Advanced SOFC Development at Redox Power Systems 04/30/2019 1:45 - - PowerPoint PPT Presentation

advanced sofc development at redox power
SMART_READER_LITE
LIVE PREVIEW

Advanced SOFC Development at Redox Power Systems 04/30/2019 1:45 - - PowerPoint PPT Presentation

Advanced SOFC Development at Redox Power Systems 04/30/2019 1:45 pm 2019 Hydrogen and Fuel Cells AMR Crystal City, VA Redox Key Contributors: Sean R. Bishop, Bryan Blackburn, Luis Correa, Colin Gore, Stelu Deaconu, Ke-ji Pan, Johanna


slide-1
SLIDE 1

Advanced SOFC Development at Redox Power Systems

04/30/2019 1:45 pm 2019 Hydrogen and Fuel Cells AMR – Crystal City, VA Redox Key Contributors: Sean R. Bishop, Bryan Blackburn, Luis Correa, Colin Gore, Stelu Deaconu, Ke-ji Pan, Johanna Hartmann, Yue Li, Lei Wang

REDOX POWER SYSTEMS, LLC 1 4/30/2019

slide-2
SLIDE 2

Outline

4/30/2019 REDOX POWER SYSTEMS, LLC 2

  • 1. High power, low cost solid oxide fuel cell (SOFC) stacks for robust

and reliable distributed generation

  • 2. Red-ox robust SOFC stacks for affordable, reliable distributed

generation power systems

  • 3. High throughput, in-line coating metrology development for SOFC

manufacturing

  • 4. Sputtered thin films for very high power, efficient, and low-cost

commercial SOFCs

slide-3
SLIDE 3
  • 1. High Power SOFC Stacks

REDOX POWER SYSTEMS, LLC

  • We are currently working

towards a 2.5 kW stack demo

  • Two “lab reformers” qualified

for > 2.5 kW

3 4/30/2019

slide-4
SLIDE 4

Natural Gas Test Facility (NGTF)

REDOX POWER SYSTEMS, LLC 4

  • Moved into new demo facility in early 2019 that is 3x larger than previous location
  • Will allow additional stack and system testing
  • Large natural gas feed capacity for a larger gas-powered reformer capable of supporting 5-6 kWe

stacks and bringing the total reforming capacity to >15 kWe.

  • Light manufacturing and engineering space as well

4/30/2019

slide-5
SLIDE 5
  • 2. Red-ox Robust Stacks

REDOX POWER SYSTEMS, LLC 5

Journal of Power Sources 195 (2010) 5452–5467

Red-ox cycles can be expected during long-term fuel cell operation

  • Interruptions in fuel supply
  • Transient SOFC operation (e.g., shutdown)

Ni-cermet anodes prone to mechanical failure during redox cycling ~69 vol% expansion of Ni → NiO Solution: All ceramic anode → small oxygen = small dimensional change (0.4 vol%)

Linear Expansion [%]

650 oC 0.4 vol% No cracks after 9 redox cycles!

4/30/2019

slide-6
SLIDE 6

All-Ceramic Anode Performance

REDOX POWER SYSTEMS, LLC 6

  • High power densities
  • ~0.75 W/cm2 @ 550°C
  • ~0.3 W/cm2 @ 450 °C
  • Acceptable electronic conductivity

Button cell data Anode electrical conductivity

4/30/2019

Red-ox Cycles: 5 cm by 5 cm cell (600 °C)

H2 on anode N2

* on

anode

slide-7
SLIDE 7

Red-Ox Cycling of Stack

REDOX POWER SYSTEMS, LLC 7

Before Redox cycling After Redox cycling Before Redox cycling After Redox cycling

144.7 W 130 W 10 cm x 10 cm stack - cycling between hydrogen and nitrogen at 600 oC

  • Some degradation in performance after red-ox cycling
  • Previous 5 cm x 5 cm tests showed 3 red-ox cycles with minimal ASR, OCV,

and seal degradation, but more cycles led to degradation

  • Future work includes continued anode structure modification

4/30/2019

slide-8
SLIDE 8

Redox can Cycle!

4/30/2019 REDOX POWER SYSTEMS, LLC 8

slide-9
SLIDE 9

$0 $50 $100 $150 $200 $250 $300 $350 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000

Cumulative Discounted Cost ($) Time (hours)

Discrete Event Simulator

NG pressure boost Reformer Water boiler Stack(s) Condenser Air Blower HX HX DC power transformer Inverter Hot pipes/valves Cold pipes/valves Electrical Cable HX: Heat exchanger Fuel Processing SOFC Operation Power conditioning and system controls Control System

Schematic of system design approximation Model output

  • Initial deployment and stack

replacements largest cost components in initial model

  • Stack replacements include failure

due to “critical events”

  • Future work includes improving

estimates of MTTFs, costs, and model utility Cost from failures on multiple installations

REDOX POWER SYSTEMS, LLC 9 4/30/2019

slide-10
SLIDE 10

4/30/2019 REDOX POWER SYSTEMS, LLC 10

$100,000 $300,000 $500,000 $700,000 $900,000 $1,100,000 $1,300,000 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000

Cumulative Discounted Cost ($) Time (hours)

$100,000 $300,000 $500,000 $700,000 $900,000 $1,100,000 $1,300,000 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000

Cumulative Discounted Cost ($) Time (hours)

Standard, no gas-backup Red-ox tolerant or gas-backup

  • Largest cost in lifetime ownership from replacing stacks every time gas

emergency shut-down occurs (even though they are fairly rare)

  • Red-ox tolerance or gas back-up system dramatically reduces lifetime cost

Manuscript in prep.

