DIRECT TESTIMONY AT EUNICE, NEW MEXICO Richard Hayes Phillips, Ph.D. Monday, January 29, 2007 Seventeen days ago I was asked to investigate the groundwater hydrology in the immediate vicinity of a site proposed by Louisiana Energy Services (LES) for a uranium processing facility. The site is located in Lea County, New Mexico, one-half mile from the Texas state line. Directly across the state line in Andrews County, Texas is a site proposed by Waste Control Specialists (WCS) for an officially low-level nuclear waste dump. The two projects are related. Test boreholes for both sites were drilled and logged by the same companies, and it has been suggested that waste from the LES site might be dumped at the WCS site. Because groundwater flow is not at all influenced by property lines or political boundaries, this study considers borehole data from both sides of the state line and presented by the applicants to the regulatory agencies in either state. Included are 199 boreholes, 78 of them in New Mexico, 121 of them in Texas. There are seven sets of boreholes; the areas covered by them are contiguous, and sometimes overlap. This study correlates lithologic descriptions from the original borehole drilling logs, and maps the structural contours of the Quaternary caliche surface and the top of the underlying Triassic red beds. Field investigation of surface geomorphology and hydrology is focused on a corridor of land 1.4 miles long and 0.5 miles wide, entirely within the State of New Mexico, situated between the LES and WCS sites. Within the study area is the convergence of three hydrologic and geomorphic regions: the Ogallala aquifer of the Llano Estacado or High Plains; the sandstone and limestone aquifers of the Edwards Plateau; and the alluvium and siltstone aquifers of the Pecos River Valley. The Ogallala aquifer and the Edwards aquifers are known to be interconnected, and have generally been undifferentiated as the High Plains aquifer. Thomas M. Lehman, in a geological study presented by WCS in its application to the State of Texas, attempts to differentiate between the two aquifers by distinguishing between the Ogallala Formation of the High Plains and the Antlers Formation of the Edwards Plateau. Where the Ogallala Formation consists partly of reworked Antlers sediments, such a distinction is difficult to establish with certainty. The Texas Bureau of Economic Geology states that: “Available maps do not unambiguously identify any area of Andrews County where the High Plains aquifer is absent.” The Bureau regards “the presence or absence of the Ogallala Formation (as) more of a public perception issue than a technical issue related to site performance.” Simply stated, groundwater does not care whether it is flowing through Antlers or Ogallala sand and gravel, but the public does care about preventing contamination of the Ogallala aquifer, which provides nearly all the water for irrigation of the High Plains from Texas to Nebraska. The Edwards Plateau is recognized as one of the largest karst regions in the United States. According to Texas Department of Water Resources Report 235, groundwater flow in the Edwards limestone and in the underlying carbonate-cemented Antlers sandstone is through fractured or jointed rocks, solution channels, and caverns. Karst topography is characteristic, and the land surface features numerous sink holes caused by the slumping or caving of the ceilings of caverns. Surface runoff extends only for short distances before it disappears into the sink holes. Groundwater in the Edwards/ Antlers aquifer flows generally in a southeasterly direction. However, where solution channels have developed, the direction of flow may be quite different than the water- table map indicates. EXHIBIT 1. MAJOR KARST AREAS OF THE UNITED STATES (From LeGrand et al., 1976)The Pecos River Valley is also recognized as one of the largest karst regions in the United States. Groundwater in the Pecos River Valley flows generally in a southwesterly direction through formation successively lower in the stratigraphic column: Quaternary and Cretaceous alluvial deposits, Triassic red beds, Santa Rosa sandstone, Dewey Lake red beds, and the gypsum and dolomite of the Permian Rustler Formation. The portion of the Pecos River Valley lying within the study area is the Monument Draw watershed. Monument Draw flows almost due south through Lea County, always within 2.5 miles of the Texas state line, crossing into Texas one mile east of Route 18, where it bends to the southwest, flows through Cheyenne, and thence to the Pecos River. In New Mexico, the surface drainage course of Monument Draw is incised and plainly visible in air photos. However, it seems to disappear underground before it gets to the Pecos River. WCS, in it application to the State of Texas (p. 2-6), acknowledges that Monument Draw “developed in response to subsurface dissolution of evaporites,” presumably the halite and gypsum of the Salado and Rustler formations. LES, in its Environmental Impact Statement submitted to the State of New Mexico (p. 3-33), acknowledges that while Monument Draw is typically dry, its maximum historical flow was measured at 575,000 gallons per minute. EXHIBIT 2. MONUMENT DRAW (From www.googleearth.com). North is to the left.
