Investigation of the technical potential of a hydrogen powered phosphoric acid fuel cell (PAFC) for CHP

Due to the fluctuating feed-in of renewable energies, controllable power plants such as highly efficient CHP plants (combined heat and power) will continue to be required to cover the residual load. Gas engines and turbines currently dominate the natural gas-based CHP market due to their low investment costs and acceptable electrical efficiency. In the event of a future fuel switch due to the energy transition from natural gas to hydrogen, fuel cell systems are becoming increasingly important due to their very high efficiency and improved dynamics in hydrogen operation and can therefore represent an alternative to gas engines and turbines. In addition to a possible fuel switch, good dynamic and full heat utilization represents an additional challenge for CHP systems. Therefore, this study aims to investigate the influence of a fuel switch from natural gas to hydrogen on the operation of a fuel cell (PAFC) in terms of efficiency, heat utilization and dynamics. It was shown that the electrical efficiency of the hydrogen-operated PAFC is significantly higher than in natural gas operation due to the omitted reformer and the associated reformer losses. In hydrogen operation, there is also no limitation of the dynamics by the reformer. Furthermore, in hydrogen operation there is a more favorable ratio of high-temperature to low-temperature heat, which facilitates the use of heat. Detailed and validated simulation models in Aspen Plus are used as the basis for this investigation.


Introduction
More than 50 countries agreed to reduce CO2 emissions to zero by 2050 or even 2045 to reach the 1.5°C target of the Paris conference [1,2,3].In order to achieve this goal, renewable energies must be greatly expanded.Since renewable energies are fed into the power grid with fluctuations, a residual load arises that is calculated from the electricity demand minus the generation of renewable energies.In order to provide this residual load and to stabilize the power grid, controllable power plants will also be necessary in the future [4,5].An efficient way of generating electricity is combined heat and power generation, in which electricity and heat are generated and used together.The future energy system places special demands on CHP systems.In addition to a possible fuel conversion from natural gas to renewable hydrogen, caused by the energy transition [6], operation that is as dynamic as possible and the most complete possible use of heat also pose a challenge for future CHP systems.Fuel cells are one of the most promising technologies for use or reconversion of H2 [7].The PAFC is one of the few commercially available fuel cell systems for stationary applications as a CHP plant and, as a mediumtemperature fuel cell, forms an interface between high-temperature and low-temperature fuel cells [8].Depending on the design, good heat extraction can be achieved with acceptable dynamics at the same time.In North America and Asia there are some systems that have proven their robustness and suitability for stationary applications [1,9,10].For natural gas, the gas engine and the gas turbine dominate the CHP market due to their low investment costs and acceptable levels of efficiency.The additional conversion of natural gas into hydrogen causes additional losses in natural gas-powered fuel cells such as the PAFC and the efficiency advantage over other technologies is reduced or lost entirely.Due to the fuel conversion from natural gas to hydrogen, fuel cell systems such as PAFC are becoming increasingly important due to the high efficiency in hydrogen operation and can therefore represent an alternative to gas engines or turbines for stationary power generation.So far, there is not much information on the operation of hydrogen-based PAFC systems and the comparison of hydrogen-based and natural gasbased PAFC cogeneration systems.Some studies only examine the admixture of hydrogen in natural gas [1].For this reason, this article examines the impact of a fuel switch from natural gas to pure hydrogen on the efficiency, performance and dynamics of a PAFC for a stationary application as a cogeneration plant.For this purpose, a detailed process engineering model of a PAFC is modelled and simulated in the Aspen Plus software.Measurement data from a 100 kW PAFC are available to validate the model.The simulation results will specifically examine and compare the efficiency, thermal performance and dynamics of the PAFC for both fuels.

