3rd Geoqus Conference - Ruhr-Universität Bochum · Simulating the Response of Faulted Geology with...

1

3rd Geoqus Conference

21.-23. August 2012

at the

Helmholtz Centre Potsdam GFZ German Research Centre for Geosciences

Sponsored by

International Lithosphere Program

Task Force VII

3rd International Geoqus Conference, 21.-23. August 2012 in Potsdam

3

Conference Schedule Tuesday, August 21st

09:00 - 09:30 Registration & Poster installation in Building H, GFZ German Research Centre for Geosciences

09:30 - 09:45 Welcome and opening of the conference

09:45 - 10:30 Wouter van der Zee (invited speaker) 3D Geomechanical Modeling using Abaqus© and JewelSuite©

10:30-11:00 Coffee break

11:00-12:00 Jens Lüke (invited speaker) XFEM in Abaqus

12:00-12:30 Martin Schoenball A mechanism for time-delayed wellbore failure

12:30-13:45 Lunch break

13:45 - 14:30 Harvey Goodman (invited speaker) Connecting Well Engineers to the Geological Prospect using acoustic driven earth models

14:30 - 15:00 Coffee break

15:00 - 15:45 Maria-Katarina Nikolinakou (invited speaker) Geomechanical modeling of stresses and pore pressures in mudstones adjacent to salt bodies

17:15 Meeting at the lobby of the Mercure Hotel

17:45 - 21:00 Boat trip (www.potsdamer-wassertaxi.de) to the icebreaker party at the Cafe/Restaurant Kleines Schloss, Park Babelsberg (http://www.kleinesschloss.de/) with cold and hot buffet, drinks in a splendid setting

Wednesday, August 22nd

09:30 - 10:15 Björn Lund (invited speaker) Modelling glacial isostatic adjustment in regions with strong lateral variations, such as in Iceland

10:15 - 10:45 Heidi Turpeinen: Investigating the response of faults to mass redistribution due to erosion and sedimentation by using fully coupled 3D finite-element models with landscape evolution tool


11:15 - 12:00 David Steedman (invited speaker) Simulating the Response of Faulted Geology with Abaqus©/CEL Code

12:00 - 13:00 Lunch break

13:00 - 13:45 Walk over the science park Telegrafenberg and visit of the historic buildings

13:45 - 15:00 Brief presentation of the poster by the authors (~3-5 minutes each)

15:00 - 18:00 Coffee break followed by discussions at the posters with cold buffet

Thursday, August 23rd

09:30 - 10:15 Tobias Hergert (invited speaker) Modelling the state of stress in the Earth's crust

10:15 - 10:45 Johannes Altmann Geomechanical-numerical model of The Geysers geothermal field


11:15 - 12:00 Jens Nüchter (invited speaker) Complex states of stress during the seismic cycle around major reverse, normal and strike-slip faults

12:00 - 12:30 Oliver Heidbach Fast visualization of 3D stress data from the Abaqus© odb-file using Tecplot 360© and the geostress add-on

12:30 Final remarks

4 3rd International Geoqus Conference, 21.-23. August 2012 in Potsdam

Poster Presentations Poster Session

Nastaran Abdolmaleki

Numerical modeling of salt diapirism at Qum Kuh, Iran University of Tehran, Iran

Sierra Boyd Stress and seismicity changes on nearby faults from The Geysers Geo-thermal Field

University of Berkeley, USA

Beatrice Cailleau Thermoelastic modelling of a backfilled salt mine Free University of Berlin, Germany

Amir Haghi Investigation of Well Casing Damage Due to "Groundwater-Withdrawn Induced Compaction" Using Advanced Numerical Simulation Method

National Iranian Oil Company, Iran

Alireza Hassanzadegan

Thermo-mechanical coupling during compaction of porous media GFZ Potsdam, Germany

Mong-Han Huang Probing the Deep rheology of Tibet: Constraints from the 2008 Mw 7.9 Wenchuan, China Earthquake

University of California, Berkerly, USA

Marie Keiding Estimates of stress changes from the 2010 Maule, Chile earthquake: the influence on crustal faults and volcanos

GFZ Potsdam, Germany

Shiyuan Li Numerical studies of the deformation of salt bodies with embedded carbonate stringers

RWTH Aachen Universi-ty, Germany

Marcos Moreno Strain changes in an earthquake cycle: Kinematic FE-Models and GPS data from the great 2010 Maule Chile earthquake


Maria C. Neves A Damage Model for the Terceira Rift Origin University of the Al-garve, Portugal

Karsten Reiter Stress Field Model of the Alberta Bais versus in-situ Stress Data GFZ Potsdam, Germany

Karsten Reiter Initial Stress State in Crustal Stress Field Models GFZ Potsdam, Germany

Holger Steffen Using ABAQUS for glacial isostatic adjustment modeling and stress analysis

University of Calgary, Canada

Rebekka Steffen Implementing faults in Glacial Isostatic Adjustment Models Using Abaqus University of Calgary, Canada

Wouter van der Wal User-defined subroutine for non linear mantle deformation in postglacial rebound

Delft University of Tech-nology, Netherlands

Michael Warsitzka Salt flow during basement extension and differential loading – viscoelas-tic deformation in analogue and numerical models

University of Jena, Germany

Stefanie Zeumann New 3D finite Element Models of the Central Andes Using Realistic Geometry



5

Welcome We are glad to have you here as a guest of the GFZ Potsdam at the 3rd International Geoqus Conference and we hope that the meeting will strengthen and widen the network of geo-scientists that use Abaqus. We hope that you will have an inspiring time during the confer-ence and also a good time in Potsdam with its wide range of tourist attractions. If there is any questions please do not hesitate to contact us.

Best regards,

Oliver Heidbach, Johannes Altmann and Karsten Reiter

Contact

Helmholtz Centre Potsdam GFZ German Research Centre for Geosciences Section 2.6 Seismic hazard and stress field Telegrafenberg 14473 Potsdam, Germany phone ++49 331 288 2814

Conference Focus Focus is the usage of the finite element code Abaqus in geomechanical-numerical modelling of lithosphere deformation and dynamics. Besides the scientific results of the model also the technical implementation into Abaqus (boundary conditions, loads, forces, physical proper-ties, usage of subroutines etc.) are of particular interest and should be integrated into the presentation. The conference will take place at the GFZ German Research Centre for Geosci-ences in Potsdam near Berlin, Germany and is financed by Task Force VII of the International Lithosphere Program (i.e. no conference fees). The conference is addressed to experienced users of Abaqus rather than pure beginners. The preliminary scientific-technical program of workshop is on the following topics:

Implementation of a reference stress state for 3D geomechanical-numerical models. Which is the best practice to implement the initial stress conditions to balance gravita-tional body forces?

Rheology such as visco-plasticity and non-linear creeping. There are multiple facilities offered by Abaqus to implement various rheologies.

How to set up a thermo-poro-elastic geothermal or CO2 storage reservoir model with Abaqus?

Combination of Abaqus with other meshing and analysis tools. What is the most conven-ient way to set up a 3D geometry of model and its discretization?

Presentation of a new 4D visualization tool (add-on Tecplot360) that allows to derive easily from the odb-file e.g. k-ratio along drilling paths, stress regime, maximum horizon-tal stress (magnitude and orientation).


Conference Venue and Ice-Breaker Location The conference takes place at the Helmholtz Centre Potsdam - GFZ German Research Centre for Geosciences on the Telegrafenberg in building H, ground floor (see campus plan next page). Walking distance from the central station to the Hotel Mercure is 5-10 minutes and from the Hotel Mercure to the conference venue app. 20 minutes. Alternatively you can take a taxi from the Hotel Mercure (5 min, ~8 Euro) or the bus number 691 from the main train sta-tion (leaves at 8:44 and 9:14 and takes 6 min., last stop is in front of the building H on the GFZ Campus).

Central train station Potsdam

Water Taxi stop


7

Campus Plan GFZ Potsdam

Hotel Mercure @ Water Taxi Stop


What is Geoqus? Geoqus is an open user group of scientist who employ the commercial finite-element software Abaqus (Abaqus Inc.) for any geoscientific application www.ruhr-uni-bochum.de/geoqus/. The user group addresses mainly academic users of universities and other scientific research institutes around the globe. Users of other finite-element software are also welcome to join Geoqus.

