TASKS AND CONTENTS:- To formulate a mathematical model for location optimization of biomass power plants, considering at least two objectives and the flow of the biomass supply chain to
INTRODUCTION
Rationale
Nowadays, fossil fuels are being promoted to be replaced by cleaner and sustainable energy sources (e.g., wind power, solar power, bioenergy, etc.) in order to address sustainable economic growth, environmental problems, and energy security in many countries [1-10] Renewable energy development is crucial for sustainable economic growth in almost all nations due to its potential to maximize natural resources and reduce greenhouse gas emissions On the other hand, the energy crisis created an unexpected twist in history, sped up clean energy, and shaped a sustainable energy network The USA enforces the IRA, the most significant investment ever for energy security and clean energy development Europarl approved the REPower EU to decrease its dependence on fossil fuels in Russia and focus clearly on renewable energy as soon as possible China has also announced ambitious plans to meet the 2030 renewable energy objective five years earlier than expected Owing to prevailing environmental and strategic imperatives, biomass is presently surfacing as a prospective alternative for effecting global energy source diversification [2, 10] According to reports from [8], renewable energy plays essential roles and accounts for over one-fifth of the global generation It is expected to become the primary source of electricity production by 2025, accounting for over one-third of the total global generation [11] Therefore, biomass energy plays an important role and will likely play a vital role in meeting the world's energy needs in the future This is the reason behind the significant increase in new investment in renewable energy worldwide over the last two decades, from 2004 to 2022 The most dramatic growth recorded was approximately 495 billion US dollars in 2022
It increased 16 times compared to 2004 It demonstrated the world's concern for clean energy These days, many renewable energy sources are accessible, including geothermal, marine, biomass, and waste-to-energy Nonetheless, the most significant
2 investment is recorded in wind and solar power Since 2004, the investment worldwide in solar energy has increased from a little over $10 billion to over 140 billion in 2019
Fig 1 1Worldwide investment in sustainable energy 1
The concentration should be on biomass and waste-to-energy technology The investment in 2011 peaked at about $21 million and decreased after reaching the threshold Because of stricter regulations and scrubbing technologies, modern incinerators emit less pollution than older ones As a result, investments in biomass and waste energy technology have decreased They peaked in 2011 at USD 20 billion but have since fallen to 4.7 billion USD in 2017 Moreover, the decreasing investment by restructuring focuses on developing available wind and solar power sources Nevertheless, in the end, the percentage of biomass investment still accounts for their essential place
1 https://www.statista.com/statistics/186807/worldwide-investment-in-sustainable-energy-since-2004/
In v estm en t in b illi o n US d o llar s
Worldwide investment in sustainable energy
Fig 1 2 Investment in biomass and waste energy technologies worldwide from 2004 to
On a global scale, biomass is the fourth most significant energy source, accounting for about 14-15% of the world's total energy consumption In developing countries, biomass is often a major energy source, contributing to an average of about 35% of the total energy supply As shown in Fig 1.3, bioenergy accounts for 2%, a small proportion of renewable energy generation by source Thus, developing bioenergy is still a potential aspect around the world Among the many types of biomass, agricultural residual biomass stands out due to its pronounced cost-efficiency relative to other alternatives Therefore, it is an opportunity for many agricultural-producing countries to develop renewable energy from biomass [1]
2 https://www.statista.com/statistics/186827/global-investments-in-biomass-and-waste-since-2004
In v estm en t in b illi o n US d o llar s
Investment in biomass and waste energy technologies worldwide
Fig 1 3 Electricity generation by source in the world 2021 3
Viet Nam is considered as one of the most dynamic and stable countries in the Southeast Asia, both in terms of political system and economic development The country began economic reform in 1986 with an evolution toward a market economy, resulting in a steady increase in Gross Domestic Product (GDP) over the years In 2002, Vietnam's GDP was a mere 44.56 billion USD, ranking 60th globally and 6th in ASEAN However, as of 2022, its GDP has reached approximately 413.81 billion USD, elevating its rank to 37th globally, reflecting a nine-time increase and a 23-step jump from 2002 According to the International Monetary Fund (IMF), Vietnam's GDP is expected to increase to about 469.62 billion USD in 2023, putting it in 35th place worldwide Vietnam has maintained an impressive annual GDP growth rate of over 6% for the past two decades
Vietnam is one of the fastest-growing economies in Asia, so the electricity demand is increasing rapidly The electricity demand significantly increased to develop the economy and serve over 99 million inhabitants Viet Nam's electricity consumption has been growing quickly, which has fueled the nation's economic development However,
3 Our World in Data- https://ourworldindata.org/
5 the rising demand for electricity due to booming industries and a growing population will outpace the construction of new power plants in the future, potentially causing severe power shortages According to data from the Ministry of Industry and Trade, the growth rate of energy demand in Vietnam is currently double the GDP growth rate Meanwhile, in developed countries, this ratio is only approximately 1 Vietnam's electrical energy consumption increased by nearly 400% within ten years from 1998-
In the next five years, electricity demand will grow 8-10% per year from 2023 Furthermore, the required electricity capacity will double The solar power market will be essential in meeting the huge energy demand Thus, Vietnam is expected to dominate the renewable energy market in the region, attracting many international investors to build new renewable energy projects and carry out mergers and acquisitions related to these projects
Fig 1.4 points out that the electricity demand in Vietnam has continuously increased from 2000 to 2022
Fig 1 4 Viet Nam Electricity Demand, 2000 to 2022
Viet Nam Electricity Demand, 2000 to 2022
Vietnam possesses abundant biomass energy sources, including agricultural waste, urban wastewater, and hydropower, in addition to renewable energies like wind, solar, and conventional hydropower These resources are widely distributed and can be utilized through cogeneration technology for efficient heat and electricity production However, if left unaddressed, the substantial biomass volume poses a significant pollution threat to human health and ecosystems Deforestation and urbanization reduce ecosystems' carrying capacities, escalating environmental conflicts The Ministry of Industry and Trade's plan for biomass power in the Mekong Delta prioritizes resources derived from sugarcane bagasse and rice husks, aiming for an installed capacity of 834 MW by 2020.
214 MW, including 50 MW of sugarcane bagasse, rice husk power 140 MW, and Wood energy power 24 MW In the period from 2021 to 2030, the total installed capacity of biomass power is 304 MW, including 30 MW of sugarcane bagasse, rice husk electricity
150 MW, wood energy power 44 MW; 80 MW of straw power (Details of the development process are stated in the Decision No 08/2020/ QD-TTg on amending and supplementing Decision No 24/2014/QD-TTg of the Prime Minister on developing biomass power projects in Vietnam)
Vietnam biomass information map (Figure 1.5) shows the difference in theoretical potential (left) and technical potential (right) of crop residues after calculating some exploitation limits which obtaining farmers' approval before purchasing crop residues and only evaluating the number of products left after they have been used for available purposes
Vietnam's abundant agricultural and forest residuals present a significant opportunity for electricity generation, with an estimated potential of 7000 Mega Watt, according to the National Electricity Development Planning period 2021-2030, Vision 2050 (NPDP8) This plan highlights the importance of biomass utilization in Vietnam's energy sector, recognizing its role in achieving sustainable and reliable electricity generation.
7 goal of biomass energy rating in electric generation is 3% in 2020, 6.3% in 2030, and 8.1% in 2050; the heat ratio produced from biomass sources reached about 17% in 2020, 14% in 2030, và 12% in 2050 According to the calculation from the Department of Electricity and Renewable Energy, Ministry of Industry and Trade, till 2035, the potential of biomass-to-electricity development from rice husk about 370MW, forest residual 3.360 MW, bagasse 470 MW, straw 1.300 MW, biogas 1.370 MW The total potential of these types is more than 9,600 MW
Fig 1 5 Vietnam biomass information map
Fig 1 6Distribution of potential biomass power by sources in Viet Nam
The Mekong River Delta is the southern critical economic region which produces more than 50% of Vietnam's rice, making it Vietnam's largest rice production region; therefore, the agricultural residual is more considerable than in other regions The Mekong River Delta makes up 33.4% of the potential for developing the biomass energy sector and has the country's highest proportion
Research objective
Mathematical modeling is crucial for optimizing the location of biomass power plants, especially when considering multiple objectives The model should incorporate the biomass supply chain and ensure that supply meets demand By incorporating these factors, the model can effectively determine the optimal locations for biomass power plants, ensuring a sustainable and efficient biomass-based energy system.
- To propose an approach to handling the multiple-objective problem
- To validate the effectiveness of the proposed model, we study a realistic case with 13 data sets of the Mekong River Delta in Viet Nam The outcomes of the model could be used as a reference in selecting locations for biomass power plants.
Research scope
The study was conducted in 13 provinces and cities in the Mekong River Delta.
