Technical Assistance Consultant s Report

Transcription

1 Technical Assistance Consultant s Report Project Number: November 2015 Republic of the Union of Myanmar: Developing the Asset Management Program for Myanmar Roads (Financed by the Japan Fund for Poverty Reduction) Prepared by the Swedish National Road Consulting AB (SweRoad); the Finnish Overseas Consultants Ltd. (FinnOC); and the Myanmar International Consultants Co. Ltd. (MMIC) For the Ministry of Construction and the Project Management Unit This consultant s report does not necessarily reflect the views of ADB or the Government concerned, and ADB and the Government cannot be held liable for its contents.

2 CURRENCY EQUIVALENTS (as of 30 November 2015) Currency Unit Kyat (MK) MK1.00 = $ $1.00 = MK1,300 NOTE In this report, "$" refers to US dollars unless otherwise stated.

3 Ministry of Construction, Myanmar Asian Development Bank Japan Fund for Poverty Reduction Consulting Services for ADB TA-8327 MYA Developing the Asset Management Program for Myanmar Roads FINAL REPORT November 2015

4 Executive Summary 1. This document forms the Final Report of technical assistance TA-8327: Developing the Asset Management Program for Myanmar Roads, which was financed by the ADB through the Japan Fund for Poverty Reduction. As part of this technical assistance, 27,000 km of trunk roads were surveyed, covering two-thirds of the total trunk road network in Myanmar, including 93% of all paved trunk roads. The survey collected geometric data, GPS data, road roughness data (IRI) and surface damage data. The road roughness and GPS data were collected using smartphone-based Roadroid software that continuously collects GPS coordinates and calculates IRI values using the phone s accelerometer. Surface damages were surveyed through visual assessments. Traffic data was collected from toll gates, other road sector studies and complemented with traffic estimations by local staff of the Ministry of Construction (MOC). Further data on road classification, surface type and other road properties was provided by MOC. All collected data was entered into a Road Data Bank (RDB) that was specifically developed for the trunk road network in Myanmar as part of this technical assistance. The RDB and all the data it contains are now available to MOC to use in their assessments and planning Road length (km) Total Total surveyed Paved (2014 data) Paved surveyed * As a result of security issues, some roads were not surveyed. 2. For purposes of assessing the condition and the budget needs for the entire trunk road network, the collected survey data was extrapolated to the non-surveyed network. The results show that 23% of the trunk road network is in good condition (IRI<4), and 44% is in bad or very bad condition (IRI>8). Paved roads are in much better condition than unpaved roads (32% good, 29% bad or very bad), while asphalt concrete and cement concrete roads are performing better than penetration macadam roads. Roads with higher traffic volumes are generally in better condition, although roads with 1,000-2,500 AADT show a slightly poorer condition as a result of the fact that these are almost exclusively Build-Operate-Transfer (BOT) roads, where the traffic volumes do not provide sufficient toll revenues to finance adequate maintenance. 100 % Condition by surface type (%) 100 % Condition by traffic volume (%) 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % AC/CC PM DBM GR ER Good Fair Poor Bad Very bad Total Note: AC=Asphalt Concrete, CC=Cement Concrete, PM=Penetration Macadam, DBM=Macadam (Metalled), GR=Gravel, ER=Earthen 0 % < >2500 Good Fair Poor Bad Very Bad Total i

5 3. As part of this technical assistance, a strategy analysis was carried out using HDM4 software in order to select appropriate maintenance strategies and predict future road conditions. The strategy analysis clearly shows that the current pavement maintenance budget of MK 100 billion per year is insufficient and will result in the gradual deterioration of the trunk road network. A doubling of the pavement maintenance budget to MK billion per year is required to avoid such deterioration over time. This is shown in the figure below by comparing current conditions of the paved trunk road network in 2015 to expected road conditions in 2020 for both the current pavement maintenance budget of MK 100 billion per year and the proposed increased budget of MK billion per year. The strategy analysis also shows that half this budget increase should be allocated to trunk roads with more than 1,000 AADT, since investments in these roads will have high economic benefits in terms of reduced road user costs. As a result of this focus on high traffic roads, these roads show a marked improvement in condition from the proposed increase in budget. Although roads with less than 1,000 AADT also show improvement, this is not as significant. 100 % Paved roads - All 100 % Paved roads - > 1,000 AADT 100 % Paved roads - < 1,000 AADT 80 % 80 % 80 % 60 % 60 % 60 % 40 % 40 % 40 % 20 % 20 % 20 % 0 % Current budget 2020 Proposed budget Good Fair Poor Bad Very Bad 0 % Current budget 2020 Proposed budget Good Fair Poor Bad Very Bad 0 % Current budget 2020 Proposed budget Good Fair Poor Bad Very Bad 4. Given that almost all roads with more than 1,000 AADT are under BOT contracts (with the exception of the expressway), the budget allocation to pavement maintenance for these high traffic BOT roads should be increased from the current estimated MK 10 billion per year to MK 60 billion per year. To facilitate this, availability payments should be introduced in these contracts, thus ensuring that contractors have sufficient revenue to finance the required maintenance activities. This should go hand-in-hand with the introduction of performance standards to monitor the performance of the contractors and apply penalties in case of poor performance. The rest of the proposed budget increase should be evenly shared between the remaining trunk roads managed directly by MOC (increasing from MK 40 billion to MK 65 billion) and the district and township roads managed by the states and regions (increasing from MK 50 billion to MK 75 billion). 5. This will require increased maintenance budget allocations from the Ministry of Finance (MOF) to MOC (both for the availability payments and for MOC managed works), as well as increased allocations to maintenance by the states/regions from their block grants. MOC needs to enter into discussions with MOF to explain the negative impact of recent budget reductions on trunk road conditions, using the results of the strategy analysis to lobby for an increase of the budget to at least MK 200 million per year for the entire trunk road network (including roads under BOT contracts and managed by states/regions). MOC will also need to review the current BOT contracts, ensuring that the roads under such BOT contracts receive sufficient maintenance investments. 6. In allocating the maintenance budget, the first priority should be to focus on roads with more than 1,000 AADT, placing seals on roads in fair condition and placing overlays on roads in poor condition, followed by the rehabilitation of roads in bad and very bad condition. The second priority are the roads with 200-1,000 AADT, firstly funding periodic maintenance and rehabilitation of roads with 500-1,000 AADT, followed by periodic maintenance of unsealed roads with AADT (e.g. regravelling), and ultimately the periodic maintenance and rehabilitation of sealed roads with 200- ii

6 500 AADT 1. The third priority should be the periodic maintenance of unsealed roads with AADT followed by periodic maintenance and rehabilitation of sealed roads with AADT. Roads with less than 50 AADT should receive only routine maintenance to keep them open. Unsealed Good Fair Poor Bad Very Bad Asphalt concrete Good Fair Poor Bad Very Bad AADT<50 Routine AADT<50 Routine 50<AADT<200 50<AADT<200 Regravel 200<AADT< <AADT<500 Penmac Good Fair Poor Bad Very Bad 1000<AADT<2500 Seal Overlay Rehab AC AADT<50 Routine AADT> <AADT<200 Cement concrete Good Fair Poor Bad Very Bad 200<AADT<500 AADT<50 Routine 500<AADT<1000 Seal Overlay Rehab PM 50<AADT< <AADT< <AADT<500 AADT>2500 AADT>2500 Overlay First Priority, Second Priority, Third Priority, Routine maintenance only 7. Using the network analysis function of HDM4, a maintenance programme was prepared for the period This programme identifies the specific sections of the sealed trunk road network to receive overlays, rehabilitation and upgrading to asphalt concrete, as well as the unsealed road sections to be upgraded to penetration macadam. The proposed maintenance programme covers 6,000 km of works, including 3,000 km of overlays, 2,000 km of rehabilitation of sealed roads, 450 km of upgrading of high traffic penetration macadam roads to asphalt concrete, and 500 km of upgrading of unsealed roads to penetration macadam standard. The total cost of the proposed works comes to MK 842 billion, spread more or less evenly between the different intervention types. Most of the works (71% of the length, 66% of the costs) involve roads managed directly by MOC. Only in case of upgrading of unsealed roads to penetration macadam standard are most of the planned works (78%) in roads managed by the states and regions. This maintenance programme should form the basis for the preparation of annual maintenance programmes, complementing the HDM4 analysis with site surveys of the prioritized roads. The multiyear programme may also form the basis for the planned ADB Highway Network Rehabilitation and Safety Investment Programme that will be prepared in This technical assistance has managed to collect a wealth of data on the trunk road network and enter this in a centralized Road Data Bank. MOC needs to designate this RDB as the central source of trunk road data, making the recently created RAMS units under the Department of Highways responsible for maintaining the RDB and updating the data through annual road condition surveys and traffic counts. The new RAMS units will require at least 9 trained officers and an annual budget to cover the costs of the annual surveys (survey equipment, vehicles and fuel, per diems for surveyors) and the data processing and data analysis (computers and software). The RAMS units will also require further training in these aspects, which will be provided under a follow-up ADB technical assistance Improving Road Network Management and Safety, which is planned to start early Lastly, the technical assistance has prepared a set of key performance indicators (KPI) to assess the performance of the trunk road network. MOC should report on these indicators on an annual basis to show progress over time. Based on the results over the next few years, MOC should also introduce the practice of publishing annual business plans that set short- and medium-term targets for these indicators and provide a description how these targets will be achieved by MOC and others working in the trunk road sector (concession contractors, state/regional governments, internationally financed projects, etc.). 1 Unpaved roads receive priority over paved roads because of the high road user costs on such unpaved roads and the high impact of periodic maintenance in reducing these costs. However, in case of upgrading of these roads to penetration macadam, priority should first be given to existing paved roads. iii

7 Preface 10. This document forms the Final Report of technical assistance TA-8327: Developing the Asset Management Program for Myanmar Roads, that was financed by the ADB through the Japan Fund for Poverty Reduction. The technical assistance was carried out by Swedish National Road Consulting AB (SweRoad / Lead partner), Finnish Oversees Consultants Ltd (FinnOC) and Myanmar International Consultants Co. Ltd (MMIC), in cooperation with the Ministry of Construction (MOC). 11. TA-8327 started in October 2013 and the original project was completed in June A nocost extension was provided up to November 2015 to carry out some additional activities. A total of 35 person-months of international consultants and 31 person-months of national consultants were used under this assignment. The following reports have been published. An Inception Report was published in November The content of the project and the Consultant s plan to carry out the tasks of the project were introduced during the kick off meeting in Naypyitaw on 12 November An Interim Report was published in October 2014 and the review of the interim phase was presented during the workshop in Naypyitaw on 3 October The Interim Report presented in detail the procurement and content of different RAMS components, used survey methods, instructions for using survey data and different data bank functions, mechanisms for road maintenance prioritization, contents of training and capacity building activities, as well as issues related to financing. This Final Report presents the summary of the works carried out by the Consultant during the whole project period of October 2013 November The Final Report concentrates on the results of the surveys and economic analyses as well as recommendations based on these. The Final Report also recommends the next steps and measures needed to guarantee the sustainability of the Road Asset Management System (RAMS). Twelve (12) Monthly Progress Reports were also published throughout the assignment. These present the activities of the Consultant during each reporting period in more detail. 12. The Consultant would like to thank the Ministry of Construction and other ministries and government agencies of Myanmar such as the Ministry of Finance, the Ministry of Transport and the Ministry of National Planning and Economic Development, together with other stakeholders for their cooperation and readiness to provide the assistance and information required. 13. The views expressed in this report are those of the Consultants and they do not necessarily reflect the views of either the ADB or the involved ministries and government agencies in Myanmar. iv

8 Contents Executive Summary... i Preface... iv A. Introduction... 1 B. Road survey and data collection... 1 C. Road data analysis... 5 Administrative classes... 6 Traffic volumes... 7 Pavement types... 8 D. Maintenance standards... 9 E. Funding scenarios F. Maintenance strategies and resulting road conditions Maintenance strategy for budget scenario 1A: MK100 billion restricted Maintenance strategy for budget scenario 1B: MK 100 billion optimized Maintenance strategy for budget scenario 2: MK 250 billion optimized Maintenance strategy for budget scenario 3: MK 400 billion optimized Operational priorities Funding requirements G. Maintenance program H. Key Performance Indicators (KPI) I. Training and capacity building J. Further steps ANNEX 1 Road Data Bank (RDB) ANNEX 2 Survey results by State/Region ANNEX 3 Road Case Matrix ANNEX 4 Road maintenance standards ANNEX 5 HDM4 calibration values ANNEX 6 Road maintenance strategies ANNEX 7 Maintenance programme v

9 Tables Table 1 Surveyed road length by State/Region (km)... 2 Table 2 Administrative classes... 5 Table 3 Traffic volume classes... 5 Table 4 Road surface classes... 5 Table 5 Paved road condition classes... 5 Table 6 Unpaved road condition classes... 5 Table 7 Percentage of surveyed paved road length by pavement width... 9 Table 8 Maintenance standards and unit costs (MK thousand) Table 9 Trunk road maintenance allocations (MK million) Table 10 Road maintenance funding scenarios (MK billion) Table 11 Proposed maintenance program Table 12 Performance indicators for road condition Table 13 Performance indicators for road capacity Table 14 Performance indicators for data collection Table 15 Performance indicators for maintenance implementation Table 16 Performance indicators for maintenance financing Table 17 Training seminars and workshops Table 18 Training materials prepared under the TA Table 19 Vehicle fleet in Myanmar for HDM4 model Table 20 Basic vehicle characteristics for HDM Table 21 Vehicle economic unit costs for HDM Table 22 Vehicle utilisation for HDM Table 23 Costs of fuel and lubricants for HDM4 (MK 1,000) Table 24 Value of time Figures Figure 1 Surveyed road length by state/region (km)... 2 Figure 2 Colour-coded Roadroid roughness results and camera icons around Mandalay and Sagaing... 3 Figure 3 Roadroid roughness results north of Kalaw with corresponding photograph... 3 Figure 4 Screenshot of the Road Data Bank... 4 Figure 5 Road condition by state/region... 6 Figure 6 Traffic volumes, surface types and road conditions by administrative class... 7 Figure 7 Surface types and road conditions by traffic category... 8 Figure 8 Traffic volumes and road conditions by surface type... 8 Figure 9 Proposed maintenance strategy under the MK 100 billion restricted budget scenario Figure 10 Works and investments under the MK 100 billion restricted budget scenario Figure 11 Impact of the MK 100 billion restricted budget scenario on road conditions Figure 12 Proposed maintenance strategy under the MK 100 billion optimized budget scenario Figure 13 Works and investments under the MK 100 billion optimized budget scenario Figure 14 Impact of the MK 100 billion optimized budget scenario on paved road conditions Figure 15 Proposed maintenance strategy under the MK 250 billion optimized budget scenario Figure 16 Works and investments under the MK 250 billion optimized budget scenario Figure 17 Impact of the MK 250 billion optimized budget scenario on paved road conditions Figure 18 Proposed maintenance strategy under the MK 375 billion optimized budget scenario Figure 19 Works and investments under the MK 400 billion optimized budget scenario Figure 20 Impact of the MK 400 billion optimized budget scenario on paved road conditions Figure 21 Investments under the different budget scenarios Figure 22 Works under the different budget scenarios Figure 23 Impact of different budget scenarios on paved road conditions Figure 24 Operational priorities for funding periodic maintenance, rehabilitation and upgrading Figure 25 Lengths and costs of proposed works by state/region Figure 26 Examples of training materials prepared under the TA Figure 27 The Road Data Bank (RDB) Figure 28 Example of road data entered into a spreadsheet before transfer to the RDB Figure 29 Example of road properties for a specific road section Figure 30 Example of road roughness information for a specific road section Figure 31 Examples of reports prepared by RDB Figure 32 Road user costs in Myanmar (MK 1,000/km) vi

10 Abbreviations AADT AC ADB BOT CC DBM DR EH ER GIS GPS GR HDM4 IC IRI JICA KOICA MK MOC NPV OL PM RAMS RDB REHAB RM RS SD TA TR UR US$ Average Annual Daily Traffic Asphalt Concrete Asian Development Bank Build-Operate-Transfer Cement Concrete Dry-Bound Macadam ( Metaled ) District Road Expressway Earthen Road Geographic Information System Global Positioning System Gravel Road or Gravel Layer Highway Design and Management software International Communication Road International Roughness Index Japan International Cooperation Agency Korea International Cooperation Agency Myanmar Kyat Ministry of Construction Net Present Value Overlay Penetration Macadam Road Asset Management System Road Data Bank Rehabilitation Routine Maintenance Region/State Road Surface Dressing / Seal Technical Assistance Township Road Union Road United States Dollar vii

11 A. Introduction 14. Myanmar has a trunk road network of 40,000 km that is managed by the Ministry of Construction (MOC). Investment in the road sector since the 1990s has been limited, with most of this funding going to construction and widening. Maintenance has received very little attention beyond routine maintenance. Although road sector budgets have increased significantly in recent years, MOC lacks the institutional and technical capacity for maintenance planning and budgeting, resulting in inadequate budget allocations to maintenance and suboptimal utilization of available maintenance funding. In response to this, the Asian Development Bank (ADB) has provided the government of Myanmar with technical assistance for developing a Road Asset Management System (RAMS) for the trunk road network. This technical assistance had the following objectives: Prepare a detailed inventory of the trunk road network and its condition Identify the maintenance needs of the different road sections Establish a Road Data Bank to centrally store all trunk road related data Review MOC s approaches to road maintenance planning and implementation Develop an assessment and evaluation system to optimize the budget allocations to different road sections and maintenance activities Prepare a multiyear trunk road maintenance programme Carry out an assessment of MOC s asset management capacity Prepare and deliver a capacity building program in road asset management Suggest a range of possible approaches for funding road maintenance in the future 15. This document forms the Final Report for this technical assistance TA-8327: Developing the Asset Management Program for Myanmar Roads. Section B starts by describing the procedure of road condition surveys and data collection applied by the TA team, while Section C provides a detailed analysis of the current situation of the trunk road system in Myanmar. The report subsequently looks at the assessment and evaluation system developed to optimize budget allocations to road maintenance, with Section D describing the different maintenance activities that were included in the analysis and Section E explaining the budget scenarios that were evaluated (current budget and potential increased budgets). The results of this assessment and evaluation are presented in Section F, describing the impact of the different budgets on the type of maintenance activities to be carried out in the different roads in order to maximise the economic benefits in terms of reduced road user costs. This section also looks at the resulting impacts in terms of road conditions for the trunk road network. Linked to this, Section G presents a multiyear road maintenance programme for periodic maintenance, rehabilitation and pavement upgrading for the period , identifying the specific road sections and maintenance activities to be included. Section H subsequently introduces a set of key performance indicators (KPI) that may be used by MOC to assess the performance of the trunk road network over time. Section I looks at the capacity building and training carried out under this TA, including the training events held and the training materials produced. Finally, Section J looks at the further steps that are needed to institutionalise the road asset management system in MOC, to improve the management capacity of MOC to use the outputs of this analysis for planning and budgeting, and to approve and implement a multiyear road maintenance programme based on the RAMS results. B. Road survey and data collection 16. Before the start of this TA, data for the trunk road network was incomplete and scattered, making it difficult to carry out an analysis of existing road assets and to estimate the maintenance needs. A significant portion of the work carried out under this TA therefore consisted of surveying the trunk road network and collecting existing data from various sources. This process started with the collection of data regarding the length of the trunk road network by administrative class and by surface type. This data was collected from MOC and used as a basis for planning the road surveys. 1

