The innovative energy systems equipped with large volume size connected with a gas turbine present
critical issues related to the surge risk especially during transient operations. In comparison with
standard gas turbines, the volume size generates different behaviour during dynamic operations (with
significant surge risk), especially considering that innovative additional components include
important dynamic constraints. The main examples of power systems including the effect of an
additional volume connected with a turbine are: fuel cell based hybrid plants, humid cycles, externally
fired layouts and innovative systems including high temperature thermal storage devices. The aim of
the work in this thesis is to cope with the surge instability in fluid dynamic machines through the
analysis of vibro-acoustic signals generated from the turbine compressor. Investigations based on
acoustic and vibrational measurements appear to provide an interesting diagnostic and predictive
solution by adopting a suitable quantifier calculated from microphone and accelerometer signals. For
this scope, a wide experimental activity was carried out with a test rig composed by a T100
microturbine connected to a modular vessel. Starting from the stable operation, the instability of surge
was produced progressively by closing a valve in the main air line while the system was equipped
with different dynamic probes to measure the vibrations during normal and surge operations. A
machine dynamical characterization was useful for a better interpretation of signals during its
transient to the surge. Hence different possible methods of incipient surge identification have been
developed through the use of different signal processing techniques in time, frequency and angle
domain. A Vibro-acoustic analysis was carried out to develop precursors which are able to highlight
the unstable operative zone approaching and to produce control data (e.g. an on/off signal for a bleed
valve) for surge prevention. Finally, four of surge precursors found were implemented into a new
diagnostic real-time software. By using the measured data, this innovative tool has proved to be able
to recognise a surge incipience condition by comparing the precursor values with a set of moving
thresholds.

In many engineering design and optimisation problems, the presence of uncertainty
in data and parameters is a central and critical issue. The analysis and design of
advanced complex energy systems is generally performed starting from a single operating condition and assuming a series of design and operating parameters as fixed
values. However, many of the variables on which the design is based are subject to
uncertainty because they are not determinable with an adequate precision and they
can affect both performance and cost. Uncertainties stem naturally from our limitations in measurements, predictions and manufacturing, and we can say that any
system used in engineering is subject to some degree of uncertainty. Different fields
of engineering use different ways to describe this uncertainty and adopt a variety of
techniques to approach the problem. The past decade has seen a significant growth
of research and development in uncertainty quantification methods to analyse the
propagation of uncertain inputs through the systems. One of the main challenges in
this field are identifying sources of uncertainty that potentially affect the outcomes
and the efficiency in propagating these uncertainties from the sources to the quantities of interest, especially when there are many sources of uncertainties. Hence, the
level of rigor in uncertainty analysis depends on the quality of uncertainty quantification method. The main obstacle of this analysis is often the computational effort,
because the representative model is typically highly non-linear and complex. Therefore, it is necessary to have a robust tool that can perform the uncertainty propagation through a non-intrusive approach with as few evaluations as possible.
The primary goal of this work is to show a robust method for uncertainty quantification applied to energy systems. The first step in this direction was made doing a work on the analysis of uncertainties on a recuperator for micro gas turbines,
making use of the Monte Carlo and Response Sensitivity Analysis methodologies to
perform this study. However, when considering more complex energy systems, one
of the main weaknesses of uncertainty quantification methods arises: the extremely
high computational effort needed. For this reason, the application of a so-called
metamodel was found necessary and useful. This approach was applied to perform
a complete analysis under uncertainty of a solid oxide fuel cell hybrid system, starting from the evaluation of the impact of several uncertainties on the system up to a
robust design including a multi-objective optimization. The response surfaces have
allowed the authors to consider the uncertainties in the system when performing
an acceptable number of simulations. These response were then used to perform a
Monte Carlo simulation to evaluate the impact of the uncertainties on the monitored
outputs, giving an insight on the spread of the resulting probability density functions and so on the outputs which should be considered more carefully during the
design phase.
Finally, the analysis of a complex combined cycle with a flue gas condesing heat
pump subject to market uncertainties was performed. To consider the uncertainties
iv
in the electrical price, which would impact directly the revenues of the system, a statistical study on the behaviour of such price along the years was performed. From
the data obtained it was possible to create a probability density function for each
hour of the day which would represent its behaviour, and then those distributions
were used to analyze the variability of the system in terms of revenues and emissions.

The growing awareness on climate change and pollution has brought to national and international policies aimed at promoting the development of innovative and environmentally sustainable energy systems. Among these systems, fuel cells are one of the most promising technologies, characterized by high energy conversion efficiencies and low emissions. In particular, hybrid systems based on the integration of a high temperature fuel cell with turbocharger-derived machinery have drawn the interest of academia and industry over the past decades. However, the complexity, fragility and high cost of these plants have slowed down their development, and only a few big companies were able to build complete prototypes. The technological challenges faced by the scientific community have highlighted the importance of simulations to design, test, control and analyse fuel cell hybrid systems.
Based on this experience, this thesis wants to expand the current knowledge on solid oxide fuel cell hybrid systems, with a particular focus on an innovative small-scale biofueled turbocharged layout, which was introduced recently within the Bio-HyPP European project. The main goal of this thesis is to determine if this kind of system can be a viable alternative to micro gas turbine-based systems, analysing its steady-state and transient behaviour in various operating conditions. To do this, it is necessary to define the system operative constraints, and to develop a control system capable of ensuring their compliance, while optimizing the plant performance. The possibility of increasing the reliability of solid oxide fuel cell hybrid systems is finally investigated, considering the implementation of surge prevention techniques and diagnostic tools.
All these activities strongly relying on simulation tools. This was possible thanks to the collaboration between the Laboratory of Applied Mathematics, Simulation and Mathematical Modelling with the Thermochemical Power Group of the University of Genoa.
After introducing the layout of the turbocharged fuel cell system, a detailed steady-state model of the plant is developed in Matlab®-Simulink® and used to design a strategy, based on the control of valves installed on the plant, able to comply with its many operative constraints. Then, an off-design performance analysis of the system is performed, considering simultaneously various conditions of power load and ambient temperature. This analysis is used to confirm the effectiveness of the proposed control strategy and to assess the capabilities of the turbocharged system.
A dynamic model is created using the TRANSEO tool to study the transient behaviour of the system. Having adopted a control strategy based on the cold bypass valve, the response of the system to a valve opening step change is analysed in order to design an effective and responsive control system, able to keep the fuel cell maximum temperature constant while complying with the system constraints. Four different controllers are designed, tested on two different load variation scenarios and compared on the basis of many parameters.
The final part of the thesis regards the development of innovative tools aimed at improving the reliability of solid oxide fuel cell hybrid system, in particular surge prevention techniques and Bayesian belief network-based diagnosis systems.
A simplified dynamic model of the turbocharged SOFC system is developed in TRANSEO, and various surge prevention techniques are tested on it: intake air conditioning, water spray at compressor inlet, air bleed and recirculation, and installation of an ejector at the compressor intake. The most effective procedures are integrated with the controller of the hybrid system and tested during a transient scenario to prevent the compressor operative point from approaching a surge condition.
Bayesian belief networks aimed at diagnosing the status of SOFC hybrid systems are developed thanks to a collaboration between the University of Genoa and the Mälardalens Högskola of Västerås, Sweden. A micro gas turbine– solid oxide fuel cell system is considered for this study, but the methodology could be easily extended to turbocharged plants. The activity is carried out simulating the system on Matlab®-Simulink® and designing the Bayesian networks on Hugin Expert. Two different diagnosis systems, one for the turbomachinery and one for the fuel cell stack, are developed and tested on stationary conditions. The second one is also tested during transients and integrated with the control system to prevent degradation of the fuel cells.
In conclusion, this thesis highlighted the great potential of turbocharged SOFC hybrid systems, showing high energy conversion efficiencies in a wide operative range in terms of load and ambient conditions. It also showed that the proper operation of the system is possible during various transient scenarios, implementing cascade controllers designed to act on a cold bypass valve to control the SOFC maximum temperature. Regarding the possibility of improving the reliability of these systems, surge prevention techniques based on compressor recirculation appeared as the most effective ones. Simulation results suggest that their integration with a surge precursors detection tool could avoid the occurrence of many potentially dangerous scenarios. The final part of this thesis showed that the durability of SOFC hybrid systems could be further improved thanks to Bayesian belief networks, which were proved to effectively diagnose the status of SOFC-MGT systems but could be applied to turbocharged plants as well.

The current global and environmental energy scenario, increasingly focused on the search for
sustainable energy solutions able to represent an effective and convenient alternative to the use
of fossil fuels, considers the energy from the sea, and in particular from the waves, a very
interesting resource, thanks to its enormous potential.
However, the great variability of the phenomenon and the complex and harsh environment where
it is intended to operate makes the identification of an efficient, robust, replicable and low-cost
technology extremely challenging. In fact, although many solutions have been proposed over the
years, today there is not a technology that is able to fully satisfy the requirements of reliability
and low costs necessary for market success.
In this context the present work aims at analysis of point absorber type wave energy converters
and the development of a facility dedicated to experimental campaign in relevant environment,
compassing in various activities such as dimensioning and design, numerical modelling,
experimental activities.
Chapter 4 shows the activity carried out regarding the wave energy converter called SeaSpoon,
developed by the Thermochemical Power Group of the DIME Department of the University of
Genoa. Several experimental campaigns have been carried out and different calculation codes
have been developed in order to evaluate the potential performance of the device in different
marine conditions, equipped with different types of profiles.
Such technology has led to the development of the SeaWHAM project, illustrated in chapters 5
and 6, whose purpose was to design, implement and characterize an artificial wave generator
installed on a quay connected to the open sea, to enable tests in a controlled but relevant
environment (unique facility in Europe). Furthermore the project targeted the design, installation
and experimental campaign on a integrated system able to convert the wave power (caught
thanks to a “point absorber” wave energy converter) in other forms, such as hydraulic, pneumatic
and electrical energy, with experimental characterisation of the entire efficiency chain.

The current global energy scenario requires specific attention to new non-fossil fuels resources. At the same time, it is impossible to identify best future technologies, characterized by good performance and cost effectiveness. Energy conversion from wave motion represents one of these sectors, with a continuous interest among last years. Research activities in this area aim to identify the most promising solutions capable of being robust, repeatable and efficient. Currently, no technology can be considered leader of the sector even if experimental activities and patents are really numerous. The topic of this thesis is a detailed analysis of the wave energy conversion system developed by the Thermochemical Power Group (University of Genoa): the Seaspoon. This device has undergone numerous changes over the years justified by various experimental and simulation activities. The thesis will then address the main steps that led to the definition of the first full scale prototype, together with its off-shore installation in 2015. Last part of the work is related to the Sea Wham project, that is related to the design and construction of an open sea concept of wave generator and to the development of the latest Seaspoon’s configuration.

