2. DATA AND METHODLOLOGY
2.2. SEISMIC REFLECTION DATA
2.2.1. THEORY OF SEISMIC REFLECTION IMAGING
The primary purpose behind acquisition of seismic reflection data is to image sub-‐
surface successions by the transmission and subsequent detection of compressional acoustic waves. The generation of acoustic waves must be repeatable to allow comparisons across the survey, have sufficient energy to propagate beyond the intended target and be safe, efficient and environmentally acceptable.
Consequently, seismic sources often comprise air guns (offshore), vibroseis or small explosives (onshore) detonated at, or just below the earths’ surface (Kearey et al., 2009). The emitted waves propagate through the subsurface and some are
37 reflected back to the surface by acoustic (geological) boundaries (reflection surfaces include bedding planes or unconformities). The remaining waves are refracted or attenuated. The proportions reflected to the surface are detected by geophone or hydrophone arrays where they may be subsequently processed for interpretation.
While seismic reflection surveys may be conducted either on or offshore, the data utilised in thesis is collected solely in marine settings, as such only offshore methodologies will be referred to here. Consequently, when references herein are made to seismic wave velocity, this refers only to P-‐wave velocity as S (shear) waves are not transmitted through fluids.
The fundamental theory pertaining to seismic reflection surveys is the defining of the acoustic impedance (z) of a material. The impedance contrast between two materials determines the relative proportions of seismic energy that are either transmitted or reflected across the geological boundary. The acoustic impedance of a material is a product of its density (ρ) and its wave velocity (v) (Kearey et al., 2009); that is,
𝑍 = 𝜌𝑣
(2.1)
Contrasts in acoustic impedance across a geological boundary control the reflection coefficient (R) of such a boundary. The reflection coefficient is a numerical measure of the effects of an interface on the propagation of waves across it. Normally it is calculated as a ratio of the amplitude of the reflected wave to the amplitude of the incident ray (Kearey et al., 2009). However relating this principal to the physical properties of the interface materials requires the stress and strain of both materials
38 to be considered. The formal solution to this relationship was derived by Zoeppritz (1919) but the widely accepted solution will be shown here (Bacon et al., 2003;
Kearey et al., 2009); such that,
𝑅 =!!!!!!!!!!
!!!!!!!! (2.2)
This simplifies to give,
𝑅 = !!!!!!
!!!! (2.3)
The velocity of seismic P-‐waves through an isotropic, homogenous substance is controlled by the elastic properties and density of the material (Sheriff and Geldart, 1982). The subsurface is rarely either isotropic or homogenous, consequently, wave velocity will vary in three dimensions depending on rock or sediment composition, porosity, fluid saturation and pressure (Bacon et al., 2003). As such, seismic reflection data must be tied to calibrated velocity models derived from well bores before it can be used to estimate the true depth of a point of interest.
Seismic data may be collected in two, three or four (time lapse) dimensional surveys. The seismic data used in this study comprises predominantly 2D seismic surveys with additional use of 3D data. No 4D (time-‐lapse) seismic data has been used and as such is included in this section in reference to its use for post injection monitoring of CO2 storage sites.
39 Two Dimensional seismic surveys are acquired as a series of parallel and orthogonal lines often kilometres apart that produce a cross section of the subsurface ((Kearey et al., 2009). The technology was first developed in the 1920’s and was refined through to the 1950’s. Interpretation of intersecting perpendicular lines allows basic models of the subsurface to be constructed by interpolation between lines.
Models however are limited by the spacing of the seismic lines as these define the scale of resolvable structures. Thus, any structures, such as channels, antiformal domes and faults smaller than the grid spacing of the survey will not be imaged.
Three Dimensional seismic surveys utilise a regular grid of multiple 2D lines with an approximate 12.5 to 25m spacing. Such spacing results in a virtually continuous 3D data cube that is viewable from any orientation. The advances in 3D seismic resolution allow small-‐scale subsurface features, unresolvable in 2D, to be mapped with a high level of detail. Additionally, the advantages of the near continuous data cube allows key horizons to be interpreted quickly across a large geographical area.
Four Dimensional seismic surveys, also referred to as time-‐lapse seismic data comprise the study of two or more 3D seismic surveys over the same reservoir or target. This aims to observe changes over time, whether as an consequence of hydrocarbon production or to observe the impact of secondary recovery techniques. Most 4D seismic surveys utilise existing 3D surveys acquired at different times over the same or overlapping area and thus require very careful reprocessing to eliminate problems. In spite of improvements in reprocessing, these surveys require a large shift in reservoir acoustic properties to be observable. Recent surveys have used permanently positioned seabed receiver arrays, which
40 significantly improves the survey repeatability and increases the detectability of subtle acoustic changes in the target reservoir or formation (Brown, 2004).
Although not used in this study, 4D seismic surveys have been identified as an important potential monitoring tool to observe the migration of injected CO2 plumes in sequestration projects as proven as proven by the Sleipner and Weyburn projects (Cairns et al., 2012; Chadwick et al., 2004, 2009; White, 2013). However, the high cost implications are seen as a barrier for large scale deployment.
2.2.1. ACQUISITION, PROCESSING AND INTERPRETATION OF SEISMIC