I found it:
http://freeoil.1111mb.com/spe/spe93439.pdf

You will see from this article that the WSO (Water Shut Off) and Horizontal Drilling wells began in 1997. It has been under way for ten years! That is, these super straws have been sucking the oil from the very top of the reservoir for ten years because just a few feet below the water to oil ratio was too high.

Methinks the collapse will be a lot sooner than a lot of people believe.

Ron Patterson

I agree with you since we have no easy way to normalize advances in production which lead to increased depletion rate. This in my opinion is as important as Export Land and as scary.

HL is a overestimate and even the conservative HL plot in the paper is probably still high.

The critical flaw if you will in HL is its blind to the technical advancement that trades increased depletion rates to maintain and increase production rates.

Can we even get a handle on this error ?

I think we can.

1991-2002 gives a URR of 178
The dogleg gives a URR of 241
Subtracting these two gives a known technical distortion
of 63 GB if you agree that the dogleg is predominantly technical advancement.

Now lets make what seems to be a save assumption and assume that the low estimate is also high because of systematic technical advances.

How high ?

A good starting point is that and additional 63GB has been extracted at higher production rates via technical advances that are not as obvious as the dogleg this gives a URR for KSA of
178-63 = 115

So I think 115GB is closer to a reasonable low estimate for the URR of KSA. Now you can consider that they did not develop some large fields and parts of Ghawar for some time so this should be too low by a bit but not all that much.
Considering that we are saying that technical advancements over decades is equal to a 5 year or so burst of technology.
You can see that even if I'm over on the technical portion of the dogleg I'm probably well under on the guess of 63GB
because of additional technical distortion. The positive correction is later discovery/late development of fields.
But even correcting for this for KSA we are still well below the lowest current estimates.

This fits with my gut feeling that technical advances may have resulted in us overestimating URR by as much as 50% even with HL. Also you cannot really trust the reserve estimates since they are also inflated about the same amount because of technical advances in fact they probably have larger errors. So the bottom up approaches even with discovery included have a bigger systematic technical error vs HL. And indeed most of the bottom up approaches come in consistently higher than HL.

So what is this URR that HL is giving. Its a combination of the real URR and progressive technical growth leading to exponential increases in extraction rate at the expense of a higher depletion rate thats not accounted for. So its really two super imposed curves. On the technical side with MRC wells its easy to say we are maxed on technology for primary/secondary recovery so the technical boost ceased over the last 10 years. So we peaked when technology peaked.
The real mid point of URR may have passed a while back in the late 1990/2000.

This systematic error if real is huge.

Next we have a pretty good case for what the worst case scenario is for the future.

1.) Technical advances will go rapidly to zero as far as their ability to increase production rate. This steady increase seems to have inflated production rates by 50%.
Say assume a exponential decline for technical reason.

2.) We may be far more depleted then has been predicted 70% globally seems reasonable. So we have say 10% of this greatly reduced global URR of about 500GB will be produced at a high rate this gives us less than two years before we see accelerated decline rates.

3.) We have Export Land.

Effectively the worst case scenario seems to indicate that we will slowly come of the current peak for about 18 months followed by and accelerated drop down to about 15GB per year
probably a lot faster than expected say 2015-2018. Of this amount export land should drop export rapidly to zero as early as say 2011-2012. They can't go to zero that fast but this puts us easily in to a really strained political situation by 2009.

The good news is this seems to be the worst case possible and even it seems to indicate we probably still have 18 months before we see accelerated decline rates in production.

Hi Memmel,

Interesting comments!

You said that the "real mid point of URR may have passed a while back in the late 1990/2000". Assuming this was in reference for KSA and assuming 1997 as a midpoint of Saudi URR, this would imply that the URR of KSA is about 165 Gb as shown by the lower red line in the chart below. The upper red line assumes that EOR actually increases recovery factor instead of just increasing the production rate temporarily.

