Sector Wide Road Programme & PIP Study

Part I Final Report

Chapter 1
I. INTRODUCTION

A. Purpose of Report

This is the Final Report of the Sector Wide Road Programme (SWRP) & Priority
Investment Plan (PIP) Study conducted for the Department of Roads (DoR). The
Study was conducted in two Parts: Part I was the overall planning study and Part
II involved the detailed feasibility study of around 800km of high priority road
improvement. The Study commenced in September 2005 with an initial
completion date in December 2006. The Inception Report for Part I of the Study
was submitted in December 2005, followed by the Part I Interim Report in early
June 2006: the submission of this latter report was substantially delayed due to
the difficulties and disruptions associated with the democracy movement during
the early part of 2006 culminating with the restoration of Parliament in May. A
series of bi-monthly Progress Reports were also submitted in February, April,
June, August and October 1 2006..

The Part I Draft Final Report was submitted in two instalments in January and
February 2007 and formed the basis for discussion and consultation initially with
the DoR and subsequently with key stakeholders in the roads sector, including
Government Ministries, the National Planning Council, other Government
Departments, and development agencies. Following these consultations and in
response to comments received, the report has new been amended and
finalised. A list of comments received and the Consultants’ responses are
presented in Annex 1. It is intended that this document now represents
Government policy for the development of the Roads Sector over the coming 10-
year period and it is hoped that it will have broad acceptance within all
Government agencies and the foreign-aid community.

Part I of the Study involved a series of ‘planning activities’ relating to the

production of a 10-Year Priority Investment Plan (PIP) comprising maintenance,
upgrading and some new construction of an expanded Strategic Road Network of
almost 10,000km. In conjunction with the development of this network, the Study
also examined the current levels of accessibility throughout the country and has
prepared a sector-wide strategy aimed specifically at improving remote area
access.

Part II of the Study, which commenced in April/May 2006, involved the Detailed
Technical and Economic Feasibility Study of over 800km of high priority sections
of the Strategic Network selected for either upgrading or new construction. Due to
overall project time constraints is was necessary to identify these roads and
commence fieldwork prior to the finalisation of the prioritisation process. The
selection process and subsequent design work is documented in a separate
series of Reports on Part II of the Study. [As with Part I, work on the detailed

1
The June and October Progress Reports were included in the Part I & II Interim Reports.

DHV - SILT - TAEC – CMS Chapter I-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

feasibility studies – and especially the fieldwork – was significantly delayed during
the earlier part of the year, resulting in delayed submission of some documents.]

This Report contains a record of the work completed during the Study, and a
summary of the main findings and conclusions of the Consultants, including work
on the development and expansion of the SRN, traffic studies, the preparation of
the PIP, and the application of HDM-4 for the development of maintenance and
upgrading programmes. Significant use has been made throughout the Study of
GIS techniques, including the development of a detailed accessibility model,
which has been used to assess the impact of individual roads or networks on the
overall levels of accessibility in the remote, non-road-served areas of Nepal.

B. Report Structure

Following this Introduction, the Report contains a further nine Chapters:

Chapter 2: Study Overview: briefly outlines the overall project objectives and
content, plus a summary of the overall Study findings:

• Part I: Planning Studies, including the Sector Wide Road Programme and
preparation of the 10-Year Priority Investment Plan (PIP); and
• Part II: Detailed Technical & Economic Feasibility Studies for upgrading
and new construction of over 800km of priority road.

Chapter 3: Existing Strategic Road Network: contains details of the existing

network, including historical development and current status and conditions. Also
contains a review of current improvement and expansion proposals resulting in
the definition of an ‘existing’ network of almost 8,000km. Existing traffic volumes
across the network are presented, based on an assembly of available count data
taken over recent years: the basis for traffic forecasting is briefly described. A
review of the existing bridges and the SRN is also included.

Chapter 4: Maintenance & Upgrading of SRN: describes the application and

use of the HDM-4 model to determine the 10-year optimised maintenance and
upgrading programme for the SRN. Specific aspects covered include revisions
and updates to the available database, calibration of the model to conditions in
Nepal (including updating the VOC model), training in the use of the program,
and its establishment within the DoR. The results indicate a total programme
(excluding committed schemes) of around Rs16 billion over the 10 year period.

Chapter 5: Development & Expansion of SRN: describes the process adopted

to identify potential additions and extensions to the SRN, including initially an
assessment of the extent of the existing network and priority ranking system to
screen and select schemes for further consideration and evaluation. In total about
3,600km of road have been selected from a long-list of around 9,000 additional
kilometres.

Chapter 6: Prioritisation of Upgrading & New Construction: provides details

of the analysis and evaluation of the proposed extensions to the SRN, including
new hill roads into remote areas, additional links in the mid-hills and, specifically,
prospects for new routes accessing Kathmandu from the west or south (Terai),
plus the upgrading of gravel or earth roads to sealed standards, and capacity
enhancements within Kathmandu Valley.

DHV - SILT - TAEC – CMS Chapter I-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 7: Rural Accessibility Standards: describes the approach adopted to

assess and measure rural accessibility, based on a detailed understanding of the
spatial distribution of population, road network alignments and the travel time on
foot from the road. A fully integrated GIS-based analysis system has been
developed allowing populations within given walk-times from a road network to be
calculated, as well as the total time taken to reach the nearest road-head from
any location. The accessibility of the existing and proposed extended SRN – in
terms of 4-hours walk in the hills and 2-hours in the Terai – has been calculated,
indicating that around 85 percent of the total population will be within the desired
targets.

Chapter 8: Rural Access Improvement Strategy: extends the above analysis

to take account of the additional impact of the local rural road networks in
extending the ‘reach’ or accessibility provided by the SRN. Overall, it is estimated
that over 90 percent of the population will be accessible by 2016 (80 percent in
the hills and almost 100 percent in the Terai).

Chapter 9: Environmental & Social Aspects: contains a brief overview of the

work undertaken in regard to the social and environmental aspects, including the
production of an Environmental & Social Management Framework (ESMF) and
recommendations for training and the strengthening of the Geo-Environmental
and Social Unit (GESU) in DoR.

Chapter 10: Priority Investment Plan and Sector Wide Road Programme:
presents a summary of the key findings from the Study, including specifically a
draft outline of a future budget estimate for the coming 10-years, presenting an
indication of the likely spending levels and sources of funding.

DHV - SILT - TAEC – CMS Chapter I-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 2
II. STUDY OVERVIEW

A. Study Objectives & Outputs

The Study was conducted for the Department of Roads (DoR) as part of the
World Bank funded Road Maintenance and Development Project (RMDP). The
Study comprised two parts which were scheduled to be undertaken over a 15
month period between September 2005 and December 2006:

• Part I: Sector Wide Road Programme and Priority Investment Plan; and
• Part II: Detailed Technical & Economic Feasibility Studies

Part I, which was effectively the ‘planning phase’ of the Study, was commenced
in September 2005 and continued though to the production of the Draft Final
Report in January 2007 and subsequent revisions to produce this final version in
April 2007. This Report now presents the Consultants overall findings and
recommendations relating to the development of the Sector Wide Road
Programme and Priority Investment Plan (PIP) for the Strategic Road Network
(SRN) and incorporates comments and observations from Government.

Part II of the Study, which was conducted in parallel from April 2006 onward,
involved the detailed technical and economic feasibility study of over 800km of
selected high-priority road for either upgrading or new construction. It is proposed
that these roads will form the basis for the next IDA-funded Loan in the roads
sector. A separate series of Reports – including Detailed Feasibility Studies for
16 individual road projects – have been produced and submitted.

B. Sector Wide Road Programme

The Sector Wide Road Programme (SWRP) has examined the overall road
system and has produced recommendations regarding the extent and nature of
the required strategic and local networks necessary to achieve the desired levels
of accessibility throughout the country. The Study has produced a balanced
strategy covering the development and maintenance of an expanded strategic
road network, plus the additional local road requirements to achieve the desired
levels of accessibility.

The analysis of the overall road network requirements has been undertaken
using a GIS-based approach that links the development and expansion of the
road networks to the current and projected population distribution throughout the
country. This approach was applied initially to the development of the Strategic

DHV - SILT - TAEC – CMS Chapter II-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Network and subsequently to the identification of areas where additional local

roads are required to provide access to all populated areas 2.

Specifically the Study has developed a detailed population density distribution

across the country based on the census populations of each of the almost 4,000
VDCs, which have then been re-distributed geographically based on the house
and village/urban clusters from the topographic mapping, to create an accurate
indication of population location on a 1 sq km grid. This allows populations to be
summed within any spatial boundary and – for this project – within catchment
areas and buffer zones around road links or networks. A detailed assessment of
the areas of influence of roads was undertaken using a terrain-dependent walk-
time model, which reflected differing walk speeds related to gradient and was
thus able to determine minimum times and optimum routes.

Detailed quantified calculations have therefore been possible of the numbers of

people benefiting from improved access – in terms of numbers within the 2 hour
and 4 hour access bands from the roads – as well as an indication of the overall
reductions in access time (measured in person-hours) to an all-weather road.

The Study did not however attempt to develop detailed local road networks, as
this is the responsibility of the relevant District Development Committee (DDC)
and is better undertaken locally, where need and priority can be better assessed.
The Study has however recommended a substantial expansion of the strategic
network – including the incorporation many local/districty roads – in order to
provide immproved and more reliable access in rural areas leading to greater
social and economic connectivity. These networks which will be complemented
by additional local roads which will be developed and funded through the DDCs.

C. Priority Investment Plan

The PIP presents a detailed strategy for the development of the SRN over the
next 10 years, including recommendations for the expansion of the network from
approximately 5,000 km of currently designated National Highway and Feeder
Road to around 9,700 km by 2016. The Study has examined both the capital and
recurrent financial implications of the proposed plan and has developed an
optimum maintenance strategy for the existing (& expanded) network.

The maintenance strategy has been developed based on an analysis of the

existing and committed SRN using the HDM-4 model, to examine and evaluate
alternative scenarios and to define an optimal approach. A number of DoR staff
have been trained in the use and application of HDM-4, including calibration of
the model to operating conditions in Nepal: the program is now set up and
operational in the Department. The need to revise and update the data on a
regular basis is emphasised, together with annual reviews of the on-going
programme.

The Study has examined a full range of possible improvements and additions to
the SRN, including new alignments in the hills, the upgrading or improvement of
existing heavily trafficked roads, increased network density and upgrading of
roads in the Terai (including links to India), additional strategic roads within the
Kathmandu Valley, and particularly access routes to Kathmandu – including a
broad assessment of the case for a Fast Track link between Kathmandu and the
Terai and Indian Border.

2
Additional local roads will be required in more densely populated mid-hills and Terai areas in
order to provide higher levels of accessibility than implied by the basic standards

DHV - SILT - TAEC – CMS Chapter II-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

In total, around 170 potential additional links in the SRN were identified, with a
length approaching 9,000 km. These were subjected to a pre-screening exercise,
which identified an initial 4,500km of road for possible inclusion in the future SRN.
These roads were subsequently further evaluated and assessed prior to inclusion
in the PIP. The analyses were primarily based on assessments of improved
accessibility throughout the country.

D. Detailed Technical & Economic Feasibility Studies

Part II of the Study includes the Detailed Feasibility Study of over 800km of
roads, mostly involving upgrading of existing earth or gravel tracks to sealed all-
weather standards. Most of the roads examined are part of a network of hill roads
in the Mid and Far Western Regions. These roads were selected from a ‘pool’ of
around 1300km of road for potential upgrading or new construction. Additionally
some further sections have been reviewed that were previously included under
RMDP and which have not yet been completed to a full all-weather standards.

The Feasibility Studies for these roads have shown that most are economically
viable, based on assumptions regarding growth in traffic demand following
improvement. In addition, the roads will provide substantial overall increases in
accessibility, with associated social and economic benefits: the initial selection
criteria were based primarily on the level of access improvements provided by the
roads.

E. Environmental and Social Aspects

In addition to the above tasks, the Study also included an extensive examination
and review of the current environmental and social impact analysis procedures as
conducted within the DoR, leading to recommendations for simplification, revision
and improvement. A review of the current institutional arrangements within DoR
in respect of social and environmental activities was also undertaken and
recommendations for necessary strengthening produced. Detailed environmental
and social impact studies have been completed for all of the selected upgrading
and new construction alignments in Part II.

A separate Draft Environmental and Social management Framework (ESMF) has

been produced and circulated for comment among key stakeholders.

F. Main Findings from Study

The overall Study contained five main areas of work, the first four of which are
reported on in this document. A separate Final Report is available for Part II
(Detailed Feasibility Studies), which summarises the conclusions and
recommendations of the 16 individual roads studied. The aspects covered in this
Report are:

• Development and expansion of Strategic Road Network;

• Assessment of Accessibility and future road network requirements;
• Development of maintenance & upgrading strategy using HDM-4; and,
• Preparation of 10-year PIP and associated budget estimates.

The separate Part II Final Report presents the results from the 16 individual
Feasibility Studies.

DHV - SILT - TAEC – CMS Chapter II-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

1. Expansion and Development of the SRN

The existing designated SRN comprises approx 5,050km of open and operational
road, comprising 15 National Highways and 51 Feeder Roads. This network
includes around 3,500km of sealed road, 950km of gravel surface, and 600km of
earthen road. Much of the earth and gravel network is seasonal in nature and/or
not available for regular traffic for significant parts of the year.

These are main national arteries which provide inter-regional connections and
links to District Headquarters, international borders, and key economic centres.
The SRN is ‘enabler’ of access at a local or district level: without the strategic
connections to the main road network (eg East-West Highway and main regional
centres), local road networks serve no useful or beneficial purpose.

The DoR has prime responsibility for the construction, development, improvement
and maintenance of the SRN.

2. Findings regarding SRN

It is evident that significant expansion of SRN, beyond the ‘official’ length of
5,050km, is already in hand: current commitments alone will extend network to
approximately 7,900km by 2010. After this, and within the plan period to 2016, a
further expansion of the network is proposed under this Study to around
9,700km. Additional projects are prioritised in terms of network connectivity,
improved accessibility (including remote area access), and network
strengthening.

Expansion of the SRN comprises a combination of the following three types of

road:

• Existing Commitments – roads planned or under construction, for which

funding is secured;
• Conversion of existing Local Roads, which may – or may not – require
improvement or upgrading, including many that are already being
maintained by DoR Divisional Offices; and
• A limited number of new proposals – roads on new alignments mostly into
currently un-served areas.

A total of almost 9,000km of potential extensions (170 schemes) were identified,

out of which around 1,800 km were selected and prioritised for inclusion in the
extended network. The initial screening – based on a multi-criteria approach –
was designed to identify schemes based on their ‘strategic’ function: the better
performing schemes were then assessed for their economic viability.

Four Categories of road were identified, over and above currently committed
schemes and existing roads already maintained by DoR:

• Remote Area Access Links – mostly in the Mid and Far West;
• Mid-Hills Linkages, especially in areas with high population density;
• Strategic improvements and additions, including access to Kathmandu;
• Improvements within Kathmandu Valley to relieve congestion.

All potential components of the 10-year PIP were assessed using HDM on other
evaluation techniques to establish their overall feasibility and priority ranking. In
general most existing commitments and on-going projects were confirmed and
most ranked relatively highly on the overall scoring system. These included the
current ADB projects, those proposed and evaluated under Part II of this Study

DHV - SILT - TAEC – CMS Chapter II-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

for potential inclusion in the next WB/IDA Loan, and the improvements and
upgrading of Terai roads proposed for Indian Government funding.

The types of improvements and works examined included the upgrading of

existing gravel or earth roads to all-weather standards; expansion of the network
into remote hill areas; additional links to Kathmandu; and strategic improvements
within Kathmandu Valley.

3. Recommendations regarding the SRN

Key elements of the expanded SRN in the period up to 2016 include:

• the completion of links to ALL District HQs;

• the upgrading of unsealed roads in Terai and main links into the hills;
• the improvement of the network in mid-hills, including inter-District links
and the development of a mid-hills corridor, within the more densely
populated areas;
• the enhancement of access to Kathmandu, including consideration of the
Fast Track and/or Bhimdhunga Link from the Prithvi Highway west of
Naubise; and
• capacity enhancement of key Kathmandu Valley roads, including the
existing Ring Road and Arniko Highway to Bhaktapur and Dhulikhel.

4. Accessibility
Accessibility has been used throughout the Study as the prime indicator of project
‘worth’ and overall impact on poverty mitigation and reductions in levels of social
exclusion. It has been demonstrated that poor accessibility is strongly correlated
with the incidence of poverty and low scores on a range of social and human
development indicators.

Measures of accessibility depend primarily on population distribution, the extent

and proximity of the all-weather road network, and the walk-time between road
and residence. The numbers of people (and percentages) with either 2 hours in
the Terai or 4 hours in the hills have been calculated, together the overall walk-
time (measured in person-hrs). It can be shown that the 2hr/4hr criterion is not
necessarily the best or only measure of accessibility, as it focuses specifically on
improving access to the more remote areas, whereas a greater impact can be
achieved through smaller improvements to a greater number of people living
relatively closer to the road.

Accessibility analysis has been used specifically for three tasks:

• Identification and evaluation of extensions to the SRN;

• Selection of roads for detailed feasibility study; and
• Forecasting of traffic levels on new and upgraded hill roads

The Study has assessed the impact on accessibility from the extensions to the
SRN. It is estimated that 65 percent of the total population live within the 2hr/4hr
time bands from the SRN at the present time, comprising 50 percent of the hill
population and 76 percent of Terai and Valley populations. The expanded
network by 2016 raises the ‘accessible’ population to 70 percent in the hills and
97 percent in the Terai – or 85 percent overall.

This indicates the very significant impact that the Strategic (Main) Road Network
has on overall accessibility. It is the initial roads into an area that have the most
significant effect as (a) they tend to be located in the more populous areas and

DHV - SILT - TAEC – CMS Chapter II-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

(b) the strategic roads are the first roads built into the non-served areas. The
inclusion (in 2016) of an additional 6,000km of local roads (representing a 60
percent increase in overall road length) raises the overall percentage served by
only 6 percent to 91 percent 3 , illustrating the diminishing returns from additional
local road construction.

5. Application of HDM-4
The objectives of the HDM-4 component of the Study were to develop an
economically optimum 10 year investment plan for all roads under DoR and to
establish HDM in DoR as a tool for expenditure planning, prioritising major
maintenance works and undertaking feasibility studies.

The HDM-4 analysis of SRN was conducted on the ‘committed’ 2010 network of
7,900km of road, including 5,400km of currently operational road, plus 2,500km
of new or additional links. The network assumed the completion of 3,000km of
‘committed’ upgrading and improvement before 2010. The analysis therefore
excluded consideration of any other upgrading in the period to 2010 as sufficient
works were already in the pipe-line.

The balanced and optimised analysis produced a total 10-year programme of

around Rs16 billion including rehabilitation, periodic maintenance and upgrading,
involving total interventions of 4,500km – divided into approximately equal
lengths between reseals, AC overlays, rehabilitation and upgrading of gravel
roads. An average regular annual maintenance expenditure of around Rs 1
billion should be added to this.

6. Review of maintenance strategies

The HDM model was also used to assess the impact and effectiveness of a
change in current Nepali maintenance practices of only using AC Overlays on
high traffic roads, with reseals applied to roads with less than 3,000 vpd. The
analysis concluded that, far a similar overall level of expenditure, a greater use of
overlays produced improved results and, most noticeably, a significantly better
(smoother and stronger) pavement at the end of the plan period. One reason for
this is that a continuing use of the cheaper re-seals does nothing to improve the
overall pavement strength or roughness, and leads eventually to a more
expensive full rehabilitation.

The Consultants recommend that HDM results should be treated as advisory and
indicative of the scale and types of intervention required. It was also noted that
the process requires continuous monitoring and regular updating of the input
data. Whilst the model produces an assessment of priorities over 10 year period,
it requires to be reviewed and re-run on an annual basis.

7. Preparation of PIP
The preparation of the Priority Investment Plan for the ten years to 2016 includes
three main items:

• Firstly, the definition of likely maintenance requirements;

• Secondly, identifying necessary upgrading of existing and potential SRN
elements (including some currently local roads); and
• Thirdly, the Construction of new (strategic) links – rural access and
network strengthening

3
Comprising increases from 70% to 80% in the hills and from 97% to 99% in the Terai

DHV - SILT - TAEC – CMS Chapter II-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The PIP presents a draft budget proposal for DoR for the next ten years,
involving the preparation of budget estimates and an assessment of potential
funding sources. The estimates indicate an overall budget of around Rs120 billion
over 10 years (at current prices), comprising: Maintenance Rs 31 billion;
Upgrading Rs36 billion; New Construction Rs27 billion; and access to Kathmandu
Rs18_billion. The annual budget is estimated to be in the range from Rs8.5 billion
today to Rs13 billion by 2016.

Key elements of the 10-year Plan include:

• a significant expansion of the Strategic Road Network including the

incorporation of substantial lengths of the local road network and the
extension of the network to serve all Districts;
• the upgrading of the more important earth and gravel strategic roads in
both the hills and Terai to provide improved all-weather accessibility;
• consideration of a new strategic access link between Kathmandu and the
Terai; and
• strategic improvements within Kathmandu Valley, including upgrading the
Ring Road and route to Kathmandu and Dhulikhel.

DHV - SILT - TAEC – CMS Chapter II-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 3
III. STRATEGIC ROAD NETWORK

A. Background

1. Function of Strategic Network

The function of the Strategic Road Network (SRN) is to provide linkage and
connectivity throughout the country. The network is critical for all economic and
social activity: it enables economic growth and development to take place,
encourages prosperity and an equitable distribution of wealth, and ensures social
wellbeing within the community together with regional and cultural integration.

Development of the SRN aims to enhance national levels of social and economic
integration, redress regional inequalities and imbalances, and reduce levels of
exclusion in the remote and more disadvantaged areas.

The SRN is a fundamental contributor to the improvement of accessibility in the

remote areas of Nepal: it provides connections to the rest of the country and
allows the development of local networks within individual Districts. The SRN
‘facilitates’ the provision of access to remote communities, rather than necessarily
provide access directly.

Nepal’s transport sector is heavily dominated by road based transport, which

accounts for almost all domestic passenger and freight movements. Air services
contribute to passenger movements to key commercial and tourist destinations,
and to the transport of both passengers and goods into remote hill areas.
Internationally, most freight movements are across the land border with India,
with internal distribution by road: limited high-value goods are carried by air.

The future and function of the SRN is thus critical to the development of the
country. Roads are a pre-requisite of most economic activity and provide links
between towns and regional centres throughout the country, and increasingly into
the more remote hill areas. In the past, economic development has been
concentrated in readily accessible areas of the Terai and around the main urban
areas: the development of the road networks in these areas has enabled and
encouraged this growth. The expansion of the network into the hills can now
provide similar benefits and opportunities to these presently remote communities.

Overall responsibility for the development and maintenance of the SRN lies with
the Department of Roads (DoR), within the Ministry of Physical Planning and
Works (MoPPW). It is the function of this Study to review the status of the SRN
and to recommend a prioritised programme for the maintenance, improvement
and expansion of the network over the coming ten years.

2. Historical Development of Network

The road network in Nepal has been developed entirely over the past 50 years
since the opening of the first road link to Kathmandu from India – the Tribhuvan

DHV - SILT - TAEC – CMS Chapter III-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Raj Path – in 1956. Prior to this there were a number tracks and trails in the Terai
linking to nearby towns in India and a limited network of roads within Kathmandu
Valley on which a few vehicles operated that been carried in over the hills.

Initially development of the strategic road network was slow and, by 1970, only a
single north-south link (330 km) had been constructed – linking Kathmandu to the
Chinese border at Kodari and to the Indian border at Raxaul. Since 1970, there
has been considerable expansion of the strategic network with the assistance of
many international development agencies.

Both India and China were instrumental in the construction of key elements of the
national network: India with significant sections of the main East-West Highway
and with the initial north-south routes from the Indian border to both Kathmandu
and Pokhara, and China with the construction of the road to the Chinese border
and the main mid-hills link between Kathmandu and Pokhara. Subsequently other
bi-lateral and multi-national agencies have been active in the roads sector, with
significant sections of strategic network being funded by the World Bank, ADB,
the US, former USSR, the British, the Japanese and the Swiss.

During the 1980s and 90s the emphasis of the aid community focussed on the
rehabilitation of the strategic network and the development of a sustainable
maintenance programme: more latterly the attention of the funding agencies has
turned to the further development of road networks to serve the remote hill areas.

In addition to extensions of the strategic network, a considerable expansion of the

local rural road network has occurred over the past 10 years, with construction
undertaken through a large number of local initiatives and agencies - often
involving significant levels of local participation. The construction of rural roads
has been seen - at all levels - as an important mechanism to accelerate the social
and economic development of remote, non-road-served areas.

This growth is best illustrated by the overall road statistics maintained by DoR
since 1970, see Table III.1 and Figure 3.1. These lengths include
estimates of the total operational road network, including both Urban and District
Roads as well as the Strategic Network. These figures demonstrate a more than
five-fold increase in the total road length over the past 30 years, from around
3,000km in 1975 to over 17,000km today – an annual increase of around 6
percent.

Table III.1: Growth in the Overall Road Network, 1970-2004 (km)

Highways Feeder Urban District Total
Roads Roads Roads
1970 1,487 679 243 322 2,730
1975 1,728 789 282 374 3,173
1980 1,967 1,603 546 824 4,940
1985 1,960 1,875 866 1,224 5,925
1990 2,111 1,822 1,098 2,299 7,330
1994 2,734 1,520 1,339 3,941 9,534
1998 2,905 1,835 1,868 6,615 13,223
2002 3,029 1,832 2,198 9,775 16,834
2004* 3,339 4,196 2,260 7,486 17,281
Note: *2004 data include the re-designation of some District Roads as part of SRN (NH&FR)
Source: DoR Road Statistics, 1998, 2002 & 2004

It should be noted that the 2004 statistics show a significant increase in the
length of the SRN as a result of the reclassification of some District Roads as

DHV - SILT - TAEC – CMS Chapter III-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Feeder Roads. These changes are not however reflected in the statistics for the
“existing SRN” presented in Section 3.2 below.

Figure 3.1: Growth in the Overall Road Network, 1970-2004

20,000

18,000

16,000

14,000
Road Length (km)

12,000

10,000

8,000

6,000

4,000

2,000

0
1970 1975 1980 1985 1990 1994 1998 2002 2004
Year

Highw ays Feeder Roads District Roads Urban Roads

Source: DoR Road Statistics, 1998, 2002 & 2004

Ignoring the effect of the re-classification in 2004, the length of the SRN has
grown steadily from around 2,000km in 1970 to a little over 5,000km today – the
equivalent of 2.5 percent per annum. The ‘other’ road network, primarily the
District Roads, has grown significantly faster – at around 10 percent pa – to reach
around 12,000km by 2002 and maybe in excess of 15,000km today.

B. Existing SRN and Need for Change

1. Designated Strategic Network

The existing SRN was designated in 1994 comprising 15 National Highways and
51 Feeder Roads, with a total length today of 5,030km. The current lengths of
each category of road are summarised in Table III.2, with a full listing of
individual roads given in Annex 3, Table A3.1a.

Table III.2: Length of Designated Strategic Road Network, 2006 (km)

Bitumen Gravel Earth Total
National Highways 2,451 396 261 3,108
Feeder Roads 1,056 563 303 1,922
Total 3,507 959 564 5,030
Source: DoR Database developed for HDM, with adjustments as noted in the Annex 030

DHV - SILT - TAEC – CMS Chapter III-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

It is noted that these lengths exclude some sections of the designated SRN that
are currently under-construction 4 and which are not open to traffic: additionally
many of the ‘earth’ roads, and some of the gravel roads, are seasonal and are
impassable to light traffic for much of the year.

2. The Existing SRN

There are however significant additional lengths of road that operate as part of
the strategic network but which are not included in the above totals. The current
"defacto" SRN comprises the open and motorable sections of the ‘designated’
network (ie excluding those sections planned or under construction), plus a
number of additional roads that are currently maintained by DoR and some that
have been built or upgraded as if they were part of the SRN. These latter roads
total around 400km see Annex 3, Tables A3.1b and A3.1c and include the
Kathmandu Ring Road a number of important roads in Kathmandu valley, the
‘Katari’ Road, Bardibas-Jaleshwar, and a number of currently open sections of
road to District Headquarters that are not officially designated as Feeder Roads –
eg sections of the roads to Khandbari, Bhojpur, Okhuldhunga, Mangalsen
(Achham) and Martadi (Bajura).

These additional roads are included in the “existing defacto” SRN of 5,428km,
illustrated in Figure 3.2, which represents the best estimate of the currently open
main road network in the country. This network forms the basis for the
development for the ‘extended SRN’ and for assessing current and future levels
of accessibility: 70 percent of the network is sealed, including all the more heavily
trafficked roads, and a further 18 percent has a gravel surface. (see
Table III.3)

Table III.3: Existing ‘defacto’ Strategic Network (open)

Surface Type: Length (km) Percentage
Sealed 3,805 70%
Gravel 985 18%
Earth 638 12%
Total 5,428 100%
Note: Includes additional ‘non-designated’ sections of road

3. Network for HDM Analysis

The HDM analysis is conducted on a network of 7,917km of road, representing
the network expected to be in operation and maintained by DoR by 2010. The
network includes the existing designated SRN of 5,030km, plus an extra 440km
(which will be completed by 2010) making a total of 5,470km, as shown in
Table III.4. In addition, a further 2,447km are included which will form part of
the future SRN by 2010, and which comprise both ‘new’ links currently under
construction as well as existing District or Urban Roads of a strategic nature and
presently maintained by DoR.

4
Excluding specifically the final sections of Surkhet-Jumla, Chhinchu-Jajarkot & Khodpe-Chainpur
which are currently under construction as earth roads under GoN funding

DHV - SILT - TAEC – CMS Chapter III-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter III-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table III.4: Existing Strategic Road Network for HDM Analysis (km)
Existing (designated) Links to be Total Network for
SRN added to SRN HDM Analysis
National Highways 3,432 1,113 4,545
Feeder Roads 2,038 1,334 3,372
Total 5,470 2,447 7,917
Note: Totals differ from Table 3.2 due to inclusion of incomplete sections of designated roads
Source: HDM Database

This network (Figure 3.3) is representative of the SRN that will exist by 2010
following the completion of those committed projects that are currently in the
pipeline and for which finance is secured. This network is however marginally
different from that used to develop the ‘extended SRN’ (see Chapter 5) and the
accessibility analysis (Chapter 6) due to the inclusion, in the HDM network, of
some additional roads presently maintained by DoR.

The full HDM network thus includes the existing operational network of roads
maintained by DoR under the Annual Road Maintenance Plan (ARMP), plus
those known committed additions and improvements (both upgrading and new
construction) that will be completed by 2010. These committed improvements
total over 3,000km (see Table III.5) and are not evaluated in the HDM
analysis as they are assumed to be implemented as planned: these schemes
include improvements to the existing (designated) SRN as well as links to be
added to the SRN. A full listing of the proposed improvements and additions is
presented in Annex 3, Table A3.2, A3.3 and A3.4.

Table III.5: Committed Improvements to the HDM Road Network (km)

Existing Total Network
Links to be
(designated) for HDM
added to SRN
SRN Analysis
Committed Improvements or
1,731 1,298 3,029
construction prior to 2010
Balance of network without
3,739 1,149 4,888
programmed improvement
Total 5,470 2,447 7,917
Source: HDM Database & DoR

The ‘Committed Projects’ can be grouped based on their funding source, as

shown in Table III.6 below and presented in full in the Annex. The ADB
schemes relate to three projects and include recurrent maintenance on the EWH,
upgrading of unsealed roads (in both the hills and Terai), construction of ‘missing
links’, and new access routes to border crossings and ICDs.

DHV - SILT - TAEC – CMS Chapter III-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter III-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table III.6: Details of Committed Projects in HDM Network – by Agency

Funding Project Length
Agency
ADB RNDP 758 Road Network Development Project (in progress)
STFP 40 Sub-Regional Transport Facilitation Project
TCP 143 Transport Connectivity Project (Core Projects)
Govt of China 26 Dhunche-Rasuwaghadi
DFID / UK RAP 92 Rural Access Project (Feeder Road Component)
Govt of India 1,262 Priority Road Infrastructure in the Terai
Govt of Japan JICA 32 Dhulikhel-Sindhuli-Bardibas
WB/IDA RSDP 676 Road Sector Development Project (RMDP2), being
prepared under Part II of this Study:
Total 3,029
Note: The WB total excludes new roads and the additional schemes – see Annex

It is noted that the World Bank (RSDP) being prepared under Part II of this Study
includes additional road sections that are not part of the ‘committed’ HDM
Network for 2010. In addition to these externally-funded ‘committed’ schemes,
there are six on-going GoN funded new road construction projects involving
access to the District Headquarters of Solukhumbu, Manang, Mustang,
Jajarkot/Dolpa, Jumla and Bahjang. The Nepal Army is undertaking some of
these works and some sections may be included in the Mid-Term Programme of
the ADB Transport Connectivity Project.

C. Network Condition and Inventory

Detailed condition data are available for the existing SRN based on a series of bi-
annual surveys of the network carried out for DoR by consultants. The most
recent survey was in 2004-5. The surveys are comprehensive and have covered
most of the accessible sections of the strategic network: the latest data relate to a
total of 4,128km of the network (2,737km of Highways and 1,391km of Feeder
Roads.

The DoR surveys have recorded both the Surface Distress Index (SDI) and
roughness using the International Roughness Index (IRI). In previous years the
DoR have reported on the condition of the network based on the results of the
SDI: this is composite index which includes cracking, potholes, ravelling, edge-
break and other pavement defects. It is a useful indicator of the pavement
condition primarily for the ‘manager’ of the road network, who is concerned with
asset preservation, but it does not necessarily reflect the condition of the road as
perceived by the road user. A section of badly cracked road can remain relatively
‘smooth’, despite having a high SDI.

The condition of existing network – as determined by the SDI – is shown in

Table III.7. This indicates that over 85 percent of the surveyed network is in
‘good’ or ‘fair’ condition: it should however be noted that this network (4,128km)
excludes many of the lower trafficked and unsealed roads that are in poor
condition. Based on these criteria, the overall condition of the network can be
shown to have improved markedly over the past decade, as a result of improved
maintenance procedures and the continuing donor-funded rehabilitation
programmes. [In 1995, less that 1,200 km was recorded as being in good or fair
condition.]

DHV - SILT - TAEC – CMS Chapter III-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table III.7: Condition of Existing Network based on SDI measurements

National Highway Feeder Roads Total

km % Km % km %

Good (SDI < 1.7) 805.1 29.4% 170.3 12.2% 975.5 23.6%

Fair (SDI 1.7 - 3.0) 1751.8 64.0% 833.2 59.9% 2585.0 62.6%

Poor (SDI > 3.0) 180.4 6.6% 387.5 27.9% 567.9 13.8%

Total 2737.3 100.0% 1391.1 100.0% 4128.4 100.0%

Source: DoR Database (2004), as reported in SWRP Inception Report (Dec 2005)

Maintenance interventions based on SDI alone will not result in the adoption of
an economically optimal programme as it takes no account of the future
performance of the pavement nor of the costs incurred by the road user over the
life-cycle of the treatment adopted. The Consultants recommend that future
maintenance interventions are better identified and evaluated using HDM-4 which
is based on a life-cycle analysis of the pavement performance, coupled with an
assessment of the overall road user costs incurred. This allows the ‘total transport
costs’ – of the road agency and road user combined – to be identified and
minimised over any given plan period.

HDM-4 uses the roughness (measured in IRI) as the key indicator of pavement
condition and determinant of road user costs: the pavement deterioration models
and intervention criteria within HDM are also based on IRI. In HDM pavement
condition is also predicted in terms of cracking, potholes and other indicators, not
just IRI, and any of these can be used to trigger intervention The roughness
index is a measure of surface irregularities, expressed in metres per kilometre
(m/km) and as recorded by a Bump Integrator (BI) calibrated to the International
Roughness Index.

The current condition of the sealed sections of the HDM Network (4,235km) is
summarised in Table III.8 and is illustrated on Figure 3.4. This shows
that over 60 percent of the sealed network (2,600km) is in good or fair
condition, with less than 10 percent in bad condition. Roughness (IRI) is not
considered to be a meaningful indicator for unsealed (gravel or earth) roads
as the condition of such roads can very widely over time and – in the case of
gravel roads – can be readily restored to lower levels through grading and/or
spot re-gravelling. Further discussion and analysis of the current and future
network condition is given in Chapter 4.

Table III.8: Roughness of Existing Network (sealed roads only)

IRI (m/km) Condition Length (km) Percentage

<4 Good 721 17%

4.1 – 6.0 Fair 1,866 44%

6.1 – 8.0 Poor 1,265 30%

>8 Bad 384 9%

All Sealed Roads 4,235 100%

Source: HDM Database (2004)

It should be noted that the definitions of good, fair, poor and bad are to a large
extent arbitrary and can be adjusted to suit individual country conditions. The

DHV - SILT - TAEC – CMS Chapter III-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

ranges of IRI selected – with break points at 4, 6 and 8 – are considered suitable
and appropriate for Nepal.

It has been necessary to re-calibrate and re-calculate the 2004 roughness data
as previously reported by DoR (see Chapter 4) as the data were clearly erratic
and were grouped in a narrow band of observations: most data were clustered
between 4 and 6, with few roads in either the good or poor/bad categories. The
adjusted data are presented in Table III.8.

D. Criteria for Inclusion in Extended SRN

It is evident that there is a need to re-classify and re-define the extent of the
existing and future (extended) SRN. The previous classification and designation
was established in 1994: since then the network has been expanded significantly
and the relative importance of some roads has changed. The last ten years has
seen the construction and upgrading of many feeder roads into the hills and a
massive increase in the length of the local (District) road networks.

This increase in the overall length of rural roads requires a clear demarcation of
function and responsibility: the DoR-sponsored strategic network should provide
access into each District (and between Districts), whilst the local access within
each District should be the responsibility of the local bodies. It is clear that the
DoR network of Feeder Roads (and Highways) into the hills should be built and
maintained to higher (all-weather) standards, befitting the strategic nature of the
connections provided. The Strategic Road Network (SRN) is the prime enabler of
access into the hills, allowing and encouraging the development of local road
networks and the consequent stimulation of the local economy.

