Capital Facilities Information Handover Standard

USPI-backed standards group face-to-face meet chez BP. Shell on handover and mounting tag counts. Chevron on ‘over ambitious’ prior art. Total’s Quantum digital twin. AmecFW’s information hub. Yokogawa on IOGP standards. USPI on ISO accreditation. Shell on broadening CFIHOS’ scope.

The handover of data and information on the ‘as-built’ state of facilities such as offshore platforms, FPSOs or refineries represents something of a holy grail for the construction industry. While parts of the constellation of technologies that is ISO 15926 have been tried by operators and software vendors, a ‘practical’ implementation that can be used by engineering prime contractors (EPC) to hand-over to owner-operators (OO) has proved elusive. The developers of Cfihos (pronounced ‘see-foss’), the capital facilities information handover standard, have set out to simplify the process, abandoning ISO 15926’s semantic web technology in favor of the exchange of a limited amount of ‘must-have’ information via standardized Microsoft Excel templates. Cfihos is owned by the USPI standards body and has backing from Chevron, Shell and BP which hosted this face-to-face meeting earlier this year.

Paul van Exel, USPI chair, presented Cfihos as a handover specification (not just a guide) of what OOs want in terms of the documents and information required to populate their information systems. Cfihos will deliver a process industry standard document with rules and principles for information handover. The specification will point to the ISO 15926 reference to data library and act as a common language for all stakeholders. Cfihos has been available to members since 2015.

Jason Roberts (Shell) provided the background to the initiative. Labor productivity is flat to downhill in oil and gas (it is rising in other verticals). Time spent on engineering design increases steadily while the oil price crash has impacted project viability. On safety, Roberts cited the Marsh study of the 100 largest losses in the hydrocarbon industry (1972-2009) which initially led Shell to think it was doing a good job. A refinery fire with fatalities was a wake-up call when it emerged that unclear ownership meant that some lines were not in anyone’s corrosion inspection systems. Elsewhere untagged valves had corroded – they too were not in the maintenance system. In answer, Shell developed EIS, an internal engineering information specification that was a forerunner of Cfihos. Feeding EIS meant addressing the problem of information handover. Supplier data comes in different formats and tools don’t talk to each other. Despite some contracts that specify ‘handover information in ISO 15926,’ nobody knows what that actually means! Meanwhile data volumes grow. Kashagan has around a million tags and as many engineering documents, even though the original design was for around 15,000!Cfihos means that OOs can tell EPCs what is required to be handed-over. Operators need to be able to tell their procurement people ‘don’t go for the cheapest contractor because their IM is a mess!’ Shell’s push for a standard was inspired by the 2010 IOGP position paper on the development and use of international standards. Roberts observed that there are big gaps in ISO 15026. But one solution is not to ‘put everything in a standard.’ It’s better to check if there are existing standards (eg ISO 13706 IS for heat exchangers) that can be used. Cfihos goes beyond data centricity, adding rules for data context, values and ranges. There is also a standard taxonomy for documents and metadata, the ‘initial discipline document type.’

Josh Vincent (Chevron) described the Cfihos philosophy as follows, ‘we build the same types of assets and buy the same types of equipment, so there is no reason we should have different ways of describing them.’ Of course, this has been tried before, notably with the Posc Caesar Association’s ISO 15926 standard, a reference data library of ‘all of the engineering terms and objects required to design, build and operate an engineering facility.’ But Part 4’s scope has maybe proved a bit ambitious. Cfihos’ scope is to be limited to around 600 items that will be needed for handover. Such scope reduction is possible since it is only necessary to describe what is in place at handover as opposed to what might be required in the design phase. The spec will also make for a more consistent way for owner operators to ask suppliers for information.

Paul Charles presented Total’s ‘Quantum’ Virtual Plant initiative, a new digital twin approach that shares the same objectives as Cfihos. Charles believes that Cfihos could need some rationalization and an improved data model. Total is prepared to share its RDL work with Cfihos to maximize alignment between the two RDLs. Combining Total’s RDL (and Exxon’s) will likely need to wait on Cfihos 1.3.

Charles Samkin presented Amec/Foster Wheeler’s (AFW, now a Wood Group company) asset information hub (AIH), another ‘digital twin’ of the physical asset. The AIH is hosted by Amec using the Aveva Net platform said to be ‘based on’ Cfihos and integrated with Amec’s software portfolio. Aveva Net serves as a single source of the truth for other compliance and integrity applications deployed at Amec. All data, documents and models share the same platform. The solution was co-developed with BP and is now used by Amec world-wide and shared with customers. Today’s handover specs are many and varied. Cfihos is expected to ‘fix the historical mess.’

