Process Knowledge and Open Process Automation

By Harry Forbes

Industry Trends

Everyone knows that the prospect of a new and more open process automation system will be “disruptive”, but if you ask people to drill down and explain exactly what they think the disruptive effects will be, it gets harder to nail this down. There are many unknown variables here, and it’s important for BOTH end users and suppliers to get a handle on them. In a recent report for ARC Advisory Service subscribers I talked about a few of the possible effects on DCS supplier business. But what about the impact on end users? Here is one possibility: much more separation between “process experts” and “automation experts”. Let me explain this a bit. Improving the performance of a process plant is a hugely important task. In process manufacturing that task is “where the money is”. Plants have pretty much no control over the costs of their feedstocks or the prices of their products. What they can do is choose feedstocks wisely and operate their processes optimally so that they add the most value through the plant’s processing. Note that these objectives have nothing to do with how much or how well the process is automated. If the optimal production plan is simple and if the plant is rock-solid stable, then a plant could run in manual and achieve optimum value. Automation is simply a means to the end -- and that end is stable and optimum plant production. Aspen HYSYS Column Analysis Likewise, to achieve this level of production, the required knowledge is really about the plant process rather than the automation. The important knowledge concerns energy and material balances, feedstock and product properties, thermodynamics, equipment properties, and production limitations. This is what is often called “deep process knowledge”. In my plant experience deep process knowledge is usually a very scarce resource, but a very valuable one. The rub is that the process operating points are set by the automation system, and the stability of the process at any operating point also depends on the automation system. This can make life difficult for a process expert who does not know all the details of DCS capability and how the automation has been implemented in a particular DCS. “Deep DCS knowledge” is also required to optimize plant operations, and that knowledge is highly vendor-specific. I argue that it is a bad thing for an automation system to create a barrier between a deep process expert and the plant. If an open process automation system can give a process expert more insight or (especially) more freedom to improve operations then it will be a big plus for its adopters. Furthermore, if the barrier is lower, than a wide range of experts, from in-house, OEMs, consultants, and others will be able to contribute to improved operations.

Engage with ARC Advisory Group