-
Table of ContentsSummary 1
1. Introduction 3
2. Building a Database for the Analysis of Road Costs 42.1
Establishing the Baseline for a New Database on Road Costs 42.2
Consolidation of Databases from Previous Studies 42.3 Addition of
Cost Data from Recent AfDB Projects 52.4 Master Database for
Analysis of Unit Rates and Cost Overruns 6
3. Analytical Approach for Road Infrastructure Costs 83.1
Standardizing the Unit Rate 83.2 Analytical Approach 8
4. Results 104.1 Establishing the Unit Cost Curve on the Basis
of Project Size 10
4.1.1 Construction/Upgrading of Paved Roads 104.1.2
Rehabilitation of Paved Roads 104.1.3 Periodic Maintenance of Paved
Roads 114.1.4 Regraveling of Unpaved Roads 124.1.5 Summary Results
12
4.2 Other Major Potential Drivers of Unit Rates 124.2.1 Location
of Road Projects 124.2.2 Origin of Contractors 144.2.3 Landlocked
vs. Seaboard Countries 16
4.3 Determining Unit Cost Trends 184.3.1 Unit Rates over Time
184.3.2 Cost Overruns/Underruns 204.3.3 Median Rates 22
5. Conclusions 255.1 Typical Road Unit Costs 255.2 Unit Cost
Overruns/Underruns 255.3 Trend in Unit Costs 265.4 General
Conclusions 26
Appendix A: Detailed Description of Database Parts for AICD
Initiative 27
Appendix B: Statistical Terms 29
Appendix C: Consolidated Data 30
AfDBMarket Study Series
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in AfricaStatistics
Department (ESTA)
May 2014www.afdb.org
A f r i c a n D e v e l o p m e n t B a n k
Acknowledgements: This report was prepared by a team compris-ing
Maurice Mubila, Chief Statisti-cian, (Statistics Department), Altus
Moolman, Consultant, (Statistics Department) and Willem Van Zyl,
Consultant (Statistics Department) under the supervision of Beejaye
Kokil, Manager, Economic and So-cial Statistics Division, and the
di-rection of Charles Leyeke Lufumpa, Director Statistics
Department.
This report was prepared by the Statistics Department in the
Chief Economist Vice Presidency of the African Development Bank.
Its findings reflect the opinions of the authors and not
necessarily those of the African Development Bank, its Board of
Directors, or the countries that they represent. Designations
employed in this ar-ticle do not imply the expression of any
opinion on the part of the
African Development Bank Group concerning the legal status of
any country or territory, or the delimi-tation of its frontiers.
While every effort has been made to present reliable information,
the African Development Bank accepts no responsibility whatsoever
for any consequences of its use.
Layout and production by Phoenix Design Aid A/S, Denmark
Mthuli NcubeChief Economist and Vice PresidentOffice of the
Chief Economist
Charles Leyeka LufumpaDirectorStatistics Department
Steve Kayizzi MugerwaDirectorDevelopment Research Department
Victor MurindeDirectorAfrica Development Institute
SUMMARY
The African Development Bank (AfDB) commissioned a study during
2010/11 to analyze road unit costs and the extent of cost overruns
in road infrastructure projects in Africa. The study focused on
three main objec-tives, namely: (i) to determine unit costs for
road infrastructure projects in Africa; (ii) to determine the
prevalence and extent of cost overruns; and (iii) to determine the
evolution of unit costs and cost overruns since the comple-tion of
previous studies specif-ically the Africa Infrastructure Country
Diagnostics (AICD) study and a related study for the AfDB on Road
Maintenance and Construction Costs in Africa.
This represents the Final Report for the study and presents the
findings of the analysis on unit costs and cost overruns of road
infrastructure projects in Africa.
Research was undertaken during 2007 and 2008, under the auspices
of the joint World BankAfDB Africa Infrastructure Country
Diagnostic (AICD), into baseline unit cost data and the causes of
cost overruns, specifically on road
projects. The research resulted in the compilation of databases
on road projects in Africa, which were used as the point of
depar-ture for the current study. The databases were consolidated
into a new structure, and a total of 26 new AfDB projects (with
Project Completion Reports (PCRs) dated 2004 or later) were added
to the database. The final database con-sisted of a total of 172
projects.
The consolidated structure divided projects into four
categories, based on the type of work that was undertaken, as
follows:
1. Regraveling or periodic maintenance of unpaved roads: This
involves reinstat-ing the surface layer of gravel roads;
2. Periodic maintenance of paved roads: This involves the repair
of minor surface defects and a seal or thin overlay, but without
structural improve-ments or geometric upgrades;
3. Rehabilitation of paved roads: This typically entails the
reinstatement of roads to
-
2
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
the original design standard, including structural repairs;
and
4. Construction and upgrading of paved roads: This typically
entails the upgrad-ing of gravel roads to paved standard, or the
addition of lanes to existing paved roads.
Analysis was performed on the projects in the database. The
sequence of enquiry followed three steps, namely (i) establishing
the position of the unit cost curve, (ii) veri-fying whether rates
that lie off the curve can be explained, and (iii) determining
whether the position of the curve is as expected (overruns) and
whether it is shifting over time (unit cost trends).
What this review has brought to the fore is that there is no
such thing as a typical unit cost. This is because (i) unit costs
are calculated by standardizing projects that are broadly similar
but which differ in their design details and specific
circumstances; and (ii) the size of the project invariably has an
overriding effect on the unit rate (economy of scale). The first
issue is largely overcome by excluding major project and
location-specific factors (e.g. bridges, taxes). The second issue
is something that anyone estimating or eval-uating roads costs
should be vigilant about.
