The Development of a Decision Support Model for Eco-Friendly Material Selection in Vietnam

: In recent years, the awareness of sustainable construction has increasingly risen in countries around the world, with the main goal being to avoid depleting energy resources and raw materials and to greatly reduce carbon emissions. Therefore, the selection of eco-friendly building materials becomes a di ﬃ cult task and choosing the best construction strategy is a complicated process. Most of the studies of the building material selection often focus on optimizing material-related green building scores with budget constraints based on the environmental impacts of those materials. However, these studies do not pay attention to the impact of sustainable materials on two important aspects of a project: The initial investment cost and the total labor-working days. Hence, this study developed a model that optimizes a material mix for buildings considering the building budget, total labor-working days, and material-related green building scores. A case study in Vietnam was conducted to illustrate the e ﬀ ectiveness of the proposed model. This proposed model provides a guidance for decision-makers in selecting approximate materials for buildings toward sustainability.


Introduction
The construction sector plays a vital role in contributing to economic growth, satisfying the needs of society, enhancing the quality of life, and providing job opportunities. However, these constructions have caused an adverse effect on the environment such as generating half of the greenhouse gas emissions, using nearly 40% of the natural resources worldwide, consuming 70% of the electrical power and 12% of potable water, and producing 45-65% of the waste placed in landfills [1][2][3]. The effect of construction on the environment becomes particularly important in developing countries, which use a large number of resources for their construction activities. Buildings, as one of the construction activities, consume a large number of natural resources, especially in the non-operational phases, such as the production of materials and the procedure for dismantling and waste disposal [4]. Especially, developing countries are increasingly speeding up the pace of urbanization in terms of the construction of infrastructure works, especially high-rise buildings [5]. Therefore, it can be said that the construction industry has a strong influence on the natural environment.
Green buildings or sustainable constructions are interpreted as using fewer natural resources and increasing the use of materials and products that have a high level of recycling or reuse [6,7]. At the same time, this practice prevents environmental degradation throughout the building's life cycle due to the manufacturing and construction processes [8]. In order to assess the sustainability of a building, it is important to identify many different factors and to examine them in a specific and transparent way. In addition, these factors are often uncertain issues, requiring an approach to provide technical Table 1. Some rating systems for green buildings across the world.

Countries
Green Building Rating Systems Sources US Leadership in Energy & Environmental Design (LEED) [29][30][31] UK Building Research Establishment Environmental Assessment Method (BREEAM) [32,33] Australia Green Star [34,35] Singapore Green Mark [36] Vietnam Lotus [26,37] Numerous studies on the optimization of building materials take into account environmental objectives. Zang et al. [38] presented an optimal method under the cap-and-trade (C&T) mechanism Sustainability 2020, 12, 2769 3 of 19 for transporting and storing construction materials to reduce carbon emissions during the project implementation towards sustainable development. A case study for the design of the new low-cost residential building was used to provide a Pareto front of the multi-criteria problem [39]. This study used the mode FRONTIER optimization tool rel.4.3.0 by developed in MATLAB through the combination of various commercial materials based on the ITACA score. Moreover, the material selection problems were investigated by different optimization techniques [39][40][41].
Park et al. [42] proposed a credit optimization algorithm to minimize LEED costs. The calculation process could be applied to both ongoing and planned LEED projects. According to Castro-Lacouture et al. [40], the optimization model based on a modified LEED system for building evaluation in Columbia could obtain a detailed purchase plan about the materials that should be used and their extent of use under constraints on design and budget. The main purpose of these sustainability rating systems is to avoid the depletion of natural resources, water, and raw materials, and to contribute to preventing the degradation of the habitat on earth.
In Vietnam, one of the most frequently discussed issues to ensure sustainable development is the restriction on the use of traditional building materials affecting natural resources and the environment. Materials mainly use burning materials, so the production of these materials greatly affects the country's resources as well as the production process that also produces CO 2 into the environment. For this purpose, the Ministry of Construction of Vietnam delivered Decision 419/QD-BXD on "Plan of actions of the construction industry to implement the Green Growth Strategy" in 2017 and Decision 280/QD-TTg, issued by the Prime Minister on 13/03/2019, on "National Program on Energy Efficiency 2019-2030 period" [27]. Therefore, sustainable building materials have emerged in the marketplace as a choice for practitioners in the construction industry. A building is a complex system consisting of many different types of construction materials. Therefore, the effect of material selection on the lifecycle energy balance of a building is clearly a complex problem. Moreover, ensuring the construction project addresses environmental goals under budget constraints is a major challenge for designers and investors.
Selecting suitable materials for buildings can help decision-makers in achieving the desired environmental, economic, financial, and social benefits. However, selecting the most appropriate eco-friendly material for a particular building application, considering Lotus scores, is not an easy task and is a great challenge for the designers and building owners. According to the above studies, the literature has produced few works that deal with this optimization problem. Therefore, in order to help the building investors or designers in selecting appropriate materials in green buildings, this study developed a multi-objective optimization model to select eco-friendly building materials considering the budget, the total labor-working days, and materials-related green building scores. The building budget was determined intensively by the building owner while material-related green building scores were calculated based on the green building rating system (i.e., Lotus-based green building rating system in Vietnam). A commercial building in Danang City, Vietnam was selected as a case study to demonstrate the effectiveness of the proposed model.
The problem of selecting the most appropriate eco-friendly material for a particular building application, considering Lotus scores, is not an easy task and is a great challenge for the designers and clients. They need to consider a range of selection criteria as well as environmental requirements including proportion of material, energy consumption, low carbon emission features, reduction of cost, and total labor-working day. As a contribution of the study, the proposed model provides decision makers such as investors and designers with a guidance of eco-friendly material selection toward green buildings.

