A little at a time, the evidence accumulates to the effect that the Hubbert curve applies across the board to just about anything that comes out of the ground.

Steve is in Australia and will respond back Saturday morning his time (Friday evening our time).

Hi Mididoctors,
Fig 4 and case 2 dynamic should be almost the same with two minor differences (first, the x axis in the phd was limited to 2200 (so the long shale oil production from the US is more obvious in Fig 4) - US shale in the scenario was assumed to have a max production limit due to limited water in the region. The other difference is that the Qatar gas production is very slightly modified - due to a very minor bug found in the model (which is now fixed).

I suspect not. I do not think psychohistory has actually been invented.

One could imagine that "above ground" factors might just help to prolong the undulating plateau for a bit, as the oil will still find the surface at some point. And, the longer the undulating plateau lasts, the more skeptical people will be about any real life changing event.

Thats quire a good question... you should cross post to the relevant CC sites

pages 154 - 156 of the thesis - B1TMessage.

Would it be possible to post the Maple source for the model as GNU Octave does very similar things?

At least Maple has source code. Excel (and note that a new version is required, on Windows only) requires the software in order to even read the algorithms used in the calculations. Though I suspect .xlsm file could be decoded without too much effort: it is supposedly a zipped XML file, and XML is plain text.

I do not understand why it is necessary to use an expensive package like Maple, or a business program like Excel (famous for amateurs creating indecipherable behemoths which are not even understood by the author) for this kind of modeling. I've done some of this sort of thing myself using both Perl and Python, either of which is quite capable of handling the job, and both of which are free, and in the case of Python easily understood by someone looking at the source code.

Hi James and others, the model is completely free from Maple now, it has excel to insert the inputs and create the graphs but the model itself lives entirely in C++ (The C++ source code is not publicly available). The model had many significant changes over it's life and began as a very maths heavy model - hence the need for and use of Maple. As the model evolved it became possible and necessary to get the calculations out of Maple and it was rewritten into C++, however due to time constrainsts the graphing and input side of things were left in Maple. Thanks to the ISF the remaining Maple components have been removed and added to the C++ code and excel macro. You can if you wish open the macro easily enough... (google excel macros for how).

...you're on the road to the Olduvai Gorge

Hi JonFreise, just download the file, and run the phd input file and change the axis yourself :-)
The historic data is not included in the projections (I ran out time). however the excel ouput creates the models production, and you can download coal statistics from my coal article, and I think I added historic gas production stats to the gas article that came out earlier this year. Fig 4 represents Case 2 dynamic in the thesis, Case 2 was my own view of what the URR is for each country, and dynamic allows supply and demand to interact - and (was intended atleast) to be the most realistic.

Hubbert did not assume straight depletion decline, he did have post-peak additions. Just not big enough.

I am still skeptical... Hubbert's guess may have been correct, but rates of production are much higher than H considered. In which case, we have a true 'blip' of production, and those optimistic addition figures are just that. And Hubbert might even have overstated additions.

Actually, I guess we should be rooting for the cornucopians, at least short term. We can use the uplift, even if temporary, to get our own ducks in a line. That's what I plan on doing, at least. Positioning in the short to mid term for the advent of long term events hearlding the end of the industrial age. And, hoping (not expecting) that we are able to turn it around in time to salvage most of our civilization.

Best hopes that the new civilization is a good one.

Craig

Yawn.

Based on the information provided, there is no way to gauge the probabilities of the answer because the uncertainties involved are not provided. You have referenced but a few. Without them, you don't know if the answer is high, low, middle of the road or pure BS.

Fill in appropriate Box and Draper quote here.

Hi Gail,

No the model does not account for these uncertainties. That is why I provide three scenarios in the PhD (and articles), Low, Best Guess, High. The model was an attempt to move the argument away from a simplistic equation, and instead attempt(albeit simply) to place more theory into why or why not the production is symmetric and if not how asymmetric.

These models are way too simple. They can only predict a KNOWN and are always ADJUSTED AFTER THE FACT as the KNOWN CHANGES.

I would be willing to bet that if you tried to account for the complexity you mention, and input the proper levels of uncertainty, you will discover why predicting something, particularly the future, is so difficult.

The backcasting you mention is something normally done as a course of business in any engineering study (or even more basic we get the "gee Johnny, that cement didn't hold up at all last time, why are we using it again this time?!" effect), and this type of quality control is hard to find on this topic. Probably for good reason, imagine how much less fun it would be if preceding the next version of the model you had to issue the following disclaimer, "So far this model hasn't worked in a predictive manner on any past data, but lets ignore that and talk about all the interesting things it predicts a few decades from now."

Gee it looks like you only want back to 1980. That is too easy!

The problem is that we can't go back in time and explain to King Hubbert the innovations that have been made in the theory of stochastic processes. You see, the probability models were only starting to get accepted in the 1960s, via books by Feller, Parzen, and Doob, to name a few. Hubbert back in the 1950's made do with some graph paper and a few heuristic models. If a few scientists were diligent about the analysis and kept at it through the years, passing on knowledge to the next generation, we would have an excellent accounting of current and projected supplies. The problem is that no one seemed to care and the effort languished as a set of heuristics, maintained only by a few of Hubbert's disciples including Deffeyes.

You talk about "what ifs", including a trillion barrels in the polar regions. Well it would seem that out of several billion people on the planet, at least a few would have the intellectual curiosity to analyze how long the supply would last.

Hi Drillo,

The assumptions and all equations can be found in my phd (which is on the oil drum).
Re bug: it sounds like the c++ code ran then died in the macro coming out which is strange (usually if it is going to fail it dies calling the executable). Was the excel input sheet the only one open at the time? was the path to the folder less than 200 characters? sorry - I haven't seen this behaviour before so I'm not sure what went wrong - feel free to email me a snapshot of the screen and I'll see if I can reproduce (and then fix) the bug.