Hamburg University of Applied Sciences Fuel Cell ... - HAW Hamburg

19 downloads 970 Views 333KB Size Report
Original manuscript published as: Fuel Cell Technology at the University of Applied Sciences Hamburg. European Fuel Cell News. Vol. 8, No 4, January 2002, p.
Hamburg University of Applied Sciences Fuel Cell Laboratory

W. Winkler

Fuel Cell Technology at the Hamburg University of Applied Sciences updated and revised version: 20 May 2005 Original manuscript published as: Fuel Cell Technology at the University of Applied Sciences Hamburg. European Fuel Cell News. Vol. 8, No 4, January 2002, p. 9 - 13.

The topic fuel cells started at the Hamburg University of Applied Sciences in 1991 in connection with the hydrogen activities in Hamburg at this time. At an very early state of research the development of combined fuel cell heat engine cycles was defined as a key issue. The target of the work is to deliver the basics of the engineering of fuel cell systems and to show product visions. This research targets strongly influenced the methodology of the research work over the years. Theoretical thermodynamic models has been developed to define the engineering targets and to be used as a very cost-efficient quality assurance tool for engineering development. Cost and market studies were used to control the development targets of the product visions. Philosophy, Methodology and Results of the Research Work Fig. 1 illustrates the used R&D philosophy. The demand on an energy converter and the allowable cost of any design are determined by the market and deliver the R&D targets of the product vision. The results of the basic research deliver the necessary database for this R&D process in engineering. The process design mainly based on thermodynamics is the first engineering step. The following design study results in a specification and an identification of the critical components. The specification allows the decision to stop or to proceed the project and delivers the target for the next step. The identification of the critical components and the modified specification initiate further activities. The possible feedback at any stage of this R&D process allows a reengineering of the process or the design or starts R&D activities of certain components. Finally the result of the last step should be a basic engineering of the system including an engineering package of the "critical components". The delivering of the necessary design tools is a task of technical faculties and research institutions in this Product Orientated REsearch (PORE) to support the product development in the industry. Market Demand

Target

Model &/or experiment

Target

Process design

Advanced process

Process design

Design tools

Fig.1 The Product Orientated REsearch approach

The introduction of combined fuel cell heat engines as a standard Design Design Critical power generation system finally study study component depends on the system's cost. One of SpecifiSpecifiSpecifithe actually most interesting combiProduct cation cation cation nations is that of a SOFC with a gas turbine (SOFC-GT). The still actual Basic Research : Material science cost studies of mature systems lead Chemistry, Physics etc. to specific cost of about 1000 US$/kW for the market entrance and to 400-500 US$/kW for developed markets. This cost will automatically decrease if the production quantity will increase. But it is important to analyse the influences on the cost effiCost

ciency of the design of SOFC-GT units for a necessary cost reduction to increase the value for the user and to strengthen the position of the SOFC-GT technology on the emerging markets of distributed generation. Fig.2 shows the main influences and their interactions. The process design is mainly based on the thermodynamics but the system cost is obviously influenced by the demands of the hardware design and of the production as well and last but not least of the market that will guarantee the produced quantity. However the process design is the general base and rules the overall design quality but its transfer in a product must be accompanied by a design analysis as well to assure a minimum risk of the product development of a cost efficient system. This is the reason that the R&D hardware design process design work at the Hamburg University of (geometry) (thermodynamics) Applied Sciences was always connected with design studies. Fig. 2 shows the interactions of the prodmarket analysis materials and handling uct development. (quantity) (production)

cost efficient system

Fig. 2 The interactions of the product development of fuel cell systems

The milestones of the R&D work at the Hamburg University of Applied Sciences had been (the for this work relevant activities are included in Italic) : 1992 First cycle with efficiency > 75 % 1993 First simplified process model (QA) - theoretical maximal efficiency ~ 80 % 1994 Complete reversible process model (QA) 50 MW SOFC-GT design study (76 % efficiency) Start experiments with planar SOFC 1st Contact US Department of Energy (DOE) 20 MW Westinghouse study (56 % efficiency) 1995 Planar stack design studies Scale up model DOE funding of SOFC-GT in USA 1996 Design review start work on tubular SOFC Westinghouse SOFC-GT design study (70%eff.) 1997 Presentation of new tubular SOFC stack design small tubular SOFC (200 K/min by K. Kendall) first Microturbines at the market 1998 Begin with mobile SOFC-GT studies cost studies (target 500 $/kW) SIEMENS buys Westinghouse (SOFC tubes) 100 kW SOFC plant (SIEMENS Westinghouse)