Research structure
The research is structured as follows: A literature review on supply chain management and studies on the location of biomass power plants is introduced in Chapter II Next,
11 Chapter III presents the research framework and mathematical model used in this study Chapter IV provides a case study in Vietnam Finally, the conclusion and future research are presented in Chapter V
LITERATURE REVIEW
Location selection for biomass supply chain and multi objectives
Location optimization is a primary problem in supply chain management and addresses essential aspects of designing renewable energy supply chain networks To identify the facility locations, a typical model, namely p-median [3], was utilized in many research, especially in defining renewable energy facilities Decision-makers deal with energy issues, especially in recent decades, have prioritized multiple aspects of sustainability (economic, environmental, and social) [4]
Cambero and Sowlati [5] developed a multi-objective mixed linear programming model for a forest-based bioenergy and biofuel supply chain, which is solved using a Pareto- generating method This model is formulated to maximize the social benefit, net present value, and reduction of greenhouse gas emissions Through a case study in the interior of British Columbia, Canada, the model's applicability is demonstrated, and the results
Research has established a positive correlation between social benefits and reduced greenhouse gas emissions To address this, Zhao & Li [6] developed a bi-objective optimization model considering logistics costs, social benefits, and environmental impacts in determining the optimal placement of biomass power plants and transportation networks Similarly, Karimi, Ekşioğlu, & Carbajales-Dale [7] formulated a stochastic optimization model to analyze economic and environmental factors in biopower supply chains, employing optimization algorithms to efficiently generate solutions.
The paper present by Coelho and Mateus [8] introduced a model based on the mix-integer linear programming for the capacitated plant location problem in reverse logistics Their study aims to minimize transportation costs and the fixed costs of installing and managing reprocessing facilities while identifying a suitable location for reprocessing facilities Martínez-Guido, Ríos-Badrán, Gutiérrez-Antonio, & Ponce-Ortega [9] proposed an approach based on mixed-integer linear programming for determining the location and design of the pelleting plants by considering the type of biomass and the most effective routes for transportation to minimize costs and environmental effects The results were obtained from the case study in Mexico, demonstrating the possibility of the model Yang, Chen, Chu, & Wang [10] found the optimal locations of plants and depots for opening by considering the objective of minimizing the transportation and total facility opening costs Disjunction and tree search, cutting plane method, and kernel search were developed to generate the solutions Sahoo, Mani, Das, & Bettinger [11] addressed optimal plant sites for biofuels and co-product production in the State of Ohio that aim to the opening and biomass delivered costs The research utilized sustainable crop residues as raw materials for biogas plants A novel methodology that integrated Artificial Neural Networks, GIS and mathematical models was developed in this work
In the study of Nayeri, Paydar, Asadi-Gangraj, & Emami [12] proposed a robust optimization model to build sustainable closed-loop supply chain by considering the uncertain parameters of transportation cost, carbon emission and demand The objective is to minimize total costs environmental impacts and maximize social impacts.The uncertainty type Epistemic is solved using fuzzy robust optimization in this study The outcome of the model can be used to decide some planning and strategic decisions such as the best supplier, transportation mode, facilities structure, and optimized flow of products inside the sustainable closed-loop supply chain Sarker, Wu, & Paudel [13] formulated a MIP model to optimize the hubs and bio-methane gas locations to minimize the overall supply chain total cost for renewable energy A Genetic Algorithm was devised to generate the solutions, and then the results were benchmarked to results from the existing solutions obtained by traditional Lingo and other comparable solvers Hariteja Nandimandalam, Amin Aghalari, Veera Gnaneswar Gude, & Mohammad Marufuzzaman [14] proposed a two-layer supply chain network model to determine the power plant location, allocation of suppliers, biomass harvesting, storage and transportation options of woods waste supply chain The objective of this model is to reduce overall costs and GHG emissions The case study results showed that this model can be utilized in many situations when it is necessary to determine the most suitable location for biomass plants
2.2 Uncertainty in biomass supply chain and proposed approach
The biomass supply chain has many unique features that make it a complex structure The seasonal availability of the biomass feedstock is one of the most vital characteristics of the biomass supply chain Uncertainty in the biomass supply might occur in supply chain flow bottlenecks and interruptions It is uncertain how much biomass will be delivered because it depends on the weather, the time of year the crop is harvested, and the availability of suitable land for replanting Other uncertainties frequently occur in the biomass supply chain include those related to biomass (conversion rates, efficiency of
15 production, fuel quality), final product demand (which can vary over time due to seasonal factors or changes in the market), cost/market uncertainties (variations in exchange rates, incentives from governments and policies), process uncertainties (disruptions or unforeseen circumstances), transportation and logistics uncertainties (price, ), and environmental effect uncertainties (production volume, transportation distance, or any other supply chain stages)
Kim, Realff, & Lee [15] proposed a two-stage mixed integer stochastic program to maximize the expected profit of different scenarios for the bio-fuel supply chain The Monte Carlo method is used to analyze the robustness and global sensitivity analysis of the nominal design Yeh, Whittaker, Realff, & Lee [16] proposed a two-stage multi- period bilevel model to maximize the profits objective of harvester and manufacturer of timberlands supply chain, respectively The first stage solved the problem of biorefinery investments, such as location and capacity and second stage solved the uncertainty parameters (biomass pricing, supplying and demanding biorefinery costs) Woo, Acuna, Moroni, Taskhiri, & Turner [17] developed an integrated approach with the combination of MCA and GIS to find the optimal biomass energy plants Forest biomass is chosen as the main biomass, and balanced economic, environmental, and social criteria are used as objectives The sensitivity is used for different moisture content levels to simulate this parameter's impact on biomass transportation costs Rabbani, Saravi, Farrokhi-Asl, Lim, & Tahaei [18] proposed a novel, multi-objective, and mixed integer linear programming model by incorporating conflicting economic, environmental, and social objectives The model was solved using two-stage algorithms consisting of augmented ε-constraint and TOPSIS This approach supports decision makers in determining suitable strategic and tactical decisions and managing bioenergy supply chain performance with appropriate trade-offs among the sustainability factors sensitivity
16 analysis is used to perform the economic, environmental and social performance as the primary, secondary and tertiary optimization objectives, respectively
Yılmaz Balaman et al [4] developed a novel optimization methodology for the sustainable design and planning of waste-to-bioenergy supply chains that involve a variety of product and feedstock types and technologies The study incorporated economic and environmental objectives into account utilizing a scenario-based fuzzy multi-objective modeling approach The objectives are to reduce total capital investment costs, maximize overall profit, and decrease greenhouse gas emissions from transportation and manufacturing Bojic, Martinov, Brcanov, Djatkov, & Georgijevic [19] introduced the p median mathematical model to determine optimal LCB plant locations Their study aims to minimize the total internal and external biomass transportation costs Gital Durmaz & Bilgen [20] developed a novel multi-stage solution methodology for biomass supply chain network The proposed multi-objective mixed integer linear programming model has two objective functions: maximization of the profit and minimization of total distance between poultry farms and biogas facilities The model can determine the optimal number, location, and size of the biogas facilities, network flow, and electricity generated Sensitivity analysis is implemented to consider the impact of biomass purchasing cost and maximum distance parameters on the decision
Saghaei, Ghaderi, & Soleimani [21] proposed a two-stage stochastic mixed integer non- linear programming model for minimizing the total cost of bioenergy supply chain while taking into account material quality uncertainty and unexpected weather interruption The results obtained from the model decide the location and size of power plants/storage and biomass flow Fattahi, Govindan, & Farhadkhani [22] developed a two-stage stochastic programming model to determine the location, type, size of technology, production level and biomass flow Computational results obtained from real-life case
17 studies in Iran show the possibility of model Ilbahar [23]formulated a fuzzy programming model for selecting the location for biomass power plants from municipal solid wastes The objective is to maximize the profit of biomass plants while deciding how many plants and how much capacity each plant needs to be installed Razm et al [24] developed a sustainable two-phase sequential approach for the bioenergy supply chain The objective is to minimize the total costs of the supply chain The uncertain parameters are solved by a robust model in the second phase In contrast, the first phase finds appropriate locations for the bio-refineries by decreasing the complexity in the computation of the problem The outcome of this model proved its effectiveness better than that of the traditional model Nandimandalam et al [14] proposed a two-layer supply chain network model to determine the power plant location, allocation of suppliers, biomass harvesting, storage and transportation options of woods waste supply chain The objective of model is to minimize the total cost and GHG emissions The case study is solved using the Lexicographic augmented ∊-constraint algorithm, which indicates that this model can be applied to many areas that need to determine the optimal location of biomass plants A sensitivity analysis was performed by implementing the impact of feedstock supply fluctuations, electricity demand, and biomass-to-electricity conversion rate of the power plant.
