CORDIS
EU research results

CORDIS

English EN
MCFC catalyst and stack component degradation and lifetime: Fuel Gas CONTaminant effects and EXtraction strategies

MCFC catalyst and stack component degradation and lifetime: Fuel Gas CONTaminant effects and EXtraction strategies

Final Report Summary - MCFC-CONTEX (MCFC catalyst and stack component degradation and lifetime: Fuel Gas CONTaminant effects and EXtraction strategies)

Executive Summary:
Reducing the carbon footprint of our society can be achieved by capturing and confining anthropogenic CO2 emissions (an immediate measure) as well as by replacing fossil-based fuels with renewable or waste-derived fuels (a more sustainable solution). Molten Carbonate Fuel Cells (MCFCs) are unique in being able to do both these things.
Thanks to their operating principle, CO2 can be extracted from a gas stream on the cathode side and renewable fuels like biogas can be converted to electricity on the anode side. To be able to profit from the characteristic advantages of MCFC power supply, combined with waste utilization and sustainability, it is essential to understand the mechanisms of degradation due to poisoning by typical gas impurities and to establish precisely – and enhance – the tolerance to residual contaminants of the MCFC, as well as improve the (cost-) effectiveness of the fuel treatment stage. This is the scope and main objective of MCFC-CONTEX.
On the fuel cell side, efforts were directed towards studying the effect of H2S on the anode side and the effect of SO2 at the cathode side of the MCFC. It emerged that in both cases sulphur combines with the electrolyte and with the anode material. The latter is probably an issue only after extended periods of poisoning time. In any case it has been demonstrated that the SO2 poisoning mechanism starts as soon as the sulphur accumulates in the electrolyte, and progressively gets worse as the sulphur is released at the anode side, where higher current densities speed up the process and favour the formation of nickel sulphides, due to the corresponding lower anodic potential. Nevertheless, the kinetics of nickel sulphide formation are extremely slow, which means that only after extended periods of time will this mechanism be of significance, whereas the electrolyte is immediately affected, resulting in almost instantaneous increase in ohmic resistance, though the magnitude of this effect is strongly dependent on the concentration of SO2 fed. In fact, at low concentrations of SO2, up to 2 ppm in a typical flue gas, even with considerable differences in flow rates between cathode and anode (which represent the concentration factor of CO2 transferred, but also of SO2 transferred) the MCFC can be regenerated to original performances. A numerical model has been developed that simulates the above effects and has been validated experimentally.
Accelerated MCFC testing protocols have been suggested: it has been proved that the open chamber configuration of the button cell simulates, in an accelerated way, the electrolyte evaporation from a regular MCFC cell. This mechanism can, in turn, be used to accelerate superimposed degradation phenomena, though the exact correlation in this case is still under investigation.
Another aim of the MCFC-CONTEX project was to develop a new continuous online monitoring device to detect biogas contaminants like sulfur components, halogenated hydrocarbons, and siloxanes, in order to improve control and reliability of an integrated MCFC system. The device developed for this challenge is based on a combination spectroscopy of a plasma induced by laser breakdown (for sulphur compounds) and high voltage discharge (for siloxanes). To reduce the lower limits of detection to below 10 ppm, a purge & trap enrichment step was developed in combination with the plasma setup, and the integrated systems were tested as a part of a field measurement campaign at a Munich wastewater treatment plant (WWTP).
Finally, extensive research has been carried out by directing in-depth characterization of commercial clean-up material (mostly activated carbon with different kinds of catalytic impregnation) towards the selection of the best candidates for clean-up of biogas from waste water treatment, and subsequent testing of these materials in a pilot plant operating in realistic conditions. Finally, based upon the outcome of this field campaign, a technical-economical feasibility assessment has been carried out of upscaling the proposed clean-up system to real world applications. It is concluded that the proposed plant is technically feasible, supplying almost the total WWTP demand of heat and the electrical demand for the biogas cleaning facility. Regarding the economic feasibility, with the considered hypothesis, the plant is feasible although with long (6-11 year) payback periods.

Project Context and Objectives:
Reducing the carbon footprint of our society is imperative, especially given environmental stress and climate change. This can be achieved by capturing and confining anthropogenic CO2 emissions (an immediate measure) as well as by replacing fossil-based fuels with renewable or waste-derived fuels (a more sustainable solution). Molten Carbonate Fuel Cells (MCFCs) are unique in being able to do both these things.

MCFCs are robust and highly flexible devices for the production of low-impact, high-efficiency power and heat. Today’s energy infrastructure is under insistent pressure to evolve and adapt to increasing demands of efficiency, rationalization and sustainability. MCFCs find their application in these challenges and can contribute to a reduction in the use of primary energy sources, reduced CO2 emissions, on-site energy production and carbon sequestration – all pressing necessities for our society, and Europe in particular.

Thanks to their operating principle, CO2 can be extracted from a gas stream on the cathode side and renewable fuels like biogas can be converted to electricity on the anode side. To be able to profit from the characteristic advantages of MCFC power supply, combined with waste utilization and sustainability, it is essential to understand the mechanisms of degradation due to poisoning by typical gas impurities and to establish precisely – and enhance – the tolerance to residual contaminants of the MCFC, as well as improve the (cost-) effectiveness of the fuel treatment stage. This is the scope and main objective of MCFC-CONTEX.



Figure 1. Schematic principle of the MCFC system components and applications dealt with in MCFC-CONTEX


MCFC-CONTEX (Molten Carbonate Fuel Cell catalyst and stack component degradation and lifetime: Fuel Gas CONTaminant effects and EXtraction strategies) aims to tackle the problem of degradation by trace contaminants from two sides:

• Investigation of the potential for active CO2 separation from power plant flue gas (generating power instead of consuming it in the process) and determination of poisoning mechanisms caused by SO2 in the tail pipe gas, also through numerical modelling and accelerated testing;
• Optimizing clean-up of biogas from waste-water treatment and natural gas to achieve tailored degrees of purification according to MCFC operating requirements.

The first line of activity requires extensive and long-term cell testing, so characterisation and accelerated tests will be carried out in parallel laboratories on MCFC components supplied through the external advisor FuelCell Energy and other manufacturers. A numerical model will be set up to simulate and predict the effects of CO2 separation and contaminant poisoning on the system’s performance. Also, one of the most difficult challenges in fuel cell research and development will be tackled: defining accelerated testing protocols that allow to assess long-term degradation phenomena in test-campaigns of reasonable lengths.
These activities are grouped in Work Package 2.

The second line of investigation (Work Package 4) entails characterization and development of clean-up materials and processes, to be narrowed down to the most cost-effective solution for utilisation in a waste-water treatment plant with a MCFC combined heat and power generator fed with the biogas coming out from the treatment process. A pilot-scale gas cleaning unit will be developed and run, and a technical-economical feasibility study will be carried out for the design of a scaled-up version of the clean-up system, fitting realistic applications.

To carry out this research, and to control and monitor the integrated system, real-time and highly accurate contaminant detection methods are necessary which have to be implemented in the biogas-clean-up-MCFC chain to monitor the fate of the harmful species and inhibit their effects. This is a crucial, supporting task and acts as scientific cement between the previous two activities, and as such is collocated as Work Package 3.

As the activities have come to a close, the outcomes of the project are:

• increased understanding of poisoning mechanisms in MCFC stacks, especially as regards their application as a retrofit CO2 separation and sequestration (CCS) technology;
• a set of tolerated operating conditions for the MCFC in the above applications;
• a numerical model for prediction of contaminant-induced degradation effects;
• preliminarily validated accelerated testing procedures;
• a prototypal clean-up system optimized for upgrading biogas from waste water treatment to MCFC requirements;
• a trace species detection system answering industrial criteria for monitoring of fuel quality and process control.

Furthermore, 7 peer-reviewed articles were published, a further 3 are submitted and 4 are under preparation.

Overall, the achievements of the project are significant and their impact and potential for further exploitation should not be underestimated:
• cross-cutting technologies such as non-invasive gas analysis and trace contaminant detection, as well as biogas clean-up system design and validation are of fundamental importance to a myriad of end-user applications and industry processes.
• the need for abatement of CO2 concentrations in the atmosphere is widely recognized and painfully necessary: before renewables can supply all our primary energy, reliance will have to be made on conventional power plants, for which solutions have to be found that mitigate greenhouse gas emissions, without compromising primary energy efficiency: using the MCFC as a retrofit CCS solution proves to be promising, though obviously certain criticalities need to be taken into consideration.