Chapters 5 and 6 describe the proposals for the expansion of the SRN and the
prioritisation of the necessary improvement works. The first requirement is to
establish the strategic nature and function of each link, which are not necessarily
dependent on either the traffic volume or the economic viability. A set of criteria
for inclusion in the SRN was established in 1994 and this remains valid today.
The second phase of the planning process determines the appropriate standards
and justification for construction or improvement – within a constrained budgetary
environment.

As indicated in the following Chapters, the expansion of the SRN includes three
elements: (a) committed additions to the SRN, including foreign aid funded
projects; (b) the re-designation of existing local or District roads as part of the
SRN; and (c) roads on new alignments.

E. Traffic Analysis

1. Traffic Data and Surveys

The DoR have been collecting traffic data on a systematic basis on the main road
network since 1988, although the records are incomplete and inconsistent. More
reliable data have been maintained since the mid-1990s when MRCU established
a series of automatic traffic count sites (using ‘traffic loggers’) coupled with
regular manual traffic counts at sites throughout the country. Data from the
loggers however are only available for the period up to 2001, since when the
automatic counts have been discontinued.

DHV - SILT - TAEC – CMS Chapter III-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter III-11

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Manual traffic counts have been undertaken on a bi-annual since the late-1990s
at 115 sites at key locations on the strategic network. The latest data available
are from the counts taken in 2004/05. It should be noted though that traffic flows
and movements have been severely disrupted in the years since 2000 as a result
of the increasing insurgency and deteriorating security situation. Over this period
there have been frequent bandhs (strikes) and closures which have distorted
traffic patterns and suppressed demand: additionally frequent security checks on
all main roads will have deterred travel and movement in many areas controlled
by the Maoists has been restricted. A considerable increase in domestic air travel
between the key commercial hubs has also been experienced in the past few
years, much of which has been attributed to the difficulties of road travel.

The Consultants undertook a series of three-day (72 hour) manual classified

counts at eight key locations on the SRN in April/May 2006 to validate the most
recent DoR counts and to provide a check on traffic volumes and vehicle mix. In
general the counts confirmed that traffic volumes are depressed below the level
that might otherwise be expected. The counts also confirmed that there has been
a very substantial increase in the numbers and proportion of motor-cycles in the
overall flow – including longer-distance movements.

The Consultants have used the 2004/05 counts as the basis for establishing
current traffic volumes on the existing road network. Reference has also been
made to the historic data in order to estimate the current growth rates and
changes in traffic composition. Specifically data from the loggers over the period
1995-2001 has been used to establish growth rates during a period of more
stable activity.

A selection of count data from the more reliable and consistent count sites are
presented in. No data are available for 2002-03 or 2005, and the counts in 2004
were seriously influenced by the security situation. The 2006 data shown are
from the Consultants counts taken in May: again these data were affected by the
disturbances.

Growth in traffic volumes at the four Central Region sites are illustrated in
Figure 3.5 (a) and (b). Long-term annual growth rates in the Central Region can
be seen to be between 5 and 8 percent, with the key sites at Nagdhunga (or
Thankot) (at the entry to the Valley) and Pathlaiya (north of Birgunj) showing the
most consistent pattern of growth. The counts at sites outside the Central Region
show extreme variability, especially since 2000.

Figure 3.5 (a): Traffic Growth at Central Region Sites (Thankot & Pathlaiya)

DHV - SILT - TAEC – CMS Chapter III-12

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Figure 3.5 (b): Traffic Growth at Central Region Sites (Mugling & Panchkhal)

Table III.9: Historic Traffic Count Data at selected sites (AADT, 1992-2006)
Western Mid & Far
Central Region Eastern Region
Region Western Region
Pathl'ya N

Itahari (N)
Bharatpur

H0140/44

H1010/12

Ataria (N)
Bard'ghat
Panchkal
H0310/1

H0502/3

H0158/9
Thankot

Walling
Charali
H0214

H0129

H0102

H0804

H1403
Babai
Year

1992 1665 329

1993 1596 353
1994 2237 360 1817 1379
1995 2482 513 1838 1492 879 936 154 163 95
1996 2684 537 2146 1620 966 1330 1111 151 188 102
1997 2606 543 2303 1731 997 1432 1125 166 216 137
1998 2913 656 2372 1761 1371 1492 1524 263 227 157
1999 2608 794 2650 1916 1134 1521 2214 545 429 191
2000 2842 901 2368 1895 1075 1630 2157 332 513 360
2001 3985 - 2969 2245 995 1605 2545 273 822 329
2002 - - - - - - - - - -
2003 - - - - - - - - - -
2004 4558 744 3169 2575 1743 1910 2382 751 655 458
2005 - - - - - - - - - -
2006 4440 1000 3042 2852 2752 199
Growth
7.8% 7.8% 4.9% 6.2%
Rate
Source: DoR counts 1992-2004; Consultants Surveys 2006
Note: All figures exclude Motorcycles (although these were included in the 2004 & 2006 counts)

2. Existing Traffic Volumes

The Consultants have based their estimates of current traffic volumes on the DoR
counts for 2004. These are the most comprehensive and consistent available
data, despite the uncertainty surrounding the period in which the surveys were
undertaken. Fully classified counts are available for 115 sites throughout the SRN
including at least one count on each designated Highway and Feeder Road. The
2004 survey was the first time that motor-cycles were included and thus the total

DHV - SILT - TAEC – CMS Chapter III-13

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

motorised vehicle (MV) count must be treated with care, especially when making
comparisons with earlier counts.

Data for the 115 sites, divided by the five major road types, are presented in
Table III.10. A total of 41 sites are located on the Core National Highways,
including the East-West Highway, links to Kathmandu, major Terai towns and
border crossings: these counts averaged a little over 3,000 vpd, with trucks and
buses each comprising 22 percent of the total, cars and ‘other’ 11 percent each,
and motorcycles 33 percent. On the ‘other’ National Highways and Feeder Roads
in the hills, traffic volumes are typically around 500 vpd, with trucks and buses
comprising 35 percent, cars and ‘others’ around 30 percent, and motor-cycles 35
percent. Traffic counts on Feeder Roads in the Terai averaged around 1,000 vpd,
with almost 50 percent motor-cycles: in Kathmandu Valley flows are substantially
higher (with some roads over 10,000 vpd) with cars and motor-cycles dominating.

Table III.10: Summary Data from DoR 2004 Traffic Counts

(Average of observed classified counts)
No of
Truck Bus Car Other M/C Total
Sites
Core National Highways 41 717 718 350 360 1051 3194
22% 22% 11% 11% 33% 100%
Other National Highways 21 92 87 79 68 175 501
18% 17% 16% 13% 35% 100%
Feeder Roads (Hills) 27 65 93 57 74 162 452
14% 21% 13% 16% 36% 100%
Feeder Roads (Terai) 18 141 132 83 214 488 1058
13% 12% 8% 20% 46% 100%
Feeder Roads (KTM) 8 808 1445 3879 1268 4522 11922
7% 12% 33% 11% 38% 100%
115

It is noted that many of the count sites are located close to settlements or urban
areas and thus may over-estimate traffic volumes due to the inclusion of localised
movements. This is particularly true for some Feeder Roads where the count has
been taken near the start of the road and may not be represent flows along the
whole length.

The Consultants have assigned 2006 traffic volumes (and associated vehicle
compositions) to all 508 links in the HDM network of sealed and unsealed roads
(7,917km). These volumes are based on the 2004 counts, inflated for 6 percent
annual growth and adjusted to remove any bias from the count location. Further
adjustments were made for sites where the 2004 counts were clearly unreliable
or not consistent with previous data.

The resultant traffic flows are illustrated in Figure 3.6, on which the band-width
indicates the daily traffic volume, including motor-cycles. It can be immediately
seen that the major traffic flows are concentrated in the Central Region and,
specifically, on the routes between Kathmandu and the Indian border. Traffic
volumes are also generally higher in the east of the country than in the west:
flows on the East-West Highway are between 2,000 and 3,000 vpd in the east,
between 3,000 and 4,000 vpd in the centre, and less than 2,000 in the west. With
few exceptions, traffic volumes on hill roads are less than 500 vpd.

DHV - SILT - TAEC – CMS Chapter III-14

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter III-15

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The lengths of road within each traffic flow band are given in Table III.11
for both sealed and unsealed roads separately. Over 80 percent of the network
carries less than 1,000 vpd: only 1,530 km of mostly sealed roads have more,
and only 130 km (wholly within Kathmandu Valley) has over 5,000 vpd. For
unsealed roads, almost all (98 percent) carry less than 500 vpd and over half
have less than 100 vpd.

Table III.11: Distribution of Road Length (km) by Traffic Volume (2006)

Sealed Unsealed Total
AADT Length Percent Length Percent Length Percent
< 100 141 3.3% 1,999 54.3% 2,140 27.0%
100 - 200 417 9.9% 919 25.0% 1,336 16.9%
200 - 500 1,262 29.8% 687 18.7% 1,949 24.6%
500 - 1,000 926 21.9% 36 1.0% 962 12.2%
1,000 - 2,000 655 15.5% 40 1.1% 695 8.8%
2,000 - 5,000 706 16.7% 0.0% 706 8.9%
> 5,000 129 3.0% 0.0% 129 1.6%
Total 4,235 100.0% 3,681 100.0% 7,917 100.0%
Source: HDM Database; 2006 Traffic Volumes

For the roads subject to detailed Feasibility Study in Part II, a more extensive
traffic investigation has been adopted, including surveys of existing movements
and considerations of trip generation potential in the areas served. Current traffic
levels on many of the roads studied are severely constrained due to the present
condition of the road.

3. Traffic Forecasts
Long-term growth rates on key links in the network are generally in the range 6- 8
percent per annum, with a doubling in traffic volumes in 10 years. Individual links
may experience higher growth – especially in urban areas and with the inclusion
of motor-cycles. Over most of the network, considerations of traffic growth rates
have little impact as network capacity is not a primary concern and only limited
sections of road are likely to approach capacity in the plan period.

Within the HDM model, traffic forecasts are based on a series of growth rates for
individual vehicle types applied to the current traffic mix on each road section.
Overall an annual growth rate of 7 percent has been adopted for all vehicle types,
except trucks (5 percent) and NMT (3 percent). These rates were determined
based on examination of recent traffic data. The results of the HDM analysis are
not particularly sensitive to the growth rates assumed, as most of the network has
adequate capacity and upgrading thresholds are best determined on an individual
scheme basis.

F. Bridges on SRN

1. Bridge Inventory
An inventory of the existing bridges on the SRN was undertaken on 2004 as part
of the RMDP. In total the inventory recorded 1,056 bridges (defined as structures
over 6m in length) on the designated SRN. The full list of bridges is presented in
Annex 3, Table A3.5, with the key facts summarised in Table 3.12. The current
inventory represents an increase of 134 over the 1996 total which included 919
bridges. Most of the increase is due to either new construction or the inclusion of
structures omitted from the earlier survey.

DHV - SILT - TAEC – CMS Chapter III-16

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Of the total number of bridges, approximately 45 percent (480 bridges) are

single-span structures 5, with an average length of less than 25m; a further 20
percent (215 bridges) are 2-span, with an average length of 30m; 25 percent (274
bridges) have 3, 4 or 5 spans with an average length of 60m; and the remaining
88 bridges are multi-span structures with an average length of over 200m.

Table 3.12: Bridges on the Strategic Road Network

Average Bridge Percentage of
Number of Bridges
Length (m) Total
Single Span 479 24.7 45%
2-spans 215 30.7 20%
3, 4 & 5-spans 274 57.3 26%
6+ spans 88 213.5 8%
Total 1,056 50.2 100%
Source: Draft Bridge Inventory, SILT, August 2005

Almost half of total bridges in the country (515) are on the East-West Highway
(Table 3.13), representing a bridge every 2km with an average length of over
60m – or 31m of bridge length per km of road. This reflects the fact that the
alignment follows the foothills and crosses a myriad of rivers and water-courses,
many of a seasonal nature. Other highways and hill roads typically have one
bridge every 8km, with an average length of around 40m (or 5m of bridge per km
of road). Feeder Roads in the Terai have approximately twice this number of
bridges (one every 4km), again reflecting the increased density of rivers and
water courses in the Terai. Hill roads tend to either follow river valleys or are
located on ridges, or higher ground, crossing fewer major water-courses and
requiring fewer major drainage structures.

Table 3.13: Incidence of Bridges on the Strategic Road Network

No of Length of Bridges Length of Av Bridge
Bridges Road (km) per 10km bridge/km Length (m)
East-West Highway 515 1,028 5.01 31.39 62.64
Other Highways 293 2,081 1.41 5.56 39.48
Hill Feeder Roads 152 1,477 1.03 3.97 38.57
Terai Feeder Roads 96 444 2.16 7.43 34.34
Total 1,056 5,030 2.10 10.53 50.18
Source: Draft Bridge Inventory, SILT, August 2005

Most of the exiting bridges have been built in conjunction with the development of
the road network and, as a result, many are 35-40 years old and increasingly in
need of maintenance and repair. Over 80 percent are reinforced concrete, with
the remainder a mix of steel trusses (including Bailey Bridges), steel plate girder,
pre-stressed concrete and stone or concrete arches in approximate equal
numbers of each. There are a limited number (8) of timber bridges.

The longest structures are on the East-West Highway in the Eastern Region: the
Koshi Barrage (1.15km with 56 spans) and the Mahuli Bridge (1.13km with 226
spans). Other significant structures are the Karnali Cable-Stayed Bridge (500m,
with a 325m main span) and three suspension bridges, each of 125m span,
across the Narayani at Mugling, the Marshyandi (Gorka) and the Bheri (Surkhet).
5
This category includes a limited number of major bridges (including arch, truss or suspension bridges)
and long causeways: three quarters of the bridges are less than the average of 25m

DHV - SILT - TAEC – CMS Chapter III-17

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

2. Bridge Condition and Maintenance

The Bridge Inventory Study also undertook a condition survey of a sample of 96
bridges – from within the total of 1056 – which were identified as being in need of
repair or rehabilitation. The selected bridges included 34 bridges on the East-
West Highway, 38 on other Highways and 24 on Feeder Roads. The damage and
defects notes ranged from ‘complete collapse’, ‘scour & siltation’, ‘corrosion of
bearings’, ‘ageing’ to ‘poor condition’: 15 cases of damage due to ‘rebels attack’
were noted.

A total of 10 elements of each bridge were inspected and ranked, including: the
approach roads, bridge deck surface, parapets, expansion joints, superstructure,
bearings, abutments, piers, protection works, and river training works. The
condition of each element was ranked on a scale of 0-4, with the extent of
damage and degree of urgency also scored on a scale of 0-4. Details of the
recommended remedial works are included in the draft Study Reports: the results
of the condition survey are presented so as to enable DoR prioritisation of the
remedial works.

A specific recommendation is made regarding the widening of the bridges on the

Pathlaiya-Dhalkebar section of the East-West Highway. There are approximately
60 bridges on this section most of which have a carriageway width of only 5.5m:
this is below the minimum for Asian Highway Standards and is insufficient for two
commercial vehicles to pass in comfort. Furthermore 10 of these bridges are in
excess of 100m in length, which adds to difficulties for opposing vehicles to pass.
A detailed study of how best to resolve the issue is recommended.

3. Recommendation
This Study endorses the recommendation of the 2005 Bridge Inventory Study to
support the continued operation of the Bridge Management System which was
originally established in DoR in 1996. Procedures for the regular inspection and
maintenance of bridges were defined, although full implementation of the cycle of
inspection, planning and maintenance has not yet been established. Effective
management of bridges requires the regular inspection, collection and recording
of accurate bridge data, together with design and construction records.

There is an on-going need for the training of engineers in the process and
procedures of bridge inspection, as well as in updating the record system.

In addition to bridges on the Strategic Network, the DoR also constructs, and is
responsible for, a number of key bridges on the local road network, which provide
valuable access and linkage into rural areas. This practice should be continued,
as DoR have the appropriate skills and expertise, although separate funding
procedures should be adopted (see Chapter 10).

It is recommended that a specific programme of bridge rehabilitation and repair

be instigated in response to the Bridge Inventory and Condition Survey in 2005,
which identified at least 100 bridges in need of remedial works.

DHV - SILT - TAEC – CMS Chapter III-18

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Chapter 4
IV. MAINTENANCE AND UPGRADING OF THE DOR NETWORK

A. Introduction
1. Objectives
This Chapter focuses on the analysis of the on-going maintenance and upgrading
requirements of the existing – and likely future – Strategic Road Network using
the World Bank’s Highway Development and Management Model (HDM-4). The
aim is to identify the extent of the optimum expenditure on maintenance over the
10-year planning period as an essential first component of future planning and
budgeting. Specifically, in regard to HDM, the Study has two objectives:

• To produce an economically optimum 10 year expenditure plan for all

roads under the DoR using HDM-4 as the primary tool for evaluation.
• To institutionalise the use of HDM-4 in DoR as a tool for expenditure
planning, prioritising roadworks and feasibility studies.
2. Network Definition
The road network forming the subject of the expenditure plan aimed to include all
roads for which DoR will be responsible over the period from 2007 to 2016. This
comprises:

• Existing links in the currently designated SRN.

• Links under construction or with committed funding which will form part of
the expanded SRN.
• Other roads which are not part of the designated SRN but for which DoR
has ‘de facto’ responsibility.
This network, comprising 506 sections with a total length of 7,917km, was
described in the previous chapter and was illustrated in Figure 3.3. In practice the
network represents the situation that will most likely exist in 2010, following the
completion of schemes currently under construction or in an advanced state of
planning, and the formalisation of many local roads as part of the SRN.

3. General Approach
The expenditure plans were assembled from two sources:

• Major works (resurfacing, rehabilitation, improvement) which could be

evaluated and optimised using HDM-4 on the basis of total transport cost
analyses.
• Other works (routine, recurrent and emergency maintenance) which are
essential to preserve the integrity of the network but which cannot be
prioritised using economic models. These costs were based on existing
DoR norms.

DHV - SILT - TAEC – CMS Chapter IV-1

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Data sources for the analyses came primarily from existing data from the DoR
Highway Management Information System (HMIS) and previous consultants’
studies.

B. Road Network Data

1. Road Inventory
DoR has an established system for dividing roads into links on the basis of
administrative boundaries, major river crossings and major towns. Where a link
is only partially sealed, it was divided into two or more analysis sections. Thus
the first link on Feeder Road F10 is designated as Link F1001. The first 9.5km of
the link is sealed and this section was designated as Section F1001A and the
remaining unsealed 8.5km as Section F1001B.

For HDM analysis a number of geometric parameters are needed, of particular

importance being horizontal and vertical geometry and pavement width. This
data is required for both sealed and unsealed roads. Road width was obtained
from the HMIS database while other geometric characteristics were derived from
various sources including a GPS survey in 2004 covering 2,000km of the network.

For sealed sections, the type and thickness of surfacing, overall pavement
strength (SNP) and pavement age are the main determinants of pavement
performance. This data was based on construction records and DCP tests
carried out during a previous PIP study in 1997.

Figure IV.1 shows the percentage of network length by pavement width.

Single lane roads predominate with major highways being two lane. There are
relatively few sections with intermediate lane width.

50%

45%
Percentage of Length of Road (km)

40%

35%

30%

25%

20%

15%

10%

5%

0%
<4 4-5 5-6 6-7 >7
Pavement Width (m) - Sealed Roads Only

Figure IV.1: Distribution of Pavement Width for Sealed Roads

Figure IV.2 shows the distribution by horizontal alignment expressed in

degrees per km of curvature (the measure used in HDM-4). Not surprisingly
there is a very high proportion with extreme values of curvature.

DHV - SILT - TAEC – CMS Chapter IV-2

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

30%

25%

Proportion of Network
20%

15%

10%

5%

0%
< 100 100 - 250 250 - 500 500 - 750 750 - 1000 1000 - 1500 > 1500
Curvature (deg/km)

Figure IV.2: Horizontal Alignment

2. Road Condition
The existing SRN has been subject to regular traffic and condition surveys for a
number of years. The last set of surveys was carried out in late 2004/early 2005.
The data has been screened for reliability and some adjustments made. This
applied in particular to the roughness data and a recalibration was carried out as
described in the Part II Interim Report (October 2006).

For sealed sections not included in 2004/5 condition survey, estimates of

roughness and cracking were based on the pavement age and surfacing type –
Surface Treatment (ST) or Asphaltic Concrete (AC).

Little surface maintenance is done on gravel and earth roads and they typically
have a high roughness. A standard value of 20m/km IRI was assumed for all
unsealed roads.

Figure IV.3 shows the current service level of the sealed network in four
classes. This uses roughness as the measure of service to road users and the
proportion of the network in poor and bad condition is much higher than previous
DoR ratings that used Surface Distress Index (SDI) as a measure of condition.
SDI is dominated by cracking and ravelling, defects that are important to road
engineers but may not be apparent to road users; a road may be heavily cracked
but have a good ride quality. Conversely there are many sections that have
received a recent reseal with consequent low levels of SDI but have a high
roughness.

DHV - SILT - TAEC – CMS Chapter IV-3

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

50%

45%

Percentage of Length of Road (km)

40%
Fair
35%

30%

25% Poor

20%

15%
Good
10%
Bad
5%

0%
<4 4-6 6-8 >8
IRI (m/km) Sealed Roads Only

Figure IV.3: Current Service Level of the Sealed Network

3. Traffic
Where available, classified traffic count data from 2004/5 was adjusted to create
2006 volumes. For links without count stations, volumes were interpolated from
adjacent counts or estimates made using the extensive knowledge of the DoR
and Consultants’ staff.

The main contribution to axle loading is the two-axle truck. Axle load surveys
have shown a considerable variation in the average loading of these trucks
depending on the location of the road; axle loads are higher on the strategic
corridor between India and Kathmandu and on certain roads in the Terai. On
feeder roads in the hills axle loads are generally lower. To cater for this variation,
3 types of truck (light, medium and heavy) were used in HDM-4 and the volume
of trucks allocated to the appropriate type following the guidelines in
Table IV.1.

Table IV.1: Allocation of Truck Types by weight

Road Type Terrain Truck VDF
Core Network All Medium 4.0
Other highways Terai Medium 4.0
Other highways Hills Light 2.5
EWH to border Terai Heavy 7.0
eeder roads Terai Medium 4.0
Feeder roads Hills Light 2.5

Figure IV.4 shows the distribution of road length by traffic level. Unsealed
roads have, generally, quite low volumes although there are a few unsealed
sections, mainly in the Terai, with over 500 veh/d. The roads with more than
5,000 veh/d are mostly in the Kathmandu Valley while the core network is mostly
in the range of 1 – 5,000 veh/d.

DHV - SILT - TAEC – CMS Chapter IV-4

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

2,500

2,000

Length of Road (km)

1,500
Unsealed
Sealed
1,000

500

-
< 100 100 - 200 200 - 500 500 - 1,000 1,000 - 2,000 - > 5,000
2,000 5,000
AADT (excl motorcycles)

Figure IV.4: Traffic Distribution

4. Ongoing and Committed Projects

As described in Chapter 3, the HDM network includes a number of links for which
construction or improvement is ongoing or funding is already assured. The total
length of committed works is 3,029km. In many cases it was decided to exclude
committed sections from further analysis using HDM. The reasons for excluding
sections were:

• overlay is ongoing – there should be no need for further major

maintenance in the 10 year plan period;
• improvement to sealed standard is ongoing and no further improvement is
needed; or
• reconstruction of existing sealed road – not further major maintenance
should be needed in the 10 year plan period.
Other committed sections were included because:

• resealing will be complete by end 2006 and further reseal my be needed

before 2016; or
• construction is to earth or gravel standard and the section may be a
candidate for improvement to sealed standard after completion of the
current project in 2010.
In total, 110 sections with a total length of 2,450km were excluded from the
analysis leaving 396 sections with a total length of 5,467km to be evaluated in the
constrained budget analysis. This, conveniently, reduced the number of analysis
sections below 400, a constraint contained in the HDM program: a later section
describes how this enables a constrained budget analysis to be performed with
HDM-4.

C. Maintenance and Improvement Standards

1. Resurfacing and Rehabilitation
Periodic maintenance (resurfacing) options followed current DoR practice. Most
sealed roads have a DBST surfacing and periodic resealing is done with a single
surface dressing. Only for the relatively few roads with higher traffic volumes are
asphaltic overlays used.

DHV - SILT - TAEC – CMS Chapter IV-5

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

At a certain level of deterioration it is impractical to reseal as this has little effect

on roughness. In these cases DoR normally scarifies and reshapes the existing
pavement, adds a new layer of granular base and new surfacing. For higher
traffic volumes, asphalt surfacing is used. Table IV.2 shows the
intervention levels applied in HDM-4 for the resurfacing and rehabilitation options.
These thresholds were an attempt to replicate current DoR policy and they are
not necessarily the economic optima. As a sensitivity test, another set of
intervention levels was applied with overlay used on the lower traffic ranges when
roughness was deemed to be unacceptably high: this is reported in Section 4.6.3
below.

Table IV.2: Intervention Levels for Resurfacing and Rehabilitation

AADT Range (incl motorcycles)
IRI Range < 1,000 1,000 – 3,000 > 3,000
<8 Seal Seal Overlay
8 – 10 Seal Rehab Rehab
> 10 Rehab Rehab Rehab

2. Road Improvement
For unsealed roads, the only major work considered was improvement to sealed
standard. In pavement terms, this normally comprises a new base layer with
DBST surfacing, though recent projects have trialled the use of Otta seals. Such
improvements are normally accompanied by improved drainage and construction
of retaining and breast walls to alleviate slope instability.

3. Unit Costs
Table IV.3 shows the unit costs of roadworks adopted for the
HDM analyses. These are simplified rates, averaged across Disticts and
between the hill and terai derived from unit rate analyses and compared with
recent contract prices. A detailed rate analysis is presented in Annex 4.1A.

Table IV.3: Unit Costs of Roadworks

Unit Cost (Rs)
Item Unit Economic Financial
SBST m2 97 105
40mm overlay m2 506 550
100mm base + DBST m2 874 950
200mm base + A/C m2 1,288 1,400
Improvement (Terai) km 3,680,000 4,000,000
Improvement (Hills) km 5,520,000 6,000,000
2
Patching m 368 400
Edge Repair m2 368 400
Spot regravelling m3 74 80

Annex 4.1B, defines the maintenance and improvement standards in terms of the
input parameters required by HDM-4.

D. Application of HDM-4
1. Model Calibration
The Consultants’ Terms of Reference required a Level 2 calibration of HDM-4.
This covers the models for road user effects (RUE) and road deterioration and
works effects (RDWE). Annex 4.2A describes the calibration of the RUE models
and Annex 4.2B the calibration of RDWE models. In the latter case, calibration
focussed on the most important determinants of agency and road user costs
which are:

DHV - SILT - TAEC – CMS Chapter IV-6

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

• cracking initiation and progression

• roughness progression
• roughness after resurfacing and rehabilitation
2. Data Import
The road network comprised 506 analysis sections. To enter this data manually
into HDM-4 would take a considerable time and would probably lead to multiple
errors in data entry. Thus the HDM-4 data import facility was used. The following
steps were followed:

• Road characteristics and traffic volumes were compiled in two separate

spreadsheets.
• Road characteristics data was imported into an Access database in the
correct HDM export/import format (Sections table).
• The traffic data was programmatically converted into the linear form
required for the Traffic table in the Access database.
• The Access database was imported into HDM-4 which automatically
checks for errors during the import process.
3. Programme Analysis
HDM-4 offers three types of analysis – project, programme and strategy level.
For this study the programme level analysis was appropriate. To allow for
constrained budget optimisation over a 10 year period, a matrix of alternatives
needs to be defined for each section. For sealed roads, the initial treatment may
be resurfacing or rehabilitation in each of 10 years giving 20 potential section
alternatives to be analysed. For unsealed sections the only alternative
considered was improvement to sealed standard. Given the large ongoing and
committed programme described above, it was considered unrealistic to consider
new upgrading projects in the first four years of the planning period. Thus,
improvements were only assigned for the years 2011 – 2016 giving 7 alternatives
(with the base case) for unsealed road sections.

In addition a base alternative (do-minimum) must be defined for each section.

For sealed roads this included pothole and edge break patching and rehabilitation
if roughness exceeds 12m/km in year 11 or later. For unsealed roads spot re-
gravelling was applied annually.

The analysis period used was 20 years and after the initial treatment a long term
periodic maintenance standard was applied. This comprised resurfacing triggered
by either cracking or roughness thresholds.

Table IV.4 shows how a series of section alternatives is defined.

Table IV.4: Example of Section Alternatives
Section Alternative Year Work Item Intervention
RES1 2007 Resurfacing IRI < 8
2008 Resurfacing Cracking > 10%
RES2 2008 Resurfacing IRI < 8
2009 Resurfacing Cracking > 10%
RES3 2009 Resurfacing IRI < 8
2010 Resurfacing Cracking > 10%
RES4 2010 Resurfacing IRI < 8
2011 Resurfacing Cracking > 10%
etc to
RES10 2016 Resurfacing IRI < 8

DHV - SILT - TAEC – CMS Chapter IV-7

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Section Alternative Year Work Item Intervention

2017 Resurfacing Cracking > 10%
Many of the section alternatives will be inappropriate for some sections; for
example the roughness may already exceed the limit for resurfacing. Manual
assignment of alternatives is a slow process and HDM-4 has the facility to copy
all the section alternatives from one section to others. This greatly speeds up the
process of setting up the analysis and, consequently, standard sets of
alternatives were applied to all sealed or unsealed sections. This is a trade-off
between set-up time (shorter) and run time (longer because many inappropriate
alternatives are analysed). However, run time is not a major problem as the
computer can be left to run while the user performs other tasks.

Annex 4.3 defines in more detail the assignment of section alternatives in HDM-4

Table IV.5 shows other parameters used in the analysis.

Table IV.5: Analysis Parameters

Parameter Value
Condition year 2006
Traffic year 2006
Start year 2007
Analysis period 20 years
Discount rate 12%
Include accident costs? No

4. Optimisation
HDM-4 provides three methods for producing optimised works programmes
under constrained budgets. All three have the same objective function – to
maximise Net Present Value for the whole network subject to financial constraint
on the total budget for agency costs. The methods are:

• Total enumeration; this produces a wholly correct mathematical solution

but it is limited to a small number of sections.
• EBM32 which is an expanded version of the Expenditure Budgeting
Module which was add-on application to HDM-III. This uses the Ahmed
method of effective gradients and is generally considered to give a nearly
optimum solution. There are certain limits to this method including the
number of analysis sections (400) and the number of alternatives per
section (17). Although the documentation claims that up to 12 budget
periods can be used it was found in practice that the limit was 6.
• Incremental benefit/cost ranking; this method has no limit on the number
of sections or alternatives tested. This method was initially applied in this
study because of the number of sections (over 500) but the results were
disappointing and there is clearly some defect in either the programming
or specification for this method.
To overcome the limitations of the software, the scope of the programme analysis
was reduced to less than 400 sections. This was done by excluding ongoing and
committed sections from the analysis, as described earlier. The number of
alternatives tested for sealed roads was also reduced from an initial 21 to 11.
This was done by combining resurfacing and rehabilitation into the same
maintenance standard and applying the mutually exclusive intervention levels
given in Table IV.2.

DHV - SILT - TAEC – CMS Chapter IV-8

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

E. Other Maintenance Expenditures

HDM-4 is suitable only to evaluate periodic maintenance and major improvement
works. Although it has a limited capability to estimate needs for routine
maintenance works (patching, edge break), it cannot analyse many of the
roadworks expenditures incurred in Nepal such as landslips.

DoR has a comprehensive definition and budget headings for different types of
expenditure:

• Routine Maintenance: Maintenance required at all times on every road

because of traffic movement as well as environment degradation. It
covers grass cutting, drain cleaning, removing debris and draining water
from the carriage way, bridge and culvert maintenance, and road furniture
maintenance etc.
• Recurrent Maintenance: Maintenance required at varying intervals
during the year with a frequency that depends mostly on volume of traffic.
It is repairing pot holes and ruts, dragging and grading on unpaved roads
where, as on paved roads, it covers repairing pot holes, patching,
repairing edges and shoulders, sealing cracks and small repairs of
structures.
• Specific Maintenance: This type of maintenance is done as required by
the road condition. The pothole maintenance in large scale, gravelling
the shoulder, construction of retaining walls, drain repair, replacing the
traffic signs and cross drains etc.
• Traffic Safety: Every Road Division is allocated some funds to improve
Traffic Safety measures and to repair the existing traffic signs, delineator
posts, road markings and other traffic safety related works.
• Bio Engineering: These works are carried out by all divisions to stabilize
road side slopes where low-cost bioengineering techniques suitable.
• Emergency Maintenance: Required for immediate opening of road
closures during disasters like flood, land slides and earthquake. It covers
removal of debris and other obstacles, placement of warning signs and
diversion works.
• Damage Rehabilitation Works: For available budget a prioritized list of
damage rehabilitation works is made. Then these works are carried out
through contractors.
• Equipment Maintenance and Mobilization: There are numerous known
locations along the strategic road network, which are highly vulnerable to
road closure during monsoon and other natural calamities. In order to
minimize the risk of long duration road closure, DOR every year prepares
an emergency action plan and deploys necessary heavy equipment,
manpower and construction materials in all potential and pre-identified
locations.
• Other Works (Miscellaneous): These recurrent costs include weighing
bridges, security arrangements of major bridges and divisional and
regional offices, road neighbour empowerment and monitoring and
evaluation works.
The need for many of these works varies considerably between road sections
and from year to year. To obtain an order of magnitude cost for planning
purposes, network average costs per km have been applied as shown in
Table IV.6.

DHV - SILT - TAEC – CMS Chapter IV-9

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Table IV.6: Annual Maintenance Expenditures

Expenditure Category Rs/km/year
Routine Maintenance 30,000
Recurrent Maintenance 35,000
Specific Maintenance 25,000
Traffic Safety 1,200
Bio Engineering 1,000
Emergency Maintenance Works 1,200
Damage Rehabilitation Works 8,000
Equipment Maintenance and Mobilization 3,500
Other Works (Miscellaneous) 2,000
Total 106,900

For the network of nearly 8,000km used for the HDM analysis, the total annual
cost of the above works is around Rs 850 million per year.

F. 10 Year Expenditure Plans

1. Unconstrained Network Needs
The unconstrained optimum solution is to select the alternative for each section
that has the highest NPV. When a road network has a large backlog of
roadworks needs the result is a very unbalanced result in terms of annual
budgets – very high in the initial year, tailing off in later years. In this analysis, the
unconstrained need had two peaks – 2007 for resurfacing and rehabilitation and
again in 2011 when upgrading of unsealed roads was first tested. This is shown
in Figure IV.5.

8,000

7,000
Cost of Roadworks (Rs million)

6,000

5,000

4,000

3,000

2,000

1,000

-
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year

Figure IV.5: Unconstrained Budget Distribution

To provide a more realistic scenario for the unconstrained needs, a light budget
constraint was applied, providing the same overall total for the 10 years but with a
more equal distribution. As mentioned earlier, the software would only accept 6
budget periods and to accommodate this limitation budgets were applied for two
year periods over the 10 year plan with unstrained budget allowed from year 11
onwards to cater for subsequent maintenance works. Table IV.7
shows the budget limitations applied and forecast expenditures in, what is called

DHV - SILT - TAEC – CMS Chapter IV-10

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

in this report, the balanced unconstrained scenario. It can be seen from the Table
that expenditure is effectively ‘capped’ at the imposed limit in Years 1 & 2, and
again in Years 6 to 8, but that the full allocation is unused in Years 3 & 4 and 9 &
10. This is an indication that the overall budget provision is adequate.

Table IV.7: Balanced Unconstrained Budget (Rs million)

Period Budget Limitation Forecast Expenditure
2007 – 2008 3,000 2,999
2009 – 2010 3,000 2,276
2011 – 2012 4,000 3,998
2013 – 2014 4,000 3,999
2015 – 2016 4,000 3,009
2017 onwards unconstrained
Total 18,000 16,282

The balanced unconstrained scenario is little different to the totally unconstrained

case in terms of the types of works applied over the 10 years – only the timing
changes with works deferred from 2007 and 2011 to later years. The two are
compared in Figure IV.6 and Figure IV.7.

7,000

6,000
10 Year Expenditure (Rs million)

5,000

4,000

3,000

2,000

1,000

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

Unconstrained Balanced Unconstrained

Figure IV.6: Road works by Cost – Unconstrained Budgets

1,400

1,200
10 Year Total Length (km)

1,000

800

600

400

200

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

Unconstrained Balanced Unconstrained

Figure IV.7: Roadworks by Length – Unconstrained Budgets

DHV - SILT - TAEC – CMS Chapter IV-11

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Although, in terms of length, there is quite even distribution between the four
categories of works, in terms of cost the programme is dominated by
improvement and rehabilitation works.

The works programme given by this scenario is shown in detail in Annex 4.4 6,
and is illustrated in Figure IV.8 showing the extent of the proposed reseal, overlay,
rehabilitation, and upgrading proposals. It can be seen that most of the network –
apart from sections with committed improvements – receives some treatment in
the 10-year period. It is also evident that the overlay is applied only to those
elements of the ‘core network’ with higher traffic volumes: ie routes to Kathmandu
and the East-West Highway in the Eastern, Central and Western Regions.

The location of works proposed in the first two 2-year periods, and the remaining
six years, are illustrated in Figure IV.9, with the resultant overall network condition
after 10 years shown in Figure IV.10. It should however be emphasised that the
precise timing and scope of works should be re-examined on an annual basis and
the priorities and work programme established accordingly. The HDM programme
is indicative only of the scale and content of the likely works.

2. Effects of Budget Constraints

Recognizing that the unconstrained budget may not be affordable, constrained
budget results were produced at 75% and 50% of the balanced unconstrained
scenario, as required in the ToR.

Figure IV.11 and Figure IV.12 show the effect on different categories of
roadworks as the budget is reduced. At 75% of the unconstrained budget,
improvement works are hardly affected while rehabilitation is much reduced. At
50% of the unconstrained budget, all categories are significantly reduced with the
exception of overlays.

