There are more than 15 sand control methods and one of the frequently used methods is the rate control method. There are mainly two cases where the rate control method can be applied: (1) permeability is significantly damaged around a well and the sand problem induced by erosion failure due to abnormally large drawdown can be controlled by the rate control method, and (2) the stability of a sand-arch built by disintegrated sand particles can be controlled by rate control. A series of laboratory tests were conducted in this work to measure the sand-arch strength using a polyaxial confining cell. The parameters changed in these experiments were the ratio of grain size to cavity diameter, grain distribution, cavity diameter, oil-water ratio, confining stress and directional stress. A similar experiment was previously conducted by the Colorado School of Minesand IKU, however, the tests conducted by the Colorado School of Mines used 20-40 mesh loose sands which are too coarse compared with actual reservoir sands, and IKU used air instead of liquid flow. Both of them used some immobile water saturation while the saturation change from 0 to 30% significantly reveals the complex phenomena of water effects upon sand-arch strength. The main objective of the laboratory tests was to separate the mechanical arch strength and the strength induced by the capillary pressure. It used a large grain size to observe large mechanical arch strength without capillary pressure effects andan intermediate grain size to observe capillary pressure effects. An analytical solution of a spherical cavity growth due to erosion of the post-failure zone is also developed to verify the observed phenomena in the laboratory tests.