DOE Tank Safety Workshop

Hydrogen Tank Safety Testing

04.29.2010 | Presented by Joe Wong, [Link].

 POWERTECH – Hydrogen & CNG Services

 Certification testing of individual high pressure components

 Design Verification, Performance, End-of-Life testing of
complete fuel systems

 Design, construction, and operation of Hydrogen Fill Stations

 Safety Studies
 Standards Development

1
 PRESENTATION

 Discuss CNG Field Performance Data

 Discuss Safety Testing of Type 4 Tanks
 Current work to support Codes & Standards
Development

2
 Storage Tank Technologies

4 basic types of tank

designs
 Type 1 – all metal
 Type 2 – metal liner with
hoop wrapped composite
 Type 3 – metal liner with
fully wrapped composite
 Type 4 – Plastic liner with
fully wrapped composite

3
 Tank Designs in Hydrogen Service
 Primarily use composite tanks for hydrogen fuel cell vehicles
 250 bar carbon fiber reinforced tank design in fuel cell bus
demonstration in 1994.
 Storage pressures increased to 350 bar in 2000
 Today, most auto OEMs have 700 bar tanks for on-board
storage
 500 km range with 5kg H2

1994 Ballard Fuel Cell Bus

4
 CNG Experience In-service Failures

 Powertech has been testing CNG storage systems

since 1983
 Powertech has maintained a cylinder failure
database through world wide contacts
 Examined CNG cylinder field failure database to
determine if trends evident
 Limited to incidents involving catastrophic rupture of
cylinders, although major leaks attributed solely to
the cylinder were included
 From 2000-2008, there were 26 CNG cylinder
failures.
 Other multiple cylinder failures attributable to
“leakage failure mode”:
 Type 1 steel pinhole leaks (<50)
 Type 4 plastic liner leak incidents (100’s)

5
 FAILURE INCIDENTS REPORTED BY FAILURE CAUSE

Data classified according to unique failure

causes:
18  Mechanical Damage – External
16
16 abrasion and/or impact
 Environmental Damage – External
14
environment assisted, typically SCC
12
No. of Failure Incidents

12
 Overpressure – Faulty fueling
10 equipment or faulty CNG cylinder
8 valves
8
6
 Vehicle fire – Faulty PRDs or lack of
6 5 PRDs; localized fires
4 3 3  Plastic Liner Issues – Man. defects
2
incl. cracking at end boss/liner
1
interface, flawed welds, liner seal
0
Mechanical Environmental Overpressure Vehicle Fire Plastic Liner Metal Liner User Error Unknown
failures
Damage Damage Issues Man. Issues Cause  Metal Liner Issues – Man. defects
Failure Cause
incl. pinhole leaks, laminations, poor
heat treat practice

6
 Why use Type 4 cylinders?

 Light weight
 Lower cost than seamless metal liners
 Less susceptible to fatigue cracks
 High toughness and elongation of liner material
 Low capital cost for manufacturing
 Capable of large diameters
 Ultra high pressure (1000bar)

7
 Large Volume Type 4 Storage

 38 feet long
tanks (11.6M)
 5300 lbs
(2,400 kg)
 42 inches
diameter
 Holds 10,000
SCM methane

8
 Type 4 In-service Failures - leaks

Learnings from CNG in-service experience

 Type 4 plastic liner leak incidents (100’s)
 Some 12,000 type 4 cylinders manufactured by 1 company ( no
longer in business)
 SWRI inspected a large number of tanks in 1997
Fleet location No of tanks Total Leaks % leaks
Salt Lake City 237 13 5.5
Las Vegas 254 2 0.8
Tacoma 492 3 0.6
Sacramento 360 4 1.1
New York 310 19 6.1
Total 1653 41 2.5

9
 Plastic Liner Issues

 The long term integrity of the connection between the plastic liner
and the metal end boss
 Different designs: mechanical connection, O ring seal, adhesive seal
 Different liner materials
 Aging effects on the plastic liner due to extreme temperatures
 Welding of plastic liners
 Permeations issues
 Liner buckling
 Static Discharge

10
 Liner Buckling
 Trapped gas between the space between the liner and the
composite
 Liner buckles inward when the tank is depressurized
 Causes fatigue cracks and mechanical damage to the end boss
interface

11
 CNG Cycle Test
 Closed loop gas cycle test
 Cracks in liner
 Poor end boss design
 Static discharge

12
 Electrostatic Discharge
 Measurement of static charge in the surface of plastic pipe
 Flowing CNG with contaminants causes high static charge
 Pinhole puncture through 6 mm pipe
 No substantial static charge buildup with hydrogen – clean fuel

13
 Standards Development

CNG Standards developed from in-service experience

 Vehicle service conditions
 End user requirements
 In-service failures / known failure mechanisms
 In-service abuse
 Collision
 Manufacturing problems
 Design problems

14
 Standard Tests for Design Qualification
Performance tests were designed and validated including:

 Ambient Cycling Test

 Environmental Test
 Extreme Temperature Pressure Test
 Hydrostatic Burst Test
 Composite Flaw Test
 Drop Test
 Accelerated Stress Rupture Test
 Permeation Test
 Hydrogen Cycling Test
 Bonfire test
 Gunfire Penetration Test

15
 Tank Testing - Hydraulic Pressure

Environmental and
chemical effects Flaw/Damage Tolerance

Powertech Cylinder Test Facilities

 Hydraulic pressure cycling up to
1,500 bar
Drop Test

16
Extreme temperature cycle test
 Tank Testing - Burst Test

Burst testing up to 2,800 bar

17
 Current Work to Support Standards
Development

 Hydrogen test validation for SAE J2579

 Fueling protocol testing for SAE J2601
 Crash Integrity of Tanks
 Localized fire testing

18
 SAE J2579 Program Overview

Powertech performed tests specified by the SAE Safety Working

Group for the purpose of validating the SAE J2579
requirements for storage of gaseous hydrogen on passenger
vehicles.