MCFC-CONTEX has provided crucial data as to the feasibility and possibilities for implementation
of the above, and this has been adequately publicized to a large audience of policymakers, industry,
research community as well as general public.

Project Results:
In this section, the main scientific-technological results and achievements of the project will be set out and explained, according to the three main lines of activity:

• MCFC characterization, testing and modelling in representative operating conditions,
• trace gas analysis and instrument development
• gas clean-up: materials characterization and system design with feasibility assessment

corresponding respectively to Work Packages 2, 3 and 4.

3.1 MCFC characterization, testing and modelling

3.1.1 Introduction

MCFCs are robust and highly flexible devices for the production of low-impact, high-efficiency power and heat. The MCFC offers high electric energy conversion efficiency (about 50 % based on the Lower Heating Value of natural gas) in a simple cycle configuration, so that it can significantly reduce the exploitation of non-renewable as well as renewable energy sources. In addition, for equal power production, a high efficiency is translated into reduced carbon dioxide emissions.

The MCFC operates at about 650ºC, thus, differently from low temperature fuel cells, no precious metal is required as the fuel catalyst. Together with production cost saving, the main consequence of this is that carbon monoxide is not a poisoning element, but, on the contrary, that it can be used as a fuel. All hydrocarbon-based fuels can be “reformed” to a mixture of hydrogen and carbon monoxide (CO). Thus, a variety of fuels can be utilized in a MCFC, such as biogas, natural gas, syngas derived from biomass or coal, landfill gas, gas obtained from industrial or agricultural by-products.

The typical structure of an MCFC is schematically illustrated in figure 2. The electrolyte is liquid and is embedded in a matrix. Ionic transfer inside the electrolyte (which close the electric circuit) is conducted via CO32- ions migrating from the cathode to the anode side.

The chemical reactions that govern the operations are:

(1)

at the cathode side, while, at the anode:

(2)
(3)

Expression (3) is commonly called a shift reaction and converts carbon monoxide and water into hydrogen, which then reacts electrochemically according to equation (2). As a consequence of equations (2)-(3), water is formed at the anode side and CO2 is needed at the cathode side.




Figure 2. Schematic representation of a MCFC

Since the CO2 required for reaction (1) is the same formed as consequence of reaction (2), anodic gas is generally recycled from the anode to the cathode. However, thanks to their operating principle, the CO2 required for reaction can also be extracted from any gas stream on the cathode side. This is particularly interesting if we take into consideration the growing concern regarding CO2 emissions from conventional, combustion-based power plants. One of the largest contributors to greenhouse gas (GHG) emissions is the power sector, which is still chiefly fuelled with fossil derivatives, in particular coal, natural gas and oil. Since it is extremely challenging to replace in a short time scale the colossal capacity for power generation based on these fuels with renewables, without affecting the security and economy of our energy supply, it is being considered to implement transitional solutions to the urgent issue of CO2 concentration in the atmosphere, proven cause of global warming. In particular, separating CO2 from the exhaust gases of these power plants and sequestration thereof (in spent oil fields, salt caverns or other potential storage spaces) can provide temporary relief to GHG effects and gain us time to implement more sustainable measures.

Conventionally, CO2 is separated by scrubbing the flue gas with solvents (ammine solutions) which is an energy-consuming process and yields non-environmentally friendly by-products. Employing an MCFC to separate the CO2 from the flue gas actually generates electricity, so that the overall efficiency of the power plant is not affected as in the case with energy intensive measures such as ammine scrubbing, by which one could imagine situations whereby the separation of CO2 actually causes more fuel to be burned and thus more CO2 to be emitted than without CCS!

MCFCs therefore can both be fuelled by renewable fuels, such as biogas from waste water treatment plants for example, as well as be employed as an active device for CO2 separation from flue gas. In both these cases, the gas that is utilized will be characterized by the presence of certain contaminants, of varying nature and degree.

To be able to profit from the characteristic advantages of MCFC power supply, combined with waste utilization and sustainability, it is essential to understand the mechanisms of degradation due to poisoning by these typical gas impurities and to establish precisely – and enhance – the tolerance to residual contaminants of the MCFC

These contaminants can have a significant impact on the effectiveness of a fuel cell. One of the most important is sulphur and its compounds. When sulphur enters an MCFC it reacts in a negative way, in the sense that it almost de-activates the nickel, which is the active material in the anode. The nickel is necessary to make the hydrogen react, but if sulphur is mixed in with the hydrogen then the sulphur reacts with the nickel preferentially over the hydrogen. So it creates a nickel sulphide species which is no longer active towards the hydrogen reaction. A gradual loss in performance of the fuel cell is the result that can lead even to a catastrophic failure if the sulphur is significant, or if exposure is longer than a certain time.

In biogas, the main contaminants are hydrogen sulphide (H2S) and silicon compounds (siloxanes). In the flue gas of a power plant, the main contaminant is sulphur dioxide (SO2), which is also the main cause for acid rain formation. Based on input from the industries and External Advisors of the project, it was concluded that for the fuel side of the MCFC it is much more cost-effective to focus on the clean-up system upstream, that upgrades the fuel to required levels of purity, and make it totally reliable, rather than to try and develop a fuel cell stack that is more resistant to these very insidious contaminants. However, so far extremely little investigation has gone into the poisoning effects of air-side contaminants. Thus, the focus of the project has been to look at the fate of SO2 within a potential CCS system application and map its harmful effects.

The results presented in the following paragraphs show how MCFC-CONTEX has gone about to investigate the mechanism of sulphur poisoning. This has been done in a harmonized approach between experimental campaigns and numerical model development. Furthermore, a cutting-edge scientific enterprise of the project has been the defining of accelerated testing protocols that could allow to assess long-term degradation phenomena in test-campaigns of reasonable lengths: considering that the MCFC is expected to reach 5 years of useful service life, test methods that allow to replicate or predict performance degradation in shorter times would greatly advance the design of integrated systems and facilitate the definition of guaranteed, safe operating conditions.

3.1.2 Contaminant-induced Degradation of MCFCs

The focus of research into the increased robustness of molten carbonate fuel cells (MCFC) – in order to facilitate their application and reduce installation, operation and maintenance costs – is to evaluate the effective tolerance of the MCFC in representative operating conditions. In this investigation, it is of paramount importance to understand the mechanisms of poisoning that take place when the MCFC is fed with biogas, natural gas or – in a radically innovative perspective – with flue gas from a combustion-based power plant (as oxidant, in order to separate the CO2 contained therein for subsequent capture and sequestration – so called CCS – while generating power). In MCFC-CONTEX this challenge is taken up and results are reported here of the extensive experimental campaigns carried out to pinpoint the effects of key contaminants and the frame of safe operating conditions for an MCFC. Above all, the efforts were directed towards studying the effect of H2S on the anode side and the effect of SO2 at the cathode side of the MCFC.

Tests were conducted at different scales: half-cells, button cells and single cells. The former are best indicated to do in-depth electrochemical studies, whereas the latter can be made to be more representative of real stack conditions. Effects have been studied of sodium-chloride and various forms of sulphur poisoning, especially H2S at the anode and SO2 at the cathode.

For example, using button cell tests, it is possible to carry out separate measurements for the anode and cathode, which allows to discriminate between fuel-side and air-side effects, respectively. Looking at figures 3a (SO2 poisoning at the cathode) and 3b (H2S poisoning at the anode), it can be evinced that the poisoning effect is predominant on the anode side, whether H2S is added to the fuel or SO2 to the oxidant. For sure, in the latter case, a slight influence on cathode kinetics is observed (see figure 3a), which is absent in the former (figure 3b), but it can be concluded that the criticality lies at the anode.




Figure 3. Polarization curves showing overvoltage (ΔE) of cell, anode and cathode with a) clean cathode gas (representing flue gas from a power plant exhaust) and after 20 h, 45 h, 75 h, and 100 h exposure to 8 ppm SO2; b) 90 min exposure to increasing quantities of H2S in the fuel (anode gas). Regeneration is performed with clean anode gas for 20 h.