Whatever aquifers underlie Lea and Andrews counties, rainwater recharge occurs primarily through perforated or degraded caliche caprock. Caliche is formed in desert soils when calcium carbonate, supplied by the wind, leaches downward to the usual depth of soil water penetration, and remains there as soil water evaporates. As windblown sand thickens the soil profile, the caliche accumulates. Caliche can rapidly harden where exposed at the land surface, undergoing brecciation, cementation, and induration. Hardpans on top of caliche profiles form caprocks, ledges, and escarpments. Calichification is a reversible process. Caliche is essentially limestone, and may develop solution features when downward percolating rainwater causes solution and removal of carbonate. Where the caprock forms a plugged hardpan horizon, rainwater infiltration is reduced, thereby accelerating runoff and flooding. But the caprock is commonly fractured into plates or blocks. Infiltrating rainwater will migrate along the caprock until it reaches an opening which allows it to move downward. The caprock is always underlain by softer or looser material which is more prone to erosion and may contain cave systems, undercutting the caprock. A whole suite of karstic landforms may develop, such as closed depressions and discontinuous drainage. Karst features in caliche have been described in southeast Texas, in south Texas, in the Ogallala caliche of the Llano Estacado, and in the Mescalero caliche of the Pecos River Valley. Depressions are still forming on the Llano Estacado in Lea County, New Mexico. The depressions are generally aligned in groups, or chains. Some are so shallow as to be noticeable only by a vegetative change, such as the chain of ten or more depressions extending west-northwestward from the proposed federal facility in Texas and into New Mexico, as shown in the black-and-white air photo dated January 14, 1996, when the site was still undisturbed, submitted by WCS in its application to the State of Texas. Such depressions may be incipient sink holes along poorly formed drainage courses. Solution of caliche would entrench the drainage and lead to enlargement of the depressions, which may coalesce to form broad swales. EXHIBIT 3. AIR PHOTO SUBMITTED BY WCS. (Cook-Joyce, Inc., Figure 6.4-26a).
Trench exposures at the Waste Isolation Pilot Plant (WIPP) site in Eddy County, New Mexico, near the Lea County line, revealed a discontinuous caliche surface presenting no barrier to the infiltration of rainwater. Photographs of the trench exposures are included in my doctoral dissertation, were submitted in color to the New Mexico Environment Department (NMED) in 1999, and are posted on my website at www.northnet.org/minstrel, Appendix WIPP. Trenching beneath a broad swale revealed numerous solution pipes in caliche, ranging from a few inches to 14 feet in diameter, many of them penetrating to the underlying red beds. Trenching beneath the WIPP-14 sink hole revealed bluish-gray gleyed sediments, indicating past ponding of the depression during which time the caliche was leached and degraded. Trenching beneath the WIPP-33 sink hole, where the caliche is entirely absent, revealed solution- enlarged holes and fractures in carbonate-cemented sandstone. In the vicinity of the LES and WCS sites, the watershed divide for both surface water and groundwater is uncertain. Thomas H. Lehman states that the Red Bed Ridge “probably acts as a regional groundwater divide” separating the Ogallala aquifer to the northeast from the Pecos River aquifers to the southwest. There is support in the geologic literature for this interpretation. The “Red Bed Ridge” is a buried structural feature, being the crest of the Triassic red bed deposits underlying the Antlers Formation, or the Ogallala Formation, or both. Lehman has provided a structural contour map of the top of the red beds in most of Lea, Andrews, and Gaines counties. The map shows the axis of the “Red Bed Ridge” coinciding with the Mescalero Ridge, a caliche caprock escarpment marking the southwestern edge of the Llano Estacado in most of Lea County. The “Red Bed Ridge” then extends southeastward into Andrews County in the immediate vicinity of the LES and WCS sites. There it is a very subtle geologic feature; its structural relief is 17 to 27 feet per mile when measured perpendicular to the axis of the ridge. EXHIBIT 4. ELEVATION ON TOP OF RED BEDS. (From Lehman, 1996, in WCS Application, Appendix 6.2-2, p. 21).
