Fractal parameters and well-logs investigation using automated well-to-well correlation

The aim of well-to-well correlation is to detect similar geological boundaries in two or more wells across a formation, which is usually done manually. The construction of such a correlation by hand for a field with several wells is quite complex and also time-consuming as well. The aim of this study is to speed up the well-to-well correlation process by providing an automated approach. The input data for our algorithm is the depths of all geological boundaries in a reference well. The algorithm automatically searches for similar depths associated with those geological boundaries in other wells (i.e., observation wells). The fractal parameters of well-logs, such as wavelet exponent (Hw), wavelet standard deviation exponent (Hws), and Hausdorff dimension (Ha), which are calculated by wavelet transform, are considered as pattern recognition dimensions during the well-to-well correlation. Finding the proper fractal dimensions in the automatic well-to-well correlation approach that provide the closest geological depth estimation to the results of the manual interpretation is one of the prime aims of this research. To validate the proposed technique, it is implemented on the well-log data from one of the Iranian onshore oil fields. Moreover, the capability of gamma ray, density, and sonic log in automatic detection of geological boundaries by this novel approach is also analyzed in detail. The outcome of this approach shows promising results.