Predicting overlooked hydrocarbons in 5000+ wells using AI
Earth Science Analytics delivered the first ever cross border machine learning project awarded by the Norwegian Petroleum Directorate (NPD) and the Net Zero Technology Center in the UK (formerly known as OGTC). The objective was to predict overlooked hydrocarbon intervals amongst thousands of wells on a basin scale leveraging cloud native big data analytics and machine learning technologies facilitated by our web-based platform EarthNET. Operators from the UK and Norway donated over 30,000 kilometres of wireline data which was used to predict missing CPI curves and identify over 300 exploration opportunities.
Predicting overlooked hydrocarbon intervals amongst 5000+ wells on a basin scale using machine learning technologies
1257 UK exploration wells, 643 Norwegian exploration wells, and 2924 UK development wells
The Norwegian Contiental Shelf (NCS)
Finding missed opportunities in wells previously marked as dry
A typical exploration team can deal with 20% of the available data before they must make a drill decision. Therefore, we tend to focus efforts on the best quality or most well-known pay zones, and miss out on those 'diamonds in the rough'. In both the UK and Norway there are numerous examples of discoveries that are triggered by forensic analysis of data in presumed dry wells that later proved to be hydrocarbon bearing.
The identification of missed pay in already producing assets or exploration wells can lead to an increase in reserves and provide ‘cheap’ oil as well as highlight new exploration targets.
AI-assisted geoscience software for predicting overlooked hydrocarbons
Our AI-driven software, EarthNET, enables you to predict previously overlooked hydrocarbons with speed and precision. With EarthNET, we successfully processed vast amounts of data from two different government repositories and re-analysed them for overlooked opportunities.
This process involved:
Building a data bank of 5000 wells
Multi- and single-well log data editing and quality control
Porosity, water saturation and multi-class lithology prediction
Net reservoir and pay identification
"We are delighted to be working with Earth Science Analytics to further investigate the application of machine learning to data analytics through using their software, EarthNET. This project will improve our subsurface understanding in the Northern North Sea - and potentially unlock previously overlooked reserves.”
Overlooked pay visualisation and ranking for over 300 opportunities
Using standard techniques for identifying and ranking pay zones, we found over 300 candidates that could be worth a second look. As a part of this project, we also delivered strategic technology and results to the OGA as part of the 2020-25 digitisation plan.
300+ instances of hydrocarbon pay in wells classified as ‘dry’
End-to-end workflow reduced from unfeasible to 3 months
Enabling exploration teams to tackle challenges that would otherwise be too time-consuming or costly