Multidimensional evaluation of a wind farm-data center hybrid energy storage system considering full life cycle carbon emissions

Wind power is one of the key energy sources for promoting emission reduction in power systems. Data centers, due to their specific electricity consumption characteristics, exhibit relatively high carbon intensity in power consumption. To explore approaches for increasing renewable energy penetration in data centers, this study takes a wind-powered data center as a representative case and performs a multidimensional evaluation of a hybrid energy storage system (HESS), considering full life-cycle carbon emissions. To mitigate the conflict between the intermittency of wind power and the stable operation requirements of data centers, an adaptive sliding window smoothing strategy is employed to smooth wind power fluctuations. Meanwhile, the power allocation and capacity sizing of storage units in the HESS are optimized using the Variational Mode Decomposition (VMD)-Hilbert transform. A consequential life cycle assessment (CLCA) is then performed to calculate the life-cycle carbon emissions of both the HESS and the integrated wind-storage power system. The results are analyzed in multiple dimensions, including storage configuration ratios, fluctuation mitigation effectiveness, and peak shaving and valley filling rates. The results show that, compared with thermal power peak-shaving regulation, the life-cycle carbon emissions of the wind-storage system are significantly lower than those of the combined wind-and-coal-fired regulation mode. Among different HESS configurations, the lithium iron phosphate (LFP) battery combined with supercapacitors under the adaptive sliding window smoothing strategy achieves the most effective emission reduction and demonstrates superior performance in fluctuation smoothing. Overall, this research promotes the low-carbon configuration and evaluation of data centers and HESS.