6
Note the final three columns show details of the alternative ‘all roads’ analysis, described in
Section 4.7 below.

DHV - SILT - TAEC – CMS Chapter IV-12

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

DHV - SILT - TAEC – CMS Chapter IV-13

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

DHV - SILT - TAEC – CMS Chapter IV-14

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

DHV - SILT - TAEC – CMS Chapter IV-15

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

7,000

6,000

10 Year Expenditure (Rs million)

5,000

4,000

3,000

2,000

1,000

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

100% 75% 50%

Figure IV.11: Effect of Budget Constraint on Roadworks by Cost

Figure IV.12: Effect of Budget Constraint on Roadworks by Length

Figure IV.13 compares the condition of the sealed network at the end of the
10 year planning period. The reduced budget apparently increases the length in
good condition; this is because more treatments are delayed towards the end of
the period.

It is evident that the proportion of the network in "poor" and "bad" condition
increases from 30 percent to over 50 percent and that in "good and fair" condition
drops from 70 percent to 50 percent. This has a significant effect not only on the
overall condition of the network (which is considerably worse at the end of the
plan period) but also on the ongoing vehicle operating costs. The overall outcome
is substantially worse on all on all counts and any reductions in the maintenance
allocations are NOT recommended.

DHV - SILT - TAEC – CMS Chapter IV-16

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

70%

60%

50%

40%

30%

20%

10%

0%
Good Fair Poor Bad

Balanced Unconstrained 75% of Unconstrained 50% of Unconstrained

Figure IV.13: Sealed Network Condition after 10 Years

3. Overlays
As shown earlier, a significant portion of the network is in fair or poor condition
(IRI > 6m/km). The current practice of using surface dressing for periodic
maintenance will not reduce roughness, only stabilise roads and prevent a rapid
increase in roughness. To reduce roughness either rehabilitation or overlay is
needed. Figure IV.7 showed a large part of the programme as rehabilitation,
a relatively expensive treatment compared with overlay (about double the cost).
To test the effect of an increased overlay programme the analysis was repeated
with a modified set of interventions that allowed overlay to be applied to any road
irrespective of traffic volume, the only constraint being to apply seals at low
roughness levels. The modified intervention criteria are shown in
Table IV.8.

Table IV.8: Modified Overlay Intervention Levels

AADT Range (incl motorcycles)
IRI Range < 1,000 1,000 – 3,000 > 3,000
<5 Seal Seal Overlay
5–6 Seal Overlay Overlay
6–8 Overlay Overlay Overlay
8 – 10 Overlay Rehab Rehab
> 10 Rehab Rehab Rehab

Figure IV.14 and Figure IV.15 show that, with the lower interventions for
overlays, rehabilitation is much reduced and sealing almost eliminated with a
consequent large increase in the amount of overlay. The overall unconstrained
10 year programme cost with a greater use of overlays is marginally higher at
Rs17.1 billion, up 5 percent compared with the previous estimate of Rs16.3 billion,
with a small increase in length (up 64km to 4,490km). The average cost of works
is increased from Rs 3.67 million/km to 3.80 million/km.

DHV - SILT - TAEC – CMS Chapter IV-17

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

12,000

10,000

10 Year Expenditure (Rs million)

8,000

6,000

4,000

2,000

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

High Overlay Threshold Low Overlay Threshold

Figure IV.14: Roadworks Expenditure with Different Overlay Policies

3,500

3,000
10 Year Total Length (km)

2,500

2,000

1,500

1,000

500

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

High Overlay Threshold Low Overlay Threshold

Figure IV.15: Roadworks Length with Different Overlay Policies

Figure IV.16, Figure IV.17 and Figure IV.18 compare network condition after 10
years with the two overlay policies. It is seen that the low threshold overlay policy
yields a significantly better result in terms of network service level.

DHV - SILT - TAEC – CMS Chapter IV-18

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

70%

60%

50%

40%

30%

20%

10%

0%
Good Fair Poor Bad

High Overlay Threshold Low Overlay Threshold

Figure IV.16: Network Condition after 10 Years with Different Overlay

Policies

100%

90%

80%
Percent in Condition Range

70%

60%

50%

40%

30%

20%

10%

0%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year

Good Fair Poor Bad

Figure IV.17: Network Condition by Year – High Overlay Intervention

DHV - SILT - TAEC – CMS Chapter IV-19

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

100%

90%

80%

Percent in Condition Range

70%

60%

50%

40%

30%

20%

10%

0%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year

Good Fair Poor Bad

Figure IV.18: Network Condition by Year – Low Overlay Intervention

4. Road User Costs

The objective of any roadworks programme is to stabilise or reduce the costs of
vehicle operation and passenger time. It is thus instructive to examine both the
gross and incremental costs to road users.

The network analysed in HDM-4 had a length of around 5,500km out of a total
network of around 8,000km, the difference being ongoing and committed projects
excluded from the analysis. The total estimated road user costs in 2007 for the
analysis network was Rs 54 billion for vehicle operating costs and Rs15 billion for
passenger time costs, a total of Rs 69 billion. If the VOC component is adjusted
pro rata to the total network length the value in 2007 is nearly Rs 80 billion which
represents about 13% of Nepal’s GDP 7. If urban and rural roads were taken into
account, VOC would amount to somewhere between 15% and 20% of GDP. This
demonstrates the importance of the road network to the country’s economy.

Table IV.9 compares the discounted (12%) costs over 10 years of the
roadworks programme for the balanced unconstrained scenario with the road
user cost savings relative to the base alternative which only had routine
maintenance in the first 10 years. As the budget is reduced the b/c ratio
increases; as resources are constrained they are concentrated on the higher
value projects.

There are, of course, user savings after year 10 but the base alternative included
rehabilitation in year 11 or later and this would distort the figures. Nevertheless,
the b/c ratios show that an adequate return is given for what can be considered a
low volume road network.

Table IV.9: 10 Year Discounted Agency Costs & RUC Savings (Rs m)
Budget Agency Costs Road User Cost Savings b/c ratio
Balanced Unconstrained 9,440 21,200 2.2
75% of Unconstrained 7,336 19,597 2.7
50% of Unconstrained 4,891 16,199 3.3

7
Latest available figure for GDP is 2005.

DHV - SILT - TAEC – CMS Chapter IV-20

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

G. Unconstrained Analysis of All Roads

The foregoing analysis has described a constrained budget optimisation of the
network, excluding sections that are subject to ongoing or committed projects. In
order to assist in prioritisation of the expansion of the SRN (described in a later
chapter) an HDM-4 analysis was carried out using the full network, including an
evaluation of those sections subject to committed projects: the objective being to
confirm (or otherwise) the feasibility and priority of these works. Due to the larger
number of sections, a constrained budget optimisation was not possible for this
network but the unconstrained results give an indication of the relative priority of
different types of work and of improvement needs on different sections.

The total unconstrained budget for the 10 year period was around Rs24 billion
compared with around Rs16 billion when the committed projects were excluded.
[The additional Rs8 billion represents the cost of the committed projects that were
excluded form the earlier analysis.] The distribution of this sum by works category
is shown in Figure IV.19 and Figure IV.20.

18,000

16,000
10 Year Total Expenditure (Rs million)

14,000

12,000

10,000

8,000

6,000

4,000

2,000

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

Figure IV.19: All Roads Analysis – Cost Distribution

3,500

3,000
10 Year Total Length (km)

2,500

2,000

1,500

1,000

500

-
Improve Gravel Road Rehabilitation Overlay Resealing
Works Category

Figure IV.20: All Roads Analysis – Length Distribution

A full listing of the results by section is given in the final three columns of Annex
4.4 and the location of the works is illustrated in Figure IV.21. It can be seen that

DHV - SILT - TAEC – CMS Chapter IV-21

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

virtually all the network is selected for maintenance or upgrading and –

specifically – the analysis confirms that almost all of the ‘committed’ projects for
upgrading in the hills (green) and Terai (blue) are confirmed as being feasible.
The exceptions to this are some of the low-volume roads in the hills which are not
shown as being justified, for example, the roads to the HQs of seven Districts:
Darchula, Bajura, Achham, Jumla, Mustang, Manang and Khotang..

The exclusion of the roads to the seven District HQs is an indication that the
upgrading of these roads at the present time cannot be justified on the basis of
the present traffic volumes. However, as has been shown elsewhere (Part II Draft
Final Report and individual road Feasibility Studies) upgrading can frequently be
supported on grounds of generated traffic, improved reliability, accessibility, and
enhanced economic and social conditions.

More importantly, the HDM-4 model supports – on traffic and operational grounds
alone – a substantial programme of sealing of roads in the Terai and the more
heavily trafficked road into the hills. As discussed above, the model also indicates
a substantial programme of (relatively expensive) rehabilitation on significant
lengths of the core network: the Consultants are of the view that a more efficient
maintenance strategy could be developed with a more extensive use of AC
overlays to prolong the life of these roads (and provide an enhanced riding
surface).

H. Road Management Systems in DoR

1. Background
Since 1993, the Department of Roads has carried out systematic network-wide
surveys of traffic and road condition. The scope of these surveys has been:

• Roughness: measurement using a TRL vehicle-mounted bump integrator

calibrated to International Roughness Index using a TRL Merlin. Data is
recorded at 1km intervals;
• surface distress: visual assessment of cracking, raveling rutting etc. This
data was recorded on a sample basis of 100m within each km;
• Traffic volume: classified volume counts of 2 – 3 days duration at around
114 locations.

Initially these surveys were carried out annually by DoR but in later years the
work was outsourced to consultants. In recent years the surveys were made
biannually, the last being carried out in late 2004 and early 2005. In addition to
the above, the there is a substantial body of knowledge within DoR about the
year of construction/reconstruction and resurfacing of pavements.

The SDI and traffic data has been used to prioritise periodic maintenance works.
The roughness data has not been used directly by DoR but has been of value to
consultants in carrying out feasibility studies.

DoR has decided to adopt HDM-4 as the primary analysis tool for planning and
programming of roadworks and in 2006 purchased the new version 2.0 of the
program. In order to provide and update the data inputs required for HDM it is
necessary to revise and expand the existing data collection activities as outlined
below, including the acquisition of automated roughness data collection
equipment. The expansion of these activities should be monitored and defined as
part of an ongoing TA support (see section 4.8.5).

DHV - SILT - TAEC – CMS Chapter IV-22

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

DHV - SILT - TAEC – CMS Chapter IV-23

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Data Collection
Surface Condition
With the adoption of HDM-4 as the analysis tool in place of the SDI prioritisation
method the format of road condition data collection and database should be
changed to be directly compatible with HDM requirements. SDI data are
recorded on a sample basis (the final 100m of each kilometre). The Consultants
have recommended that in future surface condition (cracking, ravelling, edge
break etc.) should be recorded on a continuous basis at 50m or 100m intervals
given the localised nature of many pavement defects, especially on hill roads.

Roughness

The Consultants have also recommended that an automated data capture system
should be used for roughness surveys. The most widely used low cost system is
ROMDAS with a cost of around $8,000. Such a system would enable roughness
to be recorded on the same interval as the surface condition survey.
Recommendations have also been made to tighten up the roughness calibration
procedure.

Traffic

The current traffic census system gives inadequate coverage of the existing SRN;
with the expansion of the network under DoR a much greater number of count
stations should be established. The cost of the survey need not be increased;
count duration can be reduced and counts made on a 2 or 3 year cycle. If
automated roughness data equipment is procured it can also be used for moving
observer counts which form a useful supplement to fixed location counts in
providing information on traffic distribution along links. The current vehicle
classification system should be expanded to take account of the greater variety of
vehicles being used in Nepal; 3-axle and articulated trucks, jeeps and microbuses
are not presently identified in the surveys.

Pavement Strength

Data on pavement strength is at the moment fragmentary and outdated. While

the project completion reports will help improve the database where roads are
rehabilitated and improved, there should be a systematic effort to acquire
pavement structural data for roads requiring only normal maintenance. This
could be done using either DCP or Benkelman beam testing with a target of, say,
500km per year.

Project Completion Reports

A further recommendation has been made to DoR regarding project completion

reports. Currently these focus on financial matters and neglect the technical
achievement of projects. It is proposed that supervision consultants are required
in their ToR to include the following data following substantial of a contract:

• classified traffic count following the opening or improvement of the road;

• record of pavement structure and subgrade CBR as built including DCP
tests;
• key geometric data including pavement and shoulder widths and
summary data on alignment (rise/fall, curvature).

This information should be supplied directly to the HMIS Unit in electronic format
compatible with the road database structure.

DHV - SILT - TAEC – CMS Chapter IV-24

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

Outsourcing
It is recommended that DoR continue to outsource collection of road condition
and traffic data but that data audit procedures be put in place to ensure adequate
data quality.
2. Data Processing and Storage
The HMIS unit within the DoR had, at one time, an electronic road database
using dRoad software. The DoR no longer use the dRoad software and data are
currently stored on spreadsheets.
Under a current contract, IT consultants have been engaged to develop an
extensive MIS database for DoR. This includes a road database covering
inventory, condition, traffic etc. With the decision to use HDM-4 as the analysis
tool, some of the data collection and storage formats are inappropriate and the
Consultants have provided advice on future data collection formats and the
related structure of the road database.
The new database should have some analysis capability, in particular the
identification of homogeneous road sections and transformation of data from
different tables to a structure compatible with HDM-4 network import.
3. Planning and Programming
The Consultants have provided extensive training in the use of HDM-4 to a group
of 6-7 DoR engineers, two of whom were assigned as full time counterparts. The
training involved preparation of data inputs as well as operation of HDM-4.
HDM-4 is a complex model and a user only becomes familiar with the software
after extensive use in a normal working environment i.e. using HDM-4 to produce
works programmes or feasibility studies rather than training exercises. While the
DoR staff have a basic knowledge of the theory and practice of HDM-4, it might
take 1 – 2 years of use before they become proficient. For this to happen it is
essential that there is continuity and the HDM group remain in their present posts
for the next few years.
The 10 year expenditure plans described above should be considered as a first
cut and the exercise should be repeated annually with an expanded database
and refinement of the analysis sections. As well as guiding medium term funding
needs the analyses will also identify priority works in the following 1 – 3 years.
4. Institutional Support
There will be a need for continuing support to HMIS and the HDM group over the
next few years covering the following areas:
• data collection procedures and data audits;
• operation of the road database and development of its analytical
capability;
• improved calibration of the HDM models as higher quality data becomes
available;
• use of HDM-4 for strategic expenditure planning, preparation of annual
works programmes and feasibility studies.
This support need not be continuous and could be limited to 1 – 2 months per
year of consulting services.
It is also recommended that, each year, 1 or 2 DoR staff attend the annual HDM-
4 training course run by CRRI in Delhi, normally in October. This is a 2 week
course covering all aspects of HDM-4 including road user effects, road
deterioration and works effects, model calibration and analysis.

DHV - SILT - TAEC – CMS Chapter IV-25

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

I. Performance Based Maintenance Contracts

The following section provides a brief overview of the models that are available
for administering and implementing road maintenance, based on those that are
commonly in use throughout the world. The section also contains a review of the
current Performance Based Contracts as operating in Nepal.

1. Models for Implementing Road Maintenance

There are essentially three models that can be applied to undertake basic road
maintenance activities: direct labour; regular input-based contracts (based on
time and materials or a lump sum); and output or performance based contracts.
These are described briefly below.

(a) Direct Labour

Traditionally, road maintenance was carried out directly by the road agency who
employed all the staff and owned the equipment used for maintenance. In many
countries this method was used because there was no local contracting capability
and it was therefore necessary for the government to establish its own capability
to construct and maintain infrastructure (public works department).

When well run, direct labour organizations were effective, providing flexibility and
quick response to maintenance needs, especially special or emergency works.
One of the problems was the way in which governments provided funding, usually
under several heads:
• establishment costs – salaries. pensions, housing and other employee
benefits
• capital cost of equipment
• running cost of equipment
• purchase of materials
Often there was an imbalance between the different budget heads and in some
cases, while establishment costs were met there was no funding for running
equipment or purchasing materials leading to staff sitting idle in depots and no
road maintenance being performed. Little attempt was made to monitor road
condition in a systematic way or to compare road condition with maintenance
expenditures.

(b) Input Based Contracts

In the last 20-30 years governments have been encouraged to contract out road
maintenance. In the first instance this was done by means of input based
contracts, often a schedule of rates with the contractor being paid for completed
work. Such contracts would cover either periodic maintenance/rehabilitation or
routine and recurrent maintenance, sometimes both. In most cases, contracts
were awarded annually reflecting the government accounting practices for annual
budgets.

This type of contract had the merit of encouraging efficiency by shifting the onus
to the contractor who had a financial incentive to carry out the works at the least
cost. This can, however, lead to poor workmanship and a tendency on the part of
the contractor to concentrate his efforts on work items he considers to be the
most profitable.

DHV - SILT - TAEC – CMS Chapter IV-26

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

The use of contractors does not leave the road agency free of responsibility; the
contractors completed work must be checked for quality and measured for
payment. This aspect can be carried out by consultants but is often done directly
by the agency.

Some problems have arisen from the switch to contract maintenance:

• The agency is left with a pool of staff, many unskilled labourers, which it is
politically difficult to eliminate.
• The agency has a pool of equipment with associated workshop facilities.
Too often, the staff and equipment have lingered on (at consequent cost to the
government) while road maintenance is executed by contract. The government
is, in effect, paying twice for the same service. Ideally the redundant staff should
be retired (with appropriate compensation) and the equipment and workshops
sold to the private sector but this often does not happen. One alternative, as in
New Zealand is to first corporatise and then privatize the entire direct labour
operation.

(c) Output (Performance) Based Contracts

A more recent innovation has been output based (performance based) contracts
where the contractor is paid a fixed sum to maintain all or part of the road
infrastructure to a specified level. As for input based contracts, this can cover
both periodic/rehabilitation and routine operations. Three types of PBMC might
be mentioned:
1. The contract comprises initial rehabilitation of a road in poor condition
followed by routine maintenance for a period of 3 – 5 years. In effect this is
simply an extension of the defects liability period.
2. The contract only covers routine maintenance works such as patching, drain
cleaning and vegetation control. Such a contract might be for a year or less if
funded out of a recurrent budget.
3. The contract is for an extended period(10 years, say) and the contractor is
free to decide on his own options and timing of periodic maintenance works
required to maintain the road so a specified service level (for example
roughness, cracking rut depth).
The choice of the form of contract is often dependent on the funding source. If
funded from a recurrent road maintenance budget, option 2 might be the only
possibility. In the case of external funding, the agencies normally have a time
limit on the duration of a loan and are unwilling to fund recurrent expenditure. In
this case, option 1 above might be the most appropriate with the external funding
being applied to the initial major works and the government counterpart funding
being applied to the subsequent routine works.

To adopt option 3 with a contract period of 10 years requires guaranteed funding

for the contract period. For a country like Nepal this can be problematic until
such time as the Road Fund provides sufficient income for both periodic and
routine maintenance works. For the time being it appears that options 1 and 2
must be applied if PSMC is used in Nepal.

The lengthman system which is used on many roads can be considered to be a

form of option 2; the man is paid a fixed monthly sum (his salary) to carry out a
certain level of maintenance (clean drains, short vegetation).

DHV - SILT - TAEC – CMS Chapter IV-27

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

2. PBMC in Nepal
The Consultants have reviewed two PBMC contracts that are current in Nepal,
each on heavily trafficked 75km sections of the East-West Highway to the east of
Pathlaiya.

The form of contract is the first type outlined above. The initial works comprise
an overlay and routine maintenance backlog. The PBMC starts after this work is
complete and runs for a period of about 4 years.

(a) Specifications

The PBMC component of the contracts includes performance and operational

indicators. The performance indicators are surface distress index (SDI),
roughness (IRI) and axle load control. The operational indicators comprise
specific items relating to pavement, right-of-way, structures, drainage and road
signs and marking.

(b) Performance Indicators

The use of SDI seems unnecessary. If the contractor is in full compliance with all
the operational indicators (patching, crack sealing etc) then SDI will remain close
to zero.

The indicator for roughness requires that it should not increase by more than 10%
each year. Again, if the contractor is in compliance with the operational indicators
then one would not expect the roughness to increase by more than 10% per year.
The only circumstances in which this might happen is if there is severe weakness
in the pavement structure leading to extensive deformation. This is something
over which the contractor has no control. The initial overlay is to a design given
by the client; the contractor does not have the option to provide any needed
strengthening that he considers necessary to meet the performance specification
in terms of roughness progression.

The axle load control indicator requires that no vehicle shall have an axle load
exceeding 10.2 tonnes. If the contractor were to be in full compliance with this
indicator he would need to operate permanent weigh stations at each end of the
road section and at any intermediate access points used by heavy vehicles.
Even given this, the contractor has not legal authority to prevent vehicles using a
public road. Monitoring of this indicator would require the client to operate
additional (mobile) weigh stations. This indicator seems to be impractical and
probably in contravention of legal requirements.

In summary, it is considered that the performance indicators as defined are either

unnecessary or unworkable.

(c) Operational Indicators

The operational indicators are comprehensive, though maybe idealistic given

general levels of road maintenance in Nepal. There are concerns about
monitoring these indicators which is discussed below.

(d) Performance Monitoring

The specification is based on a model document from the World Bank. This puts
most of the onus for performance monitoring on the contractor in the form of a
“self control unit”. The self control unit is responsible for day-to-day monitoring
and reporting with only random checks by the client (or his consultant).

DHV - SILT - TAEC – CMS Chapter IV-28

April 2007
 Sector Wide Road Programme & PIP Study
Part 1 – Final Report

This approach to monitoring requires a high degree of trust in the honesty of the
contractor, something that is not normal in South Asia. Without full parallel
monitoring by the client it is impossible to determine how honest the self control
unit is in reporting, say, unrepaired defects which lead to financial penalties for
the contractor. In this regard, several of the items are subject to penalties at a
daily rate and in one case the penalty is per hour. To assess these requires a
high frequency of road inspections (daily or even hourly!) while joint formal
inspections are only monthly. It is felt that some of the requirements are
impractical in terms of the necessary monitoring.

(e) Conclusion

It is considered that the PBMC specification could be considerably simplified and

reduced in scope to make them more practical in terms of fairness to the
contractor and realistic in terms of the monitoring required to assess compliance.

DHV - SILT - TAEC – CMS Chapter IV-29

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 5
V. POTENTIAL DEVELOPMENT & EXPANSION OF SRN

A. Identification & Selection of Potential Additions

1. Definition of Extended SRN
The Consultants are required to assess the future needs of the Strategic Road
Network (SRN) over the next 10 years and to make recommendations regarding
its expansion. It is suggested, in the Terms of Reference, that the future (2016)
SRN will likely include around 7,500km of road, an increase over the existing
length of around 5,000km. This was considered a realistic expansion of the
network over a 10-year period - given the present network size and the capacity
and capability of the DoR to manage and maintain the network.

This does not necessarily involve the construction of 2,500km of "new" road, as
many of the highest priority roads for inclusion in an expanded network are either
already under construction or are existing roads that are not currently designated
as part of the SRN. The additional roads thus comprise: some existing roads
currently maintained by DoR; some new (strategic) roads presently under
construction or planned; some existing local or District Roads that serve a
strategic function and could be absorbed into the SRN; and some totally 'new'
roads requiring construction on a new alignment.

The Consultants have however taken an open and unconstrained approach to the
expansion and extent of the future network. Starting from the existing ‘designated
and operational’ network, all potentially strategic additions – from all sources –
have been examined. As defined earlier (in Chapter 3) the SRN comprises the
main national and inter-regional arteries, connections to the borders, and links to
all Districts Headquarters and other centres of economic activity.

The SRN contributes to national integration and enables the required levels of
accessibility to be achieved throughout the country. This Study has examined all
links of a potentially strategic nature – ie any link which has more than a local
function. These are the roads for which the DoR has prime responsibility.

2. Current SRN
A preliminary step in the development of the future SRN was the identification of
the extent of the existing network, as described in Chapter 3. The existing
network comprises 5,030km of ‘designated and operational’ SRN – 3,108km of
National Highways and 1,922km of Feeder Roads – plus an additional 398km of
‘other’ roads considered to be part of the SRN (including some roads within
Kathmandu) and which are maintained by DoR. This defines the currently
operational SRN with a length of 5,428km, as illustrated in Figure 5.1, which is
shown against a background of the current population density distribution.

It is noted that in addition to this total, there are some further sections of the
designated SRN which are non-operational (either under construction or planned)

DHV - SILT - TAEC – CMS Chapter V-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

plus other committed roads that will form part of the future SRN. These are
considered below.

3. Identification of Long-List
The objective of this Study is to define a future SRN to be developed over the
next 10 years – the periods of the 11th & 12th Five-Year Plans. The first step in
this process of identifying potential new links was to establish the “strategicness”
of any proposed link – ie to determine whether it satisfies the criteria to become
part of the SRN. Basically a strategic road should serve traffic of national,
regional or inter-district nature: therefore if a road serves only local movements
then it is NOT strategic. This process of identifying the ‘strategic nature’ of any
road link is fundamentally different from establishing whether or not any
construction or improvement works are appropriate or required.

The following types of road are considered to meet the requirements that enable
them to be classified as strategic:

• Links to District Headquarters;

• Links to major Border Crossings;
• Links between existing strategic roads;
• Inter-District links, especially between DHQs;
• Links or access to rural airfields; or
• Links to major industrial, commercial or tourist sites.

Links that do not meet at least one of these criteria were – in general – excluded:
in most cases, they should be considered and classified as District Roads, with
no strategic function, although in many cases they may access sizeable pockets
of population and/or agriculturally productive areas.

The approach was to use multiple sources to identify potential additions to the
SRN, including all relevant previous studies and the recently produced DoR 20-
year Master Plan. The other sources consulted to identify potential network
additions included:

• the previous Priority Investment Plan Study (1997);

• the ADB Connectivity Study (2005), which contained a screening exercise
for potential new roads;
• the IDA North-South Transport Corridor Study (2004);
• the recent ADB Transit Routes Study;
• the ADB Sub-Regional Trade Facilitation Project;
• the initial planning phases of the RMDP and RNDP (FRIP); plus
• local road proposals included in the individual District Transport Master
Plans (DTMPs), the DoLIDAR local road inventory, and 20-year Plan for
rural roads.

In addition to these roads identified in previous studies – and for which indicative
alignments were available – the Consultants have developed a number of other
potential new links which could (conceivably) form part of a longer-term strategic
network.

This review resulted in the identification 170 potential schemes with a total length
of 8,780km, over and above the existing 5,430km, producing a potential SRN of
over 14,000km. These 170 schemes comprise the “long-list” of potential additions
which were subsequently evaluated and prioritised.

DHV - SILT - TAEC – CMS Chapter V-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter V-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

B. Classification and Categorisation of Roads

All the road sections forming the additional 8,780km considered for incorporation
in the SRN were included in the long-list, as one of 170 “schemes”. Each
scheme consists of either an individual link, or of a series of contiguous links, that
provides a connection between logical terminal points, such as towns, other links
or border crossings. The 170 schemes were classified into three groups, as
follows:
• Committed additions to the existing SRN – Feeder Road alignments either
planned or under construction (eg Hile-Bhojpur, Basantapur-Khandbari), or
committed potential strategic links (eg Galchi-Devighat, Manthali-Khurkot).
There are 18 schemes in this group, with a total length approaching 800km or
9 percent of the length of roads in the long-list.
• Established local or district roads that may be re-classified as part of SRN
and for which detailed alignments are available: eg the Postal Road and other
links in Terai, established inter-district District Roads, or local (potentially
strategic) roads under construction through ADB, RAP, RAIDP, etc. These
comprise 89 schemes with a length of approximately 3,170km or 36 percent
of the length of the long-list.
• Potential additional NEW links for which no detailed alignment is available:
these roads are shown as indicative linkages between two points (eg new
links to Chinese Border, Kathmandu-Terai Fast-Track, approaches to
Kathmandu from west, ‘missing’ sections of mid-hills east-west highway, links
to remaining non-connected District HQs, other inter-district links, etc). These
comprise 63 schemes with an approximate length of 4,820km or 55 percent
the total length of the long-list.
These links are shown on Figure 5.2 which illustrates the total extent of the
potential extensions to the SRN considered in the Study. A full listing of all roads
considered is included in Annex 5.1 Table A5.1a which includes also the result of
the ranking: it should be noted that the lengths quoted are initial estimates, and
may have been revised in later stages of the Study. Table 5.1 summarises the
lengths of each group of road. Most of the schemes (117) are existing roads or
committed projects, although the 63 new alignments are generally longer and
represent over half of the potential additional network length. The schemes have
been grouped into five categories – Terai Roads, links to unconnected District
HQs, links to the Chinese Border, inter-District links, and links to and within the
Kathmandu Valley. These are described in greater detail in the following sections.

Table 5.1: Classification of Potential Additional Links in SRN

No of
Length (km) Percentage
Schemes
Committed Extensions 18 791 9%
Existing Local/District Roads 89 3,170 36%
Potential New Links/Alignments 63 4,818 55%
Total 170 8,779

DHV - SILT - TAEC – CMS Chapter V-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter V-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

1. Terai Roads
The Terai roads considered are those south of the existing East-West Highway,
including those linking to Indian border crossings, sections of the east-west
Postal Road 8, and other connections to towns and villages. Most of the roads are
established earth or gravel tracks which follow existing alignments and are used
by local traffic, including cycles, rickshaws and bullock carts. No entirely new
alignments considered, although significant improvements, including construction
of bridges may be required. Many of the alignments considered form part of the
proposed Indian Government Assistance Programme for “prioritised road and
bridge infrastructure improvements in the Terai”.

The two proposed new approach roads to the International Customs Depots at
Birgunj and Bhairahawa (via Parasi), as included in the ADB-funded Sub-
Regional Trade Facilitation Project, are considered as candidates for inclusion
into the SRN.

2. Links to Unconnected District HQs

It has been a long-standing objective of successive Governments to provide all-

weather road access to all District HQs throughout the country. At the time of the
previous PIP (1996) there were 17 non-road-connected HQs: since that time work
has continued and limited access is now available to five of these 9. Construction
is on-going to extend access to nine of the 12 remaining non-connected
districts 10, leaving three districts (Dolpa, Mugu and Humla) without any currently
committed access proposals. Details of the current status of access to these 17
District Headquarters are given in Table 5.2.

It would appear from this analysis that satisfactory progress is being made to
connect most District HQs, with the possible exception of Diktel (Khotang) in the
east. The remaining HQs for which committed schemes are not in place (eg
Dolpa/Dunai and Mugu/Gamgadhi) have substantially smaller populations and
thus have a lower priority: these aspects are covered further in later sections of
this report.

The links to all the remaining non-connected HQs should be included in the long-
term strategic network, together with any existing links to District HQs that are not
classified as part of the current SRN.

8
A historic track within Nepal linking towns close to the Indian Border: much of the previous alignment
still exists although upgrading would require the construction of a number of bridges.
9
Limited access is available to Okhaldhunga since 2000 (although a bridge is required across the
Sunkoshi), Rukum/Musikot since 1998, Darchula since 2006, and Kalikot & Jumla via the Karnali
Highway in early 2007: access to the latter four District HQs is limited to tractors and the roads may
not be open for regular public traffic.
10
The remaining 12 Districts with unconnected HQs are: Sankhuwasabha, Solokhumbu, Bhojpur &
Khotang in the East; Manang & Mustang in the West; Jajarkot, Dolpa, Mugu & Humla in the Mid West;
and Bajhang & Bajura in the Far West

DHV - SILT - TAEC – CMS Chapter V-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 5.2: Current Status of Connections to District Headquarters (2007)

District/Headquarter Status
Eastern Development Region
1 Sankhuwasabwa/Khandbari Construction in progress: included in ADB/RNDP & DFID/RAP
2 Solokhumbu/Salleri Initial construction under Upper Sagarmatha Agricultural Project:
(from Okhaldhunga) local rural access initiatives to be pursued or possible ADB
funding
3 Bhojpur Feeder Road under construction through DFID/RAP
4 Okhaldhunga Initial construction by Army: feasibility study completed under
(open 2000) SWRP: selected for ‘fast-track’ implementation under RSDP

5 Khotang/Diktel Local road access from Hilepani started under DFID/RAP:

alternative access under GoN funding from Gaighat

Western Development Region

6 Manang/Chame Local access under construction with GoN funding north from
Besisahar
7 Mustang/Jomosom Access being developed by Army from Beni with GoN funding;
alternative access available from Tibet to Upper Mustang

Mid-Western Development Region

8 Dolpa/Dunai Local access track from Jarjarkot being developed by District &
local road agencies: low priority for upgrading
9 Jumla/Jumla * Access via Karnali Highway under construction by Army: initial
track open early 2007: possible upgrading under RSDP
10 Kalikot/Manma * Access via Karnali Highway under construction by Army: initial
track open early 2007: possible upgrading under RSDP
11 Mugu/Gamgadhi Local road construction proposed through DRILP & Districts:
feasibility study from Karnali Highway conducted by SWRP
12 Humla/Simikot Access available from Tibet via Hilsa border crossing: local rural
access initiatives being pursued
13 Jajarkot/Jarjarkot * Initial track under construction by Army with GoN funding:
possible upgrading through RSDP
14 Rukum/Musikot * Access track open since 1998: feasibility study for upgrading
(open 1998) under SWRP

Far-Western Development Region

15 Bajura/Martadi * Access from Sanfe commenced under RMDP: feasibility study
under SWRP: completion possible under RSDP
16 Bajhang/Chainpur Track under construction by GoN funding: feasibility study under
SWRP: selected for ‘fast-track’ implementation under RSDP
17 Darchula/Darchula * Track constructed through GoN funding: feasibility study under
(open 2006) SWRP: part selected for ‘fast-track’ implementation under RSDP
Source: Updated from 1997 PIP and WB North-South Corridors Study (2004)
Notes: * District connections previously proposed under WB/RMDP
SWRP = Sector Wide Road Programme (this study)
RSDP = Proposed WB/IDA Road Sector Development Project

3. Roads to Chinese Border

Eleven potential cross-border links have been identified including the nine
locations examined in the earlier North-South Corridor Study as discussed during

DHV - SILT - TAEC – CMS Chapter V-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

the 2003 DoR mission to PRC/Tibet: The sites included four sites in the east 11,
the existing crossing at Kodari, the proposed crossing at Rasuwagadi, plus Lo
Mantang (Mustang), Hilsa (Simikot) and Tinka La. Other potential crossings at
Nanpa La (Solukhumbu), Larke (Gorkha) and Namja La (Mugu) were not
considered feasible. The locations of all these potential border crossings are
indicated in Figure 5.2.

Only Kodari and Rasuwagadi are considered as possible through vehicle transit
routes within the 10-year plan period. The four eastern connections are not
acceptable to the Chinese (within the Qomolunga National Park) and require
lengthy (and difficult connections on the Nepali side). Lo Mantang (Mustang) and
Hilsa-Simikot could be considered as possible access routes from China to
otherwise non-road connected northern areas: it is possible that the Mustang link
could (eventually) be connected into Nepal through the Jomsom-Beni link. A
crossing at Tinka La should be reviewed as a potential tourist and pilgrimage
route but be studied in conjunction with existing and potential routes within India.

4. Inter-District Links

Expansion of the SRN potentially includes a grid of east-west and north-south

roads, involving the inclusion (and re-classification) of key local (district) roads
and the construction of links on new alignments. In the Eastern, Central and
Western Regions, a number of such inter-district roads exist or are under
construction through various donor-funded initiatives. In the Mid and Far West,
there are fewer such existing roads and a more ambitious network of new
alignments would be required.

The requirements for the links to be included are that they should connect
between adjacent districts (or more specifically district HQs), should provide
improved access to significant local populations – especially those outside the
2hr/4hr walk criteria – and should improve local connectivity (ie allow inter-District
travel between hill areas avoiding unnecessary circuitous travel via the Terai).
Roads wholly within any one district have not in general been considered for
inclusion in the SRN, unless they provide access to a significantly important
destination.

Links thus identified can become also elements in the future mid-hills east-west
highway or ‘corridor’, which – in reality – will be a combination of a series of local
links, and not a through route designed for long-distance travel. Sections of an
embryonic Mid-Hills East-West Corridor (MH-EWC) already exist in the form of
the Prithvi Highway between Kathmandu and Pokhara, with extensions west to
Baglung, and east to Dhulikhel and Nepalthok-Kurkot. Other potential sections of
the route exist in the Eastern, Mid-Western and Far-Western Regions.

Other potential east-west connections are plausible in the lower hills and/or Inner
Terai, through relatively densely populated areas to the north of the existing
EWH. It is also possible – although unlikely in the foreseeable future – that other
east-west routes could be developed further north linking between the more
northerly District HQs, although traffic demand would be low and the terrain
challenging.

5. Links to and within the Kathmandu Valley

Numerous previous studies (including the 1997 PIP) have identified the problems
associated with the restricted number of access routes between Kathmandu and
11
Olangchunggola (Taplejung), Kimathanka (Sankhuwasabha), Nangpala (Solukhumbu) Lamabagar
(Dolakha)

DHV - SILT - TAEC – CMS Chapter V-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

the Terai, the remainder of the country, and the major border crossings with India.
At present there is only one effective link between Kathmandu and the “outside
world” 12, via the Thankot-Naubise section of the Tribhuvan Rajpath. This road is
poorly aligned – being both steep and tortuous – and has limited capacity.
Current traffic volumes are approaching 4,500 vehicles per day, comprising
around 50 percent trucks and 30 percent buses. Virtually all goods and
commodities consumed in Kathmandu – and almost all bus passengers – enter
via this road.

There are four groups of alternative or additional access routes to Kathmandu

Valley:

• Completion of the Japanese Road from Dhulikhel to Sindhuli and on to

Bardibas on the EWH;
• Construction of the ‘Fast Track’ from Kathmandu to Hetauda and/or EWH, via
either a tunnel option or along the Bagmati Corridor;
• Construction of an alternative approach from the west linking the Prithvi
Highway west of Naubise with the Kathmandu Ring Road; or
• The upgrading and improvement of existing (or partially built) local roads to
provide a link of trunk road standards.

This Study has considered a number of options, which are to a degree mutually
exclusive, and some preliminary observations on the relative merits of each are
presented in a later Chapter. The scale of investment considered is however
significantly greater than that envisaged for other components of the proposed
expansion of the SRN and thus it is recommended that further detailed studies be
undertaken of the more promising options.

