Emerson Exchange 365

How to Solve Level Control Loop Problems

Mark Coughran at Emerson Exchange 2017Controlling levels in a manufacturing or production process can be challenging in some applications such as distillation column feed, column bottoms, inlet separators, reactor volume, or boiler drums. Emerson’s Mark Coughran explored ways to solve level control loop problems to smooth plant operation in his 2017 Emerson Exchange presentation.

Mark notes that level loops often have a surprising impact on plant stability and operator effectiveness. Properly designed and maintained level loops can absorb variability. However, many plants encounter level loops that generate variability (by self-cycling), propagate the variability to other vessels, or regularly cause alarms.

All components of a level loop should be considered as part of the root cause analysis. Don’t assume it’s solely the tuning of the level loop responsible for the issues with the loop. Level processes are integrating processes and can be counterintuitive to other types of processes.

The place to begin is to measure the process dynamics of the process itself. This is a requirement when applying a lambda tuning approach to the loop.

Mark shared 10 application examples that he and his fellow Operational Certainty consultants have worked on with customers. The first was a flue gas desulphurization unit softening tank. His second example was separators and coalescers in a specialty chemical plant. As found, the three coalescers where in manual mode of operation and the two separators cycling in automatic mode. Once lambda tuning was applied all were placed in automatic mode with the operators no longer having to constantly adjust the setpoint to try to maintain the level.

Mark’s third example was a power plant’s boiler drum level. The level had a constant oscillation with a constant load on the boiler. After applying lambda tuning, the oscillations were eliminated even when varying loads were placed on the boilers. Mark noted that process variable (PV) filtering should be considered part of the PID control strategy. This means that while there may be high variability on the PV, what the PID controller sees is a dampened signal significantly reducing this variability.

Another example included an LNG plant’s feed chiller and its fractionation unit’s inlet surge drum. The level loop was slowed down to absorb variability with fast response on the surrounding flow loops. The effect of these changes was to make the flow as steady as possible to the depropanizer unit.

In many cases, the process starts with fixing problems in valve response and measurement inaccuracies. Once these have been corrected, the loop tuning can be improved.

This week at the Emerson Exchange conference there is a short course on lambda tuning which you can join. If you’re not here, check out Mark’s lambda tuning whitepaper and consider available classes from the Emerson Educational Services team.

What’s the toughest level control loop you’ve encountered and how did you address it?