A syntrophic propionate-oxidizing microflora and its bioaugmentation on anaerobic wastewater treatment for enhancing methane production and COD removal

Abstract

Syntrophic propionate-oxidizing microflora B83 was enriched from anaerobic sludge.

The bioaugmentation of microflora B83 were evaluated from wastewater treatment.

Methane yield and COD removal were enhanced by bioaugmentation of microflora B83.

Hydrogen-producing acetogensis was a rate-limiting step in methane fermentation.

Methane fermentation process can be restricted and even destroyed by the accumulation of propionate because it is the most difficult to be anaerobically oxidized among the volatile fatty acids produced by acetogenesis. To enhance anaerobic wastewater treatment process for methane production and COD removal, a syntrophic propionate-oxidizing microflora B83 was obtained from an anaerobic activated sludge by enrichment with propionate. The inoculation of microflora B83, with a 1:9 ratio of bacteria number to that of the activated sludge, could enhance the methane production from glucose by 2.5 times. With the same inoculation dosage of the microflora B83, COD removal in organic wastewater treatment process was improved from 75.6% to 86.6%, while the specific methane production by COD removal was increased by 2.7 times. Hydrogen-producing acetogenesis appeared to be a rate-limiting step in methane fermentation, and the enhancement of hydrogen-producing acetogens in the anaerobic wastewater treatment process had improved not only the hydrogen-producing acetogenesis but also the acidogenesis and methanogenesis.

Graphical abstract

Keywords

Anaerobic wastewater treatment / Methane production / Hydrogen-producing acetogenesis / Methanogenesis / Rate-limiting step / Bioaugmentation

1 1 Introduction

Urban wastewater treatment plants (WWTPs) are important infrastructure for centralised treatment, standard discharge, and resource recovery of wastewater, which are related to the urban ecological environment, public health, and sustainable development. However, while improving water quality, these facilities also consume considerable amounts of electricity and emit greenhouse gases. According to estimates, the energy consumption and carbon emissions of the global wastewater treatment industry account for 3% and 1.6%, respectively (Lu et al., 2018). As an energy-intensive industry, WWTPs have remarkable emission reduction potential and are an important path to achieving sustainable development (Liu et al., 2024). Research indicates that the carbon emissions from WWTPs are closely linked to operational conditions, treatment processes and influent water quality (Zhang et al., 2023). Among them, the evaluation and selection of wastewater treatment processes are pivotal in the design and construction of WWTPs. Selecting the appropriate wastewater treatment process based on the principles of resource recovery, economic efficiency and environmental protection is beneficial to improve operational efficiency, reduce operating costs and promote the sustainable development of wastewater treatment (Huo et al., 2023; Lu et al., 2023).

Currently, existing wastewater treatment processes can be broadly categorised based on the following types: activated sludge processes, such as oxidation ditch (OD), sequencing batch reactor (SBR), anaerobic–anoxic–oxic (AAO) and its derivative processes, conventional activated sludge (CAS); membrane processes such as membrane bio-reactor (MBR); and biofilm processes, such as moving bed biofilm reactor (MBBR) and integrated fixed-film activated sludge (IFAS). Among these processes, activated sludge process is still at the heart of current sewage treatment technology (van Loosdrecht and Brdjanovic, 2014). For example, AAO, SBR, OD and their modified processes are widely used in China (Zhang et al., 2021). Meanwhile, biofilm have been increasingly used in WWTPs, and this approach can address some of the drawbacks of using the sludge processes (Mpongwana and Rathilal, 2022). Furthermore, with the demands of resource recycle and low greenhouse gas emission, emerging integrated processes and technologies have been developed, such as AAO-VMBR (Ma et al., 2023), anammox processes (Shaw et al., 2024).The type of process not only affects the wastewater treatment efficacy but also serves as a major influencing factor in carbon emissions (Xi et al., 2021). Several studies have focused on evaluating and comparing these processes based on single indicators such as cost, water quality and carbon emissions to guide the selection of wastewater treatment processes. For instance, previous studies compared the energy consumption of several common treatment processes, ranking them in order of highest to lowest as follows: CAS > SBR > AAO/AO > OD (Gu et al., 2023), and the carbon dioxide emissions of the SBR process were twice that of the AAO process, with 10% being directly emitted (Nguyen et al., 2019). For the common CAS and MBR processes, CAS had lower energy consumption and cost, while MBR had better treatment efficiency and environmental performance (Bertanza et al., 2017). However, comprehensive process evaluations and comparisons involving these emerging technologies are currently lacking.

Nowadays, wastewater treatment plants are focusing not only on the quality of the effluent and the level of operational costs but are also increasingly prioritizing carbon emissions as a key concern. The effluent quality index (EQI), operating cost index (OCI) and greenhouse gas emissions (GHG) are three objectives for evaluating sustainability performance in wastewater treatment. Thus, a comprehensive comparison that integrates the three indicators of EQI, OCI, and GHG is more scientifically sound and rational (Lu et al., 2024). However, inevitable constraints exist between the three objectives, hindering effective decision-making for process selection (Sweetapple et al., 2014; Ortiz-Martínez et al., 2021). Therefore, a multi-objective evaluation method should be adopted to comprehensively consider and balance the three objectives. Pareto Optimality decision theory is one of the methods suitable for solving multi-objective optimisation problems (Wang et al., 2023). The Pareto Optimal state represents an ideal state in resource allocation, and these optimal solutions are non-dominated, allowing decision makers to select the optimal solution based on actual conditions (Dai et al., 2023; Wang et al., 2017).

