Reservoir Geomechanics
A variety of industry oriented research projects related to the geomechanical behaviour of producing reservoirs have been and are currently performed by the Formation Physics Department at SINTEF Petroleum Research. The research is mainly directed towards the mechanical response, i.e. reservoir compaction, surface subsidence, and consequences thereof. Effort is also put into investigations of geomechanical effects on fluid drainage; such as compaction drive, and stress effects on porosity and permeability. Rock physics and rock mechanics aspects of reservoir monitoring represent an important part of our work on dynamic reservoir characterization.
Our research utilizes experimental, numerical and analytical tools. In the Formation Physics laboratory, we can study simultaneously the mechanical, hydraulic and acoustic behaviour of core samples. In addition, special tests, like sand box experiments can also be performed. We use various numerical techniques, like discrete particle (PFC from Itasca) in addition to displacement discontinuity and finite element modelling to simulate the behaviour of the reservoir from particle to field scale. In-house elastoplastic and rock physics based algorithms are applied in order to provide simple and user-friendly solutions for the clients. A special application has been the incorporation of a constitutive model for the poromechanical behaviour of chalk with water saturation effects.
A few examples of our activities within reservoir geomechanics are given below:
Core damage effects on compaction
A long-term research (funded by Shell / NAM, and previously also by Norsk Agip, Security DBS, and Elf) has identified stress release during coring as an important source of mechanical damage to the core material. This leads to reduced stiffness and strength from laboratory measurements on cores. The research has focused on quantification of core damage effects, and on the development of new laboratory techniques and a field coring tool to reduce core damage.
Synthetic sandstone has been formed under stress in the laboratory, permitting quantitative comparison between the compaction of a virgin material and an unloaded -reloaded simulated core. Discrete particle modelling supports the findings from the laboratory experiments. It was found that initial reservoir compaction may be overestimated by a factor of 2 or more using uncorrected core data. Models are developed in order to correct core compaction data for core damage.
 | Disked core of a synthetic sandstone manufactured under stress and then reloaded by simulating the stress path followed during coring. |
Trap door experiments to simulate the response of the overburden to reservoir compaction
This research (funded by Norsk Agip and Statoil) showed possible strain localization and development of surface subsidence as a function of trap door displacement, simulating reservoir compaction. Features like time-delayed subsidence and the role of stress arching are evident in these experiments.
 | Picture from a research project sponsored by Norsk Agip and Statoil where the trap door model experiment was used to study the consequences of reservoir depletion and compaction on the reservoir stress path and on surface subsidence. Sand with color layers simulates the overburden formations and a retracting trap door under the sand the compacting reservoir. |
Particle scale reservoir mechanics
Discrete particle modelling (using PFC - Particle Flow Code -from Itasca has been developed through the doctoral work of Liming Li, funded by the Norwegian Research Council. A rock like material is generated numerically by creating an assembly of particles, which may or may not be bonded to each other. Various numerical mechanical experiments can then be performed on this material. Fluid coupling has been introduced in the particle model, to permit e.g. analysis of stress effects on permeability. The dynamic nature of the PFC code permits direct analysis of elastic wave propagation. Breakable superparticles ("clumps") have been used to see the effects of grain crushing during triaxial tests. Formation of shear and compaction bands is seen as a result of such simulations. Our ambition, through an ongoing JIP ("Petrophysics under Stress"), is to calibrate the model and develop it into a "numerical laboratory" that may compute mechanical and petrophysical properties based on microscope images as an input.
 | A discrete particle representation of granular rock using spherical particles (left). Right: Stress- strain curve for a simulated triaxial test at a high confining pressure, for a cemented granular material containing breakable superparticles. The inserted plate shows the positions of broken bonds within the 2D sample after the test is completed. Notice the localization of crushed particles (black dots) into a semi-horizontal compaction band. |
Stress dependent porosity and permeability
Stress effects on porosity and permeability are analysed through experimental investigations supported by numerical (discrete particle) and poroelastic analysis.
We have developed routines for overburden correction of porosity from cores (funded by Shell). The porosity was found to change with the mean effective stress, which permits use of hydrostatic loading to reliably determine the in situ porosity. We find (for sandstone) that permeability change as a result of effective stress increase is modest and primarily related to porosity change, as long as the material behaves elastically. Beyond the elastic limit, however, large changes in permeability may occur as a result of the creation of deformation bands.
 | Stress - strain curve from a triaxial test with an outcrop sandstone, showing the (axial) permeability to decrease strongly after the rock has failed. The Kozeny-Carman prediction is based on the porosity variation recorded in the test, and shows dilating behaviour as the material fails. |
Contact: Rune M. Holt