1999 Development tubular-helix design (mass transf.) DOE new targets (80% eff., 400-600 $/kW) Program review in Japan (only tubular design) 2000 Presentation of design study of a mobile 75 kW SOFC-GT platform (efficiency 45 % (> 55%)) fuelled with diesel or gasoline; New stationary cycle design for efficiencies > 80 % First studies on maritime SOFC-GT First 220 kW SOFC-GT by SIEMENS Westinghouse at University of California (Irvine); Smallest ceramic tubes for SOFC application presented 2001 Consortium forming and application for FCSHIP and national maritime SOFC-GT, building of relations for EU networks on SOFC and fuel cell testing and standardisation; Studies on integration of FC systems in energy system 2002 Start FCSHIP (EU); Founding of FC network North Germany; Start of development of Matlab-Simulink design tools; Start of studies on FCs for aircraft applications; 2003 Start of FCTESTNET (EU) and SOFCNET (EU); Acceptance of Humboldt proposal (start April 2004); Study for Airbus on FC systems (OBOWAGS/APAWAGS) on board aircraft; Application for FC system modelling on board aircraft’s; TU Munich/NASA presented SOFC-GT on board aircraft (confirmation of results of own mobile SOFC-GT study of 2000) 2004 Further improvements on Matlab-Simulink model; Start of Humboldt funded project. Acceptance of FELICITAS and CELINA maritime and aircraft based fuel cell application studies. Re-start of fundamental analysis of 2nd law system integration Thermodynamic based process and modelling design The idea behind the thermodynamic model is very simple. The fuel cell is a "power producing burner" only defined by the ratio of the delivered power and heat. The produced heat can be used to operate any heat engine (1). The ideal generalised cycle shows that the heat recovery process for the air heating and the fuel heating is independent of the heat engine cycle. But we get a matching between the heat engine and the air heating, if we use a common gas turbine (GT) as the heat engine. The design process of such a GT cycle is now directly determined by the restrictions of the thermal stresses of the SOFC. The maximal allowable temperature difference Δϑmax between the inlet and the outlet temperature of the cathode e.g. 150 K delivers

a very high air flow for the SOFC cooling only by air. If we allow this, the waste gas loss drastically increases and the system (or electric) efficiency can become lower than the cell efficiency itself. Thus any successful cooling strategy of a SOFC integrated in the SOFC-GT system must avoid a high excess air at the outlet of the total system. work

Δϑmax

work heat fluid

work heat engine

-

fuel cell heat

Fig. 3 the generalised fuel cell heat engine cycle and the tasks of the thermal engineering

fuel processing

Fig. 4 gives an overview of the possible design strategies (2), (3), air (4), (5). One strategy is to divide fuel flue gas the SOFC module in sub-modules and to extract the heat of the SOFC • choice of heat engine and integration module by cooling down the waste • integration of fuel processing • integration of preheaters air of the first sub-module to the ⊗ restriction by ceramic cell : Δϑmax inlet temperature of the cathode of the following sub-module by a gas turbine and producing additional power. This process of an intermediate expansion (INEX) can be carried on until the last GT delivers the waste gas for the air heater and the fuel heaters. preheater

waste heat

area of thermal engineering

W. Winkler 2000

Intermediate expansion INEX :

External cooling EXCO :

• exhaust temperature ‘ SOFC waste heat

Fig. 4 Cooling strategies of SOFC modules by GT cycles

extraction (sub-systems)

