Open AccessOptimization of Biodiesel Transesterification using ANN and Fuzzy Logic
Author(s) -
Srikanth Sastry,
Bolla Govinda Rao,
G. Ravi Kishor
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.e6942.129219
Subject(s) - biodiesel , transesterification , diesel fuel , methanol , pongamia , yield (engineering) , pulp and paper industry , biodiesel production , biofuel , renewable energy , materials science , catalysis , environmental science , waste management , chemistry , organic chemistry , engineering , composite material , electrical engineering
The high energy demand in domestic sector coupled with pollution brought by extensive exploitation of conventional fuels in an industrialized world makes it mandatory to boost renewable energy sources having lesser environmental impact than non-renewable ones. In this regard bio-diesel can be considered as a more reliable resource of energy that can be used readily in the existing engines. Biodiesel is formed by transesterification reaction of alcohol and triglycerides under a catalyst. In this paper, Bio-diesel is produced from karanja (pongamia pinnata) oil in sono reactor at varied methanol-oil ratios and varied catalyst ratios. Yield was found at different molar ratios of methanol:oil (6:1; 4.5:1; 3:1), different KOH concentrations (2.0 wt %; 1.5 wt %; 1.0 wt %) and different times (15 min; 30 min; 45 min; 60 min). The biodiesel thus obtained conformed to ASTM D6751 standards. The optimum conditions of maximum yield are determined at 50o C temperature, 45 min reaction time, 4.5:1 methanol:oil ratio and 1.5% of KOH. The results obtained are well in accord with the literature. Also ultrasonic vibration used for production of biodiesel proves to be promising technique. The biodiesel thus produced is analyzed using various tests to obtain its properties. Further optimization techniques namely Artificial Neural Network and Fuzzy Logic have been applied for modeling the reaction and finding the optimum yield at different conditions. The yield predicted by using ANN and Fuzzy logic was compared with the experimental yield. The ANN and Fuzzy can precisely calculate as per the experimental data with R2 = 0.998 and R2 = 0.995, respectively.