| Popis: |
Proppant flowback from hydraulic fracturing is widespread and costly due to erosion and/or blockage of producing hydrocarbons due to proppant accumulation. One remedy is to install sand control equipment integrated into a sliding sleeve device (SSD) as part of the completion string, which raises concern about erosion during fracturing. While some installations have been successful, at least one experienced sand control failure. Computational Fluid Dynamics (CFD) was deployed to evaluate the root cause and identify more robust designs, as presented herein. Firstly, we identified the most probable causes of sand control failure during multistage fracturing (MSF) in openhole (OH) horizontals. State-of-the-art CFD simulations were then performed on the installed design using actual flow conditions (rates, slurry properties, treatment time) from a failed installation. The static CFD methodology in an initial undeformed geometry proved to be ultra-conservative, so a new quasi dynamic mesh (QDM) methodology was developed, which yielded more realistic (albeit still conservative) erosion-depth predictions. The results revealed areas for improving the design of key components, and CFD was re-run to confirm erosion resistance targets. The modifications were then implemented for a field trial. Since frack location between two openhole packers is unknown, and the frack port is located between multiple screen/SSD assemblies, one must consider annular flow across the assembly in both directions. Accurate CFD predictions of erosion of completion components versus time during MSF in OH proved challenging. The quicker static methods were useful in ruling out some components as problem areas, such as the sand control media, but proved overly conservative on other key components. The QDM methodology gave more realistic results and indicated that erosion depths in specific locations could be deep enough to possibly cause sand control failure. To reduce the erosion risk, such components were modified, and the result was a reduction in predicted erosion depths to acceptable levels. A safety factor was already built into the predictions because of two key conservative assumptions: ignored initially were 1) particle-particle interaction and 2) erosion of the reservoir wall. The former was further investigated. While waiting for the field trial results, the main conclusion thus far is that CFD is a valuable tool for diagnosing erosion failures and improving equipment design. However, it’s essential to use a methodology that realistically captures downhole conditions. Presented herein is a more robust design of a screen/SSD assembly for proppant flowback control, as well as an improved CFD methodology for diagnosing sand control failures during MSF and for identifying design improvements of completion equipment. Furthermore, the inherent challenges of controlling proppant flowback without causing erosion or flow blockage of hydrocarbons are discussed. |
