ARTIFICIAL NEURAL-NETWORK MODELING AS AN AID TO SOURCE-ROCK CHARACTERIZATION

A data-based approach is used to establish detailed and accurate geoch emical characterization in oil source rock intervals, using well logs and noisy information from cuttings. The method starts by extracting e xamples from the intervals of interest, using generalized relationship s between total organic content (TOC) and well log responses. The exam ples were used to train an artificial neural network (ANN) that extrac ted more detailed and accurate relationships between TOC and well-log responses specific to the study area. A combination of the 'quickprop' algorithm and 'Dynamic Node Creation' scheme was utilized to facilita te efficient training. The trained ANN is useful for mapping source ro ck intervals in the area of interest. This method performs satisfactor ily when applied to the Eg ret source rock of the Jeanne d'Arc Basin, offshore eastern Canada.