Using Hybrid Method in Selecting Timber Factory Location

The selection of location is vital for a timber factory to keep on functioning. It is a signifi cant decision during the setup of a business and the preparation of projects. Therefore, dual scaling method often used for selecting the timber factory location and Analytic Hierarchy Process (AHP) have been used in this study. While the AHP method and the double weighing method were previously used separately, the aim of this study was to use these two studies together in order to obtain more reliable results. For this purpose, in the Western Black Sea region of Turkey, fi ve different candidate cities were selected for the establishment of a factory site for timber production: Bartin, Bolu, Kastamonu, Karabük and Zonguldak. At the same time, a total of ten factors including raw materials, labor, market, construction costs, energy and fuel, water, transportation, tax, security and social environment were determined. As a result of the study, the hybrid method, which is based on the average of both methods, yielded more reliable results.


INTRODUCTION 1. UVOD
Business location is generally the geographical place where the organization provides the main services.The location for a manufacturing business can be defi ned as the most suitable place for carrying out the main functions such as provision, production, storage and distribution, and the related economic purposes.Location is a compulsory life space for an organization to go on running and develop.The most suitable DRVNA INDUSTRIJA 69 (3) 273-281 (2018) location for an economic business is the place where it can provide productive services at minimum cost and maximum profi t after being established.The most suitable locations for the businesses, whose main aim is to prosper and bring benefi t, are the places where they can fulfi ll these aims (Barutçugil, 1988).
The selected location is, in strict sense, "the place where business production activities are carried out".In broad sense, it is defi ned as "the most suitable place where the fundamental functions of a business, such as provision, production, storage and distribution and business income, will be at the maximum level and business cost at the minimum level, meaning that the business will be able to meet the necessary conditions for reaching its goals"; "the place meeting most appropriately the necessary technical and economic conditions for production compared to other locations"; "the place where the sum of expenses is at the lowest level"; "the place where there is no saving from the expenses and transportation costs from the selected location to a newly established location by means of replacement analysis" (İlter, 2001).
The selection of business location is a constant problem.New businesses are being established all the time.A specifi c industry can be completely replaced in 30-40 years as a result of the fact that one of the factors affecting the selection of location loses its importance while another gains importance.A factory location satisfying the ideal conditions may lose this characteristic due to several reasons such as the changing environmental factors over time, and the changes in the place and scale of demand sources.Consequently, the convenience of the place, the change of location and other alternatives are the challenges that are frequently considered in every factory (Kobu, 1989).
The location selection is an important decision when establishing a business and developing projects.Thorough analysis is required before making a decision.Likewise, this decision is a factor that shapes the cost, profi tability and running of a business organization in the future.Changing the location later is very hard and expensive.Therefore, the most affordable and profi table place should be selected among the alternatives while determining the location.
The location of an organization and its selection is one of the most important strategic issues in terms of investment decisions.The selection of location is not only important commercially but it also includes the aspects such as income distribution, local development differences, benefi ts from environmental factors and incentive schemes, exogeneity based on the gathering of business organizations in the same region that could be connected to each other.
There are several factors that infl uence the decision about the business location and that should be evaluated during this process.The factors to be taken into consideration for the selection of business location are the following: -economic and quantitative factors, -quality factors, -non-economic factors.
Economic and quantitative factors could be listed as raw material and transportation, demand centers and product distribution, labor market, wage level and all other relevant costs.Quality factors include geodetic parameters such as educational opportunities, environmental awareness of the organization, labor force quality and substructure state.Non-economic factors are the parameters related to military, political and the fi rm's own strategy (Anonymous, 1985).
There are several basic principles for selecting the business location.These principles are as follows (Üçüncü, 2003): -Requirements about factory location should be determined objectively and scientifi cally; -Characteristics of the selected location that infl uence the factory services should be identifi ed; -Selected location studies should be conducted at specifi c stages and in proper order without mixing them; -Experts and organizations should be determined, whose experience and knowledge could be benefi cial in every phase; -The decision on selecting the location should be made after a comprehensive consideration and proper evaluation of the state of location factors; -Comprehensive, suitable, complete and certain information should be obtained from various sources.

MATERIALS AND METHOD 2. MATERIJALI I METODE
Five different candidate cities, namely Bartın, Bolu, Kastamonu, Karabük and Zonguldak, have been determined in the Black Sea Region for selecting the location of a timber factory, which will produce annually 13.000 m 3 of lumber (Figure 1).
Besides, ten factors in total have been determined such as Raw Material, Labor Force, Market, Construction Cost, Energy and Fuel, Water, Transportation, Tax, Security, Social Environment, all of which would affect the investment.

Metoda dvostrukog skaliranja
Importance scores are given to the factors affecting the selection of the location between 0 and 10 (it could be between 0 and 100 or 0 and 1) according to production effi ciency and importance level.Similarly, the scores that candidate locations will be able to get from each factor range from 0 to 10.The weighted scores that candidate locations will get from each factor are obtained by multiplying the importance scores and the scores of the candidate cities.The scores of candidate locations are added separately to each candidate location, and thus total scores are found.Total scores determine the evaluation order of the candidate locations.The candidate location having the highest total weight is selected as the right location (Üçüncü 2003).

