Optimization of the Product Yields of the Pyrolysis of Palm Kernel Shell in a Fixed-Bed Reactor Using Response Surface Methodology (RSM)

• Published in: International research journal of innovations in engineering and technology 2024-03, Vol.8 (3), p.155-166
• Main Authors: Joseph, O O, Akinola, A O, Ogedengbe, T I, Akinnuli, B O
• Format: Article
• Language: English
• Subjects: Agricultural wastes; Beds (process engineering); Biogas; Coefficient of variation; Correlation coefficients; Mathematical models; Optimization; Palm oil; Polynomials; Process parameters; Pyrolysis; Refineries; Response surface methodology; Statistical analysis; Variance analysis; Waste management; Wastes
• ISSN: 2581-3048; 2581-3048
• DOI: 10.47001/IRJIET/2024.803021

Pyrolysis is a method used for the thermochemical conversion of biomass and agro wastes into three major by-products namely, biochar, bio-oil, and biogas in the absence of oxygen. Global waste generation is increasing at an alarming rate. Palm kernel shell (PKS) is among the palm wastes accumulating in many palm oil mills around the globe with the attendant problems in waste management. The utilization of coal for energy generation is more expensive and not environmentally friendly. Bio-oil refineries and industries require the utilization of minimum resources (input) to achieve optimal response (output) in the production process. However, more than 30% of the energy content derivation from the pyrolysis process is wasted due to nonoptimization of the process parameters. This study investigated the optimization of the product yields of the pyrolysis of PKS in a fixed-bed batch reactor using response surface methodology (RSM). Masses of 1 kg, 2 kg, and 3 kg were pyrolyzed in batches at process temperature variations of 200 OC, 300 OC, 400 OC, and 500 OC. Design Expert 12 software (Version 12.0.3.0 Stat-Ease Inc. MN, USA) and Analysis of Variance (ANOVA) were used for the statistical analysis. Lack-of-fit, adjusted, and predicted multiple correlation coefficients and coefficient of variation of the different polynomial models were compared to select the best-fitting polynomial model. The optimal feedstock parameters were 2.66 kg of PKS and 410.0 OC process temperature. The predicted results were 0.986 kg of biochar, 1.205 kg of bio-oil, 0.481 kg of biogas, 1116.5 cm3 of bio-oil, and 140.6 minutes. The experimental results were 0.894 kg of biochar, 1.316 kg of bio-oil, 0.498 kg of biogas, 1208.9 cm3 of bio-oil, and at 162.8 minutes through validation. The validated quadratic model was suitable for the optimization and RSM is a numerical, statistical, and mathematical tool for modeling and optimization of the pyrolysis process in a fixed-bed batch reactor.