This paper was prepared for presentation at the SPE Permian Basin Oil & Gas Recovery Conference held in Midland, Texas, 23-26 March 1998.

Well logs are the foundation on which characterization of layered reservoirs is based. However, multi-well normalization of the well logs is necessary to reduce the probability of major errors and inconsistencies in the results of the log analysis. If not removed, these inconsistencies will cause failure in any attempt to integrate log analysis results with core, well test, and production data. This paper presents a comprehensive procedure for multiwell normalization of well logs. Use of this method will ensure that the well log data can be used effectively in reservoir characterization. The advantages of multi-well normalization are illustrated in three examples of integrated reservoir studies. Example 1 illustrates the multi-well normalization process using both histograms and M-N crossplots to verify the normalization. Example 2 from a 300 well reservoir study shows the use of multi-well dual-porosity crossplot with the histogram normalization. Example 3 introduces the benefits of multi-well normalization for permeability calculations from well-logs and the effect of removing the log errors in developing a relationship between the log response and permeability. This example also illustrates the integration between well-logs and core data. Introduction Practice in using well log analysis as a part of integrated reservoir studies has shown that for the results to be accurate, consistent and comparative well-to-well, the log data require corrections' with a process called multi-well normalization. This process ensures that each logging tool reads the correct values, i.e., that the density tool accurately records formation density, the neutron log accurately reflects hydrogen index, and so forth. Experience indicates that approximately 65 to 70 percent of gamma ray logs, 50 percent of density logs, 40 to 50 percent of neutron logs, and five to 10 percent of sonic logs require some normalization to correct for variances in field calibrations of the logging tools. The normalized well log data can be effectively integrated, correlated, and calibrated with core data. The resulting correlations can be extended vertically to include layers which were not cored and laterally from well-to-well across the study area. The difference in the scale of measurement of the two sets of data must be taken into account. The core data have a scale of few cubic inches, while the log data have a scale of a few cubic meters and well test data have a scale of a few acres. Table 1 shows that even among the log measurements there are different volumes of investigation for the different tools. Multi-Well Normalization Multi-well normalization is a key activity to ensure accurate and consistent results from a multi-well log analysis study. Normalization is an iterative process that uses three tools; histograms, crossplots, and depth-based logs. These tools may be applied in the rock layers being analyzed or in nearby layers. In a marine environment of deposition, the nearby or interbeded shales may be consistent enough over a large area for normalization purposes. For instance, the Bossier shale in East Texas can be used in the normalization process. In many basins, there are often enough very low porosity carbonates with sufficient areal extent to be useful in the normalization process. This is especially true in evaporate deposition cycles. When sandstone is abundant such as in the Travis Peak & Cotton Valley of the East Texas, Prairie du Chein of Michigan, Lorelle formation of Australia, the Miocene reservoirs of the Gulf of Suez in Egypt, and many others, the sandstone itself can be used in the normalization process. P. 139^

fax 01-972-952-9435. AbstractFort Worth Basin Barnett Shale fiscal success and completion methodology have matured to the point that analog plays all over the North American continent are being actively sought and tested. It is probable that additional potential plays outside the current active areas exist, and that sufficient infrastructure may be present (or could be present) to develop them.Low permeability shale gas extraction has only become generally commercial over the last five to six years. Finding and developing this type of reservoir involves matching a number of naturally occurring parameters with an extraction process that is both detailed and capital intensive. This paper will address the natural parameters that must be present for a commercial play to be viable. It will cover the latest process and engineering improvements (stimulation and other completion issues) that have shown to improve the overall net present value of the various properties. Wildcatting strategies will be addressed, including minimizing "science" costs and reducing the time required to advance up a given learning curve.

fax 01-972-952-8435. AbstractThis paper describes methods and results of integrating modem openhole well logs and core data with old electric survey logs (ES) to determine porosity, water saturation, and net pay thickness for more than 240 wells in the San An&es Field, Veracruz, Mexico. A majority of the wells were logged with the old ES logs. As field development continued, modem resistivity logs and porosity logs were run. Several relationships were developed which allowed the old ES logs to be analyzed and the results incorporated into the reservoir description. Specifically, relationships were developed to determine porosity from the old neutron logs (GNT -both count rate and API scales), and to determine true formation resistivity (Rt) using the 64" normal logs. Log calculated porosity and water saturation were verified with available core results.