Case Study: Pertamina Hulu Mahakam reduces property modeling time by 75% with SLB Machine Learning for Property Modeling

Pertamina Hulu Mahakam utilizes machine learning to reduce property modeling time from 4 weeks to 4 days

Pertamina Hulu Sanga Sanga faced a complex challenge in its North Handil Field, which contains over 300 thin reservoirs and has 45 years of production data. The field's intricate geology and dynamic oil-water contact made it difficult to efficiently create accurate 3D property models. The customer needed a new approach to drastically reduce modeling time and improve model quality to better inform its field development plans. They partnered with SLB to evaluate a machine learning solution.

SLB implemented its Machine Learning for Property Modeling solution, powered by the EMBER hybrid algorithm, which integrates machine learning with geostatistical techniques. This solution eliminated the need for time-consuming preparatory work like variogram analysis and data de-trending. The results were significant: the new model showed an 18% reduction in average prediction error and a 75% reduction in property modeling workflow time, accelerating the process from weeks to days. This allowed geomodelers to focus on analysis and enabled SLB's customer to quickly update models with new data.

PERTAMINA Hulu Sanga Sanga