Peak Oil since 1956 ~ historical perspective

Revised April 19 2010 ~ There's been alotta talk in the media about Peak Oil, with insinuations that the recent $80/barrel prices were indicative of long term supply concerns. On the contrary, extraction of "all liquid oils" hit a global production record of 86.13-mbd in July 2006. It is hoped that our Depletion Scenarios assist in putting the Peak Oil issue in perspective. As shown in our graphs, present levels of crude extraction are forecast for some time albeit at reduced volumes post peak, but there will be lotsa oil availability well into the next century.

The models estimate an average 3.6 Trillion barrels of oil is economically recoverable. And much of the price spiking in 2005 was based on "a fear of shortages" rather than any bona fide problems. It is more likely that there is presently a shortage of refining capacity in the supply chain.

There are a ton of McDoomer sites out there that take glee in spreading misinformation about "what" will happen "when". Take care when visiting them that selling books and lucrative speaking engagements may have some part in their fear merchantry.

My acid test in determining extravagant claims is the search for their "time lines". If noticeably absent, ask these pundits when noteworthy events are to occur. Reluctance to commit to a time line allows the fraudulent to avoid challenge, application of due diligence and media scrutiny.

In quite an ironic way, it was the studies of Colin Campbell of Ireland and Jean Laherrère of France that illustrated decades of future crude production that we in turn used to dispel wrongful conclusions. At the time we did not realize that this same work was being promoted by the McPeaksters to illustrate that "the end is nigh" when in fact the present work of these geologists foretells that the decline in oil production will be spread over 90 to 130 years.

Marion Hubbert's forecast of a 1995 Peak in global oil production didn't happen; and globally recoverable oil will be threefold his original estimate. But he succeeded in bringing awareness to the realization that oil is a finite natural resource that will some day run out for practical purposes just as will copper, iron, gold, etc.

In the early 80's, future forecasts for oil production were not dramatic. It was thought at the time that hydro and nuclear generation of electricity would satisfy energy needs. As it became obvious that concern over uranium and environmentalist opposition to dams was winning the day, decisions were made to revert to fuelling global economies on the back of coal, gas and oil.

from National Geographic (1974)

In retrospect, the environmentalists won those early battles but in the end lost the war: the present atmospheric 388-ppm concentration of co2 as a greenhouse gas is directly related to the blocking of certain energy sources (nuclear & hydro) and the subsequent continued dependence on fossil fuels. Nuclear is again being touted as the saviour, but the lost lead time will result in great threats to the planet's ecology. And since the mid 90's, projections for future production of coal, natural gas and oil has skyrocketed.

Hubbert's purpose in questioning ultimate coal, gas and oil volumes was a foundation for his promotion of the Nuclear alternative. It seems that we have come full circle.

While a cornucopia of URR estimates has existed since the beginnings of the 20th Century, it was commonly expressed as a function of Reserves to annual Production ratio. For decades, BP gave us some comfort in telling us that there was sufficient oil Reserves for 40 years. But it was early graph presentations by Hubbert, Ivanhoe, Colin Campbell & Jean Laherrère that awakened discussion to the realization that oil doesn't just run out in the Year 41 - rather there is a crest followed by decades of declining extraction rates.

Marion King Hubbert started the Peak Oil discussion with his forecasts of upcoming Peaks in the USA & globally. Above, his 1974 2000-Gb URR and 111-mbd 1995 Peak Rate & Peak Date were in error, but did serve to commence the debate. For history buffs, MK Hubbert's forecasts start in 1956 (API Conference) with a forecast of a Texas Peak of 3.15-mbd in 1962. Then in 1971 forecasts (in Scientific American) that the Lower 48 had likely peaked in 1966 with a URR of 165-Gb & a Global Peak of 101-mbd in 1995 with a URR of 2000-Gb. Below, a 1997 depiction gby Ivanhoe for 75-mbd Peak in 2010 with 2273-Gb URR.