12 17. The road surveys supplied data on road roughness (International Roughness Index - IRI), road length and location (using GPS), surface type, topography, surface damage and traffic. The road roughness and GPS data were collected using smartphone-based Roadroid software 2 that continuously collects GPS coordinates and calculates IRI values using the phone s accelerometer. This was complemented by a visual survey of surface defects according to predefined damage categories 3. Road surface type, road width, topography and traffic estimations were also recorded by the survey teams. 18. Over 27,000 km of trunk roads have been surveyed between February 2014 and March This forms two-thirds of the 40,000 km of trunk roads in Myanmar. Three-quarters of the surveyed roads are paved roads (20,158 km), resulting in nearly all reported 21,360 km paved trunk roads being surveyed (93%). The remaining roads were not included in the survey due to their very poor condition (mainly earthen roads that were inaccessible by car), due to safety issues in the areas concerned, or because they were under construction. In some cases slight differences in length were encountered between the MOC reported lengths and the lengths according to GPS data Figure 1 Surveyed road length by state/region (km) Total Total surveyed Paved (2014 data) Paved surveyed Table 1 Surveyed road length by State/Region (km) State/Region Total length Surveyed length Paved length Surveyed paved length Kachin State* 3,790-0% 613-0% Kayah State* % % Kayin State* 1,914 1,171 61% % Chin State 1,972 1,739 88% % Sagaing State 4,429 3,130 71% 2,124 1,972 93% Tanintharyi State 1,356 1,234 91% % Bago Region 2,212 1,812 82% 1,731 1,711 99% Magway Region 3,523 3,066 87% 2,586 2,460 95% Mandalay Region 2,185 2,077 95% 2,065 1,964 95% Mon State % % Rakhine State 1,865 1,554 83% 1,028 1, % Yangon Region 1, % % Southern Shan State 3,876 2,593 67% 2,118 1,962 93% Northern Shan State* 5,060 2,457 49% 2,211 1,897 86% Eastern Shan State 2,096 1,471 70% % Ayeyarwady Region 2,600 2,015 77% 1,612 1, % Naypyitaw % % Total 40,116 26,963 67% 21,360 19,922 93% * Survey works in these states was restricted by the security situation. Source: Road Data Bank 19. All Roadroid survey data was uploaded to the Roadroid web service that is used to display the survey results using colour codes on google maps (both map and satellite view). With the login data, users can furthermore access the raw data and make different analyses and reports. Photographs made using the Roadroid software on the smartphones are also uploaded and are accessible by clicking on the camera icon on the maps. 2 More information can be found at 3 Defect types included cracking, edge breaks and potholes for paved roads, and corrugation, gravel loss and general damages for unpaved roads. Five defect categories were used: good, fair, poor, bad, very bad. 2

13 Figure 2 Colour-coded Roadroid roughness results and camera icons around Mandalay and Sagaing 4 Good Fair Poor Bad Very Bad Figure 3 Roadroid roughness results north of Kalaw with corresponding photograph Good Fair Poor Bad Very Bad 4 Note the good condition of the Yangon-Mandalay expressway and Yangon-Mandalay highway (green) compared to the Mandalay-Lashio-Muse highway (black). 3

14 20. As part of this technical assistance a Road Data Bank (RDB) was also developed specifically for Myanmar. The RDB includes a commercial off-the-shelf Relational Database Management System (RDBMS) and PostgreSQL. Data in the RDB is organized based on MOC s road classification that distinguishes 739 different roads, which are further subdivided into sections by state/region 5. This classification system was amended in June 2015 with technical support from KOICA. As a result, the RDB is currently being updated to reflect the changes in road codes and numbering. Data from the RDB can be exported to other systems such as HDM4 to make a strategic analysis (regarding budget needs and impact on road conditions), a network level maintenance programming analysis (to prepare a maintenance programme with specific road sections) or a project level analysis (for a specific road section). It is also possible to access the data using different Geographic Information Systems (GIS) for map preparation and visualization of data. Historical data is kept in the RDB, allowing comparison of data from various years. The RDB can also be used to directly produce different types of reports for management. The following ready-made reports are now available and can be expanded upon in the future: List of roads List of road sections Road lengths by administrative class and surface type Traffic data IRI data Surface condition Key performance indicators Figure 4 Screenshot of the Road Data Bank 21. The road condition survey data is transferred to the RDB from the Roadroid system, with visual survey data entered manually into an Excel sheet that is subsequently also transferred to the RDB. The survey data was complemented with data from MOC for the non-surveyed road network regarding length and surface type. The survey data was further complemented with traffic data from toll gates, traffic data obtained from other studies (JICA, KOICA) and traffic estimations received from the MOC district and township offices. 22. Currently the RDB contains information on road inventory, surface condition, road profile and traffic. At a later stage data can be added regarding pavement works, deflection, skid resistance, road accidents, road furniture, maintenance programmes, bridges and culverts. Multimedia files such as photographs and videos can also be added, making it possible to carry out video logging of roads 5 Within the database the roads are subdivided into even smaller sections. 4

15 to allow post-survey reviews of surface defects and other condition parameters from the office. More information on the Road Data Bank is provided in Annex 1. C. Road data analysis 23. Based on the data entered into the Road Data Bank, an analysis of the current status of the trunk road network was carried out. Given Roadroid s inaccuracy of IRI measurements for speeds below 20 km/h, it was decided to remove all roughness measurements where travel speeds were too low or where IRI values were unrealistic (IRI<2.5). This resulted in a reduction of the trunk road length having reliable roughness data to 16,820 km To facilitate the analysis and the subsequent modelling of maintenance needs using the HDM4 software, the different trunk roads were categorized according to their administrative class (C1-C4), the traffic volume (T1-T6), the surface type (P1-P6) and the road condition of paved and unpaved roads (R1-R5). To determine the road condition and traffic volume categories for those roads that were not included in the survey, the data from the surveyed roads was extrapolated to the non-surveyed roads (data on administrative class and surface type provided by MOC was used t guide this process). This has resulted in an overestimation of the road condition and traffic categories, as the non-surveyed roads are generally in poorer condition and have less traffic than the surveyed roads. The lengths of trunk roads in each category are given in the following tables. Data by state and region is provided in Annex 2 (note that this is for the full surveyed road length, before filtering out inaccurate roughness data). Figure 5 shows the road condition by state/region. Table 2 Administrative classes Code Administrative class Number Length % C1 Expressway (EH) % C2 International Communication Roads (IC) + Union Roads (UR) ,693 27% C3 Regional/State Road (RS) 59 4,568 11% C4 District Roads (DR) + Township Roads (TR) ,269 60% Total , % Note: This table is based on the old classification system. Source: MOC Table 3 Traffic volume classes Table 4 Road surface classes Code Traffic (AADT) Length % Code Surface Type Length % T1 <50 17,730 44% P1 Penetration Macadam (PM) 17,362 43% T ,105 30% P2 Cement Concrete (CC) 934 2% T ,051 13% P3 Gravel (GR) 5,566 14% T ,351 3% P4 Metalled (DBM)* 4,625 12% T ,606 4% P5 Earth (ER) 8,565 21% T6 >2500 2,272 6% P6 Asphalt Concrete (AC) 3,064 8% Total 40, % Total 40, % Source: Collected traffic data extrapolated to network Source: MOC * Generally consisting of dry-bound macadam roads Table 5 Paved road condition classes Table 6 Unpaved road condition classes Code Roughness Length % Code Gravel thickness Length % R1 IRI<4 6,769 32% R1 GR>40mm 2,313 12% R2 4<IRI<6 5,651 26% R2 30<GR<40mm 2,574 14% R3 6<IRI<8 2,725 13% R3 20<GR<30mm 2,532 13% R4 8<IRI<10 1,526 7% R4 10<GR<20mm 2,615 14% R5 IRI>10 4,689 22% R5 GR<10mm 8,722 47% Total 21, % Total 18, % Source: Road survey data extrapolated to network Source: Road survey data extrapolated to network 6 Future roughness surveys using Roadroid will need to take account of this limitation. For poor roads where travel speeds are limited, other survey procedures may be necessary (or simply stating that the IRI>10). 5

16 100 % 90 % 80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0 % Figure 5 Road condition by state/region Source: Survey data Good Fair Poor Bad Very bad 25. In order to allow a strategy analysis of the trunk road network to be carried out using HDM4 and to facilitate the analysis of the current situation, the trunk road network has been divided into different road cases. Road cases are sets of roads with similar characteristics that will have similar deterioration patterns and maintenance needs. For Myanmar, road cases are based on a specific combination of traffic class (T1-T6), condition class (R1-R5), administrative class (C1-C4), and surface class (P1-P6). For instance, T6;R1;C1;P2: refers to a section of expressway with more than 2,500 vehicles per day, with an IRI<4 and with a cement concrete surface. The length of trunk road in each road case was determined by populating a road case matrix based on the survey data from the Road Data Bank, extrapolating this to cover the entire trunk road network. Of the 720 possible road cases, only 237 cases were found to include road lengths. A copy of the road case matrix is attached in Annex 3. Administrative classes 26. The expressway, international communication roads, union roads and region/state roads are managed and financed directly by MOC. Some of these MOC roads with higher traffic volumes are under Build-Operate-Transfer (BOT) contracts with private contractors. The majority of trunk roads consists of district or township roads that are managed and financed by the states and regions (60%). 27. Roads with a higher administrative class tend to have higher traffic volumes and more durable surface types. This is not always the case, however, as there are over 100 km of district and township roads with traffic volumes exceeding 1,000 vehicles per day (especially around Yangon and Mandalay), while more than half the International Communication and Union Roads has traffic volumes below 200 vehicles per day (56%). Similarly, 47% of district and township roads have a sealed surface, while 38% of International Communication and Union Roads remains unsealed. This would imply that these roads may not be properly classified, and that the classification should not only take account of the function of the road (what the road connects), but should also consider the importance of the road in terms of traffic volumes. 28. District and Township Roads are generally in worse condition than higher level roads, but this may change rapidly with the significantly increased funding allocations from states/regions. There is not very much difference in terms of surface type or condition between the International Communication and Union Roads on the one side, and the Regional/State Roads on the other, implying that these are treated similarly by MOC. 6

17 100 % 80 % 60 % 40 % 20 % 0 % DR/TR Figure 6 Traffic volumes, surface types and road conditions by administrative class Traffic by class Surface by class Condition by class 100 % 100 % RS < ER GR DBM Good Fair Poor >2500 PM AC/CC Bad Very bad Note: ER=Earthen, GR=Gravel, DBM=Macadam (Metalled), PM=Penetration Macadam, AC=Asphalt Concrete, CC=Cement Concrete Source: Consultant s processing of data from Road Data Bank Traffic volumes IC/UR EH 80 % 60 % 40 % 20 % 0 % DR/TR RS 29. There is a clear move towards sealed and more durable surface types as traffic levels increase, moving from unsealed roads to penetration macadam and ultimately asphalt or cement concrete. Nearly half the trunk road network (47%) remains unsealed, although these roads only serve lower traffic volumes (less than 500 vehicles per day). What is surprising is that there is still a significant portion of roads with traffic volumes of AADT that remain unsealed (26%), while the majority of roads with traffic volumes of only AADT have a sealed surface. This implies that insufficient account is taken of traffic volumes in selecting roads for sealing. 30. Although all roads with more than 500 AADT are sealed, this predominantly involves penetration macadam (62%). The nature of penetration macadam and the often limited construction quality and maintenance results in this pavement type generally having a higher roughness and deteriorating more rapidly than asphalt concrete, especially where traffic volumes are higher. The use of asphalt concrete is still very limited (only 8% of the total network, 14% of the sealed network), even for roads with traffic levels exceeding 1,000 AADT where penetration macadam is considered less suitable. There are nearly 2,000 km of penetration macadam roads with traffic volumes exceeding 1,000 AADT that should urgently be considered for upgrading to asphalt concrete. However, these roads are under BOT contracts where asphalt concrete is only required for traffic volumes exceeding 2,000 vehicles per day More than 40% of the trunk road network is in bad or very bad condition. Over 6,000 km of paved roads (including nearly 1,000 km of asphalt concrete roads) have an IRI>8, indicating that rehabilitation is required. These numbers are likely to be even higher in reality since the survey was limited to the roads in better condition and the roads in poor condition are underrepresented. Road conditions tend to be better for roads with higher traffic volumes, reflecting the greater importance given to improving and maintaining busy roads. An exception is formed by the roads with 1,000-2,500 vehicles per day that are under Build-Operate-Transfer (BOT) contracts. In these roads the low toll revenues are limiting investments, resulting in worse road conditions (25% in bad or very bad condition) than the roads in one traffic category lower (500-1,000 vehicles per day) where investments are mainly from the national and state/regional budgets. BOT roads with traffic levels over 2,500 AADT are also in better condition, reflecting the higher toll revenues for these roads and the ability of contractors to invest more in maintenance. Improved financing arrangements are needed for these important roads with more than 1,000 vehicles per day to ensure that sufficient investments are made and conditions are improved. IC/UR EH 80 % 60 % 40 % 20 % 0 % DR/TR RS IC/UR EH 7 Although required by contract, this is not generally applied in practice. 7

18 100 % 80 % 60 % 40 % 20 % 0 % Figure 7 Surface types and road conditions by traffic category Surface type by traffic volume 100 % Condition by traffic volume 80 % 60 % 40 % 20 % 0 % < (BOT) >2500 (BOT) >2500 (EH) Total < (BOT) >2500 (BOT) >2500 (EH) Total ER GR DBM PM AC/CC Note: ER=Earthen, GR=Gravel, DBM=Macadam (Metalled), PM=Penetration Macadam, AC=Asphalt Concrete, CC=Cement Concrete Source: Consultant s processing of data from Road Data Bank Pavement types Good Fair Poor Bad Very Bad 32. Only 53% of the trunk road network has a sealed surface. By far the majority of sealed roads have a penetration macadam pavement. Even for roads with high traffic volumes (more than 1,000 vehicles per day), penetration macadam makes up half the road length. The quality of these penetration macadam pavements is often limited, due to improper compaction and overheating of the bitumen (using wood fires). Further mechanization of the paving process is required to ensure higher quality and durability of these pavements. Use of more durable asphalt concrete pavements is recommended for higher traffic volumes. 33. Myanmar s trunk road network currently includes nearly 4,000 km of asphalt concrete and cement concrete pavements. These pavements show better performance than the penetration macadam pavements (37% of AC/CC roads in good condition compared to 30% for penetration macadam). Asphalt concrete is often found on highways with high traffic volumes (more than 1,000 vehicles per day), sometimes as AC overlays over an existing concrete pavement. However, more than half the length of asphalt concrete roads actually involves district and township roads with low traffic volumes (less than 500 vehicles per day) Figure 8 Traffic volumes and road conditions by surface type Traffic volume by surface type (%) 100 % Condition by surface type (%) 80 % 60 % 40 % 20 % 0 % AC/CC PM DBM GR ER AC/CC PM DBM GR ER < >2500 Good Fair Poor Bad Very bad Note: AC=Asphalt Concrete, CC=Cement Concrete, PM=Penetration Macadam, DBM=Macadam (Metalled), GR=Gravel, ER=Earthen Source: Consultant s processing of data from Road Data Bank 8

19 34. Cement concrete is used mainly for the Yangon-Mandalay expressway 8, with the remainder mostly involving highways. The majority of these asphalt and cement concrete pavements (58%) are in good or fair condition, but generally because many of them are recent. Nearly a third are already in bad or very bad condition. MOC does not yet have a proper maintenance strategy in place for dealing with AC and CC pavements. 35. Of the nearly 20,000 km of paved roads surveyed, nearly two-thirds are only 12 feet wide, with 24% having an 18 feet wide pavement and only 15% having full double lane standard (24 feet or more). Although this study has not looked at the need for widening, it may be clear from these figures that a large portion of the trunk road network requires widening. This should ideally be combined with rehabilitation or pavement upgrading works. Table 7 Percentage of surveyed paved road length by pavement width State/Region Surveyed 12 feet 18 feet 24 feet or more km km % km % km % Kachin State* - - 0% - 0% - 0% Kayah State* % 35 9% 8 2% Kayin State* % % % Chin State % 12 2% 5 1% Sagaing State 1,972 1,315 67% % 62 3% Tanintharyi State % % 52 7% Bago Region 1, % % % Magway Region 2,460 1,606 65% % 62 3% Mandalay Region 1, % % % Mon State % % % Rakhine State 1,099 1,078 98% 19 2% 2 0% Yangon Region % % % Southern Shan State 1,962 1,519 77% % 111 6% Northern Shan State* 1,897 1,355 71% % % Eastern Shan State % % % Ayeyarwady Region 1,614 1,117 69% % - 0% Naypyitaw Region % % % Total 19,922 12,229 61% 4,721 24% 2,971 15% * Survey works in these states was restricted by the security situation. Source: Road Data Bank D. Maintenance standards 36. Based on the road cases that were defined, different maintenance standards have been identified that may be applied to each road case. These vary from only routine maintenance to periodic maintenance (e.g. seals, asphalt concrete or penetration macadam overlays, regravelling) and rehabilitation. Apart from these normal maintenance standards, the options also included road surface upgrading to penetration macadam or asphalt concrete standard. Each maintenance standard consists of a specific treatment and a trigger when it is to be applied. The standards assume that routine maintenance is applied to all road cases, irrespective of the standard being applied (for the routine maintenance standard, no other maintenance is carried out). 37. The maintenance standards are used by HDM4 to determine the optimal combination of treatments and triggers that results in the greatest net present value (NPV) in terms of net present benefits (economic benefits of reduced road user costs minus increased road agency costs). The different maintenance standards and their respective unit costs are listed below, together with a short description and a definition of the trigger when the maintenance standard is to be carried out. A complete list of the possible maintenance standards for each road case is provided in Annex 4. This only looks at surface type, traffic volume and road condition, since the administrative class is not 8 The expressway was constructed between 2005 and 2010 with a 2-layer, 18 inch thick cement concrete pavement reportedly able to carry 80 tons, providing a 25 foot wide (7.6 meters) two-lane carriageway in both directions. The expressway has some serious design deficiencies, however, resulting in road safety concerns. 9

