Saltdome Shower Array:
A GZK neutrino Detector For
High Energy Physics & Particle
Astrophysics
Part II: Salt Domes & Detector Details
Peter Gorham
With help from Gary Varner
University of Hawaii at Manoa
SalSA presentation, DOE HQ
1
What is needed for a GZK n detector?
Standard model EeV GZK n flux: <1 per km2 per day over 2p sr
Interaction probability per km of water = 0.2%
Derived rate of order 0.5 event per year per cubic km of water or ice

A teraton (1000 km3 sr) target is desirable!
Problem: how to scale up from current water Cherenkov detectors?
One solution: exploit the Askaryan effect:
coherent radio Cherenkov emission
Particle showers in solid dielectric media yield strong,
coherent radio pulses
Neutrinos can shower in many radio-clear media: air, ice,
rock-salt, etc.
Economy of scale for a radio detector (antenna
array + receivers) is very competitive for giant
detectors
SalSA presentation, DOE HQ
2
Saltdome Shower Array (SalSA) concept
Salt domes: found throughout the world
Qeshm Island,
Hormuz strait,
Iran, 7km
diameter
1
2
Depth
(km)
Isacksen salt
dome, Ellef
Ringnes
Island,
Canada 8 by
5km
3
Antenna array
4
5
6
7
Halite (rock salt)
• La(<1GHz) > 500 m w.e.
• Depth to >10km
• Diameter: 3-8 km
• Veff ~ 50-350 km3 w.e.
• No known background
• >2p steradians possible
• Rock salt can have extremely low RF loss, as radio-clear as Antarctic ice
• ~2.4 times as dense as ice
• typical: 50-100 km3 water equivalent in top ~3.5km =>300-600 km3 sr w.e.
SalSA presentation, DOE HQ
3
U.S Gulf coast salt domes
Hockley salt
Dome & mine
Houston
Salt origin: Shallow Jurassic
period sea, 200-150M yrs old,
New Orleans inshore Gulf coast area dried ~150
Myrs ago
Formed fairly uniform evaporite
beds ~1 km thick or more, known as
‘Louann’ salt:
94-98% halite (NaCl)
2-6% anhydrite (Calcium sulfate)
Trace Mg, Sr, dissolved gases, 1040 ppm trapped brine
Salt density (2.2) < rock (2.6)
plasticity at 10-15km depth leads
to ‘diapirism’ : formation of buoyant
extrusions toward surface
Diapirism for Louann salt ceased
50-100 Myrs ago, left stable salt
diapirs all over the Gulf coast
SalSA presentation, DOE HQ
4
Gulf coast salt domes
1.5 - 8 km sectional
axes, circular to
highly elliptical
vertical extent from
near surface to 10
km depths common
Source of oil & gas
trapped on flanks:
impermeability of
salt compared to
sediments
SalSA presentation, DOE HQ
5
Examples of Gulf coast halite purity
Salt dome
Sample depth, ft
halite %
anhydrite %
Splindletop, TX
2676
94.83
5.17
Sour Lake, TX
7290
92.48
7.52
Saratoga, TX
--
96.79
3.21
McFaddin, TX
2645
98.47
1.53
Hull, TX
706
92.15
7.85
Moss Bluff
4566
96.02
3.98
High Island, LA
3359
89.63
10.37
Grand Saline, TX
Various, mine hor.
98.0
2.0
Hockley, TX
1200, estimated avg.
