Anti-slug control solutions based on identified model

• We present new PIDF and PI tuning rules for anti-slug control at offshore oilfields.
• We propose to identify a second-order unstable model from a closed-loop step test.
• Model identification and tuning rules are tested experimentally.
• Agreement between identified and mechanistic models is surprisingly good.
• Performance and robustness of PIDF controller is comparable with H∞H∞ controllers.

A anti-slug control requires operation around an open-loop unstable operating point. One solution is to design a robust controller based on a mechanistic model. An alternative and more robust approach is to identify an unstable model of the system based on input–output data. We used a closed-loop step test to identify an unstable linear model. From this, we obtained a second order IMC (Internal Model Control) controller that can be implemented as a PIDF controller. From the asymptotes of the proposed IMC controller, we also derive a simple tuning for PI-controller. Next, we considered two types of robust H∞H∞ controller (mixed-sensitivity and loop-shaping). The proposed model identification and control solutions were verified experimentally on two different test rigs. We found that the robustness and performance of the IMC (PIDF) controller is comparable with the H∞H∞ controllers. However, the proposed IMC (PIDF) controller is easier to tune compared to H∞H∞ control.