A little more on CERA

Having listened to the CERA lecture that ProfG led us to earlier, I thought it might be interesting to check on some of their projects. For those who might be interested in doing their own "field by field" study on the supplies of oil that can be anticipated over the next five years I have copied the list from the CERA presentation discussed earlier and will give some sites where these can be looked up, for your own information. Interestingly I compared the list with that of the Megaprojects from ODAC and found more projects listed there than by CERA.  I will put their comments under CS (for Chris Skrebowski who wrote their report).
CERA has the following projects listed for 2005.
  1. Bonga is a Deepwater Project, Niger Delta this "long-delayed" offshore field is operated by Royal Dutch Shell and is anticipated to be operating in the 4th Quarter of this year and reach 225,000 bd by the middle of next year. see Chris Skrebowski has it listed  as coming on line in 2004. See also here and here

  2. Kizomba B is a Deepwater project, Angola, and is scheduled to install facilities in June 2005, and begin oil production in 2006. It is ahead of schedule. CS has it online in 2006. See here and here

  3. Albacora Leste is offshore Brazil and has been delayed (CS had it coming on stream in 2004) and is now anticipated for 3rd Quarter of this year at 100,000 bd rising to full production of 180,000 bd. Information here, from Rigzone.  There is also a note that it may produce a little less than anticipated.
    Upon completion the vessel will head to the field where it will hook to 15 producers and 15 water injectors. Peak production is expected to be 164,000b/d reached in 2009.
4. White Rose, is offshore Canada and will start in the fourth quarter. (CS has it starting then)
See here and here.  While the OGJ says:
Husky Energy Inc., Calgary, plans to start production in the fourth quarter from White Rose oil field in the Grand Banks region off Newfoundland and Labrador.

 Husky, operator of the $2.35 billion White Rose project, anticipates peak production of 100,000 b/d of oil from the field, which it estimates has probable reserves of 200-250 million bbl (OGJ Online, Aug. 12, 2004).)

  1. Thunder Horse is the Deepwater, GOMEX, set to produce 250,000 bd, which has been delayed 6 mo.
    Non-OPEC supply growth in 2006 is forecast to slow from a six-month delay in start up of the Thunder Horse oilfield in the U.S. Gulf and the loss of some production in India due to the Bombay High oilfield accident, the report said.
    See also, here and here. Note that CS had it on stream in 2005.

  2. Dharkovin, is in Iran and there is not a lot of information quickly available on it. CS does not appear to have it and possible production ranges from 10,000 bd to 160,000 bd but its current fate is not completely clear.

7.Adar Yale, (or Adar Yael) in the Sudan being developed by the Chinese.  It was initially operating at 5,000 bd, but the EIA now projects greater production

The blocks contain the Adar Yeil and Tale fields, anticipated to come online in 2005 with a combined capacity of 200,000 bbl/d. Capacity is expected to increase to 300,000 bbl/d by late 2006.
See this also.

UPDATE: Since it is a bit hard to find much information on this one (As Halfin points out CS doesn't have it) I am going to steal a site that Econbrowser gives, namely the left coaster almost at the bottom of the page.

Greater Nile Petroleum Operating Company Limited (GNPOC), a joint venture of four companies, will produce 200,000 barrels per day (bpd) at the Adar/Yale onshore fields in Central Sudan, which will raise output at the North African country to 500,000 bpd. "Next year, there will be an additional output of 200,000 bpd. Now, GNPOC's output is 300,000 bpd," one source said. Sudan's oil minister had said last September he expected oil exports from Africa's largest country to reach 600,000 bpd by 2005.
8. Primorsk, This was a bit difficult to work out, since that is the name of a Baltic Port facility which is increasing capacity to 1.2 mbd but this does not give them the oil to do that. CS has two other sites possible Priobskoye which was  Yukos (CS has it for 550,000 bd in 2004).  It could also be Prirazlomnoye Also described here  (CS has it for 150,000 in 2005)

9. Sakhalin 1 is an island off the East coast of Russia, just North of Japan.  CS has this for 250,000 bd this year.
From OGJ

With 10 of 30 wells drilled at Sakhalin I, gas production is to start in fall 2005 and oil production in 2006 (OGJ Online, July 13, 2005). The plan is for 250,000 b/d of liquids output by 2010.

