Maria Skyllas-Kazacosand

George Kazacos*

School of Chemical Engineering University of NSW, AUSTRALIA

and*V-Fuel Pty Ltd, Sydney, Australia

The Vanadium Redox Battery

- First proposed by Skyllas-Kazacos & co-workers at UNSW in 1983- First All-Vanadium Battery patent filed by UNSW/Unisearch, 1986- Extensive R&D at UNSW between 1985 and 1998- Industrial trials and demonstrations by Kashima-Kita Electric Power Corporation in Japan since 1991　　　- SEI began development of the VRB technology in 1993 and since installed more than 20 VRB systems up to 6 MWh capacity for range of applications.- Generation 2 V/Br patented at UNSW in 2001.- 2001 to present: New VRB developers emerging in Austria, USA, China and Australia (V-Fuel)

1. SAME SOLUTION IN BOTH HALF-CELLS.

•Cross-mixing of electrolytes across membrane does not lead to contamination of electrolytes.• Solutions have indefinite life so replacement costs are low (only battery stacks need replacement at end of life).

Technical Benefits

2. ENERGY STORED IN TANKS, SEPARATE FROM CELL STACK.

•Capacity increased simply by adding more solution.•Instant recharge possible by exchanging solutions•Cost per kWh decreases as storage capacity increases.

G1 VRB Capital costs per kWh versus storage time

0

100

200

300

400

500

600

700

1 2 3 4 5 6 7 8 9 10

Storage time/h

$AU

D/k

Wh

Series1

1984-1998

Main stack components are:- bipolar electrodes (usually graphite felt on conducting plastic substrate- flow-frames- ion-exchange membrane to prevent mixing of solutions

Conducting plastic substrate-carbon filled PE/PP/rubber blends

End electrodes

Bipolar electrodeFrom : S.Zhong, 1992V. Haddadi-Asl, 1995

Investigation of thermal,chemical and electrochemicalactivation processes for graphite felt electrodes.

From B.T. Sun, 1991

Wide range of commercialmembranes screened for:•electrical conductivity•permeability•chemical stability•water transfer behaviour

Novel low cost composite membrane based on “Daramic” separator.

From T. Mohammadi, 1995

Investigation of V2 O5:•chemical reduction•leaching•suspended powder electrolyis

From C. Menictas, 1993

0

10

20

30

40

50

60

Blank AP AS Fl Gl M2 P5

Additive

Indu

ctio

n Ti

me

(day

)

[SO4] = 6.50 M[SO4] = 7.00 M

[SO4] = 7.50 M

Induction times for 2.0 M V(III) solution with different total sulfate concentrations, T = 1.0 oC.

(Ref: A. Mousa)

(Ref: F. Rahman)

Demonstrations of G1 VRB Technology at UNSW

1993 to present

Emergency Back-up Battery for Submarines (1995)

Vanadium Battery Powered Solar House in Thailand (1993)

UNSW vanadium battery powered golf-cart (1996)

VRB Installations by Sumitomo Electric IndustriesVRB Installations by Sumitomo Electric Industries

ApplicationsPlace Specifications

Office building

Wind power station

Load leveling(Photovoltaic hybrid system)Golf course

Load leveling (Demonstration) 100kW x 8h

Stabilization of windturbine output(Field test)

170kW x 6h

30kW x 8h

2000

Semi-conductorfactory

1) Voltage sag protection2) Load leveling

1) 3000kW x 1.5sec.2) 1500kW x 1h 2001

2000

2001

Wind Farm Wind storage 4 MW x 1.5 h 2006

South Africa Load leveling (Field Test) 250kW x 2h 2001

Italy Load leveling (Field Test) 42kW x 2h 2001

Delivery

Battery boxes Electrolyte Tanks

Back-up power1.5MW-1Hrs/3.0MW-1.5sec VRB SystemBack-up power1.5MW-1Hrs/3.0MW-1.5sec VRB System

Courtesy Sumitomo Electric Industries

2 VRB load-leveling installations in Japan:• 200 kW/800 kWh VRB at Kashima-Kita Power Station

•450 kW/900kWh VRB at Kansai Power Station (built by SEI)

•Energy Efficiency of 80%

G1 VRB Wind Energy Storage Demonstrations

Tomomae wind farm on Hokkaido:

– 30.6 MW rated output

VRB demonstration system:- Sumitomo / J-Power- 4 MW / 6 MWh

NEDO funding for optimisation of control system

After 3 year operation more than 200,000 cycles completed

- 200

- 100

0

100

200

300

400

500

13:34 13:35 13:36 13:37 13:38 13:39 13:40Ž ž�

�o—

Í�i‚j

‚v

WT Output Stabilization Using VRB - SumitomoWT Output Stabilization Using VRB - Sumitomo

WT Output

VRB Output

Total Output

Discharge

Charge

Out

put(k

W)

Time

Courtesy Sumitomo Electric Industries

Challenges for G1 VRBHigh cost membranes used by SEI and VRB PowerEnergy density too low for mobile applicationsLimited to operation between 10 and 40 oC

• V-Fuel Pty Ltd established by original VRB inventors in 2005 tocommercialise G2 V/Br and new improved G1 VRB technologies

• REDI grant funding further R&D forG2 V/Br.

• Low cost stack technology developed and patented by V-Fuel for G1 VRB

Courtesy V-Fuel Pty Ltd

5-10 kW Stack Discharge Curves for 30 Amp Charge Current

0

10

20

30

40

50

60

70

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73

Time (arbitrary units)

30 Amp50 Amps75 Amps100 Amps120 Amps200 Amps

Note: Discharge time a function of electrolyte volume used(Data courtesy V-Fuel Pty Ltd)

Charge current = 30 Amps

(Data courtesy V-Fuel Pty Ltd)

Efficiencies and Capacity Vs I for 30 Amp Charge

0

10

20

30

40

50

60

70

80

90

100

30 50 75 100 120 200

Discharge Current (Amps)

Eff

icie

ncy

or

Cap

acity

(Am

p.h

ou

rs) Coulombic Eff

Voltage Eff

Energy Eff

Capacity

V-Fuel 1 kW/6 kWh VRBWelded Stack design

• 1 kW sealed stack design and verification completed

• 5 kW stack design verification and testing completed

• Tooling up for 5 kW stack pilotproduction

• 25-50 kW stack design underway • Government and private funding

sought to complete 50 kW stack development and manufacture

Courtesy V-Fuel Pty Ltd

•Redox Flow Batteries are only batteriesthat allow BOTH electrical recharge and “instant” recharge by mechanicalrefuelling

•Spent solutions can be rechargedovernight with off-peak power•Eliminates need for new power stationsto meet increased load from electric cars•But energy density of G1 too low

Vanadium Bromide Redox Flow Battery

G2 V/Br

Generation 2 Vanadium Bromide Redox Cell

While G1 VRB employs solution of Vanadium Sulphate inSulphuric Acid on both sides, the G2 V/Br employs a Vanadium Bromide solution in both half-cells.Higher solubility of vanadium bromide allows energy density to be almost doubled (to around 50 Wh/kg)Higher energy density opens market for electric vehicle applications with refueling station optionHigher solubility of vanadium bromide also allows lower temperature operation of Generation 2 V/Br systemNo V(V) produced, so no thermal precipitationInitial challenge to identify low cost, long-life, high performance carbon felt and membrane

G1 G2

Electrolyte V/Sulphatein both Half-cells

V/Br in both half-cells

Negative couple V3+ / V2+ V3+ / V2+

Positive couple V(IV)/V(V) Br- / Br3-

Specific Energy (energy / kg)

15-25Wh/kg

25-50 Wh/kg

Energy density (energy / litre)

20-33 Wh/l 35-70 Wh/l

Membrane Screening Studies

The following membrane types were evaluated in the V/Br cell:Microporous polyethylene separator (supplied by Mitsubishi Australia)PS11 anion exchange polysulphone membrane (V-Fuel Pty Ltd)VF02 and VF05 cation exchange membranes of 50 and 125 microns thicknesses respectively (V-Fuel Pty Ltd)

G2 V/Br Membrane RequirementsChemical stability in V/Br electrolyteStable to strong bromine oxidising agent Low electrical resistivityHigh conductivity to hydrogen ionsLow permeability to V and Br ionsLow costGood mechanical properties

Cation Exchange Membrane –Optimisation Studies

Performance a function of pre-treatmentRange of pre-treatment processed evaluatedStatic cell test results given in adjacent Table for charge-discharge current of 20 mA/cm2

