MATHEMATICAL MODELLING OF OVERALL HEAT TRANSFER COEFFICIENTS IN CONCENTRATION OF LEMON JUICE USING RECONCILED DATA

Citrus production is one of the main agro-industrial activities in Tucuman, Argentina, with a value in 2017 estimated at US$1,178 millions. Production of lemon juice concentrate requires the use of evaporation systems which are critical for the water and energy usage in all process industries. Precise knowledge of process variables is therefore extremely important for evaluation of the efficiency, economic parameters, etc. of the whole plant. In this work, data reconciliation and gross error detection were performed on data collected from the lemon juice concentration unit of an industrial plant located in Tucuman, Argentina. From reconciled data, the overall heat transfer coefficients of the four-effect evaporation line were calculated. In addition, a mathematical model was developed to predict the overall heat transfer coefficients. The parameters of the model were fitted to the experimental data and to the reconciled plant.

Addressing Higher Order Serial Correlation in Techniques for Gross Error Detection

Serial correlation present in the process measurement data may affect the performance of gross error detection (GED) techniques significantly. It has been our observation that most of the GED techniques assume that the data are not serially correlated. However, serial correlation can occur in measurement data due to delay in the process loop, signal processing elements, and other process phenomena. Performance of GED techniques depends on accurate estimation of variances of measured variables. Serial correlation in measured data increases the variance. Therefore, it is important to eliminate the effect of serial correlation in the measured data. Two approaches are proposed to handle serial correlation, one based on variance correction, and the other on prewhitening of residuals. This paper presents the performance of serial elimination techniques of higher order correlation applied in the measurement test (MT).