During periods of low demand in the first half of 2025, electricity repeatedly hit rock bottom prices on the spot market of Chinese provinces like Zhejiang and Shandong, sometimes even going negative.
This was not merely the result of abundant wind and solar power generation. It was also partly because a large proportion of electricity demand had already been met by off-market power-purchase agreements, usually for coal.
In Zhejiang during January’s Chinese New Year holiday, prices reached CNY -0.20 (US$-0.03) per kilowatt-hour – the lowest allowed – multiple times. During the Labour Day holiday in May, prices on Shandong’s spot market stayed negative for 22 hours in a row.
This is due to a problem with the decarbonisation of China’s electricity systems: despite rapid roll-outs of wind and solar power, coal is being used more than it should, rather than reserved as a flexible standby.
China’s electricity system is experiencing rocketing demand: between 2022 and 2024, annual growth in demand was between 8 per cent and 10 per cent in many parts of China, particularly in Zhejiang, Hubei and Jiangsu. At the same time, China still has 200 gigawatts (GW) of coal power capacity either under construction or permitted. The country’s existing capacity stood at 1,200 GW as of the end of 2024. We think this could reach 1,500 GW, if not more, by 2030.
Government policy documents have repeatedly designated coal power as a supporting and peak-regulation source, rather than a primary provider of baseload. Achieving this new transition role for coal units requires coordinating dispatch principles, market mechanisms and administrative arrangements. This is a major challenge for the decarbonisation of electricity.
Zhejiang: A case study in overusing coal
A National Energy Administration official put Zhejiang’s negative January electricity prices down to lower demand and surplus renewable generation.
Figures from energy data provider SP RiXin show demand across Zhejiang fell to 17 GW at midday on 27 January, two days before Chinese New Year. That meant all demand could be met by base-level generation directly ordered by grid dispatchers. Such electricity, mostly coal powered, is not affected by market prices.
Chinese provinces commonly have a “coal benchmark electricity price”, which select power plants awarded “out-of-market” commitments are permitted to charge, rather than having to compete on the wider market.
As a result, some coal power plants are scheduled to run to ensure supplies even if they produce power expensively as they will still make a profit. Meanwhile, other sources of power that would have been competitive – mainly renewables and a few coal power plants – are curtailed or shut out. These out-of-market commitments mean overuse of coal power at the system level.
We believe China should retain the minimum possible number of coal power generator units. This would broaden the opportunity for renewable generators. For example, demand for 17 GW of power could be met by 17 units running at full capacity, or by 34 at half capacity. In the latter case, the generators are already running at the minimum possible level and can’t drop down any further, meaning less room for wind or solar power integration. This is overuse of coal at the micro level, namely the generator-unit level.
Also, even at times of low demand, Zhejiang continues to import power from outside the province, 70 per cent of which is coal power. External dispatch arrangements do not respond to electricity prices or demand, with ahead-of-time plans implemented regardless.
Those imports are another example of the excessive use of coal power. Under China’s electricity-dispatch systems, each province is responsible for balancing its own supply and demand.
When local supply outstrips demand and prices start to fall, perhaps even below zero, there should be no need to use imported electricity, with its associated extra costs. When imports do occur, it doesn’t just reflect resource imbalances. It also indicates that at the macro level, China as a whole is still highly reliant on coal power.
How do negative power prices come about?
When a large share of electricity is ordered off-market in advance, there is less demand on the “day-ahead” markets. This leads to lower – potentially even negative – prices. (The day-ahead markets are part of the spot electricity market, with electricity being purchased for delivery the following day. Buyers and sellers set hourly generation levels and prices through an auction).
Analysing actual market data from the first five months of this year, we found negative prices were experienced almost all day in Zhejiang and Shandong when wind and solar power supplied 20 per cent to 25 per cent of demand. Negative prices also occurred when overall demand dropped to 55 per cent or less of peak levels with wind and solar supplying only 10 per cent, but not all day.
In April 2025, solar power was sold on Shandong’s spot markets at an average of CNY 0.02 per kilowatt-hour, far below coal’s benchmark price of CNY 0.35-0.45. This is because grid dispatch managers order generation directly, via off-market purchase agreements, rather than via market mechanisms. Such direct dispatch often meets about 50 per cent of all demand, greatly reducing the opportunities for market price-setting, and keeping the marginal payment for renewable power extremely low.
Also, markets are fragmented. Day-ahead markets do not usually allow electricity consumers to bid for power and adjust their power consumption accordingly. The three markets – day-ahead, intraday, and mid- to long-term contracts – work like independent shops, with demand unable to flow between the three.
China’s mid- and long-term electricity supply contracts are still largely based on physical delivery – the generator must actually produce the electricity and deliver it to the grid to fulfil the contract.
Electricity is a uniform good – once delivered to the grid, power from different sources cannot be distinguished. So, in theory, a generator could fulfil a contract by paying another generator to deliver the same amount of electricity. Requiring physical delivery removes flexibility and hampers optimal allocation of resources.
