Hydrogeological Investigations at Diamond Y Springs and Surrounding Area, Pecos County, Texas by Radu Boghici, B.S. Thesis Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of Master of Arts The University of Texas at Austin August, 1997 Hydrogeological Investigations at Diamond Y Springs and Surrounding Area, Pecos County, Texas APPROVED BY SUPERVISING COMMITTEE: John M. Sharp, Jr. Philip C. Bennett Brenda Kirkland George In memoriam: David J. Disney (1955 -1995) ACKNOWLEDGEMENTS This work was supported by a grant from The Nature Conservancy of Texas. The Geology Foundation financed my first trip into the field. I would like to thank Dr. Jack Sharp, my supervisor, for the opportunity to work on the Diamond Y project, and for his guidance during the phases of this investigation. I give many thanks to Dr. Phil Bennett and Dr. Brenda Kirkland George, the others members of my committee, for taking the time to read and review this thesis. Dr. Bennett's suggestions greatly improved my understanding of the springs' chemistry. I am indebted to Dr. Lynton Land for analyzing my water samples for Oxygen-18 free of charge. John Karges is commended for his help during my visits to Fort Stockton. The springs' monitoring program could not have been implemented without his cooperation. This research greatly benefited from in-depth discussions with Norm van Broekhoven, Robert Mace, Ian Jones, Barry Hibbs, and my wife, Erika. I am grateful to John Ashworth for his help during the initial stages of my research. Finally, a big thanks to Erika and son Vlad for their moral support and willingness to deal with my moods during all this time. May 1997 iv ABSTRACT Hydrogeological Investigations at Diamond Y Springs and Surrounding Area, Pecos County, Texas by Radu Boghici, M.A. The University of Texas at Austin, 1996 SUPERVISOR: John M. Sharp, Jr. This study presents the results of local hydrogeologic investigations at Diamond Y Springs and vicinity, a 660 mi2 ( 1710 km2 ) area of north-central Pecos County, Texas. The data confirm the hypothesis of Rustler aquifer waters as the chief source of flow at Diamond Y Springs today. Dissolution of halite and gypsum, base exchange, evaporation, and mixing of the two hydrochemical facies of the Rustler with recent local waters can explain the water chemistry and isotopic composition. The main controls on the regional flow pattern are: (1) the permeability contrast between the Belding-Coyanosa trough fill and the Edwards Formation, (2) the amount of cross-formational flow recharging the Edwards-Trinity (Plateau) aquifer through the Belding fault system, (3) the amount of recharge from the Rustler aquifer, and (4) the amount of groundwater pumped in the Belding-Fort Stockton and Coyanosa irrigation districts. TABLE OF CONTENTS TABLE OF CONTENTS •••.•.•.••••.•••••••.••....................•.....•....••••...•..•..•....... vi LIST OF FIGURES•••.•..•..•..•.•.••••••••......•...••••.••.•••••••••.••...•.•.•.•••...•....•.......ix LIST 0 F T ABLES •••••••••••••••.••••••••••••••••••••••••••••••••••••••••••••••••.••••••••••••••.••.••xi 1. INTRODUCTION.....••••••••••.•••••••••.••.•.•..•.•.••••..•.••••.•••...••..................... 1 A. Statement of the Problem........................................................ 1 B. Objectives of Study.................................................................. 1 C. General Nature of the Study Area......................................... 3 D. Previous Investigations........................................................... 6 E. Climate..................................................................................... 6 2. PHYSIOGRAPHY AND DRAINAGE ........•....•......•••.•...•...•...•.•.....•. 11 A. Physiography ....•.................•.............•.•..•.•.••...•............•.•......•.. 11 B. Drainage .•........•........•.•....•...•.•.•.•..•.•..•.•.•.....•..........•................. 13 3. GEOLOGY.............................................•.••.•..•....•................................ 15 A. Stratigraphy....•.•.•.•..•.•.•.•.•.•............................•.•..•.•.•.•.•.•.•....... 15 Upper Guadalupe Series ....•..............................•.•.•.••.•.... 20 Ochoan Series .............•.•...••...•........•..••.........................•.•. 25 Triassic Rocks .•....•.•.........•...................•.....................•..•.• 30 Cretaceous Rocks .•.•..........•••...............•..............•...•••.•.•.•. 31 Quaternary•••.•.•.•.•.•..........•..................•...............•••........... 37 B. Structural Geology ....•••..•.•.••.•.••••••••••••••••••••••...••••..••.•.............. 35 4. HYDROGEOLOGY ••••••.••••••••.•••.•.••.•.•••••••••••••••••••••••.••••.•.••••••••••••••••.•• 38 A. Capitan Limestone Aquifer •.........•..•....•........••.••••••••••..••...•..... 38 B. Rustler Aquifer •.•.........••••••.•••••••••••••••••••••••..•••••••••••••••••...•..•.•.•. 39 C. Edwards-Trinity (Plateau) Aquifer........................................ 41 D. Diamond Y Springs..••••••••••••••.•••••••.•.•...........•••.••...•.••••.••••••••••. 49 5. GROUNDWATER CHEMISTRY AND ISOTOPY.......................... 61 A. Methods .........................•.•••....••.•.•......................•.............••....... 61 B. Results and Interpretation••••••••••••••••••••••••••••••••••••••••••••••••••••••• 63 Chemical Processes Responsible for the Composition of Diamond Y Springs.••••••.••••.•.•••••.•••...••.• 67 C. Isotopic Analyses ...•.••.•..•.....................•••.••••••.••.••••....••••••••••••••. 71 D. Origin of Solutes.....•••.•••••••••••••••••••••••.•.............•......••.••••••••••.••. 77 6. FLOW SYSTEM MODELING.•.•..••......•...........................•...••.•.•.••.••.. 87 A. Method.s .•.••.•••••.•.•.••••••.•••••••..•••..•••...•....•••..••.••••••.•..•••••.••••••••.•.•.• 87 Boundary and Initial Conditions ...................................... 87 Finite-Difference Methods .•••.••...•.•.•••••••••••••....•••••.• 88 B. Groundwater Flow Modeling in the Diamond Y Springs Region •••.•.••.•.••••••••••.••.•......... 89 Mod.el Setup ••••••••••••••.•••.•.••••••.••.•..•.••••.•.••.••••.•.•.•.•.•••••••••••.• 90 Calibration..•••••••••.••••••.•.•••.••.•.•..•.••.•..•.••••.•...•.....•••••.•.•.•.•... 97 Mod.el Results and Discussion....•••••••.•.••••••••••••••••..•.•.•.••.•. 99 Mod.el Limita.tions ••••••••••••.•••••••••••••••••••••••••..•••••••••.•.•••••••••105 7. CONCLUSIONS ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••.••••••••••••••••••107 8. REFERENCES..•..•.•.•..•..........•...••••••.••••.•..•...........••.•..•...•.•••.•.•••.•...•...109 9. APPENDIX ........................................................................... 114 VITA••.•.•.•.••••.•..•....•....•......•..•.•.•.••...•.•.•••••.••••••••••.•••••.••.••.