Fundamentals
Some of the most important fundamentals for understanding the PAFC are explained below.The reaction equation of the fuel cell is shown in equation (1).
Hydrogen reacts with oxygen to form steam.In this process, chemical energy is converted into electrical energy and thermal energy in the fuel cell.The reaction equation can be divided into two parts.At the anode (platinum catalyst), the hydrogen is separated into hydrogen ions and electrons.
The ions can pass through the separator, which consists of concentrated phosphoric acid in a silicon carbide matrix, and thus reach the cathode.The electrons are discharged via the current collectors and then reach the cathode.At the cathode, the hydrogen ion reacts with the electrons and the oxygen from the fuel cell air to form water [11,12,13].
The operating temperature of the PAFC is about 160-220 °C and must not fall below 42 °C due to the crystallisation limit of the electrolyte [11,12].The theoretically achievable maximum electrical efficiency or cell voltage can be calculated with the Nernst equation.
In this equation, E0 represents the standard-state reversible voltage, R=8.314 J/molK is the ideal gas constant, T is the temperature, F=96485.332C/mol is the Faraday constant, and p(H2), p(O2), and p(H2O) refer to the partial pressures of H2, O2, and H2O, respectively [14].However, various voltage losses such as activation losses, ohmic losses and concentration and diffusion losses lead to a significantly reduced cell voltage during operation.The electrochemical reaction in the fuel cell is dependent on hydrogen.If natural gas is used as fuel, an upstream gas process technology is needed to convert the natural gas into a hydrogen-rich synthesis gas.First, the natural gas is purified of sulfur components in a desulphuriser (zinc oxide bed) at approx.300 °C [12,14,15].
Then the natural gas is mixed with steam and reformed in a reformer (nickel catalyst) at approx.700 °C [13,15].
The endothermic reaction has a reaction energy of 205.6 kJ/mol [12].A large part of the carbon monoxide is then converted into carbon dioxide and hydrogen in a two-stage shift reactor (copper-zinc catalyst) at 400 °C and 200 °C [13,15].
The reaction energy of the exothermic reaction is -41.2 kJ/mol [13].Due to the high operating temperature of the PAFC, the fuel cell can be operated with approx. 1 vol% CO in the synthesis gas [1,12,14].

Methods and Modelling
To investigate the impact of using pure hydrogen instead of natural gas in a PAFC, a simulation-model was built in the software Aspen Plus.Fig. 1 shows the interaction of fuel cell, gas process technology and heat extraction as well as water management for the natural gas-powered reference PAFC.The natural gas (I) is first preheated and purified of sulfur components in the desulphuriser (II).After the natural gas has been mixed with steam (III), the reforming reaction takes place in the reformer (IV).In the two-stage shift reactor (V), a large part of the carbon monoxide is converted into carbon dioxide and hydrogen.After part of the steam has been condensed out in the condenser (VI), hydrogen-rich synthesis gas enters the fuel cell stack.
Figure 1.Schematic model of a natural gas powered PAFC in Aspen Plus [1].
The calculations of the thermodynamics are based on the ideal gas method and the calculations of the reformer and the shift-reactor are based on the Gibbs equilibrium, while the calculations of the fuel cell stack are based on a stoichiometric reaction [1].In contrast to the hydrogen-powered PAFC, there are already some plants for the natural gas powered PAFC in Germany.At the ZBT in Duisburg, a 100 kW natural gas-powered PAFC is available for investigations and measurements.With the help of these measurement data additional literature data, the simulation model can be validated.A validated model contributes to improved simulation results.This simulation model then serves as a starting point for the hydrogen-powered PAFC, which has been little discussed so far.Table 1 lists some of the most important model parameters.The process diagram of the PAFC described above is thus significantly simplified and shown in Fig. 2.
Here, the hydrogen is just preheated to about 100 °C and fed directly into the cell stack stream without any further conversion steps.Since pure hydrogen is used as fuel gas, the fuel gas is almost completely used in the cell.After the hydrogen has reacted with the supplied fuel cell air, the hot exhaust gas is first used to preheat the fuel cell air and is then partially condensed in an LT heat exchanger.Unlike in natural gas operation, in hydrogen operation there is only one exhaust gas flow.The reformer-burner exhaust gas flow present in natural gas operation is completely omitted in the hydrogen-powered PAFC.The stack is cooled and the HT heat is extracted in hydrogen mode in the same way as in natural gas mode.
The only difference here is that no process steam has to be removed or provided for the reformer.