Discrete Event Simulator

Comparison of a back-up fuel gas system (standard system) and a red-ox tolerant system

slide-11
SLIDE 11
  • 3. Metrology for SOFC Coating Manufacture

Protective coating applied to the interconnect surface:

  • Barrier to Cr transport from the interconnect to the electrode (prevent cathode poisoning)
  • Barrier of inward oxygen migration to the interconnect (block resistive oxide film growth)

(Mn,Co)O4 (MCO) is a commonly used barrier coating layer

Defects in coating (e.g., porosity, cracks) inhibit coating and SOFC performance

Coating cross-section Coating surface

PNNL report ID: PNNL- 17568, May 2008 ECS Transactions, v. 68, i. 1 (2015) 1569

4/30/2019 REDOX POWER SYSTEMS, LLC 11

slide-12
SLIDE 12

Key Defects of Interest Rating

Defect Challenges it presents Likelihood of

  • ccurrence (1-5)

Severity (1-5) Level of focus (1-5) Surface dips and/or bumps Could be high ASR spots, Cr volatility 5 3 5 Thickness non- uniformity, >50% Large gradients --> variations in ASR and ability to block Cr transport, (growth of Cr oxide layer - > ASR) 4 3 4 Sample-to-sample loading variations Similar to thickness non-uniformity above (measurable by mass gain) 2 3 3 Variations in film porosity Same as above 2 3 4 Film delamination (initial) Huge ASR, Increase in Cr volatility 1 5 1 Film delamination (during operation) Huge increase in ASR, Increase in Cr volaility 1 5 2 Small Roughness, bumps, dips, scratches in substrate possible non-uniform coatings 4 2 4 Large roughness/defects in substrate non-uniform coating 1 5 1 Small scratches in film due to handling breaches in film (most likely to occur in green film) 2 5 4 mud-cracks in film breaches in film 2 4 3

4/30/2019 REDOX POWER SYSTEMS, LLC 12

slide-13
SLIDE 13

Metrology of Key Defects Approach

Measurement methods

  • Optical microscopy
  • Optical profilometry
  • Thermography

4/30/2019 REDOX POWER SYSTEMS, LLC 13

Thermography in collaboration with NREL Derek Jacobsen, Peter Rupnowski, Brian Green, and Michael Ulsh

slide-14
SLIDE 14

Coating Fabrication at Redox

  • Sprayed MCO coatings followed by typical annealing methods (reducing

atmosphere followed by oxidation to achieve oxide coating) SEM cross-section of an MCO coating

  • n stainless steel developed at Redox

4/30/2019 REDOX POWER SYSTEMS, LLC 14

slide-15
SLIDE 15

Optical imaging detects porosity and thin intentional defects

  • Stainless steel substrate with intentionally added porosity or thin coating deposition
  • Optical imaging detects more inhomogeneities in thin as compared to “defect-free”

coating

  • Optical profile detects roughness change of porous > ”defect-free” > thin coatings

Optical microscopy (grid is an image stitching artifact) Optical profilometry

4/30/2019 REDOX POWER SYSTEMS, LLC 15

slide-16
SLIDE 16

Thermography Detects Substrate Scratches

Intentionally scratched substrate with MCO coating

  • 4 scratches in stainless steel substrate
  • Optical and height profile mapping can only detect two scratches in fired film
  • Thermography detects all 4 scratches!

4/30/2019 REDOX POWER SYSTEMS, LLC 16

slide-17
SLIDE 17

Trends observed in thermal responses

“Defect-free” Redox currently performing microstructural and compositional analysis on NREL evaluated samples for feedback on thermography response origin and modeling

4/30/2019 REDOX POWER SYSTEMS, LLC 17

Temperature response

slide-18
SLIDE 18

Image Processing – Raster Removal and Defect Detection of Optical Image

Optical Image as taken with macroscope Processed image

  • Removal of raster pattern
  • Image processing highlights defects using black lines based on a

contrast or color difference

  • Future capability to count defects and quantify size and shape

MCO coated sample (with lots of bump defects)

4/30/2019 REDOX POWER SYSTEMS, LLC 18

slide-19
SLIDE 19

Image Processing: Defect Detection of Height Profile

Profilometry as taken with macroscope Processed image

  • Similar set of defects as observed in original optical profilometry image

(left), but defects are more pronounced after image processing (right) MCO coated sample (with lots of bump defects)