95.0
5.0
Avery Island, LA
Mine horizon
98.73
1.2
Cypress Creek, MS
1720-1876, 8 samples
95.61
3.97
Richton, MS
1120-1270, 8 samples
94.08
5.41
Port Barre, LA
--
99
1
SalSA presentation, DOE HQ
6
Halite & anhydrite
Pure NaCl crystals are theoretically lossless to RF via absorption
Crystal lattice defects are only mechanism for loss
Rayleigh & Mie scattering lead to attenuation over 100’s of m
Measured in situ bulk attenuation lengths can be several hundred m or
more in many salt domes, but not all (Weeks Island--water intrusion)
Chief impurity: anhydrite (anhydrous gypsum or alabaster)
Also known to have ultra-low loss at radio frequencies
Expectations: typical Louann salt will have at least several hundred
meter attenuation length if water content is low (<300 ppm)
Core samples indicate low water content in 80-90% of domes
SalSA presentation, DOE HQ
7
Halite-anhydrite salt dome structure
Morton Salt mine, Grand
Saline Salt dome, TX
~98% pure halite, 2%
anhydrite
Anhydrite banding
evident, nearly vertical
from deformation of
original salt beds
Produces negligible
effects on radio
propagation
SalSA presentation, DOE HQ
8
In situ salt dome measurements of attenuation
Location
Freq.,
MHz
Loss coefficient
Attenuation
length
Method
reference
Pine Prairie salt
dome, LA
230
<0.0042 per m (best)
<0.0105 (typical)
<0.016 (worst case)
>235m
>94m
>66m
GPR, from salt dome
flank reflections, 150200m typical one way,
very close to flank
Holser et al. 1972
Cote blanche salt
dome, LA
440
<0.0033 per m
>300m
GPR, 1245m path,
derived
Stewart &
Unterberger 1976
Hockley dome, TX
440
<0.005 per m
>200m
GPR, derived from
reflections, 350m 1-way
Hluchanek 1973
“saltdome in N.
Germany”
22.5
0.0027 per m
~370m
Dual borehole, 470m
separation
Nickel et al. 1983
Hockley dome, TX
150
300
750
<0.0039 per m
<0.0047 per m
<0.0041 per m
>256m
>213m
>243m
Transmit & receive
through salt column, 40m
thick
Gorham, Saltzberg
et al. 2001
SalSA presentation, DOE HQ
9
Borehole radar on dome flank
Pine Prairie dome, LA northern
extreme of Louisiana salt dome region
Holser et al 1972 used dipole & helix
antennas at 230MHz in a 5” diameter
sonde to map the flank of the dome
(1 microsec pulses)
Most data within 150m of edge of
dome (impurities increase close to
flank)
Flank location confirmed by retrieved
samples when flank was intercepted
Good data & SNR to 8000 foot depths,
until flank was pierced
SalSA presentation, DOE HQ
10
Salt Dome Selection & Phase I Prototype
Inputs: Surveys in 1970’s, 1980’s for Nuclear Waste Repository sites
Stringent requirements with similar needs to SalSA, large, stable dome
with dry salt, no economic usage
Richton (MS) and Vacherie (LA) domes both have excellent DOE salt core
reports
Keechi Dome in TX also appears to have no oil or gas interests
Select 3-5 salt domes, drill 1500’ borehole with 300-500 ft of salt
penetration, continuous core
Use chemical & loss-tangent measurements on core, plus borehole radar to
assess initial salt quality
Choose best of initial domes that meet requirements for three or four deep
(3km) boreholes, to install a prototype SalSA (‘Salsita’)
1-2 years’ operations to establish proof-of-concept, and discover or
confirm small sample of GZK neutrino events, then propose full array
SalSA presentation, DOE HQ
11
Current Salt Dome candidate ranking
Rank
1
Dome
Richton
US
State
MS
Volume to
3.5km depth
[cubic km
salt]
50.0
Maximum
aperture [cubic
km steradians
water equiv.] (a)
684.4
Salt crest mininmum
(left) and average
(right) depths [m] (b)
220
Positive Notes
Negative Notes
360
Shallow, extensively mapped, no
oil or gas production. Salt core
analysis shows 94% halite 5%
Numerous dropped projects. Plans for
anhydrite. Good drainage, flat cap
LPG storage caverns could revive?
region. Huge body of ONWI survey
literature. Industrial forestry on
most of cap.
some potential for flooding on dome
crest?
2
Vacherie
LA
45.4
621.7
240
300
3500' core taken by DOE, salt
analysis done; very low water
content in salt; shallow cap. Large
body of ONWI survey literature
3
Keechi
TX
32.7
447.3
90
900
No oil or gas, good drainage
around dome, good access. Near
Palestine, TX
Little survey work. Sloping cap;
requires >1000m bores on flank
4
Hainesville
TX
39.2
536.5
366
400
Well mapped; circular, flat cap,
very large, sparsely populated
Extensive oil/gas production (53 wells);
Several LPG storage caverns.