Numerous studies have adopted the Pareto Optimality concept to perform process control and optimisation for wastewater treatment through multi-objective genetic algorithms (NSGA). For example, NSGA-II was combined with a commonly used WWTP model, demonstrating the feasibility of using this algorithm to determine Pareto Optimality (Béraud et al., 2007), NSGA-II algorithm was applied in the multi-objective optimization of AAO processes, achieving greenhouse gas reduction, improved effluent quality, and reduced operating costs (Dai et al., 2023; Sweetapple et al., 2014; Muschalla, 2008). Additionally, given the complexity of the wastewater treatment process, model simulation provides an efficient approach for WWTP design and process selection (Cao et al., 2021). Among numerous simulation software, GPS-X demonstrates advantages in model generalisation, simulation and result interpretation (Jasim, 2020). The primary functions of GPS-X include optimizing and managing wastewater treatment plants, offering a theoretical basis for the modernization and improvement of wastewater processes (Wondim et al., 2023; Srivastava et al., 2024). Though the modeling could not exactly consistent with the actual WWTPs, it can provide important references for process operation. Consequently, this study opted for the GPS-X software to construct the model processes, ensuring the accuracy of the output Pareto set.

This study aims to comprehensively evaluate the operational efficiency and carbon emission reduction potential of 14 wastewater treatment processes, including mature and conceptual techniques. To achieve this objective, process simulation by GPS-X software and the concept of “Pareto Optimality” were combined for the first time, integrating three indicators of water quality, operational costs, and carbon emissions to compare different technologies. Initially, process modeling and simulation were conducted using GPS-X simulation software to obtain baseline data for the evaluation. Subsequently, a multi-objective evaluation system was established and the characteristics of the three indicators for each process under a baseline state were calculated, resulting in a set of Pareto solutions. Finally, a non-dominated sorting method was employed for multi-objective process selection to obtain the “Pareto optimal” scenario. The results of this study provide reference suggestions for future WWTPs to select appropriate processes to achieve carbon neutrality.

2 2 Materials and methods

2.1 2.1 Framework of multi-objective evaluation and selection methods for processes

The overall framework of the multi-objective process evaluation and selection method used in this study is shown in Fig.1. Taking carbon emissions, operating costs, and effluent water quality as the three evaluation criteria, GPS-X simulation software is used to model and tune more than ten wastewater treatment processes, so that each process can meet emission requirements under corresponding inlet and outlet scenarios. Based on the built-in cost accounting component and water quality accounting component of GPS-X (Hydromantics 8.0.1, Canada), the EQI and OCI of each process under different scenarios are calculated. At the same time, the carbon emissions of each wastewater treatment process are calculated using the intermediate operating parameters of simulation, and form a Pareto set with the OCI and EQI. Finally, the data set in the Pareto set are sorted using the non-dominated sorting method, and the optimal process under each scenario is determined from the Pareto level 1 solution obtained.

Fig.1 Framework of multi-objective evaluation and selection methods for wastewater treatment processes.

Full size|PPT slide

2.2 2.2 Process evaluation scope and scenario setting

This evaluation scope covered 14 wastewater treatment processes, including 7 AO/AAO processes (AO, AAO, Bardenpho, Step-AO, Step-CAS, JHB, MUCT), 2 MBR processes (AAO-MBR, AAO-VMBR), 2 biofilm processes (AO-BAF, IFAS), as well as Carrosel-OD and SBR processes. In addition, it also involved anammox process.

To improve the representativeness of the evaluation, different scenarios were created by adjusting the model’s influent water quality and effluent standards. Three influent components representing the concentration of typical wastewater pollutants in China were selected for the influent scenario. The effluent standard referred to the “Grade I–A” standard in the most widely used “Urban wastewater Treatment Plant Pollutant Discharge Standards (GB 18918-2002)” in China, while also considering the “Class IV” standard in the “Surface Water Environmental Quality Standard (GB 3838-2002)” under the improved effluent discharge standard requirements, as shown in Tab.1. Where H, M, L are used to represent the high, medium, and low conditions of influent concentration, and A and IV are used to represent the Grade I-A and Class IV discharge standards. Therefore, six scenarios were established for different influent and effluent conditions, namely as H-A, M-A, L-A H-IV, M-IV, and L-IV.

Tab.1 Scenario settings for influent and effluent water quality

Water quality indicators	Influent quality (mg/L)				Effluent standard (mg/L)
	H	M	L		Grade I-A	Class Ⅳ
COD	1000	500	250		50	30
BOD	400	220	110		10	10
SS	350	200	100		10	10
TN	85	40	20		15	10
NH4+–N	64	30	15		5	1.5
TP	15	8	4		0.5	0.3

2.3 2.3 Process simulation

All wastewater treatment processes were modeled and simulated using GPS-X software. The model input variables were influent conditions, structure parameters, stoichiometric parameters and kinetic parameters. In the model, all operational parameters are derived from the software’s built-in parameters, and optimization settings have been applied to DO, SRT, and other parameters in accordance with effluent standards. Compared to the operational parameters of actual wastewater treatment plants, these settings are within a reasonable range. The output variables were the operating results such as effluent quality and sludge volume. Intermediate variables such as inlet and outlet water quality, energy consumption, and dosage of chemicals for each structure can also be calculated and displayed. Except for vibrating membrane bioreactors and mainstream Anammox processes, the layout of other processes was established based on the sample layouts in the model library of GPS-X software. The mainstream Anammox process was established based on the existing side-stream Anammox process layout in the sample layout. The AAO-MBR process is established based on a pilot-scale study (Ma et al., 2023), and the energy consumption is reduced by one empirical value, while energy consumption is reduced by one empirical value, which is 10% of the water inflow due to reduced aeration caused by membrane vibration. Therefore, MBR components were used as a substitute and parameters are modified accordingly. The structural design parameters and operating parameters of all processes were appropriately fine-tuned based on effluent quality on the basis of actual engineering and literature reported values, and stoichiometric and kinetic parameters used default values. The schematic diagrams of two representative processes (AAO-VMBR and AO-BAF) are shown in Fig.2, while the schematic diagrams and parameter settings of all processes are provided in the Fig. S1 and Tables S1–S4, respectively.

