At the OSIsoft user conference in Paris last month we sat in on what was billed as the PI System 101 hoping to learn and report on the essentials of OSIsoft’s technology. Unfortunately this turned out to be less of a technology 101 and more of reading of the OSIsoft marketing material. So we start with our own idea of what could be a PI 101. The PI System, OSIsoft’s flagship, is a data historian. That is what a database is called in the process world. PI records data from equipment ‘tags’ i.e. streams of real-time information on temperatures, flow rates and so on from valves, rotating equipment and other assets. PI understands the plethoric data formats used in a plant, recording data as ‘time series,’ i.e. as a sequence of values and time stamps. PI does not care or know if an equipment item is functioning correctly (or at all). Moreover it can only record what is transmitted. While older kit may produce less than perfectly formed data, newer stuff may provide more context and fancier data formats. Making sense of what is captured to the historian will likely require a lot more attention and a constellation of software from OSIsoft and third parties has evolved to turn the raw data into a more usable form.
A good real-world example of PI usage was presented by Sam Scott (Talisman/Sinopec Energy). Talisman’s average asset age is 29 years. Talisman started using PI System in 2001 with a 1,000 tag subsea system. Today it has 250k tags, 50 interfaces and 50 concurrent users. PI System provides a single source of offshore data and feeds hydrocarbon accounting, laboratory systems and corrosion monitoring. Talisman’s aging assets make for a challenge in maintaining safety and production critical equipment. To combat falling production, Talisman has instigated a ‘rotating equipment excellence program’ (REEP) to improve equipment integrity and reliability. REEP covers performance monitoring, spares management, competency, contracts and audits. REEP contractor Ian Gore (CSE Controls) explained that older equipment is either poorly instrumented or unconnected, displays developed over years tend to have inconsistent interfaces and third party packages tend to focus ‘on a single piece of kit or use case.’ Enter Talisman’s ‘Spotlight’ REEP development that presents all equipment data in the same place and with the same look and feel. Spotlight calculates performance KPIs and alarm thresholds, notifying stakeholders by email. The result is better on-offshore collaboration, early detection of performance and integrity problems and a platform for condition-based maintenance. Currently REEP monitors nearly 3000 critical rotating equipment items globally, using PI AF, PI ACE, PI Performance Equations, and PI Notifications . A new Citrix environment replicates Process Book screens onto workers’ iPads.
Béla Kelemen (MOL Group) provided an entertaining exposé on the state of the industry and of ‘getting more from less’ in the face of industrial challenges in the EU. These include global competition, EU regulations and local traditions. Kelemen sees the EU as a ‘retiring old lady.’ The US, with its shale oil production, is a ‘returning rock star.’ And Asia is the hungry youngster following ‘10 years of 10% growth.’ The EU is working towards a 20% CO2 reduction, 20% more energy efficiency and 20% from renewables. ‘Who would want to be in the downstream?’
Downstream, which used to be one of the richest industries, has developed bad habits and failed to see the change that was coming. MOL is fighting back with a program to return to its ‘best in class’ status. PI is to be the backbone of MOL Group’s refining and marketing infrastructure.
Tibor Komroczki drilled down further into MOL’s usage of PI which has evolved from tactical to strategic. MOL has built a PI-based common data model representing its asset hierarchy in time and space. MOL’s new downstream program addresses the difficult economic climate described by Kelemen. Third party applications such as Sigmafine, Semafor for KPI management and Opralog’s E-Logbook are all integrated through PI. While there is less use of PI in process automation, MOL is working to fix this by blending information and automation, again by better leverage of PI. A new energy use dashboard produced a surprise, ‘the first month was completely red!’
Columbia Pipeline’s Dave Johnson showed how high availability PI data feeds its own brand CPG Enterprise Analytics. Columbia is busy now adapting its distribution networks to the US non-conventional boom. Johnson prepared a video demo showing how the system could give early warning of a failing compressor bearing (a common issue at the time). The cost of a timely replacement was much less than that of a random failure. Columbia has also deployed Transpara’s Visual KPI to its mobile devices and is in the process of rolling out iPads, these are ‘really catching on.’
Marco Piantanida and Christina Bottani (ENI) also showed how PI (and SmartSignal) helped in predictive maintenance. ENI’s onshore production from the Vald’Agri field has wells scattered around mountains and valleys. Long pipelines cause irregular flow with slugs of gas and oil causing varying water cut, liquid carry over and multiple equipment problems. Maintenance to date has been at fixed intervals but ENI is moving to a condition-based paradigm. Again PI is the key and has been interfaced with everything including 1991 DCS. SmartSignal catches most faults but additional investigations with Coresight and ProcessBook have provided warning of some imminent failures.
Reinaldo Jimenez described the PI System as Repsol’s single source of real-time data in its new production accounting system. PI is also enhancing the company’s maintenance work order management through real-time integration across multiple corporate systems. Repsol has also leveraged PI to manage bypass request approvals and renewals.
Lars Anton Mygland and Astri Hinna Fjermeros outlined how Statoil (with help from Amitec) has used OSIsoft technology extensively in its NoxTool emissions reporting tool. NoxTool was built around an existing PI system and uses PI web parts to provide a compelling view of Statoil’s assets. The authors observed that it is better to extend an existing PI system rather than build one from scratch.
A complex system of penalties and bonuses drives French utility TIGF’s gas storage business as Christophe Cuyala described. In its first year of operations TIGF incurred a €500k penalty. A new PI System, along with PI ACE (advanced calculation engine), UFL (universal file and stream data) and vCampus for third party data access means that TGIF is now getting bonuses for its reporting. More from OSIsoft.
© Oil IT Journal - all rights reserved.