Well spacing and hydraulic fracture design have an enormous impact on the economic performance of wells in shale. Key design parameters include: (a) horizontal and vertical well placement, (b) stage length, (c) cluster spacing, (d) cluster shot count, diameter, and phasing, (e) proppant and fluid type, (f) proppant and fluid volume, and (g) injection rate. Case studies from the Bakken, Delaware Basin, Midland Basin, and Montney shale plays are presented. Numerical simulation is used to evaluate alternative options and maximize economic objectives.Numerical simulations are performed with a fully integrated hydraulic fracturing and reservoir simulator. Rather than performing a ‘handoff’ between two different codes, the simulator solves all governing equations (for multiphase flow, crack propagation, non-Newtonian fluid, and proppant transport) in a single system of equations and consistent mesh (wellbore, fractures, and matrix). The workflow involves: (a) construction of an initial model, (b) calibration to field diagnostics, (c) and economic maximization with an automated optimization tool. Diagnostics include: (a) fiber DAS and DTS, (b) microseismic, (c) geochemical analysis, (d) interference testing, (e) downhole imaging, (f) production data, and (g) pressure monitoring wells.Field observations suggest significant variability in fracture geometry between basins. Height growth varies dramatically and is driven by the vertical stress profile. In one dataset, recently developed ‘viscoelastic stress relaxation’ derived stress profiles are much more consistent with fracture geometry observations than the classical Eaton’s approach. Fracture length and effective toughness show a modest degree of variability between formations. Once calibrated, the models enable economic optimization. Optimization runs suggest opportunities to improve NPV (net present value) or DROI (discounted return on investment) from 10-60%. For some parameters (such as job size and well spacing), optimization results vary significantly depending on the objective function (either DROI, NPV, or NPV/section, with or without including the cost of land). The price of oil and gas also have a major impact on the optimal design. For other parameters, such as landing depth, stage length, and perforation cluster design, optimization results are relatively robust to the objective function and economic assumptions. We show how differences between basins and differences in economic inputs and objectives lead to significantly different optimal frac designs.
