RU2542874C1 - Method for adaptive breakdown of three-dimensional grid into layers when creating three-dimensional digital geologic models of formations - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: in the method, creation or correction of grid layers is carried out by successively searching source data which characterise layering; new grid layers are created at layering boundaries based on the source data under the condition of absence, in the allowable region vertically, of grid layers created in advance, otherwise the closest grid layer on the detected layering boundary is corrected vertically, the grid layer is corrected horizontally within a triangulation grid constructed on points of intersection of the source data with the corrected grid layer in the vicinity of said point.

EFFECT: providing accurate display of non-uniformity of the geologic structure of a formation in a three-dimensional grid with the least number of layers.

2 dwg

Description

The invention relates to mining, in particular to methods for dividing a three-dimensional grid into layers, and can be used in geology when creating three-dimensional digital geological models of formations.

A three-dimensional grid is a cellular framework in the volume of which all the main stages of geological modeling are carried out. A three-dimensional grid is built inside the so-called “zone” or formation. The layer is the volume between two surfaces located above and below. To accurately reproduce the geological structure of the simulated formations, the entire volume is divided into small cells. The combination of these cells is a three-dimensional geological grid. The process of creating a three-dimensional geological grid consists in dividing the reservoir volume into small cells in accordance with predetermined rules, determined directly by the geologist conducting the simulation. The rules for dividing the volume into cells must be set both laterally (“slicing” the grid into “columns” and “rows”), and vertically (“slicing” the grid into layers) [1]. There are two known methods for breaking a three-dimensional grid into layers:

1. Proportional breakdown. With this type of breakdown, the entire thickness of the reservoir is divided into a given equal number of layers, regardless of its total thickness. Thus, with a proportional breakdown, any area of the formation contains a predetermined number of layers, however, the thickness of each layer in different parts of the formation can vary.

2. Parallel breakdown. With a parallel breakdown, the entire thickness of the formation is divided into layers of a given thickness, and the boundaries of the layers are parallel to either the roof, or the sole of this formation, or some other surface.

The number of layers into which the entire volume of the reservoir is divided depends on the necessary degree of detail of the created model, i.e. the minimum required layer thickness is set, and therefore, the more detailed the model is needed, the more layers must be specified. Then, meshing along the trajectories of wells drilled through the reservoir, the results are transferred (averaged) of the interpretation of well logs (geophysical surveys of the wells) - facies curves, lithology, porosity, oil saturation, etc. All these data directly characterize the layered structure of the formation, its heterogeneity. When they are transferred to the grid region, the layer boundaries of which do not correspond to the layered boundaries of the original well data, the initial data are distorted and especially thin layers disappear. Thus, the disadvantage of all the methods is that in the process of forming the layers, the initial information on which the further creation of the geological model will be carried out is not taken into account. The second significant drawback is that in order to achieve high accuracy of the models, it is necessary to increase the number of layers in the model grid, which, in turn, leads to an increase in the dimension of the model and complicates the possibility of its further use.

The closest way, taken as a prototype, is the method of proportional dividing the model into layers [1, adj. 3, p. 247]. The disadvantages of this method are errors when transferring the source data to the model grid and increasing the dimension of the model grid with an increase in the number of layers in order to minimize the errors of transferring the source data.

The objective of the invention is to provide a method for adequately displaying heterogeneity of the geological structure of the formation in a three-dimensional grid of the model with the least number of layers.

The problem is solved in that in the method of adaptive dividing of a three-dimensional grid into layers, the initial information on the wells, which characterizes the layered heterogeneity of the formation, is used, and the grid layers located in a certain vicinity of the layering boundaries are corrected according to the initial data.

When forming the grid layers, all source data are enumerated and a search is made for the boundaries of layering. If such a boundary is detected, a check is carried out for the presence of an existing mesh layer within the allowable area (DO). DO - this is the maximum distance at which it is possible to adjust the grid layer vertically, is determined by the user. In the absence of an existing layer in DO, a new mesh layer is created. The newly created layer extends to the entire area of the model in a proportional or parallel pattern. If there are one or several previously created mesh layers in the DO, the vertical adjustment of the nearest mesh layer to the corresponding layering boundary is carried out. To preserve the smoothness of the layer surface by means of approximation, the layer is horizontally adjusted within the triangulation grid, constructed from the points of intersection of the initial data with the corrected grid layer and certain correction values at these points. Thus, as a result of enumerating all the initial data, a three-dimensional grid is cut into layers, which completely characterizes the layering in the initial information using the minimum number of grid layers.

The essence of the proposed method is illustrated in the drawing, on which:

Figure 1. Side view of a three-dimensional grid (13). In wells (1, 2, 3), layering is characterized by the presence of (4) and the absence of (5) reservoir. Layers of the grid (6) are created at the boundaries of the layers defined in the well (1), are traced over the entire area and are adjusted (8) in the well (2), since the layering boundary is within the allowable region (7) of possible adjustment. No new layers are drawn along well (2), since the existing layering boundaries are already characterized by the existing adjusted mesh layers. In the well (3), there is a layering boundary (9), which does not have a grid layer in the allowable region; therefore, a new grid layer is created (10).

Figure 2. View of the three-dimensional grid (13) from above. The influence area (11) of a certain displacement (14) of the layer in the well (2) over the area of the three-dimensional grid. Triangulation network according to the source data (12).

Information sources

1. Zakrevsky K.E. "Geological 3D modeling." M .: LLC "CPI" Mask "", 2009.

Appendix 3. Mesh construction.

Claims (1)

A method for adaptively dividing a three-dimensional grid into layers when creating three-dimensional digital geological models of formations is that the creation or adjustment of the grid layers is carried out by sequentially sorting the initial data (ID) characterizing the layering, while new grid layers are created at the boundaries of the layered by ID, provided the absence of previously created mesh layers in the permissible vertical region; otherwise, the vertical adjustment of the nearest mesh layer is carried out to the detected layering boundary, and is also adjusted I am the horizontal grid layer within the triangulation grid, constructed at the points of intersection of the source data with the adjusted grid layer, in the vicinity of this point.

Images