The Geoqus user group provides a platform for single users, large research groups, beginners and experts to get in contact with other users of Abaqus in the geosciences for the benefit of the whole community. A mailing list is set up for this reason to which everybody can sign up and post their questions, comments, ideas, etc. Interesting questions are for example which rheological model to use, how to write subroutines, and which initial conditions to use. An-other purpose of this user group is to organize small meetings or workshops to discus tech-nical aspects of using Abaqus in geosciences. Users are also welcome to exchange scientific ideas and results based on their numerical models.

Geoqus is based on an idea of Oliver Heidbach from the German Research Center of Geosci-ences GFZ Potsdam and Kasper Fischer of the University of Bochum. Home of Geoqus is the geophysics group of the Institute of Geology, Mineralogy and Geophysics at the Ruhr-University Bochum.

Geoqus Mailing list

The Geoqus mailing list is the primary platform of communication between Geoqus members. Messages to discuss different Abaqus and Geoscientific subjects and announcements (i.e. on upcoming workshops, conferences, job positions) can be posted. Please check also the Abaqus Support Homepage if you have a more general Abaqus related question or technical problems. Please feel free to subscribe to the Geoqus mailing list. Just visit the mailing list webpage to subscribe or unsubscribe to the Geoqus mailing list. Registered users can send e-mails to all members by using the e-mail address [email protected]. The con-tents of previous e-mails of the Geoqus mailing list can be found in the Geoqus archive.


9

Abstracts in the order of the schedule Tuesday, August 21st

09:00 - 09:30 Registration & Poster installation in Building H, GFZ German Research Centre for Geosciences

09:30 - 09:45 Welcome and opening of the conference

09:45 - 10:30 Wouter van der Zee (invited speaker) 3D Geomechanical Modeling using Abaqus© and JewelSuite©

10:30-11:00 Coffee break

11:00-12:00 Jens Lüke (invited speaker) XFEM in Abaqus

12:00-12:30 Martin Schoenball A mechanism for time-delayed wellbore failure

12:30-13:45 Lunch break

13:45 - 14:30 Harvey Goodman (invited speaker) Connecting Well Engineers to the Geological Prospect using acoustic driven earth models


15:00 - 15:45 Maria-Katarina Nikolinakou (invited speaker) Geomechanical modeling of stresses and pore pressures in mudstones adjacent to salt bodies

17:15 Meeting at the lobby of the Mercure Hotel

17:45 - 21:00 Boat trip (www.potsdamer-wassertaxi.de) to the icebreaker party at the restaurant Meierei (www.meierei-potsdam.de)

Wednesday, August 22nd

09:30 - 10:15 Björn Lund (invited speaker) Modelling glacial isostatic adjustment in regions with strong lateral variations, such as in Iceland

10:15 - 10:45 Heidi Turpeinen: Investigating the response of faults to mass redistribution due to erosion and sedimentation by using fully coupled 3D finite-element models with landscape evolution tool


11:15 - 12:00 David Steedman (invited speaker) Simulating the Response of Faulted Geology with Abaqus©/CEL Code

12:00 - 13:00 Lunch break

13:00 - 13:45 Walk over the science park Telegrafenberg and visit of the historic buildings

13:45 - 15:00 Brief presentation of the poster by the authors (~3-5 minutes each)

15:15 - 18:00 Coffee break followed by discussions at the posters with cold buffet


3D Geomechanical Modeling using Abaqus and JewelSuite Wouter van der Zee E-mail: [email protected] Baker Hughes - Reservoir Development Services The use of 3D geomechanical modeling becomes increasingly more important in the oil and gas industry. The 3D geomechanical model can predict the stresses in the earth crust during the development and depletion of a reservoir. This stress state is an important input parameter for wellbore stability studies and sand production prediction analyses. Further the stress and strain fields can be used to predict subsidence of the surface during the depletion of the reser-voir. The authors will present the development and application of a 3D geomechanical model-ing system based on Abaqus and JewelSuite. The presentation outlines the workflow which starts with the structural information forming the basis for both the FE mesh and the geologi-cal and reservoir modeling. The next step is the property population of the reservoir and geomechanical model. The final step is running the reservoir model and the FE model where the coupling between the reservoir simulator results and the geomechanical results ultimately leads to coupled reservoir modeling.The system described in the presentation provides an ef-ficient way to create realistic 3D FE simulations for complicated structural settings. It is op-timized for the best resolution around the area of interest while limiting the size of the numer-ical problem to an order that can be handled in reasonable times. Most importantly, it is set up to provide exactly the same topology in the reservoir and in the FE model which allows for easy mapping of simulation results from one mesh to the other which is the basis for coupled reservoir modeling.


11

A mechanism for time-delayed wellbore failure Martin Schoenball E-Mail: [email protected] Drilling of wells is often accompanied by severe problems caused by instability of the drilled rock formation. By drilling the wellbore cavity, high stresses arise at the wellbore wall, lead-ing to the formation of breakouts which enlarge the hole to an elliptical shape. If such for-mation of breakouts occurs during drilling, the rock debris falls on the drill bit which may lead to stuck pipe problems or even abandonment of the drill string. Reasons for such time-delayed failure of the wellbore may be due to chemical fluid-rock interaction, especially in swelling clays. However, such delayed instabilities have also been observed e.g. in gneiss formations at the KTB borehole (Germany) although chemical interaction is negligible. I pro-pose to link observations of delayed wellbore failure with time-dependent brittle creep, which has been observed for many types of rocks. Following this approach, rock may fail under loads less than their short-time strength but after a long enough time span (e.g. Heap et al. 2009). This time is in exponential relation to the load applied to the rock. I implement a mod-el developed for the creation of shear bands on the basis of time-dependent brittle creep by Amitrano and Helmstetter (2006). Here, progressive damage of the formation is captured by a damage parameter D and the time-to-failure TTF. Young’s modulus E is decreased by a factor every time TTF is expired, i.e. when failure is reached. Subsequently, stresses are redistribut-ed according to the new distribution of E in the formation. Using this approach, I obtain clo-sure of the well with primary and secondary creep phases. Wellbore breakouts are formed progressively with time by chipping of a shard of rock. After a certain time the formation of breakouts comes to an end and a steady-state is achieved. References Amitrano, D., and A. Helmstetter (2006), Brittle creep, damage, and time to failure in rocks, J. Geophys. Res.,

111, B11201, doi:10.1029/2005JB004252. Heap, M. J., P. Baud, P. G. Meredith, A. F. Bell, and I. G. Main (2009), Time-dependent brittle creep in Darley

Dale sandstone, J. Geophys. Res., 114, B07203, doi:10.1029/2008JB006212.


Connecting Well Engineers to the Geological Prospect using acoustics driven earth models Harvey E. Goodman, Chevron, USA E-mail: [email protected] Since the early 1990’s, Chevron has maintained a strategy to develop technologies that enable the building of 3D rock mechanical earth volumes from acoustics dominated data sets. This effort has led to the current capability to characterize static elastic moduli of the subsurface in order to apply advanced finite element modeling tools, such as Abaqus, to define in-situ stress attributes across the geological structure. The capability to create rock mechanical properties from the standard P-wave 3D seismic grid enables numerical earth modelers to work along-side well engineers to characterize pre-drill location safe drilling mud weight windows and to forecast drill bit performance through the overburden, i.e., rate of penetration and bit life. Fur-thermore, a mature 3D formation mechanical property data set focused on the reservoir can be used to predict seismic attributes that correspond to fluid saturation. Chevron now applies this Abaqus centric technical capability to business units world-wide for wellbore stability, pro-duced solids prediction, fracture stimulation design, cuttings disposal design, bit optimization & performance prediction and seismic reservoir characterization for by-passed oil. The presentation will introduce acoustics based rock mechanics concepts, describe Chevron’s acoustics based rock property prediction technique, and present field application case histories for selected business units world-wide, including deepwater GOM, North Sea, offshore West Africa and Asia.


