Abstract:

In offshore oil platforms, choke valve erosion is a major issue. An indicator of the choke valve health state is the valve flow coefficient, which is a function of measured and allocated parameters. The allocated parameters are typically provided by a physics-based model which has been proved to be inaccurate for some operating conditions. As a consequence, inaccuracies are introduced in the evaluation of the health indicator, undermining the possibility of using it for prognostics. In this paper, we overcome this hurdle by resorting to hybrid modelling which integrates the physics-based model into an ensemble of data-driven models built using Kernel Regression (KR) methods. A local procedure which uses the historical performance of the physical and data-driven models is adopted to aggregate the different model outcomes. The proposed hybrid ensemble-based approach is verified on real measurements performed on offshore choke valves located topside at different wells.