Within the Kathmandu Valley, the DoR is presently responsible for maintaining a
network of 10 radial roads (sections of two National Highways and eight Feeder
Roads), plus a number of other urban roads of a strategic nature – including the
Kathmandu Ring Road. These are included in the existing operation network of
5,430km.

The Study has however considered the inclusion of a number of additional orbital
and radial roads in the future SRN (including the proposed Outer Ring Road),
although it is not possible to evaluate these on a comparable basis with roads
elsewhere in the country. Development of the road network to serve the
increasingly urbanised Kathmandu Valley should be the subject of a detailed
land-use and urban planning exercise.

C. Screening Criteria and Methodology

The process of defining the potential future SRN links was based primarily on a
screening process to select candidates from the long-list. This screening was
undertaken using a compound ranking, or multi-criteria analysis (MCA), process.
In this each link – or scheme comprising of a series of connected links – was
allocated scores for indicators reflecting its relative importance in the road
network. The total of scores for all indicators was used to indicate the ranking.
Details of the indicators used and the scoring system are given in Annex 5.1, with
a summary given below.

A single screening process was applied to all 170 schemes, totalling almost
9,000km, despite the differing characteristics and size of the schemes which can,

12
The only alternative external connection is via the Arniko Highway to the Chinese border at Kodari:
light 4-WD vehicles can also travel on local earth roads via Trisuli to Galchi or via Pharping and
Kulikani to Bhimphedi, but neither route is suitable for large vehicles

DHV - SILT - TAEC – CMS Chapter V-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

in other circumstances, create difficulties in making direct comparisons. However

in this case, the main concern is regarding the ‘strategic’ function of each scheme
and no direct comparison of economic viability is attempted at this stage.

The basic criteria for inclusion as part of the strategic network – as either a
National Highway or Feeder Road – have already been established in the Nepal
Road Classification System. These criteria form the basis of the screening
system, although it was necessary to add further indicators and a scoring system
related to such factors as traffic level, population density and economic activity, in
order to establish relative priorities.

The indicators used were as follows:

• Strategic connectivity – network effects, MH-EWC, etc;

• Population density served;
• Improvements to accessibility – impact on reducing in-accessible population;
• Connections to District HQ – either initial or secondary link;
• Border crossings – including potential as a transit route;
• Potential tourist or pilgrimage function;
• Serving area of economic activity, mineral extraction, rural airfield, etc
• Estimated traffic volume;
• Terrain difficulty and altitude – length of grades, max altitude;
• Poverty Index – the human poverty index for the districts traversed.
The final indicator, relating to the relative poverty of the area through which the
road passed, was added to the original indicators at the request of the DoR. After
the initial ranking it was decided to use an equal weighting for all the indicators,
as this appeared to produce the most reliable results, without any undue bias.

D. Results of the Screening and Prioritisation

From the list of 170 schemes, the highest ranked 100 schemes – with a total
length of around 5,000km – were retained as ‘candidate roads’ for subsequent
assessment for inclusion in the extended SRN. These roads form the basis for
the proposed extended SRN as illustrated in Figure 5.3: the results of the full
ranking of the 170 schemes are included in Annex 5.1 Table A5.1b, which
presents the total unadjusted ‘scores’ of each scheme.

The first step was to identify the 'committed' schemes – most of which were
ranked 'high' on the priority listing. Additional roads were then drawn from the
priority ranking, including BOTH existing District Roads and potential new
alignments. The final ranking was not based solely on the ranking obtained from
the MCA scores, as adjustments were made to produce a more logical priority list
in which, for example, competing or duplicating schemes were removed, and
some lower ranked schemes were elevated where these provided connections to
high priority schemes.

DHV - SILT - TAEC – CMS Chapter V-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter V-11

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The resultant lists (Tables 5.3 to 5.6) also include some sections of the currently
designated SRN which are not operational and which do not appear in the MCA
analysis.

The priority list of around 100 schemes was drawn up in three groups: committed
roads and other high priority schemes divided into Terai Roads and Hill Roads
(Tables 5.3, 5.4 and 5.5). The MCA approach was used initially to identify around
3,000km of additional road to be included in the extended SRN in 2016, including
800km of committed schemes, a further 800km of roads in the Terai, and
1,400km of hill roads. This total is significantly in excess of the requirement (in
the ToR) to identify an additional 2,200km of road.

In addition to these roads, further schemes have been identified for inclusion in
the extended SRN: these additional roads include those currently maintained by
the DoR, existing roads accessing major tourist or pilgrimage sites, and schemes
proposed for upgrading under foreign-aid projects.

1. Committed Schemes
There are 20 ‘committed’ schemes, with a total length of 797km 13 , including
existing roads, roads under construction or upgrading, and those for which
funding is in the ‘pipeline’. The roads are listed in Table 5.3 below in priority order
as determined by the MCA ranking. Specifically the list includes the proposed
ADB core projects from the Connectivity Study, the access roads to the two ICDs,
and links to ten District Headquarters. These schemes are shown in RED on
Figure 5.3.

Table V.3: The 20 Priority Additions to the SRN (Committed & Existing)
Status Scheme km
1 ADB (STFP) ICD (Pokhariya) - Parwanipur 10
2 ADB (USARP) Okhaldhunga - Salleri 29
3 ADB (TCP) Dhunche - Rasuwagadhi 26
4 ADB (STFP) Bhumahi - Parasi - Siddharthanagar (ICD) 29
5 ADB (TCP) Galchhi - Devighat 20
6 ADB (TCP) Khurkot - Manthali 11
7 ADB (RNDP) Basantapur-Tumlingtar-Khandbari (north of Mude)** 81
8 Exists Maldhunga - Beni 13
9 u/c (India) Shabha (Mahendranagar) - Brahmhadev 13
10 u/c RMDP Sanfebagar – Martadi (remaining section) ** 43
11 Exists (WB) Sunkoshi - Okhaldhunga 41
12 u/c (RAP) Leguwaghat - Bhojpur 66
13 u/c (RAP) Hilepani - Diktel 67
14 u/c (Army) Beni - Jomsom 80
15 Exists (GTZ) Upper Dhungeshwor - Lower Dhungeshwor** 8
16 Exists (hydro) Malunga - Kaligandaki 20
17 Exists DR Inaruwa - Duhabi 11
18 u/c in part Besisahar - Chame 65
19 Karnali Highway* Tunibagar (Rakam) - Kalikot - Jumla 132
20 Dhulikhel-Sindhuli* Nepalthok-Khurkot 32
Total 797
* Not included in prioritisation process as already included in Designated SRN
** Remaining Sections

13
Lengths (and some schemes) differ from Table A5.1A as some roads are partially
constructed and open to traffic.

DHV - SILT - TAEC – CMS Chapter V-12

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

These links have been identified in the screening and ranking process as high
priority schemes and include a number of current and proposed aid-funded
projects. Specifically, the 20 high priority schemes include:

• Access to ten currently non-road-connected District Headquarters

(totalling 604km): Khandbari, Bhojpur, Diktel, Okhaldhunga, Salleri,
Chame, Jomosom, Manma, Jumla and Martadi;
• Three Missing Links (51km): Nepalthok-Khurkot, Khurkot-Manthali and
Lower-Upper Dungeshwor (access to Dailekh);
• Access to four Border Crossings (98km): improved access to Birgunj &
Bhairawa ICDs (ADB/STFP); Dhunche-Rasuwagadhi (Chinese Border)
and Galchhi-Devighat; and Mahendranagar-Brahmadev; and
• Three Existing Roads (44km): Maldhunga-Beni, Malunga-Kali Gandaki
and Inaruwa-Duhabi
2. Other High Priority Schemes (Terai)
Twenty-five sections of road, with a length of 815km, have been identified and
prioritised within the Terai, comprising both elements of the east-west Postal
Road and north-south links between the East-West Highway and the Indian
Border. These roads are listed in priority order (based on the MCA) in Table 5.4
below: all are included in the programme of infrastructure improvements
proposed for funding by the Government of India (GoI).

It has been determined that all the Terai roads proposed for upgrading or
reconstruction under the GoI programme would, in future, become part of the
extended SRN. In total, an additional 1,500 km of Terai roads are added to the
existing SRN on this basis, including over 500km of the Postal Road. It is noted
that, overall, the GoI improvement programme includes 1,450 km of road, some
of which are part of the already designated SRN. Two sections of road in the
eastern Terai, which are currently being upgraded under the ADB-funded RNDP,
are also included in the extended SRN.

These additional links in the 2016 SRN are shown in Figure 5.3 in PINK. It can be
seen that all potential sections of the Postal Road are now proposed for inclusion
in the extended SRN, together with an additional 13 north-south connections
between the East-West Highway, main centres and the Indian border.

Table V.4: The 25 Priority Additions to the SRN in the Terai

Type Scheme km
1 Local Rd Taulihawa - Bahadurgunj 22
2 Local Rd Janakpur - Samsi 20
3 Postal Rd Jaleswor - Malangawa 40
4 Postal Rd Malangawa - Gaur 51
5 Postal Rd Gaur - Kalaiya 66
6 Postal Rd Rajbiraj - Siraha 69
7 Postal Rd Siraha -Janakpur 38
8 Local Rd Rampur - Lumbini 20
9 Local Rd Pipra - Chakarchauda 27
10 Postal Rd Gulariya - Murtiya 3
11 Local Rd Manmat - Matiarwa 28
12 Local Rd Sati - Khakraula 5
13 Local Rd Parasi - Maheshpur 12
14 Postal Rd Bhadrapur - Gaurigunj 47

DHV - SILT - TAEC – CMS Chapter V-13

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Type Scheme km
15 Postal Rd Gaurigunj - Rangeli 25
16 Postal Rd Birgunj - Jagatpur - Bharatpur 98
17 New Lumbini Circumambulatory 12
18 Local Rd Lumbini - Kakarhawa 8
19 Postal Rd Nepalgunj (Birendrachowk) - Baghauda 49
20 Local Rd Lamahi - Koilabas 34
21 Local Rd Phuljor - Tribhuvannagar 24
22 Local Rd Naya Road - Madhuwan 42
23 Local Rd Tamagadhi - Simraungadh 40
24 Local Rd Lahan - Thadi 18
25 Postal Rd Dhangadhi - Bariyarpatti 17
Total 815

3. Other High Priority Schemes (Hills)

A further 25 potential additions to the SRN in the hills, with a total length of
around 1400km (see Table 5.5), have been identified based on the MCA analysis
and in discussion with the DoR. These include a mix of new construction and
upgrading (and re-classification) of District Roads. The ‘new’ links include access
links to the three remaining non-connected District HQs – Gamgadhi, Simikot and
Dunai - and options for improved access to Kathmandu Valley from the south and
west. The two alternative options for each of the approaches to the Kathmandu
Valley are shown shaded in Table 5.5. The suggested alignments for the
additional hill roads are coloured PURPLE on Figure 5.3.

PLEASE NOTE: This ranking is based on strategic importance only and NOT on
economic viability or prioritisation. Additionally some of the schemes as shown
are mutually exclusive and require a comparative evaluation of the options.
These aspects are considers in the next chapter.

This list indicates a high priority for the provision of an additional access route to
Kathmandu Valley – to be selected from among the alternatives for access from
the West or South (taking account also of the timing and lead-time for
construction) – plus the construction of the Kathmandu Outer Ring Road, links to
the three un-connected District HQs, and a number inter-District links in the mid-
hills of eastern and western Nepal, which taken together could create a mid-hills
east-west corridor. This mid-hills corridor (MHC) should not be conceived as a
major through highway – that role is served by the existing East-West Highway –
but rather a series of local links to improve local accessibility and connectivity.

Table V.5: The Next 25 High Priority Additions (Hills)

Type Scheme km
1 New Hetauda Bypass 15
2 New Chourjahari - Musikot 53
3 New Outer Ring Road (ORR) & Mulpani-Thimi 70
4 New - DHQ Simikot - Hilsa 88
5 New - DHQ Access to Gamagadhi 100
6 New Dharke - Bhimdhunga - Sitapaila (RR) 33
OR New RR Gongabu – Kolphu (60km)
7 Local Rd Tulsipur – Bhotechour (MHC) 86
8 Local Rd Baglung – Burtibang (MHC) 105
9 Local Rd Mangalsen - Belkhet (Rakam) (MHC) 40
10 New Kaligandaki - Kusma 67

DHV - SILT - TAEC – CMS Chapter V-14

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Type Scheme km
11 New Kathmandu-Terai Fast Track 65
OR New Bagmati Corridor (90km)
12 New Brahmhadev - Jogbudha 30
13 Local Rd Jhumka - Chatara - Barahachhetra 26
14 New - DHQ Jajarkot - Dunai 144
15 Local Rd Bhojpur - Diktel (MHC) 55
16 New Hilepani -Sunkoshi – Khurkot (MHC) 43
17 Local Rd Fattepur - Kanchanpur 27
18 Local Rd Phidim - Terhathum (Myaglung) (MHC) 75
19 New Leguwaghat - Sabha 32
20 Local Rd Chatara - Dharan 13
21 New Burtibang - Musikot 145
22 Local Rd Tamghas – Sandhikharka (MHC) 30
23 Local Rd Sandhikharka – Pyuthan (MHC) 49
24 Local Rd Tandi, Ratnanagar - Saurah 7
25 Local Rd Dumre - Bandipur 8
Total 1406

The roads may be grouped into three broad categories: strategic network
improvements (Kathmandu Valley approaches, Hetauda By-pass, Kathmandu
Outer Ring Road); remote area access (the three District HQs); and mid-hills
linkages, including the potential mid-hills corridor.

Improved access to the Kathmandu Valley is clearly a major priority in the coming
decade. The sole existing route between Thankot and Naubise (and the section
within Kathmandu Valley) is operating close to effective capacity, with frequent
delays and congestion experienced: furthermore the reliance of the capital on this
single approach road has severe strategic and logistical implications, as has been
frequently demonstrated in recent years.

All of the suggested additional access routes to Kathmandu Valley – from both
the west and south – are ranked highly in the MCA, suggesting that an evaluation
of a new access route should have a high priority. There are many complex and
inter-related issues to be addressed and it is suggested that initially only one (of
the four broad options considered) should be pursued. The issues associated
with the alternatives are presented in greater detail in the following Chapter.

Examination of the population density distributions (see following sections)

reveals that significantly more people will be served though an increase in road
network density in the mid-hills rather than by extensions of the network into the
sparsely populated northern areas. Consequently, improvements to accessibility
may be more readily achieved by the development of roads in the mid-hills to
provide enhanced mobility and better connectivity to a greater number of people.

The MHC could start in the east at the Pashupatinagar border crossing with
Darjeeling 14 and traverse the whole of the country via Phidim, Terhathum,
Bhojpur, Diktel, Kurkot, Dhulikhel, Kathmandu, Pokhara, Baglung, Gulmi,
Sandikharka, Pyuthan, Dang, Surkhet, Sanfe, Silghadi, Dadeldhura and Baitadi to
the Indian Border at Jhulalghat in the Far West. The total length would be around
1,700km and the route would connect (or pass close to) 24 District Headquarters.
Around 960km of the route already exists as links in the SRN, a further 430km
follows local road alignments (either existing or under construction) and the

14
or alternatively via a new border link to the north of Phidim

DHV - SILT - TAEC – CMS Chapter V-15

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

remaining 290km would be new construction. [An alternative more northerly

alignment is available between Baglung and Surkhet via Musikot and Jarjarkot
but this would serve significantly fewer people.]

In addition to these roads that were identified and evaluated through the MCA, a
number of additional hill roads have been proposed for inclusion in the extended
SRN (see Table 5.6). These additions, totalling almost 600km, include a number
of roads that are presently maintained by DoR, some important inter-District local
roads which are open year-round and have regular bus services, and access
roads to major tourist or pilgrimage sites.

Table V.6: Additional Hill Roads proposed for inclusion in SRN

Type Road Section Length (km)
1 Local Rd Kanti Rajpath (Hetauda - Tikabhairav) 93
2 New Leguwaghat - Sabha 32
3 Exists Ridi - Wamitaksar 51
4 Local Rd Tamghas - Wamitaksar 19
5 Local Rd Gwarko - Namobuddha - Kavre Bh 42
6 Local Rd Pharping - Kulekhani 27
7 Local Rd Dhadingbesi - Gorkha 67
8 Local Rd Jogbudha - Budar 34
9 Exists Dumre - Bandipur 8
10 Local Rd Bhedetar (KRM) - Budhabare 50
11 Local Rd Budhabare - Ranke 66
12 Exists Panchkhal - Palanchowk 11
13 NEA Bhimphedi - Kulekhani 12
14 Exists Sahajpur - Barchhain (BP Nagar) 38
15 Local Rd Daiji - Jogbudha 35
Total 585

E. Conclusions regarding Expansion of the SRN

The Study has completed a thorough and comprehensive review of all potential
additional links to be considered for inclusion in as extended SRN over the next
ten years. The analysis commenced with a review of all previous studies and
recommendations, and identified almost 9,000km of potential new strategic road,
including a mix of new alignments and upgrading of existing local roads. The start
point was the existing ‘designated SRN’ comprising around 5,050km of open and
operational road, plus around 400km of ‘other’ roads considered to be part of the
existing SRN (including a network of roads within Kathmandu Valley).

The extent of the network is however in a degree of flux at the present time, as
there are a number of on-going construction and improvement programmes
(mostly foreign aid funded), in various stages of completion, which influence the
exact length of the network to be considered. Additionally, there have been many
changes and extensions to the network over the past 12 years since the last
comprehensive classification was undertaken.

The expansion of the SRN will comprise three main components: committed
schemes and existing strategic roads; the conversion of local roads to SRN; and
the construction of new alignments. The first group comprises non-SRN roads
that the DoR currently maintains, as well as committed projects including
specifically new links to District Headquarters.

The second group (local roads) includes a substantial length of existing earth or
gravel roads (in excess of 1,200km) in the Terai which are the subject of an

DHV - SILT - TAEC – CMS Chapter V-16

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Indian Government Improvement Programme: it is assumed that, on completion,

all these roads will be transferred to DoR responsibility. It is also suggested that a
number of key inter-District local roads in the hills be re-classified as part of the
SRN and transferred to the DoR.

The third group includes new roads to enhance accessibility in remote areas and
within the more densely populated mid-hills, as well as some key new strategic
links to strengthen the network and improve access to Kathmandu. These latter
schemes are potentially the most economically attractive prospects, whereas the
new roads in the hills will improve accessibility, encourage development, provide
social benefits, and reduce regional disparities.

The Study recommends the expansion of the SRN to around 9,900km over the
period to 2016. The following Chapter examines and evaluates a number of the
individual schemes, to assist in developing a prioritised list of improvements and
in determining budget requirements.

The final recommendations for the expansion of the SRN are presented in
Chapter 10, together with a detailed listing of prioritised projects. Subsequent
Chapters (7 and 8) examine the impact on accessibility of the extensions and/or
improvements to the strategic and local road networks. This analysis provides
general support for a significant expansion of the SRN (mainly though the
redesignation of key local/district roads) as the most effective means of
enhancing accessibility and providing a basis for economic growth and
development. Details of the future, expanded SRN comprising almost 10,000km
are given in section 7.4.2.

DHV - SILT - TAEC – CMS Chapter V-17

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 6
VI. PRIORITISATION OF UPGRADING & NEW CONSTRUCTION

A. Introduction

The previous chapter described the screening process used to identify additional
links to be included in the extended SRN based on their 'strategic' function. A
review has been undertaken of the proposed SRN to determine which sections
are likely to be economically viable for upgrading or new construction during the
next ten years. This prioritisation, leading to the development of the PIP, is
described in this chapter. Roads considered a priority for inclusion in the SRN are
not necessarily priorities for construction or upgrading, but there is some
correlation between being strategic and there being significant benefits from good
road condition. Strategic roads are often those that have, or potentially have,
higher traffic than other roads in the network, which gives greater potential for
significant benefits.

Some roads can be directly included in the SRN with no works required, other
than an ongoing maintenance liability, whilst others require construction or
upgrading on the basis of condition and/or traffic volume. In this chapter a review
has been made of the priority for construction or improvement to proposed
additions to the SRN based on the economic returns on the investment required
for such works. The potential projects are diverse in nature. The types of
projects that could be involved on the roads proposed for inclusion in the SRN,
and the sources of the major benefits to be derived, are:

• Capacity expansion (widening of roads) - road user benefits from higher

vehicle speeds.
• New network roads - road user benefits from reduced travel distance and
higher speeds.
• New hill roads - road user benefits from replacement of trail movements
and air transport by truck and bus traffic, plus development benefits.
• Upgrading earth and gravel roads - road user benefits from reduced
surface roughness and higher vehicle speeds.
• Rehabilitating paved roads - road user benefits from reduced surface
roughness and higher vehicle speeds.
It is not necessary to consider all roads for such works. Many of the proposed
additions to the SRN, especially the highest priority sections, are committed
roads that have already been assessed and do not require further analysis,
except where the current proposals would leave the road in an unsealed state. In
these cases the viability of upgrading to sealed standard has been examined
where possible. Some of the proposed SRN roads are currently under
construction. Again, where these works are to unsealed standard, they have
been evaluated for upgrading to sealed standard. Some roads have already been
studied in detail elsewhere and found to be viable: these have not generally been

DHV - SILT - TAEC – CMS Chapter VI-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

reassessed in this Study. The results of those that were studied for the feasibility
studies in Part II of this Study are reported in this Chapter.

Some roads cannot realistically be studied within this Study, as they require
specific, detailed information and careful analysis if the results are to have any
validity. Therefore only those where an evaluation would be useful and possible
at an acceptable level of accuracy have been studied.

Two types of analysis have been carried out:

• Population based analyses of new construction or upgrading of roads
providing access in hill areas.
• Traffic based analysis of new roads replacing or augmenting existing
roads.
Population based analysis determines the number of people that would be served
by a road, in that it provides their main access to external centres and the rest of
the road network, and from this estimates the amount of traffic that would use the
road. The method is based on assumed rates of freight and passenger traffic
generation per head of population, and the benefits of diversion from trail
movement to road transport. It can only be carried out accurately if the road
under study is the main access to a defined area. However it can also be used for
interconnecting roads, such as sections of the Mid-Hills East West Link if the road
sections serve significant numbers of people. Several of the links proposed for
the SRN have been analysed in this way as part of the feasibility studies carried
out in Part II of this Study.

The traffic based analysis determines the road used savings provided by shorter
and/or better aligned routes compared with existing route; that is, diverted traffic
benefits. This is a more conventional analysis based on vehicle operating and
time cost savings from a shorter route, and has been applied to some roads in
the core network. It is not generally applicable to roads such as the Mid-Hills
East West Link because the level of traffic that would divert from existing routes is
unknown.

Other types of analysis are not practical within the scope of this study.
Specifically capacity expansion schemes have not been studied (although the
widening of the Narayanghat - Mugling Road was studied in the Feasibility
Studies reported in Part II). Traffic levels in Nepal are for the most part low and
so there are few cases where upgrading in the form of capacity expansion is
required; the exceptions are in – and on the approaches to – the Kathmandu
Valley. In this type of project the benefits are typically very small per vehicle-km,
but can justify investment because of the high volume of traffic. Specific
information about traffic flow and speed-flow relationships over short distances is
required to estimate the benefits.

B. Evaluation of SRN Improvements and Expansion

The proposals for the expansion of SRN, as described in Chapter 5, were based
on an assessment of ‘strategic’ nature of potential links. Three types of expansion
of the SRN were identified:
• Existing Commitments
• Conversion of existing Local Roads
• New proposals
The analysis identified, by 2010, a network of around 7,900km of existing roads
and committed schemes that comprise the expanded ‘committed’ network which

DHV - SILT - TAEC – CMS Chapter VI-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

is taken as the base for further work. This committed network has been evaluated
using HDM to determine priorities for upgrading and major maintenance
interventions as described in Chapter 4.

The previous chapter established the scale and extent of the future SRN: this
Chapter now assesses and evaluates the needs and priorities for development
work on the network – in terms of both new construction and upgrading. Initially,
the following section examines the upgrading requirements of the existing
network.

Subsequent sections review the various options for extensions to the network,
strengthening of the network, and the potential for further upgrading or capacity
enhancement.

1. HDM Analysis of Upgrading

The HDM model and database have been used to identify priorities for upgrading
of existing (and proposed) gravel and earth roads that form part of the committed
network. For the purposes of this analysis ALL potential schemes (except those
actually under construction) have been evaluated, in order to confirm priorities
among the current proposals. Due to constraints of the HDM-4 model, it is not
possible to optimise the programme but merely to identify those schemes that are
assessed as feasible in the period to 2016.

The resultant schemes are listed in Annex 6.1 and are illustrated in Figure VI.1,
which indicates separately those ‘committed’ projects which are assumed to be
implemented by 2010 and those additional projects recommended for sealing
post-2011. The analysis confirms that most of the ‘committed’ projects (shown in
PINK) – including the proposed Terai road improvements – are priorities before
2016, with the exception of the final sections of the roads to Darchula and Jumla:
the forecast traffic levels on these two roads are insufficient to justify upgrading to
sealed standards on the basis of road-user savings alone 15.

The remaining schemes – shown in green – are those links in the 2010 network
that are identified by HDM-4 for upgrading in the period beyond 2011. This
category includes a number of radial roads within the Kathmandu Valley and the
sealing of remaining sections of hill road which are not included in any current
programme.

In total, almost 2,600km of earth and gravel road are proposed for upgrading over
the 10-year plan period, see Table VI.1, at a total estimated cost of Rs
13.5 billion. About 60 percent of the length is in the hills with the remainder
(1,000km) in the Terai.

Table VI.1: Roads proposed for Upgrading to sealed Standards

Number of Length Cost Average
Links (km) (Rs billion) Cost/km
Hills 71 1,598 9,569 5.99
Terai 50 980 3,908 3.99
Total 121 2,578 13,476 5.23

15
It is probable however that both would be justified if the full benefits from generated traffic were
included: both sections were included in the previous phase of RMDP and the amount of outstanding
works required to complete to sealed standards has been assessed in Part II of the Study.

DHV - SILT - TAEC – CMS Chapter VI-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

2. Evaluation of extensions beyond 2010

The previous Chapter recommended the further expansion of the SRN beyond
2010 by approximately 1,800km to a total length of around 9,700km by 2016, as
illustrated in Figure 5.3. This expansion includes the reclassification of some
exiting roads as part of the SRN and the construction of new links, both
extensions into non-road-served hill areas and new links accessing Kathmandu
Valley.

Key elements of the expanded SRN include:

• Completion of links to ALL District HQs;
• Upgrading of unsealed roads in Terai and main links into the hills (as
described above);
• Improvement of the road network in the mid-hills, including inter-District
links and the potential mid-hills corridor;
• Enhancement of access to Kathmandu – including the proposed “Fast
Track” to the Terai and/or Bhimdhunga Link to Prithvi Highway; and
• Capacity enhancement of key Kathmandu Valley roads – including the
existing Ring Road, Bhaktapur-Dhulikhel, etc
Wherever practical, schemes involving new construction have been evaluated,
based on the potential identifiable benefits. In the case of hill roads in previously
unserved areas, the evaluation is based on the populations served and the
benefits from generated traffic and transfers from air and foot travel to trucks and
buses: in situations where there is existing or potentially divertable traffic, the
benefits are based on road user savings. However in some cases, it is not
possible to identify specific quantifiable benefits and no attempt to evaluate these
projects has been made.

In total 23 proposed sections of new construction have been evaluated: nine of

these are classified as ‘remote area access’, a further nine as ‘mid-hills links’, and
the remaining five as ‘strategic network improvements’. The proposed schemes
are illustrated in Figure VI.2. Specifically, it should be noted that the schemes
evaluated include TWO options for the Kathmandu-Terai link and TWO options
for improved access from the west (Prithivi Highway). In each case these are
alternatives which need to be evaluated against each other and – in practice –
both pairs of options should be considered together as part of a study of
improved access to Kathmandu and they are, to an extent, inter-dependent.

It is also noted that not all the potential new roads have been evaluated, as in
some cases (e.g. Kathmandu Outer Ring Road) it is not possible to quantify the
potential benefits without exhaustive study.

The results of the evaluations are presented in Table VI.2 for the 23
schemes, with a total length of a little over 1,400km and a combined estimated
cost of Rs 47 billion. It can be seen that – with five exceptions – all the evaluated
schemes produce a positive return, in many cases in excess of 20 percent. The
schemes which fail to produce adequate returns are mostly the more remote
extensions to the network in the hills – Hilsa-Simikot, Nagma-Gamagdhi, Jajarkot-
Dunai and Chaujari-Musikot. The Gongabu-Kolphu-Galchhi alignment for the
improvement of the western approach to Kathmandu is also shown to be non-
feasible.

DHV - SILT - TAEC – CMS Chapter VI-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VI-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table VI.2: Evaluation of Links proposed for New Construction

Length Cost **
IRR (%)
(km) (Rs mill)
Remote Area Access
Sunkoshi – Okhaldhunga* 40 261 22.5
Okhaldhunga – Salleri 29 333 14.4
Beni – Jomsom 80 480 14.5
Upper – Lower Dhungeshwor* 8 85 13.9
Nagma – Gamagadhi* 88 1,475 7.6
Simikot – Hilsa 88 1,012 4.7
Jajarkot – Dunai 144 1,656 8.5
Sanfebagar – Martadi 57 600 34.2
Martadi – Kolti* 51 673 23.1
Mid-Hills Linkages
Diktel – Hilepani 67 402 21.6
Hile – Leguwaghat 26 156 21.1
Leguwaghat – Bhojpur 66 396 20.9
Kaligandaki – Kusma 67 771 17.7
Chourjahari – Musikot 53 610 2.6
Tulsipur – Bhotechour* 78 873 20.0
Baglung – Burtibang 105 630 35.2
Mangalsen – Sanfebagar* 38 289 15.1
Mangalsen – Belkhet* 49 576 17.8
Strategic Network Improvements
Kathmandu – Terai Fast Track 70 18,000 16.2
Bagmati Corridor 94 9,720 18.4
Hetauda Bypass 15 375 25.0
Dharke – Bhimdhunga – Sitapaila 33 2,600 15.0
Gongabu – Kolphu – Galchhi 60 5,000 3.8
Total 1,406 46,973
* Schemes evaluated in detail in Part II Feasibility Studies
** Incl VAT

Seven of the schemes examined (marked with an asterisk) are subject to detailed
Feasibility Study under Part II of the project: greater detail of the evaluation
procedures and results are available in the Part II Final Report and individual
reports on each road. The rates of return for each of the Strategic Network
Improvements are based on traffic projections assuming that only that scheme is
implemented. As noted above, some, such as the Kathmandu-Terai Fast Track
and Bagmati Corridor, are clearly mutually exclusive and only one would be
implemented. In other cases two schemes could be implemented but the impact
on traffic would lower rates of return.

DHV - SILT - TAEC – CMS Chapter VI-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VI-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

All of the remote area and mid-hills projects were assessed using the ‘population
served’ approach, traffic forecasts are made based on estimates of increases in
volumes of freight consumed and frequency of trips made per head, with and
without the road improvement. A summary of the basic assumptions is presented
in Table VI.3, which indicates the changes for populations in the immediate zone
of influence of the road (assumed at 4 hour walk) and the broader catchment
area. These estimates have been correlated against observed traffic flows on
roads into hill areas before and after improvement: the most notable change is
the increase in the number of bus passenger movements from the population in
the immediate area of influence.

Table VI.3: Estimates of Freight Consumed and Trips made with and without
improvement
BEFORE AFTER Improvement
Restricted Within 4 hour
Wider area
Access walk
Freight Consumption
40kg 60kg 80kg
(kg / head / year)
Passenger Trip Generation
0.4 0.6 1.5
(trips/year)

The approach is discussed in more detail in Annex 6.2. In general, acceptable

rates of return are obtained for roads of moderate length serving areas with
significant population densities – ie within the mid-hills. It is only the longer roads
into the sparsely populated northern mountain areas that fail to produce
acceptable returns.

The Strategic Network Improvements were analysed using a conventional road

appraisal approach, with potential benefits based on road user savings (vehicle
operating cost and passenger time savings) from shorter and higher speed routes.
Road user costs were determined for the main vehicle types for range of road
conditions and terrain types using the HDM-4 based Roads Economic Decision
(RED) Model. Estimates of costs per km were produced for different road
conditions. Although roughness varies over time it is generally acceptable to use
a single estimate that reflects average roughness over time for each road type,
consistent with an assumption of the application of typical maintenance
operations. An IRI of 4 was used as the standard measure of roughness. Road
gradient and curvature are also inputs to VOC calculations. Standard averages
applicable in different terrain types were used.

C. Network Strengthening
Five schemes involving network strengthening within the Central Region have
been evaluated. These include two alternatives for a link between Kathmandu
and the Terai, two alternatives for an improvement to the western approach to
Kathmandu from the Prithvi Highway, and a by-pass for Hetauda.

1. Access to Kathmandu
The provision of an alternative access to Kathmandu from either the south (Terai)
or west (Prithvi Highway) is possibly the single most important prospect for the
improvement of traffic conditions and the creation of a major economic impact in
Nepal over the coming decade. Reference to Figure 3.6, which illustrates the
current traffic flow pattern, reveals that the only roads with significant volumes of
traffic are in the Central Region and the dominant movement is on the indirect
link between Kathmandu and Birgunj, via Mugling and Narayanghat.

DHV - SILT - TAEC – CMS Chapter VI-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VI-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Any improvement or reduction in distance for traffic on this route will clearly have
a significant and far-reaching impact. The potential economic benefits to road
users are clearly of an entirely different magnitude to those available from
improvements to hill roads in the west, where traffic volumes are negligible in
comparison. The options for the Fast Track are considered in more detail below.

The two sets of alternatives – from the south or west – are, to an extent, inter-
dependent as both potentially serve the same (or similar) traffic. If the Fast Track
to the Terai were to be committed (and built) in the near future, then the demand
and requirement for an upgrade of the western approach to Kathmandu would be
significantly reduced. However, if a decision to construct the Fast Track were to
be deferred for at least 10 years, then the pressure and justification to improve
the western approach to Kathmandu from the Prithvi Highway would be
substantially strengthened. The overall location of the improvements are shown
on Figure VI.3.

It is clear that, of the two alternative western approaches examined, the shorter
option from Darke via Bhimdhunga is preferred. The second option – linking from
the Galchhi-Devighat road through the Kolphu Valley to enter Kathmandu near
Balaju – is substantially longer and less well aligned. Either option would produce
benefits of a similar scale and thus the shorter – and consequently cheaper –
route is preferable.

In this Study, it has not been possible to assess the impact of all the externalities
and inter-relationships relating to the timing and sequencing of the provision of
improved access to Kathmandu. It is however clear, from the analyses that have
been conducted, that the rates of return on the (more expensive) direct links
between Kathmandu and the Terai are of a similar level to those from the (more
modest) improvements to the western approach. However – if the Fast Track
option is likely to be built in the longer term – the benefits from the improvements
to the ‘western’ approach will be reduced.

The level of potential benefits from the Fast Track option is much greater. This
would suggest that, if guaranteed funding were available on a timely basis, then
the full Fast Track option should be pursued and the western approach
improvements deferred. However, if there is uncertainty regarding the ability to
complete the Fast Track within a fixed timeframe, then further more detailed
consideration should be given to the ‘western’ improvements.

2. By-Passes
There are substantial benefits available from the construction of a by-pass for
Heatuda, due to the significant volumes of through commercial traffic plus the
possibility that the by-pass route could be shorter than the more congested route
through town.

By-passes of other towns on the East-West Highway were considered but none –
with the exception of Narayanghat/Bharatpur, which is already by-passed – offer
similar potential benefits. It is suggested that the existing Narayanghat by-pass
(FR37) should be considered for upgrading, together with the relocation of the
bus stand from the centre of town and the introduction of a ban on through bus
and truck traffic.

Other potential by-passes should be considered for the border towns of Birgunj
and Biratnagar, to reduce the significant congestion on the existing approaches
to the border crossing. The situation in Birgunj could be substantially improved
with the operation of the ICD for road based traffic.

DHV - SILT - TAEC – CMS Chapter VI-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

D. Kathmandu - Terai Links

1. Previous Studies
The previous PIP Study examined a number of possible alignments for a new link
between Kathmandu and the Terai: these were subsequently reviewed by the
IDA-funded North-South Corridor Study in 2004. The options studied involved
construction on a new alignment, including tunnel sections, linking Kathmandu to
Hetauda (via Bhimphedi), as had been previously studied independently in the
early 1990s by both Finnida and the Japanese. The previous PIP Study also
evaluated the potential function and effect of the JICA-funded Sindhuli Road as
an alternative route to the Terai. Other possible options, including the upgrading
of the existing Kanti Rajpath, were reviewed and discounted.

The findings of previous PIP have been revised and updated. The basic
conclusions remain valid, with the tunnel options reducing the current distance
from Kathmandu to Hetauda (220 km via Mugling) to around 65 km. A reduction
of this magnitude – on a route carrying large volumes of commercial traffic – will
produce significant transport cost savings, together with the possibility of private
sector funding with repayment through toll revenues.

The Sindhuli Road, when complete in 2010, will provide a substantial reduction in
travel distance between Kathmandu and the Eastern Terai, although it will not
benefit traffic bound directly to Hetauda, Birgunj or India via Raxaul. The full
potential benefits to Eastern Region traffic will however only be obtained if the
road were constructed to full 2-lane bitumen standards 16.

2. The Fast Track

The objective of the development of a Fast Track route is to reduce significantly
both the distance and travel time for all traffic travelling between Kathmandu and
the Terai. For geographic, topographic and historic reasons the existing routes to
Kathmandu are both lengthy and indirect. The first route completed, in 1956, was
the Tribhuvan Rajpath, constructed by Indian Army Engineers. It follows a
tortuous ‘hill’ alignment crossing three significant ridges and has limited capacity:
the road is single lane, is in poor condition and resultant traffic speeds are low.

Subsequently an alternative route was completed in the 1970s which, in the

main, follows river valleys from Kathmandu to the Terai, via Mugling and
Narayanghat. It is less direct – being almost 90km longer – but has a
considerably superior alignment, allowing greater speeds. Despite the increased
length, the route is generally quicker (especially for commercial vehicles) and is
used by virtually all traffic between Kathmandu and the Terai.

