Tuesday, April 30, 2013

Oil Supply and Demand to 2025

Yesterday, we took a look at what 7%ish growth in China's oil demand would do if continued to 2025 - adding about another 15 million barrels/day (mb/d) to global oil demand.  Today, let's complete the exercise by looking at the other areas of the world where oil demand is growing rapidly, as well as the trends in supply (all data from BP).  We will see that things don't add up

Firstly, here are the developing regions of the world where oil demand was not squelched by the 2005-2008 oil price shock, or the 2008 financial crisis:

It's easier to see the trends in the individual regions if we show them as separate lines, rather than stacked on top of each other:

Clearly, China is the most important in terms of both level and growth in oil demand, and the Middle East is second.

Now, let's look at the growth rates (in each case, over the decade prior to the year on the x-axis):

In some cases the growth rate has been dropping on the whole (Asia outside of China), while in others it's rising (the Middle East, boosted by high oil prices).  Regardless, in each case we'll use the growth rate for the latest available decade (2001-2011), and calculate the incremental demand that each region would require if that growth rate continued to 2025.  That looks like this:

There's about 30mb/d of incremental demand overall, with China accounting for about half of it.

I stress that this is an extrapolation of the recent trend, not a forecast of what will happen.  In fact, as we'll see, this very likely can't happen.

Let's now look at global oil production.  We'll stick with the annual data from BP, which includes tar sands and natural gas liquids, in addition to crude oil, but not biofuels.  I think that's a reasonable assumption set over this timeframe - NGLs can be used to fill petrochemical demand that would otherwise have to come from oil, and I don't think biofuels can grow much more without causing intolerable food price increases (indeed that already started to happen).

Anyway, here's the data, broken into pre- and post-2005 regions:

As regular readers know, oil production hit a "bumpy plateau" in 2005, and although oil supply has increased since, the increases have been slow and fitful (red line), and only achieved by much higher prices than were required historically.  The green line shows what happens if we assume the existing slope of the plateau continues through 2025.  (Ie, this is not a near-term peak oil scenario at all, we are just assuming that the observed rate of increase of the last seven or eight years continues, and whenever oil peaks it is after 2025).  As you can see, we only get an extra 5mb/d or so this way.  Compare that to the 30mb/d we got by extrapolating the demand of developing regions.

So what's been making up the difference so far?  Primarily falls in oil demand in the developed regions of the world (here taken to be the OECD).  If we look at OECD oil consumption from 1990-2025, we get this:

OECD consumption peaked in 2005 (boundary between the blue and red regions of the data), and in my opinion is never likely to reach that level again.  Even though global peak oil has not occurred, peak OECD demand has likely occurred given the burgeoning demands in the developing world (peak oil is not synchronous).

The two orange lines show possible scenarios for future OECD demand.  The lighter top line shows what happens if we just extrapolate the post 2005 decline.  Note that this post 2005 period has not been a great time in the OECD.  The US faced an oil price shock followed by a bursting housing bubble, and is still experiencing slow growth and elevated unemployment.  Europe is even worse off, with no sign of an end to the recession there, and parts of the periphery experiencing Great Depression levels of economic damage.  Meanwhile, Japan has been stuck in permanent economic doldrums since the nineties.  Yet, if we continue this trajectory out to 2025, we only save about 8mbd.  Add that to the projected 5mbd of new supply, and we've only got 13mbd, not 30mbd.  To get things to balance purely out of OECD conservation, we need to lop off 25mbd of demand by 2025 (the heavier orange line lower down).  That's a drop of more than half in 13 years.

So it should be clear that something has to give here.  Either oil supply has to start growing a lot faster than it has been in the last eight years, or the OECD has to go on a crash conservation program, or developing regions, especially China, have to grow their consumption much more slowly than they did in the last decade (with the oil shock and the financial crisis).  Or some combination of all three.

So, either there's a financial crisis considerably worse than the 2008 one in our future, or we are going to need oil prices considerably higher than we've recently experienced to reconcile supply and demand.


Unknown said...

Gosh Stuart, this is a great analysis going in the same direction as Jørgen Randers. The scarcity of resources (because of extraction intolerable price) will slow down the GDP in developed countries. The German TV Arte showed a TV report on the former Soviet nuclear cemetery located in Arctic sea: there 3 nuclear Submarine and more then 30 nuclear reactors in deep sea loosing highly radioactive stuff... So the arctic oil frontier will not be exploited as easy as that... Another point is that governments could decide to keep oil prices low to maintain populations quiet... How long?

Michael R said...

"So, either there's a financial crisis considerably worse than the 2008 one in our future, or we are going to need oil prices considerably higher than we've recently experienced"



Luis García said...

Great analysis, Stuart! Fairly clear conclusion: "business as usual" trends are sustainable not even for 13 years unless "oil production as recently" trend changes dramatically (and very unlikely, I'm afraid)

Jim McDonald said...

Stuart how reasonable are the stories about methane-hydrate?

