thesis result tu

Faculty of Management

Tribhuvan university, 01-4332718, 01-4330818, 01-5907927 (thesis), staff login.

• MBS (Sem.) Dissertation
• MPA (Sem.) Dissertation
• Before 2078 Asar
• After 2078 Asar

MBS Thesis Status

Faculty of Management

Office of the Dean

Kirtipur, Kathmandu, Nepal

014330818, 01-4332118

01-5195568(Exam)

01-5907927(Thesis)

[email protected]

Quick Links

Thesis & Dissertations

Permanent uri for this community, communities in this community, results per page, sort options.

• Agriculture & Animal Science   3
• Education   4497
• Engineering   683
• Forestry   3
• Humanities & Social Science   8885
• Law   1
• Management   5053
• Medicine   15
• Science & Technology   1568

Collections in this Community

• Sociology   1

Examination Controller Division

Faculty of management, tribhuvan university, examinations and results.

• ONLINE CMAT FORM

• All Notices
• Master or Above

• SCHOLARSHIPS
• AFFILIATION
• TOP COLLEGES
• STUDY ABROAD
• ENTRANCE EXAM
• COURSE INQUIRY
• Kamana Sewa Bikas Bank Jobs
• TU 50th Convocation
• Lok Sewa Vacancy Calendar
• Bank Vacancy Calendar
• Pokhara Uni Scholarship

500+ vacancies in Government Sector and Lok Sewa

• By Categories:
• Teachers Service Commission (TSC)
• News on Focus
• TU Question Bank
• Medical Education Commission

TU Masters Degree Dissertation Writing Guideline

(adsbygoogle = window.adsbygoogle || []).push({});.

Tribhuvan University, Faculty of Management has published Masters Degree Dissertation (Thesis) Writing Guideline (Format) 2019. About Dissertation Guidelines These dissertation guidelines have been created as a guide to help Master?s level students establish minimum requirements, academic standards, the physical format and appearance of dissertation. The purpose is to provideacademic requirements and structural guidelines required for dissertation writing to the studentsunder the Faculty of Management (FOM), Tribhuvan University(TU). The FOM encourages the preparation of documents to be consistent with the specialized requirements prior to the submission. Submission of this document is the final step in a program leading to conferral of a Master?s degree. DOWNLOADS 1. Masters Degree Dissertation Writing Guideline 2019  [ DOWNLOAD ] READ ALSO:

TU publishes Procedures and Format for 4 Year BBS 4th Year Project Work

TU Faculty of Management Exam Schedule notice

TU Faculty of Management Exam Centers Notice

TU Bachelor 2nd year Exam routine

TU BBS 2nd Year exam routine published

TU Bachelor 1st year exam form fill up notice

Tribuvan University CMAT Entrance Exam centers

Featured news.

TU MBS Model & Old Exam Question Paper, Solutions and Syllabus

Tribhuvan University BCA Entrance Examination Results 2081 (2024)

NEB Class 12 Grade Increment (Supplementary) Exam Center 2081 2024

Tribhuvan University 50th Convocation Ceremony

TU publishes results

TU BBA Model & Old Exam Question Paper, Solutions and Syllabus

Featured articles.

Lok Sewa Aayog (PSC) Exams Reading Materials 2081

Nepal Telecom-NTC Lok Sewa Old Exam Question and Answer

Nepal Electricity Authority-NEA Lok Sewa Old Exam Questions and Answer

Officer Cadet Exam Syllabus - Nepal Army

Lok Sewa Old Past Exam Questions for Assistant and Officer Level

Lok Sewa Vacancy Calendar 2081 2082 ( All Provinces)

Featured vacancies.

Vacancy notice from Nepal Oil Corporation (NOC)

41 vacancies from Province Lok Sewa Aayog, Lumbini Province

320 vacancies from Nepal Police for various positions

Kamana Sewa Bikas Bank announces vacancy for various positions

Jobs at American Embassy; Salary: NPR 89K+ per month

Vacancy notice from Jyoti Bikash Bank

Would you like to advertise here.

Connect with us

Post graduate MSc. program thesis result notice published by IAAS, Tribhuvan University (TU)

Tribhuvan University, Institute of Agriculture and Animal Science have published roll numbers of students who have submitted their thesis work and has been marked satisfactory  in the result published by the University at their notice below:

• Tribhuvan University first announcement of Masters Dissertation Meet 2024
• Tribhuvan University published notice about abstract submission for MSc Dissertation meet 2024
• Master of Science (MSc) 1st semester exam schedule of 2080 batch by Tribhuvan University (TU)
• MSc Semester System Entrance Examination Schedule 2080
• Postgraduate level in MSc programs comprehensive exam results published by IAAS, Tribhuvan University (TU)
• TU Masters level entrance exam 2080 application form fillup extension period notice
• TU MSc Form Filling Notice
• View all Notices

• Change password
• Similarity score
• หน้าที่ 1 (0)
• หน้าที่ 2 (0)
• หน้าที่ 3 (1)
• หน้าที่ 4 (0)
• หน้าที่ 5 (0)
• หน้าที่ 6 (0)
• หน้าที่ 7 (0)
• หน้าที่ 8 (0)
• หน้าที่ 9 (0)
• หน้าที่ 10 (0)
• หน้าที่ 11 (0)
• หน้าที่ 12 (0)
• หน้าที่ 13 (0)
• หน้าที่ 14 (0)
• หน้าที่ 15 (0)
• หน้าที่ 16 (0)
• หน้าที่ 17 (0)
• หน้าที่ 18 (0)
• หน้าที่ 19 (0)
• หน้าที่ 20 (0)
• หน้าที่ 21 (0)
• หน้าที่ 22 (0)
• หน้าที่ 23 (0)
• หน้าที่ 24 (0)
• หน้าที่ 25 (0)
• หน้าที่ 26 (0)
• หน้าที่ 27 (0)
• หน้าที่ 28 (0)
• หน้าที่ 29 (0)
• หน้าที่ 30 (0)
• หน้าที่ 31 (0)
• หน้าที่ 32 (0)
• หน้าที่ 33 (0)
• หน้าที่ 34 (0)
• หน้าที่ 35 (0)
• หน้าที่ 36 (0)
• หน้าที่ 37 (0)
• หน้าที่ 38 (0)
• หน้าที่ 39 (0)
• หน้าที่ 40 (0)
• หน้าที่ 41 (0)
• หน้าที่ 42 (0)
• หน้าที่ 43 (0)
• หน้าที่ 44 (0)
• หน้าที่ 45 (0)
• หน้าที่ 46 (0)
• หน้าที่ 47 (0)
• หน้าที่ 48 (0)
• หน้าที่ 49 (0)
• หน้าที่ 50 (0)
• หน้าที่ 51 (0)
• หน้าที่ 52 (0)
• หน้าที่ 53 (0)
• หน้าที่ 54 (0)
• หน้าที่ 55 (0)
• หน้าที่ 56 (0)
• หน้าที่ 57 (0)
• หน้าที่ 58 (0)
• หน้าที่ 59 (0)
• หน้าที่ 60 (0)
• หน้าที่ 61 (0)
• หน้าที่ 62 (0)
• หน้าที่ 63 (0)
• หน้าที่ 64 (0)
• หน้าที่ 65 (0)
• หน้าที่ 66 (0)
• หน้าที่ 67 (0)
• หน้าที่ 68 (0)
• หน้าที่ 69 (0)
• หน้าที่ 70 (0)
• หน้าที่ 71 (0)
• หน้าที่ 72 (0)
• หน้าที่ 73 (0)
• หน้าที่ 74 (0)

WIND TURBINE SITE SELECTION IN INDONESIA

GALIH PAMBUDI

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF MASTER OF

ENGINEERING (LOGISTICS AND SUPPLY CHAIN SYSTEMS

ENGINEERING)

SIRINDHORN INTERNATIONAL INSTITUTE OF TECHNOLOGY

THAMMASAT UNIVERSITY

ACADEMIC YEAR 2018

A Thesis Presented

Submitted to

Sirindhorn International Institute of Technology

Thammasat University

In partial fulfillment of the requirements for the degree of

MASTER OF ENGINEERING (LOGISTICS AND SUPPLY CHAIN SYSTEMS

Approved as to style and content by

(Asst. Prof. Dr. Narameth Nananukul)

Committee Member and

Chairperson of Examination Committee

(Asst. Prof. Dr. Morrakot Raweewan)

Committee Member

(Assoc. Prof. Dr. Thananya Wasusri)

NOVEMBER 2018

Acknowledgements

The author gratefully acknowledges the financial support provided by the Excellent

Foreign Student Scholarship (EFS) for Graduate Student in Sirindhorn International

Institute of Technology, Thammasat University.

Bachelor of Engineering, Universitas Gadjah Mada, 2016

Master of Engineering, Sirindhorn International Institute of Technology, 2018

Wind farm sites are selected in spacious regions which have more output potential

within constraint resources. Due to its spacious terrain, Indonesia has great potential

for building wind power plants, providing the perfect settings to generate electricity

using wind energy. Keeping in view the reliability and sustainability of wind farm sites,

the selection of the most suitable locations for optimal result is of prime concern to

generate greater amount of energy with less utilization of resources. In this study, the

focus is on proposing a multi-criterion approach to find the most suitable location for

building wind farms. Locations from every region of Indonesia were selected based on

two levels defined by district level to province level. All districts and provinces are

considered as Decision-Making Units (DMUs) which are used to measure the

efficiency scores using Dual Data Envelopment Analysis (DDEA) method. Two levels

are defined to find the best feasible locations within Indonesia from 165 districts and

33 provinces with major focus on geographical and structural technicality of each

DMU. The results show that South Sumatra province has the highest priority potential

for the construction of wind power plants, especially in the district of Palembang. West

Papua, Papua and Maluku provinces have descending priority based on good

infrastructure accessibility and less prone to natural disaster.

Keywords : Dual data envelopment analysis, Wind Power Plant, Site Selection,

Decision Making Unit.

