If refinery capacity is struggling to return to normal as we enter the winter, and imports are rapidly declining, we will have problems with our gasoline inventory.  However, I can envision a scenario where basically all of the refinery capacity affected by Katrina and Rita is running at full capacity by Thanksgiving.  We could see crude stockpiles plummet as it becomes clear that production data has been manipulated.  Before the hurricanes, the world was likely producing closer to 82.5 mb/d than 84 mb/d, and we have now had lengthy periods where at least several hundred kb/d in the Gulf of Mexico was affected by the hurricanes, not to mention that over 1 mb/d  was out for most of this time.  

In my opinion, by the end of the winter we could have a situation where crude stockpiles are far from zero, but low enough that refineries are having trouble getting all the crude they need to run at close to full capacity.  Of course, we have the SPR, but it would make more sense to tap it for short-term disruptions rather than drain it all winter long.  It may be that upcoming times are far worse than any individual disruption could ever be, but in that case it doesn't make much sense to drain the SPR immediately.

If we have consistent crude drawdowns of more than 5 million b/d, the price of crude would likely soar.  Therefore, inventories might not get dangerously low, but only because prices went up dramatically.  That is the job of the free market.  On the other hand, if the price of refined products were to soar in lockstep, the refineries would have no incentive to slow their runs until they absolutely have to.  

It should say "5 million barrels per week", not "5 million b/d".  Sorry for any other mistakes I might have made in my post.  
I think we may get within  500k of pre-Katrina refinery production by Thanksgiving, but I don't think it will be all restored before the end of the year at the earliest.  This is an improvement on the end of 1st Q 2006 I was hearing 2 weeks ago, but is still not good news.
I am curious as to why you believe that? The MMS report has not shown significant improvements in a few weeks now. It appears that we're slowing down, not speeding up. What other data leads you to your conclusion, because it's not obvious to me why we should go from 67% shut in oil on Friday, October 28th to 33% shut in oil by December 31st. Is there word somewhere on work crews or some other factor changing?

Thanks in advance.

I think you may have misunderstood, I was talking about refinery capabilities not the production of crude from the Gulf.  It is likely to be refinery production that poses the limits on what is going to be available, at least in the short term.  I do not believe that we have tapped all the crude that could be released from the SPR and this could help, if we could get it refined.
Ah, thank you. I did misunderstand having mentally focused on the shut in for so long.