Effectively controlling the drum level in an industrial boiler helps to maximize overall energy efficiency. When the drum level varies it can not only impact boiler efficiency, but overall reliability due to unplanned shutdowns.

Emerson’s Mark Coughran shared a story with me about a major North American university that was having difficulty with one of the boilers that supplies steam to the campus.

The drum level on this boiler has alarms at +4 and -4 inches of water. It will shut off the fuel on -6 inches of water. The preferred setpoint (SP) for efficiency is 0 inches of water.

However, due to a large cycle on the drum level process variable (PV), the operators preferred to raise the setpoint and live with frequent high PV alarms.

Mark used the EnTech Toolkit, which showed that the boiler level process had an unusually sluggish response to changes in feedwater flow. The sluggish process response is not amenable to auto-tuning algorithms, trial-and-error, or shortcut tuning methods.

He noted that it is also difficult to control with a conventional proportional+integral (PI) controller; however, using derivative action (PID) can be tricky and is avoided by most people.

For these types of control issues, Mark and the Control Performance consultants recommend Lambda tuning. This tuning method provides a direct and systematic way to solve this type of problem.

Installing the Lambda tuning PID parameters immediately stopped the cycle. The operators were very pleased and have already begun experimenting with lower setpoints. The as-left data includes some firing rate changes, which are load disturbances to the drum level.

If you’re unfamiliar with Lambda tuning, see a paper Mark wrote, Lambda Tuning — the Universal Method for PID Controllers in Process Control.

You can also connect and interact with other variability and control performance experts in the DeltaV and Ovation tracks of the Emerson Exchange 365 community.