Table of Contents

Chapter Title

Signature Page

List of Tables

List of Figures

1 Introduction

1.1 Background of Propose Study

1.2 Problem Statement

1.3 Objectives of Propose Study

1.4 The Advantages of Propose Study

2 Literature Review

2.1 Literature Review

2.2 Research Gap

3 Research Methodology

3.1 Possible Factors

3.1.1 Level 1 criteria

3.1.2 Level 2 criteria

3.2 Methodology

3.2.1 Dual Data Envelopment Analysis

3.2.2 The Hierarchical Model for two Level DDEA

3.2.3 Fuzzy Primary Data Envelopment Analysis

3.2.4 Principal Component Analysis

4 Results and Discussion

4.1. Data Envelopment Analysis Results

4.2 The Hierarchical Model for two Level DDEA Results

4.3 Fuzzy Primary Data Envelopment Analysis Results

4.4 Principle Component Analysis Results

4.5 Comparison of Three Methods Result

5 Conclusions and Recommendations

5.1 Conclusion

5.2. Recommendations

Appendix A Data Resources

A.1 Districts Level Data

A.2 Provinces Level Data

Appendix B General Optimization Model in IBM ILOG CPLEX

B.1 Model on Districts Level

B.2 Model on Provinces Level

2.1 The summary of case study

2.2 Summarizes the relevant criteria in the wind farm site selection

2.3 Summarizes the relevant methods in the wind farm site selection

3.1 Comparison of the land requirement in different power plant [16]

3.2 Cost analyst for 1 m length of road infrastructure

3.3 Wind class definitions [17]

4.1 Efficiency and ranking of provinces (Level 2)

4.2 Efficiency score of districts (Level 1)

4.3 Detail of hierarchical score for provinces level

4.4 Hierarchical score for five dominant provinces

4.5 Importance degree and context free grammar on HFLTS

4.6 Pairwise evaluations of one expert in main criteria on level 1

4.7 Obtained envelops for HFLTS

4.8 Pessimistic and optimistic preference in district level

4.9 The linguistic interval, interval utilities, midpoint and weights

4.10 Pessimistic and optimistic preference in province level

4.11 The constraint of the priorities for district level

4.12 The constraint of the priorities for province level

4.13 Hierarchical Score for HFLTS

4.14 Principal Component Analysis Results

4.15 Comparison of three methods result.

3.1 Maps of provinces in Indonesia

3.2 Land cost of districts in Indonesia

3.3 Types of road infrastructure

3.4 Types of road infrastructure in Medan North Sumatra

3.5 Distance of the primary and secondary road infrastructures in Medan

3.6 Total cost of infrastructure data

3.7 Data of population in districts

3.8 Ratio of usage area

3.9 Data of electricity consumption in Indonesia’s Provinces

3.10 Data of natural disaster in provinces

3.11 Gravity loading; a. full blade; b. spar-only simplification

3.12 Blade loading cases; a. edgewise bending; b. flap-wise bending

3.13 Data of wind velocity in provinces of Indonesia

3.14 Data of total area in provinces in Indonesia

3.15 Flow Chart of the Proposed Study

3.16 Extraction of Factor analysis in district level

3.17 Scree Plot of district level

3.18 Extraction Box

3.19 Descriptive Box

3.20 Rotation Box

3.21 Options Box

4.1 Hierarchical Score

4.2 Correlation matrix on district level

4.3 KMO and Bartlett’s Test on district level

4.4 Communalities on level 1

4.5 Total variance on district level

4.6 Component matrix on district level

4.7 Pattern matrix on district level

4.8 Structure matrix on level 1

4.9 Scree plot for level 2

4.10 KMO and Bartlett’s Test for Level 2

4.11 Correlation matrix on level 2

4.12 Communalities on level 2

4.13 Total variance on level 2

4.14 Component matrix on level 2

4.15 Pattern matrix on level 2

4.16 Structure matrix on level 2

4.17 Component correlation matrix on level 2

4.18 Top five Provinces in Indonesia

5.1 Full Score of Three Methods

Introduction

The propose study in the first chapter are divided into four sections. Section 1.1

gives background of the proposed study and show the importance of the research study.

Section 1.2 contains details of problem statement of proposed study to define the issue

of the research. Section 1.3 gives the objective of proposed study presents the

framework of this study. Section. Section 1.4 The advantages of the proposed study

provide the benefit of the study to apply in the research area.

1.1 Background

Natural energy resources such as wind energy is renewable, and is freely

available which could lead to the sustainability of energy usage. Selecting the most

suitable sites which have the optimal wind energy resource is a complicated decision-

making process. It is considered as primary concern based on the sustainability and

reliability aspects. The selection of the optimal locations is very important including

several factors the topography of the area and the usage of the decision support models

could fulfill the requisites and shows the optimal outcome. It means modelling,

formulation and determining solution of the site problem that can be implemented in

establishing facilities in the selected area. Different literatures show that there are

different approaches for selecting the optimal location for wind power plant site, as

follows Haydar et al. [1] defining the optimal area in university for a station of wind

observation based on Analytical Hierarchy Process (AHP) approach. Bhatnagar et al.

[2] the establishment of gas stations and power plants using location factor as multi-

criteria. Afshartous et al. [3] to determine the location of the coast guard air station

based on Improved Optimization Model. Gamboa et al. [4] determining wind plant site

selection as a multi-criterion used social framework. Choudhary et al. [5] determined

site selection of thermal power plant based on Fuzzy DEA. As it is evident from the

previous studies, the site selection is of prime concern for establishing a facility at some

place. It needs multitude factors to be considered, making the decision hard and

required complex modeling. In this study, the method based on Dual Data Envelopment

Analysis (DEA) approach with multi-criteria is used as site selection mechanism for

wind power plants construction within Indonesia. In Indonesia the energy demand is

growing dramatically than population. At present, Indonesia have six main types of

power plant use gas, steam turbines, combined cycle, geothermal, diesel engine, and

hydro-power where fossil fuel is the major energy generation [6]. In this decades, for

genererating the electricity in Indonesia, the resources up to 96% using fossil fuel and

just 4% uses renewable energy. Hence, the government policy targets a portion of

renewable energy resources to be increased up to 17% in 2025 [7]. The current energy

policy in Indonesia is central in Fossil fuel. Decreasing of fossil fuel resources and

growing environmental concerns are challanging viewpoint in Indonesia’s energy

policy which leads to the propose of using renewable energy to increase energy

efficiency [8]. Indonesia as a archipelago country, having huge potential for wind

power generation because of high wind rate in most of the regions. The criteria of wind

turbine site selection should be selected carefully before making decisions.

Determining the potential of using the wind power in the possible region is

important. In spite of the comprehensiveness in location considered for the optimization

of wind power plants, the criteria and the method for the site selection that will be used

to compare the potential of the region must be carefully selected. Location problem

includes simulation, formulation and model in establish the facility in every region

which is likely to have multiple factors and is difficult for the analysis. The quantitative

approach must be used to determine the suitable locations.

1.2 Objectives of Propose Study

This study considers an integrated mathematical approach for location

optimization of wind plants. Determining all criteria that significantly influences the

establishment of a wind farm in Indonesia is important. The implementation of the

proposed approach to decide the most suitable location for building of a wind power

plant in Indonesia is based on a Dual Data Envelopment Analysis (DEA) for wind farm

power plant.

The advantages of this proposed study hopefully can be used as the alternative

approach to decide site selection, generally in any case and especially in wind plant

power plant. This proposed study can help improve the reseach which have correlation

with location optimization in wind farm location on the other location.

Literature Review

As a consideration of the literature, the proposed study refers several studies

which have been reviewed as a reference. Section 2.1 show the insight of the literature

review. Section 2.2 presents research gap which is used in the proposed study. Herewith

is further description of the research and the comparison of the previous studies.

Data envelopment analysis (DEA) is for analyzing the performance efficiency

of the comparable units called decision-making units (DMUs) as quantitative method.

Every DMU performs the same purpose by using ratio between input and output criteria

which are characterized by the modeled system [9]. Several references which have used

DEA for site selection such as Ertek et al . [10] for determining the efficiency of on-

shore wind turbines they provided data centric analysis. Saglam, U [11] The goal of

those paper was to evaluate quantitatively efficiencies of 39 states wind power

performance for electricity generation by using multi-criteria methods as DEA. Wu et

al. [12] in China to perform efficiency assessment of wind power plant used based on

two stage of DEA. These studies identified potential inefficicient factors and try to seek

out the factor which can improve the performance of wind farm. Azadeh et al. [9]

provided wind farm site selection under uncertainty using Hierarchical Fuzzy DEA.

Since traditional DDEA models cannot be used to combine the indicators especially in

qualitative data. Sueyoshi et al. [13] proposed an approach improvement as Range

Adjusted Measure (RAM) which is as integrated of DEA. Seiford et al. [14] proposed

the results from multi-stage DEA involved the input and output criteria which are

validated by Numerical Taxonomy and Principal Component Analysi. In this study,

the efficiency of DMUs in the selection of most suitable location for wind farm plant is

based on land cost, road accessibility, infrastructure cost, population density, supply

demand, natural vulnerability, wind velocity and total area. This research proposes a

multi-criterion apporach based on Data Envelopment Analysis (DEA) for analyzing the

most feasible wind farm site selection in Indonesia.