Fig.2 The schematic diagram of partial representative wastewater treatment process model. (A) AAO-VMBR; (B) AO-BAF.

Full size|PPT slide

2.4 2.4 Evaluation criteria

2.4.1 2.4.1 Effluent quality index (EQI)

EQI is used to evaluate the average total discharge of pollutants in daily effluent to measure whether the effluent can be safely discharged and reused. EQI is calculated by weighting the pollutants that significantly affect the effluent quality over a period of time, taking into account six indicators: COD, BOD₅, TSS, Soluble PO₄–P³⁺, NO_x (NO₂⁻ + NO₃⁻), and TKN. Their weight factors are set to 1, 2, 2, 2, 1, and 20 respectively by default in the software. The calculation formula is shown as (Eq. (1)):

(1)$\text{EQI}=Q{\sum }_{i=1}^n{w}_i{S}_i,$

where Q: Effluent flow from the WWTP, (m³/d); n: Number of pollutants considered in EQI evaluation; $w_i$: Weight of the ith pollutant in EQI; $S_i$: Concentration of the ith pollutant in EQI.

2.4.2 2.4.2 Operational cost index (OCI)

In the analysis, two parameters are used to characterize: the operating cost per ton of water and the operating cost per unit of EQI removal. The operating cost mainly includes energy costs, chemical costs, and sludge disposal costs. The energy costs include aeration energy consumption costs, pumping energy consumption costs, mixing energy consumption costs, heating energy consumption costs, and other energy costs. The total operating cost calculation formula is shown in Eq. (2). Similar to EQI, GPS-X software can directly output the OCI values corresponding to each process.

(2)$\mathrm{O}\mathrm{C}\mathrm{I}={\mathit{C}}_{\mathrm{a}\mathrm{r}\mathrm{e}}+{\mathit{C}}_{\mathrm{c}\mathrm{h}\mathrm{e}}+{\mathit{C}}_{\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{p}},$

where OCI: Operational Cost Index, measured in $/m³ or $/t EQI (per ton of EQI); $C_{\mathrm{a}\mathrm{r}\mathrm{e}}$: Energy cost, electricity fee 0.1 $/kWh; $C_{\mathrm{c}\mathrm{h}\mathrm{e}}$: Chemical cost, with carbon source cost at 2.0 $/kg,PAC(polyaluminumchloride)costat0.5$/kg, and sludge pretreatment chemical cost at 1.0 $/kg; $C_{\mathrm{d}\mathrm{i}\mathrm{s}\mathrm{p}}$: Sludge disposal cost, with sludge transportation cost at 80 $/t.

2.4.3 2.4.3 Carbon emission index (GHG)

It is characterized by two indicators: carbon emissions per ton of water and carbon emissions per unit of EQI. The scope of carbon emissions accounting involves direct and indirect carbon emissions generated by all structures from the entry of wastewater into the plant area to the discharge of effluent, excluding carbon emissions generated during non-treatment processes such as equipment repair and maintenance, material transportation, and basic work and life of employees in the plant area (Fig. S2). The carbon emissions accounting includes: (1) direct CO₂ emissions during anaerobic treatment, (2) direct CO₂ emissions during aerobic treatment, (3) direct N₂O emissions during anoxic treatment, (4) direct CH₄ emissions during wastewater treatment, (5) indirect carbon emissions generated by external carbon sources, (6) indirect carbon emissions generated by power consumption, (7) indirect carbon emissions generated by chemicals production, (8) indirect CH₄ emissions during wastewater discharge, (9) indirect N₂O emissions during wastewater discharge, and (10) direct GHG emissions during sludge anaerobic digestion. Among them, the three direct emissions of (1), (2), and (10) are calculated using the mass balance method (Eq. (3)); the rest of the processes are conveniently calculated using the emission factor method (Eq. (4)). The formulas used are referenced from the “Provincial Greenhouse Gas Inventory Compilation Guidelines (Trial)”, which is published by National Center for Climate Change Strategy and International Cooperation. For specific calculations of biochemical processes, power and chemicals carbon emissions for each structure, refer to the Text S1 and Text S2.

(3)$\mathrm{G}\mathrm{H}\mathrm{G}=\mathrm{A}\mathrm{D}\times \mathrm{E}F\times \mathrm{G}\mathrm{W}\mathrm{P},$

(4)$\mathrm{G}\mathrm{H}\mathrm{G}=\mathit{Q}\left({\mathit{C}}_{\mathrm{i}\mathrm{n}}-{\mathit{C}}_{\mathrm{o}\mathrm{u}\mathrm{t}}\right)\times \mathrm{E}\mathrm{F}\times \mathrm{G}\mathrm{W}\mathrm{P},$

where GHG: Greenhouse Gas emissions, measured in t CO₂eq/m³ or t CO₂eq/t EQI; AD: Emission source activity data, with the unit depending on the calculated emission source; EF: Emission factor, with the unit depending on the unit of activity data; $C_{\mathrm{i}\mathrm{n}}$: Concentration of the corresponding pollutant in influent, measured in mg/L; $C_{\mathrm{o}\mathrm{u}\mathrm{t}}$: Concentration of the corresponding pollutant in effluent, measured in mg/L; GWP: Global Warming Potential, where ${\mathrm{G}\mathrm{W}\mathrm{P}}_{{\mathrm{C}\mathrm{H}}_4}$ is 25 t CO₂eq/t CH₄, and ${\mathrm{G}\mathrm{W}\mathrm{P}}_{{\mathrm{N}}_2\mathrm{O}}$ is 298 t CO₂eq/t N₂O.

