With the purpose of achieving the precise and quantitative damage monitoring in the concrete medium, the methodology based on the stress wave has secured its popularity and utilized prevailingly. This study introduces a modified delay-and-sum (DAS) damage localization method and a mesolevel concrete numerical model that incorporates the inherent randomness of polygon aggregates distribution. In order to realize the localization of multiple impairments within concrete, the propagation characteristics of stress waves are leveraged to present a damaged subarea determination strategy. This strategy serves as an initial processing stage for damage localization by discerning the effective sensing paths. Additionally, the proposed damage probability distribution function integrated with damage information obtained from the effective sensing paths are employed to accurately pinpoint the location of damage in concrete. Moreover, concrete is comprehensively regarded as a three-phase composite material consisting of aggregate, mortar, and the interface transition zone. By employing mesolevel numerical models, the intricate interaction between stress waves and the internal structures of concrete is revealed. Subsequently, response signals are extracted from the mesolevel concrete models, and meticulous numerical validations are conducted to assess the efficacy of the proposed method. Furthermore, experimental tests are conducted on a concrete slab equipped with embedded tubular piezoelectric transducers. The results demonstrate that the proposed method achieves accurate damage localization with minimal artifact and interference area, surpassing the capabilities of conventional DAS methods. This study validates the potential and feasibility of the proposed method in accurately locating multiple impairments within concrete.