Importance of Data Reconciliation on Improving Performances of Crude Refinery Preheat Trains

Performance of a preheat train is evaluated through plant monitoring data (temperature and flow measurements); here the reliability of data reconciliation plays a significant role. Most preheat trains do not monitor all stream temperatures (or flow rates) and frequently guess the missing data through ‘short-cut' methods. Wrong choice of ‘short-cut' methods such as the use of linear interpolation is shown to result in misleading conclusions.  

This paper introduces a systematic data reconciliation approach through three steps. 1) Generation of missing data through a modified preheat train simulator based on Energy Fuels, 2009, 23(3), pp 1323–1337. 2) Filtering unreliable data through a ‘trusted' heat balance. 3) Grouping heat exchanger monitoring data into different time sections to identify trends under different operating periods.

The superiority of the new data reconciliation methodology is illustrated through a series of case studies. The reconciled data are then utilized to make key operational decisions to improve preheat train performance.