Saturday, March 18, 2006

Mexico Oil Shock Model

Khebab recently analyzed Mexico's oil production capacity. He provided the following curve courtesy of ASPO.

Since the chart overlays a discovery curve alongside the production curve, I figured I could try out the oil shock model on the discovery data. I initially didn't have much confidence in getting good results because the production seemed to start to ramp up much too quickly after the big discoveries of the late 1970's (note that even though Mexico discovered Canterell in 1976, the overall discovery peak remains closer to 1979, the year that production on Canterell started).

I used the common approximation, last used on the Norway data, of equal values for the fallow, construction, and maturation phases of 5 years each. Initially, I also chose the same depletion rate as Norway of 10% of remaining amount per year. The fit (dashed green curve below) looked adequate, matching the general rise and estimated fall of the ASPO data.

As the estimated fall lined up with ASPO's extrapolations exceedingly well, I reasoned that the rest of the production profile could fit rather well with judicious modulation of the extraction rate in the form of a series of oil shocks. I achieved a much better fit with the following oil shock perturbation.

Note that pre-1976, I selected an extraction rate of 0.04/year (an equivalent time constant of 25 years). Then, starting in 1976 and continuing until midway through 1983, I ramped up the extraction rate quite significantly, high enough that the effective extraction rate would drain about half of the reserves every 3 years. By 1990, the extraction rate flattened out to a steady state value 0.12.

Based on the dynamics of the curve, the late 1970's/early 80's perturbation impacted the reserves from pre-1976 as much or more than the Canterell-era discoveries.
The (Canterell) field reached an early peak in production of 1.1 million barrels per day in April of 1981 from 40 oil wells.
This amounted to less than half of total production. Maturation also had a continuing impact as additional platforms came online in the mid 1990's. Moreover, I have a feeling that the USA had as much to do with this ratcheting up of production as anything else. Remember this occurred as the lower-48 started its depletion downturn, instability happening in OPEC and the Middle East, and Alaskan oil had not quite hit a peak. As the world adjusted to the new oil economy during the 1980's, oil production from Mexico similarly stabilized.

I have to say that Mexico production looks quite strange in comparison to many other regions. It may have a lot to do with the single super-giant oil field in the mix. This does have an effect on a stochastic analysis as it adds a highly non-random and deterministic component to the dynamics. I'm rather glad I did not attack Mexico first. Remember the Alamo?