Impact of Organics and Carbonates on the Oxidation and Precipitation of Iron during Hydraulic Fracturing of Shale

Hydraulic fracturing of unconventional hydrocarbon reservoirs is critical to the United States energy portfolio; however, hydrocarbon production from newly fractured wells generally declines rapidly over the initial months of production. One possible reason for this decrease, especially over time scales of several months, is the mineralization and clogging of microfracture networks and pores proximal to propped fractures. One important but relatively unexplored class of reactions that could contribute to these problems is oxidation of Fe(II) derived from Fe(II)-bearing phases (primarily pyrite, siderite, and Fe(II) bound directly to organic matter) by the oxic fracture fluid and subsequent precipitation of Fe(III)-(oxy)hydroxides. The extent to which such reactions occur and their rates, mineral products, and physical locations within shale pore spaces are unknown. To develop a foundational understanding of potential impacts of shale iron chemistry on hydraulic stimulation, we reacted sand-sized (150-250 mu m) and whole rock chips (cm-scale) of shales from four different formations (Marcellus Fm., New York; Barnett Fm., Central Texas; Eagle Ford Fm., Southern Texas; and Green River Fm., Colorado) at 80 degrees C with synthetic fracture fluid, with and without HCl. These four shales contain variable abundances of clays, carbonates, and total organic carbon (TOC). We monitored Fe concentration in solution and evaluated changes in Fe speciation in the solid phase using synchrotron-based techniques. Solution pH was the most important factor affecting the release of Fe into solution. For reactors with an initial solution pH of 2.0 and low carbonate content in the initial shale, the sand-sized shale showed an initial release of Fe into solution during the first 96 h of reaction, followed by a plateau or significant drop in solution Fe concentration, indicating that mineral precipitation occurred. In contrast, in reactors with high pH buffering capacity, little to no Fe was detected in solution throughout the course of the experiments. In reactors that contained no added acid (initial pH 7.1), there was no detectable Fe release into solution. The carbonate-poor whole rock samples showed a steady increase, then a plateau in Fe concentration during 3 weeks of reaction, indicating slower Fe release and subsequently slower Fe precipitation. Synchrotron-based X-ray fluorescence mapping coupled with X-ray absorption spectroscopy (both bulk and micro) showed that when solution pH was above 3.25, Fe(III)-bearing phases precipitated in the shale matrix. Initially, ferrihydrite precipitated on and in the shale, but as experimental time increased, the ferrihydrite transformed to either goethite (at pH 2.0) or hematite (pH > 6.5). Additionally, not all of the released Fe(II) was oxidized to Fe(III), resulting in the precipitation of mixed-valence phases such as magnetite. Idealized systems containing synthetic fracture fluid and dissolved ferrous chloride but no shale showed that in reactors open to the atmosphere at low pH (<3.0), Fe(II) oxidation is inhibited. Surprisingly, the addition of bitumen, which is often extracted by organic compounds in the fracture fluid, can override this inhibition of Fe(II) oxidation caused by low pH. Nonetheless, O-2 in the system is still the most important factor controlling Fe(II) oxidation. These results indicate that Fe redox cycling is an important and complex part of hydraulic fracturing and provide evidence that Fe(III)-bearing precipitates derived from oxidation of Fe(II)-bearing phases could negatively impact hydrocarbon production by inhibiting transport.