EnergyGlobal Gas Transport (name changed)

LNG Carrier Scheduling System

LNG carrier connected to a coastal energy terminal

Case study. Client names are changed for confidentiality. Results are measured against each client's pre-project baseline.

$2.4 M

Annual Savings

Combined reduction in boil-off losses and demurrage fees across the 22-vessel fleet.

3 days → 2 hrs

Planning Cycle Time

Full fleet schedule generated and reviewed in 2 hours versus 3-day manual process.

98%

Contractual Compliance

Up from 91% — equivalent to recovering approximately $1.1 M in penalty avoidance.

6 → 2

Schedulers Required

Four roles redeployed to higher-value commercial and analytical work.

Case studyClient names are changed for confidentiality. Results are measured against each client's pre-project baseline.

The Challenge

Global Gas Transport was managing 22 vessels, 8 liquefaction facilities, and 14 terminal slots across a global network using a patchwork of Excel models maintained by 6 schedulers. Scenario planning was so time-consuming that suboptimal plans were regularly accepted due to deadline pressure, and contractual compliance had slipped to 91%.

Our Approach

1
Mapped and digitised the full constraint set: vessel availability windows, transit speeds, boil-off curves, terminal capacity slots, and contractual delivery obligations.
2
Built a Mixed Integer Linear Programming (MILP) model in Python using Gurobi Optimizer that simultaneously minimises total boil-off loss and demurrage exposure.
3
Developed a React + TypeScript scheduling portal with a drag-and-drop Gantt interface, allowing schedulers to review and run what-if scenarios in seconds.
4
Integrated a Node.js / GraphQL API layer connecting the optimiser to live vessel-position feeds, terminal nomination systems, and the trading desk's contract database.
5
Built automated Apache Airflow workflows that re-run the optimiser whenever a new vessel ETA deviation or terminal slot change is detected.

The Solution

We digitised the entire LNG supply-chain constraint model and solved it as a mathematical optimisation problem. The Gurobi MILP engine evaluates millions of scheduling combinations in under 90 seconds. Schedulers no longer build plans from scratch — they review, adjust, and approve computer-generated optimal plans, freeing their time for commercial negotiation.

Tech Stack

ML / AI

Gurobi Optimizer

Commercial MILP solver — evaluates millions of scheduling permutations in under 90 seconds.

Data Engineering

Python (Gurobipy)

Constraint modelling layer: encodes vessel, terminal, and contractual constraints into the MILP formulation.

PostgreSQL

Master data store for vessel schedules, terminal slot records, contractual obligations, and plan history.

Visualisation

React + TypeScript

Scheduling portal with drag-and-drop Gantt chart, scenario comparison, and one-click plan acceptance.

Power BI

Executive dashboard: contract compliance rate, boil-off trend, and demurrage exposure by trade route.

Integration

Node.js / GraphQL

API layer connecting the optimiser to live vessel-AIS feeds, terminal nominations, and trading contracts.

AIS Data Integration

Live vessel position and speed feeds used to update ETAs and trigger dynamic re-scheduling.

Infrastructure

Apache Airflow

Triggers re-optimisation automatically on ETA deviations or slot changes; orchestrates daily plan publication.

Face a similar challenge?

Send us your challenge. We will reply to arrange a conversation and scope the next step.

Start with the decision

Turn one operational challenge into a measurable data plan.

Tell us about your challenge. We will reply to arrange an assessment and identify a practical route to delivery.

Request a data assessment