Andy Davidson took over to explain how AFW has used Aveva’s Information Standards Manager (ISM) to evaluate Cfihos V1.2. ISM is a ‘powerful tool’ for managing different standards and for communicating with clients. A grid of ‘permissible attributes’ locks down what information can be exchanged, using an ‘extended version’ of Cfihos. ISM is a gateway to 2D/3D schematics, piping and instrumentation diagrams (P&ID) and other Aveva tools. ISM does data validation for attributes including units of measure. There is also a gateway to Bentley’s ProjectWise construction information management system. The hub has an interface for inputting data from the Cfihos Excel spreadsheets but, as Davidson opined, ‘Whenever I see a stack of Excel spreadsheets there is always the chance of getting something wrong.’ Such reticence over the data format that has been selected by Cfihos was shared with others in the software community.

Elbert van der Bijl presented Yokogawa’s approach to information exchange on automation projects. Yokogawa’s measurement systems for field instrumentation and production control must communicate with third party systems such as OSIsoft PI and SAP. Information exchange is usually Excel-based or on PDFs which are ‘still unfortunately the state of play.’ There is often a lot of missing information and a tendency to ‘focus on capex over opex.’ The instrumentation business also comes with its own standards, ‘Ecl@ss and BMECat. Control systems vendors deliver in their own software tools. Yokogawa has its Centum VP design suite, ADS master database and FieldMate validator. The objective being smooth commissioning, field acceptance testing and startup. Recently OOs have asked Yokogawa to ‘take care of the information lifecycle.’ This is a ‘free competitive area where all are looking to help the OO.’ For Yokogawa, standardization is driven via the IOGP’s standards subcommittee, by ISO/IEC and, in Norway, by EPIM with its STI data sheet standard.

Elsewhere, Namur’s NE150 standard for e-data exchange has been used to exchange data with Siemens’ toolset. Most of this is happening in the chemicals business where a different set of standards are being developed to do much the same things as Cfihos, making life harder for companies that work in both verticals! van der Bijl expressed some polite frustration with the fact that contractors build and model ‘logically’ and then ‘someone says hand over in Excel! What do you do?’ Then there is the issue of modern instrumentation that produces big data – maybe 500 parameters for a single flow meter (maybe too much for an OO!) There is also a lot going on in the IIoT space, notably the joint OpenGroup/Exxon Mobil standard for control systems. A ‘very aggressive approach designed to put us out of business!’

Paul van Exel (USPI) described the complex relationships between the ISO TC184/SC4 committee, Posc/Caesar and ISO 15926. The units of measure update took five years and is still not published. There are issues with different conveners, time and a lack of dedicated resources. Then there is the thorny question of who owns the standard. van Exel recommended that members make representations to ISO TC184 about their desire to see a properly maintained Cfihos standard. Progress on Cfihos is by and large on track. There is now a need to delivered quick wins and documentation, perhaps with a ‘Cfihos for Dummies’ style publication. The activity has grown to the extent that a part time project manager is required. Release 1.4 is slotted for delivery in October.

Anders Thostrup (Shell Global Solutions) led a discussion on widening Cfihos and on making it easier to understand and deploy. Scope could extend to plant and process, spare parts, document content as data (hard to do), materials management, full design data, datasheet content, 2D/3D ‘maybe for our children!’ Thostrup recently visited Sakhalin where a successful data handover leveraged an earlier version of Cfihos-type technology. ‘All stand to benefit from quality information.’

Vic Samuel (Chevron and IOGP) sees opportunity in the downturn for a renewed push for standards. Information management standards can ‘reduce exposure to legal, safety and competitive risks.’ Cfihos can be seen a fitting with the UK Government’s mandate for building information management aka ‘BIM.’

USPI is working on a memorandum of understanding for the use of Datum360’s CLS360 engineering information management tool for use in the project. Others in the software vendor community, Aveva, Phusion, Intergraph are also tracking the project closely.

Comment - Previous attempts in this space (PCA/Fiatech) have not been successful. Cfihos’ approach of a reduced information set looks promising. The use of Excel (actually CSV files) is a double edged sword. Easily understood by the engineers, but less than state of the art IT-wise. In one sample data set we spotted a cell containing ‘6 DEG C’ for a temperature value. Clear, but awkward for data ingestion! The geeks see JSON as the way forward. But the geeks have form… remember RDF?

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.