With hydro and nuclear generation pushed to the sidelines and dependence on oil back in play, concerned geologists were raising the prospect that Hubbert's 1956 time lines had to re-examined. It was found that altho Nuclear projects were making headway, Hubbert had continued his interest in questioning whether the globe's oil endowment would suffice demand in coming decades.

Too often, we saw projections that had open ended targets with no sign of production abatement. To the credit of concerned geologists, they followed Hubbert's lead with decline rates based on exhausting known URR. This allowed stakeholders, govt's & decision makers within many disciplines to view a roadmap of the future that gave some perspective to a finite resource.

That's not to say that there was not some folly in using these early depictions as a guide. They were very wrong in their estimates of both Peak Rate, Peak Date & the URR which determined ultimate exhaustion.

One of the earliest dates for Peak Oil was M King Hubbert's 1995 forecast (published in National Geographic in 1976). In 1991, Colin Campbell of Ireland suggested that the decline in oil production would commence in 1998 after a multi-year plateau. In 1995, Jean Laherrère of France forecast Peak Oil in Y2k. A common denominator among these early projections was conservatism of their URR estimates (2000-Gb. 1650-Gb & 1750-Gb respectively). In retrospect, today's TrendLines AVG of 20 recognized estimates is 3536-Gb.

These early attempts of reconciling URR with long term Production Estimates were shortly thereafter adopted by energy agencies and the oil companies in their annual Outlooks and Scenarios. IEA was the first with adoption of a 2300-Gb URR and 2020 Peak Date in its 1998 WEO. Similarly in Y2k, EIA suggested a 2016 Peak Date and 3003-Gb URR. With IEA estimating a 5-Tb URR today and EIA's 4-Tb, both agencies see Peak Date pushed back to 2030.

An unfortunate revelation is that the long term projections for most agencies and the oilco's are based on demand driven factors such as projected Real GDP Growth. This contradicts their short and medium term Outlooks that are based on bottom-up studies of flow extraction, production and refinery capacity net of field decline rates.

Above, a comparison of some early extraction projections. The shining star is Michael Lynch ('96) with a 84-mbd forecast for 2006. Ten years later is was actually 85-mbd! IEA predicted 80-mbd & EIA/DOE with 76-mbd ... all three came within 10%. The '91 target for 2006 by Colin Campbell was 50-mbd with Decline commencing in 1998 with a URR of 1650-Gb; his '97 target for 2006 was 64-mbd with decline starting in 2009 and URR of 1780-Gb. Below, a 2700-Gb URR by Jean Laherrère in 1996 with a 2008 forecast of 85-mbd & 2010 Peak Rate of 86-mbd.

We know by the current URR estimates that there is plenty of oil for the next 100 to 150 years. But it may be a leap of faith to assume that the Peak Rate will occur simultaneous with the half way point in consumption as suggested by Hubbert as we become more dependent on non-conventional sourced oils.

As we can see in one of Jean Laherrère's early depictions, while the Conventional Hubbert idealized Curve (green line) is quite symmetric, the All Liquids idealized Curve (purple line) has a much less aggressive Decline Rate reflecting the extraction challenges of the non-conventional component. If true, the extended exhaustion of resource bodes well for implementing alternative energy sources and substitution practices in the many sectors that will be affected by competition for the annually declining supply.

Colin Campbell, ASPO & the Depletion Model

Colin Campbell of Ireland has been forecasting the Peak of Oil extraction since 1991. His early studies projected a global oil decline commencing in 1998. Assuming a Hubbert Curve Model application, his Peak Date becomes a moving target as annual information comes in from oil producing nations on their changing Reserves, past consumption and future Resources. Thus, his Peak Oil Date has enjoyed a range from 1998 to 2012. It is currently 2010.

Campbell's 1991 Depletion Model can be seen amid several early Outlooks in the above discussion. At left, we see its 1996 hybrid with four scenarios. In later versions, Colin would divide crude into two categories:

a) Regular Conventional crude including traditional extraction and condensate

b) All Liquids incl Conventional plus Non-conventional crude including heavy, polar, deep water, natural gas liquids and processing gains.

