Transport Behavior of Resin-Coated Ceramic Proppants in Rough Vertical Fractures

Abstract

In hydraulic fracturing, resin-coated ceramic proppants can be added to fracturing fluid as an agent for propping the fractures. Resin-coated ceramic proppants have several advantages compared to the traditional silica sands and ceramic proppants. Resin-coated ceramic proppants can withstand much higher pressure than silica sands, while they have a lower density than the ceramic proppants. The transport behavior of resin-coated ceramic proppants in fracturing fluid is seldom investigated in the past. In this study, we aim to investigate the transport behavior of resin-coated ceramic proppants in rough vertical fractures. First, we conduct experiments to measure drag coefficients of resin-coated ceramic proppants during their settling in static water. Eight resin-coated ceramic particles are selected in the tests. Using the high-resolution images obtained from Computer Tomography (CT) scan, we measure the following for the eight resin-coated ceramic particles: bulk volume, mean diameter and volumetric fractions of three constituents making up each particle (i.e., resin coating, ceramic body, and air pockets). CT scan shows that the resin-coated ceramic particles are nearly spherical particles. High-precision electronic balance is used to accurately measure the mass of the tested particles. The densities of resin-coated ceramic particles are dependent on the volume of resin-coated ceramic particles. Three methods are applied to estimate the volume of particles. The settling velocities of resin-coated ceramic particles are measured by recording the settling process of each particle using a high-speed camera. After obtaining the diameter, density, and settling velocity of each particle, the drag coefficient of each resin-coated ceramic particle can be determined. Such determined drag coefficient is then compared to those predicted by five empirical correlations in the literature. The comparison shows that the drag coefficients of resin-coated ceramic particles agree generally well with the drag coefficient estimated by the five empirical correlations. In order to quantify the accuracy of the particle densities estimated by the three methods, we compare the densities estimated by the above three methods against the ideal particle densities estimated by the empirical drag coefficient correlations. The comparison shows that the method #3 (i.e., CT scan) leads to the minimum discrepancy between the estimated densities and the ideal particle densities. In addition, Roos and Willmarth’s correlation is shown to be more appropriate for calculating the drag coefficients of resin-coated ceramic particles than the other four correlations. Next, we conduct dynamic flow tests to examine the transport behavior of resin-coated ceramic proppants in three rough fracture models which are replicates of a beige limestone, a coarse-grained white marble, and a holocrystalline amphibole granite. In the experiments, a fluid carrying a given concentration of proppants is allowed to flow through the rough fracture model; the settling behavior of proppants and the relative area covered by the proppants in the fracture model are continuously monitored. Major influential factors on the proppants transport behavior have been examined. At a given time, the relative coverage of resin-coated ceramic proppants obtained by injecting resin-coated ceramic proppants through the top injection point is larger than that obtained by injecting resin-coated ceramic proppants through the bottom injection point. The area occupied by resin-coated ceramic proppants is much larger than that occupied by silica sands at a given time. The relative coverage of resin-coated ceramic proppants carried by slickwater is lower than that of resin-coated ceramic proppants carried by tap water at a given time. A higher flow rate can transport the proppants into deeper locations in the fractures, resulting in a lower relative coverage of proppants in the fracture models due to the limited length of the fracture models at a given time. A particle size of 20-40 mesh gives a higher relative proppant coverage in the fracture models at a given time than the 30-50 mesh resin-coated ceramic particles. Besides, at a given time, the relative coverage obtained in experiments using a 2-mm-aperture fracture model is larger than that obtained in experiments using a 4-mm-aperture fracture model. Among the three fracture models, the highest relative coverage of resin-coated ceramic proppants is obtained in Fr.4 (a replication of coarse-grained white marble). The highest injection pressure can be found in Fr.1 (a replication of beige limestone with abundant coarse fossil shells).