Peak Oil

When things go up exponentially, then smooth out -- like so: one (often) assumes a Logistic Curve will fit the data. That's the case with the H1N1 Swine Flu cases. On the other hand, if things go up then down, one (often) assumes the notorious "Bell Curve" will fit the data -- like so: That seems to be the case with Peak Oil data. That is, one assumes the worldwide production of oil increases, then decreases as supplies run out. That suggests that we might try fitting such a "Bell Curve" to the oil production numbers over the past umpteen years. This is what I get: How does it look?