Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.777779
Title: Estimation of time-lapse velocity changes from time-lapse seismic data
Author: Nguyen, Phung Kim Thi
ISNI:       0000 0004 7963 5526
Awarding Body: Heriot-Watt University
Current Institution: Heriot-Watt University
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Time-lapse velocity changes represent changes in pore-fluid and rock properties within the reservoir, reflect the geomechanical effects and provide input to an imaging correction for enhanced time-lapse interpretation. Here I present three different time-lapse velocity change estimation methods that can be applied to time-shifts from both post-stack data and also partial-stacked data. The aim of this research is to calculate robust yet stable and efficient algorithms for velocity change estimation by analytical development. Starting with the post-stack domain, I firstly develop a new method of Gaussian reconstruction that allows stable recovery of the time-lapse velocity changes despite varying levels of noise in the post-stack time-shifts. I then extend to partial-stack time-shift domain where the dependency of time-shift versus o set is explored via an extensive revision together with numerical examples before presenting the development of another robust method to extract the time-lapse velocity changes from partial-stack time-shifts. Here, I simplify tomographic inversion by using a straight-ray assumption and specialised re-gridding technique. In addition to time-shifts, in the post-stack seismic domain, amplitude changes are taken into account together with time-shifts and are inverted simultaneously by incorporating the Gaussian reconstruction method into a trace-warping algorithm. Overall, the algorithms developed in this work perform well when applied to data from the high-pressure high-temperature Shearwater field. Time-lapse velocity changes are shown to be inverted in a robust and efficient manner, without the need of a prior model or over-regularisation. The techniques are sufficiently versatile that they can be applied to different data types: post-stack time-shifts, partial-stack time-shifts or post-stack seismic traces.
Supervisor: MacBeth, Colin ; Mangriotis, Maria Daphne Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.777779  DOI: Not available
Share: