Modeling Oil Depletion Using EIA Data - The Tiger Chasing its Tail?

There are many potential definitions of Peak Oil: peak conventional, peak non-opec, date of maximum production, date of 50% Qt, peak liquids etc. We care about Peak Oil because we want to approximate the timing and estimated quantity of less liquid fuel available to a globally connected society in order to initiate mitigation and planning. Typically, this site has used EIA data to puzzle out the 'cigar' posts with moving averages, new production records (Dec 2005), etc. The March EIA Report came out recently and showed March daily production of 84,047,000 barrels per day, a decrease from the heretofore peak (Dec 2005) of 652,000 barrels per day.

On this site and others, we spend a lot of time dissecting the monthly and yearly production numbers looking for signs of a peak. But it appears that unless the EIA changes their definitions, what we are currently calling "Peak Oil" will be obfuscated (and delayed) by increasing amounts of alternative energies that are definitionally included as 'oil' in the headline number.

As long as we use EIA production numbers as the benchmark, Peak Oil will silently morph into Peak Liquids. This is relevant because the definitional layers we add on top of 'crude oil' are not equal in what they provide to society. It is also relevant in that the logistical heuristic used by M. King Hubbert was not intended to include corn and sugar cane derived ethanol in its predictive theory of oil basin depletion. The concept of Peak Oil, already not widely believed, will start to be very confusing. In essence, we need to either a) eventually adjust EIA numbers to exclude growing biofuel and coal liquid inputs or b) recognize that for practical 'peak oil societal impact' purposes, we really do primarily care about 'peak net liquid fuel', which would require categorical adjustments and a bit of algebra.

The EIA defines Oil Supply as the "production of crude oil (including lease condensate), natural gas plant liquids, and other liquids, and refinery processing gain (loss)". This includes the tar sands production in Alberta but not the Orinoco sands in Venezuela.

Other Liquids are defined as "Ethanol, liquids produced from coal and oil shale, non-oil inputs to methyl tertiary butyl ether (MTBE), Orimulsion, and other hydrocarbons." And finally, other hydrocarbons are defined as "Materials received by a refinery and consumed as a raw material. Includes hydrogen, coal tar derivatives, gilsonite, and natural gas received by the refinery for reforming into hydrogen." Although currently a small number primarily in Europe, biodiesel feedstock presumably is a refinery 'raw material' (I am uncertain whether the 26 million tons of palm oil from Indonesia and Malaysia is included in the EIA data and have assumed that it is not)

Last year, the world produced over 12 billion gallons of ethanol. At 42 gallons per barrel, that equates to 290 million barrels. The world (specifically South Africa) also produced 160,000 bpd of Coal-to-liquids which was 58 million barrels for the year. So of the 30.66 billion barrels of 'oil' produced in 2005, about 350 million of them were ethanol and coal derived, or about 1.2%.

1.2% is not a big number (yet). But keep in mind that total oil production for 2004 was 30.3 billion so the 'record setting' 2005 was higher by about the amount of ethanol and CTL. Ethanol production is up 300% since 1997 and has doubled in just the last 4 years. Continued growth at a similar pace will begin to make a meaningful contribution to the EIA totals. If ethanol advocates and alternative energy proponents get their way, these numbers will increase dramatically, and certainly outpace increases in oil production on a percentage basis. Also, the Hirsch and Bezdek report (slide 84) assume five 100,000 bpd coal-to-liquid plants a year starting in year 4 of a mitigation scenario. One could argue we are in the early stages of this now as numerous corporations are exploring syncrude options from coal. CTL is a liquid fuel, but it is not oil.

This post has 2 main points:

  1. Natural gas plant liquids (NGPL) counted as 1 for 1 in the headline crude oil production figure, only have 60% of the energy content of crude oil. Also, irrespective of ethanols energy input/output balance, ethanol only has less than 70% of the BTU per gallon as gasoline. It also requires liquid fuels (in addition to natural gas for fertilizer and nat gas or coal to steam) to harvest, transport and distribute. So the 290 million barrels of ethanol made in 2005 actually used quite a bit of oil to produce. Was the EIA correct to increase crude oil production from 30.37 GB to 30.66 GB for 2005 because of ethanol? If we are tracking "Peak Oil", or the maximum amount of oil available to a growing, demanding planet, we should care about the net (I'll save the net energy/EROI argument for a subsequent post - here I just mean net liquid fuels).

