I was using the Hubbert graph to show that my 45 billion barrel figure is in the right ball park eg we are talking about reserves well under 100 billion barrels left not the 260 the Saudis claim. Of course the issues we are discussing are not monocasual you should not draw conclusions solely based on just one graph. Your assertion that production is demand driven is perhaps based on the notion that OPEC enjoyed ample spare capacity in the 1980s and 1990s as shown by this graph( is this correct?) :

http://www.wtrg.com/EnergyCrisis/graphs/FIGURE5.gif

However capacity figures are estimates. For example as Simmons has commented upon the 1999 oil price slump to $10 occured due to the notion that there was huge spare capacity in the system; the above graph gives a figure of just over 5 million barrels/day. However Simmons reckons the true number was 2 million, the difference the so called "missing barrels" is a result of the inaccuracy of the production data that is available from the IEA etc. Simmons states in his book that all the "official" production data from OPEC, the IEA etc come from a Swiss company called Petrologistics ( not to be confused with Petroconsultants which Campbell was part of) which is based above a grocery store in Geneva. Here it is in Simmons' own words:

Source

The US was one of the first countries to peak. Alaska came on stream in 1978 and led to production increasing up until 1985 where the peak was not far below the 1970 peak (10.6mbd v 11.3mbd according to BP). Thus the first 15 years post peak can be more accurately be described as a "bumpy plateau" hence the initial decline rate was so shallow. Since 1985 the decline rate has averaged 2% per year which is still quite shallow. On the upslope of the Hubbert Curve for the US oilfield technology was relatively primitive compared to today; the fields were not overproduced via experimentation with secondary and tertiary recovery in anything like the scale they are today. Thus production cannot be expected to decline at rates like the North Sea today (low double digits). Interestingly it you look at the basic Hubbert Curve you will find that the gradient increases initially which is just what has happened with the US. Expertimentation with EOR is likely to lead to an increase in decline rates over time hence the position we have today as shown by this graph:



Of course new production comes onstream to offset some of the decline and some EOR can temporaily alleviate the steepness of observed decline rates. US oil reserves peaked in 1956 according to Hubbert and Laherrere and have been declining even since with the exception of 1968 when Prudhoe Bay was discovered. I repeat "reserve growth" is a statistical illusion.
You asked to discuss Cantarell. Cantarell's production has been artifically increased from approximately 0.9mbd in 1994 to the current 2.1 mbd plateauing since 2001 using the EOR technique of compressed nitrogen injection. One amazing fact is that the plant to produce the nitrogen for Cantarell produces half the world's production of nitrogen giving you an indication of how desperate PEMEX are to postpone the steep decline which they themselves admit will be 14% per year.
As for other examples here is a sample of UK North Sea that have exhibited the "superstraw effect", just go to UK North Sea Oil Field Production data look at the decline rates for May 2005 compared with May 2004 (the latest data):



There are plenty more with similar decline rates and of course some are increasing in production ( the Buzzard field being brought onstream next year will alleviate the rate of decline for a short period) , the examples above I chose because they have only been brought onstream relatively recently (last ten years) to demostrate how new technology which is being used in the North Sea is creating steeper production increases with steeper production decline post peak doing nothing to increase URR.
As Campbell said in the The End of Suburbia:

"there is a strange irony in this business ... the more efficient you are....all the good things that one generally sees as positive attributes....all you are really doing is accelerating the depletion and making the situation worse"

As for Saudi Arabia's OOIP I would state it to be 300 billion barrels the 700 Saudi claim has no basis in reality.

Thanks for your swift reply bobbyboy.

I think we have to make a distinction between onshore and offshore fields. You are assuming that because oil production in offshore oil fields has a sharp decline (according to you due to EOR), any EOR techniques on onshore fields also will decline sharply.

Yet you have not given any examples as to this happening. Anyway if i sum it up rightly you say that the U.S. has seen no reserve growth after its peak in 1970?

I am not saying U.S. Oil reserves didn't decline, i am saying that the decline in oil reserves would have been worse if no reserves would have been added due to reserve growth.

Second question, these north sea fields you have quoted have they had second and tertiary recovery?

Third question, can you give examples of onshore fields (because we are talking about Saudi Arabia here) except the Oman field in Yibal that show tremendous steep declines due to enhanced oil recovery?

Fourth question, OIIP in Saudi Arabia was around 500 billion barrels when the oil companies left the country and Saudi Aramco was nationalised. Do you also agree with the assesment made by the four big oil companies?

Fifth question, where did your interesting graph about depletion come from?