This thesis work is organized into three macro-chapters, each one is focused on a specific application. The descriptive organization follows the chronological order in which the case studies were addressed and follows the level of maturity of the research object of this program. Therefore, in the following chapters each single case study is presented and discussed separately. Before starting the description of the work conducted by the candidate along these years, two brief presentation are introduced to allow a better understanding of the topic and methodology at the base of the current research. Chapter 2 reports an introduction on the theme of the surge instability, which is then resumed by the discussions during each case study dissertation; while Chapter 3 is dedicated to the description of the work theorized by Greitzer in the 70s, on which current work lays its foundations.
The following list is thought to anticipate what are the main outcomes and learnings which will be encountered through the text. In this way the candidate wants to clearly summarize what can be considered as primary steps, contribution to the energy systems modeling research field and main outputs of this work:
– Mathematical descriptions for components development with integrations and improvements will be described application by application. Chapter 4.2, Chapter 5.3, Chapter 6.3.
– Description of the proposed steps followed by the methodologies used to extend compressor maps for surge simulation purposes. A mix of analytical and empirical methods will be suggested through this work. Chapter 4.4, Chapter 5.4 Chapter 6.2.
– A method for dynamic delays characterization of complex piping systems named ?-Flow approach have been proposed and applied to these studies. Chapter 4.3, Chapter 6.1.2.
– Parametric studies to investigate the effect of different parameters such as volume size, shaft inertia and equivalent lengths, on the surge cycles characteristics. Impacts of these characteristics will be analyzed in different control configurations where options are applicable. Chapter 4.6, Chapter 5.6.1, Chapter 5.6.2 Chapter 6.4.
– Simulation results, experimental comparison and validation will be presented at the end of each macro-chapter as completion of the analysis. Chapter 4.6, Chapter 5.6, Chapter 6.4, Chapter 6.5.
– The possibility to recover from a surge event acting on by pass valves will be also simulated and presented in the context of Hyper facility analysis. Chapter 5.6.3 and Chapter 5.6.4.

The purpose of this thesis is to create a transient model of a PEM fuel cell system, based on Matlab Simulink, as general, flexible and adaptable as possible, in order to be easily set on different type of systems. The object of the study is the development of the simulation tool, and its validation against literature and experimental data. An important aim of the developed dynamic semi-empirical model is to try to adopt a theoretical physics-based approach whenever possible, in order to have an accurate scientific correlation between experimental output and theoretical laws, without neglecting the accuracy that could be provided by empirical equations.
The major work is focused on the fuel cell stack modeling and involves also a large review of literature analysis concerning the simulation of PEM-FCs. In order to guarantee the adaptability of the model, taking inspiration from the latest studies in this field, a differential evolution algorithm is developed to realize the fitting process of the modeled polarization curve, by means of the stack voltage model, on experimental data. This algorithm has a strategic importance for the choice and the setting of the stack voltage equations on the real static performance of the PEM fuel cell system analyzed, with a proved error of about 2-3%. The transient behaviors captured in the model includes flow characteristics, inertia dynamics, lumped-volume manifold filling dynamics, time evolving-homogeneous reactant pressure or mole fraction, membrane humidity and thermal response of fuel cell and cooling system. From one side, the validation against literature data of Section 4 is realized after the development of a general dynamic PEM-FC system model described in Section 2 and 3, comprising all the components normally present in these systems. The comprehensive dynamic model proposed, usually not presented in literature, perform very well respect to the experimental data, comprising the thermic data and the hydration of the membrane, the most important operative parameters but also the most complex ones to simulate.
On the other side, the HI-SEA Joint Laboratory, between Fincantieri S.p.A. and the University of Genoa, allows to study a PEM fuel cell system of 8 stacks sized 33 kW each for a total maximum power of 260 kW. The adaptation and the simplification of the dynamic model to this plant layout helps to study a bigger and more complex PEM-FC system and to validate the model to the experimental data. The simplification of the dynamic model starts form the necessity to set the equations only by the commercially available data, usually limited to the datasheet information. This limitation makes the HI-SEA model less detailed but, at the same time, simpler and able to provide different important results, as the stack and cooling system thermal balances, starting from few easily obtainable data. 

The aim of this work is the assessment of the most suitable hydrogen solution for ship applications and the definition of the role of hydrogen as alternative fuel for shipping. The importance of the “Hydrogen Technologies” for ships comes from the most important social challenge that is driving innovation in the shipping sector: Environmental Challenge. The PhD research project encountered important development both from the industrial and the academic side that brought to the construction of a joint laboratory between Fincantieri and the Polytechnic School of the University of Genoa, the: HI-SEA laboratory, dedicated to the study of fuel cell system for marine application. Moreover the simulation modelling and experimental results developed during the PhD research on the PEM fuel cell and MH hydrogen storage systems, found an application in the nautical sector. The former brought to a patent and the creation of a dedicated start-up company named H2Boat, that was recognised as University spin-off.
The first part of the study define the role of hydrogen as alternative energy vector (fuel) for marine application, analysing the complex context in which it is supposed to be used. In part 2.1 a detailed assessment of the characteristics of different alternative fuels have been
conducted. The complexity of work brought to the construction of comparative models, descripted in part 2.2 that have been used to analyse the characteristic of various alternative solution. An analysis of the PEM FCS state of the art is presented in part 2.3 together with the definition of FCS design for marine application in part 2.4. The study of the hydrogen technologies considered also the definition of simulation models of fuel cell systems and metal hydride hydrogen storage system 3.2. The former has also been assessed towards experimental tests, presented in part 3.3. The models have been used to develop larger laboratory, to define correct operative parameters and FCS design.
Finally a number of application developed during the PhD study are proposed in part 4 to show the goal of the research that is still under development.

Concentrated Solar Power (CSP) hybrid gas turbine systems particularly based on the micro Gas Turbines (mGT) will be of great importance in future power infrastructure where energy security, economic feasibility and clean and efficient power generation are the key concerns. Integration of Thermal Energy Storage (TES) in CSP hybrid gas turbine systems could be a viable solution to overcome the intermittent nature of solar power, and increase the dispatchability. Based on this perception, a comprehensive analysis of both mGT cycle and TES technology should be undertaken, in order to achieve a better understanding of the behavior of TES and its interaction with other components in a hybrid gas turbine system. The present work intends to contribute to this analysis through mGT and TES system modeling and testing. This thesis is framed in two main parts: first part deals with T100 mGT modeling and second part focuses on the study of thermal storage systems. Regarding TES, detailed dynamic analysis of sensible heat storage is provided, while a preliminary study of thermochemical storage is conducted.
The mGT performance diagnosis involves the model development for steady-state simulation of T100, model validation, and application in real operating conditions at the Ansaldo Energia AE-T100 test rig. Furthermore, diagnostic application of the AE-T100 model for whole mGT cycle is discussed with the help of two case studies at AE-T100 test rig. AE-T100 model has also been applied in the real operating conditions of micro Humid Air Turbine (mHAT) system located at Vrije Universiteit Brussel (VUB), to highlight the modeling capability of AE-T100 tool as well as monitoring the recuperator performance in the VUB-mHAT cycle.
The second part of this work concerns the dynamic modeling and experimental validation of a sensible TES system at laboratory scale, which is part of the Hybrid Solar Gas Turbine (HSGT) system developed at the University of Genova. TES is modeled with the help of a two-dimensional CFD model based on the ANSYS-FLUENT code, and a one-dimensional TRANSEO model employing software designed by the Thermochemical Power Group (TPG) at the University of Genova. The experimental validation, modeling capability to present the actual thermal stratification and State of Charge (SoC) of the TES, and scope of each model are also discussed. This study also highlighted the potential of TES system based on the monolithic structures for hybrid gas turbine systems i.e. low pressure drop across the TES which are acceptable for the whole gas turbine hybrid system, modular structure of the storage and very low thermal losses.
In addition to the sensible heat storage system, ThermoChemical Storage (TCS) based on the redox cycle of cobalt oxides pair Co3O4\CoO was finally studied by the candidate during research period at Zhejiang University, China. The mathematical model which has been developed in MATLAB is based on the mass and energy conservation and reaction kinetics of the redox cycle, and has been validated against the experimental data available from literature. This work was aimed to simulate the process of thermochemical storage with less computational effort and establish a model which can be integrated with the plant components in the whole system dynamic models for CSP hybrid systems. This numerical model will help in design and optimization of the actual TCS system.
Overall this work aims to give as much as possible design parameters and performances analysis to facilitate designing and optimization of the HSGT plants coupled with TES systems. However, this thesis is limited by the analysis of the complete HSGT system integrated with TES, as it analyzed the T100 cycle and TES systems independently. Hence, the knowledge and modelling capabilities developed for mGT cycle and TES systems in this study will be merged to develop a single simulation tool for mGT based CSP hybrid systems, in the future. This simulation tool will be useful to analyze the performance, at design, off-design and transient conditions of the mGT-based CSP hybrid systems.

The aim of this thesis is investigating the role of digital controls in the modern energetic systems through the implementation of monitoring and control application in different scenarios; industrial, civil, basic and applied research. All the monitoring and control systems were entirely developed, tested and installed during the three-year PhD research.
The different applications are the following:
• monitoring system of the emissions of a biomass boiler
• control system of a hydropower plant
• monitoring and control system of an innovative Wave Energy System (WEC)
• monitoring system of level set
For every application there is a specific corresponding technological environment, for example the control system of the hydropower plant has been integrated in the complex Distributed Control System (DCS) of the plant itself. In reverse, the core of the monitoring and control system of the innovative WEC and the monitoring system of level set is a simple and cheap microcontroller.
The different researches are inserted in national and international research project including Biomass+ funded by “Maritime Italy-France Cross-Border Cooperation Program 2007-2013”, Seaspoon: verso la nuova energia del mare 2 funded by European Regional Development Fund (ERDF) and New Approach to the Optimal Control of Level Sets Generated by Partial Equations to Bridge the Gap Between Computational Mathematics and Control of Complex Systems funded by Air Force Office of Scientific Research (AFOSR).

In the present thesis, innovative energy storage systems by chemicals production employing renewable sources is investigated from the thermo-economic, enviromental and energetic perspective. Two different technologies are analyzed and compared: the “power to gas” technology and the “power to liquid” technology. Three chemicals products (two gas and one liquid) are studied as potential energy storage: hydrogen, hydro methane and methanol. The power to gas plant for hydrogen production consists of a water electrolyser and a compression system neccesary to bring the gas pression up to the transportation pressure. The hydromethane is produced in a Sabatier reactor from hydrogen and carbon dioxide; also in this case, it is neccesary to pressurize the gas for the transportation and distribution grid. A carbon capture plant installation is required in order to obtain the CO2 necessary to the reaction; as alternative the CO2 can be purchased at market price. Finally, the power to liquid system for methanol production is similar to the hydromethane? plant, consisting of a water electrolyser , a carbon capture plant and a reactor for metahnol synthesis. The methaol is produced by the reaction of H2 and CO2 in 1:3 molar ratio, and, being the methanol in liquid from at ambient temperaure, a compression system is not required. The thermo-economic study has been carried out using the W-ECoMP software developed by the Thermochemical Power Group for the time-dependent analysis of innovative plants. Due to the modular structure of the software, it is possible to easily upgrade pre-existing modules and implement new modules necessary to simulate the different systems. The analysis is focused on the European energetic scenario. In particular, the Italian and German economic scenarios are taken into account. These countries currently present the largest amount of renewable installed capacity in Europe. For each scenario, the analyses are performed varying parametrically different economic parameters (electricity price, product price, capital cost, etc…) in order to evaluate their influence on the economic viability of the system and define the best solution from the energetic, economic and enviromental point of view.