    Clearly, real businesses and individuals used some of the 30.37 GB of non-ethanol oil to produce the ethanol. So the greater the % of 'indirect oil' in the EIA totals, the more we scale up both the assets and liabilities of the planets energy balance sheet. (an extreme example would be an EROI of 1:1 for ethanol (with all energy inputs being liquids), then we might produce 50 mbpd of crude oil and 50 mbpd of ethanol for a total of 100 mbpd of EIA 'oil'. In this example however, every single drop of crude oil would be needed to produce the ethanol, so society would only have 50 mbpd of production (ethanol) at a 70% BTU rating. Obviously non-sensical. This is an extreme example, but at what point from it and our current situation (1.2%) does it become important...? (*In this example, notice that the reported production would be 100mbpd but only 50mbpd of ethanol would be available to society - at that moment we couldn't just instantaneously choose the *better* crude oil and abandon the ethanol because infrastructure would already be in place - there is a large societal time lag for Product A to Product B transformation decisions)

    Energy is the ability to do work. Our society is highly dependent on liquid fuels to do work. To trade one form of liquid fuel for another and include them both in production doesnt make sense.

  2. Though it hasnt mattered a whole lot until now, using EIA data for predictive heuristics based on geological regions may be inappropriate, as it increasinlgy includes biofuels and CTL (as well as orimulsion). The EIA can do whatever they want- they don't care if their reporting of liquid fuels conforms to M King Hubberts theories, but should those of us who study and care about Peak Oil really include corn ethanol, soy diesel and Fischer-tropsch CTL in future 'oil' analyses?
I guess another way to look at it is once Hubberts theory is validated globally, at that point all we WILL care about is net liquid fuels available to society, not whether Hubbert or anyone else was right. Said differently, if the EIA took ethanol out of their oil production numbers, I wouldnt have had to make this post.
I guess another way to look at it is once Hubberts theory is validated globally, at that point all we WILL care about is net liquid fuels available to society, not whether Hubbert or anyone else was right.

Lately, Jean Laherrere try to apply the Hubbert approach on the forecasting of the "all liquids" peak that could include conventional oil, CTL, biomass liquids, etc.:

In 30 years petroleum will have become a little-used energy source (pdf file)

I personally find his article not very convincing.

It looks to me like an exercise in speculative curve fitting (extrapolating multiple overlapping Gaussian curves based on scant data).
Personally I find this article very convincing.

First of all Laherrère has data individually for each fuel source, something neither IEA nor EIA show us (at least monthly).

Secondly, Laherrère uses two different scenarios, a probable and an optimistic one (like Hubbert did in 1956). What you get from modeling things this way is peak around 2015, either with 1 TB or with 2 TB for all liquids minus crude.

Another thing worth mentioning is this: Laherrère, Deffeyes and Campbell are all putting the 50% Qt mark for Conventional Oil in 2005. And each of them is using different source data: Laherrère - IHS, Deffeyes - O&GJ, Campbell - Exxon(?).

March06
Does EIA publish its numbers like this?

By the way, cool blog.

Unfortunately not. Their spreadsheet design seems to be from another era. I had to play with presenting their data for my own ability to understand what was going on.
Terrific post, and I look forward to the comments of Stuart, WesTexas and Khebab, and other HL experts and partisans to their responses.
So, we're definitely past Peak Oil then, once you factor the extra 1.2% out?  Westexas has convinced me that we are probably at or past peak.  This 1.2% makes a big difference, doesn't it?
The South African CTL plant has stable production AFAIK, so that has no impact on the curve.  And I am unsure about the growth curve of ethanol.  Surely spiking in the US, but I thought Brazilian was level to downward recently ?

So, not enough CHANGE in "other liquids" to affect the curve that much.  Maybe in the future.

Also tar sands are going up and they are a "non-Hubbardian" oil.

I think we should keep track of both, if possible.  For different goals, as you note.  Testing the Hubbert model vs. keeping an eye on what really matters in a practical sense.
I second that.
and a 3rd...
Two comments. First, while you're right that the limited EROEI for ethanol means that we are overestimating the "net" production from the energy sector, this is nothing new. All fuels take energy to create them. Some oil produced in some regions has rather poor EROEI. Takes considerable energy to drill in the North Sea or to awaken a played-out field with tertiary recovery. Trying to measure this for each different energy source and derate the production figures to calculate net energy would be completely impractical.

But second, look at this from the peak oil skeptic perspective. Skeptics have long predicted that as oil became scarce and prices rose, we'd see several effects. One is moderation of demand, which is coming into play now. The second is increased production of conventional oil, which may or may not be happening. And the third is increased production of alternatives, which is definitely going on, with tar sands, coal to liquid, work on biodiesel, and of course ethanol.

