With the advancement of big data and internet technologies, current standards and specifications for data center equipment rooms lack provisions addressing energy consumption in liquid cooling systems, while existing fault diagnosis methods exhibit limited generality and accuracy. This study proposes a health assessment methodology for liquid cooling system safety and energy efficiency in data centers, integrating expert knowledge, model prediction, and historical operational databases through a comprehensive scoring framework. Practical applications demonstrate that this diagnostic method effectively identifies inefficient operational states and control methods of system equipment through safety and energy efficiency metric evaluations, while assessing energy-saving potential in operations and maintenance. The results provide critical decision-making references for facility management personnel.