Research gap
Table 2 1 Review and analysis of the related literature
Biomass type Objective Main decision Uncertain parameters
Forest residues Maximize the expected profit over the different scenarios
• The size and location of the processing plants
• The biomass and product flows in each scenario
Biomass type Objective Main decision Uncertain parameters
Forest residues Maximize the profits objective of harvester and manufacturer
• Location and capacity of biorefinery
• Flows of material to the biorefineries and manufacturing facilities
• Biomass pricing, supplying and demanding
Minimize the total costs of the supply chain
• Locations for the bio- refineries
• Flow of biomass and biofuel
• Optimize the strategic design of the bioenergy supply chain
Woody biomass Minimize the total cost of producing electricity
• Location and capacity of infrastructure
• Decisions for sourcing the materials, inventory policy
Forest residues, agricultural residues, and livestock manures
Maximises the profit • The location, type, size of technology, production level and biomass flow
Balance economic, environmental, and social criteria
• The optimal biomass energy plants
Switchgrass Minimize environmental and economic performance Maximize social performance
• Suitable strategic and tactical decisions and managing bioenergy supply chain performance
Reduce total capital investment costs Maximize overall profit, Decrease
• Enhance the design and planning of multi waste biomass based supply chains
Biomass type Objective Main decision Uncertain parameters greenhouse gas emissions
Crop residues Minimize transportation cost
• Selection of an optimal lignocellulosic bioethanol plant location
Maximize the profit of biomass plants
• The location for biomass power plants
Forest residues Minimize the total cost and GHG emission
• Determine the power plant location
Minimize the total cost and CO 2 emission
• The optimal biomass power plant location and capacity
• The biomass supply chain network
This research formulates a mathematical model for multiple supply chain planning considering total cost and carbon emissions Then, the weighted max-min model is formulated to generate the solutions to handle the conflicts between objectives Although much previous research focuses on location optimization for renewable energy supply chains, there are remaining gaps that are addressed in this research:
As reviewed in Table 1, most of the existing papers on location selection biomass plants have studied biomass from forest residual, woody biomass, municipal solid wastes etc, while this research focused on agricultural residual (e.g rice husk) for power generation Most existing studies in the literature have only focused on total cost, and we could find fewer papers that quantify the supply chain's environmental impact based on its target
In this study, the environmental impact of the supply chain system regarding biomass
20 transportation is taken into account The diverse differences in residual necessitate significant analysis effort and extensive resources to model potential renewable biomass supply chain networks Therefore, this research contributes to renewable energy research by providing an aspect related to rice husk to energy that is potential research in many countries, especially in agricultural production countries
METHODOLOGY AND MATHEMATICAL MODELS
Methodology
Fig 3 1 Flowchart of the overall research methodology
Investigate the overall object of the study
Analyze the rationale and problem
Conclusion and future direction Formulate a mathematical model
Carbon emissions associated with transportation activities
Sensitive analysis Literature review and research gap Problem defination
Stage 1: Problem definition Initially, the authors investigate the overall object of the study, which includes the actuality of Vietnam and the world Finding a rationale for this research and emphasizing it is the research's primary purpose
Stage 2: Solution Method: After determining the problem and causes, the author collects the relevant documents and theoretical basis for the paper with the same problem to consider and propose the methodology and approach used to solve the problems From there, they formulate a mathematical model and a solution method The model is run and checked with a practical data set to verify its applicability
Stage 3: Conclusion and Future Direction: Analysis and conclusion of the result obtained from the model Evaluate the overall research process, analyze the advantages and disadvantages, and propose the future direction
3.1.2 Solution method for fuzzy multi-objective
The weighted max-min model for fuzzy multi-objective biomass supply chain planning optimization utilizes the weighted sum model for the fuzzy multi-objective mixed integer programming model This approach optimizes the supply chain by formulating a mathematical model that considers total costs, environmental considerations, and social benefits The model's framework involves eight steps, including defining objectives, constraints, and decision variables; fuzzifying the objectives and constraints; and solving the resulting model using the weighted max-min method This method provides a comprehensive approach to optimizing biomass supply chain planning in the presence of multiple, conflicting objectives.
Step 1 A supply chain planning problem is formulated as a multiple objective mixed integer programming according to practical constraints and decision-maker preferences
Step 2 When optimising each objective individually, the multi-objective problem is solved as a single objective problem
Step 3 At each derived solution, the respective values for each objective are determined from the results in Step 2
Step 4 From the results in Step 2 and Step 3 The value of the optimal and worst value of the negative objectives and positive objectives are calculated
Step 5 The membership function is then identified
Step 6 The weight of the objectives is identified
Step 7 Formulate the equivalent crisp model of the weighted max–min for the fuzzy multi-objective problem.
Typical problem
Vietnam is a tropical nation that provides suitable conditions for developing agriculture and forestry As a result, the country has abundant biomass resources, including agricultural and forestry wastes Biomass energy has a short circulatory cycle that the Organization of Environmental and Sustainable Development supports Utilization of this resource not only provides power for economic development and ensures environmental protection Biomass energy is a sustainable resource that is suitable for the goal of adapting to climate change, ensuring the demand for energy in the nation
A typical biomass supply chain is shown in Fig 3.4 Biomass energy can be produced in various ways, such as burning biomass for heat or electricity using steam turbines or using biomass for heat in thermal systems and generating electricity as well (this is known as “combined heat and power”), or converting feedstocks into liquid biofuels, and so on Biomass may come from different sources, such as forest residue, agricultural residue, animal waste, etc [26] The availability of biomass varies across different locations and periods, depending on factors such as climate, land use, and human activities Therefore, biomass power plants can utilize more than one type of biomass as material to operate This research focuses on using rice husk to burn biomass for heat generation to run the steam turbines
Fig 3 2 Types of materials can be converted to biomass energy (A Sugarcane bagasse
B Manures C Municipal solid waste D Rice straw E Rice husk F Wood) (Source: Internet)
For the exploitation of biofuels for biomass power development planning, many factors need to be considered: choosing the appropriate biofuel type for industrial-scale electricity production, biofuel price and plant location, which has advantages in the ability to transport and supply biomass in a stable and long-term manner at a reasonable price throughout the life of the project Favourable locations include an area with excellent health potential, near waterway and road transportation systems, capable of connecting to the electrical system and having an available water supply for the boiler
Inefficient biomass resources characterized by limited availability, dispersion, and uneconomical transportation distances from the power grid result in excessive levelized electricity production costs Consequently, these locations are excluded from consideration during analysis planning as they hinder the viability of renewable energy projects.
A biomass power plant can produce energy by heating sugar manufacturing wastes or biomass materials like bagasse, wood, rice husk, etc When utilising biomass resources, a power plant will use combustion systems to release stored biomass energy The biomass is burned to create high-pressure steam, which rotates a turbine to generate electricity Conveyors, vehicles, and storage systems are typical parts of a biomass power
25 plant that handle resources from biomass on-site Burning biomass in a boiler creates high-pressure steam that powers a generator and turbine to produce energy The boiler and steam turbine operation is the plant's heart, but other equipment used in the process includes condensers, cooling towers, exhaust gas treatment systems, pumps, fans, pipes, and more
Fig 3 3 Typical input-output of a biomass power plant [27]
In biomass supply chain design, biomass power plant locations are vital decisions for efficient design supply chain planning [11, 13, 28, 29] Besides, the decision-makers must consider other objectives, such as environmental and economic benefits Therefore, this research focuses on developing a mathematical model to define the biomass power plant locations to minimize total cost and carbon emissions associated with transportation activities mainly relying on trucks The total cost includes biomass power plant establishment, Operation and Maintenance (O&M), purchasing, and transportation costs A weighted max-min model for fuzzy multi-objectives is formulated to solve the multiple objectives problem
Fig 3 4 A typical biomass supply chain
Biomass sources come from industrial and handicraft processing facilities After being harvested, agricultural and forestry products like rice, sugar cane, and wood are transported to mills, sugar factories, and wood processing facilities Wasteful resources arise at these businesses following the processing steps For example, rice milling produces rice husks; sugar cane crushing produces bagasse; and wood processing produces sawdust, discarded shells, and wood ends Besides, there are many wood processing facilities in the country However, their capacity is small, mainly produced to serve local people
According to the investigation results, rice and rice husk activities, from production to processing, service, and the final consumer, are conducted all year round but are most numerous and focus mainly on the winter-spring crop These activities are all undertaken by private individuals who operate synchronously, from transporting rice to milling facilities and then transporting rice husks to consuming households
Demand Zones Biomass power plants
Biomass fuels, categorized as primary, can be utilized directly in combustion processes Alternatively, they can undergo chemical transformations (thermal or chemical) to produce secondary fuels, such as gas or packaged pellets These secondary fuels can then be further combusted in internal combustion engines, gas turbines, or boilers The combustion process in boilers generates steam, which powers steam turbines connected to generators for electricity production This versatile conversion pathway ensures efficient utilization of biomass resources for energy generation.
Assumption
This research examines the centralized utilization of rice husk as fuel for steam turbines and explores the potential capacity of biomass power plants within the framework of the development policy established by the Ministry of Industry and Trade of Vietnam The study considers both biomass suppliers and prospective biomass power plants, assuming that suppliers are strategically located in the center of each district to facilitate transportation to power plants, which are planned to be situated within Industrial Parks.
