I’m recovering from travel delays and lots of writing and soccer. Here is a final user story from the Profibus International North America meeting from last week.
Following up two user presentations from day one of the Profibus International meeting in Phoenix, Jim Simmons of DuPont presented one of the best, down-to-earth implementation discussions I’ve heard in my 14 years as an editor. And, despite this being a Profibus conference, he was frank in his evaluations and wound up with a mixed fieldbus system.
He was doing controls at DuPont when the company wanted to build a facility at its Fayetteville works to produce PVF polymer in 2005. The project came in $5 million under budget and 2.5 months ahead of schedule. That, in itself, is a miracle. Prior to him, someone had chosen Siemens PCS7 as the preferred DCS system, so he stayed with that.
Simmons had little knowledge of fieldbus networks at the time, but he decided to investigate them to see if they would help his controls implementation. He did a thorough analysis of potential networks. “There are many,” he deadpanned. His analysis began with looking at categories of devices–instrumentation, automated block valves, analytical devices, motors/drives and remote I/O. He then looked at the requirements for each category.
The result was Profibus PA for instrumentation (it worked well with the PCS7), AS-i for the block valves, Profibus DP or Modbus for most of the rest. The company standard for variable frequency drives was Rockwell Automation, and those devices talk DeviceNet. So he used an Anybus connector from HMS to connect Profibus DP to DeviceNet.
Some of the benefits DuPont realized included advantages of initial installed cost, plant flexibility, reduced time for project execution and time for ongoing plant changes.
Simmons concluded that there was not one perfect fieldbus for every possible application, so engineers should select wisely per design requirement. However, engineers should have to justify NOT using fieldbus on a project.