Sand production is a prevalent problem during oil and gas production from weakly consolidated or unconsolidated formations. Accurate prediction of the conditions for sand production is critical to the design of cost effective completions. Sand production is a complicated process and many factors may affect it, such as flow rate, drawdown pressure, in situ stress and strength, water saturation etc. We apply a poroelastoplastic model to investigate the effects of in situ stress, flow rate/draw down pressure, rock strength, water saturation, and formation damage on the effective stress distribution around a hemispherical cavity representing a sand arch or perforation tip formed behind a cased and perforated well completion. Using this model, conditions for the initiation of sanding were examined. Experimental results of sand production tests carried out on unconsolidated samples at high and low water saturation were analyzed using the poroelastoplastic model.
Sand grains may become detached from damaged, weakly consolidated or unconsolidated formations, enter the well, and flow to the surface. This can cause severe production-related problems such as casing damage, surface equipment erosion, formation collapse etc. The prediction of sand production is critical to the development of effective sand prevention and control strategies. However, sand production is a complicated process and many factors may affect it. In unconsolidated formations, the formation of stable arches around a perforation is widely believed to inhibit sand production [1, 2, 3]. In weak sands which can support a perforation, it has been suggested that sand production might initiate from the perforation tip due to the presence of high pressure gradients in this region [4, 5]. In either case, radial tensile failure due to radial flow through a hemispherical-like cavity may be assumed a mechanism of sand production [6, 7]. Bratli and Risnes  presented an elastoplastic model to predict the critical condition for sand arch stability. In this model, steady state fluid flow was applied, and the effective radial stress in the plastic region of a sand arch was used to judge sand initiation. If the effective radial stress became tensile, the sand arch was deemed to fail and sand grains would flow into the well. This method has been used to predict sand initiation from the sand arch [4, 5, 6]. Weingarten and Perkins  extended Bratli and Risnes' model to allow steady state nonideal gas flow. Yi et al.  extended Bratli and Risnes' model to allow transient fluid flow and predicted the critical draw down pressure for sand production. The effect of capillary forces on sand production has been observed in laboratory and field settings. It is widely known that the onset of water production in oil fields can initiate sand production. Vaziri et al.  found that several HPHT wells in a large reservoir would not produce sand before water breakthrough, even at high draw down pressures. Bianco and Halleck  observed in their experiments that stable arches only formed in two-phase saturated sands when the wetting phase saturation exceeded 3%. They further noted that the stability of these arches was compromised at high wetting phase saturations.