High Temperature PEM Fuel Cell MEA Manufacturing Advances OR Whats a Millisecond

1
High Temperature PEM Fuel Cell MEA Manufacturing
AdvancesORWhats a Millisecond?

• Raymond Puffer, PE
• (puffer_at_rpi.edu)
• Program Director, Industrial Automation
• Rensselaer Polytechnic Institute, Troy, NY
• http//www.cats.rpi.edu
• 3d MEA Manufacturing Symposium
• August 21-23, 2007
• Dayton, OH

2
Outline

• Introduce CATS
• Why fuel cell manufacturing as a focus
• Opportunities
• Challenge
• Risks
• What/why is a millisecond?
• Summary of fuel cell manufacturing projects
• DOE PEM Fuel Cell Manufacturing RD Solicitation
• NSF Fuel Cell IGERT at Rensselaer
• Conclusions
• Acknowledgements

3
Center for Automation Technologies Systems

• CATS is a Rensselaer research center founded in
  1988 based on a New York State block grant.
• Interdisciplinary environment with 40 affiliated
  faculty from 13 departments, 5 research technical
  staff, 1 business development staff.
• Research focal areas of Industrial Automation
  Smart Adaptive Optics and Autonomous Systems.
• Major program in fuel cell manufacturing

4
The CATS Focus

• The focus of our fuel cell manufacturing
  research is on fuel cell stacks, their materials
  and components, and the production and assembly
  thereof.

• Plug Power Prototype HT PEMFC 5KW system on
  display at 2006 Hanover Fair in Germany

Schematic of typical PEM fuel cell stack and
components (Woodman, 1999)
5
A Potentially Controversial Statement

• Fuel cells as we know them today will never
  achieve widespread use,

6
A Potentially Controversial Statement

• Fuel cells as we know them today will never
  achieve widespread use,
• UNLESS
• there are major reductions in MEA materials and
  manufacturing costs.

7
The Opportunity

• One simple example of the potential-
• Laptop Computers
• 4Q2006 sales of gt20M units, exceeding sales of
  desktop computers for first time
• 2007 sales projected at 91.7M units, and 137M
  units in 2010
• Assume a modest market penetration, say 20,
  thats still 27.4M stacks per year- from just one
  application, 548 Million MEAs
• Thats 52 stacks per minute on a 24/7/365 basis,
  and 17 MEAs per second 24/7/365

8
Another Example

• DOE target of 500,000 cars/year (only a 3 market
  penetration)
• That requires that one stack be assembled every
  minute on a 24/7/365 basis, 7 MEAs per second,
  250,000 m2 of GDE per day
• We simply cannot take a day or more to assemble
  an automotive fuel cell stack

Stack assembly unit process cycle times must be
measured in seconds! MEA unit processes in ms.
9
The Challenge

• The Fuel Cell Manufacturing Challenge- Any time
  you change one or more of the following you may
  have a profound impact on the viability of
  certain manufacturing processes and systems
• Fuel cell type
• Fuel cell or component architectures
• Materials
• Design tolerances
• Application
• Fuel cell size

10
The Risk

• If we do not aggressively pursue Research and
  Development of fuel cell manufacturing methods
  and systems we may well find ourselves in the
  position of leaders in the design and development
  of fuel cells, only to have the value added
  manufacturing performed off-shore.
• Our Approach- Employ modular, flexible
  manufacturing processes and systems.

11
In Order For The Industry To Succeed..

• We must develop the stack component materials,
  designs, architectures and manufacturing
  processes to totally eliminate the need for
  burn-in testing. Leak testing must be performed
  in seconds, if at all.