All these are the reasons why skeptics said that the core predictions of extreme peak oilers, of social and economic collapse, "die off", all these apocalyptic scenarios, were false. Skeptics predicted that society would adapt and find alternatives. And this is reflected in the issue you raise. As we continue to develop alternative liquid fuels, a higher precentage of our total production will be in this form. If skeptics are right, this will allow us to smoothly adjust to the decreased availability of conventional oil by finding ever more efficient ways of producing these alternatives. The magic of the market will do its work and we can sail right through the classical Hubbert peak without even noticing it.

Yes, but there is strong evidence that the total liquids supply (2005-2006) will never be exceeded, no matter what alternatives are developed. Civilization is not going to end, just some aspects of it (suburbia).
I would say the evidence is pretty weak. We don't even have good evidence that crude has peaked, and aspo thinks this won't happen until 2010. And, regardless of hydrocarbon inputs, ethanol and biodiesel are in their infancy and are obviously surging. Meanwhile, ctl and gtl is beginning, lot least in qatar.
otoh, demand continues, especially in asia but here, too. Prices seem unlikely to decline much in the near future.
Actually ASPO predicts that regular oil peaked last year (2005) but including heavy (I assume that includes tar sands), deepwater, polar and NGL takes it out to 2010.

Which measure equates to crude? Neither... Crude is probably regular plus deepwater and polar - which wouldn't be anything like as late as 2010.

An interesting debate we are not having is whether suburbia must end or not.  While it seems clear that suburbia cannot survive using the current technology, it is much less clear whether it is cheaper (measured using either money or energy) to replace suburbia with something else, or to change suburban technology.  In the US, the current population breakdown is roughly 50/25/25 for suburban, urban and rural; getting rid of suburbia requires replacing 50% of the housing stock, an enormous undertaking by either measure.  I personally suspect that it will turn out that converting the suburbs to run on electricity, improving efficiency to the necessary degree, and generating the needed amounts of electricity, will be the more successful approach.
getting rid of suburbia requires replacing 50% of the housing stock, an enormous undertaking by either measure.

No, it doesn't.  My bet is that what will happen is people will crowd together.  It's only very recently that the 2,000 or 3,000 or 5,000 SF home has become the standard for a family of three or four.  My parents had 4-5 siblings, and grew up in two or three bedroom houses.  With one bathroom.  Amazing but true.  

When people have lost their jobs or can't afford gasoline to drive to work, the last thing they are going to be doing is buying new homes.  No, instead they will be moving in with friends or relatives.  The extended family living together in a fairly small house used to be the norm.  It will be again.  

Already, people in my office with long commutes are hinting about renting the spare room in my small apartment.  (I live walking distance from the office.)  

The half-million dollar exurban house across from my sister's sat on the market for quite a while. Now there are three families living there. They all have fancy vehicles, and over Memorial Day we saw that all their friends do, too.

The inner burbs must increase in density for those that rely on feet, bikes or public transport. Zoning and codes are obstacles to density, but they can be adjusted - formally or informally. I'm wondering about infilling between houses. I used to draw townhouses, and the site planners always left addresses for the spaces between row house blocks. I couldn't imagine anyone filling in back then, but I can now.

I'm also wondering which buildings will be adapted to serve as local schools for kids too far from the mammoth schools we've been putting up.

I could see the outer suburbs becoming a spacious paradise for those that can still afford to drive that far. Exurban highways might become less-traveled and convenient for long commutes, but maintaining them will be expensive. Alan would want money for urban rail while Mr. Big would want the highways kept smooth for his Rolls-Skoda.

I've often thought how absurd it is in the historical context for me, my sister and my parents to all live in the same city and yet not live together. If everything fell apart for my family's finances, we would simply all move "home" back in with my parents. It would be quite an adjustment, but we would save tons of money and energy.

I suspect that many of the empty nesters will see their kids coming home out of necessity. This will boost communal meals, carpooling, per capita heating/lighting efficiency, etc. That is another way of demand destruction. It's very common in Europe to have kids staying at home until they get married or well into their 30s.

The single person household will be one of the first causalties of PO.

The single household is the biggest waste of resources out there.  Rather than have one TV, you have to have 3-4 TVs (depending on family size).  Everything else is duplicated the same way.  I believe that people will cluster back together in the future, but not just because of peak oil, also because of the economic impact of globalization.  
Hmm the family as the "ultimate communal" living arrangement?  Just like the hispanics locally.  We cannot build an entirly new society.  We will be stuffing more people into those houses in the future- take that to the bank per se.  This is just mass-transit-poorer american housing style- more people to share the bills with make it affordable and more economical (per person).  To think that we are going to build a whole bunch of stuff is a historically short term American view.  Look at all the other civilizations that exist and convert by recycling old structures into new ones. Look at what poor people do- not the ones on the government dole - economics in action.
My God...perhaps if families were "forced" to live together, they would have to get to know their grandparents, uncles, cousins...what a horrible fate.
Since 1950, the average size of an American home has increased by 140%, yet in the same time period the average size of an American household has decreased by a whole person.  Now, we spend only 16% of our time in the same room with another family member.

