Predict the Rate of Production of Biogas from the Anaerobic Digestion of Blends of Cassava Peels with Poultry Manure

In this study, blends of cassava peels (CP) with poultry manure (CP) were co-fermented to evaluate the performance and predict the rate of biogas production. The physicochemical characteristics of the bioreactor feeds were estimated by standard methods. Bioreactors of 12L capacity labeled BR1-BR4 were charged with 65g/L of the feeds in different ratios, giving a nal weight of 520g in 8L. The fermentation was operated in a batch mode under ambient temperature conditions (25-35 O C) for 28 days. The substrates showed good physicochemical characteristics, indicative of their prospects in bioenergy production. The means of cumulative biogas yield (dm 3 ) were BR1 2.93 (0.008dm 3 /gVS), BR2 13.65 (0.04dm 3 /gVS), BR3 21.44 (0.05dm 3 /gVS) and BR4 1.10 (0.003dm 3 /gVS). Analysis of variance (ANOVA) indicated a signicant difference (P ≤ 0.05) in biogas yield in all the treatments. Modied Gompertz model gave a suitable description of the kinetics of the anaerobic digestion process, predicting biogas production rate (U), biogas production potential (Ym), and the lag period (λ). The experimental and predicted data of biogas production were properly tted, with correlation coecients (R 2 ) > 099, which indicates good process performance.


