[OZ_DOC]

Ok so i fully undestand and agree that at some point we will not be able to take oil out of the gound any faster( peak oil), its like all your mates drinking a can of coke with straws, you can only put so many straws into the can.

What i either dont get, or possibly disagee with is why that has to happen when the amount of oil in the ground reaches half of original reserves? I dont see any reason that these two points must be the same, they may by chance be close to each other, but imho opinion the factors which affect rate of production dont have a lot to do with proven reserves. In fact that's why we all have a rant on this forum when some talking head in the media says that some newly discovered field will solve our problems.

To me peak oil is definite and pobably will occur in the next two years, however i dont see the fundamental linkage between 50% extraction and peak oil?

can anyone enlighten me?

[marek]

The production decline does not have to happen half way. It would only be the case if the actual production were symmetrical. If you produce below the idealized Hubbert curve, then that moment can happen at a higher value of cumulative production.

[bobcousins]

That's a great question. How much do you want to know about mathematical modelling? There are actually a number of ways of approaching your question depending on how you want to understand it.

To put the Peak into Peak Oil Hubbert picked the logistic curve. Normally the logistic curve applies where a growth rate is self-limiting. A classic example is population growth. It is a mathematical "feature" of this curve that maximum rate occurs at 50% exploitation. The way I think of it is as the balance point between two opposing forces - growth and depletion.

A mathematical model is chosen for two reasons - the underlying physics matches the assumption of the model, or the model generates a nice fit with the data without knowing exactly why - if the curve fits then that at least suggests the physics has the same properties as the model. I think in the case of Hubbert a bit of both applies.

There are two reasons why the point may not be exactly 50%. Of course, the maths is an ideal case and the real world is messy, that is the model assumes the parameters stay constant throughout, in reality the parameters have some variation.

The other reason is theoretical, that is the model represents the most important factor (growth is self-limited), but does not encapsulate second order features that affect the curve. (It is not obvious how oil extraction is the same as population growth). These secondary features may shift the peak earlier or later than the midpoint of depletion. However the logistic curve will still produce a curve that is close enough.

theoildrum has some excellent articles on a different model which both has a better match with underlying physical assumptions and also produces a better fit to the data. Even the likes of Deffeyes struggle to get their model to fit the data, and solve the problem by saying "the data is faulty", which is poor science, IMHO. The refined models produce more of a tail after the peak.

I hope that explains some of the modelling, if I made any errors it is because I am using a mini-keyboard. I deliberately avoided getting into the physical properties of oil extraction, in fact I believe it is more to do with economics than geology.

[Dukat_Reloaded]

To make it easy to understand, oil in the ground is much like a saturated cloth. You pick up the cloth and it's totally saturated with water, the water is quickly dripping out of it until it becomes approx half saturated, the amount of water now flowing is greatly reduced and to get the last 1/2 of the water out, you have to expend alot of time and energy to totally ring the water out of the cloth, even after ringing there would be water left but wouldn't be worth your time to expend 100x more energy to get the remaining 20% of the water out.

Basicly when an oil field has 1/2 of the oil taken out, the amount of oil OVER TIME goes into decline. You could still have 30% oil still there, but it's pumping out very slowly and your expending alot of energy trying to pump out the remainder of oil.

[TWilliam]

If I recall correctly, Hubbert's use of the logistic curve was also based, at least in part, on direct observation in the field. Maximum production from any one field tended to occur at roughly the point when about half of the ultimately recoverable oil had been removed. Based on this observation, coupled with the then-known estimates of total recoverable reserves in the US, Hubbert correctly predicted the US production peak in the early 70s. He further predicted world peak would occur in the 90s, but that event was pushed back (I believe by about 10 years or so is the estimate I've seen most) due to the drop in consumption that happened for some years following the first energy crisis (mainly because of increases in efficiency).

[Such]

It's just an assumption Hubbert made that roughly reflected aggregate oilfield behavior. Assuming a symmetric curve makes the math much more straightforward.

