Alexander Raith


Alexander Raith

PhD Student

0241 8098439
Office location:
Lochnerstra├če 4-20

PhD Project


Structure and tectonic evolution of bischofite inside the Veendam salt pillow in the Netherlands.

The aim of this work is to understand the structural evolution and the role of soft K-Mg salts imbedded in rocksalt, and to study the mechanical effects of this high mechanical heterogeneity on salt deformation. The complex, internal geometries of salt structures in the Dutch Zechstein are related to long-term creep and complex folding of layered evaporites. To understand the interaction of mechanically strong anhydrite and rocksalt layers with the much weaker K-Mg salts, with a viscosity contrast of up to five orders of magnitude, thereby is of high interest. Our study area is the Veendam Pillow, a large Zechstein salt structure on the southern Groning High. The salt structure is used for Mg-rich bischofite (MgCl*6H2O) squeeze and solution mining. For our structural study we use high resolution interpretations of industrial 3D seismic data and well logs as well as core data for chemical and microstructural analysis of the Z3 and Z4 sequences. The 3D seismic interpretation reveals a good overview of the internal salt structure due to the sonic contrast between anhydrite, halite and K-Mg salts. Thickening of the soft Z3 salts in the crest of the saltdome and strong folding of the Z3 anhydrite is visible. On basis of our results, we designed numerical simulations using the finite element package ABAQUS to estimate the displacement field of K-Mg salts during tectonic movement. First results revealed that brittle, mechanically strong anhydrite layers have a high impact on the surrounding softer salts by folding with amplitudes of several hundred meters, and that the K-Mg salts are easily squeezed and fold on a much smaller scale.


  • How did the K-Mg Salts accumulate in such great quantity?
  • What is the K-Mg salt structure in a sub-seismic scale?
  • How does a system of complex layering and folding of creeping and brittle materials evolve? 


                                     Picture_1954 - Alex_Workflow_proj

Seimic Interpretation:

  • 3D seismic interpretation of salt layers and the overburden
  • Horizons in the overburden are used to identify phases of sediment down-building and the formation of the salt structures

Core Data

  • Sub seismic geometries
  • Important to evaluate the geo-mechanical models

Finite Element Modeling:    

  • Numerical Finite Element (FE) modeling for conceptually simulating the structural evolution of the layered evaporite sequences in the Vendam Pillow.
  • Based on seismic interpretation (large-scale geometries, layer thicknesses, timespan of the salt movement) and core data (detailed sequences and layer thicknesses, rheologies)

Study Area

Picture_1955 - Alex_Seisoverview_proj

  • Rare location
  • Good dataset (3D seismic, wells and core)
  • Used for magnesium mining
Have a look at related Projects!



B.Sc. thesis:

Creation of a high resolution panorama, capturing a high density calcite vein network from a polished limestone outcrop in the Oman Mountains

See also

M.Sc. thesis:

The evolution of fault zones in brittle-ductile layered rocks.
A parametric study focussing on cohesion and fault angles using Discrete Element Modelling(DEM)

The development of normal faults in brittle ductile layered rocks is modeled in this M.Sc. thesis using Discrete Element Model (DEM) simulation. The models consist of one cemented layer inside a cohesionless granular material above a basement fault in a gravity field. I varied the cohesion of the cemented layer and the angle of the basement fault and also experimented with different packings of the material. Results show, as expected, that two structural domains exist, a graben domain and a precursor domain. In both of these domains, the increase in cohesion of the hard layer produces large differences in the structural evolution. The main parameter that determines the amount of tectonic abrasion in the fault zones is the cohesion of the brittle layer. This leads to a gradual thinning of the layer with low cohesion and development of blocks and fragments in case of a relatively high cohesion. Thus, continuity of the sheared layer is higher in the rocks with low cohesion. The structural domain also affects the continuity of the brittle layer: in the precursor domain the brittle layer is more continuous than in the graben domain.

The Discrete Element Method (DEM)
- Discrete Element Method (DEM) enables modelling of brittle fracture
- Spherical particles interacting with nearest neighbors
- Particle movement calculated by Newton‘s law
- Brittle-elastic or quasi-viscous material
- Using parallel DEM software ESySParticle

The composition is based on a two dimensional 30 cm x 40 cm sandbox setup, with one brittle layer sandwiched inbetween cohesionless layers.

Picture_1730 - MSC_Raith_Setup1Picture_2383 - MSC_Raith_setup2

Simulation process
Step 1: The box is filled with grains
Step 2: The grains subside in a gravity field
Step 3: The grains in the middle layer and at the walls are cemented
Step 4: One side of the sandbox is moved along the basement fault

Picture_1734 - MSC_Raith_Simsteps


Picture_2381 - MSC_Raith_Results1Picture_2382 - MSC_Raith_Results2


The DEM simulations give a great opportunity to investigate all kinds of geological features related to normal faulting in mechanically layered rocks. The varied parameters have an important influence on the fault geometry:
1. The structural domain has a strong influence on the fault geometry. The basement fault influences the shearing angles inside the layered rocks. This means that the fault geometry inside the layered rock not only depends on its own mechanical parameters, but is also strongly influenced by the tectonic environment.
2. The continuity of the sheared layer is determined by the cohesion and the basement fault angle. The discontinuity of the sheared layer for the same mechanical contrast depends on the structural domain. The 70° experiments show a consistently higher continuity in the sheared layer than the experiments with a 55° dipping basement fault. With increasing cohesion, the heterogeneity of deformation inside the shear zone is rising, which increases the likelihood of ruptures in the brittle layer. Concentrated strain, antithetic movement and extension in the graben domain lead to a higher discontinuity compared to the precursor domain faults.
3. The position of the first fractures in the experiments is independent of the fault domain and the cohesion for the modeled basement fault angles of 55° and 70°. In fact, cohesion determines the width of the fractures, which leads to wide fracture zones in a low cohesive layer and focused fractures in a highly cohesive layer.
4. The position of the second generation of fractures is influenced by the position of the first generation fractures, the cohesion and the basement fault angle. If the cohesion is low, no new fractures appear; the brittle layer is just eroded. If the cohesion is on an intermediate level, the brittle layer is abraded and breaks apart into small blocks, until an equilibrium position is reached, which is depending on the basement fault angle and the Coulomb angle of the cohesionless material. In case of a highly cohesive brittle layer, the position of the second-generation fractures is depending on the position of the first-generation fractures relative to the final shear angle. Subsequently, if the brittle layer is reaching too far into the hanging wall to resist the shear stress, it will break at a location inside the footwall. The block size then depends on the cohesion.
5. The fault zone is not one single shear band; it is rather a zone of parallel shear bands, which are active at the same time. These shear bands can change their positions and activity. Most of the strain is located in dominating shear bands. With increasing cohesion, the position of these shear bands is more and more dominated by the fracture position in the brittle layer. This leads to a strong influence of the initial fracture position on the resulting fault zone width for highly cohesive layers.

See also: Sandbox Experiments