The Perrier Group of America (PGA) has for several years produced excellent quality spring water from springs in East Texas.  Springs in the sandy hills of East Texas have been developed to supply pure, excellent quality, sanitary spring water with low mineral and nutrient content (i.e. sodium, chloride, sulfate, nitrate, etc.) to PGA’s bottling plants in Dallas and Houston.  As the demand for bottled spring water continues to grow, PGA must locate and obtain additional spring sites favorable for development.  Characteristics of favorable spring sites generally include:

 ·        Sufficient undeveloped property in natural condition to allow for protected source areas of springs and to ensure a long-term sanitary source of water.

 ·        Locations suitably close to plant facilities to allow for economic transport of the spring water to bottling facilities.

 ·        Settings such that development of the spring water sources will not cause adverse impacts to the springs, to the source aquifers, to streams and lakes, or to nearby water wells.

 ·        The spring sources must comply with PGA and regulatory standards and requirements, including: 

o       The spring source must be capable of producing sufficient quantity of excellent quality water, including providing reliable supplies resistant to periods of drought.
o       The spring water source must comply with TNRCC requirements for public water systems, and cannot be under the direct influence of surface water.
o       The spring source must comply with the standards of identity for “spring water” as required by the United States Food & Drug Administration (FDA) and adopted by the Texas Department of Health as specified in the 2000 Code of Federal Regulations (CFR), specifically 21 CFR Part 165.

 PGA has selected three properties in Wood County and part of Upshur County on which to evaluate potential spring water sources.  Thornhill & Associates (Thornhill) and TITAN Engineering, Inc. (TITAN) conducted cooperative investigations as part of PGA’s due diligence evaluations of the properties.  Thornhill & TITAN designed field investigations and other work tasks to meet PGA’s goals and needs of the project, including:

 (1)     Assessing the suitability of the various spring sites for producing spring water, including estimating the amount of water that can be produced from each site, evaluating the quality of water available, making preliminary determinations as to the most appropriate methods and facilities for producing spring water, and evaluating the probabilities for sources being successfully developed and approved as spring water sources with respect to PGA standards, State regulations and FDA requirements and definitions.

(2)     Collecting sufficient information and data and assessing the potential impacts of producing spring water from the subject properties, including impacts to aquifers, water levels in local wells, spring flows and related stream flows, and water quality.

(3)     Compiling the information and data to formulate the baseline studies for any future work at the subject properties, specifically for providing the information in an effective and clear manner to PGA for its use in potential spring source development.  In addition, the information is available to provide technically and scientifically sound hydrogeologic and hydrologic data to those who may have concerns of impacts that may result from PGA’s activities.

 Investigations were supervised and conducted by Thornhill and TITAN, and drilling, borehole completion, pump installation and testing services were provided by Andrews & Foster Drilling Company (Andrews & Foster).  Generally, work tasks included:

 ·        Meetings and correspondence with PGA, TITAN, Andrews & Foster, and Thornhill, including initial site visits, to discuss goals, work plans, progress and preliminary results.

 ·        Compiling and evaluating background information and data including geologic maps and cross sections, reports pertaining to the hydrogeologic conditions and ground water availability, well data and records, topographic maps, precipitation data, stream flow records and water quality data.

 ·        Field investigations including:

­       A Phase 1 Environmental Site Assessment conducted by TITAN.

­       Drilling and geologic exploration.

­       Borehole and monitoring/observation well completion.

­       Aquifer testing, including 24-hour and 7-day pumping tests.

­       Sampling of springs and test boreholes, including collecting field water-quality parameters and samples for laboratory analysis.

­       Background monitoring including flow studies (TITAN), water-level monitoring utilizing automated recorders, and precipitation monitoring (TITAN).

·        Data workup including compiling field data collected over the 18-month study period, and preparing appropriate maps, charts, diagrams, graphs and drawings to conduct hydrogeologic evaluations and to illustrate the work and results of the project.

 ·        Preparing written progress summaries and final written reports.

 Background Information

 The three subject properties are located primarily within eastern Wood County, with part of one property extending eastward into Upshur County.  The properties are about 3 to 9 miles apart, and the nearest common road to all of the subject properties is FM 2869, which is between about 0.1 and 2 miles from the entrances to the properties.  The properties are between about 6 and 20 miles from nearby cities including, from nearest to farthest, Hawkins, Winnsboro, Quitman and Mineola, and are about 100 miles from Dallas.  The nearest major highways to Dallas are U.S. Highway 80 and Interstate Highway 20, which are about 15 to 35 miles respectively from the most distant property.  For identification purposes, each of the three subject properties is referenced herein by its approximate acreage; the 250-Acre Property, the 820-Acre Property and the 1,100-Acre Property.

