Wastewater System Integration: A Biogenic Waste Biorefinery Eco-Industrial Park

Abstract

In recent years, great interest has been shown in the utilization of biogenic wastes in biorefineries as part of the concept of a circular bioeconomy. However, various challenges arise including the availability, cost and characteristics of the biogenic wastes in ensuring consistent biorefinery production processes. This work presents an optimization-based approach of a biogenic waste biorefinery eco-industrial park (BWB EIP). An indirect integration scheme is presented with a wastewater treatment plant (WWTP) acting as a centralized utility hub to treat the biogenic wastes generated from the participating plants and to supply volatile fatty acid (VFA) demanded by the participating plants through the WWTP interceptors. The objective of minimizing the VFA demanded by the participating plants from external sources is achieved. To further assess the influence of a future increase in VFA demand for one of the participating plants which is the polyhydroxyalkanoate (PHA) plant on the integration network, a sensitivity analysis is conducted. The results indicated that two WWTP interceptors are required with a 32.8% and 27.4% reduction in fresh VFA from external sources achieved through the integration network before and after sensitivity analysis. This work provides an insight into developing the framework for other BWB processes.

1. Introduction

The circular bioeconomy has gained increased attention as the key criterion in sustainable manufacturing of high-value bio-based products by utilizing biomass such as wastes and by-products effectively [1]. The advantages of a circular bioeconomy include valorisation of side and waste resources, mitigation of the impacts of climate change as well as decreased dependency on fossil fuels. A biogenic waste biorefinery (BWB) is one of the facilitating concepts in addressing the circular bioeconomy where biogenic wastes are utilized as feedstock in producing bio-based products including biopolymers, biochemicals and biofuels [1,2]. For instance, one of the main classes of bio-based plastics in current development is polyhydroxyalkanoate (PHA), a class of biodegradable and bio-based polyester that serves as an alternative to conventional fossil-based plastics [3]. PHA can be produced from biogenic wastes such as olive oil mill wastewater and cheese whey [4,5]. The bio-based plastics manufactured from PHA decompose naturally in the environment, avoiding the improper management of plastic wastes which leads to adverse environmental effects. Additionally, biofuel production such as bioethanol from wheat straw [6] as a substitute for fossil fuel can significantly reduce carbon emissions [7]. This also addresses the problems associated with fossil fuel including oil reserve depletion and oil price fluctuation [8].

Biogenic wastes can be classified into municipal waste and industrial wastes which include agro-industry waste, food-processing waste, oily waste, etc. [9]. Biogenic wastes are rich in organic and nutrient content which promote microbial growth and, thus, the mismanagement of these wastes will lead to environmental contamination. On the other hand, the organic and nutrient-rich content such as volatile fatty acid (VFA) is the main factor for the popularity of biogenic wastes in biorefineries [10,11]. VFA is the key precursor in PHA and bioethanol production as well as the production of chemicals such as vinyl acetate monomer (VAM) and terephthalic acid [12].

One of the concerns of utilising biogenic wastes is the logistic factors including harvesting, collection, pre-treatment, transportation and storage of biogenic wastes as feedstock in the BWB production process. This in turn affects the availability and cost of raw materials [8]. Additionally, considering the intrinsic factors, the constant availability, composition, dynamic behaviour and characteristics of biogenic wastes will affect the BWB production process [13]. Thus, the valorisation of multiple biogenic wastes is a feasible approach to address the aforementioned concerns, ensuring the ongoing supply and addressing the dynamic behaviour of feedstock and thus allowing large-scale and consistent BWB production [8].

Recent research has focused on a more sustainable and economical strategy towards the goal of a circular bioeconomy which is the integration of biorefineries in a management facility such as a wastewater treatment plant (WWTP) [1]. In this scenario, while treating the wastewater, the existing anaerobic digestor in the WWTP can be utilised to produce VFA, a building block in various biorefinery applications [14]. In Malaysia, there are currently more than 7000 sewage treatment plants; the effort of water reclamation from sewage sources is first implemented in an urban area, with plans for the development of reclaimed water use in other urban areas. This effort allows the bio-effluent treated at the sewage treatment plant to be further treated at a centralised water reclamation plant to produce non-potable treated water which is distributed for industrial processes and non-food-crop agricultural applications [15]. Thus, the WWTP can be further integrated into a BWB eco-industrial park (EIP) which acts as a centralised utility hub to manage a large number of resources and supply the common utilities required by the participating plants in the BWB EIP which is known as the indirect integration scheme [16]. An EIP is an industrial park which promotes cross-industry and community collaboration for economic, environmental and social benefits [17]. In recent years, research projects have focused on inter-plant water integration among several industrial plants [18,19]. The effective reuse or recycling of wastewater allows for the conservation of fresh resources such as water [19]. However, to date, there are limited studies on the VFA integration among industrial plants in an EIP.

