NYMEX Light Sweet Crude down to 63.96 04-OCT-2005:5:30EST and holding.

Okay, so I read this whole thread and first I'd like to say that I really appreciate Lynch's participation here and I wish Cambel would do the same. It goes to credibility. Having said that, what I have come away with is that the "PO crowd" believes the peak is imminent based on known reserves, production rates and technology. They also believe that, because oil has no comtemporary substitute and is unprecedented in terms of it's contribution to human energy use there is no reason to believe that another equally powerful energy source will be found. Lynch believes that the peak is much farther away based on yet to be discovered reserves, yet to be developed existing discoveries and yet to be developed technologies. Further, he believes that even when the peak is reached it will be mitigated by as yet unknown substitutes.

I'll go with Cambel, his lack of participation in this debate notwithstanding.

The only thing that Hubbert didn't really get right was the correct formulation for his depletion curves. Yet he was close enough to fittingly take full credit. Yes, he didn't take into account the possible oil shocks. But all they would have served to do was delay the inevitable.

Many analysts continue to use the Logistic curve solution to model oil depletion. This gives critics like Lynch huge opportunities to dismiss the models because of a few fundamental issues with the math. I have been carrying on a discussion with a blogger at TheOilDrum.com who continues to use that classical but flawed formulation. He probably gets the results correct but the mathematical basis is a bit suspect in that only the asymptotes match up. Do you want to call him lucky? Do want to call every famous scientist or inventor lucky for being born in the correct century? (excepting Bob Lucky at Bell Labs)

Lynch can use all the ammunition he wants but he can't puncture the math if you set up the premise correctly. See
my oil depletion model posts where I both skewer Lynch's reasoning and help us understand where the logistic model took a wrong turn.

I would expect that you can model this fairly easily based on the length of well in contact with the reservoir. With vertical wells, you might be getting 30m per well for a long oil column; single bore horizontal wells might get 10 times this, and multi branch MRC wells another 10 times for a given surface borehole. Each step will increase the cost of a single well, but you will save a fortune on surface facilities.

Obviously finding out exactly how much this value changes with well technology improvements is the hard part. There is a danger in ignoring it, though. An old onshore field in the US may have 1000 wells in it; each producing 10 b/d. A more recently developed offshore field could have just 10 MRC wells, each producing 1000 b/d, meaning both fields have been drilled up to the same degree - yet a simple estimate based on well productivity or number of wells would suggest that the second field was far less developed. This appears to be what caught out Shell in Ybal(sp?).

As far as I remember, the technology of hydrofracturing and well completion hasn't changed too much over the past few decades, so reservoir contact length may be a useful measurement; certainly more useful than number of wells or well productivity, both of which are far more subject to technological distortion.

That is exactly the concern underlying my post that Lynch answered.

BTW, yesterday with my third post the website said it wouldn't allow me to make too many in a short period of time, so I might be limited in my responses (hold the applause).

There is a large discrepancy, but you will note that almost everybody who is not using the Hubbert method gives a much higher estimate, about 3 trillion or more, and that this has been increasing over the long term. We have enough information now to recognize that the total resource base is very large, although much of it is economically unviable. (And more is shut off by political restrictions.)

I don't know what more data Simmons needs. (His use of data in the past has not been very impressive.) One almost suspects that he is using the call for more data (in the case of the Saudis) as a distraction from the poor quality of his book, i.e., he knows they won't release more data, so he always has a fall back.

In the case of market signals, there has long been a belief that markets are myopic, but it has usually proved incorrect, rather the simplisitic analysis was myopic. Since reality suggests there will be no sudden transition, it is not clear why the market would fail to miss the slow shift.
Mike Lynch

BTW, re baffled, I used to joke that energy talks fell into two categories, "Everything you know is wrong" and "What the heck is going on?"