20 considered to affect maintenance needs (although it may affect the funding and management of maintenance). Table 8 Maintenance standards and unit costs (MK thousand) Routine maintenance Surface Technique Trigger Cost/km* Cost/m 2 (sealed) AC/CC/PM Routine maintenance for sealed road Every year 1,000 (unsealed) DBM/GR/ER Routine maintenance for unsealed road Every year 1,600 Periodic maintenance Surface Technique Trigger Cost/km* Cost/m 2 SD25mm@IRI4 AC/PM 25mm Single surface dressing IRI 4 70, SD25mm@IRI5 AC/PM 25mm Single surface dressing IRI 5 70, OL40mm@IRI4 AC/PM 40 mm Asphalt Concrete overlay IRI 4 100, OL40mm@IRI6 AC/PM 40 mm Asphalt Concrete overlay IRI6 100, OL50mm@IRI4 CC 50 mm Asphalt Concrete overlay IRI4 100, OL50mm@IRI6 CC 50 mm Asphalt Concrete overlay IRI6 100, OL60mm@IRI6 AC/CC/PM 60 mm Asphalt Concrete overlay IRI6 150, OL60mm@IRI8 AC/CC/PM 60 mm Asphalt Concrete overlay IRI 8 150, OL80mm@IRI6 AC/CC/PM 80 mm Asphalt Concrete overlay IRI 6 200, OL80mm@IRI8 AC/CC/PM 80 mm Asphalt Concrete overlay IRI 8 200, PM75mm@IRI6 PM 75 mm Penetration Macadam overlay IRI 6 100, GR@30mm GR/ER 100 mm gravel layer GR < 30mm 45, GR@20mm GR/ER 100 mm gravel layer GR < 20mm 45, GR@10mm GR/ER 100 mm gravel layer GR < 10mm 45, Rehabilitation Surface Technique Trigger Cost/km* Cost/m 2 REHAB AC@IRI8 AC 100 mm Asphalt Concrete + base IRI 8 350, REHAB AC@IRI10 AC 100 mm Asphalt Concrete + base IRI , REHAB PM@IRI8 PM 100 mm Penetration Macadam + base IRI 8 150, REHAB PM@IRI10 PM 100 mm Penetration Macadam + base IRI , REHAB CC@IRI8 CC 2 layer 450mm Cement Concrete + base IRI 8 1,000, REHAB CC@IRI10 CC 2 layer 450mm Cement Concrete + base IRI 10 1,000, Pavement upgrading Surface Technique Trigger Cost/km* Cost/m 2 Upgrade PM to AC PM 100 mm Asphalt Concrete + base IRI 8 350, Upgrade to PM DBM/GR/ER 100 mm Penetration Macadam + base IRI 8 350, Upgrade to AC DBM/GR/ER 100 mm Asphalt Concrete + base IRI 8 600, * For a 24 feet wide road AC: Asphalt Concrete, CC: Cement Concrete, PM: Penetration Macadam, DBM: Dry Bound Macadam (Metalled), GR: Gravel, ER: Earthen Source: TA consultants E. Funding scenarios 38. In order to determine the optimal use of the available maintenance budget and the resulting road conditions, the available maintenance budget needs to first be defined. The budget data for the period is presented in the table below. A significant amount of the maintenance budget comes from state/regional budgets. Although these amounts are budgeted for maintenance, in practice very little is actually spent on maintenance, and then only in district and township roads. The actual amounts spent on maintenance are estimated to be in the order of MK 60 billion from MOC and MK 75 billion from state/regional budgets. Only some two-thirds of these amounts are spent on pavement maintenance, with the remainder being spent on structures and emergency maintenance. For the BOT contracts, data is not available and it is estimated that only MK 10 billion is spent on pavement maintenance. This brings the current pavement maintenance budget for the entire trunk road network to MK 100 billion per year. 10

21 Table 9 Trunk road maintenance allocations (MK million) Road type Budgeted Actual Funding source Total Pavement EH/IC/UR/RS (MOC) 126, , ,880 60,000 40,000 Union budget 69,368 68,297 55,372 60,000 40,000 State/Regional budgets 57, ,436 92, IC/UR/RS (BOT) 10,000 10,000 10,000 10,000 10,000 BOT contractors (estimated) 10,000 10,000 10,000 10,000 10,000 DR/TR (States/Regions) 106, , ,801 75,000 50,000 State/Regional budgets 106, , ,801 75,000 50,000 Total 243, , , , ,000 Source: MOC, TA consultants 39. For the strategy analysis, different alternative budget scenarios are used to identify the impact on maintenance strategies and resulting road network conditions. For this study, three different budget scenarios have been identified that were analysed using the HDM4 software. The first of these is based on current budget allocations as described above (MK 40 billion from MOC, MK 50 billion from states/regions and MK 10 million from BOT contractors), resulting in a total pavement maintenance budget of MK 100 billion per year. For this budget scenario, two alternatives have been analysed. The first alternative (scenario 1A, restricted strategy) is based on current funding sources and allocations to different road types, where the MOC budget is allocated to the Expressway, International Communication and Union Roads, and the Region/State Roads that are not under BOT contracts, while BOT roads only receive funding from BOT contractors, and District and Township roads only receive funding from the states/regions. The second alternative (scenario 1B, optimized strategy) is based on the same total budget, but with an optimized allocation to the different road types. This alternative serves to show the effect of a change in allocation of available funding. 40. Apart from the budget scenarios based on current funding amounts, two additional scenarios have been analysed. These look at the impact of an increase in budget allocations on the overall conditions of the trunk road network. Budget scenario 3 is based on an HDM4 analysis with an unconstrained budget, allowing it to determine the optimal budget allocation resulting in the highest net present value. This resulted in a budget allocation for the first five years of MK 2,080 billion, equivalent to MK 400 billion per year. Budget scenario 2 has been selected in the middle between budget scenarios 1 and 3, and aims to look at the effects of a more modest increase in maintenance budget to MK 250 billion per year. Table 10 Road maintenance funding scenarios (MK billion) EH/IC/UR/RS DR/TR Total Total Budget scenario MOC BOT States/Regions (1 year) (5 years) Scenario 1A, Restricted Strategy (current budget) Scenario 1B, Optimized Strategy (current budget) Scenario 2, Optimized Strategy (middle budget) ,250 Scenario 3, Optimized Strategy (unconstrained budget) ,000 Source: TA consultants F. Maintenance strategies and resulting road conditions 41. This section looks at the results of the strategy analysis carried out by the TA team using HDM4. HDM4 is a software that models the deterioration of the road surface over time, based on variables such as starting condition, surface type, traffic volume, topography and climate. For an analysis of maintenance strategies related to different budget scenarios, the modeling is carried out on the basis of road cases instead of individual roads. For Myanmar, each road case consists of a set of roads with similar surface type, traffic volume, administrative class and starting condition (see also Section C). For each road case, a set of possible maintenance standards are defined, indicating what type of maintenance intervention will be carried out, and at what moment (see also section D). 11

22 42. HDM4 subsequently models the road conditions over time for each possible maintenance standard and for each road case, and calculates the impact on agency costs (the costs of carrying out the intervention as well as possible savings in routine maintenance) and on road user costs (possible savings to vehicle operating costs, travel time costs, accident costs). HDM4 then calculates the net present value (NPV) of the net economic benefits (savings in road user costs minus additional agency costs) for each maintenance standard. It does this for all possible road cases. 43. In order to properly reflect the economic road user costs in Myanmar, the TA team has carried out a calibration of HDM4. The calibrated values used in HDM4 for Myanmar are presented in Annex 5. A calibration of the road deterioration models was also carried out, including a specific study to properly model the deterioration of the penetration macadam pavements that make up the majority of the paved trunk road network 44. Based on the calibrated economic costs and deterioration models, HDM4 looks at the available budget and allocates this to the road cases and maintenance standards that result in the highest net present value for the budget concerned. This means that for small budgets, it will focus on those road cases and maintenance standards that have the highest economic benefits per unit of investment (roads that are not selected will only receive minimum routine maintenance to keep them open). If larger budgets are made available, a greater number of road cases will be selected to receive maintenance (beyond minimum routine maintenance). The greater budget will also allow (more costly) higher maintenance standards to be selected that have a higher net present value, thus increasing the total economic benefits of the available budget. 45. This strategy analysis has been carried out for the three different budget scenarios identified in Section E (MK 100 billion, MK 250 billion and MK 400 billion per year). This section looks at the results of this strategy analysis. The MK 100 billion scenario is modelled first as a budget with restricted allocation (according to the funding source and responsibility for the road) and subsequently as a budget with optimized allocation. For each budget scenario the selected maintenance strategy is presented in terms of the selected maintenance standards for each different road case, and the resulting impact on road conditions is explained. The complete maintenance strategies describing the selected maintenance standards for each road case and each budget scenario, as well as their related costs, are presented in Annex 6. Maintenance strategy for budget scenario 1A: MK100 billion restricted 46. This is the current budget scenario, with a total allocation to pavement maintenance of MK 100 billion per year. As explained in section E, the allocation of this budget is according to the administrative class of the road (district and township roads are financed by state and regional governments), and whether the road is under a BOT contract (in which case the maintenance is funded by the contractor from toll revenue). As a result, HDM4 cannot optimize the budget allocation over the entire trunk road network, but only within the specific set of roads for which the funding is made available. Since the Road Data Bank did not yet have data available on the specific roads under BOT contracts, it has been assumed that all roads with more than 1,000 AADT are BOT roads (with the exception of the expressway that is managed by MOC). This is a slight simplification, but is close to reality. The maintenance standards proposed by HDM4 for the different road cases are presented in Figure 9 below. The results show that higher standards are proposed for roads with higher traffic volumes, since these result in higher economic benefits from savings in road user costs. 47. For the unsealed roads, periodic maintenance is proposed for roads with traffic volumes over 50 AADT as soon as this is considered necessary. For gravel roads this will involve regravelling, while for earthen roads this will involve gravelling and subsequent regravelling. For the metaled dry-bound macadam roads, this will involve extensive repairs to the dry-bound macadam pavement, although in the HDM4 model this has been modelled as regravelling. For roads with fewer than 50 AADT, only routine maintenance is proposed because higher maintenance standards in these roads are not considered economically justifiable due to the limited benefits in terms of reduced road user costs. 12

23 48. For the penetration macadam roads, low volume roads are similarly provided with only routine maintenance. Seals are applied to roads with higher traffic volumes (>200 AADT) that are in fair condition, as an inexpensive way to slow down deterioration. Overlays are proposed for higher volume roads in poor condition, and rehabilitation of the penetration macadam pavement for roads in bad or very bad condition. For the roads with traffic volumes exceeding 1,000 AADT, the seals, overlays and rehabilitation are only proposed for a few roads, with the others receiving only routine maintenance. The reason for this is that these roads are under BOT contracts, and the limited budget available for maintenance does not allow all these roads to receive the treatment they need. As a result, 87% of these high traffic penetration macadam roads receive only routine maintenance, with only 13% receiving the periodic maintenance and rehabilitation required. 49. For the asphalt concrete roads, inexpensive seals are proposed for roads in fair condition, and overlays for roads in poor condition. The high costs of rehabilitation for bad and very bad roads are prohibitive given the limited budget available. For the roads with high traffic volumes (>1,000 AADT), only sealing of roads in good condition is possible as the limited budgets available for these BOT roads do not allow for overlays or rehabilitation. As a result, 71% of the asphalt concrete roads with more than 1,000 AADT receive only routine maintenance. 50. For the cement concrete pavements, the high costs of overlays and rehabilitation make this uneconomical for the lower traffic volumes (<500 AADT), and only routine maintenance is recommended for these roads. Overlays are only recommended for the expressway once it reaches a poor condition. Even for the expressway, the high cost of rehabilitation of cement concrete roads and the limited funding available causes HDM4 to allocate the available funding to other roads. Figure 9 Proposed maintenance strategy under the MK 100 billion restricted budget scenario Unsealed Good Fair Poor Bad Very Bad Asphalt concrete Good Fair Poor Bad Very Bad AADT<50 Routine AADT<50 Routine 50<AADT<200 50<AADT<200 Regravel 200<AADT< <AADT<500 Seal Overlay Penmac Good Fair Poor Bad Very Bad 1000<AADT<2500 AADT<50 Routine AADT>2500 Seal 50<AADT<200 Cement concrete Good Fair Poor Bad Very Bad 200<AADT<500 AADT<50 Seal Overlay Rehab PM 500<AADT< <AADT<200 Routine 1000<AADT<2500 Seal/ Overlay/ Rehab PM 200<AADT<500 AADT>2500 Routine Routine / Routine AADT>2500 Overlay Source: Consultant s processing of data from HDM4 strategy analysis 51. The result of the limited budget and its restricted usage is that more than three-quarters of the trunk road network only receive routine maintenance during , while 20% of the network involving roads in good to poor condition receives periodic maintenance (8,000 km of seals, overlays and regravelling). Only 2% of the network (1,000 km) is rehabilitated. This is also visible in the budget allocation, with three-quarters of the pavement maintenance budget directed at periodic maintenance of roads in good to poor condition 9. Only a quarter of the budget is aimed at rehabilitation of bad or very bad roads. The budget does not allow for any upgrading. Figure 10 Works and investments under the MK 100 billion restricted budget scenario Length of works (km) Amount of investment (MK billion) BOT MOC States/Regions BOT MOC States/Regions Routine only Periodic Rehabilitation Upgrading Periodic Rehabilitation Upgrading Source: Consultant s processing of data from HDM4 strategy analysis 9 Note that the budget allocations do not reflect routine maintenance, which all roads are understood to receive. 13

24 52. The impact of the limited budget and its restricted allocation on paved road conditions is shown below 10. Overall, the available budget is insufficient to stop the gradual deterioration of the paved trunk road network. Despite a focus on roads in good to poor condition, the percentage of paved roads in good and fair condition reduces from 58% in 2015 to 44% in Due to the lack of funding for rehabilitation, the percentage of paved roads in very bad condition increases from 22% in 2015 to 30% in This effect is even more evident for the paved high traffic BOT roads, where the lack of funding for pavement maintenance causes a decrease in the portion of roads in good and fair condition from 65% to 20%, and an increase of roads in very bad condition from 14% to 53%. This is also the result of the much quicker deterioration of these roads due to the high traffic volumes, and the poor suitability of the penetration macadam pavements for such traffic volumes. Similar deterioration is visible for the roads financed by MOC and the states/regions, although the impact is more limited due to the larger budgets available. The portion of these roads in good or fair condition reduces from 57% in 2015 to 49% in 2020, while the percentage of roads in very bad condition increases from 23% to 26%. 100 % 80 % 60 % 40 % 20 % Figure 11 Impact of the MK 100 billion restricted budget scenario on road conditions Paved roads - All Paved roads - BOT > 1,000 AADT 100 % 80 % 60 % 40 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad 100 % Paved roads - MOC 100 % Paved roads - States/Regions 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad Source: Consultant s processing of data from HDM4 strategy analysis 53. The impact of this budget scenario is the gradual deterioration of the trunk road network. More importantly, this deterioration is focused mainly in the roads with the highest traffic volumes that are largely under BOT contracts. This is due to the fact that these BOT roads are not generating enough toll revenue to finance proper maintenance. The financing mechanism for these concession contracts needs to be changed, ensuring adequate financing for the maintenance of these important roads. Maintenance strategy for budget scenario 1B: MK 100 billion optimized 54. This alternative budget scenario is based on the same current budget of MK 100 billion for pavement maintenance. However, in this scenario the allocation is optimized between all roads without any restrictions regarding funding sources or responsibilities for different roads (MOC, BOT 10 HDM4 is very suitable for predicting conditions of sealed roads. However, for unsealed roads it is less suitable, also because the roughness can quickly change as a result of weather, traffic and routine maintenance. Therefore the expected future conditions focus on paved roads, with the understanding that part of the maintenance budget goes to unpaved roads. 14

25 contractors or states/regions). As a result, HDM4 proposes to allocate a far greater percentage of the budget to roads with high traffic volumes, which are largely under BOT contracts. The maintenance standards proposed by HDM4 for the different road cases are presented in Figure 12 below. 55. The budget allocation for unsealed roads is reduced since these roads have low traffic volumes and the economic benefits in terms of reduced road user costs are lower. As a result, regravelling and other periodic maintenance is only proposed for roads with more than 200 AADT. All other roads are to receive only routine maintenance. The costs savings are allocated to roads with higher traffic volumes that have higher benefits in terms of reduced road user costs. 56. For the penetration macadam roads there is a stronger focus on placing seals/overlays and carrying out rehabilitation of roads with traffic volumes exceeding 1,000 AADT. These are mostly BOT roads that did not have sufficient budget under the restricted budget scenario 1A to allow for seals/overlays and rehabilitation to be carried out in all roads. However, in this optimized scenario the funding is focused on these high volume roads since they have the highest benefits in terms of road user cost savings, and as a result the penetration macadam roads with AADT receive only routine maintenance in order to free up funds for the roads with high traffic volumes. 57. The same can be seen for the asphalt concrete roads. Where the restricted budget scenario only allowed for seals on good asphalt concrete roads with high traffic volumes due to the lack of funding from BOT contractors, this optimized scenario prioritizes budget reallocation from low volume roads to high volume roads. As a result, nearly all asphalt concrete roads with more than 1,000 AADT receive seals, overlays or rehabilitation. Asphalt concrete roads with less than 500 AADT receive only routine maintenance For the cement concrete pavements, HDM4 proposes only routine maintenance, even for the expressway. The relatively good condition of the expressway and the high maintenance needs of other trunk roads with similar or higher traffic volumes, means that economic benefits are higher in the other roads. Figure 12 Proposed maintenance strategy under the MK 100 billion optimized budget scenario Unsealed Good Fair Poor Bad Very Bad Asphalt concrete Good Fair Poor Bad Very Bad AADT<50 AADT<50 Routine 50<AADT<200 50<AADT<200 Routine 200<AADT<500 Regravel 200<AADT<500 Penmac Good Fair Poor Bad Very Bad 1000<AADT<2500 AADT<50 AADT>2500 Seal Overlay Rehab AC 50<AADT<200 Routine Cement concrete Good Fair Poor Bad Very Bad 200<AADT<500 AADT<50 500<AADT< <AADT<200 Routine 1000<AADT<2500 Seal Overlay Rehab PM 200<AADT<500 AADT>2500 AADT>2500 Source: Consultant s processing of data from HDM4 strategy analysis 59. The result of optimization of the limited budget is a strong shift in budget allocation from the MOC and state/region financed roads that have lower traffic volumes, to the BOT roads that have higher traffic volumes. The total allocation to BOT roads increases from 10% of the budget under the restricted scenario 1A, to 72% under the optimized scenario 1B. Under this optimized scenario, nearly all BOT roads with more than 1,000 AADT (96%) receive periodic maintenance or rehabilitation. This is at the expense of the lower volume roads, where over 94% receive only routine maintenance. Total coverage of periodic maintenance and rehabilitation for the whole trunk road network reaches respectively 4,500 km and 800 km. Periodic maintenance covers 11% of the total trunk road network, but covers 78% of BOT roads with high traffic volumes, reflecting the importance given to these roads due to the high economic benefits from reduced road user costs. The limited size of the available budget means that rehabilitation is still limited to 2% of the trunk road network 11 Note that the surveyed roads did not include asphalt concrete roads with traffic volumes of 500-1,000 AADT, so this was not included as a road case as no data was available. 15