It is this second multi-component definition that is commonly used in projections by the oilco's & energy agencies and in the mainstream media. The former is a study best left to academia and purists as its tracking has little to do with the dynamics in play in today's marketplace. When one buys petroleum by-product, nobody cares where it came from or its extraction cost. In the end, it all goes into the same pool.

This early 1996 effort appears to be Colin Campbell's first graphic depiction; two scenarios with differing URRs.

At left, one of Colin Campbell's early depictions of Peak Oil. In late 1996, a plateau was envisioned from 1998 to 2012 amid a URR of 1750-Gb. Below, we note that by 1997 the plateau was narrowed to '99-2008 with 1780-Gb URR.

At left, we see a Campbell/Laherrère collaboration that appeared in Scientific American. TrendLines has discovered that while the March 1998 article calls this graph their "All Liquids" version, our analysis of its URR clearly shows that its URR of 1800-Gb does not reconcile with the article's All Liquids URR of 2500-Gb but rather matches exactly with their Conventional estimate of 1800-Gb. "Oops!" The chart reflects Laherrère's logistic methodology ... not the 2000's plateau akin to Campbell ('97 & '99).

Below, we see the two graphs that should have been in the Sci-Am article. "Scenarios" reflects Campbell's view of Regular Conventional plateau-style Peak and subsequent Decline. "All Hydrocarbons" puts forth his proposal of for conventional oil, the non-conventionals and natural gas. The All Liquids component (below the orange Gas layer) is a representation of a (34-GbA) 93-mbd Peak Rate in 2009 via a 2600-Gb URR. This Outlook, part of Colin Campbell's House of Commons committee testimony in July 1999 is significant in that it's 93-mbd Peak Rate was the highest of his career (until July 2007).

The article reflects Campbell's being overly consumed by the imminent peaking of Regular Conventional Crude (RCC) and his alarmist tendencies surrounding that event. Only in the final paragraphs did the article admit EIA's projection for All Liquids of (40-GbA) 110-mbd in 2020 ... with no apparent peak in sight. The express tone was that the imminent peak of RCC would cause crisis level crude prices and foreshadow a peak of All Liquids in 12 years. But in real 1997 USDollars, today's crude price is quite reasonable. It is now 2009 and consensus for an All Liquids decline is that this event is still 14 years away...

Global production today is 84-mbd. The consensus for URR is 3.8-Tb, compared to Campbell's 2.6-Tb. Michael Lynch's 1997 position that rising crude prices would translate to ever increasing Reserves has come to fruition.

RCC (light sweet) is a very narrowly defined oil, one of eleven streams that comprise All Liquids. It was 84% of total production in 1998. Today it is 73%, and will be a mere 53% in 2030. Unfortunately, Campbell misread the growing future importance of NGL & the non-conventionals. To be fair, his stalwart monitoring of RCC was instructive: it Peaked in 2005 @ 68mbd!

Like the Hubbert National Geographic presentation in 1974, this Scientific American feature was priceless in its energy awareness value...

Above, Conventional Oil Peak Plateau (red) of 64-mbd from 2001 to 2008 with 1800-Gb URR. And also from July 1999, the Campbell "All Liquids" Peak of 93-mbd in 2009 with 2600-Gb URR.

But from the optimistic outlook of 1999, Campbell somehow got on a pessimistic streak that worsened 'til Summer 2004. By then his Peak Rate forecast had fallen from the 93-mbd in 2009 (above) to a mere 80-mbd Peak scheduled to hit in 2006.

The Achilles heel was his failure to accept the magnitude of growing Russian production and the newfound non-conventional sources. In his original work, Colin had been tracking conventional crude (and condensate). He wanted that his precise documentation of past and future production would finally validate Hubbert's 1956 prediction of a global peak of oil in Y2K. In Nov/2001 Campbell professed: "Russian production forecasts are a myth. So, barring large scale discovery or development of a secret cache of finds, this output surge will last for two or maybe three years before being overwhelmed by expanding local demand and the return of the decline curve."

