 |
| Titel |
Estimating peak dynamic strains at the ground surface and at depth during earthquake shaking: application to the safety study of a geological storage of CO2 |
| VerfasserIn |
S. Sy, J. Douglas, D. Seyedi |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250025444
|
|
|
|
|
|
| Zusammenfassung |
| Within the framework of a scenario-based methodology used to evaluate the risks related to
the geological storage of CO2, the risk posed by earthquakes to the storage safety must be
evaluated. The main aim of this article is to predict by a simple empirical method,
verified by numerical simulations, the peak dynamic strains in the reservoir during an
earthquake. This allows, following an investigation of the risk of rock rupture or
damage in the reservoir and sealing units (i.e. caprock and wells), an evaluation
of the seismic risk. However, these subsequent calculations are not carried out in
this article, which is limited to a determination of the dynamic stresses during an
earthquake.
A simplified procedure for the prediction of maximum soil strains was proposed by
Newmark [1] and later used by several authors. In this approach, the peak strain is equal to
the horizontal peak ground velocity (PGV) divided by the apparent ’propagation speed’ of
strong-motion waves, C. Using that approximate equation for strain is difficult, because C is
not known a priori and depends both on site and wave characteristics. A more recent
approach simplifies it by replacing C by β1, which is the shear-wave velocity in the
uppermost layer of the soil structure, and by including a site-specific corrective factor A
[2].
However, all these studies were limited to determining the peak dynamic strains
at the ground surface and not at depth. Thus it is necessary for us to evaluate the
site-specific corrective factor A for the geological context of studied area for the peak
dynamic strain at the surface and also to estimate a similar correlation between
strains and PGV/β at depth, where β is the shear-wave velocity in the considered
layer.
A sophisticated one-dimensional site response computer program is used to create a set
for analysis of peak strains. It calculates, for a given geological model consisting of parallel
layers and a given input accelerogram, the ground accelerations, velocities and
displacements, and the dynamic stresses and strains, in each layer. Thus, we can correlate
associated peak ground motions, in particular PGV, and the peak dynamic strains. An
approximate seismic profile corresponding to the geological structure of North Pyrenees
Basin has been used to conduct these simulations using a number of accelerograms recorded
on rock that are appropriate for the seismo-tectonic context of this region. These data are then
used for statistical analysis to find a linear correlation between the peak dynamic
strain and PGV/β. The dynamic stresses in the ground following an earthquake
can be quickly and estimating by assuming linear behaviour. The advantages and
disadvantages of such an approach and its limits related to the assumption of linear
behaviour and the fact that the 3D structure at the site is not taken account, are finally
discussed.
References
[1] Newmark, N. Problems in wave propagation in soil and rock, Proc. Int. Symp. on
Wave Propagation & Dynam. Properties of Earth Materials, Albuquerque, pp. 7-26,
1967.
[2] Trifunac M.D., Lee V.W. Peak surface strains during strong earthquake motion, Soil
Dynamics and Earthquake Engineering 15, 311-319, 1996. |
|
|
|
|
|
|