If you have been following solar project tender results lately, one set of numbers stands out: Shanghai’s 2025 incremental-project mechanism tariff is RMB 0.4155/kWh, while Gansu’s two rounds of competitive bidding cleared at RMB 0.1954/kWh. The former is more than double the latter.

Same sky, same sunlight — so why is solar power worth twice as much on one side as on the other?

The key lies in two pivotal policies from 2025 to 2026: “Document No. 136,” issued on 27 January 2025, and the “15th Five-Year Plan for Renewable Energy Development,” issued on 23 July 2026. Together they have rewritten the underlying logic of the industry.

01 Document No. 136: the watershed for renewable energy tariffs

On 27 January 2025, the National Development and Reform Commission (NDRC) and the National Energy Administration (NEA) jointly issued the Notice on Deepening the Market-Based Reform of On-Grid Tariffs for Renewable Energy to Promote High-Quality Development — known across the industry as Document No. 136. Its core objective was singular: to move renewable energy tariffs from being set by policy to being set by the market.

Three things were done.

First, wind and solar on-grid volumes are in principle to enter the electricity market in full, with tariffs formed through market trading. The “fixed tariff plus full-volume purchase” model that renewable projects have relied on for more than a decade has been torn apart.

Second, a “renewable energy sustainable development price settlement mechanism” was created as a safety cushion. For volumes included in the mechanism, compensation is paid when the market average price falls below the mechanism tariff, and the difference is returned when it rises above it — effectively an insurance floor for each project.

Third, existing and incremental capacity are treated separately. 1 June 2025 is the dividing line: capacity commissioned before it is “existing,” and capacity commissioned after it is “incremental.”

By July 2026, all 31 provincial-level regions across China had completed their mechanism tariff auctions for 2025–2026. In practice, the NEA has disclosed that the national average comprehensive on-grid tariff for renewable energy in the first half of 2026 (including mechanism settlements) was around RMB 0.3/kWh — roughly 80% of the average benchmark coal-fired tariff.

Stability is the appearance; divergence is the reality.

02 Divergence in local auctions: one sky, two fates

At its core, the gap between Shanghai’s RMB 0.4155/kWh and Gansu’s RMB 0.1954/kWh reflects differences in three factors: load, absorption capacity and outbound transmission. Installed capacity is only the entry ticket; whether the power can be sold, and at what price, is the real dividing line.

What is putting distributed solar on edge is that this safety cushion is being withdrawn quickly. Shandong has made clear that source-grid-load-storage and green power direct-connection projects, general commercial and industrial (C&I) solar, large C&I distributed solar, and household non-natural-person solar commissioned after 1 June 2026 will no longer fall within the mechanism tariff. Henan, Guangxi, Jiangxi, Anhui and other provincial regions have also issued documents excluding newly built C&I distributed solar from mechanism tariffs.

Data disclosed by Wang Shujuan, founder of Zhihui Solar, is even more sobering: among the roughly 251,200 clearing projects nationwide in the 2025–2026 mechanism tariff auctions, distributed solar accounted for nearly 250,000 of them — yet just about 9% of total cleared capacity and only 7% of electricity volume. Distributed solar is at a structural disadvantage in these auctions.

BNEF estimates that winning solar capacity in 2025–2026 totalled about 101 GW, only roughly 20% of all solar project capacity. The remaining 80% of solar projects must absorb market price volatility directly.

03 A cliff in installations: the truth behind the 66% decline

According to NEA data, China added 71.77 GW of grid-connected solar in the first half of 2026, down about 66% year on year, of which distributed solar accounted for 42.22 GW.

At least two of the reasons trace back to Document No. 136.

The first is the removal of mandatory storage co-location. The document explicitly states that configuring energy storage may not be a precondition for the approval (filing), grid connection or on-grid delivery of new renewable energy projects. This marks the end of the “policy-mandated storage” model that drove the storage industry’s rapid growth in recent years. Liu Yong, Secretary-General of the Energy Storage Application Branch of the China Chemical and Physical Power Industry Association, said in an interview: “For solar projects, there is little incentive to add storage. Originally scheduled to break ground immediately, most storage projects will now delay construction because mandatory co-location has been scrapped — while the market waits to see whether supportive policies for new-type storage emerge.”

