This paper presents a scenario-based capacity planning framework for a renewable-powered data center integrating photovoltaic (PV) and wind energy resources under uncertain and time-varying generation conditions. The main objective is to determine the optimal installed capacities while ensuring reliable supply of the data center demand under multiple renewable availability scenarios. The proposed formulation incorporates a flexible demand response strategy based on load shifting, enabling improved temporal alignment between renewable generation and consumption. A key innovation of this study is the development of a comprehensive annualized cost minimization model that simultaneously accounts for capital expenditure (CAPEX), operation and maintenance (O&M), and system investment recovery using a capital recovery factor (CRF). Unlike conventional approaches, the proposed model explicitly incorporates annual O&M costs into the economic formulation, providing a more realistic long-term planning perspective than investment-cost-only models. The model is formulated as a multi-scenario optimization framework, capturing variations in solar irradiance and wind speed. The results demonstrate that the proposed framework effectively adapts capacity allocation under different renewable conditions. The optimal capacities are found to be 2409.4 kW for PV and 4583.6 kW for wind in the base case. Under reduced PV conditions, PV capacity increases to 3168.3 kW, while under reduced wind conditions it further increases to 4034.8 kW, indicating strong substitution and flexibility between renewable sources. The total annualized system cost is approximately 1.009 million USD/year, with CAPEX and O&M contributing 827,760 USD/year and 181,490 USD/year, respectively. The obtained capacity plan ensures robust operation under all considered renewable generation scenarios.
