Predicting Volumetric Sand Production in Weak Reservoirs

Abstract

Abstract In general, sand production in oil/gas reservoirs in weak formations has an adverse impact on production costs. For efficient reservoir management, predictive tools are needed to quantitatively mitigate and control sand production. A fully coupled sand production model that integrates geomechanics and fluid flow was developed to quantify volumetric sand production during the early drawdown period (a few days), bean-up, as well as in the depletion period (a number of years). First, we compared simulation results from the proposed model to sand production data from full-scale laboratory perforation sanding tests for model verification. Then, we employed the developed model to evaluate the key variables that govern volumetric sand production. This study indicates that sand production in weak reservoirs is influenced by a number of variables. These include rock strength, flow rate, oil viscosity, and producing time. The simulation results relating cumulative sand production and sand production rate to these important variables are presented. Descriptions of the volumetric sand production model (formulations, constitutive equations, solutions procedures) are also presented. The study suggests that the proposed model may be used to generate quantitative information for predicting volumetric sand production. This provides a tool for sand-management in exploiting the weakness of sanding-prone scenarios. Introduction A sand production prediction model can assist in several aspects of field operations: sand control and management, optimal well completion design, and production optimization for reservoirs with weak formations. There exist many different types of models for predicting sand production.1,2 However, most would predict the transient and catastrophic types of sand production without quantifying volumetric sand production and sand production rate (e.g., Refs 3–7). Some models (e.g., Refs 8–14) attempt to predict the rate and quantity of solids that would be produced with the liquids in a continuous and/or episodic mode. The approaches used by these models for quantifying volumetric sand production may be classified into strain-based,8,9,10,11 erosion-based,12,13 and particle-based14 models. With the aid of laboratory and field data to calibrate some key parameters in a volumetric sand prediction model, the model could be useful for practical field applications. In this paper, a strain-based model for volumetric sand prediction is proposed along the lines of Biot's self-consistent theory and plasticity theory. To make the model a simple and practical tool for field applications, the proposed model retains only primary physics of rock failure and coupled rock deformation and fluid flow. The model does not include the effects of well configuration and completion, wellbore storage, erosion, interaction of disaggregated solid and flowing fluid, solid transport through the porous medium and any other complicated physics such as development of worm holes. In the following sections, first, we present the details of the proposed model that include coupled field equations, constitutive relations, and numerical procedures. Second, the model was verified by laboratory experiments from an industry-sponsored, sand production consortium. Third, we conducted simulation studies to examine and discuss how the key variables influence volumetric sand production in the early drawdown and the reservoir depletion periods. Lastly, some conclusions are drawn from this work.