UNCERTAINTY QUANTIFICATION AND SENSITIVITY ANALYSIS OF SEMI-MECHANISTIC MODELS FOR SAND EROSION IN ELBOWS FOR SINGLE AND MULTIPHASE FLOWS

An accurate prediction of erosion in pipe fittings is a key aspect in the design process and during real-time control strategies of fluid transport facilities. Along with CFD simulations, semi- mechanistic models are commonly used for this purpose. These models predict erosion rates in a large variety of flow conditions with reasonable agreement with existing data, but no tool is available to quantify the uncertainty of these models. In addition, due to the error in measuring the input data, an uncertainty quantification tool is also needed for model validation. In this work, uncertainty quantification and sensitivity analysis are performed to quantify the level of confidence in predictions of one-dimensional mechanistic models for erosion in elbows considering single-phase and multi-phase flow regimes. A method is proposed for this purpose accounting for the effect of several variables on erosion predictions, such as fluid flow rates, geometry, solid and fluid properties. Monte Carlo simulations were used for the uncertainty quantification while two approaches for the sensitivity analysis (the screening and variance-based methods) were used to identify which variables the predictions are most sensitive to. The most critical input parameters in terms of uncertainty are rigorously identified case by case. The prediction intervals for erosion in elbows for different gas-liquid flow patterns in vertical flow are obtained from knowledge of the measurement error in the variables provided to the model. The model predictions are also compared to experiments while accounting for uncertainty. The results of this work give new insights regarding the degree of uncertainties in predictions of erosion rates in single and multi-phase flows, providing an important tool for future experimental validations of erosion models.