[ReserveGrowthRulz]

Arctic ice melting opens up areas of exploration that haven't been available before, both practically and economically. The USGS numbers for undiscovered resources for Greenland are a perfect example, they say maybe there is a little, but maybe there is ALOT. Which is a nice way of saying no really knows quite yet. But just the idea that climate change might open up previously underexplored areas strikes me as interesting. Remember, it only requires another Ghawar somewhere, nearly ANYWHERE, to throw the entire Peak Oil theory out the window...for at least another 20 years anyway before Campbell Jr starts it up all over again.

As far as differential equations versus anything else, I am happy to spend my existence in my little niche in the world, solving the depletion problem one well at a time, specializing in small well/field economics and production/value.

I get the feeling that "depletion" as referred to around here encompasses alot more of the BIG question than things as simple as how wells and particular fields deplete. Nothing wrong with that of course, but as a practical kind of guy, I'm inclined to start small and work my way up the foodchain. What I discover the instant this process begins is that as soon as aggragatgion begins, you lose the detail perspective necessary for any normal petroleum engineer to do his job without entering in the realm of probabilities versus more deterministic numbers.

And I think the concept of "chaos" describes the oil industry of this planet much better than any particular mathematic equation.

[WebHubbleTelescope]

Ok, we have two petroleum engineers on this thread. I will make the assertion that there is nothing particularly sophisticated about devising rate equations from a finite supply of resources, randomly spatially distributed within the ground. I barely have to know anything about geology to predict what is going to happen. All that you have to assume is an exceedingly greedy population of money-seekers and some elementary knowledge of the way stochastic processes work.

Two steps are involved, a stimulus (corresponding to the discovery profile) and a response (corresponding to the extraction process). I will take the case of the USA first, but variations work equally well for other areas and global as well.

1. Discoveries in the USA followed a quadratic growth, starting in 1859 and peaking after 1930. This appears as a parabola that goes through a boom and bust.

2. About 4 time constants govern how the discoveries transform into extraction. These are what I call an average fallow period, an average construction period, an average maturation period, and an average extraction rate. Let's give each one a value of 10 years. Each of these follow a first-order Markovian rate law, which is a maximum entropy estimate. Then what you get is a quadratic convolved with a 4th-order gamma curve. It looks like the blue curve (the EIA production data in green):


I also put a gaussian curve in there as a comparison. I can get a fit with the production curve over a range of 5 orders of magnitude. Smoke that.

This is basically first principles stuff. Why don't the oil companies do analyses like this? They've been obsessing over the black gold for almost 150 years and nobody has ever shown how simply these rate equations work out?

Oh I forgot, they probably do. They just keep it locked in their desk drawers, and bring it out only for staff meetings. The oil companies basically want to keep their customers barefoot and pregnant, i.e. fat and lazy, ignorant and in the dark, willing to bleed everything dry as long as they can maximize their profits.

http://mobjectivist.blogspot.com/2006/0 ... lieve.html

[rockdoc123]

WebHubbleTelescope...what you just created as I remember was being used some twenty years ago in the research labs (back when oil companies like Gulf and Shell had large research facilities and had no problem hiring people with a maths jones to sit in an office and fiddle with data). Today most oil companies are just not that interested in how much might be left overall (exception would be BP who still has a pretty big research component)...they spend their time modeling existing production and coming up with predictions on whats left to be discovered in various parts of the world where they are working. This sort of analysis involves more subsurface information and requires less math analysis of past trends. If memory serves me correctly some of the old analyses were published in the AAPG Bulletin.

It would be interesting to look at the same analysis for somewhere other than the US. Arguably the US has been a poster child for making discoveries and getting them on-stream with little in the way of interuption or government imposed obstacles. Perhaps looking at places like Libya, Iran, Iraq etc. where there has been artificial limits imposed on production (sanctions, quotas, wars etc) would be useful.

[khebab]

WHT, your result is really interesting but if you want to compare two pdfs, in particular if it includes the gaussian case, you should use a quantile-normal or a quantile-quantile plot instead.

If you post your data (Observed, your model, gaussian) I can plot it for you.

[WebHubbleTelescope]

WHT, your result is really interesting but if you want to compare two pdfs, in particular if it includes the gaussian case, you should use a quantile-normal or a quantile-quantile plot instead.