With the use of single cells, which are sealed, it is possible also to carry out gas analysis on the exhaust gases, both anode and cathode. This yielded the highly interesting result that a few hours after SO2 starts to be fed at cathode side, H2S flows from the anode outlet (see figure 4, for two different types of cathode gas), suggesting that a transfer of the sulphur takes place, from cathode to anode, across the electrolyte – much in the same way as CO2 is transferred between the two electrodes. In fact, for the first hundred hours, no SO2 is detected at the cathode outlet, showing that the mechanisms of formation of sulphate ions (SO4=) at cathode side are very quick and well-favoured. The time delay before H2S is detected at the anode outlet represents the accumulation of sulphate ions in the electrolyte, which slowly migrate to the anode side, where they are converted with the hydrogen there, to H2S.


Figure 4. SO2 and H2S outlet compositions at OCV in a) CO2-rich cathode gas with 50 ppm SO2; and b) CO2-rich cathode gas (typical CCS conditions) with 6 ppm SO2
Therefore, the reaction mechanisms proposed are the following:

cathode (4)
anode (5)
anode (6)

In the case of the typical CCS condition, where the concentration of H2S achieved at the anode side was higher (due to the higher total flow ratio between cathode and anode side), the degradation effects seem to be more severe and the total lifetime of the single-cell was shorter, despite the fact that less SO2 was fed at the cathode side.
Considering equations (4-6) and the different flow rates in the two cases of figure 4 (CO2 rich: “KIST”, CO2 lean: “CCS”), it was then possible to calculate a mass balance:

(7)

(8)

where H2SCAT-AN is the theoretical concentration of H2S formed at the anode side due to the transfer of sulphate ions and K is the ratio between cathodic flow rate and anodic flow rate (KKIST = 2,4; KCCS = 29,6). Then:

(9)

(10)

From this result the more severe effect observed in the case of CCS conditions is evident.

It is presumed that the acute degradation effects which occur in the MCFC performance are due to the presence of H2S at the anode, according to the nickel deactivation reactions

(11)
(12)
where (11) is an electrochemical reaction with the sulphur ions in the electrolyte and (12) a chemical reaction between the catalyst and the gaseous contaminant.

Poisoning with SO2 in the oxidant therefore causes an ionic sulphur transfer mechanism that brings the contaminant from cathode to anode side, leading to poisoning effects that are practically equivalent to poisoning directly with H2S in the fuel,
Thus, the risks of application of the MCFC as a retrofit system for carbon capture and sequestration could be comparable with those of the MCFC fed with natural gas and biogas, where H2S is expected to be the main contaminant, fed at the MCFC anode. The difference in oxidant composition and flow rate are hereby the main parameters for assessing this equivalence.

It has been confirmed furthermore that there are important mechanisms occurring at all interfaces: gas-electrode and electrode-electrolyte, which convolute and depend on all tested variables, and mainly on gas composition, contaminant concentration, current density, utilization, exposure time, cell set-up. As indicated by the External Advisor to the project, FuelCell Energy, ultimately the conditions have to be tailored in such a way that less than 1 ppm of sulphur is present at the anode. Hereby, the MCFC operation profile has to be adapted in such a way that electrochemical conditions for poisoning are avoided and/or regeneration stages can be implemented for the recovery of non-contaminated MCFC performance.

Regeneration tests have also been carried out to assess the potential of recuperating original performance after flue gas SO2 poisoning, by feeding the MCFC with clean oxidant (same flue gas composition but without SO2), which could be a cost-effective measure of maintaining the system, rather than guarantee a perfectly clean flue gas at all times.
Increasing SO2 poisoning from 0.2 ppm to 4 ppm at the cathode in typical CCS composition, no significant change in performance was observed for three days up to 2 ppm of SO2. Moreover, after regeneration in clean gas for one day, the EIS and IV curves returned back to their original shape before poisoning, indicating that performance can be completely recuperated: only at 4 ppm poisoning original performance was not recoverable (see figure 5).


Figure 5. a) IV curves and b) EIS Nyquist plot before, during and after 4 ppm SO2 poisoning in CCS conditions

In conclusion, it is clear that sulphur combines with the electrolyte and with the anode material. The latter is probably an issue only after extended periods of poisoning time. In any case it has been demonstrated that the SO2 poisoning mechanism starts as soon as the sulphur accumulates in the electrolyte, and progressively gets worse as the sulphur is released at the anode side, where higher current densities speed up the process and favour the formation of nickel sulphides, due to the corresponding lower anodic potential. Nevertheless, the kinetics of nickel sulphide formation are extremely slow, which means that only after extended periods of time will this mechanism be of significance, whereas the electrolyte is immediately affected, resulting in almost instantaneous increase in ohmic resistance, though the magnitude of this effect is strongly dependent on the concentration of SO2 fed. In fact, at low concentrations of SO2, up to 2 ppm in a typical flue gas, even with considerable differences in flow rates between cathode and anode (which represent the concentration factor of CO2 transferred, but also of SO2 transferred) the MCFC can be regenerated to original performances.

On the whole, this entails that, in a hypothetical application of the MCFC as a retrofit solution to separate CO2 from the flue gas of conventional combustion-based power plants, an SO2 abatement stage has to be programmed for, and a rather accurate SO2 detection system could suffice to rely on a single abatement stage. This detection system would not necessarily have to be real-time, since it has been shown that the MCFC can handle up to three days of poisoning and still be regenerated by using clean air.

3.1.3 Numerical model development

In order to be able to predict degradation mechanisms and quantify their effects on MCFC performance, an important part of the MCFC-CONTEX project has been to set up a numerical model, based on the SIMFC code developed by University of Genoa.

After extensive theoretical analysis and experimental validation of the proposed model, the following formulation for the MCFC voltage at given conditions could be defined:




where:

V = cell potential
E = Nernst voltage
T = temperature
pi = partial pressure of component i
yi = molar fraction of component i
= Ohmic resistance
= cathode resistance, first term
= cathode resistance, second term
= anode resistance

and where the second term of the cathode resistance can be neglected if pCO2 / pO2 < 0.9 such as in the operating conditions chosen as reference in the MCFC-CONTEX project.



Figure 6. The ultimate definition of the main parameters in the MCFC model with experimental validation

The SIMFC (SIMulation of Fuel Cells) code is then able to calculate the maps of gas and solid temperatures, electrical current density, Nernst voltage, polarization, internal resistance, pressure drops and compositions and flow rates of the gaseous streams on the cell plane, also taking into account H2S poisoning at the anode and SO2 poisoning at the cathode.
Thanks to the experimentally identified parameters, a good agreement between experimental and simulated results has been obtained.

3.1.4 Accelerated testing procedures

In order to introduce molten carbonate fuel cells (MCFCs) in commercial applications, the target lifetime of a MCFC has been set at 40 000 hours. To achieve this goal it will be crucial to understand and learn how to counteract different degradation phenomena occurring in the cell. Although it is not always intended to run the cell during tens of thousands of hours, the MCFC requires a long startup time and experiments are very time-consuming. It is therefore very important to find ways to accelerate the degradation phenomena and be able to run relevant tests during shorter times.

Due to their flexibility (shorter startup) and their (spatially) uniform operating conditions, button cells are particularly suitable for development of a testing procedure for the acceleration of degradation phenomena. The button cell enables filling of electrolyte during cell operation and is also equipped with reference electrodes that facilitates separate measurements of the anode and the cathode.


Figure 7. Schematic graph showing the response of electrolyte filling on the fuel cell performance as a function of operation time.

A first step towards an accelerated test procedure was to establish whether the open chamber configuration of the button cell simulates, in an accelerated way, the electrolyte evaporation from a regular MCFC cell. Figure 7 shows schematically in what way the performance of the fuel cell is responding to electrolyte refilling over time of operation. It is possible to compensate for the electrolyte evaporation by adding electrolyte, taking the performance back to its initial value and by this procedure it is possible to quantify the evaporation rate (electrolyte loss). After a longer time the electrolyte additions is no longer enough to keep performance stable; other degradation mechanisms than electrolyte loss have appeared (corrosion, sintering etc). The same course of events will happen in a larger cell, but in the button cell, the electrolyte evaporation will be accelerated and known, Figure 8. When comparing a reference test run in the button cell (Fig. 7) with tests undertaken in exactly the same way, but with for example an added contaminant such as SO2 or H2S, it will be possible to evaluate the influence of that contaminant on degradation.



