Presentation by Concerned Citizens for Nuclear Safety (CCNS) and
Southwest Research and Information Center (SRIC):
TECHNICAL TESTIMONY OF DR. JOHN BREDEHOEFT; March 16, 1999.
Dr. Bredehoeft received a Ph.D. in Geology, University of Illinois, 1962; a
M.S. in Geology, University of Illinois, 1957; and a B.S.E. in Geological
Engineering, Princeton University, 1955. Since 1995, Dr. Bredehoeft has
worked for a consulting firm, the Hydrodynamics Group. Previously, Dr.
Bredehoeft worked for 32 years with the U.S. Geological Survey. He has
worked on a wide variety of ground water problems, including contaminant
transport. Dr. Bredehoeft has published more than 100 scientific papers,
has received numerous scientific awards. For 10 years, Dr. Bredehoeft was
a member of the National Academy of Sciences/National Research Council
Committee on WIPP.
THE HARTMAN SCENARIO
The Hartman Scenario is described as follows: A dry hole was drilled and
completed in the Bates lease (in southeastern New Mexico) in 1954. Texaco
began a waterflooding operation in 1964 near the Bates well.
[Waterflooding is a common process used in oil production in which liquid
is injected into rocks to increase the amount of oil that can be extracted
and brought up to the surface.] Later, Texaco enlarged its operations. In
its waterflooding, Texaco injected liquids at very high pressures. These
high pressures can cause hydraulic fractures underground which allow fluid
to move very rapidly.
In the early 1990s, Mr. Hartman drilled a well a 100 feet away from the
Bates dry hole. When he reached the Salado Formation at 2,200 feet, the
well blew out and water flowed out of the well for 4 days (35,000 barrels
of water). The most likely explanation, and the explanation accepted by
the court in the lawsuit in which Hartman was awarded damages, was that
Texaco's waterflooding operation had fractured an anhydrite marker bed in
the lower Salado and water had moved from the Texaco well over two miles to
the Hartman well.
HYDROFRACTING is a technology that increases the well bore permeability.
If the pressure is high enough, the rock pressure is overcome and the rock
is cracked and a fluid filled fracture is opened up. The overburden can
actually be lifted up and pumping in more fluid can extend the crack. [The
overburden is the rock layers above the fractured rock.] It is not
necessary to have very large cracks to increase the permeability
dramatically (tenths of an inch or less). Hydraulic fracturing is a common
way to increase permeability. It increases the production capacity of a
well and hydraulic fracturing is done thousands of times every year.
Hydrofracting can also be used to measure the stress state of the earth and
to give information about the state of stress along earthquake faults. In
Colorado, Dr. Bredehoeft and others started earthquakes by injecting fluid.
They also stopped earthquakes stopping the injection.
INJECTION WELLS AT WIPP
Water injection wells for secondary recovery of oil or for brine disposal
are very common in the oil business. Drilling is allowed right up to the
Land Withdrawal Area boundary. [WIPP is a square piece of land consisting
of 10,400 acres, or 16 square miles (4 miles on each side).] The WIPP
boundary is two miles from the waste panels (about the distance the Hartman
water traveled). At Hartman, the surface pressure measurement was1,000 psi
at the land surface and the water flowed for 4 days. The permeability is
more than 5 times the permeability of the marker bed at WIPP. If you had a
hydrofract at WIPP similar to the Hartman Scenario, it is reasonable to
extend the hydrofract for several miles. There will be a lot of
waterflooding and injection around the site boundary in the future. This
could move a lot of water into the repository. The 4 days of Hartman flow
would inundate 10% of WIPP. There is already some brine injection near
In Hartman, Dr. Bredehoeft used a 2-dimensional model because flow away
from the Texaco well was radial in nature and a 1-dimensional model would
not capture the flow field. If the permeability is equal, the water will
go everywhere in all directions, but if the permeability varies, it will
flow in preferential flow patterns. The marker beds are about 1 meter
thick with less permeable salt on either side. Almost all of the water
(99.9%) goes into the marker bed with the rest going into the salt. When
the water encounters the WIPP repository, which is at a lower pressure, the
repository becomes a fluid sink. If the DOE is still mining the repository
when this occurs, the pressure underground will be the same as atmospheric.
Waterflooding of the repository could occur during the operational period.
The injection rates in southeastern New Mexico are increasing. If, at the
upper end, the water that entered the repository would overwhelm the
backfill and flood the repository. This water could break through
anywhere--even in an active room. Waterflooding of mines is a common
If borehole plugs or shaft seals do not meet permeability standards, water
could move contaminants up drillholes and shafts. In the long term brine
could corrode the steel drums. This corrosion of the steel drums creates
hydrogen gas. Hydrogen gas is highly flammable. The hydrogen gas would
pressurize the repository and make it easier to transport contaminants to
the accessible environment. The pressure would fracture the marker beds
near the repository and gas and fluid could move out into the marker beds.
The Environmental Protection Agency (EPA) discounted the Hartman Scenario
in the Compliance Certification process. The EPA never agreed that the
Hartman Scenario was a hydrofract and said there was a low probability of
this occurring at WIPP. Dr. Bredehoeft believes the weight of evidence
shows that the Hartman Scenario was a hydraulic fracture. The other
consultants, especially those in the lawsuit, and even the DOE's Dennis
Powers, all agreed that the Hartman Scenario was a hydrofract. Natural
brine pockets in the Castile Formation are only pressurized at 8/10 of
lithostatic pressure, but the surface pressure of 1,000 psi at Hartman is
approximately equivalent to lithostatic. These very high pressures are not
the same as pressures in a naturally occurring brine pockets.
