9 Appendix II Underground Detector Construction at Homestake
Plans for the construction of a multiple module megaton Cherenkov
detector at the Homestake Mine have gone through a number of essential
evaluation and design stages consisting of rock strength and stability
evaluation, chamber design and layout, construction planning and
sequencing and development of budget and timetable. Here is a summary
of these steps.
9.1 Determination of Excavation Stability
The Rock Stability Group at the Spokane Research Laboratory of NIOSH
(National Institute of Occupational Safety and Health) carried out an
evaluation of the stability of large excavations as a function of
depth in the Yates rock formation in the Homestake Mine. This
involved a three-dimensional finite difference evaluation using the
FLAC3D program. These results were compared with the empirical
prediction charts of Barton and Grimstad and Barton. The conclusions
were that 50 meter diameter by 50 meter high chambers could be safely
excavated and would be stable for long term occupancy at 2150 meter
depth and probably somewhat deeper.
The Yates rock quality was determined by direct measurement of samples
taken from the accessible edges of this formation. Before excavation
begins, it is essential that core samples from various internal
sections of the proposed rock formation are measured and the
excavation reevaluated.
9.2 Construction of Multiple 100 kiloton Modules in the Homestake Mine
Using the results of the stability evaluation a group of ex-Homestake
mining engineers, (Mark Laurenti - former Chief Mine Engineer, Mike
Stahl - former Mine Production Engineer and John Marks - former Chief
Ventilation Engineer) designed an array of ten 100 kiloton water
Cherenkov chambers. The chambers are located along the circumference
of a 250 meter radius circle that is centered on the Winze 6 shaft.
The top of each chamber is connected to the 6950 ft station of the
shaft via a horizontal, radial tunnel. A similar tunnel connects the
bottom of each chamber to the 7100 ft shaft station. Fresh air will
be sent to each chamber via the top tunnel and exhaust air removed via
the bottom tunnel, thus providing independent air supplies to each
chamber.
During chamber construction, waste rock will be removed via the bottom
tunnel. This will prevent rock dust from one chamber contaminating
the fresh air supply of another chamber. Once construction is
completed, the bottom chamber to tunnel connection will be sealed. A
spiral ramp that surrounds each chamber and is used for access during
construction will then complete the ventilation loop between top and
bottom tunnels.
Each chamber will have a concrete liner. The inner surface of the
liner provides a smooth surface for the water tight plastic liner that
will separate the Cherenkov counter fill from the chamber walls. The
liner also provides additional mechanical stability for the
excavation. If necessary, drainage can be provided between the
concrete liner and the surrounding rock.
9.3 Construction Timetable and Cost
Marc Laurenti has worked out a detailed timetable and budget for the construction of these modules including initial rock evaluation coring, construction of both top and bottom access tunnels, removal of waste rock, maintenance of mining equipment, etc.
The excavation process consists of continuous repetition of three
separate tasks, (1) drilling and blasting of rock, (2) removal of the
rock rubble, and (3) installation of rock and cable bolts to stabilize
the freshly exposed rock walls. Each excavation cycle is about 10
weeks with 3 weeks for each of the above three steps. There is a
considerable cost savings in excavating three chambers at the same
time, with a three week phase shift between steps in each module.
This arrangement permits each of the three specialized crews to move
from one excavation to the next every three weeks or so and continue
using the same equipment and carry out their specialized tasks. In
contrast, using one crew to sequentially do three different tasks will
result in idle equipment for 2/3 of the time and inefficiency as they
switch from one task to another.
For the three module mode, the cost of excavating each chamber is
$14.7M. This includes $3.25M for the concrete liner and a 15%
contingency. In contrast, the cost of excavating a single module is
$16.9M including liner and contingency. The total required equipment
cost is the same for both of these construction modes.
Assuming three shifts/day and 5 days/week operation, it will take 208
weeks or 4 years to excavate each 3 module group. This time could be
reduced by going to a 6 or 7 day week. The Homestake Company
frequently operated on 6 or 7 day per week basis.
9.4 Rock Removal
Each 100 kiloton module (105 m3) involves the removal of about 416,000
tons of rock including access tunnels, domed roof, etc. For three
chambers this results in 1,248,000 tons of rock in 4 years or 312,000
tons of rock per year. Since the hoisting capacity of the Winze 6 -
Ross shaft system is 750,000 tons per year, the simultaneous
construction of three modules utilizes only 40% of the capacity of
this shaft system.