Primary biomass raw materials Burning Boiler/ rbine
Electrical or mechanical energy Converting
Fig 3 5 Block diagram converting biomass into electrical or thermal
28 For biomass power plants, biomass cogeneration (CO-GEN) is a practical and efficient way to generate electricity that allows them to achieve their emission reduction targets The electric power company typically has two objectives to consider when deciding where and the capacity of a biomass power plant to set up: minimizing the total cost and carbon emissions Because the expenses of biomass farming are spread out, transportation plays a significant role in the total cost By optimizing the use of energy sources by using fuels with low carbon, limiting the consumption of fossil fuels, but still providing steam with an extremely high temperature range, thereby using it efficiently and effectively to optimize energy and increase production efficiency Using these fuels helps reduce polluting emissions, negatively impact the environment, and improve air quality
To formulate the biomasss power plant location problem, we make the following assumptions:
• The decision-maker make decsion on the location and size of biomass power plants is risk-neutral
• The location of the biomass power plant is our main concern, so biomass harvesting and collection are not considered
• The amount of biomass lost in transportation from suppliers to biomass power plants is not considered
• The location of biomass power plants is based on the Industrial Park Therefore, only trucking is selected as the primary mode of transportation
• Biomass supplier is assumed to be located in the center of each district and potential biomass power plants will be installed in the Industrial Park
Mathematical model
The mathematical model for optimising the biomass power plant locations based on a classical p-median problem was introduced by Hakimi (1964) [3]
Table 3 1 Parameters and variables of the deterministic model
J Index set of potential sites for biomass power plant
K Index set of the power plant scale
Ck The plant capacity type k (k∈ K)
FP k Fixed cost for opening
OM k Operation and Maintenance (O&M) of a biomass power plant at site j (j∈J) with associated capacity type k (k∈K)
V The set of vehicle types
ES t The produced carbon emission c t Transportation unit cost by each type of vehicle (t∈V) d ij Distance from supplier i(i∈ I) to biomass power plant j (j∈J) α Biomass purchasing unit cost
Conversion coefficient Biomass electricity demand coefficient à Utilization coefficient
P Maximum number of plants to be opened
Variables y jk Equals 1 if a biomass power plant type k (k∈K) is opened at site j
30 x ijt The flow from supplier i (i∈ I) to biomass power plants j (j∈ J) by vehicle type t (t∈ V)
We present below a formulation of the multiple objectives biomass supply chain planning can be formulated as follows:
1 t ijt ij ijt jk k jk k i I j J t V i I j J t V j J k K j J k K
In the above model, the first objective function Z1, minimises the sum of the opening fixed costs, the Operation and Maintenance (O&M), logistics, and purchasing costs The second objective function, Z2, minimises CO2 emissions released from transportation The electricity demand must be satisfied by following Constraint (3), which ensures the
31 amount of biomass transported to satisfy 𝛽 of the current demand Constraint (4) and (5) ensures that the amount of biomass going from supplier i to plant j by vehicles t must be smaller or equal to the supplier’s capacity and the capacity of plant type k in case the plant is opened Constraint (6) ensures that the number of biomass power plants can be opened Constraint (7) requires only one biomass power plant to be opened on a site Constraint (8) requires the capacity utilized by the biomass power plants Constraints (9)–(10) are the constraints on the decision variables
Fuzzy objective
General multiple objectives for supply chain planning can be stated as follows:
As shown in Eq (11), ( ), f x 1 f x 2 ( ), f x 3 ( ), , f x n ( ) are negative objectives that want to be minimised, such as total cost, amount of CO2, etc As presented in Eq (12),
1 ( ), 2 ( ), 3 ( ), , ( ) n n n m f + x f + x f + x f x are the positive objectives that must be maximized, such as benefits, number of job offerings, and so on Furthermore, X d is the set of feasible solutions (Eq (14)) that satisfy the s constraints (as shown in Eq (13))
In many real-world problems, attaining all objectives simultaneously is not always feasible due to the constraints inherent in the system To address this, decision-makers can establish specific tolerance thresholds and membership functions [f j (x)] for the j th fuzzy objectives The membership functions are designed to increase linearly from 0 to
1, reflecting the gradual fulfilment of objectives as they move from complete non- fulfilment to complete satisfaction
32 Let f q (x) and f p (x) be the minimisation and maximisation goals, respectively The membership functions of those objectives are formulated as follows By approaching the multi-objective problem as a single objective, one focuses on optimising each objective individually Let F q − and F q + be the optimal and worst value of the negative objectives, while the optimal and worst value of the positive objectives are presented by F p + and
F p − , respectively [30, 31] Linear membership functions are shown in Fig 3, and the value of membership functions are calculated as Eq (15) and Eq (16)
Fig 3 6 Objective functions as the fuzzy number [25]
An additive model with objectives' weights is formulated as follows to consider the importance of objectives when applying the max-min approach [32] The objective function Eq (17) aims to maximize the satisfaction from objectives Eq.(18) presents the
33 constraints of models All of the weights must be positive, and the sum of them equal to
1, as shown in Eq (19) X d is the set of feasible solutions presented in Eq (20)
Model validation
To validate the correctness of the model, we are supposed to have three suppliers with the capacity and distance matrix as illustrated in Tables 3.2 and 3.3 The input data of the model includes: The demand is 745 million Kwh and two types of vehicles: small trucks and large trucks The transportation cost and carbon emission rate for small trucks and large trucks are 0.3 (USD/ton/km), 0.04 (kg/ton/km) and 0.12 USD/ton/km and 0.1 kg/ton/km, respectively
Table 3 2 Capacity of each supplier (Unit: Tonnes)
Table 3 3 Distance between suppliers and Industrial Parks
34 The location and capacity of suppliers and Industrial Parks are illustrated in Fig 3.7:
Fig 3 7 Illustration of suppliers and industrial parks
Coordinate of suppliers and Industrials Parks are shown as: S1(4,13), S2(15, 10), S3( 8, 3), F1(6, 12), F2(11,16), F3(17,1)
Table 3 4 The trade-off between the three scenarios of model validation
Units: Cost: USD, CO 2 :kg
Based on the findings of three models, it is evident that the optimal plan involves establishing a single biomass power plant in Industrial Park F1 with a capacity of 10 MW Biomass supply will be sourced from suppliers S1 and S2, with quantities of 52000 and 10741, respectively The satisfaction ratio of 0.5 indicates that this outcome aligns with the initial assumption, supporting the reliability of the model and its applicability to real-world scenarios.
The similarity in costs and total CO2 emissions between the fuzzy multi-objective model and the total cost optimization model arises from the small dataset utilized In this scenario, the total cost optimization result aligns with the optimal outcome Despite assigning equal weight to both criteria, the fuzzy multi-objective model's outputs with a small dataset tend to mirror the best outcomes from either the total cost optimization model or the multi-objective model.
CASE STUDY
Input data
The Mekong River Delta has many potentials and advantages for development It is one of the largest and most numerous Deltas in Southeast Asia and the world, and it is the largest rice-growing region in Vietnam, including area, rice yield, and productivity Contributing about 50% of rice productivity, 95% of rice exports, nearly 65% of aquaculture productivity, 60% of fish exports and nearly 70% of the country's fruits The Mekong River Delta in Vietnam is abundant in agricultural biomass; however, bioenergy development is still limited This region has considerable potential for oil and gas and renewable energy, such as wind energy, solar energy, and tidal energy in general and biomass energy in particular, to develop green energy in the Mekong River Delta The Mekong River Delta is oriented to be the centre of renewable energy in Viet Nam It can provide fully high-quality energy and meet the demand to develop economically and socially, especially in remote and coastal areas
Rice husk power plant projects have an average capacity of 10 MW per plant, all concentrated in the Mekong Delta region It is also a place with a massive demand for heat and electricity during the rice milling season while being far from sources of fossil fuels such as coal According to data from the General Statistics Office, there was a significant increase in Mekong Delta River rice production from 2006 to 2021 (see Fig 4.1) Production volume has continuously been stable since 2020 despite the impact of the COVID-19 pandemic Therefore, a positive trend is predicted over the next five years, leading to a greater abundance of biomass sources and ensuring a consistent biomass supply to support the operation of planned power plants
Fig 4 1 Rice production of Mekong River Delta
This study centres on utilizing rice husk as the primary material for operating steam turbine power plants, intending to meet 6.6% of the electricity demand
The quantity of rice husk required is estimated based on the rice production volume, utilizing a coefficient established in previous research: approximately 20% of the weight of rice constitutes rice husk [33] Used rice husk quantity is assumed to be 50% Furthermore, 1.276 tons of rice husk could generate about 1000 kWh [34]
In this research, the suppliers are assumed to be located in the central of 131 districts in the Mekong River Delta At the same time, potential biomass power plants will be installed in 68 Industrial Park in the Mekong River Delta Distances among suppliers and plants are determined based on latitude-longitude data that could be identified from Google Maps
The rice husk price in the Mekong Delta River is 26.36 USD on average The rice husk transportation cost from suppliers to power plants relies on the type of trucks: the cost is 0.2090 USD per ton per kilometre for 3.5-ton trucks (denoted as small trucks) and 0.1710 USD per ton per kilometre for 21.5-ton which are denoted as large trucks (The prices were adopted from local transportation providers quote) The selected truck is based on the available transportation situation in the Mekong Delta River, where small trucks aim
Rice production volume of Mekong River Delta
38 to carry rice husks in the daytime, while large trucks are used at nighttime Furthermore, the associated carbon emission rates for these small and large trucks are 0.3014 kg/km and 0.8886 kg/km [35], respectively The carbon emission rates per ton of rice husk are 0.0319 kg/ton/km and 0.1346 kg/ton/km
As the strategy outlined by the Ministry of Industry and Trade of the Socialist Republic of Viet Nam, there is a proposed installation of two types of power plants: one with a capacity of 10MW and another with a capacity of 20MW The investment costs for these plants have been gathered from the actual project plans of Tien Giang and Hau Giang rice husk power plants, which are 18.6 million USD and 37.2 million USD, respectively
In order to determine the annual cost= [IVC×r×(1+r) T ]/[(1+r) T -1], the investment cost (IVC) is converted, considering a projected operational span of 20 years (T) and an annual interest rate (r) of 15% Additionally, Operation and Maintenance Costs account for 7% of the investment cost annually [36]
The capacity utilized of the biomass power plant is assumed to equal 60% based on the efficiency of the burning biomass for heat or electricity using a steam turbine system.
Computational result and analysis
To ensure model validity, the proposed optimization model was implemented in IBM ILOG CPLEX Optimization Studio 12.10.0 [37] and executed on a laptop featuring an Intel Core i5 3.10 GHz processor and 4.0 GB RAM The model underwent initial validation with a simple dataset, ensuring its functionality before applying it to a larger, practical dataset.
