Why port planning fails without simulation
Container terminals are among the most complex logistics systems in the world. Vessel arrivals are stochastic — subject to weather, port congestion and carrier schedule changes. Quay crane productivity depends on vessel stow plan quality, hatch sequencing and the availability of yard transport. Yard density affects RTG cycle times, which feed back into berth turnaround, which affects the next vessel arrival window. These feedback loops are invisible in static capacity calculations.
Terminal operators facing a berth utilisation problem often invest in additional quay cranes, only to find the constraint moves to yard transport or gate throughput. Port authorities planning a new terminal extension calculate theoretical TEU throughput from berth length and crane rates, but cannot predict how the extension interacts with existing operations under variable vessel arrival patterns. Simulation captures all of this interaction before the first euro is spent.
The constraint moves: Terminal throughput is determined by the weakest link across berth, yard, transport and gate — and that link changes with traffic mix and volume. Simulation finds it before capital is committed to the wrong intervention.
The terminal zones we model
🚢 Berth & quay
- Vessel arrival & berthing (stochastic)
- Berth allocation strategy
- Quay crane assignment & shifts
- Hatch sequencing & twin-lift productivity
- Vessel turnaround time
🚛 Horizontal transport
- Straddle carrier & reach stacker fleets
- Terminal tractor & chassis pools
- AGV / automated shuttle systems
- Travel time & conflict modelling
- Fleet sizing optimisation
📦 Yard operations
- Block layout & density management
- RTG / RMG scheduling
- Dwell time by container type
- Rehandling rate prediction
- Reefer & dangerous goods areas
🚗 Gate & intermodal
- Truck arrival profile & queuing
- Gate lane throughput & OCR
- Rail terminal interchange
- Barge feeder scheduling
- Pre-gate & truck appointment systems
How a terminal simulation project runs
Terminal data collection
We collect vessel call history (arrival times, LOA, TEU), crane productivity logs, yard equipment fleet data (MTBF/MTTR), dwell time distributions by container category, and truck gate transaction times. AIS vessel data can supplement operational records where available.
Model build & validation
The simulation is built in AnyLogic — the industry-standard platform for port and terminal modelling. We validate the model against 12 months of historical berth utilisation, crane moves per hour and vessel turnaround data before running any scenario.
Scenario experiments
Investment and operational scenarios — additional berths, crane upgrades, fleet changes, gate automation, appointment systems — are tested as automated experiments. Each scenario produces throughput, utilisation, turnaround time and congestion statistics.
Investment brief & model handover
You receive a ranked investment brief with quantified throughput and cost impact per scenario, plus the executable model for ongoing planning use. Terminal planners can reuse it for annual traffic forecasts and new service calls without our involvement.
What you get at the end
Simulation vs analytical terminal planning
| Planning question | Simulation model | Analytical capacity model |
|---|---|---|
| Berth utilisation under bunching | ✓ Stochastic arrivals — bunching captured fully | ✗ Assumes uniform arrival intervals |
| Yard transport fleet sizing | ✓ Integrated with crane and yard cycle times | Estimated separately — no interaction |
| Gate congestion under peak arrival | ✓ Truck arrival profile modelled stochastically | ✗ Average throughput only |
| Expansion investment sequencing | ✓ Each phase evaluated at intermediate traffic volumes | ✗ Only steady-state target capacity |
| New crane type ROI (e.g. twin-lift) | ✓ Productivity gain in context of full terminal cycle | Crane rate only — no downstream effect |
Tools & technology
AnyLogic is the globally dominant platform for port and terminal simulation, used by the world's largest terminal operators and port authorities. Its agent-based engine handles the complex interaction between discrete vessel events, continuous equipment flows and stochastic arrival patterns that terminal modelling requires. For specialised studies — crane scheduling optimisation, barge window planning — we supplement with Python optimisation models that feed back into the simulation.
We have worked with container terminals ranging from 200,000 TEU/year regional ports to multi-berth deep-sea terminals. Models include combined vessel/yard/gate systems and specialised studies of individual terminal zones where a focused answer is needed quickly.