Peak Oil
Donate Bitcoin

Donate Paypal


PeakOil is You

PeakOil is You

The Problem With Predictive Modeling

Discuss research and forecasts regarding hydrocarbon depletion.
Post a reply

Re: The Problem With Predictive Modeling

Unread postby netfind » Sat 25 Feb 2006, 21:51:07

Given the chaotic production curves of both individual fields and entire countries, it's no wonder few people place much confidence in math curves to predict our future oil situation. But if you consider that the things that twist production curves away from smooth math curves are economic/political forces and are somewhat quantifiable, a math curve superimposed on a production history can have predictive value.
Last edited by netfind on Sun 26 Feb 2006, 10:55:56, edited 4 times in total.
User avatar
netfind
Wood
Wood
 
Posts: 14
Joined: Sun 13 Nov 2005, 04:00:00
Location: The Ozarks
Top

Re: The Problem With Predictive Modeling

Unread postby WebHubbleTelescope » Sat 25 Feb 2006, 23:01:08

$this->bbcode_second_pass_quote('netfind', 'G')iven the chaotic production curves of both individual fields and entire countries, it's no wonder few people place much confidence in math curves to predict our future oil situation. But if you consider that the things that twist production curves away from smooth math curves are economic/political forces and are somewhat quantifiable, a math curve superimposed on a production history can have predictive value.


Funny you should mention that, because that is what the Oil Shock model accomplishes. It applies a series of transfer function to a discovery curve that is then perturbed by geopolitical "shocks" that follow the real-world transients occurring in the observed oil production curves. What's surprising is that the tweaks needed to follow the production curve often turn out to be relatively small fractional changes in the extraction rate.

Current progress on the oil shock model here:
http://mobjectivist.blogspot.com/2006/0 ... d-gas.html
User avatar
WebHubbleTelescope
Tar Sands
Tar Sands
 
Posts: 950
Joined: Thu 08 Jul 2004, 03:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby netfind » Sat 25 Feb 2006, 23:47:48

"Transfer function to a discover curve" sounds a little like what I've done with some curves. I'm trying to post these charts for Russia but I'm like so all HTML illiterate it's like really really pathetic. What line of HTML do you use to add an image to a post anyway ? HELP ! Anybody.
User avatar
netfind
Wood
Wood
 
Posts: 14
Joined: Sun 13 Nov 2005, 04:00:00
Location: The Ozarks
Top

Re: The Problem With Predictive Modeling

Unread postby ReserveGrowthRulz » Sun 26 Feb 2006, 01:03:30

$this->bbcode_second_pass_quote('WebHubbleTelescope', '
')It's funny that no one in the petroleum industry wants to calibrate this analagous growth factor, and therefore predict with more certainty how much we will have in store for the future.


Hubbert, R&A, Verma and Ulmishek, Klett, all of these people having taken cracks at calibrating the effect in various environments, under various reporting rules, during various time frames.

What part of their work do you not consider "calibration"?
User avatar
ReserveGrowthRulz
Permanently Banned
 
Posts: 813
Joined: Fri 30 Dec 2005, 04:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby WebHubbleTelescope » Sun 26 Feb 2006, 01:22:10

$this->bbcode_second_pass_quote('ReserveGrowthRulz', '')$this->bbcode_second_pass_quote('WebHubbleTelescope', '
')It's funny that no one in the petroleum industry wants to calibrate this analagous growth factor, and therefore predict with more certainty how much we will have in store for the future.


Hubbert, R&A, Verma and Ulmishek, Klett, all of these people having taken cracks at calibrating the effect in various environments, under various reporting rules, during various time frames.

What part of their work do you not consider "calibration"?


Umm, like why isn't it automatically done for all predictions?

Oh, I forgot, I guess it has something to do with SEC regulations prohibiting "speculative" estimates.

Once again I have to laugh at the imbeciles running things.

If these guys were in charge at Intel instead of Andy Grove, we would still be soldering discretes on circuit boards.

No wonder we laughed at all the rockheads who flunked out of engineering school and transferred over to geology.

:) :) :) :)
User avatar
WebHubbleTelescope
Tar Sands
Tar Sands
 
Posts: 950
Joined: Thu 08 Jul 2004, 03:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby WebHubbleTelescope » Sun 26 Feb 2006, 01:31:44

$this->bbcode_second_pass_quote('netfind', '"')Transfer function to a discover curve" sounds a little like what I've done with some curves. I'm trying to post these charts for Russia but I'm like so all HTML illiterate it's like really really pathetic. What line of HTML do you use to add an image to a post anyway ? HELP ! Anybody.


On this message board, you click on the little icon that shows a tiny mountain range. It looks like this:
Image
That presupposes you have a GIF or JPEG someplace on a server already.