26 (although it covers 18% of the BOT roads with more than 1,000 AADT). The available budget does not allow for any upgrading. Figure 13 Works and investments under the MK 100 billion optimized budget scenario Length of works (km) Amount of investment (MK billion) BOT MOC States/Regions BOT MOC States/Regions Routine only Periodic Rehabilitation Upgrading Periodic Rehabilitation Upgrading Source: Consultant s processing of data from HDM4 strategy analysis 60. The impact of the optimized allocation of the limited budget on paved road conditions is shown below. Overall, the available budget is still insufficient to stop the gradual deterioration of the paved trunk road network. The percentage of paved roads in good and fair condition reduces from 58% in 2015 to 47% in 2020 (slightly more positive than under the restricted scenario). The percentage of paved roads in very bad condition increases from 22% in 2015 to 25% in 2020, which is significantly lower than under the restricted scenario. The main difference is with the BOT roads with more than 1,000 AADT, where the reallocation of available funding causes the portion of roads in good condition to increase from 35% to 71%, with the remainder in fair condition. This is at the expense of the lower volume roads, where the percentage of roads in good and fair condition reduces from 57% in 2015 to 38% in 2020 and the percentage of roads in very bad condition increases from 23% in 2015 to 29% in % 80 % 60 % 40 % 20 % Figure 14 Impact of the MK 100 billion optimized budget scenario on paved road conditions Paved roads - All Paved roads - BOT > 1,000 AADT 100 % 80 % 60 % 40 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad 100 % Paved roads - MOC 100 % Paved roads - States/Regions 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad Source: Consultant s processing of data from HDM4 strategy analysis 61. The impact of this optimized budget scenario is still the general deterioration of the trunk road network due to a lack of sufficient funding. However, the condition of the most important roads with the highest traffic volumes actually improves under this scenario, resulting in significant economic 16

27 benefits for the country. This focus on maintenance of roads with high traffic volumes shows the need to improve the financing mechanisms for the high volume BOT roads, ensuring that contractors invest more in maintenance and in keeping these roads in good or fair condition. Maintenance strategy for budget scenario 2: MK 250 billion optimized 62. This second budget scenario is based on a significant increase in the budget for pavement maintenance from the current MK 100 billion per year to MK 250 billion per year. The budget allocation is optimized without restrictions regarding funding source or maintenance responsibilities. The objective of this scenario is to show the impact of such a budget increase on the selection of maintenance standards and on resulting road conditions. Apart from a focus on high volume roads as seen in optimized budget scenario 1A, the increased budget allows for more attention to also be given to the roads with lower traffic volumes, ensuring that these do not deteriorate over time. The maintenance standards proposed by HDM4 for the different road cases are presented in Figure 15 below. 63. The increased budget allows a greater allocation to unsealed roads, with periodic maintenance of all unsealed roads with more than 50 AADT. Roads with less than 50 AADT still receive only routine maintenance since higher maintenance standards are not economically justified (the costs are higher than the benefits). 64. For the penetration macadam roads the maintenance strategy under the optimized MK 100 billion budget (scenario 1B) is extended to roads with traffic volumes of AADT. Nearly all penetration macadam roads with more than 50 AADT receive periodic maintenance or rehabilitation under this budget scenario (roads in good condition only receive seals if they have more than 1,000 AADT). Periodic maintenance consists of inexpensive seals for roads in good or fair condition and overlays for roads in poor condition. Roads in bad or very bad condition are rehabilitated. The same can be seen for the asphalt concrete roads, although here the high cost of asphalt concrete rehabilitation limits its application to roads with more than 1,000 AADT). 65. The increased budget also allows for investments in the cement concrete roads in the form of asphalt concrete overlays. Roads with medium traffic volumes are allowed to deteriorate to poor condition before the overlays are applied, while for the expressway with its high traffic volumes the overlays are already applied when the road is still in fair condition 12. The very high costs of rehabilitation of cement concrete roads makes this unfeasible under this budget scenario, even for the short expressway sections in very bad condition. Figure 15 Proposed maintenance strategy under the MK 250 billion optimized budget scenario Unsealed Good Fair Poor Bad Very Bad Asphalt concrete Good Fair Poor Bad Very Bad AADT<50 Routine AADT<50 Routine 50<AADT<200 50<AADT<200 Regravel 200<AADT< <AADT<500 Penmac Good Fair Poor Bad Very Bad 1000<AADT<2500 Seal Overlay Rehab AC AADT<50 Routine AADT> <AADT<200 Cement concrete Good Fair Poor Bad Very Bad 200<AADT<500 AADT<50 Routine 500<AADT<1000 Seal Overlay Rehab PM 50<AADT< <AADT< <AADT<500 AADT>2500 AADT>2500 Overlay Source: Consultant s processing of data from HDM4 strategy analysis 66. The result of the increased budget availability is more periodic maintenance and rehabilitation, especially in the roads with lower traffic volumes. Under this budget scenario the percentage of the trunk road network receiving only routine maintenance is reduced to 57% (twothirds of these are roads with less than 50 AADT). The remaining 43% receive surface renewals through periodic maintenance or rehabilitation. Budget allocations to high traffic BOT roads increase only slightly in absolute terms to ensure that all these roads receive periodic maintenance or 12 Note that the survey did not include cement concrete roads with traffic volumes between 500 and 2,500 AADT, and therefore no road cases have been included for these traffic volumes since no data was available. 17

28 rehabilitation (from MK 72 billion under scenario 1B to MK 75 billion per year under this scenario). The share of pavement maintenance funding going to these high traffic BOT roads reduces from 72% of the total budget to only 30% under optimized budget scenario 1B. Most of the increased budget under this second budget scenario is allocated to low traffic roads, allowing far more periodic maintenance and rehabilitation to be carried out (respectively 14,000 km and 3,300 km in total). Figure 16 Works and investments under the MK 250 billion optimized budget scenario Length of works (km) Amount of investment (MK billion) BOT MOC States/Regions BOT MOC States/Regions Routine only Periodic Rehabilitation Upgrading Periodic Rehabilitation Upgrading Source: Consultant s processing of data from HDM4 strategy analysis 67. The impact of the increased budget on paved road conditions is shown below. Overall, the increased budget is now sufficient to stop the gradual deterioration of the paved trunk road network and even improve conditions. The percentage of paved roads in good and fair condition increases from 58% in 2015 to 79% in The percentage of paved roads in very bad condition decreases from 22% in 2015 to 14% in The resulting road conditions for the high traffic BOT roads are more or less the same as under the optimized MK 100 billion scenario, with the portion of these roads in good condition increasing from 35% to 72% and the remainder in fair condition. The greatest impact of the budget increase is in the lower volume roads managed by MOC and the states/regions, where the increased budget allocation results in the percentage of roads in good or fair condition increasing from 57% in 2015 to 75% in The increased allocation to rehabilitation also results in the percentage of these lower volume roads in very bad condition reducing from 23% to 16%. 100 % 80 % 60 % 40 % 20 % Figure 17 Impact of the MK 250 billion optimized budget scenario on paved road conditions Paved roads - All Paved roads - BOT > 1,000 AADT 100 % 80 % 60 % 40 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad 100 % Paved roads - MOC 100 % Paved roads - States/Regions 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad Source: Consultant s processing of data from HDM4 strategy analysis 18

29 68. The impact of this increased budget is the general improvement of the trunk road network. Based on this budget scenario, it is estimated that the pavement maintenance budget needs to be increased to MK billion per year in order to avoid gradual deterioration of the paved road network. This requires increases in pavement maintenance funding from all sources, including MOC (from MK 40 billion to approximately MK 65 billion), states/regions (from MK 50 billion to approximately MK 75 billion) and BOT contractors (from MK 10 billion to approximately MK 60 billion). 69. For the states and regions, additional funding may come from the large road sector budgets currently available, giving priority to maintenance over construction and capacity upgrading. For MOC, increased allocations from the Union Budget are necessary to avoid deterioration of the roads it is responsible for (MOC maintenance budgets have decreased in recent years). The BOT contractors are currently dependent on toll revenue for their maintenance budgets. Although it is possible to increase toll rates, the low traffic volumes in Myanmar will not allow toll revenue to be increased to the level of the MK 60 billion per year required. To achieve the necessary budget increases for the BOT roads, the financing mechanism will need to be changed. This may be done by introducing concession contracts based on availability payments, which are financed partly from toll revenue and partly from Union Budget allocations (or other road user charges). Only by changing the financing mechanisms for these BOT roads can sufficient revenue be guaranteed to allow contractors to increase investments and properly maintain these important roads that form the backbone of the country. Maintenance strategy for budget scenario 3: MK 400 billion optimized 70. This final budget scenario is based on the unconstrained budget as determined using the HDM4 software. If HDM4 is given no budget constraints, it allocates a total budget of MK 2,080 billion during the first 5 years, equivalent to an annual budget of MK 400 billion. This is the budget allocation at which the maximum economic benefits are obtained in terms of net present value (taking into account increased agency costs and reduced road user costs). This budget scenario serves to show the additional benefits that can be obtained from a further increase in the pavement maintenance budget. As the results will show, this increase is almost exclusively allocated to higher maintenance standards, including pavement upgrading. The maintenance standards proposed by HDM4 for the different road cases are presented in Figure 18 below. 71. The budget allocation for unsealed roads is further increased, allowing for the upgrading of all unsealed roads with more than 200 AADT to penetration macadam standard. This will reduce the road user costs for these roads, resulting in transport savings. Even under this unconstrained budget, HDM4 proposes only routine maintenance for roads with less than 50 AADT since any further investments are not economically justified (due to the low traffic volumes, any savings in road user costs are lower than the related increases in agency costs). 72. The penetration macadam roads see a more extensive use of overlays instead of seals, given that overlays are more durable and result in higher long-term economic benefits. Roads with more than 1,000 AADT are upgraded to asphalt concrete once they become bad or very bad and require rehabilitation. A similar result is seen for asphalt concrete roads, although the slower deterioration speed of these pavements and the higher costs of asphalt concrete rehabilitation means that asphalt concrete roads with less than 200 AADT that are in bad or very bad condition receive only routine maintenance. 73. The budget allocations to cement concrete roads are more or less the same as under budget scenario 2, except that the higher budget allocations allow for the rehabilitation of the expressway sections in very bad condition and overlays of bad road sections with AADT. Roads in fair to bad condition receive overlays, whereby roads with lower traffic volumes are allowed to deteriorate to poor or bad condition before the overlays are applied. For the expressway with its high traffic volumes, the overlays are already applied when the road is still in fair condition. 19

30 Figure 18 Proposed maintenance strategy under the MK 375 billion optimized budget scenario Unsealed Good Fair Poor Bad Very Bad Asphalt concrete Good Fair Poor Bad Very Bad AADT<50 Routine AADT<50 Routine 50<AADT<200 Regravel 50<AADT< <AADT<500 Upgrade PM 200<AADT<500 Overlay Penmac Good Fair Poor Bad Very Bad 1000<AADT<2500 Rehab AC AADT<50 Routine AADT> <AADT<200 Cement concrete Good Fair Poor Bad Very Bad 200<AADT<500 Rehab PM AADT<50 Routine 500<AADT<1000 Overlay 50<AADT< <AADT< <AADT<500 AADT>2500 Upgrade AC AADT>2500 Overlay Rehab CC Source: Consultant s processing of data from HDM4 strategy analysis 74. The result of the further increase in budget is mainly seen in the upgrading of 1,280 km of unsealed roads to penetration macadam standard (MOC and state/region roads) and 170 km of penetration macadam roads to asphalt concrete standard (BOT roads). The share of the budget allocated to BOT, MOC and state/regional roads remains largely unchanged compared to budget scenario 2, although the portion allocated to BOT roads reduces further in favor of the roads managed by MOC and the states/regions. The budget amounts, however, are significantly higher in order to finance the proposed upgrading. Nearly half the trunk road network continues to receive only routine maintenance (mainly the roads with less than 50 AADT), while over 13,650 km receive periodic maintenance, 3,370 km are rehabilitated and 1,450 km are upgraded. Figure 19 Works and investments under the MK 400 billion optimized budget scenario Length of works (km) Amount of investment (MK billion) BOT MOC States/Regions BOT MOC States/Regions Routine only Periodic Rehabilitation Upgrading Periodic Rehabilitation Upgrading Source: Consultant s processing of data from HDM4 strategy analysis 75. The impact of this further increase in the pavement maintenance budget on paved road conditions is shown below. Overall, the further increase to the budget results in a more significant improvement of road conditions. The percentage of paved roads in good and fair condition increases from 58% in 2015 to 81% in The BOT roads continue to receive priority due to the high traffic volumes and by 2020 all BOT roads with more than 1,000 AADT will be in good condition. The lower volume roads also show a marked improvement, with the percentage of these roads in good and fair condition increasing from 57% in 2015 to 78% in The increased allocation to rehabilitation and upgrading also results in the percentage of these roads in very bad condition reducing from 23% to 14%. 20

31 100 % 80 % 60 % 40 % 20 % Figure 20 Impact of the MK 400 billion optimized budget scenario on paved road conditions Paved roads - All Paved roads - BOT > 1,000 AADT 100 % 80 % 60 % 40 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad 100 % Paved roads - MOC 100 % Paved roads - States/Regions 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % Good Fair Poor Bad Very Bad 0 % Good Fair Poor Bad Very Bad Source: Consultant s processing of data from HDM4 strategy analysis 76. The impact of this increased budget is a significant improvement of the trunk road network. The HDM4 results show that under such an unconstrained budget, the focus continues to be on the BOT roads with their high traffic volumes, underpinning the need to increase investments in these roads. The results also show the economic viability of upgrading roads to penetration macadam if traffic volumes exceed 200 AADT, and to asphalt concrete if traffic volumes exceed 1,000 AADT. Operational priorities 77. The HDM4 analysis shows that the optimized allocation of the current pavement maintenance budget of MK 100 billion (scenario 1B) results in an increased allocation to BOT roads with high traffic volumes compared to the restricted scenario (scenario 1A), and that this is done at the expense of the low volume roads. It also shows a higher budget allocation to rehabilitation of these high volume roads. These results clearly show the need to ensure proper pavement maintenance financing for roads with more than 1,000 AADT, given the high economic benefits of pavement improvements to these roads. 78. An increase in budget under scenario 2 shows this increase being allocated to lower volume roads managed by MOC and the states/regions, including the rehabilitation of bad and very bad low volume roads. A further increase of the budget under scenario 3 shows this being used almost exclusively to apply higher maintenance standards. This includes the replacement of seals by overlays and the upgrading of roads to higher pavement standards (both upgrading of unsealed roads with more than 200 AADT to penetration macadam standard, and upgrading of existing penetration macadam roads with more than 1,000 AADT to asphalt concrete standard). 21

32 Figure 21 Investments under the different budget scenarios Amount of investment (MK billion) Amount of investment (MK billion) MK 100B MK 100B MK 250B MK 400B MK 100B MK 100B MK 250B MK 400B restricted optimized optimized optimized restricted optimized optimized optimized Periodic Rehabilitation Upgrading BOT MOC States/Regions Source: Consultant s processing of data from HDM4 strategy analysis 79. The optimization of the MK 100 billion budget scenario actually sees the length of road works reduce, as more expensive maintenance standards are applied to the high traffic BOT roads (more overlays and rehabilitation of sealed roads and less regravelling of unsealed roads). The increased budget scenario 2 sees a significant increase in the length of works as a result of this increased budget being allocated to the less expensive maintenance standards for low volume roads. Budget scenario 3 sees little change in the total length of works, with the nature of these works changing and higher maintenance standards being applied (overlays instead of seals, upgrading to penetration macadam instead of regravelling, upgrading to asphalt concrete instead of rehabilitating the existing penetration macadam pavement). Figure 22 Works under the different budget scenarios Length of works (km) MK 100B restricted MK 100B optimized MK 250B optimized MK 400B optimized Minimum Periodic Rehabilitation Upgrading Source: Consultant s processing of data from HDM4 strategy analysis 80. In terms of road conditions, budget scenarios 1A and 1B see a gradual deterioration of the trunk road network. Under the optimized scenario 1B the high volume BOT roads are significantly improved, but this goes at the expense of the lower volume roads that deteriorate more rapidly. Budget scenario 2 sees an overall improvement in road conditions, with significant improvements for lower volume roads compared to scenarios 1A and 1B. Budget scenario 3 continues this trend, resulting in more than half the paved trunk road network being in good condition by

33 100 % 80 % 60 % 40 % 20 % 0 % Figure 23 Impact of different budget scenarios on paved road conditions Paved roads - All 100 % Paved roads - BOT > 1,000 AADT 80 % 60 % 40 % 20 % 0 % (1A) (1B) (2) (3) (1A) (1B) (2) (3) Good Fair Poor Bad Very Bad Good Fair Poor Bad Very Bad 100 % Paved roads - MOC 100 % Paved roads - States/Regions 80 % 80 % 60 % 60 % 40 % 40 % 20 % 20 % 0 % (1A) 2020 (1B) 2020 (2) 2020 (3) Good Fair Poor Bad Very Bad 0 % (1A) 2020 (1B) 2020 (2) 2020 (3) Good Fair Poor Bad Very Bad Source: Consultant s processing of data from HDM4 strategy analysis 81. In terms of operational priorities, the results of the HDM4 analysis clearly shows the need to focus any available maintenance budget firstly on the roads with high traffic volumes (more than 1,000 AADT), giving priority to roads with very high traffic volumes (>2,500 AADT). Within these high traffic roads, the focus should be first on periodic maintenance (overlays and seals) of roads in fair and poor condition, and subsequently on rehabilitation of roads in bad and very bad condition. This is because the net present value of the resulting benefits per unit of investment is greater in case of the overlays and seals. The economic benefits for these high traffic roads are on average at least 3-4 times higher per unit of investment than for roads with traffic volumes of 200-1,000 AADT (they may be 10 times higher in extreme cases). 82. Only once the funding for periodic maintenance and rehabilitation of these high traffic roads has been secured, should funding be allocated to periodic maintenance and rehabilitation of roads with 200-1,000 AADT. Priority should first be given to periodic maintenance and rehabilitation of paved roads with 500-1,000 AADT, followed by regravelling and other periodic maintenance of unsealed roads with AADT, and ultimately paved roads with AADT. The prioritization of unsealed roads over sealed roads is due to the rapid deterioration of these unsealed roads and the related rapid increase in road user costs if maintenance is not carried out in a timely manner. However, when upgrading unsealed roads with more than 200 AADT to penetration macadam standard, priority should first be given to funding periodic maintenance and rehabilitation of existing paved roads with more than 200 AADT. 83. Roads with less than 200 AADT should be given the third priority. Unsealed roads with AADT should receive periodic maintenance (regravelling) before funding periodic maintenance and rehabilitation in paved roads with AADT. Roads with less than 50 AADT should only receive routine maintenance to keep them open and accessible and are therefore given the lowest priority. 84. These operational priorities are presented in the figure below. It must be noted that these are general operational priorities for a budget of MK billion, and that the specific maintenance strategies presented in Annex 6 for the different road cases present a more detailed overview of the priorities for different budget scenarios. 23

