Hybrid Modeling of Sand Production from Petroleum Wellbore

本研究致力于开发一种实用、可靠的油井出砂预测工具。出砂是一种出砂随油藏流体流动的生产现象。据估计，打磨每年花费该行业数十亿美元(例如，通过侵蚀生产设施)。另一方面，少量的出砂可以通过去除井壁表面来提高井筒生产率，例如，在钻井过程中，由于地层损害而产生的井壁表面。 井筒完井有两种类型：(1)不允许出砂的防砂，和(2)允许甚至鼓励少量出砂的出砂管理。防沙成本很高，而且会降低生产率。出砂管理降低了完井成本，提高了产能，但需要在完井和井筒作业设计过程中进行可靠的出砂预测。 目前，有几种评估砂率的模型，它们要么基于连续介质力学，要么基于不连续介质力学。基于连续介质的模型在分析出砂方面很差。结果错误地依赖于数值网格设计。基于不连续介质的模型可以更准确地模拟打砂，但计算要求很高，不太适合现场规模的应用。 在这项研究中，将进行基于连续和不连续的方法的联系，以利用每个模型的最佳方面。不连续介质描述将用于井筒附近地区，在该区域，通过岩石中裂缝和裂隙的发展，预计将产生广泛的不连续响应，并在那里砂子被调动。在更远的地方，将使用连续介质描述，因为连续介质力学可以准确地描述岩石响应。这一方法使我们能够开发出一种准确、可靠的模型来评估出砂率，该模型可用于现场范围的实际应用。

This research focuses on developing a practical and reliable tool for predicting sand production from petroleum wellbores. Sanding is a production phenomenon where sand flows with reservoir fluids. Sanding is estimated to cost the industry billions of dollars every year (e.g., through erosion of the production facilities). On the other hand, a small amount of sanding can improve wellbore productivity through removing the wellbore skin created, for example, by formation damage during drilling. There are two types of wellbore completion: (1) sand control in which no sanding is tolerated, and (2) sand management in which small quantities of sanding are allowed or even encouraged. Sand control is costly and reduces productivity. Sand management reduces the completion costs and enhances productivity, but requires reliable sand prediction during completion and wellbore operation design. Currently, there are several models for assessing sand rate, which work based on either continuum or discontinuum mechanics. Continuum-based models are poor at analyzing sanding. The outcome results are spuriously dependent on the numerical mesh design. Discontinuum-based models simulate sanding more accurately, but are computationally demanding and poorly suited for field-scale applications. In this research, linkage of the continuum- and discontinuum-based approaches will be carried out to take advantage of the best aspects of each model. The discontinuum description will be utilized for the near-wellbore area where extensive discontinuum response is expected through the development of cracks and fissures in the rock and where sand is mobilized. Farther away, a continuum description will be utilized as the rock response can be accurately described by continuum mechanics. This approach enables the development of an accurate, reliable model for assessing the sanding rate that can be used in practical applications at the field scale.