According of the literature that have been reviewed, the summary of the case

study is shown in Table 2.1. Table 2.2. lists the criteria which are significant influence

in the site selection of wind farm. Therefore, the methods based on the quantitative

approach that have been used are shown in Table 2.3. Further information shows in the

describe as below:

Table 2.1 The summary of case study

Year Case Study

Saglam, Ümit

2017 efficiency assessments of 39 state’s wind power location

using A two-stage data envelopment analysis in the United

Yunna Wu, et

2016 Efficiency assessment of wind farms location using two-stage

data envelopment analysis in China

Azadeh, Ali et

2013 Location optimization of wind power generation systems

under uncertainty using hierarchical fuzzy DEA in Iran

2010 Location optimization of wind plants by an integrated

hierarchical Data Envelopment Analysis in Iran

Ertek, Gürdal

2012 Insights into the efficiencies of wind turbines using data

envelopment analysis

Table 2.2. Summarizes the relevant criteria in the wind farm site selection

Year DMU Input

Annual Land Lease

Payments ($)

Wind Industry Employment,

Auxiliary electricity

consumption,

Wind power density

Electricity

Average wind blow

Azadeh, Ali

Level 1 Land Cost

Level 2 Intensity of

occurrence,

Level 1 Population and human

labor, Distance of power

distribution networks,

Level 2 Average wind blow,

Quantity of proper geological

areas, Quantity of proper

topographical

Consumer proximity

distribution

topographical areas

Diameter of Plant

Nominal Output (kW)

Table 2.3. Summarizes the relevant methods in the wind farm site selection

PCA NT Tobit

Saglam, Ümit [11]

Yunna Wu, et al [12]

Azadeh, Ali et al [15]

Azadeh, Ali et al [9]

Ertek, Gürdal et al [10]

Where: DEA (Data Envelopment Analysis), PCA (Principal Component Analysis), NT

(Numerical Taxonomy).

The research gap of this proposed study is wind farm site selection in province

of Indonesia using multi-criteria approach based on hierarchical dual Data

Envelopment Analysis (DEA). The integrated data envelopment analysis will be

applied on two levels of DEA, the first level considers finding the best suitable province

in Indonesia and the second level focuses on sub-district within the province based on

the distance from remote areas. The possible factors used in the districts level as defined

by land cost, population in region, ratio of free usage area, primary road, secondary

road, tertiary road, and total cost of infrastructure. In the provinces level as defined by

wind velocity, population in province, total area, electricity consumption, less of land

slide, flood, earthquake and volcanic eruption. Determining the efficiency for districts

and province level based on Hierarchical Dual Data Envelopment Analysis. Hesitant

Fuzzy Linguistic Term Set (HFLTS) for determining the weight for importance criteria.

The validation of the significant criteria based on Principal Component Analysis

(PCA). Finally, comparing three methods for deciding the most suitable location for

wind turbine site selection in Indonesia.

Research Methodology

In this chapter further information about the criteria and methods used in this

proposed study is described. Section 3.1 provides description of the possible factors

which have influence to the wind farm site selection. Section 3.2 presents the methods

which are applied in this proposed study.

Based on the literature review, the proposed factors used in this study are

districts (Level 1) and provinces (Level 2) of Indonesia as shown in Figure 3.1. The

integrated model for wind farm site selection organizes the factors into two levels

defined as input and output. The optimization technique is based on Dual Data

Envelopment Analysis method to find the most efficient location. The integrated level

criteria are developed to select the most suitable location in term of province of

Fig.3.1 Maps of provinces in Indonesia

3.1.1 Level 1 Criteria

The objective of using level 1 criteria is to determine the most suitable province

in Indonesia for establishment of wind farm plant based on the efficiency of the

location. The Level 1 criteria are:

Land cost by districts in Indonesia: the land cost has become an important

criterion due to the unprecedented increase in Indonesian population, which must be

included for site selection. For selecting a wind farm site, it requires more spacious area

as compared with other energy sources. Table 3.1 shows the comparison of the amount

of the land required for the construction of each kind of facility [16]. Area required for

wind farm is up to 9900 km

/GW/year which is 283 times more than coal plant. Its

means that the land cost is the main important criterion for wind farm site selection.

Figure 3.2 shows the data of land costs in some districts in Indonesia.

Fig.3.2 Land cost of districts in Indonesia

Table 3.1 Comparison of the land requirement in different power plant [16]

Land use in km2/GW per year

Wind power plant

Hydroelectric

Natural gas

The type of road infrastructure: Good road accessibility to the constructed

facility is one of the most importance considered factor for reliable, timely

transportation and distribution of goods to and from the facility. Different road facilities

(i.e., primary, secondary and tertiary roads) have different distribution lead time which

can affect the accessibility to the facility. Data of the road infrastructure are shown in

Figure 3.3. Primary and secondary roads are main roads and can be used for the

transportation of heavy goods. On contrary to this, tertiary roads are mostly used as

connectors to the primary and secondary roads such as, small bike roads and village

roads. Hence, for timely distribution of goods to and from the wind farm need the most

effective routes. In this study, primary and secondary road infrastructures are used as a

preference indicator for wind power plant construction.

Fig.3.3 Types of road infrastructure

Relevant data, especially types of road infrastructure within each district in

every province of Indonesia, will be used. Firstly, finding the ArcGIS Maps for every

district from the Ministry of Public Works and Public Housing of Indonesia’s data

representing the infrastructures maps in every region such as types of road as well as

natural resources. By using the ArcGIS, national roads, province roads are defined as

primary road, in the maps are shown as dark red line. The district and regional roads

are considered as secondary road type represented as light red line on the maps. The

tertiary road is one of important criterion which should be careful determined and set,

due to the limited resources in ArcGIS. They can be determined by looking for village

roads or small roads which are less than 3.6 m in width. In here, the types of road

infrastructure by ArcGIS maps and google maps. Fig. 3.4. represents the difference

types of primary roads and secondary roads.

Fig.3.4 Types of road infrastructure in Medan North Sumatra

The border point between the primary and secondary roads is used to determine

the distance of the roads. Fig.3.5 shows the distance of primary and secondary roads in

Medan Region. It shows that primary road as the significance distance in Medan City

is 18.6 km and for secondary road is 9.8 km. The distance is based on a spotted location

in remote region which is suitable to establish a wind farm. Due to the good

infrastructure in the main region of Medan, there is no need for tertiary road so this

region just have 0 km of it and do not need to build additional infrastructures.

Fig.3.5 Distance of the primary and secondary road infrastructures in Medan

Total cost of infrastructure (IDR): The construction of the infrastructure

requires a lot of capital cost incurred. So, selection of the site with less incurring cost

for building new roads to access the facility is also very much important to avoid extra

expenses thus increasing the overheads of the construction projects. The data in this

study is shown in Figure 3.6 which are the sites selected by the minimum capital cost

for the construction of tertiary roads with the shortest distance. In other words, if in any

case the construction of the infrastructure is inevitable and unavoidable, tertiary roads

are given preference over primary and secondary roads. In our approach we prefer the

construction of tertiary road as compared with secondary road if the width of the

dispatching vehicle is up to 5m. As the average width of the tertiary roads in Indonesia

is up to 3.6m and we included 1.4m to accommodate the convenience of shipment. So,

the minimum cost of infrastructure is another input criterion in DDEA model.

Fig.3.6 Total cost of infrastructure data

The total cost of infrastructure for each region is collected based on Widarno,

B et al (2015). The included components are labors, materials, tools and equipment as

shown on Table 3.2. as cost analyst for each 1 m length of road. As a result, the total

area of 1 m

is approximately 150,000 (IDR) (1 USD as 13,994.25 IDR). As an example,

from Fig.3.6. consider one of regions in South Sumatra where the selected region to

build infrastructure is Pagar Alan. In Pagar Alan, there are up to 159.1 km as tertiary

road and needed to expand the road for shipping the wind power materials to the

location. The total cost for building tertiary roads in Pagar Alan is 159.1 x 1000 m x

150,000 IDR which is 23,865,000,000 IDR.

Table 3.2 Cost analyst for 1 m

length of road infrastructure

Population by district in Indonesia: for maintenance and operational

technicality in case of emergency the wind power plant should be established in regions

with easily available human resources. It can likely decrease the resources management

for the transportation cost of labors, accommodation of labors and expert availability

when needed. The choice of a centered place with easy accessibility of human resources

is an important output indicator in the DDEA modal calculation. Figure 3.7 shows the

population in districts of Indonesia.

No Component

Coefficient

Total cost (IDR)

Aggregate B class

Wheel loader

Motor grader

Vibratory loader

Pneumatic tire loader

Water tanker

Assisted tools

Total cost of labor per m

Equipment and tools

Fig.3.7 Data of population in districts

Ratio of free usage area: Areas with greater value of free area usage ratio near

to one are considered more suitable for establishment of the wind power plants. The

free usage area means the ratio between total area divided by population in each region.

The more available land is preferred and used as output in DDEA modal calculation.

Figure 3.8 shows the data of the free usage area ratio.

Fig.3.8 Ratio of free usage area

3.1.2 Level 2 Criteria

In the second level, the criterion of DDEA is to find the most appropriate

province in Indonesia for constructing wind power plants based on the geographical

and technical structures as input and output indicators for DDEA modal calculation.

The indicators in this level are mentioned as below:

Electricity consumption: This criterion based on the consumption of the

electricity which have been recorded in every province by Giga Watt per Hour (GW/h)

including general activity of electricity usage which are shown on Figure 3.9.

Fig.3.9 Data of electricity consumption in Indonesia’s Provinces

Natural disaster: The probability occurrence of natural disasters in the region

have significant impact on wind farm site selection. The damage caused by natural

disasters such as flooding, volcanic eruption, earth quakes, and land sliding have

menaces effect on site selection. Figure 3.10. shows the data of natural disasters in

provinces of Indonesia. These disasters may accumulate extra cost of maintenance, thus

increasing the maintenance and operational overheads of wind power plants. Selection

of the safe sites is very core fundamental in decision making for selecting locations for

new facility. These four main parameters (i.e., flooding, volcanic eruption and earth

quakes, land sliding) are included in the list of natural disasters as input indicator.

Fig.3.10 Data of natural disaster in provinces

Wind velocity: The wind velocity is the most important and primary criteria

which must be included in the model. Every province has different wind rate based on

the geographical features. Areas with greater wind velocity are the most suitable

locations for economic growth of energy generation. Based on the data on Figure 3.11

shows that several provinces have varieties of wind velocities. Low wind speed (LWS)

and high wind speed (HWS) are based on different configurations such as wind

resources, aerodynamics, and structural design/ analysis [17]. The aerodynamics loads

are smaller per unit length for the LWS blades but the increased span means that total

forces are closer or larger than the equivalent HWS blade. Due to that for construction

a wind farm in Indonesia should use the technology namely low speed wind turbines.

