Specifying the perforation intervals and evaluating the productivity of thin-bedded sands and shales is crucial for well completion cost optimization. This requires the accurate identification of hydrocarbon-bearing sands and their contribution to production. Relying only on borehole-imaging tools to select the productive intervals is not suitable in this lithologic type because of the difficulty of permeability quantification.
In this paper we present a technique to integrate the detection of hydrocarbon-bearing sands with water saturation information and an estimation of permeability. Hydrocarbon-bearing sands are detected by high-resolution resistivity from borehole-imaging tools combined with water saturation from openhole logs (OHL), and permeability from nuclear magnetic resonance (NMR) or a modular dynamic tester tool.
We use the geostatistical concept of indicators to convert the inputs from these tools into binary data (0 and 1) based on the best selected cutoffs for those inputs, where a value of 1 means that location is good to perforate. The results of this integration are compared to the results from the production logging tool that is sensitive to the laminated sand units for evaluating its actual productivity. The best cutoffs to select for those parameters are in good agreement with the production logging tool results. The result is a set of optimized perforation intervals consistent with all the data. In addition to the certainty percent associated with the selection of perforation intervals.
Three gas wells producing from the same formation were used to apply this technique. One of the three wells was used to select the best cutoffs. For the other two wells, we used the same cutoffs to select the best perforation intervals and determine the certainty associated with them. The correct selection of the perforation intervals from the two wells was confirmed with the production testing and production logging results.