Is that technology far enough along to have an impact in the next decade? Would it impact Chinese oil consumption?

Greg said...

Thanks for laying things out so clearly, Stuart. But I think things aren't quite as bad as they look.

I believe that it's usual in economic matters to measure from peak to peak to estimate trends. On the OECD chart, a line through the 2008 peak and current consumption (assuming that this is another local peak, as Europe's latest unemployment figures suggest) is closer to the lower orange line than the upper one.

Also, Chinese authorities recognise the need to move their economy from manufacturing to services, so China's oil consumption growth probably won't follow the historical trend.

I expect oil to induce not a crisis but depression (chronic high unemployment) in the OECD, much as we have now.

yvesT said...

You for sure are a confused piece of work ...
Any idea of the per capita oil consumption of the "service oriented" USA compared to the chinese one ?

HalFiore said...

I'm confused about the methodology used to generate the trend line in the next to last graph. it looks to me like a trend of the post-2005 data ought to give a much lower slope, though I think still >0. Someone above mentioned it being customary to go peak-to-peak, but I can't think of any reason you would do that.

OTOH, It makes more sense than what I'm seeing in the popular media these days, which is something like extrapolating the data after the low in, I guess that must be 2009. You can get a pretty impressive slope doing that!

Emil said...

Great, data-driven analysis.

And of course, we're already seeing a correction in the developing world's growth rates. India is down at 4.5%, China at 7.5%(officially), Brazil is at a pitiful 2%.

Even if OECD oil demand begun to look like the bottom line in your diagram, that would impose a great depression worse than the 1930s which would choke off any oil demand in the other part of the world due to a global crisis of epic proportions.

So we can just say, slower growth is coming to this world at a minimum.

But the most important thing to add is this: all of this assumes that oil supply can grow by 5 mb/d into 2025 which is far from certain.

So far, U.S. tight oil production has merely been able to offset other OECD oil production declines(like in Norway, UK and other places).

Much of the "oil supply" is actually not oil at all, but natural gas liquids which BP converts to oil in their database.

Something people ought to know.

Final word, Stuart, please keep these posts coming! You're a guide to many in the peak oil blogosphere by being unbiased and data-driven. Too many verbal-only doomsters who feed off fear rather than rational analysis. Keep it up!

Nate H said...

"Even though global peak oil has not occurred, peak OECD demand has likely occurred given the burgeoning demands in the developing world (peak oil is not synchronous). "

Thats not why OECD demand has peaked. Its peaked because we can't afford $100+ oil anymore. We can afford it only so long as the credit spigot stays open (same goes for China btw). The relationship to our demand peaking and non-OECD demand is secondary to credit availability.

Stuart Staniford said...

Nate - mmm, well, essentially we've been outbid for it. If it wasn't for the developing world being willing to pay $100+ for a larger and larger share, we in the west would still be getting it for <$60, and we'd be using more. At the same time, I acknowledge that lack of credit availability post the financial crisis has certainly constrained some lower income consumers who might otherwise have tried to consume more. However, had that not been the case, those additional consumers probably would have bid oil prices up even higher, and probably not have changed the supply trajectory all that much.

Stuart Staniford said...

Hal - it's essentially a straight line drawn between the two end-points. It's a little hard to see that last year or two of data because it's underneath the orange line.

Stuart Staniford said...


I understand Japan has had some initial proof-of-concept success in extracting some methane hydrate. However, I don't consider natural gas to be in short supply in general, and methane hydrate is just more natural gas. So I don't think it changes the oil picture much. What we probably will start to see is more places using natural gas as a transportation fuel, which is one way to lower oil demand.

Nate H said...

stuart - thats the thing, we really have NOT had a decline in credit availability as a nation - the central bank took over what was previously the role of commercial banks so total credit is about the same - sure the breakdown is different and the indigent now have zero access to credit - but my point is once central banks/sovereigns can no longer grow credit then demand will go down far sharper than natural decline rate of oil. The relevant question to me is -if oil goes to <$60 which i fully expect in next couple years, what does that do to production? (I ultimately think government will support oil companies who are losing money but in the intermediate term we will be awash in oil). Again, my view is we don't face a resource shortage on energy, but one of contribution - there is plenty of hydrocarbon resource/reserve out there, but very little at the prices we constructed societal institutions/beliefs around. In one sentence - we are going from spending 5% of our energy on energy to something like 10-15% - and that has major implications

Catfish said...

Oil shortages have been predicted since the mid-70s and the predictions have always proven wrong. Assuming the numbers above, several things will happen: 1) the price of oil will increase thus driving down consumption and increasing exploration/production; 2) the incentives to use substitutes will increase. Natural gas is more abundant than ever and a viable oil-substitute for most uses. Electrification of transportation will accelerate and drive down oil demand. So, i don't really see a big supply problem coming. We all dislike carbon-burning fuels and want to see clean fuels. It's high time the government tax carbon.