As the demand for energy is increasing worldwide, not only the security of energy supply and the stability of prices, but also climate change has become an important issue. The production of energy with renewable and not controllable sources like wind and solar plants has increased in a very significant way in the last 10 years. It represent a benefit from the environmental point of view, but the critical aspect of the integration of these technologies with the traditional power plant has to be faced. In fact, the renewable power plants have the priority to insert the energy produced in the national grid while the plants that use fossil fuels to produce energy has to adapt their power to satisfy the demand working at partial load and shutting down very often causing a decreasing of their efficiency. The aim of this thesis is to face this problem using the surplus of energy produced from renewable plants to generate hydrogen, which represents a chemical energy storage. Then, it will be used in a methanation reactor, reacting with the carbon dioxide, and hydro-methane (a mixing of H2 and CH4, where the maximum H2 content in volume is equal to 30%). The production of synthetic natural gas (SNG) via thermochemical conversion of biomass and subsequent methanation could be one route to address these issues. The advantages are the high conversion efficiency, the already existing gas distribution infrastructure, the wellestablished and efficient end-use technologies and the recovery of a concentrated CO2 stream without any additional cost and thus the possibility for an easy carbon capture and sequestration. Already in the 1960s, the need for the production of synthetic natural gas arose to fulfill the increasing demand of natural gas. In the following 20 years, different methanation processes were developed from coal to synthetic natural gas. However, only one commercial plant was erected in 1984 and has been producing SNG ever since. First, an overview of the motivations of this work and the existing technologies, like electrolyzer and catalytic and biological methanator was carried out to exanimate their characteristics and the possible markets where the hydro-methane can be employed. In the latter, a three-dimensional computational fluid dynamic (CFD) model of the reactor was developed, including the mixer section where the reactant are mixed before the inlet of the catalyzed section, the heat exchanger used to remove the heat produced by the methanation reaction, and the methanation reactor itself where the Sabatier chemical reaction takes place. It was validated against experimental data from the plant built in the laboratory of the Thermochemical Power Group (TPG) located in Savona campus university. The model was also utilized to improve the design of the reactor without making an huge amount of test saving money. It is worth to notice that the idea was to study the reactor working with two different cathalyzers, but the test and all the simulations was made only wirh the 3% Ruthenium on alumina cathalyzer, while the 20% Nickel on alumina was not tested until now. In future works also the second cathalyzer will be tested and other tests will be carried out to verify the simulations results. 

Solid oxide fuel cells (SOFCs) represent a promising future power generation technology, due to their extremely high efficiency, low emissions, fuel flexibility, reduced noise, and available high-grade heat that can be recovered for co-generation purposes or to produce additional electricity. When a pressurized SOFC is coupled with gas turbines in a direct cycle, the resulting hybrid system can achieve improved efficiency and increase system turn-down. The significant challenges that arise in terms of system controllability are the current limit to commercialization and drive the research on hybrid systems.
A potential advantage of SOFC/gas turbine hybridization is to operate in a range where high system efficiency and low cell degradation co-exist. High temperature fuel cells are subjected to performance degradation over time due to several mechanisms. The nature of these mechanisms is not completely understood at this time and is object of intense research. It is nevertheless clear that some operating parameters can accelerate degradation, such as high current density and high fuel utilization, which are usually necessary to ensure good efficiency. In this work, the aim was to evaluate the feasibility of operating a hybrid system so that fuel cell useful lifetime could be extended while maintaining high system performance.
A cyber-physical SOFC emulator coupled with a physical recuperated gas turbine was employed. An empirical model of degradation was developed and incorporated into the “cyber” component of the SOFC, which consisted of a numerical, real-time, one-dimensional model. The aim of maintaining real-time performance, required for hardware-based simulations, was accomplished by focusing on the effect of operating parameters on voltage degradation rather than physical degradation mechanisms. Those same operating parameters can be monitored and manipulated to reduce performance deterioration. Different operating strategies were analyzed to determine the effect on fuel cell lifetime, system performance over time, and system economic viability.
A control architecture was then designed and implemented on the physical system to test the chosen operating strategy. In the cyber-physical approach used for the test, the numerical SOFC model simulated the fuel cell electrochemistry and performance and drove the only heat source of the physical system, while the hardware components of the SOFC emulator represented the dynamic interaction (in terms of flow, pressure and heat) with gas turbine and heat exchangers. Hence, the fluid-dynamic of the gas turbine cycle was accurately captured and the integration with an SOFC stack could be evaluated. Moreover, a cursory economic analysis was performed to quantify the economic benefit in extending the fuel cell lifetime and increase system flexibility. Impact of degradation model uncertainty on lifetime and economic performance was evaluated with a response sensitivity method. With respect to state of the art knowledge, great opportunities were discovered and described for SOFC applications in gas turbine based hybrid systems.

The aim of this work is investigating the role of energy control and optimization strategies in energy district equipped with different generators, starting from the analysis of a real demonstrator: the test rig installed in the Innovative Energy Systems Laboratory (IESL) inside the University of Genova Campus located in Savona. A scientific, technological, and economic study has been carried out looking at the impact of distributed generation in the present European energy scenario. The target of this three year PhD research is to show different application of optimization and control techniques, comparing them to highlight the advantages of everyone starting from a real test case (the IESL), trying to solve issues and trying to imagine future challenges of distributed generation which is recognized as the key element of the modern energy scenario where fossil and renewable sources, traditional and polygenerative generators, suppliers and purchaser actors coexist and perform action in order to provide energy, minimize costs and reduce emissions. The work presented is complex and articulate, as it consists of many different activities. In fact, it was necessary to create a software interface to check the system under test and study the external control system with which to communicate. Secondly, it has been realized a long process of development and testing of different optimizers, compared both with simulated load curves, both with real loads. The testing work has allowed a qualitative analysis of models developed, highlighting the advantages and disadvantages of the various approaches. 

In many engineering design and optimisation problems, the presence of uncertainty in data and parameters is a central and critical issue. The analysis and design of advanced complex energy systems is generally performed starting from a single operating condition and assuming a series of design and operating parameters as fixed values. However, many of the variables on which the design is based are subject to uncertainty because they are not determinable with an adequate precision and they can affect both performance and cost. Uncertainties stem naturally from our limitations in measurements, predictions and manufacturing, and we can say that any system used in engineering is subject to some degree of uncertainty. Different fields of engineering use different ways to describe this uncertainty and adopt a variety of techniques to approach the problem. The past decade has seen a significant growth of research and development in uncertainty quantification methods to analyse the propagation of uncertain inputs through the systems. One of the main challenges in this field are identifying sources of uncertainty that potentially affect the outcomes and the efficiency in propagating these uncertainties from the sources to the quantities of interest, especially when there are many sources of uncertainties. Hence, the level of rigor in uncertainty analysis depends on the quality of uncertainty quantification method. The main obstacle of this analysis is often the computational effort, because the representative model is typically highly non-linear and complex. Therefore, it is necessary to have a robust tool that can perform the uncertainty propagation through a non-intrusive approach with as few evaluations as possible.
Hence, the primary goal of this work is to analyse and compare different uncertainty quantification approaches to understand strength and weakness of them and find the best method for application to industrial problems. Starting from the sampling methods like Monte Carlo and/or ANOVA, different approximated methods were widely described. In particular, the response sensitivity analysis, the non-intrusive polynomial chaos approach, the mid-range approximated method and the application of response surface were taken into account. All of these approaches are applied to a set of analytical test functions and engineering optimization problems to evidence strength and weakness. From these analyses, it arises that the generalized Polynomial Chaos is a good methods to apply for uncertainty quantification, especially if the least square approximation is considered. On the other hand, mid-range approximation method is not always suitable, if the optimization problem is highly constrained. In a second step, two different methodologies were applied to two different real test cases. In the first one, the least square approximation approach was applied to a turbomachinery test case, based on a thermal analysis of a high pressure turbine disk. The second one is an optimization problem design based on a heat exchanger comparing Response Sensitivity Analysis and Monte Carlo. Hence, form the results obtained in all the test cases analyzed, the generalized Polynomial Chaos seems to be the optimal methodology in terms of accuracy, however the Response Sensitivity Analysis gives good results with a significant reduction in computational time, that is an important aspect when high-complex system are analyzed. For this reason, this method was then applied to three different industrial applications:

• analysis of a thermo-economic behaviour of a micro-gas turbine through design of experiment and comparison with RSA

• study of a steady-state fuel cell hybrid system model taking into consideration the stochasticity of both design parameters and degradation model coefficients

• study of the surge problem with a dynamic approach

This study has demonstrated the effectiveness of analysing this kind of plant with a stochastic approach. In such a way, it is possible to have a better understanding of the real operation and performance of such systems. Response Sensitivity Analysis method results one of the best to study this kind of problems. In fact, the main advantage of Response Sensitivity Analysis is that it provides the possibility of treating problems with high level of uncertainty linked to their internal variables, in particular:

• it is an approximate method so that the uncertainty analysis results generated must be evaluated by comparisons with the Monte Carlo which can provide exact uncertainty results

• it shows an important saving in computational time if compared to both traditional stochastic analysis like Monte Carlo simulation and other approximated method like Polynomial Chaos, keeping good accuracy

• it can be used to investigate the direct effect of a single uncertainty on final results, in order to know what are the most influent inputs and what are the negligible ones

In this research, the Response Sensitivity Analysis method shows high accuracy when compared to the Monte Carlo, showing a difference lower than 10%. Therefore, it is a useful approach for uncertainty analysis in advanced energy system problems.

As distributed systems arise as the main and future approach in energy production, new and time-effective methods to study global configuration of small scale generation systems have to be investigated. Even though simplicity of simulation tools are improved nowadays, software is still used by high-qualified employees. This is true also in the specific field of power generation. Still, the need of specific and high-fidelity models will increase in next years. In this perspective, a dynamic model that can be used as decision support for plant operators is of interest. This work presents an innovative hybrid modelling approach to describe dynamic behavior of Solid Oxide Fuel Cell (SOFC) Gas Turbine (GT) Hybrid System. These systems are going to represent the state of the art for energy generation in a near future. Nevertheless, this PhD research started with considering Gas Turbine Combined Cycle (GTCC) framework for modeling, since GTCC represent the current best available technology in terms of performance and emissions. In particular, a dynamic simplified approach to model a Heat Recovery Steam Generator of a Gas Turbine Combined Cycle is proposed, together with its validation against field data. The initial work pursued on GTCC modeling framework has been essential to lead to a similar formulation for SOFC/GT. The approach is based on some physics-based equations and load-variable time constant supported with relations derived from system identification. To enhance applicability of the proposed tool, the modeling approach is purposely developed and run within the Microsoft Excel/Visual Basic
environment.