Extensive oil/gas prod, S flank densely
producing. In "Bubbling Bayou."
Sloping cap, requires 1000m bores on
flank. Probable LPG storage caverns
5
Chacahoula
LA
38.7
529.8
370
1000
No oil directly on cap (in 1961),
sparse population expected.
6
Butler
TX
21.1
289.3
10
450
Shallow cap. Near Palesine, TX.
Good access roads. Minimal oil or
gas production.
Sandstone quarry on flank. Two or
more gas storage caverns in 1984.
Potential for flooding.
7
Cypress Creek
MS
15.0
205.8
396
400
Extensively mapped by DOE. Flat
cap. DOE salt core analysis gives
94% halite 6% anhydrite.
Oil & gas wells on S and W flanks.
Some potential for flooding on cap
region.
NOTES:
(a) Assumes 2 pi steradians solid angle acceptance for all of volume (consistent with simulations), density of 2.18 g/cc for salt.
(b) Minimum depth is shallowest salt from surface; average is over all of crest needed to instrument the dome for a SalSA.
SalSA presentation, DOE HQ
12
Richton Dome
Richton Dome has excellent seismic,
gravity & sulfur exploration
(unsuccessful) measurements of salt
body
SalSA presentation, DOE HQ
13
Richton Dome area
Land use
primarily
industrial
forest
Plum Creek
Land Mgmt
contacted,
lease option
negotations
ongoing
SalSA presentation, DOE HQ
14
Mechanics of land use & drilling
Land use & rights studies underway, will have agreements in place for
initial phase as pre-requisite for proposal
Mineral rights owner/leaseholders will retain asset rights if oil, gas, sulfur,
etc. is discovered (unlikely but not excluded)
Surface rights owners will receive “damages” for 1 acre drilling site, and
lease agreements for duration of project
Depends on land usage, rural land: $1-2K damages typical per well
Typical $1-2K/yr lease for small well-head site (~100 sq. ft.) & right of way
Will negotiate contracts for “options” on leases for proposal
Baker-Hughes INTEQ has expressed interest in cost-sharing agreement
for prototype phase
Mississippi Office of Geology is supporting Richton dome SalSA studies
SalSA presentation, DOE HQ
15
Drilling salt domes
Shallow holes: a modest rig possible, 20-40’ truckmounted; water-well driller capable
Deep holes require large derricks, 130’ high typical,
and a 1 acre site
Bore is drilled through surface layers and “caprock” to
about 1000’ depth into salt, and must be cased with
steel liner above salt
Salt is hermetic and needs no casing or liner, is easily
drilled
Requires oil-based drilling fluids to avoid brine
formation
Borehole remains OPEN after drilling, probably for
decades at a 4” bore, and is backfilled with fluid
providing hydrostatic pressure head
Ergo: Strings will be repairable, recoverable, can be
upgraded!
SalSA presentation, DOE HQ
16
Drilling Salt Domes
Drilling costs preliminary estimates $120-150K per 1500’ bore,
$250-350K per 3.5 km deep hole
4 shallow & 3-4 deep holes: $1.2M-$2M including casing and
cores
Capital cost of dedicated drill rig ($0.8-1M) would be justified
for full SalSA, but not at this stage
rig can be sold at termination of drilling, capital re-invested in
project (eg., Don Thomas at UH has done similar)
Damage & lease costs:
Damages of order $20K in initial year
Lease costs expected to be of order $20K per yr for 3 years
Negotiations for lease options in progress
SalSA presentation, DOE HQ
17
String instrumentation: “node” configuration
Antennas (copper cylinders)
are cheap, “controller nodes”
(receiver, digitizers, data
transmitters, & pressure
housing) costly, THUS:
Use many (12) antennas per
controller node to optimize
sensitivity
12 nodes of 12 antennas
each is current choice
$100-$150K estimated per
string cost with no new
technology
pressure-compensated
controller system to be
demonstrated
