Armor damage due to wave attack is the principal failure mode to be considered when designing conventional mound breakwaters. Armor layers of mound breakwaters are typically designed using formulas in the literature for non-overtopped mound breakwaters in non-breaking wave conditions, although overtopped mound breakwaters in the depth-induced breaking wave zone are common design conditions. In this study, 2D physical tests with an armor slope H/V = 3/2 are analyzed in order to better describe the hydraulic stability of overtopped mound breakwaters with double-layer rock, double-layer randomly-place cube and single-layer Cubipod® armors in depth-limited breaking wave conditions. Hydraulic stability formulas are derived for each armor section (front slope, crest and rear slope) and each armor layer. The front slope of overtopped double-layer rock structures is more stable than the front slope of non-overtopped mound breakwaters in breaking wave conditions. When wave attack increases, armor damage appears first on the front slope, later on the crest and, finally, on the rear side. However, once the damage begins on the crest and rear side, the progression is much faster than on the front slope, because more wave energy is dissipated through the armored crest and rear side.

Mound breakwater design is evolving owing to rising sea levels caused by climate change and social concern regarding the visual impact of coastal structures. The crest freeboard of coastal structures tends to decrease while overtopping hazard increases over time. Pedestrian safety when facing overtopping events on coastal structures has been assessed considering the overtopping layer thickness (OLT) and overtopping flow velocity (OFV). This paper proposes a new method to estimate the OLT and OFV on mound breakwater crests during extreme overtopping events, based on 123 2D small-scale physical tests of conventional low-crested mound breakwaters with a single-layer Cubipod® and double-layer rock and cube armors. The new method to estimate OLT exceeded by 2% of incoming waves is based on formulas given in literature for dikes, but adapted and calibrated for mound breakwaters. The formula to estimate the OFV exceeded by 2% of incoming waves is based on the correlation between the statistics of the OLT and OFV, considering an empirical coefficient calibrated for each type of armor layer. Exponential and Rayleigh distribution functions are proposed for estimating the OLT and OFV with exceedance probabilities under 2%. Although the statistics of OLT and OFV depend on similar variables, contrary to intuition, specific OLT and OFV corresponding to the same overtopping event appear to be independent.