Autostack HFC-JU Project No.: 245142 Agenda 18 April 2011 Topic Joint Workshop on Generic Stack Design, Cost Model and Resources Requirements Venue Lucerne Date 28.06.2011 Start 14:00 Convener Telefon-Nr. Minutes taken by Telefon-Nr. Agenda from L. Jörissen +49-731-9530-605 18.04.2011 End 18:00 Participants A. Martin L. Jörissen F. Finsterwalder P. Ekdunge T. Priem B. Andreas-Schott T. Giunti S. Sibona R. Ströbel Discussion / Result V. Banhardt F. Büchi G. Tsotridis R. Zuber M. Holzmann N. Zandona C. Loevenbruck J. Sfeir U. Hannessen CC C. Navas A. Müller Remark / Action Agenda TOP 1. TOP 2. TOP 3. TOP 4. TOP 5. TOP 6. TOP 7. TOP 8. TOP 9. Welcome and Introduction to the Autostack Project OEM-Requirements for Automotive Fuel Cell Stacks European Supply Chain Analysis Stack Cost Analysis (Methodology) Coffee break Stack Cost Analysis (Preliminary Results) Draft Business Concept for a European Stack Integrator Discussion End of Meeting - 1 -
General Assembly / Workshop on generic Stack Designs Jun 28 th 2011, Luzern Autostack Final Conclusions on WP 1: OEM Stack Platform F. Finsterwalder Daimler AG
Automotive System Architecture Features: Air compressor without expander Gas-to-gas humidifier (cathode out cathode in) High power density stack An active / passive H 2 recirculation pump M p = 1.. 2 bar abs T = 55.. 95 C max (outlet) H2-blower M humidifier M anode cathode coolant Rh cathode @ max power = 50% radiator Simplified schematic 1
High Level Stack Requirements Stack nominal power (gross, continuous) (System power 80 kw) Stack Open Circuit Voltage (limit set by power electronics consider OCV > 1V @ freeze T) Minimum stack voltage (limit depending on E-motor and DC/DC converter characteristics) 95 kw (requirement) <430 V (requirement) >200 V (guidance) Weight Volume <75 kg <60 l Operating Temperature min -25 C (start capability) max +95 C (outlet temp.) Interface parameters at nominal power Pressure 2 bara Air Stoichiometry 1,6 Cathode Humidification 50% (nominal OP) 2
Stack and Plate Design Area Power density 95 kw, 1 W/cm 2 Cell voltage at nominal power 0,67 V Current density at nominal power 1,5 A/cm 2 Active area of stack 9.5 m 2 Number of cells (1-row stack is a requirement) 300.. 380 (staple height!) Active are area per cell (projected) 317 cm 2 Plate Area (a.a. = 60% of plate area) 528 cm 2 Cell pitch 1,5 mm Preliminary Plate Design (1 Path parallel Flow Field) Ci Wi 10,3 cm Ai Wo Plate Dimensions: 11,3 x 46,8 cm 2 Ao Port Region Transition Region Co 30,8 cm Active Area (aspect ratio 1:3) 3
Packaging Comparison Available Stack Box volume in OEM vehicles Daimler (E-class) Fiat (Panda) VW (Jetta) 785 mm 618 mm Top view Driving direction 865 mm driving direction Side view Driving direction 610 mm 688 mm 786 mm 610 mm driving direction 770 mm subfloor limit 4
Principle Stack Mounting Orientations stack might be tilted Driving direction critical due to crash 5
Packaging Conclusion Stack can be fitted in common box common box volume 600 mm 470 mm 450 mm 605 mm 785 mm 140 mm V = 40 l 510 mm 6
Conflicting Targets / Major Gaps Essential target is to bring down costs of stack The latter are primarily determined by the PGM content Lowering PGM content entails technical challenges / conflicting targets Improving power density w/o compromising other targets is a long-term effort Short term mitigation: Trading off power density vs. Pt-loading 7
Gap Power Density / Pt Loading Fuel efficiency targets might even be more challenging relevant for fuel efficiency relevant for top speed 8
Specific Stack Costs Projected MEA (2015) "automotive ready" 15
Specific Stack Costs Projected MEA CEA cost study" 17
