Researchers have dedicated numerous applications of machine-learning (ML) techniques for ﬁeld-scale automated interpretation of well-log data. A critical prerequisite for automatic log processing is to ensure that the log characteristics are reasonably consistent across multiple wells. Manually correcting logs for consistency is laborious, subjective, and error prone. For some wellbore logs, such as gamma ray and neutron porosity, borehole eﬀects and miscalibration can cause systematic inconsistencies or errors that might be present even after the application of wellbore and environmental corrections.
Biased or consistently inaccurate data in the logs can confound ML approaches into learning erroneous relationships, leading to misinterpretations, such as wrong lithology prediction, reservoir estimation, and incorrect formation markers.
To overcome such diﬃculties, we have developed a deep learning method to provide petrophysicists with a set of consistent logs through the multiwell automatic log correction (MALC) workﬂow. Presently, the corrections we target are systematic errors on the standard logs, especially gamma ray and neutron logs, random noises, and to a lesser extent, local formation property misreading due to washouts.
We applied the proposed method in multiple ﬁelds worldwide containing diﬀerent challenges, and in this paper, we include the results in two ﬁeld examples. The ﬁrst one covers the correction of synthetic coherent noise added to ﬁeld data, and the second example covers the correction applied to original measurements.