2.5 2.5 Non-dominated sorting method

This study refers to the Pareto optimal concept and introduces the upstream logic of the NSGA-II algorithm, the non-dominated sorting method, for evaluation. A circular-based traversal sorting method is adopted, and the “dominance” relationships between each scenario point and other scenario points were examined. The optimal scenarios in the “non-dominated” state were selected in each round, until all remaining scenarios were not mutually dominated. Since the scenarios with Pareto level 1 obtained from multi-objective sorting may not be unique, the optimal processes selected for each scenario are filtered based on the number of processes. If the number of processes is small, a direct comparison can be made, giving priority to a single criterion (e.g., carbon reduction orientation). If the number of processes is large, a normalized comprehensive indicator can be calculated by setting weight factors. Matlab software (2021 version) was used to implement the non-dominated sorting and draw the Pareto frontier.

3 3 Results and discussion

3.1 3.1 Characteristics of effluent water quality from different wastewater treatment processes

Fig.3 shows the effluent water quality evaluation results of 14 wastewater treatment processes under six influent and effluent scenarios. After parameter adjustment, the effluent pollutant concentrations of each process meet the emission requirements of the corresponding scenario (H-A, M-A, L-A H-IV, M-IV, and L-IV) as described in the Table S5. Fig.3(A) shows the considerable differences in effluent quality between various scenarios and processes. As the influent pollutant load increases, the EQI of each process notably increases. In addition, the effluent EQI of the SBR and mainstream Anammox process in the H-A scenario is substantially higher than that of other processes. The mainstream Anammox process has poor treatment performance due to the slow growth rate, environmental sensitivity and susceptibility to high organic load (Adams et al., 2024). Notably, the effluent quality of the AO-BAF process is superior under various water quality scenarios, with an average value of 0.036 kg EQI/m³. The removal of pollutants from wastewater mainly depends on the adsorption and oxidation of biomass, as well as the physical filtration effect of suspended fillers (Chang et al., 2009). Compared with traditional activated sludge processes, the biomass concentration of the biofilm formed on the inert particles is considerably higher than that in traditional activated sludge systems and can capture suspended solids through the filler medium, achieving excellent treatment efficiency in a limited residence time (Chen et al., 2017; Xu et al., 2022). In addition, the average effluent quality effects of OD process, IFAS and step AO processes are also relatively good under all scenarios.

Fig.3 Effluent water quality characteristics of wastewater treatment process. (A) The EQI value per ton of effluent for different wastewater treatment processes under six scenarios (region ① represents AO/AAO process types, region ② represents SBR process types, region ③ represents OD process types, region ④ represents biofilm process types, region ⑤ represents MBR process types, and region ⑥ represents anammox process types). (B) The distribution of EQI per ton of effluent from different types of wastewater treatment processes under the baseline scenario (M-A). (C) The removal amount of EQI per ton of effluent by each process under the baseline scenario (M-A). (D) The proportion of the EQI converted from the concentration of various pollutants in each process effluent (M-A).

Full size|PPT slide

The influent water quality of most actual WWTPs is close to level M, and the effluent limit is lower than standard A. Thus, the M-A scenario is used as the baseline scenario for evaluating the operational performance of various processes. Fig.3(B) shows the EQI of different categories of processes under the baseline scenario after preliminary classification. Overall, the biofilm process is superior to other processes in terms of effluent water quality, while the EQI of common AO and AAO processes is not ideal. Further comparison of the pollutant removal performance of each process reveals that the biofilm AO-BAF and IFAS processes have an EQI removal of 2.28 and 2.24 kg/m³, respectively, while the AO and MUCT processes are ~2.20 kg/m³. The overall level of membrane processes is relatively good, ranging from 2.23 to 2.24 kg/m³ (Fig.3(C)). As shown in Fig.3(D), if nitrogen removal is the goal, then IFAS and Step-CAS processes may be more suitable; if carbon removal is the goal, membrane processes are more suitable; while if phosphorus removal is the main goal, then SBR and AO-BAF may be not suitable.

3.2 3.2 Comparison of operating costs of different wastewater treatment processes

The operational cost evaluation results of 14 wastewater treatment processes under six scenarios are shown in Fig.4. It is evident in Fig.4(A) that the overall level of OCI per ton of water for the Class IV standard is notably higher than that for the Grade I-A Standard for the same process. That is, a strict effluent standard leads to a high corresponding operational cost. Under the high effluent standard, among the 14 wastewater treatment processes, the AO-BAF processes have the lowest overall operational cost per ton of water, is 0.18 $/m³ for scenario H-IV. Under the same scenario (H-IV), the operating costs of the activated sludge processes are 0.21 $ /m³ , which is roughly close to the results obtained from a previous study, in which the operating cost of an actual wastewater plant was 0.27 $/m³ (Zahid, 2007). The operating costs of various processes under the baseline scenario were depicted in Fig.4(B) and Fig.4(C). In this scenario, the MBR processes exhibited the highest operating costs, with an average of 0.082 $/m³.Thebiofilmprocessesfallinthemiddlerange(0.072$/m³), with operating costs slightly higher than activated sludge processes (0.059 $/m³). Biofilm processes have a strict requirement for influent SS, generally requiring SS ≤ 100 mg/L (Canler and Perret, 1994), thus necessitating pretreatment of the influent and increasing the treatment cost. Moreover, biofilm processes require specialized aeration and filter media equipment whose maintenance and replacement increase the overall process cost (Xu et al., 2022). Anammox process has the lowest operational costs, with OCI per ton of effluent of 0.047 $/m³. This low cost is due to Anammox bacteria, which are autotrophic anaerobic denitrifiers that can reduce external carbon sources and aeration inputs, thus saving operational costs to some extent (Ma et al., 2016; Xing et al., 2022).

