Challenging environments, such as deep-water, require greater effort when designing wells and completion systems. In such cases, hydrocarbon production associated risks must be evaluated from very early stages of field development. Particularly, the risk of sanding should be investigated to select and define the optimum well completion, and if required, the appropriate sand control options. This paper presents a novel approach to predict the onset of sanding by describing a case study from a deep-water gas field in Southeast Asia.

Sanding analysis generally consists of either a numerical or analytical model calibrated against laboratory tests and production histories. Here, both numerical and analytical models are implemented to achieve a robust and reliable prediction in the absence of any production data. For this purpose, the numerical model is prepared and run using a comprehensive material model considering poroelastic-plastic rock behaviour with strain hardening. The rock is modelled from a combination of single-stage triaxial and advanced thick-walled cylinder rock mechanical tests on representative samples from the target reservoirs. A poroelastic analytical model calibrated against the numerical modelling results is then used to conduct sand production predictions with sensitivity analyses for a number of well trajectory scenarios for both cased and perforated, and open-hole completions.

Although the numerical and analytical models follow different assumptions, the sanding prediction results from both models are consistent. Once a reliable validation is obtained, the analytical method can be used with confidence for sand production prediction for future wells through its simplicity and fast realization of various scenarios. The results show that a high risk of sanding exists for both cased and perforated, and open-hole completion systems for the planned deviated wells. Hence, downhole sand exclusion options are deemed to be required for the development wells.

Incorporating analytical poroelastic and numerical poroelastic-plastic sanding evaluations enhanced the reliability of sanding prediction for development wells. This approach is particularly beneficial in the absence of production data at appraisal or early development stages.

You can access this article if you purchase or spend a download.