The other strategy is to cool the SOFC module by an external cooler (EXCO). The SOFC module is the direct heat source of the GT cycle and the air and the fuel are heated by the flue gas in heat exchangers (HEX) as shown in the ’ pressure difference HEX walls fuel air generalised model. The integrated flue gas “ air inlet temperature in SOFC gas heater allows an optimisation reformer ” size of HEX surfaces of the temperature level of the cooling flows around the cell together with an integrated air heater and this avoids unacceptable thermal stresses of the cell ceramic and disturbances of the electrochemical process. W. Winkler 2000

INEX :

EXCO :

2 - n pressure ‘ SOFC waste heat extraction (sub-systems) levels (systems)

1 pressure level (system)

’ pressure difference HEX walls

maximal pressure differerence

only pressure loss

“ limit for air inlet temperature in SOFC

gas turbine outlet temperature

SOFC temperature

” size of HEX surfaces

min. 1/2,5 of ambient system

min. 1/7 of ambient system

• exhaust temperature

~ 200 °C

500 - 600 °C W. Winkler 2000

The main differences between the INEX and EXCO are listed in fig. 4 and are compared in fig. 5. The waste heat extraction ← is done in one pressure level in the EXCO design and needs up to n pressure levels in the INEX design, depending on the allowable temperature difference of the cathode, and needs thus n pressurised subsystems here. Fig. 5 Comparison of the INEX and the EXCO design

The pressure difference at the walls of the air heater ↑ is the maximal pressure at the INEX design and only the pressure loss of the module at the EXCO design. The demands on the material quality for the heat exchangers of an EXCO design is thus comparable small. The air heater outlet temperature → is limited by the SOFC (module outlet) temperature at the EXCO design and by the lower gas turbine outlet temperature at the INEX design. The size of the HEX ↓ of an INEX design is about 2,5 times smaller than under ambient conditions caused by the pressurisation at one side, but the EXCO design has up to about 7 times smaller HEX surfaces than under ambient conditions caused by the pressurisation on both sides (8). The electric efficiency of an INEX design compressor LP-SOFC module with two turbines is about 70 % similar to the simple EXCO design. But the exhaust temperature ° of the INEX design is about 200 °C and of the EXCO design is HEX flue gas HP-SOFC module about 500 to 600 °C. The EXCO turbine fuel gas design has thus the potential for a waste air air turbine combination with a steam turbine waste air 1st stage cycle (ST) and can reach an elecflue gas steam cycle tric efficiency of about 75 %. steam /water

Result : ηel ≥ 80 % possible

W. Winkler 2000

Fig. 6 The Reheat SOFC-GTsteam-turbine cycle (RH SOFC-

GT ST)

Electric Efficiency [%]

An efficiency of about 80 % is possible in combination with a reheat gas turbine process as recent studies showed. The EXCO cycle is expanded by a second low pressure SOFC-GT module and is used a first stage of the Re- Heat SOFC-GT Steam-Turbine (RH SOFC-GTST) cycle. The first GT, after the high pressure section (HP-SOFC module), is called here as "waste air turbine" to show that the waste air is used as the combustion air of the second stage. The second stage, the low pressure (LP) section, doesn't need any external gas coolers because the comparable small LP-SOFC module is cooled by a comparable high waste air flow coming from the HP-SOFC module. The waste gas boiler of the ST cycle is supplied with the flue gas of the last GT - the "flue gas turbine" - to use the waste heat of the SOFCGT modules in a most efficient way. The size of the components show that this type of cycle is interesting for a capacity > 10 MW. A similar cycle without a steam cycle is interesting for a smaller capacity and the efficiency 70 is between 70 and 75 % (5). 12000

Fig. 7 Electric efficiency of the SOFC hybrid in an aircraft application depending on part load and altitude

8000

4000 0

60

Altitude [m] ISA U = 0,7 V UfPOX+FC = 85 % POX Excess Air = 0,13504 S/CRef = 2,0185

50 20

30

40

50

60

70

Part Load [%]