10. ACG Megastructure, which is in Azerbaijan started Feb 2005, and is anticipated to average 410,000 bd this year (CS has it down for 1 mbd starting this year, as does ExxonMobil).

Well that was the 2005 list - some slippage in a number of cases.  CERA also gave us a list for 2006 and 2007, and with time I may discuss those also.  

Technorati Tags: peak oil, oil

This is the best post I've seen in a long time. Not now (I'm off for now) but later I'm going to look at each and every one and add some new ones e.g Sakhalin II coming online next year 2006, where we can expect about 1 billion barrels (total reserves) of oil from this Russian "mega-project".

In my view, this is where the rubber hits the road as far as peak oil is concerned.

Thanks, HO.
Econbrowser noted that the Sudan project, number 7 on your list, was another one missing from CS's report.
Saw this on AltEng.  Forbes thinks oil prices are a speculation bubble:


Well, Forbes is saying that to the public. But, he is saying the following to his suscribers :

  "Oil and natural gas are on their way to significantly higher levels. You can still buy select oil and gas producers that pay 12% to 14% dividends - and they pay monthly. It doesn't get better than that. There are many reasons to buy oil and gas. Unrest continues to mount in the Middle East, and the so-called terror premium in crude prices will remain until we see at least three years of peace in the fertile crescent. I think that will be a long time coming.

There is a shortfall between supply and demand, and this shortfall is growing. World demand increased 2.5 million barrels a day over the last year due to increased demand in the U.S. and Asia. India and China are industrializing at a feverish pace, and their energy appetite is increasing exponentially. On the other hand, global production is very close to a peak, and there is no "excess" production capacity left. We are at the point where the rubber hits the road, and the only rationing mechanism for who gets the available supply will be higher prices. "

Please check:

http://www.newsletters.forbes.com/servlet/ControllerServlet?Action=DisplayPage&Locale=en_US& id=ProductDetailsPage&SiteID=es_764&productID=10292900

Fernando, take care on reading too much into the demand figures.  We'll know better mid-month how Q3 went, but in Q2, a full 2.5-mbd went into stock building, mainly in the OECD.  In short, after the speculators bailed, there was nobody to buy the crude.  Over the last 18 months, the oil markets have become the playground for boomers trying to make up what they lost in the equity markets.  Notice that the Spot price is now commonly 20% greater than contract prices.  Notice that crude spot often has 20% price swings.  This is not supply and demand at work in a normal marketplace.  I fear that there are more commodity brokers in the bidding field than stakeholders.
I'm sorry to go ad hominem, but I just think Steve Forbes is an idiot.  Other than daddy's bucks, he's got no credibility.  Doesn't matter what he says.  Though since many people listen to him, he's a talking head spouting nonsense, he must be countered.
I interpreted the CERA list as intended to be illustrative whereas the CS list was intended to be exhaustive.  For this reason, I took the fact that the CERA list included a number of large projects not appearing on the CS list as an indication that CERA had conducted a more exhaustive inventory than had CS.  If so, that could be one explanation for why CERA believes that more production will be coming online than CS was expecting.
Grin - well actually if I add up the numbers I cited for total production and giving Primorsk 150,000 bd and Dharkovin 160,000 bd I get a total of 2.68 mbd which is more than the 2.4 mbd that CERA are quoting.  So I am not sure how to interpret this yet. (Bear in mind they are totalling toward a 2010 data so it would be reasonable to assume that the projects would reach their targets by that time).

Since the criteria for the two surveys were slightly different one would expect some discrepancy, and possibly some correction due to a better evaluation of the reserves with time.  Though I note that while some go up, there are also a couple that are going down.  And more that have been delayed.  It's a puzzle, but bit by bit we might assemble it. It is also something that I plan on keeping an eye on, just to see how some of these numbers pan out.

It seems to me we could really use an "open-source" bottom-up model for future supply/depletion, or perhaps a set of models under different scenarios "pessimistic", "neutral", "optimistic" or whatever. Then we could track/update/improve it over time, and others could take the base data and build their own models to test some hypothesis. I'm thinking "open-source" in the sense of open-source software in which a group collaborates to share the work of maintaining something, and the data is licensed freely to the public. Usually open source software ends up at least as good as, and maybe better than, the expensive proprietary equivalents.

