PRES22_0231.docx DOI: 10.3303/CET2294163 Paper Received: 15 May 2022; Revised: 23 June 2022; Accepted: 28 June 2022 Please cite this article as: Tigue A.A.S., Promentilla M.A.B., 2022, A TOPSIS and AHP with Spherical Fuzzy Approach for Optimal Selection of Pervious Geopolymer Mix for Heavy Metal Removal, Chemical Engineering Transactions, 94, 979-984 DOI:10.3303/CET2294163 CHEMICAL ENGINEERING TRANSACTIONS VOL. 94, 2022 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Yee Van Fan, Jiří J. Klemeš, Sandro Nižetić Copyright © 2022, AIDIC Servizi S.r.l. ISBN 978-88-95608-93-8; ISSN 2283-9216 A TOPSIS and AHP with Spherical Fuzzy Approach for Optimal Selection of Pervious Geopolymer Mix for Heavy Metal Removal April Anne S. Tiguea, Michael Angelo B. Promentillab,* aChemical Engineering Department, College of Engineering, De La Salle University, Manila 1004, Philippines bCenter for Engineering and Sustainable Development Research, De La Salle University, Manila 1004, Philippines michael.promentilla@dlsu.edu.ph Geopolymer is an emerging material that is known to have excellent properties. Numerous applications of geopolymer have been reported in various studies, for example, in the construction and building sector, nuclear waste management, and wastewater treatment industry. It gains popularity in the wastewater treatment sector as it can effectively adsorb pollutants (e.g., heavy metals). However, the performance of this material varies depending on different factors such as the type of precursors, mix formulation, process conditions, etc. Moreover, optimal selection should also consider the potential environmental impact and safety to the general public before field application. The utilization of coal fly ash and other potential material was investigated to develop a pervious geopolymer. In this study, selection of optimized geopolymer mix for heavy metal removal using TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) and AHP (Analytic Hierarchy Process) with a spherical fuzzy approach was employed. An illustrative case study is presented for the proposed decision modeling technique in optimally selecting pervious geopolymer mix. 1. Introduction In recent years, the demand for conventional concrete has increased due to rapid industrialization. Concrete is composed of cement, coarse aggregates, fine aggregates, and water. With the surge in the demand for concrete, an equivalent threat to the environment awaits. The construction industry, being one of the consumers of concrete, is compelled to look for an alternative material that would help mitigate the use of cement as a primary material. An interesting new and eco-friendly found material, geopolymer, has been gaining attention due to its excellent properties. This material has often been looked at as an alternative to cement. A study conducted by Dollente et al. (2021) showed that the geopolymer concrete has a global warming potential of 32 % less than the traditional concrete made of cement. This indicates that the use of this material in the construction industry and other sectors such as wastewater treatment for heavy metal removal is promising. Pervious geopolymer is made from a precursor, alkali activator, and coarse aggregates. The potential precursor for pervious geopolymer can be a waste material composed of aluminosilicate. The type of precursors, mix formulation, and process conditions may greatly affect the properties of the pervious geopolymer. Moreover, safety and risk to the environment before field application must also be taken into consideration. Hence, it is imperative to determine the optimal mix proportion and other factors that may affect when the product is deployed. To thoroughly evaluate the pervious geopolymer