World Oil Production and Prices

The BP World Energy Report 2016 is out. Unfortunately this report blends crude oil and condensate reserves and production with natural gas liquids reserves and production. Natural gas liquids are mostly ethane, propane and butane, they are usually recovered from gas streams using cryogenic processes, and are marketed separately (ethane is mostly used as a chemical feedstock). 

In the past, NGL recovery was neglected, or the NGL recovered were burned as "trash gas" because they couldn´t be transported mixed with the crude (a crude spiked with NGL can exceed the vapor pressure limits for pipeline, rail, and ship transport, and this can cause explosions and fires). 

Gas production and capture has also been increasing. By capture I mean the use of associated gas streams which have historically been vented or flared together with the associated NGL, such as was common practice in the Soviet Union. Thus the "oil production" reported by BP has a gradually increasing NGL volume added to it. This tends to mask the real evolution of the world´s crude oil and condensate capacity. 

The following graph shows the production and the refinery throughput reported by BP. As you can see, the refinery throughput is a pretty decent proxy for "real" crude oil and condensate production (because this stream is what´s fed into refineries and what many of us think is about to hit a peak). 

 World Oil Production and Refinery Throughput,

 and the difference between the two 

Future world oil production will increase if prices rise. It´s difficult to predict the demand-price-production triangle. This means it´s hard to predict these parameters. In the past I had access to dozens of price forecasts prepared by dozens of oil companies, banks, and governments. I can´t disclose the sources, but I can reveal they are usually wrong. 

This led me to prepare a simple algorithm to predict oil prices, which seems to work (more or less) for long term trends. 

Oil price history and forecast, in constant 2015 US dollars per barrel. 

I shaded the 60 minus region in red, to indicate that production will decline relentlessly as long as the price stays below this threshold. This is subject to minor blips caused by the ongoing oil price war between Russia, Iran, and Saudi Arabia (a war that´s spurred by the Syrian and Iraqi civil wars, in turn triggered by US and EU meddling in the Middle East).

I also shaded the 80 plus region in green to indicate that production ought to increase slightly once the price exceeds this threshold. This is purely speculative on my part, because oil companies and their banks make decisions based on their internal oil price forecasts. These forecasts are biased somewhat by the oil price of the day, but in many cases they are developed by forecasting wonks using very complex models, and company management boards tend to believe these figures even though they ought to know the forecasting game is mostly a load of bullshit (see Morgan Stanley´s forecasts in recent years to get an idea). 

This forecast is completely useless for companies interested in developing "shales", because they depend so much on the price during each well´s first year of life. However, this forecast can be used with caution for long term projects, such as Kara Sea exploration or Canadian extra heavy oil developments. It can also be used to estimate the future competitiveness of electric vehicles and biofuels.