•.••••.•••.•••••.•.•...•.120 LIST OF FIGURES Figure 1. Location of study area.......................................................... . . 4 Figure 2. Surface drainage features in Pecos County............................. 14 Figure 3. Strike-oriented geologic cross-section through Pecos County................................................................... 16 Figure 4. Dip-oriented geologic cross-section through Pecos County................................................................... 17 Figure 5. Generalized geologic map of Pecos County............................ 18 Figure 6. Regional Permian structure..................................................... 36 Figure 7. Generalized potentiometric map of Rustler aquifer................ 40 Figure 8. Piper plot showing Rustler water chemical composition....... 42 Figure 9. Map showing the areal extent and the basal structure of the Edwards-Trinity aquifer...................................... 43 Figure 10. Conceptual groundwater flow model for the Edwards-Trinity aquifer................................................ 4 7 Figure 11. Piper plot showing Edwards-Trinity aquifer water composition......................................................... 48 Figure 12. Map showing Diamond Y springs and sampling locations........................................................................ 51 Figure 13. Diamond Y Springs system hydrograph, May 1990-July 1991..................................................... 54 Figure 14. Piper plot showing Diamond Y springwater composition.................................................................. 55 Figure 15. Diamond Y Springs system conductivity, May-August 1990............................................... ........ 56 Figure 16. Diamond Y Springs system pH, May-August 1990........... 58 Figure 17. Diamond Y Springs system dissolved oxygen, May-August 1990....................................................... 59 Figure 18. Hydrochemical facies in the Edwards-Trinity aquifer...... 65 Figure 19. Saturation indices for Diamond Y waters......................... 66 Figure 20. Plot of Ca+Mg versus HC03.. ........... ............... ..... ....... .. .. 68 Figure 21. Plot of Ca+Mg-S04 versus HC03.................................... 69 Figure 22. Plot of Na-Cl versus Ca+Mg-S04-0.5 HC03................... 70 Figure 23. Plot of Na versus Cl.......................................................... 71 Figure 24. Plot of o2H versus 0180.................................................... 74 Figure 25. Plot of Cl versus Br.......................................................... 79 Figure 26. Plot of S04 versus Cl........................................................ 80 Figure 27. The model grid, boundaries and active cells.................... 91 Figure 28.Assigned conductivity domains and pumping cells.......... 94 Figure 29. Modeled rates of recharge to the Edwards-Trinity aquifer......................................................................... 96 Figure 30. Map showing calibrated and measured heads over the model area........................................... . . . . . . . . . 98 Figure 31. Simulated and interpreted water table map for the Edwards-Trinity aquifer, year 1987.................................. 102 Figure 32. Simulated water table map for the Edwards-Trinity aquifer, year 2020... ... . ..... .. . ... ... . .. . . . . . . . .. .. . .. . . ........ ... 104 LIST OF TABLES Table 1 Temperature and precipitation at Fort Stockton, 1859-1990................................................................. 7 Table 2. Freeze dates in spring and fall at Fort Stockton.................. 8 Table 3. The Permian-Quaternary hydrostratigraphy........................ 19 Table 4. Results of tritium analyses................................................... 76 Table 5.Water balance over the whole model domain (steady-state calibration runs).......................................... 99 Table 6. Pumping rates used in the transient flow simulations....... 101 1. INTRODUCTION A. Statement of the Problem Groundwater development in the Fort Stockton area, Pecos County led to lowering of potentiometric surface and to discharge cessation at several local springs, including the well-known Comanche Springs. Diamond Y Springs, which are the focus of this project, feed little ponds which are the host of a federally listed endangered species of fish (Cyprinodon Bovinus). Continued groundwater withdrawal in the area could result in flow decline at Diamond Y, thus threatening the protected species. B. Objectives of Study A preliminary search for information regarding the hydrogeology of Fort Stockton area shows an abundance of data (water-level measurements, water quality analyses, pumping records), but a lack of previous interpretation. The objective of the study is to characterize of the hydrogeology of the Fort Stockton and surrounding areas using an integrated hydrochemical and numerical modeling approach. The study employs well-established hydrological, hydrochemical, and isotopic methods to delineate the boundaries of the regional and local groundwater systems, to trace local, intermediate, and regional flowpaths, to describe the distribution of groundwater facies, and to account for groundwater origins. Basic data are used to develop a conceptual hydrologic model over the study region. A numerical groundwater flow model is developed and calibrated to estimate the amount of recharge, and hydraulic conductivity field. The results of the study should ( 1) advance our know ledge of the hydrogeologic controls on the spring-aquifer interactions in a complex carbonate-aquifer system, and (2) provide information needed for managing portions of the aquifer system. As the study area undergoes residential, commercial, and industrial development, it is clear that understanding how aquifer parameters and pumping wells affect spring discharge will be more important in the allocation of water resources and drought contingency plans of the local water authority. The eventual goal is to obtain both a scientific and utilitarian understanding of this complex, dynamic stream-aquifer system. This research is a first step toward that realization. C. General Nature of the Study Area The area encompassed by this study lies in central Pecos County, Texas (figure 1 ). The city of Fort Stockton is in the southern part of the study area, approximately 240 mi (405.6 km) east of El Paso, along Interstate Highway 10. The Diamond Y Springs are located on a 15 mi2 (38.9 km2) nature preserve owned by the Nature Conservancy of Texas (TNCT), about 10 mi (16.1 km) north of Fort Stockton, Pecos County, Texas. The investigation, however, is not confined to the TNCT preserve area, but