The design on LWS and HWS blades differ in the blade’s lengths and the magnitude of

aerodynamic loads [17]. Average wind speed in provinces on Indonesia are including

in low to medium of wind speed according to Table 3.3. Low and medium wind speed

sites are mostly classified on Class II-IV. The design of low speed wind farm mostly

based on the blades structural design where blades typically lengthened versions up to

39 m. The materials for modern wind blades are primarily glass fiber reinforced

polymer structures.

Fig.3.11 Gravity loading; a. full blade; b. spar-only simplification

A wind turbine blade in low wind speed is a cantilever which is shown on Figure

3.11. Gravity loading causes edgewise bending, as illustrated in Figure 3.12, the

direction shows reverse twice per full rotation and on the maximum loading condition

as flap-wise bending when the wind direction is perpendicular on the blade [17].

Fig.3.12 Blade loading cases; a. edgewise bending; b. flap-wise bending

Fig.3.13 Data of wind velocity in provinces of Indonesia

Table 3.3 Wind class definitions [17]

Total area: Every province has different land use activities such as industrial

zones, housing schemes, available landscapes with respect to the total area of the

province. The province which has more spacious area is preferred. Because of the

importance of the available land it is considered as a great influential indicator in site

selection as output parameter.

Fig.3.14 Data of total area in provinces in Indonesia

Population by province in Indonesia: Generally, higher population in a

province is preferred hence it implies a higher supply for electricity. Areas with more

human being are given priority in order to minimize the cost of energy distribution to

the far or bounder places.

In this study have a multi-criteria approach to find the most suitable location for

establishment of wind farms. Based on this study, the concept of location as efficiency

in sub-region is defined for wind plants location. Figure 3.15. presents the flow chart

of this proposed study. The proposed study is starting by defining of input and output

Parameter (m/s)

Class III Class IV

Average wind speed

factors that already mention before, finding the data and analyst it, then measuring the

amount of the decision-making unit to verify the amount of input and output. Location

analysis by Dual Data Envelopment Analysis (DEA) in two level and combine it to get

the rank of the DEA results, after that would like to validate the result of DEA using

principal component analysis model to verify the significant influence of the criteria to

the DMU rank. The location optimization of the wind plant by DEA model are shown

by the most suitable location in sub-region of province in Indonesia.

Fig.3.15 Flow Chart of the Proposed Study

The propose methodologies which are used in this research topics are explaining

briefly as below.

This research considering multi-criteria approach to find the most suitable

location for founding a wind farm. Hence, this propose study consider location based

on the provinces and districts in Indonesia and find the appropriate location. Every

province and district in here become the decision-making units (DMUs) which are used

to measure efficiency score. Data envelopment analysis is a non-parametric and

multivariate method to measure DMUs implementaion. In this research, the method

based on the hierarchical dual form of DEA (DDEA) is used. The DMUs are calculated

using a mathematical method as Linear programming using empirical data of inputs

and output, then measure the performance scores using the ratio between input and

output to compare the performance scopes generated. The measure of efficiency for

DMU is given by following linear programming [18].

: the efficiency of DMU,

: weight given to DMU,

: input of DMU i,

: output of DMU r.

In this section, the hierarchical of total efficiency scores from two levels of

DDEA is illustrated. In Level 1, all districts are considered, where the districts in

province k is represented by a set J

using index j

. Level 2 as provinces level, where

K provinces, and each province is given by a subscript k . Combining results

from both levels consists of three steps [9], which are explained as follows:

Step 1: Normalize in Level 1 by scaling each efficiency value by the average efficiency

J represents the number of members in set

Step 2: Calculate the combine efficiency by multiplying the scaled value of

efficiency of Level 2 (

Step 3: Scaling the value of

is the total score between two levels.

In the scope of the study, wind turbine site selection which is one of the most

important problems related to sustainable energy have many alternatives and multi

criteria decision-making. Due to the uncertainty of decision makers in criteria choices,

the hesitant decision-making approach based on hesitant fuzzy linguistic term sets

(HFLTS) is chosen for solution of the complex problem. The upper bound and lower

bound weight ratio between criteria are used for calculating on hierarchical primary

data envelopment analysis. The algorithm hesitant fuzzy linguistic term sets are

proposed by Yavuz et al [19] which provides the capability to deal with hesitancy of

decision makers in assessment. The main of HFLTS is aim to advance flexibility and

completeness of linguistic importance based on the fuzzy linguistic approach.

Linguistic term is relating to language name which is used mostly in fuzzy to define the

uncertainty relation. Context free grammar such as at most, between and so on is the

figure for dealing with uncertainty relation. This algorithm combines the linguistics

term sets with context free grammar to handle the complexity of multi-criteria problems

with hierarchical structure using fuzzy approach.

The steps of the algorithm are shown as below,

Step 1. Defining the linguistic term sets S .

S = {no importance ( n ), very low importance ( vl ), low importance ( l ), medium

importance ( m ), high importance ( h ), very importance ( vh ), absolute importance ( a )}.

Step 2. Defining the context-free grammar G

= {lower than, greater than, at least, at most, between, and}.

Step 3. Collecting the preference relations provided by experts (

Step 4. Transforming the preference relations into HFLTS.

Step 5. Obtaining the envelope between pesimistic and optimistic preference relation

= round assigns in integer number.

Step 6. Computing the pessimistic and optimistic collective preference by linguistic

aggregation.

Step 7. Build the intervals utilities for the collective preference

Step 8. Normalize the collective interval vector to get the weight scores.

Ten experts from academician, NGO on renewable energy, Integrated energy

and environmental planning and policy of Indonesia, Engineers in wind turbine project

in Indonesia, and Technical officer at ASEAN Center for Energy have been asked to

evaluate the wind turbine site selection criteria in Indonesia using their expertise by

filling the fuzzy questionnaire.

Step 9. For every input and output ( q, r ), the weight ratio v

must be bounded by L

(lower bound) and U

(upper bound) as L

The example of the weight

ratio is the relation on lower bound as pessimistic in district level between land cost and

population in region is 3.00 and upper bound is 4.20 (see Table 4.8). The lower bound

weight ratio is (land cost (LC)/population in region (PinR)) ≥ 3.00. The upper bound

weight ratio is similar as (LC/PinR) ≤ 4.20. The same procedure is carried to all criteria

to calculate the priorities.

The fuzzy set can be in combined into the primary Data Envelopment Analysis

is determined by Amy H.I Lee at al [20]. In the early stage the fuzzy analytic hierarchy

process is applied to extract expert’s questionnaire to set the pairwise comparison

values which have been introduced from step 1 to step 9. The bounded weight ratio is

designed to measure the data envelopment analysis (DEA) efficiency of a specific

DMU. DMU is a unit under evaluation in here as provinces and districts level. The

primary data envelopment analysis can be expressed by [20]:

1,... ..., R

1,...q..., Q

is the weight given to the q -th input and

u is the weight output to the

r -th output.

is the amount of the q -th input of the k’ -th DMU,

Y is the amount of

the r -th output. Q is the number of inputs and R is the number of outputs and K is the

Reducing the number of variables under study and consequently ranking and

analysis of decision-making units (DMUs). The objective of PCA [12] is to reduce

ineffective indicators and also as a ranking methodology for determination the

efficiency of different units from the results of DEA. Discussing about principal

component analysis in here using IBM SPSS for knowing the importance of component.

The illustration how to find the importance criteria is applied in district level. The first

step, knowing how many components to extract in the analysis and looking on the Scree

plot by going to Analysis menu then dimension reduction and choose factor analysis is

illustrated in Fig. 3.16.

Fig.3.16 Extraction of Factor analysis in district level

Scree plot help to look for how many components should be extracted is shown

in Fig 4.17.

Fig.3.17 Scree Plot of district level

Looking at the break seems to be at about after the first three components so the

first three components definitely look like meaningful legitimate components and then

there’s a specific estrade component and it looks like the third component might be

something worth extracting. There is a more sophisticated approach to evaluate how

many components should extract an analysis in parallel analysis.

Based on the scree plot, will be extracted total for three eigen values consisting

of two eigen values which have values greater than 1 and one eigen value close to 1

from the analysis that’s why have to do it in three steps to analyze again in dimension

reduction. Choosing analyze with correlation matrix due to the variable are measured

in different units, this implies normalizing all variables using division by their standard

deviation is given in Fig. 3.18.

Fig.3.18 Extraction Box

The next step is chosen descriptive box and checklist on Coefficient, KMO and

Bartlett’s test of sphericity, and Univariate Descriptive is shown in Fig. 3.19. Going to

get the descriptive box to look at correlation matrix on Coefficient and KMO and

Bartlett’s test of sphericity as ferocity to tell whether should actually be doing of

component analysis to begin with and would typically want to look at univariate

descriptive x in any case.

Fig.3.19 Descriptive Box

Rotation Box is chosen for the next step is illustrated in Fig. 3.20. Choosing

Direct Oblimin as the rotation method. Direct Oblimin is an approach to produce an

oblique factor rotation that means the factors solution can be actually correlated with

each other and mostly used as familiar. If the factor solution is the most appropriate an

orthogonal uncorrelated effective solution then yield can be shown as a more or less

oblique orthogonal factor solution. Correlations between the three components that

have been extracted.

Fig.3.20 Rotation Box

Another options that’s good is wanting to sort the components factor loadings

more accurately. In this case to be sorted by size which makes it much easier to interpret

a component pattern matrix is given in Fig. 3.21 on Options Box.

Fig.3.21 Options Box

After interpreting the results, the significance criteria are obtained by principal

component analysis are used to measure the efficiency of the location both on district

level and province level. The multi-criteria approach based on the hierarchical Dual

Data Envelopment Analysis in Sub Section 3.2.1 and 3.2.2 are used to measure

efficiency score.

Results and Discussion

In the proposed hierarchical Dual Data Envelopment Analysis model, 33

provinces at Level 2 and 165 districts at Level 1 in Indonesia are used to define DMUs

for wind farm sites. The data are collected from the Statistical Department of Indonesia,

Internal Ministry of Indonesia, Indonesian Agency for Meteorology, Climatology, and

Geophysics, and The National Land Agency of Indonesia. Overall data are mentioned

in the Appendix A. Measuring the data assessment based on DDEA and hierarchical

methods from section 3.2.1 and 3.2.2. Level 1 calculates for measuring the performance

score for districts level. Level 1 becomes the basic level for combining with the score

from Level 2 where is provinces level.