considering the qualitative and quantitative factors such as properties, safety to the public, and risk to the environment, this study aims to employ TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) and AHP (Analytic Hierarchy Process) with spherical fuzzy technique. Kutlu Gündoğdu and Kahraman (2019a) have studied both approaches with spherical fuzzy technique independently. Meanwhile, the modified approach presented in this study strengthens the results of the decision-makers in selecting the optimal mix proportion in developing a pervious geopolymer as it addresses the possible uncertainties in making selections by combining the two approaches- TOPSIS and AHP with spherical fuzzy technique extension. 979 2. Materials and method The figure below illustrates the overview of the study. It is divided into four parts- Development of pervious geopolymer, evaluation of criteria, calculation of weighted matrix, and selection of pervious geopolymer mix. Figure 1: Overview of the methodology The experimentation was done using various mixes of precursor and activator to develop geopolymer. After gathering the data, an assessment of the optimal mix design was done by employing techniques of AHP and TOPSIS extended spherical fuzzy. The spherical fuzzy set integration accounts for the fuzziness and ambiguity in judgments made in evaluating the linguistic scale in AHP and in rating the alternatives using TOPSIS (Kutlu Gündoğdu and Kahraman, 2019a). The criteria weights for the linguistic scale were calculated using AHP pairwise comparison while the ranking of alternatives was based on TOPSIS. 2.1 Materials Coal fly ash (CFA) was collected from a coal-fired power plant in the central region of Luzon, Philippines, and was used as received. Coarse aggregates (CA) used were gravel and were purchased from Rizal, Luzon. Chemical reagents such as sodium hydroxide micro-pearls with 99 % purity, sodium silicate with a composition 44 % SiO2, and nickel sulfate hexahydrate with 98 % NiSO4.6H2O were used in the study. 2.2 Development of pervious geopolymer Pervious geopolymer concrete samples were prepared in accordance with the following formulations as shown in Table 1. Alkali activator (AA) is composed of NaSiO3 / NaOH at a ratio of 2.5:1 (Tho-In et al., 2012). The response variables are compressive strength and heavy metal removal. Table 1: Pervious geopolymer mix proportion Mix Design Code CFA:CA ratio AA:CFA ratio NaOH Concentration (M) Mix No. 1 1:6 0.65:1 15 M Mix No. 2 1:9 0.65:1 15 M Mix No. 3 1:9 0.65:1 10 M Mix No. 4 1:6 0.45:1 10 M Mix No. 5 1:9 0.45:1 10 M Initially, CFA and AA were mixed in a mechanical mixer for 5 minutes. Then, CA was added and mixed for another 5 mins. The mixture was then poured into cylinder molds and allowed to rest for 30 mins. The molded samples were wrapped in a zip bag and were cured for 24 h at 80 °C in the oven. After oven curing, the samples were demolded and cured for 28 d at ambient environment. The compressive strength of pervious geopolymer was tested using Universal Testing Machine. For heavy metal removal efficiency, the concentration of nickel in the solution before and after passing through the pervious geopolymer was measured using Atomic Absorption Spectroscopy. 