A Trial of a Geostatistic Approach for Time-Depth Conversion

Modern technology of prestack depth migration brings directly depth section from seismic data acquired in the time domain. However, practical methods of time to depth conversion from time structure interpretation survive because of its flexibility in handling velocity and cost performance. One of the frequent problems in time-depth conversion is depth mistie at well markers. The more the well, the more complicated technique needs to be considered to solve the entire mistie. We introduce to solve the mistie issue in depth conversion by use of Geostatistics on to well check shot data, which guarantees to maintain the input data unchanged after spatial estimation. In addition to utilization of Geostatistics for conversion velocity estimation, we introduce error estimation criterion to evaluate the estimated velocity for the conversion. This criterion is evaluated comparing velocity variation at well markers. After the estimation of 3D velocity from available check shot data using Kriging approach, we evaluate velocity error distributions by means of sequential Gaussian simulation methods and extract the reliable part of velocity from the criterion. This extracted velocity data and well velocity data are integrated for final depth conversion. We estimate three realizations of velocity error distributions, which give us 3 realizations of final velocity models. The results of depth conversion by those 3 velocity models vary the depth structure within 170 meters or 7% variation at average depth of 3000 meters structure. This 7% variation can be considered as the structural uncertainty after depth conversion except for the well locations in this sturdy.