The description of the adopted framework begins with some physical considerations on global HRSG structure, and then focuses on a specific application for a real plant, i.e. a 390 MW multi-shaft combined cycle based on the AEN94.3 A4 frame. The model embodies some parameters, which are easily derived from historical data to enhance the forecasting capabilities of the software, resulting in a hybrid model, which covers a high range of working conditions. The whole model is designed to run in Excel/Visual Basic environment to extend use of simulation software to those people, who have limited experience in advanced modelling software. The model has been trained on 10 days of experimental data, in order to create the basis for accurate predictions. Therefore, the feasibility of this approach has been verified using some Gas Turbine load profiles accomplished in everyday working operations and validating the results against field data. Particular importance is given on start-up procedure and associated mechanical stresses acting on the steam turbine (ST). The currently deregulated energy market led GTCC to undergo frequent startups, a condition often not considered during plant design. Moreover, the time required for the start-up is crucial under an economical viewpoint, though it is constrained by mechanical stresses imposed to thick components by thermal gradients. Simplicity of model allows fast computation. The obtained tool can be than adopted to support decision process during plant operations. The aforementioned approach has been defined to capture dynamic behavior of SOFC-GT hybrid systems in normal operating conditions, since operating SOFC/GT will constitute the base-load of future power generation. In this case, parameters of GT are considered as well as stack performance indicators. Temperature of the solid stack – one of the main important constraints in SOFC/GT management – is determined through an empirical based function, and it can be deduced by temperature of gas leaving the stack through a thermal scalar factor. The model is then a representation of the thermal behaviour of the stack. The purpose is to build the temperature profile of the stack by using only measurable variables. Main advantage of this hybrid modeling approach regards system identification, which can be pursued through experimental measurements coming from normal operating conditions of the plant. In other words, no specific operating conditions (i.e. step responses, power ramps) are required. The development, implementation and validation of the aforementioned approach has been carried out for three real fuel cell gas turbine emulator test-rigs installed at University of Genoa (UNIGE, Italy), German Aerospace Center (DLR, Germany) and National Energy Technology Laboratory (NETL, USA), respectively. The three target systems are introduced and dynamic situations analyzed for model tuning, first, and validation, then. Relevance is given to peculiar transients where the model shows its reliability and its weakness.

Assumptions introduced during model definition for the three different test-rigs are discussed and compared. The model captured significant dynamic behavior in all analyzed systems and showed influence of signal noise on some of the SOFC computed outputs.
Finally, a possible development for implementation and application of developed modeling framework in control system environment is proposed through consideration of Model Predictive Control (MPC). The use of MPC in advanced power systems can be advantageous in controlling highly coupled variables and optimizing system operations. In this case, a cluster of SOFC/GT hybrids are taken into consideration by a supervisory controller. This manages load distribution among generation units, which are characterized by different degradation paths of the fuel cell stacks. When implementing a MPC, a critical aspect is the trade-off between model accuracy and simplicity, the latter related to a fast computational time. This is the motivation that led to consider the model developed for this research for implementation in such architecture. The reduced number of states in the model and the simple framework allow real-time performance and potential extension to a wide range of power plants for industrial application.

 The aim of this thesis is investigating the role of energy storage in energy district equipped with different polygenerative generators, starting from the analysis of a real demonstrator: the Smart Polygenerative Microgrid (SPM) installed in the University of Genova Campus located in Savona A scientific, technological, normative and economic study has been carried out studying the impact of storage in the present European energy scenario particularly through a thermeconomic approach and a new methodology to evaluate the best operating strategy of the storage combined with cogenerative and renewable generators was proposed. An entire year thermoeconomic analysis has been performed employing a software designed by the Thermochemical Power Group (TPG) at the University of Genova for the time-dependent analysis of energy systems, starting from the real demand data of the campus. The whole research is inserted in the FP7 Research Project named RESILIENT (coupling REnewable, Storage and ICTs, for Low carbon Intelligent ENergy managemenT at district level) and the same analysis operated in Savona was replicated in the other two Demosites of the Project located in Hasselt (Belgium) and Ebbw Vale (Wales), showing how different energetic and economic scenarios can affect the thermoeconomic optimum. According to this preliminary research the crucial role of thermal storage in polygenerative districts was discovered, so thermal storage facilities were studied from an experimental point of view too. A Ceramic High temperature storage test bench for gas turbine – Concentrating Solar Power plants was built and studied. Another material for high temperature storage for CSP application was studied by the candidate during his visiting research period at the Concentrating Solar Power Research group at the Energy Department of the Royal Institute of Technology (KTH) in Stockholm. The target of this three year PhD research is to show different application of storage technologies, both electrical and thermal, starting from a real test case (the Savona Campus SPM) and trying to imagine future challenges and technologies of storage which are recognized as the key element of the modern energy scenario where consumers and producers actors, fossil and renewable sources, traditional and polygenerative generators coexist and act in order to maximize the profitability, reduce emissions and guarantee energy.

In this work two different SOFC hybrid systems were taken into account and analysed. The first one is the Dry-Cycle hybrid system that RRFCS has been developing for years. The second system was proposed by TPG and is studied using models support and an emulator test rig, too.
The RRFCS system was modelled in Matlab Simulink environment using TRANSEO tool and its transient behaviour was analysed in different scenarios. During simulations, the control system and control strategies were upgraded and improved in order to keep the hybrid system within the acceptable operational envelope, complying with safety constraints, during the different transients under investigation.
At the same time, TRANSEO models were validated using experimental data from tests on the hybrid system emulator.
The TPG hybrid system was studied with different purposes. Its transient behaviour was not analyzed so deeply as made for the RRFCS system, it would have been only a repetition. The aim of the work made on this second hybrid system was to develop and analyze advanced control systems. Then, a Model Predictive Control (MPC) method was chosen to be used. Matlab MPC Toolbox was used to develop different control systems.
In the first part of this work on advanced control systems, two MIMO (multi-input multi-output) controllers were obtained: fuel cell power and cathode inlet temperature are the controlled variables; fuel cell by-pass flow, current and fuel mass flow rate (the utilization factor kept constant) are the manipulated variables. In the first case, MPC has direct access to the plant manipulated variables, in the second case MPC operates on the setpoints of PIDs which control the plant. The first MPC based control system was experimentally tested in HIL emulator against the original PID method. The results confirmed that the MPC based control system is better than the PID in limiting the FC inlet temperature maximum variations from the setpoint during fuel cell power ramps.
Then different MPC control systems (MPC 1, MPC 2, MPC 3 and MPC 4) were developed to limit the effect of wide ambient temperature variations on hybrid systems transient behaviour.

The aim of this doctoral thesis is the development of a diagnostic tool for energy systems equipped with a stack made up by solid oxide fuel cells (SOFC). In order to reach this target the work has been developed relying on the performances of a system model implemented in Matlab/Simulink®. This model was validated widely against experimental data provided within the collaborative European project GENIUS, which aimed at developing a GEneric diagNosis InstrUment for SOFC Systems.
Such a model was exploited in different ways. It was arranged in order to simulate different control strategies, i.e. either constant stack current or constant stack voltage. Then it was tuned on different experimental data in order to investigate the impact of the fuel cell geometry on system performances.
Initially, it was run to lay the basis for a monitoring and diagnosis tool for SOFC systems, which worked through an explicit classification procedure. After this first analysis, the model was used to generate a wide dataset regarding faulty operative points. As a matter of fact, fuel cells are relatively new devices and they are really expensive. Therefore there is a lack of availability of data regarding unsafe working conditions. Validated physical models could allow for the generation of
this kind of data in order to draw maps representing the behavior of the system under faulty conditions. These maps allowed for the development of pattern recognition based diagnostic tool.
Furthermore, faults simulation results will help to understand better the behaviour of this kind of
systems and to find the best signals to monitor in order to perform an efficient diagnostic procedure.
This doctoral thesis is organized as follows:
• Chapter 2, Fault detection and diagnosis of energy systems: where a brief overview of the state of art for energy systems diagnostic strategies is presented.
• Chapter 3, Modeling analysis: where the system model, which is the core of this doctoral thesis, is presented. The modeling effort and the complete model validation is reported
• Chapter 4, Fault simulation: where the results of fault simulation analysis are collected and discussed.
• Chapter 5, Supported Vector Machine fault diagnosis in SOFC systems: where an innovative approach for fault detection and identification in solid oxide fuel cell system,
based on supervised classification, is presented and discussed.
• Chapter 6, Conclusions and future development: where the whole work is summed up and possible future development is presented.

The following PhD thesis deals with diagnostics of energy systems with the main focus on data validation and reconciliation of field measurements. The DVR (Data Validation and Reconciliation) step is fundamental and necessary to carry out the further diagnostic phase with an high accuracy level. The DR problem is in general a nonlinear constrained problem but since the objective function is quadratic and the starting point of the variables (the measurements values) is usually very close to the final estimates (the reconciled values). From these considerations, a simplified approach to the Data Reconciliation problem based on the least square optimization has been developed. The Ansaldo Energia heavy-duty gas turbine operating in a combined cycle power plant has been considered as a first application of DR and GED techniques to a large scale power plant. A comparison among this new technique and the others already present in literature, in terms of accuracy and computational time, has been carried out. The final estimates are as accurate as the reconciled values obtained with the Sequential Quadratic Programming, one of the most powerful DR technique, both in linear and in nonlinear case. The most important result is in terms of computational time where the new technique has shown to be ten times faster than the others. This point is very important in case of more complex problems with a lot of unknowns and equations and above all in case of online applications. Then the GED (serial elimination) has been applied. First of all, some gross errors of different magnitudes have been imposed to some measured variables, in order to understand the minimum threshold for their correct localization and so the method sensitivity. The technique has shown to be able to detect this type of errors especially for some variables characterized by large uncertainty and with a small relationship with other measurements. Secondly historical field measurements coming from the plant DCS have been considered in order to find some gross errors that really affect measurements. A gross error in the outlet compressor temperature has been correctly identified and localized. In particular, it has been demonstrated that the common practice of using the mean of redundant sensors tends to mask the gross error in just one sensor. After that a GED sensitivity analisys has been carried out focusing on the study of the gross error detection technique limits and its behavior within the smearing zone. In this study a bias error of different magnitudes to the variables that most influence the GT has been imposed, with the aim to understand if a non high intensity error could be detected anyway. An approach using a comparison between the percentage corrections of reconciled values without gross error and reconciled values coming from serial elimination has been useful in finding a good detection criterion inside the smearing zone. Finally a monitoring strategy for HRSG based on data reconciliation was presented and tested. First of all, the relation between ambient temperature and HRSG efficiency was highlighted. A DR problem was set implementing energy balance equation around GT and HRSG control volumes respectively. The estimation of the HRSG heat losses was made using a correlation on the total energy entering the HRSG. The correlation was found in good agreement with the experimental data and suggests the presence of just random error from the measurements and so the DR applicability. It was proven that DR is a good instrument to filter data by outlier elimination, make data consistent and reduce the measurement uncertainties. The effects of these enhancements were shown on the target indicator: the HRSG efficiency.

