Elsevier

Bioresource Technology

Volume 196, November 2015, Pages 279-289
Bioresource Technology

Improved ADM1 model for anaerobic digestion process considering physico-chemical reactions

https://doi.org/10.1016/j.biortech.2015.07.065Get rights and content

Highlights

  • The physico-chemical framework of the Anaerobic Digestion Model No. 1 is improved.

  • The effects of calcium and magnesium ions during anaerobic digestion are investigated.

  • The effects of inorganic components on the production of methane are tested.

Abstract

The “Anaerobic Digestion Model No. 1” (ADM1) was modified in the study by improving the bio-chemical framework and integrating a more detailed physico-chemical framework. Inorganic carbon and nitrogen balance terms were introduced to resolve the discrepancies in the original bio-chemical framework between the carbon and nitrogen contents in the degraders and substrates. More inorganic components and solids precipitation processes were included in the physico-chemical framework of ADM1. The modified ADM1 was validated with the experimental data and used to investigate the effects of calcium ions, magnesium ions, inorganic phosphorus and inorganic nitrogen on anaerobic digestion in batch reactor. It was found that the entire anaerobic digestion process might exist an optimal initial concentration of inorganic nitrogen for methane gas production in the presence of calcium ions, magnesium ions and inorganic phosphorus.

Introduction

The “Anaerobic Digestion Model No. 1” (ADM1) developed by the International Water Association (IWA) task group is a mathematical model mainly describing the biochemical processes involved in anaerobic digestion (Batstone et al., 2002, Batstone et al., 2006). The whole process of anaerobic digestion can be divided into five steps, which are disintegration, hydrolysis, acidogenesis, acetogenesis and methanogenesis. First, composite particulate substrate is disintegrated into carbohydrates, proteins, lipids, soluble and particulate inerts in the disintegration step. Then carbohydrates, proteins, and lipids are hydrolyzed into monosaccharides (MSs), amino acids (AAs), and long chain fatty acids (LCFAs) in the hydrolysis step (Vavilin et al., 2001). Next in the acidogenesis step, MSs and AAs are degraded by acidogenic bacteria into dissolved hydrogen, carbon dioxide and volatile fatty acids (VFAs), such as propionic acid, acetic acid, butyric acid and valeric acid. Then in the acetogenesis step, these VFAs, as well as some LCFAs from the hydrolysis step, are converted into acetates, dissolved hydrogen and carbon dioxide by acetogenic bacteria. Finally in the methanogenesis step, acetates from the acetogenesis step are converted by aceticlastic methanogenic bacteria into dissolved methane, while dissolved hydrogen and carbon dioxide are converted by hydrogen-utilizing methanogenic bacteria into dissolved methane. During the whole process of anaerobic digestion, all the bacteria (or component degraders) gradually decay and become inactive. These inactive degraders maintain in the reactor as a part of substrate and involve in anaerobic digestion with the original composite particulates. ADM1 employs a set of 24 differential rate equations to describe the bio-chemical processes involved in anaerobic digestion. The disintegration, hydrolysis and bacterial decay steps are represented by first order kinetics, while all the other steps are represented by Monod-type kinetics. Besides, ADM1 includes several inhibition factors such as LCFAs, dissolved hydrogen and ammonia (Angelidaki et al., 1993, Chen et al., 2008, Schievano et al., 2010, Zonta et al., 2013). In conjunction with the rate equations, 24 dynamic state concentration variables are set for the components involved in anaerobic digestion. These rate equations can be solved simultaneously and have complete mass balances over all components in the solid, gas and liquid phases if appropriate initial values are determined for the concentration variables.

The ADM1 modeling framework is a powerful tool that has been applied in the industrial process design and optimization for wastewater treatment (Batstone and Keller, 2003, Biernacki et al., 2013, Girault et al., 2012, Mairet et al., 2011, Parker, 2005, Shang et al., 2005). Many researches of anaerobic-digestion-involved processes are also based on the application of ADM1 modeling framework, especially the process of methane production (Antonopoulou et al., 2012, Hafez et al., 2010, Parawira et al., 2008, Takiguchi et al., 2004).