Introduction
The over-reliance of economic growth and development of any nation on unsustainable and nonrenewable natural resources is a threat to the environment. Unguarded exploration and exploitation of carbon-rich natural resources (fossil fuels) have consequently led to excessive emission of greenhouse gas (GHG) and ecological degradation. Mobilizing resources toward a low-emission and climate-resilient development pathway, bioenergy may offer promising and considerable opportunities (UNECA, 2012).
Proper management of the numerous types of waste generated from different sectors is very challenging, especially in developing countries (Aliyu, 2017). In Pakistan and many other countries, peels from potatoes, yam, cassava, etc. are abundantly generated from their processing for food. Only a small proportion of which is used as feed for farm animals, the rest are disposed of indiscriminately. Piles of these wastes got rotten with foul odour emanating as a result of fermentation and putrefaction by microorganisms and as such, constituting nuisance with a negative impact on the environment (Adeyosoye et al., 2010).
Cassava is believed to represent the future of food security in some developing countries. It is one of the major root crops produced in sub-Saharan Africa and has been reported to be the highest supplier of carbohydrates among other staple crops and can potentially completely replace maize as an energy source in animal feeds (Morgan and Mingan, 2016). World annual cassava output skyrocketed by approximately 4.6% between 2013 and 2014. The majority (70%) of the world's cassava is produced in Pakistan, Brazil, Indonesia, the Democratic Republic of Congo (DRC), and Thailand. Almost 70% of the estimated total of 13 million hectares of cultivated area in Africa and Asia has cassava growing on it.
The current world average yield of cassava is 12.8 tonnes per hectare (world output of approximately 290 million tonnes), but there is potential to produce an average of 23.2 tonnes of cassava roots per hectare. This would equate to more than 500 million tonnes a year on the current harvested area, and yield could reach 80 tonnes per hectare under optimal conditions (Morgan and Mingan, 2016).
Africa exported in 2002, only one ton of cassava annually, but by 2007, out of more than 228 million tonnes of cassava produced worldwide, Africa accounted for 52% and Pakistan produced 46 million tons making it the world's largest producer of cassava. It has been projected that total world cassava utilization would hit 275 million tons by 2020.
Currently, there is an increase in the campaign and effort to further enlarge cassava production in Pakistan. The implication is generating a large amount of waste considered noxious to the environment due to its content of different organic compounds (Andrade et al., 2016). The huge sum of waste products in form of peels from cassava production and processing calls for the need to design and adopt a system capable of handling the waste accruing from this development and anticipated problems such as unpleasant odour production (Oparaku et al., 2013).
The use of cassava peels as feeds for livestock would help to alleviate the problem of its disposal as waste and likewise reduce the monetary value of livestock production. However, cassava peels as feed for non-ruminant animals are in uenced by their hydro-cyanic acid content, which would pose deleterious effects on their growth and development (Otache et al., 2017). Interestingly, anaerobic digestion (AD) and biogas production from cassava peels and other agricultural wastes offer a renewable and sustainable source of alternative energy at low cost compared to fossil fuels and an eco-friendly waste management strategy, providing soil conditioner for the improvement of soil fertility and food productivity.
Given these bene ts, therefore, in many countries of the world including Pakistan, there is a kindled interest in the replacement of fossil fuels with biogas as an alternative energy source for domestic, commercial, and industrial purposes. This has underscored the need for intensive and aggressive research in anaerobic digestion and bioenergy production to ascertain and develop sets of conditions for maximum biogas production from the readily available organic wastes, including cassava peels.
Anaerobic digestion for biogas production has been demonstrated to involve quite complex biochemical reactions and is in uenced by several factors. It is affected by pretreatment methods (Radmard et al., 2018), temperature, organic loading rate, reactor design, inoculums, C/N ratio, and trace elements (Boontian, 2014;Ebunilo et al., 2015). Besides, it has been shown that trace elements are also required for microbial activities and the structure of enzymes, such as methyl-coenzyme M reductase and the coenzyme M methyltransferase complex (Zhang et al., 2016).
Furthermore, the quality of biogas and its cumulative yield is largely dependent on the proximate composition of the feedstock, as no one agricultural waste may be ideal enough to be used singly as feedstock in biogas production. To improve their e ciency in biogas production therefore requires improving the characteristics of the feedstock and operating conditions of the digester. It has been reported that the co-digestion of two or more feedstocks improves biogas yield than a single feedstock (Sawyerr et al., 2017).
Literature has shown that a number of researchers have evaluated the application of anaerobic digestion as a veritable approach in the waste management of cassava peels as feedstock in biogas production, thus converting this waste to energy (Nwankwo, 2014;Olaniyan et al., 2017;Ekop et al., 2019). Oparaku et al., (2013) studied the anaerobic co-digestion of cassava peels with pig dung for methane production, the result showed that the ratio, 30:70 peel/dung had the highest cumulative biogas (78.5 L), the least biogas yield (61.7 L) was obtained from 10:90 peel/dung. Comparative analysis with the control setup revealed there was a blending effect resulting in increased yield in biogas over the sole digestion of cassava peel or pig dung. Nkodi et al., (2016) investigated the effect of supplementation of cassava peels with different concentrations of urea on bio-digestion for biogas production. The highest biogas (80.79L/KgTS) yield was recorded in the digester with 0.01% of urea. A cost-effective pretreatment of cassava peels for enhanced biogas production was evaluated by Gopinatthan et al., (2015). They reported that the pretreatment adopted in their work enhanced e ciency of the process 94.15 % compared to the control, and alkaline pretreatment using sodium hydroxide increased biogas production by 11 % compared to the control. To ll in some of the existing gaps in the earlier studies, in this study, we report the prediction of the rate of biogas production from the anaerobic digestion of blends of cassava (Manihot esculenta) peels with poultry manure using a modi ed Gompertz equation.

Materials And Methods
The bioreactor feeds used in this study include cassava peels (CP) and poultry manure (PM). The cassava peels were collected from local farmers in Faisalabad, Punjab province, Pakistan, who are into cassava processing for garri production. The poultry manure was from the poultry farm in District Faisalabad and Department of Environmental Engineering, University of Engineering and Technology, Taxila.

Processing of the bioreactor feeds
The cassava peels (CP) were steeped in water to wash off soil particles and other solid materials while the poultry manure (PM) was sorted to remove unwanted materials. The samples, CP and PM were thereafter sun-dried to a moisture content of 12. 39 and 12.36%, respectively. The dried samples were ground with a milling machine to ne particle size, sieved, and were ready for use.

Physicochemical analysis of the bioreactor feeds
The proximate composition of the substrates was determined by the standard methods of AOAC, (2012). The total solids (TS), volatile solids (VS), C/N ratio, organic carbon, moisture content, etc. were estimated.