39 According to table 4.1, when the model optimizes the total cost objective individually, the plants should be opened in Binh Hoa Industrial Park (BH IP), Vam Cong (VC IP), Xuan To Industrial Park (XT IP), each with a capacity of 10MW The rice husk in Cho Moi, Chau Phu, Phu Tan, Chau Thanh AG (CT AG) should be transferred to Binh Hoa Industrial Park (BH IP) while the rice husk in Long Xuyen, Chau Thanh AG (CT AG) should be transferred to Vam Cong Industrial Park and the Xuan To Industrial Park will be served by Chau Doc, Tinh Bien and Tri Ton
Table 4 1 The amount of biomass transported to opening plants in An Giang Province with the total cost optimization objective (Unit: Tonnes)
Binh Hoa Industrial Park (BH IP)
Vam Cong Industrial Park (VC IP)
Xuan To Industrial Park (XT IP)
Chau Thanh AG Large truck 556.4 50,775.8
On the other hand, only two biomass power plants were recommended for opening when the model optimized the carbon emissions objective: one in Xuan To Industrial Park with a capacity of 10MW and one in Binh Hoa Industrial Park (BH IP) with a capacity of 20MW Binh Hoa Industrial Park (BH IP) will be served by Cho Moi, Long Xuyen, Chau Phu, Phu Tan, and Chau Thanh AG (CT AG); Xuan To Industrial Park will be supplied by Chau Doc, Tinh Bien, and Tri Ton
Table 4 2 The amount of biomass transported to opening plants in An Giang with the carbon emissions optimization objective (Unit: Tonnes)
Binh Hoa Industrial Park (BH IP)
Xuan To Industrial Park (XT IP)
Chau Thanh AG Small truck 51,332.2
Using fuzzy linear programming yielded similar outcomes to the total cost optimization model It identified the same number of power plants as potential installation sites: Binh Hoa Industrial Park, Vam Cong Industrial Park, and Xuan To Industrial Park for 10MW biomass power plants Despite the equal importance of both objectives, the model prioritized minimizing total cost Additionally, a large truck was selected to meet the demand of the plants The satisfaction ratio remained at 0.5, consistent with the results obtained from the validation model using a smaller data set.
Table 4 3 The trade-off between the three scenarios in An Giang
Units: Cost: USD, CO 2 :kg.
Fig 4 2 Suppliers and biomass power plant candidate in An Giang
According to table 4.4, when the model optimizes the total cost objective individually, the plants should be opened in Tran Quoc Toan Industrial Park (TQT IP), Tan Kieu Industrial Park (TK IP), Song Hau Industrial Park (SH IP), each with a capacity of 10MW The rice husk in Cao Lanh District, Cao Lanh City, Thanh Binh should be transferred to Tran Quoc Toan Industrial Park (BH IP) while the rice husk in Thap Muoi will only be transferred to Tan Kieu Industrial Park (TK IP) and Song Hau Industrial Park (SH IP) will be served by Cao Lanh District, Lap Vo, Lai Vung, Chau Thanh DT,
Table 4 4 The amount of biomass transported to opening plants in Dong Thap with the total cost optimization objective (Unit: Tonnes)
Tran Quoc Toan Industrial Park (TQT IP)
Tan Kieu Industrial Park (TK IP)
Song Hau Industrial Park (SH IP)
Cao Lanh District Large truck 46,174.2 10,942.5
Cao Lanh City Large truck 2,475.8
Chau Thanh DT Large truck 6,269.3
On the other hand, only two biomass power plants were recommended for opening when the model optimized the carbon emissions objective: one in Tran Quoc Toan Industrial Park (TQT IP) with a capacity of 20MW and one in Tan Kieu Industrial Park (TK IP) with a capacity of 10MW The amount of rice husk for each biomass power plants are shown in Table 4.5
Table 4 5 The amount of biomass transported to opening plants in Dong Thap Province with the carbon emissions optimization objective (Unit: Tonnes)
Tran Quoc Toan Industrial Park (TQT IP)
Tan Kieu Industrial Park (TK IP)
Cao Lanh District Small truck 57,116.7
Cao Lanh City Small truck 2,475.8
43 The total cost, constituents, and amount of CO2 of the three models are presented in detail in Table 4.6 In the first model, the total cost is optimized individually, while the last model aims to optimize the carbon emissions The satisfaction ratio is 0.536, better than validation model The fuzzy multi-objective model helps to balance total cost and total carbon emissions objectives
Table 4 6 The trade-off between the three scenarios in Dong Thap
Units: Cost: USD, CO 2 :kg
As a result of the fuzzy multi-objective model, there are 3 plants opened in Tran Quoc Toan Industrial Park (TQT IP), Tan Kieu Industrial Park (TK IP), Song Hau Industrial Park (SH IP), each with a capacity of 10MW The number, location with capacities of the model are similar to the total cost optimization model But there's an substantial distinction in the types of vehicles chosen Utilizing small trucks is advised by the model in order to comply with carbon emissions objectives More small vehicles are required to meet this requirements even though the biomass transportation volume in a fuzzy multi-objective linear model and the total cost objective remain the same Consequently, the transportation cost associated with fuzzy multi-objective linear programming is the balanced cost of three models and the most miniature total CO2 despite the slight increase in total cost
Fig 4 3Suppliers and biomass power plant candidate in Dong Thap
Based on individual optimization objectives for total cost and carbon emissions, the optimal solution involves installing a single 10 MW plant in Hau Giang Rice husk from various suppliers (Phung Hiep, Chau Thanh A, Vi Thuy, Nga Bay) will be directed to meet this plant's demand However, when optimizing for carbon emissions, rice husk from Chau Thanh HG is additionally included to meet the plant's requirements The specific biomass quantities from each supplier are detailed in Tables 4.7 and 4.8 to ensure efficient transportation to the plant.
Table 4 7 The amount of biomass transported to opening plants in Hau Giang with the total cost optimization objective (Unit: Tonnes)
Suppliers Type of vehicles Tan Phu Thanh Industrial
Table 4 8 The amount of biomass transported to opening plants in Hau Giang with the carbon emissions optimization objective (Unit: Tonnes)
Suppliers Type of vehicles Tan Phu Thanh Industrial
Chau Thanh HG Small truck 9.1
The outcomes of the fuzzy multi-objective model are similar to the CO2 optimization model, with only one biomass plant opening in Tan Phu Thanh with 10MW The satisfaction ratio is 0.5
Table 4 9 The trade-off between the three scenarios in Hau Giang
Units: Cost: USD, CO 2 :kg
Fig 4 4 Suppliers and biomass power plant candidate in Hau Giang
The optimization initially concentrates just on the total cost objective According to the model, it seems to make the most economic to establish three plants: a 10MW plant in Thanh Loc Industrial Park (TL IP), a 10MW plant in Xeo Ro Industrial Park (XR IP) and a 10MW plant in Thuan Yen Industrial Park (TY IP)
Table 4 10 The amount of biomass transported to opening plants in Kien Giang with the total cost optimization objective (Unit: Tonnes)
Thanh Loc Industrial Park (TL IP)
Xeo Ro Industrial Park (XR IP)
Thuan Yen Industrial Park (TY IP)
Chau Thanh (CT3) Large truck 26,009.4
47 The establishment of biomass power plants with a capacity of 10MW each in Thanh Loc Industrial Park (TL IP), Xeo Ro Industrial Park (XR IP), and Thuan Yen Industrial Park (TY IP) is the ideal configuration when the model optimizes the carbon emissions objective The allocation details are comparable to the individually optimized total cost objective solutions, but they use smaller trucks
The ratio of satisfaction is 0.492, which is not as good as the result of the validation model The results of fuzzy linear programming balance the total cost objective, but the result of the total carbon emission is higher.Even though both models have the same demand, the fuzzy model selects a small truck to satisfy it, increasing both the total cost and the output of carbon emissions The number of power plants and the preferred locations for establishment are similar: Thanh Loc Industrial Park (TL IP), Xeo Ro Industrial Park (XR IP) and Thuan Yen Industrial Park (TY IP) are selected for biomass power plants of 10MW
Table 4 11 The trade-off between the three scenarios in Kien Giang
Units: Cost: USD, CO 2 :kg
Fig 4 5 Suppliers and biomass power plant candidate in Kien Giang
Initially, the optimization focuses on the total cost objective in isolation The model determines that it is optimal to establish two plants: one in Tran De Industrial Park (TD IP) with a capacity of 10MW and another in An Nghiep Industrial Park (AN IP) with a capacity of 10MW Regarding the transportation of rice husks, biomass from Soc Trang (ST), Chau Thanh (CT), My Tu (CT), Long Phu (LP), and My Xuyen (MX) are transported to An Nghiep Industrial Park, while rice husks from the Tran De supplier (TD) and a remains of Long Phu (LP) are designated for use in Tran De Industrial Park (TD IP) as shown in Table 4.12
Table 4 12 The amount of biomass transported to opening plants in Soc Trang with the total cost optimization objective (Unit: Tonnes)
An Nghiep Industrial Park (AN IP)
Tran De Industrial Park (TD IP)
Soc Trang (ST) Large truck 4,141.5
Chau Thanh (CT1) Large truck 26,461.2
My Tu (MT) Large truck 17,433.9
49 Long Phu (LP) Large truck 3,535.3
My Xuyen (MX) Large truck 15,024.2
Tran De (TD) Large truck 25,039.7
Long Phu (LP) Large truck 28,552.2
To minimize carbon emissions, the optimal configuration involves constructing biomass power plants in Tran De Industrial Park (TD IP) and An Nghiep Industrial Park (AN IP), both with capacities of 10MW The allocation of resources mirrors that of the separate cost optimization solutions, but with adjustments: small trucks will be used for transportation, and Ke Sach will replace My Tu as the supplier for AN IP in the same quantities.
