USER'S MANUAL FOR CORRELATOR, VERSION 5.2
Ricardo A. Olea, Robert J. Sampson · 2003
When computers first became widely available in oil companies and universities in the 1960's, everyone imagined that these miracle machines might relieve the tedium of the many repetitive tasks that occupy the time of professional geologists. One of the early dreams of stratigraphers and oil explorationists was that it might be possible to teach a computer to correlate well logs, by examining the wiggles in log traces from one well and finding a matching set of wiggles in logs from another well. In 1980, I gave a talk at the AAPG annual convention in Denver, describing all of the correlation procedures that had been attempted at the time. The talk was very pessimistic in tone, because I considered geological correlation to be an ill-defined, underdetermined, one-to-many problem, and such problems are not amenable to analytical solutions. In spite of my gloomy outlook, the next year Don Beaumont invited me to Texaco headquarters to discuss computer correlation of well logs. I repeated my AAPG presentation and again expressed my reservations about the possibility of automating such an unquantifiable procedure as geological correlation. At the conclusion of my talk,