Arrangements Non-standard Arrangements Concepts · · 2017-03-22» Utility connections and...

2/20© Clean Energy Systems Integration Lab, 2017

A Systems Perspective on SOFC

Sizing and Storage

Standard SOFC ArrangementsNon-standard Arrangements

Zero Carbon & TRL 5-6 Concepts


15

14

27

38

52

64

77

8693 97 98 99 99 100

0

10

20

30

40

50

60

70

80

90

100

110

0

300

600

900

1200

1500

1800

2100

2400

2700

3000

3300

Cum

ulat

ive

Perc

ent

# of

Inst

ance

s

Demand (%)

» Utility connections and equipment charges for 100% demand

» Average use is ≈13%» Demand is <20%, for

99% of the time» Generally predictable

demand

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12 14 16 18 20 22 24

% o

f Max

imum

Dat

a Ce

nter

Dem

and

Hour

Daily Demand Profiles

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6Day 7 Day 8 Day 9 Day 10 Day 11 Day 12

0

5

10

15

20

25

30

35

40

45

50

0 3 6 9 12 15 18 21 24 27 30

% o

f Max

imum

Dat

a Ce

nter

Dem

and

Day of Month

99th Percentile

Average

78.8%


Lead-Acid1 : $1,206k˃ Assume: $26/kW, $77/kWh, 3C discharge (50% capacity

loss), 300 cycles 7x replacement˃ 4x needed kWh capacity

Lithium Iron Phosphate Batteries1: $193k˃ Assume: $16.7/kW, $250/kWh, 15C discharge (5% capacity

loss) 1500 cycles 1x replacement˃ 1.5x needed kWh capacity

2017 Costs» Assuming 100% = 10MW» SOFC size is 2MW = 99th percentile

˃ Worst peak period: 0.8MW, 250kWh˃ Worst power surge: 5.8MW, 190kWh

» SOFC initial system cost $5,000/kW» 50k hr stack, replace twice at

$1,800/kW for 20yr plant» Net capex is = $17.2M

Capacitor: $4M˃ Assume: $12/kW, $16,000/kWh, 1,300C discharge, 1e6

cycles no replacement˃ 50x needed power capacity

Battery + Data Center Control: $125k˃ Match kWh needs, throttle data center power surges

1 Shouzhong Yi, Discussion on Lithium Iron Phosphate Batteries Used for IDCs Compared with VRLA Batteries. International Stationary Battery Conference 2016. http://www.battcon.com/PapersFinal2016/Yi%20Paper%202016.pdf2 Maxwell Rechnologies White Paper http://www.maxwell.com/images/documents/whitepaper_powerelectronicsinterface.pdf

Conclusion: peaks & surges can be handled for ≈ 1-2% of SOFC system cost

Fuel Cell

Ultra-capacitor

0

10

20

30

40

50

60

70

80

0 10 20 30 40 50 60

% o

f Max

imum

Dat

a Ce

nter

Dem

and

Minutes99th Percentile Day 1 Day 2 Day 3Day 4 Day 5 Day 4b Day 8

http://www.battcon.com/PapersFinal2016/Yi%20Paper%202016.pdf

http://www.maxwell.com/images/documents/whitepaper_powerelectronicsinterface.pdf


Simulated SOFC response: 0.04% per second

Demonstrated SOFC response: 4% per second1

1 Li Zhao, Jack Brouwer, Sean James, Eric Paterson, Jie Liu, Di Wang, “Fuel Cell Powered Data Centers: In-Rack DC Generation” 2015 http://fuelcellseminar.com/wp-content/uploads/Zhao-Fuel-Cell-Powered-Data-Centers.pdf

0 10 20 30 40 50 600.5

1

1.5

2

2.5

Time (Minutes)Po

wer

(kW

)

Rack DemandSOFC output

http://fuelcellseminar.com/wp-content/uploads/Zhao-Fuel-Cell-Powered-Data-Centers.pdf


» Fuel and water supply˃ Fuel content: pipeline natural gas, landfill or waste gas?˃ Can be net water positive.

» Fuel pre-conditioning˃ Sulfur and particulate removal

+ Won’t focus on this sub-system. Fuel stream determines best technology.

» Power conditioning˃ DC-DC conversion vs. DC-AC and AC-DC conversion˃ Solutions at all scales have similar efficiencies >97%.

