GEODM Interest in the SNOLAB Cryopit...GEODM/Cryopit 2011/08/15 Sunil Golwala Physics Requirements...

GEODM Interest in the SNOLAB Cryopit2011/08/15 S. Golwala (Caltech) for GEODM Collaboration

GEODM/Cryopit 2011/08/15 Sunil Golwala

10 Mass [GeV/c2]

SI [c

m2 ]

CDMS II

15kg @ Soudan

100kg @ SNOLAB

1.5T @ DUSEL

1

23

4

102 10310 47

10 46

10 45

10 44

10 43

10 42From CDMS II to SuperCDMS and GEODM

2

Staged three-prong program toexplore MSSM or study a signal:• decreased backgrounds• improved background rejection• increase in mass/detector and decrease in

cost/detector< 1 event misid’d bgnd at each stage

CDMS IICDMS II

∅7.5cm x 1cm ZIP0.25 kg/detector

16 detectors = 4 kg2 yr, 1700 kg-d

SuperCDMS SoudanSuperCDMS Soudan∅7.5cm x 2.5cm iZIP

0.64 kg/detector15 detectors = 10 kg*

2 yr, 4000 kg-d

SuperCDMSSNOLAB

SuperCDMSSNOLAB

∅10cm x 3.3cm iZIP1.5 kg/detector

72 detectors = 110 kg3 yr = 100,000 kg-d

GEODMGEODM

∅15cm x 5cm iZIP5.1 kg/detector

300 detectors = 1.5 T3 yr, 1.5 M kg-d

x6

x15

x15

*iZIP has 2x larger fiducial efficiency than CDMS II ZIP

10 kg

110 kg


Physics Requirements and Approach

• GEODM: Germanium Observatory for Dark Matter• Physics requirement: WIMP search with reach of 2 x 10-47 cm2 at

60 GeV/c2 from 1.5 M kg-d exposure• Approach to meet this requirement

• 1.5 tons of cryogenic Ge detectors of SuperCDMS iZIP type; 3 yrs run• each detector measures athermal phonons and ionization produced by nuclear and

electron recoils in a semiconducting Ge subtrate cooled to 40 mK• discrimination between nuclear recoils (WIMPs) and electron recoils (background) by:

‣ ionization yield: number of e-hole pairs per keV deposited energy (energy deposition density)‣ fiducial volume cut to remove surface backgrounds and detector reconstruction nonidealities

‣ ionization signal radial partition‣ ionization signal top-bottom asymmetry‣ phonon signal top-bottom asymmetry‣ phonon signal timing

• segmented into 5-kg 15-cm x 5-cm individual substrates assembled in towers • Employs shielding to reduce backgrounds to acceptable level

• low U/Th Cu cryostat and 50-cm thick Cu shield for photons• 150-cm thick polyethylene or water shield for radiogenic neutrons• deep site provides sufficiently low cosmogenic neutron flux

3


Physics Approach: Cosmogenic Neutron Backgrounds

• cosmic-ray muon spallation of nuclei in rock walls• SNOLAB or DUSEL 7400 level reduce background below necessary

levels• CDMS-II-like shield: cryostat, high-Z gamma shield, low-Z neutron moderator

4

13

where ni is the number of atoms for target element i perkg material in the detector, d!n(E)

dE is the neutron energyspectrum (equation (14)) at depth h, !i(E) is the neutroninteraction cross section [49, 50, 51] with ith element ofnatural Ge, and Fi(E) is an energy-dependent quenchingfunction [62] specific to the ith element of Ge.

We generate the muon-induced neutrons at therock/carven boundary using the formalism outlined inSection III and propagate them through the CDMS II ge-ometry described above. Since the muon veto in CDMSII has an e!ciency greater than 99.9% [15], we are con-cerned only for the neutrons produced in the rock. Wehave performed our simulations for CDMS II using twodi"erent shielding configurations. Shielding configura-tion 1 is that used in the actual experiment with 0.5 cmcopper, followed by 8.6 cm polyethylene, 22.5 cm lead,and 40 cm of polyethylene as the outer neutron shield.In shielding configuration 2, we interchanged the thickpolyethylene and lead shield positions. In this case wefound a reduction in background by about a factor oftwo over the CDMS-II shield. This reduction occurs ow-ing to the additional neutrons produced when neutronsfrom the rock interact in the lead shield. Similar obser-vations have been made in Ref. [61] and Ref. [65]. Thevisible recoil energy spectrum induced by the fast neu-trons is shown in Fig. 22 for three di"erent depths andalong with the spectrum expected for dark matter as-suming a cross section !p = 10!46cm2 and a 100 GeVWIMP mass.

