Fracture-Gradient Prediction for Offshore Wells

Abstract

Summary. A key parameter required in the design of offshore wells is the fracture gradient of the formations to be penetrated. Existing fracture-gradient correlations have been developed largely on the basis of empirical data gathered on land and in the nearshore marine environment. They have been extended to deepwater areas through an equivalent intergranular stress concept. In this study, offshore data from several operators were obtained and a new fracture-gradient correlation procedure, based on an assumption of an exponential decline in average porosity with depth in normally pressured formations, was applied. The technique is ideally suited for implementation on a hand-held calculator or personal computer. Introduction The prediction of fracture pressure gradients is as important as predicting formation pressures in well planning of today's deep wells. Unless fracture pressure is taken into account, expensive and troublesome problems develop. Lost circulation and associated problems have been disastrous in some cases. When abnormally pressured zones are encountered, the design of the casing strings to be cemented in the well is of primary importance. Also, as drilling into the abnormal zone continues, the required number of casing strings and the cost of the well increase. Hence, knowledge of the formation fracture pressure at any depth is essential for planning and drilling wells into abnormally pressured formations. Several commonly used methods are available to predict fracture pressures. Hubbert and Willis published a classic paper introducing many of the fundamental principles still widely used today. Over-burden stress gradient, formation pore-pressure gradient, and Poisson's ratio of rocks were the variables shown to control fracture pressures. Matthews and Kelly replaced the assumption of Hubbert and Willis that the minimum matrix stress was one-third of the overburden stress with a matrix stress coefficient, K, determined empirically from field data. The matrix stress coefficient was correlated to equivalent normal-pressure depth and thus depends on both depth and pore pressure. Pennebaker's correlation was similar to Matthews and Kelly's except that K was called the effective stress ratio and was correlated with depth regardless of pore pressure. Christman presented an offshore fracture-gradient prediction method based on work in the Santa Barbara channel. Results of this work were at depths less than 5,000 ft [less than 1524 m]. Eaton also published a procedure in which the ratio of minimum matrix stress to vertical matrix stress was a single function of depth regardless of pore pressure. Other correlations have also been presented. This paper extends and modifies the Eaton correlation to include offshore wells for the prediction of fracture pressure gradients where the only variables for a given area are depths and pore pressure. The effect of ocean depth on overburden stress, which is often neglected when land-based correlations for offshore areas are used, was found to be quite significant. The resultant curve-fit equations are of a form such that bulk density cannot exceed grain density when extrapolated to great depths. Also, the model requires that the ratio of effective intergranular stress to vertical intergranular stress cannot extrapolate to values greater than unity at great depths. Values for the matrix stress ratio may be computed from field data and curve fitted for the method presented, or the one developed here from Eaton's data may be used. Overburden Stress Normal formation pressure is maintained only if a sufficiently permeable pathway allows formation fluid to escape. This is shown as a simple piston model in Fig. 1. Under normal pressurization, when the "valve" is open, the pore pressure will remain at hydrostatic pressure. The "pistons" are loaded by the weight of the overburden at a given depth. Resisting this load are the vertical matrix stress, a, and the pore-fluid pressure, p. Thus,If the "valve" is closed, however, the increasing overburden stress with depth will pressurize the pore water above hydrostatic pressure and hence be abnormally pressured. Along with these compaction effects, diagenetic effects contribute to abnormal pressure in the formation. During compaction, the water loss from montmorillonite clays during conversion to illite can contribute to the pressure increase. Also, shales accept water under pressure by reverse osmosis, and the precipitates of silica and carbonates would cause the upper part of the zone to become dense and impermeable. This may also occur in other rock types. If the structure has significant dip and the pore fluid has a density less than the normal pore-fluid density of the area, as in gas wells, abnormal pressures can be encountered in the updip portion of the structure. Finally, fluid migration from a deep reservoir to a shallow formation can result in the shallow formation becoming abnormally pressured or "charged," as is common above old fields.The vertical overburden stress, u, caused by the geostatic load at sediment depth may be represented asThe bulk density is related to grain density, pore-fluid density, and porosity at a given depth byor in terms of porosity,which will allow for evaluation of the porosity from average bulk density data from well logs for any assumed grain density and fluid density. The trend of average porosity vs. depth may be represented bybecause a good straight-line trend is usually obtained on semilog plots. Because of this, it was easier to curve fit the porosity trend than the bulk-density trend. SPEDE P. 136^