+ Next talk will go into greater detail here…

» Stack˃ Important trade-offs: Size, Operating Voltage, Fuel Utilization, Lifespan,

Transient Response Rate

» Thermal management˃ Efficiency, Complexity, and Cost trade-offs in: Heat Exchanger size/type,

Reformer, Air Blower, Combustor, Waste Heat Recovery


DCDC

Combustor

Pre-reformer

Humidification Sulfur Removal

Blower

Fuel Supply

CathodeAnode

ExhaustBypass

Cell/Stack Objectives» Minimize electrical/ionic

resistance» Minimize degradation» Minimize thermal stresses» Lower manufacturing costs

System Objectives» Maximize H2 utilization» Maximize heat recovery» Minimize parasitic losses» Avoid costly high

temperature equipment» Minimize heat loses

Start-up Combustor


1. DC efficiency is directly linked to Voltage

2. Stack cost is inverse of current density

• Stack cost is ≈15% of small system, ≈45% of large system

3. Higher temperature and higher current are linked to faster degradation

4. Higher power density leads to higher thermal stress (convection limited HT)

• High air flow rates for cooling increase pressure drops, and thus blower parasitic

0

0.5

1

1.5

2

2.5

3

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5

Pow

er D

ensit

y (W

/cm

2 )

Cell

Volta

ge (V

)

Average Current Density (A/cm2)

800-V 700-V 600-V800-P 700-P 600-P

5. Degradation from 0.9V to 0.8V (a 11% loss), can cause >50% increase in heat generation

• Air handling system must be sized for end-of-life, not initial conditions


Partial Oxidation

600°C50% CH4, 45% H2O, 5% Other

700°C0% CH4, 5% CO, 20% CO2, 50% H2, 20% H2O, 10% Other

600°C79% N2, 20% O2, 1% Other

10% Energy Loss



No Energy Loss

10% Energy Gain



1000°CCombustion Products

800°CCombustion products

Adiabatic Heat Recovery


Pros: most direct control of FC conditions Cons: Tremendous heat to evaporate water, low hydrogen utilization, large heat exchangers

DCDC

Combustor

Pre-reformer

Humidification Sulfur Removal

Blower

Fuel Supply

CathodeAnode

ExhaustBypass


Pros: best efficiency, single heat exchanger, no fuel humidification, reduced cathode air flowCons: SOFC must tolerate high levels of internal reforming and higher temperature gradients, reduced controllability with ejector

DCDC

Combustor

Pre-reformerSulfur Removal

Fuel Supply

CathodeAnode

ExhaustBypass

ejector


Pros: High efficiency, minimal CH4 reforming in stack, additional control of anode inlet temperatureCons: possibly slower transient response

DCDC

Combustor

Sulfur Removal

Fuel Supply

CathodeAnode

ExhaustBypass

ejector


» Case A: Self-sufficient 2.5kW SOFC modules» Case B: Centralized air blower & desulphurization

˃ Possible safety issue if plumbing odorless methane throughout building

» Case C: Centralized fuel processing, combustion & heat exchange» Case D: Centralized plant

DCDC

Combustor

Sulfur Removal

Fuel Supply

Cathode

Anode

Bypass

ejector

DCDC

Combustor

Sulfur Removal

Fuel Supply

Cathode

Anode

ExhaustBypass

ejector

Centralized

In Rack

Centralized In Rack

Pros: Efficiency & redundancy in air blower and desulphurizationCons: More high temperature equipment in rack and 50% more heat into room

Pros: Efficiency & redundancy in air blower and desulphurizationCons: Significant heat loss between SOFC and plant


Assumptions1. 10kW rack powered by 2.5kW SOFC and 200Wh Li-ion battery2. 2MW system has 8% DC loses between SOFC & rack, uses 250kWh battery3. Centralized blower has secondary redundant blower4. Data center management can readily shift work between racks in event of module failure5. SOFC durability is 60,000 hours (2x replacement in 20 years) for 2MW and 2.5kW

systems, and 45,000 hours (3x replacement) for 1MW6. Availability: SOFC stack – 99.7%, SOFC BoP – 96% (7.5kW) –99.9% (1.5MW) Blower –

98.5%, Grid – 99.9%7. Initial SOFC cost is $9,382 for 2.5kW ($3,753/kW), $1.06M for 1.5MW ($705/kW)8. Stack replacement cost is $887 ($354/kW) for 2.5kW, $294,140 for 1.5MW ($196/kW)9. Grid cost is $0.04/kWh, $2.5/kW demand charge, $1/kW capacity reservation charge10. Fuel cost is $0.50 $/therm11. Battery cost is $250/kWh, 15C discharge, need 175kWh for 1.5MW system12. Inflation 1.9%, Financing interest 6%, SOFC O&M $50/kW per year

