Characteristics of streams and aquifers and processes affecting the salinity of water in the upper Colorado River basin, Texas

The upper Colorado River and some of its tributaries between Lake J.B. Thomas and O.H. Ivie Reservoir contain saline water (defined as water having dissolved-solids concentrations greater than 1,000 milligrams per liter). Dissolved-solids loads at nine streamflow water-quality stations increased from 1986 to 1988. The largest increases were in Beals Creek and in the Colorado River downstream from Beals Creek as a result of outflow of saline water from Natural Dam Salt Lake. The outflow contained 654,000 tons of dissolved solids and had a mean dissolved-solids concentration of 7,900 milligrams per liter. This amount represents about 51 percent of the dissolved-solids load to E.V. Spence Reservoir during 1986-88.

The concentration of dissolved solids in streamflow in the upper Colorado River, however, generally decreased in the downstream direction as flows increased in all reaches of the river and its tributaries. Diversion of low flows at a site on the Colorado River upstream from the E.V. Spence and O.H. Ivie Reservoirs decreased the dissolved-solids concentrations of inflow to the reservoirs by about 7 percent while removing only about 3 percent of the total flow. The estimated dissolved-solids concentration exceeded 1,000 milligrams per liter for water from about 233 of 386 wells and springs for which specific conductance was measured in 1986. The mean value of the estimated dissolved-solids concentration for water from the 386 wells and springs was about 1,750 milligrams per liter.

Some of the dissolved-solids content of streamflow and shallow-aquifer water was attributed to evapotranspiration in areas where water was close to land surface, in irrigated areas, and along streams. The evapotranspiration was largely from phreatophytes and was comparable to the mean streamflow from the study area during 1969- 86. However, evapotranspiration was not considered to have an effect on the salinity of water in shallow aquifers on a regional scale. Dissolution of sulfur-bearing minerals such as gypsum and pyrite in shallow aquifers contributed to salinity because of their presence in soil and shallowaquifer formations.

Mixing with brine associated with oil and gas production is a major process affecting the salinity of streamflow and shallow-aquifer water. Brines can move upward through the more than 20,000 abandoned oil- and gas-related boreholes, most of which were abandoned prior to current requirements for casing, cementing, and plugging. Leakage from pits, disposal wells, and secondary-recovery wells, which have been used to dispose of brine, a by-product of oil and gas production, also contributed to dissolved solids in streamflow and shallow-aquifer water. The ratio of brine production to oil production in the seven major oil-producing counties increased from 0.60 to 6.5 between 1957 and 1983. About 359 million barrels of brine were injected into oil-producing formations through secondary-recovery wells in 1983, compared to 11 million barrels through disposal wells.

The chemical characteristics of the saline water in streams and shallow aquifers in the study area were compared to characteristics of water that would result from the probable processes affecting the salinity of water, such as evapotranspiration, mineral dissolution, and mixing of water from streams and shallow-aquifer water with brines from deep aquifers. Dissolution of halite or mixing with deep-aquifer water was the most common cause of increased salinity in 48.0 percent of 77 water samples from shallow aquifers, as classified using salt-norm analysis; the second most common cause was the weathering and dissolution of sulfur-bearing minerals. Mixing with water from soil-mineral dissolution was classified as the principal source of chloride in 28.4 percent of 67 water samples from shallow aquifers with nitrate determinations. Trace-species/chloride ratios indicated that mixing with water from deep aquifers in rocks of the Pennsylvanian System was the principal source of chloride in 24.4 percent of 45 shallow-aquifer samples lacking nitrate determinations.