Skip to main content
European Commission logo print header

Operation of a Novel Sofc-battery Integrated hybrid for Telecommunication Energy systems

Final Report Summary - ONSITE (Operation of a Novel Sofc-battery Integrated hybrid for Telecommunication Energy systems)

Executive Summary:
An analysis of the market potential for SOFC in Telecom’s Infrastructure carried out by Ericsson Italia shows that SOFC/SNC solution has a huge worldwide market potential in the telecom sites with load up to 10 kW.
Its potential application area in the world should be approximately five hundred thousand telecom sites (of fixed and mobile network) with related annual energy consumption of around 26 TWh (energy bill of about 4 billion US dollars).
In Europe, the market potential of the solution should involve more than one hundred thousand telecom sites with related annual energy consumption of about 5 TWh. The ONSITE solution can be easily implemented in central offices of fixed network, while operative key issues might be met for the installation in the sites of mobile network such as availability of space, permits, distance from the gas grid especially for those sites located in rural areas.
Moreover, resilient grids meet growing interest for their capability of let survive critical load even in case of power fault coming from grid disturbance and disasters. To do this, such grids are characterized by
redundant apparatus and predictive control schemes.
When high value services are provided to a wide set of customers, unexpected service unavailability is source of economic losses. For this reason, the supply systems of such plants, beside the internal energy storage devices, had better to have redundancy of energy sources (e.g. electrical grid and natural gas network) and tailored power flows control strategies to optimize the equipment utilization and costs, even in the case of external energy shortage.
The ONSITE project, that ran from July 2013 until September 2017, aimed at developing a hybrid system, using two innovative technologies, Solide Oxide Fuel Cells (SOFC) and Sodium Nickel Chloride (SNC) batteries, capable of integrating production and storage devices in one system on the one hand, and of managing and controlling the energy and its exchange with the power grid on the other hand.
The SOFC/SNC hybrid system prototypes realization was the finalization of the SOFC generator development at SOFC stack and Balance of Plant (BoP) level, this in order to perform the electrical and thermal integration between the two technologies. Modelling activities aimed at dynamic SOFC stack, SNC batteries and BoP models.
The SOFC stack achieved the power (more then 10 kW combining two SOFC/SNC hybrid systems) and efficiency targets (40% electrical efficiency, 85% total efficiency in CHP mode). A first SOFC system was delivered by HTC to the project and fully characterized in the summer of 2015. The electrical and thermal efficiency were determined confirming having reached both project targets.
The system was been operated for over 500h, delivering over 1MWh of electrical energy. The system design is modified specifically on the thermal integration requirement of the project. The electrical output is modified from AC to unregulated DC to connect to the power electronics developed in this project. After design freeze three (3) dedicated ONSITE SOFC subsystem were produced to be integrated into the ONSITE system and used to perform validation tests and prototypes realization showing a good electrical and thermal efficiency, 40% and 45% respectively, but, unfortunately, an insufficient durability.
Two prototypes of SOFC/SNC hybrid system have been realized and tested:
- a 5 kW prototype combining a 2.5 kW SOFC system, two SNC batteries and a bidirectional Power Conversion System able to generate both AC and DC power for microgrids;
- a 10 kW prototype combining two 5 kW SOFC/SNC hybrid systems, a bidirectional Power Conversion System (able to generate both AC and DC power for microgrids), and an adsorption chiller (producing cold from the SOFC heat exhaust).
Both prototypes have been validated at CNR laboratories in Messina (Italy). Finally a 5 kW SOFC/SNC containerized prototype has been installed at an Italian TLC operator (fixed network).
The SOFC/SNC containerized prototype with both DC and AC electrical interfaces (to match the served load requirements) was developed for ICT equipment (Radio Base Stations, fixed networks, data centers, etc.) even integrating functionalities to offer grid services as an experimental prosumer site. The fundamental design targets were the energy conversion efficiency and the cost reduction. The former issue was addressed by developing a hybrid fuel cell/battery (SOFC/SNC) based prototype, the latter by the sizing optimization of each used device.
Moreover, starting from field tests data measurement and collection, the hybrid generator was analysed within a resilient grid scenario and later simulated to compare different battery management algorithms and assess the trade-off between battery exploitation (e.g. maximum depth of discharge) and system resiliency against grid or gas network fault.
The developed prototype showed satisfactory performance to be installed in resilient microgrids as well as applications in which the continuous availability is a critical point, such as ICT equipment and datacenters.
However, SOFC still remain a technology not mature enough to face the ICT market because of insufficient durability and high initial investment costs. Recommendation for future research is to increase SOFC stack stability against thermal stresses, to increase SOFC BoP design.
Project Context and Objectives:
The motivation for this project comes from a wide and in-depth analysis of the telecom power supply market characterized by a constantly changing dynamics. Recent R&D activities conducted by proponents, with a relevant support of FCH JU, have conducted to a relevant improvement in SOFC technology.
Despite expecting high efficiency devices, new high value services and a high number of new users, next generation telecommunication systems may require more power and energy than today applications.
To face this issue, telecom companies worldwide investigate the option of producing the power needed onsite with fuel cells taking also advantage of collateral benefits like heating and cooling - while staying grid-connected to use the grid as backup power in case of an emergency only. Fuel cells hybridized with batteries appear to be the ideal option for this application.
An analysis of the market potential for SOFC in Telecom’s Infrastructure carried out by Ericsson Italia shows that SOFC/SNC solution has a huge worldwide market potential in the telecom sites with load up to 10 kW.
Its potential application area in the world should be approximately five hundred thousand telecom sites (of fixed and mobile network) with related annual energy consumption of around 26 TWh (energy bill of about 4 billion US dollars).
In Europe, the market potential of the solution should involve more than one hundred thousand telecom sites with related annual energy consumption of about 5 TWh. The ONSITE solution can be easily implemented in central offices of fixed network, while operative key issues might be met for the installation in the sites of mobile network such as availability of space, permits, distance from the gas grid especially for those sites located in rural areas. Moreover, resilient grids meet growing interest for their capability of let survive critical load even in case of power fault coming from grid disturbance and disasters. To do this, such grids are characterized by redundant apparatus and predictive control schemes.
When high value services are provided to a wide set of customers, unexpected service unavailability is source of economic losses. For this reason, the supply systems of such plants, beside the internal energy storage devices, had better to have redundancy of energy sources (e.g. electrical grid and natural gas network) and tailored power flows control strategies to optimize the equipment utilization and costs, even in the case of external energy shortage.