The possibility of a more direct route has been discussed for many years and has
been the subject of a number of investigations. The ‘crow-fly’ distance from
Kathmandu to Hetauda for example is just 40km, compared with 130km via the
Rajpath or 220km via Mugling and Narayanghat. The more direct tunnel routes
offer the possibility of reducing this distance to around 70km. The location of the
options are shown in Figure 6.3.

Access to Kathmandu from the south is of major commercial, economic and

strategic significance. Virtually all goods and services enter Nepal from the south
via India – except for limited movements by air or across the land border with
Tibet to the north. The main border crossing with India is at Birgunj (south of

16
The current construction provides only a 5.5m roadway, which is unsuited to significant volumes of
heavy commercial traffic

DHV - SILT - TAEC – CMS Chapter VI-11

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Hetauda), where there is a connection with Indian Railways and a recently

constructed Inland Customs Clearance Depot (ICD): over 50 percent of all
Nepal’s external trade is through Birgunj 17.

Traffic volumes on the existing approach route to Kathmandu have been

increasing steadily in recent years. Volumes have increased from around 1500
vehicles per day in the early 1990s to approximately 4,000 vpd today,
representing a consistent growth of around 10 percent per annum. Any significant
reduction in the travel time between Kathmandu and the Terai will obviously lead
to further substantial increases, in addition to any continued ‘natural’ growth.

A recent study (Nepecon, 2003) specifically examined routes that did NOT
involve tunnelling. The study examined a series of options including the
upgrading of existing routes, the development local road alignments, and a ‘new’
option following the Bagmati River and terminating at Nijgadh on the East-West
Highway, approximately 20km east of Pathlaiya (see Figure 6.3). The Report
concluded that the development of a no tunnel ‘Fast Track’ concept was only
possible on the Bagmati alignment, as the alternatives were considered to be not
technically feasible due to the lengths of route involved and the difficult terrain.

It is evident that a direct link between Kathmandu and the Terai would offer
the possibility of reducing the distance to Hetauda from the current 220 km
via Mugling to around 70 km, with the same reduction of 150 km for all trips to
the Indian border at Birgunj/Raxaul and to eastern Nepal. The construction
costs will be substantial – currently estimated at US$240 million – but, given
the significant distance and operating cost savings, the project can be shown
to be economically feasible. Additionally, it is probable that the project could
also offer potential financial returns to a private investor, with the possibility of
tolls related to user cost savings.

The preferred alignment of the ‘tunnel’ route between Thankot (at the entrance to
the Kathmandu Valley) and Bhimphedi (north of Hetauda) is illustrated in
Figure VI.4. The route, as originally developed in the 1993 Study, involves two
tunnels, each of around 3.5km beneath the major ridges. The central section of
the route follows the eastern bank of the Kulekhani Reservoir: the southern
section involves upgrading the existing Feeder Road (FR19) between Bhimphedi
and Bhainse and thence a section of the Tribhuvan Rajpath (H02) to Hetauda.

It is clear that further study is required of the available options, including both
tunnel and no-tunnel alignments. There are a series of issues relating to relative
cost, impact and effectiveness that need to be addressed.

E. Further Network Upgrading

The Study has also examined the potential for further development and/or
upgrading of the SRN in the period to 2016. These potential improvements are
based on consideration of road capacity, deficiencies, missing links or other
network ‘weaknesses’.

17
Biratnagar and Bhairawa account for a further 10-15% each, other crossings to India 10%, crossings
to Tibet 8%, and the airport 6%: all by value.

DHV - SILT - TAEC – CMS Chapter VI-12

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VI-13

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Consideration of traffic volumes in Nepal (see Figure 3.6) suggests that there are
likely to be few instances where highway capacity is a major concern, requiring
investment in widening or duplication. The exceptions to this are roads within
Kathmandu Valley (see below) or the approach route to Kathmandu (Naubise to
Thankot) which carries substantial volumes of commercial traffic and is both
steep and poorly aligned. It is recommended that minor improvements only be
considered to the Thankot-Kalanki section of the Tribhuvan Rajpath in the short
term, to accommodate existing demand, until a satisfactory long-term solution is
developed for improved access to Kathmandu Valley. Longer term improvements
need to be examined is the context at the proposed Kathmandu-Terai Fast Track
and/or the construction at a new route from the west (Prithivi Highway). A futher
possibility, worthy of consideration, would be the construction of a short tunnel
(up to 1km) beneath the ridge at Thankot/Nagdhunga to eliminate the steepest
section of the approach to the summit from the west.

1. Kathmandu Valley Roads

A full assessment of the transport and highway capacity issues and requirements
within the Kathmandu Valley is beyond the scope of this present Study: however,
based on the available traffic data and an overall knowledge of the conditions, it
is possible to make some preliminary observations and recommendations.

Apart from issues wholly within the urban area, there are two broad aspects that
require consideration:
• the capacity of key radial routes, including specifically Koteshwar-
Bhaktapur-Benepa-Dhulikhel and Kalanki-Thankot; and
• the capacity and operation of the existing Kathmandu Ring Road.
The capacity problems and related issues of both the radial and orbital roads are
compounded by poor traffic management and regulation, inefficient traffic control,
a diverse mix of vehicle types, and bad driver behaviour. Many of problems
experienced could be alleviated through relatively cheap and cost-effective
measures and by improvements to driver discipline.

Notwithstanding this, the capacity of the urbanised sections of both the main
radial highways (to Thankot and Bhaktapur) will require improvement in the near
future. Relief to the Kalanki-Thankot section will be provided in conjunction with
the implementation of either the Fast Track or Bhimdhunga projects, however, a
package of short-term improvements should also be considered. These could
involve minor widening, construction of sidewalks, the provision of off-street bus
parking and, most critically, the enforcement of strict regulations regarding vehicle
parking and waiting.

Similarly, the Koteshwar-Bhaktapur corridor could benefit from improved traffic

management, regulation and enforcement. However, the traffic volumes are such
that a duplication to provide 4-lanes for through traffic, plus service roads for local
traffic, would almost certainly be justified. Just such an improvement was
recommended in the previous PIP 10-years ago and, it is understood that the
Japanese Government has recently committed to support the improvement of this
corridor..

It is suggested that the improvement should include some form of grade

separated crossing for local traffic, possibly involving the construction of under-
passes beneath the main roadway, linking the frontage roads each side.

It is probable also that, within the next 10 years, that the widening should be
extended beyond Bhaktapur initially to Banepa and subsequently to Dhulikhel. A

DHV - SILT - TAEC – CMS Chapter VI-14

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

study of the justification and timing for this improvement should be conducted in
the context of the overall pattern of urbanisation and growth potential within the
Kathmandu Valley and adjoining areas. The corridor to the east, towards Banepa
and Dhulikhel, offers the only significant opportunity for growth in the Kathmandu
Metropolitan Region.

Congestion and delays on key sections of the Kathmandu Ring Road are causing
considerable difficulties and problems for both residents and businesses alike.
Significant delays are regularly experienced at Kalanki (where almost all vehicles
entering or leaving the Valley must pass), at Balaju and near the bus park, at the
junction with Maharajgunj, and in the Chahabil area. Whilst it is recognised that
many of the difficulties and problems are a result of poor traffic management and
ill-disciplined driver behaviour – including specifically buses and micro-buses
waiting for passengers in, and on the exits to, the junction – it is probable also
that some form of junction capacity improvements will be required in the coming
10 years.

In addition to the junctions – which are the prime determinant of the capacity of
any urban road – further improvements to the overall cross-section of the Ring
Road are required to ensure that the road operates efficiently and safely. It is
recommended that the road be progressively re-modelled to provide two full
lanes in each direction, physically separated with a dividing median, and with
frontage service roads on each side to access adjacent development. A
comprehensive package of works should be envisaged, including environmental
improvements, tree planting, facilities for pedestrians, and controls on truck
movements and parking.

DHV - SILT - TAEC – CMS Chapter VI-15

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 7
VII. ACCESSIBILITY AND STRATEGY NETWORK EXPANSION

A. Concepts and Measures of Accessibility

Nepal is one of the few countries in the world with a significant proportion of its
total population living in areas not served by a motorable road. Previous
estimates have indicated that up to 40 percent of the population of the hills are
more than 4 hours walk from an all weather motorable road and 13 percent of the
Terai population are more than 2 hours walk from a road. These areas of poor
accessibility are strongly correlated with the incidence of poverty and low levels
of human development.

Measurement of improved levels of rural accessibility are being used throughout

this Study as the prime indicator of project effectiveness and overall ‘worth’.
Accessibility analysis depends on an understanding of both population
distribution and the geographical relationship to the associated road networks:
both elements may be plotted graphically using GIS techniques to measure and
quantify the effects.

The methodology developed is intended to assist with future road network

planning by allowing relationships between the existing (or future) population
distributions and any given road (or road network) to be established. The system
is fully coordinated to allow integration with road design and to enable the effects
of terrain to be modelled.

The existing population of Nepal has been plotted geographically onto a 1 sqkm
grid – representing a population density distribution – and an estimate of access
times to or from the road network has been developed using a digital terrain
model, to replicate walk-speeds across different terrain types and make
allowance for the barrier effect of rivers. It is thus possible to calculate the
populations served by any road network, within the given 2 hour and 4 hour walk
times in the Terai and hills respectively and also to compute the associated total
(person-hours) walk time to/from any network.

A principle objective of the Study is the identification of a transport network

comprising both strategic and rural roads that will bring inhabited areas of Nepal
within 4 hours walk in the hills and 2 hours walk in the Terai from an all weather
motorable road.

In 2004 under the Rural Access Improvement Project (RAIP) a preliminary

accessibility profile and maps of all Village Development Committees (VDCs) of
Nepal to all weather roads were prepared. The profile and the maps of the hill
and mountain districts as well as Terai districts were prepared based on 4 hours
and 2 hours walking time from the VDC centres to all-weather roads. The profile
and maps were prepared manually based on hard copy of 1:25000 topographic
maps. This was based on an estimated relationship between the aerial (straight-
line) distance to walking time. The aerial distance between each VDC centre the
nearest all weather road was measured and the walking time calculated. All the

DHV - SILT - TAEC – CMS Chapter VII-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

VDCs centres within 4 hours walking time in the hills and 2 hours walking time in
the Terai were considered accessible, and those outside the walking time were
considered inaccessible.

Based on the above analysis, about 39% of the people in hills and mountains
were beyond a 4 hour walk to the nearest all-weather road and about 13% of the
Terai people were more than 2 hours from an all weather road. Preliminary
analysis during this Study of these existing data, and of the information available
from DOLIDAR, revealed that out of 55 districts in hills and mountains 46 can be
classified as having a serious access problem, with over 20 percent of the
population are outside 4 hours criteria. Of the Terai districts, more than 10
percent of the population are beyond the 2 hour limit in 9 districts out of the total
of 20.

The Consultants have refined the method used by DOLIDAR by expanding the
population distribution within the individual 3913 VDC down to settlement level as
well as developing a terrain-based model to replicate walking speeds in the hills.

B. Population Density Distribution

1. Approach
An understanding of the distribution of population throughout the country is a
fundamental input to any assessment of accessibility and the associated
development of road networks to improve the current access situation. A
quantified understanding of population densities and distribution provides
valuable input to the identification of new road alignments and the calculation of
populations within influence areas.

The population data are based on the 2001 census information at VDC level.
These populations have then been “redistributed” geographically based on the
distribution of houses throughout the VDC as recorded on the Finnmap surveys
from the mid-1990s. Adjustments have been made with reference to the urban
clusters, to reflect the effect of higher population densities in the urban core
areas.

Estimates of 2006 VDC populations have been based on an extrapolation of the

growth rates between the 1991 and 2001 censuses: this provides an indication of
any on-going migration trends – including within districts (eg to the DDC centre),
between Hill Districts and the Terai, and to the major urban centres (eg
Kathmandu, Birgunj or Biratnagar). This is effectively represented by higher
growth rates in the Terai and urban areas, and lower rates (including some
negative values) in remote, rural VDCs.

The analysis procedure provides both a visual indication of density, see

Figure VII.1, and an ability to compute (& assemble) populations within any given
area. This is a basic input to the accessibility analysis and to the development
and evaluation of the extensions to the SRN.

The higher overall population densities in the Terai are immediately evident, as
well as the concentrations of population in and around Kathmandu and other
major urban centres: the relatively even population distribution throughout the
mid-hills area is also evident, together with the extremely sparse habitation in the
northern mountain areas.

DHV - SILT - TAEC – CMS Chapter VII-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

For the purposes of the analysis, a further refinement of the definitions of “hills”
and “Terai” has been adopted, based on the actual terrain rather than the District
administrative boundaries. In addition to the main contiguous area of the Terai
adjacent to the Indian border, significant other plain areas in the “Inner Terai” and
major Valleys (Kathmandu, Pokhara, Dang, Surkhet, etc) have been included, as
illustrated in Figure VII.2. As a result, the total 2006 population (25.87 million) is
divided as 11.3 million in the hills and mountains and 14.6 million in the Terai and
Valleys: the comparable 2006 figures based on the ‘Hill & Mountain Districts’ and
the ‘Terai Districts’ are 13.1 and 12.8 million respectively.

2. Methodology
The census record provides the population at VDC level. This population figure
can be utilized to develop a population distribution assuming the population is
evenly distributed in a VDC to provide an average population density by VDC. In
reality the population is not evenly distributed and VDCs differ in size by a large
margin and thus an equal population distribution within each VDC will give
misleading results.

In the present Study a different approach has been used to derive the population
density. Following are key aspects of the methodology used:

a) The 1996 & 1994 topographical mapping contains the houses as well as built
up areas for the entire area of Nepal. These locations provide an indication
of the population distribution within any given boundary. The VDC population
distribution is proportional to the map house counts and defined built-up area
markings.
b) The 2001 census population and 2006 population projection have then been
distributed among the indicated houses and the built up areas. This produces
a reasonable distribution of the population within each VDC as growth is
concentrated on the existing settlement in proportion to the numbers of map
house count and the built-up areas.
c) In case of the built-up areas, to derive the population density, a “simulated”
house is assumed at 15 x 15m size and accordingly the numbers of houses
are estimated based on the built up area. The population is then distributed
within a VDC by summing up the individual house count from map and the
simulated houses from the built up area. For each house point a population
is derived by dividing the total population by the house count.
d) To derive the population distribution, a 1 km x 1 km grid is used within which
all the house point populations are summed. This produces a population
density distribution for the whole country on a 1 sqkm grid.

C. Development of Walk-Time contours

The second important aspect in assessing accessibility is the calculation of walk-
times in different terrain types. The Consultants have developed a system based
on digital terrain modelling (DTM) using a 90m grid of spot-heights which allows
walk times and distances to be calculated and calibrated against known travel
times.

The concept of the walk-time contours has been developed based on the DTM
and walk speeds achievable in different terrain. Additionally a defined network of
major rivers has been used to represent a barrier to movement – apart from at
defined (known) bridge locations. This approach has been calibrated against
known travel times (and routes) and can be shown to produce acceptable results.

DHV - SILT - TAEC – CMS Chapter VII-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The approach adopted in this Study to assess rural accessibility is thus much
more detailed than the earlier DoLIDAR work. The method is based on actual
terrain information, rivers and bridges, instead of computing a “crow fly distance”
with subsequent adjustments, on a subjective basis, of the terrain and drainage
factors. The method adopted uses a rigorous mathematical approach which takes
account of the terrain effects (based on a 90m grid of spot-heights) and the
barrier effect of rivers. The basics of the method adopted are as follows:
• Assumption of a base travel distance per hour (effective walking speed)
for terrain with less than 12% gradient of 4 kph;
• Linear decrease in the travel speed with an equivalent increase in
gradient: e.g at 24% gradient, the travel speed will be 2 kph;
• Development of a 90m slope-grid for the entire country and assignment of
the travel time for each element of the grid based on its slope;
• Cell based modelling for computing the travel time from origin to
destination: this approach gives the least time to reach any destination
from the selected origin; and
• Modelling non-crossable major rivers as barriers with bridges as the only
access points across such rivers.
In order to model the river and bridge effects, the location of trail bridges on the
major rivers were used in the analysis. The bridge locations are taken from the
1996/94 topographical maps as well as from the MOLD Trail Bridge Division
maps. There are numerous additional minor rivers for which there is no detailed
bridge information: these rivers have therefore been excluded from the present
analysis.

The model has been calibrated in selected locations of the Nepal where average
travel time were available. The following cases were analysed.

i. Sundarijal to Tharepati: the above approach was tested for the origin at
Sundarijal up to the destination at Tharepati. The reported travel time from
the trekking guide maps is about 15.5 hours in forward direction and 18.5
hours in backward direction. The model estimated travel time in forward and
backward direction as 16.7 hours. As model uses same difficulty levels for
travelling uphill and downhill hence the travel distances are same for both the
directions. The mean error in the model and the reported travel time is about
-1.8% with forward error of 7.7% and backward error of -9.7%.

ii. Khandbari to Num: The route Khandbari, District HQ of Sankhuwashbha

District, to Num is a heavily used route accessing the northern part of the
district and the Chinese border. This trail is the only means of getting to Num
and areas beyond. The reported travel time from the trekking guide maps is
about 9.5 hours in forward direction and 10.5 hours in backward direction.
The model estimated travel time of 11.2 hours. The mean error in the model
and the reported travel time is about 10.7%.

iii. Jiri to Lukla : The route from the Jiri to Luka is a very popular trekking route
towards the Everest Region. The reported travel time from the trekking guide
is about 45.75 hours (or 7 days) in the forward direction. The model estimated
a travel time of 50.2 hours, resulting in a mean error in the model against the
reported travel time of about 8.9%

DHV - SILT - TAEC – CMS Chapter VII-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

It can be seen from the above comparisons that the model has been validated
and provides reasonable estimate of the travel time within a 10% error. There are
many factors that determine travel time in the hilly terrain, all of which can not be
modelled precisely. The approach used above provides a sound and reasonable
basis for estimating distance travelled for a given time in various parts of Nepal.

The model calculates the shortest time path across the terrain, with speed
dependent on the gradient. The model thus replicates the likely (minimum time)
path and the time contours from any location – allowing travel times and zones of
influence (ZoI) to be calculated. The results from the model’s route selection
process confirm the adoption of a ‘path of least resistance’: ie the routes selected
will – because of the relative travel speeds – tend to follow ridges, contours or
valleys, whilst avoiding the steeper land. This generally reflects the alignment of
trails established in Nepal.

D. Accessibility of Strategic Road Network

1. Overall Function of SRN
Previous Chapters of this report have discussed and described the development
of the Strategic Road Network (SRN) from its present extent of around 5,000km
to a projected 9,700km by 2016. It has also been stressed that the SRN ‘enables’
the improvement of rural accessibility, through the provision of trunk links and
access to all Districts, but that local accessibility will only be achieved through the
provision of a complementary network of local or rural roads.

Notwithstanding this, the impact of the SRN on overall levels of accessibility is

significant. Apart from in the Terai and main urban centres, the existing SRN
provides the primary – and in many cases the only - all-weather access facility.
This includes substantial areas of the country and especially the more remote
regions.

The overall access situation in the country is illustrated in Figure VII.3, based on
the currently committed strategic network. This shows all those areas of the
country that are more than 10 hours walk from the nearest road, and also time
bands up to 60 hours walk – around 8 days. These areas cover a significant
portion of the country but – it should be remembered – these areas are sparsely
populated: the total population in the areas indicated as being over 10 hours from
a road is approximately 1.9 million or 7.5 percent of the total population.

Substantial areas of the hills and mountains rely on access from a network of
“hill” roads – ie roads that connect northwards form the main elements of the
SRN: the East-West Highway, Prithvi Highway and Siddartha Highway. There are
around 22 individual ‘catchment areas’ served by such hill roads, as illustrated in
Figure VII.4, with a combined population of 9.3 million – or around 36 percent of
the national total. [The remainder of the population, including the Terai and main
urban centres, are served by the core road network.]

The role and function of these hill roads are essential for economic development
and activity within the areas served: the roads provide the economic and social
life-line for the hills, providing for the import of most goods consumed and the
export of any production. The development and upgrading of the networks
serving the Mid and Far Western Regions are the subject of the Feasibility
Studies conducted under Part II of this Study.

DHV - SILT - TAEC – CMS Chapter VII-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

2. Expansion of the SRN

The Study recommends the expansion of the designated SRN from its current
(operational) length of 5,030km to almost 10,000km by 2016. This expansion is
achieved primarily by the re-designation of the more important (and potentially
strategic) existing District or Local Roads as part of the SRN. Of the additional
4,600km of Strategic Road, over 3,300km (70 percent) are existing District or
Local Roads: less than 1,300km of new roads are proposed.

Four networks have been defined to assess the impact of the network growth on
accessibility, as illustrated in Figure 7.5:

• The Existing Designated SRN (2006) – 5,030km – comprising all sections

of the designated SRN (15 Highways and 51 Feeder Roads) that are
currently operational;
• The Existing (effective) Operational SRN (2006) – 7,360km – including
additionally all existing and open sections of road that will be incorporated
into the future SRN – including substantial lengths of District Roads;
• The Committed SRN for the Year 2010 – 8,390km – as above, plus
additional projects scheduled for completion by 2010 that are currently in
hand and for which funding is assured;
• The Future Extended SRN for the Year 2016 – 9,930km – includes the
additionally the extensions to the SRN proposed in Chapter 5.

The first two relate to 2006 and are respectively the ‘official’ and ‘de-facto’
networks as maintained by DoR; the third includes the additional committed
schemes, and those in the ‘pipeline’ with funding committed, that will be in place
by 2010; and the fourth represents the recommended extended SRN by 2016.

The composition and source of the additional 4,900km of proposed road are
presented in Table 7.1, including the completion of 306km of currently designated
(but not operational SRN).

Table 7.1: Recommended Additions to the Strategic Network (2006-2016)

Additional Committed Proposed Total

Length Operational & Pipeline Additional Additional
Percent
(km) Roads Additions Links Road
(2006) (2010) (2016) Length
Designated SRN 0 229 77 306 6%
Local Hill Roads 1,059 70 501 1,631 33%
Terai Roads 513 214 71 798 16%
Postal Roads 614 129 0 743 15%
Urban Roads 147 0 0 147 3%
New Alignments 0 388 890 1,278 26%
Total 2,334 1,030 1,540 4,904 100%

A full network description by District and individual road is presented in Annex 7,

Tables A7.1 and A7.2.

DHV - SILT - TAEC – CMS Chapter VII-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Accessibility Provided by SRN

The resultant plots of accessibility by time-band are illustrated on Figure VII.5,
showing those areas within 1, 2, 3 & 4 hours walk from the SRN. The populations
outside these areas are shown in beige, illustrating the density distribution of the
non-served (or inaccessible) population. A summary of the overall accessibility
provided by each of these four networks is given in Table 7.2, illustrating
an overall improvement from 65 percent to 86 percent within the 2 hour and 4
hour criteria.

Table 7.2: Accessible Populations to Existing & Extended SRN

SRN Length Population served
km Hill Terai Total
Existing Designated SRN 5,030 5.68 11.07 16.74
2006 50% 76% 65%

Existing / Operational SRN 7,360 6.52 13.70 20.22

2006 58% 94% 78%
Committed Additional SRN Links 8,390 7.08 14.00 21.08
2010 63% 96% 81%
Extended SRN 9,930 7.94 14.21 22.15
2016 70% 97% 86%
Note: Total Population of 25.9 million is divided into Hills 11.3 million & Terai 14.6 million

3. Accessibility for Existing Designated SRN (2006)

The Existing Designated SRN has a total length of 5,028km. The accessibility
provided by this network is 50 percent within a 4 hour walk for the hill areas, and
76 percent within a 2 hour walk in the Terai and Valley areas. Nationally, the
overall accessibility is 65 percent, which implies that 9.3 million people (36
percent of the total population) are living outside the desired accessibility norms,
including half of the total population of the hills.

4. Accessibility for Existing (effective) Operational SRN (2006)

The Existing Operational SRN has a total length of 7,360km and includes the
designated SRN plus other operational links under DoR management. The
accessibility provided by this network is 58 percent in the hills and 94 percent in
the Terai and Valley areas. The overall accessibility is increased to 78 percent,
with the bulk of the improvement in the Terai, due to the inclusion of a substantial
network of existing all-weather roads, including the Postal Road. The additional
roads considered are coloured red in upper map on Figure 7.5. The improvement
in the hills is primarily due to the inclusion of the road to Okhuldhunga in the east
and to Burtibang in the west.

5. Accessibility for Committed SRN for the Year 2010

By 2010 the Committed SRN will comprise 8,390km, with the addition of 801km
of roads to the existing operational SRN, as shown in purple on middle map on
Figure 7.5. The accessibility will be improved to 63 percent in the hills and 96
percent in the Terai and Valley areas: the overall accessibility increases to 81
percent. The improvements are mostly in the hills, resulting from the northerly
extensions to the network serving non-connected District Headquarters.

DHV - SILT - TAEC – CMS Chapter VII-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-11

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-12

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-13

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

6. Accessibility for Future Extended SRN for the Year 2016

The Extended SRN in 2016 will comprise 9,930km with the addition of a further
1,463km of roads, mostly in the mid-hills. The accessibility in the hills provided by
this network is increased to 70 percent and to 97 percent in the Terai and Valley
areas: the overall accessibility is 86 percent. The overall increase in accessibility
results form the additional roads in the mid-hills.

7. Impact of SRN on Accessibility

The above analysis indicates that the overall accessibility provided by the SRN
increases from a nominal 65 percent today (based on the officially designated
network) to 81 percent with the completion of current commitments in 2010, and
to 86 percent with the proposed expansion by 2016. Over the 10 year period an
additional 5.5 million people will – potentially – be brought within range of the all-
weather SRN: less than 4 million people (almost exclusively in the hills) will be
outside the 4 hour walk limit in 2016.

E. Effect of the Local Road Network

In addition to the SRN, there is a sizeable network of local District and Village
Roads that contribute significantly to overall levels of accessibility, although many
of these roads are of low standard and are – at best – seasonal in nature. An
analysis of the improved accessibility potential of these roads is given in the
following Chapter, together with an assessment of the additional requirements to
achieve higher levels of accessibility throughout the country.

The Consultants have identified a network of approximately 4,460km of existing

(operational) local roads (see chapter 8). The impact of this ‘existing’ Local Road
Network (LRN) is less than might be expected as most of these roads are
relatively short and are (obviously) located in areas close to and already served
by the SRN. In the remote hill areas, it is the embryonic SRN that provides the
initial access and major improvements to accessibility: the LRN complements this
and enhances access by increasing the proportion of the population within 1-2
hours of the road. Similarly a large proportion of the LRN is in the Terai, where
the basic 2 hour access is already provided by the future SRN: this does not
negate the need nor the value of the LRN, as it is required to provide the higher
levels of access expected (and justified) in the more densely populated areas.

It is however stressed that the detailed planning of the alignments and priorities
for construction of such additional local roads can only be attempted at a local
(District) level. This Study has identified those geographic areas – or inhabited
pockets of land – where the desired standards of access are not achieved and
has suggested how much additional road might be required to resolve the short-
comings.
Nonetheless – given the topography and population distribution of Nepal – it is
unrealistic to expect that road access could be practically or feasibly provided to
all settlements. As has been indicated above, some of the more remote areas are
currently 8-10 days walk from a road, and even the most ambitious road building
programme would not serve all of these areas. Indeed some of the areas are
populated because they are remote and primarily serve adventure tourism – eg
the Khumbu (Sagamartha National Park) or the Annapurna Circuit – where the
resident population supplies lodging and facilities to tourists. [These areas are
incidentally some of the richest in rural Nepal and would not necessarily welcome
road construction.]

It is therefore unrealistic – and undesirable – to attempt to achieve a 100 percent

coverage of 4 hour accessibility in the hills. There will always be some remote

DHV - SILT - TAEC – CMS Chapter VII-14

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

areas unreserved by road as it would be totally uneconomic and environmentally

disastrous to construct roads into these areas for the minimal populations
involved.

F. Assessment of Total Walk Time

An alternative indicator of the level of accessibility provided by any network can
be measured by calculating the total ‘access time’ to the nearest road from all
inhabited areas. This is expressed in terms of thousands of person-hours, which
represents the product of the number of people and their time (in hours) from the
road. Thus 10 people 1 hour from the road would have the same impact as 1
person 10 hours from the road: this measure makes allowance for the number of
people affected by a road or a road improvement.

The populations in progressive time-bands from the nearest strategic road are
given in Table VII. for the four networks examined. This indicates that
the population within 1 hour of the SRN increases from 41 percent in the
Designated Network to 59 percent in the 2016 Extended Network and, as a
consequence, the proportion over 4 hours reduces from 27 percent to 14 percent.

Table VII.3: Population within Time-Bands of SRN (millions)

Time
Strategic Road Network
Band
Hours Designated Operational 2010 Committed 2016 Extended
1 10.47 41% 13.78 53% 14.55 56% 15.35 59%
2 3.90 15% 3.89 15% 3.77 15% 3.93 15%
4 4.45 17% 3.03 12% 2.94 11% 3.12 12%
8 3.60 14% 2.59 10% 2.69 10% 2.46 10%
12 1.34 5% 1.06 4% 0.98 4% 0.69 3%
24 1.22 5% 0.83 3% 0.50 2% 0.23 1%
over
0.83 3% 0.62 2% 0.36 1% 0.01 0%
24hrs
Total 25.80 100% 25.80 100% 25.80 100% 25.80 100%

These same data are presented graphically in Figure VII. which shows
that the impact of the network extensions are particularly evident in the areas
closest to the road, reflecting the fact that the population densities are highest in
these areas.

DHV - SILT - TAEC – CMS Chapter VII-15

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Figure VII.7: Cumulative Population by Time Band

100%

90% 2016 SRN

80%
2010 SRN
70%
Operational
60%
SRN
Designated
50%
SRN
40%

30%

20%

10%

0%
1 2 4 8 12 24
Hours

The overall total walk times for the same four networks described above are
given in Table VII.4. This shows a substantial (47 percent) reduction in
the total ‘access time’ to the network nationally from 113 million person-hours
walk to 60 million, with similar reductions in both the hills and Terai. It is also
possible to express the impact in terms of average walk time: this shows an
overall reduction of 2 hours from 3 hours 20 mins to 1 hours 20 mins, with the
greatest impact in the hills (from an average of over 7 hours to less than 4 hours).

Table VII.4: Total Walk Time to access Network & Average Walk Time
[Person-Hours (thousands) and Hours]
2006 2006 2010 2016
Network Length km 5030 7360 8390 9930
Hills
Total Walk Time p-hours 83,471 69,440 54,047 42,137
Average Walk Time Hours 6.40 5.16 3.79 2.74
Terai
Total Walk Time p-hours 29,599 19,411 18,475 17,453
Average Walk Time Hours 1.03 0.33 0.27 0.20
Total – All Country
Total Walk Time p-hours 113,070 88,881 72,522 59,589
Average Walk Time Hours 3.37 2.44 1.80 1.30

DHV - SILT - TAEC – CMS Chapter VII-16

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

G. District Level Accessibility by SRN

1. Current (2006) Situation
The currently operational SRN (7,360km) provides access within 4 hours to over
80 percent of the population: the remaining 4.9 million people are distributed
through the hills of Nepal. There are 21 Districts with over 100,000 people more
than 4 hours from a strategic road. These, ranked in order of population, are
listed in the left-hand columns of Table 7.5. It should be noted that there are other
districts with higher percentages of un-served populations, including 22 Districts
with over half the population beyond 4 hours from a road. These are listed in the
right-hand columns of Table 7.5.

Table 7.5: 4-Hour Accessibility by District from 2006 Operational SRN

Districts ranked by Population Districts ranked by percentage

Unserved Unserved
Unserved Unserved
No District Percent No District Percent
Population Population
13 Khotang 238,244 100% 64 Kalikot 120,552 100%
20 Sindhuli 200,413 66% 11 Solukhumbu 111,453 100%
10 Bhojpur 199,788 97% 63 Jumla 76,364 100%
54 Rukum 154,175 75% 65 Mugu 46,387 100%
53 Rolpa 151,742 68% 66 Humla 44,180 100%
9 Sankhuwasabha 144,188 87% 62 Dolpa 30,513 100%
61 Jajarkot 143,488 99% 42 Mustang 14,292 100%
21 Ramechhap 133,808 59% 41 Manang 12,626 100%
47 Palpa 125,462 44% 13 Khotang 238,244 100%
14 Udayapur 121,782 37% 61 Jajarkot 143,488 99%
64 Kalikot 120,552 100% 10 Bhojpur 199,788 97%
69 Achham 117,623 47% 67 Bajura 108,179 93%
74 Baitadi 116,082 47% 9 Sankhuwasabha 144,188 87%
45 Baglung 114,456 40% 54 Rukum 154,175 75%
11 Solukhumbu 111,453 100% 43 Myagdi 89,512 74%
46 Gulmi 110,024 36% 1 Taplejung 95,810 68%
67 Bajura 108,179 93% 53 Rolpa 151,742 68%
12 Okhaldhunga 107,096 65% 20 Sindhuli 200,413 66%
68 Bajhang 103,694 57% 12 Okhaldhunga 107,096 65%
28 Nuwakot 102,678 34% 21 Ramechhap 133,808 59%
52 Pyuthan 102,526 45% 68 Bajhang 103,694 57%
75 Darchula 66,468 53%

It can be seen that, based on population, there is a significant cluster of 8

Districts in the east – Khotang, Sindhuli, Bhojpur, Sankhuwasabha, Ramechhap,
Udayapur, Solukhumbu and Okhaldhunga – all with over 100,000 people more
than 4 hours from the operational SRN. The total un-served population in these
Districts is in excess of 1.25 million

Similarly there are 12 Districts in the West, Mid-West and Far-West each with
over 100,000 people un-served and with a total population of 1.5 million
stretching through the mid-hills from Palpa, via Gulmi, Baglung, Pyuthan, Rolpa,
Rukum, Jajarkot, Kalikot, Achham, Bajura, and Bajhang to Baitadi in the Far
West.

This analysis presents a different complexion to one based on percentages of

non-served population. There are nine Districts with 100 percent of the population
more than 4 hours from a road, and a further four Districts with only a small
proportion of the total served by road, although many of these Districts have low
absolute populations. The Table highlights the need for improvements in the

DHV - SILT - TAEC – CMS Chapter VII-17

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

more densely populated mid-hills region, and also identifies the substantial un-
served populations in the relatively more affluent eastern areas of the country.

A full listing of the accessibility parameters in 2006 by District is presented in

Annex 7, Table A7.3, which provides data on the hill and Terai populations, the
percentages of these within the 2-hour and 4-hour limits, the total person walk-
time to the nearest road, and the average access time to the road.

The ‘severity’ of the lack of access is probably best measured by the total person-
hours walk from the road network, which allows for the incorporation of both the
time from the road and the numbers of people affected. The 25 least accessible
districts in 2006 ranked on these criteria are presented in Table 7.6 and are
illustrated on Figure 7.8.

Table 7.6: Ranking of 25 Least-Accessible Districts – 2006 Operational SRN

District Population (‘000s) Percent Man-hrs Av Walk

Dist
District Accessible Walk Time
No Hill Terai Total (2hr/4hr) (‘000s) (hours)
13 Khotang 239 - 239 0% 2,976 12.4
63 Jumla 78 - 78 - 2,936 37.5
66 Humla 44 - 44 - 2,907 65.8
64 Kalikot 122 - 122 - 2,506 20.5
10 Bhojpur 205 - 205 3% 2,391 11.7
65 Mugu 47 - 47 - 2,379 50.7
9 Sankhuwasabha 169 - 169 13% 2,126 12.6
61 Jajarkot 147 - 147 1% 2,118 14.4
67 Bajura 118 - 118 7% 1,966 16.7
11 Solukhumbu 113 - 113 - 1,838 16.3
20 Sindhuli 271 38 309 33% 1,806 5.8
54 Rukum 207 - 207 25% 1,802 8.7
62 Dolpa 31 - 31 - 1,499 47.9
53 Rolpa 227 - 227 32% 1,222 5.4
36 Gorkha 306 - 306 72% 1,124 3.7
43 Myagdi 124 - 124 26% 1,036 8.4
47 Palpa 288 - 288 56% 1,029 3.6
21 Ramechhap 225 - 225 41% 1,014 4.5
74 Baitadi 255 - 255 53% 1,006 3.9
45 Baglung 288 - 288 60% 945 3.3
69 Achham 250 - 250 53% 935 3.7
30 Dhading 373 - 373 80% 921 2.5
12 Okhaldhunga 165 - 165 35% 919 5.6
68 Bajhang 183 - 183 43% 916 5.0
14 Udayapur 178 153 331 56% 916 2.8
Note: Ranked according to walk-time to nearest SRN

These Districts can be seen to form two major ‘blocks’ in the east and west, with
the 10 worst Districts (four in the east and six in the west) being: Khotang,
Bhojpur, Solukhumbu and Sankhuwasabha in the east; and Humla, Jumla, Mugu,
Bajura, Kalikot and Jajarkot in the west. It is also noted that Gorkha and Dhading
rank in the top 25, due to substantial populations in the northern areas away from
the District HQs.

The average walk time to access the SRN ranges from a high of 66 hours in
Humla to 2.5 hours in Dhading, the lowest among the worst 25 Districts. In other
more accessible Districts, the average walk time reduces to fractions of an hour,
with a national average of 2.4 hours. Overall, the average walk time to a road
currently exceeds 8 hours (1 day) in a total of 16 Districts.

DHV - SILT - TAEC – CMS Chapter VII-18

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-19

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

2. Impact of Extended SRN in 2016

Proposals to extend the SRN to 9,930km by 2016 are described elsewhere,
resulting in an overall increase in accessibility (measured on the 2hr and 4hr
criteria) from 78 percent to 86 percent, and a reduction in the average walk-time
to access the SRN from 2 hours 30 mins to 1 hour 20 mins.

The number of Districts with more than 100,000 population more than 4 hours
from a strategic road is reduced from 21 to 6: see Table 7.7. The number of
Districts with more than half the population outside the 4-hour limit reduces from
22 to 13, and all Districts have some road access.

