Method for Fracture Gradient Prediction for Vertical and Inclined Boreholes

Abstract

Abstract A new method for the prediction of fracture gradients in deeper wells has been developed. The method is based on the principles of mechanics but uses a correlation method in the application of field data. The new, improved method (1) gives both leakoff and lost-circulation pressures; (2) works for vertical and inclined boreholes; and (3) identifies lithological effects. The result is given as a simple equation where only well depth, pore pressure, borehole angle, and lithology are needed to predict the fracture pressure. Because borehole inclination is included, the method can be used during both production and vertical-well drilling. The method was applied in a field study offshore Norway. The results show a remarkably good correlation with field observations. Introduction Detailed knowledge of formation pore pressure and fracture strength is considered the most essential element in the achievement of a successful drilling program. This becomes even more critical in the drilling of high-angle wells. Methods to predict fracture gradients are typically based on empirical correlations between fracture data, overburden data, and depth. Refs. 1 through 7 give different methods of this nature. Daines' method in particular has been successfully applied in Norway by several oil companies. All these methods work for vertical wells. Bradleys and Aadnoy and Chenevert used the equations for the stress field around the borehole and were thereby able to study inclined boreholes by means of stress transformations. The mathematical model used, the "Kirch solution" for stresses in a plate with a hole in the middle, has certain limitations. It assumes homogeneous and isotropic rock properties, linear elasticity, and a plane/strain condition. The model works well for wells deeper than 600 m [1,970 ft]. Real rock can be described as heterogeneous and anisotropic with both spatial and directional variations in all its physical properties. Because of this complexity, a mathematical model that perfectly describes real rock does not exist. It has become evident that the ideal mathematical model does not work well for field calculations because it gives too-extreme results. In view of the complexity of modeling real rocks, it is easily understood that the model is too simple. This new method is derived to bridge the gap between field measurements and the principles of mechanics. The idea is to use the principles of mechanics but to adjust certain parameters to obtain a good fit to the field-measured fracture data. The criteria for these adjustments are based on observed behavior and are not always rational from a rigorous continuum mechanics point of view. The method, however, gives good results. The new method considers borehole inclination. Therefore, fracture data from production, wildcat, and delineation wells can all be used as input. The model is well suited for computers and may be updated continuously as new fracture data become available.