The presence of particulate and gaseous pollutants at elevated concentrations in the atmospheric environment is detrimental to public health. The present study has investigated the impacts of a steelworks complex on the air quality in Port Talbot, South Wales, United Kingdom. Different offline and online air monitoring instruments were deployed to four sites around the perimeter of the steelworks (at one coastal site (Little Warren LW) and 3 inland sites placed along the length of the steelworks (Fire Station FS, Prince Street PS and Dyffryn School DS) in the study area for a four-week campaign (April 16 to May 16, 2012). Prior to Port Talbot campaign, a separate two week sampling (March 30 to April 12, 2012) was conducted at Elms Road Observatory Site (EROS) for a representation of an urban background. Gaseous and meteorological data logged during the period of sampling were also collected from the Automatic Urban and Rural Network (AURN) site at Port Talbot Margam (FS site). Hourly and daily data collected were prepared for receptor modelling using Positive Matrix Factorization (PMF) and with the use of Open Air and Lakes Environmental WRPLOT View software, windrose and polar plots were produced to show the directional emissions of particulate and gaseous pollutants. The online sampling instrument of Aerosol Time of Flight Mass Spectrometer (ATOFMS) as well as PMF solutions for Streaker and Partisol were able to identify different processing units of the steelworks responsible for pollutant emissions. The polar plots for most air pollutants revealed the steel industry as the major contributor to air pollution in the study area.

The continuous growth in wind turbine power ratings and numbers has led to increased demands in inspection and maintenance due to the more significant financial and operational consequences of unexpected wind turbine failure. The fact that wind farms are commonly located at remote sites with potentially poor accessibility means it is necessary to reduce the need for corrective maintenance through evolution to preventive and prognostic maintenance activities. Prognostic repair schedules can be employed in order to optimise maintenance and contribute to the minimisation of the overall operational costs of wind farms. The present study presents the development and qualitative evaluation of remote condition monitoring methodologies for the evaluation of the wind turbine power electronics and gearboxes. The failures of power converter and gearbox components result in significant wind turbine downtime and associated repair costs. Effective condition monitoring can enable the timely diagnosis of faults in order to prevent unexpected failures and loss of electricity production, contributing towards a noteworthy increase the reliability, availability, maintainability and safety (RAMS) of wind farms. Within this study two customised test rigs have been employed to simulate various of faults and assess the capability of RCM in diagnosing this fault effectively. In addition, field measurements have been carried out and correlated to the findings of the test rig experiments. In this study, it has been possible to identify these variables qualitatively, but the quantitative investigation is still pending and will be most likely the subject of several future studies in this field. The present thesis provides a compact summary of the analysis of the key findings of the experimental work performed within the context of the OPTIMUS FP7 European collaborative project.

Many real world problems can be formalized as optimisation problems. Yet, the environment in the real world is mostly changing over time, which makes the optimisation problems change over time. This thesis is dedicated to the problem of finding robust solutions in changing environments. By robustness, we mean that a solution not only is good for the current environment but also maintains its performance after future environmental changes. The problem of finding such robust solutions repeatedly over time is termed as Robust Optimisation Over Time (ROOT). The first contribution of this thesis is a formal definition of ROOT problems and two robustness definitions of solutions to ROOT. The second contribution of this thesis is two benchmarks developed for ROOT. The third contribution of this thesis is an algorithm framework for finding solutions repeatedly over time to ROOT problems. Finally, the fourth contribution of this thesis is a dynamic handling strategy, which is used to initialise a population of solutions right after an environmental change when evolutionary algorithms are used for ROOT problems.

