CORDIS
EU research results

CORDIS

English EN
Advanced Multi-Fuel Reformer for Fuel Cell CHP Systems

Advanced Multi-Fuel Reformer for Fuel Cell CHP Systems

Final Report Summary - REFORCELL (Advanced Multi-Fuel Reformer for Fuel Cell CHP Systems)

Executive Summary:
ReforCELL is a 47 months project focussing on developing a high efficient PEM fuel cell micro Combined Heat and Power cogeneration system (net energy efficiency > 42% and overall efficiency > 90%) based on a novel, more efficient and cheaper hydrogen reformer production unit together to the new design of the subcomponent for the BoP. The main focus of ReforCELL is to develop a new catalytic membrane reformer for pure hydrogen production (5 Nm3/h) in order to intensify the hydrogen production process through the integration of reforming and hydrogen purification in one single unit. The novel reactor will be more efficient than the state-of-the-art technology due to an optimal design aimed at circumventing mass and heat transfer resistances. Besides, the design and optimization of the subcomponents for the BoP has been also addressed.

The project brings together 11 partners covering the whole value chain, ranging from catalyst and membrane developers, reactor and system (BoP) developers, stack developers, service providers and end users.

Novel catalyst and membrane reactors for ATR membrane reactor:
Novel catalyst and membrane reactors for ATR membrane reactor has been developed in the frame of ReforCELL (TRL 4). Lower temperature (600 °C) reforming catalyst Ru based, supported on Ceria/Zirconia, has been developed achieving a good activity and long-term stability compared to the Nickel based commercial catalysts at high temperature (800 °C). Moreover fluidization regime allowed to prevent formation of hot spots increasing its life time. A final 2.5 kg batch of catalyst and an additional batch of 7 kg of inert filler, both with a defined particle size, were produced for the pilot prototype.

Micro-channel configured membrane module with channels in the feed section and porous stainless steel in the permeate section have been developed. Integrated membrane reformer experiments have been performed using a commercial Ni-based catalyst and the Ru/Ce0.75Zr0.25O2 reforming catalyst as developed by Hybrid Catalysis in the project. Compared to the conventional pack-bed reactor, which results in an equilibrium methane conversion below 40%, the continuous hydrogen removal through the Pd77Ag23 membrane shifts the equilibrium and therefore a higher methane conversion can be obtained. At W/F ratio of 27 gcath/molCH4 (GHSV = 6000 h-1), a methane conversion >95% and a H2 production rate of 4.5 Nm3·m-2·h-1 was obtained at 550 °C and a feed pressure of 6 bar.

In addition, Pd-based membranes by direct deposition of thin dense metal layers (< 5 µm) onto porous ceramic and metallic tubular supports by simultaneous ELP have been developed. Experimental tests on H2 permeance and H2/N2 ideal selectivity pointed out performances over both the project and DoE targets. For the ceramic one, a H2 permeance of ~3 x 10-6 mol m-2 s-1 Pa-1 and H2/N2 ideal selectivity of ~10000 has been obtained at 400 °C; while for the metallic supported, even after 1200 h, the H2/N2 ideal selectivity has been >150000 with a H2 permeance ~9 x 10-7 mol m-2 s-1 Pa-1 at 400 °C at lab-scale. However, the prepared Pd-Ag membranes showed a decrease of selectivity after prolonged operation at at 600 ºC due to defect formation in the membrane. Membranes were stable when working < 525 ºC.

Membrane reactor a lab-scale:
A fluidized bed multi-membrane reactor (for testing of 5 membranes) was designed and constructed by TU/e for SMR/ATR of methane at different operating conditions (p, T). First, membrane- catalyst interaction and integration strategies for the different components (i.e. sealing) were investigated. Then, membranes by TECNALIA have been integrated in membrane reactors in fluidized conditions showing that equilibrium conditions can be achieved with hydrogen recovery in the order of 30% with a single membrane (TRL 4).

Experimental results were used to validate a fluidized bed membrane reactor model allowing the overall membrane reformer design and its main characteristic. Additionally, new sealing techniques have been studied for ceramic supported membranes.

Design and manufacturing of novel ATR reformer:
An original design of ATR membrane reactor has been developed. The design of the reactor employs 15 membranes with a total 0.18 m2 of membrane area. The reactor is designed for operating at 8 bara at 600 °C with a maximum production capacity of 5 Nm3/h and capable to modulate to a minimum output of 1.5 Nm3/h of hydrogen. The novel fuel processor module was built using the membrane reactor and all needed balance of plant automated and controlled by PLC. The prototype was tested for ca. 220 hours. During the testing, the ATR reactor operated stable and reached a production with a yield of 1.7 Nm3 of H2 per Nm3 of CH4 feed. An ad-hoc model simulates the performance of the fluidized bed membrane reactor. The model can be used for calculation of membrane area and to predict the influence of different operating variables.

Integration and validation in the CHP system:
Main achievements in the frame of the system integration have been the following:
- Fuel cell relevant core components have been selected. A set of operating conditions relevant for the system and application considered have been defined, testing a short stack with these conditions to dimension the final fuel cell stack prototype to be integrated in the final system. This prototype has been manufactured, and tested before the installation into the CHP system.
- A reference case to benchmark the performances and costs of the innovative system have been define, then two different membrane reactor lay-out were investigated: one with a sweep flow at the permeate side the other with a vacuum pump. Finally this two layout have been investigated with biogas as fuel feeding.
- The selection and design of components and of the full system have been carried out taking into account the compatibility of main magnitudes (pressure, temperature...) and materials in the system, as well as costs. ATR and fuel cell stack interface and the whole system control strategy have been defined;
- Size scale up has been assessed. Technical feasibility for sizes up to 50 kWe has been verified for the main system parts (ATR, fuel cell stack, BoP) and the final specific cost (€/kW) of the whole system has been evaluated. Finally, flexibility and marketable size have been assessed.


Project Context and Objectives:
Stationary fuel cells offer a clean and efficient source of electricity in systems ranging from 1 kW up to 1 MW or more. With appropriate fuel processing technology, fuel cells are able to tap into established or accessible sources of fuels such as natural gas but also various other fuels including biofuels and bio-gases. With cogeneration or combined heat and power (CHP), efficiencies improve dramatically from 30–50 % up to as high as 80-90 % with significant primary energy savings. The application off domestic micro-CHP demonstrates (see Figure 1) that a minimum of 15% extra primary energy is required for the traditional supply when compared to the fuel cell m-CHP system. In spite of these demonstrated benefits, cost and reliability issues make the technologies’ long-term potential difficult to predict. In order to reduce costs and increase the reliability of the technology, work must be done on fuel processing design and system optimization.

ReforCELL aims at developing a high efficient PEM fuel cell micro Combined Heat and Power cogeneration system (net energy efficiency > 42% and overall efficiency > 90%) based on a novel, more efficient and cheaper hydrogen reformer production unit together to the new design of the subcomponent for the BoP.

This new high efficient PEM fuel cell mCHP system is based on:

- The design, construction and testing of an advanced catalytic membrane reactor for pure hydrogen production (5 Nm3/h) from reforming. The reactor components (catalyst, membranes, heat management etc.) were deeply investigated and optimized. The novel reactor is more efficient than the state-of-the-art technology due to an optimal design aimed at circumventing mass and heat transfer resistances.
- The design and optimization of the subcomponents (BoP) for the integration of the membrane reformer to the fuel cell stack.