covers 660 mi2 (1, 710 km2 ) around Fort Stockton (roughly, about 12 USGS 7.5 minute topographic quadrangles). Pecos County was created in 1871 from a part of Presidio County. It was named for the Pecos River, which borders the county to the north. The county is about 88 mi (142 km) from north to south, and about 108 mi (173.8 km) from east to west. It is the second largest county in Texas. Fort Stockton, the county seat, was established in 1845 and 9 years later a U.S. Army fort was established here. It served as refuge for travelers on the Old Spanish Trail. Perhaps the most important justification for a fort at this location was Comanche Springs, which flowed at about 60 million gallons of water per day (226,000 m3/day) in 1869 (Annie Riggs Museum Archives, Fort Stockton). However, largely because of ... CARLSBAD . r---­ ·\ 32'­ \ ' ~-{ \ MEXICO ~o ,,I 30 20 I 40 I ~KILOMETER Figure 1. Location of study area 4 pumping of groundwater for irrigation, Comanche Springs ceased flowing in 1952. In 1990 the county had a population of 14,675, while Fort Stockton was inhabited by 9 ,518 people. Ranching and irrigated farming are the most important economic activities in the county. Based on the Conservation Needs Inventory (1967), 67,674 acres (274 km2) was irrigated cropland, 2,759,563 acres (11,168 km2) was rangeland, 1,920 acres (7.7 km2) was water, and 12,119 acres (49 km2) was urban areas. The irrigated area declined to 25,000 acres (101 km2) (USDA, 1976). The ranching stock in Pecos County consists of beef cattle and sheep, while cotton, grain sorghum, barley, alfalfa, cantaloupe, onions, peppers, pecans, and cabbage are the main cultivated crops. Petroleum production, sulphur mining, and the automotive testing grounds are also significant in the local economy. D. Previous Investigations Adkins ( 1927) reported discharge figures for Comanche Springs and water chemistry data for several wells and for Comanche Springs, but did not comment on Diamond Y springs. Armstrong and McMillion ( 1961) were the first to refer to Diamond Y springs; they provided some discharge and chemistry data. Audsley' s ( 1956) investigation focused on the hydrogeology of the Fort Stockton area. Rees and Buckner ( 1980) investigated the Edwards-Trinity aquifer in the Trans-Pecos of Texas on a large scale. Brune (1975) presented some discharge data on Diamond Y and a history of the springs. Small and Ozuna (1993) published an assessment of the groundwater conditions for Pecos County. Finally, Veni (1991) centered his work exclusively on Diamond Y; he was the first to give a detailed account on the hydrogeology of these springs. E. Climate The climate of the study area is semi-arid and is characterized by hot summers and cool winters. Table 1 gives temperature and precipitation data recorded at Fort Stockton the period 1859 to 1990. Table 2 gives probable dates for the first freeze in fall and the last freeze in spring. TABLE 1. TEMPERATURE AND PRECIPITATION I) (Recorded in the period 1859-1990 at Fort Stockton, Texas) MONTH A VERA GE MONTHLY TEMPERA TURES (°F) AVERAGE MONTHLY PRECIPITATION (in) High Low Mean Rainfall Snowfall January 62.5 31.3 46.4 0.53 0.40 February 67.0 34.6 50.7 0.52 0.70 March 74.2 40.4 57.3 0.49 0.10 April 82.6 48.9 65.5 0.73 0.00 May 89.4 57.4 73.5 1.66 0.00 June 95.6 65 .3 80.3 1.64 0.00 July 96.1 67.4 81.6 1.68 0.00 August 95 .5 66.4 80.6 1.89 0.00 September 89.7 61.0 74.9 2.49 0.00 October 81.4 50.9 65 .9 1.34 0.00 November 70.9 39.4 54.6 0.70 0.40 December 63.5 32.6 47.7 0.68 0.20 .. •11 Data from the Nat10nal Oceamc and Atmosphenc Adnumstral!on TABLE 2 FREEZE DATES IN SPRING AND FALL 1 > (Recorded in the period 1955-1975 at Fort Stockton, Texas) PROBABILITY TEMPERATURE 24 ° F or lower 28° F or lower 32° F or lower Last freezing temperature in spring: 1 year in 10 later than 2 years inlO later than 5 years in 10 later than First freezing temperature in the fall : 1 year in 10 earlier than 2 years in 10 earlier than 5 years in 10 earlier than March 27 March 18 March2 November 10 November 17 November 30 April8 April 13 Aprill April8 March 18 March 29 November 2 October 26 November? November 1 November 17 November 11 I) After Rives (1980). In winter the average temperature is 47 °F (8.3 °C), and the average daily minimum temperature is 32 °F (0 °C). The lowest temperature ever recorded in the study area was 4 °F (-15.5 °C) and occurred on January, 5, 1972. During summer the average temperature is 80 °F (26. 7 °C), and the average daily maximum temperature is 94 °F (34.4 . °C). The highest recorded temperature at Fort Stockton was 110 °F (43.3 °C) on June, 22, 1969. The average annual precipitation in the study area is 12 in (305 mm). Of this, 8 in (203 mm) or 65 percent falls from April through September. The heaviest one-day rainfall during the period ofrecord was 3.38 in (85.9 mm) at Fort Stockton on May, 27, 1957. An average of 40 days per year have thunderstorms. Most ofthem occur during summer. Snowfall in the Fort Stockton area is rare. About half the winters have no measurable snowfall. In 25 percent ofthe winters the snowfall measures more than 3 in (76.2 mm) and is largely a short-duration event. Also, it only takes two to three days for the acummulated snow to dissipate. The heaviest one-day snowfall was more than 4 in (101.6 mm). The average relative humidity in midaftemoon is about 40 percent, and increases later in the day. The average humidity at dawn is about 70 percent. The sun shines 80 percent ofthe time during summer and 75 percent ofthe time during winter. The dominant wind blows from the south-southeast, with a high speed of 13 mi/h (21 km/h) in April. Evaporation has been monitored by the National Weather Service between 1940 and 1961 (Dougherty, 1975). The potential annual evaporation is about 109 in (2769 mm). The monthly evaporation ranges between 4 in (102 mm) in January and 12-14 in (305-356 mm) in summer when the crop demand for water is highest (Armstrong and McMillion, 1961 ). The following chapters ofthe study describe the physiography and the drainage of the study area (chapter 2), and its general geology and structure (chapter 3). Chapter 4 familiarizes the reader with the hydrogeology ofPecos County and the Diamond Y Springs, while the next two chapters discuss the springs' hydrochemistry and isotopy (chapter 5), and the development ofa groundwater flow model for the study area (chapter 6). Chapter 7 presents the conclusions ofthis research. 