If you post your data (Observed, your model, gaussian) I can plot it for you.

I'm more interested in understanding why you consider these probability density functions (PDF). The dimensions of the output function are not in terms of a dimensionless probability but of a concrete measure -- volume of production per year.

[khebab]

I'm more interested in understanding why you consider these probability density functions (PDF). The dimensions of the output function are not in terms of a dimensionless probability but of a concrete measure -- volume of production per year.

It's maybe an abuse of language of my part but it doesn't matter, even if the area under the curve is > 1 the qq-plot is still a good way to display differences between models.

[WebHubbleTelescope]

WebHubbleTelescope...what you just created as I remember was being used some twenty years ago in the research labs (back when oil companies like Gulf and Shell had large research facilities and had no problem hiring people with a maths jones to sit in an office and fiddle with data).

So why did not such a fundamental analysis get transferred to the textbooks? The lack of this kind of analysis is much like teaching first-year electrical engineering without introducing Kirchoff's Law. It's just freaking rate equations and conservation of matter. This to me is deeply troubling. I didn't take an earth sciences major, but I have taken classes in subjects such as limnology and the first thing you learn is that all freshwater lakes go through a life-cycle, birth through death. Are the petroleum engineering departments so vain and self-conscious not to even broach the subject of the life-cycle of oil? To avoid teaching that the whole thing is just a house of cards or a ponzi scheme, certain to eventually collapse, strikes me a bit irresponsible.

Today most oil companies are just not that interested in how much might be left overall (exception would be BP who still has a pretty big research component)...they spend their time modeling existing production and coming up with predictions on whats left to be discovered in various parts of the world where they are working.

And will this get out to the textbooks? Or will it get buried like the rest of the stuff, until we all say "Wha' Happ'n?"

This sort of analysis involves more subsurface information and requires less math analysis of past trends. If memory serves me correctly some of the old analyses were published in the AAPG Bulletin.

OK, so this got published. How about the results from 20 years ago?

It would be interesting to look at the same analysis for somewhere other than the US.

I have analyses for North Sea, World, lower-48, even for natural gas.
World
Lower-48
UK North Sea
Former Soviet Union
New Zealand NG
Norway
I use the same model for everything, vary the parameters a bit, and gain an understanding that I surely would not have if these rate equations were not available. We could be mining for marbles, it doesn't really matter, I see no fundamentals being taught or published anywhere. The more I look at it, this stuff with the Logistic curve and Gaussian fitting is basically rubbish. It's equivalent for me to trying to teach the response of analog electric circuits by looking at the output waveform. I would like to see the oil depletion modeling rise a step above shear empiricism.

Arguably the US has been a poster child for making discoveries and getting them on-stream with little in the way of interuption or government imposed obstacles.

Government-imposed obstacles seem like a second-order effect when put up against the greed of the human animal. Teach the first-order effects first. Just about everything in engineering is first -order effects. If you don't do the first-order stuff first, you should just give up. They don't call it first-order for nothing.

[WebHubbleTelescope]

Ok now we are talking. Will you share your equations and dataset. I have worked with Markov chains but I am not familiar with every term you are using there and certainly I cannot guess how you have structured your algorithm.

Because you likely haven't seen the source code:
http://mobjectivist.blogspot.com/2005/1 ... ogram.html

[rockdoc123]

Are the petroleum engineering departments so vain and self-conscious not to even broach the subject of the life-cycle of oil?

Uh, no.....Hubberts work on peak US production was required reading when I was an undergrad a few decades ago. Besides you have to walk before you crawl....as I've said if you don't understand the underlying science to petroleum generation, maturation, migration and accumulation then you are pretty much lost in terms of trying to ever find any of this stuff regardless of your math background....that is why universities concentrate on teaching the basic science. If you could discover all of the oil in the world simply by using math models I suspect Kurt Goedel or Mandelbrot would have already done it.

And will this get out to the textbooks? Or will it get buried like the rest of the stuff, until we all say "Wha' Happ'n?

having a black helicopter tin foil kind of day are we? There is reams of views on what is left to be found in the literature, has been since the seventies. Check American Association of Petroleum Geologists Bulletin, Bulletin of Canadian Petroleum Geology, Petroleum Geology, USGS Energy group assessment, etc.