The goal for future power generation systems, based on high efficiency and low emissions, has led the cycle innovation community to focus an extraordinary effort on the development of new advanced technologies which are able to guarantee a clean world. System integration is a necessary task which allows studying the coupling of these innovative system configurations, specifically their flexibility and their performance, in all stages of operation. Studies carried out currently, in simulation environments, in small experimental rigs and in hardware-in-the-loop-simulations (HiLS), which are able to match experimental rig and model simulations, have been of considerable importance to scope the transient behavior of these technologies and find suitable strategies to cope with the related challenges.
The “Hyper Project” at the National Energy Technology Laboratory, U.S. Department of Energy (NETL), located in Morgantown, WV is designed for studying the system integration of a complex hybrid system composed of a gasifier, solide oxide fuel cell and gas turbine generator, sometimes called an Integrated Gasification Fuel Cell system (IGFC). A hardware-in-the-loop-simulation (HiLS) procedure for a direct-fired fuel cell turbine hybrid power system has been designed to explore dynamic operation of hybrid systems and quantitatively characterize their transient behaviour. It is possible to model, test and evaluate the effects of different parameters on the design and operation of a gasifier/fuel cell/gas turbine hybrid system and quantify risk mitigation strategies.
The first implementation of emergency shut-down control strategies was based on the standard double block and bleed strategy used by industry during emergency failure. However, this strategy resulted in turbomachinery hardware failure. The primary linking event in these cases was compressor stall and surge resulting from the sudden loss of fuel, coupled with the high volume of compressed air between the compressor and turbine due to the presence of the fuel cell cathode. A new mitigation strategy involving automated ramps is proposed and described in detail to control the system from start-up to forced emergency shut-down. In addition, the control architecture shows how the virtual fuel cell model can be coupled to the real gas turbine safely, in all stages of operation. Improvements have been included to the emergency shutdown procedure, failure analyses, and the comparison of experimental data with previous results.
The direct-fired fuel cell model connected to a turbine hardware system is a very reliable technique which allows researchers to monitor the transient performance between the fuel cell and gas turbine. The big advantage of this configuration is that a more detailed model is used to predict the electrochemical transient dynamics, reducing the computational effort compared to a dynamic model of the entire system. This is an important objective, in particular related to the control of the turbine speed, when the fuel cell model drives the turbine. Since NETL is studying ultra-high efficient systems for large-size power generation, in such a context IGFC represent the unique choice to overcome 60% HHV efficiency for coal fired plants. In such a respect, this thesis focuses on large-size hybrid systems, characterized by constant rotational speed which generally avoids generator trips and the resulting shutdown caused by turbine speed variation. In addition, the cathode mass flow must be carefully controlled as well, because otherwise changes in the waste heat coming out from the fuel cell stack can also affect the turbine speed. A comparison between a decentralized control approach and a state-space feedback control, based on “classic” and “modern” pole placement technique, has been performed to tune two PIs and a MIMO controller, which are able to control the turbine speed and the cathode mass flow simultaneously. This study applies to any advanced power system where coupling between one actuator and several outputs is shown. In this specific hybrid configuration, when the turbine speed changes, the cathode mass flow changes, and when the cathode mass flow is controlled by the cold-air bypass, a strong coupling is evident on turbine speed. An experimental evaluation has been performed in the dissertation to analyze if the “classic” method has any disadvantages compared to the “modern” method, and if the maximum deviation allowed from nominal turbine operation and cathode mass flow is exceeded.

This Ph.D. thesis deals with technology development of the Seaspoon, a new energy converter of ocean/sea waves, for which patent is pending.
The potential of renewable resource is enormous and well distributed all over the globe, although latitudes between 40° and 60° in both the hemispheres show the highest energy content with average values up to 60 kW/m of wave front: regions with extended and populated coasts, and milder wave climate could take advantage exploiting such a resource.
The nature of the gravitational waves phenomenon, and the energy flux associated to it showed some interesting analogy with the wind energy sector technologies: the Savonius rotor is a robust, reliable and economical turbine able to perform well in whirly flow with extended operative conditions even though with a modest efficiency of conversion. The nature of the fluid flows induced by the orbital motion of water particles associated to sea waves allows the implementation of a passive component theoretically able to more than double the conversion efficiency of a Savonius rotor operating under the sea surface with its axis displaced horizontally and parallel to the incoming wave front.
Such a component, called Seaspoon, has been conceived in this thesis, which started the investigation both from the theoretical and experimental points of view, in order to achieve a proof-of-concept and start quantifying the performance.
The iSeaspoon has been first represented in a numerical model in the MATLAB environment, in order to fully describe and simulate the its operating conditions.
In order to validate the results of the numerical simulation and be able to later measure the conversion performance of the Seaspoon device:
– a wave flume has been designed and built at the Savona University campus;
– a physical model of the Seaspoon device has been realized according to the
geometric features supplied by the numerical simulation;
– wave train with parameters of wave height and period equivalent to the MATLAB
simulations have been generated
– an experimental campaign aimed to verify the seaspoon concept has been carried
out
The laboratory results have proved the concept, and plans for designing and building a demonstrator to be operated in open sea are ongoing.

This thesis deals with the monitoring and diagnosis of energy systems, large size power plant, as Combined Cycle Power Plant (CCPP), and premarket domestic size generators as the SOFC systems. Part of the research was carried out at Siemens Power Diagnostic Center within the Operation Support group. The main activities focus on study and test of diagnostic rules for the bottoming cycle components. Meanwhile performance calculations on components were performed, in particular with respect to the maintenance operation impact (e.g. compressor off-line washing, retuning of controlled parameter).
An analysis of preprocessing phase for field measurement validation was made. Two approaches were developed: one exploiting mass and energy balances and another using a statistical methodology basing on Principal Component Analysis (PCA). A feature selection algorithm was developed to enhance the sensitivity of the PCA based data validation.
Meanwhile in the framework of the European project Genius (GEneric diagNosis InstrUment for SOFC Systems). A tool for the monitoring and diagnosis of SOFC based systems was developed. After a validation and simulation, the algorithm was tested on line on the Galileo system (Hexis) at Eifer laboratory.

The aim of this thesis is investigating hydrogen production from large size renewable sources, considering in particular the hydroelectric facility of Itaipu (14,000 MW), located on the border line between Paraguay and Brazil.
A scientific, technological and economic study has been carried out, investigating the possibility of generating hydrogen by water electrolysis, mainly employing the so called spilled hydraulic energy, not converted into electricity since lack of users, in particular in Paraguay.
Different hydrogen generation and storage systems have been analyzed, investigating also a methodology to convert hydrogen into hydro-methane, easier to be employed in land transportation in a short to mid-term scenario. Hydro-methane is synthesized mixing hydrogen and CO2, sequestrated in CCS plants: since their absence in the Paraguayan scenario, an innovative method to produce bio-hydro-methane has been investigated, also employing the large amounts of oxygen co- produced in electrolysis process (which would be otherwise largely vent to atmosphere) for biomass gasification.
For the different plant lay-outs, a thermo-economic analysis has been performed employing two different software, one for the design point and one for the time-dependent one entire year optimization, since spilled energy used to feed electrolyzers is strongly variable throughout the year.
Optimal sizes for hydrogen, hydro-methane and bio-hydro-methane generation plants have been determined, taking into account several economic scenarios, representative of Paraguay and Europe, and investigating the influence of electricity cost on economic results.

The SOFC are very complex energy generators once they involve thermal, fluidic and electrochemical phenomena. Because of the complexity they need a set of auxiliary elements (BOP) as valves, blowers, transducers, etc. which are vulnerable to faults that can cause the stop or the permanent damage of such expensive device. To guarantee the safe operation of the SOFC systems, it is necessary to use systematic techniques, to protect the whole system or to bring the plant back to a working operation point. To accomplish this task a fault diagnosis system operating in real-time is necessary. The diagnosis system should not only allow the fault detection but the fault isolation too. In this work, a model-based SOFC fault diagnosis tool is proposed. The model-based fault diagnosis is based on comparing on-line the real behaviour of the monitored systems obtained by means of sensors and transducers with a developed model which represents the real system. In the case that a discrepancy is detected between the model outputs and the instrumentation measurements, a fault is detected. If a set of measurements is available, it is possible to generate a set of residuals (indicators) that present a different set of possible faults and thus, it is possible, in some cases, to isolate the fault. Once the set of faulty measurements are not easily found in literature, mainly because causing disturbances to a real system could be very expensive leading to the failure and damage of such very expensive devices, a new method of creation of the residual set is proposed. The innovation of this work is therefore, the residuals generation which is produced by comparing the outputs of the validated physical model against the same model running under some predefined faults. Such algorithm uses only normal measurements of process values and control input parameters. The aim of the residuals generation is the creation of a fault signature for all the simulated faults to be used later in a diagnostic tool for fault identification and for fault isolation.

This work presents a complete development, implementation and validation of a hardware-in-the-loop simulation facility for hybrid systems based on the coupling of a micro-gasturbines and a solid oxide fuel cell.
The work starts with the development of transient models for innovative energy systems in order to acquire experience and to setup robust algorithms for simulating energy systems and components. The second step is the transition from transient to full real-time modelling and this is made by exploiting real-time components, previously developed at TPG-University of Genova, to set up a complete network representing a hybrid system power plant. The Rolls-Royce Fuel Cell Systems Ltd (RRFCS) 1MW power plant has been the first system simulated with such a network of real-time components. The model has been validated against a large set of experimental data provided by RRFCS.
The last achievement is the coupling of the real-time model of a 380 kWe SOFC with a physical emulator test rig for hybrid cycles. The test rig is equipped with a 100 kWe Turbec T100 gasturbine, an anode loop with ejector, a cathode vessel and a recuperator. Using such a hardware-in-the-loop (HIL) implementation two controllers have been developed and tested, one for the cell power and one for keeping the stack temperature as constant.
The HIL layout proved to be stable and reliable and the two controllers proved to be prompt, in relation to the system dynamic, and stable.