12
Stack Burn-in Testing- Example 1 Laptops

• Stack quantity- 27,400,000 stacks per year
• Test Duration-24 hours
• Test stand cost- 30K
• Test stand capacity- 4 stacks per stand
• Test stand floor space- 25 ft2
• Test facility size- 171,250,000ft2 (4,282 acres
  or 6.7 square miles)
• Equipment cost- 205,000,000,000

13
Stack Burn-in Testing- Example 2 Laptops

• Stack quantity- 27,400,000
• Test Duration- 1 hour
• Test stand cost- 30K
• Test stand capacity- 4 stacks per stand
• Test stand floor space- 25 ft2
• Test facility size- 7,135,416ft2 (178 acres)
• Equipment cost- 8,542,000,000

14
Stack Burn-in Testing- Example 3 Automotive

• Stack quantity- 500,000
• Test Duration- 24 hour
• Test stand cost- 50K
• Test stand capacity- 1 stack per stand
• Test stand floor space- 25 ft2
• Test facility size- 34,250ft2
• Equipment cost- 68,500,000

15
Examples of Critical MEA Manufacturing Cost
Drivers

• Lack of standardization of MEA geometries (active
  area)
• Hard tooling (becomes expensive paper weights
  when MEA design changes)
• Design tolerances
• Energy consumption
• Process yield
• Capacity yield

16
Fuel Cell Manufacturing Projects

• Development of a HT (i.e. PBI Based) PEM MEA
  pilot manufacturing line
• Energy efficient processes for the manufacture of
  PEM fuel cell MEAs
• Modeling, design and development of membrane
  forming techniques for Polybenzimidizole (PBI)
  based sol-gel membranes.
• Machine vision based inspection of HT PEM MEAs.
• Adaptive control of MEA manufacturing process.
• Automated assembly of fuel cell stacks

17
Development of a HT PEM MEA Pilot Manufacturing
Line
Pilot line concept design
PBI Based MEA
From Concept to Production
POPM
Pilot production line
18
Energy Efficient Manufacturing Processes for HT
MEAs

• Partners Progressive Machine and Design (Victor,
  NY) and BASF Fuel Cell
• Sponsor New York State Energy Research and
  Development Authority (NYSERDA)
• Objectives To investigate alternative
  manufacturing processes and systems that will
  save energy, reduce costs, and improve product
  quality

• CATS laser processing testbed
• 60 W, CO2, 9.3 µm laser
• 5W, DPSS UV, 355nm laser
• Precision linear stages
• Flying optics
• Servo positioned tooling

19
Energy Efficient Manufacturing Processes for HT
MEAs

• Goals for laser process Reduce process tooling
  cost by an order of magnitude, and reduce tooling
  lead time by 80

• UV laser photo-ablation of high temperature PEM
  membrane

• Goals for ultrasonic welding Reduce weld cycle
  time by 90, increase weld strength by a factor
  of 2, reduce energy consumption by 90,
  compatible with a broad range of MEA materials.

20
Energy Efficient Manufacturing Processes for HT
MEAs

• Status-
• Highly successful feasibility study and
  experimentation.
• Laser process capabilities and limitations
  defined.
• Commercial prototype laser system being built by
  PMD under follow-on NYSERDA project.
• Ultrasonic bonding process shows promise. Further
  process development work being performed.

21
High Temperature PEM Membrane Casting

• by
• Daniel Walczyk, PhD, PE
• Associate Professor
• Department of Mechanical, Aerospace, and Nuclear
  Engineering
• and
• Center for Automation Technologies and Systems

22
Current State of the Art Nafion Membrane
Manufacturing

• DuPont uses a scaled laboratory technique, i.e.
  doctor blade casting (Curtin et al., 2002)

• Almost all other documented casting systems,
  usually low viscosity dispersions, use a doctor
  blade to cast PEM membranes

23
Problems With Doctor Blade

• Solution is exposed to atmosphere.
• Potential for solution contamination.
• Temperature variations lead to viscosity
  variations, which may result in thickness
  variations.
• Not well suited to casting of high viscosity,
  high temperature solutions

24
High Temperature PEM Membrane Materials

• Example Celtec-P membrane from BASF Fuel Cell,
  GmbH consists of a polymer polybenzimidazole
  (PBI) and phosphoric acid as electrolyte
• Handling characteristics of this membrane
• Highly viscous
• Temperature- and shear strain rate-dependent
• Corrosive
• High adherence to surfaces
• Must be dispensed at high temperatures
• Susceptible to environmental conditions

