Dr. David Snow was the first rebuttal witness for
Citizens for Alternatives to Radioactive Dumping (CARD)

Dr. Snow has degrees in geology and engineering science. Dr. Snow's thesis was on Flow Through Fractured Media. Dr. Snow has worked for 45 years as a consultant, mostly in the mining industry, on hydrology, geology, and geotechnical engineering. He has also consulted on nuclear waste disposal, including 14 months at WIPP where he reviewed WIPP's hydrology. In 1986 Dr. Snow reviewed the No-Migration Petition for the WIPP site. Dr. Snow did not think WIPP was feasible because when caverns in plastic salt close, it causes new stresses on the waste as it evolves.


Dr. Snow's work at WIPP emphasized transport in the Rustler, especially in the Culebra Dolomite. Dr. Snow was critical of the tracer test design for the Culebra. Dr. Snow believed the tracer test needed a better database of the physical properties of the Culebra in order to set up the test in a better way. The Culebra has gypsum filled fractures throughout which are leached out from east to west. Dr. Snow believes there is a lack of understanding of the evolution of the Culebra. If glassy gypsum (selenite) is found, this is proof that it has evolved by dissolution and the flowing of groundwater. The Culebra in the outcrop along the Pecos River shows evidence of dissolution of the dolomite and a complicated history of interfilling of vugs. The primary permeability has been altered so now it has a secondary permeability. The partially open fractures in the Culebra are the main reason for the permeability in that unit.

The assumption is that most brine is in the salt and anhydrite, but most transport will be in the anhydrite beds. The anhydrite beds have measured properties about one million times that of the salt. Salt has low permeability in the laboratory, but in the field, salt has six orders of magnitude more permeability. The Salado, as a whole, has more horizontal permeability than vertical permeability. Salt does not have perfectly plastic flow. Salt can also be brittle. If the stress is at a higher rate, the salt will be brittle and will fracture as it has done in some of the ceilings in the WIPP site.

Coring at WIPP has suffered because the DOE uses a single tube core barrel (commonly used for potash mining) and this seldom gives good results with fractured, weathered rock or discontinuous features. This example demonstrates DOE's lack of understanding of core. There is little intact core in the WIPP Core Library. Most of the core is rubble so there is a lot of debate about the true origin of the claylike features recovered. The Rustler characterization problem is compounded by the exploration difficulty. It is easy to miss the largest flow path, but it is possible to see residues left behind when the path moves elsewhere. There is a thin interval of clay in the core. If the highest core technology is not used, the clay will be ground up.

Dr. Powers does not believe the salt beds were present and then dissolved from the Rustler. Most of the scientific community disagrees with Dr. Powers because they know the WIPP site is located in a karst region. Dr. Powers is in error. However, dissenters who believe there is karst at the WIPP site have left the DOE and Sandia. Dr. Powers is the principle author of most of the data that relates to karst at the WIPP site.

The logging of the Exhaust Shaft lithology of the Rustler rings true to Dr. Powers' interpretation that it was formed in alluvial channels on the surface with cross bedding, laminations, pebbles, etc., which are typical of streams. But there are similar features at 3,000 feet below ground surface (bgs) that were formed by open flow in large evaporite channels. One thing unique to Dr. Powers' logs of the clay members is the presence of "slickensides." This is especially true, since Dr. Powers' logs are all horizontal.

The permeability of the Salado is very low, about 10 to 21 meters/second. The anhydrites are about 3 orders of magnitude more permeable. But what is important is how fractures propagate above a mined repository. The whole WIPP repository is not mined now.

Dr. Snow showed a series of slides and viewgraphs of the Kominko Mine near Saskatoon and the K-2 mine at Esterhazy. The slides and viewgraphs showed that after a few years there was sagging of the roof and fracturing at a 45 degree angle above the roof with breaches up to the horizontal clay partings. The typical failure at WIPP and elsewhere in salt mines will be a sudden collapse of the roof. Dr. Snow was once in a mine where a roof fall occurred about a mile away from where he was and he was almost bowled over by the blast of air through the drifts.

None of the packaging at WIPP can absorb the stresses of roof slabs that weigh hundreds of tons. Drums will be crushed, will leak, and their contents will spill out. The first line of defense--engineered barriers--will not be very effective. Fractures through the roof and floor can carry free liquids away from the repository. After the waste is crushed and mixed with brine and salt, retrieval is only tongue in cheek.

The slides showed that subsidence fractures reached well above the opening. WIPP and all mines track subsidence rates at the surface. Typical rates are about 3 to 6 inches/year. As the rooms close the rate will be about 12 inches per year below. Closure will be more and more rapid until the salt becomes brittle and the roof collapses. It is unclear what will happen above Marker Bed 138 or what the extent of the fractures above this marker bed will be. The best evidence is to go to an analog situation. It would be nice if the DOE could have explored some of the old potash mines in the WIPP area. But when the DOE had the opportunity to do this, the DOE decided against it.

In the K-2 mine slide, it is possible to see the roof deflecting, the beam bends, and there is shear of the lower member above the upper member. This slide demonstrates what is shearing the WIPP rock bolts as well. Nothing can be done about that.

