TESTIMONY OF DR. LOKESH CHATURVEDI; March 2, 1999.

Dr. Chaturvedi received a Ph.D. in Geological Science, Cornell University, 1969 and a M.S. in Civil Engineering, Purdue University, 1965. Dr. Chaturvedi is currently the Deputy Director of EEG.

STABILITY OF UNDERGROUND HAZARDOUS WASTE DISPOSAL UNITS (HWDUs): EEG's concerns about the stability of underground waste rooms are limited to Panel 1, which was excavated between 1986 and 1988. EEG has recommended abandoning Panel 1 and using a freshly excavated Panel 2 first as being the safest way to proceed. If Panel 1 is used, the Permit should require room stability evaluations to be performed in each room of Panel 1 before starting waste emplacement and an annual evaluation of the entire underground excavation. The projected safe-life of a room must be sufficient to fill the room and seal it IN THE ABSENCE OF MAINTENANCE. During waste emplacement, maintenance of the roof support systems will not be possible.

The DOE has said it would take 3 years to mine Panel 2. Each Panel contains 7 rooms. The 4 Site and Preliminary Design Validation (SPDV) rooms, which are the same size as the Panel 1 rooms, were excavated in 6 weeks. Four rooms of Panel 1 were excavated in 1 month. Therefore, 7 rooms in a new Panel could be excavated in 2 months.

If NMED does not regulate room stability, the WIPP project will be essentially self-regulated by the DOE. The WIPP Land Withdrawal Act (LWA) prescribed annual evaluations by the U.S. Bureau of Mines (USBM) in addition to quarterly inspections by the Mine Safety and Health Administration (MSHA). Congress dissolved the MSHA in 1995. After 1995 the DOE did not arrange to have annual USBM-style independent evaluations of mine stability. Although MSHA does oversee mine safety issues, their inspections focus on occupational safety issues rather than the evaluation of the structural integrity of the excavations.

Four SPDV rooms were excavated in the north experimental area of WIPP in the spring of 1983. These rooms showed rates of closure from salt creep 3 times higher than expected and showed extensive fracturing in the roof, walls, and floors. In May of 1989, after holes were drilled in the ceiling of SPDV Room 1 to install rock bolts, extensive roof fracturing was discovered throughout the SPDV Rooms. Thereafter, access was restricted to these rooms. A 700-ton pyramidal slab fell to the floor of SPDV Room 1 in February of 1991. A similar roof-fall occurred in SPDV Room 2 in June of 1994. Other roof-falls occurred in other experimental heated rooms in 1990-91. One room had been entered by personnel only 18 days before a roof-fall.

In 1991 an expert panel concluded that, without roof support, Panel 1 rooms would only last between 2 and 6 years and would need modifications and maintenance to last until 1998. By 1990 most areas in the underground had been systematically rock-bolted. In 1991 a secondary support system was installed in Room 1, Panel 1. Later secondary support systems were installed in other rooms and drifts of Panel 1. Room 7 would be the first room to receive waste. It was pattern-bolted in 1988, re-bolted in 1994, and received a third support system in May of 1998. Regular maintenance is required to replace failed bolts. Also, each bolt in Room 1, Panel 1 has to be manually de-tensioned. [The Room 1 roof has separated from the rock above it and is slowly being lowered by the "unscrewing" of the rock bolts.] With normal maintenance and without remediation, a rough estimate of a 1 to 5 year safe-life is projected for all areas in Panel 1. Salt creep is pushing the rock layers of the roof sideways against the anchored rock bolts and breaking them. The rate of bolt failure increases with time.

The DOE's plan to start mining Panel 2 within 1 year after beginning to fill Panel 1 may increase stresses on Panel 1 rooms. The DOE's goal to fill Panel 1 in 4 years is unrealistic in the initial years. [The DOE has 15 TRUPACT- IIs in service at this time. New TRUPACT-IIs will have to be manufactured in order for the DOE to meet its goal of filling Panel 1 in 4 years.] Rock bolts in a roof-fall may pierce a large number of drums. This could increase the amount of released solid hazardous and radioactive waste material. The DOE's analysis argues that a roof-fall would not result in any breached drums, but there is no evidence that WIPP drums are that robust. The DOE's analysis assumes bags of MgO backfill will cushion the drums in a roof-fall. If MgO backfill is not required in the Permit, roof-fall scenarios should be reanalyzed.

GROUNDWATER DETECTION MONITORING IN THE DEWEY LAKE REDBEDS The DOE's Permit Application and their written Comment No. 273 incorrectly states that the Dewey Lake Redbeds Formation and the Santa Rosa Formation contain little or no water near the WIPP facility and the shafts. The DOE's 1997 investigations in those areas showed better quality groundwater in the Lower Santa Rosa/Upper Dewey Lake Redbeds Formations than in any other water-bearing zone at the WIPP site. EEG recommends that in addition to the 7 monitoring wells required in the draft Permit, another monitoring well be drilled in the Lower Santa Rosa/Upper Dewey Lake Redbeds Formations near the WIPP exhaust shaft. EEG also recommends that the DOE be required to provide an accurate description of the hydrology of these formations.

