Plant Information Management at Statoil Norway
By: Dr. Stanley Port, CIMdata, Ltd.
This case study describes the implementation of a data warehouse for managing the plant life-cycle information for Åsgard B, the largest floating gas platform in the world. The data warehouse adopts the POSC/Caesar data model, and is built on the Notia software system supplied by Intergraph. This is an exciting new information management project on which much process industry attention is focused.
1. Statoil and Åsgard
Statoil is the Norwegian state-owned oil company with 17,000 employees and revenues in 1996 of NOK 106 billion (U.S. $14 billion). It has operations in 25 countries and ranks as the world's second largest net exporter of crude oil, and a leading marketer and transporter of gas.
Åsgard is one of the largest and most extensive oil/gas developments on the Norwegian continental shelf in terms of reserves, investment and technological challenges. It lies in deep water about 200 km (125 miles) off mid-Norway, near the Arctic Circle. Around NOK 33 billion (US$4.4 billion) is being invested in production facilities, wells, and subsea installations for the field. Oil production is due to start in late 1998 from Åsgard A, one of the world's largest oil production ships. This impressive facility is 278 meters (912 feet) long, has a displacement of 184,000 tons, and will be moored to the seabed. It has a processing capacity of 201,000 barrels of oil a day and can store over 900,000 barrels.
Gas exports will begin in the year 2000 from Åsgard B, the largest floating gas platform in the world. By 2007, Åsgard will be delivering some 10 billion cubic meters of gas per year to continental Europe, or roughly 15 percent of annual Norwegian gas deliveries. Hydrogen sulfide and water will be stripped out on Åsgard B before the gas is compressed and piped to land. At the same time, this platform will be able to produce 41,000 barrels of oil and 94,000 barrels of condensate a day.
Statoil places great emphasis on developing close relations with its suppliers. Good collaboration has been established with several Norwegian and international partners. The company's supplier development program has helped achieve new products which cut costs, improve safety, lower environmental impact, and open new commercial opportunities for participating companies. Strengthening relations with small- and medium-sized companies will be priorities in coming years. Apart from Statoil, the licensees in the Åsgard project include Norsk Agip, Total Norge, Mobil Exploration Norway, Neste Petroleum, Saga Petroleum, and Norsk Hydro Produksjon. Norway's Kværner group and Statoil are collaborating through an integrated Central Engineering Team (CET) of about 550 people with joint responsibility for building and completing the Åsgard B platform.
Figure 1: Åsgard Field Concept-Layout and Installations
2. Challenges and Issues
Projects with the scale of Åsgard involve efforts by numerous people in the owner/operator company, engineering and procurement contractors, an army of sub-contractors, and throughout a vast chain of suppliers. Inevitably these organizations are dispersed geographically across the whole world. The platform itself comprises many hundreds of thousands of components that must be manufactured or procured, and brought together for its fabrication. The life cycle of Åsgard B will extend for 30 years or more. Project phases start with the conceptual process design, and move through conceptual engineering design, detailed process and engineering design, fabrication, commissioning, and handover to the plant operator. Later phases include maintenance-and eventually decommissioning and disposal of the platform. During its long life, the platform is likely to undergo several changes of process or use, each of which involve engineering redesign, more procurement, and revamp.
The engineering and procurement contractors create much of the plant information, and eventually they hand it over to the custody of the plant owner/operators. Equipment suppliers, operators, maintenance contractors, and many others depend on having easy access to accurate plant information. Traditional working practices have caused fragmentation of information and activities among these groups. Inevitably the only solutions available in the past were to employ more man-hours and to create mountains of documentation in the offices of the many participants. All this, in turn, led to duplication of work, lack of control, errors, and to even more fragmentation. Any proposed change in plant configuration tends to result in after-effects, which ripple outwards to require more changes in connected components and in surrounding areas of the plant. They also affect several organizations. Whether they are designers, fabricators, operators or maintenance operatives, workers are too likely to base decisions on out-of-date information, resulting in wasted time, abortive work, or the purchase of materials and equipment which are unsuited for use. Inevitably too, it has been difficult to produce the accurate "as-built" or "as-maintained" information which is so necessary for subsequent plant activities.
The pressure is always for cost reduction, faster project completion, and for the practice of concurrent engineering. The scale and complexity of these projects, the number of participating organizations, as well as prevalent issues of safety and the environment, all combine to cry out for a better solution for the life cycle information for the plant.
