A Comprehensive Study of Sanding Rate From a Gas Field: From Reservoir, Completion, Production, to Surface Facilities

Abstract

An offshore gas field has been producing sand for a few years. Sand production has been closely monitored through acoustic flowline devices and a sand collection system installed on the platforms. Observation of sand production has triggered evaluation of whether to install surface desanders or complete future wells with downhole sand control. This evaluation requires a prediction of sanding rate over the reservoir life. The possibility of providing downhole sand control on existing wells was also evaluated in separate studies. Predicting sanding rate, particularly for gas fields, has been historically challenging, mainly because of the sporadic nature of sand production, inadequate quantification of fundamental physics, and lack of representative lab tests, and reliable field calibration. To tackle these challenges, four studies have been designed and executed, including 1) the development of a reliable log-based rock strength estimate; 2) the prediction of sanding rate over the reservoir life for a conservative well condition; 3) the evaluation of sand particle transport from the reservoir to the surface facilities; and 4) the estimate of potential erosion of platform facilities. The sanding rate prediction is based on extensive laboratory tests of four carefully selected whole cores, with gas and water flow. It has then been validated by field monitoring data from an acoustic flowline device on each producer and a sand collection system on the platforms. The studies have provided a prediction of future sand production, how much of the produced sand will be seen at the surface (and therefore how much of it will fall into the rathole), how fast various components of the surface facility will erode over the field life, and what will be the optimal completion strategy for sand control should it become necessary. They have provided input to an integrated evaluation of completion design, reservoir management, platform configuration, and field economics. Introduction For a long time sand production has been viewed as a cost source and a safety hazard for the oil and gas industry. It can erode downhole equipment and surface facilities, cause pipeline blockage, leakage, damage casing due to formation subsidence, lead to more frequent well intervention and workovers, and generate additional need for sand disposal. Since the 1980s, however, it has been consistently demonstrated that sand production could also be beneficial in both heavy oil reservoirs (Dusseault and Santarelli, 1989) as well as conventional oil reservoirs (Andrew et al., 2005). To allow sand production up to a certain level could result in a large amount of cost savings from the simplification or even elimination of downhole sand control. More importantly the removal of sand from the rock matrix could enhance the near wellbore porosity and permeability, promote oil mobility, and therefore increase production rate (Dusseault and Santarelli, 1989; Han et al., 2007). Economic benefits of avoiding complex and expensive downhole sand control have encouraged many oil and gas operators to select sanding strategies from a comprehensive evaluation of the sanding prediction, the equipment and facility tolerance, and the field CAPEX, OPEX, risk, HSE, etc. (Rawlins and Hewett, 2007), rather than simply reacting to the onset of sanding.

Sand Prediction: A Practical Finite-Element 3D Approach for Real Field Applications

“…[2] and [11]). Analytical models have however strong limitations; for instance, in presence of complex well geometries, particular constitutive behaviours of the rock or complicate subsurface environments.…”

“…A second approach [4],[7],[12],[20],[25],[18], [17] is based on analytical models which determine the critical conditions for the onset of sand production by calculating the stress state near both wellbore and perforations by means of analytical equations; these are generally derived by simplifying the geometry of the problem and the mechanical properties of the rock. These models are rather easy to use, particularly for screening and sensitivity analysis, and still continue to be one of the most used tools, also for volumetric sand evaluation in integrated process studies (see e.g.[2] and [11]). Analytical models have however strong limitations; for instance, in presence of complex well geometries, particular constitutive behaviours of the rock or complicate subsurface environments.…”