The complications involved in understanding and predicting sand production are discussed. The need to predict sanding potential on a foot-by-foot basis means that a simplifying approach is required, and this paper sets out experimental and initial modelling progress in this direction.
We have induced sand production in simple, reproducible and easily monitored systems, in order to be able to vary one parameter at a time and examine its effect. This has meant the use of synthetic rocks, moulded rather than perforated holes, and two simple geometries.
The effects of stress, flow rates, fluid viscosity and rock strength have been examined so far. For the high permeability rocks used in the experiments, the stresses in the rock produced by radial flow into the perforation tunnel have little influence on sanding; flow along the axis of the tunnel is found to be the controlling flow. This allows the sanding problem to be split into the prediction of damage around the tunnel induced by far-field stresses, and the transport of damaged material by axial flow.