The 1997 investigations were conducted to identify the source of water leakage in the exhaust shaft. [Each year, more than 14,000 gallons of lead-contaminated water collected at the bottom of the exhaust shaft are shipped off- site as hazardous waste.] A water-saturated horizon was found in the Lower Santa Rosa/Upper Dewey Lake Redbeds Formations at the same depth range where water has been found leaking into the exhaust shaft (50 to 80 feet below the ground surface). In 1985, moisture was observed during inspections of the shaft liner. When the exhaust fans were on, the moisture evaporated before reaching the underground facility. A few years after 1988, increasing moisture was adversely affecting air-sampling probes in the shaft. Video inspection of the shaft in 1995 showed a stream of water leaking into the shaft [inflow]. The DOE has stated that most of the exhaust shaft water is the result of precipitation or condensation from the ventilation system. However, the DOE's 1996 investigation showed that this is unlikely. For the evaluation period, precipitation of water in the shaft was consistently negative (meaning the dry air evaporates water from the shaft rather than precipitating it).

Dr. Chaturvedi presented footage of a 1998 videotaped DOE inspection of the exhaust shaft. Inflow was at the same depth as the 1995 video inspection. This suggests that despite evaporation, inflow has continued for at least 4 to 5 years. The DOE objected to their videotape being placed into evidence, claiming that only a Registered Professional Engineer could accurately interpret the videotape. Dr. Chaturvedi was allowed to show the videotape as a demonstration and was not allowed to narrate it. The videotape showed water flowing down the shaft wall and droplets of water being pushed upward by the exhaust airflow. The DOE claimed the videotape had been recorded under conditions of reduced ventilation.

HIGHLIGHTS OF CROSS-EXAMINATION OF DR. LOKESH CHATURVEDI:

Dr. Chaturvedi has great respect for the Ground Control staff at WIPP and believes they are competent. However, geotechnical mechanical information may or may not provide warning of instability. The DOE queried Dr. Chaturvedi on some of the statements and conclusions of the members of the 1991 expert panel examining the roof-fall. One of the 11 experts said that a 20-year life was feasible for Panel 1 with good support and repair. EEG-63, "Stability Evaluation of the Panel 1 Rooms and the E140 Drift at WIPP," stated that the Panel 1 rooms could be safely kept open without maintenance for 1 to 2 years after installation of a new roof support system. But if maintenance had to be performed during this period, it could not be done if there was waste in the rooms. Also, the newest support system in Room 7 is 10 months old already, and there may not be sufficient installation space in that room. Rooms are filled from back to front, and filled rooms are not as easy to inspect (even visually) as unfilled rooms. Roof bolts help support the roof beam but will not stop the convergence. Salt creep continues to push the roof sideways from the pillars on each side of the rooms. Room life could be extended indefinitely, but this would be costly and would require ongoing maintenance. Prohibiting use of Panel 1 in the Permit is the best thing to do. Even the DOE anticipates only using part of Panel 1.

The DOE stated that the SPDV Room 1 roof was unsupported and was allowed to fall. Dr. Chaturvedi said that no experiments were designed to test a roof-fall. The DOE stated that all Panel 1 rooms had at least 3 remotely read extensometers, that Room 1 is equipped with load cells, and that there would be some period of warning from visual inspection and remote instruments before a roof-fall. But personnel entered one experimental room 18 days before a roof-fall occurred in 1990. None of the WIPP Mining Safety and Health Administration (MSHA) reports Dr. Chaturvedi has seen have ever addressed mine structural integrity. There are no regulators performing United States Bureau of Mines (USBM) inspections at WIPP. One safety option would be to remove about 7 feet of unstable rock above each room. However, this would change accident scenarios because-- although unlikely--if rock did fall, it would fall 7 feet more than in the original scenarios. Rooms might also be better designed if they were narrower, but then equipment and 7-packs might not fit. Newly excavated rooms (33 feet wide) could safely last 5 years.

Shafts at the Santa Rosa/Dewey Lake Redbeds horizon were dry and unlined when first mapped by Powers and Holt in the mid-1980s. Water would have shown up during mapping only if sufficient time had elapsed between excavation and mapping. Measurable seepage of concern was discovered in 1995 when it affected the monitoring probe in the exhaust shaft. It is unknown how long water had been seeping into the shaft because the shaft was not inspected before 1995.

The Culebra is not the first regionally extensive water- bearing layer above the repository. The first is the Rustler/Salado interface. In some boreholes, permeability is higher for the Rustler/Salado than the Culebra. Also, the Dewey Lake Redbeds Formation has much better water quality than the Culebra.

