With current electronic designs becoming more versatile and mobile, applications that were wired and bulky before have now seen a great reduction in size and increase in portability. However, the issue is that the scaling down in size and cost of electronics has far outpaced the scaling up of energy density in batteries. Therefore, a great deal of research has been carried out to search for alternative power sources that can replace or enhance the conventional battery. Energy harvesting (also known as energy scavenging) is the process whereby ambient energy is captured and stored. The ambient energy here refers to energy that is pre-existing in nature, and is self-regenerating and has extended life time from a battery. After reviewing many possible energy scavenging methods, the conversion of ambient vibrations to electricity is chosen as a method for further research. There are plenty of different methods to transform ambient vibration to electricity, but in this research only piezoelectric and electromagnetic conversions are pursued. In order to harvest the most energy with the harvesting device, the harvester’s fundamental mode must be excited. However, this is not always possible due to fluctuations in the frequency of the vibration source. By being able to change the natural frequencies of the device, the harvester could be more effective in capturing ambient energy. In this thesis, the behaviour of the various types of energy sources is studied and the obtained information is later used to generate a vibration signal for subsequent simulation and experiments. A converter based on a piezoelectric bimorph is investigated. The resultant outputs from the design are compared to the model and the analysis is presented. The mechanical strain distributions on the beam’s surface for five different geometric structures are compared and discussed. This is followed by a discussion of the feasibility of improving the strain distribution by changing the beam’s depth (height) along the cantilever beam length. Lastly, a novel frequency tuning method, which involves applying a different effective electrical damping in different quadrants of the oscillating cycle, is proposed. The results of this analysis are presented, along with experimental results that indicate that the behaviour of the system can be changed over a limited range by changing the effective electrical damping during the oscillation cycle.

Measurements of the longitudinal and transverse direct current (d.c.) magnetoresistance of n-type InP samples (carrier density from and the alternating current (a.c.) conductivity of n-type InSb samples (carrier density from have been made at temperatures T down to 0.04 K and in magnetic fields H up to 70 kG. For H=0, the InP samples were nonmetallic. At low temperatures, the conductivity is due to nearest neighbour hopping (NNH) which is followed by variable range hopping (VRH) at lower T as described by the first, and second terms in the expression. In the NNH regime, it is necessary to plot In (p/T) against T1 and this yields values of the activation energy much larger than the traditional In p versus T-1 plots In the VRH regime, Mott's law (x = 1/4) is obeyed. Values of To obtained by considering the temperature dependence of the pre-exponential factor are found to be much higher than if the temperature dependence of this factor is ignored. Good agreement between the theory and experiment is achieved in both NNH and VRH regions if an enhanced dielectric constant is used. Magnetoresistance measurements in both the NNH and VRH regimes are analysed using the theories of Shklovskii and Efros (1984) and reasonable agreement is obtained. The anisotropy of the magnetoresistance in the NNH agrees closely with the expected H2 dependence. In the VRH, In(p(H)/p(0)) varies as T-3/4 and H2 as expected for hopping with a constant density of states at the Fermi level. The InSb samples are metallic-like in zero magnetic field. High magnetic fields are applied to shrink the donor wavefunctions (to induce the metal-insulator transition) and to locate the samples on the insulator side where the measurements are carried out. The d.c. resistivity was measured and at low temperatures was of a VRH type with 1/4 x 1/2, and T0 being magnetic field dependent. Reasonable agreement with the theory is found at high fields. The real and imaginary parts of the a.c. conductivity were measured in the frequency range of 110-105 Hz. The real part of conductivity was found to vary as where s is approaching 1 at low temperatures and high fields but decreasing as T increases. At the lowest temperatures was independent of T but at higher T the temperature dependence is stronger than the linear dependence predicted by the simple pair approximation theory. Data are interpreted in terms of multiple hopping of electrons which becomes important at high temperatures and/or low frequencies. The scaling formula; has been applied to discuss the results for the real part of the conductivity, where and are normalized values and f is a universal function obtained by Summerfield (1985). The scaling parameter -log10A is found to be 3.0 +/- 0.2.The relative dielectric constant, due to donors, calculated from the capacitive part was found to be a decreasing function as the frequency is increased and/or the temperature is lowered. At very low temperatures, depending on the magnetic field, however, a temperature-independent, but frequency-dependent behaviour is observed. The lowest temperature value of the dielectric constant was found to diverge as the magnetic field is reduced towards the metal-insulator threshold value.