2. PHYSIOGRAPHY AND DRAINAGE The surface of Pecos County is nearly level to gently undulating in the northern half and hilly to mountainous in the southern half. A network of creeks occasionally transmits surface water to the Pecos River, county's main draining feature A. Physiography The study area is in the Stockton Plateau (figure 1, p. 4), an extension of the Edwards Plateau of Central Texas. The Stockton Plateau is a plain of degradation (Adkins, 1927, p.13) where the formerly continuous limestone tablelands are being eroded. The altitude of the Plateau around Fort Stockton ranges between 2000 ft (600 m) and 3000 ft (900 m) above sea level. Relief of 200 to 300 ft (60 to 90 m) is common in the larger creeks. In the northern part of the area the Stockton Plateau is completely eroded, exposing soft shales, red beds, and evaporites. The dominant features of this region are: a flat land surface, gravel fills, and steppes vegetation. The southern part of the study area is a stripped plain of Cretaceous limestone (Adkins, 1927), presently in a stage of immature dissection. It consists of a large number of flat-top hills or mesas, whose summits lie on the same stratigraphic level so that the slope of the region is that of the rock beds, (i.e., towards E and NE). Surrounding the mesas are graded slopes and pediments. The mesas are the result of the selective erosion of about 600 ft ( 180 m) of soft shales and limestones. Three interspersed, more resistant limestone intervals (called "caprocks" by Adkins, 1927) control the local relief: the lower, middle, and upper Caprock. The lower Caprock (Upper Fredericksburg) is mostly buried under valley fill. The upper Caprock (Upper Washita) is commonly eroded, therefore the middle Caprock (Middle Washita) is dominant, and produces that "one-story" flat top profile, characteristic to the area. Although not located within the study area, the Glass Mountains serve as potential recharge areas for the Edwards-Trinity aquifer in Pecos County. The Glass Mountains are in southern Pecos County and occupy there an area of about 75 mi2 (28.9 km2). Their elevation ranges between 4,000 to 6,000 ft (1220 to 1824 m), and are made of intensely eroded and faulted limestone and dolomite. Deep creeks with thick deposits of boulders and gravel cut through the mountains. B. Drainage The main drainage feature is Pecos River which receives surface water from a number of creeks (see figure 2). All the tributaries in Pecos County are ephemeral, the most important of which are: Coyanosa Draw, Courtney Creek, Leon Creek (which unites with the Diamond Y Draw in the study area), Sixshooter Draw, and Comanche Creek. In rare occasions these creeks flow directly to the Pecos River. Usually, the high evapotranspiration rate and the streambed configuration combine to limit runoff to Pecos River. Figure 2. Surface drainage features over Pecos County 14 3.GEOLOGY This chapter presents the stratigraphy of the Permian and Cretaceous rocks under the study area, and discusses how regional structure controls the groundwater flow. A. Stratigraphy Several formations have an impact on the groundwater flow and chemistry in the study area. The Permian-Quaternary succession is summarized in table 3, and the spatial relationship between the stratigraphic members are shown in figures 3 and 4. The surface geology is shown in figure 5. Geologic units under the area of interest that contain ground water range in age from early Paleozoic to Holocene. Pre-Permian (Ordovician, Silurian, and Devonian) rocks yield water as a by-product of petroleum production activities. Head and chemistry data for these units are very limited. Fort Stockton Figure 3. Strike -oriented geologic cross -section through Pecos County (Modified from WTGS, 1962) B B' Figure 4. Dip -oriented geologic cross -section through Pecos County (modified after WTGS, 1951) 1113:' DO' '· ...... CX> _, . ~~-----,------­ ~T£RRELL I ~TY ,..~ EXPLANATION ~ ~ CENOZOIC ALLUVIUM • TERTIARY VOLCANIC ROCKS I D CRETACEOUS ROCKS Ill TRIASSIC ~~~-~j ~PERMIAN ROCKS ?,_·, ~ • PENNSYLVANIANROCKS ·,·~! --FAULT . . Figure 5. Generalized geologic map, Pecos County (modified from Small and Ozuna, 1993) TABLE 3. STRATIGRAPHIC UNITS AND THEIR WATER-YIELDING PROPERTIESn SYSTEM SERIES OR GROUP UNIT LITHOLOGY WATER­YIELDING PROPERTIES AQUIFER Quaternary and Tertiary Alluvium Unconsolidated silt, sand, gravel, clay, caliche, ~ypsum Small to large yield of fresh to brackish water Cenozoic Alluvium Cretaceous Washita Group Fredericksburg Group Trinity Group Buda Limestone "upper cae.rock" Boracho Formation "middle caprock" University Mesa Marl Finlay Formation "lower caprock" Maxon Sandstone Glen Rose Formation Soft nodular limestone, marl, hard limestone Hard, massive limestone and thin-bedded limestone with some cla:z: Soft, nodular limestone, marl, and hard limestone. Massive. ledge-forming limestone and soft, nodular limestone Fine-to coarse-grained quartz sand with silt, and limestone White, marly limestone, yellow marl Not known to yield water Yields small quantities of water. Small yield of brackish water Yields small to moderate quantities of fresh to brackish water Small to moderate yield of fresh to brackish water Small yield of water Edwards-Trinity Triassic Dockum Santa Rosa Reddish Small to Santa Rosa Group arkosic, moderate yield sandstone brackish water Tecovas Red shale, silt and sandstone Not known to yield water Permian Ochoan Series Dewey Lake Red beds Sand, shale, gypsum, and anhydrite Not known to yield water Rustler Formation Dolomite, anhydrite, halite, shale, Small to large yield of fresh to saline water Rustler sandstone, and limestone Salado Halite, Not known to Formation anhydrite, and yield water dolomite Castile Formation Calcareous anhydrite, halite, and limestone Not known to yield water Guadalupian Upper Dolomite, Small to Series Guadalupe, limestone, shale moderate yield undivided of brackish water Capitan Dolomite, Moderate to Capitan Limestone limestone, large yield of Limestone anhydrite, and brackish water sandstone .. I) Compiled from Armstrong and McM1lhon (1961) and Small and Ozuna (1993). UPPER GUADALUPE SERIES Capitan Limestone The Capitan Limestone is a sequence of carbonate rocks deposited as a reef and a reef talus. It can be as thick as 1800 ft (549 m) and consists of massive dolomitic limestone, alternating with siliceous shales, and thin-bedded limestones. The Capitan is an arcuate belt of rocks paralelling the east side of the Delaware Basin. Adams and Frenzel ( 1950) estimated the width of the reef at no more than 5 mi (8 km). In the northern part of Pecos County the depth to the Capitan is greater than 4000 ft (1219 m), while towards southwest its equivalent, the Gilliam limestone, crops out in the Glass Mountains. Towards the east the Capitan is continued into the Whitehorse Group, and in the Delaware Basin it is interfingered with the upper segment of the Delaware Mountain Group (Armstrong and McMillion, 1961) Whitehorse