The goal of oil and gas exploration using seismic methods is to accurately locate geological structures that could host such reserves. As the search for these resources tends towards more complex regions, it is necessary to develop methods to extract as much information as possible from the seismic data acquired. Prestack depth imaging is a seismic processing technique that has the capability to produce a realistic depth image of geological structures in complex situations. However, improvements to this technique are required to increase the accuracy of the final depth image and ensure that the targets are accurately located. Although prestack depth imaging possesses the ability to produce a depth image of the Earth, it does have its disadvantages. Three problematic areas in depth imaging are: the computer run times (and hence costs) are excessively high; the success of depth migration is highly dependent upon the accuracy of the interval velocity model; and seismic multiples often obscure the primary reflection events representative of the subsurface geology. Velocity model building accounts for most of the effort in prestack depth imaging and is also responsible for the likelihood of success. However, the more effort that is expended on this process, the greater the cost of producing the required depth section. In addition, multiples remain a problem in complex depth imaging since many attenuation techniques are based assumptions that may only be approximately correct and in addition require a priori information. The Kirchhoff method is considered to be the workhorse in industry for prestack depth imaging. It is a simple and flexible technique to implement, and usually produces acceptable images at a small fraction of the cost of the other depth migration methods. / However, it is highly dependent on a method for calculating the traveltimes that are required for mapping data from the prestack domain to the output depth section. In addition, it is highly dependent on the accuracy of the interval velocity model. Multiples can also be problematic in complex geological scenarios. To improve the quality of the depth section obtained from Kirchhoff depth imaging, these three issues are considered in this thesis. This thesis took on the challenge of developing new techniques for (a) improving the accuracy and efficiency of traveltimes calculated for use in Kirchhoff prestack depth imaging, (b) building the interval velocity model, and (c) multiple attenuation in complex geological areas. Three new techniques were developed and tested using a variety of numerical models. A new traveltime computation method for simulating seismic multiple reflections was tested and compared with a Promax© finite-difference traveltime solver. The same method was also used to improve the computational efficiency whilst retaining traveltime accuracy. This was demonstrated by application to the well-known Marmousi velocity model and a velocity model obtained from analysis of data from the North West Shelf of Western Australia. / A new interval velocity model building technique that utilises the information contained in multiple events was also implemented and tested successfully using a variety of numerical models. Finally, a new processing sequence for multiple attenuation in the prestack depth domain was designed and tested with promising results being observed. Improved accuracy in the depth image can be obtained by combining the three techniques I have developed. These techniques enable this to be achieved by firstly improving traveltime accuracy and computation efficiency. These benefits are then combined with a more accurate interval velocity model and data with a minimal problematic multiple content to produce an accurate depth image. These new techniques for Kirchhoff depth imaging are capable of producing a depth section with improved accuracy, and with increased efficiency, that will aid in the process of seismic interpretation.
Identifer | oai:union.ndltd.org:ADTP/222819 |
Date | January 2002 |
Creators | Manuel, Christopher D. |
Publisher | Curtin University of Technology, School of Resource Science and Technology. |
Source Sets | Australiasian Digital Theses Program |
Language | English |
Detected Language | English |
Rights | unrestricted |
Page generated in 0.0015 seconds