Quantifying Human-Driven Energy and Cost Variability in Tugboat Manoeuvres

Regulatory and societal pressure to reduce emissions is pushing ports to lower their overall environmental footprint. Tugboats, as integral elements of port operations, will have to contribute to this transition. Quantifying this human contribution in real operations is difficult because weather, traffic and towing conditions continuously vary. In this work we adopt an alternative approach: a controlled full-mission simulator experiment in which the ship model, environment and assignment are kept constant so that the participant is the only remaining variable. The aim is to make the impact of human behaviour on fuel consumption and total cost measurable and coachable. The study used a Kongsberg simulator and a harbour tug called Smit Panama. Twenty-one people took turns doing the same docking task in Bodø, Norway. Everything about the ship and environment stayed the same, so any changes in fuel use were down to how each person operated the tug. Participants were told to focus on safe manoeuvring, not saving fuel. The group included both pilots and tugboat masters. For each run, there was a commander and a helmsman, and the commander spoke out loud about their choices. The engines started at half power and all equipment was available. Every second, important details like speed, heading, engine settings, and fuel use were recorded. Because the participants had different backgrounds, they used various approaches and made different adjustments. After finishing, some said they could do better next time. Since each run was under the same conditions apart from the team driving, differences in fuel use were mainly due to how the people handled the tug. In practice, several pilots required more corrective actions to berth the tug smoothly, while some tug masters found it challenging to issue commands to a helmsman without directly feeling the vessel’s response. As a result, a wide range of approach strategies and correction patterns emerged. During the debriefings, several participants acknowledged suboptimal decisions and indicated that they believed they could improve their performance in a second attempt. These observations suggest that experience and handling style play a measurable role in operational performance. To further quantify the relationship between energy consumption and operator behaviour, the analysis was structured at three levels. Overall voyage indicators: Task duration, integrated fuel consumption, associated CO₂ emissions, and behavioural metrics provide a global benchmark of efficiency and sailing style. The use of behavioural indicators to characterise operational performance builds on established driving-behaviour analysis frameworks developed in intelligent transportation systems (e.g., Nasr Azadani Frequency distributions: The second-by-second fuel-rate data were transformed into frequency distributions of fuel rate, speed, and acceleration. These distributions typically show numerous low values corresponding to idle or slow operation, clusters at intermediate levels linked to specific manoeuvres, and occasional high-power demands. The position of the main peak, the dispersion of the distribution, and the extent of the high-consumption tail reflect how consistently participants maintained stable and efficient power use. Broad distributions and pronounced upper tails indicate frequent power adjustments, delayed corrections, and short bursts of maximum output; Peaks and valleys in time series: Energy-consumption peaks and prolonged speed valleys (idle periods) largely determine total fuel use. What-if scenarios were constructed to investigate the causes of these peaks and valleys and to assess the effect of peak shaving and smoother operation. Once their origin was identified, it was assumed that the operator could have completed the same voyage while avoiding some of these causes, allowing estimation of the potential reduction in fuel use, costs, and CO₂ emissions. The design and technical characteristics of a tugboat clearly influence its energy consumption, but this study demonstrates that the way the vessel is operated also plays a decisive role. By comparing actual fuel use with counterfactual scenarios based on improved handling, we show that relatively simple behavioural adjustments—such as limiting the use of maximum power, maintaining slightly lower average power levels, and operating more smoothly—can lead to measurable reductions in fuel consumption and CO₂ emissions, typically without significantly increasing task duration. The analysis revealed substantial variation attributable to individual handling styles, indicating clear potential for improvement. These findings suggest that behaviour-oriented training, structured performance feedback, and decision-support tools can enhance both the environmental and economic performance of tug operations. Establishing efficient operational habits today will remain valuable as tug fleets transition to alternative fuels in the future.