Conclusion The cost optimized Pt loading is influenced by: - the costs of membrane, GDL, BIP vs. Pt - the power density characteristics of the MEA Given today s price structure, the cost optimized Pt loading goes along with meetings the power density target. Lower costs of area components / higher Pt costs shift the cost minimum towards lower Pt-loadings. Power density is the paramount target. High power density allows standardization, thus economies of scale and cost reduction. 19
Standardization & Economies of Scale higher power density Cost optimized Pt loading is high higher degree of freedom (packaging) economies of scale (1) Cost optimized Pt loading moves towards lower values standardization Addressing product design targets Pt reduction economies of scale (2) Addressing ultimate cost targets Ultimate cost targets met market volume 20
Open Issues Efficiency targets at reduced Pt-loading Limited turn-down ratio of the fuel cell stack Fuel Cell Range Extender as a short-term mitigation?. to be discussed Pressure drop in stack turn down ratio) p max p min Current density / volumetric flow / load 21
Thank you very much for your kind attention Daimler AG
European Supply Chain Analysis Workshop on Cost Model and Resources Requirements Paul Scherrer Institut Stefan Kreitmeier, Philipp Dietrich, Felix Büchi 28 th June 2011, Lucerne PSI, 6. Oktober 2011 Seite 1
Objective Autostack Developing a technical roadmap for a future generic PEFC stack platform Workpackage 2.1 Inventory of the European stack component supply industry Seite 2
Goals 1. Compiling a database of European companies dealing with PEFC technology 2. Assessment of commercial availability and level of product maturity 3. Prepare basic technical data of PEFC materials and components suited for the stack platform Component properties Component cost Seite 3
Boundaries Only stack repeating components Membrane Catalyst GDL Subgasket MEA Sealing Metallic BPP Composite BPP BPP Seite 4
Boundaries Only stack repeating components Only European suppliers of PEFC stack components only representative for Europe Seite 5
Boundaries Only stack repeating components Only European suppliers of PEFC stack components Component specifications were restricted to the requirements of Auto-Stack (WP 1) Seite 6
Boundaries Only stack repeating components Only European suppliers of PEFC stack components Component specifications were restricted to the requirements of Auto-Stack (WP 1) For current and future preliminary mass production (250-25.000 stacks/year) in the time frame 2015 to 2020 Non disclosure agreement A minimum of 3 replies required for data anonymisation based on averaging Seite 7
Methodology 1. Compile an inventory of the European companies internet research stakeholder + + database Fuel Cell Exhibition 2. Data acquisition Short questionnaire Long + questionnaire + with NDA Interview with NDA Seite 8
Summary of results 1. Compiling a database of European companies dealing with PEFC technology Seite 9
Summary of component properties 1. Metallic bpp mostly fulfill the requirements 2. GDLs are supplied with adequate properties based on the Autostack requirements. 3. No evaluation for catalyst and sealing materials Seite 10
Summary of component properties 4. Gap beetween demand and supply High power density at required low Pt loading Cell pitch for carbon composite BPP usage Coating technology for metallic BPP 5. Inadequate disclosure of stack component durability 6. Limited availability of high matured PEFC components in EU Market competitiveness? Seite 11
Summary of component cost Cost specified for generic stack from Autostack 1 W/cm 2 specific power density of the MEA at 95 kw stack power* (80 kw stack net power) with 315 cells per stack with 300 cm 2 active area * DTI: 87 kw stack power for 80 kw net power Seite 12