34 Figure 24 Operational priorities for funding periodic maintenance, rehabilitation and upgrading Periodic maintenance sealed roads >2,500 AADT Rehabilitation sealed roads >2,500 AADT Periodic maintenance sealed roads 1,000-2,500 AADT 1 st Priority Rehabilitation sealed roads 1,000-2,500 AADT Periodic maintenance + Rehabilitation sealed roads 500-1,000 AADT Periodic maintenance + Rehabilitation unsealed roads AADT Periodic maintenance + Rehabilitation sealed roads AADT 2 nd Priority Upgrading unsealed roads AADT Periodic maintenance + Rehabilitation unsealed roads AADT Periodic maintenance + Rehabilitation sealed roads AADT 3 rd Priority Funding requirements 85. The HDM4 analysis clearly shows that the current pavement maintenance budget of MK 100 billion per year is insufficient, and will lead to the gradual deterioration of the trunk road network. The budget will need to be increased significantly to the order of MK billion per year. The HDM4 results also show the need to focus this budget increase on roads with high traffic volumes, notably the BOT roads with more than 1,000 AADT. Approximately half of the proposed increase in the pavement maintenance budget from MK 100 billion to MK 200 billion will need to be aimed at these high volume BOT roads, increasing the pavement maintenance budget for these roads from MK 10 billion per year to MK 60 billion per year. This may be achieved by introducing concession contracts based on availability payments instead of toll revenue. Such availability payments may be partly financed through an increase in toll revenue, but will largely have to be financed from the Union Budget (or alternative road user charges). 86. The remainder of the necessary increase in pavement maintenance budget will need to be more or less evenly split between MOC (an increase from the current MK 40 billion per year to MK 65 billion per year to be financed from the Union Budget) and the states/regions (an increase from the current MK 50 billion per year to MK 75 billion per year to be financed from the block grants provided to each state and region). The states/regions have very large road sector budgets, and the required additional funding should be easily covered through a reallocation within this budget. The required increases in allocations from the Union Budget to MOC and to the proposed availability payments under the concession contracts may be more complicated, and the national government should contemplate the introduction of alternative road user charges to complement toll road revenue. 87. This TA and other ADB consultants have reviewed the potential for introducing road user charges in Myanmar 13. A fuel tax or levy is considered the most suitable option, with a potential of generating MK billion per year based on a tax or levy of US$ 0.10 per liter 14 and an average 13 An ADB study regarding transport sector financing was prepared in 2015 (Transport Policy Note - Review of Road User Charges), and an inter-ministerial working group is currently discussing this topic. 14 It is important that the fuel levy is maintained in US$ cents, with the price in Myanmar Kyat adjusted accordingly. The fuel levy should not be allowed to devaluate over time due to inflationary pressure. 24

35 consumption of 3 billion liters per year. A heavy vehicle license fee is also highly recommended, in order to finance the repairs of pavement damages caused by such heavy vehicles. This is likely to generate approximately MK 95 billion per year. International transit charges may also be introduced to foreign trucks, generating an estimated MK 60 billion per year 15. A vehicle registration fee already exists for the 5 million registered vehicles in Myanmar (85% motorcycles), and a doubling of the registration fee is likely to generate approximately 25 billion per year 16. These road user charges may together generate approximately MK 550 billion per year, more than sufficient to cover the proposed pavement maintenance budget of MK 200 billion per year. 88. Existing toll rates may also be increased in roads with high traffic volumes, thus increasing toll revenue. This mainly involves concession contracts and this revenue may therefore be used to partially finance the proposed availability payments to the contractors. Tolls on lower volume roads provide very little revenue and should preferably be abolished and replaced by a fuel tax that features a more efficient road user charge with lower collection costs. 89. Road user charges currently flow directly to the Ministry of Finance (except for toll revenues under BOT contracts), without any earmarking for MOC or the road sector. It is recommended to introduce earmarking of these funds to ensure predictability and sustainability of funding for road maintenance. The introduction of a road fund may also be contemplated, but is not strictly necessary at this stage. 90. To cover the maintenance funding needs in the short term, the ADB is planning a Highway Network Rehabilitation and Safety Investment Program with an initial ADB contribution of US$ million (MK billion) over a 5-year period. This may fill the funding gap to a certain extent, providing time to the government for increasing its funding allocations from the Union Budget and for introducing alternative road user charges. 91. Regardless of the pavement maintenance budget available, there will be a need to optimize the allocation of that budget. This section has introduced the proposed maintenance strategies for the different budget scenarios, defining the maintenance standards to be applied to each road case. These strategies should be followed as much as possible in order to maximize the economic benefits of the available budget. Assuming that the government will be able to gradually increase the available pavement maintenance budget to MK billion per year (with support from the planned ADB project in the short term), it will be important to follow the maintenance strategy depicted in Figure 15 and described in detail in Annex 6. G. Maintenance program As part of the HDM4 analysis, a multiyear maintenance program has been prepared using the network programming function of HDM4. This is different from the strategy analysis described in section F, in that it does not look at road cases but at specific road sections. In this type of analysis HDM4 prioritizes specific road sections and the maintenance standards to be applied in those road sections, thus resulting in a list of road sections with corresponding maintenance activities to be applied over the next five years ( ). 93. This analysis has only taken into account the road sections that were included in the survey. Data for which roughness data was missing or considered unreliable, have been removed from the analysis. Since it is not possible to extrapolate this data to other road sections as was done for the road cases, the analysis only covered some 20,800 km of trunk roads (15,270 km of sealed roads and 5579 km of unsealed roads). The analysis has focused on creating a programme consisting of periodic maintenance (overlays) and rehabilitation of sealed roads, pavement upgrading from 15 This would help level the playing field in respect of the proposed heavy vehicle license fee for domestic trucks. 16 However, it is considered difficult to ensure that this funding is made available to the road sector. 25

36 penetration macadam to asphalt concrete standard, and upgrading of unsealed roads to penetration macadam standard (periodic maintenance and rehabilitation of unsealed roads were not included). 94. The resulting multiyear maintenance programme covers just over 6,000 km of works, including 3,000 km of overlays, 2,000 km of rehabilitation of sealed roads, 450 km of upgrading of high traffic penetration macadam roads to asphalt concrete, and 500 km of upgrading of unsealed roads to penetration macadam standard. The total cost of the proposed works comes to MK 842 billion, spread more or less evenly between the different intervention types. Most of the works (71% of the length, 66% of the costs) involve roads managed directly by MOC. Only in case of upgrading of unsealed roads to penetration macadam standard are most of the planned works in roads managed by the states and regions (78%). The distribution of the proposed works and related costs varies strongly between the different states and regions. Although this is to a certain degree influenced by the surveyed road length in each district, traffic levels have a far greater effect, with heavy traffic roads around cities receiving priority. The figure and table below provide an overview of the lengths and costs of the proposed works by state and region. A full list of the specific road sections included in the proposed programme and the related lengths of works and their costs is included in Annex Figure 25 Lengths and costs of proposed works by state/region Length (km) Kachin Kayah Kayin Chin Sagaing Tanintharyi Bago Magway Mandalay Mon Rakhine Yangon Shan S Shan N Shan E Ayeyarwady Naypyitaw Cost (MK billion) Kachin Kayah Kayin Chin Sagaing Tanintharyi Bago Magway Mandalay Mon Rakhine Yangon Shan S Shan N Shan E Ayeyarwady Naypyitaw Overlay Rehabilitation Upgrade Overlay Rehabilitation Upgrade Source: Consultant s processing of data from HDM4 programme analysis Table 11 Proposed maintenance program State/Region Length (km) Cost (MK billion) Overlay Rehab Upgrade Upgrade Upgrade Upgrade Total Overlay Rehab PM AC PM AC Total Kachin Kayah Kayin Chin Sagaing Tanintharyi Bago Magway Mandalay Mon Rakhine Yangon Shan South Shan North Shan East Ayeyarwady Naypyitaw Total 3,002 2, ,

37 State/Region Length (km) Cost (MK billion) Overlay Rehab Upgrade Upgrade Upgrade Upgrade Total Overlay Rehab PM AC PM AC Total MOC 2,514 1, , States/Regions , Total 3,002 2, , Source: Consultant s processing of data from HDM4 programme analysis H. Key Performance Indicators (KPI) 95. The previous sections have described the performance of the trunk road sector in different aspects, both regarding the current situation and regarding the expected future situation under various budget scenarios. Such performance measuring is often applied by road agencies as a management tool to see how well they are doing, and to set targets to be achieved in the mediumand even long-term. The targets are often linked to budget scenarios, and are used as a negotiation tool with the Ministry of Finance to show the potential positive impact that a budget increase may have (as well as the potential negative impact of a budget reduction). Under its terms of reference, the TA team is also responsible for identifying suitable performance indicators for the trunk road sector. 96. One of the most important set of performance indicators for roads is linked to the road condition, as has been widely used in this report. Ultimately it is the aim of the road agency to provide adequate road conditions for the road users. Common indicators in this respect are listed below, together with the current values of these indicators and the expected target values for 2020 under a budget scenario of MK 250 billion. These indicators look at the percentage of roads in good/fair and bad/very bad condition (this is preferable to the average condition that tells very little about the variation in conditions). Additional indicators are added for roads with more than 500 AADT as it is even more important that these roads are in good/fair condition, given the significant economic benefits related to reduced road user costs in roads with high traffic volumes. Table 12 Performance indicators for road condition Performance indicator - Road condition 2015 (Actual) 2020 (Target) Percentage of paved roads in good/fair condition 58% 79% Percentage of paved roads in very bad condition 22% 14% Percentage of paved roads with more than 500 AADT in good/fair condition 67% 98% Percentage of paved roads with more than 500 AADT in bad/very bad condition 18% 2% Source: TA consultants 97. A second set of indicators looks at the capacity of the roads to handle the traffic they have. This looks both at the nature of the road surface and the width of the roads. Proposed indicators are listed below together with the current values of these indicators and the expected target values for 2020 under a budget scenario of MK 250 billion. Targets for pavement upgrading do not show any increase as the MK 250 billion budget does not allow for pavement upgrading. Target values for road widening have not been included as this has not been modeled in this study. However, both indicators have been included here as pavement and capacity upgrading may be carried out using the development budget. Table 13 Performance indicators for road capacity Performance indicator - Road capacity 2015 (Actual) 2020 (Target) Percentage of roads with more than 200 AADT that are paved 87% 87% Percentage of roads with more than 1,000 AADT that have AC/CC pavement 51% 51% Percentage of paved roads that are 18 feet or wider 39% - Percentage of paved roads that are 24 feet or wider 15% - Source: TA consultants 27

38 98. A third set of performance indicators looks at the regular data collection to update the Road Data Bank. This includes the road condition surveys as well as traffic counts. Proposed indicators are listed below together with the current values of these indicators and the proposed targets for Distinction is made between condition surveys and traffic counts (not just traffic estimations) carried out in the complete trunk road network and just in the paved trunk road network. A period of two years is taken as traffic counts and condition surveys need not necessarily be carried out every year. However, for roads with higher traffic volumes it is recommended to repeat the surveys and traffic counts every year. Table 14 Performance indicators for data collection Performance indicator - Data collection 2015 (Actual) 2020 (Target) Percentage of roads condition surveyed in past 2 years 68% 80% Percentage of paved roads condition surveyed in past 2 years 94% 100% Percentage of roads with traffic counts carried out in past 2 years N/A 40% Percentage of paved roads with traffic counts carried out in past 2 years N/A 80% Source: TA consultants 99. A fourth set of performance indicators is related to maintenance implementation. This looks at the percentage of the paved trunk road network receiving pavement renewal each year through periodic maintenance, rehabilitation or pavement upgrading. Without regular pavement renewal, the pavement will deteriorate to a stage where deterioration suddenly accelerates and more costly repairs are needed. The total length of pavement renewal should therefore cover a certain percentage of the network each year to avoid that deterioration proceeds too far. These performance indicators also look at the degree to which the maintenance interventions follow the selected maintenance strategy as defined in this report (or the results of a future strategy analysis). As such it reflects how efficiently the maintenance budget is being used. Table 15 Performance indicators for maintenance implementation Performance indicator - Maintenance implementation 2015 (Actual) 2020 (Target) Percentage of paved roads receiving periodic maintenance - 7% Percentage of paved roads rehabilitated (including pavement upgrading) - 3% Percentage of works in line with HDM4 strategies or plan - 80% Source: TA consultants 100. A final set of performance indicators looks at maintenance financing. This looks at the total (pavement) maintenance expenditure for the trunk road network, as well as the percentage of that expenditure that is covered by road user charges (e.g. tolls, fuel tax, vehicle maintenance fees, etc.). For concession contracts, data regarding expenditure may not be readily available. However, data will still be available regarding the revenue amounts of the contractor (tolls or availability payments) and the percentage of this revenue that comes from user charges (e.g. tolls) or from Union Budget allocations. Table 16 Performance indicators for maintenance financing Performance indicator - Maintenance financing 2015 (Actual) 2020 (Target) (Pavement) maintenance budget (MK billion) Percentage of the (pavement) maintenance budget funded from road user charges 33% 17 50% Source: TA consultants 17 Based on a pavement maintenance expenditure of MK 100 billion and toll related payments to MOC of MK 3.8 billion from BOT contractors, MK 18.9 billion from auctioned tolls and assuming that the full MK 10 billion in pavement maintenance of BOT roads is financed from toll revenue. 28

39 101. Together, these different performance indicators and related targets may guide the actions of MOC and show its performance in achieving its objectives. It is strongly recommended that MOC publish these key performance indicators on an annual basis, showing progress over time. Once some experience with these performance indicators has been obtained, it is recommended that MOC introduce the publication of annual Business Plans for the trunk road network that not only describe the achievements of the previous year, but that also targets to be achieved in the short- and mediumterm. These Business Plans should furthermore include a more detailed description of the activities and improvements to be carried out by MOC with the aim of achieving the targets and improving the performance of the trunk road network, explaining how these are to be financed and implemented. I. Training and capacity building 102. The TA has given a lot of attention to training and capacity building related to road asset management. In doing so, three categories of staff were distinguished, each with their specific knowledge levels and training requirements. These are listed below together with the main skills that they require training in. i. Top managers - These need to understand the concept of road asset management and how to make use of the outputs of the RDB and HDM4 analysis in decision-making. This will help them to justify budget requests and to determine optimal allocations of available budgets to different roads and maintenance activities. a. Assessing strategies and plans prepared using RAMS b. Comparing alternatives and prioritizing c. Budgeting ii. Middle managers - These need to have an overall understanding of the technical setup and hierarchy of the road asset management system. They need to understand the logistics in terms of data collection and data analysis. They need to be able to use the RAMS as a tool for the preparation of maintenance plans and programmes, and in monitoring them. a. Coordinating data collection and surveys b. Coordinating data entry and analysis c. Preparing strategies and plans using RAMS d. Preparing maintenance programmes using RAMS iii. Operational staff - These need to be able to carry out the actual surveys and data collection, the data entry into the RDB and the operation of HDM4. They need to be able to operate the system and produce high quality outputs for higher level staff. a. Planning and carrying out surveys b. Using Roadroid and other survey equipment c. Processing and importing data into the RDB d. Extracting reports from the RDB e. Using HDM4 to analyze data 103. The TA team has focused a lot of its efforts on the operational level, specifically the data collection and surveys, as well as the data processing and data entry into the RDB. The main challenges were the changes to MOC counterpart staff, many of whom were not able to dedicate sufficient time. The TA team started by training 7 officers of the Road Research Laboratory under MOC, who assisted in the initial surveys around Yangon. Subsequently 7 officers in Naypyitaw were appointed as counterparts for the RAMS training, and received training in organizing and carrying out surveys and data processing, including training in the use of Roadroid. They were also involved in carrying out surveys around Mandalay to gain hands-on experience. The counterparts furthermore received practical training in data processing, importing data into the RDB, creating and restoring back-ups, and sectioning roads. However, only two officers were able to dedicate significant amounts of their time to this TA. 29

40 104. MOC counterparts now have a good understanding of survey procedures and can carry out surveys. They are also capable of data processing and data importing into the Road Data Bank. However, they are not yet sufficiently experienced in extracting reports from RDB which is a next training objective. MOC counterparts have also been introduced to HDM4 functions and features, but they are not yet able to use it. Training in the use of HDM4 is also one of the next training objectives In addition to the practical training to the operational level, several training seminars and workshops were organized under the TA. These were mainly aimed at the operational and middle management levels. Training of higher level management has been limited due to the fact that the RAMS was not yet up and running, with the focus under this TA lying on data collection, development of the RDB and calibration of the HDM4 software. Now that this has been largely completed, the next step will be to provide training to higher management on how to use the outputs of the RDB and HDM4 software. Reviews of the TA were presented to the Minister and staff of MOC on two occasions, and a final presentation of the TA results and findings is planned for November Table 17 Training seminars and workshops Workshop Date Participants Developing the Asset Management Program for Myanmar roads November Roadroid and RAMS workshop (Yangon + Naypyitaw) January Road Condition Surveying February Road Asset Management and Road Data Bank workshop March Road Asset Management and Road Data Bank seminar March Road Data Bank and HDM4 May Road Management System Training (HQ + state/regional offices) August Interim Review October Training of trainers October Competence pass October Roadroid and RAMS workshop (Yangon) November Roadroid international workshop November Measuring structural strength May Technical Review Workshop June Final Report Presentation (planned) November 2015 Source: TA consultants 106. As part of the capacity building under the TA, several training materials were developed and updated during implementation. A list of these materials is provided in the following table. Table 18 Training materials prepared under the TA Training material Date Capacity Development and Training Program (draft) February 2014 Instructions to Surveyors April 2014 Roadroid Survey Manual May 2014 HDM4 Manual (distributed only) May 2014 Road Database System Manual June 2014 Multiyear Training Program for MOC October 2014 IRI Data Entry Manual December 2014 Road Roughness Analysis May 2015 Macadam Deterioration Study May 2015 Multiyear Training Program for MOC June 2015 Capacity Development and Training Program (final) June 2015 Technical Review (workshop materials) June 2015 Bilingual wordlist of RAMS terms Source: TA consultants 30

