Background: Fugitive emission sources generated from leaking components are often difficult to identify and locate, especially in petrochemical processing areas, which have concentrated facilities and equipment. Methods: A large-scale search method for locating and identifying fugitive emission sources in a petrochemical processing area by using multiple intersecting open-path Fourier transform infrared (OP-FTIR) beam paths and multivariate statistical methods is proposed in this study. Multivariate statistical methods, namely principal component analysis (PCA) and correspondence analysis (CA), were applied to the measured data set and the characteristics of emission sources were summarized. Results: Styrene, 1,3-butadiene, cyclohexane, ammonia, ethylene, propylene and methanol were identified in most beam paths with mean concentrations up to 75.3 +/- 11.4 ppbv, 355.2 +/- 34.6 ppbv, 188.1 +/- 17.82 ppbv, 255.57 +/- 19.28 ppbv, 194.3 +/- 18.2 ppbv, and 94.0 +/- 16.7 ppbv, respectively. PCA extracted at least three source categories in the plant. CA provided additional information about the approximate locations of each source, and thus, future emission reduction plans can be developed accordingly. Conclusion: This study established a dynamic approach for locating fugitive emission sources in a complex petrochemical plant by applying an OP-FTIR matrix-path approach combined with PCA and CA. Future emission reduction plans can be developed according to the findings of PCA and CA. (C) 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.