Such a model might be maintained in Excel? I think a tricky thing is a good data architecture. Eg, one wants to be able to do things like discover that some field has been delayed by six months and only have to change one number to recompute all the scenarios.

I second your idea, Stuart. We should try to do scenario analysis in Excel to understand what sort of spectrum of realistic options is to be considered.
I'm rather astonished that anyone can take comfort from a growth of 2,4 million barrels per day in production over five years!  I tried to be optimistic and lumped all I could into this years and got 1.75 million for the end of this year.  That is barely 2%.  This is the kind of growth that all the pessimistic models assume.  How can you look at this and deal with depletion with anything but the most wishful thinking without seeing a train wreck?  Even at a 4% depletion rate things look grim.  Looks to me like Deffeyes was spot on.  How does new demand get taken care of.  One must assume a significant amount of demand destruction rather than new demand.  Surely this is not a comprehensive list.  Surely this just scratches the surface of the massive amount of new projects that are going to come on line in the next several years to meet the need.
What I am doing is going through the projects, year by year, that CERA said would contribute to the oil flow by the year 2010.  The ones that I have listed above are those that they had anticipated would start in 2005, and as you may note, some have and some will be delayed into 2006.

Later this week I will do the same thing for 2006 and next week probably 2007.  I don't have their projects for 2008 and 2009.  I will then do the same for the projects from ODAC that were not covered by CERA.  Then I might wander around and see if there are any more that should be considered.  

Once we have the total list in a table, it will be fairly easy to post the occasional update telling folk where we are with it (and possibly adding additional projects since there is exploratory work that should yield some).  The only thing that folk need to remember is that all these are big projects that take significant lead-time and some are sensitive to other issues that may cause delays. And since I will be anticipating production in 2010 the numbers that I will use in the tables will be the operating anticipated maximum production, rather than the step-up values as they bring the field on stream.  If folks have other suggestions I would be glad to hear them.

You might want to take a look at this to safe yourself some work.


I emailed you guys about it before but got no response.

As for the 2005 fields:

You are missing these bigger fields (above 50.000 b/d), also some smaller ones ofcourse.

Algeria, Hassi Messaoud, EOR
Iran, South Pars, Gas Condensates
Iran, Darquain
UAE, Bu Hasa, EOR
UAE, al-Dabb-iya, Rumaitha, and Shanaget
Mexico, Ku-Maloob-Zaab, EOR
Mexico, Crudo Ligero Marino
Mexico, Lankahuasa basin
Brazil, Jubarte
Trinidad & Tobago, Greater Angostura
Norway, Kristin Deepwater (oil, gas condensate)
Angola, Bomboco & north Sanha (Oil, condensate, LPG)
Congo Brazaville, Moho North, Moho South, Bilondo
Australia, Mutineer-Exeter
China, Peng Lai

The other fields mentioned:

Bonga no comments, bonga southwest is coming onstream around 2009.

Kizomba B, already on-stream and producing

Albacora Leste has 2 phases, 1 for 145.000 b/d and the second for an additional 100.000 b/d.

White rose, peak is scheduled at 92.000 b/d

Thunder horse, no comments

Dharkhovein, 2 phases, 1 already onstream this year (around 55.000 b/d) and phase 2 for next year (110.000 b/d)

Adar Yale (and tale), i agree with the EIA (300.000 end 2006)

Primorsk, no comments

ACG Megastructure. 500.000 of production ramping up until 2007. Not 410.000 b/d this year i think. also 2 additional phases later.

  I apologize for not replying to the e-mail, I was travelling at the time, and did not see it until a time I thought too late to reply. (And I suspect that the rest of the team just assumed that I had written separately).  It sounds as though you have been working the same issue as I.  With a little time (I have to give a seminar on all this tomorrow morning) I will go through your report in more detail and get back to you.

My feeling is that it is only by tracking these numbers that one can begin to see what the future holds.  But one also needs to find a variety of sources from which to get the data.  We have seen the OPEC numbers from different sources vary quite considerably within the last few months.


Looking through the list of projects in this paper I find that it is quite comprehensive.  One thing though, the daily rates quoted for most of these projects are their 100% uptime, nameplate-capacity maximum rates.  Most of the offshore projects are facility constrained (oil, water, gas, and/or total fluid, etc)and cannot really maintain these rates on a full time basis.  It is probably best to assume 90% to 95% of the quoted rates as the likely volume they add to the world market.  
No wishful thinking required.  Why do you assume 4% for ALL existing fields this year? That's simply not supported.  While some fields may decline this year at 4%, 6%, or even 10%, that does not mean that ALL fields are in decline THIS year.

