In the nuclear power industry, several design-basis accidents are critical for nuclear power plant design and licensing, with loss-of-coolant accidents (LOCAs) being particularly significant for ensuring safe shutdown, emergency cooling, and adequate containment systems. In CANada Deuterium Uranium (CANDU) reactors, a large-break LOCA causes an immediate power surge due to rapid voiding and the positive void reactivity coefficient, with break location greatly influencing severity. Inlet piping breaks, for example, can cause flow reversal, higher voiding rates, or flow stagnation. Conservative assumptions like double-ended guillotine breaks ensure bounding analyses, but for certain metrics (e.g. CANDU fuel channel integrity), partial inlet breaks may be more restrictive, necessitating critical break searches. Break size is crucial in determining mass loss, reactivity, heat deposition, and post-LOCA cooling, impacting severity and mitigation strategies. The Dynamic Probabilistic Safety Assessment (D-PSA) CANDU LOCA pilot aims to identify the most sensitive parameters in critical scenarios and demonstrate the value of D-PSA for risk-informed methods. Using stochastic generation of input parameters under uncertainty, D-PSA quantifies effective risk mitigation factors. While best-estimate analyses attempt to quantify uncertainty in figures of merit, they often impose restrictive conditions. This study integrates uncertainty analysis with component and human reliability in the D-PSA framework, applying Monte Carlo sampling of TRACE input parameters through the RAVEN framework to evaluate dynamic parameters’ impact and compare results with existing CANDU LOCA studies.