Well, Russia went on to actually surpass Saudi Arabia in annual production to become the globe's numero uno producer of oil. And worse, the new "all liquids" definition (heavy, polar, deep water, gas liquids and processing gains) attained quite substantial and unforeseen volumes with those components now making up 12% of supply.

Fortunately, he seems to have reflected and come to a realization that perhaps the "integrity" of his data releases are ultimately more important than the perception of that his bias was pushing an Agenda.

Similar to the Laherrère graphic above, this March Y2K depiction of the Conventional Peak (from the Feasta Conference) illustrates their hypothesis that a plateau-type peak (grey line) could soften the decline slope and extend the decline tail; as differentiated from the Hubbert idealized Curve (black line).

This 2001 depiction was a successor to the 1999 version above and has lived on to become the "poster child" of the Campbell/ASPO Depletion Scenario Model that was introduced in the ASPO Newsletter of May 2002.

This Newsletter Table has become the basis for the Campbell/ASPO Depletion Model. In the recent past it was fraught with arithmetic errors that led to wrong conclusions by those that depended upon them. Some errors were of the 6-mbd magnitude. In 2004 & 2005, i was ousted from several mailing lists and forums that were disturbed by my accusations that Campbell's numbers were seriously flawed and that it was my opinion that he required immediate assistance with the DM if it was to sustain the rigours of media and analyst scrutiny.

TrendLines uses this data to forecast the Hubbert Peak of Regular Conventional Oil and the Peak Oil for All Liquids. When we published that the Hubbert Peak has passed unceremoniously in the Spring of 2004, ASPO was still looking forward to a 2006 Peak. Again a mea culpa occurred and in August 2005 the Newsletter was corrected (see left) to illustrate a 2004 Peak and they kindly mentioned this website as a source for those seeking a Campbell/ASPO retrospect and development of their Depletion Model and its revisions.

The Campbell methodology includes the continual update of a global database of national figures wrt past consumption, current reserves and future resource for both Regular Conventional Oil and All Liquids. These annual, quarterly and monthly amendments can result in changes to URR that can move the respective Peak Dates of the two categories substantially. Since early 2004, the date for Conventional Peak Oil has been a moving target ranging from Y2k to 2006. Colin Campbell is one of several Peak Oil Depletion Scenario modellers that share their spreadsheets, data or results with TrendLines. By our 2007 analysis of the ASPO figures, Conventional Peak Oil occurred in April 2005 and is undergoing a 2.5% Decline. Do not be alarmed. Conventional Oil is narrowly defined and this particular Peak is monitored only by academia and purists.

Meanwhile, the one that affects us all in our daily life, ASPO's All Liquids Extraction Peak is forecast in 2010. Unless there is a plateau, from that point the volume of oil available for consumption will decline annually by a yet unknown factor, but estimated at approx 2.6% by Campbell. While the Production Peak and half-way point in consumption seem to have occurred in the same year (2005) for Conventional Oil, it appears that the half-way crossover for All Liquids will be in October 2012. In theory, this reflects that while there are over 1.4 trillion barrels of oil still to come out of the ground, the tar sands or below the oceans, the extraction will be with increased technical difficulty and expense.

I consider Campbell's coming to grips with reality in 2004 somewhat of an "epiphany moment" for him. The shame cast aside, he could now freely disseminate his current research without accusations of purposeful tainting. If not "then", well "when"? And we all watched with anticipation his subsequent ASPO newsletters.

