Rock physics for seismic and log interpretation
SINTEF Petroleum Research has extensive experience in conducting research related to rock physics for seismics. The activity embraces specialized experimental work on both reservoir and caprocks, using our advanced rock mechanics laboratory, as well as rock physical modeling of seismic properties using the laboratory data as a basis. When requested, the work can be integrated with complementary activities at the Seismics Department (seismic modeling), as well as the Reservoir Technology Department (reservoir simulations, fluid recombination) and the Basin Modeling Department (pore pressure modeling). Such integrated work has been performed at our Institute in case of for instance seismic monitoring feasibility studies. Some topics dealt with in our research are:
Representative test conditions
Theoretical models may be used to estimate the acoustic behavior of a rock. However, even complex and well developed models may omit or misjudge significant effects. Calibration on experimental data is therefore required. To be relevant, such data must be acquired at proper test conditions resembling what is encountered in situ for the rock.
The figure to the left shows the interior of our triaxial test cells, where we measure ultrasonic velocities both axially and radially on samples subject to triaxial stresses. Various pore fluids can be used during testing, including live oils, also at elevated temperatures. To control the tests and to correct the acoustic measurements, axial and radial strains are measured.
Stress and pore pressure sensitivity
The success of seismic reservoir monitoring relies on the ability to distinguish effects induced by changes in pore pressure and saturation. This calls for core testing. One fundamental question in this respect is whether the cores are representative of the rock due to possible damage induced during coring.
We have developed a methodology whereby we perform systematic measurements on synthetic reservoir sandstones formed under stress, and with properties tuned to those of a given reservoir. The figure to the left shows the significant effect of coring induced damage on the stress sensitivity of P-and S-wave velocities for a weak sandstone subject to uniaxial strain conditions during depletion.
Anisotropy
Anisotropy generally represents a problem for seismic processing. However, knowledge about the origin of observed the anisotropy, be it due to stresses or lithology, may represent valuable information about the actual formations.
We can quantify the acoustic anisotropy from measurements along different directions on the cores during triaxial testing. Moreover, we have developed a technique for manufacturing synthetic sandstones with controlled crack geometries, as illustrated in the figure to the left. This methodology has been utilized to study experimentally how various crack geometries affect the acoustic wave propagation.
Sample availability
Sample availability for acoustic measurements may be limited, particularly in shaly intervals. For instance, knowing that proper seismic parameters for the overburden are important for seismic migration, this may call for alternative sources of overburden information.
We have developed several specialized tools that can be used to measure a.o. static and dynamic moduli on small samples. The picture to the right shows our Continuous Wave Technique system, CWT, developed to measure formation velocities rig-site, even utilizing mm-sized drill cuttings.
Full waveform sonic log modeling
Sonic logs can provide valuable information along the wellbore. However, they are also sensitive to drilling induced effects caused by stresses and drilling fluids. Thus, reliably utilization of such data requires sufficient understanding of which parameters affect the log readings, and how they are related to the specific rock physical behavior.
We have developed a FD code that models the full waveform response of a formation to a sonic log tool in a wellbore. The figure to the left shows snapshots for a monopole response in a wellbore with mud invasion. Generally, both mono- and dipole sources can be modeled. The model is limited to rotational symmetry around the wellbore, and incorporates anisotropy and attenuation effects.
Frequency effects
Most laboratory velocity measurements are performed at ultrasonic frequencies, five orders of magnitude above the seismic frequencies. Attenuation will, however, render the measured velocities frequency dependent. The intrinsic velocity dispersion will be particularly sensitive to the presence of pore fluids.
The picture to the right shows a low-frequency laboratory system that we are developing to measure P-wave velocities at seismic frequencies. This allows for a quantification of velocity dispersion effects for given rocks.
Contact: Olav-Magnar Nes