TreatmentCoulombic Eff

%Voltage Eff

%

A 70% 90%

B 50% 57%

C 60% 90%

D High R N/A

E 90% 90%

(Data courtesy V-Fuel Pty Ltd)

Cation Exchange Membrane –Small V/Br Flow Cell Performance• Flow cell tests conducted atcharge-discharge currentdensity of 40 mA/cm2 with 2 M V-Br electrolyte• Typical charge-dischargecurve shown• Coulombic, voltage and Energy efficiencies calculatedas 91%, 78% and 71% respectively (Data courtesy V-Fuel Pty Ltd)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0

Time (hours)

Volta

ge (V

)

Charge-discharge current density = 20 mA/cm2Improved membrane and cell design

Bromine Complexing AgentsConsiderable development already achieved with bromine complexing agents for Zn/Br batteryMainly based on quaternary ammonium compoundsBind with bromine to form organic liquid layer and 2-phase electrolyte

Complexing Agents inZinc-Bromine Batteries

Source: http://www.zbbenergy.com/

Bromine Complexing Agent Evaluation

Bromine complexing agents evaluated were:N-methyl-N-ethyl morpholinium bromide (MEM) N-methyl-N-ethyl pyrrolidinium bromide (MEP).Tetrabutylammonium bromide (TBA)Polyethylene glycol (PG)

SolutionNumber

V(IV), Br2 concM, M

A & B concM

After 7 days at 11 oC

After 7 days at room temp

After 7 days at 40oC

1 2, 1 1, 0 Solid Solid Solid

2 2, 1 0.75, 0.25

Liquid Liquid Liquid

3 2, 1 0.5, 0.5 Liquid Liquid Liquid

4 2, 1 0.25, 0.75

Solid Solid Liquid

5 2, 1 0, 1 Solid Solid Liquid

(Grace Poon, MSc Thesis, UNSW, 2007)

SampleNumber

Solution composition

11oC for 10 days 25oC for 15 days 40oC for 10day

1 3 M V(IV) 1.5 M Br20.75 M “A”

Orange liquid organic layer

Some brown gasand orange organic liquid layer

N/A

2 3 M V(IV) 1.5 M Br20.5 M “A”0.25 M “B”

Orange liquid organic layer

Orange liquid organic layer

Some browngas and orangeliquid layer

3 3 M V(IV) 1.5 M Br20.38 M “A”0.38 M “B”

Orange liquid organic layer

Some brown gasand orange liquid organic layer

N/A

4 3 M V(IV) 1.5M Br20.25 M “A”0.5 M “B”

Orange liquid organic layer

Some brown gasand orange liquid organic layer

N/A

5 3 M V(IV) 1.5 M Br20.20 M “A”0.6 M “B”

Orange liquid organic layer

Orange liquid organic layer

Some browngas and orange liquid layer

6 3 M V(IV) 1.5 M Br20.75 M “B”

Orange liquid organic layer

Orange liquid organic layer

Orange liquid organic layer

3 M Vanadium Electrolyte withComplexing Agents

Cell cycling studies to be conductedOptimisation of cell design and electrolyte flow system required for uniform 2-phase electrolyte flow distributionImproved membrane to be further evaluated withcomplexing agents

Cost ConsiderationsG1 VRB technically proven in range of applications, but future commercial implemention linked to vanadium pricesCurrent world vanadium production inadequate to meet demand from VRB commercialisation needsRecent price instability creating insecurity for VRB investors – eg VRB Power collapseInvestment in new vanadium mines essential for VRB future

(a) Period 1980 – 2004

(b) Period : Feb 08 to Feb 09

G1 VRB Cost Estimates

VRB assumptions: $500 /kW stack cost, $5/lb V2O5 cost

• G1 VRB being commercialized in many stationary applications• New improvements in energy density of Vanadium Bromide Cell

should extend stationary applications and open mobile • New low cost, high performance membranes identified for G1 and G2• Membrane pre-treatment processes optimised for good overall energy

efficiencies without membrane degradation• Excellent performance with 5-10 kW stacks using G1 electrolyte• Membrane and electrodes compatible with G1 and G2 electrolytes• Small G2 test cells run for more than 12 months with little to no

membrane degradation observed.• Currently testing 3 M V electrolyte for G2 V/Br using bromine

complexing agent• Vanadium supply and price stability crucial for VRB implementation

Acknowledgements

Renewable Energy Development Innitiative Program for Funding Support 2006-2007

Thank You

Questions?