Reducing CO₂ emissions from fossil fuel-fired power plants is a significant challenge, technically and economically. Post-combustion carbon capture (PCC) using amine solvents is widely regarded as the most promising technology that can be commercially deployed for carbon capture from fossil fuel-fired power plants. However, for its application at full commercial scale, the main barrier is high cost increment of the electricity due to high capital costs and significant energy penalty. This thesis presents the studies on optimal design and operation of Monoethanolamine (MEA)-based PCC process and the integrated system with natural gas combined cycle (NGCC) power plant through modelling, simulation and optimisation, with the aim to reduce the cost of PCC commercial deployment for NGCC power plants. The accuracy of optimisation depends on good predictions of both process model and economic model. For the process modelling, the philosophy with its framework was analysed for this reactive absorption (RA) process. Then the model was developed and validated at three stages. In the first stage, the predictions of thermodynamic modelling were compared with experimental data of CO₂ solubility in aqueous MEA solutions. The results show the combination of correlations used in this study has higher accuracy than other three key published contributions. Then key physical properties of MEA-H₂O-CO₂ system were also validated with experimental data from different publications. Lastly, a steady state process model was developed in Aspen Plus® with rate-based mass transfer and kinetic-controlled reactions. The process model was validated against comprehensive pilot plant experiment data, in terms of absorption efficiency and thermal performance of the integrated system. The cost model was developed based on the major equipment costs provided by vendors after detailed engineering design in a benchmark report. The uncertainty of this method could be in the range of from −15% to 20%, instead of other empirical methods with uncertainty of from −30% to 50%. The cost model was integrated into the process model by coding Fortran subroutine in Aspen Plus®. Using this integrated model, the optimisation studies were carried out for the PCC process only. The impact of key variables variation was also analysed. Subsequently, the scope of this study was extended to cover different sections of the integrated system including a 453MWe NGCC power plant, PCC process, CO₂ compression trains and CO₂ transport pipeline network. For the integration of NGCC power plant with PCC process and CO₂ compression, exhaust gas recirculation (EGR) technology was investigated and showed significant economic benefit. A specific supersonic shock wave compressor was adopted for the CO₂ compression and its heat integration options with power plant and PCC process were studied. For the study on the CO₂ transport pipeline network planned in the Humber region of the UK, a steady state process model was developed using Aspen HYSYS®. The process model was integrated with Aspen Process Economic Analyzer® (APEA), to carry out techno-economic evaluations for different options of the CO₂ compression trains and the trunk onshore\offshore pipelines respectively. The results show the optimal case has an annual saving of 22.7 M€ compared with the base case. In the end, optimal operations of NGCC power plant integrated with whole carbon capture and storage (CCS) chain under different market conditions were studied. Levelised cost of electricity (LCOE) is formulated as the objective function. The optimal operations were investigated for different carbon capture level under different carbon price, fuel price and CO₂ transport and storage (T&S) price. The results show that carbon price needs to be over €100/ton CO₂ to justify the total cost of carbon capture from the NGCC power plant and needs to be €120/ton CO₂ to drive carbon capture level at 90%. The results outline the economic profile of operating an NGCC power plant integrated with CCS chain. It could help power plants operators and relevant government organizations for decision makings on the commercial deployment of solvent-based PCC process for power plans.

