Elsevier

Bioresource Technology

Volume 359, September 2022, 127473
Bioresource Technology

Short Communication
Biogas production from residual marine macroalgae biomass: Kinetic modelling approach

https://doi.org/10.1016/j.biortech.2022.127473Get rights and content

Highlights

  • Low-solids anaerobic digestion of marine macroalgae waste was conducted.

  • Pseudo-first-order, logistics, modified, double and multi-Gompertz models were used.

  • All models fit the experimental data with R2 > 0.988.

  • Multi-Gompertz fitting showed the highest R2 and the lowest RMSE and AICc values.

Abstract

Modelling the conversion of residual biomass to renewable fuels is of high relevance to promote the development of effective technological solutions. The present study compares the performance of five different kinetic models (pseudo-first-order kinetics, logistics, modified Gompertz, double-Gompertz, and multi-Gompertz) to describe the cumulative methane production during a low-solids anaerobic digestion of marine macroalgae waste. Different substrate concentrations were evaluated (0.9, 1.7 and 2.5% TS) with the best methane yield (105.2 mL CH4.g VS−1) being obtained at the highest amount of biomass. All models fitted the experimental data with R2 > 0.988. The innovative multi-Gompertz model herein proposed led to the best performance indexes for all tested experimental conditions, allowing to predict methane yields more accurately when the digestion occurs in two or more steps, as it was the case with marine macroalgae waste.

Introduction

Organic wastes have been considered an important renewable resource for biogas production through anaerobic digestion (AD) (Hassaan et al., 2021, Li et al., 2012, Panigrahi et al., 2020). The absence of lignin in marine macroalgae makes them an interesting alternative feedstock for AD compared to lignocellulosic substrates, reaching relatively higher methane yields (Allen et al., 2013, Hassaan et al., 2021). Also, macroalgae derived biofuels avoid the land competition between agricultural and energy production sectors, as occurs with other feedstocks (Allen et al., 2015, Ap et al., 2021). However, the biogas and methane yields using marine macroalgae can vary significantly from species to species and with geographic distribution, essentially due to variations in the structure and composition of the biomass (Hassaan et al., 2021, Tabassum et al., 2018).

Marine macroalgae become waste (Marine Macroalgae Waste - MMW) when accumulated at the coastal areas, thus requiring appropriate management solutions (Pardilhó et al., 2022a, Pardilhó et al., 2021, Pardilhó et al., 2022b). Although positive effects can exist if MMW are in normal quantities (e.g., incorporation into the beach dynamics or food base to the food webs), frequently such biomass tends to accumulate rapidly overwhelming the ecosystem status, creating a negative impact. In fact, presently, MMW is mostly not collected, being left unmanaged at beaches and subjected to natural decomposition, or disposed of in landfills (less desirable solution according to the waste hierarchy). The AD of this unexplored biomass allows its valorisation as an organic resource and prevents several environmental (e.g., waste accumulation, greenhouse gases release by decomposition, changes in nutrient levels and accumulation of contaminants) and economic problems (affecting tourism and other beach based activities and representing a loss of a resource with economic value) caused by MMW accumulation at coastal areas (Pardilhó et al., 2022a, Pardilhó et al., 2021, Pardilhó et al., 2022b).

As in any biological/chemical process, studying AD kinetics is essential to evaluate the feasibility of the process (as well as the design of a biogas production plant). It provides information concerning the biodegradability rate of the substrates and also bottlenecks of the process that might affect digestibility and consequently methane yield (Allen et al., 2015, Karki et al., 2022, Tabassum et al., 2018).