9.5 Equipment Cost
Since all mining equipment has now been removed, new mining equipment
will have to be purchased or leased. The required equipment, one Face
Drill, two LHD loaders, 2 Bolters, 2 Underground Support Vehicles, 2
Lift Trucks, 1 LH Drill and 2 ITH Drills, costs about $4.2M. It may
be possible to arrange for leases instead of purchasing these items.
Normal equipment maintenance has been included in the construction
cost. It is unclear whether the cost of this equipment should be
assigned to this specific task or should be part of the general
facility budget.
9.6 Choice of Depth and Depth Dependent Cost
There has been considerable discussion of depth necessary for very
large detectors and the costs associated with deep detector locations.
It is clear that the deeper the detector, the lower the cosmic ray
muon and associated particle background. It is always preferable to
have lower background. We can quantify the background limit by
specifying that there be less than one cosmic ray related event per
year within the megaton detector during the time that the accelerator
neutrino beam is on. If we assume the accelerator beam is on for one
microsecond per second, this requirement specifies an upper limit of
1.6 x 10-6 mu/m2/sec, essentially the cosmic ray flux at about
7000 ft depth. The effect of this specification is that every event
observed in the detector during the beam-on time is due to a neutrino
from the accelerator without any cuts whatsoever.
The question then is one of access and rock strength, namely, does a
specific facility have ready access to a deep location and is the
local rock structure capable of supporting large chambers. For
Homestake the answer to both of these questions is YES. The present
mine extends to 8000 ft, about 1000 ft deeper than the proposed
detector location, and the rock seems strong enough to readily permit
the excavation of large chambers.
In the appendix we provide a comparison of costs of building the
megaton Cherenkov detector at 6950 ft depth vs at the 4850 ft depth.
As indicated there, the maximum additional cost for putting the
megaton Cherenkov array at 6950 ft versus at 4850 ft is 5-6% of
excavation cost or less than 2% of total detector cost.
9.7 What Lessons About Depth Can be Learned from Previous Experience?
Detectors are located underground to reduce background in the detector
due to cosmic rays. The deeper the detector, the lower the cosmic ray
background. We have yet to have a detector that claimed to be "too
deep". The only issues are: (1) is there a substantial additional
cost associated with depth, and (2) are there technical limits
associated with rock strength, etc. that limit depth at a given
location? For many existing laboratories, depth is specified by what
is available at that facility. Only two locations, the Sudbury mine
(SNO), and the Homestake Mine (chlorine), have multiple levels available.
SNO chose to be at 6800 ft, essentially the same as the proposed
megaton detector. Since chlorine was the first underground neutrino
detector, there were no precedents and so it might be instructive to
review the sequence of events that led to its location.
In 1962, Ray Davis tested a small perchloroethylene detector in a
limestone mine in Barberton, Ohio at a depth of 2200 ft. The 37Ar
production was completely dominated by cosmic rays. That started a
search for a much deeper site. There were two possibilities in the
U.S., with Homestake the preferable one. At that time, in 1965, 4850
ft was the deepest level that the Company would agree to. At the time
the prediction for the solar neutrino signal was larger than now,
there was no thought about signal depression because of neutrino
flavor conversion and no one expected a final measurement with a 5%
statistical precision. By the early to mid-1970's it was already
clear to us that the cosmic ray induced background was too large,
given the observed signal, and that we needed a larger and deeper
detector. Unfortunately, at the time, the Company was not willing to
consider a deeper and larger detector.
The final result was that the cosmic ray induced signal is 10% of the
solar neutrino signal in the chlorine detector and the uncertainty in
that signal is the largest contributor to the systematic uncertainty.
The lesson is clear - locate detectors as deep as possible and be sure
that there is a roadmap to detector enlargement.
A detailed construction plan for the construction of three 100 kiloton
modules in four years at the 7000 ft depth in the Homestake Mine has
been developed. The total construction cost of these three modules is
about $44 M or $11M/year. In addition, there must be a one time
purchase of about $4.2 M worth of mechanized mining equipment. The
lead time in delivery of the mining equipment can be used to carry out
coring of the rock region in which the detector array is to be
constructed.
