Analytical Technique for Composite Storage Tank Design Factor

Analytical Technique for Composite Storage Tank Design Factor Specification NHA Annual Hydrogen Conference 2008 March 30 - April 4, Sacramento, CA Vladimir M. Shkolnikov, Ph. D. , CTC Kevin L. Klug, Ph. D. , CTC 1

AGENDA Ø COMPOSITE OVERWRAPPED PRESSURE VESSELS Ø DEVELOPMENTAL GOALS & REQUIREMENTS Ø ACCOMPLISHMENTS TO DATE Ø CONCERNING ISSUES Ø REQUIREMENTS VARIANCE Ø DETERIORATIVE FACTORS OF OPERATION Ø DESIGN STRENGTH RECONCILIATION TECHNIQUES Ø EXEMPLIFIED SERVICEABILITY ASSESSMENT Ø DESIGN FACTOR SPECIFICATION

Composite Overwrapped Pressure Vessels (COPV) COPV Type 2, metal-lined composite reinforced cylinder with load-sharing liner that alone can withstand the operating pressure, normally termed hoop-wrapped. COPV Type 3, metal-lined composite reinforced cylinder with load-sharing liner that alone cannot resist the operating pressure, normally termed full 3 wrapped.

DOE Goals & Requirements Develop and demonstrate viable hydrogen storage technologies for on-board and off-board applications Ø By 2010: Develop and verify hydrogen storage systems achieving 2 k. Wh/kg (6 wt%), 1. 5 k. Wh/L, $4/k. Wh, and 30 g/L; storage tank purchased capital cost $500/kg H 2; ambient -30/50ºC + full solar load; min/max delivery temperature -40/85ºC; life (1/4 tank to full) 1000 cycles Ø By 2015: Develop and verify hydrogen storage systems achieving 3 k. Wh/kg (9 wt%), 2. 7 k. Wh/L, $2/k. Wh, and >35 g/L; storage tank purchased capital cost $300/kg H 2; ambient -40/60ºC + full solar load; min/max delivery temperature -40/85ºC; life (1/4 tank to full) 1500 cycles.

COPV Type 3 Designed and produced by Hy. Per. Comp Engineering, Inc. (Brigham City, UT) for DOE project with CTC COPV-3 prototype with aluminum liner for on-board application • Capacity: 7. 75 L (0. 30 kg H 2) • Mean burst test pressure: 25. 4 Ksi • Intended service pressure: 10. 0 Ksi • Hydrogen weight efficiency: 5. 2% (6 wt% for 2010) • Hydrogen volumetric efficiency: 38. 7 g/L (30. 0 g/L for 2010) • Estimated cost: $4, 700/kg H 2 ($500/kg H 2 for 2010). 5

COPV Type 2 Designed and produced by Hy. Per. Comp Engineering, Inc. (Brigham City, UT) for DOE project with CTC COPV-2 prototype with steel liner for off-board application • Capacity: 15. 1 L (0. 44 kg H 2) • Mean burst test pressure: 15. 5 Ksi • Intended service pressure: 6. 2 Ksi • Hydrogen weight efficiency: 2. 4% (6 wt% for 2010) • Hydrogen volumetric efficiency: 29. 2 g/L (30. 0 g/L for 2010) • Estimated cost: $641/kg H 2 ($500/kg H 2 for 2010) 6

Concerning Issues Ø The DOE goals and requirements, while presenting good reference points for COPV advancement, are incomplete in terms of design criteria for verification of long-term COPV capability Ø Polymer Matrix Composites (PMC) typically exhibit enhanced sensitivity to time-variable parameters of force - temperature exposure that should be taken into account Ø Existing standards / codes don’t provide design guidance sufficient to characterize serviceability and/or properly select a design factor for COPVs subjected to changeable operational exposure. 7

Requirements Variance Criteria Material CGA [1] ISO [2] Stress ratios: Aramid FRP 2. 0 (3. 0) 2. 1 (3. 51) Burst/test Carbon FRP 1. 5 (2. 25) 2. 0 (3. 34) DOT [3] (Burst/service) Service life Glass FRP 2. 33 (3. 5) 2. 04 (3. 4) 2. 4 (4. 01) 30 years 30 + years 15 years 1. CGA C -19 FRP-3. – Guideline for Filament-Wound Composite Cylinders with Nonloadsharing Liners, Second Edition. Compressed Gas Association, Inc. , 2002 2. ISO 11119 -2: Gas Cylinders of Composite Construction Specification and Test Methods - Part 2: Fully Wrapped Fiber Reinforced Composite Gas Cylinders With Load-sharing Metal Liners, 2002 3. DOT-CFFC. Appendix A. Basic Requirements for Fully Wrapped Carbon-Fiber Reinforced Aluminum Lined Cylinders. 2000. 8

