Comparative geochemistry of flowback chemistry from the Utica/Point Pleasant and Marcellus formations
Comparative geochemistry of flowback chemistry from the Utica/Point Pleasant and Marcellus formations
复制标题
DOI：
10.1016/j.chemgeo.2020.120041
复制
发表时间：
2021-03
期刊：
影响因子：
3.9
通讯作者：
S. Welch;J. Sheets;R. Daly;Andrea J. Hanson;Shikha Sharma;T. Darrah;J. Olesik;A. Lutton;P. Mouser;K. Wrighton;M. Wilkins;T. Carr;D. Cole
中科院分区：
文献类型：
--
作者：
S. Welch;J. Sheets;R. Daly;Andrea J. Hanson;Shikha Sharma;T. Darrah;J. Olesik;A. Lutton;P. Mouser;K. Wrighton;M. Wilkins;T. Carr;D. Cole
在大约两年的时间内，从俄亥俄州的两个尤蒂卡/波因特普莱森特（UPP）现场（UPPW和UPPS）和西弗吉尼亚州的一个马塞勒斯（马塞勒斯页岩能源和环境实验室（MSEEL））现场的几口威尔斯井中收集返排/采出液样品。虽然这些地层具有不同的年龄、沉积环境、成岩历史以及地球化学和矿物学组成（即UPP比Marcellus更富含碳酸盐，Marcellus更富含硅质），但对流体种类随时间变化趋势的分析表明，总体而言，UPP和Marcellus卤水中的TDS和主要溶解元素（Na、Ca、Cl）非常相似。这些盐水中的总溶解溶质（TDS）范围约为40至250 g/L盐，一般而言，浓度随着天然气井完井和增产后时间的推移而增加。Na、Br和Cl的行为表明，这些地层的产出水特征主要来自天然地层卤水，这些天然地层卤水显示出源自蒸发海水的证据。Cl和Br之间有很强的相关性，表明这两个物种的行为保守，和这些盐水之间的相似性表明没有明显的贡献盐从岩盐溶解，因为Br被排除在岩盐结构。盐水中的Cl/Br比范围为~80至120（mg/L/mg/L）。其他元素（如K）易于在流体和粘土上的离子交换位点之间发生反应，通常在单个位点表现出保守行为，但在每个不同井场之间表现出显著差异。Sr和Ba的浓度在井场之间变化很大，并且随着时间的推移相对于Cl−而增加，表明增溶作用增加，推测来自粘土矿物的解吸或来自源地层的碳酸盐或硫酸盐的溶解。UPPW井场由于高硫酸盐输入流体而具有非常低的Ba，这导致卤水中的重晶石/天青石沉淀。相比之下，UPPS井场的Sr含量升高（~ 3500 mg/L），可能是由于在水力压裂中使用了富含Sr的循环盐水。由于Marcellus目标中的钡浓度较高（约1000 ppm，而UPP中约200 ppm），Marcellus现场的Ba浓度最高（高达10 g/L），流体中的Ba/Sr比值最高。这些观察结果表明，FP流体中的溶质来自天然盐水，在地质时间尺度上发生的水-岩相互作用，以及地下同期反应的一些贡献。结果还表明，注入流体的组成可以影响回流流体化学和可能的生产效率。
Flowback/Produced fluid samples were collected from several wells from two Utica/Point Pleasant (UPP) sites (UPPW and UPPS) in Ohio, and one Marcellus (Marcellus Shale Energy and Environment Laboratory (MSEEL)) site in West Virginia over a period of approximately two years. Although these formations have different ages, depositional environments, diagenetic histories, and geochemical and mineralogical compositions (i.e.the UPP is significantly more carbonate rich than the Marcellus which is more siliceous), analysis of trends in fluid species over time shows that, overall, the TDS and major solubilized elements (Na, Ca, Cl) in the UPP and Marcellus brines are remarkably similar. Total dissolved solutes (TDS) in these brines ranged from approximately 40 to 250 g/L salt, and in general, concentrations increased with time elapsed since natural gas well completion and stimulation. The behavior of Na, Br, and Cl suggests that the produced water signatures from these formations are largely derived from the native formational brines which display evidence of originating from evaporated seawater. There is a strong correlation between Cl and Br, indicating that both species behave conservatively, and the similarity among each of these brines suggests no appreciable contribution of salt from halite dissolution because Br is excluded from the halite structure. Cl/Br ratios in the brines range from ~80 to 120 (mg/L/mg/L). Other elements, such as K, which readily reacts between fluids and ion exchange sites on clays, generally exhibit conservative behavior for an individual site, but show significant variations among each of the different well pads.The concentrations of Sr and Ba vary dramatically among well sites, and increase with respect to Cl−over time, suggesting increasing solubilization, presumably from desorption from clay minerals or dissolution of carbonates or sulfates from the source formation(s). The UPPW well site has very low Ba due to high-sulfate input fluid, which resulted in precipitation of barite/celestite in the brines. In contrast the UPPS well site had elevated Sr (~ 3500 mg/L), presumably due to the use of Sr-rich recycled brine used in hydraulic fracturing. The Marcellus site had the highest Ba concentrations (up to 10 g/L) and highest Ba/Sr ratios in the fluids, due to the high concentration of barium in the Marcellus target (~ 1000 ppm, as compared to ~200 ppm in the UPP). These observations suggest that solutes in the FP fluids are derived from native brines, water-rock interactions that have occurred over geologic time scales, as well as some contribution from contemporaneous reactions in the subsurface. The results also show that the composition of the injected fluid can influence flowback fluid chemistry and possibly production efficiency.