COMPARATIVE ANALYSIS BETWEEN AHP MOORA AND AHP-ELECTRE METHOD FOR OPTIMAL ELECTRIC AND SOLAR-POWERED SHIPYARD SITE SELECTION

ABSTRACT


INTRODUCTION
Indonesia is the world's fourth most populous country.The Central Bureau of Statistics of the Republic Indonesia (BPS) estimates that Indonesia's population will be 278.696,2thousand people in 2023, with a population growth rate of 1.13%.However, as the population expands, so does the demand for economic energy.As a result, economic growth will rely significantly on fossil energy demand, causing catastrophic changes in our climate ( SDG 7).Transportation is the world's second-greatest CO2 emitter, accounting for 25% of total CO2 emissions [1].Greenhouse gas (GHG) emissions from the marine transportation sector increased from 977 million tons in 2012 to 1,076 million tons in 2018 (a 9.6% increase), including carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O), in CO2e, from total transportation (international, domestic, and fisheries).2012 there were 962 million tons of CO2 emissions, which climbed by 9.3% to 1,056 million tons in 2018 [2].As a result, one of the transportation sector's aspirations is for the shipping industry to become zero-carbon [1].
To achieve a zero-carbon shipping industry, the International Maritime Organization (IMO) has set a goal to reduce total GHG emissions by at least 20% per year by 2030 [3].One of the projects that can be undertaken is the development of electric ships.Indonesia can use its high sun exposure to produce electrical energy using solar cell (Photovoltaic / PV) technology, which converts solar energy into electric power [4].The sunlight obtained in Indonesia may produce 5.1 kWh per day when solar cells are used [5].According to the Ministry of Industry, there will be 250 national shipyards by 2020, with a capacity of around 1,000,000 DWT/year for new constructions and 12,000,000 DWT/year for ship repairs.71% of Indonesia's various shipyards have met the standards for production facilities [6].The objective is that by producing electric ships, it will be possible to conserve fuel, minimize pollution from auxiliary engines, and save power from shaft generators [7].Given the importance of the need for electric ships and the state of the shipyard in Indonesia, the production of electric and solar-powered ships is something that should be explored.Technological advancements will help electrical technology realize the all-electric ship [8].
Establishing an electric shipyard is one of the first steps in developing electric ship technology.This requires special attention because many shipyards around the world have failed to meet the IMO's 2030 objective.As a result, a green shipyard must be built or an existing shipyard converted to create a green ship [9].The shipping industry, often known as shipyards, is a vital component of the national economy.The hope is that the electric and solar-powered shipyard sector will be able to meet the needs of the national fleet of ships and respond to the IMO challenge for 2030.
To overcome these challenges, this research tries to solve the site selection problem of electric and solar-powered shipyards.With several alternative decisions and many complex criteria, decision-makers can find it hard to make an objective decision.Therefore, we can use mathematical models to support decisionmaking to decide the optimal site.This research tries to solve the site selection problem for an electric and solar-powered shipyard using the Multi-Criteria Decision Making (MCDM) method.The MCDM is a decision-making method for a problem that selects the best decision based on certain criteria that are often conflicting.Some MCDM methods used in site selection research include AHP, TOPSIS, SAW, ELECTRE, and SIMOS Procedure.Multi-Objective Optimization based on Ratio Analysis (MOORA) is a method introduced as an objective (non-subjective) method by Brauers et [16].In this method, the existing problems are described in the form of a hierarchy, consisting of several levels starting with goals, criteria, and alternatives [17].
Numerous previous studies used the integration between MCDM methods.Mangalan  Therefore, the purpose of this research compare the integration of weighting and ranking methods to determine the optimal location of electric and solar-powered shipyards.This research will compare the calculation results of MOORA and ELECTRE, integrated with AHP, respectively, as a weighting method [26].Comparison results are used to improve the objectivity of the electric and solar-powered shipyard industry site selection.

