Laboratory evaluation of proppant transport in complex fracture systems

Abstract

With the global trend towards exploring shale reservoirs over the past few decades, industry focus has shifted towards hydraulic fracturing. Nearly all shale gas wells are hydraulically fractured, with many treatments using slickwater, for production to be economically viable. Due to the presence of natural fractures, each shale reservoir is different; thus, the resultant hydraulic fracture network is distinct. With advances in fracture mapping and modeling techniques, the created complex fracture networks can be modeled fairly well; however, the proppant transport in these complex fracture networks is not clearly understood. Proppant transport is a central issue in slickwater fracture treatments because of the low-viscosity of the fracturing fluid. At this stage, field-scale tools are not available in the industry to estimate the propped fracture lengths and heights and to envisage whether the proppant "turns" the corner or not in presence of subsidiary fractures. Much speculation exists in the industry as to how efficiently the proppant is transported from the main fracture into subsidiary fractures, or if it is at all. An experimental study was undertaken to investigate proppant transport mechanisms in low-viscosity, high-rate slickwater hydraulic fracture treatments in shale reservoirs. Experiments were conducted in a low-pressure laboratory setting to evaluate proppant transport through a series of complex slot configurations. Different slickwater treatment scenarios were simulated by varying the pump rate, proppant concentration, and proppant size. A total of twenty-seven experimental tests were carried out using four different types of slot complexity to study the proppant transport behavior. The results of the experiments have provided some interesting insights into the nature of proppant transport and settling in complex fracture networks. While the proppant transport in the primary slot was observed to occur via traction carpet after the creation of a proppant dune, the proppant transport in secondary slots was found to be dependent on the threshold pump rate for the particular complex slot configuration. Two mechanisms were observed to be transporting the proppant into the secondary slots: 1) proppant flowing around the corner at pump rates higher than the threshold pump rate (related to the threshold velocity in the primary slot), and 2) proppant falling from the primary slot due to the effects of gravity, regardless of the pump rate. The proppant movement into the secondary slots was found to be dependent on the dune buildup in the primary slot. As a result, the secondary slots closer to the wellbore can be expected to be propped earlier in the treatment than those further away from the wellbore. Another important conclusion from this study is that the resultant fracture conductivity at the end of the hydraulic fracture treatment would vary across the fracture due to proppant segregation, especially when pumping smaller proppant sizes with broad sieve distributions. Recommendations for future experimental work are also presented. A better understanding of proppant movement in the complex fracture networks can possibly help with the designs of hydraulic fracture treatments by focusing on parameters that enhance transport and post-treatment analysis.