Figure 8. Schematic graph comparing a cell under normal conditions compared with a button cell with accelerated electrolyte losses.

In the validation tests, it was confirmed that electrolyte evaporation is accelerated and quantifiable in a button cell set-up. See for example figure 9: the shift of the EIS curve from t=432 h to t=1272 h denotes an increase of ohmic resistance only, most likely due to electrolyte evaporation. After refilling with fresh electrolyte (t=1512h), performance is recuperated and the EIS arc is shifted back, even though not completely to the original position. From t=1512h onwards, it was noticed that the shape and size of the EIS arc started to change, denoting the insurgence of a different degradation mechanism to electrolyte evaporation only.

Figure 9. EIS-curves showing the change in arc shape and size (i.e. polarization resistances) with time after t=1512h, denoting a degradation mechanism different to electrolyte evaporation

Now, with the data above, and at the time of conclusion of the Project, no certain statements can be made about the accelerating factor of electrolyte evaporation in an open-chamber configuration, less still about the acceleration of superimposed degradation effects, such as SO2 poisoning at the cathode. To do this, a benchmark needs to be made of a button cell that operates with continuous electrolyte addition that just manages to compensate the accelerated evaporation, thereby simulating stack conditions, in order to quantify exactly the acceleration of the degradation rate over time. That this is possible has already been observed by Morita et al [1], but the addition of electrolyte needs to be better distributed over time, avoiding excessive discontinuities.
The second degradation mechanism that needs to be superimposed then needs to be quantified, first in a condition of continuous electrolyte replenishment, and then in the condition of accelerated evaporation. The differences in the degradation rates observed in the latter two cases will give an estimation of the accelerating factor for – in this case – SO2 poisoning in open-chamber configuration without electrolyte replenishment.
Nevertheless, what has been achieved so far is that it has been proved that electrolyte replenishment can recuperate completely degradation that has occurred due to its evaporation, and that there may well be a reciprocal effect in the superposition of SO2 poisoning and electrolyte evaporation which needs to be assessed. Investigations into these aspects will continue beyond MCFC-CONTEX, and are of conspicuous interest due to the scientific significance of a possible proven correlation and the high potential for improved technological development resulting from a validated protocol for accelerated testing of MCFCs.

3.2 Trace gas analysis and instrument development
3.2.1 Introduction
The aim of the MCFC-CONTEX project was to develop a new continuous online monitoring device to detect biogas contaminants like sulfur components, halogenated hydrocarbons, and siloxanes. The device developed for this challenge is based on a combination of laser-induced breakdown spectroscopy (LIBS) and Raman spectroscopy.




Figure 10. Schematic presentation of the combined LIBS/Raman biogas monitoring system developed by TUM.

As LIBS is based on the laser ignition of a plasma spark in the gas sample, it was to be ensured that the whole range of possible biogas compositions is not capable of being ignited. The limited amount of O2 contained in biogas prevents it from igniting. Hence, LIBS can be applied to these gases. As the instrument is intended for application on different test sites, this whole instrument, as well as the single components need to be robust. A high-power pulsed laser (InLight, Continuum) was selected which provides pulse-energies of more than 100 mJ at 532 nm, ensuring stable laser pulse generation under all possible gas conditions. Furthermore, the laser head is designed for routine industrial application and it is small enough for implementation into a mobile instrument (see figure 10).

3.2.2 Instrument development and calibration
After the setup of the LIBS instrumentation, first experiments were focused on the determination of the limit of detection (LOD) of the relevant elements. The different components were detected in nitrogen and in argon. It was found that the LODs for the different elements in argon are significantly lower than in nitrogen. As known from literature, Ar leads to an increase of the signal intensity by increasing plasma temperature and electron density. In real-world samples, the matrix consists of about 60 vol% CH4 and 40 vol% CO2.

LIBS spectra of biogas show emission bands from recombined matrix constituents, which cover the analyte emission lines. Another problem is that biogas plasma is not optically transparent for sulfur and silicon emission, thus detection of these analytes proved not to be possible. The workaround that was developed in the framework of this project is the combination of a purge & trap enrichment step with the LIBS setup. The analytes are collected and enriched on an appropriate adsorber, which also allows for removal of the biogas matrix, when desorption is performed by argon.



Figure 11 The electric-discharge system for Si detection in biogas developed by TUM

Beyond the LIBS instrument, a new prototype instrument was developed for silicon detection. It consists of a power unit which generates an alternate current high voltage discharge. The light emitted from this plasma is focused into a spectrometer. To analyze a gaseous sample it is pumped through a small glass tube (D=7 mm, L=20 mm) with two electrodes inside. The silicon concentration is determined by the amplitude of the silicon emission line at 288.18 nm. An LOD of 1.19 mg Si/m³ in N2 was found, which is sufficient for routine application in biogas monitoring.
The main advantage of this new instrument is the smaller size and the lower price of this instrument compared to the LIBS instrument (see Fig. 11).

3.2.3 Field tests
As a part of a field measurement campaign at a Munich wastewater treatment plant (WWTP) the combination of a purge & trap enrichment step with the LIBS setup, as well as the new silicon detection system were tested. The biogas of the WWTP is untreated and contains hydrogen sulfide and siloxane contaminations. The sulfur compounds could be satisfactorily detected after enrichment on the adsorber AIRPEL 10, which is based on activated carbon. The measured sulfur amount by LIBS corresponded to the average sulfur concentration as determined by GC-MS.

Table 1. Gas composition of Munich wastewater treatment plant as analyzed by GC-MS.

Molecule Concentration
Methane ~ 60 Vol%
Carbon dioxide ~ 40 Vol%
Hydrogen sulfide ~ 200 ppm
Siloxane (D4, D5) ~ 10 mg/m³

Though the system was validated in laboratory conditions, silicon detection was not possible during the field measurement. This is attributed to the enrichment material that was utilized at the wastewater treatment plant, which probably did not desorb the species to be detected at the chosen maximum temperature and heating time, as well as due to contamination of the glass plasma tube (Figure 12).


Figure 12. Contaminated glass tube after the measurement of biogas.

The final measures of optimization are being undertaken, and the devices will be field tested at biogas producing reactors, outside the frame of MCFC-CONTEX, but with the perspective of commercial exploitation of the system, in accordance with project partner CETaqua.

3.3 Gas clean-up: materials characterization and system design
3.3.1 Introduction
Waste water treatment is one of the most important means of preserving the water cycle in industrialized countries. Spent water expelled through the municipal drains is gathered in specialized sites where the refuse flow is neutralised of its organic compounds and the water is purified in a series of processes – mostly mechanical and biological – and made drinkable again.
An important step in making this process self-sufficient in terms of energy consumption is to exploit the anaerobic digestion of the sewage sludge that results from the water purification stages. In anaerobic digestion, a mixed population of bacteria catalyses the degradation of the biopolymers (carbohydrates, protein and lipids) found in organic matter to produce biogas, primarily consisting of methane and carbon dioxide. The biogas produced in anaerobic digesters of WWTPs usually goes unused and burnt in a flare or, in some cases is only used for the digester heating. The digestion process also generates a solid residue, which can be spread on site after composting treatment, and a liquor that can be used as a fertilizer. One of the most efficient ways of exploiting and valorizing a renewable energy source such as the biogas from waste water treatment is by converting it in a Molten Carbonate Fuel Cell.
The biogas produced from the anaerobic digester is used as the fuel to generate ultra-clean electricity that can be used for the treatment plant while byproduct heat from the MCFC can be used to heat the sludge to facilitate anaerobic digestion. This Combined Heat and Power application results in up to 90% efficiency, depending on the application. Moreover, biogas is a renewable fuel eligible for incentive funding for projects in many countries throughout the world.
One of the main technical barriers to be overcome before this optimal integration can be implemented on a large scale is to adapt the biogas quality to the requirements of MCFCs in terms of contaminants which are inherent to fuel gas from organic fermentation. Among the several studied poisons, H2S is one of the most common and harmful. The accepted tolerance of the MCFC to H2S in the fuel is less than one part per million (1 ppm). Hence, in attempting to join the two above mentioned technologies (anaerobic digestion and fuel cells), a crucial and unavoidable step is gas clean-up, which has been dealt with in work package 4 of MCFC-CONTEX.
This activity has been carried out by directing in-depth characterization of commercial clean-up material (mostly activated carbon with different kinds of catalytic impregnation) towards the selection of the best candidates for clean-up of biogas from waste water treatment, and subsequent testing of these materials in a pilot plant operating in realistic conditions. Finally, based upon the outcome of this field campaign, a technical-economical feasibility assessment has been carried out of upscaling the proposed clean-up system to real world applications.

