A New Approach to Measuring Rock Properties Data from Cores Cuttings for Reservoir Completions Characterization: an Example from the Bakken Formation

摘要

The application of rock physics data for improved reservoir characterization is well known and documented and, can help to delineate fluid and facies changes down the borehole. With the advent of unconventional resource exploration & production, and the adoption of long laterals to maximize reservoir contact and recovery, use of elastic rock properties for completions characterization becomes increasingly important. Traditionally, rock elasticity has been determined through geomechanical laboratory measurements on core pieces or plugs and log derived calculations. Unfortunately, these data sets are rare (i.e. core) and inconsistent/expensive to collect (i.e. logs) along long laterals. However, cuttings are always available for the entire borehole. They provide a source of compositional and textural rock data that can be quantitatively measured to train standard petrophysical analysis in order to implement rock physics equations (e.g. Young’s Modulus, ν, λ, λ-ρ, and μ-ρ). Using electron beam (e-beam) systems, such as RoqSCAN~(TM), mineral composition data, together with high resolution textural information (e.g. pore volume, pore fabric, pore size distribution and pore aspect ratio) within cuttings are now directly measureable. This allows for input parameters (e.g. pore aspect ratio) to be directly measured within rocks rather than mathematically derived. The first phase of this project involved the collection of all available well log data for well Sidonia-106H, which penetrated the Bakken formation in Montrail county, North Dakota, by EOG Resources Inc. The well logs retrieved included a Gamma Ray, Sonic, Resistivity, Neuton Porosity and Density. These logs were loaded into a petrophysical modeling package, PowerLog~R, and through the use of stochastic modeling a mineralogical composition of the well was calculated based on the well logs input. The Statmin model also calculated kerogen, bound water and bound oil content within the Bakken formation. Based on the results from the stochastic model the formation was separated into upper, middle and lower Bakken. The resultant model output for the vertical section intersecting the Bakken formation is shown in Figure 1, below. The log clearly shows the modeled mineralogy of the upper and lower Bakken being predominately clay rich with secondary sand (quartz), while the middle Bakken is classified as dolomitic sandstone. Using the modeled mineralogical composition, PowerLog~R was then used to calculate various elastic rock properties, including Young’s Modulus and Poisson’s Ratio (Figure 1). The results from this initial log data modeling shows a clear difference in the acoustic properties within the upper and lower Bakken compared to the middle Bakken. However, though there are features with some of the derived elastic properties, λ and Young’sModulus, thePoisson’sRatiodoesnot reflect the same levelofdetail,hostingonly minordipsandpeakswithinarelativelycontinuoustrace.