Can Chinese coal data be trusted? Why we should not believe peak coal has arrived

Posted on Updated on

A few thoughts about pigs in China.

Media reports regularly claim that China consumes half of the world’s pork. However, these reports largely ignore what we actually do know about China’s consumption of pork, which is this: we do not know what China’s consumption of pork is with much accuracy.

We know this for the simple reason that China’s official statistics are completely inconsistent (see the recent paper, “Where have all the pigs gone? Inconsistencies in pork statistics in China“.)

China’s official statistics tell us a number of things about pigs, but most importantly it tells us what the supply and consumption are.

These figures should be roughly the same, with some marginal losses from the supply until you get to consumption. Yet, official statistics show that supply is more than double consumption. In fact, it was almost triple consumption in 1999, which is physically nonsensical.

So we cannot trust Chinese pig statistics.

What about coal? Here things are similar. We should Chinese coal data with skepticism, and the same goes for claims that peak coal has been reached in China.

The long history of massive revisions to China’s coal statistics

First there is the problem of historical revisions to data. China regularly revises its historical coal statistics through what it calls “Economic Census”. We can see what the revisions have been in the past by comparing historical data reported in the BP Statistical Review of World Energy in 2003 and 2013.

Below is annual Chinese coal consumption from 1965 and 2013 as reported in 2003 and 2013. (This graph is a reproduction of a similar one by Glen Peters.)


Things are perhaps much clearer if we look at the revisions in terms of percentages:


Two things leap out. China’s coal consumption at the turn of the century was later revised massively, upwards by almost 30%. In contrast, Maoist China appears to have seen massive over-reporting of coal production. Anyone with faint knowledge of China’s history will not be surprised. Over-reporting of numbers to meet delusional Mao made targets led to immense suffering.

If you were statistically minded you might conclude that there is a long term trend, one which switches from China initially over-reporting energy consumption in Maoist times to under-reporting it today.

Under-reporting today may serve many purposes. In the late 90s, the under-reporting was largely thought to be a result of central commands to shut down inefficient coal producers not being followed. Instead of shutting old coal mines down, many regions simply didn’t report the figures, an easy thing to do in a country as corrupt as China. Furthermore, there may be an incentive to deliberately under-report coal data to make it seem you are doing more about air pollution than you are. There is already evidence that Chinese cities are faking air pollution data; it would not be far fetched to imagine coal production data is faked for similar reasons.

And is the situation improving? Not really. Recent years have seen similarly high revisions. Coal output for 2007 was revised upwards by 6.2% in the 2009 Census. Meanwhile, earlier this year China revised its coal production in 2013 up 7.8%, from 3.68 to 3.97 billion tonnes. This revision, of roughly 300 million tonnes, was equivalent to one third of the annual coal consumption of America.

The inconsistent between national and provincial coal statistics

Second, there is the problem of which statistics to look at. China produces coal statistics at both the national and provincial level. They should add up to the same thing, but they don’t.

These inconsistencies were first noted by a prominent article in Nature Climate Change in 2011. This paper observed that provincial statistics imply that coal consumption is much higher and that this might result in China’s annual CO2 emissions being more than a billion tonnes higher than is officially claimed.

Further research since then has backed up that research. Ma et al. (2014) found that coal consumption, as estimated using provincial data, was at least 20% greater than it is when using national data. Similar discrepancies were found by Mischke et al. (2015).

This problem needs to be resolved, and it might be using satellite data. But until then we must recognize that China may consume 20% or so more coal than is official claimed.

We must remain skeptical of claims coal is “peaking” in China

A new meme has become popular about China. The meme is simple: peak coal has been reached. The motivations of some of the people pushing this meme are not exactly academic. Many environmental groups clearly see narrative value in the idea of peak coal in China, so they push it. If you want to stop people being able to export coal mined in America to China then it is a good idea to convince policy makers there won’t be a market for that coal.

The main evidence given for peak coal is that coal consumption supposedly fell last year. The data above shows that we should wait before even believing that this actually happened. Glen Peters put it well, Chinese energy statistics are like a good bottle of wine, they are improve with age.

Furthermore, there are economic data points coming out of China which defy belief. Official statistics supposedly tell us that China’s coal consumption fell 8% in the first quarter of this year, while GDP still went up 7%. If those statistics don’t raise a skeptical eyebrow, well, I have a bridge to sell you.

The official statistics also tell us that China’s electricity consumption only rose by 0.2% in this time. This is more than suspect. From a statistical point of view electricity generation has a big benefit: it can be measured very accurately. This makes at an excellent proxy for the growth of the economy, as Wallace (2013) showed.

Wallace also showed that discrepancies between the growth in electricity generation and GDP growth were useful indicators of China’s GDP being juked. Between 2000 and 2010, percentage growth in electricity generation was more or less uniformly greater than growth in GDP, with some notable exceptions. The fall of Lehman Brothers and the near collapse of the western financial system resulted in growth in electricity generation going negative, but miraculously this did not result in much change in GDP growth. China was clearly juking the stats.

And clearly we must ask today if they are juking the GDP numbers.

The final, and biggest problem, for the “peak coal” thesis is more obvious: China is still building lots of coal power plants.

Last year almost 50 GW of new coal power plants opened. This is like installing the entire British coal fleet every six months.

And there is more to come.

According to Coal Swarm, 116 GW of new coal power plants are under construction. A further 48 GW has been permitted, while 229 GW is under pre-permit development.

Clearly the Chinese are not acting as if peak coal has arrived, and we should not either.


4 thoughts on “Can Chinese coal data be trusted? Why we should not believe peak coal has arrived

    […] worth noting that BP and official Chinese statistics have at times been heavily revised in the past, including a 19% increase in the coal consumption figure for 2007 between […]


    […] wrote earlier in the week that we should be skeptical about China’s coal statistics. As Glen Peters has said, they are […]


    […] China’s primary energy consumption grew by 1398 million tonnes of equivalent (Mtoe). Though,if history is a guide this figure will eventually be revised […]


    […] China’s primary energy consumption grew by 1398 million tonnes of equivalent (Mtoe). Though,if history is a guide this figure will eventually be revised […]


Comments are closed.