SOFCs are expected to be about 50-60% efficient in converting fuel to electricity, and in cogeneration applications overall efficiency can be as high as 85%. They are good energy sources to supply reliable power at steady state, especially for telecommunication applications; however, due to their slow internal thermodynamic characteristics, they cannot respond to electrical load transients as quickly as desired. The realization of a hybrid system, capable of connecting production and storage devices on the one hand, and of managing and controlling the energy and its exchange with the grid on the other hand, represents the synergy of some innovative technologies, but already commercially available.
High Temperature Batteries (HTB), such as Sodium-sulphur and Sodium-nickel chloride, have no emissions, permit quiet operation and have been designed for long cycle life. SNC batteries are intrinsically maintenance free, show long life and high reliability and are fully recyclable. The choice of this kind of technology aims at exchanging thermal energy between the two devices, in order to enhance the total efficiency of the final system, as well.
The natural gas (and optionally LPG) operated SOFC and the SNC battery coud be thermally integrated. Both provide power for the end user (telecommunication systems) via a DC/DC converter – the SOFC can also charge the battery pack via an electronic DC power manager that also supplies the DC/AC converter with the power from the SOFC and the battery.
The thermal energy (waste heat) of the system can be applied for heating purposes as well as for cooling applying, e.g. an absorption cooling system.
Another advantage from the system hybridization is higher reliability and availability of the power source, even in case of a failure in one of the two integrated devices. Moreover, SOFC power generation can be modulated over a small range, if necessary to optimize battery operating point.
During peak demand the battery provides power in addition to the fuel cell, whereas the fuel cell recharges the battery during low demand periods. The key advantage of this system architecture is that the fuel cell is operated without major load variations close to constant load resulting in longer lifetime and thus reducing total costs of operation.

The overall objective of ONSITE is the development of a proof-of-concept based on SOFC/SNC batteries hybrid technology with progress beyond state-of-the-art and towards the telecom market requirements. The requirements for the Proof of Concept system were used to create a design specification for all subsystems, and finally the Proof of Concept system. Each separate subsystem has its own set of objectives that is simulated and optimized for the Proof of Concept system objectives, to reach longest possible lifetime, high system efficiency over the whole lifetime, high reliability (uptime), limited cost for operation and maintenance, and last but not least, able to compete on investment costs compared with conventional technologies. Additionally, investigations had been be carried out on the feasibility of combining the SOFC/SNC system with an adsorption heat pump, to discover the potential to boost the efficiency to a higher level and supply cooling as well.
In parallel basic research will be pursued on SOFC stack to reach FCH JU targets in terms of efficiency, duration and costs, strengthening the European position in this technological field.
On top of these activities detailed analyses of final proof-of-concept life cycle cost and total cost of ownership are foreseen exploiting routes for bringing the new technology to market.

Further major objectives of the project are listed below.
- Improvement of SOFC stack design
- Optimization of a multi-stack system topology to develop a high power SOFC system
- Design of a complete SOFC/SNC high power system operating as a whole (i.e. including balance of plant, control system) capable of matching the varying power demands of a medium size TLC station (5-10 kW, 0-100% modulation capability, etc.).
- Integration of SOFC, SNC battery pack and other balance of plant components into an application under challenging constraints, including limited volume, and stringent safety requirements and CE certification.
- Investigation of spin-off opportunities into other market sectors, especially Smart Grid, Smart Building and Rural electrification
- Development of exploitation routes for bringing the new technology to market

To these purpose, this project aimed at maximizing SOFC co-generation efficiency while supporting telecommunication network equipment supply, through the adaptation of existing SNC batteries. SNC batteries can support SOFC systems in electrical load following applications. Moreover, thanks to their thermal integration, SNC batteries can offer a number of further advantages:
- Use of exhaust heat from the battery, when operational, to augment the pre-heating of air entering the stack;
- Using a single air blower to jointly cool the battery and supply air to the fuel cell.
Reciprocally, well sized SOFC systems can provide electrical energy (through efficient generation process) and, by redirecting a minimal part (when necessary) of fuel cell off-gas toward the battery enclosure, the overall system efficiency can be further increased. In fact, this would avoid SNC battery self-discharge, usually due to the power dissipation for the temperature maintenance while no current flow is required.
Moreover, to enter the market in a short while, the integration process must bring high flexibility level to the unified device, so that different applications can be planned. For this reason, the battery customization process will lead SOFCs to different solutions tailored on specific applications.
The use of LPG with steam reforming as alternative fuel is very attractive mainly in the telecom market, as at many sites, no adequate NG infrastructure is available. The ability to operate telecome stations also at remote sites by this enlarged fuel base is a great opportunity and an important multiplier of benefits for the exploitation of the project results.
Project Results:
System requirements
The load telecom profile was studied taking into account two different possible market segmentations: data centers and radio base stations.
The analysis of a Data Center has been provided by Ericsson, due to the high power demand of the site (ca.100 kW) it was possible only to analyze the load profile. It is quite constant along all the time.
The analysis of telecom load profiles showed the details of load features from a DC point of view (TLC apparatus) and from an AC point of view (air conditioning), classifying the power consumptions in relation to season and to the different technologies adopted (GSM 900, GSM 1800 and UMTS), underlining that 80% of Telecom Operator energy consumption belongs to Radio base station sites with a size starting from few kW up to 15kW.
After having determined the suitable load, the consortium focused on defining the main specifications that each single subsystem musts satisfy in order to achieve the project targets.

