Effects of scales and extracting methods on quantifying quality factor

The attenuation values obtained in the lab are usually larger than those measured in the field. To understand this inconsistency, I propose to model attenuation from rock properties and reflection seismic data, and then compare their results. A gas well is used here for providing the rock properties and seismic properties. The preliminary results show that scaling is one of the factors that lower the attenuation values obtained from the field. In the future work, by adjusting more parameters (e.g., the frequency band), checking the received energy in the seismic data, and comparing those two Q models, I will understand more of the key influencing factor on the inconsistency, which helps optimize the acquisition design and choose a suitable method of measuring Q. Those conclusions can be used in practice and help understand the reservoir characterization. INTRODUCTION Problem Statement Consistent and accurate field measurements of Q are rare due to practical difficulties of extracting attenuation from reflection seismic data, crosswell, VSP, and full waveform borehole data. These field measurements are also inconsistent with the lab measurements. Q values estimated from seismic events are usually very high. In-situ Q in marine sediments has been estimated to be 30 in wet sand and as high as 100 and even 400 in silt and clay. Kvamme and Havskov (1989) estimate Q about 950 at 10 Hz while in Lilwall (1988), Q is between 100 and 200 in the upper 3 km of the crust. VSP data have been used to calculate Q exceeding 300 in basement rock at depths below 1.8 km. Crosswell tomography has been used to estimate attenuation in the 200-2000 Hz frequency range. Q is between 30 and 50 in soft (Vp between 2.6 and 3.0 km/s) sand/shale sequence and reaches 100 in chalk and limestone. A Q of 33 has been estimated from a high-resolution 2D seismic data over a Florida carbonate high-porosity aquifer system where VP is between 2 and 3 km/s and density is about 2 g/cc. The attenuation values obtained in the lab are usually larger than those measured in the field. A study by Klimentos (1995) is one of the few relevant to hydrocarbon exploration. It reports, based on sonic waveform analysis, that Q falls between 5 and 10 in gas sandstone of about 12% porosity (Q−1 between 0.1 and 0.2) while it may easily exceed 100 (Q−1 < 0.01) in oiland water-saturated intervals. Attenuation is large in rock with partial gas saturation and small in liquid-fill rock. Other examples also reveal that the lab-measured attenuation in the gas rock is small, as shown in Figure 1. Shen 3 Ph.D. Proposal Although those inconsistencies are observed by the above studies, their explanations have not been provided. One possible reason is that the seismic survey are usually conducted in a large spatial range. Hence the signal may not reach the target or receivers after propagating through a strong absorptive medium. Furthermore, the large scale or the low frequency in the seismic survey spatially averages the attenuation, and hence lower the estimated attenuation values. In addition, different methods used for the Q model building is another explanation for the different results estimated from the lab and the seismic measurements. Therefore, if we could fully understand the influencing factor on the inconsistency, we could optimize the acquisition parameters, choose a suitable measurement of Q, and hence help accurately predict and interpret the reservoir.