[Purpose/Significance] This study proposes an anlytical framework, which could be used to provide reference for the demonstration of the association of facilities in the planning of large facilities clusters, and help to provide a forward-looking quantitative reference for the planning and layout of large facilities, as a supplement to demand driven planning.[Method/Process] This research started with scientific outputs and applied scientometric methods to propose the large-scale scientific facilities cluster layout planning analysis framework based on the correlation of scientific research output (SOCO-LFPF)', including processes and methods such as confirming analysis objects, obtaining facility data, describing facility knowledge distribution, and analyzing facility association. This study taking the Greater bay area Comprehensive National Science Center as an example, the paper data of the same type of facilities in the world for the existing large facilities (spallation neutron source and synchrotron radiation light source) and the planned large facility X were collected, and the internal correlation of its discipline level was analyzed from multiple levels.[Result/Conclusion] The number of papers published relying on large scientific facilities continues to rise. The three types of large scientific devices have strong interdisciplinary nature, and the interdisciplinary nature of different facilities is different. The three types of facilities have different correlations. The existing facilities (spallation neutron source and synchrotron radiation light source) have high correlation and may be complementary, while the large device X has low correlation with the existing facilities. Further expert argumentation is required for layout suggestions. The method proposed in this paper is feasible and provide a forward-looking quantitative reference for the planning and layout of large-scale facilities to a certain extent.