Fracture conductivities have been extensively studied in laboratory to assess the impact of different fracturing fluids. However, if downhole shale samples are used, the comparison can be compromised or even biased due to significant sample variations caused by shale heterogeneity. The heterogeneity often leads to quite different mineralogy and mechanical properties among shale samples, even for those harvested from the same reservoir whole core. The variations in these properties have a strong impact on fracture conductivity, and create difficulties in evaluating and comparing different fracturing fluids.

In this work, the “half-core” approach is developed to obtain shale samples with almost identical properties to improve the evaluation on fracturing fluids. In this approach, shale cores are split into two half cores, which are then subjected to different fluids independently. 16 preserved shales from Barnett, Eagle Ford, Haynesville and Utica spanning a wide range of mineralogy were used to investigate the effectiveness of the approach. Mineralogy and mechanical properties on the half-core surfaces were compared. Measured and simulated fracture conductivities between paired half cores were also examined.

The half cores demonstrated similar mineralogy and mechanical properties. Among all the 16 selected shales, their paired half cores showed an average difference of 3.22 ± 1.90 % in mineralogy and of 6.89 ± 8.45 % in hardness. The differences were substantially reduced compared to the shale samples from a vertical well in the Utica Basin and those from a horizontal well in the Midland Basin, whose average mineralogical differences range from 30% to 40 %. Similar fracture conductivities were also found between the half cores subjected to a series of closure stress: difference in the measured fracture conductivities between half cores from an Eagle Ford shale varied from 1.1% to 10.7%; while the simulated conductivities between the 16 pairs of half cores presented an average difference of 15.3% ± 18.9%. The differences in both of the measured and simulated fracture conductivities between half cores were lower than the conductivity difference among shales when the traditional approach is used.

The half-core approach is experimentally proved to be effective in creating a baseline with reduced sample variation among shales to improve evaluation of fracturing fluids. The approach is applicable in both propped and unpropped fractures.