Case Net SOFC Capacity

Availability (%)

Cost ($M/yr)

Utilization (%)

Grid 0 99.9% 1.454 N/AA 2.5MW 99.9999% 1.438 52%B 2.5MW 99.98% 1.375 52%C 2.5MW 99.97% 1.148 52%D 2MW 99.58% 0.635 65%Hybrid 1MW 99.9994% 1.159 98.7%

SOFC cost data taken from a DOE report prepared by Battelle. Assessment was scaled from 5kW to 7.5kW. Annual production rates of 10,000 7.5kW units and 100 1.5MW units were assumed. https://energy.gov/sites/prod/files/2014/06/f16/fcto_battelle_cost_analysis_apu_feb2014.pdfResults are in line with 2013 PNNL Report http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-22732.pdf

Fuel Processing6%

Stack10%

Air Supply8%

Heat Transfer18%Electronics

and Controls27%

Instrumentation8%

Assembly7%

De-sulfurization

5% Install11%

PRODUCTION OF 2.5KW AT 10,000 UNITS/YEAR

$100

,000

per

yea

r

10 5

0

5

10

15Reserve capacity chargeDemand charges

Use charges

Fuel cost

O & M

Finance

Grid Only A B C D Hybrid

https://energy.gov/sites/prod/files/2014/06/f16/fcto_battelle_cost_analysis_apu_feb2014.pdf

http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-22732.pdf


» LG: estimated 64% DC efficiency with >5year life˃ Degradation of 5mOhm-cm2/khrs ≈ 0.3%

per 1000 hrs

» FCE: demonstrated 68% DC efficiency for 1500 hr˃ Degradation of 0.3-0.5% per 1000 hrs˃ Estimated >50,000 hr life ≈ 6 years

» Ceres Power: demonstrated 56% efficiency˃ Degradation of <0.5% per 1000 hrs

» Bloom: 52.3% efficiency ˃ Taken from 850lb CO2/MWh assuming CH4

input˃ Degradation unknown

https://www.netl.doe.gov/File%20Library/Events/2016/sofc/Lee-Babcock.pdfhttps://www.netl.doe.gov/File%20Library/Events/2016/sofc/Ghezel-Ayagh.pdfhttp://fuelcellseminar.com/wp-content/uploads/Mukerjee-Ceres-Powers-Steel-Cell-Technology.pdf

https://www.netl.doe.gov/File%20Library/Events/2016/sofc/Lee-Babcock.pdf

https://www.netl.doe.gov/File%20Library/Events/2016/sofc/Ghezel-Ayagh.pdf

http://fuelcellseminar.com/wp-content/uploads/Mukerjee-Ceres-Powers-Steel-Cell-Technology.pdf


» Absorption chilling» Chillers as dispatchable load» Building thermal storage» Cold-water thermal storage


» Solar PV / Wind˃ Adds to range over which FC must operate˃ Possibly extends FC life (lower current density)

» Bio-fueled˃ More expensive gas clean-up, but generally unchanged performance

» Carbon capture˃ Can improve net fuel utilization, and often can co-produce H2

˃ SOFC-MCFC hybrid, with about 8:1 power ratio captures >90% CO2

» Reversible SOFC˃ Captures excess renewable energy to create fuel, FC as a battery˃ Improvement over PEM electrolyzer + FC˃ H2 storage issues˃ Methanation as pathway to use existing natural gas infrastructure as

storage


C T

Air Exhaust

ExhaustAir

Heat

100kW

700kW

400kW

FC-GT Hybrid Challenges:1. Air flow rate & temperatures

must match2. Off-design, GT flow doesn’t

change, FC does3. Large plenum volume,

stall/surge risk


» For high efficiency you must operate at high voltage and avoid burning fuel if at all possible

» Possibility for surge power SOFC is interesting, but BoP system must be designed for peak

» Primary benefit of placing SOFC in rack is eliminating power distribution & transforming

+ Negated if a centralized UPS is deployed+ Added heat load to server room

˃ Pros: greater redundancy, same solution anywhere and any scale˃ Cons: requires fuel distribution in addition to electric distribution system,

does not transfer to non-data center applications

» More zero-carbon options with centralized system

Arrangements Non-standard Arrangements Concepts · · 2017-03-22» Utility connections and...

Documents

Transcript of Arrangements Non-standard Arrangements Concepts · · 2017-03-22» Utility connections and...