DOE states that water quality standards for the Pecos River will not be impacted by WIPP because: (1) the Pecos River is located 12 miles west of the WIPP site (Comment No. 273.X.B.6, p. 29); (2) there are no natural drainage features at the WIPP site (Id., p. 29); (3) no surface releases will occur at the WIPP (Comment No. 273.X.B.7, p. 30); (4) there is no hydrologic connection between the WIPP repository and surface water (Comment No. 273.X.B.3, p. 28); and (5) there is no driving mechanism that will allow contaminants to migrate through the salt to a groundwater unit that discharges to surface water. (Id., p. 28) Let us examine these assertions one at a time.

1. The Pecos River is indeed 12 miles west of the WIPP site (Phillips, 1987, Figure 2) [Exhibit 3].
   
   
2. The WIPP site has almost no surface runoff. This is not due to inadequate precipitation, which averages 14.2 inches per year in Nash Draw (Phillips, 1987, p. 222). Rather, the WIPP site is covered with windblown sand in the form of deflation basins and partially stabilized sand dunes. These sands are transmissive enough to allow infiltration of even the largest storms (Phillips, 1997b, p. 1). “Instead of running off, the precipitation collects in the small topographic depressions and rapidly soaks into the ground. The absence of surface runoff is characteristic of a karstland.” (Barrows, 1982, p. 9, reprinted as EEG-32, 1985, Appendix A). This is true even in semiarid climates. Apart from dune fields, deserts on impervious rocks rarely lack stream courses, however ephemeral they may be. But karst in semiarid regions is usually without them, as in the Nullarbor Plain of Australia and the Pecos River Valley of New Mexico (Phillips and Snow, 1998, p. 7).
   
   
3. Surface releases could occur at the WIPP, even before closure, by means of a blowout due to waterflooding and hydrofracture at an oil well within two miles of the WIPP repository (direct testimony of John Bredehoeft at RCRA hearings, March, 1999). At one time DOE retained a two-mile buffer (Zones III and IV) surrounding the WIPP repository in order to prevent waterflooding and hydrofracture, to prevent solution mining for potash, and to oversee the eventual plugging of oil and gas drill holes (FEIS, 1980, p. 8-4). DOE relinquished Zone IV in 1983 (Phillips and Snow, 1997, Figure 2) [Exhibit 6], thus reducing the buffer to one mile. The rationale, according to DOE, was that “the minimal amount of crude oil likely to exist within the WIPP site” made waterflooding adjacent to WIPP unlikely (EEG-55, 1994, p. 21) (EEG-62, 1996, p. xiv). There has since been an oil and gas boom in the immediate vicinity of the WIPP site. As of January 1998 there were 27 operating oil and gas wells within the old Zone IV, fifteen of them within two miles of the waste panels (Phillips and Mitchell, 1998, p. 2, Figure 1).
   
   
4. There are hydrologic connections between the WIPP repository and the land surface, namely, the waste handling shaft, the salt handling shaft, the air intake shaft, the exhaust shaft, and the ERDA-9 borehole (see maps at EEG-71, 1998, Figure 1, and CCA, Figure 3-9) [Exhibits 41 and 42]. The four WIPP shafts connect directly to the repository (EEG-71, 1998, Figure 1) [Exhibit 41], and the ERDA-9 borehole is near enough to the repository footprint (CCA, Figure 3-9) [Exhibit 42 to be within the disturbed rock zone (EEG-68, 1998) (CCA, Appendix PEER), and therefore must connect the WIPP excavations to the land surface. Ultimately, waste containment at WIPP depends upon DOE's ability to seal the shafts and plug the boreholes perfectly. There is no proven technology for plugging boreholes in salt formations, and CARD doubts that it can be done successfully. DOE attempted in 1977 to plug the ERDA-10 borehole at the Gnome site near Nash Draw. Four separate plugs were emplaced for a total length of 4430 feet (SAND 81-2034). There appears to be no record of the success or failure of the attempt.
   
   
5. There is a driving mechanism that could allow contaminants to migrate through a less than perfectly sealed shaft or borehole to overlying groundwater units. In 1981 a pressurized brine reservoir associated with hydrogen sulfide gas was encountered at the WIPP-12 borehole, one mile north of the center of the WIPP site, about 1.2 miles north of the waste panels. The brine is in the upper Castile anhydrite, 240 feet below the Salado Formation. The brine flowed to the land surface at a rate of 1500 barrels per day for forty days. Total brine outflow was 60,000 barrels, or 2.5 million gallons (DOE, 1982, TME 3148). The total volume of the WIPP-12 brine reservoir was later estimated at between 17 million gallons (EEG-23, 1983, p. 29) and 30 million gallons (EEG-22, 1983, p. 79). By comparison, about 63 million gallons would be necessary to completely fill the WIPP repository (EEG-16, 1982, p. 45). The WIPP-12 brine reservoir is estimated to underlie as much as 60% of the waste panels (cross-examination of Robert Kehrman at RCRA hearings, 02/23/99) (see also CCA, p. MASS-105, and Mass Attachment 18-5). The ERDA-9 borehole penetrated 53 feet deep into the Castile. All that separates the WIPP-12 brine reservoir from ERDA-9 and the WIPP repository is 200 feet of vertically fractured anhydrite.

