An Exploratory Study of a Wet Bulk Platform Supply Vessel using Operations Research - Single-Vessel Scheduling with Bulk Loads and Mud-Cleaning Process Plant

In order to deal with an increased demand for wet bulk on the offshore installations on the Norwegian continental shelf, a new system for handling wet bulk logistics is investigated in this thesis. The reason for the increased demand is because the production rate on an installation must be maintained as the field matures, which is done by injecting produced water into the field. Additionally, operators experience that large and unexpected cargo demands are hard to implement on pre-scheduled platform supply vessels (PSV), forcing them to use expensive, spot chartered PSVs. This thesis covers this problem in the North Sea, with offshore installations connected to Base Mongstad. The new PSV introduced in this thesis is dedicated to wet bulk operations, and has a process plant on board to clean oil-based drilling mud (OBM). There are several, equally sized tanks on the PSV that are not dedicated to carry a certain type of product, unlike ordinary PSVs. This tank philosophy makes the vessel more flexible and only limits the amount of a certain cargo it can carry to the vessel's capacity. The wet bulk supply problem is characterised by several cargoes that need to be delivered to different offshore installations from the onshore base, and some cargoes that must be transported back from the installations to the offshore base. A maritime pickup and delivery model is thought to best describe the problem, and a suiting mathematical model is thus build. A study of state of the art shows that there is performed little or no research on scheduling a PSV solely after the wet bulk demands. Therefore this thesis contributes to insight into how optimisation can be used to schedule a PSV with the aim to avoid delays. The objective function in the mathematical model is ambiguous as it minimises delay and maximises the number of available tanks at the same time. However, it is the delays at each cargo and the flexibility of the vessel in number of available tanks that are of interest. The mathematical model is implemented and run in the commercial solver FICO Xpress Optimization Suite. Different cases divided into two main sections, demand today and increased demand, are developed to investigate the capacity of the wet bulk PSV in a case study. Relevant data is identified and collected primarily from actors in the Norwegian petroleum industry to best suit the thesis problem. Results from the case study indicate that the introduced wet bulk PSV has capacity to handle large amounts of cargo, which implies that it can stay offshore for a long time. The results also show that there is no difference in letting the wet bulk PSV and an ordinary PSV handle small, frequent cargoes. It is therefore proposed that the wet bulk PSV should primarily handle large cargo demands and treatment of OBM. Treating the OBM offshore is the cause of higher delays in the results, but the delays are also explained by poor input in form of time of demand. The OBM process plant allows the PSV to stay offshore for more than a week, which makes the vessel obviate two to three returns back to base. By staying offshore for such amounts of time it can operate as a bank, holding and delivering some products that might be unexpectedly wanted at an installation. Such a vessel can save an operator money in terms of less PSVs chartered from the spot market and less fuel consumption connected to less returns to base, but also more time gained for the vessel to be available for service