I prefer it when sources list either all energy sources together (thus showing shifts in distribution) or bundle them as an aggregate. When pundits show any one energy source absent temporal or competing energy source data for context, you can only take it as a single data point.
my sense is that a lot of the additional in gas consumption is due to petrochemical sector. There have been more power generation from NG, but that's still just a smaller fraction of gas consumption in China.
my theory is that they significantly expanded chemical sector in the recent years at the expense of Euros, Koreans and Japanese. That explains why the energy consumption growth is so much higher than GDP growth. Normally, it should be the other way around,