The main idea of ReforCELL was to apply the concept of process intensification in the production of hydrogen. ReforCELL develops a novel more efficient and cheaper membrane reactor by intensifying the process of hydrogen production through the integration of reforming and purification in one single unit. While traditional reformers includes several steps for producing H2 with enough quality to feed the fuel cell stack (see Figure 2), the new concept addressed in ReforCELL reduces this process to one step (Catalytic Membrane Reactor, see Figure 3). In addition, there is a reduction of the other components in the reformer (heat exchangers) and in the BoP (auxiliary elements) when integrating the membrane reformer to the stack and building up the m-CHP system. In addition, there is a reduction of the reforming temperature.

This general objective is directly related to the development of a novel catalytic membrane reactor (CMR) for hydrogen production with:

- Improved performance (high conversion at low temperature for the autothermal reforming reaction)
- Enhanced efficiency (reduction of the energy consumption)
- Long durability under CHP system working conditions
- Clean environmental operating conditions (CO2 emissions reduced from conventional reformer).
- and including a good recyclability of its individual components and safety aspects for its integration in domestic CHP systems.

The technical objectives needed to achieve these goals with the novel multi-fuel processor (based in CMR) were the following:

- Develop an advanced catalyst able to catalyse different reforming reactions under moderate (<700ºC) conditions and resistant to sulphur compounds and coke formation and at reduced cost.
- Develop new hydrogen permeable membrane materials with improved separation properties, long durability, and with reduced cost, to be used under reactive conditions.
- To assess the large scale production of the membrane developed.
- Understand the fundamental physico-chemical mechanisms and the relationship between structure/property/performance and manufacturing process in membranes and catalysts, in order to achieve radical improvements in membrane reactors.
- To design, model and build up novel more efficient (e.g. reducing the number of steps) multi-fuel catalytic membrane reactor configurations based on the new membranes and catalysts for small-scale pure hydrogen production.
- To validate the new membrane reactor configurations, and design a semi-industrial Autothermal Reforming (ATR) prototype for pure hydrogen production.
- To improve the cost efficiency of membrane reactors by increasing their performance, decreasing the raw materials consumption and the associated energy losses.

Other technical objectives were related to the integration and validation of the multi-fuel reformer into the PEM fuel cell CHP system:

- To design the optimum CHP system (aided by simulation tools) in order to achieve a complete system able to achieve the targets in performance and cost.
- To test the reliability of the novel reactor with a Fuel Cell CHP system
- To assess the health, safety and environmental impact of the system developed, including a complete Life Cycle Analysis (LCA), of the developed system.

The ReforCELL work plan consisted on activities related to the whole product chain: i.e. development of materials/components (membranes, supports, seals, catalyst...) through integration/validation at lab-scale, until development/validation of pilot scale ATR-CMR and the proof of concept / validation of the new PEM fuel cell m-CHP system.. For a maximum impact on the European industry this research, covering the complete value chain of micro-CHP fuel cell systems, can only be carried out with a multidisciplinary and complementary team having the right expertise, including top level European Research Institutes and Universities (6 RES) working together with representative top industries (4 SME and 1 IND) in different sectors (from materials to micro-CHP developers).

The ReforCELL Project has been funded under Fuel Cell and Hydrogen Joint Undertaking. The Project started the 1st of February of 2012 and it has run for 47 months.
(See attached pdf file)
Project Results:
4.1.3. Main S&T results/foregrounds

4.1.3.1. Industrial specification of Fuel Cell CHP-System

One of the first steps in ReforCELL was the assessment of the requirements for the introduction of the novel PEM fuel cell m-CHP system, including the innovative multi-fuel processor, in the market. Therefore, state-of-the-art reactors, PEMFC, CHP systems were identified and in-deep assessment of the components and process parameters was carried out.

According to the study the typical subsystems in this kind of systems are the following:
➢ Desulphurizer
➢ fuel cells
➢ fuel processor
➢ inverter
➢ BOP
➢ Controller

The main requirements on the subsystems addressed in the project are discussed hereafter. The fuel processor must be suitable for supplying pure H2 or syngas of adequate quality for feeding the stacks involved. However, there are few commercial fuel processors having the size of interest for the project. In addition, they are very expensive due to also the use of precious metal (palladium, platinum,...).

The BoP includes all the remaining components, such as the demineraliser, pumps, valves, etc. For reasons connected to reproducibility, cost and quality, it is essential to use commercial components and not to develop them or require suppliers to develop them ad hoc.

Reference systems in the market use fuel cells with polymer membranes (LT-PEM and HT-PEM) as well as ceramic ones (SOFC). Many of the available systems have been found within the CALLUX project. Table 1 summarizes systems based on fuel cell already available in the market, or near to commercialization.


The size of most of the available systems are not directly comparable to the one developed within the ReforCELL project. These systems have an electrical production of around 1 kWe, while in ReforCELL the objective was to produce 5kWe. Nevertheless, the evaluation of these systems was a good guideline to define the context in which the ReforCELL is being developed and to understand the possibilities of its commercial success.

It is necessary to identify the market where a product should be introduced to evaluate correctly the industrial needs of this product. The energy market is very diverse and it can change greatly depending on the country taken into consideration. The following variables could be considered for the analysis (which could change from country to country):
➢ Government grants;
➢ Cost of electricity;
➢ Cost of natural gas;
➢ Standards and regulations;
➢ Average consumption of electric energy per capita;
➢ Average consumption of domestic hot water per capita;
➢ Period of heating and consumption per m2.
➢ Average dwelling size;

However, government grants of each country have not been considered in order to perform an analysis that could be applied in all countries.

The investigation has been performed taking into account 4 countries as sample representatives of the different European areas: Italy, The United Kingdom, Germany and The Netherlands. Data shown in Table 2 are relevant to the countries considered for the evaluation.

ReforCELL system has been compared with Stirling engines and internal combustion reciprocating engines (i.c.r.e) for the correct evaluation of the performances when having it in the market. Parameters such powers, cost, maintenance and payback time have been considered. Table 3 describes some main features of the systems that have been taken into account.

These kinds of systems have maximum payback when all of the electrical and heat energy produced is used. It is important that they operate as many hours per year as possible. Moreover, as efficiency is normally higher when working at full capacity rather than with reduced loads, full power cogenerator has been always considered for the analysis.

When working a full capacity using the whole heat and electricity production, the installation must supply different family-dwellings and include a back-up boiler. Any other use, for example at partial load or with standstill periods due to lack of necessity or for periodic maintenance, leads inevitably to a longer payback times. Due to the great difference between the purchase cost and that of the sale of electric energy, the system is only interesting if inserted into a context in which all of the current produced is used. Therefore 8500 total operating hours per year have been considered, with a standstill period for routine maintenance of about 250 hours. The result has been a heating season of 182 days, corresponding to 4,368 operating hours and 4,132 summer operating hours. This is the configuration that allows maximum payback, even if Stirling engines and the i.c.r.e. probably never work in these conditions. I.c.r.e. in particular require an ordinary maintenance every 1,000, 1,500 hours. The cost for boiler maintenance have been also considered which, for wall-hung boilers, is about 100 € per family.

With reference to residences and considering the use of the complete electricity and heat production, it has been seen that buildings with more than 14 dwellings have the best size for the introduction of the ReforCELL system. The proposed configuration can cover the yearly electricity requirements by also using the recovered heat. Logically, the back-up boiler is required, especially for winter heating requirements, as well as a cylinder for heat storage. The potential saving can vary from 4900 to 8900 €/year depending on the reference country.

To recover the total heat produced during the summer in order to increase yearly savings, the Stirling engines require a number of dwellings that can be compared with the ReforCELL system. At this point, they also require a back-up boiler and cylinder for heat storage. In this case, the yearly saving varies between 1,170 and 2,030 € according to the countries considered in this study. Instead, i.c.r.e. systems can satisfy 100 dwellings, requiring back-up boiler and cylinder for heat storage. In this case, the yearly saving can be around 19,700 and 35,600€ depending on the sample country. Table 4 summarizes data collected specifically from the cogenerators.