13

Geomechanical modeling of stresses and pore pressures in mudstones adjacent to salt bodies Maria Nikolinakou, University of Texas, USA E-Mail: [email protected] We employ coupled poromechanical models to study how salt bodies change the stresses and pore pressures within their neighboring sediments. We use the Abaqus poro-elastic model, and the Modified Cam Clay model through a user-developed subroutine. Salt is a viscous ma-terial and loads the mudrocks through a combination of stress relaxation and salt body ad-vance. Analyses with a simple salt geometry show that salt relaxation can lead to plastic de-formations and can induce pore pressure perturbations that extend kilometers away into the sediments. The time scale of dissipation of these perturbations is on the order of millions of years, suggesting that overpressures should commonly be present in mudstones near salt sys-tems. Because previous models have not used fully-coupled geomechanical models, they are unable to predict the interdependence between pore pressure and stress; our results may pro-vide insight into pore-pressure anomalies that have been observed in deepwater drilling near salt. Analyses of a salt-sheet advance show that the contact properties, the topography of the salt and also the presence of overpressures in the sediments determine whether the dominant mechanism of salt advance is sliding or frontal rolling. Most importantly, our analyses illus-trate that the salt topography alone can provide a good estimation of the overpressures at the salt base. Finally, analyses of actual Gulf-of-Mexico salt geometries demonstrate that the lat-eral loading from the salt is a dominant mechanism that increases the horizontal stresses in minibasins and in front of the salt body. Such geomechanical models have the potential to illuminate how deformation occurs and to predict stress and pore pressures in salt systems around the world.


Modelling glacial isostatic adjustment in regions with strong lateral variations, such as in Iceland Björn Lund, University of Uppsala, Sweden E-Mail: [email protected]


15

Investigating the response of faults to mass redistribution due to erosion and sedimentation by using fully coupled 3D finite-element models with landscape evolution tool Heidi Turpeinen , Georgios Maniatis and Andrea Hampel, University of Hannover, Germany E-Mail: [email protected] Mass redistribution on Earth’s surface creates loads that can influence the rate of crustal de-formation. We use three-dimensional finite-element modeling combined with landscape evo-lution tool CASQUS to show how surface processes may affect fault slip evolution and study the morphology resulting from the interaction between faulting and erosion/sedimentation. The models consist of a 100 km x100-km-wide and 15-km-thick upper crust with one or more embedded faults in different tectonic scenarios. A previous study has shown that surface pro-cesses may lead to accelerated fault slip on normal faults (Maniatis et al., EPSL, 2009). The results from models with single faults as well as fault arrays showed an increase of up to 15% in fault slip rate due to the effects of surface processes when compared with corresponding model runs without surface processes. Based on the previous study, we investigated the effect of erosion and sedimentation on normal faults when the far-field extension has ceased (Turpeinen et al., in review with Geomorphology). During the first phase of a model run, the fault accumulates normal slip owing to the applied velocity boundary condition, while erosion and sediment deposition are active on the model surface. In a subsequent phase, extension of the model is stopped while the surface processes remain active. The parameter study revealed that surface processes may not only increase the fault slip rate in an active extensional regime but might also prolong fault slip accumulation for millions of years even in the absence of regional extension. The amount and time interval of additional fault slip is mainly controlled by the diffusion constant as well as the fault length and dip. Our studies imply that mass redis-tribution on the Earth´s surface due to erosion and sedimentation may substantially affect the evolution of normal faults. Future modeling will investigate the evolution of thrust faults un-der the influence of evolving landscapes and the resulting mass redistributions.


Simulating the Response of Faulted Geology with Abaqus©/CEL Code David Steedman and David Coblentz, Lawrence National Laboratory, USA E-Mail: [email protected] The Source Physics Experiment (SPE) is currently being conducted at the Nevada Test Site with the goal of improving our physical understanding and ability to model how explosions generate seismic waves, particularly S-waves. Information gathered from this experiment in-cludes high-resolution accelerometer, infrasound, seismic, explosive performance, and radio frequency data. This data will advance current, state-of-the-art strong ground motion and seismic wave propagation models and algorithms toward a predictive capability. The second shot in the series (SPE-2) included detonating a chemical explosive equivalent to 1000 kilo-grams of TNT in a contained, confined environment at a depth of 45 meters in the Climax Stock granite. A weathered layer of variable thickness and two high angle faults that intersect the test shaft characterizes the geology of the shot location. This geologic scenario was incor-porated into LANL’s Geologic Framework Model (GFM) and used to model the ground shock created by a buried high explosive. The GFM facilitates development of a physical ge-ometric model with geological and geophysical data imported directly into the Complete Abaqus© Environment (CAE). Using the Abaqus© code’s coupled Euler-Lagrange (CEL) capability within an Euler domain, we model the source region, and CEL provides full physi-cal interaction between the high-deformation explosive detonation process and the Lagrange domain containing a complex user constitutive model that considers a non-linear equation of state compaction model along with three-invariant yield and failure. Shell structures explicit-ly model faults and Lagrange contacts model the interaction between the faults and the adjoin-ing rock masses, such that we could represent the large-scale geologic features observed at the site. We present the computed effects of the main identified faults on the ground response, which compare favorably in both magnitude and trend with LIDAR data collected prior to and following SPE-2.


17

Poster Presentations in Alpabetical Order Poster Session

Nastaran Abdolmaleki

Numerical modeling of salt diapirism at Qum Kuh, Iran University of Tehran, Iran

Sierra Boyd Stress and seismicity changes on nearby faults from The Geysers Geo-thermal Field

University of Berkeley, USA

Beatrice Cailleau Thermoelastic modelling of a backfilled salt mine Free University of Berlin, Germany

Amir Haghi Investigation of Well Casing Damage Due to "Groundwater-Withdrawn Induced Compaction" Using Advanced Numerical Simulation Method

National Iranian Oil Company, Iran

Alireza Hassanzadegan

Thermo-mechanical coupling during compaction of porous media GFZ Potsdam, Germany

Mong-Han Huang Probing the Deep rheology of Tibet: Constraints from the 2008 Mw 7.9 Wenchuan, China Earthquake

University of California, Berkerly, USA

Marie Keiding Estimates of stress changes from the 2010 Maule, Chile earthquake: the influence on crustal faults and volcanos


Shiyuan Li Numerical studies of the deformation of salt bodies with embedded carbonate stringers

RWTH Aachen Universi-ty, Germany

Marcos Moreno Strain changes in an earthquake cycle: Kinematic FE-Models and GPS data from the great 2010 Maule Chile earthquake


Maria C. Neves A Damage Model for the Terceira Rift Origin University of the Al-garve, Portugal

Karsten Reiter Stress Field Model of the Alberta Bais versus in-situ Stress Data GFZ Potsdam, Germany

Karsten Reiter Initial Stress State in Crustal Stress Field Models GFZ Potsdam, Germany

Holger Steffen Using ABAQUS for glacial isostatic adjustment modeling and stress analysis

University of Calgary, Canada

Rebekka Steffen Implementing faults in Glacial Isostatic Adjustment Models Using Abaqus University of Calgary, Canada

Wouter van der Wal User-defined subroutine for non linear mantle deformation in postglacial rebound

Delft University of Tech-nology, Netherlands

Michael Warsitzka Salt flow during basement extension and differential loading – viscoelas-tic deformation in analogue and numerical models


Stefanie Zeumann New 3D finite Element Models of the Central Andes Using Realistic Geometry



Numerical modeling of salt diapirism at Qum Kuh, Iran Nastaran Abdolmaleki, University of Tehran, Iran E-mail: [email protected] Numerical modeling of the salt tectonics has become an important task in geophysical investigation related to petroleum exploration, underground natural gas storage and waste disposal, and dynamic response systems. In this paper, the dynamical activity of salt tectonics at Qum Kuh located in the Great Kavir of central Iran is modeled assuming that salt and sedimentary basins behave like Newto-nian fluids. The finite element method is used to simulate the diapirs evolution and analyze the diapir behavior in geological time. The computational methodology of interface between layers is based on Lagrangian approach. For more accurate interpretation of the physical and geophysical setting the simulation is compared with surface deformation data analyzed by InSAR time-series data between 2003 and 2008 in this region. The other geological aspects such as differential loading, gravitational creep, erosion, regional extension and sedimentation have been simulated to consider their effects on diapiric rise.