Fig.4 Characteristics of operating costs for wastewater treatment processes. (A) The OCI of per ton of effluent from wastewater treatment process in six scenarios. (B) The distribution of OCI per ton of effluent from different types of wastewater treatment processes under the baseline scenario (M-A). (C) The composition of the cost per ton of effluent for each process under the baseline scenario (M-A). (D) The Composition of electricity usage costs for various processes (M-A).

Full size|PPT slide

The costs of electricity and chemicals constitute the primary components of the operational expenses of wastewater treatment plants (Kim et al., 2024). Under the baseline scenario, the sources of the overall operational expenses of 14 wastewater treatment processes are ranked in descending order of contribution rate, with chemical addition, electricity usage and sludge transportation and disposal accounting for 47.69%, 24.61%, and 27.69%, respectively (Fig.4(C)). Notably, the highest-cost processes, AAO-MBR and SBR, have higher electricity costs and chemical addition costs. The notable cost difference between processes largely depends on the amount of chemical agent applied. Therefore, WWTPs should avoid excessive use of chemicals, such as phosphate removal agents, disinfectants, exogenous carbon sources, and sludge removal agents, during the process selection and design phase to reduce operational expenses. Adopting automatic dosage and process control systems during the operational phase is recommended to ensure precise dosage, thereby reducing agent consumption and lowering operational expenses. In terms of energy consumption, the average proportions of aerating, pumping and mixing energy consumption in the 14 wastewater treatment processes for the baseline scenario are 38.51%, 17.71%, and 23.06%, respectively (Fig.4(D)). Considerable variations exist in aerating energy consumption among different wastewater treatment processes. Normally, conventional aeration tanks maintain operation at a dissolved oxygen concentration of 2.0 mg/L, with aeration power consumption accounting for approximately 50% of the total energy consumption of WWTPs (Qambar and Al Khalidy, 2022). In our study, the aeration tank energy consumption proportion of the AO and AAO processes is approximately 48.56%, which is close to it. Especially, the MBR processes require extensive aeration to provide a sufficient oxygen transfer rate due to longer sludge retention times and higher mixed liquor suspended solid concentrations in the tank, which can lead to membrane fouling. Therefore, extensive aeration is necessary to provide sufficient oxygen transfer rate, resulting in high aerating energy consumption (53.26% to 66.54%) (Tang et al., 2022). Overall, under the baseline scenario among several widely used processes in China, the total energy consumption decreases in the order of MBR > SBR > OD > AAO > AO. The main source of energy consumption is aeration energy; so effectively improving aeration efficiency or enhancing the microbial utilization of oxygen remains a key challenge for current WWTPs to reduce energy consumption and operating costs.

3.3 3.3 Carbon emission evaluation of different wastewater treatment processes

Fig.5 shows the evaluation results of carbon emission intensities for 14 wastewater treatment processes. When using the GHG per ton of water as an evaluation metric, considerable differences in carbon emission intensities are observed for the same treatment process under various influent and effluent scenarios. Overall, a strong correlation exists between the removal amount of pollutants and the corresponding carbon emission intensity (Fig.5(A)). The carbon emissions per ton of water are high under scenarios with high influent pollutant concentrations and strict effluent standards. For instance, the average carbon emission intensity for the 14 processes in the H-IV scenario is 3.6 times higher than that in the L-A scenario. Processes in the wastewater treatment must increase aeration or chemical dosage to meet water quality requirements and achieve high pollutant removal, leading to high carbon emissions. Under the same influent and effluent conditions, the GHG per ton of water for each process is similar. In the baseline scenario, the carbon emissions per ton of water for the 14 wastewater treatment processes range from 0.88 to 1.10 kg CO₂eq/m³. The carbon emission intensities for AAO-VMBR and AAO-MBR processes are relatively higher, while those for MUCT, Anammox, AO-BAF are at lower levels. Notably, the carbon emission intensities calculated in this study are higher than the average carbon emission intensities of wastewater treatment plants in Beijing (0.603 kg CO₂eq/m³) (Zhou et al., 2022). This finding is primarily due to the scope of this study, which further considers the carbon emissions associated with sludge thickening, dewatering, anaerobic digestion and transportation. It is worth noting that Anammox process has the lowest GHG per ton of water (Fig.5(B)), which is consistent with the findings regarding electricity consumption and chemical dosage for the baseline scenario. The energy consumption for the Anammox process is 0.25 kWh/m³, respectively, and the coagulant consumption is 0.89 g Al/m³, both of which are at a relatively lower level.

Fig.5 Carbon emission characteristics of wastewater treatment processes. (A) The GHG of per ton of effluent from wastewater treatment processes in six scenarios. (B) The distribution of GHG per ton of effluent from different types of wastewater treatment processes under the baseline scenario(M-A). (C) The proportion of direct and indirect carbon emissions of each process in the baseline scenario(M-A). (D) Carbon emission composition of each process in the baseline scenario(M-A).