2.3 Evaluation of criteria based on AHP and spherical fuzzy approach The diagram shown in Figure 2 illustrates the criteria considered such as toxicity (C1), life cycle analysis (LCA) (C2), compressive strength (C3), and removal efficiency (C4). The AHP integrated with spherical fuzzy set was used to evaluate the weights of each criterion. The details and the description of the criteria and the alternative considered in the study are shown in Table 2. Development of pervious geopolymer Evaluation of criteria (AHP extended to spherical fuzzy) Calculation of weighted matrix (TOPSIS extended to spherical fuzzy) Selection of pervious geopolymer 980 Figure 2: Criteria used in selecting the optimal mix formulation Table 2: Description of criteria and alternatives Cluster Element Description Goal Goal To evaluate and select the optimal mix formulation of pervious geopolymer Criteria Safety – Toxicity* The chance of harmful effects on human during preparation and deployment and is measured qualitatively Environment – LCA* The potential impact on the environment in terms of the material used and related logistics. It is estimated to be measured qualitatively Properties – Compressive Strength Obtained from the experimental result after 28 d of curing Properties – Removal Efficiency Obtained from the experimental result after passing through the column setup Alternatives Mix Design 1 In reference to Table 1 formulation Mix Design 2 In reference to Table 1 formulation Mix Design 3 In reference to Table 1 formulation Mix Design 4 In reference to Table 1 formulation Mix Design 5 In reference to Table 1 formulation The proposed AHP extended to spherical fuzzy starts by populating the matrix with the value judgment on linguistic scale to describe the relative importance of one criterion over the other. A pairwise comparison was performed as shown in Table 3. Each scale has an equivalent spherical fuzzy number that was used to calculate the weighted and normalized weight of criteria (Kutlu Gündoğdu and Kahraman, 2019b). Table 3: Linguistic scale with spherical fuzzy set (AHP integrated) Scale Code Spherical Fuzzy Number [ μ, ν, π ] Score index Very Strongly (More Important) VSM [0.9, 0.1, 0.1] 8 Strongly/Highly (More Important) STM [0.8, 0.2, 0.2] 5 Moderately (More Important) MM [0.7, 0.3, 0.3] 3 Slightly (More Important) SM [0.6, 0.4, 0.4] 2 About Equal AE [0.5, 0.5, 0.5] 1 Slightly (Less Important) SL [0.4, 0.6, 0.4] 1/2 Moderately (Less Important) ML [0.3, 0.7, 0.3] 1/3 Strongly/Highly (Less Important) STL [0.2, 0.8, 0.2] 1/5 Very Strongly (Less Important) VSL [0.1, 0.9, 0.1] 1/8 2.4 Calculation of weighted normalized decision matrix based on TOPSIS and spherical fuzzy technique On ranking the alternatives using TOPSIS, populating the decision matrix with alternatives and criteria with scores is the first step. The scores can be either quantitative or qualitative assessments. In this study, C3 and C4 (quantitative) values were obtained using the experiment and while C1 and C2 (qualitative) were evaluated using the linguistic scale shown in Table 4. 981 Table 4: Linguistic scale with spherical fuzzy set (TOPSIS integrated) Scale Code Spherical Fuzzy Number [ μ, ν, π ] Rating Ideal Best/Perfect IB [1.0, 0.0, 0.0] 1.00 Excellent EX [0.9, 0.1, 0.1] 0.73 Very good VG [0.8, 0.2, 0.25] 0.48 Good GD [0.7, 0.3, 0.35] 0.32 Slightly good/Above satisfactory AS [0.6, 0.4, 0.4] 0.22 Moderate/Satisfactory S [0.5, 0.5, 0.5] 0.18 Slightly bad/Below Satisfactory BS [0.4, 0.6, 0.4] 0.10 Bad BD [0.3, 0.7, 0.35] 0.06 Very bad VB [0.2, 0.8, 0.25] 0.03 Worst WO [0.1, 0.9, 0.1] 0.01 Ideal Worst IW [0.0, 1.0, 0.0] 0.00 The judgments made on qualitative data were transformed into a numerical score. The weighted normalized decision matrix was then evaluated based on TOPSIS and spherical fuzzy numbers using Eq(1) and Eq(2). 𝑋𝑋�𝑖𝑖𝑖𝑖 = 𝑋𝑋𝑖𝑖𝑖𝑖 �∑ 𝑋𝑋𝑖𝑖𝑖𝑖 2𝑛𝑛 𝑖𝑖=1 (1) 𝑉𝑉𝑖𝑖𝑖𝑖 = 𝑋𝑋�𝑖𝑖𝑖𝑖 × 𝑊𝑊𝑖𝑖 (2) Then, the positive ideal and negative ideal solutions are identified based on the criterion- if benefit or cost type. For the cost type, the lower the score, the better the alternative concerning that criterion. Meanwhile, for the benefit type, the higher the score, the better the alternative concerning that criterion. A measure of the separation via Euclidian distance from the positive ideal and negative ideal solution is calculated using Eq(3) and Eq(4). 