Guest essay by Mike Jonas
In January 2012, my article “Peak Oil – the R/P Ratio re-visited” was posted on WUWT. 3 years on, maybe this ever-controversial subject is worth (re-)re-visiting.
In 2012, I said that it was getting ever more difficult to increase production, and that I suspected that we were already at or close to Peak Oil, but that it was still mathematically possible that Peak Oil was many years away. Do I still think that? In a way, yes, but … well, read on …
In this article, I look at the major factors affecting oil supply, look at past oil market behaviour and how the future may develop, see what lessons can be learned from Hubbert’s Peak, and speculate on when Peak Oil will occur and what it may feel like.
Some of the controversy generated by the 2012 article came from different interpretations of “Peak Oil”, so this time, I will start with the definition and a bit more background.
Definition of “Peak Oil”
The definition I am using is simply :When the rate of oil production reaches its maximum.
· In this definition:
Peak Oil is not : “when we run out of oil“.
This is not a useful definition, because we’re not going to wake up one morning and find that yesterday’s oil has all gone.
Peak Oil is not : “when we can’t increase the rate of oil production“.
This isn’t very useful either, because “can’t” is always open to argument.
· The period over which the maximum is determined is not specified, so this still isn’t a precise definition. Certainly, any period less than a year is irrelevant. We very likely won’t know when it was until several years afterwards.
· The reason for oil production reaching its maximum is not specified, and possible reasons will be discussed below.
· I am not talking about fossil fuels generally, and I am not talking about oil and gas. I am talking specifically about oil. That does include gas liquids and “unconventional” oil, but not biofuel. (“Unconventional” oil is described later.) s` qA
· Peak Oil is not necessarily a disaster, it could even be a positive. This also will be discussed below.
One idea which surely is not open to argument is the fact that oil production will peak. Here is a long term graph of past and predicted future fossil fuel production – peak production rate is defined as 1, and the red star represents where we are now (or were recently):
Even if those future estimates are very inaccurate, it is inconceivable that fossil fuel production can keep increasing for thousands of years. The graph for oil must fit inside the graph for fossil fuel.
But predicting Peak Oil has always been an unrewarding exercise. As critics love to point out, various people have predicted Peak Oil for over a century and have been wrong every time.
Factors affecting oil supply
The principal factors affecting oil supply are:
- Price “The solution to high prices is… high prices.” – T. Boone Pickens (h/t John Garrett)
Geology is obviously a factor – geology dictates much of where the oilfields are, how much oil is present, its quality, and how difficult it is to extract.
The other factors are all interdependent, so it is arguable how much impact each factor has.
· Many if not most oilfields now belong to nation states, not to oil companies. Consequently many of the decisions affecting oil supply are political. There are also involuntary political impacts such as wars.
· Demand has grown fairly steadily for many decades, due to living standards increasing in developed nations and many other countries accelerating their economic development.
· To the extent that supply is flexible, production tends to be driven by demand.
· Price is the major balancing factor. High oil prices increase the attractiveness of bringing new sources into production, but can also suppress demand. Low oil prices increase the attractiveness of using more, ie, they tend to increase demand, while discouraging high-cost sources of supply.
· Technology is a two-edged sword. Oil production technology has seen tremendous advances since the start of the oil era, allowing ever more oil to be found and produced, thus boosting supply. But technology also helps to make oil use more efficient and other energy sources more competitive, thus reducing demand.
Should I have listed Discovery as a major factor? Maybe I should, but I chose to regard it as a subset of Geology, perhaps with Politics, Demand, Price and Technology as modifying factors.
Most people are by now familiar with Hubbert’s Peak. M King Hubbert’s theory  says that for any given geographical area, from an individual oil-producing region to the planet as a whole, the rate of petroleum production tends to follow a bell-shaped curve. In 1956, Hubbert famously predicted that US oil production would peak in about 1970. His prediction was remarkably accurate over the next 50 years.
Figure 2. Hubbert 1956 prediction vs US Oil Production . 
Predictions that are that accurate over 50 years are quite rare.
For the global bell curve, the upward slope tends to be limited by demand, ie. as much as can be sold at a reasonable price. The downward slope, for regions and fields that have reached it, tends to be limited by geology, ie. the maximum production rate consistent with maintaining reservoir viability. [NB. I’m only referring to general tendency. There are many exceptions of course.]