System modelling

The main goal was to create modelling platform allowing numerical simulation of the complete CHP system built in ONSITE project. Stationary model of the SOFC stack was the first step in the development of a comprehensive modelling tool integrating sub-models of all system components.
The starting point of solid oxide stack technology, developed by HTCeramix, and the developments planned during the project had to match with the FIAMM SNC battery to achieve the best compromise in terms of reliability and functionality of the final system.

From the current-voltage curves acquired from the experiment data, the ionic and electronic conductivities were determined. Figures 1 and 2 present charts used for determination of ionic and electronic conductivities, which are the function of the stack operating temperature.
The validation calculations for the stationary SOFC stack model were performed and the current-voltage curves generated by the model were compared with the experimental data (Figure 3). Simulation results shown that average relative prediction error was lower than 3%.
The main purpose of developing stationary model of the stack was to enable dynamic simulations in next phases of the project. For that reason, several reference conditions were defined and results of the modelling are shown in Figures 4 and 5, obtained for the model.
Advanced modelling techniques were applied to establish predictive modelling platform for both the SOFC stack and SNC battery. Such fullyphysical dynamic models provide reliable tools for predictive analysis of complex operating conditions corresponding to different mission scenarios.
Dynamic model of fuel cell stack has been be used to predict start-up and shut down procedure (Figure 6), to predict stack operation under modified working conditions and predict overall stack voltage with relative model error lower than 4% (Figure 7).
Moreover, dynamic model of SOFC stack can serve to predict stack operation with different degradation rates due to ageing, load and thermal cycling (Figure 8).
The dynamic model of the SNC battery accounts for both thermal and electrical aspects of its operation. Model developed serves therefore for thermal and electrical integration of the battery.
A dynamic model of the SNC battery has been carried out using a data-set of the thermal and electrical behaviour of the battery, as shown in Figure 9.
The resulting equivalents circuits are reported in Figure 10.
The outcome of this activity was that (due to the exothermal reaction, the Joule effect, and the battery thermal insulation) to maintain a good dynamic performance of the battery, it is necessary to have heat losses or active cooling during the discharge phase depending on the rate of discharge. This is to compensate for the internal heating effect and to maximize the charge/discharge current range.
Another result of the analysis is the evaluation of the electrical power required with and without thermal integration with the SOFC subsystem where the hot exhaust gas is sent to the SNC battery.

The work made it possible to create two models, which are integrated, in a combined SOFC/Battery dynamic module. This tool supported the analysis of alternative scenarios, including variable mission parameters, different load profiles and operational modes. Simulator has a built-in functionality allowing investigating the effects of initial conditions of the performance of the hybrid during operation in load following mode. For instance, different starting SOC can be considered to define the limiting values for which the hybrid is able to fully satisfy the power demand (Figures 11 and 12).
Moreover, the simulator can account for variable key parameters o work made it also possible to create two models, which are integrated, in a stand-alone simulator of the hybrid.
The proposed SOFC/SNC hybrid dynamic model allows predicting system inputs and outputs depending on the TLC power needs, thus making it a valuable system optimization tool.
Moreover, the simulator can account for variable key parameters of the SOFC stack with the potential effect on the performance of the complete hybrid. This includes alternative fuels, flow configurations (co-, cross-, counter-current), failure modes, different thermal properties and ambient working conditions.
The simulations of the SOFC/SNC hybrid model for various load profiles show that presented numerical tool allows to predict the system output and its proper input parameters to ensure constant and stable TLC unit operation for long periods.

Battery’s thermal integration
CFD numerical activities of thermal integration were performed in order to minimize the energy requirements of the hybrid system. In standard configuration SNC batteries are heated-up with electrical heaters. In the proposed solution hot residuals gases were used as a heat source for the battery module heating process, allowing also minimizing its thermal losses. Several technical layouts were designed and numerically tested, showing the potential possibilities of the battery module thermal integration, thus increasing overall system effectiveness.
Hot residual gases from SOFC (Figure 13) were used to heat-up and to reduce the thermal losses of the battery module. In the standard solution hot gases were released into the atmosphere, thus it was necessary to develop a model of thermo-flowing chamber for thermal integration of SNC batteries. This task required a multivariate numerical analysis.
Numerical calculations were performed using commercial software (ASNYS Fluent®) and own computational codes to selected SNC battery operating points (Figures 14, 15 and 16).
The proposed design fulfilled all the requirements, which were set during the design work. Thus, this enable the improvement of the overall efficiency of the cogeneration system by using waste gas for the heating or air for cooling the SNC battery module.
To sum up, the goal of the CFD numerical analysis was to integrate thermally battery with SOFC in the hybrid unit. Standard insulation of battery was modified by adding active thermal insulation. The active insulation includes gas ducts that surround battery. Thanks to this solution, performance of battery insulation was improved in heating mode due to flow of hot gas. The implementation of the planned work required preparation of several alternative designs and variant analysis in order to secure proper operation of battery and power generation unit.

