Production and the Hubbert Curve

Oil production seems fairly straightforward: countries pump out the oil as fast as they can and then it runs out. This is the implication between the R/P ratios (see Jargon) that are so beloved of oil companies and imply that oil will last for many decades to come. The problem with R/P ratios is that oil does not work like that.

Woodpile v Woodland

Oil production can be best understood by comparison with something such as wood. Imagine an island where there is one carpenter. The R/P ratio basis of oil usage revolves around the assumption that oil production works like a woodpile in the carpenter's backyard. Whenever he needs wood, he walks out to the pile and takes however much he requires. If things get busy and he needs more wood, he simply takes more wood from the pile. There is always enough to satisfy his needs until that fateful day when he removes the last plank and it is then all gone. The only factor in its price is demand - if fewer people want wooden things, the carpenter lowers the price to stimulate demand. If he has plenty of work on, he can increase the price and get the benefit.

Comparing this with oil, if the world has 1,050 Gb of oil remaining and we use 27 Gb a year, then dividing one by the other means that we will be able to use 27 Gb of the woodpile for another 39 years. Then the yard will suddenly turn out to be empty.

But oil does not sit in one huge hole in the ground, constantly being pumped out. Rather an oil field is a set of wells of different sizes, with new wells being set up as old ones dry out. The R/P ratio takes the view that the oil has already been found and is sitting patiently in the backyard. In reality, it is more like woodland than a woodpile.

If we imagine instead that our carpenter had to chop down a tree every time he needed to make something, the problems become more evident. Trees vary in their size, proximity and quality. Initially our man would pick those that were large, good quality and nearby. As this was relatively easy, his prices could be kept low. But, as time went on, he would have to cut more trees of smaller sizes, travel further to find them and use wood of a lower standard. This extra work would take longer and naturally result in higher prices. Eventually, unless the trees were managed and replaced, he would find himself unable to find enough wood to satisfy his customers.

But couldn't he cut the trees quicker to keep production up? He certainly could employ someone to help him (which would be like drilling more wells) but that would result in depletion occurring more quickly, and the quicker you cut away the large and nearby trees, the quicker you have to resort to the small and distant ones. New technology can only help so much; no matter what circular saw or four-wheeled vehicle you have, there's always a certain minimum time needed to cut down and drag a tree to the workshop. Production still falls, the best you can do is change the angle of the slope on the chart. Any increase in production means a gentler initial decline and a steeper subsequent one.

Oil production works in a similar way with the important distinction that, unlike trees, we cannot replace the oil we use. It is as if every tree the carpenter cut down was gone forever.

Such is oil production. Chart P1 above shows how the R/P suggests oil production might occur, as a woodpile view. An individual well might resemble that but fields are formed of many wells. Chart P2 is a more 'sophisticated' version, the one often assumed by economists that oil production can simply be increased to keep pace with consumption until the wells finally and suddenly run dry.

Chart P3 shows how it actually works, as a woodland. (In reality, because discoveries do not come along at perfectly placed times, and there are always political and economic effects, the smoothness of the curve gets somewhat warped. In particular, the right hand half tends to get stretched out (see chart P4). But the principle remains.)

The production curve for the US lower-48 (see Chart D2 in Discoveries) shows this curve in effect. Also, by looking at the production curve for the whole world (Chart P5), we can see how, up until the mid 1970s, the curve was a remarkably accurate reproduction of the theoretical curve. Then political elements interfered to balk the trend (see The 1970s Oil Shocks in Consumption). The result was economic recession and a reduction in consumption. Less oil used meant a reduction in production with the consequent ‘mountain range’ that we see in the chart. In some ways though, this reduction was actually a good thing for, if we had not reduced our production and consumption in the 1970s and 1980s, we would now be sliding down the decreasing slope and well on our way to disaster.

P5. Production Curve (World)

The Financial Aspect

Many objectors to peak oil claim that, as oil becomes more expensive, more reserves will become available because they are then financially worthwhile. For example, if the price of oil on the world markets is $100 a barrel and it costs $105 to extract it from a particular oil field, that field will not be worthwhile and therefore not classed as "recoverable resource" (see Jargon). If the oil price rises to $110, then that oil becomes viable. The argument goes that, as oil production falls, the price rises and more oil becomes available.

This is true to a certain extent (although there will not be that much extra oil available). What it ignores is the fact that the new oil will not be cheap oil. The price will remain high and continue to rise, and everything else will rise in tandem – electricity, fuel, plastics, food. What follows will be recession, unemployment, financial collapse. As is often pointed out (and often ignored) the problem that faces us is not the end of oil, but the end of cheap oil, and the end of our cheap energy society.

The Hubbert Curve

It is not pure chance that we found a curve that fitted production. The work was done by an American geologist, M King Hubbert, who predicted in 1956 that the US-48 oil production would peak in 1969. He was much reviled at the time yet he was out by just one year.

The Hubbert Curve (as it is now known), is a statistical figure formed by combining many independent producing fields into a single trend. But real oil production does does not always follow the curve. If discovery is intermittent or if the statistic covers a limited number of fields, it will look different. Many countries, for instance, display several peaks because oil has to be found before it can be produced. If the country has several discovery cycles, the production curve will mirror it.

Because the US-48 production has been uninterrupted, its production curve closely follows the Hubbert Curve, as we saw in Chart D2. Iran’s output, though initially trying to follow the curve, has been disrupted by revolutions and wars, and so is breaking into two peaks.

(To see how the Curve is actually formed, see the Hubbert Curve page.)

P6. Production Curve (Iran)

Previous Forecasts

At the beginning of this guide, I mentioned the writer who said that “I remember being told twenty years ago that there was only twenty years of oil left”. This is a common belief among uninformed people. It is where the title of this site comes from. It is also inaccurate.

There were indeed many claims during the 1970s that oil would run out by the end of the (20th) century or even before. But there was also a common prediction that oil production would peak around the end of the century, not run out, and this was the view taken by almost all reputable organisations. It now appears this prediction would have been extremely accurate if it wasn’t for the slowdown in production caused by the 1970s oil shocks. People remember the forecasts but, because they are unaware of the difference between oil peaking and oil exhaustion, they assume that all the predictions were that oil would run out by 2000.

One site on oil depletion (http://www.oildepletion.org/roger/index.htm) gives a fascinating chart of oil predictions from the past thirty years.

So many of the forecasts from the past were accurate and remain so. To correct my writer’s quotation:

I remember being told thirty years ago that oil would peak in thirty years. We are now being told again that oil will peak in the next ten years. I wonder if they will be saying “I told you so” in another thirty years!