Abstract
The aquaculture industry is one of Norway’s most significant export sectors. Most of the production
takes place at sea-based production facilities along the coast. These facilities rely on specialized
service vessels to carry out maintenance and repairs, ensuring both structural integrity and animal
welfare. Over the coming decades, the industry is expected to grow substantially. Meanwhile, to
overcome environmental challenges related to global warming, there is a need for a rapid reduction
in greenhouse gas (GHG) emissions from the industry, including those from the fleet of service
vessels. This reduction could incur large costs, making it crucial to find cost-efficient ways to
reduce emissions.
Since the climate impact of GHG emissions is determined by cumulative emissions over time, early
implementation of reduction measures is crucial. Many vessels used in aquaculture are already
equipped with battery packs, but the electric capacity is not optimally utilized.
This research introduces the Hybrid Electric Aquaculture Service Vessel Routing Problem (HEAVRP),
which involves routing a heterogeneous fleet of hybrid electric service vessels between fish farms to
perform a set of service tasks. These vessels can operate on both conventional diesel and electricity
from their installed battery packs. Batteries may be recharged at designated charging stations or
at fish farms equipped with electrical charging infrastructure, enabling emission-free operations for
parts of the service routes while retaining flexibility through conventional diesel propulsion.
The HEAVRP involves determining the routing of the fleet to the different service tasks, as well
as charging schedules and energy source allocation (diesel vs. electricity) to minimize variable
operational costs while maximizing task coverage. An arc-flow model formulation is proposed.
While small problem instances can be solved using a commercial solver, larger instances require an
alternative approach. To solve realistically sized instances, a Column Generation Heuristic (CGH)
is proposed, where promising routes are generated a priori and a path-flow model is used to find
the optimal combination of routes.
A computational study based on realistic data is conducted, considering both a realistic current
fleet and potential future fleets with varying battery capacities, reflecting expected technological
advancements. Scenarios with different levels of charging infrastructure availability are also analyzed
to assess the impact of infrastructure development on operational efficiency and emissions. Results
from solving the HEAVRP under increasingly strict constraints on fleet emissions indicate that the
cost of emission abatement is substantial. However, increasing both battery capacity and charging
infrastructure availability helps alleviate these costs.