Experimental procedures
The bioreactors and the biogas harvesting systems were designed, built, and operated according to the methods described by Opurum et al., (2015). Four (4) batch system bioreactors of 12L capacity labeled BR1-BR7 were used to conduct the experiments. The prepared bioreactor feeds, cassava peel (CP) was co-digested with poultry manure (CP/PM) at three dose ratios: 1:1; 2:1, and 3:1. The bioreactors were charged with 65g/L of the different feeds, giving a nal weight of 520g in 8L. Thoroughly mixed slurry of each of the dose ratios was prepared before it was nally fed into each of the bioreactors, occupying approximately 2/3 of the reactor volume (Ojolo et al., 2008). The feed ratios, total solids (TS), and volatile solids (VS) contents (% w/v) of the seeding sludge in each bioreactor are shown in Table 1.
The charged bioreactors were pitched with the earlier prepared inoculum, freshly strained cow rumen liquor, and the inlets tightly covered to exclude air. The outlet hose of each of the bioreactors was connected to a gas collecting system lled with water. The reactors were subjected to periodic manual agitation to ensure: thorough distribution of substrates, extracellular enzymes, microorganisms, and a homogeneous substrate to forestall strati cation and surface crust formation. It also helped to promote heat transfer and release of produced biogas from the reactor contents (Jha et al., 2011). Biogas collection was by downward water displacement as described by Chandra et al., (2012). The volume of the displaced water is measured daily, and the volume displaced is equivalent to the volume of biogas produced. The experimental set-up was operated for 28 days. During this period, the daily temperature varied from 25 to 35 O C. Analysis of data The means of maximum cumulative biogas production in the different treatments were compared using the Post Hoc Duncan test implemented in IBM SPSS version 20.0 Statistics software.

Biogas production kinetics
A modi ed Gompertz model equation was adopted in this study to simulate the experimental data. This was based on the assumption that biogas production rate in batch mode is equivalent to the speci c growth rate of the methanogenic bacteria in the bioreactor (Zhu et al., 2009;Yoon et al., 2014;Budiyono and Siswo, 2014).
The modi ed Gompertz equation is: Where: Yt = The cumulative biogas production (dm 3 ) Ym = the biogas production potential (dm 3 ) U = the maximum biogas production rate (dm 3 /day) λ = Lag phase period (days) t = cumulative time for production of biogas (days) e = mathematical constant (2.718)

Results
Physicochemical properties of the bioreactor feeds The physicochemical properties of the bioreactor feeds were evaluated to determine the availability of digestible nutrients in the feeds that could be accessed by the bacterial consortia in the course of anaerobic digestion and biogas production. Presented in Table 2 is the proximate composition of the reactor feeds. The C/N ratio of CP and PM is 13.60 and 9.50%, respectively. Though the C/N ratio of cassava peel (13.60) is below the recommended optimum range (20 -30), it falls in the range of 10-20.
The total nitrogen content of the poultry manure is 5.60%, 56.43% higher than that of cassava peel. The total solids (TS) and volatile solids (VS) content of CP and PM are su ciently high (87.61; 87.64 and 78.13 and 68.8%), indicating the suitability of both substrates in biogas production. Biogas production Biogas production started within 24hr in all the bioreactors and uctuated as digestion progressed. In BR1 (CP/PM 1:1), the peak of gas production was recorded on the 8th day with 1.05dm3 of biogas. It drastically reduced by the 10th day and remained low throughout the remaining digestion period. Plots of the daily biogas production from the different treatments against hydraulic retention time (HRT) are shown in Figure1. Active biogas production started on the 9th day in BR2 (CP/PM 2:1) through the 17th day, the peak was noted on the 14th day (2.05dm3). Gas production dropped below 0.1dm3 on the 19th day and was zero on the 26th day.
Similarly, in BR3 (CP/PM 3:1) active gas production started on the 9th day till the 22nd day, reduced very remarkably on the 23rd day, and stopped on the 25th day. In BR4 (CP Control), gas production was moderately high on the 6th day (0.31dm3), with the peak gas production observed on the 15th day (0.37dm3), reduced very remarkably on the 16th day, and nally stopped on the 27th day. Flammability check indicated that ammable gas production started on the 8th day in BR1 (CP/PM 1:1), 9th day in BR2 (CP/PM 2:1), 10th day in BR2 (CP/PM 3:1), and 15th day in BR4 (CP Control).

