MICE The Hydrogen System Safety Review Introduction
Tom Bradshaw, Yury Ivanyushenkov, Elwyn Baynham, Tony Jones, Mike Courthold and Matthew Hills Rutherford Appleton Laboratory
Overview of talks
1. Introduction (TB) Design philosophy Basic operation including hydride bed Basic design calculations Motivation and approach R&D cycle Safety review process
2. Process and Instrumentation Diagrams (MH) 3. Control System (MJDC) Overview of control system Prototype flow diagrams & control sequences Pumps and Instrumentation Implementation and hardware
4. Hall Infrastructure and R&D Cryostat (AJ) 5. HAZOP and failure analysis (YI)
Introduction
The muon ionisation cooling experiment is an pre-cursor to a neutrino factory. Its objective is to demonstrate muon cooling i.e. produce a collimated beam of muons. This is achieved by cooling and slowing down the muons in an absorber (Hydrogen, Helium or plastic) before accelerating them in an rf field. The absorber for use with liquid hydrogen or helium consists of a chamber with thin aluminium windows. This requires a hydrogen delivery system which is the subject of this review. The design of the absorber and its implementation is the subject of a separate safety review.
Safety Review Process
We are here
Objectives of R&D
There are two phases to the implementation of the hydrogen delivery system: R&D Phase where a single system is developed and tested on a test cryostat which represents an absorber. Implementation phase where the delivery system is matched to a real MICE absorber
The objective of the R&D phase is to demonstrate a safe hydrogen delivery system for the absorber. This will consist of a first model MICE hydrogen delivery system together with a test cryostat that does not have thin windows but does contain the same instrumentation.
Cross section of Absorber
This is not the subject of this review Windows are rated up to: Design pressure Test pressure Burst pressure 1.6bar 2 6.4
a) Windows are mounted off RT interface see thermal model later b) Space for change in pipe dimension close to magnet c) Large bucket at base to contain any rupture
Safety Review Process
KEK Absorber design
Past Experiences
Page Operation/Cause Result Venting/ Leak purging into operation enclosed space Air leak into system Other
81 82 83 85 86 87 89 91 93 95 97
Thawing plugs with oxy acetylene torch Sanding causing sparks Purging using blower Oxygen accumulation from air leak Smoking near hydrogen Plugged vent line Plugged line Hydrogen trapped because of plug (non cryogenic) in vent line Use of hair dryer to warm cryostat Formation of hydrogen in a uranium plant Rupture of liquifier high pressure lines venting into room
Explosion Explosion Explosion Explosion Explosion Explosion Explosion Explosion Explosion Explosion Explosion
Past Experiences
Page Operation/Cause Result Venting/ Leak purging into operation enclosed space 37% 21% Air leak into system Other
100 101 103 105 107 109 110 112
Warming up cryostat with hair dryer Reactor test cell caused shed to fill with hydrogen Rupture of Bourdon tube Combustion in reactor from leakage through valve stem Dewar leak Battery case explosion Overpressure of turbine Incorrect purging operation Percentages *
Fire Explosion Fire Fire
Explosion Explosion Plant failure Explosion 42%
16%
Hydrogen incidents listed in Control of Liquid Hydrogen Hazards at Experimental Facilities by A A Weintraub.
Hydrogen Accidents - Industrial
Number of Total Category Incidents 32 25 21 12 10 10 35 145 Percentage of Accidents 22 17 14 8 7 7 25 100
Undetected leaks Hydrogen-oxygen off-gassing explosions Piping and pressure vessel ruptures Inadequate inert gas purging Vent and exhaust system incidents Hydrogen-chlorine incidents Others Total
Source: Safety Standard for Hydrogen and Hydrogen Systems, NASA NSS 1740.16, p.A-109
Design Philosophy
We have three absorbers and have three independent hydrogen systems, this is to: Avoid consequential failures a failure or fault in one is easier to deal with than a fault on a large system This will ease the staging for MICE as only one absorber is required early on Smaller systems are easier to work on
Design Philosophy
Other considerations: Minimise venting many accidents are caused during this process MICE has to be flexible there will be many filling cycles of the absorbers and we want to minimise the amount of hydrogen that we have hence the use of a hydride bed Control system automates the filling, emptying and purging system (many accidents from ineffective purging) Must be safe in the event of a power loss or system shut-down (looking at default valve positions) No surfaces the below BPt of Oxygen this is to prevent cryopumping of oxygen on any surface that may come into contact with hydrogen in the event of a failure Safety volumes to contain gas, relief valves to prevent back flow in case of catastrophic release
RAL Codes
Hydrogen zones definition according to RAL Safety Code No.1: Zone 0: An area or enclosed space within which any flammable or explosive substance, whether gas, vapour, or volatile liquid, is continuously present in concentrations within the lower and upper limits of flammability. Zone 1: An area within which any flammable or explosive substance, whether gas, vapour, or volatile liquid is processed, handled or stored and where during normal operations an explosive or ignitable concentration is likely to occur in sufficient quantity to produce a hazard. Zone 2: An area within which any flammable or explosive substance whether gas, vapour or volatile liquid, although processed or stored, is so well under conditions of control that the production (or release) of an explosive or ignitable concentration in sufficient quantity to constitute a hazard is only likely under abnormal conditions. Intention is to have no Zone 0 or 1 regions in the design
System Overview
Gas Delivery System Hydride bed for gas storage Control Valves, pumps, alarms and indicators Buffer volume Control System Controllers Interface to the rest of MICE Test Cryostat Cryocooler Instrumentation Hydrogen volume
Window rupture
Must be safe in the event of a window rupture: Introduction of a buffer vessel limits pressure excursions Pipework sized to accommodate gas release
Effectiveness of Buffer Volume
2 1.8 1.6
Buffer vol pressure rise
350 300 250 200 150 100 50 0 temperature
Expected boil-off rate
pressure Pa
1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 5 10 tim e secs 15 20
Latent heat Power into liquid Hydrogen boiled off (kg/s) Start mass of liquid Liquid density Start pressure (bar) Rgas
446000 10179 0.022823 1.544 70.288 0.5 or 1 4157
J/kg W kg/s kg
Assumes mixing of gas - cold from absorber + buffer volume
dt Buffer vol 0.2 1 m^3
Temp in buffer calc on basis of constant Cv - this is optimistic for Tgas ~50K but pretty good for Tgas >100K For large outflow through relief valve the algorithm is not correct because the valve essentially shuts Buffer volume gives a huge safety margin over just the pipe system with vol ~ 0.1m^3 for 50m of 50mm dia pipe The buffer vessel will keep the gas warmer due to its thermal mass - this is not included - it will increase the pressure rise Typically with 1m^3 Tgas ~100K pressure rise rate is 0.1 bar/sec valve opening time of 0.1-0.2 sec would be OK
density 300K relief valve pressure outlet mass flow
0.08 1.60E+05 1.20E-02
kg/m^3 Pa kg/s
What is temperature of Outer Window ?
Thermal balance between radiation [to inner window at...