The table below provides a summary of the unit cost findings. It
should be noted that the unit rates are all expressed per lane
kilometer, i.e. a 50 km two-lane (single carriageway) road would
have 100 lane km.
The analysis of unit cost overruns shows that (i) there appears
to be a correlation between the over/underrun and the size of the
project and (ii) the estimation error (i.e. PCR value minus Project
Appraisal Report (or PAR) value) is likely to be an underestimate
(48 percent) rather than an overestimate (-15 percent).
A typical observation is that the smaller the project, the
larger the difference between the expected unit rate (PAR value)
and the PCR value. The implication is that unit rates for small
projects should be treated with some caution, although care should
be taken not to spend more resources on refining designs,
feasibility studies, and other work underlying PARs than the
ben-efit that could be expected.
In the case of the rehabilitation of paved roads, the difference
occurs both above (overrun) and below (underrun) the PAR value. In
the case of construction or upgrading of paved roads, it appears
that the pattern is for small projects to overrun,
rather than underrun. This may point to PARs being overly
optimistic.
The finding with respect to an increase in unit cost over time
is inconclusive. This may be purely because of data constraints,
i.e. a limited sample size for a specific year and standardization
issues across projects in the same class. The effect is that
statisti-cally extrapolated unit cost curves (rather than rates)
are compared.
Given these shortcomings, comparing the cost curves of those
years for which such curves can indeed be constructed, shows that
unit costs for large projects (>100 lane km) have not increased
during the last dec-ade. It might even be inferred that they have
reduced, although this is counterin-tuitive, given the field
experience of AfDB task managers, which points to increasing unit
rates.
The main conclusion from this study is that while lenders and
national road agencies will clearly benefit from having a better
understanding of unit costs and related issues, a more permanent
road unit cost database should be established (possibly under the
AICD initiative), which can then be analyzed on a systematic
basis.
Summary of Unit Rate Statistics for Different Types of Road
Infrastructure Investment (USD/lane km, rounded to 00)
Type of Road Infrastructure Investment
Regraveling/ Periodic Maintenance of Unpaved Roads
Periodic Maintenance of Paved Roads
Rehabilitation of Paved Roads
Construction and Upgrading of Paved Roads
< 100 lane km
Quartile 3 10,500 N/A 290,000 425,400
Median 9,600 N/A 180,300 227,800
Quartile 1 8,100 N/A 109,800 166,300
100 lane km
Quartile 3 12,800 72,200 130,500 162,000
Median 11,300 64,600 84,400 147,100
Quartile 1 9,600 56,900 47,400 115,900
Note: All values are given in 2006 USD.
-
AfDB African Development Bank Group / Chief Economist
Complex
3
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
1. Introduction
The African Development Bank (AfDB) and other development
agencies are con-cerned about significant cost escalations of road
projects under implementation. This is particularly important,
given the AfDBs firm commitment to scale up its support to
infrastructure development, as an opera-tional priority of its Ten
Year Strategy 2013-2022. Investing in infrastructure is a prime way
to boost economic growth, improve the social wellbeing of
populations, and promote regional integration. However, excessive
cost overrruns result in increased funding needs and act as a
constraint to development; therefore a mechanism is needed to
address this issue.
The imperative to scale up infrastruc-ture and improve the
competitiveness of African economies is being constrained by
limited data availability. Indeed, there is a scarcity of
information regarding the costs of implementing road infrastructure
projects in Africa, although significant data on the unit cost of
projects exist, both in government records and those of
devel-opment agencies in the region. There is therefore a need to
systematically review and analyze these sources, to improve the
generation of statistical data on the unit
costs of various types of road infrastructure investments.
Studies conducted about four years ago observed that cost
overruns in road infra-structure projects had become increasingly
common. The average cost overrun was 35 percent, but in a third of
the cases it could be as high as 50 to 100 percent. The assumption
has been that the increases are due to a variety of factors,
including lack of competition in the bidding process, increases in
fuel and bituminous product prices locally and internationally,
technol-ogy used in road works, contract manage-ment practices, and
the availability and quality of road construction materials.
The African Development Bank commis-sioned a study during
2010/11 to analyze road unit costs and the extent of cost overruns
in road infrastructure projects in Africa. The study focused on
three main objectives, namely: (i) to determine road unit costs for
road infrastructure projects in Africa; (ii) to determine the
occurrence and extent of cost overruns; and (iii) to determine the
evolution of unit costs and cost overruns since the time of
comple-tion of previous studies (specifically the
Africa Infrastructure Country Diagnostics (AICD) study and a
related study for the AfDB on Roads Maintenance and Construction
Costs in Africa). The cur-rent study represents the Final Report
and presents the findings of the analysis on unit costs and cost
overruns of road infrastruc-ture projects in Africa.
A four-step methodology was followed for the study: (i) first,
the project databases from the previous studies were consoli-dated;
(ii) new projects were then selected from the AfDB database, for
comparison with the previous studies; (iii) data were extracted for
each of the selected projects and added to the consolidated
database; and (iv) analysis was performed to deter-mine unit rates
and cost overruns.
This report contains four sections, includ-ing this
introduction. Section 2 describes the development of a consolidated
data-base of old (i.e. from previous studies) and newly selected
projects, which were used for the analysis of unit rates. Section 3
provides an analysis of unit rates and cost overruns for all
projects captured in the consolidated database. Section 4
sum-marizes the main findings of the report.
