PEAK OIL: Going Mainstream

[jimbob]

Hubbert took a stab at projecting the peak point of global oil production in 1956, and his findings were rejected by the majority of the experts of that time -- precisely the opposite of what you have argued: that the first cries of "peak oil" were heard some 40 years ago. I'm not aware of others besides Hubbert (I'm really only interested in the opinions of experts rather than the ravings of loons in counterculture magazines and whatnot -- which I do agree exist and surely have for some time). If you were referring to Hubbert himself, then we have no further argument there -- as long as you aren't sticking to your implication that those cries included erroneous claims that the peak was ocurring "now" (in '68, '73, '81, etc).

The interesting thing to me is that Hubbert's (basic) model seems to have been pretty accurate, as the 1956 prediction looks to have been sometime around the year 2000, as opposed to the current predictions of some time between 2010-2025.

I would say that this is more surprising than Moore's Law, which has been somewhat self-fulfilling as people have been planning the lithography "nodes" according to the prediction.

 

[Hindmost]

for information

I found a historical rotary rig website...I can't say whether it is accurate or not, but it does look reasonable. It is important to note that drilling technology changed considerably in the past 30 years and a company does not have to drill as many holes in the ground for the same amount of production...angle drilling specifically has increased production from old wells as well as finding more pocket of oil...particularly benificial in west Texas. It has also made certain oil shale deposits somewhat viable.


http://www.wtrg.com/prices.htm

This link will can provide the rig count on a real time basis and have been doing so for many years. Day rates on rigs can be substantial and the price will vary with the oil price. Modern exploration methods have essentially identified all the places on the planet that can produce oil and gas in reasonable amounts. This has made it unlikely that a giant find will occur. Giant find would be about 50 to 100 billion barrels if it is going to make any difference and even that amount would not be a panacea as it takes about 5 to 10 billion in reserves to produce about one million barrels per day at peak production for a field. (very ballpark as it varys greatly with the type of oil)

IIRC: Speculative reserves of conventional oil yet to be found in the US are about 80 billion barrels and that would be P10 type.

http://www.bakerhughes.com/

http://investor.shareholder.com/bhi/rig_counts/rc_index.cfm

http://www.pbs.org/wnet/extremeoil/science/exploration2.html

Major oil conglomerates haven't built a new refinery in the US for a long, long time--just expanded existing plants a bit. No need as existing capacity will be enough.

glenn

 

[Dymanic]

The interesting thing to me is that Hubbert's (basic) model seems to have been pretty accurate, as the 1956 prediction looks to have been sometime around the year 2000, as opposed to the current predictions of some time between 2010-2025.

Its utility as a theoretical construct was demonstrated by the accuracy with which it was able to predict the US peak. Of course, there's always that pesky discrepancy between the theoretical and the real-world, and since global politics and economics are a bit more complicated -- and these do influence production and hence the point of peak production -- predicting the global peak is a correspondingly less perfect science. The qualifying phrase, "if current trends continue" soaks up a lot of potential error. Unlike the global discovery peak, the US discovery peak was well in the past by 1957; the arab oil embargo (and the massive conservation it inspired) hadn't happened yet; OPEC and its production quotas (and the dubious reserve estimates they inspire) didn't exist.

The equation now has so many variables that it's hardly surprising to see projections all over the place, but it's worth noting that even among the minority of experts who question the validity or the applicability of Hubbert's model (favoring, say, the "undulating plateau" over the "roughly symmetrical curve") there are very few who predict an endless (let alone an endlessly increasing) abundance of oil, and (with the exception of a handful of cornucopia loons like Jum'ah) none who predict an endless abundance of cheap oil.

That last point is important, because a lot of the criticisms of peak oil theory are based on anticipation of the development of new technologies that will make it possible to extract oil from sources previously regarded as unrecoverable. As soon as one redefines "recoverable" in this way, the size of remaining reserves changes, and when those new estimates are fed into the model, the projections change. It's not a given that those updated projections are more accurate. If they ignore EROEI, they are almost certain to be unrealistically optimistic. It's easy to forget (apparently) that horizontal and deepwater methods and so forth consume more energy in the extraction process itself.