Sensitivity analysis
Sensitivity analysis is conducted to illustrate the impact of two factors: demand of suppliers and utility levels of biomass power plant
Sensitivity analysis reveals a positive correlation between demand and the number of biomass plants and their capacity As demand increases, from 2% to 7%, the number of plants required and their capacity rise accordingly This suggests that meeting higher demand necessitates the opening of additional plants and the expansion of existing capacities to utilize available rice husks.
Table 4 18 The relationship between biomass networks and electricity demand with the total cost optimization objective
Name and type of plant
2% 7 10MW: Thanh Loc IP, Tan Kim IP, Nhut Chanh IP, Hoa Binh IP, Tan
Kieu IP, Ba Sao IP 20MW: Dai Ngai IP 3% 11 10MW: An Nghiep IP, Tran De IP, Xeo Ro IP, Tran Quoc Toan IP, Tan
Kieu IP, Tra Noc 2 IP, Thot Not IP, Hoa Phu IP, Long Duc IP, Tra Kha
IP 20MW: Binh Hoa IP 4% 11 10MW: Tran De IP, Vam Cong IP, Tac Cau IP, Tra Noc 2 IP, Thot Not
IP, Hoa Phu IP 20MW: Binh Hoa IP, Tran Quoc Toan IP, Tan Kieu IP, Cau Quan IP, Tra Kha IP
61 5% 14 10MW: Tac Cau IP, Tan Phu IP, Tan Kieu IP, Thot Not IP, Song Doc
IP, Hoa Phu IP, Tan Phu Thanh IP 20MW: Dai Ngai IP, Vam Cong IP, Binh Hoa IP, Thuan Yen IP, Long Duc IP, Tra Kha IP
6% 19 10MW: An Nghiep IP, Tran De IP, Vam Cong IP, Xuan To IP, Thanh
Loc IP, Thuan Yen IP, Xeo Ro IP, Tra Noc 2 IP, Hoa Phu IP, Cau Quan
IP, Long Duc IP, Tra Kha IP, Lang Tram IP, Tan Phu Thanh IP 20MW: Binh Hoa IP, Tac Cau IP, Tran Quoc Toan IP, Tan Kieu IP, Thot Not IP
Table 4 19 Cost breakdown from sensitivity analysis of demand with the total cost optimization objective
9,593,385 7% Infeasible Infeasible Infeasible Infeasible Infeasible Infeasible
In this scenario, the greater the demand, the more fixed costs, operation and maintenance costs, purchasing costs, and transportation costs are increasing due to the associated demand The solution prioritizes opening more 20MW capacity biomass power plants and large trucks to fulfill the demand
Fig 4 8 Cost breakdown from sensitivity analysis of demand with the total cost optimization objective
Table 4 20 The relationship between biomass networks and electricity demand with the carbon emissions optimization objective
Demand Number of plants Name and type of plant
10MW: An Nghiep IP, Tran De IP, Binh Hoa IP, Tac Cau
IP, Tran Quoc Toan IP, Tan Kieu IP, Thot Not IP, Hoa Phu
IP, Cau Quan IP, Long Duc IP, Tra Kha IP
10MW: An Nghiep IP, Tran De IP, Dai Ngai IP, Xuan To
IP, Tac Cau IP, Tran Quoc Toan IP, Tra Noc 2 IP, Thot Not
IP, Hoa Phu IP, Cau Quan IP, Long Duc IP, Tra Kha IP, Tan Phu Thanh IP
20MW: Binh Hoa IP, Tan Kieu IP
10MW: An Nghiep IP, Tran De IP, Dai Ngai IP, Vam Cong
IP, Xuan To IP, Xoai Rap IP, Thuan Yen IP, Xeo Ro IP, Tan Phu IP, Song Hau IP, Thot Not IP, Hoa Phu IP, Cau Quan IP, Long Duc IP, Tan Phu Thanh IP
20MW: Binh Hoa IP, Tran Quoc Toan IP, Tan Kieu IP, Tra Kha IP
Cost breakdown from sensitivity analysis of demand with the total cost optimization objective
Fixed cost Operation & Maintainance Cost Purchasing Cost Transportation Cost
10MW: An Nghiep IP, Tran De IP, Dai Ngai IP, Long Hung
IP, Vam Cong IP, Xuan To IP, Long Giang IP, Xoai Rap IP, Thuan Yen IP, Xeo Ro IP, Tan Phu IP, Song Hau IP, Tra Noc 2 IP, Thot Not IP, Hoa Phu IP, Cau Quan IP, Long Duc
IP, Tra Kha IP, Lang Tram IP, Tan Phu Thanh IP 20MW: Binh Hoa IP, Tran Quoc Toan IP, Tan Kieu IP
10MW: An Nghiep IP, Tran De IP, Dai Ngai IP, Long Hung
IP, Vam Cong IP, Xuan To IP, Long Giang IP, Xoai Rap IP, Thuan Yen IP, Xeo Ro IP, Tan Phu IP, Song Hau IP, Tra Noc 2 IP, Song Doc IP, Hoa Phu IP, An Dinh IP, Cau Quan
IP, Long Duc IP, Tra Kha IP, Lang Tram IP, Tan Phu Thanh
IP 20MW: Binh Hoa IP, Thanh Loc IP, Tac Cau IP, Tran Quoc Toan IP, Tan Kieu IP, Thot Not IP
Table 4 21 Cost breakdown from sensitivity analysis of demand with the carbon emissions optimization objective
7% Infeasible Infeasible Infeasible Infeasible Infeasible Infeasible
Fig 4 9 Cost breakdown from sensitivity analysis of demand with the carbon emissions optimization objective
The total cost objective aims to minimize production costs, resulting in the opening of more plants with lower capacities In contrast, the carbon emission objective prioritizes reducing greenhouse gas emissions Consequently, it favors opening more smaller-capacity plants (e.g., 10MW), which reduces transmission distances and overall carbon emissions This approach balances the demand and supply while mitigating environmental impact.
Table 4 22 The relationship between biomass networks and electricity demand with the fuzzy optimization objective
Demand Number of plants Name and type of plant
10MW: An Nghiep IP, Tran De IP, Tran Quoc Toan IP, Tan Kieu IP, Thot Not IP, Tra Kha IP
10MW: Tran De IP, Vam Cong IP, Hoi An IP, Thuan Yen
IP, Tac Cau IP, Long Hau IP, Tran Quoc Toan IP, Ba Sao
IP, Tra Noc 1 IP, Binh Minh IP, Long Duc IP, Tra Kha IP
Cost breakdown from sensitivity analysis of demand with the carbon emissions optimization objective
Fixed cost Operation & Maintainance Cost Purchasing Cost Transportation Cost
10MW: An Nghiep IP, Tran De IP, Vam Cong IP, Binh Hoa
IP, Thanh Loc IP, Thuan Yen IP, Xeo Ro IP, Tran Quoc Toan IP, Ba Sao IP, Thot Not IP, Cau Quan IP, Long Duc
IP, Tra Kha IP, Tan Phu Thanh IP 20MW: Tan Kieu IP
10MW: An Nghiep IP, Vinh Chau IP, Long Hung IP, Long Giang IP, Thanh Loc IP, Thuan Yen IP, Xeo Ro IP, Tan Phu
IP, Tan Kieu IP, Ba Sao IP, Hung Phu 1 IP, Hung Phu 2A
IP, Thot Not IP, Binh Minh IP, Cau Quan IP, Lang Tram IP 20MW: Binh Hoa IP, Hoi An IP
10MW: Tran De IP, Long Hung IP, Vam Cong IP, Xuan To
IP, Tan Huong IP, Xoai Rap IP, Thanh Loc IP, Thuan Yen
IP, Tac Cau IP, Xeo Ro IP, Nhut Chanh IP, Tra Noc 2 IP, Hoa Phu IP, Cau Quan IP, Long Duc IP, Lang Tram IP, Tan Phu Thanh IP
20MW: Binh Hoa IP, Tran Quoc Toan IP, Tan Kieu IP, Thot Not IP
Table 4 23 Cost breakdown from sensitivity analysis of demand with the fuzzy optimization objective
7% Infeasible Infeasible Infeasible Infeasible Infeasible Infeasible
66 Table 4.23 and Fig 4.10 presents the change in demand to decide how many plants should be opened The outcome of fuzzy linear programming is the balanced solution of the two objectives above The number of power plants is opened based on consideration of both economic and environmental factors The more demand there is, the more biomass power plants are opened The increase in demand caused an increase in the opening of biomass power plants, which led to more biomass being utilized to adapt to the demand of these plants To align with carbon emissions, the fuzzy multi-objective model also uses small trucks to fulfill the capacity of biomass power plants
Fig 4 10 Cost breakdown from sensitivity analysis of demand with the fuzzy optimization objective
In the model, the utility levels of the biomass power plant is determined to be 60% based on the utility of the steam turbine system used in this plant To investigate more about the impact of utility on the number and capacity of opening plants In the sensitivity analysis, the effect of utility is observed on plant locations, with capacity decisions and optimal results As shown in Table 4.24, the change in utility levels causes a difference
Cost breakdown from sensitivity analysis of demand with the fuzzy optimization objective
Fixed cost Operation Cost Purchasing Cost Transportation Cost
67 in the opening of plants and capacity levels of plants However, with any utility levels, the number of open plants and their location does not change when the objective is to minimize the total cost
In the fuzzy multi-objective model, 10 MW capacity plants will increase, and 20 MW capacity plants will decrease from 70% to 90%, respectively The solutions from fuzzy multi-objective linear programming are the balance of two objectives: total cost and carbon emission The outcomes of the cost breakdown from the sensitivity analysis of utility obtained from the three models are shown in the table 4.25
Table 4 24 Number of plants with the different utility levels
Demand Utility levels Optimization Model 10MW 20MW
Table 4 25 Cost breakdown for each utility levels scenarios
4.3.3 Impact of both demand and utility levels
If the total cost objective is not considered, it could potentially lead to the selection of power plants that result in suboptimal utility levels To address this, a practical approach for future enhancements could involve introducing a threshold associated with power plant utility This threshold would serve as a criterion to ensure that the selected power plants fulfill a minimum utility requirement, thus preventing the model from allocating resources to power plant configurations that fall below an acceptable level of efficiency or effectiveness This addition could help balance optimizing various objectives while maintaining a specific performance standard In order to investigate the impact of both demand and utility, Monte Carlo simulation is used to generate 18 scenarios from demand levels and utility levels
Table 4 26 Scenerios generation and results
Demand Utility levels Optimization Model 10MW 20MW
Carbon Emissions Infeasible Fuzzy multi-objective Infeasible 80%
Carbon Emissions Infeasible Fuzzy multi-objective Infeasible 90%
Carbon Emissions Infeasible Fuzzy multi-objective Infeasible
In almost all scenarios, plants' total number and capacity levels have no effect and do not change when individually optimizing the total cost objective When the demand is 3%, the solution of the total cost objective and fuzzy multi-objective are the same The demand is 4%, and the total megawatts of all the total costs and fuzzy multi-objective opening plants are similar Still, they are different in the number and capacity of types of biomass Meanwhile, in the solution from a fuzzy multi-objective model of 5% demand, the total MW of all biomass power plants is always 200 MW, with only a change in the number of each type of biomass power plant The greater the utility, the fewer biomass power plants are opened when the model optimizes carbon emissions individually With demand at 6% and utility at 70%, the fuzzy model prioritizes open 20MW plants, and vice versa, with total cost and carbon emission objectives The utility is between 80 and 90%, and the solution becomes more balanced The increasing demand causes an increase in the required material That leads to the number of opening power plants being higher and higher with the increase in demand However, if it reaches a threshold where the biomass cannot supply enough, the solution will stop and become infeasible.