Kinetic models are useful to optimise, simulate and monitor the performance of the process under different conditions (Hassaan et al., 2021, Pramanik et al., 2019). Several kinetic models have been used to predict methane production and determine the bio-kinetic process parameters. Pseudo-first-order kinetics, logistic function and modified Gompertz are some of the models widely used to fit experimental data of AD processes (either in mono- or co-digestion) (Allen et al., 2013, Donoso-Bravo et al., 2010, Hassaan et al., 2021, Karki et al., 2022, Li et al., 2012, Panigrahi et al., 2020). Other models are also used for more specific conditions such as higher solids contents (e.g., Chen and Hashimoto model), or specific microorganisms (e.g., cone model) (Karki et al., 2022, Lima et al., 2018, Masih-Das and Tao, 2018). Generally, AD kinetic models are mostly used to determine the lag phase (time taken for microbial adaptation to substrates and beginning of biomethane production), the biomethane potential (mL CH4.g−1) and the maximum methane production rate (mL CH4.g−1.day−1) (Allen et al., 2013, Donoso-Bravo et al., 2010, Hassaan et al., 2021, Karki et al., 2022, Li et al., 2012, Panigrahi et al., 2020, Pramanik et al., 2019).

The kinetics of biogas/methane production vary from one substrate to another. For that reason, models might be suitable to predict the degradation behaviour of the substrates under study and thus different kinetic models should be compared, also, under different experimental conditions. Some studies reported in the literature showed the degradation of the substrates occurring in several stages (Allen et al., 2013, Hassaan et al., 2021, Panigrahi et al., 2020) and evidenced the existence of more than one lag phase in the process. Some authors studied the multi-stage methane production through the combination of different models, with success (Karki et al., 2022, Lima et al., 2018, Masih-Das and Tao, 2018), although no studies were found evaluating the combination of more than one Gompertz model (one of the most used as single model) to describe the process. Thus, the present study aims to contribute for the knowledge related with MMW AD and find the best model to fit the kinetics of the process, by evaluating the most common models as well as the combined solutions, including the double and multi-Gompertz, supporting the development of the technology. For that purpose, five kinetic models were used to fit experimental data resulting from MMW anaerobic mono-digestion at different substrate concentrations, namely: (i) pseudo-first-order kinetics; (ii) logistic function; (iii) modified Gompertz; (iv) double-Gompertz; and (v) multi-Gompertz.

Section snippets

Anaerobic digestion assays

The AD assays were described by Pardilhó et al. (2022a). Briefly, MMW (87% TS; 67.5% VS) was used as substrate and anaerobically digested sludge (from a municipal wastewater treatment plant) as inoculum. The process was conducted, in duplicate, at mesophilic conditions (37 °C), under magnetic stirring (80 rpm), using a constant amount of inoculum (150 mL; previously incubated to ensure the consumption of biodegradable organic matter) and a variable amount of substrate, representing different

Results and discussion

The process was operated until little or absence of biogas production, which occurred after 38 operation days for 0.9 and 1.7% TS and after 58 days for 2.5% TS. The results from the experimental assays reveal a cumulative methane production of 48.4, 83.5 and 105.2 mL CH4.g VS−1 for 0.9, 1.7 and 2.5% TS, respectively. These results showed that, under the studied conditions, an increase in substrate amount (higher TS contents) leads to the production of higher methane volumes (Pardilhó et al.,

Conclusions

Mono-digestion of marine macroalgae waste, under the range of conditions studied (0.9 – 2.5% TS), was enhanced for higher substrate concentrations, being the highest yield obtained 105.2 mL CH4.g VS−1 for 2.5% TS. Regarding kinetic modelling, the innovative multi-Gompertz model presents the best fitting performance. The lower RMSE (0.49–2.56) and AICc (−1.6–105) and the higher R2 (>0.998) values obtained indicated that it is the most suitable kinetic model to describe the studied process and

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was financially supported by: LA/P/0045/2020 (ALiCE); UIDB/00511/2020 and UIDP/00511/2020 (LEPABE); UIDB/50020/2020 and UIDP/50020/2020 (LSRE-LCM), all funded by national funds through the Foundation for Science and Technology (FCT) funds/MCTES (PIDDAC). The authors also acknowledge FCT for Sara Pardilhó’s (SFRH/BD/139513/2018) PhD fellowship, funded by national funds and the European Social Fund (ESF).

Cited by (10)

View all citing articles on Scopus
View full text