Specifying Lengths of Horizontal Wells in Fractured Reservoirs

Summary New methods have been developed to constrain optimal horizontal drilling distance in fractured reservoirs in which opening-mode fractures are dominant. Studies of opening-mode fractures in Austin Chalk outcrops and core reveal that open fractures are commonly clustered, with the distance between clusters ranging from approximately 1 m to more than 300 m, depending on the horizon in question. Aperture-size distributions follow power laws, and spacing-size distributions are negative logarithmic or log-normal. The aperture size at which fractures are open to fluids is variable and site-specific (0.14 to 11 mm). Scaling properties of fracture attributes were used to calculate fracture permeability and to constrain well-length fracture-permeability relationships. Fracture permeability depends on the scale of measurement; it has been determined at 9.2 darcies for 14 m of lower Austin Chalk core and 286 darcies for 300 m of upper Austin Chalk outcrop. Introduction The Upper Cretaceous Austin Chalk, which crops out in a swath across central Texas, is renowned as a horizontal play and is well documented as such.1,2 Most workers regard Austin Chalk reservoirs as being low-porosity, fractured reservoirs, although there is debate concerning the relative storage capacities of matrix vs. fractures. When drilling a horizontal well in a fractured reservoir, the usual aim is to intersect fractures that are capable of providing a conduit for fluid flow. Although many horizontal wells have been drilled in the Austin Chalk,3 there are still questions over where it is best to locate new operations and how to optimize three critical parameters: wellbore azimuth, vertical depth, and wellbore length.4 This paper focuses on the question of wellbore length, although information pertaining to azimuth and depth choices also has been obtained. The choice of wellbore length has, in the past, been guided by experience and by field rules established by the Texas Railroad Commission, whereby the length of wells is linked to the acreage allocation of proration units and the permissible producing rate.4 Although these guidelines are practical, they lack direct geological input. The aim of this contribution is to develop techniques in which well-length determination is based on direct observation of fracture systems in the Austin Chalk, in addition to the Texas Railroad Commission guidelines. The objective of the outcrop and core studies was to characterize the opening-mode fracture system. Aperture-size distribution, spacing-size distribution, and fracture fill were determined in each case, allowing characterization of the spatial architecture of large, open fractures. This approach enabled us to calculate fracture permeability for different well lengths and to constrain optimal drilling distance for horizontal wells. The relationship between opening-mode fractures and normal faults in the outcrop is documented, and the relative importance of fractures and faults to reservoir permeability is considered. The connectivity and vertical height of fractures, and their impact on permeability, are discussed. Study Areas Data are presented from two outcrop analogs: one is near Waxahachie, north central Texas (Grove Creek); the other is from McKinney Falls State Park, central Texas (McKinney Falls), and from two laterals of a horizontal core drilled by the Kinlaw Oil Corp. in Frio County, Pearsall field (Kinlaw core) (Fig. 1). This well is currently operated by BASA Resources Inc. Although this study relates to the Austin Chalk specifically, the techniques used are transferable and could be applied in other horizontal targets. Geology The Austin Chalk is variable in terms of mineralogy, texture, and stratigraphy in part because of the effect of a basement high, the San Marcos Arch,5 on the paleobathymetry of its depositional basin. The updip portions of the Chalk in the Austin and San Antonio regions are relatively shallow water deposits containing considerable quantities of benthic skeletal material. Deeper-water planktonic microfossils and nanofossils dominate the basin equivalents, although some benthic material was transported basinward in debris flows.5 Drake6 reports the updip portions of the chalk in Burleson County, Giddings field, to be less fractured than the downdip portions, with wells in the updip portions being poor producers. At McKinney Falls State Park, a pavement in the McKown formation is exposed where Onion Creek flows over the lower falls. The McKown formation is a lateral equivalent of the Austin Chalk and comprises chalk intercalated with pyroclastic deposits derived from Pilot Knob, a Cretaceous volcanic center 3 km to the southeast.7 The Grove Creek outcrop is stratigraphically at the top of the Upper Chalk, just below the overlying Ozan formation. The McKinney Falls outcrop is close to the overlying Taylor Marl. The horizontal Kinlaw core from Pearsall field is from the base of the lower Chalk in the Atco Member. Thus, stratigraphically and with respect to the basin architecture, the studied sites are disparate. It is not the intention of this paper to make definitive recommendations for drilling distance in the Austin Chalk based on so few sites, but rather to show with these examples how site-specific information may be used to this end. Data-Collection Methodology An important consideration in fracture studies is whether the fractures observed in a particular core or outcrop are representative of those fractures that occur in the subsurface and contribute to fluid flow. In the case of core studies, the main pitfalls surround the distinction of natural fractures from those induced by drilling or by the core-handling process. Kulander et al.8 provided a comprehensive guide to natural and induced fracture identification in cores, and their criteria were used here. In outcrop studies, the challenge is to distinguish those fractures that would have been formed in the subsurface, at an appropriate depth to be considered as a reservoir analog, from those fractures that developed during uplift and erosion. The fracture systems documented here are confined to those that exhibit partial or total mineral fill and that would have developed in the subsurface.