In this paper, a superstructure of the BWB EIP is developed to optimize the performance of the BWB EIP as shown in Figure 1. The biogenic wastes from participating plants are integrated indirectly through a shared WWTP [20]. In this case, the biogenic wastes discharged from the participating plants are sent to the WWTP which later undergo anaerobic fermentation to produce a VFA-rich stream. The VFA-rich stream is utilized as the feedstock in the production of the biorefinery plants in the BWB EIP. With that, WWTP plays a vital role in the BWB EIP based on the concept of a circular bioeconomy where resources, materials and products are circulated within the economy for as long as possible, with minimization of waste discharged to the environment [21]. The process integration in the BWB EIP also enhances economic and environmental sustainability [10]. Meanwhile, the participating plants utilize multiple biogenic waste sources generated within the BWB EIP, tackling the issue of feedstock availability, transportation and storage. The potential participating plants include industries that produce biogenic wastes such as the agricultural industry, food-processing plants, dairy industry, etc. Apart from that, the potential participating plants also include those that require VFA as the main feedstock in their biorefinery production [20].

The performance of the BWB EIP can be evaluated through the benchmarking process using overall mass targeting [22]. Benchmarking enables the comparative evaluation of process performance by establishing a standard of excellence in designing a process. Overall mass targeting is a benchmarking approach to evaluate the benchmarks for the chemical species and stream flows in part of the process or the entire process. Key performance indicators (KPIs) are used which include minimum raw materials and material utility usage, maximum desired products or by-product yield, and minimum pollutants and waste generation [22].

This paper aims to develop an optimization-based approach to the design and integration of biorefinery production plants in the BWB EIP to provide an insight into developing the framework for other BWB production. Moreover, a sensitivity analysis with respect to the effect of varying VFA requirements for one of the biorefinery production plants, the PHA plant, on the integrated network is carried out.

2. Methodology

2.1. Problem Representation

The indirect process integration within the BWB EIP through a WWTP which acts as a central utility hub is represented in Figure 2. The design of the BWB EIP is intended to treat various waste sources from a set of processes of the participating plants. The following method used in constructing the design of the BWB EIP is based on the methodology suggested by Lovelady and El-Halwagi [23]. The processes are referred to as PROCESSES = {p|p = 1, 2, …, N_Process}. For each process, a set of effluent wastewater streams from the participating plants is known as the sources: SOURCES = {i|i = 1, 2, …, N_sources,p}. Additionally, a set of process sinks is referred to as SINKS = {j|j = 1, 2, …, N_sinks,p} where a certain flowrate is required by each sink for the production processes.

During the direct process integration where the waste streams are recycled or reused, the excessive levels of pollutants can be accumulated to unacceptable levels and it can bring detrimental effects to the process performance. Thus, a more feasible approach is taken where interceptors are used to selectively remove the contaminants from the waste streams discharged by the participating plants. A set of interceptors is referred to INTERCEPTORS = {k|k = 1, 2, …, N_interceptors}. The streams exiting the interceptors will be allocated to different process sinks. Fresh sources are available to supplement the sources required by the production processes of each participating plant.

2.2. Optimization Formulation

A structural representation of the source–interception–sink is shown in Figure 3. The following optimisation formulation used is based on the study by El-Halwagi [24] and Lovelady and El-Halwagi [23]. Global software of LINGO 19.0 will be used to solve the problem and the optimisation formulation is shown in Supplementary Material S2: LINGO 19.0 Optimisation Formulation. In this paper, the objective is to minimise the fresh VFA flowrate from external sources required by the participating plants in the BWB EIP and it is given by:

$$Minimise{\displaystyle \sum }_{r=1}^{N_{Fresh}}{F}_r$$

Each source, i, is split into interception devices with one split being distributed to a waste sink. The splitting constraint is shown as:

$$F_i={\displaystyle \sum }_{k=1}^{N_{int}}{w}_{i,k}+w_{i,waste}i=1,2,\dots ,N_{sources}$$

