Abstract
This paper considers the offshore supply vessel (OSV) planning problem,
which consists of determining an optimal fleet size and mix of OSVs as well as
their weekly routes and schedules for servicing offshore oil and gas installations.
The work originates from a project with Statoil, the leading operator on the Norwegian
continental shelf. We present both a new arc-flow and a voyage-based
model for solving the OSV planning problem. A decision support tool based on the
voyage-based model has been used by planners in Statoil, and cost savings from this
was estimated to approximately 3 million USD/year. Weather conditions at the
Norwegian continental shelf can be harsh; wave heights may limit both an OSV’s
sailing speed and the time to perform unloading/loading operations at the installations.
Hence, we analyze the weather impact on the execution of a schedule and
propose robustness approaches to obtain solutions that can better withstand delays
due to rough weather. Simulations indicate that such solutions both are more robust
and have lower expected costs.
which consists of determining an optimal fleet size and mix of OSVs as well as
their weekly routes and schedules for servicing offshore oil and gas installations.
The work originates from a project with Statoil, the leading operator on the Norwegian
continental shelf. We present both a new arc-flow and a voyage-based
model for solving the OSV planning problem. A decision support tool based on the
voyage-based model has been used by planners in Statoil, and cost savings from this
was estimated to approximately 3 million USD/year. Weather conditions at the
Norwegian continental shelf can be harsh; wave heights may limit both an OSV’s
sailing speed and the time to perform unloading/loading operations at the installations.
Hence, we analyze the weather impact on the execution of a schedule and
propose robustness approaches to obtain solutions that can better withstand delays
due to rough weather. Simulations indicate that such solutions both are more robust
and have lower expected costs.