41 Figure 26 Examples of training materials prepared under the TA 107. As part of the technical assistance, the following equipment was purchased and transferred to MOC. The Roadroid licenses are only valid up till the end of 2015 and will need to be renewed. 4 Laptops 5 Roadroid devices (Samsung phones with licensed Roadroid software) HDM4 licenses Sirwey road data bank software MOC has recently created 2 RAMS units under the Road Maintenance Division of the Department of Highways. Once these units have been staffed, the training in surveying and data collection, data processing and data entry into the RDB will need to be repeated for any staff members that have not yet received the training under this TA. These staff members will also be central to the next training objectives regarding the extraction of reports from the RDB and the use of HDM4 to analyze the data. J. Further steps 109. From a situation where data was lacking and was scattered, this technical assistance has developed a Road Data Bank with a wealth of data for the majority of the trunk road network (including 93% of the paved trunk road network). It has also carried out an analysis of this data to determine the optimal maintenance strategies for different budget scenarios and to show that the pavement maintenance budget needs to be doubled to avoid gradual deterioration of the trunk road network. It has furthermore defined a maintenance programme up till 2020 that may form the basis for annual maintenance programmes for the next years. However, this is not the end, and much still needs to be done Road Data Bank - The RDB currently holds information on the entire trunk road network, including condition information for the majority of trunk roads (including 93% of the paved trunk road network). The structure allows for this to be further expanded with other data and for updated yearly data to be added. As such, the RDB forms a source of complete, up-to-date data for planning and analysis. MOC needs to designate the RDB as the main source of data for the RDB, making the Department of Highways through its recently created RAMS units responsible for regularly updating and maintaining it. In the medium term, MOC should aim to put the database online in order to make the data easily accessible for different users, and to ensure that everybody is using the same data Data collection and surveys - The road data needs to be updated and expanded to include a larger portion of the trunk road network (at least all paved roads). Problems with roughness data need to be corrected, resurveying the roads concerned. The use of Roadroid needs to be further tested and calibrated with clear procedures for surveying to avoid problems with the reliability of data in the future (related to travel speeds below 20 km/h). The licenses for the Roadroid software will also need to be renewed after December The equipment used during the surveys may be expanded to also include video logging, thus allowing post-survey analysis and validation of 31

42 pavement defects and road conditions. At a later stage, ground penetrating radar for measuring the thickness of different layers and deflectometers for measuring the structural bearing capacity of the pavements may also be contemplated for main roads. Detailed survey manuals should be prepared to ensure a common approach to the use of such survey equipment. Systems furthermore need to be put in place for regular surveying in future years to ensure data is kept up-to-date (this includes budget allocations for this purpose), since the lack of regular surveying is one of the main causes of failure of a RAMS Data processing and entry - New data needs to be entered into the Road Data Bank and existing data needs to be further checked for errors and consistency. Staff of the new RAMS units needs to be further trained to carry out the data processing and entry, including the extraction of reports from the RDB. The type of reports needs to be further expanded to fit the needs of the higher management staff in MOC. The RDB also needs to be made available online so that it may serve as a single source of data for the trunk road network, avoiding the current situation where data from different sources often conflicts. A simple end-user manual should furthermore be developed for management level to allow them to use the RDB online. The connection between RDB and HDM4 also needs to be improved, to facilitate data transfer Pavement maintenance budget - The strategy analysis carried out as part of this TA has shown that the current funding allocations to pavement maintenance are insufficient and will result in the gradual deterioration of the trunk road network. A doubling of the budget to MK billion per year is required, whereby most of this increase is directed towards trunk roads with more than 1,000 AADT. MOC will need to present these findings to the Ministry of Finance, showing the negative implications of recent maintenance budget reductions, and negotiating for increased funding. MOC will also need to amend the existing BOT contracts, introducing availability payments to ensure that contractors receive sufficient revenue in order to make the necessary investments in pavement maintenance of the roads with more than 1,000 AADT that make up the backbone of the country. Such increased payments should be linked to clear performance standards ensuring that the desired service levels are achieved Annual maintenance programmes - This TA has prepared a multiannual maintenance programme for This may serve as the basis for the preparation of annual maintenance programmes by MOC and by the states/regions (and even the concession contractors). The multiannual maintenance programme presented in this report may also be used as the basis for preparing the planned ADB Highway Network Rehabilitation and Safety Investment Programme, which will provide $ million in ADB loan funding for km of periodic maintenance and rehabilitation of trunk roads. As such it is important that MOC reviews the programme and approves it as the basis for annual maintenance plans for the next 5 years Continued data analysis - The analysis of the road data carried out as part of this TA will need to be repeated regularly. Although the current TA has done most of the calibration of HDM4, further fine-tuning is needed to ensure that the models properly reflect reality. Collection of new road roughness data may help in this process, allowing data from different years to be compared to the estimations prepared by HDM4. Staff in the recently created RAMS units will need training in using HDM4 for preparing strategy analyses, maintenance programmes, and project level analyses. Additional training will also be required in presenting the HDM4 results in a useful manner for use by high level management. It is further recommended to acquire more advanced GIS software to allow the preparation of maps and the visualization of data, ensuring a proper connection to the RDB Future planning and budgeting - This report has given a lot of attention to the development of maintenance strategies for different budget scenarios, and the development of a maintenance program up to The RAMS unit will receive training in the preparation of such plans, but the higher level management will also need training in understanding such plans and adjusting them to fit specific needs and to compensate for changing budgets and priorities Key Performance Indicators (KPI) - It is important that MOC reviews the key performance indicators prepared under this TA and approves a final set of indicators to be used to assess the 32

43 annual performance of the trunk road network. MOC should initially use these indicators to prepare annual reports showing the achievements in the trunk road network. Within the next few years and based on the results achieved in these years, it should start preparing annual business plans, setting short- and long-term targets for these indicators based on the activities it plans to carry out and the political objectives of the government. These business plans should provide descriptions of how these targets will be achieved, specifically mentioning how different activities will be financed These future steps aim to institutionalize the RAMS within MOC, to develop the management capacity of MOC to use the outputs of the RAMS for planning and budgeting, and to approve and implement a multiyear periodic maintenance and rehabilitation program. Some of the steps described above can already be initiated based on the results of this TA: recognizing the RDB as the central data source for the trunk road network, increasing the maintenance budget in line with the strategy analysis, introducing availability payments in concession contracts, and approving the multiyear maintenance programme Other activities will require further support to MOC. The government and the ADB are currently planning a follow-up technical assistance project to TA-8327, which is expected to start early This follow-up TA on Improving Road Network Management and Safety will look specifically at these issues, while also addressing road safety improvements. It is strongly recommended that the consultants under this new TA work directly with the new RAMS units under MOC, preferably being located in the same offices The recently created RAMS units will ultimately be responsible for the sustainability and proper use of the RAMS. They will also be the main recipients of the follow-up technical assistance planned for It is therefore important that these RAMS units be provided with adequate budget and staff. The new RAMS units will require at least 9 trained officers and an annual budget to cover the costs of the annual surveys (survey equipment, vehicles and fuel, per diems for surveyors) and of the data processing and data analysis (computers and software). The officers of the RAMS units should at least include the following members: Director - Reporting and presenting results to higher management, coordination with other planning units Deputy Director - Road numbering, road classification, control of the information system, verification of data, preparing annual reports, etc. Data collection / survey expert - Planning, coordination and supervision of surveys and data collection Data Bank expert - Quality assurance, verification and checking of data, data entry HDM4 expert - Importing data into HDM4, data analysis, reporting HDM4 results GIS expert - Presentation of data and results on maps, visualizing road network problems, updating GIS data for the trunk road network Survey team (3 persons) - Carrying out surveys 33

44 ANNEX 1 Road Data Bank (RDB) 121. The Road Data Bank (RDB) is the central location for all road related data. The RDB can be used to look up specific data on the road (properties, condition, geometry, traffic, etc.), to prepare specific reports directly from the RDB (annual reports, key performance indicators, traffic reports, road condition reports, etc.), to export data for the preparation of GIS maps (based on GPS referenced data), and to export data to a pavement management system such as HDM4 that in turn may be used to prepare multiyear maintenance programmes, strategy analyses (including budget requirements) and project analyses. Figure 27 The Road Data Bank (RDB) Data structure 122. The RDB structure has been created using a commercial off-the-shelf Relational Database Management System (RDBMS) and PostgreSQL. The RDB includes an executable file SROMP.exe, which contains the main road database SROMP (System for Road Operations and Maintenance Programming) that runs on the Microsoft Windows operating system (Windows 7) and that used for road data handling. A detailed Road Database System Manual that presents all the functional features of RDB, was published in June The structure of RDB is based on the administrative classification of trunk roads in Myanmar and existing road lists received from MOC. The trunk roads are further organized by each state/region. The data is stored in separate database tables according to the type of data collection. All road data is identified according to the location referencing system including road and section identification, start and end points of the whole road and of each section. The location is presented by distance from both the starting point of the road and the starting point of the individual road 34

45 section. The locations are also presented by GPS coordinates: latitude, longitude and altitude. The following data is (can be) collected and stored in the RDB (these are described in detail in the Road Database System Manual, paragraph 4.2): Road inventory data Road surface condition data Road profile data Traffic data (Pavement data) (Bearing capacity according to deflectometer measurements) (Skid resistance data) (Accident data) (Road furniture data) (Maintenance program data) (Bridge data) (Culvert data) (Multimedia files) 124. The inventory/section data includes the general properties of the roads and sections such as region/state name, administrative class, road name, road number, section id, start and end points for the whole road and each section, lengths of the whole road as well as for each section in meters and miles/furlongs, width of the carriageway and shoulders, etc. The surface condition data includes information regarding hilliness and surface type, overall condition based on visual estimations, as well as visual estimations for each of the following defects: cracking, edge damage, and potholes for paved roads; damages, corrugation, and gravel loss for unpaved roads. The road profile data includes latitude, longitude and altitude values, vehicle speed data, and IRI values. The traffic data includes average daily traffic data. If more detailed information is available, the data can be categorized by number of cars, buses, small trucks, medium trucks, large trucks and articulated trucks. Data input 125. Data input can be done directly in the RDB, but for large amounts of data it is recommended to import the data form another source. The roughness data from the Roadroid software is transferred into RDB by using Quantum GIS software. Individual road condition points collected by the Roadroid software are modified into 100m aggregated IRI values and saved by road sections in the RDB. Detailed instructions for the process are provided in a separate IRI-Data Entry Manual published in December However, this methodology is only a temporary arrangement for the IRI data processing since later updates to the Roadroid software have made it easier to transfer the data. All visual survey data including traffic information is first entered into a spreadsheet (Google tables, MS Excel, etc.). Here the data is first checked for errors and missing data before importing it into the RDB. For this it is important that the spreadsheet tables have the correct format to enable automatic transfers into the RDB. 35

46 Figure 28 Example of road data entered into a spreadsheet before transfer to the RDB 126. Multimedia files such as photographs or videos can also be entered into the RDB. This enables video logging to be included in future surveys, allowing the visual surveying to take place in the office by specialized staff. All data can be changed and corrected directly in the RDB, including the addition or removal of road sections or the updating of their properties. A road can be divided into smaller sections or different sections can be joined together Historical data can be preserved in yearly databases and is available for comparison of data collected in different years. It must be noted that when data for a new year is imported, new database tables have to be created. The storage of data from different years will enable comparison of data over time. Using the RDB 128. The main screen of the database shows the different roads by state/region and by administrative class. Each road is entered according to its road code. Data from different years can be selected by clicking on the corresponding year in the top left corner of the database. Specific roads can subsequently be selected by clicking on the road concerned, which shows the properties of that road. 36

47 Figure 29 Example of road properties for a specific road section 129. Clicking on one of the buttons at the top of the database provides the corresponding information for the road concerned. This includes information on the pavement, profile, surface condition, bearing capacity, skid resistance, traffic, road furniture, road accidents, maintenance programmes, bridges, culverts, multimedia files and general data (note that many of these do not yet have data entered, although the structure already exists in the RDB and data can be entered as soon as it becomes available). Figure 30 Example of road roughness information for a specific road section 37

48 130. The RDB is also able to prepare automatic reports for different road sections, for a specific set of roads or for the entire network. The following reports are now readily available from the RDB. The type of reports can be further expanded in future to fit the needs of higher management. List of roads List of road sections Road lengths by administrative class and surface type Traffic report IRI report Surface condition report Key Performance Indicators (KPI) Figure 31 Examples of reports prepared by RDB 131. The data from RDB can be exported to other systems such as HDM4 to carry out a strategy analysis, network level maintenance programming or a project-level analysis. The data can also be accessed by different Geographic Information Systems (GIS), which are used for map preparation, data visualization in general and geographic analysis. As part of the planned follow-up technical assistance, the RDB may be integrated with a specific GIS application. Apart from the export to other analyzing tools, the data may also be exported to spreadsheets such as MS Excel for data management and planning purposes. 38

49 ANNEX 2 Survey results by State/Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous 20,384 19,141 94% Concrete % Gravel 5,552 1,147 21% Metalled 4,627 3,539 76% Earth 8,620 2,062 24% Total 40,116 26,963 67% Code Road surface Surveyed (km) % P1/P6 Bituminous 19,141 71% P2 Concrete 780 3% P3 Gravel 1,147 4% P4 Metalled 3,539 13% P5 Earth 2,062 8% Total 26, % Code Administrative class Surveyed (km) % C1 EH 589 2% C2 IC + UR 8,988 33% C3 RS 3,375 13% C4 SR + TR 14,010 52% Total 26, % Code Traffic (AADT) Surveyed (km) % T1 <50 11,295 42% T ,201 30% T ,121 12% T ,052 4% T ,504 6% T6 >2500 1,790 7% Total 26, % Code Road condition Surveyed (km) % R1 Good 9,505 35% R2 Fair 6,519 24% R3 Poor 2,940 11% R4 Bad 1,699 6% R5 Very bad 6,301 23% Total 26, % 39

50 Kayah State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete - - 0% Gravel 100-0% Metalled % Earth 329-0% Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete - 0% P3 Gravel - 0% P4 Metalled 67 15% P5 Earth - 0% Total % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS - 0% C4 SR + TR % Total % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 >2500-0% Total % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor 31 7% R4 Bad 15 3% R5 Very bad 83 18% Total % 40

51 Kayin State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete - 0% P3 Gravel 4 0% P4 Metalled 40 3% P5 Earth % Total 1, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS 4 0% C4 SR + TR % Total 1, % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 > % Total 1, % Code Road condition Surveyed (km) % R1 Good 39 3% R2 Fair 71 6% R3 Poor 73 6% R4 Bad 80 7% R5 Very bad % Total 1, % 41

52 Chin State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete - 0% P3 Gravel 118 7% P4 Metalled % P5 Earth % Total 1, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR 57 3% C3 RS % C4 SR + TR 1,314 76% Total 1, % Code Traffic (AADT) Surveyed (km) % T1 <50 1,649 95% T % T % T % T % T6 >2500-0% Total 1, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad 111 6% R5 Very bad % Total 1, % 42

53 Sagaing Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 1,928 62% P2 Concrete 44 1% P3 Gravel 253 8% P4 Metalled 216 7% P5 Earth % Total 3, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS 171 5% C4 SR + TR 2,054 66% Total 3, % Code Traffic (AADT) Surveyed (km) % T1 <50 1,357 43% T % T % T % T % T6 > % Total 3, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad % R5 Very bad % Total 3, % 43

54 Tanintharyi State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete - 0% P3 Gravel % P4 Metalled 49 4% P5 Earth 56 5% Total 1, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR 1,034 84% C3 RS - 0% C4 SR + TR % Total 1, % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 >2500-0% Total 1, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad 67 5% R5 Very bad % Total 1, % 44

55 Bago Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 1,454 80% P2 Concrete % P3 Gravel 65 4% P4 Metalled 20 1% P5 Earth 16 1% Total 1, % Code Administrative class Surveyed (km) % C1 EH % C2 IC + UR % C3 RS % C4 SR + TR % Total 1, % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 > % Total 1, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad 108 6% R5 Very bad % Total 1, % 45

56 Magway Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 2,460 80% P2 Concrete - 0% P3 Gravel 107 3% P4 Metalled % P5 Earth 18 1% Total 3, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR 1,183 39% C3 RS % C4 SR + TR 1,361 44% Total 3, % Code Traffic (AADT) Surveyed (km) % T1 <50 1,130 37% T ,168 38% T % T % T % T6 >2500-0% Total 3, % Code Road condition Surveyed (km) % R1 Good 1,297 42% R2 Fair 1,135 37% R3 Poor 283 9% R4 Bad 76 2% R5 Very bad 274 9% Total 3, % 46

57 Mandalay Region + Naypyitaw Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel % Metalled % Earth 42-0% Total Code Road surface Surveyed (km) % P1/P6 Bituminous 2,003 81% P2 Concrete % P3 Gravel 29 1% P4 Metalled 66 3% P5 Earth - 0% Total 2, % Code Administrative class Surveyed (km) % C1 EH % C2 IC + UR 1,045 42% C3 RS % C4 SR + TR % Total 2, % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 > % Total 2, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad 204 8% R5 Very bad 211 9% Total 2, % 47

58 Mon State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel 65-0% Metalled 2-0% Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete 8 1% P3 Gravel - 0% P4 Metalled - 0% P5 Earth 46 5% Total % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS 74 9% C4 SR + TR % Total % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 > % Total % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor % R4 Bad 79 9% R5 Very bad % Total % 48

59 Rakhine State Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 1,067 69% P2 Concrete 32 2% P3 Gravel 83 5% P4 Metalled % P5 Earth 44 3% Total 1, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS % C4 SR + TR % Total 1, % Code Traffic (AADT) Surveyed (km) % T1 <50 1,042 67% T % T % T % T % T6 >2500-0% Total 1, % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor 111 7% R4 Bad 88 6% R5 Very bad % Total 1, % 49

60 Yangon Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete % Gravel 12-0% Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous % P2 Concrete 77 8% P3 Gravel - 0% P4 Metalled 16 2% P5 Earth 16 2% Total % Code Administrative class Surveyed (km) % C1 EH 48 5% C2 IC + UR % C3 RS % C4 SR + TR % Total % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 > % Total % Code Road condition Surveyed (km) % R1 Good % R2 Fair % R3 Poor 48 5% R4 Bad 30 3% R5 Very bad % Total % 50

61 Shan State (South, North, East) Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete 65-0% Gravel % Metalled % Earth % Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 4,475 69% P2 Concrete - 0% P3 Gravel 68 1% P4 Metalled 1,807 28% P5 Earth 169 3% Total 6, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR 1,672 26% C3 RS % C4 SR + TR 3,905 60% Total 6, % Code Traffic (AADT) Surveyed (km) % T1 <50 2,665 41% T ,369 36% T % T % T % T6 > % Total 6, % Code Road condition Surveyed (km) % R1 Good 2,658 41% R2 Fair 1,189 18% R3 Poor % R4 Bad 487 7% R5 Very bad 1,517 23% Total 6, % 51