Fossil-fuel fired power plants are subjected to stringent operational regime due to the influx of renewable resources and the CO2 emission reduction target. This study is aimed at modelling and analysis of supercritical coal-fired power plant (SCPP) integrated with post-combustion CO2 capture (PCC) and its response electricity grid demand constraints. Current status of dynamic modelling of SCPP integrated with PCC was reviewed to identify the gaps in knowledge. It was observed that no accurate dynamic model of an SCPP integrated with PCC had been reported in open literature. A steady state model of the SCPP integrated with PCC was developed with Aspen Plus®. The model was validated with the reference plant and it was found that the relative error is about 1.6%. The results of the conventional and advanced exergetic analysis showed that the energy/exergy consumption and the efficiency of the integrated system can be improved by recovering the avoidable exergy destruction in the whole system. Dynamic models of SCPP once-through boiler based on lumped parameter and distributed parameter approaches were compared. The distributed parameter model gave a more accurate prediction of the SCPP boiler dynamics at different load levels. Analysis of the strategies for operating the SCPP under the UK grid requirement as regards to primary frequency response was performed using the validated SCPP model. The results show that using turbine throttling approach, extraction stop or condensate stop individually was not sufficient to meet the grid requirement. A combination of turbine throttling, extraction stop and/or condensate stop can achieve a 10% increase in maximum continuous rating (MCR) of the power plant within 10 seconds to 30 seconds of primary frequency change as required by the UK grid. The dynamic model of SCPP was integrated with a validated and scaled-up model of PCC. Analysis of the strategies for operating the SCPP integrated with PCC under the UK grid requirement as regards to primary frequency response was undertaken. The results show that the stripper stop mechanism is not sufficient for the 10% MCR required for the primary response. The results show that the combination of stripper stop mechanism with extraction stop can meet the 10% MCR requirement for integrated plant operating at above 75% of its full capacity. The throttling and stripper stop configuration only barely meets the demand at full load capacity. The condensate stop combination with the stripper stop mechanism on the other hand could not meet the frequency response requirement at any load level.

In recent years there has been an increased demand for integration of renewable energy into the electricity grid. This has increased research into power converter solutions required to integrate renewable technology into the electricity supply. One such converter is a Cascaded H-Bridge (CHB) Multilevel Converter. Operation of such a topology requires strict control of power flow to ensure that energy is distributed equally across the converters energy storage components. For operation at high power levels, advanced modulation methods may be required to ensure that losses due to non-ideal semiconductor switching are minimised, whilst not compromising the quality of the voltage waveform being produced by the converter. This thesis presents several low switching frequency modulation methods based on Selective Harmonic Elimination (SHE) in order to address these two operational issues. The methods presented involve manipulating the H-Bridge cell voltages of the CHB converter to control power flow. Simulated results are supported by experimental verification from a seven level, single phase CHB converter.

Power quality problems come in numerous forms (commonly spikes, surges, sags, outages and harmonics) and their resolution can cost from a few hundred to millions of pounds, depending on the size and type of problem experienced by the power network. They are commonly experienced as burnt-out motors, corrupt data on hard drives, unnecessary downtime and increased maintenance costs. In order to minimise such events, the network can be monitored and controlled with a specific control regime to deal with particular faults. This study developed a control and Optimisation system and applied it to the stability of electrical power networks using artificial intelligence techniques. An intelligent controller was designed to control and optimise simulated models for electrical system power stability. Fuzzy logic controller controlled the power generation, while particle swarm Optimisation (PSO) techniques optimised the system’s power quality in normal operation conditions and after faults. Different types of PSO were tested, then a multi-swarm (M-PSO) system was developed to give better Optimisation results in terms of accuracy and convergence speed. The developed Optimisation algorithm was tested on seven benchmarks and compared to the other types of single PSOs. The developed controller and Optimisation algorithm was applied to power system stability control. Two power electrical network models were used (with two and four generators), controlled by fuzzy logic controllers tuned using the Optimisation algorithm. The system selected the optimal controller parameters automatically for normal and fault conditions during the operation of the power network. Multi objective cost function was used based on minimising the recovery time, overshoot, and steady state error. A supervisory control layer was introduced to detect and diagnose faults then apply the correct controller parameters. Different fault scenarios were used to test the system performance. The results indicate the great potential of the proposed power system stabiliser as a superior tool compared to conventional control systems.

As the deployment of renewable generation increases in the worldwide electrical grids, the development of distributed energy storage becomes more and more of an essential requirement. Energy storage devices connected at Medium Voltage allows for much higher powered deployments and this Ph.D. will focus on the power converter used to interface the energy storage device to the electrical grid. Multi-level converters can be used to provide this interface without huge filtering requirements or the need of a Low Frequency step up transformer. However traditional Multi-level converter topologies require a large number of electrolytic capacitors, reducing the reliability and increasing the cost. Multi-level converters constructed from a Cycloconverter Topology do not require any additional electrolytic capacitors, however the High Frequency transformer, used to provide isolation has to be considerably larger. This Ph.D. will investigate a novel hybrid converter topology to provide an interface between an energy storage device, such as a super-capacitor or battery, to the Medium Voltage grid, designed for high reliability and power density. This topology is called The Hybrid Cycloconverter Topology and is based on a Cycloconverter Topology connected to an auxiliary 3-Phase VSI. A comprehensive simulation study is carried out to investigate the semiconductor losses of this novel converter topology and compared against two alternative topologies. An experimental converter is constructed to validate the theory of operation and to justify its effectiveness.