Principal Objective Ø Improve methodology for serviceability * characterization and specification of design factors to ensure proper long-term performance of COPVs. * Prerequisite quality of service accomplished within a specified timeframe. 9

Deteriorative Factors of Operation Storage tanks will be employed for assorted engineering systems of Hydrogen Infrastructure including: Ø Over-the-road bulk transportation Ø Stationary accumulators for gas generation and/or collection sites Ø Stationary storage forecourt dispensing sites Ø As an end-user’s vehicular fuel tank. These applications imply a variety of duty cycles and loading factors which will affect long-term structural performance of composite wrap jacket of COPV. 10

Deteriorative Factors of Operation * A fabrication technique in which COPV is subjected to high pressure, causing the metal liner to yield and resulting in its compressive residual stresses. The goal of autofrettage is to increase the liner’s durability. 11

Temperature Factor Possible Effects Temperature may affect the composite jacket of COPV in several ways: • As a direct deteriorative factor of thermal-mechanical fatigue § Via increase of pressure of the contained gaseous hydrogen § Inducing additional interface pressure between the jacket and load-sharing liner due to difference in thermal expansion of the jacket and liner materials. 12

Temperature Factor Outside Ambient Temperature Off-board storage tanks can sit at any climatic zone either above ground, on, or underground. This entails a wide variety of outside ambient temperatures during operation as no temperature control is present. Filling stations in North America 13

Temperature Factor Pressure-Temperature-Density Interdependency* H 2 density vs. pressure & temperature Pressure increase due to H 2 heating * - NIST Standard Reference Database, www. nist. gov/srd/index. htm

Temperature Factor Fast Filling • Warming up The approximation is based on experimental data* regarding pressure & temperature changing over filling time and is to be utilized within COPV serviceability assessment. Approximated P, T = f (time) * - Eihusen, J. A. , Application of Plastic-Lined Composite pressure vessels for Hydrogen Storage, GD Armament and Technical Products, Lincoln, NE 15

Temperature Factor Thermal Expansion Possible outcomes The difference in thermal expansion may affect jacket-liner interface pressure either increasing or decreasing the jacket stressing and resulting respectively with: 1. Liner’s elastic instability and 2. Reduction of residual after- autofrettage pressure possibly affecting the interface bonding (when the tank is empty). Thermal expansion vs. temperature* Either case lowers liner’s functionality. * - Marquardt, E. D. , Le, J. P. , and Radebaugh, R. , Cryogenic Material Properties Database, NIST, The 11 th International Cryocooler Conference, Keystone, CO, June 20 -22, 2000.

Structural Design Reconciliation Conventional Technique Allowed stress corresponds to stress ratio specified in COPV Design Guidance Knock - down coefficients ki represent partial deteriorative influence of operational loading factors: repetitive/sustained Force (pressure), elevated Temperature, Humidity, Aggressive environment, and UV irradiation among possible others Safety factor is a function of design uncertainties: actual Operational conditions, imperfection of utilized computer Model, and scattering of Properties of a real material system utilized within COPV structure • Guide for Building and Classing Naval Vessels, Part 1: Hull and Structures, Chapter 4: Composites, ABS, July 15, 2004 • Smirnova, M. K. , Paliy, O. M. , Spiro, V. E. , Principles of Strength Criteria Specification for Ship Structures of PMC, Mechanics of Polymer Materials, Riga, Latvia, March, 1984, p. 882 -887.

