Product sampling time and process residence time prediction of palm oil refining process

Abstract

The quality of product produced in palm oil industries is important due to customer need and world demand. It is crucial to monitor and predict the quality of Refine Bleached Deodorised Palm Oil (RBDPO) produced. One of the most significant driving forces behind the demand for quality prediction of RBDPO is the reducing production cost of RBDPO. Palm oil had the highest manufacturing cost in that scenario, mainly because of the higher demand for ethanol in the pre-Treatment step. The development of rapid and non-destructive measuring technique is needed to enhance the efficiency of palm oil quality monitoring. Today in the Lahad Datu Edible Oil SDN. Bhd. (LDEO) company, the cost production of RBDPO is much higher and can be double or triple than the usual cost production due to the need of recycle process for getting the desired quality of RBDPO. The current production of RBDPO takes a lot of time. This show how important the quality predict of RBDPO in meeting the needs of the LDEO Company to reduce the need of recycle process due to the cost production and time consuming. This study aims to develop a statistical analysis that can predict the sampling time and residence time for the whole refinery process using autocorrelation and cross-correlation in MATLAB. The data analysis of Crude Palm Oil (CPO) and RBDPO is based on five parameters, percentage of the free fatty acids (% FFA), the percentage of the moisture, the deterioration of bleachability index (DOBI), iodine value (IV) and peroxide value (PV), which are used to monitor the RBDPO quality provided by the LDEO Company. For the 65 sample size data, the sampling time and residence of the whole refinery process are 2 h and 2.4 h. Both data are required in developing a tool to predict the RBDPO quality.