A Multiscale Study on The Onset of Sand Production

Abstract

Despite decades of numerical, analytical and experimental researches, sand production remains a significant operational challenge in petroleum industry. Amongst all techniques, analytical solutions have gained more popularity in industry applications because the numerical analysis is time consuming; computationally demanding and solutions are unstable in many instances. Analytical solutions on the other hand are yet to evolve to represent the rock behaviour more accurately. We therefore developed a new set of closed-form solutions for poro-elastoplasticity with strain softening behaviour to predict stress-strain distributions around the borehole. A set of hollow cylinder experiments was then conducted under different compression scenarios and 3D X-Ray Computed Tomography was performed to analyse the internal structural damage. The results of the proposed analytical solutions were compared with the experimental results and good agreement between the model prediction and experimental data was observed. The model performance was then tested by analysing the onset of sand production in a well drilled in Bohai Bay in Northeast of China. Acoustic and density log along with core data were used to provide the input parameters for the proposed analytical model in order to predict the potential sanding in this well. The proposed solution predicted the development of a significant plastic zone thus confirming sand production observed by today sanding issue in this well.