Abstract

This paper describes and critically assesses a common methodology currently used to model hydraulic fractures in geologically complex, fluvial, tight gas reservoirs. A planar 3-D fracture simulator is used with a fully coupled fluid/solid transport simulator. The model incorporates a unique data set from the Piceance basin, Colorado, which produces hydrocarbons from the Cretaceous-age Mesaverde formation. Initially, vertical variations in geo-mechanical rock properties (Young's modulus, Poisson's ratio and Biot's constant) were calculated from well logs. The results were then compared with previous work undertaken on the Mesaverde formation and carried out at the DOE/GRI MWX site. From this analysis, specific correlations were developed for rock properties derived from well logs on a foot-by-foot basis to be used in the hydraulic fracture model. Diagnostic mini-frac injection tests of individual sandstone reservoirs were used to confirm model inputs and develop a valid stress model.

Previous attempts to model hydraulic fracture growth in the Mesaverde have been hampered by a lack of detailed input data sets and the inability to accurately determine horizontal rock property variations. This paper outlines a method which uses micro-seismic/tiltmeter data to constrain and verify the model inputs. The resulting frac model is shown to have not only matched the fracture containment but also pressure matched the actual net surface pressure data in this extremely geologically complex area. From these results it is possible to get a better understanding of how fracs grow and interact with complex fluvial reservoirs, allowing operators to better optimize field well performance and completion methods in these geologic settings. Additionally, the minimum critical data required to develop such a model has been identified and will aid operators in developing their data acquisition programs. Although developed in the Rocky Mountain region, the presented technique can be extrapolated to other similar geologically complex reservoirs world-wide.

Introduction

Despite its common use, hydraulic fracturing still remains one of the most complex and least understood practices employed in the oil industry. For a complete hydraulic fracture evaluation, the engineer has to fully evaluate the well potential, as well as the effectiveness of the treatment design for creating the desired fracture. Simple layer-type reservoir models often prove to be inaccurate as the oversimplified reservoir descriptions frequently result in overestimated well productivity. Complex reservoirs require both an in-depth knowledge of fluid mechanics and the reservoir rock mechanics; requiring an inter-disciplinary approach to reservoir analysis, to generate the necessary input data for effective treatment design. Previously, a major problem for hydraulic fracture theory development was the lack of suitable models for application in heterogeneous reservoirs. Early models were two-dimensional and incorrect assumptions were often made. It is only recently that the hydraulic fracture engineer has had the tools available to effectively model and analyze the hydraulic fracture process in complex systems. This has been made possible by the advent of fully three-dimensional models, as well as improvements in reservoir analytical techniques.

The overall objective of the study is to investigate a methodology currently used to develop what practitioners consider to be accurate three-dimensional (3D) hydraulic fracture (HF) simulations of geologically complex reservoirs. Previous work has been carried out in the Piceance basin1 and other geologically similar tight gas reservoirs2 using pseudo-three dimensional models. However, the unique multi-disciplinary data set available for this investigation allows for a complete evaluation of a current "best practice" methodology. The techniques used are those presently applied to generate input data to build an accurate model to allow practitioners to simulate hydraulic fractures in complex, fluvial, tight gas reservoirs. The simulator outputs were initially matched with direct diagnostic results from fracture mapping, to help constrain the model and aid in the simulator output matching process. This work is part of ongoing research to help operators and researchers identify the minimum data necessary to effectively model and optimize hydraulic fracture treatments in geologically complex reservoirs.

This content is only available via PDF.
You can access this article if you purchase or spend a download.