The overall bell curve is the sum of its components, each of which is a bell curve. Thus a region’s bell curve is the sum of its fields’ bell curves, and each field’s bell curve is the sum of its individual wells’ bell curves. This pattern is demonstrated in, for example:
Note that the “bell curves” for the individual fields have a wide variety of shapes. Note also that once a peak has been passed, even a very large field like Johann Sverdrup (Norway’s 4th-largest discovery at 2.9bn barrels) may only produce a brief delay in the decline.
One notable exception is the USA, where very large quantities of “unconventional” oil have recently become viable thanks mainly to technological advances and high oil prices: This surge in “unconventional” oil production can be seen from around 2007 in Figure 4:
What is “unconventional” oil? A-simple definition is “oil collected by other means than simply drilling for it” . Willis Eschenbach put it differently:
The strange part is, when you open a barrel of unconventional oil to see what conventions were broken in its creation, you find it is indistinguishable from conventional oil. 
Both are right, in a sense, but I’ll stick loosely with the “unconventional” idea simply because it’s a useful word for the oil which even under current technology is much more difficult and expensive to produce than the oil we’re used to that satisfied all demand until recently. The definition is loose, but some pictures  might help:
“Unconventional” oil also includes shale oil, which can be drilled for rather than mined, but tends to be in very tight rock formations that are difficult and expensive to produce from, and tends to require techniques such as horizontal drilling and fracking .
Past Oil Production
Before I show the graph of past oil production, it will be helpful to show the graph of oil price. It’s from data in BP’s June 2014 report , so doesn’t show the recent fall back towards $40.
Figure 6. Oil Price from 1965. 
Looking at the first half of the graph –
· The first major disruption in this period occurred in 1973, when OPEC quadrupled the price of oil.
· The second was the Iranian revolution in 1979 and the Iran-Iraq war starting in 1980.
· These disruptions helped to bring on a massive recession in 1981-2, the most significant recession since the Great Depression.
· Reduced demand because of the recession resulted in the oil price fairly quickly returning to “normal”.
In the second half of the graph:-
· For a decade or so after the start of the 21st century, supply struggled to keep up with demand and the oil price went up …
· … apart from a short sharp dip at the time of the GFC.
· Not shown is the recent sharp price decline, caused by surging US “unconventional” oil production and exacerbated by Saudi Arabia’s decision to maintain high production.
But in spite of the economic booms and busts, it is clear that although the oil price has been volatile at times, oil demand and/or supply have been relatively inelastic:
Figure 7 – Oil Production.
Oil reserves are very much a function of technology and price. A generally-accepted definition  of Reserves, also called Proved Reserves or Proven Reserves, is:
the estimated quantities of oil which geological and engineering data demonstrate with reasonable certainty to be recoverable in future years from known reservoirs under current economic and operating conditions
In my 2012 article, I paid too much attention to proved reserves. That was partly because I was responding to a Willis Eschenbach article about the R/P Ratio (Reserves / Production), but mainly because in my mind I overestimated the importance of Proved Reseserves. What really matters is the Technical Reserve (“TR”, the total amount of oil that might be able to be produced at some future time), rather than the more narrowly defined “Proved Reserves”.
So – how much oil is there? That number is surprisingly difficult to find (well for me it is, anyway). The USA is it seems the only country that tries to estimate it. The U.S. Geological Survey (USGS) puts it at 732 Bbbl excluding the USA (565 conventional, 167 gas liquids) , plus 42 Bbbl in the USA (32 conventional, 10 gas liquids) . But unfortunately USGS says “Unconventional oil and gas resources, such as shale gas, tight oil, tight gas, coalbed gas, heavy oil, oil sands, may be significant around the world, but are not included in these numbers.“.
Total USA shale oil is put at 2,175 Bbbl plus “significant volumes of heavy oil in the oil sands of northeast Utah” . Outside the USA, total unconventional oil is put at 335 to 345 Bbbl, but with much information missing . [Note: The source here is Wikipedia, but (a) the data isn’t particularly controversial so Wiki should be OK, and (b) accuracy isn’t very important, only the general level.]
Total oil production to date is about 1,440 Bbbl .
A possible estimate of global initial TR is therefore about 5,000 Bbbl:
|USA gas liquids||10|
|USA shale etc||2,175|
|Non-USA gas liquids||167|
|Non-USA shale etc||345|
Peak Oil date
Looking back at Hubbert’s Peak in Figure 2, it is striking just how accurate Hubbert was for 50 years, and equally striking that after 50 years of impressive accuracy, it went badly wrong (see the production surge from 2007 in Figure 4). So – what happened, and are there good lessons to be learned?