3.3.2 Characterization of biogas clean-up material
The main objective was the selection and qualification of reactive/adsorptive gas cleaning materials and processes on the basis of test results in order to transfer into an integrated biogas-MCFC plant. Lab scale tests were carried out to characterize and evaluate the clean-up material conditions, focusing on H2S and siloxanes removal, as they are the main contaminants in biogas which can lead to the fuel cell malfunctioning. The strategy of these tests was based in varying different parameters, i.e. pollutant concentration, temperature, O2 content, relative humidity and residence time and studying their effect on the adsorption capacity of the material.
For H2S removal, adsorbent materials tested were NORIT RGM3 (activated carbon impregnated with Cu(II) and Cr(VI) salts), AIRPEL ULTRA DS (wood based activated carbon doped with KOH and KI), SICAV SV40, NORIT RB1 and PSA Adsorbent H-2-10 (virgin activated carbon), SICAV SI30K and NORIT RBAA1 (activated carbon impregnated with KOH) and MERCK (acid activated carbon), Galipur S (activated alumina impregnated with KMnO4), Zeolite ATZ (natural zeolite) and Sepiolite (natural clay).

Tests relied on the flowing of a contaminated gas carrier (with known concentration of a fixed compound, usually H2S, which best represents the contaminants in biogas) through a reactor with a fixed amount of clean-up material, and the analysis of the gas coming out. The clean-up material was characterized in terms of the time to breakthrough of the contaminant (indicating saturation of the sample, from which the adsorption capacity of the same could be derived), varying various process parameters as mentioned above. From preliminary screening, the best materials could be selected for further investigation, see figure 13.



Fig. 13: Screening results for H2S adsorption (200 ppm, T = 30°C, GHSV = 10000 h-1, h/d = 0,32)

Table 2. Characteristics of NORIT RGM and AIRPEL ULTRA DS
Manufacturer Activated Carbon Type
Doped activated carbon Metal oxide impregnated activated carbon
DESOTEC Norit
Trade Name AIRPEL ULTRA DS RGM 3
Material and composition Wood based activated carbon doped with KOH and KI Steam activated activated carbon impregnated with Cu (II) and Cr (VI) salts
Characterization (shape, size, pore size distribution) Shape Pellets Pellets
Size Particle diameter
~ 4 mm. Particle diameter
3 mm.
BET surface area (m2/g)
Micropore volume (cm3/g) ~ 1000 1028
0.4 0.3601

It can be seen from figure 13 how the activated carbon RGM takes longest to saturate, when H2S starts to be detected at the reactor outlet.
Not shown in figure 13 is the performance of AIRPEL ULTRA DS, which was taken into consideration late into the project. AIRPEL ULTRA DS claimed to have excellent adsorbing capacity for H2S in operating conditions very similar to those present in the pilot plant for field testing. From datasheet information, AC Ultra DS claimed good activity in presence of high percentages of humidity, while AC RGM1 should work best in dry conditions. Thus, great attention was in particular directed to a more in-depth evaluation of Ultra DS, compared to the performance of RGM1, in particular as regards the effects of gas hourly space velocity (GHSV, calculated as volumetric flow rate over reactor bed volume, in h-1), humidity, in the range R.H. 0-100%, and presence of trace amounts of oxygen. Table 2 shows the important properties of the two activated carbons used.

The adsorption capacity is calculated as:



Ads. cap. = adsorption capacity of adsorbent material [% w/w]
M = Adsorment material [kg]
Qin = inlet biogas flow [Nm3/h]
Qout = outlet biogas flow [Nm3/h]
[H2S]in = inlet hydrogen sulphide concentration [kg/Nm3]
[H2S]out = outlet hydrogen sulphide concentration [kg/Nm3]
t = time [hours]

Effect of Gas Hourly Space Velocity
H2S adsorption capacity of the two activated carbons present a similar behavior in terms of GHSV variation: it is strongly influenced by GHSV in the range 1,000-5,000 h-1, exhibiting a marked decrease at the increasing of space velocity (Fig. 14). The better adsorption capacity at low values of GHSV is due to the longer residence time of the gas inside the material bed, which allows the gas to penetrate better into its porous structure and to undergo the catalytic effects of the impregnation. In real-world clean-up reactors, the GHSV is usually kept below 800, so that adsorption capacities are expected to be significantly higher in bigger vessels. For lab scale tests, increasing the GHSV allows to shorten the experimental runs when qualifying a given material in terms of a given property.
The explicit functions were also calculated, allowing to estimate the H2S adsorption capacity expected for values of GHSV different from the tested ones.


RGM1:

ULTRA DS:

Fig.33: H2S adsorption capacity of AC Norit RGM1 and AC Desotec Airpel Ultra DS for different values of GHSV.


In figure 14 it can be seen that RGM has higher adsorption capacity that ULTRA DS for all values of GHSV, but the dependence is the same. These tests were carried out in dry conditions, favourable to RGM.

Effect of humidity
Differently from the Norit RGM, the AIRPEL ULTRA DS AC is functionalized in order to get its best performances at low temperature and humidified (R.H.> 70%) matrix gas conditions, and the results reflects these characteristics. In dry conditions, RGM1 adsorption capacity is more than two times Ultra DS one. Going from dry to wet conditions, RGM1 slightly enhances its performance, while Ultra DS shows a significant increase of adsorption capacity (Cads), reaching a value similar to RGM1 (Fig. 15). The increase of Cads for both the activated carbons in presence of humidity is probably due to the formation of a thin water film inside the pores, where the H2S molecules are dissolved and captured by the gas stream.

Fig. 15: breakthrough curves of AC RGM1 and AC Desotec Ultra DS in dry and wet conditions.

Effect of oxygen
Trace amounts of oxygen (<2%) are usually present in biogas, due to infiltration of small amounts of air in any stage of the process. Though this may lower slightly the calorific value of the biogas as such, it has the chemical benefit of enhancing the catalytic activity of the impregnating species in functionalized activated carbons, which may increase their capacity for H2S adsorption.
Adsorption tests at different oxygen over H2S molar ratios were performed: the capacity versus O2/H2S in a biogas matrix is plotted in Figure 16.


Figure 16. Norit RGM capacity versus O2 /H2S molar ratio of the biogas matrix samples

Oxidation reactions beside the adsorption phenomena occur, leading to a significant enhancement of performance for both NORIT and Ultra DS. It is assumed that the stoichiometric oxygen required for the reaction to occur at the carbon surface is 0.5 according to reaction (3):

H2S+0.5 O2 → S+H2O (3)

Thus, the adsorption capacity of activated carbon is highly dependent on oxygen presence, see also Table 3. This result suggests that hydrogen sulphide and oxygen adsorb on the active sites of the solid surface and then catalytically react with formation of elemental sulphur and water, which are then adsorbed in macropores.
From Table 3 it can also be evinced that the simultaneous presence of CO2 and CH4 in the gas matrix severely penalizes the adsorption capacity of the tested activated carbons. There is most likely a competitive effect towards the catalytic sites in the adsorbent between the CO2, CH4 and H2S in the biogas, so that one should pay attention in the dimensioning of a real clean-up system to the effective composition of the biogas, which could greatly affect the ultimate adsorption capacity of the loaded material and therefore its service life and maintenance intervals.