From Table 1 it is possible to understand that the i.c.r.e. is the most attractive for the final user in terms of family investment and payback time. However, they have a size limit, which is generally over 20 kWe. The necessity to install them in large residential complexes with a large number of families therefore limits diffusion. As the Stirling engines are much smaller, they can be potentially used widely. However, they have higher costs and long pay back times.

This analysis shows how the ReforCELL system is potentially more competitive, also when compared with more traditional cogeneration systems as long as maintenance costs are kept low, below 1000€/year, and have a long life time, with performance that remains stable for 10 years. Thus, duration and maintenance are fundamental for the success of the system being developed within the project.

The features that the m-CHP system developed in ReforCELL should have in order to be industrially attractive are shown in Table 5.

4.1.3.2. Novel catalytic materials

A novel ruthenium based ATR catalyst consisting of 2% ruthenium on a ceria zirconia support was developed. The synthesis of the catalyst consists of a sequential procedure where first the ceria-zirconia support, having the overall formula Ce0.75Zr0.25O2 is prepared. This support is then fractionized to the desired particle size and subsequently loaded with the active ruthenium metal. Initial tests have shown that controlled precipitation gives the best result for the support in terms of surface area and particle size.

A set of different support particle sizes was sent to TU/e for testing of the fluidization behaviour in a fluidized bed reactor. The tests showed stable fluidization behaviour for the different particle sizes. In addition to the support an inert material having the same bulk density and particle size as the support was sent to TU/e. Fluidization tests showed that there was no separation of the support and inert filler under fluidization conditions. From the tests the fraction having a particle size of 125-250 µm was chosen as the most interesting candidate for the pilot scale reactor.

The catalyst preparation was optimized for production on kilogram scale using a 10 L batch reactor (Figure 4), and a 0.5 kg batch of the catalyst having a particle size of 125-250 µm was prepared.

XRD analysis of the Ce0.75Zr0.25O2 support shows that the material exhibits the cubic structure of the ceria. The absence of monoclinic and tetragonal zirconia phases, when using the co-precipitation methods, indicates that the zirconium is incorporated into the ceria structure (Figure 5). The average crystallite size of 30 nm as derived from the line broadening is in accordance with the high resolution TEM images of the material. From the XRD and physisorption there is no difference between the materials made on laboratory or kilogram scale. This shows that the support was successfully prepared at kilogram scale. The high BET surface area of ~90 m2/g is a direct result of the small crystallite size as seen in the XRD and TEM measurements.

The ruthenium was loaded onto the support via controlled precipitation. This technique combines a good dispersion with the possibility of upscaling whereas incipient wetness has a poor dispersion and surfactant assisted routes are more difficult to scale up. The ruthenium load of 2% was confirmed by XPS analysis of the catalyst. Both XRD and N2 physisorption did not show large changes in the material after the ruthenium loading (Figure 6), which shows that the original structure of the support has been preserved during the loading.

The new catalyst was tested for its performance in a custom built lab scale reactor. During the test the conditions were varied to test the performance of the catalyst under the various conditions as can be expected in the pilot plant reactor. Figure 7 gives an overview of the performance of the ruthenium based ATR catalyst both under SMR and ATR conditions.

The graph shows that the H2 production over the catalyst is stable under the various SMR and ATR conditions, and that the catalyst operates at the desired low temperature window of 500 – 600 °C.

Cold isostatic pressing followed by crushing and sieving the catalyst was found as a suitable method for the production the required sieve fractions. A 2.5 kg batch of sieved catalyst together with 7.5 kg of inert filler was shipped to HYGEAR for testing in the pilot plant.

4.1.3.3. Membranes development

Tubular metallic supported membranes
TECNALIA has developed Interdiffusion layers with suitable surface properties and gas permeation around the target. Only a minor modification of the YSZ top layer thickness is required to ensure high enough permeance in all the supports before depositing the selective layer. The powder spraying techniques used during the project (detonation spraying, atmospheric plasma spraying) were found to be unsuitable for preparing ceramic interdiffusion barrier layers with suitable surface quality for thin Pd-Ag membranes (≤5 microns). During the 2nd period, wet deposition technique was optimized to deposit YSZ-Al2O3 based layers onto porous metallic supports obtaining layers with suitable surface quality.

TECNALIA prepared ~5 microns thick Pd-Ag membrane supported on ceramic coated Hastelloy X porous tube by direct electroless plating technique deposition (ELP). Direct PVD deposition was found to be unsuitable for obtaining defect-free Pd-Ag layers onto porous supports, and 4-5 microns thick Pd-Ag membranes on ceramic porous supports were therefore prepared by electroless plating technique. The combination of surface treated Hastelloy X support with the deposition of YSZ-Al2O3 layers by wet deposition technique provided the suitable surface quality for depositing ~5 µm thick Pd-Ag layers. TU/e tested one of these membranes in a fluidized bed membrane reactor for ATR and SMR reactions at TU/e. Single gas permeation tests during more than 800 hours at temperatures between 500 and 600 °C showed high H2 permeance (~ 1 x 10-6 mol m-2 s-1 Pa-1 at 600 ºC) with exceptional ideal H2/N2 selectivity always above 200,000 (see figure hereafter).

Prolonged operation at 600 °C resulted in the presence of defects on the surface of the membrane associated to an increase in N2 permeance with the consequence of a pronounced decrease in ideal selectivity (of 2,650 after 200 h) (see Figure 8). Catalyst interaction with the Pd-Ag layer has not been observed since H2 permeance is the same for single gas tests with empty tube configuration and fluidized bed configuration. This also implies the absence of mass transfer resistances caused by the particles. (Close to) thermodynamic equilibrium is achieved both in ATR and SMR membrane-assisted processes. However, a full conversion of CH4 is never achieved. This can be partially explained by the large amount of inert gas fed together with the CH4:H2O. This decreases dramatically the partial pressure of the other components. After all the experiments a decrease in ideal H2/N2 selectivity has been observed due to the defects created in the surface. As determined through a test with the membrane submerged in ethanol, all defects are associated to micropores in the surface, which is also observed through SEM images.

SINTEF has fabricated tubular porous stainless steel-supported membranes (8 cm long). The Pd-alloy films are made using magnetron sputtering onto silicon wafers. In a second step the film is removed from the wafer allowing the preparation of very thin membranes applied on macroporous supports. The prepared membranes show a very good selectivity (H2/N2 > 50,000 and the criteria for ReForCell is therefore met for operation temperature relevant for reforming of ethanol; T up to 450 °C (and P <10 bar).

Tubular ceramic supported membranes
~4 µm microns thick Pd-Ag ceramic supported membranes have been prepared at TECNALIA by direct ELP deposition (support: ZrO2 110 nm top layer asymmetric porous tubes). Then, they have been characterized and tested for high-temperature fluidized bed membrane reactor applications at TU/e. New high-temperature Swagelok fittings with graphite ferrules have been optimized to seal the membranes at TU/e, and a leak-tight sealing at 600 ºC for seven days under fluidization conditions has been achieved. The H2 permeance of these thin membranes (~ 5 x 10-6 mol m-2 s-1 Pa-1 at 600 °C) is at least two times higher compared to other thicker hydrogen-selective membranes reported in the literature. Both SMR and ATR have carried out with these thin-film supported membranes in fluidized bed membrane reactors, showing significant improvements in the performance compared to commercial membranes. However, the prepared Pd-Ag membranes showed defects after single gas tests after seven days at 600 °C. The cause of the pinhole formation is not clear yet, as it can be related to the membrane preparation procedure or ceramic support layer sintering.

