The UK sector of the North Sea is a mature oil and gas basin subjected to some of the harshest offshore environments with a majority of the oil and gas installations approaching or having exceeded their original design life, often specified as 25 years. It is likely that the operation of these installations will continue for a substantial period in the foreseeable future. However, the ageing nature of these installations present significant challenges to the delivery of high standards of health and safety required by the UK Health and Safety Executive. The issue of ageing installations has been shown to be an important factor in offshore incidents and accidents, leading to an increased risk of accidental loss of hydrocarbon and failures due to equipment deterioration. Two major hazards resulting from ignition of accidental hydrocarbon release are fires and explosions. Failure to minimise the effects of fires and explosions can bring about significant damage to the structural integrity of offshore installations and pose a risk to personnel safety as evidenced by the 1988 Piper Alpha Disaster and the 2010 Deepwater Horizon Explosion and Oil Spill. This thesis presents a practical tool that can be used to predict the costs, risks and service reliability of any given asset management policy for an offshore oil and gas installation. The tool is implemented using a standard Petri Net technique with already adopted and newly proposed high level extensions, and fluid flow modelling technique. The tool is further divided into two sub models that work in conjunction with one another: (1) the Petri Net based Offshore Safety System Degradation and Maintenance Model and (2) the Offshore Fire and Explosion Model based on fluid flow modelling techniques. The aim of the Offshore Safety System Degradation and Maintenance Model is to concurrently simulate the degradation, failure, inspection and maintenance of four safety systems which includes the fire and gas detection system, process isolation, process blowdown, and the deluge system. Simulation of the model generates a variety of statistics such as the estimated operational costs and unavailability associated with implementing any given asset management policy. The Offshore Fire and Explosion Model is then used to model the occurrence of a hydrocarbon leak from a process vessel located within three enclosed modules; wellhead, separation and compression, of an offshore installation. The aim of this model is to predict the frequencies of fires and explosions in the event that the safety systems previously modelled in the Offshore Safety System Degradation and Maintenance Model fail to function on demand in the presence of an ignition source. The model utilises fluid flow modelling to calculate parameters such as the hydrocarbon discharge rate, gas cloud build-up and dispersion, oil-pool build-up and reduction. These parameters can then be used to predict the magnitude of the fires and explosions in terms of the flame length produced in the event of a fire and the overpressures generated in the event of an explosion. The results and statistics generated are highly beneficial to offshore asset operations managers as they can be used to predict the number of maintenance interventions necessary to ensure safety systems are in an acceptable condition. From this, associated costs can be determined enabling offshore managers to allocate resources and budget accordingly. Finally, an optimisation study is carried out using Genetic Algorithm to identify the optimum inspection, maintenance and repair strategy for the offshore safety systems with an acceptable risk level. The methodology presented in this research considers the offshore safety systems and the processes described above in more detail compared to previous literature associated with asset management offshore oil and gas installation. Additionally, the research demonstrates the suitability of Petri Nets for integrating fire and explosion modelling within the asset management framework which is first of its kind. The model can be successfully used to predict costs, risks and service reliability, and to support asset management decisions when the model is implemented in an optimisation framework.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:748271 |
| Date | January 2018 |
| Creators | Dsouza, Serena Karen |
| Publisher | University of Nottingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://eprints.nottingham.ac.uk/49282/ |
Page generated in 0.0018 seconds