| Summary: | This thesis is a thorough examination of high-pressure enriched oxygen system design and analysis focussing on material selection and incident investigation. The aims are to develop a model to predict spontaneous ignition temperatures (SIT), enable the use of more accessible measurement apparatus, and to devise a scientific methodology to investigate oxygen incidents. Chapter 1 contains examples of incidents in high-pressure enriched oxygen and outlines current methods of material selection and incident investigation. The major flaws with the current systems are explored and the objectives of this work are identified. Chapter 2 details the current state of the knowledge in relation to oxygen incidents. Materials flammability is explored, and the importance of correct material selection is established. The criteria, standards and test methods currently used to aid this decision are assessed, and the importance of proper oxygen incident investigation is determined. Chapter 3 shows a programme of experimental work with details on the use of different apparatus for the measurement of SITs. Data acquired from the BS 4N 100 bomb test, and pressurised, and ambient, differential scanning calorimetry are recorded. The use of Thermal Desorption/Gas Chromatography is explored to identify polymer evaporation and degradation products. In Chapter 4 a simple model is developed to adapt SIT allowing calculation of the SIT of a non-metal in any pressure or oxygen concentration. Data obtained in chapter 3, and from the literature, is used to validate this model. Data on metals is also collected from the literature to be incorporated into a methodology demonstrating the kindling chain. Chapter 5 covers the development of a ‘tool kit’ for oxygen incident analysis. Understanding of ignition modes, heat transfer modelling, flow diagrams, and the SIT model are used to examine past oxygen incidents, and understand kindling chains. Chapter 6 examines ‘real life’ incidents and applies the Chapter 5 ‘tool kit’ as a clear scientific methodology, identifying likely incident ignition sources and kindling chains. Finally Chapter 7 gives conclusions, identifying the successes and limitations of this work, and points where difficulties were encountered. Areas where further scientific investigation is required are identified.
|
|---|