If you read Hubbert’s paper , you will note that he bases his estimates on very few numbers. The most important number was the total amount of oil. For the graph in Figure 2, total oil in the USA was put at 200 billion barrels (“Bbbl”). Hubbert didn’t actually know what the correct number was, so he estimated for 150 Bbbl and for 200 Bbbl. The graph for 200 Bbbl proved remarkably accurate.
Where Hubbert went wrong (50 years later) was that he did not make proper allowance for the “unconventional” oil, even though he knew of its existence. The combination of technological advance and high oil prices resulted in a massive surge in USA production of “unconventional” oil from around 2007. There was therefore a major departure from Hubbert’s predictions for the USA (Figure 4).
Put simply, all we have to do in order to correct Hubbert’s predictions for the surge in “unconventional” oil is to update the figure for total oil TR. The other principal figures Hubbert used were current production rate and rate of demand growth, and of course new values for those should be used too.
Hubbert put initial global TR (today’s TR plus all past production) at 1,250 Bbbl, and predicted global Peak Oil in 2010, based on the theory that Peak Oil occurs when about half of the TR has been produced. 2013 production rate was around 32 Bbbl per year, so using Hubbert’s theory and our new figure for TR, we can do a rough calculation on an “all other things being equal” basis, of when Peak Oil will occur (see spreadsheet ):
If global initial TR was 5,000 Bbbl and future production growth rate is 3% pa, then Peak Oil occurs around 2036. At lower growth rates (2%, 1%) the date is only slightly later (2038, 2041). The truly fascinating aspect of these figures is that a quadrupling of initial TR from 1,250 to 5,000 Bbbl only adds about 30 years to the likely Peak Oil date.
Over the last few decades, oil production has grown at around 1.3% pa  on average. In spite of the recent surge in USA “unconventional” oil production, global oil production has slowed a bit in recent years to about 1.1% pa. With USA shale/tight oil production expected to decline from 2020 , it may be difficult to maintain a positive global production growth rate from then on. Peak Oil is when the global growth rate hits zero.
The above calculation putting Peak Oil at around 2036 is simplistic, of course. It implicitly contains some big assumptions about future demand, technological advances, massive mining operations from Alabama to Utah, and a few others no doubt. It is probably reasonable to view it as giving an upper limit.
But maybe we can get more realistic if we look at some of the major factors again:
· Politics : Already, fossil fuels are under savage attack from the Greens worldwide. Hopefully, sanity will prevail, but there is clearly some risk of a politically-enforced early Peak Oil date (and peak everything else). That could be very painful indeed, a disaster on a scale that is hard to imagine. On hopefully a lesser scale, political instability in oil-producing countries is a potential threat to oil supply.
· Demand : From 2000 to 2013, production has gone up from 74,983 to 86,808 bpd , an average increase of little over 1% pa. There are increasingly plentiful and cheap supplies of natural gas, which may well eat into the demand for oil. Greater efficiencies, possibly driven by the oil price, may also help to limit demand. However, a period of global cooling, as some think likely, may increase oil demand for heating. Also, a period of renewed global prosperity would surely push up the demand for oil.
· Note: The relationship between demand and Peak Oil date is interesting. In the simple calcs  a higher production growth rate gave an earlier Peak Oil date. But under higher demand, if the oil price is strong (but not so high that it knocks demand down again), it is likely that more of the TR will get produced and hence Peak Oil date could even be later.
· Price : The full incremental cost of oil production – the cost of a barrel from new sources taking up-front capital costs into account – is now probably above $US70/bbl . This view is reinforced by the oil futures market, where oil for delivery in future years is in Contango, ie. higher than the spot price by more than the storage cost . For global oil production to keep increasing, a greater proportion of production will have to come from “unconventional” sources, and that does look likely to push the cost up further. As the cost of oil rises, so alternatives such as natural gas become more competitive.
· Technology : Will technology be able to keep ahead of demand by both (a) reducing the cost of production of “unconventional” oil, and (b) increasing the rate at which it can be produced? If it cannot, then Peak Oil occurs earlier. Technology may also work to reduce demand for oil by, for example, replacing it with gas or electricity for some uses, and by enabling more efficient use of oil.