SOFC Stack Development

The focus was to develop a single stack that is reliable and highly electrical efficient.
The operational targets of the developed stack are high mechanical robustness, high efficiency and high fuel utilization. A full-scale stack was assembled and tested under dilute hydrogen.
The stack was tested up to 1.8kWel. at 87% fuel utilisation (FU). The systems are typically designed for a slightly more conservative Utilization Factor (UF) and at a 80% UF and corrected to SMR energy input vaules and assuming 5% total BoP losses, these experiments project 57% net AC system efficiency.
The fuel cell stack and system had to comply all necessary codes of safety to be allowed being deployed for a field trial. Gas tightness in the stack is key to safety and durability. Investigations on seal ageing mechanisms were performed to develop alternative sealing materials for robust manufacturing process to ensure safety and durability. Within the project, an innovative protective layers for the seal/metal interfaces has been developed. Part of the stack design has been modified to accommodate these improvements. The improvements were monitored by detailed investigations on stack durability and robustness by stack testing on test-benches. The result is consistency in the stack manufacturing process as proven by the stack I-V curves of 5 individual stack samples. It shows high reproducibility (+/- 1.5% at nominal power) that can only be achieved by robust design and manufacturing process.
In order to establish a field trial, the SOFC stack and system has to fulfill all necessary safety requirements dictated by codes and regulation. During the project, considerable effort was spent outside it to reach this target. Finally in Jan 2015, the type-test approval from an independent certification body was received. Hence the SOFC system, including the stack, were certified and approved for a safe field-trial.
The certification activity required most of HTceramix stack development resources causing delay in the 2.5kW SOFC stack development. This delayed activity prevents the new stack being available for system characterization, and field trial. The Consortium has hence agreed to use the existing G8/80 stack. Each 2.5 kW SOFC system will contain two (2) stacks.
Assessment for LPG operation

The objective of the project is to develop a system suitable for both natural gas and LPG operation. The compatibility with LPG operation was addressed by assessing the suitability of the catalyst used in the reformer, followed by determining the modifications needed to operate the system on LPG fuel.
The reformer contains nickel based catalyst and the exact catalytic activity depends on the composition and morphology. The catalyst used in the system is commercially available and it is developed specifically for a full range of feedstock ranging from natural gas to heavy hydrocarbons. LPG is one of the feedstock that is specifically mentioned, hence no changes in the reform catalyst are required.

System design and manufacturing

To provide detailed understanding of the system operation and characteristics, a first SOFC sub-system was delivered to performed detailed analysis.
Figure 19 shows the variation of gas consumption and electrical and heat production. The regulation of the system is based on the electrical production of the stacks (Pdc). It has been varying between the nominal value (2500 W), one lower value (2000 W) and an upper value (2800 W). It is been tried to go under 2000 W.
The electrical and thermal efficiency is presented on Figure 20. At nominal power production, the electrical efficiency reaches 38% while the thermal efficiency ranges between 45 and 50%. 40% of electrical efficiency has been reached producing 2800 W of electricity.
The efficiency of the stacks is defined as the DC electricity power divided by the gas consumption passing through the reformer and the stacks. This value has been constantly close to 57% and depends on the fuel utilisation. The last is one of the parameters used to regulate the electricity power level. The fuel utilisation have been set to 68% and consequently the stack efficiency reaches 50% with an electrical power production of 2800 W.

After the test campaign, a first dedicated prototype was produced and delivered to physically integrate it with the SNC batteries (Figure 21).
After design freeze and confirmed performance, n.3 more SOFC subsystems were realized and delivered.

Thermal integration

The gas flow temperature inside the SOFC system varies from room temperature to 900°C. The high temperature flows are all integrated inside the HoTbox™ system. The exhaust gas leaving the HoTbox™ is sent to a heat exchanger to provide heat duty to the customer. Typical temperature of this gas is about 220 °C, which makes this gas stream a suitable candidate to be sent to the SNC battery.

The carried out activity allowed the Consortium to define the battery architecture (hot zone – to – cold zone separation, capacity, number of cells per string, number of strings in a battery, thermal capacity, insulation, and so on). In particular, starting from a standard FIAMM/FZsonick product, a 5 string (with 20 cells each) device was developed.
The design was addressed to build an external battery clad (and BoP) to support the thermal integration activity by exploiting the SOFC subsystem exhaust gas (Figure 22).
Electrical integration

The electrical integration analysis involved evaluation of both batteries and SOFC subsystem behaviour, as components of separated system with different interactions in different working status. A first prototype has been realized and test at HTC premises in Yverdon (Figure 23).
Integrated system

The carried out activities led to the development of:

- one 5kW SOFC/SNC hybrid proof-of-concept tested in laboratory and ready to be included in the field trials sheltered system;
- one 10kW hybrid proof-of-concept for laboratory functional tests integrated with an adsorption chiller.

Realization and test of the “base unit” 5 kW SOFC/SNC hybrid system

The “base unit” 5 kW SOFC/SNC hybrid proof-of-concept was used to supply a simulated grid connected radio base station load, with the capability to switch in an off-grid mode as a UPS (Figure 24).
After the realization, the 5 kW SOFC/SNC system was packaged in a shelter (Figures 25 and 26).
The packaged proof of concept SOFC/SNC generator system was able to supply a TLC station to moving from prototype to system able to support an “in field” trial and testing the system in a lab environment.

Telecom Qualification Tests were executed on the 5 kW SOFC/SNC hybrid systems at CNR to simulate field operation like scenarios. The system supplied laboratory equipment as during the normal operation of the TLC load to observe the components behaviour in selected situations that could determine major fluctuations of system variables.
The SOFC generator has correctly started the operations and produced power up to the design value; anyway, more errors affected the endurance test after about 12 hours, so that the internal safety routines stopped the system with its emergency control procedure. After the shutdown the SOFC sub-system was found to be damaged.