Energy (keV)0 10 20 30 40 50 60 70 80 90 100

)-1

year

-1 k

g-1

Even

ts (

keV

-910

-810

-710

-610

-510

-410

-310

-210

-110

N eutron-induced background at S oudan

N eutron-induced background at G ran S asso

W IM P s-induced nuclear recoil

N eutron-induced background at S udbury

FIG. 22: The predicted event rates for spin-independentWIMP-nucleon scattering (dotted-line) in Ge assuming aWIMP-nucleon cross-section of !p = 10!46cm2 and a 100GeV WIMP mass. Muon-induced neutron backgrounds arealso displayed for comparison, indicating the need for greaterand greater depth as experiments evolve in scale and sensi-tivity.

Using these results we determine an event rate of 0.9events/kg-year in an energy window of 10 to 100 keV forthe CDMS-II experiment operating at the Soudan mine.This rate is reduced to 0.5 events/kg-year after identify-ing those neutrons that interact with two or more crys-

tals in the CDMS-II tower. Our prediction is consistentwith the upper bound of 0.94 ± 0.38 events/kg-year thatcan be deduced from the CDMS II collaboration’s limitof 0.05 ± 0.02 neutrons detected during an exposure of19.4 kg-day [64] or 34 kg-day [63]. In Ref. [15], Kamatalso simulated the un-vetoed neutron rate in the CDMS-II detector and obtained 0.05 ± 0.02 neutrons for the19.4 kg-day exposure, in excellent agreement with ourprediction.

We can now derive the DSR appropriate to the CDMS-II experiment. As shown in Fig. 23, the experiment’ssensitivity would be limited to !p ! 10!44cm2 due tothe muon-induced fast neutron flux at Soudan and thatdepths in excess of ! 5 km.w.e. will be required to pushbeyond !p ! 10!46cm2, unless the neutron flux can besuppressed e"ectively either by further shielding and/oractive veto.

Depth (km.w.e.)1 2 3 4 5 6 7

-in

du

ce

d b

ac

kg

rou

nd

(e

ve

nts

/kg

-ye

ar)

µ

-410

-310

-210

-110

1

10

] (n

orm

ali

ze

d t

o n

uc

leo

n)

2C

ros

s-s

ec

tio

n [

cm

-4810

-4710

-4610

-4510

-4410

-4310

S hie lding configuration 1

S hie lding configuration 2

WIP

P

So

ud

an

Ka

mio

ka

Bo

ulb

y

Target sensitivity

Gra

n S

as

so

Su

db

ury

FIG. 23: The Depth-Sensitivity-Relation (DSR) derived forthe CDMS-II detector geometry for the two shielding configu-rations described in the text. The muon-induced backgroundis dominated by elastic scattering of neutrons depositing vis-ible energy in a 10 to 100 keV window. Specific points areshown, for example, at the depth of the Soudan mine wherethe CDMS-II detector has been operating. Uncertainties re-flect those present due to uncertainties in the rock composi-tion and in generating the muon-induced fast neutron flux.

It was pointed out in Ref. [65] that the nuclear recoilevent rate in coincidence with a second energy deposi-tion not associated with nuclear recoils (electrons, pho-tons, muons etc.) is a factor of 10 more than the rateof isolated nuclear recoil. We cannot confirm this state-ment (a factor of 10) with the CDMS II geometry forthe neutrons which are produced in the rock and associ-ated muons that miss the veto. This is likely due to thefact that the CDMS II detectors are segmented and muchsmaller than that considered in Ref. [65]. In our simu-lation, heavy charged particles such as pions, kaons andprotons with kinetic energy greater than 10 MeV that areproduced together with neutrons by muons in the rockare a factor of 10 less than that of neutrons and they

CDMS II design w/50-cm polyshield + lead

Mei

and

Him

e (2

005)

DU

SEL

4850

DU

SEL

7400

Need x10 at 4850 tobe safe

GEODM target2 x 10-47 cm2


CDMS II Experience for Lab Requirements

• Cleanroom: 6.8m L x 4.9m W x 5.8m H = 200 m3

• Class 1000 unoccupied, class 10000 when people working

• Anteroom (transition from dirty lab, storage): 4.4m L x 5.2m W x 3.8m H = 90 m3

• Crane: 5-ton bridge crane, free-standing, enclosed in RF/cleanroom• Not clear the RF room works very well...