Much effort and resources are being directed to reduce emissions in the most varied sectors. Combined cycles, combining gas turbine with steam power plant, or other innovative systems for energy production presented in this work are lighter polluting and more efficient than conventional gas turbines. This work is concerned with the study of innovative systems for energy production. The study includes different power sizes: from the distributed generation (500 kWe) up to large size power plants (600 MWe).
In the field of the distributed generation a detailed thermodynamic and thermoeconomic analysis on micro gas turbines is developed. Starting from the design of regenerated MGT cycle, the intercooled-regenerated layout is investigated to further increase the thermodynamic efficiency of the whole system. A detailed thermoeconomic analysis of the systems considered is carried out in order to a have a global comparison with the most used technologies in the field of distributed generation.
In the field of the large size power generation a wide investigation of energy systems for electricity production is conducted from the thermodynamic and the thermoeconomic points of view. The analysis regards an innovative concept for the co-production of electricity and hydrogen. In this work is proposed and studied an hydrogen production and purification plant integrated into an existing steam power plant. Such a framework can favour the inter-exchange of energy streams, mainly in the form of hot water and steam, that reduces the costs of auxiliary equipment. The complete systems proposed can represent an attractive approach to a flexible hydrogen-electricity co-production, which is economically sustainable if compared to a new IGCC installation. The results have been obtained through the WTEMP software, developed in the last 20 years at the Dipartimento di Macchine, Sistemi Energetici e Trasporti of the University of Genoa: WTEMP software has been considerably developed and improved in the field of the present thesis. Furthermore the most of the cost functions implemented into the software has been developed and updated.

The aim of this work was to study the sodium borohydride as source and storage method of hydrogen.
The first step was to analyze and compare traditional methods for the production, storage and transportation of hydrogen.
Then the work focused on the sodium borohydride analysis, a promising compound for hydrogen storage and transport. This study was both theoretical and experimental and concerned the hydrogen production by a catalytic reactor and the feasibility of the electrochemical regeneration of this hydride.
As a conclusion, the possible use of hydrogen and sodium borohydride were analysed.
In particular the steps of the work were:
• The study of the traditional methods for hydrogen production, storage and transport;
• The economical analysis and comparison between these traditional methods, taking into account three different scenarios: USA, Europe and Italy;
• The theoretical and experimental analysis about the sodium borohydride;
• The economical analysis of hydrogen from sodium borohydride;
• The comparison between the different safety measures during the transport of hydrogen;
• The complete economical analysis of the hydrogen, taking into account the cost of production, storage, transport and fuelling station in the three scenarios, USA, Europe and Italy;
• The comparison between different vehicles fed up with hydrogen, petrol or sodium borohydride.

Fuel processing is a fundamental step in any fuel cell system fuelled with hydrocarbons.
This work focused on the development of models of the reactors of the Fuel Processor (EFP) unit being developed by Rolls-Royce Fuel Cell Systems Ltd: the SCSO (Selective Catalytic Sulphur Oxidation Reactor) and the CPOx (Catalytic Partial Oxidation Reactor).
They have been developed in the Simulink environment, that is a graphical interface based on the Matlab environment. More precisely, They have been built as a new components of a Simulink external library named TRANSEO, which a comprehensive tool, built in the last years at TPG, and still growing. It is based on the C language and it has been used for previous works on microturbines and energy cycles: accuracy of existing components has already been proved, hence the choice to continue developing it throughout this work.
The main model subject of this thesis has been named “Standard Reacting Pipe” and it is based on a versatile C code: other reactors would be easily added, while only the Selective Catalytic Sulphur Oxidation and the Catalytic Partial Oxidation reactions are implemented now.
Both the dynamic model have been validated against experimental data: attainable results are shown to demonstrate both the accuracy of results and the validity of modelling approach.
The validated models have been used for studying the impact of different natural gas compositions on the systems: different scenarios have been considered and compared.
Finally, the development of a simple PI control system acting on the CPOx reactor outlet temperature to keep it at set-point after a step change in natural gas composition.

The Multi-Purpose Model represents a new methodology for developing model based tools for control system design and verification. The Multi-Purpose Model, as described in this paper, simulates a SOFC hybrid system – a challenging and innovative application of dynamic modelling and control.
Real-time modelling is a recognised approach to monitor advanced systems and to improve control capabilities. Applications of Real-Time (RT) models are commonly used in the automotive and aerospace fields.
Starting from existing TRANSEO components and models, a new approach to fit hybrid system application has been developed. Original C-based models have been translated into embedded Matlab functions for direct use within Matlab-Simulink. The resulting models have then been used to auto-generate C code with the Real-Time Workshop. The C-code has then been compiled to produce application specific executables.
The real-time modeling approach already applied to fuel cell gas turbine systems has here been validated against the experimental data from the micro gas turbine Turbec T100 test rig in Savona, Italy. The real-time model of the microturbine recuperator has been newly developed to fit such an application.
Two representative transient operations have been selected for verification: the heating and cooling phases of the connected volume. The results already show an acceptable agreement with measurements, and they have contributed to a better insight into performance prediction for the entire plant.

Combined cycle power plants are widely diffused all over the world and they represent the main competitor for the steam power plants, due to their highly efficiency (around 60% for last generation plants) and low hazardous emissions (NOx reduction systems, etc.)
A in-depth performance analysis of the above-mentioned plants is very important, both from the point of view of the power plant producer and of the plant end-user.
Most of the world combined cycle power plant producers (Alstom, Ansaldo Energia, General Electric, Rolls-Royce, Siemens, etc.) offer different kinds of monitoring and diagnostics systems, with the purpose of controlling power plant operations and having an on-line tool for observing the behaviour of the most important components: gas turbine, heat recovery steam generator, steam turbine, condenser, electric generator, auxiliaries.
The monitoring activity consists in evaluating continuously the productivity capacity and the efficiency of the plant, using the stream of data coming from plant instrumentation.
This activity is fundamentally different from the so-called “commissioning” or “acceptance” tests, because these are single tests used for verifying that the commissioned plant produces the predicted power, guaranteed by the contract (in the case of commissioning tests) or that the plant is degrading as scheduled (in the case of acceptance tests, in the frame of long term or service agreements). A big difference between monitoring calculations and commissioning or acceptance tests is the instrumentation precision: in the second case very precise, expensive and suitable instruments are used, while in the first one plant instrumentation is used (usually less precise and cheaper).

This PhD thesis is concerned with the development of innovative monitoring and diagnostics techniques for gas turbine cycles, focusing mainly on combined cycle power plants and, as a side application, recuperated microturbine as well. The main topics developed in this thesis are:
• Development of a Bottoming Cycle monitoring and diagnostics tool for large-size combined cycle power plants (in the frame of a collaboration with Ansaldo Energia);
• Thermo-Economics and Diagnostics of combined cycles: the development of a “thermoeconomic” monitoring tool;
• Development of a micro gas turbine monitoring and diagnostic model for the TPG experimental test-rig (located in Savona).
• Development of GT and HRSG/BOP diagnostics indicators, routines and methodologies for real data analysis (in the frame of a collaboration with Siemens AG).
This PhD thesis is organised in eight chapters as follows:
CHAPTER 1 – Introduction
CHAPTER 2 – Diagnostics techniques and combined cycle overview
CHAPTER 3 – Bottoming Cycle Monitoring Model
CHAPTER 4 – Thermoeconomic Monitoring
CHAPTER 5 – Micro-Gas Turbine Monitoring Model
CHAPTER 6 – GT diagnostics: practical applications of performance indicators for real data analysis
CHAPTER 7 – HRSG/BOP diagnostics: suggestion of methodologies and routines
CHAPTER 8 –Conclusions and Future Development

Much effort and resources are being directed to reduce emissions in the most varied sectors. Today, aviation accounts for less than 5% of human related CO_2 emissions, but projections for 2050 indicate that the growth of aviation emissions risks undermining the mitigation progress achieved in other sectors. In aeronautics, increasing fuel efficiency is considered the most effective mean of reducing emissions. Different options are being investigated, from the improvement of the air traffic system, or the adoption of more stringent regulations, to the improvement of aircraft performance. Future airplanes will be more electric and fuel cells represent promising systems to substitute the traditional gas turbine-based Auxiliary Power Units (APUs). Hybrid systems, combining solid oxide fuel cells and gas turbines are lighter than isolated fuel cells and more efficient than conventional gas turbines. This work is concerned with the study of hybrid system design for the application as aeronautical APUs. The state-of-the-art of such a fuel cell auxiliary power unit (FCAPU) was investigated. WTEMP, the software developed by the TPG of University of Genoa, was employed for the design and economic analyse. Physical, weight, and economic models were developed and the design spaces of four fuel cell system configurations were studied. A novel system was proposed (SysS), which is less complex than other hybrid systems proposed in literature, and showed interesting performance characteristics. The design study allowed identifying the n-dimensional space represented by the free variables of the FCAPU design, to identify the constraints of the design space, and to identify the influence of these variables on parameters of interest (e.g. system efficiency and weight). A study of the FCAPU operation in regional jets was performed. The influence of FCAPU design parameters on the economic performance of an airliner was investigated. The investment payback period was used as optimization parameter and optimum fuel cell systems were compared. The influence of the main parameters was analysed, allowing to conclude that a fuel cell weight reduction of 1 kg is equivalent (in terms of ticket price change) to a cost reduction of $3500-$4000. Moreover, fuel cell weights lower than 28 kg/m2 active area represent interesting systems from the environmental aspect, but economic benefits can be obtained only if the weight can be reduced to 7.7 kg/m2 , for the economic scenario under study, or to at least 8.3 kg/m2 , if the CO_2 trading is included in the analysis (30$/tCO2). For economic scenarios with higher jet fuel costs, the breakeven SOFC power density (for economic viability) reduces considerably. The value for the peak jet fuel cost in July 2008 achieved the 0.4 kW/kg (which corresponds to about 12 kg/m2 ).

The aim of this work has been the development of an optimization tool that could be a useful instrument in the evaluation of the designing, sizing and managing of the cogenerative plants. This tool is a software called ECoMP, the acronym for Economic Cogeneration Modular Program. In this name the two main characteristics of the software are integrated: the modular structure and the purpose, that is the economical optimization of the cogenerative plants.
The modular structure allows to develop the code in a useful way, while the software utilization has now become user friendly thanks to a graphical user interface.
In order to include innovative plant modules in the code, the fuel cell hybrid plants have been analyzed both from a static and dynamic point of view.
The hybrid plant analyzed is composed by a Molten Carbonate Fuel Cell coupled with a micro gas turbine.
The basics of any module are the off design performances of a plant and to do this, it has been necessary to study in depth the behaviour of the fuel cell hybrid systems.
To develop a new ECoMP module it is necessary to include in the software the off design performance curves and to obtain these curves, a new model has needed.
The structure of ECoMP is time depending and it is necessary to know the dynamic behaviour of the hybrid plants and his time response to a load variation, to a startup or a shutdown operation, to include them in ECoMP. With this purpose, a transient model has been developed.

