Eric Smistad
Technology Manager, Unconventional Resources,
National Energy Technology Laboratory,
US Department of Energy
The contribution to the US energy supply from unconventional natural gas sources, such as shale, is increasing dramatically due in large part to advanced hydraulic fracturing. To release natural gas from a shale deposit, 3 to 5 million gal of water plus hydraulic fracturing additives and proppants (sand) are pumped under high pressure down a shale gas well.
Often, this water is trucked in from remote locations. About 20-80% of this water returns to the surface as frac flowback water. Much of the flowback and produced water, or frac water, has a high total dissolved solids (TDS) level (>50,000 ppm). Although frac water is increasingly being reused in subsequent frac jobs, much of the frac water is still disposed of by deep well injection.
A significant problem in the Marcellus shale play is that the number and capacity of nearby injection sites is severely limited. For example, since Pennsylvania has only seven class II injection wells, frac water is trucked into Ohio for disposal. In other plays, such as the Barnett, water availability for frac operations is limited. In addition, recent studies have shown that within a typical county-sized shale gas development area, the supply of frac water eventually exceeds the demand for source water for subsequent hydrofracturing operations.
Cleaning options
To avoid both supply and disposal limitations, and thus to enable further shale gas development, an economical process to recover frac water as clean water and salt is required. A research project being undertaken by GE Global Research (GEGR), with funding from NETL, is examining ways to pretreat frac water for thermal recovery of clean water and a salable salt product.
Frac water recovery by thermal evaporation is commercially practiced in a growing number of shale gas applications. To avoid scaling on heat transfer surfaces and to enable reliable evaporator operation, incoming frac water must be pretreated to remove scale-causing ions such as iron and manganese as well as suspended solids and dissolved organics.
Current pretreatment methods typically use flocculation, mechanical separation (e.g., inclined plate clarifier), and filtration (e.g., filter press). Current evaporation processes yield a distilled water product and a brine concentrate, which is typically disposed of by deep well injection.
This project is focused on identifying economical pretreatment techniques to recover additional water and a salable salt (NaCl) product from high-TDS frac waters. The objective is to pretreat frac water to remove hardness, including barium and radium. Barium removal will enable the salt product to meet toxicity characteristic leaching procedure (TCLP) specifications. Radium is a key naturally occurring radioactive material (NORM). Radium removal will enable the salt product to be safely used, for example, as a road deicer, and will minimize worker radiation exposure.
Three pretreatment approaches for frac water softening were considered: chemical treatment, adsorption/ion exchange, and nanofiltration. Chemical treatment in the form of lime softening and optional sulfate precipitation for barium removal is practiced for softening chlor-alkali brine. Calcium hydroxide (lime) and/or caustic soda are added to precipitate scale-forming species. Lime softening may be conducted in a clarifier and generates a precipitate (lime sludge), which is filter pressed and may be either land filled or calcined to recover the lime.
Evaluation of the relative merits of the proposed pretreatment technologies for water and salt recovery first required a clear definition of the range of frac water feed composition to be treated. This project focused on the Marcellus shale, but included frac water sample characterization from the Marcellus, Barnett, and Woodford shales.
In addition, target purity specifications were established for pretreated frac water to ensure economic operation of the thermal equipment. Process material and energy balances were conducted for each process option to generate preliminary cost estimates. Each process option was evaluated with respect to performance and cost, and consideration was given to the adaptability of each process to mobile operation.
Research results
Work on this project began in August 2009. To date, GEGR has characterized frac water samples from seven Marcellus shale wells (Pennsylvania), two Woodford shale wells (Oklahoma), and three frac water disposal facilities in the Barnett. The samples were measured for chemical composition, including NORMS.
The research has identified frac water composition and flow rate design ranges for fixed and mobile treatment systems.
Ion exchange has been ruled out as a softening technique to remove high hardness levels (e.g. >5,000 ppm) based on cost and calculated volumes of regeneration chemicals and rinse water. Nanofiltration also has been ruled out as a softening technique to remove high hardness levels based on the high volume of waste concentrate produced.
GEGR has developed an Aspen/OLI model for chemical treatment of frac water for hardness removal, calculated material balances and costs for lime and sulfate precipitation frac water pretreatment (softening) processes, and summarized NORM composition data from 13 New York Marcellus gas wells (from NYSDEC SDGEIS report).
Studies have demonstrated the importance of removing NORM from frac water for applications involving recovery of a salt product. Processes such as calculating material balance for sulfate precipitation pretreatment of frac water were undertaken. Research has clearly defined levels of frac water 226Ra concentration for a given frac water barium concentration that would result in BaSO4-RaSO4 precipitate that is acceptable for RCRA-D landfill disposal. A paper was presented on this topic at 2010 International Water Conference.
Part of the process also served to develop scintillation counting methods and a gamma spectrometry method to measure NORM in both raw frac water and frac water that had been pretreated for NORM removal. Lab experiments were conducted that demonstrated adsorption/ion exchange and chemical treatment processes for NORM removal from frac water. Technologically enhanced NORM (TENORM) disposal options and costs have been identified.
In the lab, researchers assembled an apparatus to simulate thermal brine concentrator and crystallizer (for thermal water and NaCl recovery). As a result, preliminary tests were conducted to characterize scaling during thermal frac water concentration. This project initially targeted frac water softening as a pretreatment technique to enable higher water recovery and recovery of a salable salt product by thermal evaporation and crystallization. Preliminary investigation showed that NORM is rather prevalent in frac water, particularly in the Marcellus, and that NORM removal from frac water is an important step in recovering a salable salt product from NORM-containing frac water. It was demonstrated on the lab scale that both adsorption/ion exchange and chemical treatment are effective methods for NORM removal from frac water. GEGR conducted experimental and computational studies of the effectiveness and economics of each technique for a range of frac water composition.
Potential impacts
While direct frac water reuse (as source water in subsequent hydrofracturing operations after minimal pretreatment) is an effective method for frac water disposal, frac water recovery to generate clean water and a salable salt product is essential to long-term shale gas development.
Effective recovery and reuse of frac water may benefit the environment by greatly reducing the net consumption of fresh water. In addition, truck traffic, noise, and dust pollution could be significantly reduced in some areas. Finally, decreasing the amount of frac water currently being re-injected into disposal wells could alleviate the possible risks of long term contamination of the water supply.
