Niobrara discrete fracture network: from outcrop surveys to subsurface reservoir models

Abstract

Heterogeneity of an unconventional reservoir is one of the main factors affecting production. Well performance depends on the size and efficiency of the interconnected fracture “plumbing system”, as influenced by multistage hydraulic fracturing. A complex, interconnected natural fracture network can significantly increase the size of stimulated reservoir volume, provide additional surface area contact and enhance permeability. In 2013 the Reservoir Characterization Project (RCP) at the Colorado School of Mines began Phase XV to study Niobrara shale reservoir management. Anadarko Petroleum Corporation and RCP jointly acquired time-lapse multicomponent seismic data in Wattenberg Field, Denver Basin. Anadarko also provided RCP with a regional 3D seismic survey and a rich well dataset. The purpose of this study is to characterize the natural fracture patterns occurring in the unconventional Niobrara reservoir and to determine the drivers that influenced fracture trends and distributions. The findings are integrated into a reservoir model though DFN (Discrete Fracture Network) for further prediction of reservoir performance using reservoir simulations. Aiming to better understand the complexity of the natural fracture system I began my fracture analysis work at an active mine site that provides a Niobrara exposure. Access to a “fresh” outcrop surface created a perfect natural laboratory. Ground-based LIDAR and photogrammetry facilitated construction of a geological model and a DFN model for the mine site. The work was carried into subsurface where the information gained served to improve reservoir characterization at a sub-seismic scale and can be used in well planning. I then embarked on a challenging yet essential task of outcrop-to-subsurface data calibration and application to RCP’s Wattenberg Field study site. In this research the surface data was proven to be valid for comparative use in the subsurface. The subsurface fracture information was derived from image logs run within the horizontal wellbores and augmented with microseismic data. Limitations of these datasets included the potential to induce biased interpretations; but the data collected during the outcrop study aided in removing the bias. All four fracture sets observed at the quarry were also interpreted in the subsurface; however there was a limitation on statistical validity for one of the four sets due to a low frequency of observed occurrence potentially caused by wellbore orientation. Microseismic data was used for identification of one of the reactivated natural fracture sets. An interesting phenomenon observed in the microseismic data trends was the low frequency of event occurrence within dense populations of open natural fracture swarms suggesting that zones of higher natural fracture intensities are capable of absorbing and transmitting energy resulting in lower levels of microseismicity. Thus currently open natural fractures could be challenging to detect using microseismic. Through this study I identified a significant variability in fracture intensity at a localized scale due to lithological composition and structural features. The complex faulting styles observed at the outcrop were utilized as an analog and verified by horizontal well log data and seismic volume interpretations creating a high resolution structural model for the subsurface. A lithofacies model was developed based on the well log, core, and seismic inversion analysis. These models combined served to accurately distribute fracture intensity information within the geological model for further use in DFN. As a product of this study, a workflow was developed to aid in fracture network model creation allowing for more intelligent decisions to be made during well planning and completion optimization aiming to improve recovery. A high resolution integrated discrete fracture network model serves to advance dynamic reservoir characterization in the subsurface at a sub-seismic scale resulting in improved reservoir characterization.