Analytic Hierarchy Process (AHP) 2.2. Analitički hijerarhijski proces (AHP)
The Analytic Hierarchy Process (AHP) is a multicriteria decision making method that helps the decision-maker face a complex problem with multiple confl icting and subjective criteria (for example selecting a location or investment, project ranking, etc.).AHP is a mathematical method considering group and individual priorities, and evaluating the quantitative and qualitative variables together in the course of decision making.Several papers have addressed the AHP success stories in very different fi elds (Zahedi, 1986;Golden et al, 1989;Shim, 1989;Vargas, 1990;Saaty and Forman, 1992; Forman and Gass, 2001; Kumar and Vaidya, 2006;Ho, 2008;Liberatore and Nydick, 2008).
The use of personal judgment for decision making problems has increased on a remarkable scale recently.Efforts have been made to recognize specifi c decision making mechanisms considering the observations of the decision makers in different psychological and sociological situations through AHP.The aim of this method was to enable decision makers to make decisions more effectively (Saaty, 1980).This method has attracted considerable attention and has been applied for the solution of most decision making problems in real life.
The fi rst step in AHP is to determine the factors and sub-factors in line with the purpose of the decision maker.Initially, the purpose is set in AHP, and the factors infl uencing the purpose in line with this purpose are tried to be determined.In this stage, a survey study and opinion of the experts in this area could be obtained and applied to specify all the factors infl uencing the purpose in line with this purpose.
Psychologists argue that it is easier and more accurate to express one's opinion on only two alternatives than simultaneously on all the alternatives.It also allows consistency and cross checking between different pairwise comparisons.AHP uses a ratio scale, which, contrary to methods using interval scales (Kainulainen et al., 2009), requires no units in the comparison.The judgement is a relative value or a quotient ab of two quantities a and b having the same units (intensity, meters, utility and so on).The decision-maker does not need to provide a numerical judgement; instead a relative verbal appreciation, more common in our daily lives, is suffi cient.
Dual comparison decision matrixes are formed in order to determine the signifi cance level between each other after specifying the purpose, factor and sub-factors.While forming these matrixes, 1-9 signifi cance scale by Saaty (1990) is used.Provided that the decision made at the end of the study is infl uential for most people, dual comparison decision matrixes are formed by integrating the judgment of different people.A plenty of researchers recommend the use of geometric average method in this integration process so as to obtain consistent dual comparison matrixes (Tam and Tummala, 2001).1-9 signifi cance scale suggested by Saaty provides the best results.The other signifi cance scales such as 1-5, 1-7, 1-15 and 1-20 fail to fi nd the appropriate solution.The signifi cance scale values and meanings are explained in Table 1 (Saaty, 1980).The formation of dual comparison decision matrixes is the most important stage of AHP.According to the data by dual comparison decision matrixes, the judgments are converted into a matrix in AHP.If a ij is indicated as dual comparison score of i. and j., a ij value is obtained from 1/a ij equivalence.This characteristic is called correspondence (Saaty, 1999).After creating dual comparison decision matrixes, the following step is to calculate the priorities or weight vectors.The method requires the normalization of the comparison matrix, adding the values in each column.The next step is to divide each cell by the total of the column.Based on this normalized matrix, the overall or fi nal priorities are obtained by calculating the average value of each row.In the AHP, the pairwise comparisons in a judgment matrix are considered to be adequately consistent if the corresponding consistency ratio (CR) is less than 10 % (Saaty, 1980).In the AHP, the consistency ratio is defi ned as CR, where CR = CI/RI.To calculate the consistency index (CI), the corresponding column value in the decision matrix is multiplied by the values of the priority vectors that have emerged.After this phase, CI value (λenb-n)/n-1 is found via the solution of the equations system.Consistency rate (CR) is obtained by dividing the obtained CI values by the Random Integrity Index.RI value takes different values according to the number "n".However, calculating the eigen values and eigen vectors of this equation system is very complicated and time-consuming especially for large-scale matrixes (n>5).The methods, which are easier to calculate and will give approximate results instead of the above system of equations, are preferred in the implementation (Saaty, 2000).A common method used for calculating the priority vectors is this: Normalized matrix is found by dividing every column value into the related column sum separately, and every sequence value is averaged with reference to the normalized matrix, and these values are the importance weights found for each factor.The priority vector is formed via these weights.
Finally, the result matrix is found by multiplying weights vector and binary matrixes.Thus, the objective is accomplished by selecting the most suitable alternative for the criteria identifi ed among the alternatives.
Data were obtained by conducting a questionnaire with 14 experts in the fi eld.Those who deviated from these data have been eliminated.The average of the answers given by the remaining 11 experts was used in the study.The data of the work is also the actual data used in an investment project.

