An additional chart that is showing only the additional supply from the Wikipedia list + YTF from Tony compared to the new supply forecast by the IEA (Figure 11.6, no YTF).

On decline rates

My current reading of the tea leaves is that an initial decline rate of 4.3 to 4.5% should be applied to the discovered developed stack and this encompasses fields in build up and post-peak fields. As your excellent Figure 7 illustrates the fields in build-up will after a number of years (5±3?) pass into the post-peak domain where according to the IEA a decline of 6.7% should apply. As Rockman pointed out earlier, in decline phase 3 (essentially stripper wells) decline rates may decline to around 3%. In summary:

4.3% for 5 years
6.7% for 6 to 25 (?) years
3% thereafter

Applying 5% to the whole is probably a good approximation.

On DUD and YTF

I find it a bit confusing the way you have combined Discovered Undeveloped and Yet to Find - makes it hard to compare with IEA. Any reason for doing this?

DUD and YTF forecast capacity additions were combined because it was simpler to represent and forecast the unsanctioned/YTF supply additions as in Figs 3 and 6.

In a similar analysis that I conducted, I assumed that new open deep water and redrilled existing oil fields would decrease in oil output by 18% and 10%, respectively, 5 years after they are put into service.

Thus, if a deepwater oil field was put into service in 2005, I assumed that in 2010 that the oil field would begin an 18% decline.

Did you assume something similar which is adequately and justifyably pessimistic (realistic)?

Retsel

Some of the decline rates that I use for deepwater oil fields are very high.

Doing bottom up forecasting, country by country, sometimes requires the use of other factors. For example, Angola has many deepwater fields but Angola is subject to an OPEC quota. In addition, many of future projects in Angola may be delayed because they are uneconomic at current low oil prices.

I also place some importance on statements from state owned oil companies. In Mar 2008, Syanga Abilio, a VP of Sonangol, said that "We are doing our best to maintain our plateau of 2 million barrels, probably until 2014. Our production profile does indicate normal decline (after 2014)".
http://www.reuters.com/article/rbssEnergyNews/idUSL1883291720080318

In addition, an estimate of a country's ultimate recoverable reserves (URR) and appropriate extraction rates are used to construct a forecast. In the case of Angola, a URR of 20 Gb is used which is from Colin Campbell's Dec 2003 newsletter.
http://www.energiekrise.de/e/aspo_news/aspo/Newsletter036.pdf

Taking everything above into account, the forecast for Angola, mostly deep water now, should have a production profile similar to the chart below.

Another deep water region of significance is USA Gulf of Mexico (GoM). While it is not subject to quota, bad weather plays a role in production. In addition the majority of the remaining proved reserves of USA GoM are located in offshore Louisiana and have been in decline since 2003 at 4.25 Gb to only 3.32 Gb in 2007.
http://tonto.eia.doe.gov/dnav/pet/pet_crd_pres_dcu_R1901F_a.htm

The chart below shows of US GoM including deep water and shallow water, representing total remaining proved reserves of 3.5 Gb. Taking into account probable reserves, a remaining URR of 6.7 Gb is used for forecasting purposes.

click to enlarge