Table 7.7: 4-Hour Accessibility by District from 2016 Extended SRN

Districts ranked by Population Unserved Districts ranked by percentage Unserved

Unserved Unserved
No District Percent No District Percent
Population Population
Rolpa 144,446 64% Humla 37,995 86%
Khotang 123,871 52% Manang 9,856 73%
Palpa 122,868 43% Solukhumbu 80,586 71%
Sindhuli 110,807 36% Mugu 32,937 70%
Baitadi 107,361 42% Taplejung 92,947 65%
Bhojpur 105,986 52% Jajarkot 95,470 65%
Rolpa 144,446 64%
Dolpa 16,757 54%
Myagdi 64,311 52%
Kalikot 63,039 52%
Khotang 123,871 52%
Bhojpur 105,986 52%
Okhaldhunga 83,774 51%

The Districts with large populations not served by the extended SRN in 2016 are
all in the mid-hills – three in the eastern part of the country (Bhojpur, Khotang and
Sindhuli) and three in the west (Palpa, Rolpa and Baitadi). These are the areas in
which specific priority should be given to the development of additional local road
access.

The listing of the overall ‘least well served‘ Districts in 2016, following the
expansion of the SRN to 9,330km, is shown in Table 7.8 and is illustrated in
Figure 7.9: full details are given in Annex 7, Table A7.4. The ranking is based on
the calculated total person-hours walk-time from the road network. Almost all of
the ‘Top 25’ ranked least-accessible districts are in the mid-hills, reflecting the
higher population densities coupled with an inadequate road provision. Access to
the previously highly ranked remote northern Districts in the west of the country
will, to a large extent, have been resolved by 2016 through the proposed
extensions of the SRN.

DHV - SILT - TAEC – CMS Chapter VII-20

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VII-21

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

It is also evident that, by 2016, there will be more access deficiencies predicted in
the eastern and central parts of the country than in the west. This may be due –
at least in part – to the ongoing initiatives to improve access in the west. It is also
notable that many of the Districts close to Kathmandu are among those shown to
have poor levels of access.

The maximum average walk-time among these 25 worst Districts is now reduced
to less than a day (7 hours), with an average across the country of 1.3 hours.

Table 7.8: Ranking of 25 Least-Accessible Districts – 2016 Extended SRN

District Population (‘000s) Percent Man-hrs Av Walk

Dist
District Accessible Walk Time
No Hill Terai Total (2hr/4hr) (‘000s) (hours)
53 Rolpa 227 - 227 36% 1,032 4.5
13 Khotang 239 - 239 48% 989 4.1
47 Palpa 288 - 288 57% 955 3.3
61 Jajarkot 147 - 147 35% 939 6.4
36 Gorkha 306 - 306 75% 926 3.0
74 Baitadi 255 - 255 58% 871 3.4
20 Sindhuli 271 38 309 59% 861 2.8
11 Solukhumbu 113 - 113 29% 826 7.3
10 Bhojpur 205 - 205 48% 819 4.0
30 Dhading 373 - 373 83% 791 2.1
9 Sankhuwasabha 169 - 169 59% 782 4.6
21 Ramechhap 225 - 225 56% 768 3.4
70 Doti 229 - 29 63% 767 3.4
28 Nuwakot 310 - 310 71% 739 2.4
24 Kavrepalanchok 419 - 419 83% 719 1.7
1 Taplejung 142 - 142 35% 716 5.1
60 Dailekh 247 - 247 67% 683 2.8
31 Makwanpur 310 125 435 81% 669 1.5
12 Okhaldhunga 165 - 165 49% 663 4.0
46 Gulmi 314 - 314 77% 660 2.1
69 Achham 250 - 250 70% 635 2.5
59 Surkhet 199 115 314 74% 634 2.0
52 Pyuthan 233 - 233 67% 632 2.7
39 Syangja 334 - 334 79% 629 1.9
45 Baglung 288 - 288 75% 628 2.2
Note: Ranked according to walk-time to nearest SRN

This analysis highlights the need for the construction of a complementary network
of local roads to supplement the accessibility provided by the SRN, specifically
within the more densely populated mid-hills areas.

H. Project Selection and Prioritisation in MW + FW

The above approach has been used in Part II of the Study to identify and
evaluate the proposed network improvements and upgrading in the Mid and Far
Western Regions. The overall objective of the proposed road improvement
package is the alleviation of poverty and a reduction in the levels of ‘exclusion’
experienced in the remote hill regions in the west of the country. Access
improvements are key elements in the fight against poverty and the creation of a
more equal society.

DHV - SILT - TAEC – CMS Chapter VII-22

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The total population of the hills in the Mid and Far West is 3.46 million and
currently 2.44 million (70%) are more than 4 hours from a road: the total walk-
time to existing sealed roads is 43.2 million person-hours – or an average of 12.5
hours per person. Upgrading (or constructing) a network of 800km of roads was
estimated to reduce the ‘inaccessible’ population from 2.44 million to 1.81 million,
or 52 percent of the total population – bringing an extra 640,000 people within
the 4 hour access target.

Similarly, the inclusion of the additional 800km of all-weather road would reduce
the walk-time from 43.2 million person-hours to 25.5 million person-hours – an
average of 7.4 hours per person, implying an average 5 hour saving per head.

The populations by time-band before and after the proposed improvement of

800kms of road in the Mid and Far West are presented in Table VII.9. It
can be seen that there is a substantial increase in the population within the 2, 4
and 8 hour time-bands with consequent reductions in the number of people more
than 8 hours from the nearest road.

Table VII.9: Effect of Proposed Road Improvements – MW & FW Hills

Time Band Existing Future Percentage Change in
Percentage of of Population Population
Population
< 2 hrs 19.9% 30.8% +376445
2-4 hrs 9.5% 17.0% +260881
4-8 hrs 15.7% 22.9% +245483
8-24 hrs 38.8% 23.1% -543560
>24hrs 16.2% 6.3% -339249
Total 100% 100% 0

DHV - SILT - TAEC – CMS Chapter VII-23

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 8
VIII. RURAL ACCESS IMPROVEMENT STRATEGY

A. The Role of the SRN and LRN

The previous Chapter has examined in some detail the concept of accessibility
throughout Nepal. This was based on an understanding of the population density
distribution, walking speeds in different terrain, and the extent of the all-weather
road network. The analysis was based initially on the accessibility provided by the
existing Strategic Road Network (SRN) and then subsequently the ‘extended’
SRN that is proposed to be operational in 2016. It is proposed that by 2016, the
SRN will comprise around 9,930 km of all-weather road, linking all Districts and
providing access – within 4 hours in the hills and 2 hours in the Terai – to about
85 percent of the total population. The analysis has indicated that there are a
number of significant ‘pockets’ of population – specifically in the hills – that are
outside the desired accessibility standard: approximately 30 percent of the ‘hill’
population remain un-served by SRN in 2016.

It is evident that the extended SRN cannot – and will not – achieve universal
accessibility throughout the country: that would clearly be both impractical and
unrealistic. The SRN is intended to provide strategic linkage to and between
Districts and it will be complemented by a network of local roads providing local
access.

It is equally obvious that it is impractical to provide road access within 4 hours to

all habitation in the country: this would imply that a road to an isolated nomad in
Dolpa is worthy of consideration ahead of the construction or upgrading of other
substantially more economically viable roads elsewhere in the country. It is thus
necessary to apply a ‘cut-off’ at maybe the 90th or 95th percentile of the national
population, and accept that the remaining ‘extremely remote’ population remains
un-served.

Furthermore, some non-road-served tourist areas owe their popularity to their

remoteness: it would defeat the objective of many of the trekking areas in Nepal if
easy road access were available. Tourists pay substantial dollars in order to
travel to – and walk in – areas untouched by much of modern civilisation to the
economic benefit of the local communities: Nepal has a responsibility to the world
to preserve some of the natural areas of beauty and culture where people can
travel and experience a simpler life.

B. Local Road Network

1. Data Sources
Data for the Local Road Network (LRN) have been compiled from various
sources. The main sources have been the various DTMPs (District Transport
Master Plans), which are available in digital format for 52 of the 75 districts, plus
the Transport Layer of the 1994-96 Topographical Maps (1:25,000/1:50,000). In
addition to these two sources, significant data on the road networks developed
and/or planned by various aid agencies – eg GTZ, RCIW, RAIDP, DRILP, RAP &

DHV - SILT - TAEC – CMS Chapter VIII-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DRSP – were also compiled and projected into a similar coordinate system as
used for the SRN.

It is difficult to estimate exactly the length of the LRN as the reported and map
lengths show significant differences. Furthermore the lengths reported in the
DoLIDAR inventory are difficult to verify due to large difference in the plotted and
tabular lengths.

2. Existing Local Network

Based on the most reliable available data, it is estimated that the LRN currently
contains approximately 4,460km of operational roads, comprising 450 individual
roads in 62 Districts. There are no operational local roads in 13 Districts 18 ,
although sections of road and parts of the future Strategic Network may be under
construction or, in part, operational. A summary of the 2006 Local Road Network
is given in Table 8.1, and the network is illustrated in Figure 8.1.

Only 3 percent of the local network is sealed, a little over 40 percent is gravelled,
and the remainder is simple earth construction: as a result, much of the network,
especially in the hills, is seasonal in nature. Over half of the network (2,390 or
54% of the total) is located in the 20 Terai Districts and, of these roads, three-
quarters are all-weather gravel or black-top. In the hills, by comparison, over 80
percent of the are simple earth construction. A detailed listing of the roads in this
network is presented in Annex 8, Tables A8.1 and A8.2, by both District and
individual road.

Table 8.1: Summary of Local Road Network (2006) by Length (km)

Black Top Gravel Earth U/Constr Total Percent

Terai 87.3 1686.1 552.9 63.5 2389.8 54%
4% 71% 23% 3% 100%
Hills 61.6 196.5 1722.2 86 2066.3 46%
3% 10% 83% 4% 100%
Total 148.9 1882.6 2275.1 149.5 4456.1 100%
3% 42% 51% 3% 100%

3. Future Local Network

It is difficult to predict what will be the extent of the local network in 2016. It is
probable that there will be a substantial and continuing programme of local road
construction promoted through the local District and Village Development
Committees, with assistance from the various donor and development agencies.
These roads will provide enhanced levels of access to substantial populations
within the Districts and will be built to meet local demands for improved access –
especially in regard to local movements to markets, schools and service centres.

The Consultants have adopted two approaches to define the Local Road Network
for 2016, as illustrated in the lower two maps on Figure 8.1. The first (5,860km)
contains all existing and committed local roads and the second includes an
additional 13,500km of planned or proposed roads, taken from the available
District Transport Master Plans (DTMPs).

18
The 13 Districts without local roads are: Sankhusabuwa; Solukhumbu; Manang; Myagdi; Rukum;
Jajarkot; Dolpa; Jumla; Kalikot; Mugu; Bajura; Bajhang; & Darchula

DHV - SILT - TAEC – CMS Chapter VIII-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VIII-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Many of these roads will, however, provide only marginal increases in

accessibility (as measured with the 2 hour and 4 hour criteria), as they will be
constructed in the more densely populated areas that are already defined as
being accessible. Further construction of local roads will obviously reduce the
access time for individuals and will be of benefit to the local community and,
doubtless, additional roads will be constructed that will increase the road density
in populated areas: even today in much of the Terai, as would be expected, the
road density greatly exceeds that required to meet the 2 hour access standard.

This Study is concerned primarily with establishing the basic network of strategic
and local roads that are required to meet the Government’s accessibility
standards. The Study has thus focussed initially on the SRN – which is an
essential pre-requisite for any measure of accessibility – and secondly on those
additional local (or District) roads required to access pockets of population not
otherwise served.

The basic LRN for 2016 has therefore been defined as the 2006 network, plus all
those known additions that are presently under construction or consideration
through the main rural access development programmes. This approach
undoubtedly under-estimates the extent of the future network which will – in the
more populated areas – include many additional local access roads.

Details of the assumed 2016 LRN, by both District and individual road, are
presented in Annex 8 (Tables A8.3 & A8.4) comprising 463 roads in 63 Districts
with a total length of approximately 5,860km: it should be noted that – by 2016 –
some of the 2006 LRN will have been re-classified as part of the SRN and thus
the net gain in the local road network is more than the implied increase of around
1,400km.

The alternative version of the 2016 LRN includes an additional 13,530km of

planned or proposed local roads, making a future LRN of approaching 19,400km
and a total road network (including the extended SRN) of over 29,000km. The
additional 13,500km comprise 1,800 individual roads in 50 Districts – many in the
Terai, where they do not contribute significantly to improved access. Full listings
by both District and individual road (including the DTMP Code) are given in
Annex 8, Tables A8.5 and A8.6.

C. Calculation of Accessibility of Strategic and Local Road Networks

1. Accessibility for the Year 2006 with SRN and Existing LRN
An assessment of the impact of the Local Road Network on overall levels of
accessibility has been undertaken by adding the known and likely LRN to the
existing and proposed SRN lengths in both 2006 and 2016. The initial (2006) test
assumes an LRN length of 4,460km as illustrated in Figure VIII.1. This is the
basic ‘known and verified’ network of operational roads: as can be seen these
roads are mostly short and close to the existing SRN. Their effect on accessibility
is primarily to expand the accessible band-width adjacent to the SRN, rather than
to improve access to remote areas.

The effect on the overall population served in 2006 is shown in the top two rows
of Table 8..2, which compares the data with the ‘SRN only’
scenario. Overall accessibility is increased from 78 percent to 85 percent, with an
increase from 58 to 68 percent in the hills and from 94 to 98 percent in the Terai.
The existing LRN in the Terai is thus seen to provide almost total coverage.

DHV - SILT - TAEC – CMS Chapter VIII-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 8.2: Additional Populations served by LRN

Length Population served
km Hill Terai Total
Existing / Operational SRN 7,360km 6.52 13.70 20.22
2006 58% 94% 78%
PLUS 2006 LRN +4,460km= 7.67 14.36 22.04
11,820km 68% 98% 85%
Extended SRN 9,930km 7.94 14.21 22.15
2016 70% 97% 86%
PLUS 2016 LRN (Committed) +5,860km= 9.09 14.51 23.61
15,790km 81% 99% 91%
PLUS All known existing & 9.92 14.48 24.40
planned roads ~29,000km
87% 100% 94%
Note: Total Population of 25.9 million is divided into Hills 11.3 million & Terai 14.6 million

2. Accessibility for the Year 2016 with SRN and LRN

In 2016, the initial network tested includes the proposed Extended SRN
(9,930km) plus the 5,680km of existing and currently planned local roads, giving
a total of 15,790km. This network provides 81 percent accessibility for the hill
areas and 99 percent in the Terai and Valleys, producing an overall accessibility
of 91 percent – an increase of 5 percent compared to the extended SRN alone.

Two local networks are tested in 2016: the first (above) assesses the impact of
the known (existing & planned) LRN on the 2016 Extended SRN and the second
examines the impact of all potential local and strategic roads. As can be seen
from Table 8..2, the overall accessibility is increased from 86
percent firstly to 91 percent with the committed LRN and then to 94 percent with
all potential planned roads. Ultimately 100 percent coverage is achieved in the
Terai and 87 percent in the hills.

The “all roads” scenario for 2016 is based on an assembly of all potential road
alignments, including the planned roads in all the available DTMPs, the DoR 20-
Year Master Plan, the indicative alignments (9,000km) for additional SRN links
identified during this Study, and local networks proposed by various development
agencies. The total length of this network is approaching 29,000km, comprising
9,930km of SRN; 5,680km of existing and currently planned local roads; and
about 13,500km of planned or proposed DTMP roads and DoR Master Plan
Roads.

3. Effect on Walk-Times
The comparable data illustrating the total and average walk-times to the 2006
and 2016 networks are given in

Table 8.3, again with the ‘SRN-only’ figures for comparison. The ‘SRN plus all
roads’ scenario provides a significant reduction in the ‘Total Walk Time’ and
‘Average Walk Time’ in both the hills and Terai: the impact in the Terai is
relatively small as the opportunities for improvements are limited. In the hills there
are still around 1.5 million people more than 4 hours walk from a road (13 percent
of the hill population, see Table 8.2) – indicating that it is
increasingly difficult (and certainly uneconomic) to expand the network still further
to encompass the whole country.

DHV - SILT - TAEC – CMS Chapter VIII-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 8.3: Total Walk Time to access Network & Average Walk Time
[Person-Hours (thousands) and Hours]
Year 2006 2016
SRN
7,360km 9,930m
Length
SRN + SRN + SRN +
SRN SRN
2006 2016 ALL
Only Only
LRN LRN Roads
Hills
Total Walk Time p-hours 69,440 56,242 42,137 34,430 27,130
Average Walk Time Hours 5.16 4.03 2.74 2.05 1.40
Terai
Total Walk Time p-hours 19,411 16,020 17,453 15,218 14,605
Average Walk Time Hours 0.33 0.11 0.20 0.05 0.01
Total – All Country
Total Walk Time p-hours 88,881 72,262 59,589 49,648 41,734
Average Walk Time Hours 2.44 1.82 1.30 0.92 0.62

4. Accessibility Coverage
The geographical extent of the areas served by each of the three networks are
illustrated in Figure VIII.2, which also indicates the population distribution. It is
clear that substantial areas of the country are still not-road-served but it is equally
evident that many of these areas are supporting small or negligible populations. It
is clearly impractical to pursue the planning or construction of roads into these
remote northern areas.

However, in 2016, with the extended SRN and the known and committed LRN,
around 2.3 million people in the hills (20 percent) are still beyond a 4 hour walk
from their nearest road. This population is primarily in a limited number of
relatively densely populated areas in the mid-hills – in both the east and west of
the country. A strategy is required to fulfil the accessibility criteria in these
pockets by the concerned local authorities or the agencies working in the rural
road sector.

The ‘un-served’ populated areas for the two networks tested in 2016 are shown
on Figure VIII.3, indicating those areas where additional local road construction
will be required to meet the access needs of the population. It is interesting to
note that, despite the inclusion of an additional 13,500km of road - almost a
doubling of the network – the size and scale of the un-served areas are
remarkably similar between the two options tested in 2016.

It is evident therefore that a focussed plan for additional local road construction
should be considered in the east (Bhojpur, Khotang & Okhaldhunga) and the
west (Palpa, Syangja, Myagdi, Rolpa, Jajarkot and Kalikot). The population
densities in other non-served areas are unlikely to warrant new road construction.

DHV - SILT - TAEC – CMS Chapter VIII-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VIII-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

D. District Level Accessibility

The previous Chapter illustrated the accessibility at a District level provided by

the SRN in both 2006 and 2016. This showed that approximately 4.9 million
people (19 percent of the total population) lived more than 4 hours from the
currently operational SRN and that, by 2016, this number could be reduced to 3.4
million people (or 13 percent overall). The inclusion of the known and committed
local networks in 2006 and 2016 can reduce the un-served populations to 3.6
million (14 percent) and 2.2 million (9 percent) respectively.

The inclusion of the LRN in 2006 reduces the number of districts with more than
100,000 people outside the 4-hour limit from 21 to 14, with two main clusters of
poorly served Districts in the east (Sindhuli, Sankhuwasabha, Khotang, Bhojpur
and Solukhumbu) and the west (Rukum, Jajarkot, Rolpa, Kalikot, Baitadi, Bajura
and Bajhang). By 2016, with the inclusion of the known LRN, there are only two
Districts (Rolpa & Khotang) with over 100,000 people more than 4 hours from a
road.

Figure 7.9 (in the previous Chapter) illustrated the Districts with poor accessibility
from the SRN in 2016, highlighting those Districts where local road access was a
priority. Figure 8.4 presents similar information with the addition of the ‘known’
LRN: it can be seen from that some of the immediate problems (including areas
close to Kathmandu) are resolved, but that further expansion of the LRN is
required in the mid-hill Districts in both the east and west of the country.

A detailed listing of the accessibility indices for the SRN plus known LRN in 2006
and 2016 are presented in Annex 8, Tables A8.7 and A8.8. The inclusion of the
LRN in the analysis has a significant impact on the average walk-time to access
the road network, as shown in Table 8.4, with a 25-30 percent reduction overall,
and an even greater impact within the Terai.

Table 8.4: Average Walk-Time to access the Network (SRN & LRN)

2006 2016
Hills Terai Overall Hills Terai Overall
SRN Only* 5hr 12m 20min 2hr 26m 2hr 45m 12min 1hr 18m
SRN plus LRN** 4 hrs 7min 1hr 50m 2hrs 3min 55min
Notes: * SRN = 7,360km in 2006 and 9,930km in 2016
** SRN & LRN = 11,820km in 2006 and 15,790km in 2016

Currently, with the effective operational SRN (7,360km), the average walk-time to
access the network is 2½ hours: the inclusion of 4,460km of local roads reduces
this to less than 2 hours on average – 4 hours in the hills and 7 minutes in the
Terai. By 2016, with the extended SRN (9,930km) the national average access
time is reduced to 1 hour 20 minutes and the known LRN will reduce this to less
than an hour on average – 2 hours in the hills and 3 minutes in the Terai.

Overall this level of accessibility is considered to be a plausible standard to aim

for in Nepal over the next ten years. Access in the Terai is reduced to a nominal
level and the average in the hills (2 hours) is a reasonable target given the severe
topographical constraints and prevailing population densities in non-road-served
areas.

DHV - SILT - TAEC – CMS Chapter VIII-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VIII-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS Chapter VIII-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

IX. ENVIRONMENTAL AND SOCIAL WORKS (PART I)

Chapter 9
For Part I of the Feasibility Study the Consultants undertook three main activities
in the field of environmental and social aspects associated with rural road
development projects. The principal work consisted of the elaboration of a
framework for social and social management to be used as guidance for all
similar projects in the future. Other activities focused on the institutional
strengthening of those units of the Department of Roads which are responsible
for the incorporation and monitoring of environmental and social safeguards in
road development projects. Closely related to this activity and based on a need
assessment was the preparation of a Training Plan aiming at capacity building
and development of skill among the staff of the respective units of DoR.

A. Environmental and Social Management Framework ESMF

Road construction incurs various environmental and social impacts, including

both beneficial as well as adverse ones. Particularly in a country like Nepal,
which has young and fragile geology, widespread poverty in the mostly remote
rural regions, vulnerable communities, rich natural biodiversity, road construction
can bring significant adversities to the local, regional and national environmental
setting and socio-cultural fabric of the affected population. With foresight,
appropriate research and diligent planning and implementation of mitigation
measures and monitoring, many, if not all, of the adverse impacts can be
minimized or avoided or dealt with positive and constructive ways.
To cover and address these aspects, the Consultants were required to prepare
for DoR a general guidance document that covers environmental and social
safeguard aspects in a broad manner, giving particular attention to the practicality
for implementing mitigation measures that are in full conformity with the GoN
policies and regulations, are technically sound and suitable, take up all
requirements for land acquisition and potential resettlement, and identify
institutional arrangements as well as capacity building needs. The ESMF
document is also meant to combine the requirements of both the national
government and international donors operating in Nepal’s transport sector.
The ESMF has been written whilst keeping a number of stakeholders (planners,
survey teams, consultants, contractors) and stakeholders (line agencies, donors,
CBOs/NGOs, investors) in mind. The document intends to provide technical and
managerial inputs and guidance into the design, implementation and supervision
of countrywide rural road projects. Based on a broad literature review and
lessons learned particularly during the RMDP the ESMF also intends to raise
awareness of those responsible for planning, implementing and monitoring road
development projects with respect to common and specific environmental and
social issues related to such activities, to mitigate potential impacts and promote
opportunities to enhance foreseen benefits.

The Environmental and Social Management Framework document contains the

following sections:

DHV - SILT - TAEC – CMS Chapter 9 -1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 1: Introduction, including background and rationale for providing this

ESMF, linkage to the Sector-wide Strategic Road Network Development
Programme, methodological approach, sources of information and an analysis of
the potential users and stakeholders of this guidance document.

Chapter 2 describes the tools and procedural steps for Environmental & Social
Impact Assessment to address the environmental and social issues for all
project-related activities, and describes stepwise the corresponding management
requirements in the entire project cycle. Emphasis is given to describing the legal
framework and methodological steps for carrying out IEEs/EIAs, including
environmental auditing. It also provides the basic principles (Environmental Code
of Practice, comprehensively detailed as an Annex) for planning, implementing
and monitoring mitigation measures while considering road development
projects.
Chapter 3 reviews the legal framework - policies, regulations and guidelines of
the Government of Nepal, and those of the major donor agencies, that bear
relevance to rural road development projects. This section makes reference to
some 25 acts and the Government’s Tenth Five Year Plan, all pertaining to
various aspects of road development projects. The review refers particularly to
resettlement policies, including those supporting vulnerable communities. The
chapter discusses relevant sectoral policies and guidelines prepared by the DoR,
and makes specific reference to environmental and social policies being
compared with those of international donors, above all, of the World Bank.
Chapter 4 describes the qualitative and quantitative analysis of potential
environmental and social impacts that are commonly encountered in rural road
development projects in Nepal. This section broadly discusses both beneficial
and adverse (direct and indirect) impacts. Case studies are included to
demonstrate country-specific issues and consequences for the biophysical, social
and socio-cultural environment.
Chapter 5 delineates the Public Consultation framework that needs to be applied
while carrying out EAs for rural road development projects. This chapter also
includes the consultative procedures and participatory approach required for
different project implementation phases as outlined in the Environmental
Management Plan and in the Social Action Plan. Public disclosure of information
including employment, gender issues and controlling/ grievance resolution
mechanisms are analysed under the aspect of applicability for the present
Feasibility Study.
Chapter 6 presents a compendium of environmental and social impacts mitigation
measures in rural road development projects, particularly those being part of the
current Feasibility Study. This section provides technical aspects how to
maximize beneficial impacts and how to avoid or minimize of adverse impacts
under the typical setting in the forthcoming road development program. This
chapter identifies practical, feasible, credible and cost effective measures to
offset or to reduce adverse environmental and social impacts to acceptable level,
and ways to enhance positive impacts. It gives guidance for preparation of an
Environmental Management Plan. As applicable, also it also addresses
secondary, induced and cumulative impacts that may be associated with the
forthcoming road construction activities, and the inclusion of a Resettlement
Action Plan as required.
Chapter 7 outlines the Resettlement Policy Framework (RPF) applicable for
typical rural road development projects, focusing on land and property
acquisition, eligibility, compensation and resettlement. It assesses the
compatibility of the core principles of GoN policies with those of the main donor
agencies, identifies gaps, and gives recommendations for addressing these gaps.

DHV - SILT - TAEC – CMS Chapter 9 -2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The chapter addresses a series of current issues relating to the extent of RoW to
be considered and the associated land acquisition procedures, including the
definition of properties, households and individuals.
Chapter 8: The Vulnerable Community Development Framework describes the
requirements and options to develop support programmes for vulnerable
communities and persons, and how to integrate such planning in the design of
rural road projects.
Chapter 9 identifies the institutional arrangement for implementing the ESMF. It
identifies the key players in this process being governmental and non-
governmental agencies, as well as service providers (consultants and
contractors). It assesses the institutional arrangements procedures necessary for
implementation of the overall sector wide environmental and social management
framework under the coordinating role of the Geo-Environment and Social Unit
(GESU) of the Department of Roads.
Chapter 10: highlights various aspects for Institutional Strengthening of DoR.
This chapter identifies gaps and needs related to capacity building within the
DoR, focusing particularly on the Geo-Environment and Social Unit (GESU)
responsible for environmental and social components of road and traffic
management planning and supervision.

In separate annexes the ESMF includes (i) the ToR for this document, (ii) the
Environmental Code of Practice for 20 different sectors, (iii) a glossary of relevant
terms and (iv) a list of references/documents consulted.

B. Institutional Strengthening Recommendations, with Respect to

Environmental and Social Safeguards

Following the mandate set out in the ToR, the Consultants had various sessions
with different sections within the DoR to identify gaps and recommendations
concerning institutional strengthening. Focal aspects were the development of
mechanisms that would strengthen the capacity of the Geo-Environmental and
Social Unit (GESU) being in charge of environmental and social safeguard
aspects pertaining to road building and maintenance activities within the DoR.
Together with the Head of GESU and other officials from DoR the Consultants
developed an institutional strengthening plan which
(i) focuses on GESU’s institutional and organizational capacity situation,
relating to both individual and organizational competence and effectiveness
in addressing environmental and social problems commonly associated with
road development projects;
(ii) proposes a re-organization of the former GEU 19 to widen its capacities for
social safeguard aspects associated with road development projects (e.g.
resettlement, land acquisition and compensation, promotion of special
assistance programmes for vulnerable groups and severely-affected
families);
(iii) provides assistance for carrying out environmental assessment (EA)
studies, and actively participating in the process of EA document review and
amendments;
(iv) provides assistance in assessing and elaborating budget requirements for
the environmental and social mitigation measures (incl. resettlement) in

19
During this Project, the proposed reform has been institutionalised, resulting in the
creation of
a Social Section within the GEU, hence the name of this Unit changed into
‘Geo-Environmental & Social Unit’ GESU

DHV - SILT - TAEC – CMS Chapter 9 -3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

accord with the EA studies and, as applicable, the respective Resettlement

Action Plans;
(v) enables the Unit to coordinate with other line agencies and to
monitor/supervise the environmental and social safeguard and mitigation
measures laid out in the Environmental Management Plans described in the
EA documents;
(vi) promotes mechanisms for incorporating social and environmental safety
requirements in all of Nepal’s road development programmes (e.g. works
contracts specifications, monitoring indicators, compliance verification and
effect monitoring techniques);
(vii) devises a training plan tailored for bridging the identified gaps and
promoting skill development in the said areas.

C. Training Plan for GESU

The proposed training arrangement, being a result of the need identification

process during this FS, resulted in a document entitled ‘Training Plan for
Environmental and Social Safeguard Measures Relating to Road Development
Projects in Nepal’.

Feedbacks and lessons learned from previous and ongoing road projects in
Nepal indicate the stark need for strengthening the capacity of those engineers
and DoR staff who are in charge with procurement of services and supervising
contractors, especially with respect to ensuring safeguard provisions provided in
the respective Environmental and Social Management Plans outlined in the
project documents and, at times, in the respective clauses in the works contracts.
The same holds true for strengthening the skill of GESU staff, DoR planners,
implementers and supervisors of construction works, particularly with regard to
good practices and environmental safeguard considerations.

The objective of the Training Plan is to provide practice-oriented training and

production of easily understandable manuals, tailored to Nepal’s unique
environmental and social settings, and being complimentary to the already
existing guidelines, and being supplementary to other relevant skill development
programs which recently have been proposed to DoR.

The proposed Environmental and Social Safeguard Training Courses primarily

target the Geo-Technical & Environmental Unit (GEU) and selected personnel of
the Department of Roads. It is further envisaged to include in the training
selected service providers like contractors and domestic consulting firms.

The Training Plan proposes four training modules to be carried out as in-house
training courses 20, each involving 8-11 days, with some 20-25 trainees:
Module (1) addresses environmental and social impacts that are likely to occur in
connection with road development projects. The module is structured in four
major segments:
• Identification of environmental / social impacts: causes and effects
• Framework conditions for conducting Environmental Assessment
• Methodological approach for impact analysis
• Preparation and Review of Environmental Impact Documentation

20
with field trips/on-hand training, as applicable

DHV - SILT - TAEC – CMS Chapter 9 -4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Module (2) addresses the need to include, to the maximum extent possible, good
construction practices in all stages of planning, executing and monitoring road
development projects. The training is tailored in accordance with the
Environmental Code of Good Practice (ECoP) established in the ESMF (see
above).
• Training in basic principles of EcoP, including site selection,
management of construction plants and equipment, spoil
management, bio-engineering, erosion control, drainage, habitat
protection, work safety, health regulations, social aspects, vulnerable
groups, support programs and compensatory measures);
• Skill development and awareness raising among road engineers
involved in planning, designing, implementation and supervision of
road construction works;
• Local consulting firms and contractors to become familiar with the
requirements of ECoP;
• Consultants to adopt development of new ideas in designing and
supervision of road projects with reference to ECoP
• Provision of hand-outs and practical manuals.
Module (3) addresses the processes, quality ensuring and supervision of
procuring private sector companies to become involved in a range of services to
be provided for planning and construction works, under competitive procurement
procedures.
• Setting up a robust and transparent contract procurement and
contract management (incl. coordination processes with outsourced
activities);
• Specific environmental and social safeguard considerations to be
incorporated in the specifications and clauses of Work Contracts (incl.
practical examples, lessons learned from similar projects in Nepal,
pitfalls and specific issues);
• Setting up a database to facilitate pre-selection and/or exclusion of
service providers;
• Demonstrate instruments to minimize (costly and time-consuming)
claims by all parties involved in contracts;
• Ensuring good and anti-corruption practices during contract
performance.
Module (4) is a Training for Trainers programme, being developed in close
coordination and consultation with ongoing capacity building activities proposed
under the RMDP. Focal points are:
• update of Staff Training Plans based on need assessments
• revision of external training proposals
• design of training for impact assessment models
• effective sharing of information obtained during training activities
• integrating project-related experiences in training programmes
• training of staff in training management and skills
• preparing budgets for training programs.
For the implantation of the Training Plan, external/international trainers will be
required, recruited in international bidding, with qualifications and records
pertaining to undertake the proposed training modules. The cost estimate for the
proposed training courses amounts to approximately USD 60,000.

DHV - SILT - TAEC – CMS Chapter 9 -5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Chapter 10
X. PRIORITY INVESTMENT PLAN & SECTOR WIDE ROAD PROGRAMME

A. Overall Findings & Conclusions

1. Introduction
This Chapter summarises the key findings of the Sector Wide Road Programme
and Priority Investment Plan (PIP) Study, conducted for the Department of Roads
(DoR). The Study was carried out, with funding from a World Bank/IDA Loan,
over an 18 month period commencing in September 2005. The Consultants’
initial findings were presented in their Draft Final Report in January/February
2007 and were the subject of presentations and discussions prior to the
finalisation of this Final Report.

The Study is primarily concerned with the maintenance and development of the
Strategic Road Network (SRN) which is the main responsibility of the DoR. A key
objective of the Study is the preparation of a 10-Year Priority Investment Plan
(PIP) for the SRN, including recommendations for expansion and improvement.

Nepal is in a period of great change and is potentially poised to enter a period of

substantial and dynamic growth. The problems of the past decade, including the
long-running insurgency and frequent changes in government, have resulted in
economic stagnation and increased polarisation between relatively wealthy urban
dwellers and the rural poor.

The new Government of Nepal (GoN) is embarking on a major programme of

reconstruction and rehabilitation of facilities in the conflict-affected rural areas,
including the upgrading of poorly maintained, damaged or inadequate roads, the
construction of new roads, and the completion of projects deferred due to security
concerns.

Additionally, investment is required in and around the economic and commercial

centres, including Kathmandu, Pokhara and the main Terai towns, to position the
economy to respond to future opportunities for growth and to compensate for
past inadequacies in funding.

It is necessary therefore for the Sector Wide Road Programme to address a

broad range of concerns and priorities, in both the short and longer-term, and in
both the rural and core economic sectors. Accessibility and connectivity are the
two key parameters to describe the development of the road sector: these
encompass issues relating to the inequitable provision of road access in the
remote and generally impoverished rural areas, and the need for improved
linkages – both domestically and internationally – between the major commercial
centres.

DHV - SILT - TAEC – CMS ChapterIX-1

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

2. Regional Structure
The overall structure of the economic linkages and the associated transport
network within Nepal may be conveniently divided into three broad regions –
Central, East and West – as illustrated in Figure 10.1. The East and West are to
a large extent independent with regional centres, at Biratnagar and Nepalgunj
respectively, servicing their hinterlands.

The 'Central' Region contains the heart of the economic activity of the country,
including the main border crossings with India and the only two urban centres of
any consequence outside the Terai. The region effectively has four main nodes –
Kathmandu Valley, Birgunj, Butwal/Bhairahawa and Pokhara – which together
comprise an economic quadrangle with strong inter-linkages: Narayanghat-
Hetauda form a secondary cluster, and Janakpur services a distinct area towards
the east.

Both the Eastern and Western regions have a north-south divide between the
hills and the Terai and both are to an extent marginalised from the economic
mainstream. Secondary service centres in each are also in the Terai (or inner
Terai), with no major settlements in the hills. The East is undoubtedly richer and
more developed than the West and has greater potential.

The West has a clear focus on Nepalgunj, with a series of lesser, satellite centres
at Dhangadhi, Surkhet and Tulsipur/Ghorahi: these, in turn, have distinct sub-
regions associated with them, although all areas in the West also have a direct
relationship with Nepalgunj.

Similarly in the East, Biratnagar (with Itahari) is the main service and distribution
centre, with secondary centres at Birtamod/Chandraghadi, Dharan/Dhankuta and
Lahan.

The Central Region has a different character. Both Kathmandu and Pokhara
have their own spheres of influence, as well as the Central Region inter-
dependence. Butwal-Bhairahawa, Hetauda-Narayanghat (plus Birgunj) and
Janakpur serve the Terai and low hills, with Tansen as a subsidiary centre for the
mid-hills. The linkage of Janakpur with Kathmandu and the Central Region will be
strengthened with the completion of the Dhulikhel-Sindhuli-Bardibas Road.

3. Content of PIP for SRN

The preparation of the DoR 10-Year Priority Investment Plan for the Strategic
Road Network will comprise three main tasks:

• Firstly, a definition of likely maintenance requirements;

• Secondly, the Upgrading of existing and potential SRN elements (incl
some local roads); and
• Thirdly, the Construction of new (strategic) links – for both rural access
and network strengthening

DHV - SILT - TAEC – CMS ChapterIX-2

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS ChapterIX-3

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The focus and content of the PIP is significantly different from the plan prepared
10 years ago. The main component of the previous PIP plan, after consideration
of the maintenance needs, was the extension of the network into the non-road-
served hill areas, including the construction of new roads to connect 12 of the 17
unconnected District Headquarters. This plan has mostly been achieved,
although some of the HQs are only now being connected. It is fair to conclude
that the previous plan was followed – by Government and aid agencies alike –
although progress has been slower than planned due to difficulties in execution
and implementation (rather than funding deficiencies).

The current plan has four different components, in addition to the on-going
maintenance obligations and the completion of the previous plan elements:

• Firstly a significant expansion in the extent of the Strategic Road Network

and the incorporation of substantial lengths of existing local and District
Roads into the SRN;
• The upgrading of existing earth or gravel roads in both the Terai and hills
to all-weather sealed standards, in order to provide an improved level of
accessibility and to promote economic activity and growth;
• The consideration of major new access routes to Kathmandu Valley to
provide a significant time and distance savings between Kathmandu and
the Terai (and India), and to relieve the existing routes; and
• The improvement of key heavily trafficked routes within the Kathmandu
Valley, including the Ring Road and route to Bhaktapur and Dhulikhel.

