Sand production in the petroleum industry is a phenomenon of solid particles being produced together with reservoir fluid. This is a major problem that operating oil companies have faced for many years. To date, despite several research studies, sand production remains a nightmare for petroleum engineers.

Sanding-related problems are known to have increased production costs considerably:​ corrosion of the pipelines and other instruments, sand-oil separation cost, possible wellbore choke and repeated shut-in and clean-up of the wellbore are examples. On the other hand, a controlled sanding or even sand production invocation has proven very effective for increasing production rates, especially in heavy oil recovery, asphalt wells, and low PI wells. This thesis adopts a comprehensive approach to understand, model and, where necessary, prevent sand production. A new completion technique is presented for eliminating sand production altogether.

Even though many researchers have tried to predict sand production in the past, none suggested a comprehensive model that would take care of a range of failure and sand production mechanisms. Moreover, rare models predict sanding rate and volume along with the onset of sanding. This thesis presents a comprehensive numerical modeling of sand production whose criteria derive from the physics of sanding, taking the sequential nature of sand production into consideration.

The numerical model simulated two series of experiments on large-scale samples, one with vertical, and another horizontal well. The behaviour of the sample under loading, as well as the rate and volume of sanding, agreed reasonably well with the experimental results. This confirms that the criteria used for the model captured the essence of sanding.

An expandable completion technique is proposed which uses composite material. Two different techniques are proposed for implementation in horizontal and vertical/slanted wellbores. Even though the technique that is suggested for horizontal wellbore is deployable in all other wellbores, in order to reduce the cost and ease the installation process, another technique is suggested for vertical and slanted wellbores.

To alleviate some of the concerns and uncertainties on the interactions between an expandable liner and the medium around it, a series of experiments were conducted using hollow cylinder synthetic sandstone samples involving both fine and coarse grained sands at varying degrees of consolidation, including totally non-cemented samples. A stiffener, which was a representative of the proposed expandable completion technique, supported the central hole of a series of samples, while another series used open hole completion.

The actual influence of pore collapse on the severity of the damage to the rock could not be established, although the failure was readily noted. The experiments showed that a sample in the state of pore collapse retains its structure. In addition, the interlock between the grains is enhanced as the material compacts. The experiments showed that the material in the state of pore collapse produced little sand. The stiffener was also successful in eliminating general shear and tensile failure of the medium around the well. The experiments showed that a stiffener highly alleviated the amount of the produced sand and the magnitude of sanding, if any at all, depended on the aperture size. If the aperture size is too large, local shear failure may occur in the material at the opening location. Therefore, limited amount of sanding may be produced before stable arches develop.

Experiments on non-cemented highly porous sand-pack samples equipped with the stiffener indicated spontaneous sand production under single-phase flow. However, the situation was highly alleviated under a two- phase flow condition. As a result, capillarity played a determining role in holding the grains together and forming stable arches.

Finally, a straightforward analytical methodology is introduced which can predict the critical drawdown associated with the onset of sanding under either of single- or two-phase fluid flow. This formulation was applied to the experimental data of both weakly consolidated and unconsolidated sandstones. A remarkable agreement between the experimental observations and analytical predictions was concluded.

The proposed completion technique proved its efficacy for alleviating sand production. Additionally, the numerical and analytical modeling schemes demonstrated their capability to capture the essence of sanding. Therefore, the use of the tools presented in this thesis seems very promising for the future of sanding prediction and control.