I have analyses for North Sea, World, lower-48, even for natural gas

OK now thats good and I think it supports some of the arguments about factors other than what reserves are left to be produced controlling the way in which depletion will happen in a given country. UK north sea is a good example of that as it FSU. The proof in the pudding would be to take UK North Sea data up to the first peak run it through a model and see if you could predict the second peak without knowing anything about the subsurface, tax laws, production limits etc.

Government-imposed obstacles seem like a second-order effect when put up against the greed of the human animal

but it is in fact gov't changes to taxation in the UK that created the second spurt of exploration and the development of small fields which resulted in the subsidiary bump. My argument would be that if the Iran Libya Sanctions Act did not exist we would be seeing huge spare capacity today in both Libya and Iran. Both of these countries have only been mildly explored since ILSA was imposed. But you are right that it is important to get the math right behind the depletion model.

My question would be are we not just putting lipstick on a pig by trying to tweak the math related to depletion? Given all of the uncertainties if you did a stochaistic logistic curve analysis would not the range of results be so wide that no additional precision is gained from having a better algorithm? Again I'm not a maths person so I don't tend to get very emotional about this stuff.

[seahorse2]

Why don't the optimist ever produce a model? This is my issue with people like Lynch, Reservegrowthrulz, etc. Put a model up and show us why you are optimistic. At least the doomers like Campbell and company are willing to undergo scientific scrutiny. So, do the same. Lynch's point that doomers have been wrong in the past therefore they are wrong about the future is not a scientific argument, is a logical fallacy, and does nothing to answer the questions surrounding the eventual peak in world oil production.

As a side note, there have been several petroleum engineers on this forum that do believe the pessimist are more correct than wrong.

[dub_scratch]

Why don't the optimist ever produce a model? This is my issue with people like Lynch, Reservegrowthrulz, etc. Put a model up and show us why you are optimistic. At least the doomers like Campbell and company are willing to undergo scientific scrutiny. So, do the same. Lynch's point that doomers have been wrong in the past therefore they are wrong about the future is not a scientific argument, is a logical fallacy, and does nothing to answer the questions surrounding the eventual peak in world oil production.

Yea, you would think that if they are so confident about supply giving us the peak later scenario, a depletion model would be easy.

I think the reason why many cornucopians do not do depletion modeling is because it exposes their often generous assumptions of oil supply. Lynch would much rather say that he is too busy of do a depletion model and that oil peak is too far off to model, then to have a model showing some extraordinarily late peak date and rate. When someone does that they then are expected to show us where the oil is. And then, of course, one finds that much of the oil is imagined, oil assumed in-place. But refusing to give a depletion model as an alternative to the supply skeptics (Campbell, Defeyess, et. al.) the cornucopian is left off-the-hook about their assumptions and avoid scrutiny. No one can go after Mike Lynch and challenge his assumptions when it is not clear how much oil they contain.

[ReserveGrowthRulz]

Why do depletion modeling when oil production has been increasing to match demand, decreased in relation to demand during the oil shocks of the 70's, thereby demonstrating that when supply gets scarce enough, driving prices up, guess what happens? Consumption drops. And when consumption drops, what happened last time? Why...the Saudi's got cheesed that someone was grabbing their market share and put the entire market into a tailspin, resulting in a decade long climb back to less excess capacity and a more balanced supply/demand picture.

Modeling depletion might be an interesting technical exercise, but you aren't modeling DEPLETION but the geopolitics of oil since the rise of modern oil reporting, which means you are mostly modeling Saudi political decisions and NOT what is happening with a limited geological product being evenly developed in line with the basic function of supply and demand.

There was a recent conference back in DC somewhere which I got a report on, and interestingly enough, the Peakers stood up after a full slate of reserve growth advocates...and when the modeling Peakers were asked at what point in their modeling did they account for reserve growth contributing to the delay of their peak, the description was, "they appeared to be struck dumb and lifeless by the acceptance of reserve growth, the admission in their data that it existed and their obvious lack of accounting for it".