At Esterhazy the roof fractures went through 100 feet of overburden, breaching the integrity of the rock salt and clay seams. There were inclined fractures. The sagging roof beams relieved the vertical stress, but the horizontal stress stayed the same. Salt stress is usually lithostatic and the failure that happens is a thrust fault (fractures in one block move up and the other block moves down). The fractures reached up to the top of the salt, crossed 18 feet of mudstone (the First Redbeds) and breached the impermeable Redbeds to the dolomite aquifer. Leakage occurred along the fractures. Because of the tremendous gradient of 3,000 feet of water, eroded solution channels in the roof provided the pathway and the water drained into the mine. Sometimes water would run along a clay seam until it could force its way through the clay. Then the water carved its way through the salt into the mine. First, there was a solution channel flat along the seam and then the water dug a gorge into the panel. The subsidence is not just over one room, but over the rooms acting together as a panel. There was a steep fault from the roof into the overlying sediments.

It is impossible to know exactly what will happen over the WIPP repository, but the situation is sufficiently analogous to the examples shown here. Every anhydrite bed is bordered by clay. This thin seam of clay is where the roof breaks away and it will continue to break away at the clay seams. The most brine that has been estimated to inflow is about 800 liters/meter of room, which is about 2% saturation. This is fairly insignificant and the brine is carried away by evaporation. But the Repository has about a 1% dip to the east. The floor will be fractured to the marker bed below and brine will accumulate in the down dip region while the upper portion will be drained. Some rooms will be flooded before others.

Fracturing above the roof could go up at least 300 meters so that the breach would occur in the salt and in all the anhydrite beds. Dr. Snow could not say if the fracturing would go all the way to the Rustler, but stated that the fracturing will multiply the amount of brine inflow by a factor of 20. The fractures would also provide passages of egress for air, evolved gases, and fluids as the mine opening loses pore space due to creep. The fractures can let fluids escape from waterflooding, drilling, or the tapping of Castile brines. Finally, the subsidence fractures in the roof can facilitate the escape of fluids from the repository and it is possible the fractures could go all the way to the Rustler.


The Lowenstein report concluded that dissolution does not have to remove all the primary sedimentary features. There will still be traces of salt, clay, and other insoluble residues.

The first anhydrite layer is actually Anhydrite B (7 feet above the repository), which is 6 feet above Anhydrite A. Above those Marker Bed 138. The clays associated with the boundary with any of these anhydrites are the most important. There are thin clay seams associated with the anhydrite/salt interfaces of most of the anhydrites. There are also thin clay seams without anhydrite. Shear strength on clay partings is less than that of the salt. When friction on the partings is destroyed, the whole system sags because there is no shear strength on the partings. Failure implicit in one beam is implicit in all of them. If there are 43 marker beds up through a thousand feet of salt, failure can propagate all the way up.

Slickensides are features on a dislocation surface that reveal the relative movement of the rocks.

The speed of erosion from the water in the examples that Dr. Snow gave could be spectacular. The biggest leak was up to 8,000 gallons/minute and could dissolve the salt in days or weeks. It could excavate an area 4' x 100' x 1000' in that time. The brines entering the mine through the cover rocks was already very saturated. A little change in temperature (which increases as you go deeper) controls the dissolution rate. Also velocity controls the rate. Even if the brine is saturated, it is possible to have dissolution of the salt if the velocity is high enough.

Dr. Snow is aware that the RCRA timeframe is 80 years.

The K-2 was at a horizon that was below 90 to 110 feet of salt. Over this horizon was mudstone--the First Redbeds--which were about 30 feet thick. Over that layer was the brine saturated Dawson Bay Limestone, which was about 250-300 feet thick. The K-2 mine was about 3100 feet below ground surface (bgs). There is 1,300 feet of salt between WIPP and the Rustler and no 30-foot thick mudstone layer above the repository. There is also no 250-300 foot thick limestone layer and WIPP is at 2,150 feet bgs. In a gross sense, there is a difference in the stratigraphy between K-2 and WIPP. Also, all 10 panels of the whole WIPP repository are about the size of a single panel at K-2.

The DOE claims that only 40% of the total subsidence will occur during the 80 years of the RCRA timeframe. Dr. Snow stated that that figure is not necessarily shown in the DOE documents and that the figure would have to be computed.

There are about 40 marker beds above WIPP. WIPP hydrology needs better delineation even with all the wells that have already been drilled. The Culebra fractures are also vertical. The fractures will probably not be found so the frequency, orientation, apertures etc., of the fractures in the Culebra are not known well enough to do transport modeling. The willingness of the DOE to find what is necessary to do the characterization in the zone between the repository and the surface remains incomplete.

Subsidence fractures may transect large parts or all of the Salado. Shafts, boreholes, etc. will be breached by fractures. The sealing of the shafts and boreholes is immaterial if fractures compromise them. Brine in the down dip rooms compromises the panel seals even though it is not claimed that the panel seals are barriers to movement. The technology of borehole sealing is debatable and has been researched for many years. An experiment to verify the ability to seal boreholes is impossible. Salt causes corrosion of the sealing materials. There is an effect of shear on the boundary with stiff and compliant sediments. A small leak along the contact will soon become a solution passage.

In Esterhazy, the German consultant said that every potash mine ultimately floods. Ralph Crosser in Carlsbad also said they all leak. Clay seams and anhydrite beds leak water. Some leaks are small, but some continue over a long period of time and cause a progressive inundation of the workings.

Subsidence from roof-falls would not go to the surface from a single room, but perhaps subsidence could from an entire panel or the whole repository.

The groundwater in Mexico probably would not be contaminated by WIPP, but it is the quality of water in the Rio Grande that is at risk.

Back to Summaries