This is why I have a problem with the CERA forecast.  I suspect it was put together without Murphy's law in mind - no hurricanes, no strikes in Nigeria, no wars, etc.  Plus there is (in my mind at least) a 95% chance that any of these mega projects will be delayed - usually by years.  The Bonga Project in Nigeria was intitiated, at least from an exploration standpoint, in 1993.  Oil may start flowing from there in late 2005.  Then again, it might be 2006.
They do have a "delay and disruption" scenario intended to account for that. It estimates 11mbpd extra supply by 2010, instead of the 16mbpd under the normal scenario.
David Yergin actually said that problems with oil production are not below ground but above ground. I guess it's a way to cover his ass.
From the past productions, does somebody know the probability of a new project to be accurate (on time with the right numbers)? the problem is that you don't really know your field output before a few months into production.
Almost all mega-projects, and essentially all offshore projects are designed for a specific capacity and flat production for specific period of time.  Usually they are designed to have a comfortable cushion to be able to achieve the plateau rate.  However, there are usually big uncertainties as to how long the project can stay on plateau or whether or not other wells will be needed to maintain the plateau rate.  Also, there are always major uncertainties about schedule and cost for these projects.  As a rule 90% will be over budget and behind schedule.  The higher the oil price, the more likely that this will be true, as the projects are planned under certain assumptions about costs and availabilty of resources.  As the world oil price rises, costs increase dramatically and resources dwindle( rigs, engineers, trained construction laborers, steel, wellheads, undersea pipeline layers, etc.).
Bubba,

Thank you for this insight into your industry.  I questioned (in a very poor way in another session!) how costs change with respect to production rate.  It sounds like upgrading, improving, expanding withdrawel rate WHEN prices are rising is often a money losing proposition.  

This is due to higher cost for everything related to extraction and IS NOT offset by the additional oil brought to the surface.  Huge costs for an incremental increase in production.

Did I understand this correctly?

Also, and I think you and J. have touched on this before, does this mean raw material costs are rising faster than energy (oil & NG) prices?

Don't get me wrong here.  When the price of oil rises producing companies make out like bandits.  However, for mega-projects that take 6-10 years to bring on, the conditions that these projects are planned under rarely if ever exist when the projects actually come to fruition.  When commodity prices rise, costs rise, and inefficiencies show up everywhere. Consequently, projects universally come in late and over budget.  But in the end the projects make much more money than they would have if they had been executed efficiently in a low-commodity-price environment.
Look at Oil & Liquids Capacity to Outstrip Demand Until At Least 2010: New CERA Report for information about the CERA "delay & disruption" scenario that Stuart pointed out above.

There's not much there, unfortunately, nor was there much in the multimedia presentation we got access to lately from the CERA analysts. However, making an elementary point here, overall depletion in existing conventional crude oil sources is inexorable and never delayed while new projects are subject to the exigencies of investment, market price, technology, geology and politics. This must have been part of why Jean Laherrere thought that all new projects should come with an associated "probability of successful [projected] production [by the target date]" calculation.