19

Stress and seismicity changes on nearby faults from The Geysers Geothermal Field Sierra Boyd, University of Berkeley, USA E-Mail: [email protected] Much research has focused on the abundant shallow seismicity associated with geothermal steam extraction and fluid injection in geothermal environments. Induced earthquakes can be damaging to nearby populations. Thus far, the largest magnitude earthquake at The Geysers Geothermal Field, California, is a Mw 4.6. Another concern is whether geothermal energy production can affect the state of stress and trigger earthquakes on nearby faults beyond the reservoir. The Geysers steam reservoir lies in a complexly faulted antiform bounded to the northeast by the Collayomi fault zone and to the southwest by the Maacama fault zone, both right lateral faults of the San Andreas system. An analysis of seismicity rates along these faults reveals changes that could provide insight into this question. For example, earthquake swarms have been noted in the vicinity of the nearby Maacama Fault subsequent to geother-mal production at The Geysers. Simple models of deformation associated with The Geysers will be studied to provide a first-order estimate of the magnitude of likely stress changes on nearby faults beyond the reservoir. Mossop and Segall, 1997 and 1999, model volume strain at The Geysers during various time periods and show that surface deformations recorded from 1975-1996 are consistent with volume contraction within the reservoir. In this study, the geo-thermal reservoir will be modeled using a source subjected to a negative pressure to simulate volume contraction. Nearby faults will be modeled using planes of various orientations con-taining 2-km square patches that can move relative to each other. Constraints will be imposed on the faults such that motion, if any, will be consistent with regional faulting. In addition, the effects of poroelasticity on the resulting stress and deformation of nearby faults will be ex-plored. References Mossop, A., and P. Segall (1999). Volume strain within The Geysers geothermal field, J. Geophys. Res. 104,

B12, 29,113-29,131. Mossop, A., and P. Segall (1997). Subsidence at The Geysers geothermal field, N. California from a comparison

of GPS and leveling surveys, Geophys. Res. Lett. 24, 14, 1839- 1842.


Thermoelastic modelling of a backfilled salt mine B. Cailleau(1), D. Becker(2), T. Dahm(3), S. Shapiro(1), and D. Kaiser(4) 1 Free University Berlin, Institute of Geophysics, Berlin, Germany. E-mail:[email protected] 2 University of Hamburg, Institute of Geophysics, Hamburg, Germany. 3 GFZ Helmholtz-Zentrum Potsdam. Previously at (2) 4 Federal Institute for Geosciences and Resources (BGR), Hannover, Germany E-Mail: [email protected] The project investigates the deformational effect of cyclic thermal loading on rock surround-ing a salt mine cavity. A cavity in an abandoned salt mine in North Germany has been back-filled by salt concrete for consolidation. The process of backfilling was monitored with high resolution to track thermal expansion and contraction of the rock due to hydration of the fill-ing and solidifying material that leads to stress increase and decrease in the rock. To better understand the formation of cracks measured by acoustic emissions (AE), thermoelastic mod-eling has been developed. The presentation will describe the model set-up and show the pro-cedures to analyze the output for comparison with the data. The combination of AE data and numerical modeling allow better constraining the type, location and temporal occurrence of cracks. Noteworthy is a Kaiser effect: the onset of AE activity during a cycle can be predicted knowing the maximum stress of the previous cycle. However, there is deviation for late cycles due to e.g. macrocracking which would require further modelling including growing damage zones.


21

Investigation of Well Casing Damage Due to "Groundwater-Withdrawn Induced Compaction" Using Advanced Numerical Simulation Method R. Alkhamis and A. H. Haghi E-Mail: [email protected] Subsidence is a geologically hazardous phenomenon that can be aggravated by human such activities as long term water, oil, and gas extraction from underground resources or other min-ing activities. Abstractions from aquifers cause the aquifer-maintaining forces to lose their state of equilibrium whereby land starts to settle and ultimately subsides. One of the conse-quences of exploiting groundwater is the rising of well casings. Actually, it is not the well casings that rise but the ground around them that subsides. Subsiding forces can damage well casings and this can foist repair costs, or the wells might even need to be replaced by new ones. In this paper, finite element methods in two and three dimensional modes as well as ABAQUS software were used to investigate the impacts of subsidence from groundwater ex-ploitation and drawdown on well casings. This investigation was based on data obtained from a real basin. The results obtained through geo mechanical simulations using ABAQUS soft-ware have shown that the mechanism by which subsidence "bowls" are created cause two forms of rupture in the casings. At the edges of the bowl, surface sediments might slide on soft clay toward the center of the well and cause the casings to bend along their way. Around the center of the bowl something different happens. Considerable compaction of soft clay over time causes the distance between the surface and the bottom of the well to decrease, which results in the buckling of the casings and, in some cases, causes the casings to stick out of the ground.


Thermo-mechanical coupling during compaction of porous media Alireza Hassanzadegan and Oliver Kastner, GFZ Potsdam, Germany E-Mail: [email protected]

This research addresses the effect of thermally induced, intrinsic stress fields under external loading on the mechanical properties of porous rocks. During injection of cold water into a geothermal reservoir temperature and pore pressure within the reservoir changes which results in changes of stresses acting on the reservoir and surrounding rocks. A combination of tem-perature increase and pressure change would effect on thermo-mechanical behavior of the porous rock. In hydrostatic compression tests we observe a residual strain during load-ing/unloading cycles. The amount of residual deformation is an increasing function of tem-perature, while the corresponding elastic moduli of the sample exhibit an inversion effect: At low effective pressure, the bulk modulus decreases with temperature, while beyond a charac-teristic pressure, it increases. In our contribution we try to relate the thermo-mechanical be-havior to the effects of externally induced stresses and intrinsic thermal stresses. We employ an abstract FEM model to rationalize our hypothesis: a circular hole in a plate under thermal and mechanical loads. The change of pore shape is important since it influences in transport properties of porous rock such as porosity and permeability. The interrelation between ther-mal expansion coefficient and bulk modulus, and the path dependence of heat transfer pro-cesses govern the temperature effect on granular rock and changes in pore geometry.


23

Probing the Deep Rheology of Tibet: Constraints from the 2008 Mw 7.9 Wenchuan, China Earthquake Mong-Han Huang and Roland Bürgmann, University of Berkeley, USA E-Mail: [email protected] The time-dependent surface deformation after a large earthquake reflects the response to the redistribution of stresses induced by the earthquake and can be used to probe the viscous strength of the lithosphere. The 2008 Mw 7.9 Wenchuan earthquake occurred at the eastern flank of the Tibetan Plateau, and its postseismic deformation give us an opportunity to exam-ine the long lasting question of whether the growth of the Tibetan Plateau is by brittle crustal thickening or by lower crustal flow. We use ABAQUS to generate a 3D finite element model of viscoelastic relaxation with lateral heterogeneity applied for the calculation of the 1.5 years postseismic displacement. The inferred 35-km-thick lower crustal flow with viscosity of 4 ×1018 Pa s under eastern Tibet can explain bothspatial and temporal patterns of the postseismic displacement. A layered dislocation model is also tested for contributions of afterslip on the down-dip extension of the rupture or a shallow detachment. The best fitting results require more than a half meter slip below 25 km which might be below the brittle-ductile transition zone. In the SW Longmenshan, the lower crustal flow model can explain the near- to far-field deformation in space and time. In the NE, the lower crustal flow alone can-not produce the localize deformation observed there. Consequently, afterslip on the shallow part of the fault plane appears to contribute to the deformation along this strike-slip dominated portion of the rupture.

Figure 1. The Wenchuan coseismic displacement. The white arrows are the GPS recorded coseismic displacement. The colors in the triangles are the vertical displacements. The beach ball diagram shows the focal mechanism of the main shock, and the circles are the aftershocks color coded with depth. The white lines outline the coseismic surface rupture. The inset indicates the study area location at the edge of the Tibetan Plateau.


