Tomorrow's Petrophysics: Blending Big Data, Artificial Intelligence and Human Insight (Kjetil Westeng, Aker BP)


Kjetil Westeng from Aker BP


Nils Andre Aarseth, Peder Aursand and Tanya Kontsedal (all AkerBP)


The rapid advancements in the petrophysics sector over recent years have transformed the tools and methods available to professionals. New databases such as OSDU provide enhanced access to cross-disciplinary data for cross-platform use. This has paved the way for a more systematic approach to data combination, presenting numerous novel opportunities. At the same time the artificial intelligence revolution is affecting almost all aspects of life, and certainly it provides us with many opportunities in the analytical work of petrophysics.

Embracing these new possibilities will allow the users to:

Enhance their efficiency and effectiveness.

Produce results of a superior quality.

Ensure a higher degree of consistency in their findings.

Leverage diverse datasets for innovative interpretation results.

Achieve a more cohesive and integrated working process.

However, as we relentlessly refine our methods, tools, and workflows, certain considerations remain important:

It’s important to plan for the human involvement in the workflow where appropriate, ensuring that the process remains engaging and relatable.

There’s potential in exploring the synergy between AI/automation and human input. Could their combined strengths offer benefits that neither could achieve alone? How can this be done most efficiently.

Given that multiple inputs might exist for the same parameters, finding strategies to combine them effectively is essential.

Recognizing instances when results might be unreliable is key, with a focus on factors like data uncertainty and the overall dataspace. Can be done automatically and with good QC tools.

Making steps towards more transparent AI is vital, ensuring that the relationship between inputs and outputs remains clear and understandable. The goal is twofold: to optimize the appropriate use of the models and contribute to growth in domain knowledge and expertise among users, a challenge often faced with black box solutions.

Lastly, practical and flexible implementation remains a core focus. It’s imperative to ensure these advancements and methodologies are seamlessly integrated into the daily routines of users. This work delves into our experiences as we adapt our existing workflows from the ground up and pioneer revolutionary subsurface approaches, highlighting the opportunities and challenges we’ve encountered.


Kjetil Westeng is an Advanced Petrophysicist at Aker BP. He earned his MScdegree in petroleumtechnology in 2003 from NTNU, Norway and began his career as a field engineer at ResLab Sampling the same year. In2005, he joined Ødegaard as a petrophysicist, and by 2007, he took on a role as a Rock Physicist/Petrophysicist and soon Senior Rock Physicist at Schlumberger. Since2015, Westeng has been working at Aker BP, where he has played a key role in leading the digital transformation in the use of well data by leveraging his knowledge of petrophysics, Python programming, and datascience. His responsibilities include developing innovative petrophysical interpretation methodologies, designing algorithms, and creating interdisciplinary solutions to tackle complex challenges in the industry.