-
4
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
2. Building a Database for the Analysis of Road Costs
Significant data on the unit cost of projects exist in the
records of governments and donor agencies in the region. Under the
auspices of the joint World BankAfDB initiative of the Africa
Infrastructure Country Diagnostic (AICD), research was undertaken
during 2007 and 2008 into unit cost data. This then served as a
baseline against which future improve-ments in infrastructure
services could be measured. The research also analyzed the causes
of cost overruns, specifically on road projects.
The research resulted in the compilation of databases on road
projects in Africa, containing cost data and allowing
cate-gorization of the type of work conducted. These databases were
used as the point of departure for the current study.
2.1 Establishing the Baseline for a New Database on Road
CostsAfricon Limited completed a study on unit costs during 2008 as
part of the joint World BankAfrican Development Bank AICD
Initiative. The main aim of the study, which focused on 24
countries in Sub-Saharan Africa, was to provide a set of baseline
unit cost data against which future infrastructure improvements
could be measured. This was expected to provide a more solid and
empirical foundation for prioritizing investments and design-ing
policy reforms in the infrastructure sectors in Africa.
As part of this study, a database of infra-structure projects in
Africa was compiled, covering roads, water, sanitation, and
elec-tricity. The database was populated from Project Completion
Reports (PCRs), con-tractual documents from donor agencies
participating in the study, and publicly
available data from governments. For roads, the database covered
115 projects.
The database consisted of two parts, namely: (i) a general
section common to all projects, which captured the major exogenous
variables; and (ii) an infra-structure-specific section, adapted
for each sector (e.g. roads, water, and elec-tricity) and for
different infrastructures and technologies within each sector. A
detailed description of these parts is presented in Appendix A.
In 2007, Africon Limited was appointed to carry out an addendum
to the AICD assignment which was then underway. The topic of the
addendum generally related to the main unit cost study, i.e. an
investiga-tion into the causes of recent cost overruns,
specifically on road projects.
Whereas the main AICD study sought to determine the final cost
of projects that had already been completed, the Addendum examined
the reasons for cost changes in ongoing projects, relative to the
original engineers estimates. The purpose of the Addendum was not
to analyze and improve the project appraisal and procurement cycle
per se, but rather to identify and mitigate the causes of the
recent wave of cost overruns (i.e. during the time of the AICD
study).
A total of 24 AfDB projects were selected for the purposes of
the study. Project details were obtained on the following: brief
project description; timeframe (listing milestones such as Specific
Procurement Notices (SPNs) issued, pre-qualification, tender
issued, tender received, tender evaluation approved, contract
signed); key project events; procurement; project com-ponent
increases; changes in explanatory
variables; summary impacts on tender price; and
classification.
Since the purpose of the 2007 AfDB study was to investigate cost
overruns and not unit costs, it was not a requirement to enter
these project details into a general database. Rather, the focus
was on analyzing these projects on an individual basis.
2.2 Consolidation of Databases from Previous StudiesThe
databases developed for the AICD study (2008) and the Addendum to
AICD (2007), differed slightly in terms of their structure and the
data captured. This was owing to the divergent objectives of each
study. There was a need to consolidate these databases into one
structure, taking due account of the specific objectives of this
report. The consolidated structure con-sisted of nine major data
fields, as listed below:
1. Project information (general infor-mation describing the
project and its location);
2. Procurement information (e.g. International Competitive
Bidding (ICB) or National Competitive Bidding (NCB));
3. Number of bidders, plus the name and nationality of the
winning bidder
4. Project dates (dates for approval and signing of loan,
signing of contract, expected and actual contract com-pletion
dates);
5. Type of project (road type and sur-face type);
-
AfDB African Development Bank Group / Chief Economist
Complex
5
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
6. Activity (category describing the type of work performed, as
dis-cussed below);
7. Inputs & Costs: a) as per Project Appraisal Report (PAR)
and b) as per Project Completion Report (PCR);
8. Size descriptors (number of lanes and road length); and
9. Currency (currency of costs, con-version factor to USD and
USD CPI factor, for conversion to 2006 USD).
All projects from the 2008 AICD study, as well as from the 2007
AfDB study, were captured in the new database structure for the
purposes of this report. It should how-ever be noted that some
projects did not have data for all the fields in the database.
Projects listed in the database were further grouped into four
categories, based on the type of work that was performed. These
four categories, as well as a definition of each, are as
follows:
1. Regraveling or periodic mainte-nance of unpaved roads. This
is the activity of reinstating the sur-face layer of gravel roads.
This term was also applied when there were substantial gravel road
rehabilitation activities;
2. Periodic maintenance of paved roads. This involves the repair
of
minor surface defects and a seal or thin overlay, but without
struc-tural improvements or geometric upgrades. These contracts
come in a range of light to heavy upgrades. Since the term periodic
mainte-nance is used by road agencies for several different types
of activity, some contracts with this activity in their title were
reclassified (mostly to rehabilitation/reconstruction);
3. Rehabilitation of paved roads. This typically entails the
reinstatement of roads to the original design stand-ard, including
structural repairs; and
4. Construction and upgrading of paved roads. This covers the
upgrading of gravel roads to paved standard, or the addition of
lanes to existing paved roads.