RESEARCH METHODS
There are three alternative decisions for shipyard site selection as shown in (Figure 1).Alternatives 1 and 3 are in Paciran District, Lamongan Regency, East Java Province, while alternative 2 is in Serang Regency, Banten Province.Alternative location 1 has an area of 38 Ha and is located in Sidokelar Village, Paciran District.Currently, alternative location 1 is a shipyard owned by PT Dock Pantai Lamongan (DPL), a shipping company whose major business is handling large ships for maintenance and repair.Site 2 is located in Bojonegara, Serang, Banten.Currently, alternative location 2 is a shipyard built in 2021 by PT Armada Bangun Samudera (ABS) in the PT Gandasari Energi region.Alternative site 3 is in the Paciran District of the Lamongan Regency of East Java Province.Alternative location 3 is a shipyard of PT Lintech Seaside Facility (LSF), a PT Lintech Duta Pratama subsidiary.The framework stage used for electric and solar-powered shipyard site selection is shown in (Figure 2).First, we determine the parameters for making industrial site selection decisions.Then we collect the data of each parameter for the alternative decisions.Data collection is performed using a field survey for primary data and a desk study for secondary data.After the data is complete, we analyze the data to obtain a decision matrix.Next, the weight for each parameter is calculated using the AHP method.The ranking of decision alternatives is performed using ELECTRE and MOORA.Finally, we compare and analyze the results of these two methods.

Industrial Site Parameter Establishment
The assessment of relevant criteria is essential to determine optimal site selection.Moreover, the solar and electric shipyard industries must minimize environmental and social damage at the lowest feasible cost with optimal site location.In this study, twenty-one (21) parameters are used to determine the location of the electric and solar-powered shipyard industry based on regional characteristics and data availability, while also satisfying a set of constraints and requirements: Industry Law No. 3 of 2014, Ministerial Regulation No. 13 of 2017 concerning the National Spatial Plan, and Minister of Industry Regulation No. 30 of 2020 concerning Technical Criteria for Industrial Designation Areas (KPI).Parameters to determine the location of the electric and solar-powered shipyard industry include slope parameters, soil type parameters, rainfall parameters, flood-prone potential parameters, landslide-prone potential parameters, sustainable food agricultural land designation (LP2B) parameters, land area parameters, land transportation availability parameter, railway station availability parameter, toll road availability parameter, port availability parameter, airport availability parameter, regional dock availability parameter, raw water source availability parameter, availability of sewage disposal plants, availability of regional transmission lines, parameter water depth conditions, parameter water wave conditions, parameter availability of maritime universities, parameter distance of settlements from industry, and parameter suitability of regional superior industries (Table 1).(Figure 3) depicts the conditions for identifying the location of the electric and solar-powered shipyard industries.

Desk Study and Data Gathering
The desk study aims to obtain various kinds of references related to the selection of the location of the electric and solar-powered shipyard industry.To provide an overview and verify the data from the desk study, a field survey was then conducted in the alternative location area of the electric and solar-powered shipyard industry.

Data Analytics
Data analysis aims to determine the best location mathematically by using the AHP MOORA Integration method and the AHP ELECTRE method.Parameter weights were first calculated using the Analytic Hierarchy Process (AHP) [19], and then the MOORA and ELECTRE methods were used to rank potential locations for the electric shipbuilding and solar shipbuilding industries, calculating the indexes of suitability and incompatibility between pairs of alternatives.Weighting using the AHP approach can provide the optimal value of each index, calculation formula, and statistical data, as well as a quantitative transformation method for processing qualitative indices to achieve a scientific and fair evaluation of the indices [27].After the weighting is completed, the MOORA method and the ELECTRE method are implemented.ELECTRE is a collection of decision-making approaches that present sequential and superior alternatives to make trustworthy decisions [28].Weighing using AHP weights [19], [29] and ranking using MOORA and ELECTRE methods are part of the AHP MOORA integration method and AHP ELECTRE integration method.

