Abstract—Renewable-driven direct-current optimal load shedding (DC-OLS) requires a model that is interpretable to operators, data driven under continuous forecast errors, sensitive to severe security failures, and computationally tractable. This paper develops a budgeted KDE-ϕ-HMCR-RS-OLS framework for that purpose. Robust satisficing (RS) replaces ambiguity-radius tuning with an admissible shedding budget. A one-dimensional KDE reference family with ϕ-divergence reweighting gives a smooth finite-sample representation of renewable uncertainty and a tractable blockwise reformulation. Higher-moment coherent risk (HMCR)distinguishes the severity of reserve exhaustion and line overload once violations occur. The paper derives a continuous distribution dual template, the main finite-sample reformulation, and Kullback–Leibler and Pearson χ2 specializations, and develops a stabilized bundle algorithm for benchmark-scale instances. Results on RBTS and RTS79 show that the proposed method improves security relative to deterministic DC-OLS while avoiding the full preventive cost of heavier robust baselines.