Peak coal in China? Not so fast

China’s coal consumption officially fell by 2.9% last year for the first time in 14 years. Is this evidence of “peak coal” in China as some are already claiming or a temporary blip?

Let’s begin with an obvious problem. China’s coal demand officially declined 14 years ago. In other words, just prior to the biggest coal binge in human history China’s coal demand was officially flat. This sounds hard to believe. And it is.

Between 1996 and 2000, China’s coal consumption grew by only 0.3%. However, as Glen Peters has pointed out, BP has significantly revised China’s coal consumption figures between 1996 and 2000. And the same is true for China’s official statistics.

Officially China’s coal consumption declined massively in the late 1990s. However, this was only in the original statistical estimates. New data resulted in massive revisions to those early estimates. And this is clearly a long term problem. I reproduce a fascinating graph from Peters’ below.

Furthermore, the causes of the late 1990s “decline” in coal consumption must be borne in mind. Coal consumption almost certainly did not decline at all, but was simply under-reported (see Peters et al. 2007). There are potentially many reasons for this. Among them is the apparent failure of China to actually close small or illegal coal mines. Mines that were officially off the books were still producing coal, even though that coal was not appearing in official statistics.

This may be happening today. China is now officially closing large numbers of small coal mines, but it is unclear how many are actually being closed.

Similarly, illegal coal mining appears to occur on a large scale. Greenpeace recently found one “14 times the size of London” in Qinghai province. However, unsurprisingly, there appears to have been no attempt to properly quantify the extent of illegal coal mining.

A final problem with China’s coal statistics is that they are not internally consistent. A recent comparison of CO2 emissions using official national and provincial level statistics found a difference of 1.4 billion tonnes (Guan et al. 2012), which would translate into close to a 10% difference in coal production.

This is a consistent problem with Chinese statistics. Attempts to reconcile them with each other can quickly lead you to attach significant error bars to them. For an important example, see research on pork consumption statistics (e.g. Yu and Abler 2014).

So, clearly this decline in coal consumption should be treated skeptically simply because the underlying statistics are unreliable.

Not only that, but we must put this decline in the context of a significant fall in growth in the industrial drivers of coal consumption.

This is most noticeable in steel, which is the biggest source of coal consumption outside of the electricity sector. World Steel Association estimates for 2014 indicate that after 3 decades of year on year growth, China’s steel production did not increase last year.

This is shown in the graphs below I’ve produced using WSA data. Officially, China claims that crude steel production increased by 1.2%. This is growth, but still, it’s the slowest growth in over 30 years.

In addition, reports indicate that China’s demand for steel actually fell last year, and that it is being forced to increase exports to make up for the shortfall in national demand.

The same goes for two other key drivers of coal consumption. Electricity production, which makes up around 50% of China’s coal consumption, only grew by 4% last year. This is the lowest annual growth rate in the last 30 years, with the exceptions of 1997 and 1999. Typical growth rates were above 10% in the last decade.

Cement production growth is not entirely clear, but it tells a similar story. China is officially stating that it grew by 2.3%. However, USGS estimates indicate that it grew by 3.3%. Either way, this growth is much lower than the historical norms. A 2.3% increase would be lower than any year since 1990, and it is far below the typical 10-15% seen in the 2000s.

Cement statistics are similar to coal, they are prone to revision. Initial USGS estimates for 2013 coal production was 2.3 billion tonnes, which represented a 4.1% increase over 2012 levels. However, their more recent estimates revise the 2013 figure to 2.42 billion tonnes, which further revised the growth figure to 9.5%. So, again, we will have to wait for more reliable numbers.

But these low steel, electricity and cement growth figures imply that China’s industrial growth, and probably GDP, is lower than officially stated. But, again, the reliability of GDP figures has long been questioned.

So, can we take seriously claims that coal consumption in China has peaked or is about to? Almost certainly not.

The peak coal thesis appears to rest on a single data point, and this data point rests largely on China’s industrial growth slowing massively. And who is going to bet that China’s economy continues to slump?

But more importantly we must look to what China is still doing: building huge numbers of coal power plants. As Armond Cohen has calculated, the majority of new electricity generating infrastructure built in China last year were coal power plants.

China opened 47.3 GW of new coal power plants last year.

This is not the behaviour of a country that is going to peak coal use any time soon. If the majority of new power plants China built last year will burn coal, then talk of an imminent peak in Chinese coal use is little more than wishful thinking.