Structural Design Reconciliation Advanced Technique Durability under assorted loading: for Pijk fraction of service exposure to ijk loading conditions (i - category of operational exposure regarding j - load case under k - temperature) Durability under ijk exposure in terms of Kinetic Theory of Fracture Accumulated micro-damage (deterioration) with Bailey’s Integral

Serviceability Assessment Technique Experimental Verification Actual loading: • Single loading rates 0. 3… 13. 0 MPa/s (43. 5… 885 psi/s) • Cycling frequencies 0. 03 … 0. 50 Hz • Maximal cyclic stress (0. 6 … 0. 8) SU • Temperature 19 … 30°C (66 … 86°F) Base-line loading: • • Fatigue data actual and standardized Cycling frequency 0. 1 Hz Stress range 0. 1 • Temperature 20°C (68°F) * - Lavrov A. V. and Shkolnikov, V. M. , Experimental Research of Fatigue of Thickwalled PMC Structures, Structural Application of PMC, Russian Society of Naval 19 Engineers, v. 510, St. Petersburg, Russia: 1991, p. 4 -15.

Serviceability Assessment Technique Influence of Loading Waveform Typical cyclic loading waveforms Glass FRP fatigue performance under • Cycling frequency 0. 1 Hz • Stress range 0. 1 20

Serviceability Assessment Technique Single (Burst Pressure) Loading Evaluation Burst pressure in function of loading profile & rate 21

Serviceability Assessment Exemplified Operational Profiles * Per CGA C-19 – 2002. FRP-3 – Guideline ** Once per week *** Fraction of service life (%) except as otherwise noted 22

Serviceability Assessment Prorated Influence of Service Loading Severe loading conditions

Serviceability Assessment Prorated Influence of Service Loading Mild loading conditions

Serviceability Assessment Prorated Influence of Test Loading * - Autofrettage pressure is included into all loading profiles and is referred as Load case #1 Per Test Protocol of CGA C-19 – 2002. FRP-3 – Guideline

Serviceability Assessment Prorated Influence of Test Loading Test #1 26

Serviceability Assessment Prorated Influence of Test Loading Test #3

Serviceability Assessment Summary * Continuous service with stress ratio δP = 3. 5 ** To endure 30 -year continuous service 28

Serviceability Assessment Summary (continued) Ø Loading parameters significantly affect length of COPV service life and should be taken into account Ø Notable mismatch between computed and specified stress ratios substantiates the need for improvement of the design methodology Ø The nomenclature of design qualification tests should be selected corresponding to assigned COPV operation Ø Not being reconciled against actual / assigned operational conditions COPV’s serviceability might be either greater or lower than required. 29

Design Factor Specification Ø It is possible to specify the knock-down factors on a pro-rata basis employing the proposed serviceability assessment technique Ø Safety factor should be added to cover the multiple uncertainties in loading parameters, PMC properties, and COPV computer modeling Ø Safety margin should be aligned with a selected loading profile; when this is close to actual or anticipated loading conditions, the safety margin might be reduced. 30

Conclusion Remarks q The outlined technique meets the demand is applicable regarding both operation and testing q The technique is also capable to specify a testing protocol adequately to the required COPV durability and to support COPV health monitoring in service q Dependable data on fatigue parameters of employed structural materials are needed to fulfill the serviceability assessments q The technique will somewhat complicate the conventional design reconciliation procedure; however, this extra effort will be paid off with increased performance predictability and ensured operational safety.

ACKNOWLEDGEMENTS The presented work is performed under R&D project sponsored by the U. S. DOE, award # DEFC 36 -04 GO 14229. The sponsorship and guidance for this project provided by Monterey Gardiner and Paul Bakke of the DOE are gratefully acknowledged. The presented work has benefited from support of David K. Moyer and Eileen M. Schmura of CTC, former and current managers of the project, respectively. 32

Questions / Comments? CONTACT INFO: Dr. Vladimir M. Shkolnikov E-mail: shkolniv@ctc. com Tel: (814) 269 -2639 Dr. Kevin L. Klug E-mail: klugk@ctc. com Tel: (910) 437 -9904 Thank you 33

Back-up 34

Project Relevance Hydrogen Regional Infrastructure Program in Pennsylvania Hydrogen Fuel Cells Infrastructure and Technology (HFCIT) Program Multi-Year Research, Development and Demonstration Plan (MYRD&DP) 35

COPV-3 Prototype The selected COPV-3 has an aluminum (6061 -T 6) liner fully wrapped by fiber/epoxy PMC. The intermediate modulus Toray T 1000 12 K carbon fiber is chosen for PMC wrapping jacket. 36

COPV-2 Prototype Commercial grade Toray T 700 12 K and Epon 828, an epoxy resin were selected for the PMC jacket. High strength chromium molybdenum steel (34 CRM 04) alloy is the liner’s material.