Abstract
The total reliance on fossil fuels has led to the depletion of its sources and environmental
challenges. This has necessitated the research to develop renewable and sustainable sources of
energy that are environmentally benign and can substitute for dirty fossil fuels. As a bright
means of generating second-generation biofuels from readily available and low-cost feedstocks
like lignocellulose materials, there is lofty anticipation reposing biogas to alleviate the harmful
effects of fossil fuels and reduce the challenges of the energy and also improve ecological impact
globally. Agricultural and agro-industrial residues are becoming major challenges because of
improper disposal, use, and management methods. Anaerobic digestion of lignocellulose
materials from agricultural and agro-industry wastes for bioenergy recovery and pollution
control is a bright option for a greenhouse gas mitigation alternative and is regarded as a
sustainable energy production and waste management technique. Lignocellulose materials are
primarily cellulose, hemicellulose, and lignin, making them a complicated organic polymer
crystal structure. Lignin and hemicellulose in lignocellulose materials are strongly embedded in
the cellulose, leading to lignification and crystallization, making cellulose more inaccessible to
microorganisms and the cell wall recalcitrant during the hydrolysis stage of the anaerobic
digestion process. Due to this, the energy recovery rate from lignocellulose materials without
pretreatment is about 20%, with a longer retention time. To overcome these challenges,
pretreatment of lignocellulose materials with different techniques before anaerobic digestion is
required to enhance digestion and reduce the lag time. An appropriate pretreatment technique
alters the lignocellulose materials' structural arrangement and chemical characteristics, making
them accessible to the anaerobic digestion microorganisms, thereby improving the biogas and
methane yields. This study considered Arachis hypogea (groundnut/peanut) shells, a lignocellulose
material with very little engineering application reported, as feedstocks. The shells were
pretreated with mechanical (particle size reduction), combined (particle size reduction and
nanoparticles), thermal (conventional heating), and chemical (acid and alkali) pretreatment
techniques at different pretreatment conditions. The influence of the pretreatment techniques was
examined on the morphological structure, crystallinity, and functional groups using Scanning
Electron Microscopy (SEM), X-ray Diffraction (XRD), and Fourier Transform Infrared
spectroscopy (FTIR) analysis. The pretreated Arachis hypogea shells were digested in a
laboratory-scale batch digester at mesophilic conditions to study the effect of pretreatment
regimes on biogas and methane yields. Furthermore, Response Surface Methodology (RSM),
Artificial Neural Network (ANN), and Adaptive Neuro-Fuzzy Inference System (ANFIS)
models were developed to forecast the biogas and methane yields of the pretreated Arachis
hypogea shells. The process parameters selected as input parameters were retention time,
temperature, and pretreatment conditions, while the outputs were biogas and methane. A linear
polynomial regression equation was employed to predict the biogas and methane yield and
ascertained the selection of process parameters that produce optimum yields using the RSM.
Analysis of Variance (ANOVA) performance metrics were utilized to determine the significance
of the model. During the ANN and ANFIS prediction, the observed data were divided into
training and testing data sets. Training of the models was carried out with 70% of the data, while
the remaining 30% of hold-out data were used for model testing. The developed models were
then validated with the experimental data. The performance of the developed models was
investigated with Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE),
Mean Absolute Deviation (MAD), and Correlation Coefficient (R2) performance metrics. The influence of pretreatment techniques showed that particle size reduction does not
significantly affect the morphological, crystallinity, and functional groups of Arachis hypogea
shells. On the contrary, thermal, acid, and alkali pretreatments exhibited significant effects on the
morphological, crystallinity, and functional groups of Arachis hypogea shells. It was discovered
that the severity of the pretreatment techniques was not the same on the substrate, and each
treatment technique has different conditions for optimum biogas and methane yields.
Biomethane yield was improved by 23.96% in a conventional thermal pretreatment at 100 °C for
30 min. Combined treatment of 6 mm particle sizes and 20 mg of Fe3O4 nanoparticles enhanced
biogas and methane yields by 80.59 and 106.66%, respectively. Acidic pretreatment with H2SO4
at a treatment condition of 0.5% (v/v) for 15 min at 90 °C increased biogas and methane yields
by 12.90 and 178.00%. Alkali pretreatment using NaOH enhanced methane yield by 67.59%
when 3% (w/w) of NaOH was applied for 15 min at an autoclave temperature of 90 °C. The
influence of particle size reduction was considered on organic dry matter biogas (oDMBY), fresh
mass biogas (FMBY), organic dry matter methane (oDMMY), and fresh mass methane (FMMY)
yields, and the results showed that the expected yield would determine the choice of particle size.
Modelling the influence of particle size reduction with RSM showed a coefficient of
determination (R2) of 58 – 63%, whereas the ANFIS model produced an R2 value of 92 – 96%.
ANN and ANFIS models developed were validated with experimental biogas and methane yield
data from combined pretreatment techniques. The ANN prediction results indicated that the
developed models predict similar results to the observed results with RMSE, MAPE, MAD, and
R2 of 98.06, 93.68, 98.98, and 94.96%, respectively. ANFIS models also forecasted biogas and
methane yields of combined pretreatment techniques on Arachis hypogea shells with RMSE,
MAPE, MAD and R2 values of 98.77, 94.77, 98.75, and 98.50%, respectively. The experimental biomethane yields from thermally pretreated Arachis hypogea shells were used to validate the
RSM and ANN models developed. The R2 was observed to be 0.7393 and 0.9754 for RSM and
ANN models.
Biogas and methane yields of Arachis hypogea shells were enhanced with all the
pretreatment techniques considered at specified conditions as established by this study. It can be
inferred from these results that pretreatment of Arachis hypogea shells before anaerobic
digestion will improve energy recovery from lignocellulose materials, assist in greenhouse gas
mitigation, and encourages the application of anaerobic digestion as a sustainable waste
management technique. These pretreatment techniques can be applied commercially and serve as
the baseline for other lignocellulose materials with similar morphological structures. The
performance of the developed models showed that the use of RSM to predict biogas and methane
yield when pretreatment techniques are involved is average. Still, the performance metrics of the
developed ANN and ANFIS models are acceptable as predictive models for biogas and methane
yield of pretreated Arachis hypogea shells digested in the batch digester at mesophilic
temperature. Applying these models at an industrial scale will save time and make the anaerobic
digestion of lignocellulose materials more economical.
Keywords: Energy, Greenhouse gas, Renewable energy, Lignocellulose materials, Arachis
hypogea shells, Anaerobic digestion, Pretreatment techniques, Biogas, Methane, Modelling.