Rainwater accumulation in the cowl region, located at the base of the windshield, can lead to serious HVAC performance degradation, corrosion, and passenger discomfort if not effectively drained. Traditional physical validation methods are often time-consuming, costly, and limited in diagnostic insight. This study presents a simulation-driven methodology for evaluating and optimizing HVAC cowl box drainage performance during the early design phase. Using STAR-CCM+, a multiphase Volume of Fluid (VOF) approach was implemented to visualize water flow behavior under static and dynamic conditions. Design variants were assessed by modifying drain tube geometry (shape, size, and placement) and cowl surface features, such as baffle positioning. Results showed that inadequate drainages were primarily due to stagnation zones, shallow slopes, and drain locations prone to clogging. Water film accumulation near the HVAC inlet was accurately predicted, highlighting potential ingress paths under high water load scenarios. The optimized configuration demonstrated improved drainage performance and reduced risk of water ingress. This methodology enables proactive identification of failure zones, supports robust design iteration, and reduces dependency on late-stage physical testing. Overall, the proposed simulation framework contributes to improved vehicle water management and long-term HVAC system reliability.