A blowout in Hartman's well in the lower Salado was associated with
anhydrite marker beds. Fluids moved a few miles. A similar situation
could occur at WIPP and could move very substantial amounts of fluid in a
very small timeframe in terms of the amount of drilling around the site
boundary. Such a situation could occur during WIPP's operation thereby
endangering workers, or during the closure period after waste operations
HIGHLIGHTS OF CROSS EXAMINATION OF DR. BREDEHOEFT
EPA said Dr. Bredehoeft's model was unrealistic and they could not
replicate it. Also, the DOE claims that for the Hartman Scenario to occur
at WIPP there would have to be a hole in the casing at Marker Bed 139 to
allow water to get into this marker bed. But oil and gas wells surrounding
the WIPP site are finished much lower than the WIPP horizon [Marker Bed
NEFF Federal Well #3 is a water-injection well near WIPP. The DOE argued
that NEFF injects at 7,000 feet, so water would have to come up 5,000 feet
or there would have to be a hole near the WIPP horizon in order for water
injection to affect the WIPP repository. However, the New Mexico oil and
gas regulators found that NEFF #3 injection activities would affect potash
reserves, which are at 1,300 feet. The EPA also said there was a low
probability of this event happening. Dr. Bredehoeft did not calculate the
probability of this event happening, only what the effects would be if it
occurred. Wells near WIPP are supposedly finished to higher standards than
the wells in the older well fields. However, waterflooding still occurs
frequently in southeastern New Mexico. It would be prudent for NMED to be
notified when there will be waterflooding or brine injection near WIPP.
There are significant geological differences between the area where
Hartman's well is and WIPP. However, the Salado Formation and its marker
beds are located both at WIPP and at Hartman's well. The Salado Formation
is where the blowout occurred.
There are disagreements between Dr. Bredehoeft and the EPA on how far a
hydrofract would go. The EPA uses BRAGFLO to calculate the extent of the
fractures. [BRAGFLO is a numerical computer model, developed by Sandia
Labs for DOE, used to predict pressures in the repository. BRAGFLO has
only been used for WIPP modeling; it has never been used by industry.] Dr.
Bredehoeft compared BRAGFLO with the Linear Elastic Fracture Model (LEFM)
and showed that BRAGFLO underestimated fracture magnitude by a factor of 3
to 5. LEFM is an accepted model used by the oil industry.
Blowouts still occur in southeastern New Mexico despite regulations and
monitoring. In particular, many leaks are associated with water flooding.
Oil operators in one field put together a committee to deal with blowouts
because it is such a widespread problem. The leaks usually are of large
quantities of fluid.
If injection is made at high pressure and the fluid encounters a marker
bed, a crack can be created, which increases the permeability of the marker
bed by 5 orders of magnitude. If pumping continues, the fracture may
extend almost indefinitely.
Naturally occurring brine reservoirs exist in the Salado Formation, but
they are no more than a 1,000 gallons. These naturally occurring brine
reservoirs could not have accounted for all the fluid that blew out the
A study undertaken 7 to 8 years ago modeled the rise of water levels that
were occurring in the Culebra south of WIPP. This study concluded that the
probable cause of the rise of water levels was a nearby water injection
Marker bed 139 is connected to the WIPP repository by means of the
Disturbed Rock Zone (DRZ).
Even the DOE's Performance Assessment (PA) indicated that borehole plugs
would degrade fully within 200 years. At that time, what is left of the
borehole plugs will have the permeability of silt.
There are several formations below the Salado that produce oil and gas near
the WIPP site. Since 1990, 140 oil and gas wells have been drilled within
two miles of the site boundary and 47 more are planned for that area.
Twenty-seven of the wells are within 2 miles of the waste panels, which is
about the distance the brine traveled from the Texaco well to the Hartman
NEFF Federal #3 is about 1.5 miles from the site boundary and it is not the
only injection well near the site. There are 4 brine injection wells
within 4 miles of the boundary, and one is operating above lithostatic
pressure. NEFF Federal #3 is operating above lithostatic pressure at about
500 feet below the repository.
It is estimated that gas generation from the drums will take several
hundred years or more before it could create fractures and push brine and
contamination out into the marker beds. But if large quantities of fluid
entered the repository very quickly, increased gas generation could occur
narrowing the timeframe for contamination of the marker beds.
There is only limited information on the permeability structure of the
Culebra. One way to fill in the information blanks is to use the Monte
Carlo Method. [The Monte Carlo Method is used in modeling to allow random
chance to occur.] The Monte Carlo Method does not work well for one cell,
but works well if there are many cells. The Monte Carlo Method is a
commonly used procedure, but needs confidence limits put around the
The casing material in the boreholes is steel and can be corroded by the
salt. This may open up pathways through which fluid can escape. If a leak
occurs between the outer casing and the formation at a pressure about 1,410
psi on the land surface, the leak could create a hydrofract in the range of
300 to 400 feet below the repository. The leak could cause hydrofracturing
in any marker bed above that level. The leak does not have to be located
exactly at Marker Bed 139 in order to reach that marker bed.