One of the recurring issues in software is the tension between using ‘best of breed’ applications from different suppliers and sticking with a single developer’s ‘platform’. But what happens when the developers of a whole range of software tools merge. Does this make for an interoperable platform? Or will the merger/acquisition just mean rebranding? One thing is for sure, it was smart of Aveva to buy PI, grab its massive client base and extend the use of its own refinery-focused simulators.
Speaking at the 2022 Aveva World in San Francisco, CEO Caspar Herzberg traced the three strands of what is now Aveva. These go back over 50 years with Schneider/Invensys SimSci (1967), Aveva CAD Centre (1967) and OSIsoft PI (1980). These are now ‘grouped around an industrial strategy of data sharing’, providing ‘end to end AI-infused engineering and operations solutions’, all delivered from the cloud via Aveva Connect and Data Hub. Herzberg’s flagship use cases were suitably environmentally conscious, coming from CCS, water and renewables. Herzberg stressed Aveva’s ‘close relationship’ with Schneider Electric. This is in fact an ongoing situation. The deal was only completed in January 2023 and, as widely reported, the plan is to ‘preserve Aveva’s business autonomy and future R&D investments’.
Andrew McCloskey drilled down into the ‘new frontiers’ of the industrial software landscape, presenting Aveva’s toolkit as underpinning a digital twin initiative. He suggested a maturity scale for the twin from ‘full DT’ to FDT+ (the unified operations center) and the ‘natural next extension’ that is (will be?) the industrial metaverse. McCloskey illustrated the use of a laser scan backpack from Navvis for data collection and creation of the 3D point cloud model, leveraging ‘infused AI’. Aveva’s toolset performs the heaving lifting, extracting contextualized information from multiple sources. ‘Vision AI’ is said to catch what sensors and the human eye might miss. An ‘early proof of concept’ of the industrial metaverse was said to show the ‘incremental, yet amazing value for collaborative engineering and remote operations’. The IMV is ‘not about the goggles’, it works on any device and is, or will be, ‘highly relevant’ for remote operations and collaborative engineering.
Bart-Jan Ruules presented ‘the world’s most comprehensive digital twin’, that Shell is creating for its refining operations. This leverages industry standards (Ruules cited ISO 15926, ISO 14224 and Cfihos) along with Shell’s own corporate standards for engineering design. All of which is being rolled into an asset data model, illustrated by a labyrinthine flow chart spanning design, build, operations, maintenance and on to decommissioning. The massive model is said to simplify the data lifecycle, supporting early (pre-Feed) data delivery requirements and drive ‘integrated end to end thinking’. Returning to the green theme, Shell has invited Aveva to participate in its Energy Transition Hub* as part of a ‘deeper collaboration’.
* ‘A new digital platform open to exchange emissions data, sharing best practices and exploring low-carbon solutions that work for you’.
Rob Kennedy gave a slightly different take on Wood’s digital twin to the one* we reported on in our last issue. Again ‘right-to-left thinking’ is needed to determine the strategic capital investments required to transform both project and operating performance. This has resulted in Wood’s ‘best-in-class’ digital twin ecosystem built with wall-to-wall Aveva components, notably with OSIsoft PI, now rebranded as ‘Aveva PI’. A data integration layer is built with Aveva Asset Information Management. Kennedy recommends budgeting for the digital twin in project costs and scheduling.
* This was built for Turkish Petroleum’s Sakarya Gas Field and leveraged technology from Hexagon. In a short email exchange, Kennedy told Oil IT Journal ‘While Wood has a number of digital twin solutions (such as our own Virtuoso) our main focus is on advisory and implementation services – where we remain technology independent and work with a range of suppliers and partners to deliver best fit solutions that deliver our clients’ desired outcomes’.
For a comparison of Aveva and Hexagon’s AIM tools see Gartner.
Japanese refiner Eneos has built its digital twin using internal systems and the Aveva Connect cloud. Here, a 3D point cloud model consolidates data from Informatica, J5 (maintenance), Excel and SAP. The prototype digital twin was built and evaluated for one unit in Eneos’ Kawasaki refinery. Aveva products were selected based on cost and features. The solution allows for equipment search by tag in Aveva Net Dashboard.coredownload.inline.pdf . Intelligent P&IDs are hyperlinks to documents, point cloud items and process data in PI Vision. A ‘visual reporting’ function allows for color-coded P&ID displays grouping equipment (e.g.) by operating temperature. Switching between PCM and AIM* allows tags and markups to be identified in the 3D model. The link to the J5 maintenance database from Hexagon allows for maintenance work planning. Work areas can be plotted on a map, again with color coding to show work types and ‘keep out’ areas. The prototype digital twin has been verified with various facilities management scenarios and deemed ‘very effective in improving the efficiency and quality of operations’. For analytical work that deals with huge amounts of information, integration with big data/AI is required. This has been used to gain new insights into corrosion prediction from point cloud data. Other digital technologies such as OCR and robotic process automation accelerate digitization and reduce costs by adding attribute information to documents and 3D viewers.
* Point cloud manager and asset information manager.
