A Marone 121917 1 Outline Testing Scope Schedule Number of quenches Stored energies Burst disc rupture Contributing factors Quench Summaries Estimates Short Term Solutions Redesigned relief stack ID: 695207
Download Presentation The PPT/PDF document "Vertical Testing Quench Recovery" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Vertical TestingQuench Recovery
A. Marone 12/19/17
1Slide2
Outline
Testing ScopeSchedule
Number of quenches
Stored energies
Burst disc ruptureContributing factorsQuench SummariesEstimatesShort Term SolutionsRedesigned relief stackCold helium storageLong Term solution
2Slide3
Dewar Assembly
Tophat
Lambda Plate
Quench relief valve 5600g/s (ANSYS CFX model)&burst disc (5700 g/s)
Inner Dewar
Outer Dewar
LARP Magnet
17000 Lbs.
Design Conditions
35 PSI differential across Lambda plate.
Dewar Relief set at 50 PSIA.
Entire Dewar designed for 85 PSIA.
Lambda plate carrying 18,000 Lbs.
3Slide4
Testing Scope
Jan. ‘17 Mirror magnet ( full length, only one of 4 coils) successfully tested with no problems.Quench venting was not required ( almost no pressure buildup)
Approx. 1.4 MJ stored energy at highest quench.
First Prototype in vertical
dewar now.Testing stopped to resolve rupture disc issue.Results need soon for D.O.E. reviewSecond prototype due at B.N.L. in April ’18
Production magnet tests (27) starting in June ‘19
As of now 20-25 training quenches expected per magnet ( production).
Stored energies
4.5 MJ @ 16,470 Amp ( operating current)
5.3 MJ @ 17,890 Amp ( ultimate current)
4Slide5
Burst Disc Rupture
With automated quench venting burst disc ruptures occurred approx. 12 seconds after quench.
Quench vent system capable of 500-600 g/s.
Contributing factors;
Initial energy extraction estimate 50%-60% (actual 20%-25%)Recent coil temperature simulations (E. Ravaioli, LBNL) show significantly higher temps in the inner coil (direct contact with helium), 25-50 K higher.Rapid pressure rise through critical pressure to rupture pressure (approx. 3 seconds) does not allow enough time to transition to supercritical helium where energy transfer rate would be lower than film boiling.Spare burst disc tested (rupture @ 47
psia
)
This result provides good confidence that a pressure spike was not missed by the data collection and that the
dewar
pressure sensor was reasonably accurate.
5Slide6
Quench Summary
Quench
I
(A)
Energy into coil (MJ)Helium mass flow @ 2seconds after rupture(g/s)215440
2.96
1750
3
15916
3.20
2150
6Slide7
Quench #2 (11-13-2017)
Current at quench; 15440 Amps (3.93 MJ)Energy deposited into coil; 2.96 MJ ( 25% extraction)
Time from Quench to Burst disc rupture; approx. 13 s.
At time of burst disc rupture; rating 40 psig @ -320F
Estimated flow into quench tanks 500 g/s ( from gage pressure readings).1 ½” Anderson –Greenwood relief valve set at 30 psig never relieved.Calculated mass flow 2 seconds after ruptureThrough burst disk ( 1400g/s)Through quench system ( approx. 350 g/s)
7Slide8
Quench 2 Graph
8Slide9
Quench #3 (11-15-2017)
Current at quench; 15916 Amps (4.18 MJ)Energy deposited into coil; 3.20 MJ ( 23% extraction)
Time from Quench to Burst disc rupture; approx. 15 s.
At time of burst disc rupture; rating 40 psig @ -320F
Estimated flow into quench tanks 550 g/s ( from gage pressure readings).Anderson –Greenwood relief valve, set pressure reduced to 10 psig (valve actively relieving at time of rupture). Calculated flow 390g/s.Calculated mass flow 2 seconds after ruptureThrough burst disk ( 1650g/s)Through quench system ( approx. 500 g/s) 4 seconds after rupture flow rates match that of quench #2 , 2 seconds after rupture.
9Slide10
Quench 3 Graphs
10Slide11
Estimations
The best estimations would be from data already collected and that will be collected.
Extend fast data logger to capture the entire pressure rise and fall.
Mount
dewar pressure sensor as close to dewar top hat as possible.Add pressure sensor to relief stack.Add another pressure sensor to the quench line ( measured distance from an existing sensor) to better attain flow rate.Rough estimates can be made assuming film boiling at relief conditions, and initial coil temperatures from simulations.
Balancing heat flux through inner coil insulation, with heat flux of boiling helium at a given temperature can give an estimate.
Estimate at quench #3 parameters (comparison)
Estimate at 5.3MJ heat input ( 17900A no energy extraction).
