As China emerges from one of the most serious epidemics of the century – even as much of the rest of the world remains in a coronavirus crisis – the country’s energy demand and emissions are beginning to return to normal.

Update 30 March 2020: The analysis contained in this article was first published in mid-February and was subsequently updated several times as new figures became available. This analysis remains in the first sections of the article, below.

Now, as of the end of March and based on the most recent data, including official government figures, it is possible to gauge the accuracy of this initial analysis. There is a detailed update and reflection on what is known so far at the end of this article.

Overall, the figures reinforce my estimate that China’s carbon emissions fell by around 25% over a four-week period, as outlined below, equivalent to around 200m tonnes of CO2 (MtCO2). Demand slowly returned to normal levels over an extended seven-week period, bringing the reduction so far to around 250MtCO2, with emissions some 18% lower than usual levels.

As before, however, the Chinese government’s coming stimulus in response to the disruption could outweigh these shorter-term impacts on energy and emissions, as it did after the global financial crisis and the 2015 domestic economic downturn.

A country in shutdown

Every winter, during Chinese new year, the country closes down for a week, with shops and construction sites closing and most industries winding down operations. The holiday has a significant short-term impact on energy demand, industrial output and emissions.

The blue lines on the chart below show how coal-fired power generation typically drops by an average of 50% in the 10 days following the eve of Chinese new year, marked as zero on the x-axis.

This year, shown in red, the usual fall in energy use has been prolonged by 10 days so far, with no sign of rebound. This is because the annual holiday was extended to give the government more time to get the epidemic under control – and demand has remained subdued, even after the official resumption of work on 10 February.

Daily coal consumption around the Chinese new-year period at six generating companies reporting daily data, in 10,000 tonnes per day. X-axis shows days before and after Chinese new year eve, which falls on various dates in the second half of January or in February. Source: Analysis of data from WIND Information. Chart by Carbon Brief using Highcharts.

In the four-week period commencing 3 February this year, average coal consumption at power plants reporting daily data fell to a four-year low, with no sign of recovery in the most recent data, covering Sunday 1 March.

The short-term effect has been equally dramatic across a range of other industrial indicators, shown as 28-day averages in the figure below. The top left chart shows coal throughput at the main coal port, Qinhuangdao, which fell to the lowest level in four years in the four weeks to 1 March.

Similarly, refinery operating rates in Shandong province, the country’s main centre for oil refining, fell to the lowest level since autumn 2015 (below left), indicating a sharply reduced oil demand outlook. Furthermore, as expected, underlying demand for oil products, steel and other metals has fallen much more than output, resulting in record-high stockpiles, which will put pressure on production going forward.

Strikingly, all indicators of industrial capacity utilisation – coal power plants, blast furnaces, coking, steel products, refineries – deteriorated further in the week commencing 10 February, when business was officially expected to resume.

The rebound in industrial operation and domestic fossil fuel consumption has proven to be slow, with the first signs of the resumption of activity evident in the national aggregate data only in the past week, but still with a long way to go. This is not for lack of trying though, as some cities have reportedly even resorted to mandating factories to use more electricity, whether or not they have the personnel to resume production, in an effort to doctor a resurgence in power demand. While anecdotal, this is testimony to the massive pressure on local officials to jumpstart the economy.

Taken together, the reductions in coal and crude oil use indicate a reduction in CO2 emissions of 25% or more, compared with the same two-week period following the Chinese new year holiday in 2019. This amounts to approximately 100MtCO2 – or 6% of global emissions over the same period.

One exception to the wider downturn has been primary steel production, which kept running through new year and the extended holiday. In contrast, production of the main steel products – a closer proxy of demand – is down a quarter, hitting the lowest 14-day level in five years. Unless demand rebounds fast, blast furnaces will have to shut down as well given limited capacity to hold stocks and a souring demand outlook.

There is further confirmation of the reduction in fossil-fuel use in satellite measurements of NO2, an air pollutant closely associated with fossil-fuel burning. In the week after the 2020 Chinese new year holiday, average levels were 36% lower over China than in the same period in  2019, illustrated in the right-hand panels below.

Demand-side impacts

Although the short-term impact of the current crisis is large, in terms of reduced energy demand and industrial emissions, the longer-term direct effect of factory closures could be much more limited.

Apart from the annual Chinese new year holiday, shutdowns of a week or more are not uncommon in China.

Moreover, shaving 25% off energy consumption and emissions for two weeks would only reduce annual figures by around one percent. China also has very substantial overcapacity in all of the major CO2 emitting industries, meaning production volumes – and emissions – can catch up rapidly after a shutdown, if the demand is there. 

Any sustained impact on fossil-fuel use would come from reduced demand, which initial indicators suggest could have a major impact. For example, February car sales are forecast to fall by 30% below last year’s already depressed levels.

If consumer demand is reduced – for example, due to unpaid wages during the crisis cascading through the rest of the economy – then industrial output and fossil-fuel use might not recover, even though capacity is available to do so.