I'm no good at this new fangled "math" thing that seems to be all the rage these days, but wouldn't that mean that in the 1970s, at the rate it was/is going, at about 10 million per 5 years, there were 10 million barrels being produced? Or in 1950 there were less than 1 million being produced? Just wondering how the graph would look if extended from the 1900s or so...
Something like this.
There was a big change in 1973-1974 during the oil shock, and again after the 1979 oil shock. The growth rate prior to 1974 is much steeper than after 1984. (That decade in between was a mess). The average growth rate through the 60's was 8% per year. For the past 20 years, the growth rate has been only 1.5%.
Here's another way to look at exactly the same data, but without the exponential graphing paper.
Now it looks like we're into another change, to a negative growth rate. The question is... how fast will it be? 1.5%? 8%? higher? lower? in between?
I think another question, though, would also be, is the change in growth necessarily from ... something... environmental reasons? What I mean is, could the change in the growth rate be related to a group of men in a dark board room saying "Mwahaha let's only produce X barrels this year, that way the prices go up and we get rich, rich rich!"? I dunno. It's 6 and I'm tired and have been up for 19 hours so anything I say is going to sound very... silly.
Well, that certainly seems to be what happened in the oil shock of 1973-1974, so it certainly should be something to look into.
On the other hand, we know that peak oil (and decline) is close. A few reasons...
1. No more oil refineries being built. Why? Don't need any more if the oil supply is already at maximum.
2. All drilling rig and crews, even old ones, are at 100% usage. Why? Because it is taking that much drilling to simply keep production where it is now.
3. Oil discoveries peaked about 40-50 years ago, and we're now only finding about 10-20% of that each year. Why? Because we're extremely good at finding oil, and we've found all the super giant fields, and probably most of the giant fields. There isn't that much left to find.
4. Saudi Arabia keeps on talking about increasing supply, but they have actually decreased production. Why? They probably can't increase production much without ruining their fields. They also probably know they are at or near their own peak, but they don't want to alarm the world.
5. Over 50 oil producing countries are already past their peaks, and into declines. More countries are moved from the pre-peak column to the post-peak column each year. Why? Because the various countries have been very effective at finding oil, and extracting it fast. Once a critical mass of countries is on the decline post-peak side, the increasing pre-peak side will be overwhelmed.
So, while I think that it could be a group of men in a dark room, these men would only be slightly delaying the inevitable decline.
Take a look at that graph up above that shows the 8% growth rate up until 1973. That growth was not sustainable. In fact, after a few more years of that kind of growth, the rate would have had to change anyway. The group of men (back then) didn't do anything that nature would not have done anyway... although they did it quite a bit faster (decreasing supply), and got rich in the process.
By the way, with 4% growth per year, that doesn't mean that the growth began only 25 years ago. That's due to the compounding effect. A 4% growth rate would extend back over a hundred years, and still be at over a million barrels per day.
See pretty graph below.
Hey, 19 hours is a long day. Go to sleep!
2007-05-18 08:45 pm (UTC)
Finding a curve that fits a few existing data points and then thinking that extending this curve is a good predictor of the future is unconvincing. If it worked I would be making a fortune from the stockmarket!
The uncomfortable fact is that you could fit many different curves to the data in a variety of different ways and end up with many different "predictions" - take your pick according to when you want peak oil to be (or even if you want a peak at all).
The anonymous guest brings up a fine point. One has to understand a set of data before one can fit a curve to it.
We know that oil production will peak. The evidence is overwhelming that we are in a "growth recession", so to speak, for oil. Fitting a curve to the for the depletion of a natural resource is nothing like fitting a curve to money (stocks) which are ephemeral and based primarily on the belief. So, while it may be easy to say this data and curve are dismiss able, they are not.
The fit of the curve is my best estimate of where the supply will go, if there are no unusual disruptions. Clearly, as a best prediction, I expect that the actual curve will be above or below what I drew. I would say that I'm about 90% confident that the actual curve will end up being within 1% of through the end of 2007.
Predicting the future is a matter of taking all the best data and making a hypothesis. Natural resource depletion is very predictable. King Hubbert did it for oil in the lower-48 US states 13 years before it happened.
2007-05-19 11:13 am (UTC)
Interesting that when you want to show oil DEMAND going up you fit a straight line (or even an exponentially increasing one), but when you want to show SUPPLY peaking you fit a polynomial over just a few years that has such a peak. And yet the SUPPLY and DEMAND historical data points ARE EXACTLY THE SAME! (supply=demand, basic economics). Same data, two different curves depending upon what it is you want to show...
A straight line, or exponentially expanding one, are not my projections. Rather, they are projections of the oil industry, EIA, and others, such as Cambridge Energy Associates. Notice, these are projections for the future.