Group The Whitehorse Group consists of five formations: the Grayburg Formation, Queen Formation, Seven Rivers Formation, Yates Sandstone, and Tansill Formation. These are back-reef time equivalents of the Capitan Limestone. Grayburg Formation The Grayburg Formation is the basal sequence of the Whitehorse Group. Under the study area the Grayburg is about 350 ft (107 m) thick, and consists of dolomite, sandy dolomite, and limestone with sandstone intercalations (Armstrong and McMillion, 1961). The top of the formation is sometimes marked by a brown limestone layer. Anhydrite may be present, and is common near Bakersfield (figure 5). The depth to the Grayburg Formation ranges between 1300 to 2800 ft (396 to 853 m) (Armstrong and McMillion, 1961). Queen Formation The Queen Formation overlies the Grayburg Formation, and both display similar lithologies: dolomite interbedded with red and grey sandstone and anhydrite. The Queen usually contains more anhydrite than its neighbour below and is known to be about 400 ft (122 m) thick (Armstrong and McMillion, 1961). Seven Rivers Formation The Seven Rivers Formation is present throughout the Delaware Basin and adjacent Northwestern Shelf and Central Basin Platform (Page and Adams, 1940; King, 1942; Tait and others, 1962; Hills, 1972). It ranges from approximately 800 ft (244 m) thick in the Midland Basin to 370 ft ( 113 m) or less on the Central Basin Platform. The Seven Rivers Formation exhibits regional lithologic variations. The Seven Rivers is up to two thirds halite within the northeastern Delaware Basin. On the Central Basin Platform, it is mainly anhydrite and gypsum (King, 1942), which were deposited in a restricted back-reef area of the Capitan complex. In the reef area and at the southern edge of the Central Basin Platform, the unit is primarily dolomite (Mear and Yarbrough, 1961; Hills, 1972). The formation is described by McNeal and Hemenway (1972) as tan, brown, and gray fractured, microcrystalline dolomite layered with thin beds of gray, shaly siltstone, dark brown to black shale, and streaks of white to gray, finely crystalline gypsum. Yates Formation Conformably overlying the Seven Rivers Formation, the Yates, occurrs throughout the eastern Delaware Basin and on the Central Basin Platform (Page and Adams, 1940; Tait and others, 1962). The formation was originally described as the Yates Sandstone, an 80-(24 m) to 125-ft (38 m)-thick sandstone bed in the Yates oil field of Pecos County (Gester and Hawley, 1929). However, this bed does not everyehere include all strata between the Seven Rivers and Tansill Formations. Mear and Yarbrough ( 1961) formally proposed that the Yates Formation be expanded to include the entire interval. The Yates is thickest (340 ft [104 m]) near the western edge of of the Central Basin Platform in Ward County, and thinnest (80 ft [24 m]) at the southeastern edge of the platform, in northeastern Pecos County. The Yates Formation displays a marked lithologic heterogeneity. For instance, within the inner shelf, the Yates is primarily fine red sandstone with minor shale, anhydrite, and halite interbeds; however, the southern edge of the platform the formation contains mostly dolomite and subordinate amounts of sandstone and anhydrite/gypsum. Tansill Formation This uppermost formation of the Guadalupian Series in West Texas was deposited throughout the study area. The Tansill rests conformably on the underlying Yates Formation (Page and Adams, 1940). Within the Delaware Basin the Tansill is typically 100 to 150 ft (30.5 to 45.5 m) thick and it grades into the massive reef facies of the Capitan limestone (Deford and Riggs, 1941). Although "Tansill" has been used in subsurface terminology since the 1920' s, Deford and Riggs ( 1941) formally defined the unit and provided a description of the type section located near Carlsbad, New Mexico. In the Delaware Basin, the Tansill Formation contains varying amounts of anhydrite/gypsum, dolomite, halite, and sandstone (Page and Adams, 1940). On the western edge of the Platform, underlying the Diamond Y area of interest, the Tansill is gray to brown microcrystalline dolomite with anhydrite inclusions. Eastward, away from the reef, the formation becomes progressively more anhydritic (Mear and Yarbrough, 1961 ). OCHOAN SERIES Ochoan strata are represented by four formations (in ascending sequence): Castile, Salado, and Rustler Formations and Dewey Lake Redbeds. Castile Formation The Castile was named by Richardson ( 1904) from exposures at Castile Springs in northern Culberson County, Texas. As mapped by Dietrich and others (1983), the Castile Formation in Texas forms a broad, east-dipping, north to south outcrop belt (Gypsum Plain) extending from the Apache Mountains in the south to the Texas-New Mexico border north of Culberson County, Texas. The overlying Salado Formation is exposed in the eastern part of this outcrop belt, but because of the pervasive hydration of anhydrite and dissolution of gypsum and halite in the near-surface zone, the two formations cannot be easily differentiated. The Castile Formation, 1000 to 2100 ft (305 to 640 m) thick, extends throughout the Delaware basin (King, 1942; Bachman, 1984). In the eastern part of the basin, the formation thins to a few feet of basal strata, grading laterally into uppermost Capitan reef carbonates, deposited along the Central Basin Platform (Hills, 1972). The Castile Formation is composed of rhythmically interlaminated salts of different solubilities (mostly calcite­anhydrite, but also anhydrite-halite interlaminations) and beds of anhydrite/gypsum, halite, limestone, and minor amounts of very fine terrigenous elastics. The Castile is easily identifiable by its distinctive alternating laminae of calcite and anhydrite. In core, individual lamina pairs have been correlated as far as 70 mi ( 113 km) across the basin and are interpreted to represent nonglacial varves (Snider, 1966; Anderson and others, 1972). Other rock types are also present in the Castile Formation. Halite beds, the principal large-scale marker beds of the formation are present only in the northern and the eastern parts of the Delaware Basin. Salado Formation The Salado Formation crops out in the eastern part of the Gypsum Plain. King (1942) and Adams (1944) suggest that although some Salado rocks are exposed, most of the formation is truncated by an angular unconformity at the base of the overlying Rustler Formation. They noted that the Salado thickens downdip to the east. Kroenlein ( 1939) suggested that the present westernmost extent of the formation approximates its original depositional limits. Lang (1935) originally defined the formation and named it after Salado Wash in northern Loving