 The land surface in the study area is characterized by gently rolling to hilly topography, with sandy soils.  The northernmost properties, the 250-Acre and 1,100-Acre tracts, are located within the drainage basin of Little Cypress Creek, while the 820-Acre tract is located along the main stem of Big Sandy Creek.  Ground-water discharge including numerous seeps and springs provides the baseflow for these perennial streams and other tributary drainages in the area.  Small reservoirs impounding flows of these creeks are located upstream and downstream from the subject properties.  Much of the land on and near the properties is covered with evergreens and deciduous vegetation, though some has been cleared by foresting and some cleared for pastureland.  The average annual precipitation for the area is approximately 44 inches.

 The Sparta Sand is the uppermost (i.e. youngest) geologic unit and aquifer present in the study area, except for Alluvium deposits in major stream channels.  The Sparta crops out as isolated outliers overlying the Weches Greensand.  The Sparta Sand is less compacted in its outcrop areas than the older formations in the area, and weathers to deep, very loose, light gray sandy soil.  The Sparta Sand is generally considered excellent for infiltration, and springs issuing from the base of the Sparta provide significant contributions to the base flow of Big Sandy Creek.  The Sparta Sand ranges up to about 250 feet thick in Wood County. The Weches is about 50 to 75 feet thick, and separates the Sparta from the Queen City.  The Weches Greensand is characterized by interbedded glauconitic sand and clay, and commonly contains secondary deposits of limonite, or “ironstone” which is altered by weathering to hematite.  The Queen City crops out across most of the southeastern two-thirds of Wood County, and is composed of alternating and interbedded layers of sand, clay and shale with some lignite seams.  Numerous seeps and springs discharge along the contact planes of the sand beds and the underlying clay layers.  The sand layers in the Queen City are typically thinner and less laterally consistent than those of the Sparta Sand, and the Queen City locally contains more clay.  The maximum thickness of the Queen City Sand is about 400 feet in Wood County.  The primary and most productive aquifer in the area is the Carrizo-Wilcox aquifer, which is tapped by most of the local moderate- to large-capacity wells.  Locally, this aquifer is tapped by wells between about 240 and 1,250 feet deep.  The springs on the subject properties generally issue from horizons within the upper 60 feet of the Queen City or Sparta sands, and are completely separate from the Carrizo-Wilcox aquifer, which is locally separated from the Queen City by the overlying Reklaw Formation.

 The aquifers are recharged as part of the precipitation that falls on the outcrops infiltrates the soil layers and percolates to the generally shallow water table.  Within outcrops of sandy units such as the Carrizo, Sparta and Queen City, recharge amounts typically range from 25 to 50 percent of average rainfall.  Because the Queen City and Sparta aquifers are full to overflowing in Wood and Upshur counties, much of the water that percolates to the water table moves only relatively short distances and discharges to the numerous springs and seeps that occur within the many tributary drainages and major streambeds in the county.  This discharge, or rejected recharge, averages about 50,000 acre-feet per year, or 31,000 gallons per minute (gpm) in Wood County.  An additional 50,000 acre-feet per year or more flows downdip and recharges the deeper portions of the aquifers.  Development of shallow spring sources could not impact the quantity and quality of water in the deeper parts of the aquifer. 

 Recharge areas for individual spring sites are typically very localized.  The shape of the water table conforms to a suppressed version of the land surface topography, so ground water flows generally from the hilltop areas toward the drainages.  Depths to the water table range from land surface at spring and seep areas to about 45 feet beneath hilltops.  Typical ground water flow velocities in these aquifers range from less than one foot per day (ft/day) to more than 10 ft/day in sand zones.  Most of the ground water discharged in Wood and Upshur counties is discharged via springs, seeps and as evapotranspiration to the oftentimes densely vegetated areas.  Only a small amount of water is pumped from wells or boreholes, which mostly tap zones deeper than those contributing to spring flows. 

 The quality of water available from the shallow portions of the Sparta and Queen City aquifers is generally excellent, with low total dissolved solids contents and no undesirable mineral or nutrient (i.e. nitrate) contents.  As with most East Texas springs in similar settings, iron, manganese and low pH are concerns.  Studies evaluating iron occurrence in aquifers in Wood County indicate that the shallow zones, generally from the water table to an altitude slightly below the base of the larger stream valleys, are relatively iron-free (Broom, 1968).  Commonly, iron concentrations in samples collected directly from spring and seep areas are due to the presence of suspended solids in samples, or to reduced conditions caused by organic and vegetative litter at the discharge (and sampling) points.  Ground water in deeper zones and aquifers is typically more highly mineralized than in those zones contributing to spring flows.

Summary of Spring Assessment Results

 PGA personnel and landowners identified several springs on each property for which assessments were conducted.

 PGA conducted several rounds of sample collection from the natural spring openings prior to the additional field investigations conducted herein.  Actual springs orifices are typically difficult to delineate for individual springs, as discharge areas are often characterized by large seep areas with numerous small spring flows at the base of steep drainage slopes that are covered with geologic and plant debris and dense growth of vegetation.  Aggregate flows of several springs and seeps at one site typically formed significant flows, however, the natural conditions cause measuring total flows to be difficult.