Micro-channel supported membranes
SINTEF has developed microstructured membrane modules that reduce gas phase diffusion limitations and that increase the membrane area to reactor volume ratio compared to traditional tubular reactors.

Since concentration polarization effects are expected to be reduced in such modules, a high space–time-yield is anticipated due to the supplied high volumetric surface area for reaction and membrane separation. An example of a membrane module employing a micro-channel configured feed section given in Figure 9-a).

In the ReforCELL project the development has focused on the following topics:

Stability studies of microchannel-supported thin Pd-alloy films: the long-term stability of various designs of microstructured membrane modules has extensively been investigated. In the experiments, the H2 permeation performance and stability of the modules are verified over a period of up to 50 days. Operation of micro-channel modules that employ a stainless steel plate with apertures on the permeate side results in a large settling of the film into the permeate section; ultimately this will result in a membrane failure. The operation limits are ~450 °C and pressures up to 5 bars.

Integration of the microchannel feed section with a porous stainless steel (PSS) support: For pressures above 5 bars a porous metallic support is introduced for sufficient stabilisation of the thin Pd77Ag23 films (see Figure 9-b)). For such a module, a hydrogen flux of 195.3 mL·min-1·cm-2 was obtained at 5 bars and 450 °C. The module shows a very good stability up to the highest feed pressure applied of 15 bars at a H2/N2 permselectivity > 39.000. The temperature stability is improved by the introduction of YSZ IMDBL, and selective operation has been obtained for 160h at 550°C.

Development of enlarged microchannel supported module: A larger micro-channel membrane reactor with micro-structured plates with dimensions ca 17 times larger than the lab-scale module has been developed.

Membrane-enhanced reforming reactions applying the microchannel modules: Both non-integrated sequential and integrated membrane reformer process design have been evaluated in terms of membrane module stability and performance under varying methane steam reforming conditions. A catalyst bed has been applied into the feed section channels, see Figure 9-c. The main results are summarized below:
- During non-integrated membrane and catalyst tests a stable membrane operation was obtained at 400 and 450 °C. Even though only one membrane separator stage is applied, the membrane module is able to operate at a H2 recovery factor (HRF) of ~42% at 400 °C and ~44% at 450 °C.
- Integrated membrane reformer systems were tested up to 8 bars and 550 °C during periods of up to 75 days applying either a commercial Ni-based catalyst or the Ru-based catalyst developed by HYBRID. The Ru-based catalyst is less prone to coke formation than the Ni-catalyst and more active at low temperatures.
- The performance of the Ru-catalyst integrated reactor system is limited by the H2 removal rate through the membrane. This contradicts to the Ni-based catalyst which limits the performance.
- Compared to the conventional packed-bed reactor with an equilibrium methane conversion < 40%, the continuous hydrogen removal shifts the equilibrium and therefore a higher methane conversion is obtained. At a W/F ratio of 27 gcath/molCH4 (GHSV = 6,000 h-1), a methane conversion >95% and a H2 production rate of 4.5 Nm3·m-2·h-1 was obtained at 550 ºC and a feed pressure of 6 bar, see Figure 10.

Further improvements to the reactor design can be implemented in terms of membrane surface area per catalyst volume by a decrease in channel dimension from the currently applied width and height of 1 mm. This parameter can be exploited to tune the membrane surface area to the catalytic activity per volume unit of the applied catalyst.

Manufacturing of dense-metal membranes for integration into prototype unit
As requested by HYGEAR, 34 membranes have been prepared at TECNALIA and sealed at TU/e using Swagelok with graphite ferrules (see Figure 11). The membranes consist of 4-5 µm thick Pd-Ag layers deposited onto ZrO2 110 nm top layer asymmetric porous tubes prepared by direct ELP deposition. The final length of the membranes after sealing is of 19-20 cm.

Analysis of production costs and scale up of the membrane production technology unit
A cost analysis of an up-scaled production of Pd-based membranes has been performed. Depending on the production method and type of support applied (ceramic and metallic) a cost per square meter of membrane ranging from 3000-5000 euro/m2 has been assessed.

In the case of metallic supported 4-5 microns thick Pd-Ag membranes prepared by direct ELP deposition at TECNALIA, the metallic support is the main cost of the total of the membrane. In the case of the ceramic supported 4 microns thick membrane prepared by ELP, the palladium and the ceramic support have very similar cost.

On the other hand, the total cost for production of membranes by magnetron sputtering (two stage process) developed at SINTEF is 5,000 euro/m2 (for 5 micron thick Pd-based membranes). In this case, the cost of the support is the main constituent of the overall membrane cost (~50%).

Non-Pd alloys
For the first generation of non-Pd membranes, pure V based composite layers (Pd/V/Pd) were developed using DC power supplies by PVD magnetron sputtering at TECNALIA. The membranes were tested at 300 °C and were fragilized after H2 exposure.

According to theoretical calculations, amorphous Zr30Cu60Ti10 has similar H2 perm as Pd at T > 600 K. Pd/Zr30Cu60Ti10/Pd made by sputtering at SINTEF shows however a significant reduction in hydrogen flux with temperature cycling caused by recrystallization during testing at 450 °C. The work on non-Pd alloys was terminated shortly after M18.

4.1.3.4. Lab-scale ATR-CMR fuel reformer

The selected lab scale reformer is the fluidized bed reactor. TU/e has developed the lab-scale Fluidized Bed Membrane Reactor (to accommodate up to 5 membranes). This reactor has been tested up to 6 bar and up to 650 °C with commercial membranes and commercial catalysts. Tecnalia delivered more than 10 Pd-alloy membranes, deposited on both ceramic and metal supports. Hybrid Catalysis synthetized a new generation of catalyst, based on Ru and supported on CeO2-ZrO2 matrix. TU/e has tested the mechanical stability of such particles in fluidized bed operation and no loss of particle size has been observed even under high temperature operation in bubbling fluidization regime.

The Pd-alloy membranes manufactured by TECNALIA were sealed using commercial standard Swagelok connectors (316 SS) together with graphite gaskets for 3/8” OD tubes provided by CHROMalytic TECH(nology) Pty Ltd. The graphite gaskets were sized to the outer membrane diameter of 10.1 – 10.5 mm and pretreated with a membrane dummy to the standard Swagelok connector. Figure 12 shows a schematic of the sealing (a) and a picture of the sealed tubular Pd-alloy membrane with graphite gaskets to a standard Swagelok connector (b). The bottom part of the connector has been specially designed for membranes that will be immersed vertically in the fluidized bed to avoid gas holdup below the membrane. In case of integration in packed bed reactors, simple Swagelok caps could be used.

The membranes were tested for single gas permeation, mixed gas permeation and for steam reforming and autothermal reforming of methane in fluidized bed membrane reactors.

In Figure 13-a the hydrogen flux through a sealed membrane at different hydrogen partial pressures and different temperatures between 380 – 600 °C is shown after the stability tests were performed. The hydrogen permeation rate is increasing with increasing transmembrane partial pressure difference and temperature, as expected. The tested membrane shows an almost perfect linear behaviour for the pressure exponential factor n = 0.5 (R2 >0.995) which is typical for Pd-alloy membranes at low pressures, if bulk-diffusion through the membrane is the rate limiting step according to Sieverts’ law [5]. The membrane parameters for the tested membrane have been determined at 10 kJ·mol-1 for the activation energy (∆Eact) and 6.93·10−8 mol m-2 Pa-0.5 s-1 for the pre-exponential factor (P0) using the plot of the logarithm of the permeance against the reciprocal temperature (shown in Figure 13-b).