It should be noted that as of January 1998 there were 177 operating oil and gas wells within two miles of the WIPP site boundary, and 47 more had been planned and located (Phillips and Mitchell, 1998, p. 2, Figure 1). DOE plans to prevent any drilling at the WIPP site for 100 years after closure, longer than the duration of a RCRA permit. But for the record, it seems inevitable that, after institutional controls are lost, someone will drill through the karstic Rustler aquifer, through a waste panel, and into a pressurized brine reservoir, thereby breaching the WIPP repository, without ever reaching the underlying oil and gas horizons. That is, if the site has not already been breached by hydrofracture.

It is important to identify where contaminated water escaping from the WIPP repository would reach the accessible environment. Two regional groundwater discharge points are known: Laguna Grande de la Sal in Nash Draw, and the brine springs at Malaga Bend on the Pecos River. Phillips (1987, pp. 219-222, 232-235) showed through evaporation analysis that groundwater discharge to Laguna Grande de la Sal is about nine times the amount of groundwater discharge at Malaga Bend.

DOE states correctly that “Nash Draw is the nearest major geomorphic feature” to the WIPP (Comment No. 273.X.B.3., p. 27). Nash Draw is one of the largest karst features with surface expression in the world. Bounded by cliffs, Nash Draw is a closed drainage basin, 18 miles long, 5 to 10 miles wide, and 200 feet deep, formed by the coalescence of thousands of sinkholes (Phillips, 1997a, p. 12). DOE agrees that Nash Draw is an undrained physiographic depression resulting from differential solution of the Rustler and upper Salado (Comment No. 273.X.B.3, p. 27). The eastern rim of Nash Draw, called Livingston Ridge, reaches within one mile of the WIPP site boundary.

The WIPP site lies within the Nash Draw drainage basin. The lowest point in the basin, both topographically and potentiometrically, is Laguna Grande de la Sal in Nash Draw. It is a salt lake with no outlet at the surface or underground; it loses water only by evaporation, which is why it is a salt lake (Phillips, 1987, pp. 216-218). The karst springs that drain the Rustler Formation reach the surface at Laguna Pequena (Phillips, 1987, Plate 23) [Exhibit 43], the most copious inlet to Laguna Grande de la Sal, where an inflow of nearly 400 ft3/sec was measured following a four-inch rainstorm (Phillips, 1987, pp. 227-231). There is no other apparent surface runoff into either lake, and so the regional water balance may be expressed as follows:
 

E - P = I
 
where E = evaporation from the lake surface, P = precipitation falling directly onto the lake surface, and I = groundwater inflow to the lake. The USGS estimates that brine evaporation equals 90 inches (7.5 feet) per year in the vicinity of Laguna Grande. Precipitation equals 14.2 inches (1.18 feet) per year in Nash Draw. The natural extent of Laguna Grande, as mapped by Robinson and Lang in 1934, when potash mining began in Nash Draw, was about 2,120 acres (9.23 x 10⁷ ft²) [Exhibit 44]. Net evaporation from Laguna Grande would equal 5.84 x 10⁸ ft³/yr. At least this amount of water drains from the Rustler aquifer into Laguna Grande, and an equal amount of infiltrating rainwater must reach the Rustler Formation (Phillips, 1987, pp. 219-222).

In a karst terrain such as the Nash Draw watershed, there is almost no surface runoff; drainage is almost entirely underground. Thus the regional water balance may also be expressed this way:
 

P - I = E
 
where P = precipitation, I = infiltration, and E = evapotranspiration. From analysis of USGS topographic maps, the size of the Nash Draw watershed may be estimated at 226,000 acres, or 9.84 x 10⁹ ft². Potentiometric contour maps indicate that this topographic divide approximates the groundwater divide. If precipitation equals 1.18 ft/yr, then precipitation falling on the watershed is 1.16 x 10¹⁰ ft³/yr. The infiltration rate of 5.84 x 10⁸ ft³/yr would equal about 5% of annual precipitation, and so the rate of evapotranspiration would be about 95% (Phillips, 1987, pp. 222-224).

DOE claims that if more than 90% of precipitation is lost to evapotranspiration, then “infiltration below the surface is negligible.” (Comment No. 273.X.B.4, p. 28). To the contrary, an infiltration rate of only 5% results in a salt lake 2,120 acres in extent. It should be noted that Laguna Grande is only eight miles from the WIPP site, not ten miles as DOE suggests (Comment No. 273.X.B.3, p. 27).

DOE does admit that “intense local thunderstorms may produce runoff and percolation.” (Comment No. 273.X.B.4, p. 28). EPA states that about 75% of total annual precipitation results from intense thunderstorms between April and September (EPA Docket # A-93-02, Item # III-B-3, p. 88). Phillips (1987, pp. 84, 86) stood in one of these thunderstorms on September 18-19, 1985. He observed five feet of standing water in the WIPP-33 sinkhole, carried there by a disappearing arroyo. The water sank into the sand within days, leaving behind a “bathtub ring” of organic debris to record the high-water mark. Phillips also observed a brand new arroyo appear on the landscape, only to disappear into another sinkhole previously identified by hand augering (Phillips, 1987, Figures 29 and 30) [Exhibits 45 and 46]. These field observations of rapid rainwater recharge are proof that karst processes are active today. WIPP-33 is the westernmost of a chain of four sinkholes (Phillips, 1987, Figure 5) [Exhibit 47], indicative of an underground flow path beneath them. The easternmost sinkhole is within 1000 feet of the WIPP site boundary.