If those projections are more accurate, it's because the shape of the curve itself has changed; the actual peak point has moved to the right, increasing the steepness of the downslope. It could buy some time, and it's tempting to see that as an entirely good thing. We could use that time to really get on the ball with alternative energies and conservation and optimization and all that, smoothing the inevitable descent. But it also increases the pressure to be successful in those mitigation efforts while at the same time making it even more difficult to muster the political will to implement them, because it will surely create the impression in the minds of a lot of people that the crisis has been solved by technology rather than merely delayed.

 

[JihadJane]

Former Chief Scientist lends voice to "Peak Oil" warnings
Sir David King says politicians have "heads in the sand" over oil supply risks

 

[lomiller]

The interesting thing to me is that Hubbert's (basic) model seems to have been pretty accurate, as the 1956 prediction looks to have been sometime around the year 2000, as opposed to the current predictions of some time between 2010-2025.

I would say that this is more surprising than Moore's Law, which has been somewhat self-fulfilling as people have been planning the lithography "nodes" according to the prediction.

I don’t find it surprising at all. Hubbert’s basic model is something’s of a tautology for certain resources.

For a sufficiently profitable resource new investment yields production, which yields profits which yields even more investment. This means that as long as returns are roughly linear with new investment you get an exponential ramp up of production.
This isn’t dissimilar to what happens with Moores law expect that in the case of Moores law there is also exponentially diminishing returns on investment.

At some point, however, you will no longer yield significant returns on new investment. In the case of raw material resources, and probably some other items what happens is some other resource becomes cheaper so nearly all investment in the existing resource stops. With the cessation of new investment you get an exponential decline in production as the already developed resources run out.

Any time you get a exponential increase followed by an exponential fall you get a point someplace near the middle where you have produced ~1/2 of your total. Hubbert in effect has causation backwards. The peak at 1/2 total production is a result of the exponential climb then decline not a predictor of it.

 

[jimbob]

Lolmiller, I'd quibble with your last statement about cause and effect. It can still be used as a predictor, which would reflect the fundamentals, as outlined in your earlier post.

If you look at real curves:

You see that the actual shape is a bit more rectangular towards the peak, but despite this, and with several significant oilfields undiscovered at the time, (I guess) his curve gave a pretty good predictor of 45-years in the future, as opposed to 55-65 years.

That was what I was finding surprising, not the shape, but that although real curves are more complex, the original prediction was pretty close.

 

[Dymanic]

In the case of raw material resources, and probably some other items what happens is some other resource becomes cheaper so nearly all investment in the existing resource stops. With the cessation of new investment you get an exponential decline in production as the already developed resources run out.

The logic of that doesn't track for me. The concern over Peak Oil is that it represents the end of cheap oil: it's not that there aren't any other sources that could do for us what oil does, it's that there aren't any that could do it as cheaply. Ultimately, what we're concerned about is the money. If some other resource becomes cheaper, and if the reason for that is that oil has become more costly to extract and hence prohibitively expensive, then it seems peculiar to consider the problem to have resolved itself.

Hubbert in effect has causation backwards. The peak at 1/2 total production is a result of the exponential climb then decline not a predictor of it.

That doesn't make sense to me either. You don't seem to be disputing whether the peak occurs at 1/2 of total production. Once it is established that that is in fact the case, then -- whether the reasons have more to do with economics or geology or engineering -- why can that observation not then be used as a predictor? Whether you're saying that what you're predicting is the peak or what you're predicting is the halfway point, doesn't it really become a matter of accurately predicting where that (those) will occur?

 

[lomiller]

The logic of that doesn't track for me. The concern over Peak Oil is that it represents the end of cheap oil: it's not that there aren't any other sources that could do for us what oil does, it's that there aren't any that could do it as cheaply.