Contract by the National Renewable Energy Laboratory (NREL),

working with the Society of Automotive Engineers International
(SAE).

19
 Program Overview

Tests to validate SAE J2579 composed of three parts:

I. Verify that the vehicle storage validation tests specified in J2579

can be performed by a test facility

II. Verify that vehicle storage systems that have failed in past vehicle
service would not pass the J2579 tests

III. Verify that vehicle storage systems that have not failed in past
vehicle service will either: 1) pass the J2579 tests or, 2) fail the
J2579 tests only when the reasons for failure are understood and
would be expected to occur in vehicle service.

20
 Validation of Test Sequence
Safety Concerns

Powertech SAE J2579 Test Setup Safety Considerations

21
 Validation of Test Sequence
Test Equipment Requirements

Powertech has designed and built 2 parallel test setups for

the pneumatic sequence specified in SAE J2579

Each test setup consists of the following equipment:

1. Environmental chamber for the test tank
2. Hydrogen gas pre-cooler
3. Hydrogen gas flow control system (inlet & outlet)
4. Hydrogen compression
5. High-pressure hydrogen storage (88MPa)
6. Low pressure hydrogen storage (1MPa)

22
 Validation of Test Sequence
Safety Concerns

The potential for tank rupture or high-volume hydrogen release

must be accounted for in test setup
Powertech has designed & validated a safety system incorporated
into the SAE J2579 test setup
The Powertech safety system will:
1. Allow excessive hydrogen leakage to be safely vented
2. Contain a hydrogen ignition/detonation
3. Not contain a tank rupture
As an added safety measure, all vehicle fuel systems under test
must meet the requirements of the hydraulic test sequence
prior to undergoing the pneumatic test sequence

23
 Localized Fire Testing –
Purpose of NHTSA Program
• Since year 2000, leading cause of CNG cylinder failures is vehicle fire,
and single leading cause of vehicle fire failures is localized fire effects

• Objective is to verify effectiveness of a localized flame test developed

previously in a Transport Canada study
• Objective to be achieved by the “…evaluation of various fire protection
technologies that will reduce the risk of cylinder failure during a vehicle
fire”
• The localized fire test developed for Transport Canada involved meeting several
precise time and temperature criteria occurring on a tank surface as defined by an
OEM, and was found not to be adaptable to evaluating the performance of various
fire protection technologies
• A more versatile flame impingement test was developed based on vehicle fire data
• Maximum temperature exceeding 900ºC
• Duration of 30 minutes (duration of tire fire)
• Fire length that is 25% of the length covered by a standard 1.65m fire
24
 Thin Thermal Wrap Material –
Fire Test

25
 200 bar Fire Test –
Heat damage to intumescent epoxy after 30 minutes at 1000ºC

26
 Localized Fire Test Programs – Conclusions

• There are:
• Protective coating and wrap systems that work
• Remote fire detection systems that work

• Protective systems are available that are:

• Cost-effective
• Minimal added weight
• Minimal added wall thickness

• Systems can protect against fires even more localized, and

even longer duration, than the fire source used in the
testing program
27
 Type 4 Composite Tank Collision Damage

Tanks mounted on CNG bus roof. The Tanks mounted on CNG bus roof. The
bus impacted a low overhead, bus impacted a low overhead,
collapsing the roof. Tank still collapsing the roof. Tank punctured,
exceeded minimum burst pressure. released gas but did not rupture.

28
 Hydraulic Crush Test (150,000 kgf)

Used hydraulic ram to

attempt crush of
pressurized hydrogen
tank

Test ended at 150,000

kgf when reinforced
concrete wall on
opposite side of ram
broke

29
 2 Ton Drop Impact on Pressurized Tanks

30
 2 Ton Impact on Tank

 Battelle Program funded by NHTSA

Type 3 and Type 4 tanks
 vertical and horizontal impacts
 350 and 700 bar tanks

31
 Multi-Client 70 MPa Hydrogen Fast Fill Study

Outputs of the study to support J2601:

 Minimum fueling time at each ambient condition to safely fill all fuel
systems
 Pre-cooling levels for each ambient condition
 Energy required for pre-cooling
 Temperature gradients throughout the fuel system
 Durability of fuel system under extreme fueling conditions
 Performance data of station components (flowmeter, flow controller,
nozzles, hoses, compressors, etc.)

Consortium members: Air Liquide, BP, Nippon Oil, Sandia (US DOE),
Shell, Iwatani, Chrysler, Ford, GM, Nissan, Honda, Toyota.

32
 70 MPa Hydrogen Fast Fill Test Facility

Fueling Station Simulator

Fuel System Chamber

-40C to +50C

Ground Storage Chamber

875 bar -40C to +50C

33
 Multi-Client 70 MPa Fast Fill Study
OEM-1 Fuel System

34
 SUMMARY

 In-service experience with CNG tanks have provided input

into the development of CNG & Hydrogen tank standards
 Studies are underway to provide data to standards being
developed by organizations such as SAE, ISO, and CSA
including:
 Tank Safety
 Fire safety
 Fueling protocol
 Impact resistance

35