If Peakers can't even accept some basic other THINGS happening which negate depletion ( using the term loosely which appears the norm around here ) modeling, and don't account for these obvious things, than what point is there in modeling? Optimists don't have to produce a model because they have ALWAYS been right, since Pennsylvania in the 1870's, the club of rome(?), the CIA reporting in the 80's, since Hubbard declared the peak a decade ago, since Campbell picked it in the 80's,you name it, they have all been wrong, and while the IDEA isn't a bad one, saying it happened last Thanksgiving and we all missed it doesn't make the credibility of the idea any higher, for certain.

[dub_scratch]

Optimists don't have to produce a model because they have ALWAYS been right, since Pennsylvania in the 1870's, the club of rome(?), the CIA reporting in the 80's, since Hubbard declared the peak a decade ago, since Campbell picked it in the 80's,you name it, they have all been wrong, and while the IDEA isn't a bad one, saying it happened last Thanksgiving and we all missed it doesn't make the credibility of the idea any higher, for certain.

Oh, I see. Since the "optimists" are ALWAYS right, oil must be infinite and a peak will never be reached. Humans will produce 1 gazzilion barrels per day and yet still not reach the top. Because you see, if that is not true, then your argument above is a logical fallacy.

[WebHubbleTelescope]

And will this get out to the textbooks? Or will it get buried like the rest of the stuff, until we all say "Wha' Happ'n?

having a black helicopter tin foil kind of day are we? There is reams of views on what is left to be found in the literature, has been since the seventies. Check American Association of Petroleum Geologists Bulletin, Bulletin of Canadian Petroleum Geology, Petroleum Geology, USGS Energy group assessment, etc.

I will lay this out very carefully because my concerns involve a combination of the practice of "framing" along with negation of arguments that the industry condones via paid consultants. First consider how Mike Lynch has become one of the most visible debunkers of peak oil theory out there. And what does he use as ammunition? Lynch uses the mathematical formulation of Hubbert curves to allow peak oil advocates to effectively shoot themselves in the foot.

How does he do this? Well, he starts with the Logistic and Gaussian curves that the traditional analysts use and starts poking holes in how they get formulated. His favorite argumentative weapons are "physically impossible symmetry" and "causality violations". The argument about symmetry is quite simple because all one has to do is look at real-world depletion curves and see that they are not symmetric. Lynch points out that the curves should not be symmetric based on his own studies. Yet, depletion modelers continue to use them. Score: Lynch 1, Modelers 0

Lynch's causality argument says that the Logistic and Gaussian curves have tails at time 'minus infinity', which is a physicallly impossible condition considering that oil was only discovered in the 1800's. Peak oil analysts have nothing to counter this other than some magical truncation that occurs in the curves, without any supporting explanation. Score: Lynch 2, Modelers 0

Now, I would consider these minor blemishes on the overall peak oil theory, but Lynch builds a strong rhetorical argument building from this foundation of negating theories. This paper is a good example of how he pulls together the arguments:
The New Pessimism about Petroleum Resources:Debunking the Hubbert Model (and Hubbert Modelers)
Lynch probably enjoys doing this and probably feels better than out-and-out lying about things (it's the same reason that people don't cheat at crossword puzzles). Lynch probably just sits back and laughs in the way he uses the exact models as taught to petroleum engineers and geologists (as you just admitted, RockDoc), and then basically shreds and decimates the arguments. Listen again, the stuff you were taught in textbooks (Hubbert curves, etc) is being turned upside down and used by Lynch to "disprove" peak oil or at least cast suspicions into its relevancy.

Why does this work? You just have to look at the psychology of people and their fierce need to believe in the status quo. Any bit of debunking is enough to set the whole theory to collapse in these people's minds, especially the sharp ones (and former high-school debaters), who tend to revel in such nit-picking matters. A bit off-track but think back to how GW Bush's entire questionable National Guard record was vindicated when a set of fraudulent memos were exposed. Even though the majority of the evidence pointed out that Bush avoided most of his service, the few flaws in evidence presented sunk the entire investigation. And we have a corrupt Bush administration as a result.