Definition of the Optimization Problem
The problem in this study aims to build an eco-friendly material selection integrated platform that will enable building owners and designers to optimize decision making through knowledge

140
A given set of X = [xi,j] includes the decisive variables that represent the proportion of material 141 ith in category jth to maximize the material score in the green building assessment systems (i.e., Lotus-142 based green building assessment system in Vietnam [27]). Let S be the set of building system (i.e., 143 wood, brick, steel, cement, flooring, glass). Based on the above-mentioned problem, the proposed 144 model should meet three main objectives, as shown below.

145
Maximize the rating score of eco-friendly materials (RS) in the building.  A given set of X = [x i,j ] includes the decisive variables that represent the proportion of material i th in category j th to maximize the material score in the green building assessment systems (i.e., Lotus-based green building assessment system in Vietnam [27]). Let S be the set of building system (i.e., wood, brick, steel, cement, flooring, glass). Based on the above-mentioned problem, the proposed model should meet three main objectives, as shown below.
Maximize the rating score of eco-friendly materials (RS) in the building.
Constraint to: Minimize the initial investment cost (IC) for the building.
Constraint to: Equations (3) and (4), and Minimize the total labor-working days (LWD) that need to complete the building.
Sustainability 2020, 12 where LB: Limited material cost ($) LP: Total limited labor-working days c i : Unit price of i th material t i t i : Standard hours to finish i th material d j : Weight of j th category of material x ij : Rate of i th material in j th category z j : The Lotus scores obtained in j th category of the material. N(j): the maximum number of categories in the system j. a%: The percentage corresponding to each criterion of the sustainable materials in the Lotus-based system. l nj : The minimum rate of j th material based on the requirement of the design. u nj : The maximum rate of j th material based on the requirement of the design.
(Minimum and Maximum rate of j th material are determined based on green building standards in Vietnam or requirements of building owners).
r i : The type of sustainable material. According to the Lotus-based system [27], sustainable materials are divided into three categories such as (1) fast-growing renewable materials (grown and harvested within 10 years); (2) reusable materials or materials with at least 10% pre-consumer recycled components; or (3) 5% post-consumer recycled components, or materials with Volatile Organic Compounds (VOC) content <30 mg/L; where r i = 1 is a sustainable material and r i = 0 is an unsustainable material.
The problem in this study is to select the proportions of materials in buildings that can achieve the maximum of material-related green building scores within the budget constraint and the total labor-working days. The general objective function simultaneously accounts for the minimum value of material cost and total labor-working days to reach the desired Lotus scores as below In summary, the resulting model is comprised of the general objective function (9) subject to constraints (2)-(4), (6), and (8). To reveal, the existing trade-off is between building material cost, total labor-working days, and Lotus scores.