The first test with five tubular membranes prepared by TECNALIA has been performed with conventional SMR. Figure 14 and Figure 15 show the obtained retentate compositions together with the achieved methane conversion. It can be seen that methane conversion increases with increasing temperatures, where the extracted hydrogen stream includes a noticeable amount of CO. Nevertheless, the hydrogen purity still remains above 99.98 % for all cases. During the experiments it was observed that the amount of CO in the permeate stream increased from the first day of tests to the second day of tests at the same operation conditions. An increase of the temperature also leads to an increase of the CO impurity in the permeate stream.

Table 6 gives a summary of the different operating conditions used in the SMR tests. The results clearly show the increase in the hydrogen recovery at higher temperatures due to the increased hydrogen permeability of the membranes at higher temperatures. During the first day of tests the equilibrium conversion was reached as expected, but remarkably during the second day of experiments, the methane conversion was higher at the same operating conditions and increased even slightly beyond the equilibrium conversion. Comparing the equilibrium conversion of methane and the conversion obtained during the experiments, especially in the second day, a significant increase has been detected. Analysing the surface of the membranes and the sealing after these tests, it was observed that the sealing was damaged, which could be the reason for the CO increase in the permeate gas stream. The membrane surface was free of defects, which assures that the membranes can survive under reforming conditions for this test duration.

In a second test, ATR of methane has been performed using the five tubular membranes prepared by TECNALIA, and the retentate stream composition and methane conversion are shown in Figure 16. The feed conditions have been different to the test performed with the REB membranes, so the results obtained in terms of conversion of methane cannot be directly ascribed to the different membranes, while the hydrogen recovery and separation can.

Comparing the results obtained for ATR reforming without and with the TECNALIA membranes, which are shown in Table 7, the methane conversion has increased by 7.2% from 89.5% to 96.7%. The total amount of produced hydrogen has been increased by 7.7% resulting in a total flow rate of hydrogen of 2.71 l/min. The hydrogen recovery and the purity of the hydrogen stream are still comparable.

4.1.3.5. Design and manufacturing of novel ATR Reformer

The following objectives regarding pilot scale ATR membrane reactor were achieved:
➢ Design of ATR membrane reactor
➢ Design and assembly of test setup for evaluation of the performance of membrane reactor
➢ Conceptual design and implementation of automated controls for membrane reactor
➢ Building and testing the pilot scale membrane reactor
➢ Completion of the final model of ATR membrane reactor
➢ Evaluation of potentials and markets for membrane reactors


Based upon the improved membranes an ATR membrane reactor was developed. The reactor employs 15 membranes and about 0.4 m in length (see Figure 17). The reactor was designed for operating at 8 bara at 600 °C with a maximum production capacity of 5 Nm3/h and capable to modulate to a minimum output of 1.5 Nm3/h of hydrogen. The reaction section includes an array of thermocouples giving information about the axial and radial distribution of the temperatures inside the reactor.
The novel fuel processor module was built using the membrane reactor with all needed balance of plant mounted into a skid for ease of transportation as shown in Figure 18. The fuel processor is completely automated and controlled by a smart controller. The system allows for remote monitoring and control by a separate computer.

The membrane reactor was tested for different operating conditions with ranging pressure, temperature and steam-to-carbon ratio. Difficulties were found in fragility of the supports of the membranes which failed early in the testing phase. Improved protocol for handling, assembling and testing the membranes has been defined. The remaining tests were performed at lower flows and without extraction of hydrogen. During the testing, the ATR reactor reached a stable production of 1.7 Nm3 of H2 per Nm3 of CH4 feed. Flow patterns, heat transfer limits and integration possibilities of the reactor were identified. During testing, the performance was checked in the SCADA panel view (see Figure 19). Several parameters (e. g., pressures, temperatures, flow rate, valve position, etc.), can be also monitored by the graphical control panel.

A model was developed based on the results to simulate the fluidized bed membrane reactor. The model can be used for calculation of membrane area and to analyse the influence of some variables such as reactor temperature, heat loss and operating load. The model also evaluates the required NG, air and steam flow rates to achieve the H2 production target, operating at autothermal condition.

Another model of the complete fuel processor including peripheral components of the balance of plant was also developed to describe the ATR-MR system performance. The system includes the ATR membrane reactor and the auxiliary units to recover the heat source from permeate and retentate streams. The model analyses the influence of some variables such as reactor temperature, heat loss and operating load, allowing an optimization procedure (CAPEX minimization) of the heat exchanging network. The results of ATR-MR system model provide a sort of maps of the required operating condition as a function of the H2 production target. By decreasing the hydrogen output, the O2/C ratio increases, and the requirements for feed of steam relative to the NG feed (S/C) decreases.

Technologies developed in ReforCELL could open doors for commercialization of new products. A market study for economically relevant countries in Europe shows that for units in the size range of 3-8 Nm3/h of hydrogen, a potential market size of nearly 2.25 M units a year can be estimated. It is essential for potential developers to consider specific geographical areas and market segments which are particularly favourable for initial introduction of the systems. The reforming catalyst could compete with commercially available solutions in the sectors of small reformers for stationary fuel cells or hydrogen generators for Hydrogen Refuelling Stations (HRS). Membrane technology may offer in the short term period solutions in markets for dehydrogenation, high value chemicals, and hydrogen upgrading or extraction from mixtures. The ATR membrane reactors could have a role in the diffusion of distributed power generation with fuel cell technology.

The characteristics of the membrane reactor make it especially suitable for their use in combination with stationary fuel cell power generators:
➢ Compactness
➢ Operation at lower temperature
➢ Delivery of low pressure, high purity hydrogen
➢ Proven for stationary applications

The focus is on economies where the technology becomes economically attractive due to a number of factors:
➢ Energy price structure characterized by an elevated electricity to gas price ratio
➢ Prolonged cold seasons
➢ Policy framework favourable for CHP and adoption of innovative technologies for distributed generation

4.1.3.6. Integration and validation in CHP system

The technical objective of this activity was the integration and validation of the m-CHP system including the manufacturing and test of the fuel cell stack prototype to be integrated in the m-CHP. The activities on the fuel cell stack focused on the evaluation of the performance and short term durability of PEMFC short stacks in operating conditions representative of the application and compatible with reformer and system requirements. The activities included the selection of relevant core components available as commercial products or at prototype level, the definition of sets of operating conditions relevant for the application considered, the test of a short stack applying these conditions to get additional information for the dimensioning of the system. Performance tests have been performed as well as some durability test to evaluate the performance short term stability. Finally, manufacturing of the prototype to be integrated in the final system, and the test of this stack have been carried out.

Short stacks (6 or 8 cells) have been made with Membrane Electrodes Assemblies (MEAs) adapted for reformate fuel operation. Operating conditions have been defined based first on state of the art and then on partners information and system requirements for the fuel cell temperature, fuel and air gases pressures, flow rates, humidification, and of course fuel composition. The impact of fuel composition has been evaluated comparing pure hydrogen with mixtures including mainly hydrogen, carbon dioxide (CO2) and some carbon monoxide (CO). Effect of adding some air (air bleeding operation, in the range 0.5 to 2% of air in the fuel) to reduce the impact of CO has been also studied. Two operating modes for fuel feeding have been tested: fuel circulation with fixed stoichiometric ratio and tests to check the fuel cell behaviour in dead end mode with purges.