The score of efficiency at the provincial level are shown in Table 4.1. The

province efficiency represents the priority of each province based on the location

Table 4.1 Efficiency and ranking of provinces (Level 2)

Efficiency No Province

East Kalimantan

East Nusa Tenggara

Southeast Sulawesi

South Sumatra

DI Yogyakarta

West Kalimantan

Central Sulawesi

West Sulawesi

West Sumatra

North Maluku

North Sulawesi

Central Kalimantan

DKI Jakarta

South Sulawesi

South Kalimantan

Riau Islands

Central Java

West Nusa Tenggara

North Sumatra

Bangka Belitung Islands

The 165 districts efficiency and rankings as Level 1 from 33 provinces of

Indonesia are given in Table 4.2. It shows that the most suitable district for establishing

a wind power plant is in Palembang, one of district in province of South Sumatra. The

location of this districts is on the remote of the province, one of the public facilities as

good transportation infrastructure to ship wind power plant materials by both river and

road transportation is mainly advantages. The geographical location also giving benefit

to the regions due to Palembang is less occur able to natural disasters. The wind rate as

natural resources with a decent average wind speed that can be used for economical

electricity generation.

Table 4.2 Efficiency score of districts (Level 1)

Rnk Province

Lhokseumawe

Subulussalam

Tebing Tinggi

Tanjung Balai

Pematangsiantar 0.190 115

Bukit Tinggi

Singkawang 0.158 129

Bengkayang 0.267 88

Rokan Hilir

Banjarmasin 0.380 63

Sungaipenuh

Barito Kuala 0.621 32

Lubuk Linggau

Rejang Lebong

Kotamobagu 0.164 124

Pangkal Pinang 0.164 125

West Bangka

East Belitung

Tanjungpinang

South Jakarta

Central Jakarta

East Jakarta

West Jakarta

North Jakarta

Tasikmalaya

Kulon Progo

Gunung Kidul

Sorong City

Probolinggo

Both efficiency between Level 1 and Level 2 are combined as hierarchical score

using hierarchical model for two level DMU in sub section of 3.2.2. Hierarchical Score

means the final score of hierarchical for two level of district and province as the

combination of efficiency score by both levels using DDEA. The ranking of province

is based on full efficiency. The result shows that the best location is in The South

Sumatra province, especially in Palembang district. West Papua, Papua, and Maluku

provinces also have high efficiency scores which are shown on Figure 4.1.

Fig.4.1 Hierarchical Score

Table 4.3 Detail of hierarchical score for provinces level

Hierarchical

West Sulawesi Polewali

North Maluku Morotai

Riau Island

DI Yogyakarta Gunung

The methodology for combining the results of two levels between province and

district levels using three steps procedure from section 3.2.2. The hierarchical scores

show the Dual DEA Results based on province level. Firstly, collecting two levels

efficiency. Then using first step to normalize every district with average districts in one

province. Repeating for every province, in here showing for 5 dominant provinces

which will discuss further with another methodology in the following sub section. After

normalizing the district efficiency, then combining with province level by multiplying

it with province efficiency which will give the hierarchical score for every district. The

last step is scaling the hierarchical district score with the highest one in each province

to get the maximize one. Finally, the hierarchical score for each province is determined

by averaging the results from the third step. The detail results are shown in Table 4.4.

Table 4.4 Hierarchical score for five dominant provinces

Pagar Alam 0.136

Prabumulih 0.256

Manokwari 0.301

Due to a lot remaining for the expert to decide with their subjective judgement

and expertise. Ten experts have been informed with the objective information and asked

to fill the significance of decision-making criteria using their expertise. After the

importance degree and the context free grammar are built in the first and the second

steps which are shown in the Table 4.5, then collecting preference relations were

collected from experts. The fuzzy questionnaire based on importance degree and

context free grammar to apply with the criteria in Level 1 and Level 2 are designed. In

here we do not show all relations matrices here, we show one example for all steps. The

illustration here is one of seven main criteria in district level. For province level, we

show the result as well in the following steps to combine the results by HFLTS to get

the hierarchical score using data envelopment analysis method.

Table 4.5 Importance degree and context free grammar on HFLTS

Importance Degree

Context free grammar

No importance (n)

Very low importance (vl)

greater than

Low importance (l)

Medium importance (m)

High importance (h)

Very high importance (vh)

Absolute importance (a)

The expert evaluation data shows in Table 4.6. is the one of expert evaluation

of the main criteria in district level with respect to the goal. Firstly, shows as discreate

sets and then converted to intervals. For example, the first expert preference the land

cost (LC) in relation to population in region (PinR) is “at least low importance” in

relation of linguistic terms and can be expressed in the discreate set as {low importance

(l), absolute importance (a)} as the interval set term [l,a], similarity for all relation term

set between every criteria in one expert linguistic evaluations. These evaluations are

proposed for ten experts for every level. After converting the relations term to interval,

the data were collected to determine envelops based on expert evaluations which are

shown in Table 4.7.

Table 4.6 Pairwise evaluations of one expert in main criteria on level 1

Expert1s Linguistic Evaluations

PinR at most h

Table 4.7 Obtained envelops for HFLTS

In the interval set for every evaluation represent the pessimistic in left hand site

and optimistic in right hand side as [P,O]. In here we show the calculation for

pessimistic and optimistic preference using two operations. The scale of the importance

degree is shown in Table 4.5. to the linguistic terms. Table 4.8. shows the pessimistic

and optimistic values. For instance, we show one of the examples for pessimistic and

optimistic preference by land cost (LC) with respect to Population in Region (PinR)

criteria is calculated as follows:

Pessimistic preference.

Optimistic preference.

Table 4.8 Pessimistic and optimistic preference in district level

3.0 4.2 3.2 4.0 2.7 4.0 1.9 4.2 2.6 4.2 1.4 2.4

3.4 4.4 3.0 3.6 3.7 4.6 3.0 4.1 2.4 2.7

2.0 2.8 1.6 2.6 -

2.1 3.2 2.0 4.3 1.6 3.5 1.0 1.4

2.1 2.8 2.4 3.0 2.8 3.9 -

3.5 4.0 4.7 5.1 1.2 2.2

1.8 4.1 1.5 2.3 1.7 4.0 2.0 2.5 -

3.0 4.0 1.3 2.0

1.9 3.4 1.9 3.0 2.5 4.4 0.9 1.3 2.0 3.0 -

3.6 4.6 3.3 3.5 4.6 5.0 3.8 4.8 4.0 4.5 4.4 4.7 -

The next step is looking for the linguistic intervals. The linguistic intervals are

calculated by using the average of pessimistic and optimistic values. For example, using

in one criterion as land cost (LC) as follows:

(2 1 3 1 4 0 5 4 6 4)

+ + + + + + + + +

(6 2 4 3 4 2 6 5 6 4)

(3.00 3.20 2.70 1.90 2.60 1.40) ,

(4.20 4.00 4.00

4.20 4.20 2.40)

(2.467), (3.833

The linguistic intervals are converted to interval utilities as known as the value

to get the midpoint by the average between pessimistic and optimistic values. The

weight value is obtained by normalizes the midpoint.

The linguistic interval, interval utilities, midpoint and weights of all seven

criteria in district level are given in Table 4.9.

Table 4.9 The linguistic interval, interval utilities, midpoint and weights

interval utilities

Midpoints Weights

m,-.117 h,-.267

m,-.217 h,-.500

After getting the weight ratio for every criterion in district level. We need to

look for the efficiency by using the ratio as constraint of criteria. Table 4.8. is ratio

relation of criteria in for province level is given in Table 4.10.

Table 4.10 Pessimistic and optimistic preference in province level

5.1 5.5 4.3 5.4 3.2

4.2 4.7 3.5 4.7 3.1 5.3

3.9 2.5 3.5

4.2 3.4 4.5 3.5 4.8 3.1

2.5 3.1 2.7

1.4 3.4 2.2

2.5 3.5 2.9 3.5

2.8 4.3 2.7 4.1 3.1 4.6 3.5 4.5

3.4 1.8 3.2

3.8 3.9 2.4 3.3 3.1 3.5

1.5 2.6 2.5 4.6 1.9 3.2 1.9 2.2

2.7 2.9 2.7 3.4

1.2 2.5 1.8 3.5 1.4 2.9 2.7 3.9 3.1 3.3

2.9 1.9 3.7 1.5 2.5 2.5 2.9 2.6 3.3

3.150 3.308 2.342 3.142 2.517 2.267 4.233

Pairwise comparison matrix is performed based on the fuzzy aggregation in

Table 4.8 for district level and Table 4.10 for province level. The constraints show the

lower bound and upper bound values as pessimistic and optimistic priorities in fuzzy

matrix, for showing the example of the priority range in district level as Step 9 in Sub-

Section 3.2.3. The constraints of the priorities for each criterion are given in Table 4.11

for district level and Table 4.12 for province level. Due to the space constraints in here

just for showing one example of the constraints for district level as Palembang district.

South Sumatra as representing for province level. The same procedure is applied for

each region in district and province level to get location efficiency.

Table 4.11 The constraint of the priorities for district level

Table 4.12 The constraint of the priorities for province level

Hierarchical DEA is run to evaluate the total score between district and province

level for wind turbine site selection after getting ratio of weight by HFLTS as seen in

Table 4.13, the result shows that, considering expert judgement on the importance of

significant criteria, South Sumatra as the most appropriate location for establishing

wind turbine power plant, following by west Papua, Papua, Maluku, and East of Nusa

Tenggara, respectively.

Table 4.13 Hierarchical Score for HFLTS

No Province

3.6238 13.1317

Prabumulih 0.0001

Manokwari 0.1218

Based on the scree plot in Fig. 3.17., will be extracted for total three eigen values

consisting of two eigen values which have values greater than 1 and one eigen value

close to 1 from the analysis that’s why have to do it in two steps to analyze again in

dimension reduction. Choosing analyze with correlation matrix due to the variable are

measured in different units, this implies normalizing all variables using division by their

standard deviation.