• Cleanroom sits on following:• 30-cm Soudan cavern floor, sloped for drainage• 10-cm level slab• 1/2-cm polyethylene layer• 1-cm steel plate floor of cleanroom

• Cleanroom has mezzanine for shield storage; 5 tons lead shield + 2 tons polyethylene

• No radon abatement aside from• purging of shield volume and cryostat inside when open• purged storage of innermost components

5


GEODM/Homestake 7400 Design

6

transition from cavern

detector assembly roomclass 100 6m x 4m x 6m(detector can shown)

shield and cryostat∅5.6m x 5.6m

shield lid storage

rollup door forconstruction access

cryo/vac eqpt room

DAQ/electronics room

control room

25 m

10 m

12 mexperimental room

class 10000


Modifications for GEODM @ SNOLAB

• Cleanliness:• SNOLAB = class 2000 cleanroom. Obviates GEODM class 10000 cleanroom.• RF barrier to minimize pickup of noise from rest of lab

• Full enclosure or barrier at entrance to hall? Experience suggests unnecessary?• Still need internal class 100 radon-abated cleanroom and temporary structure

over experiment during detector payload installation.

• Space• Cube hall: 18m x 15m x 15m• Cryopit: ∅15m x 15m• Shortened 25m dimension compensated by greater overhead space at SNOLAB,

in-hall crane, use of drifts for noncritical utilities. • DUSEL required experiments to be self-contained inside predefined lab modules;

not a requirement at SNOLAB• e.g., cryo/vac, DAQ/control could go into drifts or mezzanine level

• No other major issues; likely cost-neutral or less expensive thanks to excellent SNOLAB infrastructure.

7


Timescale

• SuperCDMS SNOLAB currently expecting start of science operations at start of FY15• Getting project-specific R&D funds in FY11• Begin construction of facility, long-lead-time purchases in FY12• Must do experiment-specific R&D, pass CD milestones, and keep cost reasonable

• GEODM desired timeline for SNOLAB siting• Start project-specific R&D phase of GEODM in ~2015• Construction start ~2017/2018• Science operations start ~2020• Takeaway: need a hall of size similar to cryopit or cube hall in latter years of

decade.

8

SNOLAB would improve the present sensitivity for dark matter WIMPs by more than two orders of magnitude. The recent report to HEPAP from the PASAG has recognized the importance of this project in the field and recommended that it proceed under all funding scenarios considered. Figure 1 shows how the various phases of CDMS would dovetail together.

Activity Name 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

CDMS IIOperations

4kg, 4E-44 cm2Expected Sensitivity

SuperCDMS SoudanDetector R&DConstructionOperationsExpected Sensitivity 10 kg, 5E-45 cm2

SuperCDMS SNOLABR&DCritical Design Milestones

CD-0

CD-1

CD-2/3

CD-4

ConstructionSNOLAB facility

Ge Towers

OperationsPartial Payload, 2 years

Full Payload, 3 years

Expected SensitivityExpected Sensitivity 100 kg, 1E-46 cm2

GEODM...

Figure 1 CDMS projects and WIMP-nucleon cross section reach by calendar year

SuperCDMS SNOLAB R&D The SuperCDMS SNOLAB project is projected to cost $20-30M, based on previous experience and top-down estimates. Thus it will follow the DOE Critical Design process, although it might be managed from within DOE OHEP. We anticipate a CD-0 (Mission Need) in fall 2011, and a Major Item of Equipment (MIE) in the FY2013 budget. We anticipate that NSF will play a major role in this project as well, as they have in all previous iterations of CDMS. Given the expected scale of this project, and the limited funding likely to be available for a national direct detection experiment program, it is essential to address technological risks with an aggressive R&D program prior to the start of fabrication. FNAL and SLAC have submitted DOE Field Work Proposals (FWPs) to perform part of this risk assessment. The collaboration will soon submit a proposal to the NSF for the remainder of the program. The R&D program funded by DOE and led by SLAC will focus on demonstrations of:a) 100 mm diameter Ge crystals of suitable quantity and quality available from industrial vendorsb) a Ge fabrication rate suitable for a cost-effective 100 kg experimentc) Ge iZIP prototypes that meet specifications for background rejectiond) Design of a mechanical and cold electronics tower system adequate for 100 mm iZIPs The R&D program funded by DOE and led by FNAL will focus on the design of:a) A cryogenics system capable of operating 100 kg of detectors at 50 mKb) A shielding system (passive and active) that produces detector background levels suitable to

achieve 1E-46 sensitivity The R&D program funded by NSF will focus on:

SNOLAB would improve the present sensitivity for dark matter WIMPs by more than two orders of magnitude. The recent report to HEPAP from the PASAG has recognized the importance of this project in the field and recommended that it proceed under all funding scenarios considered. Figure 1 shows how the various phases of CDMS would dovetail together.

