Issue Building bottom-up demand models for data centres, advanced industry and electrification
Summary
Recent plans made progress by counting additional demand separately, but there is still room to reflect each field's technical traits and uncertainty in more detail. Much technology-driven efficiency arises from market competition and cost incentives, and changes too quickly for macro models or sectoral variables to capture. Semiconductor complex demand in particular rests on firms' investment intentions and could swing with the future chip market. AI data centre and advanced industrial complex demand is still estimated mainly from business plans, contracted power and case data, unlike electrification. The author recommends first building up research and tracking technology indicators, such as accelerator share, cooling type, process power intensity and charging patterns, before developing bottom-up models in stages.
Classification
Evidence 1
- 국회미래연구원 국가 전력수요 예측시스템 진단과 고도화 방향 제언 국회미래연구원 no link — bibliographic entry pp. 17, 18 2026-06 accessed 2026-09-30
Constituent trends 2
Directly linked signals 0
No objects.
Relation types: constitutes
Public id: fm-8cd72767e98a
