Understanding the Mechanism of Long-Term Fracture Conductivity Decline

Abstract Fractured well performance largely relies on fracture conductivity that is created during hydraulic fracturing. As the well starts producing oil and gas, the decrease in reservoir pressure, formation stress impact, and proppant quality all contribute to fracture conductivity decline. This conductivity decline can result in a production decline of the fractured well and is sometimes detrimental to the well's performance. The conductivity decline can be caused by formation rock creeping, proppant embedment, or proppant crushing. In any case, these changes reduce fracture width, proppant pack permeability, or both. Understanding the decline mechanism and its impact on well production is vital for optimizing fracture design. This experimental study uses the standard API conductivity cell to systematically test short- to long-term fracture conductivity behavior. Samples from a Permian Basin unconventional reservoir were tested for this study. Fractures were created by saw-cutting the samples to create two smooth surfaces. Fracture conductivity tests were performed using dry nitrogen to isolate mechanical effects. The tests used 100-mesh regional sand at different concentrations. Unpropped fracture conductivity was also tested to separate the effect of rock properties from the proppant properties on fracture conductivity decline. Conductivity was measured from 500 to 6,000 psi, and pre- and post-test sieve analyses were used to quantify proppant degradation. An exponential decline correlation was applied to the short-term data to estimate zero-stress conductivity and to characterize stress-dependent conductivity reduction. The results show a clear separation between unpropped and propped behavior. The unpropped fracture exhibited low initial conductivity and only minor long-term reduction, consistent with limited creep-related narrowing of the fracture aperture during the 50-hour tests. Propped fractures exhibited higher initial conductivity but also greater stress-dependent and time-dependent losses. Higher proppant loadings produced higher zero-stress conductivities and higher decline rate, indicating that the higher the proppant loading, the more susceptible the fracture is to deformation under elevated stress. Sieve analysis confirmed grain crushing at higher stresses, supporting the observed conductivity degradation. These experiments provide clear evidence of how both stress and time contribute to fracture conductivity loss. The data show measurable creep behavior in both unpropped and propped fractures, with higher proppant loadings maintaining greater conductivity despite larger time-dependent changes. The results help explain how conductivity evolves after fracture closure and can be used to improve the prediction of long-term fracture performance in unconventional reservoirs.