County, Texas. Ranging from 380 to 700 ft (91.5 to 213.5 m) thick, the Salado Formation was deposited both in the Delaware Basin and on bordering shelf areas. King (1942) estimated the formation's minimum original thickness to be up to 1000 ft (305 m) near the basin margins, and as thick as 2000 ft (609.5 m) in the east-central part of the Delaware Basin. Snider ( 1966) noted a maximum thickness of 2530 ft (716.5 m) in the eastern part of this basin. Progressive eastward (from outcrop) dissolution of thick Salado halite beds and truncation of the upper Salado Formation in the western and west-central parts of the Delaware Basin probably account for most regional thickness variations (Anderson et al., 1978). The Salado Formation is a more diverse assemblage of rock types than is the Castile. In the study area, the Salado Formation is predominantly halite with lesser amounts of anhydrite, dolomite, sandstone, and siltstone. Locally in the northern part of the Delaware Basin, potash salts, such as poly halite occur (Adams, 1944 ). Terrigenous elastics occur mostly in the lower part of the formation. The greatest differences between the Salado and Castile Formations in west-central Delaware Basin are ( 1) absence in the Salado of thick intervals of interlaminated calcite and anhydrite, (2) occurrence in the Salado of dolomite, and (3) greater abundance of terrigenous elastics in the Salado (Hentz et al., 1989). Rustler Formation The Rustler Hills and the adjacent plains to the east of Culberson and western Reeves Counties compose the outcrop belt of the east-dipping Rustler Formation (Dietrich et al., 1983). The north-to-south-trending outcrop belt ranges from 5 to 12 mi (8 to 19.5 km) wide and extends from the Apache Mountains in the south to southeastern New Mexico in the north. The Rustler Formation (Richardson, 1904) is the youngest unit of the Ochoan Series that contains bedded evaporites. The Rustler ranges from 250 to 670 ft (76 to 204 m) thick, and is distributed throughout the Delaware Basin and adjacent shelf areas (King, 1942). The Rustler thickens to the south (Snider, 1966). Where the lower part of the Rustler Formation crops out in the Rustler Hills, it is composed of dolomite, dolomitic limestone, limestone breccia, gypsum, and mudstone with minor siltstone and sandstone near the base (Hall, 1952; Tunnell, 1952). The complete Rustler section, preserved in the subsurface, can be divided into an upper 150-to 175-ft (45.5 m to 53.5 m) unit of anhydrite/gypsum and a lower dolomite, anhydrite, sandstone, and shale unit (Adams, 1944). In the eastern part of the basin, dolomite decreases in abundance. Halite replaces some anhydrite in the upper and lower divisions at the northeastern edge of the basin. Dissolution of anhydrite/gypsum-bearing members and concurrent collapse of overlying beds produced brecciated strata that disrupted the Rustler succession. An average of up to 30 percent of the original thickness has been lost through dissolution (Hentz et al., 1989). Dewey Lake Redbeds This formation lies conformably atop the Rustler (Page and Adams, 1940; Tait et al., 1962), but is exposed in only a few isolated areas between the Rustler Hills and the western limit of Quaternary surficial deposits that cover most of Reeves County. Eifler (1976) and Dietrich (1983) mapped a few outcrops of the Dewey Lake in the westernmost Reeves County and along the Pecos River to the north. Page and Adams (1940) first described the unit from a subsurface section located near Dewey Lake, an alkali lake in northern Glasscock County, Texas. The Dewey Lake Redbeds are at the top of the Permian section throughout the Delaware Basin. The overlying Upper Triassic Dockum Group rests unconformably on the Dewey Lake in some areas, but elswhere the contact is gradational, suggesting that Lower and Middle Triassic deposits are included in the Dewey Lake (McGowen et al., 1979). The Dewey Lake is thickest in structurally low areas along the eastern and southern edges of the Delaware Basin (Adams, 1944), and thins towards the west. Bachman (1984) noted that the formation attains a maximum thickness of 560 ft ( 170. 7 m) in the nort­central part of the basin. The Dewey Lake Redbeds are a homogenous reddish­brown mudstone and siltstone that commonly display gray reduction spots (Eager, 1983). Minor gypsum is locally present as cement, secondary crystals and vein fills. TRIASSIC ROCKS The Triassic of Delaware Basin was described by Cummins ( 1890) as Dockum Beds from the type locality at Dockum in western Dickens County. Adkins ( 1927) partitions the Dockum strata into two members: a lower one (known as the Tecovas Formation) and an upper one (the Santa Rosa Sandstone). The total thickness of the Triassic in Pecos County ranges between 0 and 1550 ft (0 and 472.4 m) (Armstrong and McMillion, 1961). The Tecovas Formation is a fine-grained elastic sequence composed of red shale, silt, and sandstone. The Tecovas underlains much of the western part of Pecos County. Its eastern boundary is roughly described by a north-south­trending line, passing between Grandfalls and Fort Stockton. The Santa Rosa Sandstone overlies the Tecovas Formation. The Santa Rosa is a reddish-brown, arkosic, micaceous, and conglomeratic sandstone. Armstrong and McMillion ( 1961) add a third, uppermost member to the Dockum succession which they consider as "the exact equivalent of the Chinle Formation of the Colorado Plateau region". It is constituted of fine-grained elastic deposits, and locally overlays the Santa Rosa Sandstone. CRETACEOUS ROCKS The Cretaceous in the study area is represented by elastic and carbonatic rocks of the Trinity, Fredericksburg, Washita, and Terlingua Groups. Trinity Group The Trinity is divided into the Glen Rose Formation and the Maxon Sandstone. Glen Rose Formation The Glen Rose is composed primarily of thin-bedded, marly, white limestone ledges and yellowish-brown, partly sandy soft marl. It overlaps Permian strata, and it is 200 to 400 ft (61 to 122 m) thick. Maxon Sandstone This sequence is also known as the "Trinity Sand", which is a widely used term by well drillers, geologists, and most well owners in Pecos County. Armstrong and McMillion (1961) include the Maxon in the "Trinity Sand" together with the "Basement Sands" (King, 1930) and with the "Basal Cretaceous Sandstone" (Adkins, 1927). The Maxon Sandstone is a fine-grained, brown or yellowish-brown sandstone and sand with limestone fragments. A conglomerate layer about 8 in (20.5 cm) thick crops out at the base of the Maxon 30 mi (48.5 km) south of Fort Stockton (Armstrong and McMillion, 1961). The Maxon sandy member is present throughout Pecos County, and the total thickness of the formation averages 90 