Hans Jud thought 160 Gb was a good estimate of KSA URR
http://www.aspo-portugal.net/Articles/SA-Oilprod_field-by-field_V2.pdf
http://i129.photobucket.com/albums/p237/1ace11/4HLAll.jpg
I'm currently assuming URR=185Gb
http://i129.photobucket.com/albums/p237/1ace11/fig10r.gif

click to enlarge, - the source of the underlying chart without the red is from one of Robert Rapier's stories

You also said "We may be far more depleted then has been predicted 70% globally seems reasonable".

The summary of this report by the German based Energy Watch Group by the Guardian
http://www.guardian.co.uk/oil/story/0,,2196435,00.html
says that "the EWG study relies more on actual oil production data which, it says, are more reliable than estimates of reserves still in the ground. The group says official industry estimates put global reserves at about 1,255 gigabarrels - equivalent to 42 years' supply at current consumption rates. But it thinks the figure is only about two thirds of that." Two thirds of 1,255 Gb is 840 Gb. If about 1,100 Gb is produced to date that means almost 60% depleted globally, which is less than your 70%.

The report should be released in full, in just over an hour, from here
http://www.energywatchgroup.org/Reports.24+M5d637b1e38d.0.html

Maybe this report might have a new source/method for estimating Saudi URR!

Thanks Ace !

Yes 1997 for KSA makes sense according to what I'm saying.
And I believe the numbers your presenting from my technical discount concept. They may be a little bit high 10-30GB but
KSA is a bit problematic since a lot of fields where developed late but on the other hand when developed more advanced technology was used since they where developed later. So the numbers your presenting are close to slightly high. My estimate is I feel a good guess at the minimum.
This leaves about 30GB that needs accounted for. I think a lot of these are still ghost barrels.

The key point is no method presented to date as adequately discounted our advances in technology allowing increased extraction rate. Using a known period when reserves did not change much but technology was applied i.e the infamous dogleg indicates that the technology effect may account for 50% of the increase in production rate. The reverse is we are depleting the oil supply 50% faster today then we where say 30 years ago.

HL is actually measuring two entangled effects. Technical advances and actual depletion. The relentless technical advances hide a significant amount of depletion as we have gotten better and better at keeping the oil production high.
HL gets the peak date right but it can be off by as much as 50% on the URR estimate to the high side.

For primary and secondary recovery MRC or Christmas Tree wells aka super straws represent a maximum extraction method we really cannot do much better. The current peak is then actually a maximum in our ability to reasonably apply technology to maintain and increase extraction rates. The world peak i.e 50% of URR is in the past I'd guess we hit 50% of global URR in 2001-2002.

Thanks for the numbers Ace and I really think people should seriously consider this argument. On of its strengths is it answers the people that claim technology will save us.

It already has.

Our worst case Saudi scenario (fastest decline + fastest rate of increase in consumption) was that net exports hit zero in 2024, in 17 years (middle case was 2031).

In regard to worst case URR's based on HL, the lowest estimate I could come up with is about 150 Gb (C+C), which totally discounts the recent dogleg up. Note that including NGL's would increase the estimate. Looking at just C+C, my personal guesstimate is between 150 and 200 Gb.

In any case, if we look at the totality of the HL data set from 1991 to 2007, it certainly looks like 2003, 2004 and 2005 are outliers, which, as I have pointed out ad infinitum is what we saw on the Texas data set.

I should also point out that Khebab and I did not discount anything in the May, 2006 Texas/Lower 48 paper. We used production data from 1991 to 2005, and the Texas model, to warn that Saudi Arabia was on the verge of a production decline.

It all comes back to Saudi as the "swing" producer-- meaning we can expect all data based upon KSA's "swings" in production to also swing back and forth as to URR.

As for me, I see induced swings in production in connection with the 2003 invasion of Iraq; and again in March 2004 and March 2006 to drop prices in the USA in time to influence our elections (is there someone who doesn't believe KSA cares about who controls Congress and the White House?).