Full size|PPT slide

As illustrated in Fig.5(C), the direct and indirect carbon emissions from various treatment processes under the baseline scenario are compared to further elucidate the sources of carbon emissions and identify directions for carbon reduction. The direct carbon emission levels of 14 processes are similar, with a range of 0.56–0.62 kg CO₂eq/m³, because the pollutant removal amounts of each process are similar under the same scenario. However, considerable differences are observed in the indirect carbon emissions among various processes, ranging from 0.25 to 0.54 kg CO₂eq/m³, primarily due to differences in chemicals and energy consumption among different processes. In the simulation results, the direct carbon emissions from processes are higher than their indirect emissions, indicating that the carbon emission intensity of wastewater treatment processes is predominantly driven by the direct emissions from the biochemical treatment phase, consistent with the findings of a previous study (Li et al., 2017). Furthermore, the detailed analysis of carbon emissions sources reveals that the direct emissions of CH₄, indirect electricity emissions and indirect chemical agent emissions are substantial, accounting for 36.47% to 52.85%, 23.65% to 36.92% and 0.12% to 3.90% of the total, respectively (Fig.5(D)). The indirect carbon emissions can be reduced through energy harvesters that utilize energy from the environment for power (Huo et al., 2023). The differences in direct CH₄ emissions from the biochemical process among various processes were negligible, revealing an average value of 0.463 ± 0.005 kg CO₂eq/m³. In all processes, the Anammox process has smaller agent inputs and energy consumption for aeration, exhibiting good potential for carbon reduction. Notably, Anammox is widely recognized as a low-carbon and energy-efficient technology due to its high energy utilization efficiency and low carbon footprint (Arora et al., 2021). Despite the extensive study of Anammox, its application in mainstream wastewater treatment remains challenging due to its slow growth, environmental sensitivity, and strict requirements for influent loadings (Ahmad et al., 2023).

3.4 3.4 Multi-Objective comparison under different scenarios

The comparison and selection of wastewater treatment processes are the foundation for the design, operation and upgrading of wastewater treatment facilities. Identifying the optimal process under various influent and effluent scenarios is necessary to enhance the multifaceted performance of the wastewater treatment facility. The evaluation indicators include GHG with the removal of unit EQI, OCI with the removal of unit EQI and EQI per ton of water to minimize the differences in the same process under different scenarios, and to enhance the comparability among different processes. The comparison and selection of processes under different scenarios were conducted using a multi-objective non-dominated sorting approach. Fourteen wastewater treatment processes had formed a total of 8 Pareto levels under 6 scenarios. The Pareto front with level 1 is shown in Fig.6, and the ranking results of the other levels are shown in the Table S6. The Pareto frontier for level 1 consists of 14 scenarios, including five AO-BAF process scenarios, five AAO-VMBR process scenarios, two AAO-MBR scenario, one Step-AO scenario, one AAO-MBR, and one MUCT scenario (Tab.2). Among them, AAO-VMBR and AO-BAF are nearly the optimal processes under all influent and effluent scenarios. Especially when the effluent standard is more stringent, such as Class IV standard, where most processes reach their limits and require more extreme operating conditions to improve water quality, leading to increased operational carbon emissions and costs, AO-BAF and AAO-VMBR still stand out from other processes due to their superior effluent quality, lower operational costs, and lower carbon emissions. Currently, regions like Jiangsu Province in China have adopted Class IV as a reference for setting local standards. In the long run, as environmental protection requirements for water become more stringent, it is imperative to raise the effluent standards for wastewater treatment. The “Pareto optimal” scenarios identified in this study occur in six high water quality standard scenarios, indicating that stricter standards do not necessarily result in a disadvantage in the comparison and selection process. However, based on objective data, there is still a significant increase in operational costs and carbon emissions overall. From these results alone, AAO-VMBR and AO-BAF processes are the top candidates for consideration under high emission standards. The BAF process exhibits superior performance in scenarios of high discharge, a characteristic that is intrinsically linked to its operational parameters. It can be seen from Fig.3 that the effluent water quality of AO-BAF process is better under each scenario, and the removal EQI of AO-BAF process reaches 2.28 kg/m³, which is the highest among all processes. Considering the common scenarios, AO-BAF process is the optimal process. Moreover, this process is relatively low in energy consumption (Huang et al., 2023), which was also proved in this study.

Tab.2 The optimal processes for different influent and effluent scenarios

Influent quality	Effluent standard	Optimal process
H	Grade I-A	AAO-VMBR, AAO-AMX, AO-BAF
H	Class Ⅳ	AO-BAF
M	Grade I-A	AO-BAF, Step-AO, MUCT, AAO-MBR, AAO-VMBR
M	Class Ⅳ	AO-BAF, AAO-VMBR
L	Grade I-A	AAO-VMBR
L	Class Ⅳ	AO-BAF, AAO-VMBR, AAO-MBR

Fig.6 The Pareto front of selected wastewater treatment processes.

Full size|PPT slide

Anammox processes was eliminated in the comparison. Notably, the reported successful applications of anammox process are mainly achieved through coupling with partial nitrification and denitrification to achieve stable nitrite supply (Adams et al., 2024). However, these coupling processes are currently difficult to be accurately simulated, thus this study may underestimate the potential of anammox process application. Moreover, the Pareto ranking of level 1 also includes the MUCT and Step-AO processes under the baseline scenario, while emerging technologies such as IFAS have not shown absolute advantages. Therefore, under current conditions, the optimization and upgrading of wastewater treatment processes can be fully studied within the existing mature technologies.

