Replacing Your DCS for the Very Last Time

Recently I gave a speech to a European group about ExxonMobil’s initiative to replace many of its existing process automation system with a new kind of automation system. Many of the units in ExxonMobil’s downstream and chemical operations are automated with quite old (30+ years old) DCS technology. ExxonMobil realizes that replacing these systems will be a huge, complex, and (yes) expensive undertaking. One aspect of ExxonMobil’s objectives was to make sure that “never again” would they face this type of situation. They want to ensure that going forward their automation systems could be replaced or augmented incrementally with much less system-level engineering than is now required. They want to simplify the task of automation system upgrades to the point where it is no longer that big a deal. Today’s automation systems usually require a high degree of system-level analysis and planning in order to perform software and/or hardware updates. There are 2 major reasons for this. First is that each Distributed Control System (DCS) controller manages dozens or even hundreds of process measurements and control loops. If the controller is to be shut down, even for a few seconds, process engineers need to evaluate the possible impact on operations. The number of measurements, calculations, and (especially) actuators driven by a single controller makes this analysis more difficult. The second major reason updates are difficult is concern about interoperability. If you are operating a plant with DCS hardware or software that is only partially updated, you have to understand if doing that will cause any interoperability issues between the “new” and “old” system technologies being employed. Newer DCS products have improved their ability to inter-operate this way, but often older DCS products require that the entire DCS be operating at the same revision level. This can mean that a complete plant shutdown is the only time when the DCS can be updated. In some plants this type of shutdown/turnaround/outage occurs only every 3-4 years. How is ExxonMobil addressing these 2 challenges? For the controllers, they are planning to shrink the scope of controllers back even to control of a single loop. If you have a controller with a single measurement and a single output to a field device, a competent process engineer should be able to assess the impact of a shutdown without difficulty. For interoperability, ExxonMobil is planning to use a very mature (and publicly defined) software technology that has proven to have the required capabilities in many existing installations. There is much more to say about this topic. So I have created an ARC report based on the content of my European speech. If you are a manufacturing end user and would like to receive a complimentary copy of the report, please send your full contact information to hforbes@arcweb.com and request the report entitled “ExxonMobil’s Quest for the Future of Process Automation”.