Shale-gas production using hydraulic fracturing of mostly horizontal wells has led to considerable controversy over water-resource and environmental impacts. The study objective was to quantify net water use for shale-gas production using data from Texas, which is the dominant producer of shale gas in the U.S. with a focus on three major plays: the Barnett Shale (~15,000 wells, mid-2011), Texas-Haynesville Shale (390 wells), and Eagle Ford Shale (1040 wells). Past water use was estimated from well-completion data, and future water use was extrapolated from past water use constrained by shale-gas resources. Cumulative water use in the Barnett totaled 145 Mm(3) (2000-mid-2011). Annual water use represents ~9% of water use in Dallas (population 1.3 million). Water use in younger (2008-mid-2011) plays, although less (6.5 Mm(3) Texas-Haynesville, 18 Mm(3) Eagle Ford), is increasing rapidly. Water use for shale gas is <1% of statewide water withdrawals; however, local impacts vary with water availability and competing demands. Projections of cumulative net water use during the next 50 years in all shale plays total ~4350 Mm(3), peaking at 145 Mm(3) in the mid-2020s and decreasing to 23 Mm(3) in 2060. Current freshwater use may shift to brackish water to reduce competition with other users.

We compared water use for hydraulic fracturing (HF) for oil versus gas production within the Eagle Ford shale. We then compared HF water use for Eagle Ford oil with Bakken oil, both plays accounting for two-thirds of U.S. unconventional oil production in 2013. In the Eagle Ford, we found similar average water use in oil and gas zones per well (4.7-4.9 × 10(6) gallons [gal]/well). However, about twice as much water is used per unit of energy (water-to-oil ratio, WOR, vol water/vol oil) in the oil zone (WOR: 1.4) as in the gas zone (water-to-oil-equivalent-ratio, WOER: 0.6). We also found large differences in water use for oil between the two plays, with mean Bakken water use/well (2.0 × 10(6) gal/well) about half that in the Eagle Ford, and a third per energy unit. We attribute these variations mostly to geological differences. Water-to-oil ratios for these plays (0.6-1.4) will further decrease (0.2-0.4) based on estimated ultimate oil recovery of wells. These unconventional water-to-oil ratios (0.2-1.4) are within the lower range of those for U.S. conventional oil production (WOR: 0.1-5). Therefore, the U.S. is using more water because HF has expanded oil production, not because HF is using more water per unit of oil production.

No abstract