Central AsianResearch Index

ECONOMIC AND LEGAL MODEL FOR IMPLEMENTING ARTIFICIAL INTELLIGENCE IN NUCLEAR POWER PLANT EQUIPMENT MAINTENANCE

An economic and legal model is developed to support management decisions on implementing an artificial intelligence (AI)-based advisory system for predictive diagnostics of rotating equipment at a nuclear power plant (NPP). The model links functional boundaries, data governance, independent validation, computer security, human oversight and change management requirements with life-cycle costs and investment indicators. An integrated realized-performance coefficient is proposed to account for model quality, equipment coverage, data availability and personnel use of recommendations. The assessment is performed for a hypothetical water-water energetic reactor unit rated at 1200 MWe (VVER-1200) in constant 2026 prices. In the baseline scenario, the project’s net present value (NPV) is USD 2.648 million, its internal rate of return (IRR) is 29.33%, and the discounted payback period is 4.90 years. The break-even realized-performance coefficient is 43.32%. The reduction in the levelized cost of electricity (LCOE) is USD 0.075/MWh and should be interpreted as a localised operational effect rather than a radical improvement in NPP competitiveness. A staged implementation approach is justified, with scaling allowed only after technical, legal, and economic criteria have been confirmed using pilotoperation data.

Сведения

Журнал
NNC RK Bulletin
Опубликовано
2026 · № 3, С. 204-209
Язык
Английский
ISSN
1729-7516 · 1729-7885
Издатель
Republican State Enterprise "National Nuclear Center of the Republic of Kazakhstan
Аффилиации
Kazakhstan Nuclear Power Plants, LLP