A 3D modeling approach to complex faults with multi-source data
Introduction
Building a 3D geological model from field and subsurface data is a major instrument in geosciences, involving natural resource evaluation and hazard assessment. Faults are very important in geological modeling because they partition space into regions where stratigraphic surfaces are continuous. It is important to generate faults and to determine how faults terminate onto each other before considering other geological surfaces (Caumon et al., 2009, Euler et al., 1998).
General procedures and guidelines have already been proposed to build a model made of faults and horizons from typical sparse data, and to describe a typical 3D modeling workflow based on meshes (Tertois and Mallet, 2007, Caumon et al., 2009, Ritz et al., 2012). However, the faulted and fractured nature of the geology increases the complexity in determining the subsurface geologic framework (Wu and Xu, 2003). A major difficulty in elucidating geological structures is that the structure is largely unexposed; the fault-related uncertainties in the subsurface can significantly affect the numerical simulation of physical processes.
The improved methods are presented, such as the implicit structural modeling, that allow for including unconformities, thin and/or pinched-out layers in the models but that cannot explicitly localize slip along horizons (Caumon et al., 2013, Durand-Riard et al., 2013, Steckiewicz-laurent, et al., 2013). By using a combined cognitive and geostatistical approach, the resultant 3D model is more consistent with current geological observation and understanding (Royse, 2009). A 3D parametric fault representation has been proposed for modeling the displacement field associated with faults in accordance with their geometry (Cherpeau et al., 2010, Laurent et al., 2013). Holden et al. (2003) describe a stochastic model for the reservoir structure that may be used to represent the uncertainty of the structural model. The model may be conditioned using seismic and well data, while still allowing efficient simulation of realizations. Some structural uncertainty modeling techniques either allow for geometrical changes and keep the topology fixed, or create realistic stochastic fault networks with different topologies where the number and features of faults are changeable by taking into account fault-related uncertainties induced by subsurface imaging and interpretation ambiguities (Cherpeau et al., 2010, Cherpeau and Caumon, 2012).
However, there is a gap between research papers presenting case studies or specific innovations in 3D modeling and the objectives of a typical class in 3D modeling (Caumon et al., 2009). Because of the sparse samples and complex geological environment, the process of 3D modeling generally requires extensive manual intervention. In the case where fault and horizon mesh spacing are not closely matched in the preliminary modeling stages, parts of a horizon margin can project through, or stop short of, the fault surface. Although they are normally only visible at large magnification, the horizons are manually re-cut to the fault surfaces, and any overlap slivers are deleted (McCormac, 2009).
The 3D geological modeling is a process which implies to dynamically update and improve the models following the “construction-simulation-revision” (Wu and Xu, 2014). Fault modeling is one of the most difficult stages in this process. Three major problems may arise: (i) how to interpret the fault data acquired from existing measuring instruments to create a fault network; (ii) how to efficiently deduce the shape of fault in the area where samples are insufficient; and (iii) how to improve and evaluate the accuracy of the 3D fault model. For this, depending on the existing multi-source data (e.g., boreholes, 2D/3D sections, field investigation data), we provide key concepts and methodologies to be applied to different stages and tasks during fault modeling, with a specific focus on the spatial shape deduction and data analysis. The methods proposed in this paper make it possible to create models from sparse samples, and enhance the degree of automation, while reducing manual intervention.
The following contributions are presented in this paper:
- •
The definition of fault characteristics which allows automatic deduction of the shapes of faults (Section 2.3)
- •
A framework of fault modeling that makes the process of simulation geological structures more distinct, regardless of whether the available geological data are sufficient or not (Section 3).
- •
A fault deduction method that can be used to get further fault data so as to construct fault models by inferring the hanging wall and footwall lines (section 4.3) after displacement calculation (Section 4.2).
- •
A fault cutting mechanism that can supplement the available fault points on the location where faults cut each other (Section 4.4).
We demonstrate the application of the proposed 3D fault modeling approach and the implemented GeoSIS system in Tangshan, China, showing that the approach can be applied to broad and complex geological areas (Section 5).
Section snippets
Fault data generation and processing
Because geological phenomena are variable and uncertain, the geological data are incomplete and heterogeneous. The data that we use as input parameters of the 3D models come from diverse sources, therefore we refer to them as multi-source data. The data integration should unify the input data format and coordinate system, and address ambiguity or inconsistency through check and adjustment based on the judgment and reinterpretations of geologists. We consider three different kinds of
The framework of fault modeling
The fault modeling process requires to make interpretational choices and different approaches to structure modeling, deduction and uncertainty quantification can be applied. A workflow of fault modeling, called the framework of multi-source increasing, in the GeoSIS system we developed, may be decomposed into the following steps (Fig. 2).
Fault network analysis
Because all the discussed features require proper interpretation from indirect measurements of the incomplete data and demand judgments based on domain knowledge, there is a need for a case study that would consider fault location, dimensions and uncertainty of network connectivity (Arnold et al., 2013, Wu and Xu, 2014). The fault control-points form fault lines, and these lines form a fault network which is able to explicitly describe the topological relationship of faults. Further, through
Modeling in Tangshan, China
In this section, we present a concrete example to demonstrate the application of the proposed method to 3D modeling.
The geological structure in Tangshan, China is very complicated. The rocks are folded several times and there are 74 faults in the study area. The major strike direction of the faults is NE. Normal faulting is the dominant stress regime. The dip range is from 45° to 80°. The 3D model consists of 27 horizons. The fundamental geological multi-source data from the area (5.8×5.6k m2)
Summary and conclusions
In view of existing multi-source data, a workflow of fault modeling is developed in the GeoSIS system. It can integrate various samples to construct fault models through the processes including fault network analysis, fault deduction, data fusion, and fault cutting. By using a fault-based interpolation and remeshing the horizons, an accurate 3D geological model can be constructed.
On the basis of traditional fault elements (e.g., strike, dip, and displacement), we have added new characteristics,
Acknowledgments
This research was financially supported by China National Natural Science Foundation of China (Grant nos. 41430318, 51174289, and 41102180), China National Scientific and Technical Support Program (Grant nos. 201105060-06 and 2012BAB12B03), Beijing Natural Science Foundation (4142015), National Geological Survey Program (shui[2012]-01-035-036), Fundamental Research Funds for the Central Universities (2010YD02), Innovation Research Team Program of Ministry of Education (IRT1085) and State Key
References (29)
- et al.
Hierarchical benchmark case study for history matching, uncertainty quantification and reservoir characterisation
Comput. Geosci.
(2013) - et al.
Stochastic simulations of fault networks in 3D structural modeling
C. R. Geosci.
(2010) - et al.
The relationship between displacement and length of faults: a review
Earth—Sci. Rev.
(2005) - et al.
A parametric method to model 3D displacements around faults with volumetric vector fields
Tectonophysics
(2013) - et al.
Geometry, segmentation pattern and displacement variations along a major Apennine thrust zone, central Italy
J. Struct. Geol.
(2005) - et al.
Integrating complementarity into the 2D displacement discontinuity boundary element method to model faults and fractures with frictional contact properties
Comput. Geosci.
(2012) - et al.
Analysis of the relationship between displacements and dimensions of faults
J. Struct. Geol.
(1988) - et al.
An approach to computer modeling and visualization of geological faults in 3D
Comput. Geosci.
(2003) - et al.
An effective method for 3D geological modeling with multi-source data integration
Comput. Geosci.
(2005) - et al.
Scaling of fracture systems in geological media
Rev. Geophys.
(2001)