25
System Concept Casting Tool Investigation

• Gravity-Fed, Single Blade Film Casting Method
  works, however solution is exposed to critical
  environmental factors (i.e. air)
• Reverse Roll Extrusion - Solution stickiness
  doesnt allow for uniform separation between
  rollers
• Stencil Printing with Slot Die Precise casting
  possible, but shrinkage of membranes after
  hydrolyzation leads to distorted geometry
• Stencil Printing onto GDL Material is not
  easily stenciled and shrinkage during hydrolysis
  causes curling of GDL. However, mechanical
  adhesion to GDL is good and curling may be
  controllable (e.g., GDL-membrane-GDL sandwich)

26
Modeling, Design and Development of Membrane
Forming Techniques for PBI Membranes.
Empirical Viscosity (?) Model Carreau Yasuda
Time Temp. Superposition
Dynamic Viscosity Reduced Viscosity Empirical
Viscosity Curve
27
Membrane Characterization
Temperature-controlled parallel plate rheometer
with nitrogen atmosphere
28
Membrane Casting Slot Die Design
Design in CAD
Simulation of pressure and velocity in die (CFD)
Prototype die on casting line
Model Validation
29
Proposed System Concepts
30
Casting Tool Investigation 10 cm Slot Die

• Desired
• Film thickness 400 µm
• Velocity 1.5 mm/s
• Temp 120 C
• Input Pressure 2 bar
• PPBI membrane of IV 4.0

• Experimental results
• Pressure drop 2 bar
• Film thickness 39010 µm
• 98 uniformity across the width
• Slot Die Extrusion Uniform, defect free film
  are realizable if all process parameters are well
  controlled (i.e. substrate temperature)

31
DOE PEM Fuel Cell Manufacturing RD Solicitation

• Manufacturing Research and Development for
  Hydrogen and Fuel Cell Systems
• Funding Opportunity Number DE-PS36-07GO97012
• Announcement Type Initial
• CFDA Number 81.087 Renewable Energy Research and
  Development
• Issue Date 07/19/2007
• Application Due Date 10/10/2007, 1159 PM
  Eastern Time
• See http//www1.eere.energy.gov/hydrogenandfuelcel
  ls

32
DOE Solicitation Topics

• Topic 1 Alternative Electrode Deposition
  Processes
• Topic 2 Gas Diffusion Layer Fabrication
• Topic 3 Novel MEA Manufacturing
• Topic 4 Process Modeling for Fuel Cell Stacks
• Topic 5 Process and Device for Cost Effective
  Testing of Cell Stacks
• Topic 6 Manufacturing Technologies for High
  Pressure Composite Tanks

33
Integrated Graduate Education and Research
Traineeship (IGERT)

• Rensselaer has been awarded an NSF IGERT program
  in Fuel Cells.
• Five year program valued at up to 7M.
• 20 faculty staff
• 6 academic units
• 5 research centers
• Support for 28 PhDs over five years. 19 selected
  to date.
• Four new graduate level courses
• Research in Materials, Manufacturing Modeling,
  with an emphasis on entrepreneurship
• All students serve an internship in industry or
  National Lab.
• Support for overseas research.
• U of PR (Mayaguez), NYSERDA, National Labs,
  Industry

34
Conclusions

• We cannot wait until we know all the answers to
  address key fuel cell manufacturing issues.
• There will be a technology tipping point that
  will result in an exponential growth of demand.
• To minimize risks employ modular, flexible
  manufacturing processes and systems.
• Major advances are required to make fuel cells
  viable on a wide-spread basis.
• We need to change how we prepare scientists and
  engineers to enter the fuel cell industry.

35
Acknowledgements

• BASF Fuel Cell, GmbH
• Progressive Machine and Design
• NYS Energy Research and Development Authority
• National Science Foundation
• NYS Office of Science, Technology and Academic
  Research
• Robotics Industries Association
• Kuka Robotics
• Applied Robotics