Abstract Sand production and foamy oil flow are the two key factors contributing the success in cold flow production in Alberta and Saskatchewan. However, two mechanisms have been treated and studied separately as geomechanics and multiphase flow problems, respectively. Special experiments were designed to combine these two processes, and conducted to study their interaction. The experiments involved flow of dead and live heavy oil in undisturbed heavy oil sand cores. It was found that gas nucleation in heavy-oil is the major factor in causing initiation of sand production in oil sand. This finding is consistent with field observation. A model for sand production in heavy oil reservoir is developed based on field and experiment observation. This model includes the effects of geomechanics and gas exsolution phenomena such as strength of oil sand, stress distribution in the reservoir, solution gas diffusion, foamy oil gas and fluid phase properties. Introduction Sand production and foamy oil flow are interrelated mechanisms in primary production (cold production) of heavy oil reservoirs in Alberta and Saskatchewan. Massive sand production could cause excessive deformation in oil sand and overburden resulting in detrimental effects on the production facilities. However, sand control measures tend to reduce the oil production rate. Numerical studies have been conducted to predict sand production in heavy-oil reservoirs (Wong et al. 1994; Geilikam et al. 1994; Geilikam et al. 1995; Fang and Wong 1996). However, limited experimental work has been performed to study the sand production in oil sand. Trembaly et al. (1996; 1997) used computer tomography imaging technique to examine the sand production process in sand pack columns using dead oil injected at a constant rate. They observed that a channel-like cavity was developed and evolved under a critical flow pressure gradient. However, there is no experimental reported study on sand production using natural oil sand cores and live heavy oil. The main objective of the paper is to investigate the effects of bitumen, oil sand interlocked structure, pressure gradient, and gas exsolution on sand production near a perforation in heavy-oil reservoir. The first part of this paper describes testing material, testing equipment, test details and results. The second part focuses on interpretation and analysis of test results and observation, followed by conclusions. Details of mathematical models used in the analysis of test results are presented in Appendix. TESTING MATERIAL AND EQUIPMENT The oil sand cores for the experimental study presented in this paper were recovered at a depth of 424m from an observation well (3–66-4-W4M) at a site near Cold Lake, Alberta. Core sampling was carried out using a conventional rotary core barrel of 89 mm inside diameter. Cores recovered were frozen at site and kept inside PVC tubes in a freezer. Prior to any testing, the frozen cores were x-rayed for sample selection. A high-pressure (70-MPa capability) stainless steel triaxial cell was used to conduct the sand production tests. The top and bottom platens used to provide confining pressure to the test core were modified to allow fluid injection and sand production (Fig. 1).