Lehman and Rainwater have provided a map covering 25 square miles in Andrews and Lea counties. The map shows where the caliche caprock is at or near the land surface. Not surprisingly, this roughly coincides with the axis of the “Red Bed Ridge.” The map indicates that the ridge lies directly underneath the WCS site in Andrews County, and to the northeast of the LES site in Lea County. The map also shows 38 boreholes at 35 locations on the Flying “W” Ranch, 34 of which function as groundwater piezometers. From the detailed lithologic logs appended to Lehman’s report I was able to construct contour maps of the caliche surface throughout the area, and of the groundwater table northeast of the divide. EXHIBIT 5. MAP OF CAPROCK CALICHE. Shown in white. (From Lehman and Rainwater, 2000, in WCS Application, Appendix 6.2-1, p. 23).
Figure 1 leaves little doubt that the WCS site is located directly on the caprock divide, and implies that the LES site, located to the west of the WCS site, lies southwest of the caprock divide. Figure 1 identifies the locations of 17 boreholes that struck water. Of these, according to Lehman, 13 were in the Antlers Formation, three were in the Ogallala Formation, and one was in caliche. Three boreholes struck water directly beneath the caprock divide, and two struck water southwest of the caprock divide, which raises doubts as to whether the structural divide and the groundwater divide are one and the same. Southwest of the divide, caliche in 5 of 15 boreholes was described as “poorly developed” and/or “discontinuous,” presenting no barrier to rainwater infiltration. In all five boreholes the Antlers and Ogallala formations are absent. All five boreholes bottomed in the upper 15 feet of the Triassic redbeds, and might have struck water if the drilling had been deeper. FIGURE 1. CALICHE SURFACE AT FLYING "W" RANCH
Figure 2 maps the water levels of all 17 wells that struck water, whether in the Antlers Formation, in the Ogallala Formation, or in caliche. Smooth contour lines can be drawn for the water table no matter where in the stratigraphic column the water was encountered. Thus the distinction between the porous sands and sandstones of the Antlers and Ogallala formations, although fascinating to stratigraphers, is of little consequence from the standpoint of groundwater hydrology. Contaminated water escaping to the northeast of the LES or WCS sites would flow downgradient, perpendicular to the contour lines, in a generally southeastward direction, into the Ogallala aquifer and toward the karstic limestone aquifers of the Edwards Plateau. FIGURE 2. WATER TABLE AT FLYING "W" RANCH
The WCS site in Andrews County, Texas, and the structural ridge on which it is located, has been investigated in earnest. About 130 soil borings were drilled into an area about 10,800 feet long and 3,600 feet wide, more than one borehole per 7 acres. This has enabled me to construct contour maps of the caliche surface and the top of the underlying red beds in appropriate detail. By contrast, the investigations of the LES site in Lea County, New Mexico, and the corridor of land between the LES site and the state line, have been inadequate. Only 14 soil borings were drilled into the LES site, and another 14 southeast of the LES site, 28 boreholes in all, to investigate an area of about 870 acres, or one borehole per 31 acres. This has forced me to construct contour maps of the caliche surface and the top of the underlying red beds in far less detail than I would have liked, but it is the best that I can do. New Mexico deserves an environmental investigation as thorough as the one in Texas, and the New Mexico Environment Department (NMED) should require it. There are absolutely no boreholes north of the LES site, so there is no way of knowing whether contaminated water escaping the LES site could flow northward, or what would happen if it did. To make the best of an unfortunate situation, I am submitting to NMED, for the record, by concise transcriptions of the borehole data from which my maps were derived. On Figures 3 and 4, the waste disposal areas are depicted with rectangular lines: the LES evaporation ponds to the northwest, and the WCS disposal facilities to the northeast. Boreholes that struck water are depicted with bulls-eyes. FIGURE 3. CALICHE SURFACE AT WASTE DISPOSAL SITES