11Slide12
Quench #3 Estimate
Helium
Coil insulation
.002”
Kapton & .011” Epoxy Fiberglass
Avg. Coil Temp 125K
q insulation= q helium boiling
)
from Breen &
Westwater
Correlation
Mass flow = 2850g/s
Courtesy E. Ravaioli
(Calculated uniform coil temp. 102.7K)
12Slide13
5.3 MJ Input Estimate (17,900 Amp no E.E.)
Helium
Coil insulation
.002”
Kapton & .011” Epoxy Fiberglass
Avg. Coil Temp 170K
q insulation= q helium boiling
)
from Breen &
Westwater
Correlation
Mass flow = 4100g/s
Courtesy E. Ravaioli
(Calculated uniform coil temp. 130.3K)
13Slide14
Bottom Line on Estimates
After a quench, dewar
conditions become very dynamic and transient.
Estimates should be conservative simplifications of this process.
Quench #3 Comparison;4 seconds. After burst disc rupture a sustained venting of about 1800g/s.At rupture; (Burst disk + vent + relief valve) = 3100 g/s, this provided rapid depressurization of the dewar.Calculation estimate 2850 g/s.5.3 MJ quench estimate ( no energy extraction) = 4100 g/s
Based on what has been shown so far;
Relief system should be designed for > 5000 g/s
Quench recovery piping should also be able to handle this flow rate
Recovery storage should be able to accommodate all helium contained in the
dewar
(long term).
14Slide15
Short Term SolutionsProposed Relief Stack
Redesign relief stack to protect burst disc.
Increase to piping to 3” IPS. ( from 2” IPS)
(1) Cash C-776 ( 2”) relief valve 25 psig Set point
Capacity 1380 g/s @ 30 psig.(1) Magnetrol 2” fast acting solenoid valve.Capacity 1400g/s @ 30 psig.(1)
Magnetrol
3” fast acting solenoid valve.
Capacity 2700g/s @ 30 psig.
Total relieving capacity 5480g/s @30 psig ( 6700g/s @ 40 psig)
3” Burst disk (40 psig)
Capacity 5750g/s
Support structure stress analysis is ongoing.
15Slide16
Solenoid Valve OperationOption #1 only (vent all helium to atmosphere).
Each valve independently controlled.
Both valves open on quench detection signal.
Closing of valves operator controlled.
For first quench valves will remain open for entire quench.Once pressure data can be reviewed, valves may be able to be closed when the flow is within quench recovery capability.With Option #2 (cold buffer storage)
Independent valve control.
Valves would open on pressure signal ( P. Joshi).
16Slide17
Vent Stack AnalysisEnvironment
3” Sol. Thrust (204 lbs.)
Burst disc thrust ( 291 lbs.)
2” sol. Thrust (96 lbs.)
Relief valve thrust (116 lbs.)
17Slide18
Vent Stack AnalysisWeights and Burst Disk T
hrust
18Slide19
Vent Stack AnalysisWeights and 3” Sol. & Burst Disc.
This condition yielded the highest stresses of all
valve combinations.
19Slide20
Short Term SolutionsCold Storage
Use Test Dewar 3# ( about 6 feet away) as cold helium storage.
Connect in series with test
dewar
with a VJ transfer line.Would add 2.2 cubic meters of cold helium storage.This would allow for about 30KG of additional cold helium storage (@ 30psig and 9 K)Liquid Helium contained in test dewar about 87kg.Most plumbing already set up.
Need transfer line between
dewars
and additional Transfer line to keep some liquid in the
dewar
.
20Slide21
Cold Storage
21Slide22
B.N.L. Safety Committee Review
Proposed solutions were presented to the B.N.L. Safety committee 12/15/17.Permission granted to go forward with options #1& 2.
Upon completion of manufacture and installation, final review of documentation with select members of the committee.
Granted permission is for (2) magnets( MQXFAP1 & MQXFAP2), another review will be needed for production magnets.
22Slide23
Long Term SolutionsIncrease Recovery Capability
Increase piping size to 6” to quench valve, then 10” out to recovery tanks.
Pressure drops for 150
ft
( including K-losses)@ 3 kg/s…..4.5psi@ 5 kg/s…..12 psiMove 30,000 gal. tank closer to test facility.Possibly add another existing tank (on site) to lower final tank pressure.
23Slide24
Summary
Now
Increase relief capacity to 5480 g/s @ 30 psig to avoid burst disk rupture.
Increase fast logger window to capture entire event.
Add second dewar as cold storage (8 weeks).Need transfer line between dewars.Can significantly reduce helium loss through relief system.
Long term
Collect data on quenches as they occur.
Upgrade recovery line to 10”
Will reduce pressure buildup in test
dewar
.
Move 30,000 gal tank closer to facility.
24Slide25
Backup Slides25Slide26
26Slide27
Film Boiling Correlation Linear curve fit.
27Slide28
Sample CalculationQuench #3 Estimate ( by iteration)
Page #1
28Slide29
Sample CalculationQuench #3 Estimate ( by iteration)
Page #2
29Slide30
Sample CalculationQuench #3 Estimate ( direct solve)
30