ft (27 .5 m). Fredericksburg Group The Fredericksburg Group is represented in the study area by the carbonatic Finlay Formation. Finlay Formation The Finlay Formation consists of massive, coarse-grained limestone and sandstone about 40 ft ( 13 m) thick. In the vicinity of Glass Mountains, the Finlay is truncated, and that area the overlying Washita Group rests directly atop of Maxon. Washita Group In Pecos County the Washita Group is divided into three members: the University Mesa Marl, Boracho Formation, and Buda Limestone. Rocks of the Washita Group crop out extensively throughout northern and eastern Pecos County, forming the typical mesa relief in the area. University Mesa Marl The University Mesa Marl is about 200 ft (60 m) thick and consists of soft, nodular limestone, marl, and hard massive-ledge forming limestone (Small and Ozuna, 1993). Boracho Formation The Boracho Formation, about 150 to 200 ft (45 to 50 m) thick, conformably overlies the University Mesa Marl in Pecos County. It consists of hard massive limestone, thin-bedded limestone and soft, nodular limestone with some clay. Buda Formation Buda Formation totals about 140 ft (40 m), and is represented by very hard, thin to thick bedded limestone in its upper third, a middle argillaceous thin to thick bedded sequence, and a bioclastic, coquinoid interval. Terlingua Group Boquillas Limestone and Gulflan Rocks, Undivided The lowest unit of the Upper Cretaceous is the Boquillas Limestone which consists of limestone, marl and shale, flaggy, light gray, grayish-orange in colour. Armstrong and McMillion (1961) report a 250-ft (76 m) thickness for this interval. The uppermost Cretaceous rocks in the study area are approximately 300 ft (91 .5 m) of marl, shale, and argillaceous limestone. Barnes ( 1983) suggests a possible correlation with units of Central Texas: Taylor, Austin, and upper part of Eagle Ford. QUATERNARY Cenozoic Alluvium The Cenozoic alluvium consists of unconsolidated gravel, sand, silt, clay, and caliche. Its thickness varies from 10 ft (3.5 m) in the creeks near Glass Mountains, to more than 1000 ft (350 m) in the Coyanosa area, where the alluvium fills a deep trough (Armstrong and McMillion, 1961 ). B. Structural Geology The most prominent structural· features under the Pecos region are the Delaware Basin and the Central Basin Platform (figure 6). The Diamond Y area lies on the eastern edge of the Delaware Basin, at the hinge with the Central Basin Platform. Armstrong and Mc Million ( 1961) noted that the structure of the Cretaceous rocks in Pecos County closely reflects the structure of the older beds. Generally, Cretaceous rocks dip away from the center of the Marathon thrust belt at a rate of 5 to 10 ft I mile ( 1 to 2 m I km). North and northeast of the Glass Mountains rocks dip towards the northeast, whereas east of the Glass Mountains they dip eastward. In the Diamond Y area the Cretaceous beds dip to the north at a rate of about 30 ft I mile (18.5 m I km), and strike roughly east-west (Veni, 1991). Dissolution of Ochoan evaporites by Cretaceous seawater caused the Permian beds to collapse and form a deep north-south trough located 2-3 mi (3.5-5 km) west of Fort Stockton (figure 5). Cretaceous, Tertiary, and Quaternary rock and sediments gradually filled the trough, and underwent subsidence, faulting and folding. The troughward tilt of the basal Cretaceous beds indicates that the trough was deepened by post-Cretaceous movements. Figure 6. Regional Permian structure (after Small and Ozuna, 1993) 36 Armstrong and McMillion, (1961). state the relatively undisturbed conditions suggest that post-Cretaceous folding is not related to earlier tectonic events. They report post-Cretaceous faulting and describe the Belding fault system (figure 5) located 14 miles (23.5 km) southwest of Fort Stockton. No studies have discussed the region's fracture patterns but, according to Veni (1991), joints and other faults in the study area have a primary N20-30E orientation and a secondary east-west trend. In addition, the interpretation of two seismic sections shot in the Diamond Y area suggests at least two other fault systems. They are located right under the Diamond Y Springs Preserve (figure 5) and affect both Permian and Cretaceous strata. 4. HYDROGEOLOGY The main aquifers in the study area are: Capitan Limestone, Rustler, and Edwards-Trinity (Plateau). Several pre-Permian formations yield water as a by­product of oil and gas exploitation activities. They are not discussed here. A. Capitan Limestone Aquifer The confined Capitan Limestone aquifer is hosted by the vuggy carbonates of the Capitan Limestone. The aquifer parallels the edge of the Delaware Basin in an arcuate strip along its eastern and northern margins, from the Guadalupe Mountains (southeastern New Mexico) to the Glass Mountains (southwest of Fort Stockton). The depth to the aquifer in Pecos County is about 4000 ft (1200 m), which explains why it remains largely undeveloped in the area. In Pecos County only two wells penetrate the aquifer. One of them flows about 1000 gallons per minute (3890 I/min) from a producing interval of about 3200 to 3600 ft (975.5 to 1097 m) below land surface. The owner of the well states that the casing ruptured soon after completion, thus making possible contamination with water from the Rustler aquifer. The other well, located in southern Pecos County, is plugged. B. Rustler Aquifer The Rustler aquifer, both confined and unconfined in the study area is in the carbonates and evaporites of the Rustler Formation. It yields brackish to saline water to stock, irrigation, and oil recovery wells. Groundwater in the Rustler aquifer occurs under artesian conditions in northern, western, and southern Pecos County. In all these areas, the Triassic Tecovas Formation is the overlying confining layer for the Rustler aquifer. In northeast Pecos County, the Tecovas is truncated by Cretaceous rocks. Here the Rustler is unconfined. Most production in the county's 31 Rustler wells comes from solution openings or fractures in the Rustler dolomite (Armstrong and McMillion, 1961). Wells that do not tap into the solution cavities are acidized to increase yield. Recharge to the Rustler aquifer is by precipitation on its outcrop in Culberson County, and on the Rustler-equivalent formations that crop out in the Glass Mountains (Armstrong and McMillion, 1961 ). Figure 7 depicts heads in the Rustler in Pecos County and neighboring areas. The arrows represent flowlines inferred from head distribution, and show a centripetal pattern: flowlines originating in the recharge areas converge to Diamond Y Springs. Discharge from the Rustler takes place mainly through wells (some of which Modified from Richey, 