However, many deficiencies of the original ADM1 have been noted since its publication. First, the ADM1 does not accommodate complete component mass balances over the nitrogen and carbon components, which can result in the discrepancies between the carbon and nitrogen contents in the biomass and those in the composite particulate material (Blumensaat and Keller, 2005). Second, the ADM1 does not consider too much physico-chemical processes which are not directly mediated by microbes but can affect the bio-chemical processes of anaerobic digestion (Horiuchi et al., 2001, Loewenthal et al., 1989, Loewenthal et al., 1991, Mikkelsen and Keiding, 2002). The most critical physico-chemical processes omitted in the ADM1 are the solids precipitation processes caused by metal ions. The main reason of excluding the solids precipitation processes from the ADM is that the range of precipitating ions is wide, which leads to a large number of precipitate types (Ekama et al., 2006). Also, the presence of some types of metal ions may have inhibition effects on the precipitation processes involving other metal ion types. In addition, precipitates formed by same ions may exist two different forms: amorphous precipitates and crystalline precipitates, which have different formation mechanisms, precipitation kinetics and the rate-limited factors (Koutsoukos et al., 1980). Third, the ADM1 does not consider phosphoric acid and phosphate as components involved in the physico-chemical processes of anaerobic digestion. It has been shown in many studies that metal ions and phosphate may cause solids precipitation, such as struvite (MgNH4PO4), affect the pH of the anaerobic digestion circumstances and have strong effects on bio-chemical processes as well as physico-chemical processes (Britton et al., 2005).

The deficiencies of the original ADM1 will unavoidably limit the ability of the model to precisely represent the changing rates of some of the components and to correctly predict the final concentrations of these components. Recently, many studies have focused on developing a more complicated Anaerobic Digestion Model (Fedorovich et al., 2003, Musvoto et al., 2000a, Musvoto et al., 2000b, Sotemann et al., 2005a, Sotemann et al., 2005b). However, most of them do not consider critical inorganic components and physico-chemical processes in all three phases. In this study, the original ADM1 is extended to incorporate more inorganic components, such as metal ions and phosphates, and more physico-chemical processes, such as the association/dissociation processes of carbonate and phosphate ions and the solids precipitation processes of metal ions. Integrating a more detailed physico-chemical framework into ADM1 can enhance its abilities of keeping track of the change of circumstances pH value and each component involved in anaerobic digestion. The modifications to the original ADM1 may contribute useful information for its further development. After the model validation, the improved ADM1 is used to investigate the effects of some dissolved metal ions and inorganic components on the whole process of anaerobic digestion. Findings of the investigation may be useful for the design and scale-up of anaerobic digestion units for waste water treatment and biogas production processes.

Section snippets

Methods

The structures of the modified AMD1 are shown in Table 1, Table 2, Table 3, Table 4, Table 5. The entire model can be categorized into two bio-chemical framework and physico-chemical framework and contains totally 47 dynamic state variables representing the concentrations of 47 components in three phases during anaerobic digestion. In addition, the model describes 38 possible bio-chemical and physico-chemical processes involved in anaerobic digestion. The kinetic rate of each process is

Model validation

In order to testify the accuracy and the predictive ability of the physico-chemical framework of the modified model, the model outputs were compared with experimental data measured at 25 °C in the work of Musvoto et al. (2000a). As can be seen in Figs. 1 and 2, the model results accurately predict the changing trends of inorganic components and pH value in the batch reactor, and exhibit good agreement with the experimental results. The comparison demonstrates that the model is able to accurate

Conclusions

The original ADM1 has been modified by improving its bio-chemical framework and integrating a more detailed physico-chemical framework. The modified ADM1 was validated by a set of experimental data and used to investigate the effects of dissolved calcium and magnesium ions, inorganic phosphorus and nitrogen on anaerobic digestion in batch reactor. The modifications improved the ADM1’s ability to keep track of the biogas production in gas phase, the pH value in liquid phase and the precipitates

Acknowledgements

We acknowledge the National Science Foundation (CBET 1138734, CHE 1230803) for financial support.

References (30)

Cited by (28)

  • Modification of anaerobic digestion model No.1 with Machine learning models towards applicable and accurate simulation of biomass anaerobic digestion

    2023, Chemical Engineering Journal
    Citation Excerpt :

    Therefore, the simulation of anaerobic digestion process by theoretical or numerical simulation is an effective means to explore the optimal regulation of anaerobic digestion process [12]. Since the end of the last century, many scholars have devoted themselves to the mathematical model of anaerobic digestion [13], in order to digitize the reaction process of anaerobic digestion [14,15]. After repeated practice and comparison by many scholars, it is found that the ADM1 proposed by International Water Association (IWA) in 2002 has higher accuracy and operability [16].

View all citing articles on Scopus
View full text