Weighting in the Analytic Hierarchy Process (AHP)
• Step 1. Compare each criterion in pairs to establish the criteria weight.The precedence scheme outlined in  Both elements are equally essential Two factors have the same impact on achieving an objective.

3.
One component is marginally more significant than the other Experience and discernment provide marginally more support than other factors.

5.
One component is more essential than the other.
Experience and discernment strongly favor one element over another.

7.
One element is more essential than all others.
One of the most powerful elements is maintained and predominates in practice.

9.
One of the most essential components of the other components.
With the maximum degree of dependability, the evidence favoring a task relative to another.

2,4,6,8 Consideration values between two adjacent values
This value is assigned when two compromises exist between two options.

Inverse
If activity "i" receives one more point than activity "j," then "j" has the opposite value of "i."Data source: Adapted from "How to make a decision: The Analytical Hierarchy Process" by T.L. Saaty (1) • Step 3. Computes the weight of synthesis by adding each column within the same row of the comparison normalization result matrix.
• Step 4. Calculates the eigenvalues by multiplying each matched matrix column in the same row by a predetermined criterion number.
• Step 6. Determines the significance of every factor by dividing synthesis weight by priority weight.
• Step 7. Determine the highest Eigenvalue () by dividing the number of criteria by the total number of importance values.
• Step 8. Determines the consistency of use to ensure that discernment for decision-making is highly consistent.

Multi-Objective Optimization based on Ratio Analysis (MOORA) Method
Brauers was the first to introduce the MOORA technique [11].This approach is a multi-objective optimization technique that can be utilized to effectively resolve several challenging decision-making issues.
The key benefits of the MOORA approach are simplicity and ease of implementation, low mathematical calculations, and relatively short execution time [32].
The MOORA technique seeks to maximize two or more competing objectives while satisfying a set of constraints by ranking several possible solutions.The MOORA method entails five main steps, including establishing objectives to identify evaluation attributes, creating a decision matrix, normalizing the decision matrix, deducting the desired maximum and minimum values, and ranking [11].
• Step 1. Construct the decision matrix that shows the performance of different alternatives against various criteria, like all multi-criteria decision-making approaches.
Where   is the performance measure of  ℎ alternative on  ℎ criterion,  is the number of alternatives ,and  the number of criteria.
• Step 2. Decision Matrix Normalization: Normalize the decision matrix using where   is the value of alternative  ℎ from the  criterion and is always in the range [1,0].
• Step 3. Construct the weighted decision matrix after determining the criterion weights.In this work, the criterion weights are determined using the AHP approach.Latest knowing the criterion weight, we multiply the weight of each criterion with the normal decision matrix and form the weighted normal decision matrix.
• Step 4. Determine the assessment values by finding the difference between the sum of beneficial and non-beneficial (cost) criteria as given in Equation ( 8).
• Step 5. Rank the assessment values in decreasing order to get the global rank of the alternatives.