Summary of component cost MEA: Agreement to DTI study for lowest cost target. Production cost [per kw stack ] Mean value Lowest value @ annual production rate Production capacity [units] 2010 124 71 100.000 units 10 300 k/a 2015 62 14 1 mio units 0.1 1 mio/a 2020 44 14 4 mio units n.a. DTI study Production cost [per kw stack ] 2010 74 @ annual production rate 300.000 units B. D. James, J. A. Kalinoski, K. N. Baum, DTI, Contract Nr. DE-AC36-08GO28308, Virginia, USA, 2010. 2015 2020 23 14 0.9 mio units 2.4 mio units Seite 13
Summary of component cost Membrane: Moderate cost dependency on production rate Production cost [per kw stack ] Perfluorinated Hydrocarbon @ annual stack production rate 2010 n.a. 15 1000 2015 1 10 10.000 2020 n.a. n.a. 100.000 DTI study B. D. James, J. A. Kalinoski, K. N. Baum, DTI, Contract Nr. DE-AC36-08GO28308, Virginia, USA, 2010. Production cost [per kw stack ] 2010 2015 40 6 @ annual stack production rate 1000 30.000 Seite 14
Summary of component cost GDL: 50% lower cost compared to DTI study in 2015 Production cost [per kw stack ] Mean value Lowest value @ annual stack production rate 2010 13 6 1000 2015 3.5 1.5 10.000 2020 n.a. n.a. 100.000 DTI study Production cost [per kw stack ] 2010 16 @ annual stack production rate 1000 B. D. James, J. A. Kalinoski, K. N. Baum, DTI, Contract Nr. DE-AC36-08GO28308, Virginia, USA, 2010. 2015 7 30.000 Seite 15
Summary of component cost BPP: Composite plates more expensive than metallic Both metallic and composite plates are several factors more expensive compared to DTI study Production cost [per kw stack ] Carbon BPP Mean value Carbon BPP Lowest value Metallic BPP Mean value Metallic BPP Lowest value @ annual stack production rate 2010 87 32 44 37 300 2015 42 27 18 15 3000 2020 20 12 11 8 30.000 DTI study Production cost [per kw stack ] 2010 2015 Carbon BPP 2.8 2.3 Metallic BPP 13 B. D. James, J. A. Kalinoski, K. N. Baum, DTI, Contract Nr. DE-AC36-08GO28308, Virginia, USA, 2010. 3 @ annual stack production rate 1000 30.000 Seite 16
Summary of component cost Only a limited number of suppliers with high maturity level of their product Cost and technology assessment difficult Seite 17
Thank you very much for your kind attention PSI, 6. Oktober 2011 Seite 18
Status Work Package 4 Luzern 27.6.2011 Preliminary results
1. Deliverable 4.1 Expertise and Resources André Martin Consulting 1/17/2012 AMC 1
General assumptions of the business model Combination and integration of expertise and resources of research, supply industry and OEMs. Development of a joint technology roadmap and transparent collaboration scheme based on strategic partnerships. Improvement of economies of scale by combining volumes of several OEMs and applications based on joint platform concept. Establishment of a dedicated development structure with focus on stack integration, validation and production. Recruitment of experienced staff with solid track record in stack development to accelerate learning curves. 2
Core competences of stack integrator Core competences: Stack development and design (concept, interfaces, specification, IP ) Component specification and validation (MEA, BPP, BoP..) Stack prototype build, testing and validation Supply chain management Manufacturing/assembly Support to core: Sales & Marketing Project- and program management Quality system Infrastructure and IT Finance & Admin Overall staff demand: 25 40 depending of progress 3
2. Deliverable 4.2 Market Study André Martin Consulting 1/17/2012 AMC 4
Common OEM specification Operating point 1,5 A/cm²@ 0,675 mv Pt-loading 0,15 mg/cm² Stack-power 95 kw, scalable Operating temperature < 95 C Operating pressure < 2bara Voltage 220-430 V Power density < 60l / 75kg Cost 40 /kw @ 100000 5