The end of cheep oil is actually a decidedly different topic from Hubbert peak oil which describes a very specific curve, one that isn’t likely to be followed when oil is expensive enough to restrict demand but still cheaper then competing energy sources.

That doesn't make sense to me either. You don't seem to be disputing whether the peak occurs at 1/2 of total production. Once it is established that that is in fact the case, then

The conditions for the second half of that curve would be that oil gets replaced by some other cheaper energy source. At the moment that doesn’t appear to be happening, so we can’t be at a Hubbert peak. This doesn’t preclude a non-Hubbert “plateau” where ever increasing prices make marginal oil sources profitable enough to bring to market despite the high cost of doing so.

Again a Hubebrt peak is fairly specific, production peaks and then begins to decline precipitously. We may have already hit peak global oil production or at least be close to it, but the rapid decline predicted by Hubbert is nowhere in sight.

 

[Dymanic]

The end of cheep oil is actually a decidedly different topic from Hubbert peak oil which describes a very specific curve, one that isn’t likely to be followed when oil is expensive enough to restrict demand but still cheaper then competing energy sources.

It's not clear to me where we leave off talking about the point of peak oil production and begin talking about the consequences of that event. Because demand and price always exert a strong influence on production, I see the two as inseparable at every point. You appear to agree, and have made a number of arguments based on precisely that. I don't understand why you would then turn right around and refer to them as decidedly different topics.

The conditions for the second half of that curve would be that oil gets replaced by some other cheaper energy source.

I wonder how that argument looks when we apply it to the peak of US production which took place in 1970. What cheaper energy source was it again that replaced all that domestic oil?

Again a Hubbert peak is fairly specific, production peaks and then begins to decline precipitously.

To the extent that that's true, it's because the Hubbert curve is a theoretical construct, one which would not be expected to provide anything but a rough approximation, and very prone to variation when given different sets of inputs.

 

[Hindmost]

The end of cheep oil is actually a decidedly different topic from Hubbert peak oil which describes a very specific curve, one that isn’t likely to be followed when oil is expensive enough to restrict demand but still cheaper then competing energy sources.




The conditions for the second half of that curve would be that oil gets replaced by some other cheaper energy source. At the moment that doesn’t appear to be happening, so we can’t be at a Hubbert peak. This doesn’t preclude a non-Hubbert “plateau” where ever increasing prices make marginal oil sources profitable enough to bring to market despite the high cost of doing so.

Again a Hubebrt peak is fairly specific, production peaks and then begins to decline precipitously. We may have already hit peak global oil production or at least be close to it, but the rapid decline predicted by Hubbert is nowhere in sight.

What is important to note about Hubbert's prediction was that fields would have be developed and pumped at unrestricted extractions rates continuously throughout the life of the fields--essentially all the oil that can be pumped out is pumped out. Hubbert did not predict there would be a precipitous decline past the peak. He actually expected a reasonably symetric curve.

The world's oil fields have not had this type of full flow production and therefore the peak will be more difficult to focus on a specific time. This is why a "flat" top to peak oil is expected. With enhanced oil recovery, the downslope should actually be more gradual as the extraction rates can be higher---but the energy return per barrel will be less at is takes more energy to get the oil out. Enhanced recovery can damage fields however, and the downslope is not guaranteed to recover more oil.

http://www.theoildrum.com/story/2006/11/20/91748/298


glenn

 

[lomiller]

It's not clear to me where we leave off talking about the point of peak oil production and begin talking about the consequences of that event. Because demand and price always exert a strong influence on production, I see the two as inseparable at every point. You appear to agree, and have made a number of arguments based on precisely that. I don't understand why you would then turn right around and refer to them as decidedly different topics.

I suspect you don’t quite realize what claim you are making, which would account for you not understanding arguments against us. Hubbert wasn’t just arguing there would be some peak in oil production he was arguing that once that peak came we would see a decline in production that mirrored the previous rapid increase in production. It’s the second half of that claim that differs from mainstream estimates of oil production.