Table 3. Typical adsorption capacities of activated samples in different gas compositions

Gas Matrix Relative Humidity (RH) Scap [w/w %] @ 1 ppm
N2 0 1.4
35 2.5
70 2.2
N2/CO2 0 0.28
35 2.10
70 1.04
CH4/CO2 0 0.12
35 0.52
70 -
N2/CO2/O2 35 4.32
70 4.46

3.3.3 Pilot plant operation of selected clean-up materials
Based on the above, and many other laboratory characterization tests, a field testing campaign was carried out at the waste water treatment plant of Matarò, in Spain, operated by Aqualogy (external advisor to MCFC-CONTEX) and under the guidance of project partner CETaqua.
Table 4 shows the basic parameters of this WWTP.
Table 4. Mataró WWTP general parameters
HE Sewage flow Sludge flow Biogas production
230,000 28,500 m3/day 339 m3/day 5,750 m3/day
The mixed sludge is digested anaerobically at mesophilic conditions (37ºC) for approximately 20 days in a two 3,500 m3 digesters. The produced biogas is then accumulated in a gas holder and is used either as a fuel in the digesters boiler or sent to thermal facilities near the WWTP.
Some adjustments and modifications were required in order to prepare the Mataró pilot plant for the field tests to be conducted for MCFC-CONTEX, which took place between November 2013 – April 2014.

Biogas composition and MCFC fuel inlet requirements
Table 5 shows the average value for the biogas composition, obtained from several analyses carried out between 2009 and 2010. This composition is considered to be kept constant in these years.
Table 5. Biogas analysis (average data 2009 and 2010)
Type Compound Unit Average
Main gases CH4 % 63.8
CO2 % 32.0
N2 % 2.4
O2 % 0.7
Organo-sulphurated compounds H2S ppm 2,600
Methyl-S ppm 0.2
Ethyl-S ppm 1.2
DMS ppm 0.6
DMDS ppm UDL
CS2 ppm 0.4
Linear hydrocarbons C-8 ppm 0.6
C-9 ppm 1.5
C-10 ppm 8.9
Aromatic hydrocarbons Toluene mg/m3 1.6
Limonene mg/m3 1.0
ALL BTEX mg/m3 4.4
Siloxanes TMOH mg/m3 0.2
D4 mg/m3 3.4
D5 mg/m3 6.4

Configuration
The raw biogas coming from the gas holder goes through the pilot plant following the next configuration:
a) First, the biogas is introduced into a biotrickling filter, where the main removal stage is carried out since over 90% of the H2S content is supposed to be removed in this step.
b) Then the biogas enters the polishing stage in which H2S is removed completely as well as the other pollutants. It consists of the following process units:
i. H2S adsorption filters: the biogas is first driven through two adsorption filters in series, in a lead/lag configuration. Two adsorbent materials will be tested: in the first set (November 2013 – January 2014) an impregnated activated carbon, selected from the laboratory characterization tests and in the second set (March 2014 – May 2014) an iron oxide, where the main mechanism is a catalytic oxidation in both cases.
ii. Biogas drying: a heat exchanger cools the humid biogas and then re-heats it again, allowing most of its water to be removed by condensation as well as some VOC’s and siloxanes through solubility.
iii. Siloxanes adsorption filters: at the end of the polishing step, the tempered biogas enters two adsorption filters, also in series and in lead/lag configuration, where other pollutants such as siloxanes, VOC, and the remaining H2S (if any) are physically adsorbed.
Finally, the clean biogas is fit for further utilization in a MCFC. However, the aim of the pilot plant is to deliver reliable data about technical performance and verify the selected adsorption technologies, so no fuel cell is installed and the clean biogas is returned to the WWTP main pipeline.

Flow sheet
Figure 17 depicts a simple flowsheet including the most relevant pieces of equipment of the three parts of the plant. Only one filter of each lead/lag configurations is shown, for a simpler representation of the whole process:

Figure 17 Plant flowsheet
Figure 18 shows the real lay out of the plant.
.
Figure 18 Plant global picture
The ultimate performance of the H2S removal stage loaded with Airpel Ultra DS is shown in figure 19, where the inlet ([H2S] SP2) and outlet concentrations of the clean-up reactor ([H2S] SP3) are plotted as a function of time in the final stage of the pilot test, where the reactor started to saturate.


Figure 19. Hydrogen sulphide concentration at SP3 (red) and relation Cout/Cin (green)
At a removal efficiency threshold of 95% (Cout/Cin) the loaded amount of Airpel Ultra DS had adsorbed 81% of its weight in H2S, a remarkable result, showing the effectiveness of this material and proving the outcome of labscale tests where it was shown that the trace presence of oxygen, high humidity and a low GHSV all contribute to increasing the adsorption capacity of this activated carbon.

3.3.4 Technical-economic feasibility assessment of scaling up the clean-up system
This final activity had two different objectives: on one hand, to give a technic-economic assessment of typical case studies and, on the other hand, to carry out a preliminary design of the scaled-up biogas cleaning facility.

Technical-economic assessment
Using a simulation-based calculation tool developed by CETaqua (BiogApp Tool), four case studies were considered, selecting Wastewater Treatment Plant size and biogas pollution, using a MCFC as an end-use – see Table 6.

Table 6. Parameters of the four case studies
SIZE [PE] BIOGAS [Nm3/h] H2S [ppm] VOSiC [mgSi/Nm3]
CASE 1 225,000 240 2,500 10
CASE 2 450,000 480 2,500 10
CASE 3 225,000 240 250 10
CASE 4 450,000 480 250 10
Two sizes of WWTP were considered, based on the person-equivalent of waste water treated, and considering two levels of contaminant loading.

The heat and electrical demand self-sufficiency in WWTP is evaluated. Heat needs can be covered during the three of the four trimesters of the year, although in the cold months a little amount of natural gas consumption is needed (7% of the WWTP heat demand).
Big plants require higher investment due to larger power fuel cell purchasing. However, the higher OPEX (in relative terms) corresponds to small plants. Both CAPEX and OPEX increase in the WWTP with polluted biogas (2500 ppm) as a consequence of the biotrickling filter, both for investment and maintenance. The results are schematized in tables 7-9.

Table 7. Case studies comparison: Technical results
Parameter [Units] CASE 1 CASE 2 CASE 3 CASE 4
Treatment
H2S removed load [kg H2S/day] Biotrickling filter 17.52 35.04 - -
Polishing stage 4.38 8.76 2.19 4.38
Moisture removal [kg H2O/day] 273.6 273.6 544.8 273.6
VOSiC removed load [kg Si/day] 0.04 0.04 0.08 0.04
End-use electricity
Installed electrical power [kWe] 800
(2x400) 1450
(3x400+1x250) 800
(2x400) 1450
(3x400+1x250)
Electric power production [kWe] 733 1450 733 1450
Average load [%] 92 100 92 100
Total electricity consumption [GWhe/year] 5.9 11.8 5.9 11.8
Net ECS electric energy produced [GWhe/year] 5.73 11.34 5.97 11.81
Net electric energy produced [GWhe/year] 5.37 10.61 5.65 11.17
End-use heat
Heat Production Thermal power [kWt] 608 1203 608 1203
Total heat production [GWht/year] 4.76 9.41 4.95 9.80
Heat Consumption Total heat consumption [GWht/year] 4.52 9.04 4.52 9.04
Heat covered with biogas [GWht/year] 4.04 8.07 4.20 8.41
Covered heat [%] 89.28 89.28 93 93
Heat covered by NG [GWht/year] 0.48 0.97 0.32 0.63
Heat surplus [GWht/year] 0.72 1.34 0.75 1.39
Average purchased Natural Gas [Nm3/day] 10 27.5 9.6 26.4
Flared biogas [Nm3/h] 25.7 56.8 16.8 39
Flared biogas [%] 10.71 11.83 7.00 8.12
Availability
Overall availability 89.28
Electric power production has been considered as the maximum power that the fuel cell is able to give. In Cases 1 and 3, the biogas flow loads the 92% of the fuel cell capacity, which is why the power production is lower than the installed power.
Net ECS electricity produced corresponds to the total fuel cell electricity production, whereas the net electricity produced considers the energy balance between the fuel cell electricity production minus the electricity energy consumption of the treatment plant (Biotrickling filter (BTF) + Chiller + Adsorbent filters) and the fuel cell itself (5% of its production).
With this last value, it can be seen that the fuel cells almost cover the electricity demand of the WWTP, since the net electricity produced is just a bit lower than the electricity demand. On one hand, big plants lead to higher losses. On the other hand, low polluted plants lead to lower losses since the BTF is no needed.
Regarding the end-use heat, it must be taken into account that the values are considered as a yearly average. It seems that the fuel cell could be able to cover the heat demand, but fuel cell availability must be taken into account as well as that, in the first trimester (lower temperature), natural gas is needed in order to supply the heat demand of these three months.
Therefore, natural gas consumption is considered as the needed in the first trimester as well as the thermal energy consumed during the no-operation periods of the fuel cell (availability 93%), but its consumption is very low compared to the rest of values.
Flared biogas corresponds to the biogas not used during the no-operation of the fuel cell as well as the excess (Cases 1 and 3). In polluted plants, more biogas is flared since the biotrickling filter availability is also considered.