Erin Jones (ExxonMobil) teamed with Peter Townson (IOGP/Cfihos) to present a practical implementation of Cfihos for data surveillance. The aim of the exercise is to populate RED, ExxonMobil’s implementation of Aveva AIM and ISM. RED* promises validated data-centric engineering surveillance, improve data handover to operations, and an engineering data foundation for global asset analytics. RED seeks to expose data that may be siloed in an EPC’s system. Data now passes through Aveva and on into RED for checking and cleansing. A minimum viable product was rolled out in2022. Data now goes on to maintenance as an ‘O&M engineering data system of record’. RED is said to be a highly-searchable and integral data- centric element for project information and handover, contrasting with earlier document-centric project delivery.
* More on ExxonMobil’s use of Cfihos alongside RED in our last issue.
Trey Lowe, CTO, outlined Devon’s data-driven production operations. Devon has thousands of sites across the US, with over 10,000 tanks. An enterprise-wide data and digital investment has seen the deployment of AI, RPA and a move to the cloud. In 2020, Devon’s SWPM* hydraulic fracturing technique was granted a patent. Devon monitors over 11 million data points in PI Vision. Again, the environmental side of the business was highlighted. Devon is using its comprehensive monitoring system to reducing Scope 1 & 2 emissions and has joined the Oil & Gas Methane Partnership. A 65% reduction in methane emissions is planned for 2030. Along with the emission reduction, PI Vision has driven operational efficiencies. Image analysis turns cameras into sensors that capture data in hazardous areas, augmenting existing sensors and performing tasks such as checking pump rod rotation. PI AF and PI Vision are the foundation for advanced analytics and computer vision applications. Exception-based surveillance has accelerated time to value for advanced analytics and machine learning projects.
* Sealed (offset) wellbore pressure monitoring.
Kenneth Guidry showed how BP is extending its extensive PI System infrastructure to facilitate its ‘journey to net zero’. BP has almost 4 million PI tags running on 60 PI servers across its four business areas. Modeling of CO₂ emissions in refinery operations uses Aveva Unified Supply Chain to drive emission reduction. Adding the CO₂ modeling capability resulted from a strong ‘co-innovation partnership’ that has helped understand the options that could reduce CO₂ refinery emissions. The USC package includes flow sheets of a linear programming optimizer. Guidry also presented BP’s digital twin developed for the Azeri Central East (ACE) development. Here the Aveva AIM digital twin collects plant data for viewing in PI Vision. The twin blends data from laser scans, tags data, 3D Model, documents, SAP, and Bentley APM. More on ACE here with a cool time lapse video of construction.
Prabhas Bhat and Matthew Richardson teamed to present Plains’ experience with the Aveva portfolio. Plains software ‘journey’ spans Valmet, Metso, Telvent, Schneider Electric, OSISoft and Aveva – ‘first as customer, now as strategic partner’. Plains has replaced a diverse, tag and Excel-based system with an ‘enterprise class’ solution based on PI and Aveva’s Enterprise Pipeline Management System. This is now ‘transforming the way we work’. A flagship development, Plains leak detection suite, was built using the PI SDK/API, working on data sent to PI from leak detection systems. This has captured ‘synergy value’ between PI and the EPLMS. Development was led by a operations-led PI System community of practice which has redefined the IT/OT relationship. While there is a role for traditional IT, ‘OT needs to own the PI System’.
Sheri Williams explained how Accenture can help with Aveva-based digital twin development. Williams reported that digital maturity is poor across all industries* with many firms in ‘pilot paralysis’. Williams advocates achieving greater maturity by deploying ‘bionics’, ‘cobots’, real time multilingual translation, the cloud, digital threads and twins. And, naturellement, the Aveva data hub and metaverse. A key player on the digital shop-floor is the robotics integrator, a subject matter expert on automation and robotics.
* However, unlike previous analyses of digital progress, in Accenture’s current evaluation, oil and gas is actually leading a rather underperforming field!
Aveva’s Norton Green invited visitors to visit the innovation labs and ‘take a walk in the Aveva Industrial Metaverse’ in the form of a Youtube video. Here, a robotic avatar gesticulates at a plant model (at one point it seems to be counting on its fingers!). There is no audio so its hard to see exactly what is intended. Underwhelming!
Chris Kahrs presented a joint venture between Microsoft and Aveva to apply deep learning to control a distillation unit autonomously. This is presumably the ‘AI-infused’ aspect of the portfolio that Caspar Herzberg referred to in his keynote. The demonstrator applies Microsoft’s Bonsai low code AI demonstrator alongside Aveva DynSim to optimize operations of a digital replica of the plant in the cloud. The results appear quite convincing, but not perhaps ready to ‘infuse’ the whole portfolio!
Other presentations of note include Saudi Aramco’s development of a refinery-wide closed-loop real-time optimization, built with Aveva Process Optimization software on a PI System infrastructure. Another Aramco presentation covered a reliability and integrity management solution developed at its 4IRC center, leveraging PI data and Aveva Asset Predictive Analytics. For this ‘fourth industrial revolution’ development, Saudi Aramco evaluated the alternative of developing in-house applications. This was rejected due to its development and maintenance complexity. KBR presented its ‘complete O&M digital twin’ as built for the BP ACE development (see above) presented as a 3 way joint development by BP, Aveva and KBR.
View the Aveva World presentations here and visit the Aveva World minisite.
© Oil IT Journal - all rights reserved.