3. The Solution
The solution required to support Åsgard was a single source of accurate, up-to-date, and consistent plant data, accessible electronically by any organization that has justifiable cause to need it, from wherever they are located. The data store must be dynamic, to cope with the mounting volume of information and continual changes in the types of data required at any stage in the lengthy plant life cycle. Since it is impossible to foresee the nature of all the computer applications that will draw upon and use the data, the solution must:
- Integrate data from a wide variety of legacy sources and organizations
- Separate the data from the computer applications that create, update or use it
- Permit applications to interoperate, so users can exchange and share the data
At Statoil, they interpreted these as requiring an open and flexible solution based on international standards.
POSC-Petrotechnical Open Software Corporation-is a not-for-profit, vendor-neutral membership corporation founded in 1990. Its mission is to benefit the exploration and production industry by establishing, maintaining, and promoting specifications to be used as standards for the sharing of information through the asset life cycle. Statoil is one of the major sponsors of POSC, some others being BP Exploration, Chevron, Elf Aquitaine, Mobil, Saudi Aramco, and Texaco. POSC now has 130 members and operates from offices in London and Houston.
Statoil, Kværner, Norsk Hydro, Saga Petroleum, Aker Engineering, and Det Nortek Veritas were among the original sponsors of CAESAR Offshore. This is a joint industry project for data standardization initiated in Norway in 1993. It had the objective to reduce the life cycle cost and development time of offshore oil and gas production facilities.
In 1994, CAESAR Offshore was reorganized and a collaborative project was agreed with POSC, resulting in the birth of the POSC/CAESAR project. This extended from 1994 to 1997, was located in Oslo, but international in scope. Since 1995 Brown & Root, BP, and Elf have also sponsored the project, and since 1997 Intergraph, IBM, Oracle, Lloyd's, Shell, ABB Technology, and UMOE Technology have joined the project as sponsors. Its stated vision is to:
- "Reduce the life cycle cost and development time of oil and gas facilities by improving sharing and exchange of information among those involved in the facilities' life cycle."
- Produce agreed standards for digital descriptions of facility products
- Understand and facilitate the use of available technology for implementing the standards
- Encourage and assist adoption of the standards and technology by business
- The Product Data Model (P/C-PDM), which defines the underlying data terminology and structure
- The Reference Data Library (P/C-RDL)--previously called "class library"-which is a dictionary defining standard data items that are relevant for petrochemical assets, including all the various materials, activities, facilities, and relationships between things. The RDL is designed to be extensible.
BP, Shell, and Brown & Root decided to use the P/C-RDL as basis for their data warehouse in the ETAP (Eastern Trough Area Project) offshore development project in the UK sector of the North Sea. The partners have stated that they expect the full benefit of the ETAP data warehouse to be a 10 percent saving of lifetime asset costs. The plan was to have their data warehouse populated by the end of 1997. The POSC/CAESAR project has also entered into a collaboration with PIPPIN, an European Union Esprit IV funded project, which is a joint industry initiative to generalize the work of the ETAP data warehouse.
VÅV Take-Up Project and the Data Warehouse
Three owner/operators formed the VÅV Take-Up Project--a cooperation among the following Norwegian offshore development projects:
- Varg--operated by Saga Petroleum
- Åsgard--operated by Statoil
- Visund--operated by Norsk Hydro
The objective of VÅV was to implement a data warehouse based on the P/C data model. A data warehouse draws on raw data from different sources throughout a company and from external sources, and organizes it into ready-to-use business data. A data warehouse differs from conventional data storage because it aims to be:
- Integrated: Source data is integrated and stored in the warehouse in a standardized neutral format-in this case using the P/C model. Integration is essential because business decisions are generally based on engineering data generated in different departments and organizations, and by several disciplines of people.
- Subject-oriented: Data from fragmented systems is stored in the warehouse by subject, so it can be retrieved, analyzed, and used more easily.
- Time-variant: This means data is unalterable by users after it has been submitted for storage, though new data including plant modifications or changes may be added from various applications, and time-stamped.
- Non-volatile: Data is stored as read-only. Users can extract data and analyze or manipulate it, without altering the data store itself. Data is additive.
Notia is Intergraph's data management and data warehouse product. Taking advantage of its STEP track record and sponsorship of the P/C project, Intergraph has developed Notia as an open product data manager, specifically to handle the life cycle of plant and process data. It is an implementation of the P/C-PDM and P/C-RDL based on Intergraph's AIM product data management system. Notia leverages the power of OMF (Object Management Framework), a market-leading engine for product data management from Metaphase, Inc.
Beginning early in 1997, the VÅV Take-up Project set up an agreement with Intergraph to supply and implement the Notia software. Kværner had just been appointed prime Engineering and Procurement Contractor for Åsgard B, and it was likely that this company and its partners would begin amassing design data for the project in earnest by September 1997. Intergraph duly supplied its Notia software and the acceptance tests were completed on schedule by the key date of August 20, 1997.