 Field geologic evaluations included drilling several exploration holes at each site to try to determine the depth, extent and geologic and hydrologic character of zones feeding the springs.  Boreholes were completed at sites exhibiting apparent favorable conditions including sufficiently thick saturated sand sands.  At five of the seven sites on the 1,100-acre property no boreholes were completed due to insufficient occurrences of sand.  Boreholes were completed at all sites drilled on the 250-acre and 820-acre properties.  Testing of sites with completed boreholes included pumping tests, sampling and water-level monitoring.

 The springs on the 250-Acre and 1,100-Acre properties issue from shallow parts of the Queen City Sand, while the springs on the 820-Acre Property flow from the base of the Sparta Sand.  The geologic and hydrologic characteristics of the Sparta Sand sites are typically more favorable for developing spring sources via boreholes than for those sites in the Queen City Sand.  Tables A, B and C provide summaries of results from the spring source assessments, and results for each property are discussed below.

The results and conclusions provided herein are based on field data collected during a period between May 2000 and November 2001.  Precipitation records indicate that rainfall near the properties was about 11 inches less than normal from January 1999 through August of 2000, and no precipitation occurred from mid-June through mid-September.  Therefore, testing conducted during the summer of 2000 should provide a relatively conservative approximation of property production capabilities.  Precipitation from November 2000 through March 2001 substantially exceeded the average, and was reflected in water-level rises and spring-flow increases.  The long-term capabilities of spring sites and their resistance to drought can only be more accurately determined with continued monitoring of precipitation, water levels and spring flows over periods of several climatic changes.

Conclusions

·        An abundance of spring flow is available - Discharge to springs and seeps makeup large portions of the flow in Little Cypress Creek and Big Sandy Creek.  Flows of individual springs are typically difficult to delineate and are relatively small, as the discharge areas are characterized by large spring and seep areas and steep drainage slopes are covered by weathered geologic debris and vegetation.  Aggregate flows at each site are significant and measures should be taken to enhance conditions in discharge areas, so that total flows can be identified and measured.

 ·        Boreholes are the most favorable method for spring water production – At suitable sites, boreholes allow for more effective, efficient and sanitary collection of spring water.  Due to springs typically discharging in the bottoms of stream drainages and the occurrence of fine, loose sandy material and organic matter at land surface and the shallow subsurface, completing alternative spring catchment structures may be difficult.

 ·        Springs produce sanitary, excellent quality water – Each of the springs tested produces water with very low total dissolved solids concentrations similar to the excellent quality water produced at other PGA sites.  Each test borehole generally produced the same excellent quality of water.  Iron was detected in samples collected directly from most of the springs, while no iron was detected in any of the borehole samples.  The presence of iron in samples collected directly from the springs is likely due to suspended sediment in samples, or due to the reducing environment caused by the natural organic or vegetative debris at the spring discharge areas.  Filtrate testing indicated that water collected from boreholes is sanitary and not under the influence of surface water.

 ·        Spring zones are separate from primary targeted producing aquifers and zones – The water-bearing zones contributing to the springs on the properties are typically less than about 60 feet deep.    Most nearby wells tap primary targeted zones and aquifers that are typically between about 100 and 1,200 feet deep beneath the properties.  Additionally, the contributing or recharge areas for the spring areas are typically localized, and do not include large areas extending beyond property boundaries.

 ·        Spring water production will not cause adverse impacts – Capturing spring water via boreholes or other catchment methods will not cause adverse effects to the hydrologic conditions in the area.  Spring flow reduction will be minute compared to overall flows in the streams.  Borehole production will not cause any measurable draining of local aquifers, nor will it cause detectable water-level declines in wells near the property.

 ·        Sparta Sand sites are more favorable for borehole production than Queen City Sand sites – In East Texas, favorable target areas for borehole production of spring water are typically located in isolated outlier outcrops of very sandy geologic units such as the Carrizo Sand and the Sparta Sand.  The 820-Acre property is located within this type of area.  Testing indicated potential successful borehole completion at 3 of 4 spring sites.  Due to thinly bedded alternating layers of sand, clay and shale, the Queen City Sand poses more difficulty in utilizing boreholes for spring water production.  The 250-Acre and 1,100-Acre properties are located atop the Queen City Sand outcrop.

 ·        Spring water can successfully be developed from one or more properties – Spring water sources complying with PGA’s quality standards and State and Federal regulations can be developed on one or more of the subject properties.  The 820-acre property appears to be the most favorable site for producing spring water via boreholes.  For the other two properties, borehole production of spring water will not be as favorable, and other methods such as horizontal boreholes or collection boxes may be difficult to install.