What can we do about this? Simple. Come up with better models than the Logistic curve or the Gaussian. These are basically empirical relationships that prove nothing and have no foundation. Instead, teach something that works. Do you realize that an EE professor would be laughed out of a classroom if he showed a Gaussian curve as the output to a electrical circuit? Yet this is exactly what the Hubbert modelers do when they keep on showing Logistic or Gaussian curves. I groan at the bad math, while Lynch calls them on the B.S. But the Campbell's and Laherrere's would never ever give up their dearly loved Logistics formulations because of their selfish pride. This is what Lynch counts on as well. That they will keep on showing the Logistic curves and Lynch will keep on debunking. Lynch has basically framed the Hubbert modelers as inept mathematicians. Jeez, even CBS gave up their inquest when they could not counter fraudulent memos.

I say let's turn it around and call Campell and Laherrere on their own models, and get some decent ones -- models that don't violate causality. Then we can effectively ignore Lynch, because he will have no legs to stand on. Or else, he will eventually resort to out-and-out lying and we can call him on that.

Rockdoc, What you call "tin-foil hat" theorizing on my part is basically discounting the modus operandi of the entire right wing establishment. If we can just beat down the framing and negative arguments practiced by people like Lynch, we might have a chance to stop digging a deeper and deeper hole.

Or you can laugh and pass my whole approach as a phony strawman. But then you have to remind yourself of who we have as president. This framing stuff works and you may have drunk the kool-aid.

[seahorse]

Reservegrowth,

It would be foolish for a world based on oil not to be concerned with the issue of peak oil, when "peaking" is a given at some point in the future. Even Lynch agrees that some day world oil production will peak, and I assume you agree also.

Lynch would argue that the rising cost of oil prices will eventually drive the world to an alternative, therefore, predicting the peak is unnecessary. That assumes the market will react. There are serious reasons to question this reasoning.

You argue that optimist don't need to produce a model bc they have always been right. They have been right that the world has not peaked yet, but they have been wrong regarding individual regions - for exampel, the optimist were wrong about the United States, Norway, UK, Mexico, Indonesia, and any number of other places that are now past their peak. In these cases, the optimist thought the pessimist were crack pots, but the crack pots were proven right.

The modeling that makes sense to me is a model similar to what was done by Taskforce Unity. If you are not familiar with his model, you should take the time to read it. This model is done by estimating world oil demand, known production coming online, and known or estimated depletion rates of existing fields. It doesn't rely on the unknown URR. Based on his model, world oil peak is about 2012.

If world oil production were to peak in 2012, it would be impossible for the world market economy to find an alternative in a mere six years. Several people, namely Hirsch, have done an analysis on how long it would take to transistion the American economy off of oil. His opinion is it would take 20 years of lead time of transistion to avoid serious economic consequences. This is supported also by the research of the oil majors, which state that oil will remain the dominant energy source for at least the next 20 years.

All this to say that models are an extremely important part of any business, particularly oil. The date of peak is extremely important to us all. Therefore, optimists ought to consider producing a model also. I'm not sure your optimistic opinion is shared by all the oil majors. BP (Beyond Petroleum), Exxon and others now runnning webpages asking for solutions to the long term problem of oil supply issues. For the majors to suggest and raise these questions implies they have at least some models, that they aren't just guessing.

[ReserveGrowthRulz]

Seahorse, I have never said I wasn't concerned with peak oil, as a petroleum engineer there are certain things which I have the fortune to actually watch, and taking care of wells and fields is a good practical basis for how I might go about drawing conclusions about the rest of the planet, if on a grander scale.

Your general conclusions about what Lynch assumes strikes me as reasonable. The flaw that appears is when you say"Lynch assumes markets will react"...of COURSE he assumes that, it evident throughout world history....the quote I like is "the stone age didn't end because of lack of stones". England NEVER ran out of coal in the early 20th century unlike the Coal Doomers predicted at the time. Assuming markets WON'T react strikes me as more than a little off base for what history shows and what has happened in the oil and gas business if only during the last hurricane which screwed up supplies...which doubled the price....which caused more than one person to cut back driving here in the States.

I am not familiar with TaskForceUnity but I'll be more than happy to go look. The first questions I will try and answer will be how well he handles unconventional resources, his rate of discovery and field size distributions for basins which haven't yet been developed, Lapthia and anything in the Arctic are the first two examples which pop to mind, and how he determines how reserve growth interferes with a conceptual depletion of a given field. Remember, this is a game of the extreme's changing the average for the entire planet, another Ghawar hidden in Siberia and its game over for modeling any particular past event. But I'll go check out his stuff.

