Optimal configuration of supply logistics for remote oil and gas fields

The objective of this thesis has been to compare alternatives to improve the supply service for remote locations offshore. First the alternative of running a conventional supply service has been assessed, where the platform supply vessels (PSVs) sailed directly from a land depot to supply the installations with commodities. The alternative to the conventional supply service was to have two converted cargo ships in use as storage units located at the oil and gas field which deliver cargo to the PSVs. For the alternative with the storage units the ships switch position, when one is in operation the other one goes to shore and stock up on supplies. PSVs are expensive and if one can reduce the number of ships in operation it will grant great savings for the operator of the oil and gas field. Having storage units located in the proximity of the installations offshore will reduce the sailing distance of the PSVs to a fraction of the original sailing distance from a land depot. While the cost of the PSVs will be reduced one has a cost increase due to the storage units. Finding the point at which the storage units can become profitable has been the essence of this thesis. A case study has been made to compare the two different setups for oil and gas activities outside Jan Mayen. It was expected a maximum activity level with three installations in operation. The case study has been made more general by adding up to five additional installations to get a better look at the savings of having the storage units with more installations to service. To minimize the costs of each setup it was made a mathematical formulation for each of them. For the regular setup the minimum cost of routing the supply vessels was found from a land depot, given a required supply frequency of the installations. The port in Kristiansund has been chosen as a suitable land depot for this case study. In the setup with storage units the routing from the storage units with PSVs are considered in addition to the cost of the storage units. To minimize the routing cost from the storage units the mathematical model had to consider different locations of the units. Both of the setups have the same input data as a basis to get a good comparison. The models are solved for a weekly planning period. Some basic design characteristics of the storage units have been made. It was concluded that open hatch bulk carriers would be most suited for the operation. The vessels to be converted should have a deadweight capacity of around 57 300 tons. It is estimated here that the cost of one storage unit will be 53 000 USD per day. This price may not be very accurate and is based on many assumptions due to lack of assessable cost information, but it has been tried to make a conservative estimate. The mathematical formulations of the models have been solved with the optimization software Xpress. For the regular supply service from land the total cost of the supply service ranged from 590 000 USD per week for three installations and up to 1 550 000 USD for eight installations. The results showed that there was a linear increase in cost when increasing the number of installations to supply. Compared to the setup with the converted cargo ships the total costs ranged from 916 000 USD per week for three installations to 1 137 000 USD per week for eight installations. In this case the cost per installation gets lower with the number of installations to supply. When solving the model for the supply service from land one gets larger ships in operation which can visit as many installations as possible on a route as they can only make one trip per week. Compared to the case with the storage units one gets a few smaller PSVs in operation which can sail many routes per week. It has been found that the concept with the storage units would be profitable with 6 installations or more to service in this case. With less than 6 installations to service it would be more expensive to have storage units than running a regular supply service from land. Generally it can be concluded that having the storage units could be profitable if there is enough installations to service. The other factor with the biggest effect on the problem is the distance to shore, however, finding at what distance from shore the storage unit becomes profitable was not included in this study. The final conclusion of the work here is that the use of storage units could reduce the cost of the supply service, and is an interesting concept that should be studied further