These splits are mixed before entering k^th interceptor which is shown as:

$$W_k\times Y_k{}^{in}={\displaystyle \sum }_{i=1}^{N_{sources}}{w}_{i,k}\times y_i{}^{in}k=1,2,\dots ,N_{int}$$

The intercepted streams are distributed to process sinks and can be written as:

$$W_k={\displaystyle \sum }_{j=1}^{N_{sink}}{g}_{k,j}k=1,2,\dots ,N_{int}$$

The distributed streams are mixed before entering j^th sink which is shown as:

$$G_j={\displaystyle \sum }_{r=1}^{N_{fresh}}{F}_{r,j}+{\displaystyle \sum }_{k=1}^{N_{sink}}{g}_{k,j}j=1,2,\dots ,N_{sinks}$$

where F_r,j is the f^th fresh water flowrate and g_k,j is the k^th interceptor stream fed to the j^th sink, respectively.

The constraints for the fresh streams and pollutant load and composition from the interceptors:

$$G_j\times z_j{}^{in}={\displaystyle \sum }_{r=1}^{N_{fresh}}{F}_{r,j}\times y_r{}^{fresh}+{\displaystyle \sum }_{k=1}^{N_{sink}}{g}_{k,j}\times y_k{}^{out}j=1,2,\dots ,N_{sinks}$$

The flowrate of the total f^th fresh stream is shown as:

$$F_r={\displaystyle \sum }_{r=1}^{N_{sinks}}{F}_{r,j}r=1,2,\dots ,N_{fresh}$$

The discharge of unused material from the sources is shown as:

$$Waste={\displaystyle \sum }_{i=1}^{N_{sources}}{w}_{i,waste}$$

The following constraints are added for non-negativity flow rates:

$$F_{r,j}\ge 0wherer=1,2,\dots ,N_{fresh}andj=1,2,\dots N_{sinks}$$

$$Waste\ge 0$$

3. Results and Discussion

3.1. Case Study

A BWB EIP is set up by considering participating plants located in proximity [20]. The participating plants involved utilize VFA in their production processes and produce biogenic wastes that contain VFA sources or unreacted VFA from their production processes. A WWTP treats the biogenic wastes and supplies VFA through anaerobic digestion to sustain the demand of each participating plant. The WWTP consists of interceptors with different allowable waste inlet impurities which prevent the overload of an interceptor with excessive levels of impurities. A fixed removal ratio is given for each interceptor. This results in different VFA outlet purities produced from the interceptors which can satisfy the demand of the participating plants.

The selected participating plants include palm oil mill, PHA plant, bioethanol plant, terephthalic acid plant and vinyl acetate monomer (VAM) plant. Note that the participating plants are not limited to the abovementioned plants, whereby the VFA sources or unreacted VFA generated in the biogenic wastes from these plants can be recycled, benefiting the plants in the BWB EIP.

3.2. Assumptions of Case Study

Several assumptions made for the design and integration of the BWB EIP were determined as follows:

• The anaerobic digestion carried out in the WWTP produces only acetic acid. Thus, acetic acid is assumed to be the VFA-enriched stream [20];
• The biogenic wastes produced and VFA requirements of the participating plants are based on the production capacity of the respective plants. The production capacity of each plant and the stoichiometric information are summarized in

  Table 1. Production capacity and stoichiometric information of potential participating plants in BWB EIP.

  Plant	Production Capacity	Reference	Stoichiometric Information	Reference
  Palm oil mill (palm oil mill effluent produced)	27.08 t/h	[25]	VFA concentration in 100L POME is 1900 ppm.	[26]
  PHA	50,000 t/y	[27]	The PHA yield from VFA is 48%.	[28]
  Bioethanol	30,000 t/y	[29]	The VFA conversion to bioethanol is 81%.	[20]
  Terephthalic acid	610,000 t/y	[30]	The VFA conversion to terephthalic acid is 80%.	[31]
  VAM	330,000 t/y	[32]	The selectivity of VAM based on VFA is 99%.	[33]

  ;

3.

The wastewater source and sink data for the BWB EIP are shown in

Table 2. Wastewater source and sink data for the BWB EIP.

Plant	Source (ton/h)	VFA Impurity (%)
Palm oil mill	15.91	58.74
PHA	0.51	42.15
Bioethanol	1.77	57.30
Terephthalic acid	10.62	42.00
VAM	7.40	16.50
Plant	Sink (ton/h)	VFA Impurity (%)
PHA	1.16	28.10
Bioethanol	9.30	38.2
Terephthalic acid	53.09	38.2
VAM	46.69	15

with the calculation shown in Supplementary Material S1: Calculation for Source and Sink Data. Wastewater source refers to wastewater produced by the participating plants including the unreacted VFA while sink refers to the amount of VFA required by the participating plants. The VFA impurity refers to other acids, alcohols and water in the respective streams;

4.

