An Integrated SMR–S-CO2 Energy System for High-Performance Data Centers: Dynamic Simulation and Performance Evaluation
The rapid expansion of artificial intelligence (AI) has significantly increased the power demand of high-performance data centers. This study constructs an integrated energy system based on Small Modular Reactors (SMRs). The system adopts a supercritical carbon dioxide (S-CO2) Brayton cycle that directly supplies power for data centers. This work outlines the deep coupling of three core modules: power generation, S-CO2 energy storage, and waste heat absorption refrigeration driven by residual heat. The design enables coordinated optimization and cascaded utilization of nuclear multi-energy flows. We first build a full thermodynamic model for the whole system, and then formulate dynamic operation scheduling strategies. A 24 h full-condition simulation is carried out. The simulation object is a 125 MW SMR with a 100 MW data center in an off-grid island operation mode. Simulation results for the key performance indicators are as follows: The system cycle thermoelectric conversion efficiency reaches 45.00%. Compared with equal-capacity SMR units equipped with traditional steam cycles, the efficiency rises by 12.5 percentage points. The overall comprehensive energy efficiency hits 82.21%, and the system exergy efficiency is 74.30%. The system achieves a completely self-sufficient power supply without grid support. Its load power deficit rate is only 1.73%. Two operation modes dominate daily system operation: surplus power charging for peak shaving accounts for 60.6% of total runtime, while energy discharging to fill power deficits equates to 38.4%. Waste heat refrigeration requires roughly 93% of the data center’s total cooling demand. This research provides a new technical framework for low-carbon and sustainable construction of next-generation high-performance data centers. The integrated system we propose provides a replicable zero-carbon off-grid energy technical route. It can serve large computing hubs constructed under China’s national “East Data, West Computing” strategy.
Authors
- Miangang Tang (ORCID: https://orcid.org/0000-0002-3014-1854)
- Qiang Xu (ORCID: https://orcid.org/0000-0002-2252-0838)
- Xiaoyuan Chen (ORCID: https://orcid.org/0000-0002-9816-6724)
- Dianchuan Xing
- Xiyang Ma
Institutions
- Sichuan Normal University (CN)
Publication Details
- Journal
- Processes
- Published
- 2026-07-16
- DOI
- https://doi.org/10.3390/pr14142311
- Primary Topic
- Thermodynamic and Exergetic Analyses of Power and Cooling Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00