1985 Approximate extent of Rustler aquifer 0 20 MILES EXPLANATION • 3587 Well location and water level elevation (feet above sea-level) General direction of groundwater flow in the Rustler aquifer Figure 7. Selected wells, water levels and inferred flow directions in the Rustler aquifer, Pecos, Reeves, Ward, and Crane counties have been flowing for years) and by upward leakage into the Edwards-Trinity (Plateau) aquifer. Rustler water quality is variable in the study area. In the Leon-Belding irrigation area wells yield a brackish (TDS=1500 mg/I), water, as opposed to the saline waters (TDS as high as 80,000 mg/I) issuing in wells in northeastern Pecos County. The few uniform characteristics of this water include a high calcium concentration (>500 mg/I), and low bicarbonate ( <200 mg/I). Hydrogen sulfide is also present. Piper plots showing the Rustler water chemical composition are shown in figure 8. A detailed discussion of the Rustler hydrochemistry is given in Chapter 5. C. Edwards-Trinity (Plateau) Aquifer The Edwards-Trinity is the most important aquifer in the study area. It underlies most of Pecos County, as well as a portion of Reeves County and small parts of Culberson and Jeff Davis Counties (figure 9). The more permeable units in the Edwards-Trinity are the lower Cretaceous sands and limestones which are hydraulically connected with the overlying Cenozoic alluvium. Therefore, the Edwards-Trinity and the Cenozoic alluvium are commonly treated as a single, unconfined aquifer. 0 0 0 .. 0 " 0 Mg $04 40 0 40 80 .... . 0 o' . .'~ ,. . . . 80 Ca 20 40 60 80 Na 20 HC03 40 60 80 Cl Figure 8. Piper plot showing Rustler water chemical composition EXPLANATION -IOOO--Base of the Edwards-Trinity aquifer, in feet above sea level Figure 9 . Areal extent and basal structure of the Edwards-Trinity aquifer, Trans-Pecos, Texas. The spatial relationships of the Cretaceous with the older Permian strata are shown in figure 3 (p. 16) and figure 4 (p. 17). In much of western and southern Pecos County the Cretaceous rests upon Permian and Triassic red beds. The shaly Tecovas Formation (Triassic) is a semi-pervious layer which limits flow between the Edwards-Trinity aquifer and the deeper, artesian, brackish-to­saline Rustler aquifer. East of the meridian passing through Fort Stockton and Grandfalls, the Tecovas is truncated, and the Edwards-Trinity directly overlies the Rustler. As stated above, dissolution of Ochoan evaporites during Gulfian time produced a collapse feature (trough) west of Fort Stockton which runs northward. It is filled with permeable alluvium, limestone, dolomite, and evaporite rubble. This produces a permeability contrast between the trough fill and the adjoining rocks to the east and west, and has an important impact on the flow system configuration. Direct recharge to the Edwards-Trinity aquifer by precipitation and return flow is very small because of the high potential evapotranspiration (Armstrong and McMillion, 1961 ). Recharge by precipitation only occurs after periods of steady rainfall and during winter when evapotranspiration is low. This inference is also supported by water chemistry data from the Texas Water Development Board files and by data collected for this study. Water samples from wells located in fertilized agricultural areas show no traces of nitrate. Much of the precipitation falling over the mountainous areas becomes runoff. In creeks and ravines it reaches faulted Cretaceous rocks or karstic features (such as the sinkholes on the M.R. Gonzales property east of Fort Stockton) and infiltrates. Armstrong and McMillion (1961, their plate 14) show a spectacular photograph of the now dry Comanche Creek Reservoir running into a sinkhole. Cross-formational flow is perhaps the largest contribution to the aquifer's recharge budget. Computer simulations of flow suggest at least 75,000 3 acre-ft (9.25·107 m) water recharge the Edwards-Trinity (Plateau) aquifer from deeper sources every year. This recharge occurs principally near or through the Belding and Diamond Y fault systems. Intensive development of the Edwards-Trinity aquifer for irrigation started in 1940. Previously, the groundwater flow system was essentially at steady-state (discharge approximately equaling recharge). Prior to irrigation, discharge occured through springs, by evapotranspiration in areas of shallow water table, and by baseflow to the Pecos River. The average flow at Comanche Springs was 31,000 acre-ft (3.82·107 m3) a year between March 1941 and February 1948 (U.S. Bureau of Reclamation, 1956). The combined flow of other springs in Pecos County was estimated to have been 17,500 acre-ft (2.15·107 m ) a year (U.S. Bureau of Reclamation, 1956). Based on the same source of information, baseflow to the Pecos River was in the order of 30,000 acre-ft 3 (3.7"107 m) a year as the flow of the Pecos River gained about or 36,000 acre-ft 3 ( 4.44· 107 m) per year while passing Pecos County. The water issuing out of the springs usually evaporated, or evapotranspired, or seeped into the ground, so that runoff of springflow to the Pecos was (and is) negligible (Armstrong and McMillion, 1961). After 1940, discharge through wells became important in the aquifer budget. In 1958, the total discharge from the aquifer was estimated to be 3 to 4 3 times the average annual recharge with about 120,000 acre-ft (l.48·108 m) being yielded to wells (Armstrong and McMillion, 1961 ). The quantity of water pumped for irrigation, stock, an municipal use went down to about 77 ,000 acre­ 3 ft (95107 m) in 1974 (Rees and Buckner, 1980). The conceptual model for groundwater flow within the Edwards-Trinity aquifer is illustrated in figure 10. Groundwater in the western part of Pecos County flows towards the north-northeast, in the central part it flows toward northeast, and in the eastern part of the county it flows towards the east. Much of the recharge is "funneled" through the trough towards the Pecos River. ETF 504 20 20 40 .' ·. ~· 40 . . . s . so~· ·~. oo :0 . 60 ...· a".'. ·. · G . . . .. . .' 80 80 ".' . ($:) o" <9. . . ·0· . o. . . . . . '? . ·..· ·..·o · . . · ~ 40 60 ~ ~ 40 60 ~ Ca Na HC03 Cl ETS 0. 504 20 40 60 80 20 40 60 80 Ca Na HC03 Cl Figure 11. Piper plot showing Edwards-Trinity aquifer water composition 48 The trough also acts as a barrier for flow from west to east. This reduces the fresh water input east of Fort Stockton meridian. The water chemistry in the Edwards-Trinity aquifer reflects the local geology, structural controls, and the flow regime. Sulfate and chloride are the dominant ions, with high concentrations of sodium, calcium, and magnesium. Where the aquifer is underlain by the semi-permeable Tecovas Formation, the water is fresh to brackish (TDS