Interception data for the BWB EIP are shown in

Table 3. Interception data for the BWB EIP with 70% VFA impurity removal ratio.

Interception	Maximum Inlet of VFA Impurity (%)	Maximum Outlet of VFA Impurity (%)
1	45	13.5
2	55	16.5
3	60	18

. A fixed removal ratio of 70% is applied for each interceptor. The maximum inlet of VFA impurity refers to the biogenic waste streams generated from the participating plants while the maximum outlet of VFA impurity refers to the VFA-enriched streams supplied to the participating plants;

5.

The fresh VFA is supplied at 20% impurity from external sources;

6.

In anaerobic digestion, a mix of wastewater will not cause any inhibition of reaction;

7.

The products produced by the participating plants are not included in the study of this paper. The treated wastewater and secondary sludge produced by the centralized WWTP are sent for further treatment which is also not within the scope of this study [20];

8.

Sensitivity analysis is conducted to assess the influence of a future increase in VFA demand for the PHA plant on the integration network. Due to the current low production capacity of the PHA plant, an assumption of a 20-times increase in VFA demand for the PHA plant is made by maintaining the VFA impurity. The new sink data of the PHA plant are shown in

Table 4. New sink data of PHA plant.

Plant	Sink (ton/h)	VFA Impurity (%)
PHA	23.20	28.10

.

3.3. Results and Discussion

An initial case study is shown in Figure 4. Biogenic wastes from the participating plants are sent to the WWTP which produces VFA-rich streams. The insufficient VFA that is required by the participating plants is supplied through fresh VFA from external sources. The result shows that the fresh VFA demanded is 74.03 ton/h with two interceptors selected in the WWTP.

As the PHA production is projected to increase in future years [34], a sensitivity analysis was conducted to assess the influence of a future increase in VFA demand for the PHA plant on the integration network. A solution to the case study is shown in Figure 5. In this solution, the minimum fresh VFA from external sources required by the BWB EIP was 96.07 t/h. Similarly, two interceptors were present, namely, interceptor 1 and interceptor 2 as shown in Table 3. This indicates that the two interceptors were sufficient to selectively remove the impurities present in the biogenic wastes in order to supply the VFA demanded by the participating plants. Thus, all biogenic wastes were utilized and no waste was generated in the BWB EIP.

Table 5. Summary of fresh VFA required in the BWB EIP.

Case Study	Total VFA Required (ton/h)	Fresh VFA Required from External Sources (ton/h)	Percentage Difference (%)
Before sensitivity analysis	110.24	74.03	32.8
After sensitivity analysis	132.28	96.07	27.4

illustrates the percentage difference between the total VFA sinks required by the participating plants and the fresh VFA from external sources. By assuming the total VFA sinks required by the participating plants are originally supplied by external sources, it was compared to the fresh VFA as obtained in the integration network. In both solutions before and after sensitivity analysis, a reduction of 32.8% and 27.4% of fresh VFA required from external sources was achieved. Thus, it was shown that the integration of BWB processes in a BWB EIP which promotes material exchange could potentially reduce the fresh resources demanded by the participating plants. The solution of the case study after conducting the sensitivity analysis is considered the final outcome. This is due to its more robust network design as compared to the initial solution. Moreover, considering the increase in future demand for the PHA plant, future expansion costs as well as future piping reallocation can be avoided. The sensitivity analysis can be further carried out for all the participating plants within the BWB EIP, preventing the redesigning of the integration network due to an increase in future demand.

The integration of the biorefinery plants in a BWB EIP allows the utilization of biogenic wastes produced within the BWB EIP where the logistic issues of feedstock are addressed. Considering plants located in proximity, it allows the feasibility of material exchange through pipelines [20]. This allows the ease of feedstock handling and transportation, further ensuring the cost-effective and consistent supply of feedstock for the BWB production. Moreover, the integration network with a centralized WWTP allows the recycling of biogenic wastes within the BWB EIP. This reduces the fresh VFA from external sources in the BWB production and also potentially reduces the environmental impact of disposing the nutrient-rich biogenic wastes to the environment such as eutrophication. Additionally, the integration network is believed to save costs. According to Goh et al., with a business correlation between the centralized WWTP and the participating plants, wastewater is sent to the WWTP without any charges as the VFA produced will later be sold to the participating plants [20]. This is followed by a reduction in wastewater treatment operating costs.