I do not make the assumption that $6/gal gasoline and diesel in the States would make it impossible to find an alternative in 6 years. Why? Because alternatives are already here, just not particularly economic. Give them $6/gasoline to compete with and its Katie bar the door, everybody and their mother will drive a hybrid or alcohol car, there will be solar panels on every roof in the desert SW and people won't be whining about how much federal land gets torn up to use Shells new extraction method across the Rockies.

Peak oil is still a concern, I would agree....but how MUCH of a concern is a quite different question. People who would normally be worrying about an asteroid strike destroying civilization seem to have latched onto Peak Oil as their next best hope, but I imagine they are still cheering on the asteroid in other forums.

[seahorse]

ReserveGrowthRulz,

Hope I get this link right, but it is to the model by Taskforce_Unity. If I get it wrong, do an author search for TaskForce_Unity.

Here's the link:

http://www.peakoil.com/fortopic11688.html

I will later try to post the Hirsh article for the DOE done in February of 2005. There is probly a forum discussion on his paper. His paper is titled "Mitigating the Effects of Peak Oil." In that paper, he discusses transitioning from oil to other resources using currently available technology. Its worth reading.

You stated earlier that England never ran out of coal before switching to oil, however, their coal reserves, what's left, could not power their society now, neither could wood. They don't have enough of either. They could always revert back to the stone age, but that's what the pessimist are worried about.

Last, let's assume that "peaks" in oil are caused solely by politics or economics, and not by a lack of resources. It seems that we are peaking now, with the formation of the BRIC alliance (Brazil, Russia, India and China), Chavez in Venezuela, Bolivia, an ever unstable Nigeria, Iran, Russian flexing its muscle with its natural gas to Europe, etc., it seems the West is closely approaching another peak in fossile fuels by that definition - and that the Gulf of Mexico, Alaska, the North Sea, can't save us.

Whether fuel/energy is expensive or unavailable due to political, economic, or depletion issues, doesn't really matter to the common man like myself. The fuel peaks for any of the above reasons, becomes expensive and is therefore painful. The solution to these "peaks" in oil, whether the peaks are caused by any of the above reasons, is to find an alternative, and for that to happen now, so that the Mullahs are no longer a threat and neither is Chavez, we need the government to lead the way to an alternative, bc its not happening fast enough on the private side.

[ReserveGrowthRulz]

I went through the thread you mentioned, and read some of the comments, but the original PDF listed at the top of the page comes up as a dead link, and the link to dieoff is just the Ordovician Gorge(?) arguement which has us 7 years into a huge electrical blackout and anarchy by now. I can't tell from the comments exactly what the guy was predicting and exactly how, but it looks like it generated lots of interest anyway. I'll try rooting around some more this evening and see if I can find more of his actual work rather than just people talking about it.

I would speculate that coal in England probably couldn't satisfy their current energy requirements, but that wasn't my point. My point is, they still mine some coal TODAY, nearly a century after they were terrified of running out, but while they once FEARED running out, they didn't. So the coal age really hasn't ended for them, any more than the stone age ended for lack of stones. At each major energy change on the world stage, something else has come along and changed the entire picture.

The guy who knew this was all going to happen WAY before most Peakers showed up to go the other way was Ahmed Yamani. He took a look around and realized that the instant the rest of the world figures they are being held hostage to OPEC and semi-stable repressive regimes in the Middle East as well as having their economies jacked into oblivion over oil, they would do something about it, and Yamani and his buddies would be sitting around trying to figure out how to get $5/BBL for their product in the ground because it wasn't worth anything given current world demand. I wonder sometimes if the Saudi's have forgotten what he knew decades ago.....

As far as government leadership, I've got no objection to it. Myself, I figure if they just raised fuel taxes to make the price of the refined product $7/gal or so, the market would handle the rest. Do that to the biggest consumer nation in the world, and then just stand back and see what happens to demand.

[seahorse]

Reservegrowth,

Sorry for the not-so-good link. Hopefully this is the direct link to the report by Taskforce_Unity.

http://www.peakoil.nl/images/ponlreport.pdf