Interesting reading material:
http://realestate.msn.com/improve/Article.aspx?cp-documentid=353659
http://www.post-gazette.com/pg/05080/474759.stm
http://www.ci.cambridge.ma.us/~cdd/data/demo/city/hhsize_2000.html

Nice articles...thanks for sharing.
My pleasure.
Not an accurate reflection.  The previous generations in the United States had more children.  This changed around 1900 or so.  Like all other developed countries, US residents have moved toward a declining population, we just keep expanding due to immigrants and their families which still tend to be larger, but after a generation they will follow the same trend.  This is the market at work, on a very large scale.  Whereas having many children used to be an advantage, it simply is too expensive to have many children anymore and is in fact a disadvantage.  

The advent of suburbia is a result of more land (and secondarily resources) available for a smaller number of people.  Granted, US population has still increased overall, but our surplus of land still remains.  

Your concept of the extended family moving in together is reasonable, but I wonder whether you have taken into account the counterbalance that as it gets more expensive to commute long distances, suburban houses will become cheaper, ironically being a great place to live for those who don't have to commute.  We're not just going to tear the suburbs down.  And neither can current urban environments support all of the people living in suburbia (even if they move in with family members, there are limits).  I guess we could live like people in India or something, but I find that to be unlikely.  

I think we'll see a move toward some consolidation.  Suburbs that are an extreme distance away from urban centers will languish and die (I guess those who can afford to live out there will be able to have a massive estate, rather than a small 1/4 acre plot), but those within reasonable distance 20 miles or so, will not.  Greatly we're going to see increases in transportation efficiency.  We can improve our transportation system by a lot and reduce a lot of our use of fossil fuels.  People haven't even really started to move away from gas guzzlers.  We clearly have much room for improvement.  

Will we soon be talking about "peak suburbia?"

Much, if not most, of the new housing in the US is suburban or exurban, but there are already some counter-trends happening around the country. A number of cities, especially along the West Coast and in New York and Chicago, are seeing lots of new high-rise residential towers go up. "Transit villages" are springing up around inner-urban and suburban communities. Some of the mature "Edge Cities", like Hacienda Business Park in the Bay Area and Tyson's Corner in Virginia, are diversifying their land uses beyond just offices to add dense housing, hotels, and the like. Finally, a number of suburbs which have an old-fashioned district are trying to revitalize them by "main street"-type programs. These trends, along with the others people mentioned, could become more pronounced as the oil peak (or the liquid peak) is passed. A traditional virtue of cities, which is to maximize access while minimizing energy and transport costs, may once again shape urban form much more strongly than in the past 50 years or so.

It is true that we have a huge fixed capital investment in suburban buidings and infrastructure, but it is also true that that infrastructure will require huge ongoing investment if it is to stay in a state of good repair. If oil peaks, transportation investments (of whatever is left to invest) could be shifted from highways to reviving rail services, as many of the rail lines still exist. Transit routes could be established, or reestablished, between new nodes.

Buildings, too require reinvestment; Stewart Brand, in his "How Buildings Learn", noted that buildings need a big reinvestment at an age of about 40 years, as roofs, siding, plumbing, and other components wear out. Post-peak, some buildings, expecially the most remote and energy-intensive ones (i.e. offices without operable windows) may no longer be worth the trouble, and abandoned or dismantled to build other buildings. Other single-use buildings may be adapted to new uses, and "grayfields" -- suburban parking lots -- may become the sites of new mixed-use settlements.

Will we soon be talking about "peak suburbia?"

Let's hope so!

A traditional virtue of cities, which is to maximize access while minimizing energy and transport costs, may once again shape urban form much more strongly than in the past 50 years or so.

Let's hope so!!!

 Whether you like surburbia or not there are useful structures and possibly a few house will be removed and the land converted to food  gardens.   In PO why would anyone think that there will be this big bucket-o-money to run off and build utopia.  If I was in the desert SW (USA) well then yes surburbia could  be the future ghost towns of 2200. I get a sense of "suburbia hatred" from so many people,  we all get to make the best of it, what ever that is, and in the future those surbubia tract houses might make clothing, etc.  Who`is to say for sure, why waste a built building?
it's a good bet "suburbia" is going to be redefined, it's unrealistic to say that we can call it now, what that new definition will be.
I would have to expect that many Bedroom Communities that are looked on as Suburbs now, will start to develop more active market centers (ie, food/essentials) again, and will create increasing local jobs in doing so.  As said above, people will likely re-converge and make use of existing housing stock that is convenient to these re-forming Towns.