The proposed BWB EIP focuses mainly on VFA integration among the participating plants. Additionally, to date, there are limited studies on non-water exchange streams in an EIP. Based on the world’s leading industrial symbiosis at the Kalundborg Eco-industrial Park, the collaboration of thirteen public and private companies allows a robust exchange network to be constructed including water, energy and material exchange [35]. Thus, this work provides an insight into other material exchange, as well as water in the BWB EIP.

3.4. Project Findings

The major contribution of this paper is the utilization of multiple biogenic wastes in the BWB processes via the circular bioeconomy approach. The integration of several plants with a centralized WWTP further provides insight into material recycling within the BWB EIP. The integration network could reduce the fresh resources required within the BWB EIP. Additionally, the integration network could potentially solve the issues of feedstock availability, cost, transportation, etc., in the BWB processes. Additionally, by considering future aspects when designing the network such as a future increase in feedstock due to product demand, an efficient integration network can be constructed. Thus, the redesigning of the integration network or future expansion costs can be avoided.

This approach could bring a positive impact in developing the framework for other BWB processes including the production of organic acids and chemicals (ethanol, methanol, succinic acid, etc.) as well as biofuel (biohydrogen and bioethanol) [36]. These production plants can be integrated into a BWB EIP to promote waste valorisation, a more environmental sustainability process, and an increase in resource efficiency.

3.5. Limitations and Future Work

One of the limitations of this work includes the inconsistency of biogenic waste characteristics and availability in the integrated BWB process. This might be due to the decisions in the industries which alter the manufacturing processes, further influencing the availability and properties of the biogenic wastes [37]. This will further affect the quality of VFA produced during the anaerobic digestion process. Thus, it is necessary to ensure the continuous supply of biogenic waste sources within the BWB EIP for the production of VFA which is the main feedstock being used in the production processes of the participating plants in the BWB EIP. One way to ensure the ongoing supply of biogenic wastes is through a storage facility where the biogenic wastes can be stored in large storage areas during periods of high waste discharge.

Additionally, as the biogenic waste sources are process- and location-specific, several factors such as the waste sources’ availability, sourcing distances and demand should be taken into consideration [37]. This is agreed with by Adams [38], where a more sophisticated technology is required as a result of the wide versatility of waste sources in the BWB process. The technology should also be technically and economically feasible in addressing the aforementioned issues. During the implementation and development of the BWB EIP, efficient planning and evaluation of the participating plants, characteristics of the biogenic wastes, interception data of the WWTP interceptors, etc., are required. This is to ensure the effective valorisation of biogenic wastes to fulfil the demand of the VFA feedstock required by each participating plant.

Moreover, in a BWB EIP, the wastes generated from one industry can be utilized by other industries. While an EIP emphasizes material exchange, it also focuses on energy exchange within the participating plants, resulting in economic benefits for all involved. For instance, in the Kalundborg Eco-industrial Park, the residual heat produced in the manufacturing process is used to distribute heat to businesses and households in the city of Kalundborg [35]. However, this paper presents the material exchange, mainly VFA. Thus, it is recommended to analyse the BWB EIP with a more comprehensive approach including water integration specifically to recover the useful nutrients in wastewater. In addition, the economic and environmental feasibilities of the BWB EIP are necessary to be carried out.

4. Conclusions

An optimization study for the integration of participating plants in a BWB EIP using a source–interception–sink approach is presented. This is illustrated using a five-plant BWB EIP case study with a WWTP consisting of interceptors to selectively remove contaminants from the biogenic wastes produced by the participating plants. The study focuses on material integration, intending to minimize the fresh VFA required in the production processes of the participating plants within the BWB EIP. A sensitivity analysis on the effect of a future increase in VFA demand for the PHA plant on the integration network is conducted. The key takeaways of this work are as follows:

• The integration network is able to reduce up to 27.4% of fresh VFA from external sources with two WWTP interceptors required after conducting a sensitivity analysis.
• The integration network could potentially address the challenges associated with the logistic issues of feedstock, availability and cost of feedstock in the BWB production processes.
• Future work is recommended including consideration of the availability and dynamic behaviour of biogenic wastes, storage facilities, efficient planning and evaluation of biogenic wastes, as well as economic and environmental analysis.
• This work can be further developed and applied to energy exchange and the exchange of other non-water streams within a BWB EIP.