Review of Helium Venting Analyses

1
Review of Helium Venting Analyses

• Chris Tutt
• AMS-02 Project Manager

2
Review of Hazard

• Hazard to be addressed is release of asphyxiant
  gas into an occupied area.
• Three major helium reservoirs within AMS-02
  payload and GSE.
• 2500-liter main helium dewar inside the payload.
• 1000-liter master dewar used for filling main
  dewar.
• 1000-liter transfer dewar used for filling master
  dewar.
• Venting analysis focused on main dewar as
  enveloping case, but results for all others are
  similar.

3
Main Helium Dewar

• Main Helium Dewar has two major components
• Helium Tank contains the cryogen itself
• Vacuum Case provides vacuum space around the
  tank.
• 3 bar burst disk on helium tank defines the
  Maximum Design Pressure for the system.
• Nominal operating pressure is 16 mbar.
• All hardware will be extensively tested prior to
  arrival at KSC.
• Structural analysis shows high margins for MDP.
• All welds will be inspected per MSFC-STD-504C.
• Both items will be proof pressure tested.
• Both items will be vacuum leak tested.

4
Main Helium Tank
5
Vacuum Case
6
GSE Dewars
7
Tank Overpressurization

• Two mechanisms for overpressurization of tank.
• Failure to remove nominal boil-off from tank.
• Large external heat source increases boil-off
  beyond pumps capability to remove.
• First scenario requires weeks to reach burst
  pressure, so does not present safety hazard.
• Only possible heat source for second scenario is
  ambient atmosphere and requires air leak into the
  dewar vacuum space.

8
Cause of Leak

• Two leak scenarios were considered in analysis
• Loss of Vacuum (LOV) Total loss of vacuum
  caused by large breach of Vacuum Case. Requires
  major accident
• Forklift tine breaks through VC outer cylinder
• Payload dropped during lifting operations
• Large hardware falls on payload from significant
  height.
• Maximum Credible Leak (MCL) 3 leak through the
  double O-Ring seals in the VC upper and lower
  support rings.
• Leak size defined by Payload Safety Review Panel
  and used for all payload bay leak analyses.
• LOV can be prevented operationally, so following
  discussion will focus on MCL.

9
VC O-Ring Seals
10

• Each sealing surface has double O-ring seals.
• In MCL, both seals are assumed to have failed in
  the same location.

OUTER CYLINDER
11
MCL Defined Leak Size
12
Venting Analysis Overview

• Venting Analysis consists of three basic steps
• Calculation of time required for tank to reach 3
  bar burst pressure after leak begins.
• Calculation of mass rate of flow of helium
  leaving the main tank.
• Calculation of oxygen levels in surrounding
  external space.
• Only the third step is location dependent.

13
Time to Burst

• Pre-burst pressure rise modeled as isochoric
  heating.
• Heat flux from MCL calculated to be 655.5 W.
• Helium thermodynamic properties from NIST
  handbook.
• Results driven by initial temperature and fill
  level.
• Calculated times range from 54 min (1.9K, 95
  full) to 86 min (1.7K, 80 full).

14
Typical Pressure Rise Profile
15
Time to Empty Tank

• Mass flow out of the tank modeled as isentropic
  expansion through choked nozzle.
• Results driven entirely by fill level.
• The more helium in the tank, the longer it takes
  to empty.
• Below 90 full, helium becomes two-phase prior to
  reaching 1 atm.
• Calculated times range from 131 min after burst
  (80 full) to 239 min (95 full).

16
Typical Mass Flow Profile
17
Exterior Volume

• Analysis done for four KSC spaces
• Space Shuttle Processing Facility
• Canister Rotation Facility
• Canister
• Payload Changeout Room
• Results driven by three factors
• Volume of external space
• Air refreshment rate
• Gas diffusion model

18
Original Gas Diffusion Assumption

• Original discussions assumed that helium vapor
  would rise to the ceiling of external volume.

19
Current Gas Diffusion Assumption

• At GOWG in November, GSRP requested analysis be
  redone assuming gas spreads evenly throughout the
  room.
• Model used GSFC algorithm provided by SHOOT team.
• Based on helium sensors measurements from
  Tevatron accident at Fermilab.
• Venting memo describes results using second
  assumption.

20
GSFC Algorithm
Incoming Air
Ventilation Exhaust
Incoming Helium

• Helium enters control volume at R m3/s.
• Ambient air removed by ventilation system at Q
  m3/s.

21
Oxygen Concentration Levels in PCR
22
Oxygen Concentration Levels in Canister
23
Oxygen Concentration Levels in SSPF
24
Oxygen Concentration Levels in CRF
25
Overall Trends

• PCR falls briefly below 19.5, but high
  ventilation rate allows rapid return to safe
  levels.
• Canisters small volume can be rapidly
  overwhelmed, but is not normally a manned volume
  when door is shut.
• SSPF/CRF volumes are large enough and ventilation
  rates are fast enough that oxygen level never
  falls below 19.5.

26
Proposed Safety Controls

• GSE should be monitored for signs of temperature
  or pressure rise while AMS-02 is in manned area.
• If leak observed, personnel should be removed
  from vulnerable areas.
• Entire PCR
• Elevated structures within SSPF and CRF
• Oxygen sensors should be used to determine safety
  of atmosphere prior to reentering any area after
  venting event or opening Canister in the PCR.
• Additonal vent lines and building modifications
  should not be necessary.