It is however important to note that, during the first four or five years of the plan,
most of the expenditure will be on projects that are already committed (or are in
the funding ‘pipeline’): it is only in the latter years of the plan – post-2011 – that
new schemes will be included.

The expansion of the network is accomplished by the construction of some new

roads, but more significantly through the ‘conversion’ of existing local or District
Roads and tracks to SRN status. The new roads include extensions into remote
non-road served areas, key missing links especially in the hills, and new strategic
access routes to Kathmandu. However many of the potential new strategic links
merely involve the re-designation of existing operational roads or else follow the
alignment of an existing local road or track. In these cases, the inclusion of a road
in the SRN may involve only a marginal additional costs – ie the additional on-
going maintenance obligation and possible subsequent upgrading to sealed
standards.

A significant programme of upgrading earth and gravel roads to sealed standards

in both the hills and Terai is proposed. Specifically this involves the improvement
of a number of Terai roads (including the Postal Road) to all-weather standards,
together with the provision of the necessary bridges and drainage structures.
These Terai roads improve access for substantial populations in the areas close
to the Indian Border and link the border towns more effectively than the East-
West Highway. In the hills, despite the comparatively small traffic volumes, it is
recommended that the initial ‘trunk’ sections of the north-south road networks are
also upgraded to sealed standards to accommodate existing and potential traffic
flows.

This also marks a change in the previous policy which stressed the extension of
the network to serve all areas of the country: the current emphasis is directed
towards a ‘consolidation’ of the network and the provision of improved standards

DHV - SILT - TAEC – CMS ChapterIX-4

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

and greater reliability on the critical (core) elements of the hill road network.
These sections of road are fundamental to the provision of improved access to
whole regions, as well as providing direct access to more people in the
immediate vicinity.

Reference to the population density distribution map reveals that – away from the
major urban centres – there are significant concentrations of population in both
the Terai and mid-hills areas (see Figure 10.2). In the latter case there is a ‘band’
of population in and around the key ‘hill towns’, with particular clusters in the area
south and south-west of Pokhara, in the Inner Terai Valleys (Dang and Surkhet),
in the hills north of Kathmandu and Dhading, and to a lesser extent in the Far
West and East.

Most importantly, it is evident that there are minimal populations in the areas to
the north of the existing (and proposed/committed) roads: there is thus no case to
be made for further northerly extensions to the network and future network
development should concentrate on an intensification of the network in populated
areas and improvements to the serviceability and reliability of existing roads.

This is a significant change in policy to that of the previous PIP in the mid-1990s,
which placed a higher priority on the provision of basic fair-weather extensions to
the network rather than upgrading to higher standards: this reflects the changed
circumstances now prevailing and the fact that most of the basic initial access
roads have been completed or are under construction.

It is recognised that the provision of improved, reliable, all-weather access not

only reduces operating costs for existing traffic but, more significantly, leads to a
substantial growth in the numbers of passenger trips made and volumes of goods
consumed. The opportunity to make quick, comfortable and affordable trips on a
reliable basis can be seen to transform travel habits and lead to greater economic
activity and social interaction.

B. Strategic Network

1. Extent of Network
The existing ‘designated’ SRN comprises around 5,030km of road – 3,100km of
National Highway and 1,920km Feeder Road. In addition, there are a further
400km of open and operational road (maintained by DoR) that will form part of
any future SRN. The existing network thus comprises 5,430km of road, of which
70 percent (3,800km) has a bituminous surface and the remainder is either gravel
(18 percent) or earth track (12 percent).

It is notable that there are substantial additional lengths of road that are either
currently maintained by DoR (as if they were part of the strategic network) or are
committed improvements that will form part of the SRN by 2010. As described in
Chapters 3 and 4 above, the HDM analysis of the maintenance and upgrading
requirements of the SRN has been based on an assumed network of 7,920km of
road in 2010, including all schemes presently in the ‘pipeline’, which have been
taken as committed. The following discussions on the maintenance budgets are
based on this road length.

In regard to the development and expansion of the SRN, the analysis is based on
a current (2006) network of 7,360km, which includes ALL roads currently being
maintained by DoR; an expansion of 1,030km to 8,390km by 2010; and an
additional 1,540km, mostly proposed by this Study, to create a total of 9,930km

DHV - SILT - TAEC – CMS ChapterIX-5

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

by 2016. This network is illustrated in Figure 10.3, showing the phasing of the
expansion: a full listing of the links is included in Annex 7, Tables A7.1 and A7.2.

The current 2006 network includes an additional 2,330km of non-designated SRN

roads that are currently maintained by the DoR Divisional Offices: included in this
total are 116km of urban roads within Kathmandu, substantial lengths of Terai
roads (including most of the Postal Road), and a number of existing schemes in
the hills built under various aid or development initiatives. The additions by 2010
(1,030km) are mostly ‘committed’ schemes in the hills and the remaining links in
the Terai that are part of the package of Indian-funded improvements.

The remaining additions prior to 2016 (1,540km) comprise mostly the schemes
identified and prioritised within this Study, including a series links within the mid-
hills corridor, connections to the remaining non-road-served District HQs (Humla,
Mugu and Dolpa), new and improved access routes to Kathmandu, and network
improvements within Kathmandu Valley.

These latter improvements – including upgrading of the existing Ring Road,

dualling of the Kathmandu-Bhaktapur-Dhulikel Road, and the possibility of an
Outer Ring Road (ORR) coupled with improvements to radial routes – cannot be
realistically assessed within scope or remit of this strategic study, and will require
detailed specialised evaluations. In particular the ORR should be evaluated in the
context of the overall Urban Development Strategy for Kathmandu Valley, as it
will have a fundamental influence on the patterns of growth and urbanisation in
the Valley – in much the same way as the existing Ring Road did 25 years ago.
The proposed ORR is much more than a ‘road project’ as it will determine land-
use and affect land values over a broad area: it may also lead to the continuing
proliferation of unplanned low-density developments that are inefficient and
expensive to service, and which will destroy the last vestiges of open space and
agricultural production within the Valley.

2. Regular Annual Maintenance

The first “call” on the budget in any year should be for the continuing and on-
going maintenance obligations, referred to here as Regular Annual Maintenance
and including Routine, Recurrent, Specific and Other Maintenance. The costs of
each are estimated on a per km basis across the whole network.

• Routine: including grass cutting, drain clearing, debris removal, signs and
markings, etc;
• Recurrent: minor repairs carried out on a cyclical basis, including pot-hole
repairs, patching, edge-break repair, shoulders, etc;
• Specific: more substantial works as dictated by road condition – eg
gravelling shoulders, building or repairing retaining walls, etc; and
• Other: traffic safety, bio-engineering, emergency maintenance, drainage
rehabilitation, equipment repair and servicing,

The total annual costs of these regular maintenance liabilities over the 10 Year
plan period are summarised in Table 10.1, based on the assumed increases in
the length of the SRN, from 7,360km in 2006 to 9,930km in 2016, plus an
allowance for non-SRN roads and strategic urban roads that are not otherwise
included. The ten-year total is Rs 11.1 billion, increasing from Rs 1.09 billion to
Rs 1.27 billion per year over the period.

DHV - SILT - TAEC – CMS ChapterIX-6

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS ChapterIX-7

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 10.1: Regular annual maintenance (Rs million)

Total Cost
Rate
Type of Maintenance Unit over 10 Percent
(Rs/km)
years
Routine Rs/km/year 30,000 3,104 28%
Recurrent Rs/km/year 35,000 3,621 33%
Specific Rs/km/year 25,000 2,586 23%
Other Rs/km/year 16,900 1,748 16%
Total 11,059

Funding for the Regular Annual Maintenance should unequivocally be available

through the Road Fund, itself funded through a fuel levy, tolls, and vehicle
registration and licence fees. It is noted however that adequate funding is not
currently available from this source.

3. Periodic Maintenance
Periodic maintenance involves the resealing, overlay or rehabilitation of existing
roads on a cyclical or demand responsive basis. The future periodic maintenance
requirements for the SRN in Nepal have been assessed using the Highway
Development & Management Tool (HDM-4) as described in Chapter 4. This
model optimises the maintenance and improvement (upgrading) interventions on
the network, by minimising the life-cycle costs incurred by both the road agency
and road users.

The model has identified an overall economically optimum 10-year investment

programme for the strategic network of 7,900km. The evaluation assumes the
completion of approximately 3,000km of committed schemes (new construction or
upgrading) by 2010 and further assumes that no additional interventions will be
required on many of these (2,450km). In total, HDM-4 recommends interventions
on 4,124km of road over the 10 years at a total cost of Rs15.7 billion, including
almost 3,000km of periodic maintenance – reseals, overlays and rehabilitation –
at a cost of Rs9.5 billion. Around 1,130km of earth and gravel road are proposed
for upgrading to sealed standards, after 2010 (see below).

The results are summarised in Table 10.2 for the full 506 links considered in the
HDM analysis. It can be seen that 110 links (2,450km) are assumed to have
committed improvements by 2010 (and thus no further interventions before 2016)
and 79 links (1,343km) have no intervention programmed.
Table 10.2: Summary of Periodic Maintenance and Upgrading
(Optimised, Balanced & Constrained) – 2007-2016
Number of Cost Average
Length (km)
Links (Rs billion) Cost/km
Upgrade to Seal
Hills 45 871 5,226 6.00
Terai 17 262 1,047 4.00
Total 62 1,133 6,273 5.54
Rehabilitation 95 1,132 5,357 4.73
Overlay 100 955 3,611 3.78
Reseal 60 904 485 0.54
Committed Schemes 110 2,450
No Intervention 79 1,343
Total 506 7,917 15,725

DHV - SILT - TAEC – CMS ChapterIX-8

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS ChapterIX-9

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The approach adopted in the model has been designed to replicate current
practice in Nepal, involving specifically the application of AC overlays only for
roads with higher traffic flows (ie in excess of 3,000 vpd). The implication of this
approach is (and has been) the extensive use of relatively low-cost reseals on
much of the network. Whilst this approach helps conserve the integrity of the
pavement, it only has minimal effect in reducing roughness or in adding strength:
as a result, over time, it will become necessary to undertake significantly more
expensive rehabilitation works.

The proposed 3,000km programme of periodic maintenance described above

contains a substantial proportion of major rehabilitation (1,132km) representing
almost 60 percent of the total periodic expenditure. This is as a consequence of
the past practice of applying only reseals (DBST or SBST) which do little – in the
long term – to enhance the quality or strength of the pavement

The alternative approach is to adopt the initially more expensive AC overlay on

any road with traffic in excess of (say) 1,000 vpd. This has the effect, for a
marginal increase in overall cost, of providing a smoother surface whilst also
adding strength to the pavement, with the result – as demonstrated in Chapter 4
– that vehicle operating costs are reduced and the network at the end of the
analysis period is in a substantially better overall condition.

It is therefore recommended that the current practice be reviewed and a greater

use of AC overlays be introduced. This has a beneficial impact on the overall
efficiency of the management of the road network – although the initial costs of
the works are higher: it may therefore be necessary to retain the involvement of
foreign aid during the transition period whilst the new practices are being
introduced.

[The Consultants are also of the opinion that the costs of AC overlays will reduce
if the volume and certainty of continuity of AC work were to improve, encouraging
contractors to invest in the plant and equipment.]

4. Upgrading to Sealed Standards

The above analysis has also identified 1,130km of earth and gravel road for
upgrading to sealed standards – 870km in the hills and 260km in the Terai. It
should be noted that these upgradings are additional to the currently committed
programmes – including the Indian funded upgrading in the Terai and the current
WB and ADB proposals. Because of the scale and extent of the current road
improvement programmes planned over the next four years, the HDM model was
constrained to allow upgrading to occur only after 2010.

In addition to the optimised analysis described above, which was constrained to

less than 400 links, an unconstrained test was conducted on the full network, to
determine the ‘true’ relative priorities of all schemes (excepting those presently
under construction), including those previously considered as committed.

This fully unconstrained analysis of the full network identified a total of almost
2,600km of existing (or proposed) gravel and earth roads to be upgraded to
sealed standards over the 10-year plan period. The total cost of this upgrading is
estimated at around Rs13.5 billion. The analysis confirms the inclusion of
substantial lengths of ‘committed’ improvements, plus additional lengths in the
hills and in and around the Kathmandu Valley. The location of the projects are
shown on Figure 6.1, which separately identifies the ‘committed’ and additional

DHV - SILT - TAEC – CMS ChapterIX-10

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

schemes. Of the total, 1,600km are in the hills and 980km are in the Terai, as
summarised in Table 10.3.

Table 10.3: Recommended Upgrading to Sealed Standards

Number of Length Cost Average
Links (km) (Rs billion) Cost/km
Hills 71 1,598 9,569 5.99
Terai 50 980 3,908 3.99
Total 121 2,578 13,476 5.23
Note: Based on ‘all-links’ analysis and includes committed schemes

A full prioritised listing of the roads recommended for upgrading is included in

Annex 10 as Table A10.1: the ranking is based on the NPV/C 21 as a measure of
project ‘worth’. Any road with a positive NPV – which is equivalent to an IRR in
excess of 12 percent – is worth constructing: beyond this, the higher the NPV/C,
the better value of the road scheme. A list of those schemes that failed to achieve
a positive NPV is also included (Table A10.2).

The above analysis is a purely technical economic evaluation based on current

traffic volumes. It includes NO allowance for social factors nor for increased
economic activity (and associated increased traffic) induced by the improved
levels of access provided.

It is noted that not all of the roads selected for detailed feasibility study, or
subsequently recommended for upgrading, are included in the prioritised list,
including Gokuleshwar-Darchula, Sanfe-Martadi, Sanfe-Mangalsen-Belkhet, and
final sections of the Karnali Highway to Jumla. These results are however purely
a reflection of the exceedingly low levels of present traffic and an evaluation
procedure that takes no account of generated traffic nor diversion from other
modes. As would be expected these non-feasible sections are the most remote
sections of road, serving relatively small populations: the ‘trunk’ sections of the
hill roads (and all the Terai roads) are all shown to be feasible.

The specific analyses of the roads in the Detailed Feasibility Studies (Part II of
this Study) take into account additionally the potential for significant changes in
trip making (and freight consumption) resulting from the road improvement: as a
consequence all of the roads selected for upgrading have been shown to be
feasible. Details of the IRRs calculated are included, where appropriate, in
Table A10.2.

5. Expansion/Extension
The above analysis is concerned with the maintenance and upgrading of the
network of 7,900km that will exist in 2010. This is the prime concern of the DoR
and should form the “core” of any future budget provision. However the
Consultants have also examined the potential for expanding the SRN beyond this,
and for the inclusion of additional roads in the network: these additional roads
include both ‘new’ construction and the inclusion (and re-classification) of existing
local roads.

In practice, the ‘existing’ 2006 network as maintained by DoR has been defined
to comprise 7,360km of road (see para 10.2.1.above) and the 2010 network
additionally includes a further 1,030km of committed roads, producing a total of
8,390km. The difference in the network lengths compared with the HDM analysis
results from the inclusion of additional roads in the Terai and within Kathmandu
21
The ratio of Net Present Value (= Present Value of benefits minus PV costs) to Capital Cost

DHV - SILT - TAEC – CMS ChapterIX-11

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Valley (116km) that were not included in the HDM analysis due to a lack of
available data.

It is recommended that an additional 1,540km of road be considered for inclusion

in the SRN by 2016, making a total network of 9,930km. These additions include
both the re-designation of existing local roads and new construction. The addition
of local (District) roads does not necessarily involve any direct cost, although the
overall maintenance liability will be increased and the roads will need to be
assessed for upgrading. In addition to current commitments and proposals, it is
estimated that a further 700km of roads (500km in the hills and 200km in the
Terai) will require upgrading to sealed standards towards the end of the 10-year
plan period.

6. New Construction
Only limited amounts of truly new construction are proposed in the forthcoming
10-year period, beyond that which is presently committed. Over the past decade
considerable lengths of new road have been built (or planned) both at the District
level and as extensions of the SRN to District HQs. The emphasis of the plans
now is on the consolidation and upgrading of these roads to provide reliable and
sustainable access. This may involve the transfer (and upgrading) of locally built
roads and tracks to the SRN, placing the responsibility for on-going maintenance
at the central (DoR) level.

The exceptions to this that involve new construction are the links to the remaining
three District HQs for which no “commitments” presently exist (Humla, Mugu and
Dolpa – see below), ‘missing’ sections of the mid-hills corridor and other mid-hills
links, and new access routes to Kathmandu. The provision of road access to all
District HQs has been a long-standing priority of successive Five-Year Plans,
although it has been recognised, from the previous PIP (1997) onwards, that
these linkages are unlikely to be economically justified as the roads serve
insufficient populations.

(a) Remote Area Access

Following the completion of present commitments (and schemes in the ‘pipeline’),

three District HQs will remain unconnected: Humla, Mugu and Dolpa. The
economic benefits of constructing these roads are low – see Chapter 6 – and it is
recommended that initial access tracks be developed through local District-based
initiatives, with DoR involvement for upgrading considered only at a later date.

It is proposed to provide motorable access to Simikot (Humla District HQ) from

China via Hilsa. An 88km track is presently under construction and, when
complete, this will enable basic goods and foodstuffs to be provided from the
Chinese side, where a motorable road exists. The provision of a road access
from within Nepal would involve a lengthy route and very substantial construction
costs, and cannot be justified once a link via China has been established. It is
recommended that access within Nepal continues to rely on air services from
both Surkhet and Jumla.

The feasibility of a 88km access road from Nagma on the Karnali Highway to
Gamgadhi (Mugu District HQ) was considered in Part II of this Study. The road
was not shown to be economically feasible (IRR = 7.6%), even with the higher
levels of trip making and goods consumption, on account of the low population
density and extremely small numbers of people served. The area however has
considerable tourist potential and if this were to be developed, the demand for

DHV - SILT - TAEC – CMS ChapterIX-12

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

access and support services could help justify the road – even if the main tourist
access remained by air.

It is however recommended that the road be initially constructed through local

efforts and initiatives, with funding through DRILP 22 : subsequent upgrading
should be considered only post-2016. At this stage however it is important to
ensure that the alignments (horizontal and vertical) adopted are suitable to allow
subsequent upgrading to Feeder Road Standards.

A local road is under construction to connect Dunai (Dolpa District HQ) with
Jajarkot using local initiatives and District-level funding: it is probable that this will
take many years to complete. Construction of this road was shown (Table 6.2) to
have a very low rate of return (IRR = 8.5%) and it is recommended that the
present locally-based approach be continued and that upgrading be considered
only at a later stage.

(b) Mid-Hills Corridor and links

A series of additional links in the mid-hills totalling around 650km are proposed,
which would connect between District HQs and link up existing sections of the
potential east-west corridor. A number of these were evaluated in Chapter 6
(Table 6.2) and most would appear to produce a healthy economic rate of return,
due primarily to the relatively dense populations in the areas served and the
connectivity provided between the centres in the hills. Some other schemes,
which were not evaluated as they scored low on the original ranking, have been
excluded. A list of the proposed additional links is included in Annex 10, Table
A10.3, which shows also the IRRs, where calculated. For roads not specifically
assessed an estimate (high / medium / low) of the economic return is shown,
based on the population served.

(c) Access to Kathmandu

Improved access to Kathmandu has been identified as the single most effective
and significant potential project in the transport sector during the plan period. It
offers the opportunity of reducing the travel time between Kathmandu and the
Terai by between 4 and 5 hours, with a distance saving of around 150km. Such a
reduction would revolutionise the current trading and travel patterns between
Kathmandu and the Terai and would provide very substantial benefits for the
significant volumes of mostly commercial traffic.

There are two groups of options – from the west linking from the Prithvi Highway
to Kathmandu and providing relief to the Naubise-Thankot section; and more
extensive schemes linking Kathmandu to the Terai either directly via a series of
tunnels or else via a longer route following the Bagmati River corridor. Either
alignment would be expensive – estimated at US$40 million for the western
approach and US$240 million for the Fast Track (tunnel) option – although each
would likely be economically viable due to the substantial time and vehicle
operating cost savings available.

As both groups of options are addressing the same central issue, construction of
both is mutually exclusive. The shorter cheaper western approach routes resolve
the immediate capacity and strategic issues associated with the single steep
approach route to Kathmandu, but these do not offer the same scale of benefits
that are associated with the direct link to the Terai and the 150km distance saving.
Initial construction of the western approach would detract from the viability of the
22
DRILP = Decentralised Rural Infrastructure and Livelihoods Programme; ADB / SDC

DHV - SILT - TAEC – CMS ChapterIX-13

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

direct Fast Track route – and would itself not be feasible if it were assumed that
the Fast Track were built within the next 15-20 years.

The issues involved in the selection of the preferred option to improve access to
Kathmandu are complex and inter-related and, without doubt, any one of a
number of schemes would be viable. No clear decision can be made without a
detailed combined study of all the options, although it may be safely concluded
that it would be counter-productive to pursue more than one option.

One of the key issues is the scale of the investment required: it is probable that
the “best” option – in the long term – will be the most expensive option, but if
adequate finance is not forthcoming then it may be preferable to pursue a sub-
optimal solution.

The size and nature of the project would make it an ideal candidate for private
sector involvement through some form partnership or BOT-style financing, with
the potential of cost recovery through tolls. The following analysis assumes that
the Fast Track forms part of the future DoR budget, whereas in practice funding
may come from elsewhere.

7. Network Strengthening and Improvement

In addition to the upgrading from gravel to sealed surface and the proposals for
extensions for the SRN, a limited number of capacity deficiencies and potential
improvements have been identified, mostly within the Kathmandu Valley. For
those within the urban area, it has not been possible to conduct a complete
evaluation, although international experience would indicate that very substantial
benefits will accrue from these schemes – if they can be implemented with no or
minimal land acquisition and resettlement

The urban schemes require specialised evaluation, taking account of the specific
traffic mix, road and junction capacity, traffic demand and growth, the needs of
public transport, and road safety. The other proposed projects will also require
detailed individual feasibility studies to take account of their specific features. The
schemes recommended for further detailed study are:

• Kathmandu-Bhaktapur-Dhulikhel: initially the Koteshwar-Suryabiniyak

section to be upgraded to dual 2-lane highway with frontage roads for
local traffic and access; improvements to junctions and facilities for public
transport; second phase widening to 4-lanes to Banepa and Dhulikhel.
• Kathmandu Ring Road: upgrade and improve to dual 2-lane urban
standards with frontage roads (as required) to provide for local access,
parking, etc; preparation of detailed traffic management plan to regulate
and control public transport services, truck movements and parking;
evaluation of alternative improvement proposals for key junctions to
include traffic management measures, local widening and at-grade
channelisation improvements, as well as possible grade-separated
solutions.
• Kathmandu Outer Ring Road and construction/upgrading of additional
radial routes: a plan for a comprehensive road network to serve the outer
areas within the Kathmandu Valley should be prepared in conjunction
with an overall urban development strategy for the Valley to ensure the
coordinated provision of services and a rational and efficient distribution
of land-uses. The future strategy and plan for the whole Kathmandu
Valley should be developed with the overall national context and policy
structure regarding decentralisation and balanced regional growth, and to
ensure environmental sustainability and the protection of agricultural land.

DHV - SILT - TAEC – CMS ChapterIX-14

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

An Outer Ring Road should NOT be developed in isolation nor as a

means to stimulate land development in the outer areas.
• Koshi Bridge and replacement of Barrage: a detailed investigation of the
replacement of the existing (barrage) crossing of the Sapta Koshi in
Eastern Nepal; two broad alternatives are available: a bridge-barrage-
high dam at Chatara where the river leaves the hills, or a lengthy multi-
span bridge on the line of the East-West Highway, approximately 5km
north of the existing barrage.
• Hetauda By-Pass: consideration should be given to the options for a by-
pass of Hetauda, including the potential impact of the construction of the
Fast Track.

8. North-South Corridors
The previous WB/IDA funded N-S Corridors Study (2004) examined a total of
nine potential crossing points to China in the context of developing north-south
transit routes between India and China. The Study concluded that, although the
volume of trade was currently small, there was significant potential for expansion.
The Study specifically recommended the development of a second border
crossing at Rasuwagadhi to supplement the existing road crossing at Kodari,
which should also be retained and expanded. There are however constraints at
Kodari – both in the approach routes and land available at the border – which
constrain the future capacity. There are commitments on both the Nepali and
Chinese sides to the development of Rasuwagadhi: the Chinese have completed
their access road to the border and are currently funding the development of an
18km link within Nepal, to link with the existing road network north of Dhunche.
Within Nepal, the ADB Connectivity Project will upgrade the access to Dhunche
and complete a missing link to connect to the Prithvi Highway west of
Kathmandu.

It is expected that the Rasuwagadhi crossing could be operational by 2010. In

conjunction with the new and upgraded approaches from Nepal, it is possible
also to develop transit facilities and/or an export/import processing zone on the
ample available flat land near Devighat (south of Trishuli).

The N-S Corridor Study strongly recommended that investment in north-south

potential transit routes be focussed on the development and upgrading of these
TWO locations only. To be effective and to attract potential transit traffic, the
routes need to be developed to a high standard (eventually capable of taking 40ft
container traffic) and investment in more than two routes would be counter-
productive: it is better to concentrate the available resources on the provision of
two routes, until such time as the level of demand is assessed and the available
capacity becomes an issue.

Additionally, the N-S Corridor Study considered the possibility of further cross-
border connections in the west and recommended that two potential locations be
considered for LOCAL movements: at LoManthang in Upper Mustang and at
Hilsa, to provide access to Simikot the District HQ of Humla. These crossing
locations are currently being developed for local traffic and road access is
available from China to otherwise non-road-connected areas in Nepal. In the
longer term, these crossings might from part of a through route, but the distances
and standards of road involved make their use for transit traffic improbable.

In addition to transit trade, it is considered that bi-lateral trade from Nepal to both
India and China is equally important, as this leads directly to domestic economic
benefits. Such bi-lateral trade with China will obviously benefit from investments

DHV - SILT - TAEC – CMS ChapterIX-15

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

in north-south transits routes, which should be designed to provide facilities to

handle domestic import and export traffic.

9. Bridges on the Strategic Network

The DoR Bridge Unit is responsible for the construction and maintenance
of bridges on the SRN: a separate budget is provided on an annual basis to
cover both recurrent and development (capital) expenditure. The Bridge
Unit maintain their own inventory of bridges and this forms an adequate
basis to establish annual maintenance requirements, together with a
refurbishment or replacement programme as required.

The Recent (2005) Bridge Inventory Study and Condition Survey identified
around 100 bridges in need of significant rehabilitation and repair. A specific
budgetary item should be included to provide for on-going annual rehabilitation.
Separate allowance should be made for the construction and maintenance of
bridges (as requested) on the local and District road networks.

In general adequate bridges exist on most of the existing designated network of

15 National Highways and 51 Feeder Roads. Earlier sub-standard bridges on
some of the more heavily trafficked sections of the core network have been
replaced over the past 10-15 years and – with a few notable exceptions where
bridges have failed in a dramatic fashion – most of the existing bridges are in
adequate condition, although minor periodic maintenance may be required to
repair parapets, erect warning signs, etc.

The recent Bridge Inventory and Condition Survey identified approximately 100
bridges (10 percent of the total) as requiring substantive repair or rehabilitation –
ie more than routine maintenance. These works should be programmed and
prioritised and an on-going inspection regime established. An annual budget
allocation of Rs1 billion is proposed to cover this item.

As noted in the Survey, the bridges on the ‘Russian’ section of the East-West
Highway between Pathlaiya and Dhalkebar which are of substandard width (5.5m
between kerbs) and do not allow commercial vehicles to pass – except at a crawl.
Due to the length of some of these bridges, this creates severe difficulties and
dangers to traffic, to pedestrians and to slow-moving vehicles. Given the ever
increasing traffic volumes – and thus conflicts between opposing vehicles – the
situation will deteriorate rapidly and a solution will be required.

For the lower volume hill roads, bridges are generally constructed in conjunction
with upgrading and improvement projects in accordance with the overall design
requirements to achieve an all-weather connection. It is suggested that a set of
guidelines be developed to assist with the selection of the appropriate form of
crossing (ford, culvert, causeway, floodway, submersible bridge, RCC bridge,
steel truss bridge, etc) in conjunction with the forecast dry-weather and flood
conditions and anticipated traffic volumes. It is observed that in some cases
bridges on low-volume roads have been over-designed and permanent structures
provided where simpler causeways might have been appropriate.

The proposed (Indian-funded) programme of improvements to Terai roads will

require the provision of a substantial number of bridges (or causeways) across
major rivers and water-courses. This applies particularly to the east-west Postal
Road, which runs parallel to the Indian Border and crosses at least as many
rivers as the East-West Highway: to be effective, the improvement package

DHV - SILT - TAEC – CMS ChapterIX-16

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

should include the bridges or causeways necessary to provide all-season

serviceability.

On average, existing roads in the Terai (EWH and Feeder Roads) have over 4
bridges in each 10km of length (1 bridge every 2.4km), with an average length of
40m, excluding major bridges. It can this be deduced that the ‘Indian’
programme, involving the upgrading of 1,500km of road, will require over 600
bridges. Assuming that half of these may exist (or that their construction may be
deferred), the requirement for new bridges could amount to Rs6.6 billion
(approaching US$100 million), which is similar in scale to the estimated road
improvement cost.

The proposed development and expansion of the SRN will require a number of
additional major bridges across some of the more important rivers in the country.
The proposed extensions to the SRN will require at least sixteen major bridges,
as indicated, across the following eleven rivers:

• Arun – Leguwaghat (Hile-Bhojpur)

• Tamur – Phidim-Myaglung
• DudhKoshi – Hilepani-Diktel
• SunKoshi – Gaighat-Diktel; Katari-Okhaldhunga; Khurkot-Mantali
• Kaligandaki – Kusma; Beni; Beni-Jomsom (2)
• Bheri – Chhinchu-Jajarkot
• Sharada – Botechaur-Tulsipur
• Rapti – Nepalgunj-Bagauda
• Karnali – Belkhet-Rakam
• Tila – Surkhet-Jumla
• Sinja – Surkhet-Jumla

Elsewhere bridges should be constructed in conjunction with the upgrading of

roads in accordance with the proposed guidelines: it should be noted that bridges
are not an essential pre-requisite for an all-weather road – causeways, vented
causeways or floodable/fordable bridges are an acceptable solution in many
locations, where the periods of closure may be limited to a few hours or even
days following heavy rain during the monsoon.

Beyond the present commitments, approximately 890km of new road are

proposed which will likely require around 200 additional bridges at a cost of Rs4.5
billion (US$60 million).

It is noted also that the DoR constructs bridges – on request from the Ministries
or DDCs – on the local road networks which are not specifically the responsibility
of the Department. In many ways, this is a perfectly legitimate function for the
DoR to perform, as the Department has the necessary skills and expertise to
design and construct (or supervise construction) of these bridges. However, the
funding and prioritisation of these bridges should be strictly the responsibility of
the relevant DDC (or Ministry), unless the structure is on a road alignment that
will eventually form part of the SRN.

C. 10-Year Priority Investment Plan

The above sections have detailed the anticipated maintenance and development
work proposed on the SRN over the next 10-year plan period to 2016. This has
included regular and periodic maintenance; upgrading and surfacing of unsealed
roads; expansion and extension of the network (through both new construction

DHV - SILT - TAEC – CMS ChapterIX-17

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

and the re-designation of existing roads and tracks); and network strengthening
and improvement.

1. Proposed Budget (2007-2016)

This section summarises and combines the proposed expenditures on each of
these items to produce a consolidated budget estimate for road works for the
next 10 years. The results are summarised in Table 10.4, showing the indicative
levels of expenditure under the main budget headings for each year from 2007 to
2016: a more detailed breakdown is given in Annex 10, Table A10.4. [Please
note that these estimates excludes any establishment or overhead costs.]

The total expenditure over the 10 years is estimated at Rs120 billion (US$1.7
billion), with annual budgets increasing from around Rs8.5 billion today to Rs13
billion in 2016. All costs are presented in current (2006) values.

Table 10.4: Summary of Draft DoR Budget (2007-2016) Rs million

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total

SRN Length (km) 7,360 7,781 8,202 8,390 8,647 8,903 9,160 9,417 9,673 9,930

Length maintained by
8,596 9,099 9,605 9,859 10,192 10,528 10,869 11,215 11,566 11,923
DoR (km)
Regular Annual
919 973 1,027 1,054 1,089 1,125 1,162 1,199 1,236 1,275 11,059
Maintenance
Periodic Maintenance 2,331 2,416 2,171 2,228 1,364 1,435 2,171 2,246 1,850 1,929 20,141

Improvement & Works:

Upgrade 2,068 3,027 3,243 4,324 5,147 5,193 3,017 3,070 3,867 3,258 36,215
New Construction 2,611 2,938 2,938 2,588 2,964 1,833 2,693 2,693 2,693 2,693 26,644

Kathmandu Access 0 0 1,000 1,000 1,000 3,000 3,000 3,000 3,000 3,000 18,000
Sub-Total (All
4,679 5,965 7,180 7,911 9,111 10,027 8,710 8,763 9,560 8,952 80,859
Improvements)
Design & Superv’n, etc 474 582 685 749 850 930 831 842 913 871 7,727

Total by Year 8,403 9,936 11,063 11,942 12,415 13,516 12,874 13,050 13,559 13,027 119,785

The overall maintenance expenditure is estimated at Rs31.2 billion, representing

28 percent of the total 23 . The Regular Annual Maintenance (Rs11.1 billion) is
calculated on a per km basis for a network length increasing from 7,360km in
2007 to 9,930km in 2016, plus a continuing allowance for non-SRN roads
maintained by DoR. The Periodic Maintenance estimate of Rs20 billion is based
on the HDM-4 output, referring to a network of 9,700km, plus an allowance for
the additional road length and on-going bridge maintenance.

The largest portion of the total budget (Rs36.2 billion or 32 percent of the total) is
allocated to the upgrading of earth and gravel roads to all-weather sealed
standards. This figure includes, in the early years, a significant proportion of
‘committed’ expenditure, including projects funded by ADB, World Bank and the
Indian Government (in the Terai).

The estimated expenditure on new construction (Rs26.6 billion or 24 percent of

the total) includes committed projects of the ADB, World Bank and JICA plus, in

23
Percentages are of total ‘works’ costs (Rs112 billion), ie excluding Design, Supervision & TA

DHV - SILT - TAEC – CMS ChapterIX-18

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

the later years, a programme of new links in the mid-hills and into remote areas
as proposed in this Study.

The construction of a new access to Kathmandu would potentially dominate the

overall expenditure in the latter years, estimated at Rs18 billion (US$240 million),
representing 16 percent of the ten year budget.

The overall composition of the budget is illustrated in Figure 10.4 below: the left-
hand graph indicates the proportion of the ‘works’ budget allocated to each sub-
heading and the right-hand graph suggests the source of funds. Specifically, the
bottom ‘slab’ of expenditure in right-hand graph shows the funds that should be
raised by the Roads Board, which remains relatively constant at around Rs3.2
billion per year, through to 2016. The second slab (in green) illustrates the funds
that are already committed (or will shortly be committed) by the various foreign
aid agencies: this can be seen to taper off from 2011 onwards, as current
commitments expire.

DoR Annual Budget Requirement DoR Annual Budget Requirement

by Type (2007-2016) by Source (2007-2016)

14,000
14,000

12,000
12,000 Kathmandu
Kat hmandu Access
Access
Annual Budget (Rs million)

10,000
Annual Budget (Rs million)

10,000

New 8,000
8,000
Const ruct ion Potential PIP &
Upgrading

6,000
6,000
Committed
Upgrade Projects

4,000 4,000

Periodic
2,000 Maint enance 2,000 HDM - Annual &
Regular Periodic
Maintenance
Annual
0 Maint enance 0
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Regul ar Annual Mai ntenance Per i odi c Mai ntenance Upgr ade HDM - Annual & Periodic Maintenance Committed Projects
New Constr ucti on Kathmandu Access Potential PIP & Upgrading Kathmandu Access

Figure 10.4: DoR Annual Budget Requirements (2007-2016)

(excludes Design, Supervision & TA)

The third slab of expenditure (in orange) represents the projects identified in this
PIP and is synonymous with the potential ‘funding gap’ for additional (new)
projects. The final block of expenditure (in red) indicates the potential scale of the
funding required for the Kathmandu-Terai Fast Track: it is expected that private
sector – or public-private partnership (PPP) – funding would be sought for this
project.

The total distribution of funds by type and source are given in Figure 10.5 below,
which illustrates the scale of the overall Fast Track project (in red) and the
dominance of the expenditure on upgrading (yellow). It can also be seen that the
Roads Board should be funding around 30 percent of the overall ‘works’ budget,
although it is recognised that the Road Fund is not yet fully operational. As a
result and, in part due to the recommendation to adopt more widespread use of

DHV - SILT - TAEC – CMS ChapterIX-19

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

the initially more expensive AC overlays, it may be necessary for DoR to continue
to use aid money to support their maintenance activities.

Summary of DoR Budget (2007-2016) Summary of DoR Budget (2007-2016)

(Rs million) (Rs million)
Regular HDM - Annual &
Kathmandu
Maintenance, Kathmandu Periodic
Access,
11,059, 10% Access, Maintenance,
18,000, 16% Periodic 18,000, 16% 31,200, 28%
Maintenance,
20,141, 18%

New Constr, Potential PIP &

Upgrade, Committed
26,644, 24% Upgrading,
36,215, 32% Projects,
36,958, 33%
25,901, 23%

Figure 10.5: Distribution of Overall Budget by Type & Funding Source

2. Composition of the PIP

Within the above categories – and in addition to the maintenance expenditure –
there are six main components of the proposed 10-year Plan for the SRN:

• Terai Roads Upgrading – mostly with Indian Government funding

• Access to Rural Areas – but more distant links have low IRRs
• Mid-Hills Links – good returns, unless serving only small population
• Strategic Network Improvements – eg Fast Track, Bhimdhunga
• Kathmandu Valley – high priority – need specialised evaluations
• Bridge Construction, Repair and Rehabilitation

(a) Terai Roads Upgrading

The overall proposals contain a substantial expansion of the SRN in the Terai
and specifically the area between the East-West Highway and the Indian Border.
The Indian Government funded improvement programme proposes the upgrading
of 1,450km of roads to all-weather sealed standards, all of which are assumed to
from part of the SRN by 2016. In total 1,500km of road (including 500km of the
Postal Road) will be added to the SRN. The HDM (all-roads) analysis of the
existing DoR road network recommended the upgrading of 980km of Terai roads
at a cost of Rs3.9 billion: the Indian programme proposes 1,450km at a cost of
Rs8 billion. This is assumed to be a “committed” element in the future roads plan
and has not been considered or prioritized further.

