Modelling and Validating a Deoiling Hydrocyclone for Fault Diagnosis using Multilevel Flow Modeling

Decision support systems are a key focus in research on developing control rooms to aidoperators in making reliable decisions, and reducing incidents caused by human errors. For thispurpose, models of complex systems can be developed to diagnose causes or consequences forspecific alarms. Models applied in safety systems of complex and safety critical systems, requirerigorous and reliable model building and testing. Multilevel Flow Modeling is a qualitative methodfor diagnosing faults, and has previously only been validated by subjective and qualitative means.This work aims to synthesize a procedure to measure model performance, according to diagnosticrequirements, to ensure reliability during operation. A simple procedure is proposed for validatingand evaluating Multilevel Flow Modeling models. For this purpose expert statements, a dynamicprocess simulation in K-spice, and pilot plant experiments are used for validation of two simpleMultilevel Flow Modeling models of a deoiling hydrocyclone, used for water and oil separation.