System integrity and holistic risk understanding
MetadataShow full item record
- Institutt for marin teknikk 
The objective of an offshore topside process facility is to maximise production of the clean and marketable products: oil, natural gas and condensates. Safe and reliable operation is an absolute requirement to remain competitive in the petroleum industry. Therefore, to mitigate the risk of costly production upsets, process equipment that is subject to deterioration must be adequately maintained during its life cycle.Several analyses techniques and risk assessment methods are used as a basis for maintenance planning and prioritisation. However, these analyses are mostly static in nature, and are therefore unable to account for dynamic information that contributes to changes in the risk picture. Moreover, integration between risk assessments and maintenance management systems is limited.A separator is used as a case study to exemplify the main elements in the thesis. Internal maintenance and modifications of enclosed vessels, such as separators, require shutdown. Additionally, it is also challenging to accurately determine a separator?s current technical condition during operation. In many cases, man entry of the vessel is the only possibility to verify the internal state. However, condition monitoring methods can capture physical quantities that may indicate the current technical condition. Qualitative evaluations by operation and maintenance personnel may also be used to verify the measurements. In order to ensure comprehensible communication of risk between management and personnel involved in daily operation, a state classification system may be used to describe deterioration of equipment and systems.Planned production stops such as turnarounds, and unplanned shutdowns due to unexpected events are costly affairs. In general, only ?must do? maintenance and modification task are performed at these stops to minimise the production loss. However, by having a holistic overview of the technical integrity of the platform equipment, operational downtime can be utilised more constructively. If equipment is expected to fail before the next scheduled shutdown, unplanned downtime can be utilised to prevent these failures before they occur. However, a structured approach to dynamic risk assessment is required to obtain the necessary state information.The purpose of this study is to illuminate the elements in the maintenance management process where state information can be utilised to a greater extent, in order to facilitate a proactive approach to maintenance planning and prioritisation. Thus, it is suggested how dynamic information can be captured, updated and utilised for maintenance purposes. A conceptual model is developed to assist in answering the following questions:? Given the current equipment state, what is the appropriate action?? Based on the deterioration trend, what is the estimated time to failure?? Is it better to perform corrective actions at first opportunity or to wait until turnaround?? If several equipment units need corrective measures at the forthcoming turnaround, which ones have first priority? The aim of the model is to provide operational decision support, as well as continuous maintenance prioritisation of equipment and failure modes. A key enabler for the model is holistic risk understanding, which can be obtained by combining static preliminary studies such as FMECA, with on going assessments that indicates the equipment?s current technical condition. Based on condition monitoring data, prognostic estimates of time to failure can be performed. The thesis provides suggestions of how trend extrapolation may be used for such purposes. Aggregated deterioration estimates and consequence classification may assist in ranking failure modes and equipment by maintenance priority. Additionally, diagnostic procedures may identify root causes that trigger abnormal parameter behaviour, and thereby facilitate informed decision-making for operation and maintenance tasks. Ultimately, this thesis proposes a proactive methodology where state information enables shutdowns to be treated as opportunities to reduce the total amount of unplanned downtime.