Predictive accident modeling approach inrelation to workover systems
MetadataVis full innførsel
Hydro carbon releases are the main contributor to the major accident risk on oil and gas platforms, and the Petroleum Safety Authority Norway (PSA) has thus set a target for reducing such releases. Traditionally topside equipment has been the main focus of study in risk analysis, but to obtain the reduction goals it is important to focus on drilling and well intervention as well. This is due to the complexity of the systems, and the lessons learned from several accidents during such operations. To improve the reliability of offshore operations it is essential to focus on safety barriers installed to prevent hydrocarbon releases, and find methods to control and monitor such barriers. The system hazard identification, prediction and prevention (SHIPP) methodology is a new process accident model proposed by S. Rathnayaka, F. Kahn and P. Amyotte in 2011, with properties to analyze safety barriers and predict their performance. The main objective of this report is to give recommendation related to the applicability of the SHIPP methodology on a workover operation to improve management of risk during such operations. This report is based on a literature study on the SHIPP methodology and related methods, and a study of a workover procedure, previous workover accidents and related articles and standards. The SHIPP method is studied to obtain a good understanding of the approach to perform a critical assessment and apply it on a complex operation. The study of workover operations/systems are mainly executed to obtain input to the SHIPP method. Applying the SHIPP method on a complex workover operation proved to be a challengedue to the constantly changing barrier picture and lack of data. To only include one barrier configuration the sub-operation tubing retrieval is considered in the case study,and are chosen due to the related high. The first part of the SHIPP analysis was to identify potential hazards and accident scenarios. The main identified contributors to the risk during the operation are swabbing or trapped gas resulting in a kick when the tubing is pulled out of the well, operational challenges related to the amount of equipment and interfaces with other systems/operators, BOP failures and loss of hydrostatic head. The philosophy of the SHIPP methodology is that accidents often are a result of a sequence of failures occurring during a short period of time, and propose a generic accident sequence model with performance barriers to control and mitigate theescalation of abnormal events. An accident sequence with performance barriers are constructed for the workover operation based on the initial hazard analysis, and the safety barriers included in the rig-up for the heavy workover operation are allocated to the different performance barriers. Further the SHIPP method use fault trees and event trees to predict the system performance based on the hazard analysis and observed abnormal events in the system. In the case study fault trees are constructed for the first performance barriers in the accident sequence (release prevention and dispersion prevention barriers) based on the literature study and hazard analysis. The result includes human and operation factors, mechanical factors and latent conditions. Asimilar hazard study is performed for the remaining performance barriers, but fault trees are not constructed due to lack of detailed information on the top side systems. The hazard identification analysis identifies the BOP and kill mud as the most influentialcomponent with regards to controlling upsets in the well. A deep set plug functioning asthe primary barrier is also a critical component, but has a lower probability of failure duethe low complexity. The BOP is especially important since it is responsible for several functions, and is the only component in place with a shear function to cut the tubing.Thus the BOP is studied in more detail, and the results identify the BOP control system as the main contributor to BOP failures. The control system is an essential part in avoiding accidents, and should be paid special attention with regards to testing, maintenance and monitoring during operations. The objective of the SHIPP method is to predict a system’s performance by studying accident precursors, and prevent incidents rather than relying on control and mitigation.This is obtained by predictive calculations and updating mechanisms based on fault and event trees and recorded observation from the system. In the case study the predictive calculations are performed based on recorded incidents collected by the Health and Safety Executive (HSE) in the UK due to lack of specific observations on the studied operation. The results can thus not indicate the performance of the system in the case study, but are included as an example on how the calculations are performed. Further updating was not accomplished. The results of this report cannot give a definite conclusion on the application of the SHIPP method, but is a critical assessment of the method, and creates a foundation for further studies. It does however indicate that the SHIPP method could have a positive effect on the safety during workover operations, though some drawbacks are identified.The main concerns being: lack of coverage on occupational and external hazards, lack of data on workover systems in general and the complex and changing barrier picture. The method does however evaluate a wide range of potential hazards, and cover a whole accident scenario from the initiating event to damage control and emergency response. Another advantage is the structure of the barrier model which provides a goodperspective of the barriers, though the split between performance barriers may cause interfaces between the barriers to be overlooked. The predictive ability is the main new aspect in the SHIPP method, and improves the focus on accident precursors and recording of abnormal events. The clear definitions of abnormal events and the concrete use of the recordings improve the surveillance of the safety systems, and ensure better follow-up of the identified accident precursors. In relation to workover operations the focus on barrier performance and collection of data could be beneficial, and though the method is not fully applicable on such operations, the mind-set could have a positive effect on the reliability by better monitoring of the safetylevels in the operational phase and a better overview over the safety functions and possible hazards. It is however better fit to assess complex systems working continuously than assessing a complex and continuously changing operation. The main outputs from the case study are identified potential hazards during the operation and the framework to continue the SHIPP process, which could be useful input to other analysis, or a foundation for further work on the SHIPP method.The focus further should be on collecting and estimating data to utilize the method as intended. This is not only beneficial for application of the SHIPP method, but is necessary to be able to improve the reliability during workovers in general.