Source–Load Coordinated Multi-Stage Low-Carbon Planning for Data Center Clusters Considering Construction Sequence

Planning data center clusters requires explicit consideration of construction sequencing and load growth over the planning horizon to avoid redundant early-stage capacity and premature investment while preventing capacity shortages and deterioration in power-supply reliability during later stages. This study proposes a source–load coordinated, multi-stage, low-carbon planning method for data center clusters. First, representative daily operating scenarios for the four seasons are generated. Second, the regional energy-planning cycle is aligned with the projected load growth of the data center clusters, and a tiered carbon trading mechanism is incorporated into a multi-stage optimization model that minimizes the total life-cycle cost. Comparative scenarios with different carbon constraint intensities are then evaluated to determine the construction schedule and capacity allocation of each technology. The results demonstrate that the proposed method can support the low-carbon expansion of large-scale data center clusters by coordinating investment timing, capacity deployment, and carbon reduction requirements.