Actual and Optimal Hydraulic-Fracture Design in a Tight Gas Reservoir

Abstract

Designing hydraulic-fracture stimulation to optimize well productivity requires a carrier fluid with a suitable leakoff coefficient and knowledge of the initial reservoir pressure and permeability, the closure-stress magnitudes in the reservoir and in the bounding formations, and rock properties such as the Young's modulus and the Poisson's ratio, which can be determined from cores. When key parameters are left unknown, the hydraulic-fracture stimulation is likely to be severely suboptimal.This study integrates pressure-buildup and production transient analyses with microseismic surveys and the recorded pumping schedule to estimate the aforementioned parameters in previously fractured wells in a tight gas reservoir that have been on production for up to 8.5 years.The first well drilled and completed in the block included a pressure-buildup test that enabled accurate estimation of the initial reservoir pressure and permeability. A post-fracture buildup test was also conducted, and annual pressure-buildup tests in the subsequent 6 years showed continuous changes in the fracture morphology, with fracture conductivity decreasing by a factor of three and fracture length increasing by approximately 50%. Many of the subsequent wells were drilled in two patterned well clusters, each designed to account for fracture-propagation behavior indicated from a microseismic survey. A comparison with an optimal hydraulic-fracture design that was intended to maximize well productivity indicates that most of the well stimulations were suboptimal, with rate and cumulative production approximately one-half that of an optimized design on the basis of the same proppant mass.The observed changes in fracture conductivity and length over time were not anticipated. Because such data are rarely recorded, the variations in fracture morphology may be fairly typical and should be of considerable interest to pressure-transient analysts.The production-data analysis shows the difficulties in determining formation and fracture parameters when the transient response lacks radial flow. The fracture-treatment analysis shows a comparison between the actual fracture treatment and one designed to maximize well productivity, and clearly illustrates the potential for well improvement through the use of modern hydraulic-fracture-design principles.

A numerical investigation on the performance of hydraulic fracturing in naturally fractured gas reservoirs based on stimulated rock volume

“…In moderate permeability reservoirs, the fracture accelerates production without impacting the well reserves. In low-permeability reservoirs, hydraulic fracture contributes both to well productivity and to the well reserves, because in such reservoirs the well would not produce an economic rate without the hydraulic fracture (Pang et al 2016). In homogenous reservoir with planar fracture geometry, there exists an optimum fracture dimension that would provide maximum reservoir performance for a given amount of proppant (Wei and Economides 2005).…”

Effect of Pump Schedule on Fracture Geometry and Shape during Frac Packing Job

“…First, using reservoir simulation and according to geological conditions, well pattern, well spacing and well density the optimal fracture half-length and fracture conductivity for the well were found 16 m and 200~300 μm 2 .cm respectively; then the following steps were performed:…”

“…When key parameters are left unknown, the hydraulic-fracture stimulation is likely to be severely suboptimal [2].…”

Prediction of Refracturing Effect of Tight Gas Reservoirs Based on Bayesian Inversion Algorithm

“…Foreign research on the selection of refracturing well layers is earlier, and after a lot of theoretical research and practice, a prediction method with considerable practical significance has gradually been formed. Some scholars have applied the typical curve method, statistical method, and artificial intelligence method to the selection of refracturing wells and layers and applied numerical simulation technology to compare the three methods [ 15 , 16 ]. It was later concluded that the artificial intelligence method was the best forecasting method used at the time.…”