My full length video presentation (available on Google Video, 48 minutes) should point out that the supply curve and demand curve always end up matching. I think that same concept is in my 10 minute edited summary video.http://localfuture.org/peak_oil_summary_transcript_20060722.htm
A few years ago, it looked like growth of 4% might be possible. If oil futures contract buyers and sellers believed this, then it would have depressed the price. Based on the last 20 years, 4% would have been a very optimistic projection, but not out of line with pre-1973 levels of 8%. In this case, belief in future supply is a major component in the price level.
A 1.5% growth projection makes more sense, since that does square with increases since 1983. Of those predicting 1.5% supply growth, supply is now be in a shortfall of about 3 mbpd (as opposed to 8 mbpd).
Supply and demand do tend to go together, but do not fit exactly, depending on the definition of "supply". There are four definitions and measures to keep in mind.
1. Extraction rate - how much are we getting out of the ground during any given time period
2. Consumption rate - how much are we burning or otherwise consuming during any given time period
3. Supply - how much is available for sale during a given time period
4. Demand - how much is actually purchased during a given time period
Economists tell use that supply always equals demand, but this fails to point out that there may be "excess capacity", meaning the extraction rate can be increased, OR there may be an increase in the storage tanks of the oil producers, which is not being sold.
To learn more about supply and demand for oil, and how that impacts prices, the article and accompanying screen cast on inelasticity in oil supply and demand are good starting points.http://valuesystem.livejournal.com/13169.html
In a nutshell, extremely inelastic supply and demand causes volatility in the price level. Change in either will drastically move the equilibrium point. As oil extraction goes into decline, this becomes a major issue.
"Demand will be going up, but it will be constrained by supply," Mulva said. " I don't think we are going to see the supply going over million barrels a day and the reason is: Where is all that going to come from.
As price increases, the consumption level decreases.
Ben Bernanke mentioned yesterday that ratio is around 10 to 1... a 10% increase in price lowers consumption by 1%.
2007-05-19 12:34 am (UTC)
Maybe, but the EIA chart of 5/7/07 speaks for itself!
We are at plateu stage now.
2007-05-19 08:59 pm (UTC)
Looks to me like we were also on a "plateau" in '74 and '79.
Welcome to the new plateau!
2007-05-19 09:04 pm (UTC)
whoops, forgot to also mention the minor plateaus of '84, '92, '98, '01, etc...
2007-05-22 12:07 pm (UTC)
You make a good point about those previous plateaus (plateaux?). But consider all the factors that are "conspiring" to occur during this particular, unprecedented plateau:
1. This plateau comes within the timeframe of peak predictions given by Professor Deffeyes (2005 crude + condensate), Colin Campbell and ASPO (2005, conventional), Ali Samsam Bahktiari, former of Iran's national oil company (2005-2007), and others.
2. This plateau coincides with the highest oil prices in over twenty-five years.
3. This plateau comes forty years after oil discoveries world-wide peaked (just as the peak in US production occurred forty years after the peak in US discoveries).
4. This plateau "coincides" with the brutal fact that the largest oilfields on the planet -- some producing oil for fifty years -- are now in decline: Cantarell (MX), Burgan (KW), Daqing (CH), and yes, Ghawar (SA).
5. This plateau is accompanied by a decided lack of spare capacity, which over the last 5 years has decline from 10 million to fewer than 1 million barrels.
6. Apparently, there is no new oil forthcoming from Saudi Arabia: there extraction rate has been in decline for two years. They claim it is "voluntary."
7. Wars, skirmishes, revolts are "happening" to break out in the oil producing regions of the world during this "plateau," including an invasion by the by-far most oil-hungry nation into the by-far most undeveloped oil-rich region of the middle east.
8. Oil and fuel prices are at all-time highs, demand shows no sign of slackening, but instead of pumping more oil to service this demand and make a killing, we're at a PLATEAU.
This is called "preponderance of evidence."
2007-05-22 12:09 pm (UTC)
goodness, I wish I could fix those spelling errors now ("there"? ugh.)
Well he would be making a worthwhile point if the Labour Theory of Value on which his argument rests had not been thoroughly refuted by the Austrian School of economics through Carl Menger's theory of marginal utility more than a century ago.
It is sad it is coming to this. I don't want to be vindictive about how our country has squandered its wealth and its oil. Hopefully the peak is a long nearly-horizontal line. Even if that means rising prices for increased extraction costs.
I've read "oil on the brain" recently. It goes in to how much deeper they drill in Texas these days to find oil. Have they tried drilling that deeply in Ohio or Pennsylvania? That is where the oil industry started, and the region was once a shallow ocean. Just like the Texas oil fields.
I don't know about the depth of drilling. My understanding is that if oil is found at a certain level, then it usually isn't found in a much deeper level. In any event, the USA is only extracting 5 million barrels per day, compared to about 10 million barrels per day in 1970. Even if there were additional oil to find in the states, it would not prevent peak oil. I suspect that if the oil companies thought there was oil to find in Texas, Ohio, or Pennsylvania, then there would be crews drilling for it.
