Permit Module V (V.C.1, pp. V-1 and V-2) states that the Detection Monitoring Program (DMP) for groundwater contamination shall consist of seven wells -- six in the Culebra dolomite (WQSP-1, WQSP-2, WQSP-3, WQSP-4, WQSP-5 and WQSP-6) and one in the Dewey Lake Redbeds (WQSP-6a). No other geologic strata are to be monitored, and no other test well locations are contemplated.
The six test wells completed to the Rustler aquifer are all in the Culebra dolomite. This reflects DOE's erroneous concept of the Culebra as a confined aquifer, bounded above and below by impermeable anhydrite beds. There is ample evidence that the Rustler is recharged by rainwater (Phillips, 1997b, all) (Phillips and Snow, 1997, pp. 1-4) (Phillips, 1998a, pp. 4-7, 15, 17, 19) (Snow, 1998a, p. 10), and that all members of the Rustler are involved in groundwater transport (Phillips, 1998a, pp. 2-3, 11-12, 28-30)
The six Culebra test wells are located at random, in a hexagonal array, surrounding the WIPP repository (CCA, Figure 3-9). Six monitoring wells, some of them hydraulically upgradient from the hazardous waste facility being monitored, might not be considered sufficient for an ordinary landfill, and are surely insufficient for the national nuclear waste dump. The random locations of the test wells (a perfect geometric pattern) might be appropriate if the Rustler were a porous, homogeneous, isotropic medium (e.g. sand and gravel) in which groundwater flows predictably and uniformly downgradient. They are not appropriate in a fractured, heterogeneous, anisotropic medium with solution-enhanced groundwater pathways, such as the karstic Rustler aquifer. In karst, groundwater flows through discrete channels comprising a small fraction, typically 0.1%, of the total rock volume. Test wells, unless specifically located to intercept the groundwater channels, are likely to miss them.
There is ample evidence of karst in the Rustler at and near the WIPP site (Phillips, 1987, Chapters II, IV, V and VII) (Phillips, 1997a, all) (Phillips, 1998a, pp. 3-4, 19-20, 26) (Phillips, 1998b, all) (Snow, 1997, pp. 11-13) (Snow, 1998a, pp. 11-12). Potential groundwater pathways from the WIPP repository to Nash Draw have been identified (Phillips and Snow, 1997, pp. 8-14) (Phillips, 1998a, pp. 2-3, 7-13), and groundwater travel times as short as ten years have been calculated along these pathways (Phillips and Snow, 1998, pp. 10-20) (Snow, 1998b, pp. 8-11). None of the WQSP test wells are known to intercept these groundwater pathways, and therefore they cannot be relied upon to detect groundwater contamination (Phillips and Snow, 1998, pp. 19-20). Other locations, known to be in or near these groundwater pathways, should be monitored also, to wit: H-3, DOE-1, H-11 and H-7 in the Culebra, and H-3, WIPP-13 and WIPP-25 in the Magenta. If none of these Magenta test wells exist, then NMED should require DOE to drill them.
The proposal to monitor only one test well in the Dewey Lake Redbeds (WQSP-6a) is a token gesture. At WQSP-6a the Dewey Lake produced 12 gallons per minute of potable water, and so it is a good choice, although the Dewey Lake was more productive at other locations (Phillips, 1998b, Table 2) and should be monitored at these also.
NMED states that the Detection Monitoring Program “is necessary to demonstrate compliance” with environmental standards. It is therefore incumbent upon NMED to require that groundwater be monitored at the test wells most likely to detect contamination. Failure to do so would run the risk that a breach of containment would remain undetected until much of Nash Draw becomes contaminated.