CERA's estimate of 2.5 mbd new production for 2005 means at least a 3% increase in production, and probably slightly more.  So it seems that oil production may increase slightly this year.

These projects are mostly long lead-time investments, and were planned 3-5 years ago for oil in the $20-30 range.  If oil stays above $35-$40, we should expect LOTS of additional new production to be announced in the next 12 months for 2008-2010.  Of course, CERA says that $40 is the likely equilibrium price for the near future, but that recessionary economic shocks might push it lower.  It remains to be seen if the oil companies will use a $25 or $40 price when evaluating new production proposals.

A lot of production now is offshore, and offshore fields decline very fast. See the data for various North Sea fields here, with decline rates well over 10%. See graphs here. The whole North Sea domain is now declining over 10% and closer to 15%.

So 4% overall for existing production is probably pretty kind. Eg the EIA reports that FIP in Saudia Arabia is declining at 5-12% annually.

I suspect that what is assumed for decline rates on the FIP is going to be the leading source of uncertainty in a prediction. The problem of gathering and reconciling a list of future projects is a soluble one - it's just work in doing it.

Here are some initial thoughts on requirements for a good data architecture:

  • Data should be stored in a format suitable for general use by any researcher without specialized or custom software (probably Excel).
  • Data format should allow for easy updating of country by country production statistics as new production data becomes available on a monthly basis.
  • Data format should support projections based on varying decline rates of the FIP for different countries.
  • Data format should allow more complex decline projections than a straight exponential where data quality supports a more detailed model.
  • Data format should support global tweaking of decline rates from "optimistic" to "pessimistic" in a way that allows one multiplier to adjust all the individually estimated country decline rates.
  • Data format should support modeling new fields at varying levels of detail (either a flat mpd for the life of the project, or a sculpted profile where an estimate of that is available).
  • Data format should support incorporating individual projected field volumes into global estimates in a way that allows easy time translations and upswing stretching due to project construction delays.
  • Data format should support easy transition of a field from being part of FUD to being part of the monthly country production statistics.
  • Data format should be self documenting and allow easy graphical inspection of exactly what is being assumed for any country or any field.
  • Data format should support multiple researchers maintaining it in parallel with only modest editorial effort in consolidation.
  • Consideration should be given to versioning, as data will be a living repository that changes over time.
Any other requirements?

I am willing to put some time into this.

The latest data for UK oil production from the UK Department of Trade
and industry at:-
give the 2005 monthly declines over the corresponding month in 2004 as
Jan -12.53%
Feb -11.54%
Mar -20.95%
Apr -15.86%
May -17.65%

For the five months that is -15.75%
an average loss of 307900 barrels/day

What did CERA expect UK production to do?

Sounds good, im willing to share a part of my data (the project list). Which happens to be in excel. I have 165 of projects in the list and another 50 potential ones.
A couple of questions:

1) When Stuart says "a lot of production now is offshore, and offshore fields decline very fast... [North Sea example] -- I always see this cited as a typical example. What tells us generally that offshore fields always decline very fast? Is it the difficulty of applying EOR techniques that can be used in onshore fields? How will this affect future light sweet crude output from (for example) the Gulf of Guinea?

2) For Rembrandt, I am interested in your Assumptions Section C, wherein "There are no political, economical or natural events that take oil production offline. This happens quite regularly...." -- Do have an equivalent to the CERA "Delay & Disruption" scenario? Also, your numbers on Iraq (where I have been tracking production) seem a bit high.
So the great J, in a comment I can't find at the old site, discussed this at length - because the platforms are really expensive to maintain, and because they have good technology for seeing and sucking out the oil, the field developments are designed to operate that way. Once the field gets pretty old in the tooth the big operators sell it off to smaller companies who specialize in vacuuming out the last bits. After they are done, it's over. It's not like on land where it's economical to put a nodding donkey and let it sit and slowly decline for decades more. Alaska and GOM field profiles that I've seen look roughly similar, as do the estimates for Cantarell.

Here's an interesting discussion of Prudhoe Bay.