Generalized Reduced Gradient-Based Nonlinear Optimization Model
The Generalized Reduced Gradient (GRG) method proposed by Lasdon et al. (1974) [43] is one of the most popular methods to solve problems of nonlinear optimization. The GRG is an extension of the reduced gradient method to accommodate nonlinear inequality constraints. The main idea of this method is to solve the nonlinear problem dealing with active inequalities. The variables are separated into a set of basic (dependent) variables and non-basic (independent) variables. Then, the reduced gradient is computed in order to find the minimum in the search direction. This process is repeated until the convergence is obtained [44].
The nonlinear program to be solved is assumed to have the form (10) Minimize. f (x) (10) Constraint to g i (X) = 0, i = 1, . . . , m where X is n-vector and u i , l i are given lower and upper bounds u i > l i . The fundamental idea of GRG is to use the Equation (11) to express m of the variables called basic variables, in terms of the remaining n-m non-basic variables. This is also the way the Simplex method of linear programming operates. This study use the Generalized Reduced Gradient (GRG) method in the standard Microsoft Excel Solver [45]. This tool has been proven in use over many years as one of the most robust and reliable approaches to solving difficult NLP problems. Figure 2 presents the flowchart of the proposed model that is based on the GRG nonlinear optimization algorithm [43]. The GRG method is a generalization of the reduced gradient method that can deal with nonlinear constraints and arbitrary bounds on the variables.  The proposed model was implemented in the add-in function in Microsoft Excel [45] that is shown in Figure 2. The input data can be classified into five specific groups. The first group is the requirements for sustainable materials that include technical standards of materials, and unit prices of materials. The second group is the design parameters of the building that consist of the number of eco-friendly materials and the requirements of minimum and maximum material usage. The third group is the parameters related to the investor that are the estimated budget and expected construction time. The fourth group is the parameters related to the labor working days that requires for each type of materials. The last group is the score parameter related to the Lotus green building assessment system. Figure 3 shows experimental setting of the optimization model.  In Vietnam specifically, the Vietnam Green Building Council (VGBC) has developed the Lotus-213 based certification system with the ability to be applied to all types of construction projects [27]. The 214 Lotus-based system is used to assess buildings regarding sustainability scores, indoor environment In line with this development trend, the demand for construction materials will increase by about 217 Figure 3. A snapshot of implementation of the proposed model.

Lotus-Based Green Building Rating System
In Vietnam specifically, the Vietnam Green Building Council (VGBC) has developed the Lotus-based certification system with the ability to be applied to all types of construction projects [27]. The Lotusbased system is used to assess buildings regarding sustainability scores, indoor environment quality, and materials and resource usage. Green buildings are concerned by building stakeholders. In line with this development trend, the demand for construction materials will increase by about 10% per year, according to the Vietnam Association of Building Materials. By offering a common goal of saving construction materials, Lotus's material portfolio encourages solutions are beneficial to buildings because they maximize the reuse and recycling materials. Figure 4 shows the Lotus certified projects increase over the years. The Lotus-based system has built up a strong reputation in green building certification systems in Vietnam. The Lotus SB (Small Buildings) assesses new constructions and major refurbishments of non-residential buildings with a GFA of less than 2500 sqm. Figure 5 presents GUI of Lotus SB V1 in the certification system. and major refurbishments of non-residential buildings with a GFA of less than 2500 sqm. Figure 5 223 presents GUI of Lotus SB V1 in the certification system.   According to the Lotus certification system, selections of eco-friendly materials encourages the use of local materials. The Lotus system also encourages the use of materials with a high level of recycling, a fast regeneration cycle, and low emissions to reduce their impact on the environment and on the indoor air quality of the building. Lotus's material portfolio includes three main goals: Minimizing the use of materials produced from new sources of extraction; encouraging the use of sustainable and low energy materials throughout the life cycle; and reducing construction waste. During the evaluation process, the Lotus system gives a high score for items that use fast renewable materials (planted and harvested within 10 years), reused materials, or materials with at least 10% pre-consumer recycled or 5% post-consumer recycled components. This system also recommends the use of materials with low VOC content to ensure the health of the occupants of the building.