𝑆𝑆𝑖𝑖 + = ��(𝑉𝑉𝑖𝑖𝑖𝑖 𝑛𝑛 𝑖𝑖=1 − 𝑉𝑉𝑖𝑖 +) 2� 0.5 (3) 𝑆𝑆𝑖𝑖 − = ��(𝑉𝑉𝑖𝑖𝑖𝑖 𝑛𝑛 𝑖𝑖=1 − 𝑉𝑉𝑖𝑖 −) 2� 0.5 (4) Lastly, the performance score which was based on the relative closeness to the ideal solution was calculated using Eq(5) and the ranking of alternatives follows. 𝑃𝑃𝑖𝑖 = 𝑆𝑆𝑖𝑖 − 𝑆𝑆𝑖𝑖 + + 𝑆𝑆𝑖𝑖 − (5) 3. Results and discussion 3.1 Properties of pervious geopolymer Table 5 shows the measured compressive strength of pervious geopolymer which ranges from 1.02 MPa – 2.09 MPa and the recorded removal efficiency which ranges from 96.0 – 98.1 %. Table 5: Pervious geopolymer mix proportion Mix Design Code Compressive Strength (MPa) (C3) Removal Efficiency (%) (C4) Mix No. 1 2.09 96.7 Mix No. 2 1.13 96.8 Mix No. 3 1.02 96.0 Mix No. 4 1.25 98.1 Mix No. 5 1.20 98.1 The removal efficiency was observed to be high for all runs. The possible mechanism of nickel removal is due to precipitation. At higher pH, nickel can precipitate as nickel hydroxide (Ni(OH)2). Considering the environment of pervious geopolymer which is known to be alkaline, this phenomenon may have occurred. Moreover, a study 982 by Escudero et al. (2017) showed that nickel precipitates completely at a pH of 11. This further supports the results of this study. On the other hand, the compressive strength has been observed to be low for all samples. The size of the coarse aggregate may have been a factor that can be considered in future works. These two properties of pervious geopolymer were used in the succeeding analysis for the optimal selection of mix formulation using TOPSIS and AHP integrated with spherical fuzzy numbers. 3.2 Criteria weights based on AHP and spherical fuzzy approach Analytic Hierarchy Process is a multi-criteria decision analysis tool introduced by Saaty (1987) wherein both quantitative and qualitative factors are considered in modeling the complexity of the decision problem hierarchically (goal, criteria, and alternatives) to derive weights or priorities using pairwise comparison. Another approach to integrating with this tool is the use of a spherical fuzzy set number to quantify the qualitative data. Table 6 shows the pairwise comparison of each criterion using the linguistic scale with numerical value as tabulated in Table 3. The transformed data with the equivalent fuzzy set and the synthesized criterion with the derived weight is shown in Table 7. Table 6: Pairwise comparison matrix for criteria Toxicity (C1) LCA (C2) Compressive Strength (C3) Removal Efficieny (C4) Toxicity (C1) AE MM STM MM LCA (C2) AE AE AE Compressive Strength (C3) AE AE Removal Efficiency (C4) AE Table 7: Synthesized criteria weights μ ν π Normalized Toxicity 4.106 0.031 20.187 0.367 LCA 1.413 0.026 11.775 0.214 Compressive Strength 1.347 0.020 11.520 0.209 Removal Efficiency 1.347 0.020 11.520 0.209 3.3 Decision Matrix Weights based on TOPSIS and spherical fuzzy technique TOPSIS approach integrated with spherical set was also used to evaluate and rank the mix formulation based on the desired criteria. Table 8 shows the summary of the decision matrix. This provides us an overview of the mix formulation with the corresponding scores derived from an experimental and qualitative estimation of scores. Table 8: Summary of decision matrix Mix Design Toxicity* LCA* Compressive Strength (MPa) Removal Efficiency (% Removal) Mix No. 1 GD AS 2.09 96.7 Mix No. 2 VG VG 1.13 96.8 Mix No. 3 EX VG 1.02 96.0 Mix No. 4 S S 1.25 98.1 Mix No. 5 AS EX 1.20 98.1 Tables 9, 10, and 11 showed the results of calculation steps for each alternative. Lastly, the resulted rank is shown in Table 12. The final ranking showed that the optimal mix formulation of