Crossflow Membrane Technology for Crude Palm Oil Treatment

Crude palm oil (CPO) is refined to quality edible palm oil by removing objectionable impurities such as free fatty acid (FFA), phospholipids, trace metals and colouring pigments which are detrimental to the flavour, odour, colour and stability of the oil. Conventional refining involves degumming, bleaching and deodorization steps. The energy-saving membrane technology is a physical separation process that can offer an alternative method to improve the conventional refining method by reducing energy cost and minimizing the waste disposal problem. The objectives of this research study are to investigate the influence of membrane on the quality of CPO, to evaluate the operating conditions such as transmembrane pressure and feed flow, to study the performance of membrane with time, to establish the appropriate cleaning procedures for membrane, and to compare the quality of membrane-processed oil and conventional-processed oil as well as the storage stability of the oils. CPO was filtered through micro filtration (MF) and ultrafiltration (UF) ceramic membranes. Two different pore sizes of MF membranes (0.2 um and 0.45 um) and two pore sizes of UF membranes (20 run and 50 run) were used in the study. Comparison study was conducted for CPO treated with 0.2 um membrane. Ceramic MF membranes with pore sizes of 0.45 and 0.2 um rejected about 14% and 56.8% of phosphorus, respectively. The 0.2 um membrane removed more than 80% of the iron. Ceramic UF membranes with pore sizes of 50 and 20 run rejected about 60% and 78.1% of phosphorus, respectively. The 20 run membrane reduced about 60% of iron content. All membranes (MF and UF) showed no influence on carotene, FFA and fatty acid composition (FAC). The MF membranes (0.2 and 0.45 um) showed similar trend where the permeate flux for the membranes increased with average transmembrane pressure and feed flow until it reached a certain limit where the flux declined with increasing pressure and feed flow. Both membranes showed rapid flux decline during the initial stage, but stabilized for the period of 5 hr. Cleaning process was achieved by using Alconox detergent and acid/alkalis solutions. The effect of pressure on flux for the 20 nm UP membrane showed similar trend with the MF membrane, but after 0.9 bar, the increased in flux was only slowed down rather than declining. The flux for the 50 run UF membrane increased proportionally with pressure and no sign of flux decline was observed at 2.9 bar. The increase in flux at higher feed flow was only observed for the 20 nm membrane. Concentration polarization could have occurred for the 50 nm membrane where the flux declined at higher feed flow. Cleaning for the UP membrane was more difficult than the MF membranes. The cleaning process involved cleaning with Alconox detergent, acid/alkalis solutions and soaking with solvents (hexane and isopropanol). Reduction of phosphorus for oil treated with 0.2 um MF membrane was comparable with commercial bleached oil. FFA and carotene content were reduced after deodorization process. Carotene content and colour reading for membrane-processed oil was slightly higher than conventional-processed oil. The membrane process was unable to remove oxidation products as it was observed that the FFA and peroxide value were increased during the storage period.

PDF FK_2001_16_A.pdf Download (1MB)

Actions (login required)

View Item