The benefits we experienced in commissioning and starting up the first plant were repeated when the second unit was brought into operation in late 2002. Faster startups allowed us to deliver the chemical intermediates for Teflon production earlier than expected.
Practice predictive maintenance
AMS Device Manager provides information about the operating condition of the field devices and associated equipment, which is the basis for predictive maintenance now that the plants are operating normally. Although there isn’t a great need for predictive maintenance of the instrumentation in a new plant, problems can develop, especially with valves, and online monitoring by the software conveys a valuable “heads up” if the condition of a device begins to deteriorate. This early warning enables maintenance technicians to determine how long the device is expected to last and when it should be replaced, considering its importance.
Advance knowledge of a device’s condition is especially useful at our plant; chemical hazards can make it difficult to get to an instrument or to break into a line while the plant is operating. For such procedures, it’s necessary for all personnel to be in protective gear, which can be cumbersome and uncomfortable to work in. We can now try to pinpoint the cause of a problem before maintenance personnel suit up to go into the plant.
In the past, it was standard operating procedure to go out, pull a valve and take it back to the shop for repair if it didn’t seem to be performing properly. Frequently, no problem was found and the valve was reinstalled. If the problem persisted, the valve might be pulled a second time. This doesn’t happen anymore.
Now, when a valve appears to be malfunctioning, the first thing we do is compare the valve’s current performance with its signature (a baseline taken when the valve was new) to determine if and how the basic characteristics have changed and whether that change is affecting the overall operation (Figure2). Is it slowing down the process? Is it affecting the quality of the product? The answers to these questions can help determine whether any action needs to be taken. If there is nothing wrong with the valve, the technicians look elsewhere. Most importantly, maintenance decision-making now is based on factual information and not supposition.
It’s not necessary to be in the control room to communicate with these devices. Recently, I was in Maryland on Thanksgiving when an instrument in the newer plant failed. It was replaced with a spare transmitter that had a slightly different configuration. Although the plant was not fully operational, it was scheduled to run over the weekend, so they called me. Using my personal laptop computer, I was able to connect with the instrument via our corporate network, scale it according to the stored data on the device it replaced, and put it in service – all from more than 500 miles away and in less than 15 minutes.
Without question, this versatile software has earned a place in the arsenal of weapons we have to eliminate processing problems and ensure smooth, efficient production without unnecessary delays. It has saved us time and maintenance dollars while permitting increased plant uptime.
Warren Way is a control systems engineer for E.I. DuPont deNemours Co., based in Fayetteville, N.C. He is a mechanical engineer with 35 years of experience at DuPont. E-mail him at email@example.com.