Earth Science Analytics has completed a Missed Pay Prediction project using machine learning and big data analytics in the Northern North Sea.
Earth Science Analytics has completed a Missed Pay Prediction project using machine learning and big data analytics in the Northern North Sea.
In collaboration with the UK Oil and Gas Authority, the Oil and Gas Technology Centre and the Norwegian Petroleum Directorate and partner TAQA Bratani Limited, Earth Science Analytics has conducted and completed a case study on missed pay identification over 5000 plus wells in the Northern North Sea.
The developed workflows consist of a comprehensive semi-automated quality control (QC) and well log editing steps followed by the implementation of machine learning algorithms to predict well log properties of interest e.g. porosity, HC saturation, lithology, reservoir and pay flags. The machine learning step itself includes various QC and uncertainty quantification measures ensuring the quality of the predictions. Earth Science Analytics has further developed visualization and analysis tools to interrogate and validate the findings with the necessary geological sense checks.
Earth Analytics CPO Ehsan Naeini, said: ‘Discovering missed pay zones has the potential to postpone costly decommissioning projects and provide new exploration targets. Missed pay can result from a multitude of factors including poor data quality and coverage, insufficient logging runs, poor petrophysical evaluation at time of discovery, a focus on the primary pay zone, and loss of knowledge due to company mergers or buyouts.’
