I notice that Mike Lynch has been posting on the energyresources Yahoo! group lately. This could be a good place to observe (and participate if you follow the guidelines of the group) in a constructive and interesting debate.

After re-reading Lynch's paper, The New Pessimism about Petroleum Resources: Debunking the Hubbert Model (and Hubbert Modelers), and noticing the following conclusion,



I would not characterize Lynch as denying PO entirely or denying that conventional oil is a finite resource. He merely disputes the claims by Deffeyes, Campbell, and others that PO will arrive very soon, i.e., within five or ten years. I would put him with Peter Odell in the "PO later" category. Both Odell and Lynch have hinted that later might be about thirty years from now.

Hi, guys, I'm sorry if I have been neglecting you. The board sent me a kind invitation to join, but then I got swamped with travel, etc. (For some odd reason, everyone is interested in oil these days.)

My interpretation of high oil prices right now is that investors (mutual funds, etc.) are pouring money into commodity index funds, thinking a) that market tightness could be coupled with new supply disruptions and b) longer term, capacity can't be replaced. Since oil production has been rising sharply the past 2 years, it is hard to see how geology can be driving high oil prices. (Capacity constraints are another matter.)

Hubbert modeling, to my thinking, is curve-fitting, nearly the same as trying to find an equation that fits automobile production, the stock market, etc. Lots of things fit S-curves, bell curves, and so forth, but it doesn't mean that they are predictive.

Modeling oil production at the global level is, most of the time, a question of modeling oil consumption. In the end, you need to consider, in a general way, the resource base for an area and what is required to replace production and/or increase it (or not). Prices and government policy play a major role in that, country by country.

I hope this is coherent.
Mike Lynch

Why would it be in the interest of an oil company to discredit peak oil to keep the oil price low? Wouldn't it be more rational (if unethical) to insist the peak is actually going to be here any day now (just not our fields) to keep prices artificially high on speculation and fatten profits?

JoeW wrote:

Dude, you reminded me of my calculus day!
Actually Rolle's theorem from elementary functional analysis predicts a peak lol
Lets see if I remember it correctly
Let f:[a,b]->R function differentiable in the closed interval [a,b]
If f[a]=f[b], then there is a c point where the first derivative of the function is equal to zero
But if a function is differentiable, and has a derivative equal to zero, then that point is an extremum (maximum or minimum)
So think of f as the production to date, its first derivative is the rate of production (i.e. how much they are pumping from the ground per unit time e.g. day). So what does that mean? That there is a point where the production will be maximum. Before that point, the production rate will de increasing, after that it will be decreasing till it goes to zero.
Come to think of it, it is not even college math, probably senior high math.
Not even an economist can beat both geology and math

Mr Lynch, some of us around here have been addressing those issues. Anyone with a computer and a decent fitting package can see WHY EIA modelling is absurd (had to say that, sorry) and how one can generate multiple compatible predictions. Having said that, shouldn't one adopt a play safe stance and assume that the Peak is here (even if it is not).
My position on using Non linear regression tools on the time series that BP published ... is that the peak is going to occur in this decade.
Simple mathematical arguments (i.e. peak production starts from zero, has to go down to zero, time is finite, production is a mathematically continuous curve) shows why a peak has to exist.
BTW most of the modeling approaches (even the ones at ASPO) would have been thrown away if submitted in any decent medical journal (my field) due to statistical inadequaceis

What we know:
- world production is a sum of individual field productions;
- once a field goes into depletion there is no way back
- new oil fields are small in size
- discoveries have peaked in the 80s;
- reserve growth is going flat since 2003;
- reserve growth is mainly on paper (few appraisal wells)
- demand is growing rapidly
We don't need to do any curve fitting to see that we may have a problem!

Hey khebab,
Did I get the right theorem?
Is it Rolle's, or the extreme value the one that guarantees a peak?