Estimates of stress changes from the 2010 Maule, Chile earthquake: the in-fluence on crustal faults and volcanos M. Keiding (1), O. Heidbach(1), M. Moreno(1), J. C. Baez(2), D. Melnick(3), and N. Kukowski(4) (1) Helmholtz Centre Potsdam GFZ German Research Centre for Geosciences, Germany, (2) University of Concepcion, Chile, (3) University of Potsdam, Germany, (4) Friedrich-Schiller-Universität Jena, Germany

E-mail: [email protected] The south-central Chile margin is an active plate boundary where the accumulated stress in the subduction interface is released frequently by megathrust earthquakes (Mw>8.5). The Maule earthquake of February 27 2010 affected about 500 km of the plate boundary produc-ing spectacular tectonic deformation and a devastating tsunami. A compilation of pre-, co-, and post-earthquake geologic and geodetic data offers the opportunity of gain insight into the processes that control strain accumulation and stress changes associated to megathrust events. The fore-arc deformation is primarily controlled by the stresses that are transferred through the locked parts of the plate interface and the release of stresses during megathrust events. During a great interplate faulting event, upper plate faults, rooted in the plate interface, can play a key role in controlling fluid pressurization. Hence, the hydraulic behavior of splay faults may induce variations of shear strength and may promote dynamic slip weakening along a crustal fault. Furthermore, the co-seismic stress transfer from megathrust earthquakes can severely affect nearby volcanos promoting eruptions and local deformation. InSAR and time-series of continuous GPS in the aftermath of the Maule earthquake show evidences of activation of the NW-striking Lanalhue fault system as well as pressure increase at the Antuco volcano. We build a 3D geomechanical-numerical model that consists of 1.8 million finite elements and incorporates realistic geometries adapted from geophysical data sets as well as the major crustal faults in the region. An updated co-seismic slip model is obtained based on a joint inversion of InSAR and GPS data. The model is used to compute stress changes in the upper plate in order to investigate how the Maule earthquake may have affected the crustal faults and volcanoes in the region.


25

Numerical studies of the deformation of salt bodies with embedded carbonate stringers Shiyuan Li, S. Abe and J. L. Urai, RWTH Aachen University, Germany E-mail: [email protected] Large carbonate or anhydrite inclusions are embedded in many salt bodies (so-called rafts, floaters or stringers) and these respond to the movements of the salt in a variety of ways, in-cluding displacement, folding and fracturing. The movement and deformation of those em-bedded carbonate or anhydrite bodies is a process which is not fully understand yet. No nu-merical study yet has investigated the brittle deformation of individual stringers during the initial phases of salt tectonics. We presented numerical models of deformation of salt body with embedded stringers using a case study from the South Oman Salt Basin. We investigated by Abaqus package (finite element models) how differential displacement of the top salt sur-face induces salt flow and the associated deformation of brittle stringers (including both brit-tle and viscous material properties). In our research, the main work was to use and develop techniques in Abaqus to make models of rheological behaviour of salt tectonics and brittle or ductile behaviour of carbonate stringers embedded in salt. A series of techniques were used in order to make successful models. We simplified the geometry from seismic data, define car-bonate or anhydrite and rocksalt material, model passive tectonic process through applying boundary conditions, model large displacement through adaptive remeshing technique and python script, model brittle fracture through iterative scheme for stringer breaking and adap-tive remeshing techniques. The simplified model offers a practical method to investigate complex stringer motion and deformation. Models suggest that brittle stringers can break very soon after the onset of salt tectonics. The extension can make brittle stringer to boudinage and fracturing and compression can make brittle stringer bending and thrusting. Models suggest that viscous stringers have folding and extension deformation. Results also show the internal structure of salt body and stringer fracturing or deformation are strongly dominated by the geometry or material properties of models.

References Urai, J. L., Schléder, Z., Spiers, C. J. & Kukla, P. A. 2008. Flow and Transport Properties of Salt Rocks. In: Littke, R., Bayer,

U., Gajewski, D. & Nelskamp, S. (eds) Dynamics of complex intracontinental basins: The Central European Basin Sys-tem. Berlin Heidelberg, Springer Verlag, 277-290.

Li, S., Abe, S., Reuning, L., Becker, S., Urai J.L. & Kukla, P.A. 2012. Numerical modeling of the displacement and defor-mation of embedded rock bodies during salt tectonics - a case study from the South Oman Salt Basin, in: Alsop, I. (eds), Salt tectonics, sediments and prospectivity. Geological Society Special Publication, 363: 503-520.

Figure 1: Brittle stringer deformation in a de-forming salt body in a compressive environment

Figure 2: Viscous stringer deformation in a deforming salt body in a compressive environment


Strain changes in an earthquake cycle: Kinematic FE-Models and GPS data from the great 2010 Maule Chile earthquake Marcos Moreno, John Bedford, Marie Keiding, Dietrich Lange and Oliver Heidbach E-mail: [email protected] The Maule earthquake of 27th February 2010 (Mw=8.8) affected about 500 km of the Nazca-South America plate boundary in south-central Chile producing spectacular crustal defor-mation and a devastating tsunami. This earthquake ruptured a megathrsut segment that has been monitored with a dense space-geodetic network and state-of-the-art seismic experiments before and after the event, providing not only one of the most detailed models of a subduction interface but also a high resolution image of the kinematic behavior of inter-, co- and postseismic deformations. Here, we use estimates of Earth’s surface displacements measured by a dense GPS network to constraint kinematic simulations of plate interface seismic and aseismic slip. Then, we explore the interrelation of the slip patchworks and compare them with aftershock seismicity and discuss the physical implications. We estimated slip dis-placements (dip-slip and strike-slip) along a curved fault by using an inversion method based on FEM-generated Green's functions. Our spherical FE-model incorporates topography and bathymetry data, as well as the geometry of the plate interface and the continental Moho, both derived by combining geophysical data. The fault plane is a deformable contact surface con-sisting of quadrilateral elements whereas the rest of the mesh is composed of tetrahedral ele-ments. The use of a finite element model that introduced the main geometrical complexities of the Chile subduction zone also allowed us to compare the spatial relation of slip patterns with major tectonic features of the forearc.


27

A damage model for the Terceira Rift origin Neves, M.C., Luis, J.F., Lourenço, N., Miranda, J.M. E-Mail: [email protected] We present a model aiming to provide a consistent explanation for the Azores Linear Volcan-ic Ridges (LVRs) growth and Terceira Rift birth. Admitting that the transport of magma through the brittle part of the lithosphere occurs via fractures we try to determine the most likely paths of fracture development. For that we build a 3D representation of the brittle litho-sphere and underlying ductile mantle, driven by plate boundary forces applied at their edges. The brittle layer is governed by an elasto‐plastic rheology with progressive damage where fractures are assumed to be analogous to localized shear bands. The ductile mantle underneath is modeled as a viscoelastic layer. The base of the elasto‐plastic plate is assumed to roughly coincide with the brittle‐ductile transition defined by the 400°C isotherm. Two geometries are modeled representing two moments of the Azores history. The first simulates the tectonic setting at the beginning of the Azores triple junction migration, shortly after the East Azores Fracture Zone (EAFZ) ceased to work as a transform fault. The second incorporates the pre-sent‐day Mid Atlantic Ridge (MAR) segmentation and represents the plate geometry in the last few Ma. Apart from the MAR segmentation the geometry of the models is identical and contains two parts, Eurasia and Africa, which interact through the EAFZ and the Gloria fault. The stress distribution and the damage patterns are predicted using the Abaqus/explicit finite element program. Numerical instabilities are typical of elasto‐plastic models subjected to lo-calization processes, resulting in non‐unique solutions that are mesh dependent and have no physical meaning. To suppress such instabilities and overcome the numerical difficulties we take on two main options. First, we neglect the lithostatic state of stress and the gravitational body force. The advantage of the model being 3D is that the mantle flow is calculated explic-itly but the lithosphere is in fact deforming under approximately plane stress. This assumption implies that we can only consider depth‐independent yielding. Second, the damage evolution law parameters are artificially adjusted to guarantee the stability of the solution. The approach taken here is to adjust a small number of damage parameters in order to produce convergent solutions and modeling results that fit the observational constraints. Considering damage zones as analogues to LVRs we propose an evolutionary model for LVR development that ultimately leads to the creation of the Terceira Rift. The predicted damage zones are in good agreement with the predominant directions of LVRs and faulting found in the Azores and match the rotation of principal stresses along the Terceira axis, as supported by morphological observations and earthquake focal mechanisms. The modeling results thus confirm previous tectonic interpretations in which magma transport in the Azores region is mainly controlled by lithospheric architecture and tectonic stress. This work is a contribution to PTDC/MAR/108142/2008 MAREKH research project. Pest‐ OE/CTE/LA0019/2011 ‐ IDL


