Reservoir characterization based on seismic spectral variations Yanghua Wang 1 ABSTRACT The seismic frequency spectrum provides a useful source of information for reservoir characterization. For a seismic profile presented in the time-space domain, a vector of the frequency spectrum can be generated at every sampling point. Because the spectrum vectors at different time-space locations have different variation features, I attempt for the first time to exploit the variation pattern of the frequency spectrum for reservoir characterization, and test this innova- tive technology in prediction of coalbed methane (CBM) re- servoirs. The prediction process implicitly takes account of the CBM reservoir factors (such as viscosity, elasticity, cleat system, wave interference within a coal seam, etc.) that af- fect the frequency spectrum, but strong amplitudes in seis- mic reflections do not necessarily show any influence in clustering analysis of spectral variation patterns. By calibrat- ing these variation patterns quantitatively with CBM produc- tions in well locations, we are able to characterize the spatial distribution of potential reservoirs. INTRODUCTION In this study, I propose to exploit the variation patterns of seismic frequency spectra for characterizing the spatial distribution of po- tential reservoirs. Considering, for example, a 2D seismic profile in the time-space domain, a vector of the frequency spectrum can be generated at every sample point. Such frequency spectral vectors at different time-space positions have different variation patterns. If using different colors to present the time-space points with different variation patterns, one can create a 2D color image. This colorful 2D picture has the potential to highlight the reservoir anomalies. As a demonstration, I apply this technology to predict the spatial dis- tribution of coalbed methane (CBM) reservoirs, which is an impor- tant unconventional energy resource (Shuck et al., 1996; Bachu and Michael, 2003; Peng et al., 2006). The “data” I use in prediction are the frequency spectra. Thus, the first step of the entire procedure is to raise an extra di- mension. For a conventional 2D seismic profile in the time-space domain, I generate a 3D data cube in which the third dimension is frequency. This data cube in time-space-frequency domain is often called a time-frequency spectrum. The second step is cluster- ing analysis, which is a dimension-reduction process that reduces the data dimensions from three down to two. The resultant image is presented again in the time-space domain. The third step is characterization which calibrates the indexes of spectral patterns with known CBM production and predicts the CBM spatial distribution. I generate the time-frequency spectrum using the matching pur- suit technique (Wang, 2007, 2010). However, because there is no straightforward relationship between the CBM content and the seis- mic strength, one cannot make a quantitative characterization of CBM spatial variation based on various types of amplitudes either in the frequency domain or the time domain. Therefore, in this study I propose for the first time to predict the spatial distribution of coal seam and its methane content, based on the variation characteristics of the seismic frequency spectrum. The essential message here is that it is the spectral variation along the frequency axis, rather than the amplitude of each indivi- dual frequency component, that plays a key role in reservoir char- acterization. A strong reflection in the seismic profile, either in the time or frequency domain, does not necessarily show any influence in spectral-variation-based reservoir characterization. STUDY AREA Coalbed methane used to be a mining hazard, which has now been converted into an environmentally friendly fuel. As a signifi- cant energy resource, CBM is cleaner than any other fossil fuel. It has been claimed that the worldwide resources of methane trapped within the cleat and fractured coal seams are greater than the total reserves of all known conventional natural gas fields (Bachu and Michael, 2003). Manuscript received by the Editor 1 September 2011; revised manuscript received 26 May 2012; published online 27 September 2012. 1 Imperial College London, Centre for Reservoir Geophysics, Department of Earth Science and Engineering, London, U. K. E-mail: yanghua.wang@imperial .ac.uk. © 2012 Society of Exploration Geophysicists. All rights reserved. M89 GEOPHYSICS, VOL. 77, NO. 6 (NOVEMBER-DECEMBER 2012); P. M89–M95, 7 FIGS., 1 TABLE. 10.1190/GEO2011-0323.1