Geochemical phenomena between Utica‐Point Pleasant shale and hydraulic fracturing fluid

Correspondence Jason P. Trembly, Institute for Sustainable Energy and the Environment, Department of Chemical and Biomolecular Engineering, Ohio University, Athens, OH 45701. Email: trembly@ohio.edu Abstract This study evaluated geochemistry between the Utica-Point Pleasant shale and reservoir/hydraulic fracturing fluid mixtures under simulated reservoir conditions in a batch reactor system. Analytical techniques were utilized to monitor fluid composition with time along with preand post-trial shale microscopy and phase identification analyses. Formation of iron-based precipitate was evident through results from fluid and material analyses. Ferrous iron was the predominant iron form found in the aqueous phase, with oxidation to ferric iron and subsequent precipitate formation. Geochemical modeling further supported ferric iron was the favorable phase for precipitation.


| INTRODUCTION
U.S. tight oil has had a tremendous impact on stabilizing global oil prices and increasing energy security. Such light crude is recovered from low permeability rock formations, primarily shale, using unconventional methods such as horizontal drilling with hydraulic fracturing.
As of March 2019, the Energy Information Association (EIA) reports approximately 60% total U.S. oil production comes from tight oil. 1 While unconventional wells have shown promise in terms of oil production, economic viability is still uncertain due to rapid production decline leading to shorter well life. Peak performance of an unconventional oil well occurs during the first quarter of production, with up to 74% production decline after 1 year. 2 Solutions explored to counteract production decline include refracturing or other enhanced oil recovery methods. Simulation studies have been developed to evaluate the effects of stimulation techniques on production to determine if profitable recovery is feasible. [3][4][5] Although simulation of refracturing and enhanced oil recovery methods have shown promise, these techniques do not fully account for phenomena encountered in unconventional wells.
Various approaches have been utilized to mimic downhole phenomena. [6][7][8] An approach by Luo evaluated confinement effects on hydrocarbon bubble point. This study found octane and decane confined in nanoporous media possess two distinct bubble point temperatures, with lower/higher bubble point temperatures differing ±15 K in comparison to respective bulk properties. 6 By understanding various in-situ well phenomena, proper recovery analysis can be performed and modifications to hydraulic fracturing processes can be developed to increase well productivity and lifetime. 9 Aqueous phase chemical reactions are another key phenomenon occurring in shale reservoirs. Such reactions occur during well completion, when hydraulic fracturing fluid (HFF) mixes with rock and formation water, potentially causing precipitate formation, more commonly referred to as scale. Scales form when dissolved solid concentrations result in supersaturation, resulting in formation of insoluble solids. 10 If scale formation occurs in a critical flow path, that is, fracture network or hydrocarbon bearing pores, hydrocarbon flow and production will be hindered negatively impacting well production. The most probable time frame for reservoir scaling is during the shut-in period of the completion process. 11 During this period, the HFF is in a nearly stagnant flow regime and at peak reactivity. Therefore, prevention of scale formation during initial well completion requires further consideration.
Addition of chelating agent or scale inhibitor to the HFF is a common method to prevent scale formation. Prevalent scaling ions found in shale reservoirs include Ba 2+ , Ca 2+ , Sr 2+ , Fe 2+/3+ , P 3− and S 2−. [11][12][13][14] Chelating agents bind to metal ions forming soluble complexes preventing formation, while scale inhibitors adsorb onto growth sites or incorporate into scale crystal lattice hindering or preventing further growth. 15 found the occurrence of iron precipitation as a potential source for production decline in some wells. 24 These studies provide the basis for current and future research into unconventional reservoir geochemistry with respect to in-situ scale formation during initial completion. However, these studies are tailored to well-developed formations such as the Marcellus, Bakken and Eagle Ford and at this point to the author's knowledge, no studies exist involving the Utica-Point Pleasant (UPP) shale.
The UPP shale formation has been proven productive in eastern Ohio, western Pennsylvania, and northern West Virginia and consists of two separate members deposited during the middle Ordovician period. The upper member, the Utica shale, is composed of black calcareous shale with total organic carbon (TOC) up to 3.5%. This upper member has seen tremendous development in eastern Ohio and is known for producing significant quantities of wet shale gas. The underlying member of the Utica, the Point Pleasant shale, consists of interbedded limestone and shales and is organic-rich with TOC content up to 4-5%.
The Point Pleasant shale is the primary oil target of the UPP due to its higher organic content. Across the play, the Point Pleasant shale While the UPP OW has the potential to be a significant tight oil reservoir, it has yet to be fully developed.

| X-ray diffraction mineralogy
Portions of the UPP shale core samples were pulverized into powder.
The powder was used to determine mineral composition via X-ray diffraction (XRD; Rigaku Ultima IV). XRD patterns were captured at 2θ between 10 and 70 with a scan rate of 1 /min. Approximate weight percentages were calculated using the relative intensity ratio (RIR) method with the XRD pattern.

| Synthetic fracturing fluids
The simulated HFF used in this study contained chemicals and concentrations similar to the formula used in the initial Brookfield 3H well completion, the HFF composition is shown in Table 1. For experimentation, two synthetic HFF compositions were tested. To simplify aqueous system chemistry, scale inhibitor was omitted from HFF solutions to create a favorable precipitation environment. Iron control was included in the Trial 1 HFF solution, with its omission in Trial 2 and 3 HFF solutions. Industrial chemicals used include CC-1, SC-2 and FR-1 obtained from Halliburton and Ecopol-FEAC and Ecopol 2000PTW from Economy Polymers.

