A data driven health assessment model for high pressure output pumps in LNG terminals

• Published in: IOP conference series. Earth and environmental science Vol. 446; no. 4; pp. 42064 - 42069
• Main Authors: Zeng, Yuyun, Wang, Xiaoshang, Xie, Guangyao, Liu, Jingquan
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.02.2020
• Subjects: Condition monitoring; Feature extraction; High pressure; Liquefied natural gas; Pressure; Pumps; Signal monitoring; Statistical analysis; Terminals; Vaporization; Vibration; Vibration analysis; Vibration monitoring; Wavelet analysis
• ISSN: 1755-1307; 1755-1315
• DOI: 10.1088/1755-1315/446/4/042064

High pressure output pumps are critical equipment in the vaporization and output system of LNG terminals. Health management helps improve efficiency and reduce cost of the maintenance of high pressure output pumps, thus guaranteeing the efficient productivity of LNG terminals. In order to develop health management system for high pressure output pumps, a data driven health assessment model based on online condition monitoring data of the pumps is proposed. Time domain and frequency domain features are extracted from the monitored vibration signal by statistical analysis and wavelet packet decomposition respectively, and a health index is constructed based on T2 and SPE statistics given by PCA results of the extracted features. The proposed model is validated based on monitoring data of high pressure output pumps in Qingdao LNG terminal. Results show that the calculated health indices are good indicators of the health status of the pumps, and are able to detect potential fault in their early development stages.