Applications of Surface Lineaments and Fractures for Optimal Horizontal Well Drilling and Production Potential Estimation

Abstract

Abstract This paper presents new techniques using surface lineaments and fractures to estimate the optimal drilling azimuth and the production potential for a horizontal well in a naturally fractured reservoir. Surface lineament and fracture data were collected and digitized. The surface linear features were mapped from aerial photos and/or satellite images in the midcontinent region. Probability functions of natural-fracture characteristics (orientation, length, and spacing) were obtained by best-fitting the surface lineament and fracture data. A new procedure is presented to determine the optimal azimuth for a proposed horizontal well in natural fracture systems so that it will intersect a maximum number of fractures. A new analytical inflow performance equation is derived to estimate the production potential of a horizontal well intersecting multiple random fractures. The new techniques presented in this paper provide a cost-effective approach for optimizing the orientation of a horizontal well and for estimating its production potential using surface lineaments and fractures. The probability functions of natural-fracture characteristics derived from a statistical analysis of the extensive surface lineaments and fractures collected in this study can be used for stochastic characterization of naturally fractured reservoirs. Introduction Horizontal drilling is becoming an increasingly effective technology for hydrocarbon exploration and production. It is particularly attractive in producing oil and gas from naturally fractured reservoirs. A horizontal well may provide a much higher production rate than a conventional vertical well if it is drilled perpendicular to natural fractures. However, in cases where a gas cap and/or an underlying aquifer is present, a horizontal well may be drilled parallel to the natural fractures to avoid extraneous gas and/or water production. The key to the success of a horizontal well drilling program in producing a naturally fractured reservoir is to obtain a realistic description of the subsurface natural fractures. Unfortunately, subsurface natural fracture data are rarely available and very expensive to obtain. A complete description of subsurface natural fracture systems in real reservoirs may be impossible. Many efforts have been made to use surface fracture data mapped from outcrop studies to derive the probability distributions of natural fracture characteristics. After being conditioned against a limited amount of subsurface fracture data acquired from cores, logs, and seismic interpretations, these distribution functions can be used to probabilistically optimize the azimuth, location, and attitude of a horizontal well to maximize its production potential. To optimize the azimuth of a horizontal well bore, one must have the knowledge of the number of fracture sets prevalent in the reservoir, and the spacing and preferred orientation of each set. To accurately assess the production potential of a horizontal well bore intersecting a natural fracture system, more detailed fracture data, as well as reservoir rock and fluid properties, are required. Orientation may be the best understood characteristic of natural fractures. Studies show that fracture orientations have a normal or Arnold's hemispherical normal distribution for two-dimensional data and a Fisher-Von-Mises distribution for three-dimensional data. Fracture lengths have been found by various investigators to have a log-normal, exponential, or power law distribution. The distribution functions which provide best fits to many fracture aperture data include normal and log-normal distributions. As the most extensively investigated characteristic of natural fractures, fracture spacing is found to have an exponential, log-normal, Weibull, Gamma, or power law distribution. Furthermore, fracture spacing appears to have a linear relationship with bed thickness when rock beds are relatively thin. This linear relationship has been used by many to develop techniques for estimating subsurface fracture spacing from core data. P. 923