Most of the newer technologies allow for the field to reach it's peak production very fast, and then plateau for a period of time before beginning an exponential-like decent. The tail in many fields can carry on for a long time.

We still have stripper wells pumping oil out of the ground here in Akron Ohio after 120 years.

[nero]

$this->bbcode_second_pass_quote('bobcousins', 'T')hat's a great question. How much do you want to know about mathematical modelling? There are actually a number of ways of approaching your question depending on how you want to understand it.

To put the Peak into Peak Oil Hubbert picked the logistic curve. Normally the logistic curve applies where a growth rate is self-limiting. A classic example is population growth. It is a mathematical "feature" of this curve that maximum rate occurs at 50% exploitation. The way I think of it is as the balance point between two opposing forces - growth and depletion.

A mathematical model is chosen for two reasons - the underlying physics matches the assumption of the model, or the model generates a nice fit with the data without knowing exactly why - if the curve fits then that at least suggests the physics has the same properties as the model. I think in the case of Hubbert a bit of both applies.

There are two reasons why the point may not be exactly 50%. Of course, the maths is an ideal case and the real world is messy, that is the model assumes the parameters stay constant throughout, in reality the parameters have some variation.

The other reason is theoretical, that is the model represents the most important factor (growth is self-limited), but does not encapsulate second order features that affect the curve. (It is not obvious how oil extraction is the same as population growth). These secondary features may shift the peak earlier or later than the midpoint of depletion. However the logistic curve will still produce a curve that is close enough.


theoildrum has some excellent articles on a different model which both has a better match with underlying physical assumptions and also produces a better fit to the data. Even the likes of Deffeyes struggle to get their model to fit the data, and solve the problem by saying "the data is faulty", which is poor science, IMHO. The refined models produce more of a tail after the peak.

I hope that explains some of the modelling, if I made any errors it is because I am using a mini-keyboard. I deliberately avoided getting into the physical properties of oil extraction, in fact I believe it is more to do with economics than geology.

[OZ_DOC]

Thanks for the answers guys. Once again correct me if im wrong but what i think your all saying more or less, is that peak oil will occur at around 50% of extraction, as shown by historical data and observations. But mathematically speaking peak oil and 50% extraction are not inextricably linked, ie they are not guaranteed to occur in conjunction but they will approximately align.

regarding the cloth wringing analogy, it was exactly this sort of thought experiment that made me doubt the 50% rule. If you think about it you can probably extract 60-70% of the water from a towel with very minimal effort, and then it is the last 30-40% which requires exponentially rising physical effort to extract fully. In fact come to think of it wringing out a towel is a very good analogy for oil extraction isnt it. The initial extraction is simple, just like a gusher of oil the water literally pours out of the towel. gentle wringing (pumping) will extract a significant portion of water (oil). Finally no amount of forceful wringing (EOR) will extract all the water (oil) and you are left with unrecoverable residual water (oil).

The only reason i raised the question is because a few times PO has been raised in the media (usually mainstream) i have heard it defined as the point at which we have extacted 50%of oil fom the ground. I wondered how true this was or whether it was just a simplified definition for the mathematically challenged in the community.

[Auntie_Cipation]

I think it's also worth mentioning again, that when TSHTF is not tied precisely to the peak as well. It could (and probably will, in my opinion) happen before the technical peak of production. TSHTF as soon as demand is sufficiently beyond supply to cause a wobble in the house of cards that is the finances of the industrial world. So, even if production has a few years of increase left, if increasing demand outstrips it (or tries to) then the troubled times begin.

I don't think it's as simple as "high demand = high prices = demand destruction = now all is well again." High prices WILL cause demand destruction in the form of reduced discretionary travel. shipping, and other uses, but I don't think that will be enough demand destruction to lower the prices sufficiently to keep all the NON-discretionary uses of gas and oil from collapsing in on themselves, financially.