TDDB lifetime projections at operating voltages for backend of line (BEOL) dielectrics have been based on accelerated testing at high fields and extrapolation to operating conditions based on electric field dependent dielectric wear-out models. A major complication in this approach is the impact of line-edge roughness (LER) which can lead to inaccuracies in lifetime projection due to creation of electrical field in-homogeneities in the test structure as well as non-uniform charge percolation effects. This paper shows a more robust approach of lifetime projection to operating voltages that uses as input experimentally measured LER statistical characteristics, an assumed field acceleration model and percolation model. The veracity of the assumed acceleration model is verified post priori by its ability to fit or describe the data set in entirety. The paper studies the impact of LER when the underlying acceleration model is Root-E (RE) as well as Power-Law (PL) and shows that the power-law model exhibits unique characteristics with respect to LER impact due to an absence of a characteristic field in power-law.