It should be noted that the above four cat-egories are fairly
broad, and may contain a wide variance with regard to technical
standards. For example, new road con-struction (category 4) can
include stand-ards such as gravel base and asphalt layer, gravel
base and seal layer, bitumen-treated base and asphalt layer, and
concrete base. However, in order to conduct a statisti-cal
analysis, a sufficiently large sample is required (i.e. data
points), and a compro-mise was therefore reached between the number
of categories and level of detail per category. Furthermore, the
data available in PCRs often did not provide a detailed technical
description of the type of design
and work conducted. For these reasons, the four categories
indicated above were used for the purpose of this study.
2.3 Addition of Cost Data from Recent AfDB ProjectsThe previous
studies (discussed above) contained projects up to about 2007.
Therefore, in the identification of new projects for addition to
the database, the focus was generally on post-2007 projects.
During 2010, the AfDB availed a list of 44 projects (approved
since 2007) for consideration in the study. The list was then
evaluated based on the availability of their project documents. For
most of the projects, PARs were available but not PCRs. As the PARs
did not contain all the data required for the database informa-tion
that was also needed for the analysis of unit rates and cost
overruns it was decided to investigate a wider range of
projects.
Subsequently, it was decided to list all pro-jects for which
PCRs dated 2004 or later (although the actual projects could have
been completed much earlier) were availa-ble from DARMS (Documents
and Records Management System). A further criterion was that such
projects should not have been included in the previous studies. The
list contained 26 projects and was approved by AfDB for addition to
the database, and for use in analysis of road infrastructure costs
and cost overruns. The list of 26 new projects is indicated in
Table 2-1.
-
6
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
PCRs and PARs for the 26 projects indi-cated above were
extracted from DARMS. All relevant data were obtained from these
documents and recorded in the database. Some of the projects
comprised two or more types of road infrastructure invest-ments
(e.g. rehabilitation of paved roads, as well as the upgrading of
paved roads). In such cases, each road infrastructure
investment was captured separately in the database, resulting in
more than one entry for the specific project.
2.4 Master Database for Analysis of Unit Rates and Cost
OverrunsThe final database of projects used for the analysis of
unit rates consisted of projects
from the 2008 AICD study, the 2007 AfDB study, as well as the
new AfDB projects identified during 2010. A total of 172 pro-jects
were included in the final database, as indicated in Table 2-2.
Table 2-1: List of New Projects Selected for Addition to the
Database
Number Country Project Name
1 Benin CotonouPorto Novo Road Rehabilitation Project
2 Botswana Trans-Kgalagadi Road Project
3 Burkina Faso Second Road Program
4 Cameroon Road Improvement in the West, Littoral and South
Provinces
5 Chad DjermayaMassaguet Road Construction Project
6 Chad Road Rehabilitation and Maintenance Project
7 Ethiopia AlemgenaButajira Road
8 Ethiopia Road Maintenance and Rehabilitation Project
9 Ghana AchimotaAnyinam Road Rehabilitation Project
10 Morocco Road Project III
11 Lesotho LikalanengThaba Tseka Road Upgrading - Lot1
(LikalanengCheches Pass)
12 Lesotho MpharaneBela Bela Road Upgrading Project
13 Madagascar Road Rehabilitation and Maintenance Project
14 Malawi MsuliraNkhotakota Road Project
15 Malawi Road Maintenance and Construction (ROMAC II)
Project
16 Mauritius South-Eastern Highway Project
17 Mozambique PembaMontepuez Road Rehabilitation Project
18 So Tom and Principe Second Road Maintenance Project
19 Swaziland Transportation Sector Project
20 Swaziland Two International Roads Project
21 Tanzania HimoArusha Road Rehabilitation Project
22 Tunisia Classified Road Network Development Project II
23 Tunisia Classified State Road Network Development Project -
Phase I
24 Tunisia Classified State Road Network Rehabilitation
Project
25 Uganda KyoteraMutukula Road Upgrading Project
26 Uganda Rural Feeder Roads Maintenance Program
-
AfDB African Development Bank Group / Chief Economist
Complex
7
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
Table 2-2: Projects Included in Final Database
Data Source
Number of Projects
Regraveling/ Periodic
Maintenance of Unpaved Roads
Periodic Maintenance of
Paved RoadsRehabilitation
of Paved Roads
Construction and Upgrading of Paved Roads Total
2008 AICD Study 37 4 51 23 115
2007 AfDB Study 2 0 7 13 22
New projects from AfDB identified during 2010 3 2 14 16 35
Total 42 6 72 52 172
Note: As described above, some of the new projects entailed more
than one type of road infrastructure investment. In such cases,
each investment was cap-tured separately, resulting in more than
one entry for the specific project. Therefore a total of 35 new
projects are reflected in Table 2-2, and not 26 as indicated in
Table 2-1.
-
8
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
3. Analytical Approach to Road Infrastructure Costs
3.1 Standardizing the Unit RateThe Terms of Reference for the
study identified three topics that should be investigated, namely
(i) what typical road unit costs are; (ii) whether the actual unit
rates overran what was initially anticipated and if so, to what
extent; and (iii) whether there is a trend in unit cost rates. The
unit rate applied in this study is the cost per lane-kilometer. The
lane-kilometer is the product of the number of lanes and the road
length.
Road design standards differ between juris-dictions, and road
works do not necessarily fall neatly into the four types of project
clas-sifications identified above. Projects were classified based
on the description of the works in the PCR and the project
contract. The main adjustment made to compensate for physical
differences was to exclude all major bridges and structures.
Physical characteristics that were not standardized included the
use of specific materials and differing geometric standards.
