Typical costs for adding oil production capacity have mostly been in the range of $5000 to $10,000/b/day over the past five years, but there are many reasons to believe that mean costs will be twice this over the next five years, except in the Persian Gulf, where this range still seems valid.
Global upstream oil annual investment over the past four years has averaged about $60B, which would seem to suggest a mean growth in production capacity of perhaps 7 Mb/d over the next few years, as there is a 3-6 year lag between investment and production. However, over 1M/b/d of this new development is required to replace that lost by depletion, as about one third of the world's oil wells are past peak and declining at the rate of 3-4%/year, while most of the rest (except for some in the ME) are operating at capacity.
Demand grew by 2.7 Mb/d last year, and the IEA most recently predicts demand this year will increase by another 1.5 Mb/d, as there are no indications a global recession will occur this year. Such demand growth seems improbable, as it would severely test the limits of global production, pipeline, tanker, and refinery capacities and push prices to the point of limiting demand growth.
The mean annual oil price increase over the past decade has been about 9% plus inflation. Over the past six years, the mean annual rate of increase has been about 18% plus inflation. This high rate of increase does not seem likely to be sustained for another year, even though supplies will be tighter, simply because the price is finally getting to the point that it is beginning to limit demand growth. Hence, I suspect demand will grow by 1 Mb/d this year and prices will rise at half last year's rate of increase, or about 17% in nominal dollars.
It seems improbable, based on recent trends, that oil production capacity would increase more than 2 Mb/d annually for the next two years while demand continues to grow as fast as the price permits, so oil prices should continue to rise at a similar rate for the following two years. A radically different scenario, as in falling oil prices, would require implausibly rapid increases in investment in production, pipeline, tanker, and refinery capacities.
Of course, the high prices will stimulate increases in development investment (recent trends suggest 20%/year), which should result in a more rapid increase in production in several years, and the prices by then will be to the point of inducing a depression in the U.S. and possibly a few other countries. The combined result will be a temporary stabilization (or perhaps a drop) in price, but for only a few more years, after which point it will become increasingly more difficult to keep up with global demand, and prices will continue to rise, probably at ~7%/year plus inflation for perhaps another 5 years. The high price will eventually turn growth negative, particularly as large-scale renewables and GTLs come on line, but it seems unlikely that the price will ever again drop below $75/bbl after 2012, as the global thirst will continue to strain the supply.