I would suggest that a view of KSA production that identifies those periods when we can reliably believe that KSA was maximizing production (like in 1991, 2003, and early 2004 and 2006) based upon external events, could provide the most reliable set of data points. In every other case, the data could be because KSA was "swinging."

I would also suggest that the limited surges in 2003, 2004 and 2006 would likely be not a sustainable maximum, but the absolute maximum-- if KSA could have bumped production more in 1Q 2006 I believe that they would have, to give their BFF GWB more leverage in Congress. Just remember what happened to gas prices between Aug. 15 2006 and Nov. 1, 2006.

Regardless of what the actual number is, though, the USA should be acting, from a policy standpoint, as if the lowest number were the actual number, at least to generate some margin of safety.

And 150 may still be high since its not discounting more gradual changes in technology. The dogleg period is obviously the result of technical changes not some sudden large discovery and it obviously causes HL to produce a higher URR estimate.

Less obvious has been the continuous technical advancement over decades that has exactly the same effect. Inflating the URR reported by HL. This is because the URR determined by HL is actually two variables the rate of technical improvement in extraction capability and the real geologic depletion rate. Technical improvements are a two edged sword since they increase depletion rate but in general they make the final decline far steeper causing a large systematic overestimate of URR until technical advances no longer help.

The technology to extract oil has increased at a astounding rate topping out with the MRC wells in use today for primary and secondary recovery. We are still advancing to some extent in final recovery processes.

Assuming that this technical advance is effectively at its peak over the last few years means that we are no longer getting large production rate gains from technology.

The actual geologic peak was passed some time ago probably around 2000-2002 for global oil production. Increases in production rate since then have been because of aggressive technical extraction. Over the last few years we hit peak technology the peak production rate has nothing to do with the actual URR.

HL is still valid for predicting peak production but its URR estimates are probably invalid. The US happens to work primarily because technological advances in extraction within the US have been relentless. If we had peaked in technology earlier we would have seen that HL estimates for the US where to high.

These dogleg regions are one of the few ways we can get a good handle on the technical effect on extraction rate. And they point to a fairly large technical boost factor on the order of 50% higher extraction rates vs stagnant technology.
This means the real depletion rates are up to 50% faster than prediction by HL or more correctly the URR may be over estimated by up to 50%.

Now the flip side of the coin is that we can now probably maintain a high production rate to about 80-90% of URR but the decline rates are incredibly steep.

The technical peak then production a production profile close to a square wave on the post peak side.

So HL is still valid for determining peak since it is measuring both technical advances and real depletion rates.

Its simply a coincidence that the lower URR estimates also concide with time periods that prices where low or declining and thus technology was not being applied at a feverish pace. So these quite periods are a good time to calculate the real geologic URR.

This means HL used during times of stable low or falling prices is able to get the real URR since technical advances and aggressive drilling are much lower during these times.
KSA is tough because they actually shut in a lot of good production and this messes with the numbers.

If I'm right I find it funny that peak oil production is actually related to technology and has little to do with the geologic peak except its well after we have passed geologic peak. And its no surprise that the two are not offset by all that much since geologic declines are driving the technical advancements but the critical factor is they lead the technical decline in production rate. But its critical to understand whats happened the last few years is we have hit peak technology and can no longer mask the geologic decline.
Technical peak is effectively a ceiling or flat line and its boost effect goes almost directly to zero governed by the deployment rate. Most of the NCO's have rejected advanced methods in the last few years.

Look at the average production rate vs geologic depletion over time and you see we have been highly successful in keeping production rates high at the expense of faster depletion. Almost every method used to account for URR falsely includes this as a higher URR. No one has correctly subtracted the effect.

So in closing it seems we are actually 5 years or more post geologic peak and have lasted this long on a last ditch technical burst. If true then we will soon see some fairly steep global decline rates generally inline with natural depletion rates for fields that have been extracted using advanced technology these can be steep on the order of 15%.
This probably equates to the decline in overall production rate quickly approaching 5-10% globally over the next several years.

Peak Oil production rate has nothing to do with 50% URR except that it happens well after we have passed 50% URR.