pervious geopolymer based on the criteria presented is mix design no. 5 which is composed of CFA / CA ratio of 1:9, AA / CFA ratio of 0.45:1, and sodium hydroxide concentration of 10M. Table 9: Transformation of qualitative data with spherical scoring function Mix Design Toxicity* LCA* Compressive Strength (MPa) Removal Efficiency (% Removal) Mix No. 1 0.32 0.22 2.09 96.7 Mix No. 2 0.48 0.48 1.13 96.8 Mix No. 3 0.73 0.48 1.02 96.0 Mix No. 4 0.18 0.18 1.25 98.1 Mix No. 5 0.22 0.73 1.20 98.1 983 Table 10: Fuzzified and normalized decision matrix Mix Design Toxicity* LCA* Compressive Strength Removal Efficiency Mix No. 1 0.33 0.21 0.67 0.45 Mix No. 2 0.49 0.46 0.36 0.45 Mix No. 3 0.75 0.46 0.33 0.44 Mix No. 4 0.19 0.17 0.40 0.45 Mix No. 5 0.22 0.70 0.39 0.45 Table 11: Weighted normalized decision matrix Mix Design Toxicity* LCA* Compressive Strength Removal Efficiency Si+ Si- Pi Weights 0.12 0.04 0.14 0.09 Mix No. 1 0.18 0.10 0.08 0.09 0.118 0.171 0.591 Mix No. 2 0.28 0.10 0.07 0.09 0.140 0.113 0.446 Mix No. 3 0.07 0.04 0.08 0.09 0.226 0.062 0.216 Mix No. 4 0.08 0.15 0.08 0.09 0.127 0.208 0.621 Mix No. 5 0.12 0.04 0.14 0.09 0.061 0.225 0.786 Table 12: Resulted rank for optimal selection Mix Design CFA/CA AA / CFA Sodium Hydroxide Concentration Rank Mix No. 1 1/6 0.65 15 M 3 Mix No. 2 1/9 0.65 15 M 4 Mix No. 3 1/9 0.65 10 M 5 Mix No. 4 1/6 0.45 10 M 2 Mix No. 5 1/9 0.45 10 M 1 4. Conclusions Pervious geopolymer developed in this study uses coal fly ash as precursor. The properties of geopolymer obtained from the experimental results are compressive strength and removal efficiency. Together with this data, qualitative factors such as toxicity and life cycle assessment have also been considered. These factors were used to select the optimal mix formulation of pervious geopolymer. Considering quantitative and qualitative data, aggregation of these data has been made possible because of the new technique that was integrated into this study - TOPSIS and AHP integrated with spherical fuzzy set. This shows that this approach is a straightforward tool that can be used in multicriteria decision-making that aims to minimize uncertainties in making selections. Of the five alternatives, the mix design that has been favored based on the set criteria was the mix design no. 5 with a composition CFA/CA ratio of 1:9, AA/CFA ratio of 0.45:1, and 10 M sodium hydroxide concentration. Acknowledgments The authors are thankful to Geopolymers and Advances Materials Engineering Research for Sustainability (G.A.M.E.R.S.) and Chemical Engineering Laboratory for their assistance while conducting the study. References Dollente I.J.R., Tan R.R., Promentilla M.A.B., 2021, Life cycle assessment of precast geopolymer products, Chemical Engineering Transactions, 88, 799-804. Escudero, G., Espinoza, E., Rao, F., 2017, Chemical precipitation of nickel species from waste water, International Research Journal of Pure and Applied Chemistry, 15(2), 1–7. Kutlu Gündoğdu F, Kahraman C., 2019a, Spherical fuzzy analytic hierarchy process (AHP) and its application to industrial robot selection, Intelligent and Fuzzy Techniques in Big Data Analytics and Decision Making, 988-996. Kutlu Gündoğdu F., Kahraman C., 2019b, Spherical fuzzy sets and spherical fuzzy TOPSIS method, Intelligent and Fuzzy Systems, 36(1), 337-352. Saaty R.W., 1987, The analytic hierarchy process- what it is and how it is used, Mathematical Modelling, 9(3- 5), 161-176. Tho-In, T., Sata, V., Chindaprasirt, P., Jaturapitakkul, C., 2012, Pervious high-calcium fly ash geopolymer concrete, Construction and Building Materials, 30, 366–371. 984 PRES22_0336.pdf A TOPSIS and AHP with Spherical Fuzzy Approach for Optimal Selection of Pervious Geopolymer Mix for Heavy Metal Removal