When an engineering prime contractor (EPC) builds a large facility, be it an offshore FPSO or a power station, huge data management issues arise. These are compounded when the construction company hands the project over to the Owner Operator (OO). Much thought has gone into technical data models which set out to standardize the way engineering data is recorded and served-up to the different stakeholders. These have had a limited impact, partly because the EPC industry is dominated by a few large companies who have their own internal processes and ‘standards’ and who may be reluctant to share these with competitors. Companies are spending large amounts on data collection and clean-up. Another trend is the increasing communications between the EPC and OO early in the project’s life. By integrating – or at least planning for integration – early in the project, the drama of a ‘big bang’ handover can be attenuated.
Michiel Molenaar described information handover on Shell’s $ 4 billion Nigerian Bonga deepwater project, implemented by no less than five EPCs located in 3 time zones. The Bonga Strategic Asset Management Solution covered standards setting and data management for handover from the EPCs to the Owner Operator. The project involved establishing standards, classifications, a ‘Hand-over Guide’ and quality program. Lessons learned included the need for contractual clarity and the need to communicate through frequent meetings.
Statoil’s Norwegian Kristin field development was designed to be a ‘paperless project’ according to Bjorn Henrik Magnus. Working with EPC Kvaerner, Statoil set out to get rid of paper-based working practices and to replace them with electronic systems. The project applied to suppliers – there is ‘no paper to or from suppliers to Kristin’! The Kristin plant data warehouse was built around Intergraph’s SmartPlant. The software allows for impressive 3D navigation through the data warehouse.
Pete Mayhew, Information Manager with EPC AMEC described the challenge of handing over 2000 major equipment items, 200k tags and 2.5 million attribute values. AMEC deploys the STEP data model inside its ORBISS data warehouse. Previous issues of performance, security and bandwidth have largely been solved. Software can call up data from the data warehouse and link into the 3D model. IM is no longer a ‘by-product of technology’. Mayhew concludes that ‘it is business requirements and quality that are the challenges’.
Ewan Botterill expressed a contrarian’s view of asset management which is “not a cycle”. Different tools and approaches are required at different times in life of a plant. Engineering used during construction is different from maintenance and needs its own tools. Also the EPC and OO have different views of data. “Getting a fully integrated, clean, fixed data set to populate a data warehouse may never happen – as a project evolves, so does the volume and completeness of the data.” Rather than a single source for all data, Bottrill advocates a ‘single point of access to data’ with as much data as possible moved from EPC to OO systems on handover. This is colorfully termed the ‘evergreen digital plant’. Aveva leverages ISO 15926 , XML Template and RDL via Microsoft .NET and Sharepoint – BizTalk. At least 50% of the cost of integrated asset management goes into content services.
Ian Bishop (Foster Wheeler) notes that the expected benefits from data sharing have so far failed to materialize – although the original arguments still hold true. Standards and web technology have moved forward but while there are a number of web-based IT solutions, we have not yet achieved a true collaborative, sharing environment. Factors weighing against this are a focus on reducing costs and a ‘document centric’ approach. The Holy Grail of the single data base is lost! We still need both document and asset-centric views tolerant of low bandwidth global access. Engineers are risk averse and slow to change – being document-centric is a comfort to some. We need to build an engineering critical path based on key information, managing key information with documents and accepting that non-critical data will lag.
© Oil IT Journal - all rights reserved.