Compared to previous studies, this study is the first comprehensive evaluation of wastewater treatment processes conducted from three perspectives: carbon emissions, effluent quality, and operational costs. However, there are also some limitations. Take AAO process as an example, since this study did not account for the economic benefits associated with the reuse of water, the OCI value is comparatively higher than in other studies (0.21 $/m³) (Ni and Wang, 2024); Compared to previous studies, the AAO process in this research exhibits differences in terms of carbon emissions(2.25 kg CO₂eq/m³) for the H-IV scenario, which is mainly due to variations in aeration rates and chemical usage (He et al., 2023); the carbon emission of the OD process is 1.07 kg CO₂eq/m³ for the H-IV scenario, which is similar to the research conducted by (Masuda et al. (2018) (0.92 kg CO₂eq/m³) . Based on the results of this study, there are some issues could be considered in the future. First, the main parameters for process simulation, such as structure size, stoichiometric parameters, and kinetic parameters, are mainly based on the default values in GPS-X software, which can be calibrated and localized with extensive data from actual wastewater plants in the future studies. Secondly, equal consideration is given to optimization objectives. When the weights of optimization objectives are not equal, for example, if carbon reduction is the objective, indicator weight factors need to be determined or fuzzy evaluation methods need to be used to assist decision-making. Lastly, the non-dominated sorting algorithm is a “reverse algorithm”, which filters “non-optimal” scenarios layer by layer. Multi-objective optimization of key parameters for each process can be done to obtain a global optimal solution in the future.

4 4 Conclusions

This paper aims to provide a multi-objective process comparison method for the selection and design of future WWTPs under the context of carbon neutrality, and, for the first time, comprehensively evaluate the performance of traditional and emerging wastewater treatment processes in terms of water quality, cost and carbon emission benefits. Biofilm processes outperform activated sludge and membrane processes in terms of effluent quality, while Anammox process is more economical. The direct emissions of CH₄ from wastewater biological treatment were the main source of GHGs, followed by indirect emissions from chemicals and electricity consumption. Furthermore, the optimal processes under different scenarios were recommended through the non-dominated sorting approach. The key to achieving carbon neutrality for WWTPs lies in focusing on reducing energy and material consumption, thereby lowering carbon emissions. Currently, the built-in models in GPS-X software are mainly used, and some of them differ from the processes in full-scale WWTPs. In the future, more emerging treatment processes can be considered, further operation parameters of actual WWTPs and emission factors in specific regions can be integrated to optimize the process comparison for practical applications.

References

[1] Gunaseelan V N. Anaerobic digestion of biomass for methane production: a review. Biomass and Bioenergy, 1997, 13(1): 83–114

[2] Gou M, Zeng J, Wang H Z, Tang Y Q, Shigematsu T, Morimura S, Kida K. Microbial community structure and dynamics of starch-fed and glucose-fed chemostats during two years of continuous operation. Frontiers of Environmental Science & Engineering, 2016, 10(2): 368–380

[3] Nielsen H B, Uellendahl H, Ahring B K. Regulation and optimization of the biogas process: propionate as a key parameter. Biomass and Bioenergy, 2007, 31(11): 820–830

[4] Bhunia P, Ghangrekar M M. Statistical modeling and optimization of biomass granulation and COD removal in UASB reactors treating low strength wastewaters. Bioresource Technology, 2008, 99(10): 4229–4238

[5] Feng J, Wang Y L, Ji X Y, Yuan D Q, Li H. Performance and bioparticle growth of anaerobic baffled reactor (ABR) fed with low-strength domestic sewage. Frontiers of Environmental Science & Engineering, 2015, 9(2): 352–364

[6] Öztürk M. Conversion of acetate, propionate and butyrate to methane under thermophilic conditions in batch reactors. Water Research, 1991, 25(12): 1509–1513

[7] Lange M, Ahring B K. A comprehensive study into the molecular methodology and molecular biology of methanogenic Archaea. FEMS Microbiology Reviews, 2001, 25(5): 553–571

[8] Rajhi H, Puyol D, Martínez M C, Díaz E E, Sanz J L. Vacuum promotes metabolic shifts and increases biogenic hydrogen production in dark fermentation systems. Frontiers of Environmental Science & Engineering, 2016, 10(3): 513–521

[9] Stams A J, Plugge C M, Mirna M C. Electron transfer in syntrophic communities of anaerobic bacteria and archaea. Nature Reviews. Microbiology, 2009, 7(8): 568–577

[10] Worm P, Stams A J M, Cheng X, Plugge C M. Growth- and substrate-dependent transcription of formate dehydrogenase and hydrogenase coding genes in Syntrophobacter fumaroxidans and Methanospirillum hungatei. Microbiology, 2011, 157(1): 280–289

[11] Zheng G, Li J, Zhao F, Zhang L, Wei L, Ban Q, Zhao Y. Effect of illumination on the hydrogen-production capability of anaerobic activated sludge. Frontiers of Environmental Science & Engineering, 2012, 6(1): 125–130

[12] Daniel S L, Keith E S, Yang H, Lin Y S, Drake H L. Utilization of methoxylated aromatic compounds by the acetogen Clostridium thermoaceticum: expression and specificity of the co-dependent O-demethylating activity. Biochemical and Biophysical Research Communications, 1991, 180(1): 416–422

[13] Wang L, Zhou Q, Li F T. Avoiding propionic acid accumulation in the anaerobic process for biohydrogen production. Biomass and Bioenergy, 2006, 30(2): 177–182

[14] Gallert C, Winter J. Propionic acid accumulation and degradation during restart of a full-scale anaerobic biowaste digester. Bioresource Technology, 2008, 99(1): 170–178

[15] Mohan S V, Rao N C, Prasad K K, Sarma P N. Bioaugmentation of an anaerobic sequencing batch biofilm reactor (AnSBBR) with immobilized sulphate reducing bacteria (SRB) for the treatment of sulphate bearing chemical wastewater. Process Biochemistry, 2005, 40(8): 2849–2857

[16] Marone A, Massini G, Patriarca C, Signorini A, Varrone C, Izzo G. Hydrogen production from vegetable waste by bioaugmentation of indigenous fermentative communities. International Journal of Hydrogen Energy, 2012, 37(7): 5612–5622

[17] McInerney M J, Bryant M P. Anaerobic degradation of lactate by syntrophic associations of Methanosarcina barkeri and Desulfovibrio species and effect of H₂ on acetate degradation. Applied and Environmental Microbiology, 1981, 41(2): 346–354