Economical results
Table 8. Case studies comparison – CAPEX
Parameter CASE 1 CASE 2 CASE 3 CASE 4
CAPEX [k€]
BTF 184 194 - -
Polishing 212 322 212 322
Fuel Cell 3840 7410 3840 7410
Civil & Electrical works 405 803 405 803
Subsidy 0 0 0 0
TOTAL 4641 8729 4457 8535

BTF and Polishing OPEX include media purchasing and disposal; water, electricity and other chemicals consumed, but not their maintenance. It is shown aggregated with maintenance of Fuel Cell in the next row.

Table 9. Case studies comparison – OPEX

Parameter CASE 1 CASE 2 CASE 3 CASE 4
OPEX [k€/year]
Natural Gas 1.46 4.31 1.46 4.31
BTF 6.37 12.74 0.00 0.00
Polishing 10.67 21.95 6.74 14.1
Maintenance (BTF + Polishing + FC) 80.56 143.61 72.17 136.98
Manpower 4.11 8.22 4.11 8.22
TOTAL 103.18 190.82 84.48 163.61
SPECIFIC OPEX [c€/Nm3 biogas]
Natural Gas 0.07 0.1 0,07 0.1
BTF 0.3 0.3 0,00 0,00
Polishing 0.51 0.52 0.32 0.34
Maintenance (BTF + Polishing + FC) 3.83 3.42 3.43 3.26
Manpower 0,20 0.2 0,20 0.2
TOTAL 4.91 4.54 4.02 3.89
SPECIFIC OPEX [c€/kWhe]
Natural Gas 0.03 0.04 0.03 0.04
BTF 0.12 0.12 0.00 0.00
Polishing 0.20 0.21 0.12 0.13
Maintenance (BTF + Polishing + FC) 1.50 1.06 1.29 1.24
Manpower 0.08 0.08 0.07 0.07
TOTAL 1.92 1.78 1.51 1.48

Table 10. Case studies comparison – economic viability

Parameter CASE 1 CASE 2 CASE 3 CASE 4
ECONOMIC VIABILITY
NPV (15 years) [k€] 619 1884 1342 3020
IRR [%] 9 10 11 12
Payback period (years) 11.44 10.44 9.62 9.16

Small plants (Cases 1 and 3) and big plants (Cases 2 and 4) can be differentiated mainly by their CAPEX. Big plants need a higher investment as they have to treat more biogas and, consequently, more energy can be produced. Therefore, the main investment is focused on the type of fuel cell.
Regarding the specific OPEX, both in terms of c€/Nm3 and c€/kWhe, it shows that bigger plants have lower values corresponding to the biogas treatment and the maintenance compared with the small plants.
As for the economic viability, the more profitable case is Case 4, with a payback period of less than 10 years, whereas the less viable case is Case 1, with more than 11 years of payback period:


Figure 20. Payback period for the Case studies


The most economically viable case is a big WWTP with low polluted biogas with a payback period below 10 years, whereas the least one is a small WWTP with a high polluted biogas with a payback period of more than 11 years.

3.3.5 Scale-up of the proposed system

After studying the different case studies from the technical and economical point of view, Case 2 will be further studied in this section. Case 2 has been selected because it corresponds to the nominal capacity of the WWTP of Mataró:
Table 11. WWTP and biogas characteristics
SIZE [PE] BIOGAS [Nm3/h] H2S [ppm] VOSiC [mgSi/Nm3]
CASE 2 450,000 480 2,500 10

Biogas treatment plant definition

Figure 21 shows the P&ID of the proposed design, following the ANSI/ISA-S5.1-1984 (R 2009) and ISO 10628 (R 1997) rules. The raw biogas coming from the gas holder goes through the treatment plant following the next configuration:
a) First, the biogas is introduced into a biotrickling filter, where the main removal stage is carried out since over 90% of the H2S content is supposed to be removed in this step.
b) Then the biogas enters the polishing stage in which H2S is removed completely as well as the other pollutants. It consists of the following process units:
i. H2S adsorption filters: the biogas is first driven through three adsorption filters in series, in a lead/lag/lag configuration. This design aims to optimize the operation as well as to reduce the operational costs (although the investment is a bit higher due to the necessity of the third filter). Biogas always goes through two of the adsorbent filters: the lead filter adsorbs all the H2S while the lag filter is used in case there are still traces of pollutant. When the first one saturates, the second filter becomes the lead one and third filter becomes the lag one, while the first one is being changed. This way, the treatment plant has 100% of availability and it is ensured that the fuel cell will not be harmed.
ii. Biogas drying: a heat exchanger cools the humid biogas and then re-heats it again, allowing most of its water to be removed by condensation as well as some VOC’s and siloxanes through solubility.
iii. Siloxanes adsorption filters: at the end of the polishing step, the tempered biogas enters two adsorption filters, also in series and in lead/lag configuration, where other pollutants such as siloxanes, VOCs, and the remaining H2S (if any) are physically adsorbed.
Finally, the clean biogas is ready to enter the fuel cell. fuel cell.


Figure 21. P&ID

Finally, it is concluded that the proposed plant is technically feasible, supplying almost the total WWTP demand of heat and the electrical demand for the biogas cleaning facility. Regarding the economic feasibility, with the considered hypothesis, the plant is feasible although with long payback periods. Nevertheless, it is very affected by the economic conditions and the regulatory framework in each country, so each case shall be studied specifically.

Potential Impact:
The activity in MCFC-CONTEX has provided key information to a large number of interested parties who are interested in the forefront, green and efficient technology investigated within the project. Progress has been constantly publicized as well as the immediate and perspective benefits arising from the development in all three lines of activity.

After publication of a feature article in EU-Researcher (2013, September issue, pp. 46-48), a monthly magazine for the promotion and dissemination of scientific research in Europe, negotiations have been completed to post and describe the project on a platform for policy makers, experts, and front-line figures for the discussion and debate of current challenges and instances of best practice (http://www.adjacentgovernment.co.uk/). The space is still online and active and a feature article by the project coordinator was also published in their magazine, issue May 2014.

MCFC CONTEX has further been disseminated at the FCH-JU General Assembly and Project Review days (Brussels, November 2013) and at the Project Cocktail of the European Fuel Cell “Piero Lunghi” Conference (Rome, December 2013). Furthermore, several articles have been published, both through peer-reviewed channels and directed to the general public, results have been presented and discussed at national and international conferences, and the final works have been submitted as articles for publication or are under preparation as reported in table 2 below.

Table 2 Dissemination of MCFC-CONTEX results

When What Where
17 May 2010 Oral presentation “MCFC fed with biogas: effects of H2S and CO” ASME-ATI-UIT Conference on Thermal and Environmental Issues in Energy Systems, Sorrento, Italy
21-22 Mar. 2011 Oral presentation “Electrochemical Impedance Study of the Poisoning Behaviour of Ni-based Anodes at Low Concentrations of H2S in an MCFC” International workshop on molten carbonates and related topics, in Paris
22-25 Mar. 2011 Poster describing the activities of WP3, and the first analytical steps ANAKON 2011, Zürich/CH
24-25 May 2011 Oral presentation “Biogas clean-up for HTFC systems” Workshop “Large Fuel Cell Systems and System Components” in Bruges
18 Jul. 2011 Poster describing the MCFC-CONTEX project and the first analytical steps Sustainable Energy Supply of the Future MSE, Munich/D
Aug. 2011 Article published “MCFC fed with biogas: Experimental investigation of sulphur poisoning using impedance spectroscopy” International Journal of Hydrogen Energy Vol 36 (16), pp. 10311-110318
9-11 Jul 2012 Oral presentation “Detection of Contamination in Biogas using Laser Induced Breakdown Spectroscopy and Raman Spectroscopy” Workshop of DASp and the A.M.S.El. Mainz/D