• Step 1. Constructing the decision matrix [13], [23],
The Determination Matrix is a matrix that displays the factor values for every alternative.Each column of the Decision Matrix represents the value of all factors associated with a particular alternative.While each column of the Decision Matrix displays the value of each alternative for a given factor, each row displays the value of each alternative for that factor.The Decision Maker establishes the Decision Matrix's values.
is the normalized matrix of the problem's base matrix, with  = 1,2,3, … ,  and  = 1,2, … , .The base matrix to be normalized is   .Each  represents a row of the matrix, whereas each  represents a column.
• Step 3. Weighting the normalized matrix, When the  matrix has been normalized, each column is multiplied by the weight (  ) resulting from the AHP weighting.Thus, the Weight-normalized matrix is  =  × , which is represented as follows: • Step 4 Determine the set of concordance and discordance on the index, by estimating the ranking relation, the concordance index and discordance index for each alternative pair are calculated.For each pair of alternatives k and l (,  = 1, 2, 3, . . . .,  and k is not equal to l) the set of j criteria is divided into 2 subsets namely concordance and discordance.
To ascertain the values of the elements in the concordance matrix, one must add the weights contained within the concordance set, namely: To determine the value of the elements in the discordance matrix, divide the greatest difference between the criteria included in the discordance set and the greatest difference between the criteria.The variance is: • Step 7. Calculating concordance and discordance dominant matrix Calculating the dominant concordance matrix, the f matrix as the dominant concordance matrix can be built with the help of the threshold value, namely by comparing each concordance matrix element value with the threshold value ( ≥ ).The threshold value () is (−1) (13) where  = , so the matrix element f is determined as follows:  = 1   ≥  and  = 0   <  Calculating the dominant discordance matrix.g matrix as the dominant discordance matrix can be constructed with the help of values, namely by comparing each discordance matrix element value with the threshold value (  ≥  ).With threshold (d) value is • Step 8. Determine the aggregate dominance matrix Matrix e as the aggregate dominance matrix is a matrix in which each element is the result of the elements of matrix f and the corresponding elements of matrix g, so it can be written as follows: e kl= fkl x gkl (15) • Step 9. Elimination of less favorable alternatives Matrix E specifies the ranking of each alternative, so if  = 1, then Ak is a superior alternative than Al.Therefore, the row of matrix E with the fewest instances of  = 1 can be eliminated.Therefore, the row of matrix E with the fewest occurrences of  = 1 can be eliminated.Therefore, the greatest option is the option that dominates other options.

AHP Weighted Calculation
Based on the explanation of step 1. the weighting method with AHP, the decision maker constructs a comparison matrix.Next, the matrix is normalized as in step 2 Equation (1).By applying the AHP weighting calculation steps 3 to 7, The Synthetic Weight Value, Eigenvalue, Priority Weight Value, And Value of Interest are obtained in Table 4.In the initial step of calculating the  value, the maximum eigen value obtained in the preceding calculation step is used to test the consistency of the calculated value.Then proceed with the final calculation to verify the consistency value, which is to calculate the  value.
Calculating the value of CI using Equation (4), the initial counting procedure is as follows: Using Equation ( 5), the  value is computed as the final step in the AHP method application.1.63 is used as the  value.The calculation of the  value is as follows: Consequently, the consistency test is valid since the consistency ratio (/) is less than 0.1

AHP -MOORA Integration
Step 2. Decision Matrix Normalization: normalize the decision matrix using Equation (7).Step 3. Construct the weighted decision matrix after determining the criterion weights.In this work, the criterion weights are determined using the AHP approach.Step 4. Determine the assessment values by finding the difference between the sum of beneficial and non-beneficial (cost) criteria as given in Equation (8).
Step 5. Rank the assessment values in decreasing order to get the global rank of the alternatives.

AHP -ELECTRE Integration
This result and discussion will describe the outcomes of applying a combination of AHP and ELECTRE methods to aid in the selection of the optimal location for the electric and solar-powered shipyard industry.The subsequent step is to normalize the decision matrix using Equation ( 9).The results are then displayed as shown in the Decision Matrix Normalisation Table 8.The weight of each parameter that has been analysed using AHP is then used to multiply by the normalisation value of the Decision matrix using Equation (10).Then the results are displayed in the Weighted Normalisation of Decision Matrix table as in the Table 9.The concordance index set could be obtained as shown in the Table 10.
The discordance index set could be obtained as shown in the Table 11.The construction of the concordance matrix was processed by determining the element values of the concordance index set using Equation ( 11  From the results of the table, it can be seen that candidate L1 has the highest total value, so candidate L1 is decided to be the best location for the electric and solar-powered shipyard industry using the Electre approach with AHP weighting.

