Uncertainty quantification for cuttings transport process monitoring while drilling by ensemble Kalman filtering. (May 2017)

• Record Type: Journal Article
• Authors: Frangos, Michalis
• Is Part Of: Journal of process control. Volume 53(2017)
• Page Start: 46
• Page End: 56
• Publication Date: 2017-05
• Subjects: Oilfield drilling process monitoring -- Cuttings transport -- Uncertainty quantification -- Bayesian inference -- Ensemble Kalman filtering -- Real-time process monitoring; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2017.02.008
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract : Highlights: A real experiment was conducted to collect wired drilled pipe oilfield data. Cuttings transport monitoring is formulated as a statistical inverse problem. A model for oilfield cuttings transport is proposed and described in detail. Ensemble Kalman filtering was used to make real-time process estimations. We draw a conclusion on process uncertainties from limited data at the moment. Abstract: The aim in this work was to develop a method to draw our conclusions for the uncertain quantities of interest in a real problem in oilfield, namely the cuttings transport problem, given the limited data available. The cuttings transport process is subject to disturbances and is influenced by various factors in a stochastic and uncertain manner. In addition, the drilling environment is very complex to understand fully, and to model efficiently and accurately. An experiment was conducted to monitor the process in real-time via time-series measurements of distributed pressure transducers along the drillstring. A mathematical model describing the dynamic process of cuttings transport is developed, which aims to capture the dominant characteristics of the process, without attempting to model reality perfectly; modeling and measurement errors are represented by uncertainties in model inputs, parameters, and states. The proposed model is described in detail, and is incorporated in the Bayesian framework (via ensemble Kalman filtering) to make our best estimate about the location and amount of cuttings transported along the wellbore in real time, given the available data. The estimations of the uncertain process parameters are described.