It has been possible to evaluate the performance of the stack with different gas feeding conditions and to validate that the operating points in terms of cell current/voltage are in the expected range (vs. reference case definition). The impact of some fuel composition has been evaluated considering the voltage losses (in %) versus pure H2.

Within the operating range of 0.35/0.45 A/cm², the polarization curves showed that replacing pure hydrogen by a synthetic reformate including CO2 and some CO has a negative impact but limited to 5% if CO concentration is less than 10 ppm. Air bleeding allowed recovering half of the voltage losses in the same current density range.

The results achieved in dead end mode with pure hydrogen show that the stack can operate with a homogeneous behaviour of the cells with a selected rhythm for the purges, needed to remove the inert gas and water.

More extreme conditions have also been check with particularly the impact of low relative humidity, low pressures, low hydrogen flows and high CO contents (>50 ppm) to check the possible effect of issues related to the system management or to the processor. In parallel, performance stability has been also evaluated by different load cycles, with a day/night type current profile and then following a more specific daily profile planned for the application (Figure 20).

Finally the prototype stack has been designed (specific end-plates), assembled and tested under pure hydrogen. 85 cells stack has been proposed (targeting more than 5 kW at nominal point and 8 kW maximum electric power).

Effect of some parameters important for the system operation has been checked showing no impact of stack temperature between 65°C and 70°C, no impact of fuel relative humidity, and around 20 mV/cell gap at nominal point for air pressure between 1.2 and 1.5 bars.

Under the nominal operating conditions and pure hydrogen, the prototype showed the same performance (average cell voltage) as the short stack (Figure 21).

Another objective of this activity, was focused on the definition of the optimized system lay-out implementing the membrane reactor developed within ReforCELL and the PEM stack. The first step consisted in the definition of a reference case to benchmark the performances and costs of the innovative system. Two reference cases have been considered: the first based on a steam-methane reforming reactor while the second on an autothermal reformer (ATR). The main performances of the two reference cases are summarized in Table 8. Results are consistent with available information of commercial systems based on this technology.

The second step consisted in the evaluation and comparison of different configurations in order to maximize the system performances. In particular, two different membrane reactors lay-outs have been investigated: the main difference being the adoption of a sweep flow at the permeate side to reduce the membrane surface area instead of a vacuum pump. For both cases, several design parameter of the membrane reactor (i.e. temperature, S/C ratio, feed and permeate pressures) as well as other components (i.e. fuel cell current density, burner temperature, etc.) have been considered in order to define the optimal operating conditions.

Example of results is summarized in Figure 22, where calculated electric efficiency and membrane surface areas for different operating conditions of the membrane reactor are reported. In general, the higher is the membrane surface area, the higher is the electric efficiency. Therefore, a preliminary economic assessment is necessary to determine the optimal system design. It can be noted that the maximum electric efficiency is about 40.5%, which is lower than project target (i.e. 42%). The target efficiency can be achieved by increasing the fuel cell conversion efficiency (i.e. reducing the current density), but this option is not worth economic wise, because the additional cost of the fuel cell wouldn’t be balanced. Cases with vacuum pump show even less electric efficiency because of the pump power consumption.

These two layouts have been also investigated with biogas as fuel feeding. The main difference between natural gas and biogas is in the concentration of the methane: in biogas it is about half of an average European natural gas composition. Two different biogas compositions were considered and evaluated: a typical landfill and anaerobic digester compositions.

Efficiency variation for the three different considered fuel compositions and assuming vacuum pump lay-out as function of membrane area is shown in Figure 23.

An important result is about the flexibility of the system with respect to the fuel composition. The membrane reactor can handle and separate pure hydrogen even from diluted methane as in biogas. Biogas requires higher membrane reactor pressures and membrane surface areas because of the dilution, which reduces the permeation driving force. In the case of vacuum pump the efficiency penalty is between 3% and 4%, while it reduces to 1% for sweep case. The net electric efficiency with the same reactor design of NG case ranges from 34% to 39% which is lower than NG. Energy penalties in flexible conversion system dealing with different natural gas occur also in commercial systems.

Based on the optimized system layout fed with NG, BoP components have been investigated. The selection and design of components and of the full system have been carried out taking into account the compatibility of main magnitudes (pressure, temperature...) and materials in the system, as well as costs.

The best solution identified for natural gas compressor is a COLTRI system coupled with a buffer tank provided by ROTH CYLINDERS. While for compressed air, a BOGE compressor has been identified.

The Hexiburner designed by CATATOR, which is a combination of a catalytic burner and a plate heat exchanger, have been selected to recover energy from retentate gas.

The interface between the fuel processor and the fuel cell stack system is driven by a vacuum pump proposed by H2Systems, while the air to the fuel cell stack cathode is supplied by Vairex VRB8-35 air blower and humidified by a PermaPure FC200-780-10. The cooling circuit, which keep the stack temperature into the required range, is equipped with a Swep B5 18 plates heat exchanger and the cooling water is circulated by a Grundfos Magna3 pump.

To manage the CHP system, a PLC is used. It is composed by two identical controllers (IOBLOCK): one communicates with fuel processor PLC, and the other is used to measure each single cell voltage. Concerning interface components (mechanical and electrical interface) and the system control strategy, they have been defined.

The final integration and test of the system have not been finalized due to Soprano liquidation.

On the other side, system size scale up has been assessed, investigating the three main components of the CHP system (ATR, fuel cell stack, balance of plant); technical feasibility and economical aspects have been taken into account. The maximum size considered is 50 kWe according to European Directive 2012/27/EU on energy efficiency.

Once technical feasibility for sizes up to 50 kWe has been verified, the final specific cost (€/kW) of the whole system has been evaluated. The best size should be a system suited to supply 50 kWe. The specific cost decrease (€/kW) of the whole system liked to size increase is around 78%.

Furthermore the flexibility of the CHP system scaled-up have been evaluated, confirming the feasibility of partial load operations (as the system developed within the project) ensuring load decrease until 40%.

Finally, which size is more marketable across Europe have been assessed: it is preferable small size systems (10-20 kWe) for smallest buildings with a few dwellings, ensuring bigger sizes (30-50 kWe) with a modular approach.

4.1.3.7. LCA and safety analysis

The polymer electrolyte membrane fuel cell (PEMFC) micro combined heat and power (m-CHP) system investigated in the ReforCELL project was assessed by means of a LCA. The general objective is to perform a LCA which evaluates the environmental burdens of conventional PEMFC m-CHP systems (both steam methane reforming (SMR) and autothermal reforming (ATR)) over their whole life cycle (“cradle to grave”) and to compare them with the ReforCELL developed technology system. To have also an idea of the position of the PEMFC m-CHP systems in comparison with other technologies, their impact is also compared with the impact of a natural gas conventional CHP, an alternative where electricity and heat come from the average mix (European electricity mix and country average heat mix) and a last one where electricity and heat come from the green available technology (called “GAT”, wind power and solar thermal). Finally, there are 6 systems compared in the detailed LCA:

➢ Fuel cell micro-CHP conventional systems with steam methane reforming (SMR)
➢ Fuel cell micro-CHP conventional system with autothermal reforming (ATR)
➢ Fuel cell micro-CHP ReforCELL developed technology: autothermal reforming with membrane reformer (ATR MR)
➢ Natural gas fuelled CHP system (50 kWe)
➢ System without CHP conventional (using electricity from the grid and an average heat mix)
➢ System without CHP good available technology (GAT, using wind power and solar thermal)

