膨胀的
地质学
可塑性
磁导率
机械
岩土工程
压力(语言学)
多孔性
岩体分类
孔隙水压力
材料科学
复合材料
生物
物理
遗传学
哲学
语言学
膜
作者
Victor N. Nikolaevskiy,Michael J. Economides
摘要
Abstract High-stress concentration and substantial weakness of loosely consolidated to unconsolidated petroleum formations may cause considerable mechanical damage both during the drilling and the subsequent reservoir fluid production. Damage in our context is the appearance of a cloud or swarm of rock defects (microcracks) and the opening or, even, closing of preexisting pores or cracks. A physically and conceptually adequate idea is the mathematical modeling of the phenomena using an elastoplastic approach, which accounts for stress-sensitive matrix failure and solid/mass movements. Volume deformation because of dilatant interaction with shear is also accounted for. Damage is characterised by the increase of a zone of plasticity augmented by internal-to-the zone phenomena. The emergence of plasticity does not necessarily mean total formation failure or loss of wellbore stability. Such a conclusion can be drawn only with the determination of displacement (i.e., strain) fields. Because problems such as the ones envisioned here are complex and their tracking depends on a number of parameters, computer modeling is necessary along with, at times tedious, laboratory determination of rock parameters. In the absence of measurements, published "typical" data for similar rocks can be used. The parameters include initial porosity and pore pressure, permeability, the elastic moduli, cohesion, and the friction and dilatancy coefficients, which change with changing stress. Problems described in this paper include the near-well damage effects because of high stress anisotropy and pore collapse. In this case local strains are determined. A second problem that is considered is the impact of well curvature and deviation from the vertical and the effects of crossing layers. Finally, sand production and some thoughts on water-shale interaction during drilling are addressed.
科研通智能强力驱动
Strongly Powered by AbleSci AI