9.8 Comparison of Costs at 4850 ft versus 6950 ft
There are two depth dependent costs, the cost of hoisting rock and the
cost of rock and cable bolting. To estimate this effect, we determine
the difference in costs between identical chambers built at the 4850
ft level (the bottom level of the Ross shaft, the upper hoist system,
and the beginning of the Winze 6, the lower hoist system) and the 6950
ft level. The direct manpower costs for hoisting the extra 700 meters
in the Winze 6 are about $0.30/ton. The power costs add another
$0.20/ton for a total of $0.50/ton or $208,000 per 100 kiloton
module, where shaft maintenance costs have not been included.
The incremental rock support costs are more difficult to determine.
The cable bolting planned and budgeted for these modules is far
greater than required. This was done to insure that the chambers
would have a minimum 50 year occupancy. A similar approach to
corresponding excavations at the 4850 ft level might result in exactly
the same bolting pattern and thus the same cost. Another approach
would scale the bolting cost by the difference in rock stress between
the two levels. The rock stress in the Homestake and Poorman
formations, the formations that have been extensively studied in mine,
are rather surprising. The measured vertical stress Sv= 28.3 ×
D kPa, where D is the depth in meters, is exactly what is expected
for a fluid of density 2.9 (the rock density). The horizontal stress
is very direction dependent. Along the high stress axis Sh1 =
14,328 + 12.4 × D kPa, while along the low stress axis, Sh2
= 834 + 12 × D kPa. Presumably, the high horizontal stress
results from the rock folding that resulted in the upbringing of the
gold ore deposit to the surface and thus its discovery.
We assume that the effective stress at 6950 ft is about 35% greater than the corresponding
one at the 4850 ft level. Since the total cost of the cable and rock bolts is
$910,000 and the related labor, including benefits, is about the same,
we assign a depth dependent cost increase of $630,000 for rock support.
Combining this with the increase in hoisting costs gives a total of
$838,000 or 6% of the total construction cost.
Note that this is less than 2% of the complete detector cost.
However, there are three offsetting costs that reduce the cost of
constructing the Cherenkov detector array at 6950 ft vs. at 4850 ft.
The first of these is the water fill. The total water fill for the
megaton detector is 250 million gallons. Removing that much water
from the local streams would be quite significant, especially given
the present drought conditions in the area. Instead, we plan to use
the water that is being pumped from the bottom of the mine at the 8000
ft level. This water will be purified to remove any light scattering
or absorbing material and any radioactive contaminants. Since the
mine now pumps out about 350 gallons per minute, we will require about
1.4 years worth of water distributed over the construction time of the
entire detector. For a detector at the 6950 ft level, this water is
only pumped up 1000 ft while for a detector at 4850 ft, the water must
be pumped up about 3100 ft. The cost savings here is about 1/4 of
the increase in rock hoist cost or about $50,000.
The second offsetting cost is that of cooling the Cherenkov detector.
Operating the detector at 10oC gives 1/4 the photomultiplier
noise of operation at 20oC. Since the rock temperature at the 4850
ft level is over 35oC and still higher at 6950 ft, cooling will be
necessary at either depth. The mine has an enormous refrigeration
plant (2400 ton capacity) at the 6950 ft level, with a fairly short
path for the coolant from the refrigeration plant to the detector. A
detector at the 4850 ft level will either require a new refrigeration
plant at that level or the installation of 2000 ft of vertical coolant
piping in the mine shaft. We have not estimated the cost of either of
these steps, but they are clearly very substantial.
The third offsetting cost is that the level structure at 4850 ft does not readily lend
itself to the construction and ventilation system described above.
If the upper detector access is at 4850 ft then the lower, rock
removal tunnel is at 5000 ft. Unfortunately, there is no ventilation
exhaust system at that level and waste rock would have to be raised
in order to get it into the hoist system. The alternate approach,
putting the top access at 4700 ft, would require additional excavation
in order to provide the necessary tunnels for the upper access.
The material in this section was assembled and compiled by Kenneth
Lande based on work done by a number of senior mining engineers who
previously were in charge of mining operations at the Homestake Mine.