In thinking about how to synthesize all the key drivers of oil production/demand, it seems like the period between now and 2017 is of particular interest - the contracted-for plateau for the new contracts in Iraq is seven years out, and they are being signed over the next few months. Furthermore, if the meteoric rise in the Chinese car fleet were to continue, then it would reach the size of the US fleet somewhere around 2017.
So it seems of interest to try to construct uncertainty envelopes of some of the key variables and then try to fit them together into a range of reasonable overall scenarios.
Probably the largest uncertainty in the evolution of the global economy over that timeframe is the uncertainty about Iraqi oil production. In the graph above (click for a larger version), I give my subjective guesstimate of the 90% confidence interval, with a "Low Projection" for if things go very badly from the perspective of Iraqi oil production, and a "High Projection" if things go very well.
The reasoning behind the "High Projection" is as follows: the country in the past produced a maximum of 3.5mbd and likely doesn't have much more capacity to produce and distribute/export the oil than that. The giant megaprojects in Iraq required for the al-Shahristani plan would take three years if they were in Saudi Arabia, so let's allow four as the best case in Iraq, given it's a much more difficult environment to operate in than Saudi Arabia, which has far more infrastructure and has been stable for a long time. So then the idea is that things continued to get fixed over the next couple of years up until the 3.5mbd level has been reached. Production plateaus out there for a while, until all the various elements of the al-Sharistani plan start to come together, and all projects hit their plateau together at the start of 2017. For lack of basis for making a more complex plan, I just linearly interpolated between these various constraints.
Hopefully, the reader will agree that it's hard to imagine things going much better than that.
For the low projection, the assumption set is that the elections in March result in renewed civil war, that the US leaves despite the renewed unrest, and the country descends into increasing fighting, with only a small amount of oil exported intermittently. Obviously, the smooth estimates in my "Low Projection" above are just a general indication of what this might look like. In reality, in this scenario, oil production would be very noisy month to month, but it's not likely there would be none, as the combatants would need to sell some to fund the war. I don't think this scenario is all that likely - in particular I doubt it would be politically feasible for the Obama administration to leave if the country were descending into chaos, and there's a limit to how chaotic things can get as long as there's a heavily armed US military force in the country. So that scenario forms my lower bound on Iraqi oil production.
In between these two scenarios, lie all the possibilities like it taking much longer than seven years to get the Iraqi fields to plateau, given the difficulties of managing such a large set of projects all happening in the country at the same time, or that some of the IOC's have overbid on plateau, given the incentives to do so, such that the eventual total is less than 12mbd until further fields have been brought on line, or that a new government decides it's imprudent to produce the oil at relatively high depletion rates by Middle East standards and decides to go slower. Somewhere in that range is the most likely scenario, which is hard to project with much precision at this time.