| Produced water
Produced water was collected from the Brookfield 3H well in February 2018, approximately 6 years after onset of production.
Water collected at this stage of production was deemed to be similar in composition to the well's formation water. To protect against oxidation, the water was continuously purged with nitrogen during collection and storage. Water composition was analytically quantified using Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES), Ion Chromatography (IC) and pH measurements. Chloride concentrations were calculated using the Geochemist's Workbench Spreadsheet module using dissolved ion results from ICP to determine chloride ion concentration to achieve solution charge neutrality. The pH of the water was found to be 5.64 and the Brookfield 3H produced water ion composition is shown in Table 2.  Table 2) via ICP-OES and Fe 2+ was determined via ultraviolet-visible spectroscopy (UV-Vis; Hach DR6000) using the 1,10-phenanthroline method. Fe 3+ concentration was determined via difference between total iron, determined via ICP-OES, and Fe 2+ concentrations.    Figure 3 with data provided in Table S1.

| Absence of scale inhibitor
To develop an understanding of scaling phenomena between UPP shale and associated HFF, a trial was performed using the fluid composition ( greater than the median value for Marcellus shale flowback water (41 mg/L). 29 This elevated sulfate concentration increases the likelihood calcium and strontium will form sulfate precipitates in UPP wells. Additionally, barium also forms sulfate scales and has the lowest solubility product of the three divalent cation sulfates. In the Brookfield 3H well barium has the lowest concentration of the scaling ions (2.32 mg/L) and orders of magnitude lower than the median flowback concentration for Marcellus shale wells (164 mg/L). 29 Although at lower concentrations, barium still has a high probability to precipitate due to its low solubility product and elevated sulfate content of the reservoir water.
Iron is another component which may precipitate based upon produced water concentrations. In the Brookfield 3H well, the iron concentration (129 mg/L) is greater than the median flowback concentration for Marcellus shale wells (29.7 mg/L). 29 Additionally, when iron is in the ferrous state and oxygen is present, there is a high likelihood for oxidation to the ferric state. In conjunction with ferric hydroxide's low solubility product, iron precipitation is favorable in the Brookfield 3H well.
Data for these scaling ions was normalized to solution chemistry at trial initiation (i.e., Day 0) and used to assess formation of insoluble  Table S2.
In the absence of scale inhibitor, Ba 2+ , Ca 2+ and Sr 2+ appear to be in chemical equilibrium. Dissolution of S 2− and P 3− occurs initially, followed by chemical equilibrium after 6 days. P 3− can be conservatively assumed to be orthophosphate (PO 3− 4 ). However, with low initial P 3− concentration, it is difficult to discern if the source is via dissolution of phosphorus containing minerals or experimental/analytical error. Ba 2+ , Ca 2+ and Sr 2+ prefer to form sulfate-based precipitates as shown in Equations (1) through (3). 35 Solubility of sulfate products from these ions decreases with increasing cation atomic radius, with and Sr 2+ all failing to form sulfates in this trial, lack of sulfate is likely limiting formation of these scales. This hypothesis is further supported by the increase in elemental S 2− throughout the trial. A conservative assumption can be made that elemental S 2− is likely sulfate formed as a byproduct of pyrite oxidation, Equation (4). 36 Normalized Fe 2+ , Fe 3+ and total Fe data are shown in Figure 6.
Fe 2+ concentration decreased during the trial; however, the solution pH range of 3.7-5.5 is lower than the reported pH range for Fe 2+ precipitation, occurring between 6.0 and 7.5. 39,40 However, solution pH range is within precipitation pH range for Fe 3+ , occurring between 2.5 and 3.5. 40,41 These observations indicate the iron control is chelating Fe 3+ , as lower concentrations are expected based upon the ferric hydroxide solubility product. These results indicate Fe 2+ undergoes oxidation to Fe 3+ , which then precipitates.
As the HFF is aerated, this is the likely Fe 2+ oxidation source.
Although oxygen could be removed from HFF via inert purging, such oxygen removal techniques would be considered cost prohibitive in an industrial setting. 40 Once in the Fe 3+ state, hydrolysis occurred forming ferric hydroxide. 41 The reaction for Fe 2+ oxidation and Fe 3+ hydrolysis are shown in Equations (5) and (6)

| Absence of scale inhibitor and iron control
Additional trials were performed without scale inhibitor or iron control (Ecopol-FEAC), to assess the inhibiting effect of the iron control agent. Normalized ion and pH results in the absence of iron control were consistent with the previous trial, with iron as the only precipitating specie. Figure 7 presents the solution pH with time, showing Fe 3+ precipitation favored over Fe 2+ , while normalized iron results are shown in Figure 8 with data provided in concentration was found to be lower in the absence of iron control (Trial 1), as minimal Fe 3+ remained in solution at trial's end.
Pre-and post-trial analyses were performed via SEM/EDS relocation. EDS results were used to qualitatively assess precipitation or dissolution at the shale surface (i.e., signal intensity was used to assess shale surface elemental composition changes). Figure 9 presents SEM/EDS images for a pre-and post-trial sample at the same location.
EDS results indicate iron to be the only element with noticeable change after the trial. Some noticeable differences can be seen in shale's surface structure, appearing smoother after exposure to the HFF. This smoothing appearance is likely caused by dissolution of car-    Figure 11 through Figure 13, respectively with data provided in Tables S4 and S5

| CONCLUSIONS
This research provides a preliminary assessment of UPP well geochemistry during initial completion activities. Experimental trials along with pre-and post-trial material analyses indicate iron scaling is a concern in UPP shale wells. During experimentation, Fe 2+ was found to oxidize to Fe 3+ , which then hydrolyzed forming a precipitated scale.