If you don't, you will have to upload the GIF image (or JPG) as well. If you don't have a server readily available, you can use flickr.com or the one I prefer, ImageShack, at this site:
http://img113.echo.cx/

Upload your image and then copy&paste the URL into the image prompt above ^ Image

I look forward to anything you can add to the discussion.
User avatar
WebHubbleTelescope
Tar Sands
Tar Sands
 
Posts: 950
Joined: Thu 08 Jul 2004, 03:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby ReserveGrowthRulz » Sun 26 Feb 2006, 03:16:03

$this->bbcode_second_pass_quote('WebHubbleTelescope', '')$this->bbcode_second_pass_quote('ReserveGrowthRulz', '')$this->bbcode_second_pass_quote('WebHubbleTelescope', '
')It's funny that no one in the petroleum industry wants to calibrate this analagous growth factor, and therefore predict with more certainty how much we will have in store for the future.


Hubbert, R&A, Verma and Ulmishek, Klett, all of these people having taken cracks at calibrating the effect in various environments, under various reporting rules, during various time frames.

What part of their work do you not consider "calibration"?


Umm, like why isn't it automatically done for all predictions?

Oh, I forgot, I guess it has something to do with SEC regulations prohibiting "speculative" estimates.



So now you are back to pretending that the SEC reporting regulations have anything to do with reserve growth in Russia? As for things being done for all predictions....things are done...they are called reserves, they are defined by the SPE and they are calculated by reasonable and semi-consistent rules. Reserves ARE predictions...but you already know that, and want to blame the SEC for two sets of predictions being different over a span of years? Must be nice to work in an industry where you can't inprove on your performance given a couple years more practice and better technology.....how do those slide rules work for performing computational analysis nowadays?

$this->bbcode_second_pass_quote('WebHubbleTelescope', '
')Once again I have to laugh at the imbeciles running things.

If these guys were in charge at Intel instead of Andy Grove, we would still be soldering discretes on circuit boards.

No wonder we laughed at all the rockheads who flunked out of engineering school and transferred over to geology.


The engineers I went to school with didn't flunk out to geology, they became math teachers. Attanasai isn't a geologist, he's an economist. Verma is a petroleum engineer. And both of them know that the SEC doesn't have anything to do with reserve growth in Russia, a fact which apparently escapes people with an agenda to disprove the idea in the first place...and I'm guessing neither could care less about Intel or Groves because the guy probably couldn't find, drill or produce oil without a map pointing him to Ghawar and a petroleum engineer to hold his hand so he doesn't fall down a well and hurt hisself.
User avatar
ReserveGrowthRulz
Permanently Banned
 
Posts: 813
Joined: Fri 30 Dec 2005, 04:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby WebHubbleTelescope » Sun 26 Feb 2006, 11:37:41

$this->bbcode_second_pass_quote('ReserveGrowthRulz', '
')So now you are back to pretending that the SEC reporting regulations have anything to do with reserve growth in Russia?


As for Verma, here is a telling quote (USGS):
$this->bbcode_second_pass_quote('', 'V')olga-Ural and West Siberia provinces show most of their reserve growth in the first 5-7 years after discovery and little or no growth thereafter. It is difficult to compare the growth in Russian fields with those of the U.S. fields where growth continues even after 90 years, because in Russia oil fields are first evaluated over a 5-7 year period before being produced whereas in the U.S. both the evaluation and production of fields start shortly after their discovery.


Read again, Verma says they "show most of their reserve growth in the first 5-7 years after discovery and little or no growth thereafter." If the guy didn't mean to say this, well, he shouldn't have said it. He makes it look like Russia does very nicely with their estimation techniques.

[sub]And don't give me any of that baloney about other EOR methods not being included in these estimates unless you can say what fraction this includes. [/sub]
User avatar
WebHubbleTelescope
Tar Sands
Tar Sands
 
Posts: 950
Joined: Thu 08 Jul 2004, 03:00:00
Top

Re: The Problem With Predictive Modeling

Unread postby rockdoc123 » Sun 26 Feb 2006, 12:01:33

$this->bbcode_second_pass_quote('', 'R')ead again, Verma says they "show most of their reserve growth in the first 5-7 years after discovery and little or no growth thereafter." If the guy didn't mean to say this, well, he shouldn't have said it. He makes it look like Russia does very nicely with their estimation techniques.


Comparing US and Russian exploration is an apples and oranges argument. In Russia the traditional exploration technique was to drill an exploration well, if it was a discovery the field was delineated by further drilling prior to bringing the field on production, often they ended up with wells spaced at 100 acres prior to bringing the field onstream. As a consequence it is not surprising they have a better idea of ultimate reserves early on. In the US and many other places fields are brought on production as soon as they believe an economic reserve limit has been met....sometimes this can be with very few wells into the field. This type of development makes a lot more sense from the standpoint of economics....taking advantage of the time value of money.

The Russians simply have more information.....which is what reserves growth is all about.
User avatar
rockdoc123
Expert
Expert
 
Posts: 7685
Joined: Mon 16 May 2005, 03:00:00
Top
Previous
Post a reply

Return to Peak oil studies, reports & models

Who is online

Users browsing this forum: No registered users and 0 guests



cron