On the other hand, the battery and power conversion system demonstrated the effectiveness of the 48V bus voltage stability both in case of load variations (both ramp and step like variations) and grid failure events (disconnection and restoration), even without dangerous fluctuation amplitudes (spikes or overshoots) during the transients.

Tests at communications provider facilities

The aim of the activity was the test of the whole system, in a real environment for the project applications and satisfying the load requirements, even strengthen the telecommunications system availability.
Due to the failures of all the SOFC generators installed both in laboratory and in the shelter after some operating hours in previous activities, to carry out the field trials, two commercial SOFC generators were used.
The containerized system was transported to the telecommunications station site and the field trials were carried out. After the operations to adapt the power and data connection to assure the continuous service of the pre-existing equipment (Figure 27).
That prototype system was used to supply a sheltered grid-connected telecommunication equipment load, with the capability to switch in an off-grid mode as a 48V DC UPS.
Two smaller units separately acquired replaced the SOFC system developed within the project; all the remaining devices (power converters, control system, cooling system, energy storage system) were successfully tested to comply with the operator requirements.

To demonstrate the effect of the hybrid system integration (despite the limitation of the possibility of using a 600W SOFC system instead of the 2.5kW one as per the design calculation), the load consumption was observed and logged in the two both configurations (standard TLC system and TLC system with hybrid supply integration).
In the former case, an eight-day logging of the load consumption (three single-phase load readout and the total power from the grid) is reported in Figure 28.
After those measurements completion, the (on average use) same load was observed in the latter configuration to visually represent the effect of the hybrid system on the power (and single-phase current) fluctuation, since the power compensation algorithm let envisage a smoother trend of the 3-phase AC load behavior. In Figure 29, it is possible to see a lower current and power peak-to-peak variation. Further, even more evident, the AC voltage is strongly more stable due to the presence of the internal voltage following algorithm of the power inverter that mitigate the effect of load variation directly on the system insertion point (POD) to limit the power demand from the grid.
In parallel, to evaluate the contribution of the developed system, the SNC batteries current (discharge over time) to compensate for the difference between the load power consumption, and the SOFC generation and the (feedback limited) grid power, was measured and reported in Figure 30. Those plots show the slow discharge at about 53V over time. It was impossible to recharge the battery due to the very small power from the SOFC with respect to the load.
For this reason, (the average load power significantly above the SOFC actual generation) this beneficial effect would have been more evident by using the 2.5kW SOFC system as in the initial system design of the project.
In summary, the addressed tasks and achievement are reported:
Validation of operating performance of key system components: Reached (Every component worked as expected in single device test)
Integration of the SOFC subsystem, the adapted SNC battery, the mechanical and electrical Balance of Plant components and the controller into a system to be tested in the lab: Reached
Test of the integrated fuel cell power generation systems in a laboratory environment to check they meet requirements: Reached (The system was tested in lab and worked properly until the SOFC generator failure. After the check the single components were sent to FIA in order to be integrated in a shelter. A new one SOFC generator will be delivered by HTC to replace the damaged one.)
Demonstration of operation and systems response under start-up, power cycling, thermal cycling, and normal and emergency shutdown conditions: Reached
Set-up / adaptation of the telecom site for safety and continuous operation guarantee: Reached (The developed system provides UPS function both for 48V DC, as per the expected target of the project, and 400V AC loads)
Integration of the SOFC subsystem, the adapted SNC battery, the mechanical and electrical Balance of Plant components and the controller into a system for field trials: Reached (Despite the replacement of the SOFC generator, due to the failures of 5 SOFCs in previous activities, the selected system (n.2 300 W SOFC generators working in parallel) has been validated in a real telecom site.
Test of an integrated hybrid fuel cell based power generation system at the selected telecom site: Reached

Realization and test of the 10 kW SOFC/SNC hybrid system

The activity on 10kW hybrid proof-of-concept (combination of two “base units” 5 kW SOFC/SNC hybrid systems) comprises the functional and integrity verification of each subsystem (SNC batteries, power conversion devices, adsorption chiller) and the final prototype test.
Starting from the already developed SOFC/SNC hybrid systems and the system design, the activity was:
- To check the bidirectional AC/DC converter (inverter delivered by VEC) to support the whole power deliverable to the grid;
- To couple the two “base units” 5 kW SOFC/SNC hybrid systems from the electrical and thermal point of view, i.e. voltages and currents regulation to distribute the power request from the load, battery charge and discharge strategy to optimize the storage capacity utilization;
- To realize a unique controller (hardware and software) to coordinate the two component subsystems controller.
In order to approach the overall system proof-of-concept tests each component was separately tested to verify their integrity and correct functionality, as expected in the design phase.
In particular, these preliminary laboratory tests involved:
- The SOFC modules to verify the stacks integrity, the internal controllers functionality and the exhaust gas utilization feasibility;
- The DC/DC boosters functionality and their capability of working in parallel;
- The controlled charge and discharge of the batteries through the DC/DC bidirectional converters;
- The inverter to connect the two modules to the grid as a single hybrid system. A dedicated activity was carried out to develop a 10kW bidirectional inverter able to support the two modules both during the system start-up and in each expected operating mode.

The 10 kW SOFC/SNC prototype was coupled with an adsorption chiller to arrange the laboratory setup for the experimental assessment of the benefits (in terms of energy conversion efficiency improvement and optimization of the operation point of every subsystem) coming from the integration (Figure 31).
The selected adsorption chiller is a commercial unit from Invensor, using silica gel/water as working pair.
CNR selected and acquired (specifically developed and realized for the ONSITE prototype) two DC/DC step-up converters to connect the two SOFC subsystems to the 400Vdc bus that are expected to support this operation mode.
Despite the expected capability of working in parallel, when the two boosters were connected to create a single 400Vdc bus for the hybrid 10kW system, that connection led to the irreversible malfunctioning of one of the two boosters.
For this reason, there are no operational data of those devices together, and the laboratory tests went on by using the only working hybrid 5kW SOFC/SNC hybrid system.