Stress Field Model of the Alberta Basin versus in-situ Stress Data Karsten Reiter, Oliver Heidbach, Inga Moeck, GFZ Potsdam, Germany E-mail: [email protected] The recent stress field in the Earth's crust is an issue related to earthquake hazard, understand-ing of current tectonic processes as well in a more applied topic, in the context of the energy supplying industry. The World Stress Map project (Heidbach et al., 2009) delivers stress in-formation about the Earth's crust from plate scale first-order stress field (plate boundary forces) over second-order intra-plate stress field (mountain belts, post-glacial rebound), up to local third-order stress field variations. An extension of the WSM database, using magnitudes of in-situ stress data, the Q-WSM (leak-off tests, hydro-fracturing so on) is under construction (Zang, Stephansson, Heidbach, & Janouschkowetz, 2012). However, for the most regions, both databases have loose date distribution, which could not be used to dedicate third order questions about the stress field variations and they still provide point wise information of parts of the six independent components of the stress tensor. A larger region is never been explored or modelled, based on a very dense network of quantitative in-situ stress measure-ment (magnitudes and stress orientations). Many studies still researched regional stress field pattern in the Alberta Basin, which contribute understanding of first- and second-order fea-tures (e.g. Bell, Price, & McLellan, 1994). The Alberta Basin in general is part of the Western Canadian Sedimentary Basin, which is a retro-arc foreland basin, caused by accretion of mi-cro-continents, developing the Rocky Mountains as part of the Cordillera. The risen curved fold-and-trust belt (FTB) is convex toward the sedimentary basin. The basin has a wedge shape, thickest sediments close to the FTB with about 5500 m, sediments tapers out at on the Canadian Shield. A large scale crustal and upper mantle 3D-geomechanical-numerical model of the Alberta Basin and the surroundings is constructed to describe continuously the full stress tensor. Each stress field model has to be tested against model-independent constraints, like GPS measurements, fault slip rates or others. In-situ stress measurements are the most likely data, because they deliver the most direct informations of the stress field and they pro-vide insights into different depths, a major benefit compared to usually indirect surface infor-mations. For this study, a very dense distributed dataset is available from the Alberta Basin and surrounding. The data base contains data from leak-off tests and hydro-fracturing, con-tributing minimum principal stress values, as well as vertical principal stress values and quali-tative values from borehole breakouts. A new work flow will be tested during this study, from the generation of the geometrical model via geological modelling software gOcad, allowing incorporation of a great variety of geological and geophysical data. After construction, the model has to be meshed with finite element using HyperMesh. The numerical modelling code Abaqus allows testing of initial body forces and different boundary conditions. Finally the visualization tool Tecplot 360 allows directly comparison between model outcome and meas-ured data. This workflow has little restrictions, related to variously input data, resolution, comparison methods, and possibilities of data output and transfer. References Bell, J. S., Price, R. A., & McLellan, P. J. (1994). Chapter 29 - In-situ stress in the Western Canada Sedimentary Basin. In G.

D. Mossop & I. Shetsen (Eds.), Atlas of the Western Canada Sedimentary Basin (pp. 439-446). Alberta: Canadian So-ciety of Petroleum Geologists and the Alberta Research Council.

Heidbach, O., Tingay, M., Barth, A., Reinecker, J., Kurfeß, D., & Müller, B. (2009). The World Stress Map Based on the Database Release 2008, equatorial scale 1:46,000,000. Geochemistry Geophysics Geosystems (Vol. 4, p. 1). doi:doi:10.1594/GFZ.WSM.Map2009

Zang, A., Stephansson, O., Heidbach, O., & Janouschkowetz, S. (2012). World Stress Map Database as a Resource of Rock Mechanics and Rock Engineering. Geotechnical and Geological Engineering, 22. doi:10.1007/s10706-012-9505-6


29

Initial Stress State in Crustal Stress Field Models Karsten Reiter, Oliver Heidbach, Tobias Hergert E-mail: [email protected] The recent stress field in the Earth's crust is an issue related to the understanding of earth-quakes, tectonic processes in general as well as reservoir and drill hole mechanical issues. Geomechanical-numerical models allow 3D estimations of the crustal stress state. The initial stress state of the model is a key issue, when gravity is applied to the model, before simula-tion starts. Afterwards comparison of the model output with model-independent constraints, like GPS measurements and fault slip rates are important tests. In-situ stress measurements are the most likely test data, because they deliver the most direct information besides in 3D. A large scale crustal and upper mantle 3D-geomechanical-numerical model of the Alberta Basin and the surroundings is constructed from to describe continuously the full stress tensor. Dif-ferent initial body forces and boundary conditions will be tested against hundreds of in-situ stress measurements form the Alberta Basin and surroundings in order to derive a best-fit model. From the general assumption, crustal stress is comprised from tectonic stress and the reference stress. The first is applied to the model as constraints, (body) forces or pressure, whereas the reference stress or the so called initial stress state ensures that only small defor-mation occur, when gravity is applied to the model. Sheorey (1994) derive a spherical analy-sis without any topography, faults or tectonics. The found stress ratio k for the Young modulus E is described by the following equation: k = 0.25+7 E (0.001 + (1/depth)). The out of it calculated stress ratio curves are confirmed by data from deep wells like the KTB. To ensure balance of the gravity forces to the horizontal stress, the model is firstly embedded in a laterally and vertically extended model frame with conical shape, which would converge in the earth centre. The laterally extended part has a very high Poisson Ratio whereas the exten-sion below has a very high Young modulus; the boundary constraints allow laterally only ver-tical movement and only horizontal movement at the bottom. Gravity is applied to the ex-tended model and stress ratio curve is checked. In the second step, stress from the boundary to the actual model is applied together with gravity to the real model alone, iteratively until only a few settlement occurs. The stress curve is double checked until success; otherwise proper-ties of the model frame from the first step are varied. References Sheorey, P. R. (1994). A theory for In Situ stresses in isotropic and transverseley isotropic rock. International

Journal of Rock Mechanics and Mining Sciences Abstracts, 31(1), 23–34. Elsevier. Retrieved from http://www.sciencedirect.com/science/article/pii/0148906294923124


Using ABAQUS for glacial isostatic adjustment modeling and stress analysis Patrick Wu, Holger Steffen, Rebekka Steffen, University of Calgary, Canada E-Mail: [email protected] ABAQUS has been a powerful tool in glacial isostatic adjustment modeling. However, it can-not be used as is. Mainly designed for engineering applications, some equations such as the stiffness equation are not adequately solved for geoscientific problems. Wu (2004) provided a detailed description for glacial isostatic adjustment (GIA) investigations when incompressibil-ity is assumed and a non-self-gravitating earth is considered. We will review these methods and show how to implement them in ABAQUS. Furthermore, we will discuss how to correct the stress in GIA investigations. This poster will complement the talk by Björn Lund and the poster by Rebekka Steffen et al. References Wu, P., 2004. Using commercial Finite element packages for the study of earth deformations, sea levels and the

state of stress, Geophys. J. Int. 158(2), 401-408, doi:10.1111/j.1365-246X.2004.02338.x.


31

Implementing faults in glacial isostatic adjustment models using ABAQUS R. Steffen, P. Wu, H. Steffen, D.W. Eaton E-mail: [email protected] Flexural stresses induced in the lithosphere during glaciation, are released near the end of deglaciation as earthquakes along pre-existing faults. In order to get a better understanding of the relationship between glacial loading/unloading and seismic activity, a model for glacial isostatic adjustment (GIA) is extended by including a fault structure. Solving this problem involves the development of three models using the finite-element method. The results show stable conditions along the fault during glaciation and deglaciation. After the end of the deglaciation period, the fault starts to move, and fault offsets of up to 36m are obtained for a fault at the centre of the ice sheet. In the following time, seismic activity along the whole fault but also only in the lower part is observed. Seismic activity decreases and stops after 5,000 to 6,000 years using a glaciation history similar to northeastern Canada.


