(715a) Geochemical Impacts on Mineral Scaling in Unconventional Shales: Interplay between Mineralogy, Additives, and Base Fluids

Unconventional oil/gas shale production has led to significant reductions in CO₂ emissions over the last decade while also providing US energy independence. Though advances in hydraulic fracturing have resulted in extraction of resources previously thought unobtainable, this stimulation technique leaves significant resources underground. Part of this is due to the low porosity/permeability of the rock, and mineral scaling caused by injection fluids. Contrary to conventional oil/gas extraction, unconventional extraction relies on high-pressure injection of fluids (acid, additives, and base fluids) and proppants to fracture rock, ease extraction of hydrocarbons, and maintain apertures of newly formed fractures. Additionally, most wells have shut-in periods between the end of injection and start of production that can last weeks to months. The injection and shut-in of fluids results in a wide array of chemical reactions to the rock that are both beneficial and deleterious. The increase of porosity/permeability due to acid injection at the beginning of a stimulation is beneficial, but can lead to deleterious mineral scaling as solution conditions change.

We have adopted a multi-faceted approach to identifying types of mineral scaling, its causes, and its effects on the shales caused by stimulation chemicals. Our approach includes batch reactions using ground and intact core material, synchrotron-based spectroscopy/imaging, reactive transport modeling, core flood experiments with contrast CT imaging, and acoustic monitoring of un/reacted samples. By using these complementary in situ/ex situ techniques, a clearer picture of what is occurring downhole in the stimulated rock volume is obtained.

Batch reactors for two major shale plays, Marcellus and Permian Basin, were conducted using shale and fracture fluid recipes from each respective geological area. Time-resolved and pressurized experiments were conducted using both ground shale (150-250 μm) and intact cores (1 cm x 1 cm). Experiments were conducted following a standard fluid injection sequence and proper mixing ratios at 80^oC for 3-weeks to mimic field practices. Post-experiment analyses using synchrotron-based techniques show that pyrite contained in the shale partially dissolved and the released Fe(II) oxidized resulting in the precipitation of a new mineral scale in the form of Fe(III)-(hydr)oxides. The types of Fe(III)-bearing particles formed varies depending on the carbonate concentration (i.e. buffering capacity) of the shale. These experiments were conducted using doubly de-ionized (DDI) water. Another set of experiments conducted using Permian Basin samples and water from the field, both municipal freshwater and clean brine, showed significant sulfate mineral precipitation on clay-rich samples as well as Fe(III)-bearing mineral precipitates when clean brine was used. Our results show that mineral scaling occurs on both the fracture surface and within the shale matrix caused not only by alterations to the matrix but also by reactions between the injection fluid additives and the shale.

We leverage these observations to inform forward, predictive models merging the principles of transport and transformation. Such reactive transport software capabilities have been developed largely in application to contaminant hydrology, but the same principles are readily adapted to embed process understanding of mineral scaling into models encapsulating the open, transient conditions of unconventional shale stimulation. This parlay of molecular scale information into environmentally relevant context via forward models offers a means of describing how a variety of mineral scaling reactions combine with the unique condition of each reservoir to impact reactive alteration of the shale, the attendant fluid transport properties and ultimately production. Thus far, we have built a suite of multi-component reactive transport simulations using the CrunchFlow software which encapsulate the mineralogy of Marcellus and Permian shales, and the principle fluid additives to reproduce the temporal evolution of reactive fronts developing in the shale as constrained by our synchrotron work. The models demonstrate that the solubilization of bitumen is vital to convert Fe(II) released via oxidative dissolution of sulfides to Fe(III) and thus acts as a mediator of the extent to which iron mineral scaling occurs. Sulfate minerals, including barite and celestite, have also been built into these forward model simulations, and suggest a close relationship between the depth of reactive fluid penetration, the content of easily solubilized carbonates, and the accumulation of this class of mineral scaling. Such predictive capability offers a rapid diagnostic of the effects of a given fracturing fluid composition on a given shale play and ultimately offers the potential to reduce the need for re-stimulation and thus the volume of water consumed during the production life cycle.

Reactive core flood experiments are the logical extension of batch reaction tests and provide data for matching of reactive transport models. Reactive core floods are, by their nature, long experiments as it takes time to inject a suitable amount of fluid into the low permeability system. We have examined clay-rich and carbonate-rich samples from the Marcellus shale. Samples have a distinct microcrack and matrix network. Alterations to shale were measured using computed tomography of samples before, during, and after experiments. Scanning electron microscopy was used for detailed views of the shale fabric before and after exposure to reactive fluid. Samples tested to date show a reduction in gas accessible porosity following reaction. A corresponding reduction in liquid permeability was measured. SEM-EDS (energy dispersive spectroscopy) of rock surfaces shows an iron-containing precipitate on and near fracture surfaces and in the shale matrix of the clay-rich sample indicating partial dissolution of pyrite and/or ferruginous dolomite followed by precipitation of iron (hydro)oxide. Extensive imaging reveals fracture filling with migrated and/or precipitated fine particles. Significant barite scale growth was detected on the reacted surfaces of the carbonate-rich sample. Halite and (hydro)oxide precipitates resulting from geochemical reactions between the basin-specific injectant and shale minerals were also noted. The carbonate-rich sample experienced substantial calcite dissolution and a corresponding decrease in its bulk density and microcrack openings. That is, an altered zone was apparent over the roughly first ½ of the length of the core. It is clear that fracture fluid composition should be optimized based on intrinsic shale and resident brine chemistries.

Shale alterations not only result in new mineral scaling, but result in mechanical alterations to fractures and rock matrix, such as porosity/permeability increases, mineral scale precipitation, and fracture closure due to proppant embedment. To monitor these changes to the stimulated rock volume in situ, it is necessary to develop seismic signatures in the laboratory that benefit the interpretation of field measurement changes in velocity. We conducted time- lapse rock physics experiments on clay-rich (carbonate-poor) Marcellus shales to characterize the acoustic velocity and permeability responses to fracture acidizing and propping. We measured acoustic P- and S-wave velocities together with permeability before and after inducing chemical alterations of the sample fractures. We also complemented the physical measurements with microstructural imaging using X-ray μCT and SEM techniques. Experiments show that S-wave velocity is a key geophysical observable, and in particular, the S-wave polarized perpendicular to fractures as it is sensitive to fracture stiffness. Results show that both acidification and propping decrease the fracture elastic stiffness. This effect is stronger when fractures are acidized making the monitoring of proppant efficiency be potentially masked by chemical alteration. Furthermore, fracture permeability is compromised by the softening of fracture surfaces due to acidification. Conversely, the problem is minimized when fracture propping is highly efficient (i.e., majority of the fractures are propped). These contrasting effects on fluid flow in combination with similar seismic attributes indicate the importance of experiments on basin-specific shales so as to improve rock physics models that incorporate the coupled effects of hydraulic fracturing and acidification. This is necessary for a correct interpretation of seismic velocity signatures of flow pathways in unconventional shales.