Due to the damaged booster, the SOFC generator test was performed with a single SOFC subsystem connected to the remaining booster. The tests aimed to verify the ability of the SOFC/SNC hybrid system to demonstrate operation and systems response under start-up, power cycling, thermal cycling, and normal and emergency shutdown conditions.
The SOFC/SNC hybrid prototype was switched-on according with the following procedure:
1. Control system ON
2. Inverter ON – 400 V AC grid connection
3. DC/DC 48 V ON – telecom load enabled
4. Batteries ON - 24 hours warm-up and charging up to 60% SOC
5. SOFC generator ON
6. Normal operation ON

As visible in Figure 32, the SOFC generator was started, the warm-up procedure was completed in about 20 hours. Afterwards, the generator was operated at increasing load values.
The 10 kW SOFC/SNC prototype has been coupled with an adsoption chiller to arrange the laboratory setup for the experimental assessment of the benefits (in terms of energy conversion efficiency improvement and optimization of the operation point of every sub-system) coming from the integration.
The selected adsorption chiller is a commercial unit from Invensor, using silica gel/water as working pair.
The positioning of the adsorption chiller and the ONSITE hybrid system for the assessment of the energy benefits coming from the integration (improved efficiency of the chiller, improved energy conversion efficiency of the overall system, optimization of the working point of each device).
The system was experimentally characterized to assess their performance map under typical working boundaries (Figure 33).
Data from experimentation were used to develop a dynamic model in TRNSYS environment. The model was setup to provide electric and cooling energy to a reference telecommunication shelter, which is a new application for SOFC-CCHP systems. The numerical model was used to perform yearly simulations, aiming at the evaluation of optimized system layouts, in terms of energy efficiency and CO2 emissions reduction.
The integrated system was designed with the aim of exploiting the cogeneration capabilities of the SOFC to operate an adsorption thermally driven chiller, increasing the overall efficiency by utilisation of the as much as possible the energy produced. A schematic of the system is shown in Figure 33. The heat produced by the SOFC is provided to a thermally driven chiller that, in turns, produces the cooling power needed for the load. A sensible heat storage unit employing water as storage media was included, to guarantee a more constant inlet temperature to the chiller and decoupling the energy production of the SOFC from the request of the chiller. Indeed, while the SOFC can modulate the current produced with a dynamics of 2 A/min, which means a thermal energy production quite stable, the daily as well as seasonal variation of the load is much more marked. Adding a storage tank to the layout removes the need for the SOFC to rapidly follow the cooling demand. Indeed, the storage tank stores the excess heat produced by the SOFC when the cooling load is low, that can be eventually used for peak demand. Hence, start-up time and precise regulation of part load is not an issue for the proposed solution. A back-up heater was considered as well, to be operated when the request of hot water from the chiller exceeds the available one, produced by the SOFC. Condensation and adsorption heat of the thermal chiller, instead, are released into the environment by means of a dry cooler.
The electric energy produced by the SOFC is used to cover the electric load of the system and to drive the auxiliaries of the adsorption chiller or the adsorption-compression unit. Excess generated power is sent to the grid.
The experimental activities and the yearly base simulations demonstrated that the multi-generation approach employing an adsorption chiller can achieve high energy savings: the maximum achievable global efficiency ranges between 0.5 and 0.6. CO2 emissions avoided are up to 38 t/y.

Most of the objectives were reached – all in particular for the BoP components.
Unfortunately a long term operation was not possible due to failure of the SOFC subsystems.

Potential Impact:
One of the most important topics that currently are on top of the Telecom Operators agenda is to develop environmentally better and finally sustainable solutions. Energy consumption and CO2 emissions reduction are a worldwide goal. Telecom networks constitute a major sector of Information and Communications Technology (ICT) with a tremendous growth, in terms of adoption of new technologies, in the already developed countries, and in terms of coverage in the developing ones. Taking this into account, the ways to manage and save energy are definitely the key aspects that allow the Telecom Operators to achieve operational effectiveness.

Furthermore, energy cost is a significant portion of operating costs and continues to increase. Telecom Operators, both in developing and mature markets, are realizing the necessity to reduce energy cost related to the operational expenditure (OPEX) and to reduce the dependency on diesel or fossil fuels to reduce emissions and ensure sustainable future growth.
Sustainable development is a global strategic goal, which seeks to achieve economic growth that promotes a fair and just society while conserving the natural environment and the world’s scarce, non-renewable resources for future generations.

The European Union has targeted a reduction of 20% CO2 emission for the year 2020. A small but nevertheless significant part of this energy reduction scheme concerns the Telecom industry and ICT that participates in a direct, indirect and systematic way. Characteristic examples are green networks, smart buildings, smart grids, Intelligent Transportation Systems (ITS), energy efficient electronics (OLEDs, photonics, nanotechnology) and the application of embedded systems towards low carbon and energy efficient technologies.

In order to respect the propositions above, the European Telecoms Network Operators’ Association (ETNO) has prepared a paper of intent called “The Sustainability Charter”, with the aim of being able to give an important contribution to make the sustainable development happen.

The Sustainability Chapter contains many propositions and intents, related to different activities. The point 4) Procurement says: ”To implement efficient management of resources, energy use, waste, emissions reductions, environmentally friendly process and product requirements; eliminating use of hazardous materials; observation of human rights and labour conditions.”

