Time-lapse seismics is known to be a very effective monitoring technique for the subsurface fluid movement and saturation changes, as well as for geomechanical phenomena [Snieder et al., 2007]. The integration of seismic and reservoir engineering is now becoming state-of-the-art [Boutte, 2007] while the number of applications is steadily increasing [Staples et al., 2006]. Among the future challenges to the use of time-lapse seismics is the integration with geomechanics [Landrø, 2006]. The improvement of time-lapse seismic technology [e.g. Tang et al., 2007, Aarre et al., 2007] allows for better and more accurate data acquisition, that in turn allows to "see" effects previously difficult to detect. The effects of geomechanics on time-lapse seismic data have been described in detail by a number of publications [Hatchell and Bourne, 2005; Sayers and Schutjens, 2007; Cox and Hatchell, 2008; Kristiansen and Plischke, 2010]. The overall impact of reservoir exploitation on the changes in seismic response includes the following aspects: (1) Fluid saturation effects, that are based upon: (a) dependence of density on fluid saturation; and (b) dependence of bulk moduli on fluid saturation (Gassmann, 1951). This is the key effect sought in time-lapse seismics, as it allows remote monitoring of the fluid migration in the reservoir. Mainly, two effects are sought in data hopefully depending on the above saturation changes, i.e.: - time shifts, i.e. changes in reflector location in time as a consequence of changes in velocity, and mainly: - impedance changes, i.e. reflectivity changes, as impedance is the product of velocity and density, both changing with fluid content. (2) Pressure (effective stress) effect: this is the first, well known geomechanical effect, often referred to in the literature as pressure effect, but it is actually a dependence on effective stress. It is generally observed that the velocity decrease is very strong in presence of effective stress decrease (expansion), while velocity increase is relatively mild under stress increase (compaction) [e.g. Hatchell and Bourne, 2005]. This asymmetric behaviour is often explained in terms of crack opening under stress release conditions. In addition to the above well-known effects, other important time-lapse phenomena depending on geomechanics have been highlighted in the recent past, i.e.: (3) Seismic velocity changes in the overburden: in presence of significant arching and stress redistribution above the reservoir, a large region can undergo large velocity drops in presence of effective stress decreases. This effect can have a strong effect on time-shifts, with possible 3D (lens) effects causing also lateral time-shifts [Cox and Hatchell, 2008], as well as on changes in apparent impedance contrasts caused by wavelet interference phenomena. In this contribution we present an approach that can take into account all phenomena described above, by integrating the results of a reservoir model, of a geomechanical model and of a full-waveform seismic model - see Figure 1. This integrated approach can lead to a calibration of both the reservoir and the geomechanical model also on the basis of the time-lapse seismic data. A key role is taken by a suitable constitutive model linking stress changes/strains to seismic velocity changes, allowing for a different non-symmetric behaviour in compaction and dilation. (Fiture presented) We tested the proposed procedure on a synthetic model (Figure 2) where the reservoir is assumed to be fully depleted. The velocity model was constructed applying a simplified relationship (Calvert, 2005) to the effective stress field computed with an elasto-plastic geomechanical model (Isamgeo). The difference in velocity between base and monitor is shown in Figure 3, highlighting the complex spatial pattern resulting from stress redistribution. The Sem2Dpack seismic software (Ampuero, 2008) was then used to simulate the base and monitor seismic acquisitions. Examples of waveform snapshots are shown in Figure 4, highlighting amplitude and time differences resulting from geomechanical deformation.
A study of geomechanical effects on time-lapse seismics
Cassiani, G.
Primo
;Vignoli, G.;
2012-01-01
Abstract
Time-lapse seismics is known to be a very effective monitoring technique for the subsurface fluid movement and saturation changes, as well as for geomechanical phenomena [Snieder et al., 2007]. The integration of seismic and reservoir engineering is now becoming state-of-the-art [Boutte, 2007] while the number of applications is steadily increasing [Staples et al., 2006]. Among the future challenges to the use of time-lapse seismics is the integration with geomechanics [Landrø, 2006]. The improvement of time-lapse seismic technology [e.g. Tang et al., 2007, Aarre et al., 2007] allows for better and more accurate data acquisition, that in turn allows to "see" effects previously difficult to detect. The effects of geomechanics on time-lapse seismic data have been described in detail by a number of publications [Hatchell and Bourne, 2005; Sayers and Schutjens, 2007; Cox and Hatchell, 2008; Kristiansen and Plischke, 2010]. The overall impact of reservoir exploitation on the changes in seismic response includes the following aspects: (1) Fluid saturation effects, that are based upon: (a) dependence of density on fluid saturation; and (b) dependence of bulk moduli on fluid saturation (Gassmann, 1951). This is the key effect sought in time-lapse seismics, as it allows remote monitoring of the fluid migration in the reservoir. Mainly, two effects are sought in data hopefully depending on the above saturation changes, i.e.: - time shifts, i.e. changes in reflector location in time as a consequence of changes in velocity, and mainly: - impedance changes, i.e. reflectivity changes, as impedance is the product of velocity and density, both changing with fluid content. (2) Pressure (effective stress) effect: this is the first, well known geomechanical effect, often referred to in the literature as pressure effect, but it is actually a dependence on effective stress. It is generally observed that the velocity decrease is very strong in presence of effective stress decrease (expansion), while velocity increase is relatively mild under stress increase (compaction) [e.g. Hatchell and Bourne, 2005]. This asymmetric behaviour is often explained in terms of crack opening under stress release conditions. In addition to the above well-known effects, other important time-lapse phenomena depending on geomechanics have been highlighted in the recent past, i.e.: (3) Seismic velocity changes in the overburden: in presence of significant arching and stress redistribution above the reservoir, a large region can undergo large velocity drops in presence of effective stress decreases. This effect can have a strong effect on time-shifts, with possible 3D (lens) effects causing also lateral time-shifts [Cox and Hatchell, 2008], as well as on changes in apparent impedance contrasts caused by wavelet interference phenomena. In this contribution we present an approach that can take into account all phenomena described above, by integrating the results of a reservoir model, of a geomechanical model and of a full-waveform seismic model - see Figure 1. This integrated approach can lead to a calibration of both the reservoir and the geomechanical model also on the basis of the time-lapse seismic data. A key role is taken by a suitable constitutive model linking stress changes/strains to seismic velocity changes, allowing for a different non-symmetric behaviour in compaction and dilation. (Fiture presented) We tested the proposed procedure on a synthetic model (Figure 2) where the reservoir is assumed to be fully depleted. The velocity model was constructed applying a simplified relationship (Calvert, 2005) to the effective stress field computed with an elasto-plastic geomechanical model (Isamgeo). The difference in velocity between base and monitor is shown in Figure 3, highlighting the complex spatial pattern resulting from stress redistribution. The Sem2Dpack seismic software (Ampuero, 2008) was then used to simulate the base and monitor seismic acquisitions. Examples of waveform snapshots are shown in Figure 4, highlighting amplitude and time differences resulting from geomechanical deformation.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
