Geometry of Hydraulic Fractures From Modestly Deviated Wellbores

Abstract

Summary Model studies of the geometry of hydraulic fractures from modestly deviatedwells revealed that the rough surfaces seen previously are formed by thecoalescence of a number of starter fractures. At higher deviation angles, thestarter fractures do not link. The overall effect is a loss of wellproductivity resulting from restricted communication length and reducedeffective aperture. This paper presents some guidelines for paper presents someguidelines for minimizing this adverse effect. Introduction Historically, most hydraulically fractured wells have been vertical. An increasing number of deviated wells now require hydraulic fracturing, in somecases because of environmental awareness and legislation, which can restrictthe number and location of drillsites. and in others because of the developmentof offshore. low-permeability fields. Productivity, from a fractured deviated well is expected to be lower than that from a fractured vertical well. Fornormal reservoir depths. the axis of an ideal vertical well lies in the preferred fracture plane, defined as the plane perpendicular to the directionof the minimum horizontal rock stress and the fracture communicates with the wellbore over the entire perforated interval The axis of a deviated well, perforated interval The axis of a deviated well, however. rarely lies in thepreferred fracture plane. and the wellbore may intersect the fracture only overa limited reservoir interval. The limited cross-sectional area of fluids toenter the wellbore from the fracture leads to high local flow rates and hence additional pressure losses and reduced productivities. particularly inhigh-rate gas productivities. particularly in high-rate gas wells The economicsof South Ravenspurn, a southern North Sea gas field, depended heavily, on effective hydraulic fracturing to achieve high productivity increases. Thefield shape and platform locations were such that all parts of the field couldbe accessed by build-and-hold wells with deviations up to about 30 deg. Suchdeviations were likely to lead to limited-entry problems. An alternative was todrill S-shaped wells to return wells to near vertical through the reservoirinterval. Doing so increases well productivity and reduces the number of wellsrequired, but also increases drilling costs and risks. The closer the wellreturns to the vertical and the higher the step-out, the higher the cost of drilling an S-shaped well. An experimental study was conducted to definemaximum inclinations through the reservoir interval as a function of well direction for test conditions, representative of South Ravenspurn. This paper describes the experiments, discusses the results, and makes recommendations forthe fracturing of deviated wells of moderate step-out. Experiments Tests were conducted on 1/4-in.-diameter model boreholes in 12-in. cubes ofplaster of Paris (hydrostone) containing silica flour to increase strength andto decrease permeability. The cubes were loaded with a permeability. The cubes were loaded with a vertical stress of 1 = 2. 1 00 psi, a maximum horizontal stress Of 2 - 2, 1 00 psi, and a minimum horizontal stress Of 3 = 700 psi. Thefield effective s were about 6.000 psi vertical and 2,000 psi minimum psivertical and 2,000 psi minimum horizontal. The intermediate stress was unknown. The preferred hydraulic fracture plane, by definition, was perpendicular to eAppendix gives full experimental details. The communication length of thefracture with the wellbore, defined here as the length of wellbore in contact with the hydraulic fracture, depends on the well azimuth, measured from thedirection and the well deviation, (Fig. 1). A wellbore lying in the preferredfracture plane has a O degs azimuth. The following tests were carried out on both open and cased bore holes: vertical well -B=0 Deg.; =90 deg. with 10, 20, and 30degs; and 0=30* with a=O, 10, 20, and 30degs. The wellbore pressure wasrecorded at 10 Hz. After testing, each sample was examined in acomputerized-tomography (CT) scanner and then opened to reveal the fracturegeometry, which was subsequently photographed. The communication length of thephotographed. The communication length of the fracture with the wellbore wasmeasured. Results and Discussion Fracture Communication With the Wellbore. Tables 1 and 2 give the results ofthe open- and cased-hole experiments, respectively. The measured communication lengths are significantly longer than the minimum theoretical length, L, for aplanar fracture perpendicular to a3. From geometrical considerations, L-d/(sin ct sin 6), where d is the wellbore diameter. JPT P. 742