ABSTRACT: Centrifuge tests were performed to study sand production problem that is a major issue in many oil and gas recovery projects. The tests simulated seepage-induced failure around a multi-perforated vertical well. Two different scenarios were considered: (1) perfs evenly spaced along the casing within a homogeneous sand layer, and (2) ditto but a shaley seam was introduced in the middle of the layer. Data collected during the experiments included time dependent variation of pore pressure within the pay zone, flow rate, sanding rate, and tracking of the enlargement of the cavity around the well. Numerical modeling was used to better interpret and support the experimental results. Overall, the study performed shows that (1) the location of perfs affects the mode and magnitude of sand production and the concomitant productivity, and (2) long-term productivity can be improved through managed sand production. These findings are consistent with some field cases where extraordinary increases in production were noted due to sanding.
1 INTRODUCTION
Management of sand production related problems is complex and is in desperate need of optimization. The conventional approach has been to err on the conservative side by employing any effective measure that prevents or minimizes sanding. Such measures commonly include installation of liners and gravel packs around wellbores right after drilling. Depending on the depth of the reservoir and length of the payzone, the installation cost can be quite high. There are also ongoing maintenance (e.g., workover) issues to contend with as gravel packs and filters become progressively plugged and that lowers the overall productivity (Kenter and Curfie, 1998). The primary design issue in situations where excessive sand production cannot be tolerated is to determine when to install sand control measures as often sanding does not become a problem until late in the life of the reservoir when significant depletion leads to squeezing of the material towards the opening. By deferring the placement of packs and filters, not only a better use of the capital spending is made but also the productivity of the well is not unnecessarily compromised.
There are a number of key issues that field operators are concerned with when it comes to sand production. In some fields (e.g., offshore) any sand production is considered too much and every feasible measure must be taken to prevent it. The question in these situations is when should sand exclusion measures be installed. In less sensitive cases where operators are prepared to tolerate some sand production (sand bursts) in order to remove the skin damage, the question is how to make sure that sand production will not get out of control. Then there are cases where operators intend to exploit sanding to boost long term productivity. While this has been exercised with some success, the field performance has not been consistent. In some cases, very little improvement in productivity has been experienced, perhaps because not enough sand was produced or perhaps sanding could not be contained within the pay zone. In other cases, sand production either be came continual or culminated in water production, possibly due to absence or break up of the cap rock. In cases where sanding did successfully yield higher productivity, the principal mechanisms could not be uniquely identified as such the same success could not be repeated.
Take the case for the Burnt Lake project (Yeung, 1995). In this case, in a pilot test, sanding resulted in over two orders of magnitude increase in productivity. Based on a relatively fast communication established between an injection and observation well that were over 500 m apart, several possibilities have been specul
