An Operations and Maintenance Information Open System Alliance

Problems With Information Handover

Digital plant information is now seen as valuable an asset as the physical plant itself. Accurate capital project information handover is vital to enterprise profitability. Following handover, plant information must be kept accurate and accessible for the lifetime of the facility, which typically spans 30-50 years. Inaccurate or missing plant information causes lost revenue due to missed project deadlines, unnecessary material purchases, design rework, start-up delays, production cutbacks, and unplanned shutdowns. This business use case focuses on the periodic information handovers of information from capital project systems for use in plant operations and maintenance (O&M) systems. This handover activity is vital to bringing a new plant on line within time and budget as it drives the compilation and validation of the records necessary to pass into and support the operational life of the plant, together with their transfer of ownership from the project team and acceptance by the plant owner.

Essential to quickly bringing a plant into production and enabling plant O&M personnel to safely operate the plant, is populating O&M systems with essential information about the as-designed engineering structure of a plant with the required operating parameters, the as-procured product data, and the as-built serialized assets. The specific information needs of O/O, EPC, and OEM personnel are quite different. The O/O is concerned about the long-term needs for the life of the facility, the EPC focuses on the capital project design and construction phase, and the OEM is concerned with the requirements and delivery of products which they either Make-to-Stock or Make-to-Order.

Studies by FIATECH, estimate that the traditional execution by an EPC of handover is typically quantified as less than 0.3% of the project, but that a further 2-4% of the project cost is required by the O/O to manually correct and key in required O&M information into O&M systems. This means that handover traditionally only equates to $1M on a $400M project for the EPC, but the O/O then has to spend an additional $8-16M in "hidden" data entry and validation costs. Fiatech studies have shown that periodic, structured, non-proprietary, automated information exchanges from EPC systems to O&M systems are estimated to save 60% of this O/O hidden cost, resulting in a savings of between $5-10M on a $400 million project. To maximize these savings, there are four major problems which must be addressed.

Problem 1: Big Bang Approach to Information Handover

The first barrier to overcome is the big bang approach to information handover from the EPC to O&M personnel. O/Os often experience unexpected delays in startup if information is not periodically assembled, delivered, and reviewed during the project. Often, capital projects wait until startup/handover to dump the design and information to O&M personnel to populate their systems. Since this occurs at the end of the project, cost pressures can be extreme due to earlier problems. As a result, it can be tempting to cut handover spending to hit the project budget. It is not uncommon for an O&M organization to have the burden of finding and validating data to populate their systems after startup.

Recommendation Both EPCs and O/Os should follow the periodic, phased handover methodology captured in NIST's Capital Facilities Information Handover Guide (CFIHG) to improve the efficiency and quality of information handovers throughout the capital facility life-cycle. The CFIHG provides a framework for the definition and delivery of information packages to be transferred among participants in capital facility projects. Information handover strategies shall be based on identifying the information created in each phase that will be needed downstream and how it will need to be used. It is essential that the facility life-cycle information strategy and the handover requirements be established before project initiation so that contractual requirements for a continuous information handover can be defined.

Problem 2: Unstructured Data Exchange Formats

The second problem is the use of unstructured data exchange formats for information transfer from engineering to O&M personnel. Traditionally, most plant engineering information is available to O&M solely in unstructured document-oriented formats that cannot be readily machine-interpreted. Examples of these unstructured formats include PDF, JPEG, TIF, Microsoft Excel and Microsoft Word. While these formats have value for the human reader, they are of limited value to populate O&M systems since it is difficult for computer software to extract information such as the tag identifier from a P&ID diagram represented in a PDF format. Since manual data entry from these unstructured documents is now required, data integrity problems are common between engineering and O&M systems, which can result in operating mistakes. Another resulting problem is the existence of redundant equipment files, created by both maintenance and engineering, which are not synchronized, creating recurring engineering and maintenance errors. This can result in unplanned shutdowns, delayed maintenance projects, and incidents.

Recommendation Plant information should be generated as structured, computer-readable, and O&M-meaningful that can be linked to derivative unstructured document. Plant structural information should be generated by engineering systems from logical P&IDs and PFDs, and placed into computer system readable formats such as XML or RDF. The extracted structural information packages should be defined as specified in the CFIHG handover plan section and should include the equipment/tag identifiers with requirements such as operating envelope setpoints, I/O port connections with process flow information, as-built serialized asset data, and associated plant breakdown structures.

Problem 3: Proprietary Data Exchange Formats

The third problem is the proprietary format of data exchange handover information. As specified in CFIHG, handover requirements - content description and exchange format - should be defined in the contract between EPCs and O/Os. Even though structured formats are utilized by an EPC, unless the information is originally created with knowledge of the final desired handover format, it may be difficult and expensive to convert. A study conducted by the Construction Industry Institute (CII) in the United States in the early 1990s suggested that the effort to convert facility drawings developed manually or in an unstructured CAD format to a structured model was ineffective in controlling construction costs or schedule, while the use of information-rich models during design did result in such benefits.

Recommendation Specify that outputs from EPC systems shall be formatted accordingly to ISO 15926 and OpenO&M-standardized formats using ISO 15926 and PCA/JORD Reference Data. ISO 15926 employs a generic data model that is supplemented with OpenO&M templates and a Reference Data Library (RDL) to support standardized transfers of information packages throughout the complete life cycle of a facility. Specifying only ISO 15926 and OpenO&M-standardized formats provides a standardized definition of data exchange packages for accessing and presenting model data in any way the end user requires.

To be consumed by O&M execution environment systems, the reference data transferred in OpenO&M ISO 15926 RDL templates need to be converted to the standard for registry data - the MIMOSA Common Conceptual Object Model (CCOM) XML format.

Note The Joint MIMOSA/PCA O&M Special Interest Group is working to properly incorporate MIMOSA CCOM concepts in ISO 15926 so that key O&M information can be unambiguously exchanged based on existing O&M information models.

Problem 4: Manual Data Exchange Methods from Engineering and Construction Systems to O&M Systems

The final problem to address is the manual data exchange methods from Engineering and Construction Systems to O&M systems. Plant engineering and construction systems that contain updated Plant Breakdown Structures, P&ID and PFD as-designed and as-built data need to periodically publish revisions to O&M onto a safe and secure information bus using a Service-Oriented Architecture (SOA) methodology and subscribe to revisions to the as-maintained state of equipment from this same information bus.

Recommendation All systems should support the OpenO&M Information Service Bus Model (ws-ISBM) SOA architecture, which can be implemented by major Enterprise Service Bus (ESB) suppliers to provide systemic notifications of information changes in a safe and secure transport with guaranteed delivery.

Recommendation Include the use of an OpenO&M-ISO 15926 Transform Engine on an ISBM, transforming ISO 15926 content received by the Active O&M Registry from Engineering to MIMOSA CCOM XML format with the use of the OpenO&M Web Service Common Interoperability Registry (ws-CIR) for index lookups and returning this data back to the Active O&M Registry system via the ws-ISBM.

Recommendation The OpenO&M best practice also includes the use of an Active O&M Registry which subscribes to the output from the Transform Engine and provides a centralized GUID registry for all O&M data dictionary items (REG-DICTIONARY), data dictionary taxonomies (REG-TAXONOMY), system structures (REG-LOCATION), serialized assets (REG-ASSET), and product templates with product models (REG-PRODUCT).