US oil reserves peaked in 1956. After that date the general trend was that production exceeded discoveries, this occured for the world in 1979, this graph by Campbell & Exxon demonstrates what I am talking about:


http://www.readerweekly.us/issue/308/POGrowingGap-308.gif

The production line moved ahead of discoveries in 1979 and has been so ever since with the exception of 1986 (the bars on this graph are for two years so this does not show up here). For reserve growth to occur as opposed to reserve shrinkage discoveries in a year would have to exceed production in that year, is this not clear?



Cook is operated By Shell and has not.
Dunbar is operated by Total and has.
Elgin is operated by Total and has not.
Hudson is operated by Dana (UK indepedent) and has.
Shearwater is operated by Shell and has not.

The issue is not simply one of EOR but one of the effect of technology on production in general, the way oil fields are being produced is far more sophisticated than in the past. This has allowed production to be ramped up far quicker than previously which allows the costs to recovered quicker (good from a financial viewpoint). The issue is that models that fitted in the past such as the basic Hubbert Curve will not be good predictors due to earlier peak date, higher peak production, steeper gradients on both the up and down slope.



I offer you the Minas & Duri fields in Indonesia, both are onshore giant fields and have seen steep production declines after EOR techniques have been used. For Minas (Indonesia's and South East Asia's largest field) waterflooding is used. For Duri (Indonesia's 2nd largest field) steam injection has been used since 1985 in an attempt to maintain pressure (largest steamflood in the world). Both are operated by Chevron (so you can't blame it on inefficient state oil companies). Minas declined from 168kbd in 2000 to 109kbd in 2003 (15.1% annual average decline) to 100kbd in 2004 (8.3% decline) whereas Duri went from 275kbd in 2002 to 204kbd in 2003 (25.8% decline) to 110kbd in 2004 (46.1% decline! ) . Interestingly Chevron's website claims production of 309kbd and 220kbd respectively yet thats not what this says. Chevron's solution is to start injecting chemicals into the fields to try and stem the decline. Basically Indonesia's oil production is falling off a cliff presently causing serious social unrest. Indonesia (Asia's only OPEC member which peaked in 1977) became a net oil importer in 2004 primarily as a result of the steep declines seen at these two fields (black is production, red is consumption) :



Presently production for Indonesia is down to 0.9mbd (against a quota of 1.45mbd) and looks likely to continue this trend although there is one field (Cepu, 170kbd possible peak production) coming onstream when Exxon and Pertamina can agree terms (its been delayed for four years so far).



The american oil companies' estimates were based on the technical knowledge of the fields at the time, the advanced reservoir modelling that Aramco uses today was not available to them, certainly OOIP of 500 billion barrels is nearer the mark than Saudi's 700 claim. However I still would say that it is too high (about 200 billion barrels too high) as not all the complex details of the fields were known to them at the time (of course they knew about the production problems from the early 1970s ramp up). Essentially they assumed Saudi's oil resource was the size of the US' when this is not the case.



A Merrill Lynch oil analyst called Steven Pfeifer.

[quote="bobbyboyThe production line moved ahead of discoveries in 1979 and has been so ever since with the exception of 1986 (the bars on this graph are for two years so this does not show up here). For reserve growth to occur as opposed to reserve shrinkage discoveries in a year would have to exceed production in that year, is this not clear?
[/quote]

We have a misunderstanding of terms here i think.

What i mean is that in the period after the U.S. Peak Reserves did increase due to technology/new understanding but production was higher then combined technology/new understanding reserve increases (reserve growth) and discovery. Thus NET reserves decreesed.


As for the rest of your information, i rest my case you have stunned me with the indonesian examples. I think you might be very right with this technology thing.

New question, can you name any other specific fields you have knowledge of in which this fast decline happened very quickly (preferably onshore). (Just the names I will look them up myself). I intend to study this further in detail.

Many thanks for your splendid information.

As for Saudi Arabia, i have nothing to say anymore. Ill see if i can find some time to analyse your 300 billion barrel claim next week.

Thankyou very much for the compliment.



Yes what you are saying makes sense to me, we have simply defined our terms differently. "Reserve growth" is a reification to the degree seen in the official data. Advances in technology/new understanding can add to URR particularly CO2 EOR. However technology and new reservoir understanding works both ways- they can lead to a shrinkage of URR. Gains where they do exist are offset by other technologies such as horizontal "bottle brush" drilling that have been devoloped and as Simmons has warned may very well lead to a shrinkage in URR for a field not an increase due to the rate sensitivity issue as a result of overproduction as alleged for Ghawar. Similarly new reservoir understanding can lead to a decrease in URR for a field simply due to greater heterogeneity in the source rock than previously thought lowering the recovery factor. Thus we must look at the net effect on URR.


Agha Jari in Iran, Samotlor In Russia, Statfjord in Norway, Slaughter in Texas,US and Brent in the UK should keep you busy.