Flying over southern Jersey, I'm struck by the Cookie-cutter Housing Webs that cover old farm-fields..  There will be some bizarre Ghost Towns, and if the population shrinks (as it must), then the material wealth in these hollowing places will become a new 'Raw Material' for salvaging into later waves of new-home construction.  Miles of Copperwire, Steel Studs, Aluminum Doorframes, Glass, PVC

I have to wonder if the scraps from this great overproduction we've been part of might not allow a smaller population to have a somewhat abundant supply of many necessary resources for the materials we commonly work with.  Will we have to mine for more HardDrives, or are the trillions of used and dead ones going to be reworked into the next ones?

Will there be a new industry in Mining Landfills for all the precious material we looked on as junk over the last 60-80 years?

Saw on National Geographic Channel last night that the Puente Hills landfill (Los Angeles County, California) is making 50 MW of electricity from methane emissions ("enough for 100,000 homes").  At that rate we may not want to break them apart.
This PDF file has much more detail on LA County landfill gas recovery, and is probably better than "I saw on tv ..."
I worked on a video at one such Methane CoGen site in Massachusetts (10yrs back or so).. Yep, it works.  Don't know the efficiencies, but the engineer told me that they water the piles when they need to increase production.
"An interesting debate we are not having is whether suburbia must end or not."

Odd. I would have called it a boring topic that is constantly being discussed - even in posts about Hubbard linearization.

Well, I admit that it's boring in the form that it's usually expressed: "Peak oil means suburbia will die." That's not a debate, that's a prediction that precludes any discussion. And a prediction that, I think, is usually predicated on a whole set of unstated assumptions. It's certainly more interesting if you start from a different perspective: "Under what conditions could suburbia survive?"

Just as one example, the stock answer says you have live close enough to shopping to walk or bike. But outside Denver, where the sun shines and the wind blows, most suburban houses have enough energy available to them if they could capture and store it to run a small, efficient electric vehicle the several miles needed for shopping every couple of days. What technologies make that possible? How soon will they be available? Or you could consider it in reverse. There's lots of energy available on the roof of the Safeway, is it enough to run electric delivery vans to take groceries to the customers who order online?

Another stock answer is that suburban housing requires too much energy for heating and cooling. Ground-source heat pumps are very efficient, are feasible retrofits in the Denver suburbs, and use electricity from any source in place of today's common natural gas heating fuel. Combine that with much better insulation. Is it good enough?

Suppose electrification is possible. Colorado is rich in potential wind and solar resources, has enough coal to meet that augmented electricity demand for decades, and significant uranium and thorium that could run CANDU reactors (thorium needs a bit of seeding); does suburbia survive in Colorado but not in New Jersey? Not to pick on New Jersey in particular, but they seem to have a lot of people and darned few energy resources. Still, is balkanization of the US along energy-rich and energy-poor lines possible?

I fully accept the first point re energy costs of all fuels, but I think that the second part gets to the heart of a significant number of posts.  It is the point that skeptics often miss and it is the concern that any transition takes time.  Yes we have seen a significant growth in ethanol in the US, and might see this providing 10% of what we need.  Similarly we might get about the same from bio-diesel.  Since these can both be generated from relatively small plants at relatively low investment, and little permitting time, that might get up part way to a solution, fairly rapidly.

However for more significant programs, be they oil shale, tar sand, nuclear power, CTL or whatever, there is a very significant lead time required to acquire the funds, get the permits etc etc.  Without the awareness and incentive to initiate and accelerate those programs, there is quite likely to be a period where supply available will fall and there will be considerable pain in the world economy.

I'm glad this is being discussed.

Personally, I think of "peak oil" as conventional oil + condensates. That's it. As the world moves toward substitutes, as it is doing, we are open to all sorts of monkey business regarding "peak liquids". Now, nothing compares with the EROEI of conventional oil. This fueled the creation of industrial civilization. However, in order to define conventional oil, you need to be quite specific, for example, about the API. Is Saudi heavy crude that can't be as easily refined "conventional oil"? Depends on what your agreed upon definition is.

The EROEI of all liquid substitutes, including "extra heavy oil" like that from Orinoco, is worse both in the production & refining. Same with the tar sands and ethano