There may be a modest spike in global coal demand this year as China force feeds economic activity because of the 19th National Congress (they do it every time) but growth will taper off in 2018 onwards.
So 2017 will prolly be the peak for CO2 emissions (up a little on 2016), with small falls in 2018 and 2019, but accelerating declines thereafter.
Using only partial data, the Global Carbon Project estimates that emissions rose about 2% this year.
The chart shows different possible pathways to limiting the global temperature rise to 2 degrees C. The longer we take to start reducing emissions, the faster they'll have to fall to reach zero in time prevent a rise of more than 2 degrees C.
Is such a sharp rate of decline from 2020 onward plausible? Possibly.
The costs of renewables continue to decline. In the US, whole-cost renewables are cheaper than new coal, and in some locations cheaper than the operating cost of existing coal. In China, wind and solar are close to the total cost of new coal. By 2019, new wind will beat new coal, by 2020, new solar will beat new coal. And that's without taking account of the indirect costs of air pollution imposed by coal. In Japan, solar will beat new coal by 2023, in India solar already beats new coal, though BNEF predicts the crossover in 2020. (Source: BNEF)
That's a tipping point. It means no new coal power stations will be built after 2020 or so. In fact, since everybody is aware that renewables are falling in cost, it means that any coal power stations started now will be financially unviable when they are completed in 5 or 6 or 7 years time, because by then the whole-cost of renewables will be less than the running cost of coal. No rational person will build new coal power stations. By the mid 2020s, demand for coal for power stations will be falling by 10% per annum (Coal for iron & steel will still be used, but even there, alternative technologies will likely be cutting demand) Since burning coal is responsible for 40% of total CO2 emissions, this suggests that emissions will be falling by 3% per annum just from that.
China has mandated that 10% of new car sales in 2019 will be EVs or plug-in hybrids. And that percentage will rise each year after 2019. Because of the structure of the mandate, an effective 4% of sales will be EVs/PHEVs in 2019, and 15-20% by 2025. This rise will likely be mirrored in most car markets because of the rapid decline in the cost of EVs. So by 2025, oil demand will also be falling 2% per annum, and that rate of decline will only accelerate. By 2030, oil demand (for petrol and diesel) will be falling by 10% per annum. This will reduce total emissions by another 3% per annum.
Meanwhile, the world will go on warming. And politicians everywhere will be introducing or raising carbon taxes, forcing ICEVs off the roads, urging and compelling remaining carbon-emitting sectors such as iron & steel, cement, air travel and sea transport to de-carbonise.
So yes, I think after 2020, the decline in carbon emissions will accelerate and we will reach zero emissions (or as close as) by 2050.
Have CO2 emissions already peaked?
Global CO2 emissions could be peaking
The 4% club