<![CDATA[Fuel distribution to Diesel Power Plants (PLTDs) in archipelagic regions faces significant logistical challenges due to dispersed demand locations, limited vessel capacity, and varying fuel consumption. In Raja Ampat Regency, the existing round-trip distribution system results in low vessel utilization and high transportation costs. This study proposes a two-phase optimization framework integrating the Clarke and Wright Savings Algorithm (CWSA) with inventory simulation and sensitivity analysis to improve distribution efficiency and ensure operational feasibility. Secondary data obtained from PT PLN (Persero) UP3 Sorong include monthly fuel demand, storage tank capacity, daily fuel consumption, vessel capacity, and inter-island distance matrices. The CWSA was applied to generate feasible distribution routes under vessel capacity constraints, while inventory simulation validated delivery schedules based on safety stock, storage capacity, and delivery intervals. The proposed approach reduced the total sailing distance from 6,147.14 to 2,680.30 nautical miles, representing a 56.4% reduction, increased average vessel utilization to 90.75%, and decreased operational distribution costs by Rp346,684,000 per planning period. Furthermore, inventory simulation confirmed continuous fuel availability above safety stock levels throughout the 30-day planning horizon. These results demonstrate that the proposed framework provides an effective decision-support tool for improving fuel distribution efficiency in remote archipelagic regions.]]>
