A volumetric sand production prediction tool has been developed, targeting rapid evaluation of production strategies for fields where sand risk may be an issue. The tool incorporates a semi-analytical model, based on hollow cylinder laboratory tests with radial flow where sand mass is monitored as a function of confining stress, borehole deformation and flow rate. Field data has been used as input to this tool in studies were different production strategies were evaluated as to their effect on predicted amount of produced sand. In particular, drilling and completion strategies are easily compared in terms of their associated sand risk: well deviation and perforation pattern. Similarly, the user can easily change input of hydrocarbon production strategy in terms of planned drawdown levels against assumed depletion evolution. Since geomechanical and petrophysical parameters are input as logs, the producing intervals of a whole well can be treated in one run, yielding a sand risk log as output.
The SandPredictor software, developed to assess the expected sand production rates for different oil and gas wells, using field data. The software implements a semianalytical model, based on a 1-D physical model taking into account erosion of a poro-elastic material under fluid flow and confining stress [1]. In addition, the model couples the stress state and pore pressure gradient at the borehole to the expected sand rate, through a constant parameter and a function of the excess stress (the difference in stress beyond the rock failure stress normalized by the latter). The software allows for intuitive selection of the required parameters, simply by clicking on the appropriate button. The sand model is tailored to field data input by repeating the sand rate calculation for all depths input directly in log format. The well geometry, including inclination, azimuth and completion (open hole, oriented or non-oriented perforations) together with the expected drawdown and depletion scenarios can be varied to quickly highlight the impact on sand rate. In this study, we demonstrate the use of the software on fictitious data assembled from real field log files and calculate sand rates for given draw-down and depletion scenarios. The expected sand production onset stress is calibrated from dedicated laboratory sand production tests on field cores, but the expected sand rate is also compared to the case where the onset stress is obtained from in house correlations [2]. The effect of varying input parameters such as unconfined compressive strength, in-situ stresses, well inclination and completion strategy are investigated. Outputs are cumulative sand mass and sand rate for the whole well as a function of time and as a function of depth. This type of study is easy to set up, provided that good quality log data is available, together with dedicated rock mechanics and/or hollow cylinder sand production tests to calibrate log interpretation on measured values for individual depths. Moreover, the calculation is near instantaneous, making it easy to go back and vary some of the input parameters to run a sensitivity study.
