Premium
AUTOREGRESSIVE PATTERN RECOGNITION APPLIED TO THE DELIMITATION OF OIL AND GAS RESERVOIRS *
Author(s) -
BOIS P.
Publication year - 1980
Publication title -
geophysical prospecting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.735
H-Index - 79
eISSN - 1365-2478
pISSN - 0016-8025
DOI - 10.1111/j.1365-2478.1980.tb01243.x
Subject(s) - autoregressive model , facies , geology , trace (psycholinguistics) , entropy (arrow of time) , petroleum engineering , pattern recognition (psychology) , mineralogy , statistics , computer science , mathematics , artificial intelligence , paleontology , thermodynamics , linguistics , philosophy , physics , structural basin
A bstract It is often difficult to precisely determine the boundaries of oil and gas reservoirs in the horizontal and vertical directions. Autoregressive pattern recognition is used to reveal lateral facies variations and to specify reservoir boundaries. The method is based on the calculation of autoregressive coefficients for short trace sectors between the top and bottom boundaries of the reservoir. These coefficients are determined by applying the maximum entropy principle. Once these coefficients are known, an estimate can be made of the power spectra of these trace sectors. Then these coefficients will act as characters for a pattern recognition algorithm. A decision criterion is used to distinguish the trace sector corresponding to layers impregnated with oil or gas and those impregnated with water. Two examples are given to show how autoregressive pattern recognition allows to accurately delimit gas or oil reservoirs.