Bobbyboy wrote:


I think you have to be very careful in making broad brush projections of EOR effectiveness from either of Minas or Duri. They are different oil types but both are problematic from a production standpoint. Minas is light oil (35 API) but it is quite waxy. The pour point is very high (38 degrees C) which means at surface conditions the oil would be solid. Dead oil viscosity is measured at 14 centistokes and at reservoir conditions the viscosity is still about 5 centistokes (anything over 1 is considered to be high viscosity). What this means is that any type of EOR will likely be less effective than if the oil were of low viscosity and low pour point. The chemicals Chevron are using are likely pour point depressants....although I am not sure.
In the case of Duri the crude is not waxy (pour point is around 10 degrees C) but instead is lower API (20). As a consequence the dead oil viscosity is very high....300 centistokes and even at reservoir conditions the dead oil viscosity is around 14 centistokes. Steam flooding is required in order to decrease apparent viscosity of the oil. Again because of the viscosity issues EOR would be less effective than in a conventional oil.
I think Shakespeare has mentioned on several occassions that it is quite dangerous to try and generalize about field behavior. The example of Duri and Minas is a classic case where one can get completely wrong ideas based on simply looking at production profiles at the exclusion of understanding reservoir/fluid mechanics. Unfortunately nothing in the oil business is simple

also just for clarification:



I think what you meant to say was greater heterogeneity in the reservoir rock. Recovery factor would be independant of anything to do with the source rock.

I have to disagree. The fields are very applicable to our general understanding of future oil production, this is due to the stage we are at in this industry. Simmons sums it up best:


Simmons Interview

Of course all 5000 odd oil fields in the world are unique. Try and understand the big picture, fields can post peak production suffer steep decline rates whether due to EOR or natural decline. Saudi Arabia's five giant fields that have produced over 90% of Saudi's production thus far are likely to suffer the same fate due to being overproduced at various stages in their field histories. As Simmons puts it:


Simmons Interview



Yes you are correct; greater heterogeneity in the reservoir rock means a lower OOIP and consequently a lower URR.

I am afraid that you confuse oil grade in terms of price it will fetch with oil properties in terms of how it performs in the reservoir....the two are not entirely linked one to one. For instance Minas blend is high quality sweet crude and there is huge demand for the waxy stuff....it's just a bugger to produce. On the contrary a lot of the high sulphur and other nasty stuff crudes (vanadium as an example) have quite low viscosities and are relatively easy to produce....EOR in such oils can be much better (all other things being equal) than in sweet oils with higher viscosities.





Well as I have said a few times there is absolutely no proof that the Saudis have damaged their reservoirs. Simmons claims that 30% water cut in of itself is serious....this is not true. There are many reservoirs around the world that produce at water cuts up to and slightly exceeding 90%...as long as you have an economical way of dealing with the water and are not experiencing by-passing, Bob's your uncle. Indeed the Aramco presentation suggests that water production in Ghawar and other fields such as Shaybah has remained fairly constant over the last number of years. This suggests the horizontal smart wells have successfully held water coning to a minimum.
saudi aramco. The only reason I can think of that one might expect a steeper decline following EOR would be if capillary pressure effects basically shut off remaining oil behind water filled pore throats....which of course might be solved using surfactants. But I stand by the view that you cannot lump all reservoir behavior together and suggest that EOR creates rapid depletion...it just isn't true.

I don't know why people say the Saudi fields have been overproduced - their depletion/offtake rates are in general very low. The Ain Dar/Shedgum area of Ghawar has produced 27 billion barrels, 40% of OIIP, its highest rate was 2.5mb/d. Interestingly the 1970s estimates were that only 40% would be recovered thus if they were correct the fields would be dead by now...but it is still churning out 2mb/d at 36% water cut according to Aramco which suggests it still has a fair bit of life left. They hope to recover 51Gb or 75% OIIP...even taking a conservative estimate of 35Gb ultimate recovery gives a depletion/offtake rate of 2.6%/year at peak production. Prudhoe bay was 4.2%/year at plateau, Forties 7.4%, Brent 9.6% (!!!), Yibal 4.3%. Shaybah is at 1%/year, Haradh incriment 3 will be 1.7%, for these ones you have to trust their feild reserve estimates, but for Ain Dar/Shedgum the cumulative production doesn't lie (well not by much lol). I would be very suprised if there wasn't a correlation between offtake rates and fast post-peak decline...need to do some more research here.