Results of Comparative Analysis
In this research, two different MCDM integration methods namely AHP-MOOORA and AHP-ELECTRE were successfully adopted to determine the optimal alternative location for the electric and solarpowered shipyard industry.After the fulfillment of the proposed comparative study, the following observations can be made: • The resulting ranking results are the same both using the AHP-MOORA and AHP-ELECTRE methods.
• Alternative location 1 is the best location based on the results of AHP-MOORA and AHP-ELECTRE analyses.
• alternative location 3, is in third place based on the results of AHP-MOORA and AHP-ELECTRE analyses.
• Among all other methodologies, AHP-MOORA is the simplest algorithm.It is merely a matter of summing up and contrasting positive and negative factors.
• Compared to AHP-MOORA, AHP-ELECTRE requires a substantial amount of computation.Given the additional information, it is evident that the computation of the set of concordancediscordance intervals, matrix intervals, and index matrix intervals requires additional time.

CONCLUSIONS
This research attempts to solve the problem of selecting the best location of three alternative locations (L1, L2, and L3) for an electric and solar-powered shipyard.It takes twenty-one ( 21) parameters (Table 1 to Table 21) based on regional characteristics, data availability, and satisfying constraints and requirements.Two different MCDM integration methods, AHP-MOOORA and AHP-ELECTRE, were adopted to determine the optimal alternative location.Based on the calculation using the AHP-MOORA and AHP-ELECTRE methods, both have the same results.Alternative location 1 (L1) has the first place as the best location, alternative location 3 (L3) has the second place, and alternative location 2 (L2) has the third place.Using a combination of these methods ensures that location 1 (L1) is the best location that best meets the criteria.Future research may consider adding other criteria that influence shipyard site selection more significantly.In addition, other MCDM methods with an ensemble approach can be used to integrate weighting or ranking methods to obtain more comprehensive decisions.
et al. used MOORA for ranking and Simos procedure for weighting [18].Cahyapratama & Sarno integrate Simple Additive Weighting (SAW) as a ranking approach and AHP as a weighting method in the vocalist selection process [19].Parkhan & Vatimbing used AHP and TOPSIS integration in choosing a shipyard location [20].Fatema et al. used AHP combined with the General Feature Extraction Technique (GFEA) for choosing the location of a trauma center [21].Kumar et al. combined AHP and MOORA to optimize the characteristics of electrical discharge machining [22].Another study combined ELECTRE with AHP in the decision-making process for the cycling routes in Franciacorta for sustainable tourism [23], [24].Prahesti et al. compared the AHP-ELECTRE and SAW method to give school recommendation based on criteria that student wants with applying [25].The results of these studies showed that combined weighting and ranking techniques gave an improvement in decision-making objectivity.

Figure 1 .
Figure 1.Alternate sites for the shipbuilding sector that use solar and electricity

Figure 2 .
Figure 2. The framework stage for this research

Figure 3 .
Figure 3. Parameters For Determining The Location Of Electric And Solar-Powered Shipyard Industries

|||||
), with the following result:The construction of the discordance matrix was processed by determining the element values of the discordance index set using Equation (12), with the following result:The step of calculating the threshold value before determining the dominant concordance matrix and the dominant discordance matrix.By using Equation (13), the threshold value  is obtained and Equation (14), the threshold value  is obtained as follows: 1) = 0,6083 Then the matrix f as the dominant concordance matrix is as follows:Then the matrix g as the dominant discordance matrix is as follows:The step to determine matrix e as the total dominant matrix.Using Equation (12), matrix e is obtained as follows:Due to the fact that  is a value of zero, the classification is determined by the values of C and D. Listed below are the results of the ranking.

Figure 4
is a summary of the AHP -MOORA ranking results and the AHP -ELECTRE Integration results.

Table 2 . Priority Arrangement Table Degree of concern Specific details Comprehensive Overview
1.
. Eur J Oper Res 1990; 48, 9-26• Step 2. Normalize a pairwise comparison matrix by aggregating the value of each matched pair matrix column and dividing each value in the column by the sum of the corresponding columns.

•
Step 2. Normalize the decision matrix, in this phase, each parameter is converted into a comparable value.The formula is used to normalize any   values [13].

Table 3
of the Decision Matrix.