In electricity markets, the price of power is usually set by the rates of the last power plant needed to meet demand. This “marginal generator” usually produces the most expensive power because the cheapest generators are used first. Such a setup balances supply and demand for the greatest overall benefit.
In a coal-reliant system, prices should in theory fluctuate in line with the costs of fuel for the marginal coal-power generator. But in reality, many provinces have spot prices much lower than that theoretical value, as well as wild fluctuations.
This is because generation from more expensive coal-power plants is prioritised (through off-market commitments), removing the base demand that should be met on the market by low-cost generators. So, lower-cost renewable generators end up becoming, at points, the marginal generator. When demand is ultra-low, those low or even negative prices set the price for the market as a whole.
Negative prices in China and abroad
Negative electricity prices also occur in the US and the European Union (EU), but there are significant differences in their cause and frequency when compared to China.
In the EU, electricity exchanges and dispatch authorities are generally independent of each other. The players in the exchanges determine the quantity and location of electricity to be supplied. The dispatch authorities are not involved in this trading process but are responsible for the electricity-balancing market and physical dispatch. Self-dispatch is commonly used: generators themselves determine how much electricity to supply given the prices on offer.
Dispatch results from market trading, rather than government orders. As the ancillary service and energy markets are separate, inefficiencies can result.
For example, if a generator signs a mid- to long-term contract at a price that is higher than the market price at the time of delivery, economics would dictate it stops generating power and simply buys electricity on the market with which to fulfil its obligations. But when power plants opt to generate the power themselves, costs go up for the entire system. This runs contrary to the system-wide goal of minimising costs.
Under the US Independent System Operators arrangement, electricity markets and dispatch are closely linked. Dispatch authorities use an optimised model and bids from market actors to determine a plan for the following day’s generation, and adjust that plan in real time. Prices for electricity and ancillary services are co-optimised, with the aim of minimising overall costs.
In comparison, China’s provincial spot market trials reveal an imbalance between authorities and responsibilities: dispatch authorities have strong powers, as in the US, but their duties are more akin to those of their EU counterparts: to ensure physical balance. There is no goal of overall optimisation, and generators are not empowered to come up with their own dispatch arrangements.
With the market and dispatch uncoupled, market pricing often fails to reflect supply and demand realities. The low and negative prices seen in China are largely due to market mechanisms being hampered by non-market arrangements, rather than too much renewable energy in the system.
The impact of negative prices
Negative electricity prices prevent wind and solar power from earning their market value, which necessitates external support to sustain renewable-energy investments.
In response, this year the Chinese government announced reforms in how renewables are sold to the grid. Contract for Difference-type mechanisms should ensure income for wind and solar generators.
Many bodies have interpreted these changes as meaning more of a role for market pricing. In reality, the existing, entirely market-based method of determining revenue for renewable generators is going to be replaced with subsidies and taxes, based on expected pricing.
It is more a case of creating time for follow-up improvements to market rules, than of implementing actual market reforms. But the new arrangements may reduce the internal impetus for such reforms and mean a return to the traditional model: all short-term pricing depends on long-term costs, and dispatchers decide who generates. The outcomes remain to be seen.
If we are to avoid a rollback of market reforms, we have two options. First, make more use of markets. Use markets in dispatch, increase flow between the different markets, and let market actors respond to price signals, borrowing from the EU’s experience. Option two would be an overall optimised dispatch system based on price or cost signals. This would do away with off-market purchase agreements, closely coupling trading and dispatch – as is done in some competitive markets in the US.
Simply put, the current policy provides space for further improvements to market mechanisms and rules, but those reforms are still needed.
The outlook for 2030
China’s electricity system could be described as half pulling ahead, half lagging behind. Demand is pulling ahead, with sustained growth. The management lags, however, with higher levels of administrative intervention and centralisation overriding market forces: the pace of investment, the power generation mix, price levels and market access are all under high levels of administrative control.
If China is to eliminate frequent negative prices on the spot market, it needs to prevent the overuse of coal power. Low or negative prices, overuse of coal and lack of flexibility are three aspects of one problem. The generation mix needs to be improved, with operational decisions made in line with market prices or cost signals, to avoid overuse or excessive retention of coal power.
Given the scale of existing and planned coal power generation, China may have 1,500 GW of coal power by 2030. In that scenario, coal’s share of the electricity mix will hold steady at 55 per cent and CO2 emissions from the power sector will increase slowly, by about 0.2 per cent to 1 per cent a year.
This is a continuation of what we have seen for the past 5 to 10 years. But it does not align with the national goal of achieving peak carbon emissions before 2030. It may, however, be a “moderate deviation” deemed acceptable by policymakers. The impetus for future electricity-sector reforms may come from changes of international rules, and the increasing priority given to climate issues in domestic politics.
This article was originally published on Dialogue Earth under a Creative Commons licence.
