.28 TW

China's solar capacity reached 1.286 TW at the end of July, passing 1.285 TW of coal for the first time and taking 31.5% of installed power capacity.

China's solar capacity has overtaken coal for the first time. The National Energy Administration said on 1 September 2026 that cumulative PV capacity reached 1.286 TW at the end of July, edging past 1.285 TW of coal-fired capacity and taking around 31.5% of the country's 4.078 TW of installed generating capacity. It ends a century in which coal was China's largest power source by capacity.

What China's power mix looks like at the crossover

The gap is one gigawatt out of 4,078. Treat it as a threshold being crossed, not as a lead.

Capacity is not generation, and the difference is the whole story

Solar produced 802.4 TWh in the first seven months of 2026, up 15.5% year on year, equal to roughly 13% of national electricity consumption. Coal-fired plants produced about 2.5 PWh in the first half alone and held 49.7% of total generation. A solar fleet and a coal fleet of the same nameplate size do not deliver the same electricity, because one runs when the sun is up and the other runs when it is told to.

Liu Zhiqiang, deputy director of planning and development at the China Electricity Council, called the crossover a milestone in the country's low-carbon transition while cautioning that solar's intermittency and lower utilisation hours mean coal remains important for system support in the near term. That caution is the operative part of the statement.

How fast this happened

China had 17.45 GW of PV at the end of 2013, roughly 1.4% of a fleet then about 63% coal. The fleet has expanded almost 74-fold since. It passed 130 GW in 2017, 250 GW in 2020, 392 GW in 2022, 609 GW in 2023 and 887 GW in 2024, reaching around 1.2 TW at the end of 2025. Two earlier crossovers preceded this one: total renewable capacity passed coal in 2023, and combined wind and solar passed thermal power in 2025.

The global framing matters for anyone buying modules. IRENA puts world solar capacity at around 2.4 TW at the end of 2025, so China's fleet alone was roughly half of it, and solar accounted for about three-quarters of all new renewable capacity added worldwide that year.

Why the milestone arrives during a contraction

China installed about 71.8 GW of PV in the first half of 2026, down roughly 66% from 212.2 GW a year earlier. Part of that is a base effect — 2025 was inflated by a rush to finish projects before market-based renewable pricing arrived — but grid-connection conditions have tightened and project economics have weakened. Average PV utilisation fell to 91.4% in the first half from 94%, meaning nearly one kilowatt-hour in eleven that the fleet could have produced was not taken.

Manufacturing is contracting alongside it. China Photovoltaic Industry Association data show polysilicon output down 9.8% year on year in the first half and module output down 35.1%. Wang Bohua, former secretary-general of the association, described the sector as in a deep adjustment.

The SOLTECH view

The headline reads like an argument for cheaper panels. The numbers underneath it are an argument for the opposite, and we would rather say so now than have clients discover it at quotation stage.

• For a villa or building owner in the UAE: this does not change your system, your DEWA tariff or your payback. What it does change is the direction of module pricing. Module output in China fell 35.1% in the first half while capacity kept being written off — that is a market clearing its surplus, and prices rise on the way out of a glut, as they already did when TOPCon cell prices jumped 13% in August. If you have a quotation with a price valid for 30 days, the risk is now on the upside. See how system cost breaks down in Dubai.
• For an EPC contractor: the utilisation figure is the one to internalise. A 91.4% national utilisation rate means China is now building transmission, storage and grid-forming capability rather than more panels. Gulf procurement will follow the same path, and the tender language will shift from megawatts to dispatchability. Contractors who can price a solar-plus-storage package credibly will win work that pure-PV contractors will not be shortlisted for.
• For an investor: be careful with the word "overtakes". Coal still supplies half of China's electricity and solar supplies about 13%. Anyone building a thesis on Chinese decarbonisation off the capacity number is using the wrong denominator. The generation crossover is a different event, years away, and it depends on storage and transmission rather than on module shipments. See how we frame project economics.

What we would not claim: that any of this is predictive for the Gulf. China's problem is a saturated grid absorbing too much midday solar. The GCC's problem is the opposite — demand peaks in the afternoon with air conditioning, and there is nowhere near enough solar on the system yet. The lesson to import from China is the sequencing, not the numbers. They built the panels first and the flexibility second, and are now paying for that order. DEWA's phase seven procures both at once, which is the better model.

Sources: pv magazine — China's solar capacity hits 1.28 TW, accounts for 31.5% of total power capacity, 3 September 2026, National Energy Administration of China, IRENA — Renewable capacity statistics.