User-defined subroutine for non-linear mantle deformation in postglacial rebound Wouter van der Wal, TU Delft, Netherland E-Mail: [email protected] The effect of stress-dependence (dislocation creep) in models for postglacial rebound has been shown to improve fit with observations compared to models with diffusion creep only. Because diffusion and dislocation creep can occur simultaneously in mantle materials, they both need to be implemented in postglacial rebound models. A uni-axial creep law in Abaqus can simply be implemented in the user-defined subroutine CREEP. Individual stress tensor elements are not available in this routine but the von Mises equivalent stress is. However, it is not immediately obvious that individual strain rate elements can be derived from it. Also it might not be clear how the CREEP routine can be extended in case of a large number of ele-ments each with different creep properties. Here the CREEP subroutine is tested in a simple axisymmetric model and combined with another subroutine to load an external database. The CREEP subroutine with a non-linear flow law with stress-exponents 1 and 3 is tested against the standard creep laws in the material model definition (option *CREEP, LAW=STRAIN) for an axisymmetric model. Differences in von Mises stress and displacement are negligible. This suggests that also the combination of diffusion and dislocation creep can be correctly implemented with the CREEP subroutine. This subroutine is subsequently used in a 3D model for a spherical Earth subjected to glacial loading. In order to implement lateral varying creep parameters the creep parameters are computed from olivine flow laws with local pressure and temperature derived from seismology and stored to a file. The user subroutine UEXTERNALDB reads from this file the properties for each element which are then used in the creep subroutine to compute the strain increment for the same element. Creep parameters can have very small numerical values, therefore future work should include tests for numeri-cal stability. For more realistic stress-dependence access to individual stress tensor elements is required, which can be implemented using the user subroutine UMAT.


33

Salt flow during basement extension and differential loading – viscoelastic deformation in analogue and numerical models Michael Warsitzka, Jonas Kley and Nina Kukowski E-mail: [email protected] In many geological simulations, rock salt can be considered as a pressurized fluid between rigid basement and rigid or elasto-plastic cover sediments. Within this system, salt flows in response to hydraulic pressure gradients due to differences in elevation or differential loading. In order to investigate flow regime and structural development during early salt diapir evolu-tion, we combine analogue experiments and numerical modelling. In a first step, we used dy-namically scaled sandbox experiments containing viscous silicone putty (Newtonian behav-iour) and brittle granulate obeying Mohr-Coulomb failure criterion. Basement displacement and syn-kinematic sediment accumulation were applied to trigger viscous flow. In a subse-quent 2D numerical modelling study conducted in ABAQUS/Standard, we intend to (1) re-produce and validate analogue models using equal material parameters and conditions, (2) investigate flow kinematics and structural evolution including conditions of natural salt basins (3) evaluate the stresses acting during salt flow. The numerical model set-up basically in-volves three compartments: a rigid basement overlain by a viscoelastic layer simulating sili-cone or salt, respectively, and an elastoplastic layer, which mimics the brittle cover. Deforma-tion is either purely gravity-driven or due to definition of a displacement boundary condition at the rigid basement. Advanced numerical simulations will include differential loading result-ing from syn-kinematic sedimentation and sediment compaction. At first, numerical models will be spatially restricted to a single salt structure, while subsequent models should resolve regional salt flow on a basin scale. Analogue experiments show that bulk salt flow reflects an interplay between downward flow driven by basement subsidence and upward flow driven by sedimentary differential loading. First preliminary results of numerical experiments reveal that the overall flow regime in the viscoelastic material is similar to that observed in analogue experiments.


New 3D Finite Element models of the central Andes using realistic geometry Zeumann, St., Sharma, R., Jahr, T., Jentzsch, G. E-Mail: [email protected] In the frame of the German research project „Mass transport and mass distribution in the sys-tem Earth“ (SPP 1257, funded by the German Research Soc.) we concentrate on a better un-derstanding of the geophysical processes in the South American subduction zone. Therefore, high resolution geodynamic 3D models are developed by the Finite Element Method (FEM) with ABAQUS. The geometry plays an important role for the processes along subduction zones. Thus, models of the central part of the Andes are developed with geometries taken from a well constrained gravity model. The models consist of 16 different units with various parameters e.g. densities, Young’s modulus, etc. For different model rheologies (elastic, vis-cous-elastic, etc.) the stress, strain and deformation field is calculated to investigate the effect of viscosity and plasticity. The comparison between the resulted stress field and earthquake regions shows that the distribution of earthquakes is strongly correlated with stress field pat-terns along the Andes. The horizontal deformation field fits the observed GPS data. Common-ly for subduction zone modelling is to fix the edge of the upper plate and apply velocity to the subducting plate only. However, various investigation results from different groups show the importance of the horizontal upper plate moving for developing high topography like the An-des. Therefore, we investigate this effect to our models.


35

Abstracts in the order of the schedule Thursday, August 23rd

09:30 - 10:15 Tobias Hergert (invited speaker) Modelling the state of stress in the Earth's crust

10:15 - 10:45 Johannes Altmann Geomechanical-numerical model of The Geysers geothermal field


11:15 - 12:00 Jens Nüchter (invited speaker) Complex states of stress during the seismic cycle around major reverse, normal and strike-slip faults

12:00 - 12:30 Oliver Heidbach Fast visualization of 3D stress data from the Abaqus© odb-file using Tecplot 360© and the geostress add-on

12:30 Final remarks


Modelling the state of stress in the Earth’s crust Tobias Hergert and Oliver Heidbach E-mail: [email protected] The state of stress in the Earth’s crust is an issue in understanding and assessing natural haz-ards and geotechnical problems. The crustal state of stress is modelled by incorporating the major sources of stress such as gravity and plate tectonic forces and by accounting for stress perturbations arising from faults and spatially varying material properties. Technically, a crus-tal reference stress state is established using the mechanism of compaction, on which the plate tectonic forces are applied. The model set-up and approach pursued are presented for a geomechanical model of northern Switzerland, where low deformation hampers the detecta-bility of faults and the assessment of their kinematic behavior and of their influence on the stress field. The modelled displacement field provides the sense of slip on particular faults and the relative importance of the faults is inferred. Probable fault extensions over hitherto known fault lengths are proposed based on modelled vertical displacements. Fault-related perturbations of the regional stress field can be identified based on the modelled stress field. Stress perturbations depend on the amount of slip on a fault, the coefficient of effective fric-tion on a fault, the vicinity to a fault, the geometry of a fault and the tectonic regime. The role of a potential decoupling horizon below the sedimentary wedge is increased uplift and a more compressional state of stress in the sediments.


37

Geomechanical numerical model of The Geysers geothermal field Johannes B. Altmann1, Oliver Heidbach1, Roland Gritto2

1 Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences 2 Array Information Technology E-mail: [email protected] Knowledge of the absolute 3D stress state and its spatio-temporal evolution due to fluid injec-tion and depletion is critical for our understanding of enhanced geothermal systems. The poroelastic coupling and the thermo-mechanical process during hot water extraction and cold water injection are the key processes that contribute to the stress changes in geothermal reser-voirs. Due to massive heat extraction in The Geysers geothermal field, California, during the 80’s, reservoir pressure dropped down. In order to stabilize the reservoir pressure and thus productivity, pipelines were constructed to pump the average amount 1500 liter of wastewater per second into the reservoir. In the following time, M > 4 events occurred more frequently. This raises the question whether those events are man-made or natural. In addition, The Geysers geothermal reservoir is located in an active tectonic region in between the Collayomi and Maacama fault systems. Thus, also the regional tectonics in this area influence the state of stress and lead to stress changes. In order to account for both effects of stress changes (man-made due to fluid injection and extraction, tectonics), we set up a 3-D geomechanical numerical model with poroelastic rheology of The Geysers geothermal field. Besides the geometry of the geo-thermal reservoir, the geomechanical numerical model contain also the major fault system and rupture planes of M > 4 earthquakes occurred after the start of the wastewater injection project. We use the Finite Element Method to solve the resulting partial differential equations by the commercial software Abaqus. The geomechanical numerical model enables us to investigate stress changes due to both the tectonic setting of this area and the operation (water injection and production) of the geothermal field. The model quantifies the 3-D absolute stress state as well as its spatio-temporal changes due to water injection and production, tectonic loading and occur-ring earthquakes. We are able to differentiate between those processes that lead to spatio-temporal stress changes, thus to calculate which process has the major effect on the state of stress. To describe the effect of opening cracks during fluid injection, and thus fluid diffusing much faster in these directions, we make the permeability dependent on the pore pressure. This we have implemented via a user subroutine in Abaqus. Here, we show the model setup, first results of the modeling of the stress field of The Geysers are as well as the idea behind linking of permeability and pore pressure.