Observations of atmospheric corrosion on aluminium alloy AA2024-T3 were made within droplets of NaCl solution and simulated ASTM ocean water. High speed in-situ synchrotron X-ray tomography has been used to observe the initiation and propagation of corrosion sites. Tomographic scans had a temporal resolution of 300s per full tomographic scan, providing a non-destructive visualisation of corrosion. Sites initiate at numerous surface morphologies, including at intermetallic inclusions, under surface deposits and salt crystals, and also at sites with no obvious micron-scale microstructural features. It was observed that corrosion sites grow discontinuously and inhomogeneously when conditions are constant, consistent with bursts of dissolution followed by some repassivation. The nature of corrosion products was investigated with Raman spectroscopy and EDX. The dominant corrosion product was Dawsonite for both droplet solutions, however the distribution of corrosion products differed between the two types of droplet. NaCl droplets spread more, resulting in fewer, larger corrosion sites. ASTM ocean water droplets showed multiple small corrosion sites within the droplet owing to the presence of Hydrotalcite around the droplet perimeter. Corrosion development under wet-dry cycling conditions was different for NaCl and ASTM ocean water droplets. Under NaCl droplets, no growth or initiation was observed during “dry” phases. However, ASTM ocean water remains partially wet in “dry” periods as some of the salt constituents in ASTM ocean water remain hydrated. As a consequence, corrosion site growth is able to continue.

The key product from anaerobic digestion (AD) is biogas, which is used to generate heat and/or electricity. Anaerobic digestion involves degradation and stabilisation of the feedstock by microorganisms, leading to the formation of biogas and a digestate residue, which is used as a fertiliser. Contaminants entering the system in the feedstock may limit biogas yield if functioning of the methanogenic archaea is disrupted. Digestate quality may also be compromised. Approximately 40% of the current UK biogas plants are farm-based, using manures and slurry as the main feedstock. Veterinary medicines are excreted in dung of treated animals and if used prophylactically, concentrations in the faeces or urine may be significant. Contaminated animal waste is therefore likely to be used as a feedstock in AD. Since digestates are commonly used as fertilisers, contaminant loading and fate must be understood to prevent transfer into crops, reductions in soil microbial activity, increased antibiotic resistance and detrimental effects to livestock if digestates are applied to fodder crops or pasture. It is unlikely that the use of veterinary pharmaceuticals will be reduced whilst livestock farming is intensifying and the current demand for meat is growing, therefore understanding the operational processes of AD that influence the persistence of commonly used veterinary medicines and subsequent toxicity are crucial to minimising potentially detrimental effects. Research was undertaken using laboratory-scale digestion vessels to quantify the effect of the commonly used veterinary antibiotics, oxytetracycline and tylosin, when added to naïve (organic) cattle dung or to slurry from a conventional dairy farm. Anaerobic digestion units were spiked with either oxytetracycline or tylosin at low (environmentally realistic) and artificially high concentrations, either at start-up (day 0) or once the system was producing gas (day 15). Biogas production was measured and gas collected every 5 days to quantify the temporal effect of the antibiotics on methane production. Oxytetracycline and tylosin significantly reduced both biogas quality and quantity, with the extent of the effect differing with each feedstock. In organic cow dung, the low (4.33 mg L-1) and artificially high (86.63 mg L-1) concentration of oxytetracycline added on day 15 to organic cow dung caused an overall drop in biogas production of 12% and 25% respectively, whilst the same concentrations incorporated at start-up caused a drop of only 4% and 18% respectively. Both the low and artificially high concentrations of tylosin added on day 15 caused a 33% drop in biogas production, whilst the same concentrations incorporated on day 0 caused a drop of 15% and 42% respectively. In conventional dairy slurry, low (4.33 mg L-1) and artificially high (86.63 mg L-1) concentrations of oxytetracycline caused an overall 3% and 10% drop in biogas production respectively, with tylosin amendment causing a decrease in total biogas production of 7% and 22% respectively. Feedstock origin affected biogas production and quality when the system was challenged by antibiotic inputs. These data highlight the complex interactions that can occur between feedstock and exposure to veterinary pharmaceuticals.