SalSA presentation, DOE HQ
18
Fat dipole results in salt
120 MHz
370 MHz
180 MHz
530 MHz
Gain, dB
50 ohm
feedpoint
coupling
SWR (predicted)
SWR (measured)
4” diameter by 30 inch length, copper
Usable from 50MHz to 1 GHz (better than
model predicts)
Single mode from 50-350MHz
SalSA presentation, DOE HQ
Frequency, Hz/MHz
19
Basic string architecture
NEMA 3R
38" x 21" x17"
String
12 nodes
armor
tape
Insulated
conductors
Stainless tube
Fibers
Node = 12 antennas
and center housing
SalSA presentation, DOE HQ
20
GEISER
(Giga-bit Ethernet Instrumentation for SalSA Electronics Readout)
GEISER Philosophy
Set low threshold
Fill Gb/s ethernet link
Event build at surface
Pure digital transmission
Trigger/Event building
No custom, fast trigger
Exploit telecomm
Event building on PC farm
SalSA presentation, DOE HQ
21
GEISER Data flow
GEISER approach:
Digitize the “mud”
in downhole
Pan for gold at the
surface
Trigger
packets sent
via FM/local
radio
4-deep analog buffering:
Node/String Time stamps
100ms latency/hit
>99.999999…% livetime @ 1.5kHz
Event request
Data Transfer
RF in
Continuous
Hold at
1.5kHz
(>2.4s)
Internal FPGA
Buffer RAM
Digital Cell system
for data collection
64kb/event
1.6kHz (100baseT)
16kHz (GbitEthernet)
SalSA presentation, DOE HQ
22
In hole digitization
Digitizer n’ Readout, In-situ Transient Observation in Salt
[D’RITOS]
Massively parallel ADCs
• 50ms conversion
• 7x256 samples/event
• 50ms readout
(40MHz)
• 100ms total latency
4-deep analog
buffering for each
antenna channel
3rd generation switched-capacitor
array (SCA) architecture
6
Reference
timing
Channel
SalSA presentation, DOE HQ
23
Readout board
D’RITOS
LNA, 2nd-stage
amps
HV-lvDC regulation on separate board
Trigger, bi-directional fiber-link
LNA, 2nd-stage
amps
RF conns
SalSA presentation, DOE HQ
24
Radio Cherenkov testbed system
Salt, 25 tons
Liquid Scintillation
counters (MACRO)
Antenna layer
Shown exposed
Goal: to detect first coherent radio Cherenkov emission signals
of natural origin, from muon-bremsstrahlung showers
Standard hodoscope tagging combined with antenna array
SalSA instrument development: up to 196 antenna channels!
SalSA presentation, DOE HQ
25
First Observation of Cosmic-ray muongenerated Radio Cherenkov signals
Average of ~10K events selected for showers, out of 230K (2mo. data)
Signal antennas & time determined by track fit from scint. Counter
Backgrounds taken from out-of-cone and out-of-time data
We see strong enhancement due to ensemble of ~200 GeV
muon bremsstrahlung showers
SalSA presentation, DOE HQ
26
Summary
The SalSA concept
intellectual fruit of two OJI awards, Saltzberg & Gorham
Strong HEP motivation to study & use GZK neutrinos
We have gone about as far as we can without a prototype array
Salsita will position us for a full-scale proposal within 2 years
Capable of discovery and/or confirmation of GZK flux
Pathfinder for full scale detector, built around the prototype
We solicit your advice & guidance!
OJI awards have mentored us both to this stage
We offer SalSA as a next generation Energy Frontier HEP instrument
SalSA presentation, DOE HQ
27
Neutrino Flavor/Current ID
Charged current
(SM: 80%)
Neutral current
(SM: 20%)
e
25% hadronic +
75% EM shower at
primary vertex; LPM
on EM shower
Single hadronic
shower at
vertex
m
25% hadronic at
primary, 2ndary
lepton showers,
mainly EM
Single hadronic
shower at
vertex
t
25% hadronic at
vertex, 2ndary
lepton showers,
mainly hadronic
Single hadronic
shower at
vertex
~2 km
1018 eV nm
Charged/neutral current & flavor ID possible on subset of SalSA events
At least 20% of GZK CC events will get first order flavor ID
For non-SM high neutrino cross sections, NC events can interact twice
SalSA presentation, DOE HQ
28