Technological possibilities for the separation of H2 from CO2 · Membrane Separation Technology Dai‐ichi Hotel, Tokyo (東京) 28th of September 2012. Japan and the Netherlands.

www.ecn.nl

Technological possibilities for the separation of H2 from CO2

Jaap Vente

Symposium for Innovative CO2Membrane Separation Technology

Dai‐ichi Hotel, Tokyo (東京)28th of September 2012

Japan and the Netherlands

A long joint history

Japan‐Netherlands exchange in the Edo Period (www.ndl.go.jp/nichiran/) 

Dejima Dutch Trading Post (出島 1641 – 1853)

Dejima in the Literature

• A very well written and recent novel 

Netherlands in Japan today

Locations

Petten

Amsterdam

Eindhoven

Wieringerwerf

Beijing

Brussels

Mission:With and for the market, we develop knowledgeand technology that enable a transition to a sustainable energy system.

R&D fields

Energy Efficiency &

CCS

Policy Studies

EnergyEngineering

Environment

Wind Energy Solar Energy Biomass

H2 and separation technologies

Hydrogen production

• Annual production of H2

~70x106 metric tons or ~0.7x1012 Nm3 and growing with ~ 7%/yr

• 50% is produced by steam reforming of methane• Often high purity is demanded

• need for the separation of H2 from CO2

(www.world‐nuclear.org/info/inf70.html)

Elements of the hydrogen economyH2 production plant (Praxair)

H2 – CO2 Separation Drivers

• Low cost high purity ‘green’ H2 with low CO2 food print

• De‐carbonized fuel

• CO2 as valuable feedstock 

• Electricity production with Carbon Capture and Storage.

Pre-combustion capture

Selections to be made:

•Membranes vs. sorption•Separation Enhance Reactors vs. various unit operations

(integrated)   (non‐integrated)•Hot vs. cold•H2 selective vs. CO2 selective

Pre‐combustion vs. post‐combustion.Where is the win?

•Always high P(CO2) so plenty of driving force! Up to 20 bar cf. < 0.2 bar in post‐combustion

•Similar about P(H2)

Solvents at low temperature

• Demonstration phase, maturing quickly

Selexol scrubbing Eagle project, J‐Power, KitakyushuFukuoka Prefecture, Japan

Physical solventCatch‐Up Vattenfall, Buggenum, Limburg, The Netherlands

MDEA scrubbing, Puertollano Plant, Elcogas, Spain 

Sorbent at high temperature

• Integration with reaction  Separation Enhanced Reactor.• Higher conversions at lower costs

• SEWGS• ECN – technology 

• ALKASORB Stability >5000 cycles of loading and unloading

• Full process demonstrated at 20 kWth scale.

SEWGS Scale-up

caesar.ecn.nl

ECN leading in thin filmdense metal membranes

Demands on the membrane

Performance demands•High flux•High selectivity•Long lifetime

Operational restraints •High temperature (400°C and higher!)•Large pressure drop (Pfeed 20 – 40 bar, Pperm 1 – 5 bar)

Materials options

• High performance  thin layers• High pressure drops  support system• High temperature  rule out polymers

Two basic designs for separation layer• Nanoporous: hybrid and ceramic • Dense metal

Two basic materials for support• Ceramic (Al2O3)• Metallic, stainless steel 

Focus of today

H2 Transport in dense metal membranes

1. Bulk diffusion2. Adsorption of hydrogen molecules3. Dissociation into atoms4. Absorption of atoms5. Diffusion of atoms6. Recombination7. Desorption

CO2

H2OH2

H2 H2

H2

H

HH

H1

2

3 54 6

7

Pd alloy membrane

High pressure Low pressure

CO2

H2OH2

H2 H2

H2

H

HH

H1

2

3 54 6

7

Pd alloy membrane

High pressure Low pressure

Pressure exponent n

Layer thickness l

Activation energy EactH2 flux: JH2 = (PoH2/l).exp(-Eact/RT).(PH2fn-PH2p

n)

Temperature activated process

Manufacturing the dense metal membrane

Two distinct approaches

Electroless plating

Dalian Institute of Chemical Physics

Worcester Polytechnic Institute

Energy research Centre of the Netherlands 

Magnetron sputtering

Southwest Research Institute

SINTEF

Sputtering route

(1) Magnetron sputtering on Si wafer

(2) Pull‐off alloy 2 μm thin foil

(3) Wrap around tubular support

(4) Membranes on 50cm length scale

Electroless plating route

(3) Sequential reduction and deposition of metal ions.

(4) Alloying step by heat treatment

(1) Support pore size reduction

(2) Seeding of support 

Lab test module

• Single tube (~10 cm long)• ~40 cm2

• Hydrogen flow: 0.15 ‐ 1 Nm3/h• Up to 70 bars

Scaling up to pilot size

The Chieti plant in Italy

Hysep® Pilot Module

• 13 membranes, 26 seals, L =70 cm• 0.4 m2 surface area• H2 production > 6 Nm

3/h

• Equipped with ECN technology 

• World leader

Simplified process scheme

• 2 stages of reaction and separation • 3 installed membrane module• 20 Nm3/h of hydrogen

First series of tests

CO2 feed ~  6 ‐ 6.8     mol%CO feed ~  1.1 ‐ 2.8 mol%H2Ofeed ~   50‐ 57    mol%H2 feed      ~  26‐ 35    mol%CH4 feed    ~  6‐ 12       mol%

0 200 400 600 800 1000 12000.0

0.4

0.8

1.2

1.6

2.0

2.4

time [h]

0

1

2

3

4

5

6

7QH2 [mmol m-2*s-1*Pa-0.5] GHSV*103 [h-1]

Low pressure feed: 11 – 12 bar

Continued testing

0 200 400 600 800 1000 1200 14000.0

0.4

0.8

1.2

1.6

2.0

2.4QH2 [mmol m-2*s-1*Pa-0.5]

time [h]

Improved insulation

Operating variables

Variable QH2 H2 production H2 recovery H2 purity

GHSV  ? ‐ ↓ ?

T reformer ‐ ↑ ↑ ↑

T membrane ↑ ↑ ↑ ↑

Pfeed  ‐ ↑ ↑ ?

Ppermeate ‐ ↓ ↓ ?

Test results

• Successful scale up of a factor 100:from 40 to 4000cm2

• High H2 purity > 99.94%• Production up to 8 Nm3/h H2

(exceeding demands)

• Total test period of over 1500 hrs.

• Operating conditions:• T = 450 °C• P = 25 bar

Prototype HydrogenSeparation Modules

• To full length multi tubular (1/8 to 1/2m

2)• Up to 8 Nm3/h• Up to 25 bar

ChamplainFrom Pilot to Demonstration

• Consortium under construction• Three modules  three suppliers  one dream• Once more one step increase in– Size, production rate, duration, and conditions

More parties are invited to join as:

• Full consortium partner

• Associated participant

• Collaborative contributor

Acknowledgements

• Sintef: Thijs Peters, Marit Stange, Rune Bredesen

• Tecnimont KT: Annarita Salladini, Gaetano Iaquaniello

• My colleagues at ECN working on:SEWGS & HySep

Our sponsors

EU

NL

35

Jaap Vente+31 – 224 56 [email protected]

www.hysep.com