Natural language processing (NPL) was employed to streamline the examination of historical well documentation using smart search terms, for use in CO₂ storage developments. Extracted information was then pushed to the Wellbarrier™ well integrity life cycle solution to enable objective well integrity risk assessment.

Repurposing fields and existing wells for CO₂ storage developments is often seen as a more attractive option than drilling new wells, mainly due to the cost and schedule reductions it offers. Key technical aspects that determine the viability of repurposing existing wells, from a well integrity perspective, include addressing the challenges posed by legacy abandoned wells, and utilizing digital tools for data mining. Additionally, standardizing well barrier schematics is crucial for performing qualitative risk assessments throughout the life of the field.

The first step in assessing the integrity of the legacy wells for CO₂ storage was to acquire and verify the accuracy of historical well records, from both public and company data sources. The key documents examined were well integrity reviews, well completion reports, casing and cementing reports, and daily drilling reports (DDRs).

To optimize the substantial effort typically involved in reviewing hundreds of DDRs for well barrier qualification information, natural language processing (NLP) was employed. This provided a dashboard with smart search terms for key integrity information, such as pressure tests and cement jobs. The extracted data was then pushed into the Wellbarrier solution via its API to automatically generate well barrier schematics, saving an estimated month of man days across 11 wells.