Data reconciliation of an industrial petrochemical plant case study: Olefin plant (Hot section)

The quality of online measured operational data is usually not satisfactory for the performance monitoring of olefin plants, due to the low accuracy of measuring instrument. Data reconciliation (DR) is data preprocessing method which can improve the accuracy of measured data through process modelling, optimization, and can also be applied for gross error detection together with the statistical test method. In this study, DR and gross error detection are used to an industrial olefin plant. DR simulation results showed that the relative root means squared errors of the primary flow measurements, namely the inlet mass flow rate of feed to F-101, F-102, and F-104 are reduced by 40.3%, 13.4%, and 21.4%. After that, the authors applied Global Test (GT) and serial elimination strategies to gross error detection and validation in the measurement of existed olefin plant and was successfully detected and validated by onsite inspection of the olefin plant. (C) 2020 Elsevier Ltd. All rights reserved.