Complex states of stress during the seismic cycle around major reverse, normal and strike-slip faults Jens-Alexander Nüchter and Susan Ellis E-Mail: [email protected]

In seismically active regions, the crust and the mantle are subjected to stress cycles controlled by repeated earthquake activity. Determining the spatial and temporal changes in stress field magnitude and geometry during the seismic cycle is essential in assessing seismic risk and the stress build up on a particular fault. Most estimates for the magnitude of stress disturbance during the seismic cycle have been based on seismological observations, which cannot con-strain stress changes from regions below the fault that do not radiate seismic energy. Inver-sion models on transient postseismic surface deformation provide information about the proc-esses beneath a fault, though the results are generally non-unique, and most studies focus on processes active in the mantle. The structural record of metamorphic rocks exhumed in seis-mically active regions provides detailed insight into the imprint of postseismic processes and conditions active beneath a seismically active fault with unmatched resolution, but informa-tion is typically restricted to a single outcrop. We use numerical models constrained by the rock record to address the critical information gap regarding stress changes near the lower front of an active fault on a larger scale. The models consist of a 300 km wide and 30 km thick crust underlain by a 70 km thick mantle. The strength of the crust is limited by fric-tional-plastic yield described by the Mohr-Coulomb criterion for optimally oriented slip. Where temperatures are sufficiently high, thermally activated creep described by a flow law for quartz relieves stresses to below frictional yield. The strength of the mantle section is de-scribed by a flow law for olivine. Throughout the experiments, the models are loaded by grav-ity and appropriate kinematic boundary conditions to simulate an extensional, compressional or transpressional tectonic regime. After an initial loading step, a series of seismic cycles with constant recurrence intervals are simulated. The model faults are represented by straight con-tact surfaces that reach down to the base of the crust. Earthquakes are triggered by a short term reduction in fault friction to values below the internal friction of the crust, so that the contact algorithm self-determines the depth to which the fault slips. The postseismic periods are simulated by setting the fault friction to high. The numerical results are consistent with coseismic stress changes determined from the rock and from seismological data. Coseismic stress redistribution is accompanied by a deflection of the principal stresses within a wide region in the crust. During the postseismic period, all stress tensor components recover to-ward their pre-earthquake state. This causes a gradual restoration of the magnitude and the orientation of the principal stresses, and progressive reloading of the fault. Stress transfer from the middle into the upper crust by localized postseismic creep has a similar importance for reloading of the fault as the kinematic boundary conditions. The numerical results imply that the understanding of the processes active beneath a fault is crucial for the estimation of stress evolution in the upper crust during the seismic cycle.


39

Fast visualization of 3D stress data from the Abaqus© odb-file using Tecplot 360© and the add-on Geostress Oliver Heidbach and Dietrich Stromeyer, GFZ Potsdam, Germany E-mail: [email protected] A time-consuming step during the assessment of 3D geomechanical-numerical model is the visualization of the 3D stress tensor. To accelerate this process one can use the software Tecplot 360 that directly imports the binary odb-output file from Abaqus. However, the standard scalars and vectors provided by Abaqus are often not the ones needed in Geoscienc-es. Thus, we developed an add-on for Tecplot 360 that calculates for all nodes in each time step the desired values from the six independent components of the stress tensor given by Abaqus. We present which stress values are available and show the efficiency of this add-on in terms of computation time. In particular these values are: Orientation (trend, plunge) and magnitude of the principal stresses, differential stress, minimum and maximum horizontal (SH and Sh) and vertical stress (Sv), mean horizontal stress (SHmean), the stress ratio k (k=SHmean/SV), the tectonic index, the Regime Stress Ratio (RSR) which is a contiuous value from 0-3 for the stress regime (Simpson 1997) and the Fracture Potential that is defined as the ratio between the differential stress in the model and the critical differential stress at failure with given values for the Mohr-Coulomb failure criterion (Connolly and Cosgrove, 1999). Once the new values are calculated they can be displayed on any surface cutting the model. The surfac-es can be draged on the fly through the model. Furthermore, the values on the appropriate surfaces can be exported as xyz value-tables in ASCII format for further processing or combi-nation with other geo-data, e.g. using GMT. The real advantage is, that these surfaces must not be planar, but can have irregular shape and it is also possible to e.g. import 3D fault struc-tures as surfaces and map the values onto these. The visualization is not limited to surfaces, but one can display the values also e.g. with 3D iso-surfaces or along curved borehole trajec-tories.

Fig.1: Left: Dialog box shows values that are provided and calculated from the geostress add-on within Tecplot 360 for every node in the mesh in each time step. Middle: Figure shows the distribution of the stress regime in a contiuous scale from normal faulting (0) via strike-slip (1.5) and thrust faulting (3) on a horizontal plane at 1 km depth. Rigth: Orientation of the maximum horizontal stress. SH. References Simpson, R. W., (1997): Quantifying Anderson's fault types. J. Geophys. Res. 102, 17909-17919. Connolly, P. und Cosgrove, J., (1999): Prediction of static and dynamic fluid pathways within and around

dilational jogs. In K. J. W. McCaffrey, L. Lonergan undJ. J. Wilinkson, eds., Fractures, Fluid Flow and Mineralization, Volume 155: Special Publications: London, Geological Society, 105-121.

40 3rd World Stress Map Conference, 15.-17. October 2008 in Potsdam

List of participants First Name Name Institute Country E-Mail

Nastaran Abdolmaleki University of Tehran Iran [email protected]

Johannes Altmann GFZ Potsdam Germany [email protected]

Jonathan Bedford GFZ Potsdam Germany [email protected]

Sierra Boyd University of Berkeley USA [email protected]

Jeremy Brown University of Stanford USA [email protected]

Beatrice Cailleau Free University of Berlin Germany [email protected]

Dave Coblentz Los Alamos National Laboratory USA [email protected]

Harvey Goodman Chevron Houston USA [email protected]

Amir Haghi National Iranian Oil Company Iran [email protected]

Alireza Hassanzadegan GFZ Potsdam Germany [email protected]

Oliver Heidbach GFZ Potsdam Germany [email protected]

Tobias Hergert Karlsruher Institute of Technology Germany [email protected]

Mong-Han Huang University of California, Berkerly USA [email protected]

Oliver Kastner GFZ Potsdam Germany [email protected]

Marie Keiding GFZ Potsdam Germany [email protected]

Shiyuan Li RWTH Aachen University Germany [email protected]

Björn Lund University of Uppsala Sweden [email protected]

Jens Lüke Simulia Germany CSE Germany [email protected]

Fabian Magrie GFZ Potsdam Germany [email protected]

Georgius Maniatis Leibniz-Universität Hannover Germany [email protected]

Marcos Moreno GFZ Potsdam Germany [email protected]

Maria C. Neves University of the Algarve Portugal [email protected]

Maria-K. Nikolinakou University of Austin USA [email protected]

Jens Nüchter University of Bochum Germany [email protected]

Ove Stephansson GFZ Potsdam Germany [email protected]

Simona Pierdominici GFZ Potsdam Germany [email protected]

Karsten Reiter GFZ Potsdam Germany [email protected]

Sebastian Rehde University of Freiberg Germany [email protected]

Andreas Schenk Karlsruhe Institute of Technology Germany [email protected]

Martin Schoenball Karlsruhe Institute of Technology Germany [email protected]

Sebastian Specht University Potsdam Germany [email protected]

David Steedman Los Alamos National Laboratory USA [email protected]

Rebekka Steffen University of Calgary Canada [email protected]

Holger Steffen University of Calgary Canada [email protected]

Heidi Turpeinen Leibniz-Universität Hannover Germany [email protected]

Wouter van der Wal Delft University of Technology Netherlands [email protected]

Wouter van der Zee Baker Hughes Netherlands [email protected]

Michael Warsitzka University of Jena Germany [email protected]

Jeoung Seok Yoon GFZ Potsdam Germany [email protected]

Arno Zang GFZ Potsdam Germany [email protected]

Stefanie Zeumann University of Jena Germany [email protected]

Date post:	07-Sep-2018
Category:	Documents
Upload:	doxuyen
View:	216 times
Download:	0 times

3rd Geoqus Conference - Ruhr-Universität Bochum · Simulating the Response of Faulted Geology with...

Documents