You're quit right, a more general form is the Mean Value Theorem:

If f (x) is a differentiable function with a continuous derivative and if a < b are any two points then there is a point c between a and b at which:
f'(c)= (f(b) - f(a)) / (b - a)

Doesn't it say at least one point?
Guarantees at least ONE maximum, maybe more but totally rules out a plateau as some of the economists think. It is due to the finite nature of oil. Builds up and declines.
So from a mathematical standpoint the following are the only possible outcomes:
1) Single Peak
2) Multiple Peaks (but with decline)
3) A plateau and decline
If there is no break in consumption due to high prices .... single peak, otherwise a multiple small peaks and then nose dive.
There is no way the no-sayers can refute that. They can try by saying that the production is not continuous, singly differentiable ... but this is stupid. Oil is liquid and it flows and at first approximation the flow is laminar hence differentiable
Why hasn't anyone thought about this before? (I only thought about when I read the initial post, it clicked memories from the 1 st day out of High School

I would tend to agree with you that it is curve fitting, or in other words an empirical prediction. However empirical predictions are not bad things, if they're all you've got, they're better than nothing. There are however, fundamental differences between a hubbert curve and empirically predicting the stock market. Efficient markets (which stock markets aim to be) are inherently unpredictable. You cannot say the same thing about oil production.



Of course modelling production is the equivalent of modeling consumption and vise versa. That is basically saying that when production starts decreasing, there will either be demand destruction and or higher prices. It certainly doesn't mean that we will always find the oil necessary to meet future predicted demand at current prices.

The suggestion that I'm on an oil company payroll is entertaining; many of the pessimistic views come from people with oil company support. Matt Simmons comes to mind. And some oil companies seem to disagree with me, in whole or part (Exxon, Total).


As I pointed out in my article in Quarterly Review of Economics and Finance (and elsewhere), NONE OF THE MODELS HAS A GOOD RATE OF SUCCESS. Investment data is poor, drilling data is poor, field size data is uncertain. Neither volumetric, engineering or economic models work very well.

Production has to start at zero and end at zero, but that tells us nothing about the shape of the curve in between. In fact, Hubbert admitted that himself. You can (and often do) have multiple peaks, sharp peaks, long rising or falling plateaus, etc. Depends on how you define the region, first, but also geologic, geographical, political and economic factors. The ASPO models, Laherrere's creaming curves, Deffeyes' curve do not factor that in.

Khebab, most of your comments are wrong or irrelevant. There is no sign that reserve growth is dropping, reserve additions are replacing production, and fields (and regions) can decline and then recover, happens quite often.

Mike Lynch

What makes you thing reserve growth has been falling for 30 years?
Mike Lynch

Actually, you misquote me, I didn't say 'new physical laws' but they are claiming to have developed methods such as the creaming curve and parabolic fractals (new methods, or improved applications) that have given estimates of URR that are stable and robust. That is my dispute.
Mike


Who, what, why, where, and how? Sounds like the ISS theorem strikes again!

Read their work. Campbell, Deffeyes, Bentley and others dismiss the notion that economics effects oil supply, implying that 'economists' think higher prices will increase the existence of oil in the earth. (They are often unclear in this as so many others.) Examples are rampant in their work, and I cited some of them in my article (quoted above).
Mike Lynch

I disagree with many of your opinions, but thank-you for joining the discussion!

Mike,

I and everyone here appreciate your input. Its absolutely critical that you participate in this debate. I would hope and ask that you go the the experts forum and look at my challenge questions to you and answer them. Though my intro is too personally harsh, it was intended to get a response from you to answer the questions. Hopefully, you will do so. Also:
(1) what is the explanation for the high oil prices right now? Is it pure speculation? Or, is it as SA says a "down stream" problem which, if so, seems it would take several years to get the necessary refineries built and put into operation;
(2) What was your take on Simmons recent book "Twilight in the Desert"?
(3) I would like to hear your take on the Hirsch report to the DOE about mitigating the effects of peak oil (February 2005);
(4) Recently, SA was quoted in the news as saying OPEC would be unable to meet IEA oil production forecast after the next 10-15 years. Do you believe this is the case and if so, does or should this change the IEA's forecasted peak date for world oil?
(5) Also, your take on the American natural gas issue would be much appreciated. Is North American natural gas peaking? If so, what is the estimated date of peak and will LNG terminals be in place in time to remedy the situation.

Thanks for being here.