The work carried out during these three years regards the analysis and the development of innovative systems for energy production with low CO2 emissions.
The study includes a wide range of power sizes: from the distributed generation (1 MWe) up to large size power plants (600 MWe).
In the field of the distributed generation a detailed thermodynamic analysis on pressurised SOFC hybrid systems is developed: the high efficiency of this technology can permit to reduce considerably the specific CO2 emissions. Starting from the design of Rolls-Royce Fuel Cell Systems Ltd patented system (Agnew et al. 2003), different options are included to further increase the thermodynamic efficiency of the whole system; moreover different options to separate CO2 in pressurised SOFC hybrid system are considered and deeply investigated. A detailed thermoeconomic analysis of the innovative systems considered is carried out in order to a have a global comparison with the most worldwide used technologies in the field of distributed generation (reciprocating engines, small simple cycle gas turbines).
In the field of the large size power generation a wide investigation of energy systems for electricity production is conducted from the thermodynamic and the thermoeconomic points of view. The analysis regards combined cycles, integrated gasifier combined cycles and integrated gasifier solid oxide fuel cell: in all these systems the impact that CO2 separation has on the thermodynamic performance and on the economics is investigated.
An innovative concept for the co-production of electricity and hydrogen is proposed in this work: an hydrogen production and purification plant integrated into an existing steam power plant is studied. Such a framework can favour the inter-exchange of energy streams, mainly in the form of hot water and steam, that reduces the costs of auxiliary equipment. The complete systems proposed can represent an attractive approach to a flexible hydrogen-electricity co-production, which is economically sustainable if compared to a new IGCC installation.
The results have been obtained through the WTEMP software, developed in the last 20 years at the Dipartimento di Macchine Sistemi Energetici e Trasporti.

The aim of this work is the development, the installation of an experimental hybrid system emulator composed by a microturbine and a SOFC emulator, and the execution of a series of experimental test to investigate the matching between turbomachines and high temperature fuel cells. The test rig will be installed at the Rolls-Royce UTC research centre of the University of Genoa in the laboratory located at Savona. The test rig will be founded by the FELICITAS, Large-SOFC european projects and some other foundings.

The part of work in charge at the University of Genoa for the FELICITAS project is the assessment of feasibility of the use of a SOFC in combination with a microturbine like APU for naval application. The work of the University will be parted to different work-packages in synergy whit other partners.

The developed test rig will also be used within the Large-SOFC project in the study of the components and subsystems of an hybrid energy system. The results will be useful to design and develop components and sub-systems suitable for SOFC units, both atmospheric and pressurized. The work intends to address the issues of efficiency and effectiveness of a commercial microturbine recuperator.

The rig is also used for educational purposes: for master and Ph. D. theses, for lessons and seminars, to provide a real machine to the students, for a better comprehension of the theoretical study and to illustrate the problems that arise with the running of a real plant.

The work carried out in these three years focused at the beginning on the development of apt simulation tools for high temperature hybrid systems steady state and transient analysis. In particular, modelling activities focused on SOFC based hybrid systems and in particular on Rolls-Royce Fuel Cell Systems Ltd design; modelling activities were carried out throughout the whole three year span as more detailed information on the system were made available. The models developed were subsequently used to evaluate performance in design and off design condition as well as provide information for the definition of a feasible control strategy, first step in the development of a robust control system. With regard to this issue, the anodic loop free response to step and sinusoidal variation of fuel and current has been studied with the aim of determining the extent of risk of hydrogen starvation, and hence anode oxidation, of the stack. The performance of the system was finally also simulated with fuels different from natural gas. This part of the work, carried out in the frame of the Felicitas European Project focused on studying the effects of providing multi fuel capabilities to such a system in terms of changes in both system layout and performances.

Nowadays part of the scientific research is focused on renewable sources to overcome the higher and higher worldwide energy requirements and the consequent increasing pollution. An example is represented by the hydrogen economy and its main application: the fuel cells.

For stationary power generation the most suitable ones are the Solide Oxide Fuel Cells (SOFC) which operate between 850 ?C and 950 ?C at high pressures. In a power plant powered by fuel cells, like the 1 MW Rolls-Royce Fuel Cell Systems Ltd developing product, there are two main loops: one that supplies air to the fuel cell stack and the other one that delivers the optimum fuel mixture to it. The work described in the thesis deals with the design and development of the main components in the anode loop such as ejector, reformer and fuel cells.

The ejector has been preferred to conventional blowers because less expensive and does not need any maintenance. Moreover it heats the entraining fuel up by mixing it with the stack off-gases. The recycle is very important also because keeps a constant ratio of steam to carbon avoiding the deposition of carbon on the fuel cells that can be damaged. The preliminary design of the ejector has been conducted developing a first 0-D model to calculate the basic sections such as nozzle, mixing duct and diffuser outlet. Afterwards the Computational Fluid Dynamics (CFD) has been employed to optimize the design of the ejector. The results have been validated in a dedicated test rig.

The reformer is a reactor which target is to convert methane into hydrogen that can be used by the fuel cells to deliver power. For its design a 0-D model has been built to calculate the kinetics (reforming and shifting) and then a first prototype has been built and successfully tested. A CFD model has been then built and validated against experimental data and it has been used together with a 0-D model to optimize the design and to increment the heat and mass transfer between catalyst and fuel bulk. The reformer, then, has been built and tested in custom-built test rig.

The last step of the work regards the SOFC modelling at different temperature and pressure conditions and using different fuel compositions. The results have been validated experimentally and used to optimize the layout of the fuel cells.

All the components built within the thesis have been coupled together and the overall anode cycle has been tested successfully in the 10 and 80 kWe demonstrators.

Fuel processing is a fundamental step in any fuel cell system fuelled with hydrocarbons. Where natural gas is the primary fuel, the process includes desulphurization, pre-reforming of higher hydrocarbons and reforming of methane. Additional fuel processing methods to support off-gas combustion, anode protection from Red-Ox and steam reforming priming were investigated. In a SOFC system pre-reforming and reforming catalyst and in particular fuel cell anodes can not tolerate the sulphur content of untreated natural gas, resulting for the former in a strong activity and life reduction and for the latter in a significant drop in voltage, even if reversible depending on temperature and exposure time.
Higher hydrocarbons (C2-C6) present in the natural gas have a decomposition temperature much lower than methane so that feeding a higher hydrocarbons rich natural gas to a high temperature SOFC system may result in carbon deposition in any point where that thermal cracking temperature is exceeded. Therefore for commercial system very low sulphur and higher hydrocarbons concentration in methane has to be obtained and natural gas has to be maintained below critical temperature before any mixing with a steam or oxygen rich stream.
Different natural gas desulphurization processes have been investigated and only three processes have been taken into consideration and compared on the basis of their suitability and applicability to a SOFC system in terms of operating conditions and volumes: passive adsorption, hydrodesulphurization and selective catalytic sulphur oxidation (SCSO). Based on data reported in the literature, SCSO resulted to be the more attractive and suitable process but required a further investigation. For this reason a study on SCSO process kinetics has been carried out in order to understand in more detail the basics of the process and the influence on the performance of the main variables controlling the process. SCSO technology also includes high capacity sorbent beds able to trap sulphur oxides formed in the upstream catalytic section. Both catalytic and sorbent bed volumes have been estimated from information reported in the literature. Sorbent beds are required to be warmed up to their operating temperature with hot air during the start-up phase and after SCSO catalyst light off are purged with natural gas: a dynamic model has been developed in order to investigate the formation of flammable mixtures in the entire sorbent beds volume and to evaluate the volume of an inert gas eventually required before the introduction of any flammable gas to avoid any safety issue.
Different natural gas pre-reforming processes have been also investigated and compared on the basis of their applicability to a SOFC system. Some experimental tests have been carried out using a high nickel content natural gas steam pre-reforming catalyst in order to understand its activity towards ethane and propane hydro-cracking combining the fuel with a small amount of synthesis gas containing hydrogen from a catalytic partial oxidation reactor.
An experimental study have been performed to investigate the kinetics of thermal cracking of the higher hydrocarbons present in natural gas, and in particular the thermal decomposition apparent activation temperature of each species on the basis of their conversion and the effect of total pressure and residence time on it.
In a SOFC hybrid system anode off-gas burner has also the function during start-up to warm the system up to its operating temperature: either high natural gas auto-ignition temperature or high system temperatures do not allow using natural gas as fuel for lighting the off-gas burner and for sustaining it in all the conditions during which either natural gas autoignition temperature is not achieved or higher hydrocarbons thermal cracking temperature is exceeded. Natural gas catalytic partial oxidation is a reforming technology which is able to generate a synthesis gas containing mainly hydrogen and carbon monoxide as fuel, then with a reduced autoignition temperature and with a negligible content of higher hydrocarbons, by partially oxidizing hydrocarbons in the natural gas over a catalyst with air added in a proper ratio. A thermodynamic analysis of this process has been performed in order to understand the effect on the equilibrium conversion and composition of the main variables controlling the process. Through a similar thermodynamic study it has been possible to evaluate critical temperatures for carbon formation due to Boudouard reaction for any catalytic partial oxidation reactor operating condition. Furthermore an experimental study has been carried out to investigate the spontaneous ignition temperature and combustion of weak syngas mixtures in air, representative of off-gas burner ignition air to fuel ratio.
During start-up and shut down operations of the SOFC system in a certain range of temperature, non-explosive or low energy fuel gas (safe gas) containing hydrogen has to be supplied to the anode side to keep it in a reduced state avoiding its re-oxidation due to the oxygen leaking through the electrolyte from the cathode side. Leakage through a fuel cell has been studied applying the principles of diffusion through porous media, since leakage flux determines the volume of safe gas required. In order to find an alternative to storage of pre-mixed safe gas, different options to generate this stream with the level of flammables desired by using natural gas have been considered and compared through a thermodynamic analysis.
Therefore the main functions of an external fuel processor to support a SOFC hybrid power generation system have been considered and different solutions have been investigated. In the end, several fuel processor general layouts have been proposed as an attempt of combining and integrating the different fuel processing technologies presented.

The aim of this work is the development of a transient model of a solid oxide fuel cell hybrid system focusing the attention on a wide experimental validation of the component models and on control system development.
The first part of the thesis is devoted to the anodic side of the plant presenting the fuel cell and the anodic ejector transient models used for plant simulations. Because of the critical performance of this single stage ejector, this work reports a wide validation carried out with an innovative experimental test rig developed at the University of Genoa.
Then, attention is focused on the cathodic side of the system for the component model validation using two experimental test rigs at unsteady conditions. The first one, developed at the University of Genoa to study the early start-up of hybrid systems emulating the turbomachine flows with a real time model, has been used for the experimental validation of volume, pipe, thermal capacitance and valve models. The second one is an hybrid system emulator with the turbomachinery and the recuperator developed at the U.S. DOE-NETL of Morgantown (WV-USA). It has been used to study load step changes and to validate the whole cathodic side of the system.
Finally, the validated component models have been used to develop the complete hybrid system transient model. After a preliminary temporal characterization of the phenomena, the plant model has been equipped with a new control system able to meet at any time the global plant power demand, supporting global load variations and avoiding dangerous or unstable conditions.

