Research Projects

India’s nuclear power program with a current operational 25 reactors across seven nuclear power plant (NPP) sites, and an installed capacity of 8.8 GW, aims to reach 100 GW by 2047 by adding few more reactors (PIB 2025). Three coastal NPPs—Tarapur Atomic Power Station (seismic Zone III), Madras Atomic Power Station (MAPS, seismic Zone III), and Kudankulam Nuclear Power Station (seismic Zone II)—and two planned sites at Kovvada (seismic Zone II) and Jaitapur (seismic Zone III) are in seismically active coastal belts. Despite stringent seismic design standards, cascading events, such as those causing coolant loss in historical nuclear disasters (Three Mile Island, 1979; Chernobyl, 1986; Fukushima Daiichi, 2011), pose risks of core meltdown and radiation exposure. The Fukushima disaster, triggered by the 2011 Tohoku earthquake and tsunami, led to radioactive releases, evacuating over 100,000 residents and incurring costs of approximately $500 billion USD. The current INDHAN project proposes integrating offshore wind turbines (OWTs) as emergency backup power to enhance the seismic resilience of coastal NPPs. Inspired by the seismic resilient performance of Kamisu OWTs during the 2011 Tohoku event, the project leverages OWTs’ displacement-sensitive dynamics (Bhattacharya and Goda 2016). Recent research by the lead PI team (Kolli et al. 2023) advocates OWTs as backup power for MAPS at Kalpakkam, southern India. The INDHAN project employs advanced digital twin (DT) technology to develop reliability-centric OWT designs, ensuring robust coolant power supply to mitigate seismic risks. OWTs’ tall, flexible structures enhance dynamic response but face challenges from soft seabed soils prone to liquefaction, seismic risks near the “Ring of Fire,” cyclones, tsunamis, corrosion, fatigue, and tower instability. DT models address these by integrating real-time sensor data with AI/ML and Physics-Based Machine Learning (PBML), achieving 97% accuracy in predicting failures like fatigue and corrosion. Bayesian updating and surrogate models, such as Gaussian Process, reduce computational costs by 70%, managing uncertainties in soil properties, environmental loads, and material behaviour. Real-time monitoring, exemplified by Horns Rev 3, enables proactive maintenance, reducing operation and maintenance (O&M) costs, which constitute 30% of the levelized cost of energy (LCOE), ensuring reliable backup for NPPs. Structural reliability is assessed using non-linear limit state functions, per DNV-ST-0126 (2016), encompassing Ultimate Limit State (ULS), Fatigue Limit State (FLS), Serviceable Limit State (SLS), and Accidental Limit State (ALS). Simulation-based Structural Reliability Analysis (SRA) employs high-fidelity 3D-Finite Element Models (3D-FEM), requiring approximately 50,000 time-history simulations for comprehensive loading scenarios (Vorpahl et al., 2013), which increases computational costs. Simplified models (Van der Tempel, 2006) and reduced load cases (Kühn, 2001) compromise accuracy. Surrogate models, including Kriging, Polynomial Chaos Expansion, and Neural Networks, balance accuracy and efficiency (Alizadeh et al., 2020). Predictive accuracy depends on the surrogate model’s fidelity, while computational efficiency is influenced by analysis methodology, experimental design, parameter space dimensionality, and training data volume. The INDHAN project develops a DT framework for real-time OWT reliability analysis, integrating seismic, geologic, and environmental data to evaluate resilience seismic events and corelated impacts. By pioneering DT-driven OWT-NPP integration tailored to India’s coastal challenges, INDHAN enhances NPP safety, supports India’s 30 GW offshore wind target, and establishes a global benchmark for nuclear resilience using sustainable wind energy.