The videotape probe was lowered from a spot more than 30 feet above the ground surface, and that amount has to be subtracted from the "50 to100" feet below ground surface described as the area of water seepage in the shaft. The 1997 reports agreed that the depth of water found by drilling in the Santa Rosa/Dewey Lake Redbeds Formations was the same as the exhaust shaft flow. The exhaust shaft sump once accumulated almost 2,000 gallons of water in about a week (though the amount varies).

FLOW PATHWAYS IN THE SALADO: In 1995 water had moved all the way down the exhaust shaft below the repository floor to Marker Bed 139 (MB139)--a permeable zone under the repository--and had moved along this marker bed for several hundred feet. These interbeds and marker beds are continuous layers that run across the entire Delaware Basin for tens of miles--sometimes for hundreds of miles. Salt will not completely seal off the waste. There are pathways for fluid movement. The Delaware Basin has about 43 marker beds and other thinner beds of anhydrite, clay, and other rock. [Anhydrite is a white to grayish or reddish mineral of anhydrous calcium sulfate, CaSO4.] Salt is plastic, but- -because of the layers--the salt formation is divided.

The hydraulic conductivity of each of the marker beds and interbeds is different from the halite. [Halite is rock salt.] The conductivity of MB139 is about a million times higher than the salt. MB139 is an undulating layer up to 5 feet thick 3 to 5 feet below the floor of the repository and is modeled as a zone through which water and gas may leak laterally from the repository. Flow pathways from the repository are not controlled by salt but instead by salt mixed with clay, clay layers, and fractured anhydrite layers. MB138 is about 20 feet above the roofs of the waste rooms. MB138 and MB139 are fractured anhydrite and clay. Anhydrite is more conductive than salt, and fracturing makes it more conductive still. Fractures can remain open in anhydrite, and there can be openings where the anhydrite interfaces with the salt.

EEG has said inflow in the shaft from the Dewey Lake Redbeds Formation can be traced to recharge. The DOE has said that the recharge comes from ponds created on the surface for fire control. More information is needed on this subject, but if the Dewey Lake Redbeds Formation is saturated for more than 80 acres, as indicated in the 1997 reports, recent recharge from the ponds is probably not the source. The Santa Rosa Formation is extensive east of the site but pinches out at the center of the site, to the west. Flow in the Santa Rosa and Dewey Lake Redbeds Formations is generally south, but specific data on its course is limited. Only part of the WIPP site is characterized with respect to this water-bearing zone, and there is not enough data available to accurately model the Santa Rosa and Dewey Lake Redbeds Formations as a potential pathway. Because of the lack of data, it cannot be unequivocally stated that the Culebra is the most transmissive unit above the site.

Definition of the water table at WIPP has long been a major question. Leonard Konikow of the United States Geological Survey (USGS) has said that if you do not know the water table, you do not know the hydrology of the site. Based on the limited available data, the DOE defined the water table as being in the Dewey Lake Redbeds Formation. However, there is no data in the Permit Application on this formation. The Magenta and Culebra dolomites are saturated in every test well on the WIPP site, and all members of the Rustler are water bearing at some locations. [Dolomites are deposits of a light-tinted mineral (CaMg(CO3)2), either gray, white, or pink.]

The surface of the WIPP site is covered by sand, and there are no well-developed surface drainage patterns at WIPP. The sands are transmissive to rainwater. Directly beneath the sands is the MESCALERO CALICHE FORMATION, which is generally continuous but does have gaps. It is possible for rainwater to penetrate the caliche through these gaps to reach the lower strata. Dr. Chaturvedi did not know whether the Magenta and the Culebra are connected in nearby wells. In areas where there are wells in the Culebra but not in the Magenta, it is uncertain whether the two are connected. USGS and Sandia National Laboratory investigations state that recharge for the Rustler and Dewey Lake Redbeds is north, northwest of WIPP. WIPP-33 (a well) is drilled in a sinkhole. WIPP-14 (also a well) is also drilled in a closed topographic depression, but Dr. Chaturvedi has not seen evidence of it being a sinkhole.

As the floor of the repository buckles and shifts from salt creep, small fractures could be created over time, but these fractures probably would not go down 700 to 800 feet vertically to possible brine pockets at the Salado/Castile interface. The most likely path for releases of hazardous materials from the repository is through fractured interbeds. If a roof-fall connects the waste rooms to Anhydrite Layers A or B (7 to 13 feet above the repository), the anhydrite layers could become pathways. Then a borehole could allow contamination to travel higher. The marker beds connect the repository with the shafts and all the other boreholes in the basin that cross them.

Permeability increases in any disturbed area but decreases very quickly as you move away from an excavation. The only way to increase the permeability away from the excavations is with water or gas. Although the Culebra is probably the most prolific water-bearing zone, the Magenta, Dewey Lake Redbeds, and the Rustler/Salado contact are other potential pathways.

Dr. Chaturvedi does not believe that any waste has ever been brought to WIPP. Bill Bartlett, also of EEG, checked out rumors of nuclear waste in the repository with radiation detection equipment and found nothing.