Model Settings
The Danang International Airport Customs Headquarters was used as a case study to assess the applicability of the proposed model. Figure 6 presents the 3-D view of the experimental building that consists of five stories and a total floor area of 2450 m 2 . The main building structures are reinforced concrete frame, brick walls, and reinforced concrete roofs.  The Danang International Airport Customs Headquarters was used as a case study to assess the 240 applicability of the proposed model. Figure 6

256
Considering the above-mentioned facts, the selected actual building case study, named

257
International Airport Customs Headquarters in Danang city, Vietnam ( Table 2). The total estimated Considering the above-mentioned facts, the selected actual building case study, named International Airport Customs Headquarters in Danang city, Vietnam ( Table 2). The total estimated budget for the materials was $6,000,000, the total labor-working days was expected to result in 9000 days, and nine materials were considered including: Cement, steel, bricks, ceramic brick, glass, wood, floors, tiles, and paint. Table 3 presents the information on the materials used in the construction, including the material types, the quantity of the materials (d j ), the unit price of the materials (c i ), the standard days (t i ), and the sustainable material types (r i ). This information was derived from the construction company organization.  This building can reach a score of 13 in the green building rating system as its materials satisfy the requirements of the Lotus system. Table 4 shows information related to green building assessment in the Lotus system. The first category (M1) related to the structure materials. The building is evaluated with scores of 1, 2, and 3 if percentages of the sustainable structure materials are 40%, 60%, and 80% of the total material cost of the buildings, respectively. The second category (M2) refers to non-structural walls. The building achieves scores of 1, 2, and 3 as sustainable materials of the non-structural walls are 40%, 60%, and 80% out of the total material cost of the building, respectively. For the third category (M3) of window and door materials, if 40% of the windows and doors are made up of sustainable materials, 1 point is awarded, and if 80%, then 2 points are awarded. For the category of flooring (M4) and roofing materials (M5), the rating point system is as per the window and door materials. The last category corresponds to paint and overlay. If 50% of the material has a low volatile organic compound (VOC) level, the system gives 1 score.

Analytical Results and Discussion
The proposed model seeks to maximize the number of material-related green building scores by fulfilling the Lotus requirements. Table 5 summarizes the optimization results that indicate the optimal percentage of each material type (i.e., x ij column) for each building component. Table 5 shows that the building can reach a score of 13 in which the percentage of material of the Cm 1 , Cm 2 , Cm 3 were 35%, 41%, and 24%, respectively. In order to provide decision-makers with more information in their material selections, a set of score scenarios was presented in Table 6 in which the Lotus score varied from 7 to 12 and total labor-working days and building cost were changed accordingly. Table 6 shows the optimal solution with the best materials, as well as the extent of their use on a case-by-case basis. It is easy to realize that when the constraint conditions are adjusted, the main structural materials of the building, such as steel, cement, and bricks, have the biggest changes in the selection process of materials, as they are materials that account for a large proportion of construction work in Vietnam. These materials greatly affect the cost as well as the construction time; therefore, changing only one Lotus score for these materials also significantly affects the construction cost and project completion time.
In contrast, in the non-structural wall category, the changing of types of sustainable bricks has the most impact on the Lotus score in comparison with all other remaining materials. Bricks are the first material to be changed when the Lotus score fluctuates, because until they meet the sustainability criteria, the proportions of materials in the other categories change to ensure the achievement of the targeted Lotus score. In contrast, in the non-structural wall category, the changing of types of sustainable bricks has the most impact on the Lotus score in comparison to all remaining materials.  The correlation among the Lotus score and the project cost and the total labor-working days is shown in Figure 7. Points A-G illustrate the opportunity to obtain a Lotus score with the lowest budget and the lowest number of the total labor-working days. In addition, Figure 7 reveals that in order to achieve a higher Lotus score, more budget should be used, and the total labor-working days must be larger. However, in order to examine the correlation among them, each specific pair should be analyzed; these results are presented in Figures 8 and 9.

