Estimation of changes in water column velocities and thicknesses from time lapse seismic data

Abstract

A B S T R A C TSea-bed diffractions are frequently observed for several of the fields in the Norwegian Sea and the Barents Sea. This is a challenge in time lapse seismic analysis, since diffracted multiples are difficult to remove by processing and therefore is a major source of poor time lapse data quality. In this work we test if the diffractions can be used for enhanced 4D interpretation. By analysing the time-shift of the sea-bed diffraction hyperbola between the base and monitor it is tested if changes in water velocity and tides can be estimated.Two models using time lapse diffraction analysis are tested: the first one simply adds time-shifts for the two branches of the diffraction hyperbola and this average time-shift is then used to estimate the water velocity change. The other method uses an inversion method based on the diffraction equation for a point diffractor to estimate the velocity change. In-line common-midpoint shifts are estimated by subtracting the time-shifts of both hyperbola branches followed by direct inversion. The diffraction based time-shifts are compared to time-shifts estimated by standard cross-correlation of the sea-bed reflection. The averaging method gives slightly higher uncertainties, while the inversion using an exact traveltime equation gives similar uncertainties compared to the sea-bed reflection method.

Sea‐bed diffractions and their impact on 4D seismic data

“…However, we observe on the data that the time‐shift after tidal corrections is not constant. Osdal and Landrø (2010) showed two different examples of time‐shifts of 0.7 ms and 1 ms. The variation in water velocity and temperature with depth are measured by lowering a probe from a seismic vessel.…”

“…Six measurements from the 2006 survey (called the base survey in the following) and two from the 2008 survey (monitor survey) show average velocities of 1487 m/s and 1484–1485 m/s. The measured velocity profiles and uncertainties can be seen in Osdal and Landrø (2010).…”

“…The water layer time‐shift is corrected by first using tidal correction followed by a swath‐wise smoothing of the estimated time‐shifts for the sea‐bed reflection or the estimated time‐shifts in the overburden, typically from cross‐correlation. However, it is likely that tuning between the sea‐bed reflections and sea‐bed diffractors might disturb this time‐shift correction and make it a challenge to estimate water velocity changes (Osdal and Landrø 2010). Landa (2010) emphasized the possibilities one has from seismic diffraction analysis and Osdal and Landrø (2010) suggested using sea‐bed diffractors to correct for water layer changes.…”

“…However, it is likely that tuning between the sea‐bed reflections and sea‐bed diffractors might disturb this time‐shift correction and make it a challenge to estimate water velocity changes (Osdal and Landrø 2010). Landa (2010) emphasized the possibilities one has from seismic diffraction analysis and Osdal and Landrø (2010) suggested using sea‐bed diffractors to correct for water layer changes. Using a deeper unbiased reflector for water layer corrections might be an alternative that can reduce the impact from a rough sea‐bed for non‐compacting reservoirs, where the overburden velocity is assumed unaffected by changes at the reservoir level.…”

Diffraction imaging of sub‐vertical fractures and karst with full‐resolution 3D Ground‐Penetrating Radar

“…Ground-Penetrating Radar (GPR) is often used to characterize fractures and karst in near-surface carbonates (Liner and Liner 1995;McMechan et al 2002;Pipan et al 2003;Theune et al 2006;Nuzzo, Leucci and Negri 2007). Typically GPR signals can reach 10-30 m depth in clay-free carbonates but practice shows that sub-vertical fracture and karst networks are difficult to interpret due to chaotic scattering.…”