Activity Name 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

CDMS IIOperations

4kg, 4E-44 cm2Expected Sensitivity

SuperCDMS SoudanDetector R&DConstructionOperationsExpected Sensitivity 10 kg, 5E-45 cm2

SuperCDMS SNOLABR&DCritical Design Milestones

CD-0

CD-1

CD-2/3

CD-4

ConstructionSNOLAB facility

Ge Towers

OperationsPartial Payload, 2 years

Full Payload, 3 years

Expected SensitivityExpected Sensitivity 100 kg, 1E-46 cm2

GEODM...

Figure 1 CDMS projects and WIMP-nucleon cross section reach by calendar year

SuperCDMS SNOLAB R&D The SuperCDMS SNOLAB project is projected to cost $20-30M, based on previous experience and top-down estimates. Thus it will follow the DOE Critical Design process, although it might be managed from within DOE OHEP. We anticipate a CD-0 (Mission Need) in fall 2011, and a Major Item of Equipment (MIE) in the FY2013 budget. We anticipate that NSF will play a major role in this project as well, as they have in all previous iterations of CDMS. Given the expected scale of this project, and the limited funding likely to be available for a national direct detection experiment program, it is essential to address technological risks with an aggressive R&D program prior to the start of fabrication. FNAL and SLAC have submitted DOE Field Work Proposals (FWPs) to perform part of this risk assessment. The collaboration will soon submit a proposal to the NSF for the remainder of the program. The R&D program funded by DOE and led by SLAC will focus on demonstrations of:a) 100 mm diameter Ge crystals of suitable quantity and quality available from industrial vendorsb) a Ge fabrication rate suitable for a cost-effective 100 kg experimentc) Ge iZIP prototypes that meet specifications for background rejectiond) Design of a mechanical and cold electronics tower system adequate for 100 mm iZIPs The R&D program funded by DOE and led by FNAL will focus on the design of:a) A cryogenics system capable of operating 100 kg of detectors at 50 mKb) A shielding system (passive and active) that produces detector background levels suitable to

achieve 1E-46 sensitivity The R&D program funded by NSF will focus on:


Lab Interface Requirements

• Following are lab interface requirements developed for DUSEL design, quickly adapted for SNOLAB.

• More careful and SNOLAB-specific assessment needs to be done.• Esp. need to consider hoist trip requirements: very lab specific.

9


Power/Cooling

• 173 kW for experiment• Cooling• 100 kW directly by

chilled water

• 73 kW by cavern ventilation; need additional chilling capacity?

• Emergency power• 80-100 kW UPS to enable

carryover to generator• Generator-supplied

80-100 kW to keep system cold during outage• Avoids cryogenic cycle time

ItemPower Draw Cooling Method

4 pulse tube coolers 60 kW chilled water

DR circulation pumps 40 kW chilled water

class 100 cleanroom air handler 5 kW cavern vent.

front end electronics 30 kW cavern vent.

back end electronics 20 kW cavern vent.

control room 5 kW cavern vent.

back end/control room air handler 13 kW cavern vent.

Total expt 173 kW

10


Ventilation/Radon Requirements

• Modest amounts of cryogenic liquids• few x 100L LN, LHe

• Need more during cooldown to assist cooling

• Radon mitigation required during opening of cryostat/detector installation• LN boiloff purges radon from between high-Z shield and poly and inside poly

during normal operation

• Implementation of radon abatement:• Need ~ 30 m3 radon-abated volume around inner can for detector insertion• Need similar sized volume over cryostat during inner can insertion• Existing radon abatement systems (Ateko) provide 150 m3/hr at 1 mBq/m3 using

15 Bq/m3 input air for ~$200k and ~50 kW. Achieved 150 mBq/m3 at 1 air change/hr for NEMO-3, probably limited by leaks.

• Even with ~ 500 Bq/m3 and higher output radon level due to higher change rate (10-60 changes/hr) seems likely could achieve < 1 Bq/m3 in 30 m3

• Need to detail power requirements, hazards; but low duty-cycle for use, anti-phased with experiment cryo/electronics power needs, so doesn’t increase total power draw.

11


Floor Loading and Hoisting Needs

• Dominant weights:• Copper shield: 100 tons, segmented• Polyethylene or water or liquid scintillator shield: 125 tons, segmented

• Floor loading:• Close-in shield: ~250 tons spread over ~6 m diameter circle = 10 tons/m2

• Crane• Need a 20-ton crane to hoist shield parts for opening of cryostat.

