Determining Fracture Pressure Gradients From Well Logs

Abstract

An expression for computing fracture pressure gradient in Gulf Coast sands as a function of overburden pressure, pore pressure, porosity, Poisson's ratio, and well depth is developed using Biot's stress/strain relationships. Then, using field data, an empirical relationship is established between Poisson's ratio and the shaliness of the sand as derived from well logs. Introduction While drilling a well, it is often necessary to increase mud weight to contain high pore pressures. A potential hazard arises because an increase in mud weight may cause one of the exposed formations to fracture, resulting in loss of circulation. Lost circulation is a dangerous and troublesome problem, often taking several days to correct, and in some cases leading to the loss of a well. The usual preventive measure is to set a protective casing string. When and where to set casing are important decisions governing the cost and safety of drilling a well, and these decisions clearly require an accurate knowledge of pore pressure gradients and fracture pressure gradients. Considerable work has been done in the last few years in the area of determining fracture pressure gradients. The numerous methods proposed vary greatly and point up the fact that the oil industry still needs an accurate and reliable answer to the problem. Following the work done by Hubbert and Willis in 1957 and Matthews and Kelly in 1967, Eaton in 1969 published a technique to predict fracture pressure gradients. In each of the above methods, the formation properties that influence the fracture pressure either are ignored or are assumed to be monotonic functions of depth. Matthews and Kelly used an empirical "matrix stress coefficient," which is a monotonic function of depth, to represent formation properties. Eaton introduced Poisson's ratio as an important elastic constant and made use of Poisson's ratio as an important elastic constant and made use of the concept of a variable overburden. However, he again resorted to general trends to describe Poisson's ratio as a function of depth. In their treatment of formation properties, both methods are essentially the same. For preparing a drilling plan, these rather general approaches for representing the influence of formation properties are valuable to the drilling engineer. Once the general trends have been established in a given field, these methods can be used to estimate fracture pressures before the well is drilled. Thus maximum mud weights can be reasonably forecast; and with some knowledge of the pore pressure trend in the field, casing points can be roughly estimated. However, it will be shown that the fracture pressure gradient can vary significantly at a given depth in the same geological field. Some means of identifying weak and strong sands, while the well is being drilled and logged, would therefore be valuable. For example, an exposed weak sand just below a casing string severely restricts mud weights during attempts to drill as far as possible into an overpressured transition shale. Conversely, there is no point in setting a casing string deeper to protect a strong sand. point in setting a casing string deeper to protect a strong sand. JPT P. 1259