Why efficiency measures don't always have the effect you'd expect
What simulation research into complex tanker rotations reveals about capacity, coordination, and the limits of local optimization.
A terminal that loads and unloads faster looks like a win for the whole system. Simulation research into parcel tanker rotations across multiple terminals shows that this isn't a given, and that the real solution more often lies in coordination than in local improvement.
One ship, multiple terminals, one system
Parcel tankers typically don't carry a single cargo to a single terminal. They carry multiple part-cargoes across several terminals within the same port area. That turns the planning of their rotations into a system-level question: the performance of one terminal is tied to what happens at the others, and to the behavior of the ships shuttling between them. Researchers from Erasmus UPT and Systems Navigator built a simulation model to study this interaction.
Why chemicals, and why a generic model
The research focuses on chemical products. Oil tankers typically carry a single product to a single terminal and deal with price-driven waiting time at anchor, which makes multi-terminal complexity less visible. Chemical transport, where vessels usually call at several terminals in the same port area, is better suited to studying these rotations. The model was deliberately kept generic, inspired by existing shipping patterns but not built around one specific port.
What the model represents
Two clusters are connected by a fleet of parcel tankers: a smaller western cluster of four terminals and a larger eastern one of seven, with eight vessels circulating between them, each with twenty-eight separate cargo tanks. That asymmetry between the clusters is deliberate, because it is what makes the system's response to change interesting. The base case was calibrated against interviews with operators and shipping companies until its output sat in a realistic range: vessel utilization just under 89 percent and an average parcel delivery time of 29.4 days.
Growth is manageable, a shifted balance is not
The system absorbs moderate volume growth reasonably well. At 10 percent more volume, vessel utilization rises while pressure on the system stays limited, though pressure on individual terminals rises faster. Push further and it runs into a wall: at 30 percent growth with an unchanged fleet, utilization has hit its ceiling, and no slack is left anywhere. What is striking is how easily that wall moves, since adding a single ship, from eight to nine, restores the balance almost at once. That is easier said than done, though: terminals and shipping companies are independent parties, and if a ship is drawn to one port to relieve pressure, there is a ship not sailing somewhere else.
The more striking result concerns not how much cargo moves, but where. The base case assumes an even split between the two clusters. Move it to 55/45 and the western cluster needs around 30 percent more storage capacity. At 60/40 the requirement roughly doubles, at 65/35 the system breaks down altogether. A ten-point shift in market share between two port clusters is hardly exotic in a sector reorganizing around the energy transition, yet it strains terminal capacity far more than a comparable rise in total volume would.
Faster isn't always better for the system
The second finding may be the most surprising. Measures that appear to improve efficiency at one point, shorter transfer times between terminals, shorter preparation time for loading and unloading, do reduce pressure at that point. However, they do not always reduce pressure across the system, and sometimes the pressure simply shifts elsewhere. This helps explain why the sector is cautious about high expectations for standalone efficiency projects: local improvement doesn't guarantee a system-wide one. (Quote uit interview, nog van Rienk krijgen)
Take the time vessels spend moving between terminals within a port area. Cutting the additional time allowed for that movement from four hours to none does what you would expect at first glance: ships pass through the port faster and system-wide storage requirements fall. But the two clusters respond differently and not in proportion, the western requirement first dropping and then rising again while the eastern rises throughout. The ships themselves do slightly worse, since moving through the port faster leaves less opportunity to pick up newly generated parcels, so utilization falls from just under 89 percent to 84 percent. Innovations that save time genuinely do save time; how that saving propagates through the rest of the system is neither linear nor reliably predictable.
What does work? Coordination between terminals
The researchers first tested the opposite. Adding terminals to the eastern cluster leaves system-level indicators largely unchanged, but terminal by terminal the storage requirement goes up, because spreading the same flow across more parties means each holds buffer for cargo it may or may not receive. Fragmentation is mildly good for the system and bad for the terminals within it, precisely the configuration in which no party has an incentive to fix anything.
What the research does point to is coordination between terminals, ranging from actual consolidation to better-aligned planning. The effect on storage is substantial: replacing the four western terminals with a single one drops the western storage requirement from 196,000 to 122,500 cubic meters, while average parcel delivery time falls from 29.4 to 27.7 days. It is not free, since vessel utilization drops and parcels wait longer, so the gain at the terminals is partly paid for by the ships.
The core finding: ships and terminals are inseparably linked
The thread running through the research is the interdependence between ship and terminal operations. In the open market, neither party has structural control over the other: ships don't coordinate which terminal they call at, and terminals don't determine which ships visit them. That is exactly why simulation is needed to understand what happens when one link in the system changes.
Workshop at Erasmus!
Systems Navigator and Erasmus UPT are preparing a workshop later this year to explore the research findings and their practical implications together with participants from the sector. Want to stay informed about the workshop, or learn more about how simulation can make system interactions visible in your own logistics chain? Contact Systems Navigator.
Comments