The second is a marked increase in revenue uncertainty for renewable projects. Solar output is concentrated at midday, when power supply rises sharply and prices fall noticeably, while there is almost no generation during the high-priced evening peak. Once fully exposed to the market, actual settlement prices are pushed down by real-time market curves. BNEF estimates that solar’s actual market price averages RMB 0.127/kWh below the mechanism tariff.

For the storage industry, the shift from mandatory co-location to on-demand configuration is an unavoidable path, though short-term pain is hard to avoid. Several storage companies have expressed similar views to Jiemian News: large-scale storage projects may take a hit in the short term, but over the longer run policy favours the development of C&I storage and behind-the-meter storage. Moving from policy-driven to market-driven, and from price-driven to value-driven, are the keywords of this transition.

04 The 15th Five-Year Plan: a 3,500 GW target and a new landscape

If Document No. 136 is the watershed, the 15th Five-Year Plan for Renewable Energy Development, issued on 23 July 2026, is the roadmap.

The plan sets out that by 2030 China’s total installed renewable energy capacity will reach about 3,500 GW, with combined wind and solar capacity exceeding 2,800 GW. For distributed solar, new distributed renewable capacity during the 15th Five-Year Plan period will exceed 300 GW. Renewable energy must progressively move from being the main source of installed capacity to being a main power source responsible for electricity supply.

The plan also lays out market-based pathways: large-scale bases participating in the electricity market as a whole (with integrated bundled trading for desert, Gobi and wasteland wind and solar bases), support for joint bidding by renewable projects (using intraday wind-solar complementarity to smooth output fluctuations), and the normalisation of multi-year green power trading. Multi-year green power agreements reached 60 billion kWh in 2025, with Greater Bay Area users consuming Inner Mongolian green power and Yangtze River Delta users sourcing Guangxi green power for the first time.

In its interpretation published on 30 July 2026, the NEA summed up the 15th Five-Year Plan period as “the first five-year period of renewable energy’s full entry into the market.” Document No. 136 in 2025–2026 is only the prologue; the real market-based test is still ahead.

05 A rebuilt investment logic: from betting on policy to betting on absorption

Before Document No. 136, the investment logic for a solar project was simple: find a roof, sign a contract, lay the modules, collect the electricity fee. Fixed tariffs, calculable returns, controllable risk.

After Document No. 136, that arithmetic is void. Investment logic is switching from “betting on policy” to “betting on absorption.” The first layer is securing mechanism tariffs: with only about 20% of capacity winning them, landing a mechanism tariff means locking in baseline returns. The second layer is securing green power demand: the mechanism tariff is only a floor — what really determines the profit ceiling is the green premium, and multi-year green power agreements, PPAs and direct green power connection are all tools that turn “selling electricity” into “selling green value.” The third layer is securing absorption channels: outbound transmission corridors, local load, storage peak-shaving and virtual power plant aggregation are no longer optional but mandatory.

The revenue model has changed completely too. In the past: revenue = generation × fixed tariff. Now: revenue = mechanism volume × mechanism tariff + market volume × spot average price + green power volume × green premium − storage and aggregation costs. Generation is only one variable; tariff volatility, green premiums and absorption costs can each become a black hole for profit.

06 A final word

China’s solar industry is not dead — but the rules of the game have changed completely.

From “finding rooftops” to “finding buyers,” from “betting on grid connection” to “betting on tariffs,” from “counting electricity fees” to “counting total returns.” Every link in the chain is being reshuffled.

The two pivotal documents of 2025–2026 — Document No. 136 and the 15th Five-Year Plan — together mark out the starting line for a new phase. Over the next five years, whoever survives marketisation, secures long-term buyers for green power demand and gets absorption channels working will be the winner.

All that is left for the solar industry is real capability.