Sanding Prediction Validated by a Perforated Test in a Sandstone

Abstract

This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper SPE 158255, ’Sand Prediction by Different Criteria and Validation by a Perforated Test in a Sandstone,’ by Y. Wang, SPE, Petro-Geotech, and E. Papamichos, SPE, Sintef Petroleum Research, prepared for the 2012 SPE Heavy Oil Conference Canada, Calgary, 12-14 June. The paper has not been peer reviewed. Sand production from a perforated sample was determined by the onset of a significant discrepancy between strains in orthogonal directions. The onset was analyzed with three sanding models—shear-failure, cohesive-tensile-failure, and effective-plastic-strain (EPS) models. The EPS model provided results closest to testing. Equivalent plastic displacement (EPD) is proposed as an additional critical parameter that enables determining the sanding onset directly. Introduction Predicting sanding is increasingly important in completion and production engineering. If the sanding point is surpassed, engineers must determine if, when, and what sand-control strategy is required, and these determinations critically affect the decisions regarding the type of completion selected, such as fracturing, horizontal well, open hole, or cased hole. Engineers also need to evaluate the benefit of oil-production enhancement if no sand control is implemented. All these issues can depend on the accuracy and reliability of the sanding prediction. Even though a Mohr-Coulomb (M-C) criterion is used widely for the critical-shear-failure or sanding conditions, it has been considered too conservative; therefore, a Lade model is proposed. Study Three criteria, reflecting different sanding mechanisms, are applied extensively to sanding problems in the field, and, consequently, different critical sand pressures or drawdowns are generated. Use of these criteria to determine the sanding point from a single test or field condition can yield more questions than answers; therefore, defining an adequate criterion could be quite helpful. The use of stress-oriented criteria usually is simple, requiring fewer parameters, but such criteria can be too conservative, and, more importantly, those required parameters cannot be measured and monitored directly. The models used with stresses often are determined statically (i.e., they cannot be linked uniquely to the deformation), whereas extensive deformation can be linked proportionally to sanding, which is essential to determine the critical point both physically and mathematically. Another issue of the use of stress-oriented criteria defined by the M-C and cohesive-sanding models is that the parameters cannot be measured or monitored in the laboratory or in the field; instead, they are acquired from an indirect strength measurement because stresses or strengths can only be interpreted, not measured. Alternatively, a deformation-related criterion is proportional to well unloading and the required input can be monitored and measured directly.