Tight-Gas-Sand Permeability Estimation From Nuclear-Magnetic-Resonance (NMR) Logs Based on the Hydraulic-Flow-Unit (HFU) Approach

Abstract

The crossplot of porosity vs. Klinkenberg permeability (PERM) for 378 core samples, drilled from tight gas sands in the Xujiahe formation of the Anlu region-central Sichuan basin, southwest China, showed that tight-gas-sand permeability cannot be estimated effectively from porosity directly because only a poor relationship can be found between core-derived porosity and permeability because of the strong heterogeneity, especially for reservoirs with dominant microfractures (with porosities lower than 6.5%). However, the problem can be solved by introducing the HFU approach. In this paper, the 378 core samples were divided into five types on the basis of the difference of the flow-zone indicator (FZI), and then relationships of rock porosity and permeability were established for every type of core sample. By virtue of the analysis of the expression of FZI and the classical Schlumberger Doll Research (SDR) Center model, a novel technique used to obtain FZI from NMR logs was proposed and a corresponding model was established. The model parameters were calibrated by use of the laboratory NMR measurements of 54 plug samples taken from the Xujiahe formation. Carried out on the experimental data sets, this procedure can be extended to reservoir conditions to estimate consecutive formation permeability along the intervals through which NMR logs were acquired. The processing results of field examples illustrate that the calculated FZI values from field NMR logs match very well with the core analyzed results; the absolute errors among them are within the scope of 60.15. Moreover, permeability, estimated by use of the proposed technique and the core analyzed results are consistent. However, the calibrated SDR model is exclusive to the cases where formation permeability ranges from 0.2 to 0.4 md. To improve permeability prediction with the SDR model, many more core samples drilled from formations with dominant microfractures needed to be tested for laboratory NMR experiments to calibrate the SDR model for each HFU.

Application of nuclear magnetic resonance permeability models in tight reservoirs

“…The same technique can be used for calculating permeability by using the following equation for FZI: …”

“…The samples were collected from the Montney, Nordegg, and Doig formations from the Western Canadian Sedimentary Basin (WCSB) in Alberta, Killam area in central Alberta, Wapiti in northwestern Alberta, Mist Mountain and Red beds in southwestern Alberta, Lathom area in southern Alberta, York River sandstone formation in Quebec, and Weyburn Middale CO 2 Monitoring and Storage Project in Saskatchewan, Canada. Samples were also collected from the Ordivician quart reservoir in the Bir Ben Tartar concession in south Tunisia, Teslimkoy and Kesan formations in Turkey, Xujiahe formation of the Anlu region, central Sichuan basin in southwestern China, and one non‐specified carbonate formation . The data from Xujiahe, Weyburn, and the one non‐specified carbonate formations were adopted from other literature .…”

“…Samples were also collected from the Ordivician quart reservoir in the Bir Ben Tartar concession in south Tunisia, Teslimkoy and Kesan formations in Turkey, Xujiahe formation of the Anlu region, central Sichuan basin in southwestern China, and one non‐specified carbonate formation . The data from Xujiahe, Weyburn, and the one non‐specified carbonate formations were adopted from other literature . This section describes the series of experiments that were designed for the core plugs from the remaining formations.…”

Hydraulic-Unit-Based Fuzzy Model to Predict Permeability from Well Logs and Core Data of a Multi-Layer Sandstone Reservoir in Ulleung Basin, South Korea

“…Soto et al (2001) developed a fuzzy model that predicted permeability using FZI and HU grouping, and showed its improved prediction accuracy. Shahvar et al (2009) Xiao et al (2013) estimated permeability from NMR logs by applying HU grouping using FZI. Many studies have successfully predicted the permeability of uncored wells using ANN or fuzzy models with HU groups, but the models have been applied to uncored intervals or wells with the same depositional environment as the core data.…”

“…Hydraulic-unit-based Fuzzy Model to predict permeability from well logs and core data of a multi-layer sandstone reservoir in Ulleung Basin, South Korea formation in southwestern Iran. Xiao et al (2013) estimated permeability from NMR logs by applying HU grouping using FZI. Many studies have successfully predicted the permeability of uncored wells using ANN or fuzzy models with HU groups, but the models have been applied to uncored intervals or wells with the same depositional environment as the core data.…”