Seismic Investigation Of Deep 2D, 3D And 4D Structures Using Body Wave Interferometry, Computer Simulations And Semi-Conventional Processing

Abstract

The main focus of this dissertation is retrieval and imaging with virtual body wave reflections extracted from ambient noise record. The advantage of these virtual reflections over surface waves, which require inversion, is the direct relationship between the recorded arrival times and the depth of the imaged interface. However, the virtual reflection retrieval from a seismic noise record can be problematic due to limited amount of body wave sources in the ground. Furthermore, body wave interferometry is sensitive to the 3D distribution of these sources as oppose to effectively 2D distribution of surface wave sources. The work presented here is focused on the effect of source distribution on the virtual body wave arrivals. We generate models and synthetic seismograms for various source distributions and apply interferometric technique (i.e. crossand auto-correlation) to retrieve the synthetic virtual direct, reflected and refracted arrivals. We then analyze source distributions that lead to the distortion of these arrivals and rise of spurious arrivals that do not directly correspond to the modeled subsurface structure. Finally, we are able to use the conclusions of this analysis for interpretation of the real data to differentiate between the arrivals directly related to the subsurface and artifacts. In addition to the main focus of the dissertation, we are presenting work on seismic basement reflections in South-East New Mexico. We have applied ex- tended correlation technique to extend the vibroseis 3D recording to image deep structures (potentially to crust-mantle discontinuity). We are able to recover a 3D structure associated with the Precambrian basement using conventional processing applied to the extended recording and propose a collaboration between the industry and academia to enable the latter to systematically probe for the deep structures using oil and gas provided datasets. This collaboration will expand the existing database of deep reflections and in some cases can provide a high resolution depth maps of these structures. The last chapter of this presentation is part of a summer internship project in Shell Oil Company (Houston, TX). This project involved seismic 4D simulations of subsurface response in an enhanced oil recovery (EOR) field. We have established that frequent 4D seismic repeats are necessary in order to confidently interpret dynamic changes in the subsurface properties (e.g. pressure, fluid saturation, temperature) beneath oil production fields that undergo steam injections .