Plant Information Management (PIM) in Åsgard
In January 1997 a workgroup for implementing the Åsgard data warehouse was set up under its project manager, Mr. Bjørn Henrik Magnus. The team comprises experts in IT and in technical information logistics, as well as piping, instrumentation, electrical, and mechanical engineers. This team is responsible for developing applications on top of Notia needed for the physical implementation of the Åsgard project. The focus of the team has been to quickly achieve real and measurable benefits by improving information flow and work processes that depend on Life Cycle Information.
The P/C model holds data in a highly normalized form. This means, for example, that data to describe a pump comprises many data objects, each having few attributes, but the objects are linked by many relationships. It is the extensible P/C-RDL that gives meaning to the generic P/C-PDM and ensures it can be used for practical application in the oil/gas industry. When populated with data that describes all the materials, equipment and activities associated with Åsgard B, the data warehouse becomes the single source of plant information usable by all concerned with the platform during its entire life cycle. The prime source of data is the Åsgard B legacy database. However, all the data-whatever the source-needs to be checked for quality and referential integrity, and converted to P/C Part 21 format, for feeding to Notia. After that, it becomes possible to extract from Notia the data required for any compatible computer application, and this can emerge in P/C Part 21 format.
Notia provides a Query client to allow users to set up search criteria, while a Browser client displays the results of a query. Intergraph has also delivered a Web-based query browser to allow Åsgard participants and suppliers, wherever they are located, to connect to a Web server. This, in turn, is linked through a firewall to the data warehouse.
At present, the P/C data model does not have entities for graphical data, so the data warehouse currently does not contain drawings such as P&IDs or CAD models of the plant. However, the P/C-PDM with its expanding RDL is rather broader in scope than the current STEP Application Protocols for the process industry.
Storing data has many advantages over the traditional approach of storing documents in hardcopy or electronic form. Documents are needed for many purposes and much of the data they contain is inevitably duplicated many times. Whole new document versions are formed and stored even when only a tiny proportion of the underlying data has changed. Sometimes each document is held simultaneously in several formats. While it is possible to reduce the number of documents, it is unrealistic to eliminate them. So the Åsgard team commissioned Intergraph to add facilities for capturing, storing, and retrieving documents in various formats in the data warehouse. This was done within the Åsgard LCI project using Intergraph's AIM system, which is seamlessly integrated with Notia. (See Figure 2) Manufacturer catalogues are held in AIM as compound documents.
It has been anticipated that a wide variety of computer applications will be developed by project partners to draw and operate on the data from the data warehouse, and perhaps to send more data back to it. The first application developed by the Åsgard Data Warehouse team is for dynamic datasheet production. Currently, 1,000 datasheets for different variants of equipment models have been developed, where each datasheet contains all the data necessary for engineering and purchasing. Such applications can communicate with the data warehouse either through P/C Part 21 file formats, or more directly through the Notia API. The Åsgard team is also working with Intergraph to develop tools for creating business objects; these are aggregations of data extracted from the data warehouse (or for sending to the warehouse) in a form that is compatible with applications.
4. Åsgard User Perceptions and Justification
The Åsgard LCI Data Warehouse project is responsible for the data warehouse technology, its functionality, and for loading standard data and document files. The CET operates the data warehouse and offers support and training to the Åsgard suppliers, and forwards requests to the LCI Data Warehouse project for additional data needed in the warehouse.
The Åsgard LCI Data Warehouse has a 30-strong team. One of its major functions is to extract and convert legacy data to Part 21 format so it can be imported into the data warehouse. Integrating this data and controlling its quality before feeding it into the data warehouse is accomplished with the mapping tools available, but this work has been challenging and time consuming. At the time of writing (February 1998), almost 1 gigabyte of data is loaded. As the project engineering proceeds, this volume is likely to spiral upwards into the terabyte range.
With Åsgard B now well into its detail design phase, Statoil is rolling-out a new version of its P/C data warehouse in February 1998. This will provide Web access for a multitude of partners-designers, contractors, and major equipment suppliers around the world. Most users will access the data warehouse to get information on equipment and plant items in datasheet form. They can have considerable confidence in the accuracy, timeliness, and completeness of the data, and it will be in a usable standard format.
Users can navigate through the Notia system to find items by many different means. These include the P/C material classification, the NORSOK-standard nomenclature, by manufacturer, plant model or variant, or by a dynamic datasheet (structured data) such as a valve datasheet. Navigation among documents and their revision histories, files, projects, and contracts are all handled as standard AIM object classes. This functionality has been identified as an area providing a major cost benefit to the project.