In my 1995 review of oil supply forecasting methods, I noted that nearly all produced a near-term peak and decline, which then had to be revised into the future. And the Hubbert method, by mistakenly taking URR as a fixed number, was particularly susceptible. Campbell and Laherrere have mistakenly argued that their method produces 'robust' estimates of URR. They have then revised the peak prediction repeatedly, almost always outward. So, I am hardly surprised (though I think they are perhaps disingenuous in ascribing the change to deepwater.)
Mike Lynch

Yes, but they stay within a 10-year range. That isn't a whole lot of sliding, in my opinion.

I had that problem too - it's not you, it's the server and the forum software. The time-stamp on the messages are getting mangled, sometimes resulting in a post being stamped for several hours into the future. The problem is that if you have a miss-stamped 'future-post', the forum software won't let you post again until after that time passes.

It is when you've been sliding for ten years.

Laherrere has already abandoned the pre-2010 peak and now looks for an "all liquids peak" it to come c.2015. At least he's made an effort to find out where he went wrong, admit it, and try to correct for it. Campbell, meanwhile, is simply running out of time - I doubt he could push his peak date back any farther.

I think you are cutting too wide a swath with the “no sudden transition” knife. While on the one hand I agree that oil isn’t going to suddenly run out, meaning that there is “no sudden transition” away from the product itself, I disagree that there would be “no sudden transition” in the price of that product. Once it is apparent that production has peaked (and provided this isn’t clouded by political or other non-geological events which confuse people as to whether or not production can be recovered) the market will rapidly bid up the price of oil. If you think the market is speculative now, wow, imagine what that scenario would generate.

Hyper expensive oil due to falling production will not be good for the overall health of the economy. So in this scenario, we are forced to perform the Herculean task of transitioning away from oil at the worst possible time – when everything is getting much more expensive and the economy is contracting. That to me means the market is indeed myopic and has failed us in a catastrophic way.

The only hypothetical scenario I see where the market is able to give us enough lead time is where the price of oil continues to rise even as production is rising. ie. Demand continues to significantly outstrip production. And, importantly, production must continue to rise “enough” for at least 10 years, and the price must also rise along with it.

Really, there are only two scenarios under which we are going to transition away from oil: production starts declining indicating we have no choice. The other is that the price of oil gets sufficiently high enough that alternates are pursued instead. Obviously the latter is the preferable scenario, because we have more time.

Hmm, hey, maybe that’s what is happening today?? Maybe the current price isn’t speculation based after all!

Actually, according to spike/Mike Lynch's "Crying Wolf: Warnings about oil supply" paper, Campbell (in 1986) predicted PO as around 1990, twenty years earlier than his current prediction.

Sorry, my point is that most data available doesn't provide this information. For example, rigs available, not type of rig, wells drilled, but not type. Not sure what would make a good analogue (feet/well?).
Mike Lynch

Campbell orginally said in a 1989 article that oil was peaking that year. In his 1991 book, he put the peak in 1992, etc., etc. Since this conforms to my theory, that the method constantly yields a near-term peak which always has to be pushed outward and upward, and contradicts his argument, that his method produces URR and peak estimates that are fairly precise and don't need to be revised, it would seem to affirm my arguments.

In recent years, he has downplayed the argument that his results are precise, and admitted to revision: http://www.oilcrisis.com/campbell/assessments.htm
But he doesn't note that he is reversing his earlier insistence that his method was robust and would result in reliable predictions, contrary to my arguments.
Mike Lynch

My point (the one I'm attempting to make, however imperfectly) is that (according to resource economics and historical experience), you never suddenly fall off a table from cheap to expensive resources. Rather, there is a large continuum of deposits and the movement is relatively gradual, from $2 oil to $2.15 oil, not $2 oil to $25 oil. (Costs, not prices.)

Political interference (kicking the big oil companies out of the P Gulf in the 1970s) can cause a big discontinuity, of course, but I doubt we'll ever see worse than the 1970s.

So if costs are rising gradually, you have lots of warning and time to invest in other things (including slightly more expensive oil).
Mike Lynch