Characterization of laminations and thin beds is important for understanding sedimentary processes. Laminae are defined as layers less than 1 cm thick and beds as layers thicker than 1 cm. The frequency of laminations and beds, also called the lamination index, gives an indication of sedimentary energy and provides important clues for understanding the depositional process within a geological framework. In thinly bedded conventional reservoirs, quantification of thin beds is vital to accurately determine volumetrics. The degree of lamination in unconventional reservoirs, on the other hand, can significantly affect hydraulic fracture initiation and propagation.
Resistivity images of the borehole are high-resolution measurements that precisely capture the orientation and various properties of the laminations. Conventional analysis relies on the manual picking and counting of boundaries observed as sinusoids on the borehole image. This process is time consuming and user dependent. New image processing techniques enable automatic and fast extraction of lamination properties and statistics. A new workflow offers the possibility to perform this analysis at relevant scales for the layer thicknesses and to focus the characterization on particular layers such as ash beds.
The first step of the workflow consists of estimating the lamination orientations in a sliding window and is based on a method called the Hough transform. Resistivity curves are then extracted from the borehole image, parallel to the orientation. The algorithm is robust to gaps in the image (as in the case of wireline images) and small depth discrepancies. In addition, the extracted resistivity curve is not influenced by the presence of small scale textural features, fractures and breakouts. This curve is then decomposed in the frequency domain into different scales, corresponding to different thickness ranges. For each scale, lamination and bed boundaries are identified and their properties are determined (layer contrast and resistivity). Finally, the lamination index thicknesses are computed. The calculation uses the true stratigraphic thickness index and is thus independent of the well deviation. These properties can then be used in facies analysis or correlation.
The workflow is illustrated by examples from both conventional and unconventional reservoir settings. Beds at different scales are discriminated in a conventional sand-shale sequence. Bed thickness arrangements are readily identified, such as thickening or thinning upward sequences valuable for stratigraphic studies. A second example is from the unconventional Eagle Ford shale play. In this example, the sequence can be classified into two categories, sedimentary beds/laminations and individual conductive layers (i.e. clay rich, altered ash or pyrite rich). The density and statistics are separately computed for both categories. In addition, highly conductive beds that are thinner than the image resolution are characterized using an equation derived from tool response modelling and validated by comparison with core photographs. The final bed/lamination density is accurately calculated and can be used as potential input to hydraulic fracture models or to build stratigraphic framework (sequence or correlation based) in 3D reservoir static models.