Environmentally friendly solutions such as SOFC/SNCB will surely have opportunities for application and growth. Electricity delivered by the grid, produced in big central power plants but which cannot be used, brings to the waste heat and in addition there are losses in the grid. It brings to a “fuel-to-consumer” electrical efficiency of only around 40%. The situation is totally different with ONSITE’s SOFC/SNCB concept. Electricity generation efficiency approaches 40% and - since the generation is at the consumer’s site – the extra heat generated by the solution is not wasted, but is valuable thermal energy to be used for heating or cooling through absorption chiller in case of telecom sites with power consumption higher than 20 kWatt: this allows total efficiencies of being above 60% for cooling scope. In conclusion, SOFC/SNCB concept needs less fuel and is therefore environmentally and ecologically more friendly. Additionally, the simultaneous production of electricity and heat will allow SOFC/SNCB solution to be competitive in the market of the trigeneration systems.

In order to assess the potential accessible market of the ONSITE concept in the Telecom industry it is best to start with an analysis of strengths, weaknesses, opportunities and threats (SWOT analysis):

Strengths: characteristics of the business or project that give it an advantage over others.
Weaknesses: characteristics that place the business or project at a disadvantage relative to others.
Opportunities: elements that the project could exploit to its advantage.
Threats: anything that could put the success at risk.

For ONSITE the analysis is the following:
- Proven efficiency of technologies that compose the SOFC/SNCB solution: overall CHP (combined heat power) system efficiency is approximately slightly above 80%
- Thanks to its modular concept the SOFC/SNCB is a flexible solution (composition in subsystem SOFC generators of 2,5kW and individual batteries)
- The solution can solve the problem of energy peak demand. In fact, fuel cell alone answers slowly to the energy peaks, the integration with SNCB batteries allows to meet the peak demand thanks to the utilization of the high temperature batteries with high dynamics. Additionally SNCB batteries can as UPS in telecom sites.
- traditional fuels (natural gas, LPG) can be used
- Reliability and durability are still issues for SOFC
- Cost is still high and it is not sure that competitive cost will be achieved soon. CAPEX must decrease.
- The solution is still in a state of concept, although it is already clear that it will develop a shelter/container.
- For a massive roll-out the solution should be reengineered by reducing size and weight of the ONSITE apparatus

- Big market potential in the Telecom industry: energy costs and the EU target of sustainability for 2020 are slowly forcing many Operators to adopt more efficient or even green solutions.
- SOFC/SNCB + biofuel is a potential sustainable solution (low emissions).
- Capability of the current system of using the thermal energy produced by the system in telecom site for cooling scope (telecom sites with power consumption >20kWe)

- Sites selection should take into account connection to a gas grid, environmental constraints , lack of space for the equipment and permits issue can be further constraints.
- The connection to a gas grid may be an issue especially in rural areas in terms of feasibility and costs.
As alternative solution, thanks to the fact that SOFC can convert liquid natural gas (LNG) and liquid petroleum gas (LPG), sites not connected to the gas grid can be served. However, it requests a significant increase in operative costs (opex) due to need of recurrent activities of refueling.
- In case of off-grid sites, there is the strong competition from established, well proven, comparably cheap and mature technologies like e.g. diesel generators. They can be worse in opex, but the total cost of ownership (TCO) can be low due to low investment cost of diesel generators.
- Regarding the capability of the current system to reuse the thermal energy (“waste heat”) produced by the system for air conditioning, an additional investment is necessary for the purchase of an heat pump/absorption chiller and other components.
- Moreover, according to Directive 2004/8/EC aimed to increase energy efficiency and develop high efficiency cogeneration of heat and power, energy efficiency certificates (White Certificates) are given for CHP (combined heat power) solutions.
In order to get White Certificates for SOFC/SNCB hybrid solution, it would be necessary:
o to integrate SOFC/SNCB hybrid solution with a heat pump or an absorption chiller
o to meet the requirements of high efficiency cogeneration defined by Directive 2004/8/EC.
Below potential operative/commercial issues related to the implementation of SOFC/SNCB hybrid solution are highlighted:

- Lack of space for the realization of the system especially in the case of urban sites of mobile network implemented as e.g. roof-top.
- In case the Telecom site is not in proximity to the gas grid, certified high length pipes are necessary to bring the gas to the machine and this implies additional costs that might impact significantly on TCO and maybe will not be approved. An alternative solution to the network grid connection could be based on the usage of LNG or LPG tanks. However, it requests some increase in operative costs (opex) due to need of recurrent refueling activities.
- Incentives: it is likely critical to get white certificates for this type of plant (issue related to the use of the thermal energy produced for heating/cooling purposes in a Telecom site), unless the solution is integrated with a heat pump apparatus and meets the requirements of high efficiency cogeneration defined by Directive 2004/8/EC.

Worldwide total addressable market

Addressable market is focused on radio sites of Mobile Network (1 kW-10/12 kW) and small central offices of fixed network (2-10 kW)

Radio sites of Mobile Network (1 kW-10/12 kW)

We estimate there are about 3.4 million telecom towers installed in the world, growing at a cagr of 4.1% to 2020 by taking into account that different Operators are co-locating their own equipment at the same site.

From the estimation we have excluded the small cells where our solution doesn’t find application (small cells are based on small Telecom equipment, maximum transmission power around 5-7 Watt) and are aimed to provide a small radio footprint, which can range from 10 meters within urban and in-building locations to 2 km for a rural location).

We estimate that the amount of radio sites may increase, but slightly in the near future: over 4 million by the end of 2020.