CONCLUSION, FUTURE DIRECTION
Conclusion
Driven by rising population and industrialization, the demand for renewable energy sources has surged Among these, biomass, a renewable energy source, stands out as the primary resource for Vietnam's future energy strategy.
30 years Hence, the location of the opening plant also plays an essential role in determining the effect of the operation Suppose the plant is not opened in the optimal location In that case, it leads to many problems because of very high fixed costs, transportation costs, and lack of feedstock, especially when seasonal availability is a significant characteristic of almost all biomass resources We formulated an optimization model for location optimization and weighted max-min for fuzzy-objective problems for handling the conflict between total cost and carbon emissions objectives
The decision is made based on the importance of the objective that needs to be optimized Therefore, for each different objective, we can explain why the optimization model suggests other locations and capacity levels There are no best solutions for the optimization model; it depends on the objectives that decision-makers need to achieve Our model's efficiency is illustrated using a case study in the Mekong River Delta in Vietnam As a result, it has been determined that 25 biomass power plants should be established in the Mekong River Delta to ensure all of the region's rice husk is utilized.
Limitation
This research has some limitations that expose the opportunity for future research To begin with, the model’s illustrations are based on a restricted dataset Therefore, future studies could involve running the model on multiple datasets to yield a broader range of insights Secondly, we assume that the two objectives, total cost and carbon emissions, hold equal importance However, these weights could be refined and justified This can
72 be achieved by employing a multi-criteria decision-making (MCDM) technique such as the Analytic Hierarchy Process (AHP) [38].
Future direction
Additionally, this research can be extended by considering different objectives such as total cost and CO2 emissions for the whole supply chain [39], economic, carbon emissions and social rejection objectives [40] etc Furthermore, given the favourable natural conditions of the Mekong River Delta, it presents an ideal environment for developing complex transportation models that incorporate multiple modes of transport, including both trucks and inland waterways [41] A potential future exploration could include additional transportation modes beyond those currently considered This expansion could lead to a more comprehensive and realistic representation of the transportation landscape, enabling a more accurate assessment of the supply chain dynamics and optimization potential
T D Nguyen, N.-H Do, and D D Nguyen, “ Weighted max-min model for fuzzy multi-objective biomass power plants location optimization problem”, in the Book Series of EAI/Springer Innovations in Communication and Computing (Print
Weighted max-min model for fuzzy multi-objective biomass power plants location optimization problem
Thanh Dat Nguyen 1,2[0009-0006-1426-8384] , Duc Duy Nguyen 1,2,*[0000-0001-7553-5913] , Ngoc-Hien Do 1,2 [0000-0001-9322-3518]
1 Department of Industrial Systems Engineering, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong
Kiet Street, District 10, Ho Chi Minh City, Vietnam
Vietnam National University Ho Chi Minh City (VNU-HCM) is located in Linh Trung Ward, Thu Duc City, Ho Chi Minh City, Vietnam The contact information for VNU-HCM is as follows: Email: nguyenthanhdat.sdh212@hcmut.edu.vn; duy.nguyen@hcmut.edu.vn; hienise97@hcmut.edu.vn.
Abstract Renewable energy significantly reduces a nation's reliance on fossil fuels, enhances energy security, and promotes sustainable development, among other positive impacts Strategic selection of facility locations is the most important in designing a supply chain network To develop a sustainable supply chain, we propose a multi-objective mixed integer programming approach that optimises the location of biomass plants while simultaneously considering total costs and environmental considerations Recognising the uncertainty and varying importance of objectives, we formulate a weighted max-min fuzzy model to handle vague input data and diverse objective weights The proposed framework is illustrated by employing a practical dataset from a province in the South of Vietnam, a supply chain network comprising 11 suppliers and 6 potential biomass power plant locations The analysis incorporates two categories of biomass power plants and two kinds of trucks
By implementing our model, decision-makers gain the ability to optimise facility locations and effectively manage supply chain performance
Keywords: Biomass, Sustainability, Fuzzy Optimization, Supply Chain Management, Renewable Energy
Nowadays, fossil fuels are being promoted to be replaced by cleaner and sustainable energy sources (e.g wind power, solar power, bioenergy, etc.) in order to address sustainable economic growth, environmental problems and energy security in many countries [1-10] Owing to prevailing environmental and strategic imperatives, biomass is presently surfacing as a prospective alternative for effecting global energy source diversification [2, 10] According to reports from [8], renewable energy plays essential roles and accounts for over one-fifth of the global generation
It is expected to become the primary source of electricity production by 2025, accounting for over one-third of the total global generation [11] This is the reason behind the significant increase in investment in renewable energy technology since 2004
As shown in Fig 1, bioenergy accounts for 2%, a small proportion of renewable energy generation by source Thus, developing bioenergy is still a potential aspect around the world Among the many types of biomass, agricultural residual biomass stands out due to its pronounced cost-efficiency relative to other alternatives Therefore, it is an opportunity for many agricultural-producing countries to develop renewable energy from biomass [7]
Fig 1 Electricity generation by source in the world 2021 4
Vietnam has maintained an impressive GDP growth rate of over 6% per year over the past two decades As a result, the electricity demand significantly increased to develop the economy and serve over 99 million inhabitants Viet Nam's electricity consumption has been increasing quickly over time, which has fueled the nation's economic development The National Electricity Development Planning period 2021-2030, Vision 2050 proposed the goal of biomass energy rating in electric generation is 1% in 2020, 2.1% in 2030, and 8.1% in 2050
Agricultural and forest residuals in Vietnam are diverse and continuously increasing, potentially generating about 7000 Mega Watt, as reported in the National Power Development Plan for 2021-2030 (NPDP8) However, these residuals are seen as natural waste, and being wasted is more dangerous to the cause of environmental pollution The Mekong Delta produces more than 50% of the nation's rice in Vietnam, making it Vietnam's largest rice production region 5 ; therefore, the agricultural residual is more considerable than in other regions The Mekong Delta region makes up 33.4% of the potential for developing the biomass energy sector and has the country's highest proportion The rice production of 2021 is 43.79 million tonnes, so rice husk production is also significant However, its traditional uses are as a fertiliser additive, in stock breeding rugs, as cooking fuel, and in landfill or paving applications Therefore, using the rice husk to generate electricity could benefit the Mekong Delta's natural environment
A typical biomass supply chain is shown in Fig 2 Biomass energy can be produced in various ways, such as burning biomass for heat or electricity using steam turbines or using biomass for heat in thermal systems and generating electricity as well (this is known as “combined heat and power”), or converting feedstocks into liquid biofuels, and so on Biomass may come from different sources, such as forest residue, agricultural residue, animal waste, etc [12] The availability of biomass varies across different locations and time periods, depending on factors such as climate, land use, and human activities Therefore, biomass power plants can utilize more than one type of biomass as material to operate This research focuses on using rice husk to burn biomass for heat generation to run the steam turbines In biomass supply chain design, biomass power plant locations are key decisions for efficient design supply chain planning [1-4] Besides, the decision-makers also need to consider other objectives, such as environmental and economic benefits Therefore, this research focuses on developing a mathematical model to define the biomass power plant locations to minimise total cost and carbon emissions associated with transportation activities mainly relying on trucks The total cost includes biomass power plant establishment, Operation and Maintenance (O&M), purchasing, and transportation costs A weighted max-min model for fuzzy multi-objectives is formulated to solve the multiple objectives problem
4 Our World in Data-https://ourworldindata.org/
5 General Statistics Office-https://www.gso.gov.vn/
Demand Zones Biomass power plants
Fig 2 A typical biomass supply chain
Leveraging the p-median problem, we formulated a mathematical model for supply chain design, introducing capacity constraints through a generalized capacitated p-median problem Recognizing the dynamic nature of demand, we incorporate adjustable facility capacities to optimize resource utilization Additionally, considering environmental sustainability, we integrate carbon emissions into our cost-optimization objective To handle multi-objective challenges, we employ a weighted max-min model, providing decision-makers flexibility in prioritizing objectives and accommodating varying preferences This study's practical significance lies in promoting renewable energy adoption and offering a versatile tool for supply chain optimization across diverse industries and regions.
