ICE 2022


One common issue found when working with well log data is the irregular abundance/availability of the different parameters that may be registered. This is especially true when working with datasets collected in different campaigns that may span through many years, even decades. Artificial Intelligence may be useful to fill gaps in the original database, resulting in a more complete, standardised one. The process can be performed iteratively, successively populating the database with missing parameters, starting with those for which there are more available training data and whose results are more reliable. In this work, we present an example in which we filled an incomplete dataset consisting of wells provided by the UK National Data Repository (NDR) of the Oil & Gas Authority (OGA). The performance of some of the most commonly used artificial intelligence methods (random forest, multi-layer perceptron, gradient boosting etc.) was tested varying their hyperparameters until reaching an adequate result. A segmented analysis of the predictions was carried out to determine whether the regression reliability varies depending on properties such as lithology, stratigraphic unit, location, etc. If the predicted values are accurate enough, filling the gaps also helps to improve the training and application of classifiers as more observations may be included.