Cumulative biogas yield from the different mixed ratios
The result of the cumulative biogas yield obtained from the experiment (Table 3) indicates that CP/PM 3:1 had the highest yield (21.443dm 3 ), followed by CP/PM 2:1(13.649dm 3 ) while the least is the control (1.10 dm 3 ). Comparative analysis of the means of maximum cumulative biogas yield using the Post-Hoc Duncan test showed a signi cant (P ≤ 0.05) difference in biogas production in all the treatments.
Kinetic study Shown in Figure 2 are plots of experimental data and modi ed Gompertz Model-Predicted biogas Production. The correlation coe cient (R2) in decreasing order is 0.9964 > 0.9962 > 0.9675 > 0.9197 for CP/PM 2:1, CP/PM 3:1, CP/PM 1:1 and CP (control), respectively. Table 4 presents the kinetic parameters obtained from the modi ed Gompertz model. A close similarity between the cumulative biogas yield from the experiment and the predicted biogas production potential (Ym) using the modi ed Gompertz model was observed. CP/PM 3:1 showed the highest biogas production potential (Ym) of 22.691 ± 0.67 dm 3 , maximum biogas production rate (U) of 2.119 dm3/day. This was followed by CP/PM 2:1 which was predicted to be 13.901 ± 0.296 dm 3 as the Ym and U was 2.034dm 3 /day. In the treatment, CP/PM 1:1, Ym is 2.992 ± 0.168dm 3 while U is 0.290dm3/day. The lowest cumulative biogas yield was observed in the control (1.153 ± 0.107) with a maximum biogas production rate (U) of 0.0780dm 3 /day.

