Virtual outcrop models of petroleum reservoir analogues: a review of the current state-of-the-art
Virtual outcrop models of petroleum reservoir analogues: a review of the current state-of-the-art
复制标题
石油储层类似物的虚拟露头模型：当前最先进技术的回顾
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
D. Hodgson
中科院分区：
文献类型：
--
作者：
J. Pringle;J. Howell;D. Hodgetts;A. R. Westerman;D. Hodgson
地下储层模型是基于计算机的岩石物理参数（如孔隙度、渗透率、流体饱和度等）的表示。鉴于这些参数的直接测量仅限于少数威尔斯井，因此有必要外推其分布。由于地质是岩石物理学的一级控制因素，因此了解相及其分布对于预测储层质量和结构至关重要。用于地下的大多数储层建模系统基于地震导出表面的相关性来定义储层区。然后，井数据用于进一步定义亚地震尺度层位，并确定网格单元中表示的区域属性。了解两个亚地震表面的分布和它们之间潜在的非均质地质仍然是一个重大的挑战。此外，由于典型的网格单元尺寸是c。50 - 200 m2，很难将小规模的异质性结合起来。因此，关键的是使用关键地层层位和内部相分布的实际值。沉积相是岩石物理的基本控制因素。然而，相尺度的非均质性使用当前的地震方法是不可分辨的，并且井数据提供很少或没有关于超出井筒的3D几何形状的数据。对现代沉积事件的研究可以对沉积过程和相分布之间的联系给出一些指示（例如，Kenyon等人，1995年）;然而，保存的沉积结构也强烈地受到随时间变化的可容纳空间的控制（Jervey，1988年）。基于验证的实验（例如，Kneller & Buckee，2000年）和基于过程的建模（例如，艾格纳等人，1989; Peakall等人，2000）进一步说明了沉积机制和相结构之间的联系。然而，这样的模型通常在远小于典型场的尺度上，并且更适用于放大研究（Nordhal等人，2005; Ringrose等人，2005年）。露头研究长期以来一直被用作研究类似物和了解油田的机制（Collinson，1970; Glennie，1970; Breed & Grow，1979）。一旦从地下数据中得出沉积体系的类型和油气田的可容纳历史，就可以识别适当的露头类似物（例如，亚历山大，1993年）。合适的类似物是那些在地质学上与正在研究的系统相当的，并且在足够大的区域上具有良好的三维露头暴露，以捕获所需的异质性规模（Clark & Pickering，1996）。因此，露头模拟研究是提高对储集相结构、几何形状和相分布的认识的关键途径。已进行了露头类似物的定性研究，最近还进行了定量研究。传统的定量研究（例如，德雷尔等人，1993; Chapin等人，1994年; Bryant & Flint，1993年; Clark & Pickering，1996年; Reynolds，1999年）一直专注于收集露头数据，以通过随机、基于对象的方法填充井间储层模型区域（Floris & Peersmann，2002年）。然而，从传统的露头研究中提取有用的数据可能很困难，特别是当它需要与石油工程数据库集成或以3D方式可视化时。此外，代表穿过固体地质的地形切割的露头是2D的，虽然很少有例子显示具有不同方向的固体地质的多个截面，但仍然需要地质专业知识来充分理解和解释这些物体的3D性质。此类工作可能还需要地质统计数据处理，以克服露头方向和尺寸问题（Geehan &安德伍德，1993年; Vissa & Chessa，2000年），但理想情况下，数据应进行三维重建。准确的三维重建是唯一的方法，可以定义三维目标砂体的参数，如河道弯曲度，连通性和连续性。这些参数是对烃生产的关键控制，包括波及效率（Pringle等人，2004 a; Larue & Friedmann，2005）。用于以3D表示地质的软件通常用于对地下储层进行建模。本文将展示最近的数字数据采集技术的进步如何通过获得准确和定量的露头模拟数据集来帮助解释储层地质学家，从而帮助并可能修改其储层模型。
A subsurface reservoir model is a computer based representation of petrophysical parameters such a porosity, permeability, fluid saturation, etc. Given that direct measurement of these parameters is limited to a few wells it is necessary to extrapolate their distribution. As geology is a first order control on petrophysics, it follows that an understanding of facies and their distribution is central to predicting reservoir quality and architecture. The majority of reservoir modelling systems used for the subsurface are based on correlation of seismically-derived surfaces to define reservoir zones. Well data are then used to define further, sub-seismic scale horizons and determine the zone properties which are represented in grid cells. Understanding the distribution of both sub-seismic surfaces and potential heterogeneous geology between them remains a significant challenge. Furthermore as the typical grid cell size is c. 50-200 m2 it is challenging to incorporate small-scale heterogeneities. It is critical, therefore, to use realistic values for both key stratigraphic horizons and internal facies distributions. Depositional facies is a fundamental control on petrophysics. However, facies scale heterogeneities are not resolvable using current seismic methods, and well data provide little or no data on 3D geometries beyond the well bore. Studies of modern sedimentary events can give some indication of the link between depositional processes and facies distribution (e.g., Kenyon et al., 1995); however preserved depositional architecture is also strongly controlled by changes in accommodation through time (Jervey, 1988). Laboratory-based experiments (e.g., Kneller & Buckee, 2000) and process-based modelling (e.g. Aigner et al., 1989; Peakall et al., 2000) further illustrate the link between depositional mechanism and facies architecture. However, such models are typically on a scale that is far smaller than the typical field and are more applicable to upscaling studies (Nordhal et al., 2005; Ringrose et al., 2005). Outcrop studies have long been employed as a mechanism of studying analogues and understanding petroleum fields (Collinson, 1970; Glennie, 1970; Breed & Grow, 1979). Once the type of depositional system and the accommodation history of a hydrocarbon field are derived from subsurface data, appropriate outcrop analogue(s) can then be identified (e.g. Alexander, 1993). Suitable analogues are those that are geologically comparable to the system that is being studied and also have excellent 3D outcrop exposure over an area that is large enough to capture the scale of heterogeneity required (Clark & Pickering, 1996). Outcrop analogue studies are thus a key way of improving understanding of reservoir facies architecture, geometry, and facies distributions. Outcrop analogue studies have been undertaken both qualitatively and more recently quantitatively. Traditional quantitative studies (e.g., Dreyer et al., 1993; Chapin et al., 1994; Bryant & Flint, 1993; Clark & Pickering, 1996; Reynolds, 1999) have been focused on the collection of outcrop data to populate inter-well reservoir model areas by stochastic, object-based methods (Floris & Peersmann, 2002). However, it can be difficult to extract usable data from traditional outcrop studies, especially when it needs to be integrated with petroleum engineering databases or to be visualized in 3D. Furthermore, outcrops which represent a topographic cut through solid geology are 2D and while rare examples show multiple sections through the solid geology with different orientations, geological expertise is still required to fully understand and interpret the 3D nature of the bodies. Such work may also need geostatistical data manipulation to overcome outcrop orientation and size issues (Geehan & Underwood, 1993; Vissa & Chessa, 2000) but ideally the data should be reconstructed in 3D. Accurate 3D reconstruction is the only way that parameters such as channel sinuosity, connectivity, and continuity of target sandbodies in 3D may be defined. Such parameters are a key control on hydrocarbon production, including sweep efficiency (Pringle et al., 2004a; Larue & Friedmann, 2005). Software for representing geology in 3D is routinely used to model subsurface reservoirs. This paper will show how recent digital data capture technique advances aids the interpreting reservoir geologist by obtaining accurate and quantitative outcrop analogue datasets to aid and perhaps modify his reservoir model.