307
The correlation among the Lotus score and the project cost and the total labor-working days is 308 shown in Figure 7. Points A-G illustrate the opportunity to obtain a Lotus score with the lowest 309 budget and the lowest number of the total labor-working days. In addition, Figure 7 reveals that in 310 order to achieve a higher Lotus score, more budget should be used, and the total labor-working days 311 must be larger. However, in order to examine the correlation among them, each specific pair should be analyzed; these results are presented in Figures 8 and 9.  with the amount of money at $252,708; $257,537, and $263,229, respectively. Obviously, to obtain a 334 higher score with an increase of 1-2 points, the cost increases from $5,000 to $6,000. This is a relatively 335 large sum just for the cost of materials.

336
In Figure 9, a tendency that can be seen is that to achieve a higher Lotus score, greater total labor-337 working days must be employed. Based on the change of total labor-working days for each Lotus 338 point, the selection of sustainable materials also affects the construction strategy, the construction 339 methods, and the construction technology. For instance, by using sustainable materials and applying 340 new construction technology, employees may be unfamiliar with the operation and construction 341 process. Therefore, labor productivity may significantly reduce.

342
As shown in Figure 10, the change of Lotus score is recorded according to the corresponding 343 costs and the total labor-working days. The results show that sustainable materials are more 344 expensive and the labor workforce for construction must be more. This allows investors to consider 345 and select materials from suppliers that supply the same category of materials to achieve their goals 346 under budget constraints, along with the total labor required.

347
Lotus score Total labor-working days . Correlation between Lotus score and total labor-working days.

329
As shown in Figure 8, the project can achieve Lotus scores of 7, 8, and 9 with a budget ranging 330 from $5,513,011 to $5,516,463; to obtain a score of 10, a budget at $5,520,122 is required, with a 331 difference of $3,000-$4,000 to achieve a higher score. However, to increase from a score of 10 to scores 332 of 11, 12, and the highest score of 13, the budget must be increased by $10,000 for each Lotus score,

333
with the amount of money at $252,708; $257,537, and $263,229, respectively. Obviously, to obtain a 334 higher score with an increase of 1-2 points, the cost increases from $5,000 to $6,000. This is a relatively 335 large sum just for the cost of materials.

336
In Figure 9, a tendency that can be seen is that to achieve a higher Lotus score, greater total labor-337 working days must be employed. Based on the change of total labor-working days for each Lotus 338 point, the selection of sustainable materials also affects the construction strategy, the construction 339 methods, and the construction technology. For instance, by using sustainable materials and applying 340 new construction technology, employees may be unfamiliar with the operation and construction 341 process. Therefore, labor productivity may significantly reduce.

342
As shown in Figure 10, the change of Lotus score is recorded according to the corresponding 343 costs and the total labor-working days. The results show that sustainable materials are more 344 expensive and the labor workforce for construction must be more. This allows investors to consider 345 and select materials from suppliers that supply the same category of materials to achieve their goals 346 under budget constraints, along with the total labor required.

