Seismic performance of reservoir embankments rehabilitated using maehagane materials: applicability of effective-stress and total-stress strength parameters

Japan has a large number of aging irrigation pond embankments, many of which were constructed empirically and therefore exhibit uncertain internal structures and material properties. Under current Japanese guidelines, Level 1 seismic performance is primarily evaluated using slope stability analyses with shear strength parameters derived from effective stress. However, effective stress-based evaluation requires an appropriate representation of pore-water pressure, including earthquake-induced excess pore-water pressure, which can be difficult to quantify in embankments. This study investigates the influence of using effective stress- and total stress-derived strength parameters on the Level 1 seismic evaluation of irrigation pond embankments rehabilitated using the maehagane method, in which a low-permeability upstream core zone is installed to improve seepage control and slope stability. Two candidate core-zone materials were examined: conventional compacted clay and PSA-stabilized high-water-content clay prepared as a simplified laboratory analogue of dredged sediment, using the same base clay. Consolidated undrained triaxial compression tests with pore-water pressure measurements and falling-head permeability tests were conducted to determine the mechanical and hydraulic properties of the materials. Seismic stability was evaluated using two-dimensional circular-slip limit-equilibrium analyses, while seismic deformation behavior was assessed using two-dimensional dynamic finite-element analyses. The limit-equilibrium results show that the calculated safety factors based on total stress-derived parameters were consistently higher than those based on effective stress-derived parameters under their respective stress frameworks, and that the PSA-stabilized core satisfied the adopted design requirement only in the total stress-based evaluation. The dynamic analyses showed that maehagane rehabilitation reduced crest settlement, with the PSA-stabilized core exhibiting slightly smaller upstream settlements. Normalized crest settlement exhibited a clear tendency with the safety factors calculated using total stress-derived strength parameters, whereas no clear relationship was observed with the effective stress-based safety factors. Settlement increased sharply as the total stress-based safety factor approached unity, indicating a nonlinear increase in deformation near the limiting equilibrium condition. These results demonstrate that the choice of stress framework and associated shear strength parameters can substantially influence the calculated Level 1 seismic performance of low-height irrigation pond embankments rehabilitated using the maehagane method. Under the specific material, loading, and analytical conditions examined in this study, PSA-stabilized high-water-content clay also showed potential as an alternative core-zone material for improving seismic stability and deformation resistance.