State-of-the-art in stationary offers room for maneuver Stationary PEFC Stacks Moderate power density (~0.4 W/cm2) Medium cell pitch (~ 5 mm) High endurance ~ 20 000 h under stationary conditions) Turn down ratio 1:3 Automotive PEFC Stacks High power density (~ 1 W/cm2) Low cell pitch (<1,5 mm) Reasonable endurance < 10 000 h under stationary conditions) Turn down ratio ~ 1:20 Several stationary applications have complementing requirements. Transfer of automotive design concepts can deliver substantial benefits. Stack design needs to be robust to allow sufficient flexibility. André Martin Consulting 1/17/2012 6
Assessment identified technical compliance level André Martin Consulting 1/17/2012 7
Large potential based on technical synergy ~ 0,8 1,0 billion market potential even @ low penetration rates André Martin Consulting 1/17/2012 8
Possible Stack Classification According to Stack Voltage: Standard Voltage levels: 12 V, 24 V, 48 V, 96 V According to design single cell voltage: High efficiency: Automotive: High power: 0.75 V 0.67 V 0.62 V According to BoP-Effort: Low effort: Automotive: High effort: ambient pressure 2 bar abs 3 bar abs 1/17/2012 André Martin Consulting 9 1/17/2012 AMC 9
Stack design examples using the same cell hardware (active Area 300 cm 2 ) High efficiency stack for 24 V applications: Number of cells: 32 Average single cell voltage: 0.75 V Pressure level: ambient Area specific power density: 0.5 W/cm 2 Stack nominal power: 5 kw Automotive: Number of cells: 314 Average single cell voltage: 0.67 V Pressure level: 2 bar abs Area specific power density: 1.0 W/cm 2 Stack nominal power: 95 kw High power density stack for 96 V applications: Number of cells: 155 Average single cell voltage: 0.62 V Pressure level: 3 bar abs Area specific power density: 1.3 W/cm 2 Stack nominal power: 61 kw 1/17/2012 10 1/17/2012 AMC 10
Success story consumer markets 1. Outline and objectives of the project Standardized cell and battery design have reduced cost dramatically.
Anticipated overall market launch for FCEVs and other applications is now more certain Assumed roll-out scenario of FCEVs in Europe *: 2015 100.000 FCEVs 2020 1.000.000 FCEVS Other applications have the potential to add 10 000s of units from 2015 Hence production volume potential of at least 50.000 units is assumed as baseline after ramp-up * A portfolio of powertrains for Europe (McKinsey Study) 12
Use of platform concepts to help economies of scale early on Standardized platform concepts are allowing superior scale effects in early phases of commercialization for both, OEMs and suppliers. In the long-term, novel processes and materials will be required to achieve ultra-low Pt-loading targets and thus achieve the required scale effects at growing market volumes. High power density Economies of scale (1) Pt - reduction Marktvolumen Improved flexibility (Packaging) Standardization Economies of scale (2) 13
Deliverable 4.3 - Draft business plan André Martin Consulting 1/17/2012 AMC 14
Massive savings are possible when accumulating volumes through a common platform concept Savings by no of vehicles when arriving next level of scale effect: A B = 6,5 m /1000 B C = 28,3 m /10000 C D = 46,5 m /50000 D E = 166,6 m /100000 Total stack cost Production rate 15
P & L - sensitivity analysis over three scenarios Total Cash Flow Scenario 1 58 M Scenario 2 64 M Scenario 3 72 M Breakeven 10 200 4000 50000 16
5. Summary and conclusions André Martin Consulting 1/17/2012 AMC 17
Summary and conclusions Economies of scale are much more important for early cost reduction than Ptloading. Pt-loading is a long-term issue for higher production volumes in combination with advanced technology while observing efficiency limitations. High power density of the stack is key for platform concept and for achieving cost targets in mid term. Market introduction of fc vehicles will require massive investment before and until reaching market penetration (optimum production rates). Therefore, common platform concepts can help mitigate and substantially reduce market introduction cost. 18