Conversion of data appropriate for AHP method 2.3. Pretvorba podataka prikladnih za AHP metodu
The data, which were obtained as a result of dual scaling and an algorithm developed in the study, are turned into data sets to be used by the AHP method.The formulas are the following: Where: a ij -data set obtained via dual scaling Mak(a ij ) -the highest number in the data set t ij -normalized data for AHP Min(t ij ) -the lowest number in the data set z ij -data used for AHP.

REZULTATI
The factors, determined beforehand for Timber Factory, should be weighted before dual scaling method.The loads of the necessary factors have been identifi ed so as to determine the factory location during the installation of the timber factory (Table 2).The data were obtained by averaging the data recommended by 11 experts.
The data in the study are primarily evaluated in regard to Dual Scaling method.The results obtained via Dual Scaling method are presented below (Table 3).The data were obtained by averaging the data recommended by 11 experts.
The data sets are made usable for AHP by using normalization formulas developed in the following step (Table 4).One factor favors over another.

Extreme importance ekstremno važnije
The evidence showing one factor favoring over the other has a high reliability.Dokazi na osnovi kojih se favorizira jedan kriterij ili alternativa u odnosu prema drugome potvrđeni su s najvećom uvjerljivošću.

2,4,6,8 Intermediate values međuvrijednosti
The values between two successive judgments to be used when compromise is necessary.
Vrijednosti između dviju uzastopnih prosudbi koje će se primijeniti kada je potrebno napraviti kompromis.Step 1: The matrixes in Table 5 have been found as a result of forming the priority matrixes of the candidate cities for each factor.Step 2: The column values of every matrix are added up and they are divided into the data in that column (Table 6, 7).Step 3: Necessary coeffi cients for raw material factor to be used in the main matrix are obtained by fi nding the line averages of these new values (Table 7).
The exact matrix is found as follows by repeating the same process for the other factors (Table 8).
After the matrix is found as a result of comparison between the factors among the candidate cities, a new matrix is similarly formed in the consequence of priority comparisons applied among the factors themselves (Table 9).The loads of the factors have been identifi ed as shown in Table 10 by conducting similar processes as in step 2 and 3.

Step 4:
After multiplying the last two matrixes, the values indicate which city stands out in the selection of the factory location (Table 11).
It has been determined, with 22.08 %, that it would be most appropriate to establish the planned timber factory in Bolu.

DISCUSSION AND CONCLUSION 4. RASPRAVA I ZAKLJUČAK
It is clearly stated in the literature that multicriteria decision making techniques and the results obtained using AHP are more effective when used together (Kurttilaa et al., 2000;Gürbüz et al., 2013;Okello et al., 2014).The aim of this study was to show that the result is more effective when combining It can be seen that Bolu achieved this result based on some important factors such as raw material, labor force and market.It is striking that both methods used in the study gave similar results.It is understood in the evaluation method that Bolu is proceeding with 22.66 % compared to other cities, where both methods were averaged (Table 12).
Although the ranking seems the same resulting from both methods, the increase in Zonguldak, Karabük and Kastamonu stands out, while there is a decrease in Bolu and Bartın according to the AHP method.The increase is especially dramatic and remarkable in Zonguldak and Karabük.
A similar result arising from the evaluation of the candidate cities appears in the comparison of the loads.The anticipated loads for the factors that will determine the location are such as to affect the results directly.The most important factor in dual scaling method, raw material, is the factor having the highest weight (0.1488) in accordance with AHP.The factor having the lowest weight, Water, has the lowest weight in AHP method with 0.0429.
It is very important to go through the details again before selecting the timber factory location.All the alternatives must be assessed before determining the location where large-scale factories, requiring large investments, will be established.As a result of this study, applying the AHP method after the implementation of dual scaling method will cause the planners check their point of view.Such evaluation will provide reconsideration of the factors regarded as less important.
This study will provide a different viewpoint for the selection of timber factory location.This approach has been applied for the fi rst time in this area.By taking the average of the results obtained by these two methods at the end of the study, the evaluation according to these data will lead the decision makers to make decisions in a more reliable way.
The results of this study show that not only can the timber factory location be selected, but in other areas it can be used to determine the factory location.The location of the paper mill, fi berboard mill and chipboard mill can also be selected.

Table 2
Values of factors to be used in dual scaling method Tablica 2. Vrijednosti kriterija koji će se primijeniti u metodi dvostrukog skaliranja

Table 4
Conversion of data sets to data to be used by AHP Tablica 4. Skupovi podataka pretvoreni u podatke kojima će se koristiti AHP metoda

Table 10
Determination of factor loads Tablica 10.Određivanje opterećenosti kriterija Considering the results of dual scaling, Bolu is a good option for the timber factory planned to be established with the value of 546.99 (23.25 %).Bolu has ranked fi rst with 0.2208 (22.08 %) as a result of the evaluation via AHP.

Table 12
Comparison of the results obtained by two methods and hybrid method Tablica 12. Usporedba rezultata dobivenih zasebnom primjenom metode odlučivanja i hibridne metode