Fig.4.2 Correlation matrix on district level

Looking at the correlation on Fig 4.2 between Primary and Secondary road have

positive correlation 0.795 they seem to hang together when the primary road is needed

in wind turbine site criteria, Secondary can also necessary. But there is also some

negative correlation such as Land cost and population which are different in usual but

cannot expect too much such a thing as slightly not significant. We can see a lot of

positive correlation mostly on Primary, Secondary, Tertiary and Total Cost of

Infrastructure that very consistent, Overall have a lot positive correlation but there are

also have some negative correlation that’s not going to be real straight forward one

component extraction effects on the scree-plot. That true as three component extraction.

Fig.4.3 KMO and Bartlett’s Test on district level

The Bartlett’s test of sphrericity will be non-significant because see on Fig. 4.3.

in this case is statistically significant basically telling that at least one statistically

significant correlation matrix. On the Kaiser-Meyer-Olkin measure of Sampling

Adequacy is also more effect size measure is determining whether use principal

component analysis or not. 0.695 or up to 0.70 or higher is great the lower point is on

less than 0.40 this is the rule time that generally used. Overall on the correlation matrix,

KMO that are over than 0.40 and The Bartlett’s test is statistically significant this will

make confidence to perform the component analysis on district level.

Fig.4.4 Communalities on level 1

The Communalities is output from SPSS that shows the extraction based on

three components being extracted as shown on Fig. 4.4. Communalities represent

variance that have been counted from Component analysis. We can see that ratio of free

usage area have the largest amount of variance that being explained by component

analysis solutions as 99.6% of the significant criteria on wind turbine site selection,

following with total cost infrastructure and tertiary road, respectively. Overall the

communalities are good due to more than 50 % for each criterion.

Fig.4.5 Total variance on district level

The real important thing that should be interpret on column extraction sums of

squared loadings that have been extracted from the three components factor solutions

as given in total variance on Fig. 4.5. These are the eigenvalues 3.375 for the first

component, 1.835 for the second component, and following by the third component is

1.158. Overall the extraction sums of squared loadings have more than 1. The

cumulative percentage of variance and these are the rotated component solution

eigenvalues. SPSS technically calls rotations sums of squared loading as the

eigenvalues in the rotated component solutions which is oblique on scree plot that we

used for the first step of the analysis.

Fig.4.6 Component matrix of district level

The component matrix on Fig. 4.6 of component loadings is basically the

extraction method based on the unrotated solution just for showing the initial value and

we don’t need to interpret it.

Fig.4.7 Pattern matrix on district level

When we have oblique rotated component solution, we really want to use the

pattern matrix are given in Fig. 4.7. The pattern matrix is to help to identify the nature

of the components and what have here in the first component is total cost of

infrastructure is 92.2%, tertiary road is 92.2%, secondary road is 90.3%, and primary

road is 87.8% all loading nicely on this first component so wind turbine site selection

criteria seem to hang together to trade together as significant criteria. But that’s not

always exactly true because these factor loadings are not on all 0.95 but they are high

enough to suggest a pretty strong pattern. These the rest of the listed criteria as land

cost and ratio of free space don’t seem to load very strong only the exception would be

ratio of free space what do you use as a statistically significant component loading.

The second component has two major loadings that are land cost is 91.2% and

population 89.9% then it has negative component loading and if look at total cost of

infrastructure and tertiary road. its component loading to the first component and the

second component its one’s positive and one’s negative either one is very high positive

in first component. Mostly we choose positive value for both components and have a

significant decision or both have difference value on less of negativity.

The third component has one major loading that is ratio of free space area as

highly positive 99.8%. It is totally hanged to trade as significant criteria. So, we can

conclude that majority the percentage of significant criteria have more that 87% as in

the first component is total cost of infrastructure, tertiary, secondary and primary road,

for the second component is land cost and population, and the last component is ratio

of free usage area.

Fig.4.8 Structure matrix on level 1

The last table is the structured matrix which is actually the correlation between

each variable in the analysis and that sequence of respective component from the most

significant criteria to less significant criteria as given in Fig. 4.8.

On the District level we can conclude that have seven significant criteria which

influence on wind turbine site selection in Indonesia such as total cost infrastructure,

tertiary road, secondary road, primary road, population and land cost and we reduce

ratio of free space based on analysis of principal component analysis.

Due to same steps to do analysis using principal component. In province level

we can interpret the results directly.

Fig.4.9 Scree plot for level 2

Based on the scree plot which have been shown on Fig. 4.9, we can see on the

oblique shows on the first three component which have eigenvalue more than 1. Due to

that 3 components look like have meaning components and we can use it on extraction

Fig.4.10 KMO and Bartlett’s Test for Level 2

The KMO and Bartlett’s results test is given in Fig. 4.10 show that for KMO

measure of sampling adequacy is 0.660 is good enough to determine using principal

component analysis. The Bartlett’s test of sphrericity will be non-significant because in

this case is statistically significant basically telling that at least one statistically

significant correlation matrix. Overall on the correlation matrix, KMO that are over

than 0.40 and The Bartlett’s test is statistically significant this will make confidence to

perform the component analysis on province level.

Fig.4.11 Correlation matrix on level 2

Fig. 4.11 shows the correlation matrix on province level. Looking at wind

velocity column we can see that majority have positive correlation such as wind

velocity with population, electricity consumption, earthquake, volcanic eruption and

landslide, but have some negative correlation with total area and flood. Correlation

between wind velocity and population have positive 0.419 as dominant. Also looking

at other column, we can say as generally, overall have a lot positive correlation but

there are also have some negative correlation. It is meaning that more than one

component has extraction effects on the scree plot. We use three components for

extraction analysis.

Fig.4.12 Communalities on level 2

The results of variance being showed in the communalities table is given on Fig.

4.12 We can see that majority have good value of extraction. Population have the

highest amount of variance as 84.2% and wind velocity even have less amount still

more than 10 % for saying as a statistical criterion on 42.3%.

Fig.4.13 Total variance on level 2

The rotation sums of squared loadings are the most important thing to interpret

as the extraction results from three components factor solutions which are given in Fig.

4.13. These are the result 2.666 for the first component, 1.837 for second component

and the last as 1.118 for third component. These results show the position of component

after rotated component as scree plot on the early analysis.

Fig.4.14 Component matrix on level 2

The component matrix shows the initial component as unrotated solution as

given in Fig. 4.14. In here just show the early step before being rotated.

Fig.4.15 Pattern matrix on level 2

The pattern matrix helps to identify oblique rotated component solution is given

on Fig. 4.15. In first component we can see the positive loadings is population,

electricity consumption, earthquake, wind velocity, land slide and volcanic eruption.

The second component have 5 major loadings as volcanic eruption, landslide, flood,

earthquake and population. Wind velocity and electricity consumption have decreased

on few negativities and have not impact on changes. On third components just have two

loadings positive as Total area and earthquake but majority on negativity neither one is

very high. Overall, we can conclude that each criterion has positive loadings even in

just one component and have an impact on solution as the group of components. Due

to that, we can conclude that all of criteria have statistically significant and we do not

need to reduce the criterion.

Fig.4.16 Structure matrix on level 2

As we have discussed the function of structure matrix on Fig. 4.16 to show the

sequence of influence criteria. Looking at last criterion on total area, even on first and

two components give negative loadings but have a great trend on last component as

87% its look like increasing trend and have been impacted by rotated component

Fig.4.17 Component correlation matrix on level 2

Component correlation matrix is correlation between each component based on

rotated component solutions is given in Fig. 4.17. Looking at component 1 shows that

have correlation with component 2. Due to negativity on component 3 so component 1

have not correlation with it. Differently at component 2 have positive correlation for

component 1 and component 3. It is showing the good relation from rotated component

We can conclude that population, electric consumption, earthquake, wind

velocity, flood, volcanic eruption, landslide, and total area as the significant criteria

which are influencing on province level of wind turbine site selection in Indonesia.

Following results of significance criterion, multivariable ranking method namely

Principal component Analysis (PCA) is used for verifying a hierarchical DEA result.

The PCA ranking result is given in Table 4.14.

Table 4.14 Principal Component Analysis Results

The top five are chosen based on the three methods which are discussed.

Hierarchical data envelopment analysis (DEA) is one of multi-variable approach which

can measure efficiency score, in here the result represents the total score of the province

combine with district level. Hesitant fuzzy linguistic term set is used for measuring

uncertainty criteria which can influence in site selection, in this study the judgement,

expertise and advisement by the expert needed to evaluate the importance of the

criterion. After getting the hesitant criteria which have been proven, validation is

needed to validate the significance criteria. The top five suitable locations for

establishing wind turbine power plant in Indonesia are South Sumatra, Papua, West

Papua, Maluku and East Nusa Tenggara provinces, respectively as shown in Fig. 4.18

as geographically location.

Table 4.15 Comparison of three methods result.

The Table 4.15 shows that although using three methods differently still giving

the same priority ranking. Expert judgement can help with multi complex decision and

uncertainty condition. The fuzzy DEA results based on the expert advice giving the

same priority as South Sumatra have the highest priority to build a wind farm. It shows

that the importance criteria relation with specific bound weight are optimal used. The

significance criteria which are obtained based on principal component analysis shows

that the ratio of free usage area and total cost of infrastructure are highly influence to

the results in district level. The ratio of free usage area in South Sumatra is high. It

shows that more space area in one region is advantages. The availability of the

infrastructure of primary road and secondary road in each region have improved in

South Sumatra as primary concern. In this region do not need to build some additional

infrastructure in tertiary road. It can be decreasing the total cost of infrastructure. The

minimize total cost of infrastructure is preferable to influence the efficiency score. In

the province level some criteria such as population, electricity consumption and less of

natural disaster are the most influence to the total efficiency score. The availability

human resources in South Sumatra shows that higher spreading population is in

Palembang district. It can decrease the resources management for the transportation and

accommodation cost of labors. Electricity consumption as the demand in South

Sumatra is high are needed to show the amount of electricity distribution to the center

area. The establishment of wind farm in South Sumatra can help as the alternative

energy resources to fulfill the electricity demand. The third influence criteria in South

Sumatra is less of natural disaster. The South Sumatra as geographically located in the

Sumatra Island based on the occurrence of the disaster shows that in this region have

less of landslide, earthquake and volcanic eruption. Due to the reasons that South

Sumatra can be decided as the most suitable location to build a wind farm power plant.