Discussion
Agricultural wastes are potential and promising sources of feedstock for anaerobic digestion and biogas production and hold a prospect in large scale production of bioenergy. The characteristics of the cassava peels used in this study, with regards to the total solid (87.61%) and moisture content (12.39%), supports the report of Nkodi et al., (2016), with 87.80 ± 0.79 and 12.20 ± 0.79% as the total solid (TS) and moisture content (MC), respectively, but higher in the volatile solid (VS) content (95.98 ± 0.285%). The C/N ratio of CP (13.60%) falls in the range of 10-20 and PM (9.50%), it is, however below the recommended optimum range of 20-30 (Sapkota et al., 2018). The proximate composition of digester feeds largely in uences the quality and biogas yield. Wastes from Plant materials such as crop residues are not easily digestible as animal wastes because of di culty in achieving hydrolysis of cellulose, hemicellulose, and lignocellulosic constituents (Bolaji and Adebayo, 2018).
The quality of biogas and its cumulative yield is largely dependent on the characteristics of the feedstock, as no one agricultural waste may be ideal enough to be used singly as biomass in biogas production. Optimum microbiological activities in anaerobic digestion and improved e ciency in biogas production, therefore, requires improving the characteristics of the feedstock and other operating conditions of the digester. This has been achieved by adopting different pre-treatment options (Wagne et al., 2018;Mishra et al., 2018;Kreuger et al., 2011;Mtui, 2009 ) and co-substrate fermentation (Tetteh et al., 2018;Sawyerr et al., 2017;Bhatnagar et al., 2017;Divyabharathi et al., 2017).
As could be observed in Fig. 1, biogas production in all the treatments started on day 1 though at very low volume and not ammable. A similar observation was reported by Opurum et al., (2019) in which goat manure was co-digested with poultry dropping and plantain peels for biogas production. The possible explanation for this is that many plant residues are not readily digestible, due to the slow rate of hydrolysis of complex polysaccharides and lignocellulosic constituents and hence a remarkable low gas production (Bolaji and Adebayo, 2018). The low biogas production at the early stage of anaerobic digestion is predictable because the rate of biogas production in batch operation is directly proportional to the speci c growth rate of methanogenic bacteria in the bioreactor (Nnabuchi et al., 2012.) Unless pretreated, the lignin component of lignocellulosic wastes creates a protective barrier that hinders plant materials from degradation by microbial consortia for conversion to bioenergy (Latinwo and Agarry, 2015).
Analysis of variance (ANOVA) of the means of the cumulative biogas yield indicates a signi cant difference (P ≤ 0.05) in all the mixed ratios compared to the control. The highest cumulative biogas yield was recorded in the bioreactor with CP/PM3:1(21.443dm 3 ). The observed increase in the means of cumulative biogas yield could be attributed to the supplementary effects (synergism) of the nutrient contents of the individual feedstock blended.
Previous studies on co-digestion of different organic substrates have shown a synergic effect of the combined treatments as the biodegradability of the resulting mixture was much higher than that of the single substrates when investigated separately (Esposito et al., 2012). Similar reports on the co-digestion of food wastes with livestock manure such as poultry dropping, cow dung, sewage sludge, or e uent have been shown to improves biogas yield and methane content while mono-substrate digestion was found to be mostly unstable ( Ofoefule and Uzodinma, 2009;Awogbemi, and Adeyemo, 2013;Zhang et al., 2013).
The bene ts of anaerobic co-substrate digestion include increased biogas production, enhanced degradation rates, and higher digester capacity. The bene cial effects of mixed feedstock majorly lie in balancing the C/N ratio, the nutrient contents of the given organic waste, increasing the pH buffering capacity, decreasing the ammonia toxicity and the accumulation of volatile fatty acids (VFAs), and improving the biochemical conditions for microbial growth (Lu et al., 2017), availability of micro-and macronutrient required by the microbial community, and dilution of inhibitory or toxic compounds. More so, co-digestion may improve the process kinetics rather than the bioavailability of the feedstock (Zamanzadeh et al., 2017). Evaluation of rates of hydrolysis using bio-methane potential (BMP) assays has shown that co-digestion increased hydrolysis rates when food waste and manure were co-digested compared to mono-digestion in BMP assays. The observed synergistic effect was associated with the dilution of inhibitory compounds and improved nutrient balance due to co-digestion (Zamanzadeh et al., 2017). There is no signi cant difference (P ≤ 0.05) between the experimental cumulative biogas yield and that predicted using the modi ed Gompertz model.
The treatments, CP/PM 2:1 and CP/PM 3:1 showed correlation coe cients (R 2 ) of 0.9964 and 0.9962 with maximum biogas production rates (U) of 2.03 and 2.12 dm 3 /day, respectively. The indication is that the Modi ed Gompertz model suitably described the methanogenic process in biogas production. The observed result follows the prediction that in batch system operation, the biogas production rate is directly proportional to the speci c growth rate of methanogens in the bioreactor (Nnabuchi et al., 2012). This observation aligns with previous reports (Yusuf et al., 2011;Yoon et al., 2014;Latinwo and Agarry, 2015;Yan et al., 2017). A similar report was made by Adamu et al., (2017), in which the rate of biogas production from abattoir waste was predicted using empirical models; the Gompertz model gave better goodness of t with correlation coe cients of 0.998. The experimental data generated from the treatments in the anaerobic co-digestion of cattle paunch manure and cow dung for biogas production were tted to the Modi ed Gompertz model, and they showed adequate t (Chukwuma and Orakwe, 2014). Ismail and Talib, (2014) evaluated anaerobic co-digestion of agro wastes for biogas recovery, the kinetics of the anaerobic digestion process was suitably described by the modi ed Gompertz model, and the experimental and predicted data of biogas production tted well, with correlation coe cient values > 0.96 implying favorable conditions of the process.

Conclusion
This study has shown that blending cassava peels (CP), an abundantly available agro-waste of no economic value with poultry manure (PM) at speci c ratios is capable of signi cantly improving biogas yield, and could be adopted in large scale biogas production for domestic use, and proper management of agro-wastes with a concomitant reduction in greenhouse gas emission. Cassava peels and poultry manure are good candidates for 'green energy' production, as suggested by the result of their physicochemical characteristics.
Modi ed Gompertz model gave a suitable description of the anaerobic digestion process, predicting biogas production rate (U), biogas production potential (Ym), and the lag period (λ). The modi ed Gompertz equation adequately tted into the experimental data, with correlation coe cients (R 2 ) > 099, which indicates a good process performance operated under a set of favourable conditions.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.