Identification and characterization of Hydraulic Flow Units in the San Juan Formation, Orocual Field, Venezuela

Abstract

This thesis focuses on the integration of core and well log data in order to provide a petrophysical characterization of the Hydraulic Flow Units (HFU) in the San Juan Formation, Orocual Field, Venezuela. We used three separate approaches to correlate/ integrate these data from Orocual Field. The first method uses the concept of Hydraulic Flow Units (HFU). The second method uses non-parametric transformation and regression, and the third method estimates permeability using polynomial and power law-type functions. We provide a comparison of these methods - and we believe that the HFU and non-parametric regression techniques give the most representative results. In this study we have core data for four wells and well log data for sixteen wells. The core permeability data were adjusted (in this work) to compensate for the Klinkenberg effect (a low pressure gas flow phenomena) and to relate porosity and permeability from laboratory to in-situ (net overburden stress) conditions. To model the stress correction we use the original reservoir pressure of 7470 psia and a two-point stress model. For well logs, we use a quality control method based on depth shift, depth match, environmental corrections, and core-log correlation. Finally, core permeability data were correlated to multiple well log suites (density, neutron, gamma ray, and resistivity). The HFU concept provides a deterministic approach for combining the geological attributes with available petrophysical data to delineate the reservoir into "units" of similar fluid flow characteristics. The HFU method does not presume a relationship between variables. The non-parametric regression/transform technique is based on using variable transformations to generate relationship between dependent and independent variables. This iterative, non-parametric procedure is called the Alternating Conditional Expectation (or ACE) method. Additional deterministic models based on polynomial, power law, and modified power law relations are also used to estimate permeability. The methodologies demonstrated in this work can be applied to any reservoir system having a significant population of permeability measurements and well log data.