The main contributors to the impacts of the ReforCELL system are the extraction and distribution of the natural gas (especially for resources), the direct emissions from reforming (regarding climate change), the auxiliary boiler and the purchased electricity. The m-CHP production, maintenance and end-of-life are very low contributors to the impacts, except for human health. The ReforCELL developed technology has lower impacts than the conventional production of electricity and heat, but higher impacts than a theoretical optimistic case where wind power and solar thermal are used. Among CHP systems, the ReforCELL developed technology has slightly lower impacts than the other for climate change and resources consumption while they all have similar impacts for the other indicators (slight differences are visible but due to the uncertainty on these indicators, they cannot be considered as different). These results are obtained with a dimensioning for 14 dwellings (for the 5 kWe CHPs). For the PEMFC m-CHP scenarios, important amounts of heat have still to be produced conventionally with an auxiliary boiler. A different dimensioning would give different conclusions as shown in a sensitivity analysis: if the CHP system is dimensioned to fulfil a higher fraction of the needs in heat and the surplus electricity is injected to the grid, the results become better. Indeed, inject electricity to the grid enable to avoid conventional electricity production and represent a benefit. Other sensitivity analyses have been performed on the electricity mix choice, the electric efficiency of the ReforCELL system, the energy use for the manufacturing, the treatment of the ReforCELL m-CHP at the end-of-life or the type of heat considered for the conventional supply but they do not show high influence on the conclusions, or at least not for all indicators. Normalized and weighted results show the importance of climate change and resources indicators for systems such as those studied.

Regarding the safety analysis, HyGear identified and evaluated specific safety reactor/membrane parameters for CMRs using tools such as HAZOP and heat and mass transfer transport reaction models, whereas ICI identified and evaluated the safety parameters on the complete system.

Potential Impact:
4.1.4. Potential impact

The European Commission has set the target of greenhouse gas emissions reduction by 80-85% compared to 1990 levels for the year 2050. The energy landscape in the EU is the key focus in realising this goal. Carbon efficient technologies like fuel cells have a large potential in reducing emissions in Europe, and therefore combined heat and power systems should be one of the focus areas. On the other hand, one of the key objectives of the Lisbon Agenda is to increase the productivity and competiveness of European enterprises by investing mainly in key industrial sectors, covering the primary sector and raw material industries up to the knowledge industry focusing on technological research, design and development. This development is encouraged by the challenge to generate more and more power generation from renewables and investments to modernise the electricity grid. Future energy systems need to be equipped with new ways of complementary supply, such as power generation from natural gas. Moreover, long term solutions such as permanent backup supply and efficiency in order to save primary energy and a reduction of fuel imports are needed to increase the energy security. The ReforCELL projects fits just this description and focuses in particular on groups where academia and industry partners jointly strive for cooperation and exploitation in the field of hydrogen production and CHP systems. Next to that the ReforCELL project is also valued because of its contributions to the development of a hydrogen economy transition.

The implementation of new and more efficient micro-CHP systems is called by the need of complying with demands for more efficient and cost saving solutions from both the general public as well as policy makers. Of all of the objectives set out by the European Union, energy efficiency seems to be the one most difficult to realise. Central attention should be given to fuel cell powered distributed generation since it shows comparatively higher energy efficiency than conventional sources. Both in residential and industrial sector, there is a great potential for combined heat and power utilisation.

The benefits of the use of CHP systems for the customer lay in increased reliability and be sought in terms of energy and economic savings. The fuel processor developed in ReforCELL is based on low temperature reforming with the use of perm-selective membranes for separation of hydrogen. High selectivity of the membranes and high conversion rates in the system lead to a more efficient use of raw materials and minimization of by-products. Distributed generation from fuel cell systems show a huge energy saving and at the same time avoid transmission losses. Next to that, the technology reduces local emissions of GHG. Using natural gas as the main source, the existing infrastructure can be used. Depending on the fuel used (e.g. biogas), GHG emissions can be eliminated altogether by this technology.

Not only the end user experiences benefits when using the micro-CHP systems, the industry will benefit in terms of selling products with a higher value and at larger volumes compared to conventional boilers. In integration with the electricity network (see Figure 24) the technology shows a very high potential for grid balancing in terms of a power mix and a larger use of renewables and electric heating solutions, due to its flexible modulation options and its high efficiencies.

The production of CHP systems for residential purposes is stimulated by the European Union as well as the National Governments. The initial aim in the European Union is to reduce residential CHP system costs below 5000 €/kWel. It is expected that market entry and commercialization of this new technology will be successful when the systems have a maximum pay-back time of about five years or when subsidy projects are implemented to support the market accessibility. It is expected that micro-CHP costs will drop significantly once companies’ production volumes increase to small-series and eventually fully industrialized proportion.

Not only in Europe, but also in other parts of the world such as Japan, South Korea and the USA, stationary fuel cells have already commercialized and the industry is gaining traction. An obstacle for the European market is that the industry needs to reduce production costs and offer competitive pricing in order to successfully commercialize this new technology.

The technical approach implemented in the ReforCELL project aims to reduce the commercial, technical and environmental risks in a way that ultimately leads to a better product and a replacement for current technologies. The consortium members of ReforCELL investigated a harmonized approach to reformer development and the accompanying requirements and parameters for small scale CHP.

The industry is anticipating three major phases of the technology learning process that correspond with cost decrease as formulated as an objective above. The first is standardization, estimated to be up to 500 units cumulative production per company. This already could create a significant cost reduction. It is expected by the players in the industry that costs will drop 40% coming from stack production and added system components. Even today, some European system developers who have already made the step towards commercialization achieved a cost reduction of 25 %.

The second step is industrialization, which would account for up to 10,000 units cumulative production per company. Costs are expected to decrease even further with 60 % due to stack and added system costs. At this point in time installation costs are expected to stay the same. A further cost reduction is caused by semi-automation of the production and assembly process.

The final phase is considered to be mass-market production and can be identified by a cumulative production per company of 10.000 and more. In this phase, an even further volume range benefit for stack producers is expected. A shift from batch production to completely automatic manufacturing lines is foreseen.

During these different phases also maintenance costs and the costs for stack replacements are expected to reduce significantly (respectively by 60 % and 50 %).

Savings in primary energy use and therefore overall CO2 emissions are expected from CHP systems in which natural gas is converted into hydrogen and then into electricity and heat. Therefore the European Union and national governments have supported the implementation of CHP in the last couple of years even to the extend in which some member states support so heavily that the electricity supply system operates with up to 65 % overall efficiency compared to the average 33 % overall efficiency on average in the EU.

In 2007 there were only one thousand PEM based CHP systems installed in residential houses, with back then, a perspective of sixty thousand in 2010. The total market size could be as large as seven million systems per year (based on the number of boilers sold every year). The future outlook and turning point will be dependent on the PEM costs and legislation.

CHP systems can be implemented in many industries and are suitable for many applications, residential, tertiary as well as industrial. The combined heat and power systems can be beneficial for all economic sectors and therefore have a good opportunity to increase its market potential. Next to this, a tight collaboration with boiler manufacturers is necessary since a combination of conventional boilers and micro-CHP systems would result in the best economic option. Moreover, boiler manufacturers already have established contacts and customers in the market that will enable an easier entry for CHP systems.

Commercialization is expected to take five years from now onwards. Early adopters of this technology will be universities and academia and governmental buildings, but it is also expected that, depending on the pay-back time, commercial applicants will adopt quickly due to the benefits in terms of energy bill. Studies have shown than micro-cogeneration systems have a pay-back time of eight to nine years for 1 kWel CHP systems and larger plants have a pay-back time of three to four years for 1 MWel CHP systems. The CHP systems developed in ReforCELL are expected to have a pay-back time of five years.