Data on lane widths were not always avail-able. The 3.50m
appears to dominate the data, but there are lane width outliers
from 3.25 m to 3.80 m. The unit rate does not compensate for width
variations. For regraveling and periodic maintenance of paved
roads, all roads were single carriage-way (two lanes). For the
construction of paved roads, about 4 percent of projects were dual
carriageway (four lanes), while for the rehabilitation of paved
roads, dual carriageways comprised about 1 percent.
Financial adjustments made to standard-ize projects entailed
excluding all (i) fea-sibility, environmental, design and other
studies; (ii) social mitigation costs (e.g. relocation costs);
(iii) supervision and audit services; and (iv) taxes. The value
considered is the latest available contract
cost, either obtained from the PCR or the project contract. The
2008 AICD data and 2010 AfDB data (i.e. the current study) are
typically based on PCR values. The 2007 AfDB data mostly entailed
projects that were on-going and that had experienced unexpected
increases from the point of project appraisal or loan approval,
prior to the contract being completed. For the purposes of this
report, contract values were standardized in the same manner as for
the 2007 AfDB and 2008 AICD studies.
Unit rates are expressed as United States Dollar (USD) per lane
kilometer. All con-tracts were standardized to 2006 USD value. This
was the convention used in the previous two studies (which supply
the bulk of the material considered in this study). The most common
currency in the combined database is the UA (AfDB Unit of Account),
followed by the FCFA and the USD, but the USD is favored as the
currency to which most readers can readily relate.
3.2 Analytical ApproachIn analyzing the four types of road
project, a similar sequence of enquiry is followed. It entails
three steps: (i) establishing the position of the unit cost curve;
(ii) verifying whether rates that lie off the curve can be
explained; and (iii) determining whether the position of the curve
is as expected (overruns) and whether it is shifting over time
(unit cost trends). These steps are further elaborated below.
Step 1: The projects in the particular cat-egory are reviewed,
highlighting the con-tribution of each of the three studies to the
universum and indicating any major differences between the three
data subsets. The convention applied is to plot unit rates
by showing the relationship between the rate and the project
size (lane length). This is a crucial understanding: in most cases,
the main explanatory variable for the unit rate is economies of
scale. The smaller the project, the more disproportionate is the
relative contribution of project overheads such as Preliminaries
& General (site establishment).
Step 2: After project size, the other major potential drivers of
the unit rate are con-sidered. It is expected that these variables
could help to explain any deviation from the economies-of-scale
curve already estab-lished. (a) Regional characteristics such as
geography, climate, business practice and state of the contracting
industry are cap-tured by grouping projects into Northern
(Maghreb), Western (largely ECOWAS), Central (largely ECCAS),
Eastern (EAC and surrounds), and Southern Africa (mostly SADC). (b)
The origin (home country) of the main contractor is con-sidered, to
determine whether outlier unit rates can be traced back to the
prevalence of home country vs regional or contractors from further
away. (c) It is often argued that a countrys accessibility to the
sea affects its cost of construction. This prop-osition is tested
by comparing unit costs between landlocked countries and those with
seaboards.
Step 3: Movement of the unit cost curve could be short-term,
i.e. the unit rate achieved exceeding the rate expected by the
roads agency; or longer-term, i.e. that rates are drifting upward
over time. In the short-term, overrun refers to a project
completion amount exceeding the sum originally foreseen. Completion
amounts can be the actual completion, as recorded in the PCR, or
the contracted amount. The original amounts can be either as stated
in the PAR or the loan agreement. This
-
AfDB African Development Bank Group / Chief Economist
Complex
9
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
study applied the PCR values for comple-tion and the PAR values
for the expected amount. In other words, overrun defines the extent
to which the contracted amount was exceeded.
The short-term overrun issue relies heavily on relevant project
documentation being available. Of the three studies, only the 2007
AfDB investigation attempted to unpack the reasons for overruns and
the current report therefore largely reflects overrun findings of
that study. Here, cost overruns were assessed against the reference
project amount as per the loan approval, or where this was
unavailable, the PAR.
Road projects considered were contracted through 20002010. For
the unit rate trend investigation, for most of the project types,
there are insufficient data points to explain possible long-term
upward trends in unit rates. In no cases are there directly
com-parable projects (in terms of size, country/location, etc.)
which would enable a firm conclusion to be drawn on changes in unit
rates. Unit rate trends are therefore cal-culated by comparing
best-fit statistical simulations of different time periods.
The unit rate data are subject to large var-iances. This may be
the result of projects being incongruously categorized
together,
even though they may have some unique characteristics. Or it may
be due to vari-ances in contracting conditions; or a num-ber of
other considerations. This report therefore presents the unit rates
not as a firm value, but as ranges of values that indi-cate the
confidence interval around the calculated rate. The unit rates are
presented as median (not average) rates, bounded by first and third
quartile intervals. One set of statistics is provided for projects
smaller than 100 lane km (typically subject to a large variance)
and those larger than 100 lane km (typically demonstrating fairly
small variance).
-
10
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4. Results
4.1 Establishing the Unit Cost Curve on the Basis of the Project
Size4.1.1 Construction/Upgrading of Paved RoadsThe database
includes 52 construction or upgrading of paved roads projects,
about a quarter each from the 2007 AfDB and AfDB 2010 studies, and
about half from the 2008 AICD study. Of these, 47 have usable data
(i.e. contracted values).
