For oil and gas focused attendees, the highlight of the IRM UK Enterprise data and business intelligence conference held last month in London was the presentation on a ‘robust common master data foundation for oil and gas’ made by Tullow Oil’s Mesbah Khan and Chris Partridge of Boro* Solutions. The upstream presents a complex landscape of vendors, contracts and assets. Each contractor has a different system and interoperability is a problem. Tullow’s approach has been to focus on consistent terminology and clean data across drilling, development, operations, finance and legal.
The Boro methodology is a ‘Clean’ and ‘Pure’ data architecture developed for the oil and gas vertical. ‘Clean’ describes the overall process of data consolidation, loading, enrichment, assimilation and ‘novation.’ Further deconstruction of the awkward acronym reveals a sequence of data cleansing and repackaging into a single ‘Pure’ model sans duplication—the holy grail of a single source of the truth.
‘Pure’ in turn breaks out as a ‘precise, unambiguous, re-usable and extensible’ view of the data world that seeks to build bridges between islands of transactional data. This is achieved using a semantic layer and a foundational ontology. The latter builds on Matthew West’s work for Shell on extensional 4D (time based) data modeling such that, for instance, petroleum agreements can be modeled as objects that vary in time and spatial extent.
Further ontologies describe complex relationships between joint venture partners, social entity, business units and group legal structures.
After his talk we quizzed Partridge on how Boro integrates the standards space. He pointed us to a presentation showing the evolution of Boro’s industrial ontology from roots in ISO 15926 Part 2, ‘Ideas,’ the International defense enterprise architecture specification, and the US ‘DoDaf’ Department of defense architectural framework.
Ontological modeling is not for the fainthearted. Prior work for Shell on the Boro website has it that, ‘ontological understanding needs to be separated from epistemological and implement-ational gloss’ and that ‘collaboration between conceptual information systems modelers and those involved in philosophical ontology is potentially fruitful.’
Tullow’s ontology derives more prosaically from IBM Maximo. Other industry data models (Witsml, PPDM, Norsok) appear to have been found wanting. The Tullow development was a four person-year project. The IRM-UK presentation was made with Prezi, a snazzy alternative to boring old PowerPoint. Visit IRM-UK and Boro Solutions.
* Boro—Business object reference ontologies.
© Oil IT Journal - all rights reserved.