Treffeisen, Torben (2021)
Fault Representation in Reservoir-Scale Hydro-Mechanical Finite Element Models.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00017419
Ph.D. Thesis, Primary publication, Publisher's Version
|
Text
20201230_Treffeisen_Dissertation.pdf Copyright Information: CC BY 4.0 International - Creative Commons, Attribution. Download (9MB) | Preview |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Fault Representation in Reservoir-Scale Hydro-Mechanical Finite Element Models | ||||
Language: | English | ||||
Referees: | Henk, Prof. Dr. Andreas ; Backers, Prof. Dr. Tobias | ||||
Date: | 2021 | ||||
Place of Publication: | Darmstadt | ||||
Collation: | XIX, 104 Seiten | ||||
Date of oral examination: | 17 December 2020 | ||||
DOI: | 10.26083/tuprints-00017419 | ||||
Abstract: | Hydro-mechanical reservoir models are used to obtain quantitative insights into the spatial distribution of stress, strain and pore pressure. Recent studies have shown that different approaches to incorporate faults into such reservoir simulations have a profound impact on the modeling results. Since faults are a key feature in the subsurface affecting both the hydraulic and mechanical behavior of a reservoir, their proper implementation in the numerical model is crucial. Fault representation has to accurately model the effect faults have on (1) fluid flow and (2) the local stress field. However, a fault is not just a discrete geological feature but rather a fault zone with a complex geometry and various rock units with distinct material properties. This small-scale heterogeneity can hardly be represented in reservoir scale finite element models considering the typical grid size used in these simulations. Thus, fault representation in reservoir-scale hydro-mechanical simulations has to be based on simplifications and upscaling techniques. To improve decision making and help in choosing the right fault representation, knowledge about the different effects each simplification and each approach used to incorporate faults has on the modeling results is necessary. This thesis focuses on different approaches of fault representation with a single upscaled set of material properties in reservoir-scale hydro-mechanical finite element models. The main objectives are (1) Implementing the fault geometry with respect to the finite element grid properly (2) Addressing the scale differences between the internal heterogeneity of the fault zone (centimeters to meters) and the typical size of the calculation cells of the numerical grid (meters to tens of meters) accurately (3) Assigning fault material properties to the numerical models, which stem – if available at all – from rock mechanical testing on core samples with a diameter of a few centimeters and therefore require upscaling and merging techniques In order to meet these challenges three research articles were published, each based on simple generic fault zone models. The approaches analyzed to represent faults in reservoir-scale hydro-mechanical include a regular rectangular grid, a grid geometry adapted to the fault geometry as well as fault representation by contact elements. Fault representation as volumetric weak zones is investigated for different grid geometries, fault dip angles as well as different mesh resolutions inside the fault zone. In addition, the impact of different elastic and frictional fault zone properties is assessed. Differences and similarities in the calculated stress and strain patterns as well as the pore pressure field obtained from different fault implementation strategies are discussed and general recommendations concerning the implementation of faults in hydro-mechanical reservoir models are given. Fault representation as either volumetric weak zones or contact elements leads to significant differences in the stress and strain patterns in the vicinity of the fault zone (< 50 m). While fault dip is not of critical importance for fluid flow, it has a significant impact on the stress perturbation induced by the fault. Another important finding is that the mesh resolution has to be considered very carefully as – particularly in combination with a rectangular grid – interlocking effects and serious errors can occur. If, however, the focus of a modeling study is not in the vicinity of the fault zone, a rectangular grid with the appropriate mesh resolution allows for faster and easier model generation in comparison to a curvilinear grid adapted to the fault geometry. Regarding material parameters, Young's modulus and cohesion assigned to the fault zone have the most significant impact on the modeling results, while the internal friction angle and Poisson's ratio play a subordinate role. Overall, this thesis provides recommendations and guidelines to improve fault representation in reservoir simulations. The goal is to gain more realistic simulations and thus, more reliable modeling results to improve forecasts, lower costs and reduce risks during subsurface operations. |
||||
Alternative Abstract: |
|
||||
Status: | Publisher's Version | ||||
URN: | urn:nbn:de:tuda-tuprints-174196 | ||||
Classification DDC: | 500 Science and mathematics > 550 Earth sciences and geology 600 Technology, medicine, applied sciences > 620 Engineering and machine engineering |
||||
Divisions: | 11 Department of Materials and Earth Sciences > Earth Science > Engineering Geology | ||||
Date Deposited: | 28 Jan 2021 07:13 | ||||
Last Modified: | 28 Jan 2021 07:13 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/17419 | ||||
PPN: | 476568013 | ||||
Export: |
View Item |