Table 4-1: Construction/Upgrade of Paved Road Projects
Sample
Study/Region North West Central East South Total
2007 AfDB - 3 - 7 2 12
2008 AICD - 4 2 14 2 22
2010 AfDB 1 1 2 3 6 13
Total 1 8 4 24 10 47
Table 4-2: Rehabilitation of Paved Road Projects Sample
Study/Region North West Central East South Total
2007 AfDB 1 5 - 1 - 7
2008 AICD - 8 2 8 33 51
2010 AfDB 3 3 3 2 2 13
Total 4 16 5 11 35 71
As shown in Figure 4-1, for large projects ( 100 lane km), the
unit cost rates of the three studies distribute similarly. For
smaller projects (< 100 lane km), there are some outlier
projects, especially in the 2010 AfDB and 2007 AfDB studies that
result in very spiky curves compared with the AICD 2008 study
curve, which shows a less prominent diseconomy of scale.
Figure 4-1: Construction/Upgrading of Paved Road Projects per
Study
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
0 100 200 300 400 500
USD/lane km
lane km
AfDB 2007
AICD 2008
AfDB 2010
Swaziland MR3 Bypass
Mauritius South-Eastern Highway
4.1.2 Rehabilitation of Paved RoadsA total of 72 projects were
classified under the rehabilitation of paved roads inter-vention.
The majority (two-thirds) were from the 2008 AICD study, with 14
added in this study. Unit rates could not be calcu-lated for one
project, leaving a universum of 71 projects, as indicated in Table
4-2.
-
AfDB African Development Bank Group / Chief Economist
Complex
11
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.1.3 Periodic Maintenance of Paved RoadsThe combined data set
includes six pro-jects for the periodic maintenance of paved roads
(Table 4-3). Four of these are from the 2008 AICD study and two
were intro-duced in the 2010 AfDB review.
Figure 4-2: Distribution of Paved Rehabilitation Projects per
Study
Figure 4-3: Periodic Maintenance of Paved Road Projects per
Study
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
AfDB 2007
AICD 2008
AfDB 2010
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
AfDB 2007
AICD 2008
AfDB 2010
Table 4-3: Periodic Maintenance of Paved Road Projects
Sample
Study/Region North West Central East Southern Total
2007 AfDB - - - - - -
2008 AICD - 3 1 - - 4
2010 AfDB 1 - - - 1 2
Total 1 3 1 - 1 6
From Figure 4-3, it can be seen that the unit rates of the two
2010 AfDB projects are substantially lower than the rates from the
2008 AICD study. The lower of the two 2010 unit rates is a
resurfacing project in Morocco and the other a resealing project in
Malawi.
The projects that were added to the overall database by the 2010
AfDB study distribute very similarly to those of the 2008 AICD
study (Figure 4-2). The projects from the 2007 AfDB study tended to
be more expensive if smaller than 500 lane km. It should be
recalled that the 2007 projects were specifically selected because
they had resulted in overruns.
-
12
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.1.4 Regraveling of Unpaved RoadsFor unpaved roads, the major
intervention identified in the three studies is regrave-ling. There
are 42 projects in the database, of which 37 yield useful unit
rates (Table 4-4). The bulk of projects are from the 2008 AICD
study, while two were added in the AfDB 2010 review. Reliable unit
rates could not be calculated for the two regraveling projects from
the 2007 AfDB study.
Table 4-4: Regraveling of Unpaved Roads Projects Sample
Study/Region North West Central East South Total
2007 AfDB - - - - - -
2008 AICD - 31 3 1 - 35
2010 AfDB - 2 - - - 2
Total - 33 3 1 - 37
Figure 4-4: Regraveling Projects per Study
Figure 4-5: Regional Distribution of Construction/Upgrading of
Paved Road Projects
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
0 200 400 600 800
USD/lane km
lane km
AfDB 2007
DRC projects
AICD 2008
AfDB 2010
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
0 100 200 300 400 500
USD/lane km
lane km
North
West
Central
East
South
The addition of the two AfDB 2010 projects enhances an
understanding of the cluster of unit rates below 100 lane km
(Figure 4-4). Even with the addition of these pro-jects, the same
economy-of-scale curve that appears for other project types still
does not appear here. Even if the outlier projects are ignored, the
curve remains flat.
4.1.5 Summary ResultThe unit cost curve confirms the econ-omies
of scale proposition, that the smaller the project, the more
dispro-portionate is the relative contribution of project. In other
words, smaller con-tracts with road size < 100 lane Km have
higher unit costs compared to larger projects.
4.2 Other Major Potential Drivers of Unit Rates4.2.1 Location of
Road Projects4.2.1.1 Construction/Upgrading of Paved RoadsFor
construction/upgrading of paved roads, half of the projects are in
East Africa and a fourth in Southern Africa (Figure 4-5). The unit
rate variance is large (i.e. there is a high degree of scatter in
the data) for small projects in both these regions. The most
outlying small project is the Mbabane bypass (Swaziland) a project
that over-spent notoriously. The outlier small project from the
2007 AfDB study is an 11 km project in East Africa (Mauritius).
That country has a sizable contracting industry, which means that
the unit rate cannot be explained simply by it being an island
state.
-
AfDB African Development Bank Group / Chief Economist
Complex
13
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.2.1.2 Rehabilitation of Paved RoadsFor rehabilitation of paved
roads, half of the projects are in Southern Africa, 22 percent in
West Africa, and 15 percent in East Africa (Figure 4-6). The
distribution of unit rates reflects the typical econo-mies-of-scale
curve, with higher unit rates for small projects and lower rates
for larger ones. As was previously noted in the 2007 AfDB and 2008
AICD reports, the variance is particularly high for projects of
fewer than 100 lane-km. For paved rehabilita-tion projects, the
initial peakiness (unit rates from about USD500,000/lane km) is
ascribed to projects in the West Africa subregion, specifically
four in Benin and two in Ghana.