Actually paradoxically you are the one confused here. EOR is most effective where as you so eloquently put it "it's just a bugger to produce". Duri illustrates this beautifully. Primary recovery was only able to extract 8% of estimated OOIP. Steamflooding since has allowed a further 18% to date raising production from 80kbd to 300kbd at peak. It remains to be seen how much more can be recovered- probably another 5% giving a recovery factor of 31% (this is half what Chevron were promising). Compare this to Prudhoe Bay where every conceivable new technological advancement in oilfield extraction has been chucked at the poor dear, the result? See for yourself:

http://www.markswatson.com/Depression12_html_b6d3864.png



Here it is in the words of the chief reservoir engineer for Chevron at the time, Bill Messick under oath in the 1974 Senate subcommitee closed door hearings:



Simmons Interview

If that doesn't convince you there have been serious reservoir management issues at Saudi Arabia's fields I don't know what will.



If you have read Simmons' work objectively you will know that the issue of water cut is one of a tipping point. When water cut in a field reaches this level well productivity falls off a cliff, the cut will quickly go from 40% to 80%- a serious issue. The SPE papers Simmons references in Twilight in the Desert describe when this has occured, additionally the horizonatal wells that for example Ghawar now exclusively uses shut off at 80% as part of the water management of the field. Thus Ghawar going from 4.5mbd to under 2mbd in a matter of months is not out of the realms of possibility.



Cantarell-nitrogen injection. Result 2.1mbd in 2005 to 1mbd in 2008 (22% annual average decline rate).

Duri-chemical steam injection. Result 275kbd in 2002 to 110kbd in 2004 (37% average annual decline rate).

Samotlor-water injection. Result 7.0mbd in 1980 to 3.0mbd in 1981 (57% decline rate)

Ghawar-water injection. Result 4.5mbd in 2005 to 2mbd in 2006 (56% decline rate)?

If such a thing happened, would there be any warning sign that one could see from outside Aramco?

Cantarell was producing around 1mb/d before nitrogen injection and it will probably fall back to that. It was expected that the field would decline steeply - you can't squeeze 2mb/d out of a heavy oil field of 15-20Gb or so for long.
http://bioage.typepad.com/photos/uncategorized/cantarell.png

Pemex hopes to recover 70% of OIIP by gravity drainage as a huge nitrogen cap forms though this will be a slow process if doable.

Duri - a field with only 8% recoverable with primary recovery is again a very different beast. Schlumberger says:
Schlumberger

Like Cantarall this feild was producing at much lower rates under primary recovery - 40kb/d vs 300kb/d with steam. I think this is an important difference, where EOR is used later in the field life for a quick hit, as opposed to water injection for pressure support.

Samotlor is the poster child for overproduction. The Soviets were squeezing as much as they could out of it, 7mb/d from a URR of 20-25Gb. Thats 10% depletion/offtake per year which was done by overpressuring the reservior with water. Ghawar has only been produced at about 2%/year. I don't think any of these fields are very good analogues for Ghawar.

No but we can infer the possibility that they are anticipating such an outcome (or it is occuring at present) from what they are doing and saying. Just look at the attempt to double their drill rig fleet (which will not occur due to the rigs simply not being available in the time scale required).

BusinessWeek

I would add that they are paying up to three times the going rate and offering three year contractual guarantees; this sort of behaviour is unprecedented since the shortages in the 1970s. The need for the rigs is a sign of falling well productivity in Ain Dar, Shedgum and North Uthmaniyah (the productive parts) to offset this more wells are being drilled. The EIA themselves quote am Aramco executive on decline rates back in 2004:


Coupled with the Saudi admission back in March 2003 that they had peaked in their giant fields:


Deffeyes Interview

Putting two and two together we can see that Saudi production is declining and has been for over two years. Yes Shaybah is offsetting some of the decline but it is not enough.



All oilfields are unique, the only fields that would truely be analogous to parts of Ghawar would be the likes of Abqaiq and Berri. The point is that steep decline rates can occur in a wide variety of different fields with different recovery techniques. The 2% depletion rate you quote for Ghawar is very misleading. Firsly it takes Aramco's claimed remaining reserves (70 billion barrels) at face value. Secondly Ghawar is very much a Jekyl and Hyde field. It should be analysed as such with the productive part(Ain dar, Shedgum and North Uthmaniyah) and the unproductive parts(Fazran, South Uthmaniyah, Hawiyah and Haradh) separated. For example at Ghawar's peak in 1990 it was producing 6.5mbd. Of this 5.3mbd was from the productive part with remaining reserves (without resorting to artificial lift) of 20 billion barrels at the time in this part of the field giving a depletion rate of 9.7% comparable to the antics that went on by the Soviets at Samotlor in 1980. Thus this is clear evidence of damaging the field's reservoir via overproduction with water injection.

Love that quote, couldn't agree more.

Additional rigs means they need more oil, but they can need more oil for two distinct reasonse: to offset depletion and/or to significantly increase production.

The Saudi's have stated that they are going to significantly increase production. So I see no inconsistencies here. (That doesn't preclude the fact that they may be trying to make up depletion too.)