E-mailed note on Fuel Gas Pressure Control received from Stuart of BP Whiting, copied below with Stuart’s and the refinery’s permission

ProControl had earlier analyzed Whiting’s pressure-header test data (data from two pressure transmitters located a long ways apart) using DISCOVER, and had sent pressure control tuning recommendations to Stuart (a Revelations Course control engineer graduate) for each transmitter. The furthest-away transmitter exhibited many additional seconds of deadtime. We were told that this plant’s existing (in-house) tuning program yielded overly-aggressive, cycling, response for the closer of the two transmitters. Stuart’s note follows:

Bob,

Our refinery fuel gas header pressure controller has been using the tuning constants you provided from Discover for a few months now and they are working great. Discover provided vastly different tuning for the two different transmitters that this controller can use for the PV. Because of this, I have implemented adaptive tuning that loads the controller with the proper tuning constants depending on which transmitter it is using. We control from the transmitter with the better dynamics whenever we can, but we still have good control when we must use the other transmitter.

The system set point is 48 PSI and with the old tuning the typical variation of the pressure was +- 1 PSI, with large disturbances causing deviations of +- 2 PSI or more. Now the pressure is controlled to well within +- 1/4 PSI even for the peak disturbances when using the better transmitter and to +- 1/3 PSI when on the other transmitter. The much better pressure control has improved the control of furnaces all over the refinery and we no longer have to put furnace controls in manual because of pressure swings in the fuel gas system.

Discover's tuning recommendations are right on. Thanks.

Stuart

Process Control Engineer

BP Whiting Refinery