80

90

100

The increase of system simulation work made it necessary to identify a new computing platform instead of Basics and Excell for using the in house developed calculation tools. The choice was MATLAB-Simulink. The last version was used to calculate

part-load operation of an aircraft’s SOFC-GT power generator (6). Fig. 7 gives an example of the operation of an SOFC-GT system onboard aircraft during a mission between ground operation (0m) and its ceiling altitude (12000 m) and shows the capacity of developed MATLAB-Simulink tool (6). Stack design and experiments The result of the first phase of SOFC stack design was a design concept of a welded sheet metal interconnector. The SOFC was beard by welded naps inside of the interconnector with an internal manifolding (7). Insulating glass powder was used as a sealing material. The free stretching SOFC seemed to be an interesting option, because of the opportunity to use commercial metallic alloys instead of very specialised alloys or ceramics. The first hot tests in a self built test rig showed that the ceramic material of the SOFC was not destroyed after temperature cycles in a single stack with interconnectors of the high temperature alloy HA 230. HA 230 was used as an available example of the typical thermal expansion characteristics of metallic alloys and not because of any requirements of a long term stability. The thermal cycling could have been reproduced but the produced power decreased during the thermal cycles however the open circuit voltage was comparable stable. The reason of this behaviour was identified as increasing cracks in the electrodes around the naps. Fig. 8 shows a damaged anode after thermal cycling. A design review of the concept showed that this cycling problem seemed not to be solved in an comparable easy way at reasonable cost. Thus the tubular design was included in the design review an the general influences have been compared (8). Fig. 8 Planar SOFC with metallic interconnector after thermal cycles The surprising result was that nearly all relevant parameters indicated the tubular concept as the most promising design. The complete perimeter of a tube surrounding the flow channel can be used as electrode while maximal about only 50 % of the perimeter of the flow channel of a planar design can be designed as electrode. A result of the design analysis was the tubular SOFC design proposal, as shown in fig. tube element interconnector stack cooling tube 9. combustion zone

conducting air nozzle

air

+

-

Fig. 9 Proposed tubular SOFC stack design

The proposed tubular SOFC design has the benefit of forming a cascade of the total rotary symmetry of the SOFC tube electrochemical reaction and a SOFC integrated cascade and netzwork element network in a very simple way. hexagon spacing with highest power density high thermal integration Different ways of a system integration of small diameters with highest power density the stacks are possible (9), (10). The combustion of the depleted fuel is integrated into the stack. The free expansion of the SOFC tubes is possible and helps to avoid a nickel felt

fuel

insulating nozzle plate

W. Winkler 1999

complicated design or a complicated and expensive matching of the used materials of all stack components. The short sealing length is a further benefit of any tubular design. Hexagonal spacing and small tube diameters allow a high power density and a short start-up time as needed for mobile applications. The use of a small number of components and only completely rotary parts of the SOFC tubes helps to control the manufacturing cost. There is also very strong influence of the power density of the SOFC stack on the balance of plant cost that cannot be neglected (11). Studies on the mass transfer in SOFC flow channels indicated that very small diameters or the integration of a helix can deliver high mass transfer coefficients in a very small volume as necessary for a high power density. The equations for calculating the Nusselt numbers of tubular-helix heat exchangers have been used to calculate the Sherwood number of a tubularhelix SOFC (12). Successful model experiments have been made to proof the concept, by measuring the change of electrical conductivity in the secondary loop of a test rig carrying porous tube models caused by the mass transfer of salt ions (13).

2

3

Fig. 10 view of the micro reformer test rig at Hamburg University of Applied Sciences, 1 fluid distribution, 2 evaporator, 3 test rig

The use of diesel or kerosene within fuel cells in aerospace and maritime application 3 2 need a fuel processing however SOFCs allow a relative simple approach. This may occur by partial oxidation (POX) of by steam reforming. Recent presented calculations (6) showed that the only POX 1 operation can reduce the electrical cycle 1 efficiency by up to 20%. The necessary compactness of the fuel cell system and its lightweight design indicates experiments with steam reforming in micro-devices. Fig. 10 gives a view of the test-rig designed and built for this purpose. The work done after a re-design for this application within the EU project FELICITAS. Design studies The necessary design study as a proof of concept of the cycles has been already done in 1994 for the ECXO cycle as shown in fig. 11. In the centre of the main floor an axis is formed by the generator, the gas turbine, the waste heat boiler, the flue gas condenser and the stack. The new technology of the SOFC leads to a four pass design of all components directly connected with the SOFC. The four pressure vessels containing the SOFC are arranged around the mentioned axis generator to stack. Each vessel can be supplied with air and fuel separately via the heat exchanger systems. The four compressors are electrically powered. The compressors and the heat exchangers are on the ground flour (14).