I wonder how that argument looks when we apply it to the peak of US production which took place in 1970. What cheaper energy source was it again that replaced all that domestic oil?

Imported oil

 

[JoeTheJuggler]

You might notice that the decline is gradual. Peak production doesn't mean zero production immediately afterwards. In fact I understand that the top of the curve tends to ba a bit flatter than this simplistic model.

The UK (North Sea) produced peak oil before 2008.

This is true, but imagine the graph of oil consumption/demand.

If we don't make some drastic changes soon, demand for oil in say 2030 will not be anywhere near as low as it was in 1950. In fact, demand will far outstrip supply after we reach the peak.

That is, even if there is a substantial flat spot at the peak, the effect of ever increasing demand would be the same as if demand were constant and production dropped off precipitously.

 

[jimbob]

This is true, but imagine the graph of oil consumption/demand.

If we don't make some drastic changes soon, demand for oil in say 2030 will not be anywhere near as low as it was in 1950. In fact, demand will far outstrip supply after we reach the peak.

That is, even if there is a substantial flat spot at the peak, the effect of ever increasing demand would be the same as if demand were constant and production dropped off precipitously.

Indeed, if the peak demand is kept up, then the oil supply runs out faster in the longer run.

In the exchange below I think the answer is that oil demand can't exceed supply, because the price will rise until the demand goes away.

The optimist's slant on this is taken by drkitten, and I think it is a fair summation of this argument that as the price rises, technological development will be encouraged, and eventually a substitute will be found and the price will fall (cf the Simon-Ehrlich wager discussion).

I tend towards the view "that past performance is no indication of future performance" and indeed that the Simon-Ehrlich wager probably means very little, looking at the how volitile these commodity prices are.

To flesh out your statement a little, Joe, the US DOE is projecting an increase in demand to about 130% of current levels by 2030 with an "unidentified" component making up about 43/105 of this supply - In this case I read "unidentified" as meaning "hoped for without much evidence of its existnace".

Page#2 and page#7 of this report (pdf)

http://www.eia.doe.gov/conference/2009/session3/Sweetnam.pdf
suggest that the demand for "oil" is expected to increase to about 130% of the current demand by 2030. (about from about 80 to 105-Million Barrels/day in 2030) with an "unidentified" component of 43-million barrels.

This seems quite a lot.

ETA: What are the copyright rules on using images from US Governmental websites? That's why I didn't post the graph.

Joe the Juggler has answered for me.

The existance of strategic oil reserves is why I used the word "approximately".

I was responding to DrKitten's statement that it took 200-years to reach this level of production and would take about another 200-years for it to reach zero.

The problem isn't the amount produced, it is the difference between supply and demand.

Production does indeed look to be a symmetric curve. But (potential) demand ("need") isn't.

Of course the price will rise until demand falls and meets supply. However there could be some bad effects from this, for example on food prices.

Yes but "(potential) demand" doesn't exist. Potential demand for anything is infinite -- that's one of the fundamental axioms of economics. Wants are unlimited. If oil were free, everyone would drive Hummers -- or helicopters --- or both.

China would like to burn a lot more oil than it does, but can't afford it. The United States would like to burn a lot more oil than it does, but can't afford it. Estonia would like to burn a lot more oil than it does, but can't afford it. As the supply drops and the cost of production rises, Estonia (and the USA) will be able to afford less and less oil, which will mean people will get rid of their helicopter, then their Hummer, then their SUV, and eventually learn that commuting by bicycle isn't all that bad.... The big train companies, which can outbid me for oil but which also use it much more efficiently will still be burning oil and handle the long distance travel needs that I would do in my own private helicopter-Hummer-Learjet if oil were free.

What I should have said, is the amount to sustain the current economic activity our way of life.

 

[Dymanic]

Hubbert wasn’t just arguing there would be some peak in oil production he was arguing that once that peak came we would see a decline in production that mirrored the previous rapid increase in production.

A roughly symmetrical curve, yes. What does that even have to do with the post you are responding to here?

It’s the second half of that claim that differs from mainstream estimates of oil production.