347
Lotus score Total labor-working days Figure 9. Correlation between Lotus score and total labor-working days.
As shown in Figure 8, the project can achieve Lotus scores of 7, 8, and 9 with a budget ranging from $5,513,011 to $5,516,463; to obtain a score of 10, a budget at $5,520,122 is required, with a difference of $3000-$4000 to achieve a higher score. However, to increase from a score of 10 to scores of 11, 12, and the highest score of 13, the budget must be increased by $10,000 for each Lotus score, with the amount of money at $252,708; $257,537, and $263,229, respectively. Obviously, to obtain a higher score with an increase of 1-2 points, the cost increases from $5000 to $6000. This is a relatively large sum just for the cost of materials.
In Figure 9, a tendency that can be seen is that to achieve a higher Lotus score, greater total labor-working days must be employed. Based on the change of total labor-working days for each Lotus point, the selection of sustainable materials also affects the construction strategy, the construction methods, and the construction technology. For instance, by using sustainable materials and applying new construction technology, employees may be unfamiliar with the operation and construction process. Therefore, labor productivity may significantly reduce.
As shown in Figure 10, the change of Lotus score is recorded according to the corresponding costs and the total labor-working days. The results show that sustainable materials are more expensive and the labor workforce for construction must be more. This allows investors to consider and select materials from suppliers that supply the same category of materials to achieve their goals under budget constraints, along with the total labor required. In any process of choosing sustainable materials for building construction, the decision is not easy, and it is often difficult to choose the best construction strategy. This is further complicated by the simultaneous consideration of project budgets, total labor-working days, and green building 353 certification. In order to address this problem, the model in the study proposed an optimal model to will provide decision-makers with blueprints to improve the green performance of construction 359 projects.

360
This study aims to build an eco-friendly material selection integrated platform that will enable 361 building owners and designers to optimize decision making through knowledge of materials type,

Conclusions
In any process of choosing sustainable materials for building construction, the decision is not easy, and it is often difficult to choose the best construction strategy. This is further complicated by the simultaneous consideration of project budgets, total labor-working days, and green building certification. In order to address this problem, the model in the study proposed an optimal model to select the rate of eco-friendly materials that achieved a minimum initial investment cost and total labor-working days, which is different from traditional design, and could maximize the score of construction project via Lotus-based system at the conceptual stage of building projects. Optimization results facilitates decision-makers to obtain a detailed material procurement plan. This will provide decision-makers with blueprints to improve the green performance of construction projects.
This study aims to build an eco-friendly material selection integrated platform that will enable building owners and designers to optimize decision making through knowledge of materials type, construction strategy, and Lotus-based green building certification system in Vietnam. The results of an experimental building in Danang city, Vietnam confirmed the importance of the materials that form the main structures of the building including bricks, cement, and steel. They are materials that account for a large proportion of the construction works in Vietnam. Therefore, an adjustment of the proportion of these materials is greatly affecting the Lotus green building scores, project cost, and completion time of the project. Hence, this study suggests that if stakeholders in the project aim to adjust the sustainable goals for the building, they should not adjust the materials in this category; instead, they should adjust the materials of non-structural walls, particularly considering different types of eco-friendly bricks. This suggestion is consistent with the construction works in Danang city because bricks are still the main materials used in dividing spaces in buildings.
At the same time, with increasing awareness of sustainable construction in Vietnam, eco-friendly bricks (e.g., adobe bricks and foam concrete bricks) are being produced at lower and lower costs. This will help stakeholders to easily change the proportion of bricks in the building when they want to change sustainability goals via Lotus score, while the cost and time of construction will not be remarkably affected. The research results also show the importance of the specifications of sustainable materials on the market, because without these parameters, it will be a great challenge to evaluate sustainable buildings based on the Lotus standards. Therefore, state management agencies need to adopt specific regulations to ensure that manufacturers provide data or information about the composition of materials to help investors, private units, consultants, and construction units to choose materials that ensure the sustainability of buildings. In summary, the proposed model based on multi-objective optimization approach provides a guidance for decision-makers in selecting approximate materials for buildings toward sustainability.
The challenge for increasing the use of eco-friendly construction materials in developing countries is that each material is managed individually by different manufacturers, making it difficult for building owners, designers, and construction companies to access detailed information about the types, prices, standards, and environmental impacts of products certified as green building materials. Future work should include the advanced model based on a Web-based platform that could collate data easily to modify the established optimization model. Moreover, if a client can easily employ a program by utilizing only the limited information available during initial construction planning, accessibility can also be improved for users aiming to achieve green building certification.