Prediction of Fracture Pressures for Wildcat Wells

Abstract

Summary To date, fracture pressures have been predicted using Gulf coast-derived empirical formulae.1-3 Exploration is extending to high latitudes and deep waters, and the cost of drilling wells in these areas is becoming exceedingly high. The inability of the empirical relationships to predict adequately fracture pressures in areas other than the Gulf coast has prompted a reevaluation of the problem. Using laboratory-derived physical properties of typical sedimentary rocks, and taking Hubbert and Willis'4 minimum fracture-pressure model a step further, a hypothesis is proposed for predicting fracture pressures in a wildcat well after the first fracture test in compact formation. The model requires that pore pressures, overburden pressures, and lithology are known; with this information, fracture pressures may be predicted accurately at any point within the drilled hole. If the overburden-pressure gradient can be extrapolated and pore pressures estimated, and if lithologies are continuously available, fracture pressures actually may be predicted a lag time after the formation has been drilled. Initial testing of the model indicates that an accuracy greater than 95% may be obtained consistently. Introduction The cost of drilling wells in deep waters and high latitudes is becoming exceedingly high. Deep wildcatting in areas of poor geological control can be extremely hazardous and costly without adequate pore-pressure and fracture-pressure information. If abnormally high pore pressures are encountered, another casing string may be necessary; if the pressure zone is shallow in relation to the target, completion of the well could be jeopardized. Of prime importance in these wells is an accurate assessment of kick tolerance. This requires knowledge of the fracture pressures at any depth in the open hole. In the Gulf coast and other areas that have undergone extensive drilling, fracture pressures are predicted using empirical formulae.1-3 These can be applied with confidence in other areas of similar geological and tectonic regime only when sufficient drilling has allowed calculation of the necessary empirical constants. However, the absence of any method for predicting fracture pressures outside these areas has necessitated the use of those empirical formulae, with the general result that actual fracture pressures can be very different from calculated pressures. This is mainly due to the application of the empirically derived constants (usually representing the stress ratio), which are unrelated to the wildcat area. Accurate information on the in-situ principal stresses is vital for the solution of the fracture-pressure problem. None of the empirical formulae can predict accurately stresses in poorly explored regions. A hypothesis is proposed for resolving and extrapolating the local principal stresses after the first fracture test in compact formation. Compact is defined here as the point at which the sediment can transmit an applied stress through the grain contacts. Along with other pertinent data usually calculated on rank wildcats (i.e., overburden gradients and pore pressures), fracture pressures then can be obtained for any point within the drilled hole. Kick-tolerance calculations then become more realistic when they are based on fracture-pressure calculations for that specific well, so that in the event abnormal hole conditions are encountered, the chances of completing the well are greater than when relying on formulae containing unrelated empirical constants.