I also hope that the rate of peak and decline doesn't screw things up to quickly. Colin Campbell, the top oil geologist around peak oil, seems to think that when we do peak, it may be a gradual change, and it may look more like a plateau. He believes that the oil prices will go up and down, and each time they are up for a prolonged period of time, they will lead to anything from a "growth recession" to an all out global recession. I believe his film, "Peak Oil: Imposed by Nature" goes into this.
This question about Texas and the midwest would be best answered by a domestic oil geologist, although I don't personally know any.
2007-05-22 06:59 pm (UTC)
Below 20,000 feet, oil cooks to natural gas
Go read Kenneth Deffeyes book "Hubbert's Peak" for a primer on oil formation. Generally, sedimentary rocks containing keragin that are lowered to between 20,000 and 25,000 feet "cook" into oil from the extreme heat and pressure. Below 25,000 feet, oil breaksdown into natural gas. So when you hear folks say "We'll drill deeper.", it assumes that there IS oil much father down. And that violates the laws of physics.
Also, as you alluded to, if the oil companies thought there was oil in Texas, Ohio or Pennsylvania, there would be crews drilling for it. This country has an EXCELLENT USGS which has surveyed the country for energy resources. We have been intensely studying US geology for the past 175 years for coal and oil. We know the landscape, we know the geology underneath the landscape. The big ones are all gone. What's left are little ones here and there. We'll get them but they will be too expensive to burn. What will we use them for? One word: Plastics.
|From: valuesystem |
2007-05-22 08:29 pm (UTC)
Re: Below 20,000 feet, oil cooks to natural gas
Thanks for the reminder. I believe I have Deffeyes book on the shelf. Sounds like time for me to freshen up on it. I personally have a difficult time staying focused on peak oil, and holding down a full time job, family, etc. which requires a totally different kind of thinking.
I agree with your post, and am intrigued by your plastics comment. It may very well be that hydrocarbon molecules from petroleum will eventually be used only as feed stocks.http://en.wikipedia.org/wiki/Petroleum#Uses
I am curious if there are substitute sources of those hydrocarbon molecules that might be easier to get at -- perhaps from biological products?
Any reason why a binomial fit is appropriate to this case?
A binomial fit will predict an early decline, and will trail off the sides on the early part of the graph, and doesn't reflect any attempt at modeling oil production (logistic curves, error curves, et cetera).
A more fitting tool is a low pass filter that does away with trends on a shorter scale than a year.
I chose a binomial fit because of the various fits available in Excel, it was the one that most closely followed the trend.
A bionomial can be used as an approximation of a normal curve, as long as it is zoomed in to the maximum point. I am not suggesting that the curve be extended much beyond the graph, since the binomial curve (parabola) will diverge from the normal curve.
Of course, we don't know if we are at peak yet, or if the curve will be normal. I suspect that in the long run, it will not be normal, but lopsided, with a fairly steep drop off. This is due the the law of diminishing returns, and the difficulty in pumping from a ever increasing number of ever smaller, declining fields.
Another good question someone asked me last week was how fast the production would go to zero. If the curve is geometric decline, then it could take as long as the run up, so 150 years or so. Again, I suspect that will not be the case, because of the very different nature of the decline vs. the run-up.
All oil fields must eventually reach a point where the total energy used in making the finished energy product (gasoline, diesel, fuel oil, etc.) is more than the energy in the raw resource itself. That suggests that oil fields will all eventually be uneconomical, from both the monetary and energy standpoint. Looking at the continental USA is probably a good approximation of what could happen worldwide. I wonder if the peak of global oil production will be as dramatic and as the U.S. peak, where the rate of decline of the percentage growth was so remarkable. I'd like to see the month by month figures from about 1965-1975 for that.
As with most things, only time will tell. One of the few things that we know for sure is that there will be a peak in the amount of energy we can recover from petroleum resources. Based on the various evidence at hand, I would not at all be surprised if we are at that point right now.
As far as the best fit curve, I'm intrigued at whether Excel can do a good normal curve. I'll try to do a few fits on that the next time around. I suspect that Colin Campbell's graphs are some of the most accurate, so that would a good place for anyone to look for some more ideas.http://www.tushar-mehta.com/excel/charts/normal_distribution/
2007-11-05 01:13 pm (UTC)
just what i needed
i'm eric. joining a couple boards and looking
forward to participating. hehe unless i get
There are good levels of distraction, and bad levels of distraction. Such is life.
2007-11-06 04:07 am (UTC)
just what i needed
i'm eric. joining a couple boards and looking
forward to participating. hehe unless i get
Yes, distraction is a problem. It is part of the "cultural trance". Try to keep it real.
2007-11-18 07:17 am (UTC)
Hi! Хороший проект =)))
Хотелось бы принять посильное участие в его жизни!
"New post! Hi! A good project =))) We would like to take all possible involvement in his life!" - From Google Translation, Russian to English
Well, thank you. I'm glad you found it helpful!
Ну, спасибо. Я рад, вы нашли ее полезной! (Google Translation, English to Russian)