The Effect of Fracture Surface Roughness on Propped Fracture Conductivity Using 3D-Printed Fracture Surfaces

When rocks are fractured in tension, the fracture surfaces created are rough, with a wide range of surface morphologies possible. In previous studies of propped fracture conductivity using fractured samples, the fracture surface topography was found to have a strong influence on fracture conductivity and stimulation efficiency. Fracture surface patterns (relatively uniform, randomly rough, step changes, ridges and valleys) strongly affect propped fracture conductivity. Different types of surfaces can result in propped fracture conductivities differing by an order of magnitude or more for identical proppant loading conditions. To generate quantitative correlations including surface topographic effects, consistent samples with well-defined surfaces should be used in the experiments. However, when using actual rock samples to create realistic fracture surfaces by fracturing them in tension, the surfaces created are never the same, even using small samples all taken from the same block. This lack of repeatability in fracture surfaces greatly complicates identification of the effects of the rough surfaces on propped fracture conductivity. To overcome this, we created repeatable rough fracture surfaces using 3D-printing technology. First, we geostatistically generated a numerical depiction of a rough fracture surface. Then the surface was printed with resin using a 3D-printer. The hardened resin model of the rock sample was used to make a mold, which was in turn used to create a rock sample made of cement. High strength cement was used so that the samples had similar mechanical properties to unconventional reservoir rocks. With this methodology, we created multiple samples with identical surface roughness and features, allowing us to isolate and test other parameters, such as proppant size and concentration. Fracture conductivity tests were conducted using a modified API conductivity cell and artificial rock samples that are nominally 7 inches long and 2 inches wide. A well-established protocol to generate propped fracture conductivity as a function of closure stress was employed to test three different proppant concentrations on identical rough surfaces. For all three experiments, 100 mesh sand was used. The study demonstrates how proppant concentration affects propped fracture conductivity behavior in a systematic way.