Fig.4.18. Top five Provinces in Indonesia

Conclusions and Recommendations

Wind energy as natural energy resources is a renewable, freely available and

environmentally compared with other sources of fossil fuel, such as coal, and oil. In

this study, three methods are proposed to decide the most suitable location based on the

multi criteria approach. The hierarchical DDEA to determine the integrated efficiency

scores of DMUs between the district level and the province level. The Fuzzy Data

Envelopment Analysis is used for measuring the bound weight ratio for specific DMUs

based on the expert judgement, advice and expertise. The validation based on principal

component analysis to know the significance of criteria which are influences to the

wind turbine site selection in Indonesia. The possible factors used in the districts level

as defined by land cost, population in region, ratio of free usage area, primary road,

secondary road, tertiary road, and total cost of infrastructure. In the provinces level as

defined by wind velocity, population in province, total area, electricity consumption,

less of land slide, flood, earthquake and volcanic eruption. This method was applied to

33 provinces and 165 districts of Indonesia. The final result shows that the South

Sumatra province has the highest efficiency score which is the most economical

location for constructing a wind farm as given in Fig. 5.1. The most significant criteria

which influence on wind turbine site selection based on principal component analysis

in district level is ratio of free usage area, following by total cost of infrastructure, and

tertiary road. Population, electricity consumption and total area influence in province

level. This study is a milestone for policy makers, government and private stakeholders

in decision making for selecting the most suitable sites for wind power plant

construction in Indonesia. The hierarch DEA results can be used to assist decision

maker on selecting the most suitable wind farm site. The proposed approach can be

considered as an alternative solution, and an early study for policy makers.

Fig.5.1. Full Score of Three Methods

Further improvement could be on criterion specification, which includes social,

environmental, economic, and technical aspects. The final site selection will be more

practical, if opinions from experts, policy makers, government, and private stakeholders

are also considered in the analysis. Collecting more specific data is better approach to

improve the advance analysis in wind turbine site selection.

[1] A. Haydar, E. Senol and K. Eylem, “Multi-criteria selection for a wind

observation station location using analytic hierarchy process,” Renewable

Energy, vol. 29, pp. 1383-92, 2004.

[2] R. Bhatnagar and A. Sohal , “Supply chain competitiveness: measuring the impact

of location factors, uncertainty and manufacturing practices,” Technovation, vol.

25, pp. 443-56, 2005.

[3] D. Afshartous, Y. Guan and A. Mehrotra, “US Coast Guard air station location

with respect to distress calls: a spatial statistics and optimization based

methodology,” European Journal of Operational Research, vol. 196, p. 1086–96,

[4] G. Gamboa and G. Munda, “The problem of wind farm location: a social

multicriteria evaluation frame work,” Energy Policy, vol. 35, p. 1564–83, 2007.

[5] D. Choudhary and R. Shankar, “An STEEP-fuzzy AHP-TOPSIS framework for

evaluation and selection of thermal power plant location: A case study from

India,” Energy, vol. 42, pp. 510-21, 2012.

[6] R. Dutu, “Challenges and policies in Indonesias energy sector,” Energy Policy,

vol. 98, pp. 513-519, 2016.

[7] M. H. Hasan, W. K. Wuzammil, T. M. I. Mahlia, A. Jannifar and I. Hasanuddin,

“A review on the pattern of electricity generation and emission in Indonesia from

1987 to 2009,” Renewable and Sustainable Energy Reviews, vol. 16, no. 5, pp.

3206-3219, 2012.

[8] S. Mujiyanto and G. Tiess, “Secure energy supply in 2025: Indonesias need for

an energy policy strategy,” Energy Policy, vol. 61, pp. 31-41, 2013.

[9] A. Azadeh, S. F. Ghaderi and M. R. Nasrollahi, “Location optimization of wind

plants in Iran by an integrated hierarchical Data Envelopment Analysis,”

Renewable Energy, vol. 36, no. 5, pp. 1621-1631, 2011.

[10] G. Ertek, M. M. Tunç, E. Kurtaraner and K. D, “Insights into the efficiencies of

on-shore wind turbines: A data-centric analysis,” in International Symposium on

Innovations in Intelligent Systems and Applications , Trabzon, 2012.

[11] Ü. Sağlam, “A two-stage data envelopment analysis model for efficiency

assessments of 39 state: wind power in the United States,” Energy Conversion

and Management, vol. 146, pp. 52-67, 2016.

[12] Y. Wu, Y. Hu, X. Xiao and C. Mao, “Efficiency assessment of wind farms in

China using two-stage data envelopment analysis,” Energy Conversion and

Management, vol. 123, pp. 46-55, 2016.

[13] T. Sueyoshi and M. Goto , “Measurement of Returns to Scale and Damages to

Scale for DEA-based operational and environmental assessment: how to manage

desirable (good) and undesirable (bad) outputs,” European Journal of

Operational Research, vol. 211, no. 1, pp. 76-89, 2010.

[14] L. M. Seiford and J. Zhu, “Modeling undesirable factors in efficiency evaluation,”

European Journal of Operational Research, vol. 142, no. 1, pp. 16-20, 2002.

[15] A. Azadeh, A. R. Golkhandan and M. Moghaddam, "Location optimization of

wind power generation–transmission systems under uncertainty using

hierarchical fuzzy DEA: A case study," Renewable and Sustainable Energy

Reviews, vol. 30, pp. 877-885, 2014.

[16] R. Ramanathan, “Comparative risk assessment of energy supply technologies: a

data envelopment analysis approach.,” Energy, vol. 26, pp. 197-203, 2001.

[17] R. H. Barnes, E. V. Morozov and K. Shankar, "Improved methodology for design

of low wind speed specific wind turbine blades," Composite Structures, vol. 119,

pp. 677-684, 2015.

M. Toloo, M. Barat and A. Masoumzadeh, “Selective measures in data

envelopment analysis,” Annals of Operations Research, vol. 226, no. 1, p. 623–

M. Yavuz, B. Oztaysi, S.C. Onar and C. Kahraman, "Multi-criteria evaluation of

alternative-fuel vehicles via a hierarchical hesitant fuzzy linguistic model,"

Expert Systems with Applications, vol. 42, p. 2835-2848, 2015.

A.H.I. Lee, H.Y. Kang, C.Y. Lin and K.C. Shen, "An Integrated Decision-Making

Model for the location of a PV Solar Plant," Sustainability, 7, 13522-41, 2015

Data Resources

Secondary Tertiary

Total Cost of Insfrastructure (Rp)