Differs from mainstream estimates, or from cornucopian estimates? (Are you supposing that those are the same?)

What cheaper energy source was it again that replaced all that domestic oil?

Imported oil.

Progress. Now: why was it that imported oil suddenly represented the cheaper energy source?

 

[jimbob]

I suspect you don’t quite realize what claim you are making, which would account for you not understanding arguments against us. Hubbert wasn’t just arguing there would be some peak in oil production he was arguing that once that peak came we would see a decline in production that mirrored the previous rapid increase in production. It’s the second half of that claim that differs from mainstream estimates of oil production.

This is only a blog but the graphs are sourced, so looks pretty reliable to me at first glance.

Thee plots for both the North Sea Oil and the US (especially the North Sea) *do* look symmetrical so far to me.

 

[lomiller]

A roughly symmetrical curve, yes. What does that even have to do with the post you are responding to here?

It’s what Hubbert predicted and what you are attempting to defend. Hence my point about you not really understanding the claim you are making.

Differs from mainstream estimates, or from cornucopian estimates? (Are you supposing that those are the same?)

No, mainstream estimates. Real geologists looking at what oil is really in the ground and how much it would cost to extract.

Progress. Now: why was it that imported oil suddenly represented the cheaper energy source?

Because it’s cheaper to extract. This does nothing to prove your contention that the immanent exponential drop in production predicted by Hubbert has any merit.

 

[Dymanic]

It’s what Hubbert predicted and what you are attempting to defend. Hence my point about you not really understanding the claim you are making.

One of us appears not to understand the claim I'm making. I hope this misunderstanding doesn't reduce to quibbling over the shape of the curve at the top or the definition of "precipitous".

No, mainstream estimates. Real geologists looking at what oil is really in the ground and how much it would cost to extract.

Whether or not I understand the claim I'm making, it looks like I'm still not clear on whether I understand the claim you are making, because I've been assuming that (unlike those "real geologists" who DO include geological constraints) you are using a strictly demand-based approach. Maybe we both ought to try to clarify. I'll go first:

As far as I can see, very few of the "real geologists" appear optimisic about the prospects for maintaining (let alone increasing) the rates of extraction we've seen in recent decades. On the other hand, none of them predict a sudden and dramatic decline in production, either. I therefore predict a peak in production (possibly a lingering at a peak plateau) followed by a decline which, being neither a vertical cliff nor a gentle downward slope, would conform reasonably well to a Hubbert curve at a low level of resolution while at a higher level of resolution reflecting a lot of market volatility (as I noted at the top of page three in this thread).

Because it’s cheaper to extract.

And why is it cheaper to extract?

This does nothing to prove your contention that the immanent exponential drop in production predicted by Hubbert has any merit.

Remember now, we're not talking about the imminent drop in global production, we're examining an argument you made: "The conditions for the second half of that curve would be that oil gets replaced by some other cheaper energy source" by seeing how it holds up when applied to Hubbert's 1956 prediction that US production would peak by 1970 (the merits of which appear self-evident from this vantage point in time).

 

[lomiller]

One of us appears not to understand the claim I'm making. I hope this misunderstanding doesn't reduce to quibbling over the shape of the curve at the top or the definition of "precipitous".

Hubbert peak oil is all about the shape of the curve, so you can’t begin to discuss it without such “quibbling”.

Remember now, we're not talking about the imminent drop in global production,.

Hubbert peak’s which you are clearly defending specifically predict such a drop off so yes in fact you are defending the idea of an immanent drop.


Allow me to quote my original post you initially took issue with:

The end of cheep oil is actually a decidedly different topic from Hubbert peak oil which describes a very specific curve, one that isn’t likely to be followed when oil is expensive enough to restrict demand but still cheaper then competing energy sources.

If you have abandoned the peak oil curve what exactly are you trying to argue, because you now seem to be advocating my position when you started off by disputing it. Goalpost move much...

we're examining an argument you made: "The conditions for the second half of that curve would be that oil gets replaced by some other cheaper energy source" by seeing how it holds up when applied to Hubbert's 1956 prediction that US production would peak by 1970 (the merits of which appear self-evident from this vantage point in time).