I had concerns with some ambiguities in the published work and was extremely fortunate to be one of few with whom he shared his documentation. My position was that, aside from the plethora of arithmetic errors, we were indeed witnessing the passing of the Hubbert Curve in early 2004. But in April 2004, Campbell was mysteriously showing Conventional or Regular Peak Oil was pushed ahead to 2005 and all Liquids moved back to 2007. I shortly found that whispering "the emperor has no clothes" is grounds for exodus from Peaksterville. Alas, my whispers got much louder. Over the ensuing months we saw the Peaks move to and fro, sometimes with rationalization ... and sometimes not. Tho his arithmetic may not be the best, new information is introduced in the ASPO Monthly Newsletters. A consequence is that the extraction Peak Rate and Peak Year thus move around like a yo-yo; but that is the nature of dealing with the scientific method. Further, folks must understand that Colin's work actually relates to four peaks. The Hubbert extraction peak and the half-way consumption Peak of Regular Oil. And those same two peaks for "all liquids".

The perennial ultra conservative ASPO Projection is compared to Jean Laherrère and other analysts at the top of our Peak Oil Issues page.

Historic Tracking of (ASPO-IE) Colin Campbell Depletion Model: 1989-2011

March 28 2012 delayed FreeVenue public release of Dec 27th MemberVenue guidance ~ Today's update adds Colin Campbell's May/2011 Outlook. It re-confirms his position All Liquids peaked @ 85-mbd in 2008 (despite EIA data to the contrary) and is founded on a 2,52334-Gb URR (up 89-Gb from last year). The chart tracks all the production profile revisions over his career. Its forecasts of Peak Year have ranged from 1989 to 2012. In fact, December marks the 22nd anniversary of Campbell's initial All Liquids declaration that oil had indeed peaked. To be accurate ... a sub-peak. In Dec/1989, he declared All Liquids production had reached its physical limits @ 66-mbd and would never again attain the 67-Mbd Peak back in 1979.

Campbell's estimates for Peak Rate span from that virgin call of a 66 Mbd sub-peak in 1989 to his 2008 forecast of a 97 Mbd peak in 2010. His underlying All Liquids URR estimates range from 1575-Gb (1989) to 2900-Gb (2002). TRENDLiners may have notice my last three annual chart revisions have excluded Campbell's 1991, 1996, 1997 & 1998 projections. I determined those studies forecast Regular Conventional Oil ... not All Liquids, and only led to unnecessary confusion. His current (2011) forecast for RCO can be compared to the only three other such projections for light sweet crude at my Scenarios venue.

The highlighted years of distinction are: 2008 (highest peak 97-Mbd), 2002 (2900-Gb URR high), 2011 (current update), 2004 (Colin Campbell's dark days call: 80-Mbd peak coming in 2006) & 1989 (Campbell's initial 66-Mbd scenario which declared that All Liquids would never breach its 1979 record). Because the Depletion Model newsletter graphic ends in 2050, it was not readily apparent that five of Campbell's early All Liquids projections failed to exhaust his designated URR. The 300-yr outlook resolution view of this chart exposes the errant methodology of the Depletion Model in 1999, Y2k, 2002, 2003 & 2004. These profiles have been corrected via compensating plateaus or "doglegs".

See how the 2010 ASPO Depletion Model measures up against other failed outlooks in our Invalidated Scenarios presentation & compared agin Tier-1 URR estimates.

click here to see how the latest (2011) Campbell Depletion Model measures up against the only other three studies addressing Regular Conventional Oil (light sweet crude)

May 5 2011 delayed FreeVenue public release of Dec 27 2010 guidance @ our MemberVenue ~ Since 1996, ASPO has been selling the illusion above. There purpose is to fuel gloom merchantry. Until we exposed their agenda in 2007, nobody knew the magnitude or consequences of backdating, nor that the "Future Discovery" omitted 368-Gb of non-conventional resource. For some perspective, Trendlines Research offers the reality check below. Same data. Very different future. Click here for our opinion on ASPO's agenda-driven backdating strategy.

The EIA & Peak Oil

We mentioned above that IEA published its first feature on the topic of Peak Oil Depletion in 1998. With a URR of 2300-Gb, it forecast a Peak Rate of 112-mbd in 2020.

At left, we see the EIA's first addressing of Peak Oil, back in Y2k. Among other scenarios, this graph foretells Peak Oil in 2016 with a Peak Rate of 96-mbd via a 3003-Gb URR.

