Resumo

ALZATE E., Guillermo; BRANCH B., John William; SUAREZ B., Oscar Mauricio  e  VEGA S., Carlos Andres. WELL LOG CORRELATION WITH ARTIFICIAL NEURAL NETWORKS USING SEVERAL CURVES OR LOG PARAMETERS. Bol. cienc. tierra [online]. 2006, n.19, pp.87-99. ISSN 0120-3630.

Well logs correlation if it is performed by traditional way (visual inspection) is a complex activity which it takes long time and a rigorous commitment from one or more human experts. This paper presents a computational tool based on artificial neural network (ANN) to correlate logs among wells and also the results for a field study case. It is presented the process of well logs correlation and its requirements of qualified people and time specially for hydrocarbon reservoirs with complex geology and many wells. Finally, it is presented a computacional tool and its application. The final results for a five wells’ logs study case show a good performance, a fast and close approach to the results obtained by an human expert using the conventional visual inspection procedure. The developed tool based on ANN presents good versatility to correlate several well log curves beginning from the identification in any well of oilfield or reservoir of a lithologic marker previously picked on a reference well. The final result is a reliable and effective computer tool to correlate a stratigraphic marker from a well to another one in a hydrocarbon reservoir or oilfield.

Palavras-chave : Well Logs Correlation; Artificial Neural Network; Backpropagation.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )