Reducing Loop Interactions and Optimizing LNG Plant Operations

I discovered Nick Denbow’s Industrial Automation Insider blog this weekend and added it to my list of automation and process industries-related blogs. He had a post about university students studying the feasibility of using liquefied natural gas (LNG) as a fuel source in short sea shipping.

I mention this since I had just gotten my hands on an LNG-related presentation that Emerson’s Mark Coughran is giving at the upcoming Emerson Exchange technical conference. You may recall Mark from some of our past process optimization-related posts. Mark’s presentation is titled, Solving Loop Interactions in a New LNG Train Using Combined Emerson Tools.

This LNG producer had several key level loops operating in manual mode, which consumed a portion of the operators’ attention to maintain process stability, quality of propane and butane, and overall throughput. Many of the flows had large variability and the pressure loops were slow or oscillatory.

These loops included a drum level controller on a heat recovery steam generator, a De-Aerator level controller in the utilities plant, and all the controllers on the Cryogenics units. Mark notes that it’s typical when starting a new plant based on a new design that the interactions between loops have not been fully considered and that controllers may be initialized with default values.

The De-Aerator control is critical to boiler production and reliability. The operators were operating this level loop in manual and were constantly fighting interactions between pressures and flows. The controls involve a pressure loop on the low-pressure steam and cascaded flow and level controller on the polished condensate line into the drum surge tank. Mark and the team optimized the performance by tuning the pressure loop first and making its response the fastest. Using the Entech Toolkit to measure process dynamics and to help identify controller PID parameter settings, they were able to tune the loops to avoid oscillatory loop response and separate the dynamics and interactions between loops. After testing the performance, the loops were returned to automatic and cascade modes of control.

In the presentation, Mark will share how they tuned fuel-gas supply pressures and user pressures to hold their setpoints better and not to interact by applying Lambda tuning and slowing down the response of the upstream supply pressure loop. Other process optimization improvements were done to an incinerator airflow loop on a sulfur recovery unit, inlet surge drum level on a fractionator unit, and liquefaction unit inlet flow and pressure controllers.

In each case, Mark presents the simplified piping and instrumentation diagram (P&ID) and the original trends of the loops operating in automatic mode before optimized loop tuning applied. He then describes how the operators expect the process to perform. Next, he shares what process dynamics were measured, and the trends after the new PID parameters were placed in service. He concludes with the results of these changes and the suggested educational course of learning for the engineers and operating staff to maintain optimized loop performance.

The result of this optimization effort was to move all of the controllers into automatic mode and give the operators the confidence to change setpoints and maintain automatic mode even during startup and shutdown sequences. With daily revenues over a million U.S. dollars, these improvements in throughput, quality, and overall process stability quickly paid for the work performed.

If you’re coming to the Emerson Exchange and are responsible for the performance of the loops in your plant, this is a presentation you’ll not want to miss.

GreenPodcast.gif MP3 | iTunes

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Leave a Reply