The Schlumberger Information Solutions (SIS) 2006 Forum was subtitled ‘breakthrough team performance’ but perhaps a better title would have been ‘The Petrel Show’. The plenary session covered familiar ground. Industry is on a roll, but constrained by lack of drilling rigs and other resources. The National Oil Companies are on the up, while the Internationals are constrained by lack of personnel. Everyone is planning to hire—but where these folks are to come from is uncertain. A solution to the people shortage, conveniently for SIS, is IT-led productivity gains. Hence the enthusiasm around Petrel which has built great momentum as a tool for integration of geoscience and engineering data. The ‘fly’ in the Petrel ointment is its own integration with other interpretation tools. Petrel files proliferate like Excel spreadsheets—making the data manager’s job hard.
Two other themes from the conference are ‘openness’ and web services. Openness to Schlumberger means the Ocean API and the Seabed data model. But don’t imagine that SIS is going open source. The Ocean API is for the development of plug-ins to Petrel (and in the future Osprey, Merak, and Avocet) and the Seabed data model has been published as a textual description of the database, rather than a DDL. But understand that third party developers using Ocean are impressed with its functionality. Likewise, the exposure of the Seabed intellectual property has raised some approving eyebrows in the industry.
On the web services front, the situation is less clear. SIS’ Petrel, Ocean and Seabed flagships do not appear to expose much in the way of web services that could be leveraged by third parties. Indeed there is a natural tension in a services-oriented architecture (SOA) which could for instance, move control of the workflow from Schlumberger’s infrastructure to a client’s script. SIS’ Joe Perino put it succinctly—‘SIS has embarked on a move toward SOA and will offer, for instance, versioning and results management as a service. But it is up to operators to push us to use open systems.’
Charles Johnson, MD of Microsoft’s manufacturing industry unit, described the partnership with Schlumberger. Microsoft is to offer Excel-based spreadsheet access to data in Schlumberger’s data bases, just like the Microsoft/SAP ‘Duet’ product. Johnson noted Microsoft’s arrival in the high performance computing (HPC) marketplace with Windows Compute Cluster Server 2003 (CCS) and the 64 bit edition of Windows XP. This was greeted with some skepticism and Johnson admitted that those requiring 64 bit support and robust debuggers will have to wait on Vista.
Stephen Whitley offered an insight into SIS’ IT architectural vision. Project teams include the developer, a business person and an ‘architect’ to assure conformance with Schlumberger’s Technology Blueprint Vision (TBV). According to the TBV, all products ‘must be open and interoperate’. There is focus on standardization. SIS ‘will adhere as closely to ProdML as possible.’ The roadmap favors software that exploits multi core processors and ‘wide scale XML-based data frameworks’. The desktop shift to Windows .NET was ‘a tough decision for SIS, traditionally a Unix company.’ Enterprise computing is to leverage OpenSpirit and Microsoft BizTalk and will ‘adhere to BEPL2.’
Jevon Williams presented Shell’s hydrocarbon development planning workflow (HDP). This leverages LiveQuest along with an in-house tool for uncertainty management. A ‘Smart Workflow System’ keeps an audit trail of decisions made. The solution is ‘blue printable’ and can be deployed in other contexts. Projects use Shell’s Active Directory to build and notify teams. OpenSpirit is leveraged to build a Petrel project.
The Petrel Papers
Petrel is now integral to Petroleum Development Oman’s (PDO) subsurface asset management as Talib Al-Ajmi explained. PDO now has 100 active models and 125 licenses (for 200 users). Usage spans petroleum engineering, reservoir engineering and petrophysics (seismic interpretation is under development). Petrel offers good integration of fracture data, mud loss data, PLT (production logs), pressure, seismic well etc. This lets PDO highlight areas where efficiencies are possible. Full field models let PDO keep facility maps up to date by including satellite photos, pipeline data etc. Petrel has proved an ideal collaborative environment for data and interpretation QA/QC and managing uncertainty.
Arun Narayanan presented the Petrel data management roadmap which ultimately will combine flat file performance with the features of a database. The Petrel data management roadmap extends out to 2008 when a Petrel ‘DBX’ database will interface with Seabed—bringing full units and coordinate reference system support.
Seismic in Petrel
Peter Diebold described Shell’s evaluation of Petrel’s seismic interpretation (PSI). Datasets over 10GB are common in Shell, often with a mix of 2D and 3D data. Diebold interspersed his evaluation with proselytizing for a paradigm shift from what Diebold describes as the old-fashioned ‘paper on screen’ (PoS) technology—advocating instead a modern ‘sparse’ 3D approach. Petrel was evaluated in this context across a matrix of 8 ‘themes’ and was tested all over world on different live projects. The results showed that Petrel was great for ‘quick look’ seismic interpretation and integration with non seismic disciplines. Users commented that they had ‘never seen volume interpretation come so naturally’. Petrel’s workflow management, multiple realizations and its use in interpretation quality assurance were positive points. On the downside, Petrel’s lacunae in data management were deemed ‘a huge issue’. Legacy data integration was problematical and OpenSpirit proved ‘slow and unpredictable’. Shell has recommended limited deployment of PSI, for example in the New Business Ventures group. PSI requires training and a mindset change for the PoS ‘gang’ found in middle management.
This article has been summarized from a 14 page report produced as part of The Data Room’s Technology Watch reporting service. For more on this subscription-based service please visit the Technology Watch home page or email firstname.lastname@example.org.
© Oil IT Journal - all rights reserved.