The main objective of this thesis is the system reliability study of different Hybrid
Systems through the construction of models that represent the time-to-failure of the entire
system based on the life distributions of the components, subassemblies and/or assemblies
(?black boxes?) from which they are composed.
To do this, visual calculation codes will be employed through a graphical representation
to describe the interrelation between the components of the system.
The main reason of this study is to predict and improve the future reliability performance
of these new (advanced) systems. These results are very beneficial for assuring the safety
and productivity of installations and, moreover, they provide input to other associated
studies, in maintenance planning, inspection scheduling, spare holdings and many other
activities where the reliability performance of plant, systems and equipment is of concern.
This study led to the RAM analysis of atmospheric and pressurised fuel cells integrated
with gas turbines.
The main objective is to define and understand how the system lay-out affects the RAM
Analysis of the plant. In fact the direct connection of the gas turbine with the Fuel Cells in
the pressurised plant, the presence of high temperature recuperators in the atmospheric
ones and the use of blowers or ejectors could strongly vary the reliability and availability
of the system.
Moreover, a comparison between the two typologies of high temperature fuel cell
hybrid systems above-mentioned will be made also in order to highlight how operational
parameters variations, as for instance temperature, pressure, current density, influence the
fuel cell and system reliability.
Finally, this thesis analyses which plant, among those mentioned, is more sensitive to
fuel cell reliability fluctuations and the most critical components for each kind of system.

During this last three years a newly designed micro gas turbine combustor (denominated ARI100 combustor), aimed at non-standard fuels application, has been developed at Ansaldo Ricerche s.p.a. facilities.
Centred on the development of the ARI100 combustor, the present thesis as for objective the review and development of numerical tools to support the design of small gas turbine combustors, with special emphasis on the CFD high swirling turbulent flow modelling, and on the detailed chemical kinetic analysis.
During the combustor design phase, where accuracy and low computational cost must be carefully balanced, CFD can become computationally expensive if detailed turbulence and combustion modelling is required. To obviate this fact, a simplified CFD analysis, based on 2D-axisymmetric modelling, and later verified by a 3D model, is proposed to quickly evaluated changes in the combustor design. Furthermore, a Chemical Reactor Modelling tool has been developed to decouple the fluid dynamic global reacting flow analysis from the detailed chemical kinetics analysis. Post-processing CFD results based on global kinetics, major pollutants formation can then be efficiently evaluated through the Chemical Reactor Modelling approach, while using complex multistep mechanisms.
Finally, the numerical procedure established for the ARI100 design control, has been partially validated by the ARI100 cold test rig results. Although further validation from the incoming combustion test rig is still required.

The work focused on the development of TRANSEO, a new simulation tool based on MATLAB environment for transient and dynamic analyses of advanced cycles. The modular structure of TRANSEO allows a wide range of cycle layouts to be studied in detail at steady-state or dynamic operating conditions.
Starting from previous approaches for a flexible and general interface of the components, a new interconnecting plenum protocol has been developed and adopted. A special time-characterisation procedure allows the user to easily monitor the results and ensures the accuracy required by the type of simulation. The components are organised in a library and different models are available for each of them: the dynamic, the lumped-volume and the off-design models. Different component models can be mixed in the same cycle simulation, according to the user?s requirements.

Attainable results are shown to demonstrate both the accuracy of results and the approach adopted. The test-cycle is constituted by a conventional recuperated microturbine cycle (Bowman TG-45), whose experimental data were available for code validation. TRANSEO was then applied to some advanced microturbine cycles, such as the Externally Fired micro Gas Turbine cycle (EFmGT), the Closed Brayton Cycle (CBC) for space applications, the Molten Carbonate Fuel Cell Hybrid System Simulacrum (?Emulator?).

The results provide useful information on the transient behaviour of such cycles and the characteristic time delays of the main cycle components.

In order to perform reliable, innovative gas-turbine cycle calculations, a new flexible computational tool for the complete analysis (convective and film) of high-temperature gas-turbine stages has been proposed. It consists of a calculation procedure incorporating a new analytical procedure for the investigation of the blade cooling and a new one-dimensional modelling of the expansion through the blading, the two calculation systems working synergistically.

The calculation approach is a ?stepped? approach. The blade cooling analysis is, in fact, applied to single strips of the blade in the cordwise direction, thus basing the evaluation on the local values of the heat transfer coefficients, mainstream gas flow temperature, film effectiveness, internal blade geometry, etc. The one-dimensional expansion modelling parallels the blade cooling analysis and considers the elementary steps of the expansion in the downstream direction, the effects of cooling and mixing being evaluated at each intermediate station.

The investigation of the blade cooling starts with the knowledge of the blade-row geometry required to run the simulation. The effect of the internal blade geometry (i.e. cooling channel geometry) on the cooling performance is considered through the cooling-system technology-level parameter Z. The procedure is able to automatically set a proper value for Z based on semi-empirical data: the Z setting is therefore determined at the level of existing technology.

Once Z is determined, the internal blade configuration is established, making it possible to calculate the pressure losses of the coolant and the behaviour of a given blade configuration in an environment consisting of a new cooling medium and/or a new hot gas composition. This means that valuable information on blade-design requirements for alternative fluids can be obtained. It is also possible to study how Z should be altered, for example by varying the hot gas temperature, in order to avoid an excessive coolant mass flow or pressure drop.

The new code has been applied to the evaluation of an existing high-temperature gas-turbine stage. The investigated stage consists of the film-cooled nozzle and first rotor of an aero-derivative gas turbine from a European manufacturer.

Comparisons with a simpler semi-empirical calculation method, assuming constant blade temperature along the span, indicate broad agreement.

However, the full potential of the code lies in its ability to respond to a variation of input data (e.g. heat transfer coefficients, blade internal geometry, etc.) along the blade chord. This feature has yet to be exploited, and will be the subject of future work. In the future use of the code, in fact, this will lead to the optimum distribution of the cooling flow over chordwise distance, and may enable the total cooling flow to be minimised.

The new method presented should give more reliable evaluations than the semi-empirical ones, particularly for the study of innovative cycles, where conventional fluids and operating conditions cannot be assumed in the performance calculations. It is hoped that such a full development of the code will provide a useful design-tool for the turbine designer faced with the problem of blade cooling.

High temperature Fuel Cell-Gas Turbine (FC/GT) hybrid systems have exceptional potential to improve both the local and global environmental impact of power generation.

Their near-zero emissions and low noise will allow them to be used in populated areas, reducing the community?s dependence on overhead electrical transmission and distribution systems and leading to improvements in the quality of life. Furthermore, their potential to generate electricity at a greater efficiency than conventional Combined Heat Power (CHP) plants, even on relatively small scales, makes them economically attractive .

The key point in the economic and technical development of FC/GT power generation plants is the possibility of predicting their performance at both design and off-design conditions. Since these systems are currently under development and only demonstration plants have been tested, simulation plays an important role in their design. In particular, the simulation of the SOFC electrochemical reactor (which is a key component of the plant) is of critical importance, and is under intensive study at present. Different types of simulation models have been proposed for Solid Oxide Fuel Cells (SOFCs), and they can be classified as (i) models based on the interpolation of experimental data of reactor behaviour under different operating conditions; (ii) approaches based on balance equations in the form of macroscopic balances, i.e. finite equations that simply express a balance between inlet and outlet flows of mass and energy, and allow the evaluation of the average values of the physical-chemical variables (i.e., temperatures, concentrations, etc.), and the electrochemical performance of the SOFC reactor as a function of the operating conditions, and (iii) modelling approaches based on balance equations written as local balances in the form of partial differential equations. These equations require numerical integration along the SOFC geometrical co-ordinates, allowing a detailed evaluation of the distribution of the physical-chemical variables (temperature, gas compositions, electrical current density, etc.) within the SOFC.

Considering that the aim of this work is the study of Hybrid Systems, including the SOFC model as a module in the simulation of the whole plant, even if the simplified approaches are less accurate than the detailed ones, this loss of accuracy is sometimes acceptable within the framework of the overall plant evaluation. In these cases a significant reduction in the geometrical input data and computational time required can be achieved while obtaining reasonably accurate results. On the other hand, it has to be said that simplified models do not allow a profound investigation of the status of the SOFC reactor, and there are several significant cases where critical issues, such as hot spots and/or steep temperature gradients, dangerous for the stability of the materials involved, can arise and not be detected with the simplified models. These issues, and in particular the limitations of the different types of simulation tool, are investigated in the present work, with particular reference to the pressurised Solid Oxide Fuel Cells coupled to a recuperated gas turbine with variable speed control. Critical aspects such as the light-off and the extinction of the SOFC, and the sensitivity of the plant model to the accuracy of the simulation in this respect have been investigated at both the design point and at off-design conditions.

This work was split into several phases. Firstly a macroscopic fuel cell model was developed for the study of different cell geometries and sizes. Afterwards, the design point and off design performance of the cell model were studied to develop an even more detailed model. Then the behaviour of the SOFC reactor was minutely investigated under different operating conditions using the two different models and the results obtained were compared to define a range of reliable simulation results. Finally, a Hybrid System model was developed considering all the components of the plant in detail. The two different Fuel Cell models were integrated into the Hybrid System simulation tool.

In the last part of this study the response of the plant was investigated with particular attention to the influence of the control system. The effect of the approach to the Fuel Cell simulation was also considered at this point.
Finally a preliminary thermoeconomic analysis of different Hybrid Systems (size and lay-out) was performed.

Part of the work concerning the development of the design and off-design model of a SOFC Hybrid System has been carried out in direct collaboration with Rolls Royce and with the support of Turbec, Turbomeca and Alstom Power.

The work dealt with the complex and interesting topic of the thermoeocnomic and environomic analysis and optimisation of energy systems, with the development of TEMP code, whose acronym stands for ?Thermoeconomic Modular Program?.

This approach represents the first attempt to make thermoeconomics easy to use and to exploit for any users involved in power plant design and management.

The TEMP approach allows the internal and global thermoeconomic analysis to be carried out automatically. The influence of various parameters such as equivalent operating hours, price of fuel, revenues, emission taxation, emission abatament costs, etc? on the best plant configuration (as far as thermoeconomic objective functions are optimised by a non linear optimisation algorithm, named ?Direct Thermoeconomic Optimisation?) can be quantified numerically. Moreover, also the exergoeocnomic analysis was developed and included in TEMP, so that the thermoeconomic diagnostics of complex energy systems can be carried out straightforward.

The work dealt with the complex and interesting topic of the thermoeocnomic and environomic analysis and optimisation of energy systems, with the development of TEMP code, whose acronym stands for ?Thermoeconomic Modular Program?.

This approach represents the first attempt to make thermoeconomics easy to use and to exploit for any users involved in power plant design and management.

The TEMP approach allows the internal and global thermoeconomic analysis to be carried out automatically. The influence of various parameters such as equivalent operating hours, price of fuel, revenues, emission taxation, emission abatament costs, etc? on the best plant configuration (as far as thermoeconomic objective functions are optimised by a non linear optimisation algorithm, named ?Direct Thermoeconomic Optimisation?) can be quantified numerically. Moreover, also the exergoeocnomic analysis was developed and included in TEMP, so that the thermoeconomic diagnostics of complex energy systems can be carried out straightforward.