SIMULATION OF RISER DISCONNECTION IN STOCHASTIC WAVES
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2350602Utgivelsesdato
2014Metadata
Vis full innførselSamlinger
- Institutt for marin teknikk [3502]
Sammendrag
Ever since the first oil was discovered in Norway in 1971 the production rates have varied. The development has gone from few wells with high production rates to several wells with lower production rates. It is expected that the production rate will continue to decrease in the future. Taking this into consideration, it is apparent that operations which will enhance the production rate are important. Workover operations are operations where the well is cleaned out, which enhances the production rate. Moreover, Workover operations are conducted with a Workover Riser Systems. The Workover riser system is a long slender structure stretching from the vessel down to the sea bed, which is exposed to loads from the environment and the platform. If a storm is coming and the operational limits are expected to be exceeded, the system has to perform a planned/normal disconnection. The disconnection takes place between the Emergency Disconnect Package (EDP) and the Lower Riser Package (LRP). This is done to prevent damage to the components, especially the well head. The criteria for conducting disconnection are defined in the operation envelopes. When a planned/normal disconnection is performed there are some critical scenarios that should be avoided. On such scenario is collision between the EDP and the remaining structure at the sea bed, the LRP. The scope of this Master Thesis is to study this event, collision between these two components. The purpose is to study the risk of contact between the Emergency Disconnect Package and the Lower Riser Package and propose a strategy for reducing it. To study the collision event the computer program RIFLEX has been applied. Two different models have been developed, both are based on the model applied in the Project Thesis. One model on -348m water depth and one model on -996m water depth. Several analyses and disconnection events have been conducted. The first set of analysis conducted is the different disconnection events. These events are disconnection at heave displacement top, heave displacement bottom, half way up the heave displacement (max velocity) and a random in time. The vertical displacement of the EDP after disconnection is studied. Additionally, a mean study has been performed in order to determine the trend for the vertical displacement after disconnect. Based on the different disconnection events, a probability study has been performed to study the frequency of hits. To enable for the probability study a MATLAB script has been developed. This checks if the EDP is within the limits of the LRP. Moreover, a correlation study has been performed. This has been done to determine if there exists a relation between the vertical velocity 30 seconds in advance of disconnection and at the actual disconnection timing. The reason for studying a 30 seconds time interval is because this is the time it takes for the electrical signal to be sent from the platform down to EDP. Finally, the last set of analyses which have been completed is disconnection analyses with riser lift up. When the disconnection is done the riser is lifted up 2-4 meters. A set of random disconnection events with retraction for both the shallow and deep water model has been conducted. The results from the different disconnection events show that when the disconnection is performed at the heave displacement top, the EDP will experience a negative vertical displacement immediately after disconnection. Simultaneously as the EDP is disconnected the riser is locked to the vessel at the top. For disconnection at heave displacement top it is locked at a higher location in the riser, this leads to a longer effective length of the riser underneath the vessel. Thus, the mean vertical displacement is below the location of the LRP. The results from the mean study show that the trend is an immediate negative displacement after disconnect. Additionally, the probability study shows that disconnection at the heave displacement top, has the highest mean percent of time when the EDP is within the LRP limits. On the other hand, the results from disconnection at the heave displacement bottom show that the EDP will have displacement in positive z-direction, and the mean vertical displacement is above the location of the LRP. The mean percentage of time the EDP is within the limits of the LRP, is zero. When the disconnection is performed in the middle of the heave displacement, the average value of the mean vertical displacement lies at the initial position of the EDP. The mean percent of time the EDP is within the LRP limits is the second lowest. The results from the random disconnection show that the mean percent of time the EDP is within the limits of the LRP, is the second highest. Therefore it can be concluded that the most critical disconnection is the disconnection at the heave displacement top. The second most critical is the random disconnection. However, to avoid collision the most optimal disconnection event is disconnection at the heave displacement bottom.Finally, riser disconnection analyses with a 0-4 meter lift-up have been performed for the deep and shallow water model. Before the disconnection is performed the riser is retracted 2-4 meters to avoid collision. The results from these analyses show that when the riser is lifted up either two or four meters there is a significant clearance between the EDP and LRP. However, the analysis was also done with zero and one meter retraction, which illustrated that the number of hits is lower for the deep water model than the shallow water model. This may have to do with the length of the riser that is exposed to the current. The deep water model will therefore have a bigger horizontal displacement. From these results it can be concluded that if the riser is lifted up a minimum of two meters, it is likely that the EDP will never collide with the LRP.