Within the last decades the importance of sustainable treatment technologies, such as constructed wetlands (CWs) and vegetated ponds, has raised due legislation (e.g. WFD), directing toward green infrastructure to mitigate water pollution. The efficiency of pond and CW treatment systems depends on the internal hydrodynamics and mixing interactions between water and aquatic vegetation. In order to contribute to the current knowledge of how emergent real vegetation affects solute mixing, and physical flow characteristics in full-scale aqueous systems, an understanding and quantification of those processes and interactions was sought under the: i) natural seasonal vegetation and flow rate variation in two CWs, and ii) physical flow characteristics in overall six different full-size treatment units. To address these issues, outdoor tracer field studies were undertaken in each treatment unit. Regarding the seasonal plant variation, an intelligent automated tracer injection system was developed to achieve autonomous remote measurements in two CWs, vegetated by Phragmites australis, in different seasons and flow rates. Experiments involved measurements of longitudinal mixing, physical flow characteristics and vegetation characteristics in different plant ages and various discharges. It was shown that seasonal vegetation variation influences the longitudinal mixing coefficient by up to four times, and the physical flow characteristics by increasing the flow resistance and creating stagnant backwaters at the end of plant cycle, achieving reduction of the peak concentration by three times. Longitudinal mixing decreased with discharge in all plant ages. Furthermore, it was shown that internal design (i.e. bed topography or vegetation distribution) overwhelm the seasonal plant variation effects on mixing and flow characteristics. Moreover, relative comparison of outlet configuration, inflow conditions, and internal features, between the six different treatment units demonstrated an increase in residence time by up to three times. Results underlined the importance of investigating hydrodynamics and physics of flow in full-size units to enhance treatment efficiency and predictions of water quality models.

Digitized by Kansas Correctional Industries

Due to high regeneration energy demands for amine absorption processes for post-combustion CO2 capture, alternative technologies such as adsorption processes using solid adsorbents have been considered. Other practical issues such as corrosion of equipment and loss of solvent can be avoided with adsorption processes. Fixed bed adsorption processes, in which CO2 adsorption and adsorbent regeneration are performed successively in a vessel packed with adsorbent, are the most common adsorption processes. However, in fixed bed temperature swing adsorption (TSA) processes, large columns and long heating and cooling times would be needed. Fixed bed pressure swing adsorption (PSA) processes use electrical energy, which is more expensive than thermal energy in a power plant. Therefore, the feasibility of moving bed adsorption processes including fluidised-bed, co-current and counter-current systems is investigated. In these systems, the adsorbent continuously circulates from a CO2 adsorber to a regenerator. The adsorbents considered are a supported amine adsorbent, activated carbon and zeolite 13X. Numerical simulations of moving bed TSA cycles for CO2 capture have been carried out. The effects of influential parameters in the process have been assessed via sensitivity analyses. It was found that counter-current beds with supported amine adsorbent give the best overall performance. Compared to an amine absorption process, it was found that a moving bed TSA process without heat integration requires the same heat consumption per unit mass of CO2 captured. There is a potential for a lower heat consumption in moving bed TSA processes if, similarly to amine absorption processes, heat integration is carried out or if the CO2 working capacity of the adsorbent can be increased.

Digitized by Kansas Correctional Industries

An aerosol chemical ageing (ACA) system has been developed for artificially processing atmospheric particles. An aerosol flow tube coupled to a scanning mobility particle sizer (SMPS) and an aerosol time-of-flight mass spectrometer (ATOFMS) have been proposed to study the heterogeneous reaction (HR) between two oxidants, ozone (O\(_3\)) and hydroxyl radical (OH), and three different single organic aerosol proxies: oleic acid (OL), maleic acid (MA) and bis(2-ethylhexyl) sebacate (BES). The ACA system operates under conditions equivalent to ambient processing times of 1 and 20 days with respect to ozone and OH ambient levels, respectively. The study provides evidence for the ageing process of organic aerosol. The physical characterisation of aged particles suggests the formation of volatile products resulting in appreciable decrease in particles size and mass. The chemical study, however, shows that the properties of the aged particles are sensitive to the oxidant and organic material identities, the particle size and the oxidation environment. The applicability of the developed ageing technique has been tested on real atmospheric particles, however, it is concluded that extrapolating laboratory procedures to ambient atmosphere may be challenging due to the complexity of the real atmosphere, particularly, the urban atmosphere.