The design study of the 50 MW unit showed that the SOFC technology can be pretty good integrated in the existing technology. Only the SOFC module and its interfaces are the completely new designs. The study already showed that the specific size of these plants is equal to actual CCGT plants. But a new deSOFC-Module Generator Gas turbine 76 % efficiency velopment of system components as microturbines and new stack designs Stack as well will generate new applications with very small capacities as discussed below. Flue gas condenser

Waste heat boiler

Fig. 11 The design study of a 50 MW SOFC-GT plant

The temperature gradients of 200 K/min that can be realised by thin Heat exchanger tubular SOFC, as shown by K. Kendall et. al. (15), gives the impression that thin tubes can solve the start-up problem of SOFC in mobile applications. The second benefit of thin tubes in general is the high power density > 1 kW/l that can be expected. The development and the market introduction of the microturbines shows the availability of small sized turbines down to 25 kW capacity (16). Compressor

fuel tank

SOFC-vessel

power electronic

Fig. 12 Design study of a mobile 75 kW SOFC-GT power train system

These developments initiated the own study about the feasibility of batteries mobile SOFC-GT systems based on the EXCO cycle. These results led microturbine to the conceptual design as shown in fuel prefig. 12. The necessary information heater SOFC module about the dimensions of the car was el. motor taken from an actual compact midclass car. The pressurised SOFC module including the air and the fuel heaters is under the hood of the car. The microturbine is flanged directly at the SOFC module. The batteries, the power electronic and the fuel tank are positioned in the mid and at the rear of the car. One interesting result of the design study is that the placement of the components left a still available space. Thus the failure tolerance of the system design is comparable high because there is still a reserve of space available that could be used for possible corrections if the further developments should show that some assumptions had been too optimistic (17). generator

General Thermodynamics and Innovation Studies Beside the direct engineering oriented work there had been always a need to control the modelling results by more basic and simple thermodynamic models. Fig. 3 above showed this principle. The first results - already published in (1), (2,), (3), (18) - showed the possibility of using the reversible process model to identify general design principles. Simple modifications of the reversible structures as using exergetic efficiencies (ratio of real work to reversible work) or to modify the structure slightly by using a heat exchanger instead a system of reversible heat engines helps to identify influences on design and operation caused by the system architecture and not by a certain behaviour of the component. However the influence of the excess air on the cell operation is very limited, but its contribution to the losses of the en-

tire system and the cost is enormous. A high excess air e.g. needs a high rate of exchanged heat to restrict the losses this leads to high cost of the heat exchangers and a reduction of the heat recovery leads to a strong increase of heat losses. This principle is general and independent of the technical solution of the single components itself. Energy saving strategies and related product development need an input about the thermodynamic background of the process they want to influence and control. Again the approach developed for fuel cell systems can be used, if we identify the reversible structure within complex architectures. The system integration of the fuel cell system was a first step in a broader approach of technologies by these methodology (19). The use of reversible process chains allows in a simple way the estimation of the potential of new technologies and their thermodynamic value can be evaluated by a chart (20). It becomes obvious that the re-use of energy is the main outline for a more reversible strategy. Batteries are thus complementary and not competing technologies for fuel cells because they recover electric power in an adequate system architecture and fuel cells equalises the irreversibility by its power generation. Thus hybrid fuel cell cars are products with a reversible architecture (20). The analysis of overseas technology e.g. reported in (21) the definition of economic targets for development (22) and contributions to develop a European standardisation strategies (23) are based on the application of these very basic thermodynamic considerations. They are a scientific base to identify early coming developments of technology and to develop strategies for their implementation and use. The practical application of these activities becomes obvious if we consider that actual power trains in automobile application have an amount of about 80% mechanical and about 20% electric and electronic parts. The use of hybrid fuel cell cars would lead to systems with only about 10 % mechanical parts. The relation to the workplaces in the European automotive industries explains the societal need of applied thermodynamics.

