MCDM-guided Planning of Grid-side Storage: Dualtechnology Complementarity for Peak Shaving and Frequency Regulation

High-penetration renewable generation causes deeper net load peak-valley gaps, intensified ramping, and squeezed system inertia/fast reserves, demanding second-level power and short-Term energy buffering. This paper proposes an end-To-end planning framework that couples contextualized multi-criteria evaluation with dual-Technology energy storage's collaborative capacity sizing. CRITIC-TOPSIS provides objective baseline weights/rankings, PROMETHEE forms peak-shaving/primary frequency regulation scenario preferences, and Monte-Carlo quantifies robustness stably revealing complementarity between power-Type LFP and energy-Type VRFB. In planning, renewable forecast intervals are mapped to time-series up/down primary reserve envelopes. Verifiable indicators (peak reduction rate, ramping smoothness) are embedded in the linear capacity model as soft penalties. For a 96×15-min typical day, the model outputs differentiated power/capacity configurations, achieving 100% primary frequency regulation coverage (up/down), cutting system peak from 1000 to 967 MW, and reducing total net ramping by 20%. Linking evaluation and planning via verifiable engineering indicators, the framework supports replicable deployment across regions.