DOE plans to emplace hazardous waste in steel drums in direct contact with salt, the most corrosive host rock imaginable. Upon closure, WIPP will be a wet repository due to steady inflow of brine. The tunnels themselves are subject to salt creep; the floors heave, the roofs collapse, the walls cave in. Already a 1500-ton slab of rock salt has fallen from a ceiling in one of the WIPP experimental rooms (Carlsbad Current-Argus, 02/05/91, pp. A-1, A-4). These rooms have since been barricaded, with no access for inspections (affidavit of Jack
Parker, 01/26/99, p. 9). The roofs in Panel 1, the area proposed for waste emplacement, have already experienced failure (direct testimony of Jack Parker at RCRA hearings, 03/05/99), and are presently supported by 13-foot bolts, wire mesh, expanded metal, channel steel, and point-anchored threaded rebar (EEG-71, 1998, p. 3). There is a 220-foot-long open fracture, up to three inches wide, in the ceiling of Room 7, and an 180-foot-long network of open fractures, up to three inches wide, in the ceiling of Room 6 (personal observation, 03/08/99) (see also EEG-71, South 1600 Access Drift) [Exhibit 51and Exhibit 52]. No one can say for certain that a roof fall will not occur in Panel 1 during the time required for waste emplacement, and so, out of concern for worker safety, NMED should order that Panel 1 not be used for waste disposal.
An original premise of WIPP was that salt would flow like a plastic, thus encapsulating the waste and isolating it from the environment. Experience has shown otherwise. Even if new waste panels are excavated, the roofs will eventually collapse. Retrieval of waste would involve crushed drums under tons of fractured salt, with contaminated brine dispersed throughout. The volume of contaminated salt might be many times greater than the volume of the original waste. Because some of the waste (RH-TRU) is too hot to handle, retrieval would have to be attempted remotely, by machine. The waste would have to be packaged and hauled to another dump site. In short, retrieval of waste would be impractical. If any miner says it can be done, it is because he expects never to be so required (Snow, 1997, pp. 1-2).
Contamination at a monitoring well would suggest that the entire groundwater pathway from the WIPP repository to the test well had become contaminated. Corrective action would require the pumping of contaminated water from a number of test wells drilled directly into the groundwater pathway and the injection of clean water into wells upgradient. Such action would be futile because the source of contamination would be continuous, the waste being irretrievable.
What makes karst hydrology so relevant to RCRA proceedings is the speed of groundwater transport. Phillips and Snow (1998, pp. 16-19) have calculated groundwater travel times as short as ten years from the WIPP repository to Laguna Grande de la Sal. DOE claims that contaminants in groundwater would be retarded. However, this conclusion is not based on sorbing tracer tests in the field. Rather, DOE has performed laboratory analysis upon a few surviving blocks of dolomite taken from an otherwise completely shattered interval of Culebra dolomite at H-3b3 (Core 31.33, Box 62) [Exhibit 53]. Clearly this is not representative of conditions in the field (Phillips, 1998b, pp. 3-4).
Under karst conditions, the conservative assumption is that there is effectively no filtration, that contaminants will travel at the speed of water. Contamination would begin to arrive at Laguna Grande de la Sal as soon as the groundwater can carry it there. Contaminants would concentrate in the lake sediments until flushed out by major flooding. There is a low, but discernible topographic divide between Laguna Grande de la Sal and the Pecos River. This topographic divide is partly breached by an irrigation canal, the elevation of which is 2960 feet. Field observations indicate that the evaporite crust of Laguna Grande de la Sal has killed all vegetation up to an elevation of 2960 feet, the same elevation as the irrigation canal. The top of the evaporite crust records the high-water level for the salt lake. Thus the irrigation canal can be a conduit for overflow discharge from Laguna Grande de la Sal to the Pecos River in times of major flooding. The irrigation canal is 0.4 miles long and reaches the Pecos River 3.25 miles east of the town of Loving, New Mexico; hence it is known informally as the Loving Canal. If this canal should carry contamination from the salt lake to the Pecos River, it is here and downriver that actual victims would be affected (Phillips, 1987, pp. 301-303) (Phillips and Snow, 1998, p. 20).