2-5 Sep. 2012 Poster presentation “Hot gas H2S removal using activated carbons” ISCRE-22 International Symposium on Chemical Reaction Engineering, Maastricht, Holland
Dec. 2012 Article published “Strategies and new developments in the field of molten carbonates and high-temperature fuel cells in the carbon cycle” International Journal of Hydrogen Energy Vol. 37 (24), pp. 19345-19350 (2012)
Dec. 2012 Article published “Electrochemical Impedance Study of the Poisoning Behaviour of Ni-based Anodes at Low Concentrations of H2S in a MCFC” International Journal of Hydrogen Energy Vol 37 (24), , pp. 19312-19318
Dec. 2012 Article published “Experimental and theoretical analysis of H2S effects on MCFCs” International Journal of Hydrogen Energy, Vol 37 (24), December 2012, pp. 19329-19336
4-7 Mar. 2013 Poster presentation describing the MCFC-CONTEX project and WP3 results, poster award ANAKON 2013, Essen/D
Jun. 2013 Article published “Membranes and molten carbonate fuel cells to capture CO2 and increment energy production in natural gas power plants” Industrial & Engineering Chemistry Research, 2013, 52 (26), pp 8755–8764
Sep. 2013 Oral presentation “Experimental procedures for accelerated aging tests using MCFC button cells” Int. Workshop on Molten Carbonates and related issues in Gyeongju (Korea) and publication (IJHE)
Sep. 2013 Oral presentation “Experimental and theoretical analysis of SO2 effect on MCFCs” Int. Workshop on Molten Carbonates and related issues in Gyeongju (Korea) and publication (IJHE)
9.-11. Dec. 2013 Poster presentation describing the MCFC-CONTEX project and results 11. Dresdner Sensor-Symposium, Dresden/D
11-13 Dec. 2013 Oral presentation "The SO2 influence on the kinetics of MCFCs" European Fuel Cell Conference 2013, Rome
11-13 Dec. 2013 Oral presentation “Adsorptive removal of H2S in biogas conditions for high temperature fuel cell systems” European Fuel Cell Conference 2013, Rome
11-13 Dec. 2013 Poster presentation “Performance of a biogas-powered SOFC pilot plant: description of the biogas treatment and optimisation of the fuel cell operational conditions” European Fuel Cell Conference 2013, Rome
Joint Publication CET+ENEA
Feb. 2014 Article published “Effects of sulfur contaminants on MCFC performance” International Journal of Hydrogen Energy 39 (2014) 12242-12250
Joint Publication UNIGE+ENEA
Feb. 2014 Article published “Experimental analysis of SO2 effects on Molten Carbonate Fuel Cells” International Journal of Hydrogen Energy 39 (2014) 12300-12308
23 Apr. 2014 Oral presentation “High temperature fuel cells: innovative applications and contamination issues” IEA – ANNEX 25 Stationary Fuel Cells Meeting, Trento (Italy)

2-3 June 2014 Oral presentation “Electrochemical impedance study of the poisoning behaviour of Ni-based anodes at low concentrations of H2S in an MCFC” 10th International Symposium on Electrochemical Impedance Analysis, Borovetz (Bulgaria)
June 2014 Article submitted “Performance of MCFC fed with simulated flue gas” International Journal of Hydrogen Energy (2014)
June 2014 Article submitted “Adsorptive removal of H2S in biogas conditions for high temperature fuel cell systems” International Journal of Hydrogen Energy (2014)
June 2014 Article submitted “MCFC for CO2 separation – an analysis of feasible operating conditions ” International Journal of Greenhouse gas control (2014)
Joint Publication KTH+ENEA
July 2014 Article in preparation "The SO2 influence on the kinetics of MCFCs" Joint Publication UNIGE+ENEA
July 2014 Article in preparation “Hydrogen sulfide removal by impregnated activated carbon at ambient temperatures" To be submitted to Canadian journal of chemical engineering
July 2014 Article in preparation “Plasma Emission Spectrometry with Electrical Discarge for the Detection of Siloxanes in Biogas” To be submitted to Analytical and Bioanalytical Chemistry
July 2014 Article in preparation “Detection of Sulfur and Silicone Compounds in Biogas by Laser Induced Breakdown Spectroscopy” To be submitted to Analytical and Bioanalytical Chemistry

The activity in MCFC-CONTEX has created the possibility to provide key information to a growing number of interested entrepreneurs who are keen to invest in forefront, green and efficient technology. Through the various channels of relations, often the coordinators were contacted to know about the potential of MCFC technology for specific applications, often related to distributed generation using natural gas or biogas.
These are the parties that ENEA have been in discussion with as regards the installation of new MCFC systems in Italy, often for feeding with biogas, which have led to negotiations with FuelCell Energy Soultions GmbH (FCES) currently at various degrees of commitment:

• ATR of Castelfranco Veneto (Lorenzo Cusinato) have signed an agreement with FCES and are completing a first order
• Net Srl of Rho (Paolo Longoni) are negotiating the supply of a 800kW system for the exhibition centre Milano Fiera in occasion of EXPO2015
• The municipality of Ariccia are discussing a 800 kW system fed with biogas from the anaerobic digestion of the organic fraction of municipal solid waste (OFMSW)
• INFN (National Institute for Nuclear Physics) of Frascati are discussing a 400 kW system for their computing centre
• University Parthenope of Naples are discussing a 400 kW trigenerative system
• The municipality of Terni is interested in a 400 kW system fed with a mix of landfill gas and OFMSW
• The municipal consortium of Rio Marsiglia (Daniel De Ferrari) are interested in a system to be fed with landfill gas
• Hydrogen Energy srl (Felice Basile) is in touch with FCES for a commercial agreement and possibly supply of a system

A dedicated encounter between the MCFC CONTEX consortium and the key European players in MCFC industrialization – namely Fuel Cell Energy Solutions GmbH with their shareholder Fraunhofer IKTS – was organized at the end of the project (June 26, 2014). The objective was to run through the major technical achievements of the joint effort, exchange views on their significance for present industrialization and discuss the opportunities for further exploitation and future collaboration. The trace gas analysis device and the technical-economical assessment of the scale-up of the clean-up system were particularly interesting for immediate take-up in industrial development, and two main areas of further cooperation were defined: trigeneration systems with H2 production for H2 refueling, and generally research activities on new material, components combination and new manufactory processes. A key challenge in the future remains to improve robustness Vs impurities, but other interesting topics also require deeper investigation: utilization of MCFC for naval application in slow motion operation; utilization of MCFC as electrolyzer; utilization of MCFC as trigeneration in industrial chemical plants.

In July 2014, the results of pilot experimentation were presented to the managers of the WWTP of Mataró (Spain) and to the Consell Comarcal del Maresme, the county public authority in charge of wastewater management. In September 2014, it is expected to present the results of pilot experimentation and the whole project to Aqualogy (second external advisor in MCFC-CONTEX).

Overall, the achievements of the project are significant and their impact and potential for further exploitation should not be underestimated:

• cross-cutting technologies such as non-invasive gas analysis and trace contaminant detection, as well as biogas clean-up system design and validation – both developed within MCFC-CONTEX – are of fundamental importance to a myriad of end-user applications and industry processes
• the need for abatement of CO2 concentrations in the atmosphere is widely recognized and painfully necessary: before renewables can supply all our primary energy, reliance will have to be made on conventional power plants, for which solutions have to be found that mitigate greenhouse gas emissions, without compromising primary energy efficiency: using the MCFC as a retrofit CCS solution proves to be promising, though certain criticalities need to be taken into consideration, which have been identified in MCFC-CONTEX.

MCFC-CONTEX has provided crucial data as to the feasibility and possibilities for implementation of the above, and this has been adequately publicized to a large audience of policymakers, industry, research community as well as the general public.


List of Websites:
http://mcfc-contex.enea.it/
See contacts scientific representative

Project information

Grant agreement ID: 245171

Status

Closed project

  • Start date

    1 January 2010

  • End date

    30 June 2014

Funded under:

FP7-JTI

  • Overall budget:

    € 4 132 858,94

  • EU contribution

    € 1 841 832,55

Coordinated by:

AGENZIA NAZIONALE PER LE NUOVE TECNOLOGIE, L'ENERGIA E LO SVILUPPO ECONOMICO SOSTENIBILE