4/30/2019 REDOX POWER SYSTEMS, LLC 19

slide-20
SLIDE 20

Thermal transport modeling

Key observations:

  • Spatial variation in IR images even

when there is no excitation

  • Thermal map “reversal” when a

specimen is excited vs. non-excited Recent Progress:

  • Concept of model defined (see left

image)

  • Coating and substrate properties

(e.g, thermal conductivity, heat capacity, and density) collected and/or predicted (includes coating porosity function) Thermal transfer parameters model Sample with coating on top

4/30/2019 REDOX POWER SYSTEMS, LLC 20

slide-21
SLIDE 21

Long-term ASR of “defect-free” coating exhibits reasonable performance

“Defect-free” coating 650 oC

  • ASR at ~0.037 cm2 for 1000 h (a 2nd measurement resulted in ASR

~0.048 cm2 for 350 h)

  • Achieved M2.2 (<0.05 cm2 for 1000 h at 650 oC)

ASR Temp.

4/30/2019 REDOX POWER SYSTEMS, LLC 21

slide-22
SLIDE 22

Long-term ASR of intentionally defective coatings

ASR Temp.

  • Thin coating exhibits

high ASR that increases from 0.06 cm2 to 0.1 cm2 (66%) with time

  • Porous coating has low

ASR, which also increases with time from 0.024 cm2 to 0.029 cm2 (21%)

  • Porous coating exhibits

a promising initial ASR, though high porosity may lead to more Cr volatilization

4/30/2019 REDOX POWER SYSTEMS, LLC 22

slide-23
SLIDE 23
  • 4. Sputtered Thin Film SOFCs
  • Thin electron-blocking layer expected to increase Redox GEN1 Ni-cermet

cell power density by >2x

  • Electron-blocking layer eliminates electronic leakage through ceria based

electrolyte → ~40% increase in open circuit voltage

  • Thin-ness of electron-blocking layer adds negligible resistance
  • Takes advantage of high performance Redox GEN1 cell platform

4/30/2019 REDOX POWER SYSTEMS, LLC 23

slide-24
SLIDE 24

GDC Buffer Layer Deposition

GDC film KDF YSZ film GDC electrolyte

  • Successful deposition of GDC buffer layer with over 1 m/hour deposition rate
  • n lab-scale system
  • Required development of pre-sputter parameters and improvement of

deposition conditions (e.g., Ar and O2 pressure and sputtering power)

  • GDC film deposition still being developed to ensure deposition of dense,

robust film (see next slides on oxidative stress) GDC deposited on GEN1 SOFC sample with YSZ layer previously deposited by KDF SEM of fractured cross-section

4/30/2019 REDOX POWER SYSTEMS, LLC 24

slide-25
SLIDE 25

Buffer Layer Annealing in Air

As deposited After 600 oC 1 h anneal

  • GDC film cracked substantially

after annealing

  • YSZ layer appears to retain

integrity Surface Surface Cross-section GDC film YSZ film GDC electrolyte

4/30/2019 REDOX POWER SYSTEMS, LLC 25

slide-26
SLIDE 26

Source of Film Fracture

42 43 44 45 46 47 48 49

Alumina substrate peak GDC peak shift after annealing XRD spectra peak shift Annealed As-deposited Color lightening after annealing This is a 0.19% chemical expansion →stress ~ 0.0019 * 250 GPa = 0.5 GPa →δ ~ 0.015 (from graph at left) GDC chemical expansion Consistent with loss of oxygen vacancy color centers

  • Film fracture after

annealing most likely due to

  • xidation driven

stress

  • Deposition

parameters being tuned accordingly

4/30/2019 REDOX POWER SYSTEMS, LLC 26

slide-27
SLIDE 27

Scale-Up Sputtering Process

High Rate!

4/30/2019 REDOX POWER SYSTEMS, LLC 27

slide-28
SLIDE 28

Summary

4/30/2019 REDOX POWER SYSTEMS, LLC 28

  • Good progress toward 2.5 kW stack demonstration
  • Expanded capabilities in new, larger natural gas test facility
  • Fabricated large format cells and all-ceramic anode stack with

promising red-ox stability

  • Cost modeling predicts significant decrease in lifetime cost for

red-ox tolerant stacks

  • Optical, height profile, and thermography metrology

techniques shown to detect key defects in MCO coatings

  • Thermal modeling and image analysis software in development

to aid in defect detection

  • Successfully deposited GDC buffer layer with sputtering,

identified significant chemical expansion effect to be mitigated with process optimization

slide-29
SLIDE 29

Acknowledgements

4/30/2019 REDOX POWER SYSTEMS, LLC 29

NETL program managers Seth Lawson and Venkat Venkataraman

  • DE-FE0026189
  • DE-FE0027897
  • DE-FE0031178
  • DE-FE0031656