In fact, on one hand there is a strong increase of the number of radio base stations (especially in emerging markets), on the other hand several Operators choose to co-locate their equipment in the same physical site, by realizing the tower sharing in different ways:
o Tower sharing between Operators is where Operators permit each other to install active equipment on their respective towers under a bilateral agreement. This sharing is also of more use to established Operators than new entrants. The sharing costs of shared infrastructure between multiple Operators, sometimes via an intermediary, can bring a significant opex reduction for the Operators. This is a driver in more penetrated and mature markets, where price competition and margin pressure are big features of the market.
o Sale-and-leaseback structures can be used to establish tower sharing between Operators or mainly between an Operator and an existing Towerco (that is a company that is owner of the infrastructure, without providing Telecom services and managing network performance). Towers are transferred to a specialist Towerco and space on them is “leased” back to the party that contributed them. In this way Operators may benefit from monetizing real estate assets by a sale and-leaseback with a specialist tower company. For Telecom Operators, this is a way of raising capital for investment in other markets, or for new technologies

TowerXchange estimates in 2016 approximately three and a quarter million telecom towers in the world, of which around 60% are owned and operated by a Towerco valued at over US$250bn. In detail, China hosts the world’s largest and newest tower industry, China Tower Company (CTC), which owns more towers than the rest of the global tower industry combined after the recent purchase of the towers from the three Chinese mobile Operators.
In case of co-siting, where one Operator hosts other Operators and/or a tower company (owner of the sites infrastructure) hosts different Operators, the range of power consumption of a typical radio site is from 3 kW to over 12 kW according to the number of Operators sharing the site infrastructure.

Small central offices of fixed network (2-10/12 kW)
By taking into consideration the current amount of worldwide fixed broadband (in 2016 around 800 millions), the number of telephone lines (in 2016 about 1 billion) and the amount of fixed lines handled on average by each small central office, we estimate that currently the number of small central offices of fixed network (up to 10 kW) in the world is in the range 330.000-390.000.
However, it is difficult to give a precise estimation of the amount of worldwide small offices with power absorption up to 10 kW.
In fact, many Operators are used ,where it is feasible, to reuse the infrastructure of their small central offices, by hosting additional Telecom loads, such as radio or fixed equipment of other Operators, or their radio base stations.
It brings to a significant increase of the power absorption in the range 12-20 kW, that means to exceed significantly the power of 10 kW. However, the modular concept of ONSITE would allow to provide that power by adding more modules.

In terms of number of sites where the implementation of the solution SOFC/SNCB is feasible (due to gas grid distance, availability of space, permissions, etc), we estimate the following figures worldwide:
n° of radio sites of Mobile Network (1 kW-10 kW) : approximately 260.000-280.000 around 8-10% of the total amount, with an annual consumption energy of about 10-15 TWh (value estimated)

n° of small central offices of fixed network (2-10/12 kW) : approximately 220.000-240.000
about 60% of the total amount, with an annual consumption energy of about 16 TWh (value estimated)
For the European market we estimate the following figures:
n° of radio sites of Mobile Network (1 kW-10 kW) : approximately 55.000 around 8-10% of the total amount, with an annual consumption energy of over 2 TWh (value estimated)

n° of small central offices of fixed network (2-10/12 kW) : approximately 45.000
about 60% of the total amount, with an annual consumption energy of about 3 TWh (value estimated)
Moreover, we have identified three main types of Telecom Operators on which it is worth focusing on.
Surely the impact of SOFC/SNCB change among the three types of Operator:
Type 1 Fixed Operators with central offices
For a “small” central office of fixed network (2-10 KW) the annual energy costs is around 3k-14k€.
Type 2 Fixed/Mobile Operators with both fixed and radio sites
Type 3 Mobile Operators with all radio technologies installed in their radio sites: GSM, UMTS/HSPA, LTE. Annual energy cost per each radio site is around 5000€. In case of one radio technology missing (GSM missing), annual energy cost per each radio site is around 4000€.

All the three types of Operators could open interesting application scenarios for SOFC/SNCB solution.

Main dissemination activities:

The EERA/ONSITE project workshop on Hybrid Energy and Energy Storage Systems took place at Ericsson in Rome on September 21-22, 2017 following the first workshop in San Sebastian in March 2017. The main objective of the workshop was to stimulate an in-depth discussion on the use of different energy and energy storage technologies in combination. The workshop was organised in four sessions that were introduced by an expert followed by a discussion. Moreover, interesting results from Ericsson and the main results from the ONSITE project were shown.

The workshop was jointly organized by EERA JP Energy Storage and ONSITE project gathering close to 50 participants. In addition, the EERA JP Fuel Cells & Hydrogen and Fuel Cells and Hydrogen Joint Undertaking also contributed to the presentations and the discussion. A report summarizing the main conclusions is under preparation.

Dissemination by:
- mailing list (EERA, FCH JU working groups, other)
- EERA JP Energy Storage website:
- FCH JU website:
- CNR ITAE website:

Magazine articles

Generatori ibridi per il futuro - Platinum Novembre 2016, pag. 98 [IT]
Hybrid generators for the future - Platinum November 2016, p. 98 [EN]

A magazine article has been recently published on International Innovation issue N. 173
both electronically ( and on
hard copies.
The impact of the article will be maximized:
- Mailing hard copies to a contact list;
- Linking from ONSITE website direct to the article page within the digital magazine;
- Posting the high resolution version on ONSITE website;
- Distributing the low resolution PDF to ERICSSON wider stakeholder database;
- Linking the article via social media such as Twitter (

List of Websites:
Public website address:
The project website was maintained and updated to have a fruitful exchange platform for the project partners and to give visibility to the public project documentation.
The virtual space was extended to collect test data, project deliverables, and meeting presentations.
Moreover, an access counting service was activated to determine the project dissemination impact through the web; the website registered about 8000 visitors during the project development.