A Probabilistic Design Framework for Semi-Submerged Curtain Wall Breakwaters

Semi-submerged curtain breakwaters are increasingly favored to protect marinas and other microtidal basins, yet they are still almost exclusively designed with deterministic wave transmission equations. This study introduces a fully probabilistic design framework that translates uncertainty in wave climate and water level design parameters into explicit confidence limits for transmitted wave height. Using Latin Hypercube Sampling, input uncertainty is propagated through a modified Wiegel transmission model, yielding empirical distributions of the transmission coefficients Kt and Ht. Our method uses the associated safety factor required to satisfy a 95% non-exceedance criterion, SF95. Regression analysis reveals the existence of a strong inverse linear relationship (R = −0.9) between deterministic Kt and the probabilistic safety factor, indicating that designs trimmed to low nominal transmission (e.g., Kt ≤ 0.35) must be uprated by up to 55% once parameter uncertainty is acknowledged, whereas concepts with greater transmission require far smaller margins. Sobol indices show that uncertainty in Hm0 and Tp each contribute ≈40% of the variance in Ht for a tide signal standard deviation of ση = 0.16 m, while tides only become equally important when ση > 0.30 m. Model-based uncertainty is negligible, standing at under 8%. The resulting lookup equations allow designers to convert any deterministic Kt target into a site-specific probabilistic limit with a single step, thereby embedding reliability into routine breakwater sizing and reducing the risk of underdesigned marina and port structures.