While it portrays a Decline Rate of 2%, it came with dire warnings that if Peak Rate was higher, the Decline Rate would likely be in the order of 6 to 10%.

To measure progress, 2006 supply was actually 31-mbd, 2-mbd ahead of that date's target on this graph. While the forecast is accurate thus far, EIA has since reduced its High Price Scenario for 2016 to 34-Gb/yr (93-mbd). Because EIA's URR has increased to 4036-Gb with time, their Peak Rate is now 102-mbd in 2030.

EIA's reliance on Saudi Arabia to "make-up" the void on its long term Outlooks is gradually being based on geology rather than Demand targets.

Lynch 2006

Jean Laherrère's view of Peak Oil

Jean Laherrère of France has been a practitioner of linearization models to project URR for two decades. His most recent version illustrates that since 2005 a new paradigm is apparent, indicating URR may be in the plus 3-Tb domain.

Since past consumption is only 1074-Gb and annual usage is 31-Gb/yr, Hubbert curve theory would lead us to believe that the half-way consumption point will be in Autumn 2020. We do not know for certain, however, that this date will also be the Peak Production Date. The "low hanging fruit" was picked during the "Cheap Oil Age".

As we enter significant extraction of non conventional from this point, there is no guarantee that traditional production rates can be maintained. If that is the case, the Peak Oil will precede the half-way point and the decline curve will take a less aggressive slope than we have been seeing with regular oil fields. The USA saw a 6% Decline Rate after its 1970 Peak. Globally we may see something in the order of 2% if the Hubbert Peak strikes early and/or Reserve Growth continues at its present horrid pace of 160-Gb/yr. Either way, there seems to be lotsa oil forthcoming until at least 2125.

Some noteworthy misses and musings:

In '56, M King Hubbert forecast that the USA's Lower48 Peak would be in 1968. He was only two years out in this attempt.

In '71, he moved that date to Jan 1st 1967. We know today that he was mistaken by three years.

MK Hubbert appeared before a US Senate Committee in '74 proposing that Peak Rate had likely passed by "unceremoniously" in the Lower48 on July 1st 1965. The base graph used has been overlaid with the actual production to 1971. Note the rounded Hubbert Peak to its left.

Thus, in 1974 MK Hubbert believed his data showed a Peak in 1965. He was mistaken by five years in his final attempt.

Linearization is a great guide; but it is not a predictor of Peak due to its gross resolution.

While MK Hubbert's '74 Testimony had its date failing (six years), going back to the 1956 USA Lower48 prediction yields similar Peak Rate and Peak Date errors.

desousa tod 70221

This is a 1974 graph by Hubbert showing a 111-mbd Peak in 1995 with 2-Tb URR. He reconfirmed this in 1976 publishings.

This 1999 effort by James MacKenzie of the World Resources Institute includes three scenarios for Regular Conventional Crude. Even its lowest projection of 77-mbd in 2007 for 1800-Gb was somewhat optimistic compared to the actual unfolding: 68 in 2005.

Pierre-René Bauquis, presently with IFP, provides this 2002 view of his own 1999 Outlook plus those of Colin Campbell, IEA & Shell 2000.

His 97-mbd Peak in 2020 is still valid. He has since raised it to 100.

As an intuitive exercise, if i plot the blue Hubbert Curve as shown as if there was no Middle East crisis, with its 1995 Peak of 111-mbd, it yields a conventional oil URR of 1.85-Tb. It is constructed with both a 4% growth & decline rate. This compares very well with Colin Campbell's present 1900-Gb determination.

Lynch 2006

Top

Peak Oil since 1956

Peak Oil Home • Peak Oil Since 1956 • URR/EUR • World Production Records • Scenarios • Freddy Hutter's Peak Scenario-2500 • Barrel Meter • Gas Pump • Saudi Arabia Outlook • NPC submission

Top FreeVenue Home • Peak Oil • Economics • Climate Change • Elections
Peak Oil @ FreeVenue