[18] De Bok F A M, Plugge C M, Stams A J M. Interspecies electron transfer in methanogenic propionate degrading consortia. Water Research, 2004, 38(6): 1368–1375

[19] Friedrich M, Springer N, Ludwig W, Schink B. Phylogenetic positions of Desulfofustis glycolicus gen. nov., sp. nov., and Syntrophobotulus glycolicus gen. nov., sp. nov., two new strict anaerobes growing with glycolic acid. International Journal of Systematic Bacteriology, 1996, 46(4): 1065–1069

[20] Sekiguchi Y, Kamagata Y, Nakamura K, Ohashi A, Harada H. Syntrophothermus lipocalidus gen. nov., sp. nov., a novel thermophilic, syntrophic, fatty-acid-oxidizing anaerobe which utilizes isobutyrate. International Journal of Systematic and Evolutionary Microbiology, 2000, 50(Pt 2): 771–779

[21] Bruns A, Cypionka H, Overmann J. Cyclic AMP and acyl homoserine lactones increase the cultivation efficiency of heterotrophic bacteria from the central Baltic Sea. Applied and Environmental Microbiology, 2002, 68(8): 3978–3987

[22] Schoenborn L, Yates P S, Grinton B E, Hugenholtz P, Janssen P H. Liquid serial dilution is inferior to solid media for isolation of cultures representative of the phylum-level diversity of soil bacteria. Applied and Environmental Microbiology, 2004, 70(7): 4363–4366

[23] Martins M, Faleiro M L, Barros R J, Veríssimo A R, Barreiros M A, Costa M C. Characterization and activity studies of highly heavy metal resistant sulphate-reducing bacteria to be used in acid mine drainage decontamination. Journal of Hazardous Materials, 2009, 166(2–3): 706–713

[24] Marchaim U, Krause C. Propionic to Acetic-acid ratios in overloaded anaerobic-digestion. Bioresource Technology, 1993, 43(3): 195–203

[25] Ahring B K, Sandberg M, Angelidaki I. Volatile fatty acids as indicators of process imbalance in anaerobic digestors. Applied Microbiology and Biotechnology, 1995, 43(3): 559–565

[26] Van Lier J B, Martin J L S, Lettinga G. Effect of temperature on the anaerobic thermophilic conversion of volatile fatty acids by dispersed and granular sludge. Water Research, 1996, 30(1): 199–207

[27] Liu R R, Tian Q, Yang B, Chen J H. Hybrid anaerobic baffled reactor for treatment of desizing wastewater. International Journal of Environmental Science and Technology, 2010, 7(1): 111–118

[28] Zhu G F, Li J Z, Wu P, Jin H Z, Wang Z. The performance and phase separated characteristics of an anaerobic baffled reactor treating soybean protein processing wastewater. Bioresource Technology, 2008, 99(17): 8027–8033

[29] Altaf M, Naveena B, Venkateshwar M, Kumar E V, Reddy G. Single step fermentation of starch to L(+) lactic acid by Lactobacillus amylophilus GV6 in SSF using inexpensive nitrogen sources to replace peptone and yeast extract–optimization by RSM. Process Biochemistry, 2006, 41(2): 465–472

[30] Turki S, Kraeim I B, Weeckers F, Thonart P, Kallel H. Isolation of bioactive peptides from tryptone that modulate lipase production in Yarrowia lipolytica. Process Biochemistry, 2006, 41(2): 465–472

[31] Federation W E. American Public Health Association. Standard methods for the examination of water and wastewater. American Public Health Association (APHA): Washington, D C, USA, 2005

[32] Dubois M, Gilles K A, Hamilton J K, Rebers P, Smith F. Colorimetric method for determination of sugars and related substances. Analytical Chemistry, 1956, 28(3): 350–356

[33] Li J, Zheng G, He J, Chang S, Qin Z. Hydrogen-producing capability of anaerobic activated sludge in three types of fermentations in a continuous stirred-tank reactor. Biotechnology Advances, 2009, 27(5): 573–577

[34] Kalia A, Rattan A, Chopra P. A method for extraction of high-quality and high-quantity genomic DNA generally applicable to pathogenic bacteria. Analytical Biochemistry, 1999, 275(1): 1–5

[35] Angelidaki I, Alves M, Bolzonella D, Borzacconi L, Campos J L, Guwy A J, Kalyuzhnyi S, Jenicek P, van Lier J B. Defining the biomethane potential (BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays. Water Science and Technology, 2009, 59(5): 927–934

[36] Pullammanappallil P C, Chynoweth D P, Lyberatos G, Svoronos S A. Stable performance of anaerobic digestion in the presence of a high concentration of propionic acid. Bioresource Technology, 2001, 78(2): 165–169

[37] Liu Y, Whitman W B. Metabolic, phylogenetic, and ecological diversity of the methanogenic archaea. Annals of the New York Academy of Sciences, 2008, 1125(1): 171–189

[38] Hill D T, Cobb S A, Bolte J P. Using volatile fatty-acid relationships to predict anaerobic digester failure. Transactions of the ASAE (United States), 1987, 30(2): 496–501

[39] Hill D T, Holmberg R D. Long chain volatile fatty acid relationships in anaerobic digestion of swine waste. Biological Wastes, 1988, 23(3): 195–214

Acknowledgements

This work was supported financially by the National Natural Science Foundation of China (Grant No. 5148141), and the State Key Laboratory of Urban Water Resource and Environment (Harbin Institute of Technology) (No. 2016DX06).

Funding

RIGHTS & PERMISSIONS

2016 Higher Education Press and Springer–Verlag Berlin Heidelberg