Figure 4-6: Regional Distribution of Paved Rehabilitation
Projects
Figure 4-7: Regional Distribution of Periodic Maintenance of
Paved Road Projects
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
North
West
Central
East
South
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
North
West
Central
East
South
4.2.1.3 Periodic Maintenance of Paved RoadsFor periodic
maintenance, as shown in Figure 4-7, the difference between the
rates of the 2010 and 2008 studies may be attributable to the fact
that the projects are widely distributed over the continent. A
further explanatory factor could be that the two AfDB 2010 projects
predate 2000 (i.e. it is expected that unit rates have increased,
even in real terms). Nevertheless, the data sample (six projects)
is so small that it can-not be concluded with certainty that these
projects unit rates are in fact outlying.
-
14
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.2.1.4 Regraveling of Unpaved RoadsFor regraveling roads, the
three apparent outlier projects are all in Central Africa,
specifically the Democratic Republic of the Congo (DRC). These were
all emergency projects in Katanga Province. The cluster of small
(less than 25 km road length) West African projects relates to the
rural access program in Burkina Faso. For purposes of calculating
the summary statistics below, this program is excluded because it
has a dominating effect in the overall calculation.
Figure 4-8: Regional Distribution of Periodic Maintenance of
Paved Road Projects
Figure 4-9: Paved Road Construction/Upgrading Projects by Origin
of Contractor
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
0 200 400 600 800
USD/lane km
lane km
AfDB 2007
AICD 2008
AfDB 2010
-
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
0 100 200 300 400 500
USD/lane km
lane km
Home Country
Regional Country
Europe
China
Other East
Other
N/A
4.2.1.5 Summary ResultIn all four test scenarios, the location
of the project does not seem to signifi-cantly influence the
distribution of unit rates.
4.2.2 Origin of Contractors4.2.2.1 Construction/Upgrading of
Paved RoadsFor construction/upgrading of paved roads (Figure 4-9),
Chinese contractors account for a significant share of projects.
However, except for one possible project (the 11.1 lane km
Mauritius South-Eastern Highway Project), they are not obviously
responsible for outlier contracting values. The widest variance in
contract values may rather be ascribed to local contractors.
-
AfDB African Development Bank Group / Chief Economist
Complex
15
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.2.2.2 Rehabilitation of Paved RoadsAlthough there is no
discernible pat-tern in contractor origin versus unit rate (Figure
4-10), three of the outlier con-tracts identified above were
carried out by European contractors and one by a Chinese
contractor.
Figure 4-10: Paved Rehabilitation Projects by Origin of
Contractor
Figure 4-11: Paved Road Periodic Maintenance Projects by Origin
of Contractor
-
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
Home Country
RegionalCountry
Europe
China
Other East
Other
N/A
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
Home Country
RegionalCountry
Europe
China
Other East
Other
N/A
4.2.2.3 Periodic Maintenance of Paved RoadsSimilarly for
periodic maintenance of paved roads (Figure 4-11), it is unlikely
that the origin of the contractor played a significant role in the
unit rate variance.
-
16
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.2.2.4 Regraveling of Unpaved RoadsAs might be expected for a
type of inter-vention that does not require a high degree of
sophistication or major capital invest-ment, regraveling contracts
are dominated by domestic contractors (Figure 4-12). When the three
outlier projects (DRC) are excluded, projects in landlocked
coun-tries have very similar unit rates to those in seaboard
countries (Figure 4-13). There is however still no clear pattern to
make a conclusion.
Figure 4-12: Regraveling Projects by Origin of Contractor
Figure 4-13: Construction/Upgrading of Paved Road Projects
Landlocked vs. Seaboard Countries
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
0 200 400 600 800
USD/lane km
lane km
Home Country
Regional Country
Europe
China
Other East
Other
N/A
2,000,000
1,800,000
1,600,000
1,400,000
1,200,000
1,000,000
800,000
600,000
400,000
200,000
0
0 100 200 300 400 500
USD/lane km
lane km
Landlocked
Seaboard
Power (Landlocked)
Power (Seaboard)
4.2.2.5 Summary ResultIn all four test scenarios, the origin of
the contractor does not seem to sig-nificantly influence the
distribution of unit rates.
4.2.3 Landlocked vs. Seaboard Countries4.2.3.1
Construction/Upgrading of Paved RoadsFor construction/upgrading of
paved roads, there is no apparent difference in the distribution of
unit costs for landlocked countries versus those with sea
coasts.
-
AfDB African Development Bank Group / Chief Economist
Complex
17
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.2.3.2 Rehabilitation of Paved RoadsFor rehabilitation of paved
road projects, the unit rate patterns are very closely cor-related
between landlocked and seaboard countries (Figure 4-14), with no
difference in the distribution of unit costs.
Figure 4-14: Paved Rehabilitation Projects Landlocked vs.
Seaboard Countries
Figure 4-15: Paved Rehabilitation Projects by Landlocked vs.
Seaboard
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
Landlocked
Seaboard
Power (Landlocked)
Power (Seaboard)
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
Landlocked
Seaboard
4.2.3.3 Periodic Maintenance of Paved RoadsThe project sample is
too small to draw a conclusion about the unit cost comparison
between landlocked and seaboard coun-tries (Figure 4-15).