Figure 3 shows a caliche surface sloping steadily southwestward, dropping 60 feet in 0.75 miles between the facilities, and another 60 feet in 1.2 miles south of the facilities. Above the break in slope, infiltrating rainwater could be expected to evaporate or to flow downhill along the surface of the caliche caprock until joints, fractures, or solution features allow the water to pass through the caliche profile. Below the break in slope is a pathway within which the lithologic logs for all 12 boreholes describe the caliche as soft, weak, loose, poorly cemented, semi-cemented, semi-consolidated, or friable (easily crumbled). This is clearly a pathway of preferential rainwater infiltration. At the head of this pathway are two boreholes which struck water, and a structural depression 27 feet deep in the caliche. Similar structural depressions in the caliche appear near Windmill Hill, also near a borehole which struck water; and within the southeast corner of the LES site downhill from borehole MW2, which produced water within 24 hours of the completion of drilling, a fact not revealed in the Environmental Impact Statement (p. 3-37). Borehole MW-2 is located on-site, northeast of the proposed evaporation ponds. Figure 4 depicts, east of the Texas state line, a relatively flat-topped red bed ridge, 1200 to more than 3000 feet wide, pockmarked by at least seven structural depressions. One of these structural depressions, 12 feet deep in the red bed surface, lies directly underneath the WCS federal waste disposal facility. Another structural depression, 19 feet deep in the red bed surface, coincides with the 27-foot-deep depression in the caliche surface described earlier. To the south and west are four more deep depressions in the red bed surface, one of which has cut into the edge of the flat-topped red bed ridge; the other three coincide with the pathway of soft, loose, poorly cemented caliche described earlier. The break in slope in the red bed surface at this location is much more abrupt than in the caliche. Southwest of the flat-topped ridge, the red bed surface drops 50 feet within a distance of 1000 feet, then drops only another 75 feet in 1.67 miles. One other structural depression was mapped in both the caliche and red bed surfaces; this one, located within the eastern boundary of the LES site, is behind a chain link fence and not open to field investigation. FIGURE 4. RED BED SURFACE AT WASTE DISPOSAL SITES
The fact that the proposed WCS waste disposal sites in Andrews County are located on a flat-topped red bed ridge pockmarked with structural depressions demonstrates that contaminated water escaping the WCS facility could flow in either direction. It could flow northeastward into the Antlers or Ogallala formations and thence to the karstic aquifers of the Edwards Plateau, or it could flow southwestward into the structural trough of Monument Draw and thence to the karstic aquifers of the Pecos River Valley. In the case of the LES evaporation pond site in Lea County, the applicants have provided a contour map of the water table (Appendix B, p. 69) showing that contaminated water escaping the LES site could flow southward toward Monument Draw. To say that there is no hydrologic evidence that water could flow northward toward the Ogallala aquifer is a truism, because there are no soil borings or test wells located north of the LES site. Guided by the available data, I conducted a field investigation of readily accessible land located entirely in New Mexico, between the LES and WCS sites. The investigation focused on two features: a structural depression located approximately 1.2 miles north of Highway 176 and very near the state line, shown in some air photos as holding muddy water; and Baker Spring, located approximately 1.4 miles north of Highway 176 and 0.35 miles west of the state line. Both are identified on USGS topographic maps showing Baker Spring as holding water perennially, and the structural depression as holding water ephemerally. GO TO NEXT PAGE FOR EXCERPTS FROM SLIDE SHOW