62 Ayeyarwady Region Surface 2014 length (km) Surveyed (km) Surveyed (%) Bituminous % Concrete 19-0% Gravel % Metalled % Earth 362-0% Total % Code Road surface Surveyed (km) % P1/P6 Bituminous 1,614 80% P2 Concrete - 0% P3 Gravel 54 3% P4 Metalled % P5 Earth - 0% Total 2, % Code Administrative class Surveyed (km) % C1 EH - 0% C2 IC + UR % C3 RS 108 5% C4 SR + TR 1,588 79% Total 2, % Code Traffic (AADT) Surveyed (km) % T1 < % T % T % T % T % T6 >2500-0% Total 2, % Code Road condition Surveyed (km) % R1 Good 1,219 61% R2 Fair % R3 Poor 99 5% R4 Bad 45 2% R5 Very bad 185 9% Total 2, % 52

63 ANNEX 3 Road Case Matrix T1 <50 AADT T AADT T AADT T AADT T AADT R1 <4 IRI P2 Cement Concrete R2 4-6 IRI R3 6-8 IRI R IRI R5 >10 IRI R1 <4 IRI P6 Asphalt Concrete R2 4-6 IRI R3 6-8 IRI R IRI R5 >10 IRI R1 <4 IRI P1 Penetration Macadam R2 4-6 IRI R3 6-8 IRI R IRI R5 >10 IRI P4 Dry-Bound Macadam (Metalled) R1 <4 IRI C1 EH - C2 IC/UR ,432 C3 RS ,287 C4 DR/TR , ,661 14,012 C1 EH - C2 IC/UR ,537 C3 RS ,669 C4 DR/TR , ,899 C1 EH - C2 IC/UR ,700 C3 RS C4 DR/TR ,722 C1 EH - C2 IC/UR C3 RS C4 DR/TR C1 EH - C2 IC/UR C3 RS C4 DR/TR C1 EH T6 C2 IC/UR ,629 >2500 C3 RS AADT C4 DR/TR - Total ,295 4,829 2,150 1,113 3, ,017 2, , ,181 1,117 1,100 4,339 40,116 EH: Expressway, IC: International Corridor, UR: Union Road, RS: Regional/State Road, DR: District Road, TR: Township Road AADT: Average Annual Daily Traffic, IRI: International Roughness Index Source: Consultant s processing of data from the Road Data Bank R2 4-6 IRI R3 6-8 IRI R IRI R5 >10 IRI R1 <4 IRI R2 4-6 IRI P3 Gravel R3 6-8 IRI R IRI R5 >10 IRI R1 <4 IRI R2 4-6 IRI P5 Earthen R3 6-8 IRI R IRI R5 >10 IRI Total 53

64 ANNEX 4 Road maintenance standards P6 Asphalt Concrete Traffic AADT IRI<4 4<IRI<6 6<IRI<8 8<IRI<10 IRI>10 SD25mm@IRI4 SD25mm@IRI5 OL40mm@IRI6 REHAB AC@IRI8 <50 OL40mm@IRI4 OL40mm@IRI4 OL60mm@IRI6 REHAB AC@IRI10 OL40mm@IRI6 OL40mm@IRI6 OL60mm@IRI >2500 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 REHAB AC@IRI8 REHAB AC@IRI10 REHAB AC@IRI8 REHAB AC@IRI10 REHAB AC@IRI8 REHAB AC@IRI10 REHAB AC@IRI8 REHAB AC@IRI10 REHAB AC@IRI8 REHAB AC@IRI10 REHAB AC@IRI10 REHAB AC@IRI10 REHAB AC@IRI10 REHAB AC@IRI10 REHAB AC@IRI10 REHAB AC@IRI10 P1 Penetration Macadam Traffic AADT IRI<4 4<IRI<6 6<IRI<8 8<IRI<10 IRI>10 SD25mm@IRI4 SD25mm@IRI5 PM75mm@IRI6 REHAB PM@IRI8 REHAB PM@IRI10 <50 OL40mm@IRI4 OL40mm@IRI6 OL40mm@IRI4 OL40mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 REHAB PM@IRI >2500 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI4 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 SD25mm@IRI5 OL40mm@IRI4 OL40mm@IRI6 OL60mm@IRI8 PM75mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 PM75mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 PM75mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 PM75mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 PM75mm@IRI6 OL40mm@IRI6 OL60mm@IRI6 OL60mm@IRI8 REHAB PM@IRI8 REHAB PM@IRI10 REHAB PM@IRI8 REHAB PM@IRI10 REHAB PM@IRI8 REHAB PM@IRI10 UPGRADE AC REHAB PM@IRI8 REHAB PM@IRI10 UPGRADE AC REHAB PM@IRI8 REHAB PM@IRI10 UPGRADE AC REHAB PM@IRI10 REHAB PM@IRI10 REHAB PM@IRI10 UPGRADE AC REHAB PM@IRI10 UPGRADE AC REHAB PM@IRI10 UPGRADE AC 54

65 P2 Cement Concrete Traffic AADT IRI<4 4<IRI<6 6<IRI<8 8<IRI<10 IRI>10 <50 REHAB > REHAB REHAB REHAB REHAB REHAB REHAB REHAB REHAB REHAB REHAB REHAB P3, P4, P5 Gravel Roads/Dry Macadam/Earth Traffic AADT GR > 50mm 40mm > GR >50mm 30mm > GR >20mm 20mm > GR >10mm 10mm > GR <50 UPGRADE PM UPGRADE PM UPGRADE PM UPGRADE PM UPGRADE PM UPGRADE AC UPGRADE PM UPGRADE PM UPGRADE AC UPGRADE PM UPGRADE PM UPGRADE AC For descriptions of the maintenance standards, see Table 8. Source: TA consultant UPGRADE PM UPGRADE PM UPGRADE AC UPGRADE PM UPGRADE PM UPGRADE PM UPGRADE AC UPGRADE PM UPGRADE AC 55

66 ANNEX 5 HDM4 calibration values Table 19 Vehicle fleet in Myanmar for HDM4 model Vehicle class for HDM4 Make and model Motorcycle Honda/2007 Passenger Car Toyota Crown (2004) / Toyota Lexus GS305 (2005)/ Toyota Mark II (2001) Minibus Toyota Probox / Honda Fit (2007) / Suzuki Alto (2009) Large bus Mitsubishi Rosa(2005) / Mitsubishi Rosa (2004) / Hino (2000) Light Truck Toyota Townace (2000) / Toyota Liteace (2004) / Madza Bongo (2003) Medium Truck Isuzu Elf (2006) / Toyota Dyna (2002) / Mitsubishi Canter (2004) Heavy Truck Hyundai/2000 Semi-Trailer/ Trailer Hyundai/2000 Source: TA consultant Table 20 Basic vehicle characteristics for HDM4 Vehicle Class Operating weight (tonnes) PCS E No. of Axles No. of Wheels No. of ESAs No. of passengers Workrelated trips (%) Motorcycle % Passenger Car % Minibus % Large Bus % 2-axle Light Truck % 2-axle Medium Truck % 3-axle Heavy Truck % 4+axle Heavy Truck % Semi-trailer/Trailer % Source: TA consultant Vehicle Class Table 21 New vehicle cost Vehicle economic unit costs for HDM4 Cost per Cost per hour of hour of repair/ vehicle maintenance crew Cost of new tyre Passenger in work time delay cost per hour Passenger in nonwork time delay cost per hour Annual overhead Motorcycle Passenger Car 10, Minibus 33, Large Bus 140, axle Light Truck 24, axle Medium Truck 29, axle Heavy Truck 50, axle Heavy Truck 57, Semi-trailer/Trailer 86, Source: TA consultant 56

67 Table 22 Vehicle utilisation for HDM4 Vehicle Class Annual vehicle km Annual working hours Vehicle life (years) Annual Interest (%) Motorcycle 4, % Passenger Car 25, % Minibus 25, % Large Bus 95, % 2-axle Light Truck 53, % 2-axle Medium Truck 53, % 3-axle Heavy Truck 70, % 4+ axle Heavy Truck 67, % Semi-trailer/Trailer 73, % Source: TA consultant Table 23 Costs of fuel and lubricants for HDM4 (MK 1,000) Item Cost/Litre Petrol 0.69 Diesel 0.66 Oil (car) 1.80 Oil (other) 0.80 Source: TA consultant Table 24 Value of time Type Value Work 1.35 Non-work 0.44 Source: TA consultant Figure 32 Road user costs in Myanmar (MK 1,000/km) Source: TA consultant 57

68 ANNEX 6 Road maintenance strategies Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T1;R1;C2;P1; T1;R1;C2;P3; T1;R1;C2;P4; T1;R1;C2;P5; T1;R1;C3;P1; T1;R1;C3;P5; T1;R1;C4;P1; T1;R1;C4;P2; T1;R1;C4;P3; T1;R1;C4;P4; T1;R1;C4;P5; T1;R1;C4;P6; T1;R2;C2;P1; T1;R2;C2;P3; T1;R2;C2;P4; T1;R2;C2;P5; T1;R2;C3;P1; T1;R2;C3;P5; T1;R2;C4;P1; T1;R2;C4;P2; T1;R2;C4;P3; T1;R2;C4;P4; T1;R2;C4;P5; T1;R2;C4;P6; T1;R3;C2;P1; T1;R3;C2;P3; T1;R3;C2;P4; T1;R3;C2;P5; T1;R3;C3;P1; T1;R3;C3;P5; T1;R3;C4;P1;

69 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T1;R3;C4;P3; T1;R3;C4;P4; T1;R3;C4;P5; T1;R3;C4;P6; T1;R4;C2;P1; T1;R4;C2;P3; T1;R4;C2;P4; T1;R4;C2;P5; T1;R4;C3;P1; T1;R4;C3;P3; T1;R4;C3;P5; T1;R4;C4;P1; T1;R4;C4;P3; T1;R4;C4;P4; T1;R4;C4;P5; T1;R4;C4;P6; T1;R5;C2;P1; T1;R5;C2;P2; T1;R5;C2;P3; T1;R5;C2;P4; T1;R5;C2;P5; T1;R5;C3;P1; T1;R5;C3;P3; T1;R5;C3;P4; T1;R5;C3;P5; T1;R5;C4;P1; 1, T1;R5;C4;P2; T1;R5;C4;P3; T1;R5;C4;P4; T1;R5;C4;P5; 3, T1;R5;C4;P6; T2;R1;C2;P1; T2;R1;C2;P3; 23 GR@30mm GR@30mm 0.72 GR@30mm 0.72 T2;R1;C2;P4; 144 GR@30mm GR@30mm 4.70 GR@30mm

70 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T2;R1;C2;P5; GR@30mm 1.20 GR@30mm 1.20 T2;R1;C3;P1; T2;R1;C3;P2; T2;R1;C3;P4; 109 GR@30mm GR@30mm 3.30 GR@30mm 3.30 T2;R1;C3;P5; 35 GR@30mm GR@30mm 1.15 GR@30mm 1.15 T2;R1;C4;P1; 1, T2;R1;C4;P3; 91 GR@30mm GR@30mm 2.75 GR@30mm 2.75 T2;R1;C4;P4; 128 GR@30mm GR@30mm 3.91 GR@30mm 3.91 T2;R1;C4;P5; GR@30mm 3.70 GR@30mm 3.70 T2;R1;C4;P6; T2;R2;C2;P1; OL40mm@IRI OL40mm@IRI T2;R2;C2;P3; 121 GR@30mm GR@30mm 3.85 GR@30mm 3.85 T2;R2;C2;P4; 213 GR@30mm GR@30mm 7.40 GR@30mm 7.40 T2;R2;C2;P5; GR@30mm 6.11 GR@30mm 6.11 T2;R2;C3;P1; OL40mm@IRI OL40mm@IRI T2;R2;C3;P2; T2;R2;C3;P4; 75 GR@30mm GR@30mm 2.20 GR@30mm 2.20 T2;R2;C3;P5; 35 GR@30mm GR@30mm 1.15 GR@30mm 1.15 T2;R2;C4;P1; OL40mm@IRI OL40mm@IRI T2;R2;C4;P3; 134 GR@30mm GR@30mm 3.93 GR@30mm 3.93 T2;R2;C4;P4; 108 GR@30mm GR@30mm 3.26 GR@30mm 3.26 T2;R2;C4;P5; 103 GR@30mm GR@30mm 4.12 GR@30mm 4.12 T2;R2;C4;P6; OL40mm@IRI OL40mm@IRI T2;R3;C2;P1; OL40mm@IRI OL40mm@IRI T2;R3;C2;P3; GR@30mm 4.46 GR@30mm 4.46 T2;R3;C2;P4; 198 GR@30mm GR@30mm 7.59 GR@30mm 7.59 T2;R3;C2;P5; GR@30mm 3.68 GR@30mm 3.68 T2;R3;C3;P1; OL40mm@IRI OL40mm@IRI T2;R3;C3;P2; T2;R3;C3;P4; 32 GR@30mm GR@30mm 1.01 GR@30mm 1.01 T2;R3;C3;P5; 35 GR@30mm GR@30mm 1.16 GR@30mm 1.16 T2;R3;C4;P1; OL40mm@IRI OL40mm@IRI T2;R3;C4;P3; 198 GR@30mm GR@30mm 6.31 GR@30mm 6.31 T2;R3;C4;P4; 70 GR@30mm GR@30mm 2.27 GR@30mm

71 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T2;R3;C4;P5; GR@30mm 5.14 GR@30mm 5.14 T2;R3;C4;P6; OL40mm@IRI OL40mm@IRI T2;R4;C2;P1; REHAB PM@IRI REHAB PM@IRI T2;R4;C2;P3; GR@10mm 3.14 GR@10mm 3.14 T2;R4;C2;P4; GR@10mm 4.28 GR@10mm 4.28 T2;R4;C2;P5; GR@10mm 2.51 GR@10mm 2.51 T2;R4;C3;P1; REHAB PM@IRI REHAB PM@IRI T2;R4;C3;P2; OL80mm@IRI T2;R4;C3;P4; 29 GR@10mm GR@10mm 0.96 GR@10mm 0.96 T2;R4;C3;P5; 35 GR@10mm GR@10mm 1.16 GR@10mm 1.16 T2;R4;C4;P1; REHAB PM@IRI REHAB PM@IRI T2;R4;C4;P3; 292 GR@10mm GR@10mm 9.89 GR@10mm 9.89 T2;R4;C4;P4; 73 GR@10mm GR@10mm 2.63 GR@10mm 2.63 T2;R4;C4;P5; 175 GR@10mm GR@10mm 6.87 GR@10mm 6.87 T2;R4;C4;P6; T2;R5;C2;P1; REHAB PM@IRI REHAB PM@IRI T2;R5;C2;P3; GR@10mm GR@10mm T2;R5;C2;P4; 149 GR@10mm GR@10mm 5.40 GR@10mm 5.40 T2;R5;C2;P5; GR@10mm 6.17 GR@10mm 6.17 T2;R5;C3;P1; REHAB PM@IRI REHAB PM@IRI T2;R5;C3;P2; T2;R5;C3;P4; 169 GR@10mm GR@10mm 5.87 GR@10mm 5.87 T2;R5;C3;P5; 35 GR@10mm GR@10mm 1.22 GR@10mm 1.22 T2;R5;C4;P1; REHAB PM@IRI REHAB PM@IRI T2;R5;C4;P2; T2;R5;C4;P3; 422 GR@10mm GR@10mm GR@10mm T2;R5;C4;P4; 289 GR@10mm GR@10mm GR@10mm T2;R5;C4;P5; 333 GR@10mm GR@10mm GR@10mm T2;R5;C4;P6; T3;R1;C2;P1; T3;R1;C2;P2; T3;R1;C2;P3; 8 GR@10mm 0.55 GR@10mm 0.31 GR@10mm 0.31 UPGRADE PM 3.08 T3;R1;C3;P1; T3;R1;C3;P4; 7 GR@30mm 0.41 GR@30mm 0.41 GR@10mm 0.23 GR@10mm

72 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T3;R1;C3;P5; UPGRADE PM T3;R1;C4;P1; OL40mm@IRI T3;R1;C4;P3; 76 GR@10mm 2.49 GR@10mm 2.49 GR@10mm 2.49 UPGRADE PM T3;R1;C4;P4; 24 GR@10mm 0.79 GR@10mm 0.79 GR@10mm 0.79 UPGRADE PM 9.24 T3;R1;C4;P6; T3;R2;C2;P1; 448 SD25mm@IRI SD25mm@IRI OL40mm@IRI T3;R2;C2;P2; T3;R2;C2;P3; 20 GR@10mm 0.71 GR@10mm 0.71 GR@10mm 0.71 UPGRADE PM 7.70 T3;R2;C2;P5; 12 GR@10mm GR@10mm 1.04 UPGRADE PM 4.62 T3;R2;C3;P1; 113 SD25mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T3;R2;C3;P4; 17 GR@10mm 0.53 GR@10mm 0.53 GR@10mm 0.53 GR@10mm 0.53 T3;R2;C3;P5; 35 GR@10mm 1.15 GR@10mm 1.15 GR@10mm 1.15 UPGRADE PM T3;R2;C4;P1; 411 SD25mm@IRI SD25mm@IRI OL40mm@IRI T3;R2;C4;P3; 61 GR@10mm 1.92 GR@10mm 1.92 GR@10mm 1.92 UPGRADE PM T3;R2;C4;P4; 44 GR@10mm 1.38 GR@10mm 1.38 GR@10mm 1.38 UPGRADE PM T3;R2;C4;P6; 112 SD25mm@IRI SD25mm@IRI OL40mm@IRI T3;R3;C2;P1; 137 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T3;R3;C2;P2; OL60mm@IRI OL60mm@IRI T3;R3;C2;P3; 20 GR@10mm 0.68 GR@10mm 0.68 GR@10mm 0.68 UPGRADE PM 7.70 T3;R3;C2;P5; 76 GR@10mm GR@10mm 6.56 UPGRADE PM T3;R3;C3;P1; 21 OL40mm@IRI OL40mm@IRI OL40mm@IRI T3;R3;C3;P4; 4 GR@10mm 0.13 GR@10mm 0.13 GR@30mm 0.25 GR@10mm 0.13 T3;R3;C3;P5; 35 GR@10mm 1.15 GR@10mm 1.15 GR@10mm 1.15 UPGRADE PM T3;R3;C4;P1; 159 OL40mm@IRI OL40mm@IRI OL40mm@IRI T3;R3;C4;P3; 39 GR@10mm 1.28 GR@10mm 1.28 GR@10mm 1.28 UPGRADE PM T3;R3;C4;P4; 55 GR@10mm 1.80 GR@10mm 1.80 GR@10mm 1.80 UPGRADE PM T3;R3;C4;P6; 112 OL40mm@IRI OL40mm@IRI OL40mm@IRI T3;R4;C2;P1; 42 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T3;R4;C2;P2; OL80mm@IRI OL80mm@IRI T3;R4;C2;P3; 22 GR@10mm 0.74 GR@10mm 0.74 GR@10mm 0.74 UPGRADE PM 8.47 T3;R4;C2;P5; 124 GR@10mm GR@10mm UPGRADE PM T3;R4;C3;P1; 6 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T3;R4;C3;P4; 11 GR@10mm 0.38 GR@10mm 0.38 GR@10mm 0.38 GR@10mm 0.38 T3;R4;C3;P5; 35 GR@10mm 1.20 GR@10mm 1.20 GR@10mm 1.20 UPGRADE PM