12


Dimensions

• What dimensions set our needs for hoist trips?• Detectors? no, 1.5T Ge only ~ 1m3, will come down in multiple trips• Cryostat? yes, 77K can ~1.6m diameter & height. Can make cryostat aspect ratio

flat to accommodate. Vacuum shell will need a split flange and its diameter will be limited. Dogleg parts long and thin.

• Dilution fridge? no, long and thin.• Dilution fridge plumbing? no, can be disassembled into parts that can be joined

underground• Shielding material?

• High-Z: will need to segment into blocks anyways, hoist capacity will set limit• Scintillator panels? Long thin, light transport sets natural limits anyways.

• Pumps and chillers? They will fit. Could always be disassembled and reassembled.• Air handlers: may need to disassemble• Structural elements: clean room walls, long structural pieces• Electronics? Can be disassembled.• Purge cabinets: can be assembled underground

13


Major Hazards

• Major hazards• Lifting• Cryogenic hazards during cooldown• HV for PMTs• Vacuum loss• Fire hazard due to plastic scintillator veto (~6 m3 of plastic scintillator)• Water tank if water moderator shield• Fire, chemical hazard due to liquid scintillator neutron veto if implemented• Electrical/water interaction if instrumented water moderator shield

14


IT Requirements

• IT: 50 TB/day data rate, rather demanding• because # sensors/mass is not being reduced. Details:

• 16 bit digitization → 2 bytes/sample, 5000 samples at 1 MHz, 64 channels/detector = 640 kB/event/detector.

• Assume 10 detectors on average in each event → 6.4 MB/event.• Event rate ~ 0.3 Hz in CDMS II, x 300 in mass, so 100 Hz → 640 MB/sec, 50 TB/day• Probably a substantial overestimate:

EM bgnd will be reduced because of U/Th reductions needed to reach desired radiogenic n bgnd, self-shielding of inner detectors will reduce rate

• Need 10 Gb/s ethernet, few days local storage needed ($100/TB today).• Calibration mode is greater challenge: ~100x faster for 10% of the time , but don’t

need to keep up with rate in real time. Still, would increase needs by x10. Need to select energy region of interest online to cut down rate.

• Send all data to surface ~immediately for local backup, fast analysis• Transfer data off-site immediately, need to buffer in case of network outage.• Few days local storage, CPU to do quick offline diagnostics• Need 10 Gbits/sec out of SNOLAB to off-site storage

• Need routine remote control access (on- and off-lab)

15


Other Space Needs

• Underground materials manufacture, storage• No: cosmogenic activation not an issue

• Surface needs• Storage of materials between arrival and transport underground;

Largest parts: cleanroom walls, water tank raw materials• Office space for up to 5-6 people, facilities for collaboration meetings• Machine and electronics shops for utility work, but no substantial machining or

electronics design or manufacture on-site.

16


Staging 1: Dirty

• Preparation and sealing of main experimental hall, but not final clean• Bring in construction materials for RF barrier and inner class 100

cleanroom, cryostat vacuum can, refrigerator, cryocoolers before room becomes clean.

• Cryo/vac, DAQ/electronics, control room can be constructed in parallel or afterward. • Once cryo/vac and DAQ/electronics rooms are ready, this equipment can be

brought down and installed in parallel with inside-clean-room operations. • Not as complicated, no cleanliness requirements except in making feedthroughs.• Includes radon abatement system up to feedthrough ducts to clean room

• Construct class 100 cleanroom, do all mechanical infrastructure work for cryostat, shield, refrigerator, cryocoolers (i.e., frame to hold fridge, mechanical structure to hold cryostat and shield materials, etc., but not the cryostat itself yet)

17


Staging 2: Clean

• Go class 2000 in experimental hall, class 100 in inner cleanroom• Reclean all items inside cleanroom.• Begin construction of shield, cryostat, and refrigerator/cryocooler.• Bring in cryostat cans. Maintain in radon-free storage.

• Could bring them in earlier if transported in and stored in airtight, purgeable containers

• Radon abatement by purge from gas canisters, not radon-abated air.

• Plumb radon-abatement system into class 100 cleanroom and cryostat cover tent (through HEPA filters)

• Install cryostat cans• Bring in detectors in batches through transition and store in class 100

cleanroom in radon-free storage (purge cabinet)

18

GEODM Interest in the SNOLAB Cryopit...GEODM/Cryopit 2011/08/15 Sunil Golwala Physics Requirements...

Documents

Transcript of GEODM Interest in the SNOLAB Cryopit...GEODM/Cryopit 2011/08/15 Sunil Golwala Physics Requirements...