Well planning remains one of the most demanding aspects of drilling engineering. When a Fortune 100 company wanted to improve well planning by providing drilling engineers with a comprehensive, improved view of well offset data, they faced a daunting task: connecting data scattered across multiple locations and analyzing it simultaneously. Data quality issues – such as inconsistencies in well naming across systems – made data integration a challenge; even drilling engineers found it difficult to aggregate and access all data about each well in one searchable location. In addition, the company’s well sensor data contained vast amounts of detail, but it had not been mined to detect events such as kicks, lost circulation and other drilling problems.
The Maana Knowledge Platform significantly simplified the process of data aggregation, preparation, access and analysis by crawling data from multiple technical sources, including WellView databases, rig sensor files and surface parameters, downhole log files, technical documents, risk assessments and well problem databases. Using the platform’s natural language processing and machine learning capabilities, the company’s subject-matter expert indexed the various data types in these disparate sources, made pertinent connections between the datasets, ran analytics and rendered the data searchable and relevant.