73 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T3;R4;C4;P1; 138 REHAB REHAB PM@IRI REHAB PM@IRI T3;R4;C4;P3; 31 GR@10mm 1.02 GR@10mm 1.02 GR@10mm 1.02 UPGRADE PM T3;R4;C4;P4; 20 GR@10mm 0.66 GR@10mm 0.66 GR@10mm 0.66 UPGRADE PM 7.70 T3;R4;C4;P6; REHAB AC@IRI T3;R5;C2;P1; 96 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T3;R5;C2;P2; T3;R5;C2;P3; 120 GR@10mm 4.03 GR@10mm 4.03 GR@10mm 4.03 UPGRADE PM T3;R5;C2;P5; 85 GR@10mm GR@10mm 7.34 UPGRADE PM T3;R5;C3;P1; 11 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T3;R5;C3;P4; 153 GR@10mm 5.01 GR@10mm 5.01 GR@10mm 5.01 UPGRADE PM T3;R5;C3;P5; 35 GR@10mm 1.15 GR@10mm 1.15 GR@10mm 1.15 UPGRADE PM T3;R5;C4;P1; 247 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T3;R5;C4;P3; 48 GR@10mm 1.57 GR@10mm 1.57 GR@10mm 1.57 UPGRADE PM T3;R5;C4;P4; 65 GR@10mm 2.13 GR@10mm 2.13 GR@10mm 2.13 UPGRADE PM T3;R5;C4;P6; REHAB AC@IRI T4;R1;C2;P1; T4;R1;C3;P1; T4;R1;C4;P1; T4;R2;C2;P1; 192 SD25mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T4;R2;C3;P1; 120 OL40mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T4;R2;C4;P1; 140 SD25mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T4;R3;C2;P1; 36 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T4;R3;C3;P1; 43 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T4;R3;C4;P1; 186 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T4;R4;C2;P1; 7 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T4;R4;C3;P1; 13 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T4;R4;C4;P1; 70 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T4;R5;C2;P1; 26 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T4;R5;C3;P1; 99 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T4;R5;C4;P1; 66 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R1;C2;P1; SD25mm@IRI SD25mm@IRI OL40mm@IRI T5;R1;C3;P1; 9 - SD25mm@IRI OL40mm@IRI OL40mm@IRI T5;R1;C3;P6; OL40mm@IRI T5;R1;C4;P1; 16 SD25mm@IRI SD25mm@IRI SD25mm@IRI SD25mm@IRI

74 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T5;R2;C2;P1; SD25mm@IRI SD25mm@IRI OL40mm@IRI T5;R2;C3;P1; 26 - SD25mm@IRI SD25mm@IRI OL40mm@IRI T5;R2;C3;P6; 86 - SD25mm@IRI SD25mm@IRI OL40mm@IRI T5;R2;C4;P1; 39 OL40mm@IRI SD25mm@IRI OL40mm@IRI OL40mm@IRI T5;R3;C2;P1; OL40mm@IRI OL40mm@IRI OL40mm@IRI T5;R3;C3;P1; 24 - OL40mm@IRI OL40mm@IRI OL40mm@IRI T5;R3;C3;P6; 54 - OL40mm@IRI OL40mm@IRI OL40mm@IRI T5;R3;C4;P1; 14 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T5;R4;C2;P1; 51 - REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R4;C3;P1; 15 - REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R4;C3;P6; REHAB AC@IRI REHAB AC@IRI T5;R4;C4;P1; 4 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R5;C2;P1; 40 - REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R5;C3;P1; 71 - REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T5;R5;C3;P6; REHAB AC@IRI REHAB AC@IRI REHAB AC@IRI T5;R5;C4;P1; 42 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T6;R1;C1;P2; T6;R1;C2;P1; SD25mm@IRI OL40mm@IRI OL40mm@IRI T6;R1;C2;P6; 403 SD25mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T6;R1;C3;P1; 18 SD25mm@IRI SD25mm@IRI SD25mm@IRI OL40mm@IRI T6;R2;C1;P2; OL50mm@IRI OL50mm@IRI T6;R2;C2;P1; OL40mm@IRI OL40mm@IRI OL40mm@IRI T6;R2;C2;P6; SD25mm@IRI SD25mm@IRI OL40mm@IRI T6;R2;C3;P1; 6 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T6;R3;C1;P2; 5 OL60mm@IRI OL60mm@IRI OL60mm@IRI T6;R3;C2;P1; 98 - OL40mm@IRI OL40mm@IRI OL40mm@IRI T6;R3;C2;P6; OL40mm@IRI OL40mm@IRI OL40mm@IRI T6;R3;C3;P1; 2 OL40mm@IRI OL40mm@IRI OL40mm@IRI OL40mm@IRI T6;R4;C1;P2; 2 OL80mm@IRI OL80mm@IRI OL80mm@IRI T6;R4;C2;P1; 41 - REHAB PM@IRI REHAB PM@IRI UPGRADE AC T6;R4;C2;P6; 29 - REHAB AC@IRI REHAB AC@IRI REHAB AC@IRI T6;R4;C3;P1; 2 - REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI T6;R5;C1;P2; REHAB CC@IRI T6;R5;C2;P1; 100 REHAB PM@IRI REHAB PM@IRI REHAB PM@IRI UPGRADE AC

75 Road case Length (km) Scenario 1A: MK 100 billion restricted Scenario 1B: MK 100 billion optimized Scenario 2: MK 250 billion optimized Scenario 3: MK 400 billion optimized Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) Standard Cost (MK billion) T6;R5;C2;P6; 55 - REHAB AC@IRI REHAB AC@IRI REHAB AC@IRI T6;R5;C3;P1; 29 - REHAB PM@IRI REHAB PM@IRI UPGRADE AC For explanations on the road cases, see Table 2, Table 3, Table 4, Table 5 and Table 6. For descriptions of the maintenance standards, see Table 8. Source: Consultant s processing of data from HDM4 strategy analysis 65

76 ANNEX 7 Maintenance programme Road Road name RDB Start End Length of works (km) Cost of works (MK billion) code Sections Overlay Rehab Upgrade Upgrade Total Overlay Rehab Upgrade Upgrade Total (miles/furlongs) PM AC PM AC Ayeyarwady DT162 Pa Thein-Ngwe Saung Road /0 29/ DT165 Kyain Pin Sae-Set Kawt- Dana Phyu -Zalun Road /0 27/ DT204 Hin Tha Da-Do Yar - Daunt Gyi- Da Na Phyu Road 10 0/0 0/ DT205 Da Nu Phyu- Thaung Gyi Road /0 24/ SR59 Ma Euu Pin-Twan Tay Road /0 23/ UR20B Yangon -Pa Thein Road /4 80/ UR8A Pa Thein - Mon Ywar Road /0 74/ Bago DT53 Nyaung lay Pin - Pa Zun Myaung - Shwe Kyinn 10 0/0 12/ DT57 Pyay-Pout Kaung-Taung Gu 40 40/0 80/ IC25A Yangon - Maw La Myin - Dewe - Myeik 10 60/5 86/ IC25F Sit Taung Bridge Approach 10 0/0 6/ IC41 Yangon - Taungoo - Mandalay Highway Old Road /0 200/ NC7E Shwe Bon Thoor - Sin Del - Padaung - Ohn Ship /1 46/ UR8B Pa Thein - Mon /5 179/ UR9B Yangon - Pyay - Mandalay /6 193/ UR9E Pyay City Out Bound Road 10 0/0 13/ Kayin IC10B Tha Htone-Ba Ahn-Kokkareit-Myawaddy Road /1 23/ IC10F Tha Htone-Ba Ahn-Kokkareit-Myawaddy Road /0 41/ IC10G Tha Htone-Ba Ahn-Kokkareit-Myawaddy Road /2 9/ IC10H Tha Htone-Ba Ahn-Kokkareit-Myawaddy Road /0 103/ NC3C Hte Lone - Ta Tar Kyae Road 10 0/0 9/ TV70 Hteepoekalone Myinegyinguu Maethayor road 10 0/0 13/ Magway DT59 Min Bu - Sa Linn - Ta Nyaun - Sate Phyu Road /0 45/ DT61A Gway Cho - Chauk - Sate Phyu Road /5 399/ DT71 Sin Paung Wal - Taw Nyaung Pin Road 10 0/0 16/ IC23B Monywa - Pale - Gangaw Road /0 120/ IC24A Kalay - Gangaw Road 10 0/0 8/ IC32 Chaung Oo - Pa Koak Khu Road 10 6/4 10/ SR19 Pa Koak Khu - Mon Ywa Road /5 24/

77 Road Road name RDB Start End Length of works (km) Cost of works (MK billion) code Sections Overlay Rehab Upgrade Upgrade Total Overlay Rehab Upgrade Upgrade Total (miles/furlongs) PM AC PM AC SR27B Pyaw Bwel - Ywar Mon - Kan Pyar (Magway) Road 40 62/4 75/ SR31 Nay Pyi Taw - Taung Nyo - Kan Pyar (Magway) Road /6 66/ UR10 Magway - Min Bu Road 10 2/4 5/ UR12A Min Bu - Ann - Sittwe /0 38/ UR7A Pa Kouk Khu - Kyauk Htu - Kan Gyi - Min Tap Road 80 62/6 77/ UR8C Pa Thein - Mon Ywa Road /0 274/ UR9C Yangon - Pyay - Mandalay Road /7 226/ UR9C Yangon - Pyay - Mandalay Road /0 391/ Mandalay DT81 Mandalay-Moe Kote Road /0 54/ DT83 Meik Hti Lar-Ma Hlaing-Taung Tha Road /4 30/ DT84 Myin Gyan-Myit Thar-Yay Won Road /3 56/ DT85 Thar Si-Tha Pyay Wa Road 10 0/0 9/ IC36 Bagan-Nyaung Oo-Myin Gyan Road /0 19/ IC38A Mandalay - Lar Shio - Myit Kyi Na Road /0 55/ IC39A Meik Hti Lar-Taung Gyi-Kyine Ton-Tar Chi Late Road /6 35/ IC41 Yangon - Taungoo - Mandalay Highway Old Road /3 421/ SR27A Pyaw Bwe-Ywar Taw-Net Mouk-Kan Pyar Road /0 14/ UR4 Mandalay-Ta Kaung-Banmar-Myit Kyi Na Road 10 0/0 4/ UR9D Yangon-Pyay-Mandalay Road /2 397/ UR9D Yangon-Pyay-Mandalay Road /2 457/ UR9D Yangon-Pyay-Mandalay Road /0 510/ Mon IC10A Tha Hton - Ba Annn - Kokareit - Mya Waddy Road 10 0/0 8/ IC25B Yangon - Maw La Myine - Dewi - Meik Road /4 190/ IC25C Yangon - Maw La Myine - Dewi - Meik Road /6 96/ IC25G Yangon - Maw La Myine - Dewi - Meik Road 10 6/5 8/ IC25H Yangon - Maw La Myine - Dewi - Meik Road 10 0/0 3/ IC8B Maw La Myine - Zar Tha Pyin - Ein Du Road 10 17/7 27/ IC9A Than Phyu Za Yat - Bayar Thone Suu Road 10 0/0 10/ SR24A Mudon - Myawaddy Road /0 9/ TV213 Theinseik - Wiyor - Laykay Road / Naypyitaw IC41 Yangon-Taungoo-Mandalay Highway Old Road /1 296/ Rakhine DT109 Buu Thee Taung - Maungdaw Road /0 14/ SR37 Than Dwal - Gwa Road 50 40/4 55/

78 Road Road name RDB Start End Length of works (km) Cost of works (MK billion) code Sections Overlay Rehab Upgrade Upgrade Total Overlay Rehab Upgrade Upgrade Total (miles/furlongs) PM AC PM AC TV230 Kanyin Tann - Ala Than Kyaw - Anguu Maw Road /0 49/ UR18 Ann - Pa Dae Kyar - Ma Ei Road 40 33/5 41/ Sagaing DT35 Yay Oo - Ka Lay Wa Road /0 12/ DT37 Shwe Bo - Kyaunt Myaung Road /0 16/ DT38 Shwe Bo - Yay Oo Road 20 9/7 17/ DT45 Sin Kone - Wun Tho - Pin Lae Bu Road /0 18/ DT45 Sin Kone - Wun Tho - Pin Lae Bu Road 60 41/0 50/ DT46 Mon Ywar - Ywar Thar Gyi (Shwe Bo) Road /0 36/ DT47A Sabai Nant Thar - Kant Balu - Kyun Hla Road 10 0/0 9/ DT47B Sabai Nant Thar - Kant Balu - Kyun Hla Road 10 18/0 25/ IC14 Mandalay - Sagain - Mone Ywar Road /1 82/ IC15 Mone Ywar - Yar GYi - Kalay Wa Road /4 115/ IC16 Kalay - Kyee Kone - Ta Muu Road /0 93/ IC17 Myinn Mu City Outbound Road 10 0/0 1/ IC19 Mon Ywa City Outbound Road 10 81/2 88/ IC23A Monywa - Pale -Gangaw Road /0 5/ NC5J Ma Lae - Ya Ma Nay - Zee Kone - Kar Boe /4 33/ SR10 Mandalay - Sagaing - Shwebo Road /0 69/ SR11 Mandalay - Sagaing - Mon Ywa - Yay Oo Road /1 116/ TV114 Taung Maw - Korr Linn - Winngyi Road /0 13/ TV120 Mone Ywar - Tharsi - Naung Gyi - Shwe Bo /0 37/ UR3 Mandalay - Shwebo - Myit Kyi Nah Road /0 43/ UR3A Mandalay - Shwe Bo - Myit Kyi Nah Road /0 141/ Shan East DT149 Tar Lay-Par Show-Kyaing Lup /0 25/ DT155 Mine Tone-Pon Par Kyin /0 34/ DT158 Mine Satt-Tar Chi Late /0 60/ IC39BC Meikhtila-Taungyi-Kyine Tone-Tar Che Late /4 378/ IC39CD Kyaing Tone-Tar Che Late Part /3 102/ SR55B Nan San-Lin Khae-Tar San-Mine Tone /0 160/ Shan North DT129A Mae Han - Mine Rae - Kyae Thee Road /0 47/ SR54B Pyin Oo Lwin - Mogoke Road /1 50/ TV325 Namt Poung - Man Pan - Noung Lai ROad /4 32/ TV377 Kookhine - Tarmoenye - Monesee - Tar Par Road /1 44/

79 Road Road name RDB Start End Length of works (km) Cost of works (MK billion) code Sections Overlay Rehab Upgrade Upgrade Total Overlay Rehab Upgrade Upgrade Total (miles/furlongs) PM AC PM AC Shan South DT128 Shwenyaung-Nyaung Shwe Road 10 0/0 7/ IC39B Meikhtilar-Taungyi-Kyinetong-Tarchelaik Road /6 230/ IC39B Meikhtilar-Taungyi-Kyinetong-Tarchelaik Road /0 278/ IC42A Loilin-Pankaytu-Thipaw Road 10 0/0 1/ SR55A Namsan-Moenae-Linnkhae-Minesatt Road 70 82/0 89/ TV274 Taung Gyi-Hammsee-Naungyar Sine Road /0 28/ TV310 Ayetharyar-Nyaungshwe-Namtpan-Ponlong Road /0 57/ Tanintharyi IC25D Yangon-Mawlamyine-Dawei-Myeik 20 57/4 69/ IC25D Yangon-Mawlamyine-Dawei-Myeik 80 0/0 3/ IC25E Yangon-Mawlamyine-Dawei-Myeik /0 74/ Yangon DT116 Dagon - Thi La War Instural Zone Road 10 0/0 13/ DT118 Kan Bae - Pyaw Bwe Lay - Da La Road 10 0/0 9/ IC41A Yangon - Taungoo - Mandalay Road (Old) /0 14/ SR18B Bago - Tha Net Pin - Khayan - Than Lyin Road /7 69/ SR40 Htan Ta Pin - Thi Dar Aye 10 0/0 8/ SR43 No.2 Highway Road /0 31/ SR44 No.3 Highway Road 10 0/0 9/ SR45 No.4 Highway Road 10 6/0 12/ SR46 No.7 Highway Road (City Orbit Road) 10 0/0 1/ SR47 Twan Tay - Kawt Hmuu - Kwan CHan Kone Road /7 27/ SR48 Kwan Chan Kone - Day Da Ye' Road 10 0/0 8/ SR49 Hlaing Thar Yar - Da La - Twan Tay Road 10 0/0 11/ TV251 Ma Kyi Kan - Hnat Aww San - Kyauk Htaw Road 10 0/0 15/ TV252 Kawt Hmuu - War Blauk Thouk - Tha Yet Taw Road 10 0/0 9/ TV256 Hlegu - Phaung Gyi Road 10 0/0 10/ TV257 Hlegu - Sa Da Lin Road 10 0/0 6/ TV264 Khayan - Yay Kyaw - Aung Thar Dan Road 10 0/0 10/ TV266 Yay Twin Kone - Myo Chuang - Ywar Thar Road 10 0/0 9/ TV267 Tha Htay Kwin - Pa Gan Taung - Maung Ma Road 10 0/0 4/ TV268 Kwan Chan Kone - West Taw Kayan - Boe Din Road 10 0/0 11/ TV270 Thone Kwa - Ka Da Pa Na Road 10 0/0 5/ UR20A Yangon - Pa Thein Road /0 17/ UR9A Yangon - Pyay - Mandalay Road /3 70/

80 Myanmar Overlay Rehabilitation Upgrade PM Upgrade AC

81 Kayin Overlay Rehabilitation Upgrade AC

82 Sagaing Overlay Rehabilitation Upgrade PM Upgrade AC

83 Taninthayi Overlay Rehabilitation Upgrade PM

84 Bago Overlay Rehabilitation

85 Magway Overlay Rehabilitation

86 Mandalay Overlay Rehabilitation Upgrade AC

87 Mon Overlay Rehabilitation Upgrade PM Upgrade AC

88 Rakhine Overlay Rehabilitation Upgrade PM

89 Yangon Overlay Rehabilitation Upgrade PM Upgrade AC

90 Southern Shan Overlay Rehabilitation

91 Northern Shan Overlay Rehabilitation

92 Eastern Shan Overlay Rehabilitation Upgrade PM

93 Ayeyarwady Overlay Rehabilitation Upgrade PM

94 Nay Pyi Taw Overlay Rehabilitation