South KoreaEnergy and utilities2026 – ongoingKorea Gas Corporation (KOGAS)
AI LNG send-out optimization in KOSPA
KOGAS reported supply forecast errors in the 3% range after adding AI to KOSPA
Problem
Controllers relied on experience to adjust LNG supply. More variable gas-fired generation made these decisions harder, while confidential generation plans limited access to demand data.
Approach
KOGAS added a model trained on historical operations to its KOSPA pipeline analysis system. It calculated send-out volumes for each terminal against production costs and pressure requirements. Regional generation plans supplied through an existing interagency system improved forecasts without disclosing individual plants' plans.
Results
Takeaway
Access to aggregated generation plans was part of the forecast improvement. Controllers used the model's scenarios to support decisions.
Categories
- Industry
- Energy and utilities
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLOptimization
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The report described KOGAS's deployed optimization model and interagency data sharing. It attributed lower forecast errors to incorporating generation plans; financial savings remained forecasts.
Full details are in the original sources.