With so little experience available on data warehouses in this industry, the Åsgard team finds it difficult to predict the data volumes, the number of users, or system performance. However, in addition to the CET in Oslo, Statoil expects the following contractors to become users of the data warehouse:
- Emtunga, Sweden--Living quarters and heli-deck
- SaiRos, Italy--Separation and stabilization module
- Reinertsen Vigor, Central Norway--Gas hydration module
- Kværner Oil & Gas, Norway--Gas treatment module and gas compression module (engineering and procurement performed at Kværner Groupe Sofresid, Paris)
- Kværner Process Systems, Norway--MEG regeneration module
- Daewoo, Korea--Fabrication for the hull
Statoil estimates that data volumes will be about one hundred times larger than if the equivalent data was stored in traditional relational tables. This may surprise those unfamiliar with STEP-like data warehouses, The multiple arises mainly because the P/C product model with its highly normalized object structure results in huge numbers of simple objects, with numerous relationships among them. The volume of data also multiplies due to the accumulating effect of holding multiple versions of data for an object throughout the plant lifecycle.
Currently, an equipment datasheet is built and displayed within one second, but this is on a lightly loaded system. The team accepts that they may have to implement plenty of additional server power to handle the processing demand from users. However, they are unshakable in their view that the ability to hold a single source of project information and to provide any authorized user with a simple means of access from any location will result in far less duplication of engineering effort. It will also result in much less waste and rework, and more speed throughout the whole plant life cycle. In support of this view, the team points out that 85 percent of an offshore platform is composed of fairly standard products and equipment. But until recently, the lack of data standards and any easy means of managing data with modern tools has meant there has been little chance to curb the explosion of project documentation-with its attendant proliferation of versions, revisions, and copies.
The first version of the data warehouse is focused towards handling the structured information and documentation for standard catalogue-type equipment approved for use on Åsgard B. By making this information available from a single, central, verified source, the project contractors and suppliers will gain the following benefits:
- Help to reduce the variants of equipment used in constructing the platform
- Reduce time in requisitioning standard equipment
- Eliminate the need for suppliers of standard equipment to send out identical documentation to multiple sites and maintain all technical details on every data sheet
- Eliminate the administration of technical document and information flow for standard equipment
- Substantially reduce the volume of documentation handed over for plant operations
5. Conclusions and Future Expansion
Mr. Magnus believes the main incentive for implementing the LCI data warehouse in Åsgard was the wish to implement a standards-based structure for plant information (POSC/Caesar) and Web-based information sharing and exchange over the Internet. Conventional PDM systems handling proprietary formats could not meet these requirements, and so were not considered
The early emphasis in this project has been on data for standard equipment and materials. In future, they will concentrate more on tag information and on handling complex engineering data such as instrument loop information and electrical distribution networks. Integration of 2D P&IDs and 3D CAD models with the data warehouse to provide a more effective user interface for navigating the data warehouse is planned for 1998. In addition, data quality control, merging, and consolidation will be prioritized. More software releases will be made during 1998 with additional functionality being rolled out during 1999 and 2000. Statoil's ambition is for the LCI data warehouse to become the foundation for their STID portfolio of technical information systems by the time Åsgard B goes operational during 2000.
While the management of data in the P/C standard data model will reduce the proliferation of documents, there is a pragmatic realization that documents will not be eliminated entirely for a long time. This project therefore combines the concept of the seamless integration between a data warehouse (Notia) with the conventional data and document management (AIM) system. Also, Statoil's plans to provide document redlining and an integration with SAP/R3, which is used for financial and logistics information in Statoil.
Owner/operators are the custodians of the plant information on behalf of a vast number of widely dispersed users. It is clearly in their interests to maintain the information as an accurate and accessible model of their real facility. The worldwide efforts that have gone into developing data standardization and techniques for data modeling are already immense. The technology required is highly complex and expensive, and the dedication of the people who build and implement data warehouses should not be underestimated. From this case study, it is clear that there are huge, complex, and expensive engineering projects that are ripe to benefit from this technology. There are surely urgent and compelling business reasons for implementing the data management technology, and the financial benefits must be carefully studied and understood. Organizations and their management must grasp the opportunities that are available today.
Åsgard is clearly becoming an important milestone in the development of plant information management. It has needed clear sight and courage on the part of Statoil and its business partners to progress this far with data warehouse technology.
This is a reprint of one Case Study from fifteen presented in CIMdata's PDM Case Studies Subscription Service. You may purchase the PDM Case Study Subscription Service directly from CIMdata, Inc., by telephoning their office at 734-668-9922 or by sending a fax to 734-668-1957
More information on Notia is available from: Intergraph Corp., Huntsville, AL 35894-0001, Tel: 205-730-2000.
The U.S. offices of the Petrotechnical Open Software Corporation are at: 10777 Westheimer, Suite 275, Houston, TX 77042. Tel: 713-784-1880.