Voltage source inverter (VSI) is the traditional power converter used to provide variable voltage and frequency from a fixed voltage supply for adjustable speed drive and many other applications. However, the maximum ac output voltage that can be synthesized by a VSI is limited to the available dc-link voltage. With its unique structure, the Z-source inverter can utilise shoot-through states to boost the output voltage and provides an attractive single-stage dc-ac conversion that is able to buck and boost the voltage. For applications with a variable input voltage, this inverter is a very competitive topology. The same concept can equally be extended to the two-stage matrix converter, where a single Z-source network is inserted in its virtual dc-link. The topology formed is, thus, quite straightforward. Its modulation is, however, non-trivial if advantages like voltage buck-boost flexibility, minimum commutation count, ease of implementation, and sinusoidal input and output quantities are to be attained simultaneously. This thesis presents two novel space vector modulation methods for controlling a three-level Z-source neutral point clamped VSI to enable the use of a boost function. The second of the two space vector modulation methods is then adopted and applied to a three-level, two-stage matrix converter with a Z-source network inserted in its virtual dc-link to increase the voltage transfer ratio beyond the intrinsic 86.6\% limit. Simulation results are supported by experimental verification from two laboratory prototype converters.

This research deals with optimising a supercapacitor-battery hybrid energy storage system (SB-HESS) to reduce the implementation cost for solar energy applications using the Genetic Algorithm (GA) and the Support Vector Machine (SVM). The integration of a supercapacitor into a battery energy storage system for solar applications is proven to prolong the battery lifespan. Furthermore, the reliability of the system was optimised using a GA within the Taguchi technique in the supercapacitor fabrication process. This is important to reduce the spread in tolerance of supercapacitors values (i.e. capacitance and Equivalent Series Resistance (ESR)) which affect system performance. One of the more important results obtained in this project is the net present cost (NPC) of the Supercapacitor-battery hybrid energy storage system is 7.51% lower than the conventional battery only system over a 20-years project lifetime. This NPC takes into account of components initial capital cost, replacement cost, maintenance and operational cost. The number of batteries is reduced from 40 (conventional – battery only system) to 24 (SB-HESS) with the inclusion of supercapacitors in the system. This leads to reduction cost in the implemented hybrid energy storage system. A greener renewable energy system is achievable as the number of battery is reduced significantly. An optimised combination of the number of components for renewable energy system is also found. The number of batteries is sized, based on the average power output instead of catering to the peak power burst as in a conventional battery only system. This allows for the reduction in the number of batteries as the peak power is catered for by the presence of the supercapacitor. Subsequent efforts have been focused on the energy management system which is coupled with a supervised learning machine – SVM, switches and sensors are used to forecast the load demand beforehand. This load predictive-energy management system is implemented on a lab-scaled hybrid energy storage system prototype. Results obtained also show that this load predictive system allows for accurate load classification and prediction. The supercapacitor in the hybrid energy storage system is able to switch on to cater for peak power without delay. This is crucial in maintaining an optimised battery depth-of-discharge (DOD) in order to reduce the rate of battery damage thru a degradation mechanism which is caused from particular stress factors (especially sulphation on the battery electrode and electrolyte stratification).

In the coming decades, electrical energy networks will gradually change from a traditional passive network into an active bidirectional one using concepts such as these associated with the smart grid. Power electronics will play an important role in these changes. The inherent ability to control power flow and respond to highly dynamic network will be vital. Modular power electronics structures which can be reconfigured for a variety of applications promote economies of scale and technical advantages such as redundancy. The control of the energy flow through these converters has been much researched over the last 20 years. This thesis presents novel control concepts for such a structure, focusing mainly on the control of a Cascaded H-Bridge converter, configured to function as a solid state substation. The work considers the derivation and application of Dead Beat and Model Predictive controllers for this application and scrutinises the technical advantages and potential application issues of these methodologies. Moreover an improvement to the standard Model Predictive Control algorithm that include an intrinsic modulation scheme inside the controller and named Modulated Model Predictive Control is introduced. Detailed technical work is supported by Matlab/Simulink model based simulations and validated by experimental work on two converter platforms, considering both ideal and non-ideal electrical network conditions.