Summary
Horizontal wells with multiple hydraulic fractures have become a standard completion for the development of tight oil and gas reservoirs. Successful optimization of multiple-fracture design on horizontal wells began empirically in the Barnett Shale in the late 1990s (Steward 2013; Gertner 2013). More recently, research has focused on further improving fracturing performance by developing a model-derived optimum. Some researchers have focused on an economic optimum on the basis of multiple runs of an analytical or numerical model (Zhang et al. 2012; Saputelli et al. 2014). With such an approach, a new set of model runs is necessary to optimize the design each time the input parameters change significantly. Running multiple simulations for every optimization case might not always be practical. An alternative approach is to develop well-performance curves with dimensionless variables on the basis of the performance model. Such an approach was the basis for unified fracture design (UFD) for a single fracture in a vertical well (Economides et al. 2002). However, a similar systemized method to calculate the optimum for a horizontal well with multiple hydraulic fractures was missing.
The objective of this study was to develop a rigorous and unified dimensionless optimization technique with type curves for the case of multiple transverse fractures in a horizontal well—an extension of UFD. The mathematical problem was solved in dimensionless variables. Multiple fractures include the proppant number (NP), penetration ratio (Ix), dimensionless conductivity (CfD), and aspect ratio (yeD) for each fracture, which is inversely proportional to the number of fractures. The direct boundary element (DBE) method was used to generate the dimensionless productivity index (JD) for a given range of these parameters (28,000 runs) for the pseudosteady-state case. Finally, total well JD was plotted as a function of the number of fractures for various NP. The effect of minimum fracture width was studied, and the optimization curves were adjusted for three cases of minimum fracture width.
The provided dimensionless type curves can be used to identify the optimized number of fractures and their geometry for a given set of parameters, without running a more complicated numerical model multiple times. First, the proppant mass (and hence, NP) used for the fracture design can be selected on the basis of economic or other considerations. For this purpose, a relationship between total JD and NP, which accounts for the minimum fracture width requirement, was provided. Then, the optimal number of fractures can be calculated for a given NP using the generated type curves with minimum width constraints.
The following observations were made during the study on the basis of the performed runs:
For a given volume or proppant, NP, total JD for multiple fractures increases to an asymptote as the number of fractures increases. This asymptote represents a technical potential for multiple fractures and for high proppant numbers (NP≥100), with a technical potential of 3πNP. Below this asymptote, the more fractures that are created for a fixed NP, the larger the JD. In practice, minimum fracture width constrains the fracture geometry, and therefore maximum JD. For the case when 20/40 sand is used for multiple hydraulic fracturing of a 0.01-md formation with square total area, the optimal number of factures is approximately NP25. Application of horizontal drilling technology with multiple fractures assumes the availability of high proppant numbers. It was shown mathematically that the alternative low proppant numbers (NP≤20 for the previous case) are impractical for multiple fractures, because total JD cannot be significantly higher than JD for an optimized single fracture in the same area. This means that low formation permeability and/or high proppant volumes are needed for multiple fracture treatments.
Semi-analytical model of multiple fractured horizontal well across a discontinuity in a linear composite reservoir
