Over the course of the last four decades, the rotor diameter of Horizontal Axis Wind Turbines (HAWTs) has undergone a substantial increase, expanding from 15 m (30 kW) to an impressive 240 m (15MW), primarily aimed at enhancing their power generation capacity. This growth in blade swept area, however, gives rise to heightened loads, stresses and deflections, imposing more rigorous demands on the structural robustness of these components. To prevent sudden failure and to plan effective inspection, maintenance, and repair activities, it is vital to estimate the reliability of the rotor blades by considering all the forces (aerodynamic and structural dynamics) acting on them over the turbine's lifespan. This research proposes a comprehensive methodology that seamlessly combines fluid-structure interaction (FSI) simulation, Kriging model/algorithm and Adaptive Kriging Monte Carlo Simulation (AKMCS) to assess the reliability of the HAWT rotor blades. Firstly, high-fidelity FSI simulations are performed to investigate the dynamic response of the rotor blade under varying wind conditions. Recognizing the computationally intensive nature and time-consuming aspects of FSI simulations, a judicious approach involves harnessing an economical Kriging model as a surrogate. This surrogate model adeptly predicts blade deflection along its length, utilizing training and testing data derived from FSI simulations. Impressively, the Kriging model predicts blade deflection 400 times faster than the FSI simulations, showcasing its enhanced efficiency. The optimized surrogate model is then used to estimate the flap wise blade tip deflection for one million wind speed samples generated using Weibull distribution. Thereafter, to evaluate the reliability of the blades, statistical modeling using methods such as Monte Carlo Simulation (MCS), AKMCS is performed. The results demonstrate the faster convergence of AKMCS requiring only 21 samples, as opposed to 1 million samples for MCS with minimal reduction in the precision of the estimated probability of failure (Pf) and reliability index (β). Demonstrated on the backdrop of an IEA-15MW offshore reference WT rotor blade, the proposed methodology underscores its potential to be seamlessly incorporated into the creation of WT digital twins, due to its near real-time predictive capabilities for Pf and β assessments.