Simulation and Results
In the following, the efficiency as well as the power output and dynamics of the natural gas powered PAFC are examined and compared with the hydrogen powered PAFC. Figure 3 shows the electrical efficiency of the PAFC for both fuels hydrogen and natural gas.The net electrical efficiency is calculated from the electrical cell efficiency and the fuel utilisation efficiency minus the electrical losses due to compressor power, transformation losses and other energy demand losses.The good partial load behaviour of the fuel cell can be seen for both fuels.The electrical cell efficiency is increased in the partial load range due to lower voltage losses, which also initially increases the electrical system efficiency in the partial load range.Below about 50 % load, the comparatively constant losses and the fuel cell's own consumption dominate the efficiency and the electrical efficiency of the system decreases.Furthermore, it can be seen from the diagram that the electrical efficiency of the hydrogen powered PAFC (49 %) is significantly higher than the electrical efficiency of the natural gas powered PAFC (39%).In natural gas operation, part of the synthesis gas produced is used to provide the reformer heat.The conversion of natural gas to hydrogen in the reformer is associated with losses, so that in the end less fuel can be converted in the fuel cell.In a hydrogen powered PAFC, the reformer as well as the shift reactor are completely omitted, which means that the fuel can be completely converted in the fuel cell, resulting in a higher electrical efficiency.Figure 4 shows the thermal outputs of the PAFC for both fuels.It can be seen that the natural gaspowered PAFC can provide more heat overall.It should be noted, however, that this is due in particular to the increased low temperature (LT) heat in natural gas operation.In natural gas mode, the hot exhaust gas from the reformer burner also exists in addition to the hot exhaust gas from the fuel cell.In hydrogen mode, no reformer is necessary, so only the exhaust gas stream from the fuel cell exists.Since the LT heat comes from the cooling and condensation of the exhaust gas streams, the LT heat is reduced in hydrogen operation.The high temperature (HT) heat is obtained from the stack heat and it is almost the same for both fuels.The influence of the gas process technology not only has a great impact on the electrical and thermal performance and efficiency of the PAFC, but also on the power control and dynamics of the system.By adding additional component parameter such as mass, volumina, heat conductivity and heat capacity as well as PID controller, the stationary PAFC model was expanded to a dynamic model.With the help of the extended dynamic model of the PAFC, it was possible to carry out investigations about the load change behaviour of the PAFC.The load change speed was varied between 4 kW/min (66.6 W/s) and 16 kW/min (266.6 W/s). Figure 5 shows the course of the reformer temperature during the load change.
It can be seen that the deviation of the reformer temperature from the setpoint increases with increasing load change speed.Since the reformer burner is operated with the anode off-gas of the fuel cell, there is a strong coupling of fuel cell power and reformer power.If the power of the fuel cell is increased with a constant fuel gas supply, less anode-off gas is available for the burner and the reformer-burner power decreases.By a controlled increase of the fuel gas supply, the system can be adjusted so that sufficient anode-off gas is available for the burner for the new operating point.The fuel gas control is not arbitrarily fast, so there is a small time delay between fuel cell load change and fuel gas adjustment.To ensure a continuous and controlled reaction in the reformer, the reformer temperature should be kept as constant as possible.For this reason, a load change rate that is as low as possible (e.g. 2 kW/min) is selected in practice.In hydrogen mode, there is no reformer, which means that the reformer does not limit the dynamics.In this case, the dynamics are determined by other components (e.g. the stack).However, a significant improvement of the dynamics with regard to the load change and the cold start can be expected.

Conclusion
To investigate the influence of a fuel switch from natural gas to hydrogen on the operation of a PAFC, a detailed model of the PAFC was modelled in the software Aspen Plus and validated with measurement and literature data.The simulation showed that the electrical efficiency of the hydrogen-powered PAFC (49%) is significantly higher than the electrical efficiency of the natural gas-powered PAFC (39%).Furthermore, it was shown that the hydrogen-powered PAFC provides less LT heat compared to the natural gas-powered PAFC.This results in a higher HT/LT heat ratio.Since the LT heat often cannot be used completely because of the low temperature level of approx.55 °C, the hydrogen-powered PAFC results in a simplified and thus potentially more complete heat utilisation.The dynamic simulation showed that the dynamics of the PAFC is significantly limited by the gas process technology and especially by the reformer.By eliminating these components, a significant increase in the dynamics of hydrogen-powered PAFCs can be expected.All in all, the use of hydrogen in a PAFC results in improved performance and dynamics.

Figure 2 .
Figure 2. Schematic model of a hydrogen powered PAFC in Aspen Plus.

Figure 3 .
Figure 3. Electrical efficiency of a PAFC for natural gas und hydrogen (Aspen Plus).

Figure 4 .
Figure 4. Thermal power output of a PAFC for natural gas and hydrogen (Aspen Plus).

Figure 5 .
Figure 5. Reformer temperature of a PAFC-system during a load change (Aspen Plus).