The accessibility analysis has confirmed the benefits of road improvements and
upgrading in areas with high population densities. Basically this includes the
whole of the central and eastern Terai (from Birgunj eastwards), the areas around
Bhairahawa and Lumpini (Rupandehi and Kapilbastu), and to a lesser extent to
the west of Nepalgunj. The Postal Road should be completed – including bridges
or causeways across major water-courses – over these sections, together with
additional and improved north-south linkages.

(b) Access to Rural Areas

The remote area access schemes involve the continuation of the programmes
initiated during the 9th and 10th Plans to provide road access to the non-road-

DHV - SILT - TAEC – CMS ChapterIX-20

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

served District Headquarters: much of the expenditure now proposed is for the
upgrading of the initial earth or gravel tracks that have been opened over the past
10 to 15 years. It is recognised that there are significant additional benefits to be
achieved through the provision of regular and reliable access, which are manifest
by more frequent trip-making and increased overall consumption of goods.

Notwithstanding this, the current and forecast traffic volumes on these roads are
low in absolute terms and the scale of the vehicle operating cost savings alone
would not be enough to justify the construction or upgrading. However there are
additional economic and social benefits, brought about by the release of the
currently suppressed demand for movement which appears as ‘generated’ traffic
following the completion of the improvement works.

The main trunk routes into the hills, which serve very substantial populations
(upwards of 600,000 people in many cases), are prime candidates for upgrading
as the benefits of improved levels of access are widely distributed. However the
improvement of the more distant sections of road – which serve significantly
lower numbers of people – cannot be justified so easily. There is thus a limit as to
how far (literally) the Government should go in building roads into the remote
areas: it is clear that there are diminishing returns from building roads to remote
settlements with populations numbered in hundreds not thousands.

(c) Mid Hills Links

The construction of additional roads in the mid-hills – which are relatively densely
settled – is however shown to be economically viable and a network of additional
links between Districts and specifically District Headquarters are recommended
for improvement to sealed standards: most of these proposed links for inclusion
in the SRN follow the alignment of existing local-level District Roads.

It is thus possible to create a Mid-Hills East-West Corridor (as highlighted in

yellow on Figure 10.3) that serves or passes close to 23 District HQs and
connects with the Indian Border and Darjeeling in the east, and Jhulalghat in the
west. It is not intended that this route would (in any way) serve as a long-distance
east-west highway – that function is taken by the existing East-West Highway in
the Terai – but it would serve as a link between adjacent Districts and an
alternate route to the main centres in the hills, including Kathmandu and Pokhara.

The Mid-Hills Corridor is estimated to be almost 1,700km in length, of which over

1,000km currently forms part of the existing SRN. Most of the remaining length
comprises either existing local roads or new links under construction: the only
section not currently planned is the 45km section along the Sun Koshi between
Ghurmi and Khurkot in the east.

(d) Strategic Network Improvements

The construction of the Fast Track between Kathmandu and the Terai would
undoubtedly be the single largest project ever undertaken in the transport sector
in Nepal and potentially the most significant in terms of reducing overall transport
costs and influencing the growth and development of the country. The project is
remarkable in its scale – involving the construction of 70km of new high standard
road with two tunnels of approximately 3.5km each – and its potential impact on
the transport network in the country.

For the full benefits of the project to be realised, it is recommended that the
design standards and alignment are not compromised. The intent is to create a

DHV - SILT - TAEC – CMS ChapterIX-21

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

high standard route with the shortest practical distance between Kathmandu and
the Terai: for the route to be effective in achieving its objectives it must be
possible for all traffic to travel at speeds of at least 50kph (preferably 80kph). This
will allow Hetauda to be reached within an hour and Birgunj within two hours from
Kathmandu.

Due to the scale and importance of the project, it is clearly essential that the best
alignment is selected as it will remain a key element in the road network for
generations to come. It is therefore important that a full and detailed assessment
of the alternatives is undertaken before a final decision is taken and construction
committed. The scale and cost of the project also requires that an innovative
approach to funding be adopted, including joint public and private partnerships,
and the key multi-lateral donors.

(e) Kathmandu Valley

Kathmandu is the dominant urban centre in the country and a major contributor to
overall economic activity. The urban area has grown dramatically over the past
decade and the transport conditions have deteriorated significantly, as a result of
increased population pressure and a very substantial growth in vehicle numbers,
including an inexorable increase in the numbers of motorcycles.

The dilemma posed by the unrelenting growth of Kathmandu is in contrast to the

oft quoted vision for Valley, which aims to:

• limit future growth by decentralising activities to alternative locations (by

promoting secondary towns);
• preserve the last vestiges of agricultural land from urbanisation;
• reduce pollution by the introduction of strict environmental controls; and
• promote the Valley as a centre for education, central level administration
and tourism.

Kathmandu is now beginning to experience serious traffic congestion, leading to

delays at key locations throughout much of the day. The congestion is caused not
only by the increases in vehicle numbers but, more significantly, by poor traffic
management and driver discipline. There is a critical need to address these
problems and to produce a coordinated package of improvement measures,
including some physical measures to address capacity deficiencies.

Within Kathmandu Valley, two specific strategic improvement projects have been
identified: the widening and upgrading of the Kathmandu-Bhaktapur-Dhulikhel
corridor and a package of improvements to the existing Ring Road.

The corridor east to Bhaktapur and beyond to Banepa and Dhulikhel serves the
only available axis for significant development and growth within, or around, the
Valley. It can thus be expected that traffic growth in this corridor will be sustained
and current travel conditions will deteriorate unless significant improvements are
made. It is recommended initially that the first 10km to Bhaktapur be widened and
improved to provide a four-lane divided highway, with frontage roads to service
adjacent development and local traffic.

Similarly the Ring Road should be widened to provide dual 2-lanes throughout,
together with frontage roads and specific facilities for public transport (buses,
micro-buses and tempos): particular attention should be paid to the design and
management of the junctions, which are the current main problem areas.

DHV - SILT - TAEC – CMS ChapterIX-22

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

(f) Bridge construction, repair and rehabilitation

A separate budget provision for bridge construction and rehabilitation is required,

as the regular maintenance procedures do not prioritise or make allowance for
expenditure on bridges. The recent inventory and condition survey has identified
at least 100 bridges in need of remedial works and a comprehensive programme
should be developed in response to these needs.

It is common DoR practice for bridge construction to be pursued independently

from road construction or upgrading, except where both are funded by donors. In
conjunction with the extensive upgrading works proposed over the next ten years
(including the re-designation of substantial lengths of District Road, especially in
the Terai), it is expected that there will be a substantial demand for new bridge
works. An allowance has been incorporated in the budget estimates, based on
the frequency and length of bridges on the existing network.

D. Programme Details – Priority Schemes

A comprehensive list of Priority Road Schemes – including both upgrading and
new construction – has been developed based on the analysis, priorities and
evaluations presented above. The first step in the process is the identification of
all currently committed schemes for which funding is assured, followed by a
prioritized list of other projects ranked on the basis of their preliminary economic
– and to an extent social – evaluation. The full list is presented in Table 10.5 and
the components are described below.

A key element in the overall programme is the construction or upgrading of links

to District Headquarters. In total access improvements (either new links or
upgrading) are proposed to 24 District HQs:

• the Committed Projects provide improved access to six Districts (3 new

connections and 3 upgrades);
• upgrades to sealed standards are proposed for nine Districts;
• new connections to Feeder Road standards are proposed for five Districts;
and
• new connections as simple earth tracks are recommended for four
Districts.

DHV - SILT - TAEC – CMS ChapterIX-23

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 10.5: Priority Road Schemes for Upgrading and/or New Construction
Scheme Length Traffic Cost NPV/C IRR FS Comments and/or Committed Funding Source
(km) (ADT) Rs (m)
A. COMMITTED SCHEMES
Basantapur-Mude-Chainpur-Khandbari 96 1,920 ADB – RNDP - Construct access to District HQ
Galchhi-Devighat-Syabrubesi 79 760 ADB – Connectivity - Upgrade access to Border Crossing
Khurkot-Manthali-Tamakosi 64 660 ADB – Connectivity
Phidim-Taplejung 87 704 ADB – Connectivity - Upgrade access to District HQ
Pokhariya-Parawanipur (Birgunj ICD) 10 320 ADB – SRTFP
Bhumahi-Parasi-Bhairahawa ICD 30 928 ADB – SRTFP
Baitadi-Satbanjh-Gokuleshwar 74 80 518 0.7 18% * WB/IDA – RSDP Ph1
Khodpe-Jhota-Chainpur 110 120 770 0.2 16% * WB/IDA – RSDP Ph2
Bangesimal-Rakam-Tila River (km135) 135 270 675 large * WB/IDA – RSDP Ph3
Sun Koshi-Okhaldhunga 42 160 270 0.7 22% * WB/IDA – RSDP Ph4
Hile-Leguwaghat-Bhojpur 92 1,320 DFID: RAP Feeder Road
Nepalthok-Khurkot 32 1,920 Govt of Japan/JICA
Kathmandu-Bhaktapur 10 1,000 Govt of Japan/JICA
Syabrubesi-Rasuwagadhi 26 520 Govt of China
1400km of Terai Roads 1,500 9,000 Govt of India
Sabha-Bramadev 13 260 Govt of India
Total 2,400 21,545
B. UPGRADING SCHEMES - EXISTING DoR ROADS (to Sealed Standards)
H03 Barabise - Kodari 8 645 48 9.2 Connection to Border: included with Fast Track Study
F13 Bhalubang – Chakchake – Rolpa (Liwang) 108 30-300 650 7.6 Upgrade access to District HQ
H11 Ameliya – Tulsipur 30 130-400 178 7.1 Upgrade section of National Highway
F30 Panchkhal – Melamchi 23 200-400 135 4.8 Upgrade access to Tourist Area
F47 Chhinchu – Devisthal – Jajarkot 107 150-300 642 2.3 * Complete construction & Upgrade access to District HQ
H18 Maldhunga-Beni 13 200 78 2.1 Upgrade access to District HQ
H11 Tulsipur – Shitalpati – Salyan 54 350-500 322 1.5 * Upgrade access to District HQ
F14 Chakchake - Pyuthan 25 65 150 1.1 Upgrade access to District HQ
H11 Shitalpati – Musikot 86 120 511 0.9 28% * Upgrade access to District HQ
F133 Tulsipur – Purandhara – Botechaur 86 50 516 0.2 14% * Upgrade link between District HQs
F48 Lower Dhungeshwar - Dailekh 28 120 218 20% * Upgrade access to District HQ
H20 Sanphebagar-Martadi 57 100 302 35% * Upgrade access to District HQ
H25 Mangalsen-Belkhet/Rakam 51 50 509 18% * Upgrade inter-District & inter-Regional linkage
H25 Safebagar-Mangalsen 38 100 256 15% * Upgrade access to District HQ
Total 714 4,515 NB: Costs as per HDM evaluation
C. NEW CONSTRUCTION (to FR Standards - sealing subject to FS)
Karnali Rajmarg (km135 to Jumla) 118 1,180 Completion & upgrading Kalikot-Jumla (118km)
Martadi-Kolti 51 673 23% * Upgrade GTZ track
Hilepani-Diktel 67 402 22% Provide access to District HQ (upgrading of RAP track)
Beni-Jomsom 80 480 15% Upgrade of track constructed by Army
Okhaldhunga-Salleri 50 333 14% Upgrading of Green Road / Army construction
Total 366 3,068
D. NEW CONSTRUCTION (as Earth Track mostly through local efforts)
Jajarkot-Dunai 144 1,656 9% Track opening by local efforts - access to Dolpa District HQ
Simikot-Hilsa 88 1,012 5% Track opening by local efforts
Besisahar-Chame 65 780 On-going DoR construction
Nagma-Gamgadhi 88 1,306 8% * Track opening by local efforts (DRILP): FS under SWRP-PIP
Total 385 4,754
E. UPGRADING OF LOCAL ROADS TO SRN STANDARDS
Baglung-Burtiwang 90 630 35% Upgrade ADB RIDP Road
Kaligandaki-Kushma 67 771 18% Requires Feasibility Study
Sandhikharka-Tamghas 31 217 Upgrade inter-District link
Sandhikharka-Pyuthan 49 343 Upgrade inter-District link
Tamghas-Warmitaksar 19 133 Upgrade inter-District link
Pharping-Kulekhani-Bhimphedi 24 240 Upgrade inter-District link & alternate access to Kathmandu
Bhedetar-Ranke 115 805 Upgrade inter-District link
Okhaldhunga-Khurkot 43 430 Construct inter-District link
Phidim-Myaglung 75 750 Upgrade inter-District link
Bhojpur-Diktel 55 550 Construct inter-District link
Daijee-Jogbuda-Budar 68 476 Upgrade inter-District link
Total 636 5,345
F. NEW CONSTRUCTION TO ACCESS KATHMANDU
Kathmandu-Terai Fast Track ** 65 18,000 16% Major access improvement between Kathmandu & Terai
Sitapaila-Bhimdhunga-Dharke ** 33 2,600 15% Alternative acces improvement to Kathmandu from West
G. OTHER STRATEGIC ROAD IMPROVEMENTS
Kathmandu Ring Road Upgrade 28 2,800 Provide dual carriageway, frontage roads & junction imp'vts
Bhaktapur-Banepa-Dhulikhel 20 1,000 Dual carraiageway upgrade
Narayanghat-Mugling Improvement 36 220 32% * Widen to 7.5m formation (Asian Highway Standards)
Hetauda By-Pass 15 450
Koshi Barrage Replacement ** 50 1,270 Assumes 50km diversion & 350m cable-stayed bridge
Notes:
* FS undertaken under SWRP-PIP Study
** Requires Deatiled Feasibility Study
All COSTS are indicative
DHV - SILT - TAEC – CMS
NPV/C & IRRs quoted as available
ChapterIX-24
April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

1. Committed Schemes
There are 16 schemes that are assumed to be committed and for which funding
is assured and which are mostly expected to be constructed over the next 3-4
years. These schemes are therefore not included in the subsequent prioritisation
process, although they are included in the overall budget estimates.

The 16 committed schemes – Table 10.5, Section A – comprise:

• six schemes included in three on-going ADB projects - Road Network

Development Project, Connectivity Project (three Core Schemes), and
Sub-Regional Trade Facilitation Project;
• four upgrading schemes selected for ‘fast-track’ implementation under the
forthcoming WB/IDA Road Sector Development Project (RSDP);
• one DFID-funded Feeder Road;
• one new and one improvement scheme funded by JICA;
• one Border Crossing link funded by the Government of China; and
• the upgrading of 1500km of existing roads/tracks in the Terai in three
priority packages over 9 years, plus a new connection to the Mahakali
Barrage at Bramhadev, funded by the Government of India.

Excluding the Indian programme in the Terai, these schemes comprise around
900km of upgrading and new construction and will represent the bulk of the
foreign-funded involvements on the SRN over the next 3-4 years. The Indian
programme will significantly improve access within the Terai areas over the next
10 years and will account for a substantial proportion of the proposed expansion
of the SRN.

The total cost of the committed works is approximately Rs21.5 billion (US$305
million), with the Terai Roads accounting for Rs9 billion (US$130 milion), over 40
percent of the total.

These committed schemes include the completion of an initial road access to

three District HQs – Khandbari (Sankhasabhuwa), Chainpur (Bajhang) and
Bhojpur – plus the upgrading of access to three more: Taplejung, Baitadi and
Okhaldhunga.

It may be necessary subsequently to consider upgrading the roads to Khandbari

and Bhojpur to sealed standards, as well as the link from Baitadi to Jhulaghat (on
the Indian Border): the latter should be considered in conjunction with a possible
border bridge over the Mahakali River.

2. Upgrading Schemes
The highest priorities among the non-committed schemes are the 14 upgrading
schemes – Table 10.5, Section B – totaling over 700km at an estimated cost of
Rs4.5 billion (US$65 million). These comprise the upgrading to sealed standards
of the most important and strategic sections of the existing (and extended) SRN
which are presently unsealed. The priorities may be grouped as follows:

ƒ Barabise-Kodari;
ƒ Bhalubang-Chakchake-Rolpa & Chakchake-Pyuthan;
ƒ Amelia-Tulsipur-Salyan (Sitalpati)-Musikot;
ƒ Panchkhal-Melamchi;
ƒ Chhinchu-Devisthal-Jajarkot (incl construction of final section);
ƒ Tulsipur-Purandhara-Botechaur;

DHV - SILT - TAEC – CMS ChapterIX-25

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

ƒ Lower Dhungeshwar-Dailekh;
ƒ Maldhunga-Beni;
ƒ Sanfe-Martadi; and
ƒ Sanfe-Mangalsen-Belkhet

An additional nine District HQs (highlighted above) will be connected by a sealed

road. The selection and ranking of the schemes has been based on the NPV/C
as calculated though the HDM-4 analysis or IRR for individual schemes subjected
to detailed feasibility study. Feasibility Studies were undertaken for all or parts of
eight of the 14 schemes, as shown by an asterisk on Table 10.5.

This programme will have upgraded and sealed the access roads to all of the
currently road-served District HQs, with the exception of Darchula: upgrading of
the final section (from Gokuleshwor) should be examined subsequently.

3. New Construction
The priorities for new construction involve the completion and possible upgrading
of 366km of existing (or under-construction) access tracks to a further five District
HQs – see Table 10.5, Section C. The cost is estimated at a little over Rs3 billion
(US$44 million). Feasibility Studies are required to ascertain the timing of any
subsequent upgrading to sealed standards.

ƒ Karnali Highway (from km135-Kalikot (Manma)-Jumla;

ƒ Hilepani (Sunkoshi)-Diktel (Khotang);
ƒ Okhaldhunga-Salleri (Solukhumbu);
ƒ Beni-Jomsom (Mustang); and
ƒ Martadi-Kolti.

The priority for four of these five schemes has been established during this study
(see Chapter 6): the fifth scheme was not evaluated at it entails the completion of
the Karnali Highway from km135 (Tila River, Kalikot Border) to Jumla. Initial
construction of part of this section is on-going through the Nepal Army and the
final section to Jumla was constructed through local contracts under RMDP.
Completion of the Karnali Highway is a core element of the overall strategy and it
should be accorded a high priority.

A second access is simultaneously being developed by GoN to Diktel (Khotang)

from Ghaighat. This project has a lower priority than the alternative from Hilepani,
as it serves fewer people and requires an additional bridge over the SunKoshi. It
should be retained, as part of the extended SRN, to serve the local population
and link the two District HQs.

An additional four schemes (total length 385km) with a lower priority and lower
economic returns are also proposed for new construction – see Table 10.5,
Section D. These schemes, listed below and connecting to the final four District
HQs, should be pursued only as earth tracks at present, built mostly through local
efforts and initiatives. The preliminary cost estimate is Rs4.8 billion (US$68
million). These should only be considered for upgrading to full Feeder Roads
Standards at a later date.

ƒ Jajarkot – Dunai (Dolpa)

ƒ Simikot – Hilsa (Humla – connection to China)
ƒ Besisahar – Chame (Manang – GoN Construction)
ƒ Nagma – Gamgadhi (Mugu)

DHV - SILT - TAEC – CMS ChapterIX-26

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Budgets for the construction of the three roads within the Karnali Cluster should
be channeled through DoLIDAR and associated local funding options (eg DRILP)
or local Districts funds. Construction of the Chame road should continue directly
through GoN.

4. Upgrading of Local Roads

In addition to the above improvements and extensions to the existing SRN and
links to District HQs, the PIP includes recommendations for the improvement or
upgrading of eleven local roads (totaling 636km) to Feeder Road Standards, at
an indicative cost of Rs5.3 billion (US$76 million). These should be subject to
detailed feasibility study to determine priorities and appropriate standards. The
links considered mostly form part of the proposed mid-hills east-west corridor and
are designed to improve inter-District connectivity and accessibility to densely
populated hill regions. The schemes are listed below and described in Table 10.5,
Section E.

ƒ Baglung-Burtiwang (upgrade RIDP earth track)

ƒ Kaligandaki-Kushma (construct new link)
ƒ Sandhikharka-Tamghas (upgrade local road)
ƒ Sandhikharka-Pyuthan (upgrade & complete local road)
ƒ Tamghas-Warmitaksar (upgrade local road)
ƒ Pharping-Kulekhani-Bhimphedi (upgrade local road)
ƒ Bhedetar-Ranke (upgrade local road)
ƒ Okhaldhunga-Khurkot (construct new link)
ƒ Phidim-Myaglung (construct new link)
ƒ Bhojpur-Diktel (complete/upgrade RAP track)
ƒ Daijee-Jogbuda-Budar (upgrade local road)

5. New Strategic Links

The Study has identified and confirmed the overall priority for the construction of
improved access between Kathmandu Valley and the Terai. There are two broad
options to be considered: a “Fast Track” direct link on a new alignment with the
potential of providing significant reductions in time and distance; and smaller
schemes to remove worst constraints on the existing route and to improve the
access to the Valley from the west. The adoption of either option should be
subject to detailed and exhaustive Feasibility & Funding Studies that are beyond
the scope of the present Study.

The Study however recommends that a full detailed Engineering Feasibility Study
be undertaken to establish realistic and updated cost estimates for the “tunnel”
options as these will undoubtedly produce the ‘best’ long term solution with the
greatest potential benefits to the country. The longer (and geologically vulnerable)
routes along the Bagmati Valley are likely to produce lower benefits.

If required, improvements to the western approaches to Kathmandu would be

best achieved through construction of a new 30km link from Darke on the Prithvi
Highway, via Bhimdhunga, to Sitapaila on the Ring Road. The alternative route
via the Kolphu Valley is significantly longer and offers no compensatory benefits.

The Study recommends that the following schemes be pursued to improve

overall access to Kathmandu Valley:

ƒ Kathmandu-Terai Fast Track (tunnel to Hetauda)

ƒ Sitapaila-Bhimdhunga-Dharke

DHV - SILT - TAEC – CMS ChapterIX-27

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

6. Other Strategic Road Improvements

Although the Study has not specifically focussed on urban transport issues, it is
evident that two major strategic road schemes within the Kathmandu Valley have
a high priority on a National level. These are:

ƒ Upgrading of the Kathmandu-Bhaktapur-Banepa-Dhulikhel corridor;

ƒ Improvements and widening of existing Ring Road

It is understood that funding has been committed by the Government of Japan for
the widening and improvement of the Kathmandu-Bhaktapur road to four-lanes: it
is recommended however that the upgrading be extended in subsequent phases
to Dhulikhel to service the likely traffic growth resulting from increased
urbanization in the corridor and the completion of the Sindhuli Road.

It is additionally recommended that funding be sought from other donors for a

comprehensive upgrading of the Ring Road, with specific attention to traffic
management, road safety, public transport requirements, and junction layout: the
improvements should be accomplished with minimal land-take.

The present SWRP has investigated the improvement of Narayanghat-Mugling to

meet Asian Highway Standards. This can be achieved with minor widening and
realignment, in conjunction with the next major maintenance intervention. As part
of the improvement of the main route between the Kathmandu and the Indian
Border (and as a key location on the EWH), it is recommended that a by-pass to
Hetauda be constructed in order to remove heavy through traffic from the town
centre.

Consideration should also be given to the construction of an alternative crossing

of the Sapta Koshi in Eastern Nepal to replace the existing Koshi Barrage which
is Indian owned and operated, is on the Indian border, and has an uncertain
structural future. It is the only link to the eastern areas of Nepal. Two alternatives
have been examined: one at Chatara close to where the river leaves the hills and
the other on a more direct route passing immediately north of the Koshi Tappu
Nature Reserve and involving a longer bridge crossing.

7. Priority Ranking
The 14 upgrading schemes (B) have the highest priority after the committed
schemes. These should be followed by the new construction (Schemes C) and
local road upgrading (Schemes E). Other non-prioritised projects should not be
entertained until these programmes are complete or funding is assured. The
lower priority remote area access schemes – listed under (D) – should be
pursued under local initiatives.

E. Sector Wide Road Programme

The key objective of the overall Study is the preparation of a Sector Wide
Investment Plan for the Road Sector to provide all-weather motorable access
within four hours walk of remote settlements (ref: ToR, page 1). The Study has
addressed this requirement and has completed a comprehensive analysis of
accessibility throughout the country which has determined the most cost-effective
and efficient means of achieving the desired levels of accessibility – within the
bounds of practicality.

Determination of accessibility is dependent on three key factors:

DHV - SILT - TAEC – CMS ChapterIX-28

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

• Firstly the location and distribution of population;

• Secondly the location and extent of the all-weather road network; and
• Thirdly the time taken to walk between home and the nearest road.

The Study has addressed all three aspects and has developed a GIS-based tool
that enables the accessibility of any particular road or network of roads to be
determined mechanically. Two measures of accessibility have been adopted in
the Study: the first – as required in the ToR – measures the percentage of
population within the desired access-bands (ie four hours walk in the hills and two
hours in the Terai); and the second calculates a total measure of accessibility as
the product of the number of people times their (time) distance from the road. The
output of the second measure is in terms of thousands of person-hours – with ten
people one hour from a road being equivalent to 1 person ten hours from a road.

It is considered that this second measure is a very useful additional indicator of

overall accessibility, although the 2-hour and 4-hour criteria are retained as the
primary indicator.

The approach adopted has only been possible through the use of GIS techniques
which allow the computation of vast numbers of individual calculations. Previous
attempts at measuring accessibility have been based primarily on manual
methods and use of topographic mapping, with resultant constraints on accuracy
and applicability.

The first key task was the preparation of the population density distribution maps
for the whole country. These are based on the population data abstracted from
the census at VDC or Municipal ward level. The populations within each of the
almost 4,000 VDCs have then been re-distributed geographically, within the VDC
boundary, based on the location of individual houses or house clusters taken
from the 1:25,000 topographic maps (which are in turn based on the 1995 air
photography). The distribution was adjusted to take account of the households
within the urban clusters based on an assumed house size: the adjustments and
assumptions were verified by means of a reality-check on household size.

The second task was the definition (in GIS coordinates and attributes) of the
existing and proposed road networks. Data were collected initially for the SRN
and the proposed extensions, and subsequently also for the local (District level)
networks. These data have been compiled into the series of networks presented
in Chapters 7 and 8: considerable effort was applied to the verification of the local
road network data, which were obtained from records maintained at DoLIDAR
and from the individual District Transport Master Plans (DTMPs). Supplementary
data were also obtained from the various donor agencies active in the rural roads
sector.

The analysis regarding the impact of the local road network is presented at two
levels: initially those roads for which the alignment and condition has been
verified and secondly all potential road alignments taken from the comprehensive
(but unchecked) inventory of 22,000km of road kept by DoLIDAR. The ‘verified’
network contains around 4,460km of operational road in 2006, increasing to
5,860km by 2016 on the basis of known commitments from various donor
agencies. [The unchecked network, when coupled with the SRN, creates a total
network of around 29,000km of potential road alignment.]

The third and final stage was the development of the walk-time model which is
based on walking speeds which are dependent on the gradient developed from a

DHV - SILT - TAEC – CMS ChapterIX-29

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

digital terrain model (DTM) based on a 90m grid of spot heights, coupled with the
barrier effects of rivers. This walk-time model develops a shortest path algorithm
between any two points, thereby allowing the development of time contours,
zones of influence, and catchment areas: when coupled with the population
density distributions, it is possible to compute populations within any given area
and also the time taken to reach the nearest road from any location.

Application of these modelling techniques has allowed the impact of the existing
SRN, and the proposed extensions, on overall levels of accessibility to be
determined. The results of this analysis are presented in Chapter 7. The existing
designated SRN (5,030km) provides access – as measured by the 4 hour and 2
hour criteria – to 65 percent of the total population: 50 percent in the hills and 76
percent in the Terai. The extension by 2016 to 9,930km increases the overall
accessibility to 86 percent – 70 percent in the hills and 97 percent in the Terai.

Similarly, the proportion of people within 1 hour of a road increases by 50 percent

from 10.5 million people to 15.4 million, and those more than 12 hours from a
road reduce from 8 percent today (2.05 million) to around 1 percent (0.24 million).
These are clearly very significant impacts, but illustrate also the difficulty (and
impracticality) of providing the 4 hour level of access to ALL people.

Analysis conducted at a District level indicates that the most severe problems –
measured in terms of person-hours walk-time from a road – are currently
experienced in the remote northern Districts in the east and west of the country:
this is due primarily to the distances involved to reach the nearest road. However,
once the initial access has been provided into these areas, the priorities for
additional road construction are no longer in these remote regions, but rather in
the more densely populated mid-hills Districts – including a number of Districts
relatively close to the Terai or to Kathmandu.

The overall impact of the proposed expansion of the SRN between 2006 and
2016 is a reduction in the average walk-time to reach a road from 2½ hours to 1
hour 20 minutes, with a reduction from over 5 hours to less than 3 hours in the
hills and from 20 minutes to 12 minutes in the Terai.

The additional impact of the inclusion of the local road network on the 2 hour and
4 hour accessibility is comparatively minor, as most local roads are built in areas
close to (and served by) the SRN and most are relatively short. Local roads do
however improve accessibility by reducing the access time to the SRN, but this is
primarily in areas that are already served.

The effect of the LRN is however significant in terms of the additional reduction in
walk-time to reach a road, as most of the local roads are constructed in areas
with a high population density. In 2016, the LRN produces an overall reduction of
30 percent in walk-time (from 78 minutes to 55 minutes), with an even more
dramatic reduction in the Terai (from 12 minutes to 3 minutes).

The overall analysis has however identified a limited number of pockets of high
density population that are NOT served by either the SRN or the LRN. It is
recommended that local road programmes be developed in these areas to define
the best and most appropriate form of local road to meet the observed access
deficiencies.

The District analysis has identified initially those Districts with current poor levels
of accessibility, and these are basically in the remote northern areas in the east
and west of the country. By 2016, with the extension of the SRN, the picture is

DHV - SILT - TAEC – CMS ChapterIX-30

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

significantly different: the worst Districts are those with higher population
densities in the mid-hills and many are in fact close to Kathmandu. This requires
a different response and different interventions. Consideration of the currently-
known local road network, and proposed extensions, addresses the deficiencies
to an extent, but a significant proportion of the “least accessible” Districts remain
in the mid-hills region. It is recommended therefore that a programme to improve
local road provision in these worst-effected Districts be instigated.

It is clear also from the analysis that there are relatively few people living in the
northern mountainous regions of the country that are beyond the area served by
the proposed extensions to the SRN. There is thus no prima facie case for more
northerly extensions of the network. [It is also noted that much of the population
of these northern areas is there to support the tourism industry which, in turn, is
based on the beauty and remote, inaccessible nature of the region. Construction
of roads into these areas – which are mostly National Parks or Conservation
Areas – would indeed be counter-productive and damaging to the fragile ecology
of the area.]

F. Part II Detailed Feasibility Studies

Part II of the Study involves the Detailed Feasibility Study of around 1,060km of
road for upgrading or new construction. The original ToR and Contract specified
a total of 600km, of which approximately 400km would be upgrading and 200km
would be new construction. Selection of the roads was carried out in the early
phases of the Study and was targeted specifically at improving levels of
accessibility in the remote hill areas: roads with on-going projects were excluded,
as were those with funding committed from other donors. For practical scheduling
reasons, it was necessary to select the roads for study prior to the completion of
the ranking and prioritisation for the extension of the SRN.

It became evident early in the selection process that there was limited potential to
identify entirely new alignments for road construction: in almost all cases, some
form of local road or track existed, constructed either by the local DDC or under
some other donor-funded local road initiative. It was also evident that the priority
need was for the upgrading of previously constructed earth or gravel tracks to an
improved (and more reliable) all-weather standard. Substantial lengths of new
road and track had been built – both as part of the SRN and as local roads – but
a large proportion were in poor condition and not available for general traffic for
much of the year.

The initial screening process eliminated many potential roads in the Central and
Eastern Regions as these were mostly included in on-going or proposed projects
funded by either ADB or DFID. An initial list of around 640km was finally agreed
and approved by DoR on 25th June 2006, although survey work had commenced
following the resolution of the conflict and the restoration of democracy in late
May. An additional 168km of road for upgrading was requested during September
2006, bringing the total to 815km. The Consultants were further obligated to
review and update the designs for four sections of road (236km) which had been
included in the previous phase of the RMDP but which had not been completed
due to the conflict situation.

The location of the roads studied are illustrated in Figure 10.6: it can be seen that
most are in the hills in the Mid and Far Western Regions of Nepal, plus one road
in the East and the improvement also of the Narayanghat-Mugling section of the
main route between Kathmandu and the Terai.

DHV - SILT - TAEC – CMS ChapterIX-31

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

The Consultants’ approach to the roads in the Mid and Far West has been to
consider them as a ‘network’ aimed at improving the overall levels of accessibility
and integration, by improving to sealed standards a mix of both north-south and
east-west routes. The objective was to improve all-weather accessibility to the
disadvantaged and excluded rural hill populations.

Despite the relatively low existing traffic volumes on the roads selected (which is
primarily due to the very poor condition and unreliable service offered), the
Consultants consider that upgrading to sealed standards will produce net benefits
from the increased traffic levels to be expected after improvement. Specifically
passenger trip making can be shown to increase significantly with the availability
of faster and more frequent bus services that become practical with a sealed all-
weather road, and more people will be within easy reach of the services provided.

Similarly, the volume of freight imported into the hills can be expected to increase
as individual consumption rises with increased proximity to well stocked and
supplied markets. [Improved access tends to have minimal impact on agricultural
production in the short term and, in any event, the potential volume of exports is
not significant in relation to the volume of goods being imported: typically imports
in the hills are around 10 times the volume of exports – and thus any produce
exported gets a ‘free ride’ as back-haul cargo on otherwise empty trucks.]

The results of the analysis of the 15 upgrading roads, the one ‘new construction’,
and four additional ‘review and update’ roads are presented in Table 10.6. All the
upgrading roads are shown as having a positive NPV and IRR in excess of 12
percent. The exception is the one new construction road (Nagma-Gamgadhi)
which is not justified at the present time: it is recommended that this road be built
as an earth track using a local participatory labour-based approach. All the
upgrading roads were evaluated with a 3.5m Otta Seal which, in all cases,
produced a superior result than a gravel surface.

The methodology for the ‘review and update’ roads was firstly to identify the
extent of works that was left ‘undone’ (based on a review of the original contract
and the payments made for works completed) and then to re-cost these works
based on current rates: this – plus the cost of a new base course and Otta Seal –
was then taken as the works required to complete the road to a sealed standard.

The widening of Narayanghat-Mugling to Asian Highway standards – 6.0m, plus

0.75m shoulders – is justified in conjunction with an AC overlay throughout.

The schemes are justified in most cases because of the exceedingly poor
condition of the present roads, which suppresses traffic demand: the provision of
an improved all-weather road surface has been shown elsewhere to have a
significant impact on traffic levels and in the type of vehicle operating.

The total cost of the 20 road schemes (1,060km) is estimated at Rs7.8 billion or a
little over US$100 million.

DHV - SILT - TAEC – CMS ChapterIX-32

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

DHV - SILT - TAEC – CMS ChapterIX-33

April 2007
 Sector Wide Road Programme & PIP Study
Part I Final Report

Table 10.6: Summary of Findings of Detailed Feasibility Studies

Buildings

affected
directly
Constr Cost
Length Traffic Rate of
Road Section Status Cost per km NPV
(km) ADT* Return
(Rs m) (Rs m)

Upgrading Roads:
Satbanj-Jhulalghat 37 76 Open 5 247 6.72 3.8 12.2
Satbanj-Gokuleshwar 54 96 Open 6 413.7 7.66 297.6 21.6
Gokuleshwar-Thaktholi 41 74 Rough 1 296.9 7.73 52.9 14.6
Khodpe - Jhota [- Chainpur] 78 128 Open 5 548.1 7.03 153.9 15.9
Sanfe - Ekadighat [- Martadi] 15 121 Rough 0 135.4 6.64 280.9 34.2
Martadi-Kolti 52 66 Track 84 595.5 11.5 525.5 23.1
Sanfe-Mangalsen 38 93 Rough 0 255.5 5.84 53.7 15.1
Mangalsen-Belkhet 51 51 Track 25 509.4 9.93 212 17.8
Lwr Dungeshwar- Dailekh 28 113 Open 24 217.8 7.84 121.8 20
Chhinchu-Pokhare 25 119 Open 0 109.2 4.24 72.4 20.1
Tulsipur-Purandhara- Part
55 111 0 8.7
Botechaur open 772.1 78.7 14.1
[6.5]
Tulsipur-Cement Works 25 511 Open 0
Tulsipur-Salyan 63 137 Open 0 339.7 5.53 384.5 24.9
Sitalpati-Musikot 86 180 Rough 10 613.6 7.15 791.6 28
Mugling-Narayanghat 34 3,400 Open 194.4 5.76 744.5 32.2
Not
Sun Koshi-Okhaldhunga 42 164 4 231.3 5.77 175.4 22.5
linked
New Construction:
Foot
Nagma-Gamgadhi 88 58 12 1,305.6 14.63 -343.9 7.6
Trail
Roads previously examined under RMDP (Phase 1):
Thaktoli-Darchula 31 30 Track 182.6 5.37 9.6 12.8
Foot
Sanfe-Martadi (from km29) 28 87 166.6 4.38 460.7 41.9
Trail
Part
Surkhet-Kalikot 132 270 488.8 3.62 2,378.2 59.6
Open
Part
Pokhare-Chhedegadh 45 149 206.2 3.38 842.8 55.3
Open
Total: 1048 km 7,829 7.47
Notes: Cost per km for individual roads excludes major bridges; all costs exclude VAT
* ADT on year of opening

DHV - SILT - TAEC – CMS ChapterIX-34

April 2007