Commissioning-stage poor performance of control loops in process plants

Isolating the causes of poor performance observed at the time of commissioning can prove to be difficult especially when oscillations are observed, as these can propagate and hence mask their source. An approach that combines an analysis of the overall loop performance index, power spectra, physical properties and cross-correlation is proposed to try to distinguish between poor performance caused by tight coupling due to improper system design, or by a loop fault. A case study is presented to clarify the analysis process