| dc.description.abstract | Oil and gas well stimulation by hydraulic fracturing necessitates large volumes of water, in turn generating copious amounts of produced water, which is most commonly disposed via deep-well injection. Purifying produced water to a level where it can be reused for additional fracturing addresses the dual challenges of fresh water sourcing and its environmentally conscious disposal. In this manuscript, we systematically investigated fit-for-purpose treatment of produced water for fracking, using the Permian Basin as a test case. Highly saline (~200,000 mg/L total dissolved solids) and turbid (~ 80 NTU) produced water from the Delaware Basin was treated with a combination of FeCl3 as the primary coagulant aided by an anionic polymer to remove suspended solids and iron over a range of environmentally-relevant temperatures via isothermal jar testing. Robust, user-friendly techniques based on mobile phone video capture, optical microscopy, and digital image analysis were developed to accurately characterize floc morphology, size, and settling velocity to interpret experimental data on suspended solids removal. Conformational changes of the polymer over the 4 – 44 °C temperature range were inferred from viscosity and hydrodynamic size measurements providing clues to its performance characteristics. Settling velocities of flocs conformed to a model incorporating their fractal nature, average size, and the viscosity of the suspending medium (i.e. produced water). Juxtaposing the anionic polymer with the hydrolyzing metal-ion coagulant effectively destabilized the suspension through sweep coagulation and adsorption and inter-particle bridging removing ≥ 98% turbidity and ≥ 97% iron allowing reuse of the produced water for hydraulic fracturing. Although bench-scale laboratory experiments established the efficacy of coagulation-flocculation-sedimentation for suspended solids and iron removal over a wide range of temperatures, we recommend additional larger-scale testing to evaluate process performance under actual field conditions before possible widespread implementation. | |