References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)

W. Winkler, Brennstoff-Wärme-Kraft 45 (1993) Heft 6. p. 302 - 307. W. Winkler, Brennstoff-Wärme-Kraft 44 (1992) Heft 12. p. 533 - 538. W. Winkler, Proceedings 2nd IFCC. NEDO, Kobe, Japan. 1996. p. 397 - 400. W. Winkler, Proceedings 3rd EUROPEAN SOLID OXIDE FUEL CELL FORUM in Nantes. 1998. Ed. Philippe Stevens. Oral Presentations. p. 525 - 534. W. Winkler, H. Lorenz, Proceedings 4th EUROPEAN SOLID OXIDE FUEL CELL FORUM in Lucerne. July 2000. p 413 - 420. W. Winkler, P. Nehter: Proceedings 9th International Symposium on Solid Oxide Fuel Cells. American Electrochemical Society, Quebec, Canada. 2005. p. 229 - 239. W. Winkler, J. Koeppen, Fourth Grove Fuel Cell Symposium Commonwealth Institute in London. Journal of Power Sources 61, (1996) p. 201 - 204. W. Winkler, European Fuel Cell News, Vol. 5, No 2, April (1998), p 9 -1 4. W. Winkler, J. Krüger, Journal of Power Sources 71, (1998), p 244 - 248. W. Winkler, J. Krüger, M. Sax, R. Telle, Proceedings 3rd EUROPEAN SOLID OXIDE FUEL CELL FORUM in Nantes.1998.Ed. Philippe Stevens. Posters p 245 - 254. W. Winkler, Proceedings 6th International Symposium on Solid Oxide Fuel Cells. American Electrochemical Society, Honolulu, USA. 1999. p. 1150 - 1159. W. Winkler, Journal of Power Sources. 86 (2000), p 449 - 454. W. Winkler, H. Lorenz, Proceedings 7th International Symposium on Solid Oxide Fuel Cells. American Electrochemical Society, Tsukuba, Japan. 2001. p. 1089 - 1094.

(14) (15) (16) (17) (18) (19) (20) (21)

(22) (23)

W. Winkler, Proceedings First EUROPEAN SOLID OXIDE FUEL CELL FORUM. 3 – 7 October 1994. Ulf Bossel. Lucerne. 1994. p. 821 - 848. T. Alston, K. Kendall, M. Palin, M. Prica, P. Windibank, Journal of Power Sources 71, 271 (1998), p 271 - 274. W. Winkler, European Fuel Cell News, Vol. 6, No 2, July 1999, p. 8 - 11. W. Winkler, H. Lorenz, Proceedings 7th International Symposium on Solid Oxide Fuel Cells. American Electrochemical Society, Tsukuba, Japan. 2001. p. 196 - 204. W. Winkler, Brennstoff-Wärme-Kraft 46 (1994) Heft 7/8. S 334 – 340 W. Winkler, Proceedings FUEL CELL WORLD in Lucerne. Marcus Nurdin editor. July 2002. p. 339 - 355 W. Winkler, Proceedings of FUELCELL2005: The 3rd International Conference on Fuel Cell Science, Engineering and Technology May 23 – 25, 2005, Ypsilanti, MI W. Winkler, Analysis of actual RTD on SOFC architectures in Japan and US and conclusions. Revised and expanded version 15th December 2004. Report written for the EU funded thematic network SOFCnet W. Winkler, Proceedings of IIR Conference: Quantifying the Costs & Commercial Viability of Fuel Cell Technology for Power Generation, London November 1998. W. Winkler, 1st CEN-CENELEC Annual Meeting 2005. 'European standardization : Advancing the Lisbon Strategy'. Wednesday 8th June 2005. Assembly Hall of the Upper House of the Hungarian Parliament, Budapest