With all these conflicting factors, the future for oil looks difficult to pin down.
A Peak Oil date cannot, I think, be predicted with any reasonable certainty. Although it is mathematically possible for Peak Oil to occur after say 2040, it looks unlikely to be as late as that, and may well be much earlier.
Major advances in “unconventional” oil technology are needed to forestall Peak Oil for more than a few years. Technology has always delivered in the past, and it is reasonable to assume that it can continue to deliver in future.
But there is another possibility : The price of oil is still likely to increase. If alternative energy sources such as natural gas (or even nuclear or solar energy) are available in sufficient quantity, and if the technology is developed for them to compete for some uses of oil, then Peak Oil occurs early and everyone benefits.
So Peak Oil doesn’t have to be regarded as a disaster. With any luck it will be a massive non-event that no-one even notices.
Prediction is very difficult, especially about the future . So I won’t make any predictions. Instead, I will offer some scenarios (don’t take them too seriously, consider them as using “artistic licence”, a bit like climate model “projections” but hopefully more realistic):
Scenario 1. The marginal cost of oil (driven mainly by “unconventional” oil) stays high, while alternatives such as gas become increasingly available at attractive prices. Heating at first, then transport, start converting to the alternatives. Peak Oil occurs early, but the total amount of energy available just keeps increasing while prices remain fairly stable, so there is net economic benefit.
Scenario 2. US shale oil production increases sharply until 2020, then declines. Together with other new sources of supply, the US shale can more than keep pace with declines in existing fields until 2020, but then it is found that other “unconventional” oil sources simply cannot ramp up production enough to keep pace. Total oil production declines quite sharply. In this scenario, Peak Oil occurs around 2020 and the oil price soars. There is then a painful transition period while other fuels scramble to make up the shortfall.
Scenario 3. Technological advances continue to reduce the marginal cost of “unconventional” oil production and to allow ever-increasing production rates. Technological advances also allow increasingly efficient uses of the oil, fuelling economic advances world-wide. A lot more “unconventional” oil is discovered in various parts of the world, and oil production continues to increase steadily until well past 2040. Oil production peaks around 2050, but gas, coal and nuclear keep the energy supply increasing for many more years after that, while solar technology matures and eventually everything (except possibly air transport??) runs on nuclear and solar.
Scenario 3 should be the most likely, but if our politicians continue to get ever more adept at stuffing things up, then the unpreparedness of scenario 2 may be the best we can hope for.
As always, I will be happy to be proved wrong!
Mike Jonas (MA Maths Oxford UK) retired some years ago after nearly 40 years in I.T.. He worked for BP in the 1960s and 70s, including 3 years in Abu Dhabi.
 OilPrice.com “Declining Fossil Fuel Supplies and the Energy Trap“, Tom Murphy, 31 Oct 2011. http://oilprice.com/Energy/Energy-General/Declining-Fossil-Fuel-Supplies-and-the-Energy-Trap.html
 Data is from: BP Statistical Review of World Energy June 2014.
 BP: Oil Reserve Definitions.
 US Energy Information Administration (EIA): Petroleum & Other Liquids.
 Norway crude oil production 1970-2013. Data is from the Norwegian Petroleum Directorate (NPD).
 heatingoil.com Unconventional Oil Reserves in and Around the US.
 wattsupwiththat.com Conventional Wisdom, Unconventional Oil, Willis Eschenbach, February 2, 2013.
 Paleofuture article We’ve Been Incorrectly Predicting Peak Oil For Over a Century
 The Bakken Shale Oil Field, North Dakota. http://bakkenshale.net/
 USGS Global Estimate for Undiscovered, Technically Recoverable Conventional Oil and Gas Resources US Department of the Interior Press Release.
 USGS U.S. Oil & Gas Reserve Growth Estimates US Department of the Interior Press Release.
 Wikipedia Oil Reserves in the United States.
 Wikipedia Unconventional Prospective Resources.
 Business Insider Australia Breakeven Oil Prices For America’s Shale Basins. Keystone XL Project Draft Supplemental Environmental Impact Statement (Figure 1.4.6-8 Mining Breakeven Oil Price. Note also Figure 1.4.4-8 showing US shale/tight oil production expected to decline after 2020)
 Niels Bohr, from The Quotations Page http://www.quotationspage.com/quote/26159.html
 Futures Magazine Contango crude markets.