The paper is structured as follows: A literature review on supply chain management and studies on the location of biomass power plants is introduced in Section II Next, Section III presents the research framework and mathematical model used in this study Section IV provides a case study in Vietnam Finally, the conclusion and future research are presented in Section V
Location optimization is a primary problem in supply chain management and addresses important aspects of designing renewable energy supply chain networks
In order to identify the facility locations, a typical model, namely p-median [13], was utilised in many research, especially in defining renewable energy facilities The p-median was adopted by Bargos et al [1] to find the optimal locations for opening new and/or expanding milling facilities for sugar and ethanol production in Brazil A heuristic greedy algorithm and standard branch-and-bound were developed to generate the solutions Teixeira et al [2] formulated a location-allocation model for developing Brazil's forest biomass power plant First, suitable zones for biomass activity were identified by applying fuzzy logic, and then an MIP was formulated for facility location optimization considering the availability of raw materials and the existing transportation modes Sahoo et al [3] addressed optimal plant sites for biofuels and co-product production in the State of Ohio that aim to the opening and biomass delivered costs The research utilised sustainable crop residues as raw materials for biogas plants A novel methodology that integrated Artificial Neural Networks, GIS and mathematical models was developed in
78 their work Sarker et al [4] formulated an MIP model to optimize the hubs and bio-methane gas locations to minimize the overall supply chain total cost for renewable energy A Genetic Algorithm was devised to generate the solutions, and then the results were benchmarked to results from the existing solutions obtained by traditional Lingo and other comparable solvers
Since sustainability involves the “Triple Bottom line” concept, including economic, environmental, and social sustainability [14], much research considers more than one object: economic and environmental objectives [5, 6, 10]; economic, environmental, and social benefits [9], and so on Specifically, Karimi et al [5] developed a stochastic bi-objective optimization model to optimize biomass cofiring supply chain planning while balancing conflicting objectives of minimising total costs and total carbon from coal acquisition and transportation Data about biomass and coal plants in North and South Carolina was used to illustrate the feasibility and effectiveness of the model In the same context of co-firing power generation, Aixia Chen and Yankui Liu [10] developed a robust fuzzy optimization model to identify new cofiring power plant locations among existing coal-fired plants in Heilongjiang Province in China The model aims to optimize total cost and minimize carbon emissions from burning coal, biomass preprocessing and transportation Nandimandalam et al [6] formulated a multi-objective mathematical model for location-allocation in biomass-to-energy supply chain planning problems The first objective is to minimize total supply chain costs: power plant establishment costs, biomass harvesting costs, inventory costs, transportation costs and so on Another objective is minimising the GHG emissions from all processes in the network Along with economic and carbon emissions objectives, Xu et al [9] also consider social rejection objectives when designing an animal waste-sourced biogas supply chain
This research formulates a mathematical model for multiple supply chain planning considering total cost and
LIST OF INDUSTRIAL PARK IN MEKONG RIVER DELTA
An Nghiep Industrial Park An Nghiep IP
Tran De Industrial Park Tran De IP
Dai Ngai Industrial Park Dai Ngai IP
Vinh Chau Industrial Park Vinh Chau IP
Long Hung Industrial Park Long Hung IP
My Thanh Industrial Park My Thanh IP
Vam Cong Industrial Park Vam Cong IP
Binh Hoa Industrial Park Binh Hoa IP
Hoi An Industrial Park Hoi An IP
Xuan To Industrial Park Xuan To IP
My Tho Industrial Park My Tho IP
Tan Huong Industrial Park Tan Huong IP
Long Giang Industrial Park Long Giang IP
Xoai Rap Industrial Park Xoai Rap IP
Thanh Loc Industrial Park Thanh Loc IP
Thuan Yen Industrial Park Thuan Yen IP
Tac Cau Industrial Park Tac Cau IP
Xeo Ro Industrial Park Xeo Ro IP
Kien Luong II Industrial Park Kien Luong II IP
Long An International Port Industrial Park Long An International Port IP
Phuoc Dong Industrial Park Phuoc Dong IP
Cau Tram Industrial Park Cau Tram IP
Dong Nam A Industrial Park Dong Nam A IP
Tan Kim Industrial Park Tan Kim IP
Tan Kim II Industrial Park Tan Kim II IP
Nhut Chanh Industrial Park Nhut Chanh IP
Hoa Binh Industrial Park Hoa Binh IP
Long Hau Industrial Park Long Hau IP
Duc Hoa I Industrial Park Duc Hoa I IP
Thuan Dao Industrial Park Thuan Dao IP
Xuyen A Industrial Park Xuyen A IP
Vinh Loc II Industrial Park Vinh Loc II IP
Phuc Long Industrial Park Phuc Long IP
Tan Duc Industrial Park Tan Duc IP
Duc Hoa III Industrial Park Duc Hoa III IP
An Nhat Tan Industrial Park An Nhat Tan IP
Tan Do Industrial Park Tan Do IP
Phu An Thanh Industrial Park Phu An Thanh IP
Hai Son Industrial Park Hai Son IP
Tan Phu Industrial Park Tan Phu IP
Nam Thuan Industrial Park Nam Thuan IP
Viet Phat Industrial Park Viet Phat IP
Sa Dec Industrial Park Sa Dec IP
Tran Quoc Toan Industrial Park Tran Quoc Toan IP
Tan Kieu Industrial Park Tan Kieu IP
Song Hau DT Industrial Park Song Hau DT IP
Ba Sao Industrial Park Ba Sao IP
An Hiep Industrial Park An Hiep IP
Giao Long Industrial Park Giao Long IP
Tra Noc I Industrial Park Tra Noc I IP
Tra Noc II Industrial Park Tra Noc II IP
Hung Phu I Industrial Park Hung Phu I IP
Hung Phu II Industrial Park Hung Phu II IP
Thot Not Industrial Park Thot Not IP
Khanh An Industrial Park Khanh An IP
Hoa Trung Industrial Park Hoa Trung IP
Song Doc Industrial Park Song Doc IP
Binh Minh Industrial Park Binh Minh IP
Hoa Phu Industrial Park Hoa Phu IP
Dong Binh Industrial Park Dong Binh IP
An Dinh Industrial Park An Dinh IP
Binh Tan Industrial Park Binh Tan IP
Cau Quan Industrial Park Cau Quan IP
Long Duc Industrial Park Long Duc IP
Tra Kha Industrial Park Tra Kha IP
Lang Tram Industrial Park Lang Tram IP
Song Hau KG Industrial Park Song Hau KG IP
Appendix 2 1 The amount of biomass transported to opening plants in Bac Lieu with the carbon emissions optimization objective (Unit: Tonnes)
Suppliers Type of vehicles Lang Tram Industrial Park
Appendix 2 2 The trade-off between the three scenarios in Bac Lieu
Fixed cost 2,971,563 2,971,563 2,971,563 O&M Cost 208,009 208,009 208,009 Purchasing Cost 2,323,990 2,323,990 2,323,990 Transportation Cost 417,765 510,601 510,601 Total cost 5,921,327 6,014,164 6,014,164 Total CO2 328,837 77,934 77,934
Appendix 2 3 The amount of biomass transported to opening plants in Vinh Long Province with the total cost optimization objective (Unit: Tonnes)
Suppliers Type of vehicles Hoa Phu Industrial Park
Vinh Long City Large truck 365.4
Appendix 2 4 The trade-off between the three scenarios in Vinh Long
93 Fixed cost 2,971,563 2,971,563 2,971,563 O&M Cost 208,009 208,009 208,009 Purchasing Cost 2,322,653 2,322,653 2,322,653 Transportation Cost 390,435 477,199 477,199 Total cost 5,892,661 5,979,425 5,979,425 Total CO2 307,325 72,836 72,836
Appendix 2 5 The amount of biomass transported to opening plants in Tra Vinh with the total cost optimization objective (Unit: Tonnes)
Suppliers Type of vehicles Long Duc Industrial Park
Chau Thanh (CT ) Large truck 23,858.9
Tra Vinh City Large truck 1,240.4
Appendix 2 6 The trade-off between the three scenarios in Tra Vinh
Fixed cost 2,971,563 2,971,563 2,971,563 O&M Cost 208,009 208,009 208,009 Purchasing Cost 2,244,289 2,244,289 2,244,289 Transportation Cost 328,832 401,906 401,906 Total cost 5,752,694 5,825,768 5,825,768 Total CO2 258,835 61,344 61,344