The main socio-economic impacts (benefits) of these CHP systems are lower energy costs and lower emissions. Wide diffusion of micro-CHP systems can help set up a smart-grid revolution. Smart grids are considered a more efficient way of producing electricity. Another impact of this technology development is that a higher technology level is needed to acquire the micro-CHP systems and therefore it will lead to an investment in research and knowledge. Commercialization will have economic impacts in terms of employment. At this moment over 100.000 people in Europe are working on cogeneration. New developments and research in the CHP sectors and hydrogen production sectors will lead to an increase in the global share of these sectors and the possibility to export the technology outside Europe.

In order to create a scenario in which distributed generation has emerged as the preferred energy generation and a high share of the energy is generated from renewables, a policy commitment to distributed generation needs to be high and especially policy schemes to push fuel cell powered distributed generation. The main prerequisites should be a high commitment to distributed energy generation, a high share of renewables available, an advanced, integrated and pan-European smart grid, a high spark spread and a high price of CO2.

In order to define the market potential for stationary fuel cells, three markets can be identified, primary markets for residential fuel cell CHP solutions where a switch to the micro-CHP technology is easily made because of already existing connections to the gas grid, conversion markets which are also interesting but where switching costs might be higher due to existing non-gas heating solutions and technologies, and at last the tertiary markets in which commercialisation of fuel cells is difficult due to niche segments with very specific power and heat requirements such as biogas and biomass solutions.

When considering commercial market segment, both residential and non-residential categories need to be taken into account. The demand potential of stationary fuel cells (heat driven fuel cell CHP solutions) in commercial buildings is limited by conventional heating solutions such as solar PV, which may show direct savings to the end users.

With regards to residential buildings, heating structures in apartment buildings in many EU countries is expected to change. In Germany, the UK and Italy especially, gas usage is expected to increase. In fact, the UK is considered the most attractive market for stationary fuel cell systems in apartment buildings.

1/2-family dwellings make up by far the biggest share in the European building stock in terms of units, accounting for 73% of the total building stock in Germany, 65% in the UK, and 67% in Italy and Poland. Gas in the most prevalent solution in the UK, where approximately 80% of buildings are heated with gas –fuelled technologies. A similar dependence on gas can be found in Italy, where approximately 60% of 1/2-family dwellings use gas as a primary heating solution. In Germany, gas remains the most frequently used heating sources, but with a share below 50%. In Poland, due to the proliferation of district heating, gas only accounts for 7% of 1/2-family dwellings’ heating choice. The addressable market for fuel cell technologies is determined by three main factors: the development of the building stock, driven by construction of new buildings; heating technology installations in new buildings (including the further expansion of the gas distribution grid); switching of heating technologies in the building stock.

The largest primary market for stationary fuel cells in the ½-family dwellings segments lies in the UK., where approximately 792,000 gas boiler replacements are due in 2012 (see Figure 25). Assuming an average size of the fuel cell system of 1 kWel, the total addressable primary market is approximately 792 MWel. In 2030, the market is expected to increase to 904,000 replacements and 904 MWel. The size of the primary marked in Germany and Italy is very similar, with more than 400,000 units annually. Poland is the smallest potential primary market with approximately 40,000 units annually, increasing to 70,000 units.

On the other side, the apartment building sector is the largest in the commercial market segment, accounting for 55% of total building stock across all focus markets. The largest primary markets for stationary fuel cell technologies in apartment building remain the UK, Italy and Germany. Poland’s gas share in apartment buildings is significantly superior to the gas share in 1/2-family dwellings. In the four focus markets, there is an estimated annual primary market potential of 1.69 GWel installed capacity and conversion market potential of almost 0.59 GWel. Until 2010, the primary market potential could reach 1.77 GWel, whilst the conversion market may increase to 0.62 GWel (Figure 26). The figure below shows the potential market (primary and conversion markets) in numbers of units and as a total installed capacity per country. Compared to the UK and Italy, Germany shows a relatively small market with regards to annual exchanges of heating technologies. However, Germany is expected to reach UK levels in 2030. With capacities of more than 810 MWel and 670 MWel respectively, the UK and Germany are seen as the most promising potential markets, followed by Italy and Poland.

When considering the non-residential structures, consisting of agriculture, commercial, storage and industrial buildings, there is a lot of diversity between the different segments. The heating structure is highly diverse per industry and per country, dependent on the specific requirements. However, two overall conclusions can be made: market specific heating structures influence heating technologies to a very large extent and differentiation between rural and urban buildings should not be underestimated.

Depending on the evolution of non-residential buildings, which is compared to the residential sector, more significant on average, it is expected that the commercial sector will show the largest market potential in terms of annual capacity. However, due to the complexity of customer settings and purchasing decision-making processes, the European stationary fuel cell industry has set the main target on systems that are primarily designed for the residential (customer) sector.

Associated community societal objectives

Employment: The energy efficiency sector in Europe as a whole employs over half a million European citizens with over 100,000 of those employed in cogeneration. Under the risk that the EU CHP industry could become a net importer of CHP systems from Japan or USA, the European Hydrogen Fuel Cell Technology Platform defined future strategic R&D and deployment strategies for this sector (now implemented by the FCH JU) aimed to help this sector in maintaining a large global share. REFORCELL will contribute herewith by providing decisive step-changes in the CHP sectors and hydrogen production sectors that will not only safeguard its global market position and employment but also enable it to enter new markets and create new workplaces.

Quality of Life & Health: Smaller and more compact MRs (and m-CHP), with lower hazardous inventories, will lead to safer and more comfortable working conditions.

Environmental issues: (1) More efficient use of raw material resources and minimization of by-products formation (less wastes) due to the high selectivity and conversion rates. (2) Fostering of green transportation and energy supply technologies due to lower H2 and m-CHP prices.


List of Websites:
4.1.5. Project public website and contact

The address of the public Website of the Project as well as relevant contact details.

Project public website with further information of the about the project and consortium and main contacts details are detailed hereafter:

www.reforcell.eu

Project manager: Dr. José Luis Viviente
e-mail: joseluis.viviente@tecnalia.com

Technical manager: Associate Prof. Fausto Gallucci
e-mail: F.Gallucci@tue.nl

Dissemination manager: Dr. Giampaolo Manzolini
e-mail: giampaolo.manzolini@polimi.it

Exploitation manager: Dr. Leonardo Roses
e-mail: leonardo.roses@hygear.nl


List of all beneficiaries with the corresponding contact name and associated coordinates

Nº Participant short name Contact name E-mail
1 TECNALIA José Luis Viviente joseluis.viviente@tecnalia.com

2 TU/e Fausto Gallucci f.gallucci@tue.nl

3 CEA Sylvie Escribano sylvie.escribano@cea.fr

4 POLIMI Giampaolo Manzolini giampaolo.manzolini@polimi.it

5 SINTEF Rune Bredesen rune.bredesen@sintef.no

6 ICI Carlo Tregambe carlo.tregambe@icicaldaie.com

7 HYGEAR Leonardo Roses leonardo.roses@hygear.nl

8 SOPRANO * - -

9 HYBRID Erik Abbenhuis H.C.L.Abbenhuis@tue.nl

10 QUANTIS Simone Pedrazzini simone.pedrazzini@quantis-intl.com

11 JRC Georgios Tsotridis Georgios.TSOTRIDIS@ec.europa.eu

* SOPRANO is under liquidation

Project information

Grant agreement ID: 278997

Status

Closed project

  • Start date

    1 February 2012

  • End date

    31 December 2015

Funded under:

FP7-JTI

  • Overall budget:

    € 5 546 194,57

  • EU contribution

    € 2 857 211

Coordinated by:

FUNDACION TECNALIA RESEARCH & INNOVATION