Land Cost (m^2) (IDR) Population Ratio of Free Usage Area

1 Lhokseumawe

International Samudera Pasee Port

0.000961954

2 Banda Aceh

Ulee Lheue Port

0.000260768

Kuala Langsa Port

0.001471452

4 Subulussalam

Tapak Tuan Port

Sabang Port

0.004018174

Belawan International Port

0.000107485

7 Tebing Tinggi

Kuala Tanjung Port

0.000182583

8 Tanjung Balai

Tanjung Tiram Port

0.000650507

9 Pematangsiantar

0.000200176

10 Padang Sidempuan

Angin Sibolga Port

0.000508371

Indonesian Port II

0.000795235

12 Bukit Tinggi

13 Payakumbuh

14 Pariaman

0.000773586

0.001119644

16 Pekanbaru

0.000739306

0.006147211

Roro sei Paknik Port

0.015205654

19 Rokan Hilir

Bandar Seribu Kubah Port

0.014195962

Tanjung Buton Port

0.020327493

Pelita Jambi Port

22 Sungaipenuh

0.003863805

23 Merangin

12540000000

0.023334964

24 Sarolangun

0.019972805

10140000000

0.019968846

26 Palembang

0.000238504

27 Pagar Alam

28 Lubuk Linggau

0.001928203

29 Prabumulih

0.001339522

0.012681875

31 Bengkulu

Bengkulu Pelindo Port

0.019223685

33 Rejang Lebong

0.006106364

0.011721471

35 Kepahiang

0.004604689

36 Bandar Lampung

Panjang Port

0.000253694

0.000381894

38 Pesawaran

0.004131816

39 Tanggamus

Piers Attorney Port

0.004759526

Mesuji Port

0.007219262

41 Pangkal Pinang

Balam Base port

0.000440483

0.009676137

43 Belitung

Tanjung Pandan Port

0.015064762

44 West Bangka

Muntok Port

0.015695255

45 East Belitung

Belitung Port

0.022880782

0.000932455

Bintan Lagoon International Port

0.009404433

48 Tanjungpinang

Sri Bintan Putra Port

Jagoh Port, Dabo Singkep

0.025916902

Tanjung Batu Port

0.003851233

51 South Jakarta

Tanjung Priok Port

7.30194E-05

52 Central Jakarta

4.69952E-05

53 East Jakarta

6.40404E-05

54 West Jakarta

5.56929E-05

55 North Jakarta

Patimban port

7.16703E-05

0.000120614

58 Sukabumi

0.001701338

59 Tasikmalaya

Cirebon port

7.65235E-05

61 Semarang

Tanjung Mas Port

0.000230531

Kartini Port

63 Pekalongan

Nusantara Port

0.000918491

64 Surakarta

8.33336E-05

65 Magelang

0.000874189

66 Yogyakarta

7.97317E-05

0.000540854

0.000556588

69 Kulon Progo

0.001431459

70 Gunung Kidul

0.001910713

9 Bangka Belitung Islands

71 Surabaya

Tanjung Perak Port

0.000124929

72 Pasuruan

Pasuruan Port

0.000179592

0.000962251

75 Probolinggo

Probolinggo Port

0.001582136

76 Tangerang

9.82831E-05

Karangantu Port

0.001237855

Binuangeun Port

0.003022535

0.000452853

80 Pandeglang

Labuan Port

0.002411539

81 Denpasar

Indonesian, Benoa Port

0.000202178

0.000758174

83 Buleleng

Buleleng Port

0.001693459

0.001878388

85 Klungkung

Nusa Penida Port

0.001486817

Lembar Port

0.000149914

0.006560921

0.011331574

89 East Lombok

0.013183075

6.03268E-05

0.013819754

Atapupu port

Aimere port

0.010114736

95 Southwest Sumba

Waikelo Port

0.004923854

96 Pontianak

0.000165556

97 Singkawang

0.002189248

98 Bengkayang

Teluk Suak Port

0.018115845

Dwikora Port

0.022756128

100 Kubu Raya

0.011666625

101 Palangka Raya

Sampit Port

0.009619972

102 Seruyan

Sigintung Port

103 Gunung Mas

10830000000

0.079523375

104 South Barito

0.072640819

105 Pulang Pisau

0.073659563

106 Banjarmasin

Trisakti Port

0.000113263

107 Banjarbaru

0.001712835

108 Balangan

12810000000

0.015468298

109 Barito Kuala

0.009883012

110 Tabalong

Semayang Port

0.016318037

111 Balikpapan

0.000881824

112 Samarinda

TPK Palaran Port

0.001040055

113 Bontang

Tanjung Laut Port

0.002520514

15300000000

115 Tarakan

Malundung Port

0.000402057

ASDP Manado Port

Bitung Port

0.001386097

118 Tomohon

0.001184512

119 Minahasa

Amurang Port

0.000337316

120 Kotamobagu

0.000564359

Pantoloan Port

0.001099374

122 Parigi Moutong

Tinombo Port

0.011573264

123 Donggala

Donggala Port

0.014808754

124 Banggai

0.027215227

Laut Poso Port

0.029833263

126 Makasar

Soekarno Hatta Makasar Port

Tanjung Ringgit Port

0.001404486

Awerange Port

0.005927867

129 Parepare

Indonesian Port 4

0.004098198

131 Kendari

0.000908998

Murhum Port

0.001452581

Nusantara Raha Port

0.008581761

Kolaka Port

0.016092558

135 Wakatobi

0.005185824

136 Gorontalo City

0.000414754

137 North Gorontalo

Anggrek Port

0.013724984

138 Bone Bolango

0.012621633

139 Pohuwato

Marisa Port

0.031105696

140 Boalemo

Tilamuta Port

0.010754041

Mamuju Port

0.017200453

Palipi Port

0.005775744

143 Polewali Mandar

Tanjung Silopo Port

0.003460092

0.014956338

145 North Mamuju

Pasangkayu Port

0.014791713

0.000802178

Kobi Sadar Port

0.051159097

0.003066602

0.080904472

Namlea Port

151 Ternate

A Yani Port

0.000522286

Sanana Port

0.026924368

153 Morotai

Morotai Ferry Port

0.039281012

Trikora Tidore Port

0.015951479

155 Halmahera

Tongute Port

156 Manokwari

Manokwari Port

0.016740203

Fak Fak Port

0.132854391

159 Sorong City

Sorong Port

0.002411024

160 Bintuni

Bintuni Port

0.276366927

161 Jayapura

0.068786799

162 Merauke

Merauke Port

0.200835771

163 Biak Numfor

Laut Biak Port

0.018800442

Nabire Port

0.068012791

0.071307552

Wind Velocity

Consumption

Volcanic eruption

General Optimization Model in IBM ILOG CPLEX

/*********************************************

* OPL 12.5.1.0 Model

* Author: Galih Pambudi

* Creation Date: Mar 17, 2018 at 11:23:37 AM

*********************************************/

//define indices

//the set of district name

//the set of input criteria

//the set of output criteria

//range of decision making unit in districts

//number of input

//number of output

//define parameter

// Data of inputs

// Data of outputs

// measurement of the DMU efficiency

//assert refDMU in DMU;

//decision variables

// variable of DMU efficiency

// variable of lambda value

//objective function

// minimize input

input constraint

output constraint

// shows report for DMU

"DMU Efficient"

"DMU Not efficient"

"lambda="

// Loop to measure efficiency for all DMU

// To implement Flow Control

thisOplModel

//generating the current model instance

//writeln("District ",dmu);

// modifying

lower bound of output

//one of CPLEX Optimizer’s MP algorithms to solve

postProcess

//to control and manipulatethe

* Creation Date: Mar 17, 2018 at 12:23:37 AM

//total of districts

//DMU reference

//total amount of input criteria

//total amount of output criteria

SheetConnection

"Theses Data.xlsx"

//data connection

with excel data

"I"

//read input table

"OPCA"

//read output table

//teta to SheetWrite(sheetData,SheetWriteConnectionString);

"wo"

"wi"

* Creation Date: Mar 17, 2018 at 18:23:37 AM

//the set of province name

//range of decision making unit in provinces

//writeln("lambda=",lambda);

"Province "

* Creation Date: Mar 17, 2018 at 19:23:37 AM

//total of provinces

"X"

"Y"

Similarity score (Including citation)

• For Students
• Find a Course
• Undergraduate
• Postgraduate
• Lifelong Learning
• Professional Development
• Springboard+
• International Students
• Micro-credentials
• Apprenticeships
• Contact Admissions
• Discover our Impact
• Postgraduate Research
• Research Journey
• Research Support Services
• VP Office for R&I
• Contact Research and Innovation
• Partnering with Us
• Communities
• The TU Dublin Foundation
• Technology Services
• Oifig na Gaeilge
• Campus & Estates
• About the University
• Faculties and Schools
• Our Campuses
• University Calendar
• Policies and Forms
• Governance and Compliance
• Data Protection

Uh-oh! Internet Explorer is out of date.

You are currently running an old version of Internet Explorer that does not support some of the features on this site. For a better experience upgrade to Microsoft Edge.

Or try an alternative browser:

• Student Services & Support
• Examinations

Exam Results

City release dates September 2024

Aungier Street, Bolton Street, Grangegorman Students can access Exam Results once released here .

Please do not attempt to access results ahead of the release date as this may result in your pin being blocked.

An email will be sent to your TU Dublin email address when your results are available for viewing. Your PIN to view your results will be unblocked when your results are available. Please contact your local Examinations Office by email if you have any queries.

Examination Results will be available on Monday 23 September 2024 at 2pm

• Students can access results released here

The Blanchardstown Campus TU Dublin, has completed an upgrade to their Student Record Management System (SRMS) and therefore the Student Self Service Account will change in look, feel, and is accessed by students through single sign on (SSO).

What does this mean for students?

As a TU Dublin Blanchardstown Student, you will have a TU Dublin email account and can access a wide range of Microsoft Office 365 services.

Your Student Self Service Account is now associated with your existing TU Dublin Office 365(email) and the username and password details will allow you to login more easily and securely to your Self-Service Account.

To access results online use your student email and password. If you have any difficulties accessing your student email e.g. if your password has elapsed visit password recovery to update. If you need a password re-activation re-set please contact [email protected] .

Viewing Exam Scripts

Applications have now closed.

Examination Recheck Review

Examination Grade Information

If you have F (Fail) OR NP (Not Present) OR I (Deferred) grades OR a GPA of less than 2.00 you must apply to take/sit supplemental exams.

D Grades are Compensatable Fails. It may be necessary to sit some of your D grades as well as F, NP and I grades in order to pass the year (i.e. to bring your GPA up to 2.0). If you have no F, NP or I grades in any module but have a GPA less than 2.00 you must also apply to repeat some or all of those modules in which you have a D grade in order to pass the stage (i.e. year of course).

CA’s, Project/Placement contacts

It is also necessary to apply for a module even where the Autumn assessment relates to a Placement, Project or Continuous Assessment. You may be required to sit/take the assessment and/or some element of Continuous Assessment, as identified by the co-ordinator of the course.

Details are available from your school office.

Registration for August supplemental exams is now open.

Applying for Repeat Examinations – Closing date no later than Friday 12 th July @ 5pm

• How to Apply for Repeat Examinations

1. If you have D, F and NP grades you must apply online to repeat these by logging back into your examination results and clicking on the re-apply link below your list of examination grades.

2. If you have I grades as well as F or NP grades, you must apply by logging back into your examination results and click on the re-apply link below your list of examination grades.

3. If you only have ' I' (deferral) grades to sit you must apply Autumn Examination /Assessments 2024 Application Form

4. If you only have 'D' grades that you may need to repeat, and no F or NP grades, you must apply Autumn Examination /Assessments 2024 Application Form

When applying online via your exam results webpage, please take care to select the correct course with the Section that applies to you as explained below.  You must tick only one box for each module you wish to take. If you tick the wrong box, we will be unable to record your results correctly. When you have selected the correct course, click Register, then click Make Payment to complete your application.

NOTE: The “Make Payment” button must be clicked to complete the application.

B. Placement, Projects and 100% Continuous Assessments

Remember it is necessary to apply for a module even where you deferred the original sitting and even where the assessment is a Placement, Project or Continuous Assessment. Once you apply as explained in A. above, please check with your School office for details of the Placement, Project or Continuous Assessment.

The fee payable for repeat modules is €100 regardless of the number of modules.  This fee is non-refundable, and must be paid on application. Exam results cannot be entered until payment is made.  There is no charge for deferred sittings.

  D. Exam Dates and Timetables

Supplemental/Repeat exams will be scheduled from 12 – 23 August 2024 (inclusive).

Timetables will be published online by the end of July.

Exam results will be released on Monday , 23rd September 2024 after 2pm

The Tallaght Campus TU Dublin, has completed an upgrade to their Student Record Management System (SRMS) and therefore the Student Self Service Account will change in look, feel, and is accessed by students through single sign on (SSO). 

As a TU Dublin Tallaght Student, you will have a TU Dublin email account and can access a wide range of Microsoft Office 365 services.

• Students can access Exam Results once released here .

To access your results online uses your student email [email protected].

If you have any difficulty accessing your student email (e.g. you need a password/activation re-set), please contact [email protected] .

If you have any queries re registering for the next academic year please contact [email protected] .

Appeal Form