I see...

How can it possibly be that you keep insisting you are not arguing anything about the shape of the Hubbert curve when the specific thing you are taking issue with is my statement his curve is unlikely. This isn’t just a goalpost move it’s completely non-sequitur.

 

[Dymanic]

Hubbert peak oil is all about the shape of the curve, so you can’t begin to discuss it without such quibbling.

As I've already stated, the Hubbert curve is a theoretical construct. No one, not even Hubbert himself, would expect real-world events to conform exactly to that curve. I don't know if you've read the original work, but a key phrase is "if current trends continue".

If you have abandoned the peak oil curve what exactly are you trying to argue, because you now seem to be advocating my position when you started off by disputing it.

We do seem to be talking past each other here, and at this point I'm not even sure just exactly what we do and do not agree on. Again, I made my position clear at the top of page three of this thread in a post made on March 27:

If we could expect a nice, steady, predictable decline in oil production, it would be a lot easier to make whatever transition we need to make. Unfortunately, it seems more reasonable to expect markets to become increasingly volatile. Production falls off, price goes up, producers scramble to produce more, consumers scramble to consume less, industries scale back; lots of interest in alternatives -- and because of all those things happening at once: supply goes up, price eases some, everybody relaxes a little; it was just a hiccup. Less interest in alternatives. I could see things lashing back and forth like that for quite a while.

Perhaps in your mind a model like Hubbert's is useful as a tool for prediction (or validated post-hoc) only if the actual data conform very rigidly to the model. I don't; never have; and am not aware of having made any statement to that effect.

Now: You still haven't returned the favor I asked for by clarifying just what your position is. Shall I assume that your stance is that oil production will continue to increase indefinitely if driven by sufficient demand, or have I misread you entirely?

How can it possibly be that you keep insisting you are not arguing anything about the shape of the Hubbert curve when the specific thing you are taking issue with is my statement his curve is unlikely.

Oil industry analysts in 1956 also considered Hubbert's curve "unlikely" where US domestic production was concerned, and continued to feel that way until somewhere around the mid-seventies when it became apparent that production was indeed following that curve. Have YOU compared the actual production numbers to see how they compare to Hubbert's curve? I would describe that as a reasonable fit, but not perfect (the main discrepancy being that the actual decline was initially somewhat steeper than Hubbert predicted). Would you agree? If so, would you agree that it would also have been a reasonable-if-not-perfect fit if the actual peak had been flatter rather than sharper?

 

[jimbob]

This is true, but imagine the graph of oil consumption/demand.

If we don't make some drastic changes soon, demand for oil in say 2030 will not be anywhere near as low as it was in 1950. In fact, demand will far outstrip supply after we reach the peak.

That is, even if there is a substantial flat spot at the peak, the effect of ever increasing demand would be the same as if demand were constant and production dropped off precipitously.

Coming back to the following post be drkitten

Yes but "(potential) demand" doesn't exist. Potential demand for anything is infinite -- that's one of the fundamental axioms of economics. Wants are unlimited. If oil were free, everyone would drive Hummers -- or helicopters --- or both.

China would like to burn a lot more oil than it does, but can't afford it. The United States would like to burn a lot more oil than it does, but can't afford it. Estonia would like to burn a lot more oil than it does, but can't afford it. As the supply drops and the cost of production rises, Estonia (and the USA) will be able to afford less and less oil, which will mean people will get rid of their helicopter, then their Hummer, then their SUV, and eventually learn that commuting by bicycle isn't all that bad.... The big train companies, which can outbid me for oil but which also use it much more efficiently will still be burning oil and handle the long distance travel needs that I would do in my own private helicopter-Hummer-Learjet if oil were free.

By the above reasoning, food demand could never exceed supply, because the price would rise to equalise supply and demand, so fammine wouldn't exist.

Or am I missing something?