-
18
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.2.3.4 Regraveling of Unpaved RoadsFor regraveling roads, there
is no clear pat-tern (Figure 4-16).
Figure 4-16: Regraveling Projects Landlocked vs. Seaboard
Countries
Figure 4-17: Construction/Upgrading of Paved Road Projects Unit
Rates over Time
USD/lane km
lane km
Landlocked Seaboard Power (Landlocked) Power (Seaboard)
0
5,000
10,000
15,000
20,000
25,000
30,000
0 200 400 600 800
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
0 100 200 300 400 500
USD/lane km
lane km
pre-2000
2000
2001
2002
2003
2004
2005
2006
2007
4.2.3.5 Summary ResultFor at least two test cases, it appears
unit rates between landlocked and seaboard countries strongly
correlate without any major influence on unit cost
distribution.
4.3 Determining Unit Cost Trends4.3.1 Unit Rates over
Time4.3.1.1 Construction/Upgrading of Paved RoadsFigure 4-17 shows
the trend in unit rates over time (i.e. the movement of the unit
cost curves). Good-fit curves can be plotted for the period prior
to 2000 (grey), and for 2003 (green), 2004 (yellow), and 2006
(orange). For projects larger than 400 lane km, the 2006 curve
exceeds the others, but the occurrence of projects of this length
is limited, and firm statistical conclusions cannot be drawn. For
projects of 200 lane km and less, the variance of unit rates is
high, resulting in an apparent decrease in unit costs over
time.
-
AfDB African Development Bank Group / Chief Economist
Complex
19
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.3.1.2 Rehabilitation of Paved RoadsFigure 4-18 demonstrates
the evolution of unit rates over time for rehabilitation of roads.
Three trend lines are shown for years that had a relatively large
number of projects: pre-2000 (grey), 2004 (dark yellow), and 2006
(orange). The 2004 and 2006 curves virtually overlie. Unexpectedly,
they are both slightly lower than the pre-2000 curve.
Figure 4-18: Paved Rehabilitation Projects Unit Rates over
Time
Figure 4-19: Paved Road Periodic Maintenance Projects Unit Rates
over Time
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
pre-2000
2000
2001
2002
2003
2004
2005
2006
2007
2008
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
pre-2000
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
4.3.1.3 Periodic Maintenance of Paved RoadsThe project sample is
too small to make a conclusion about the unit cost comparison over
time (Figure 4-19).
-
20
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.3.1.4 Regraveling of Unpaved RoadsThe projects universum is
dominated by projects from 2005 and 2006, with no subsequent
projects recorded. There is no clear pattern of unit rate changes
over time (Figure 4-20).
Figure 4-20: Regraveling Projects Unit Rates over Time
Figure 4-21: Construction/Upgrading of Paved Road Projects Unit
Rate Overruns/Underruns (PCR vs. PAR)
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
0 200 400 600 800
USD/lane km
lane km
pre-2000
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
-120%
-100%
-80%
-60%
-40%
-20%
0%
20%
40%
60%
80%
100%
0 50 100
150
200
250
300
350
400
450
500
% Over/Under (Contract vs PAR)
lane km
4.3.1.5 Summary ResultIn cases with wide distribution of sam-ple
data, it appears that for smaller pro-jects the unit cost rate
reduces over time while for larger projects it increases.
4.3.2 Cost Overruns/UnderrunsThe summary result above is borne
out by comparing the expected unit rate (as per the PAR) with the
actual contracted rate.
4.3.2.1 Construction/Upgrading of Paved RoadsCost over/underrun
data are available for 24 of the projects (Figure 4-21).
Significant overruns (>20%) occurred in seven pro-jects and
significant under-expenditure (< -20%) in five projects. The
major overrun-ning projects were: the Cameroon Road Improvement
Project; Swaziland Two International Roads Projects; the Lesotho
MpharaneBela Bela Road Upgrading Project; Swaziland M3 Bypass
Project; and the Ghana Akatsi-Dzodze (Akatsi-Akanu Section)
Project. The Tanzania KigomaLusahunga Project (308lane km) is the
only large project that overran significantly.
It would therefore appear that small pro-jects are more
susceptible to cost overruns than are larger projects.
-
AfDB African Development Bank Group / Chief Economist
Complex
21
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
4.3.2.2 Rehabilitation of Paved RoadsOver/underruns were
generally not signifi-cant except for very small contracts, which
in any case are subject to much wider price variances. In the case
of rehabilitation of paved roads (Figure 4-22), there appears to be
a clear correlation between the degree of over/underrun and the
size of the pro-ject: the smaller the project, the larger the
difference between the expected unit rate (PAR value) and the
contracted value.
Figure 4-22: Paved Rehabilitation Projects Unit Rate
Overruns/Underruns (PCR vs. PAR)
Figure 4-23: Paved Road Periodic Maintenance Projects Unit Rate
Overruns/Underruns (PCR vs. PAR)
-40%
-20%
0%
20%
40%
60%
80%
100%
120%
0 200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,000
% Over/Under (Contract vs PAR)
lane km
-50%
-40%
-30%
-20%
-10%
0%
10%
20%
0 50 100
150
200
250
300
350
400
% Over/Under (Contract vs PAR)
lane km
4.3.2.3 Periodic Maintenance of Paved RoadsFrom Figure 4-23, it
can be seen that con-tracts for the two 2010 AfDB projects were
concluded substantially below the rates originally estimated.
4.3.2.4 Regraveling of Unpaved RoadsIt was only possible to
calculate cost over-runs for the two 2010 AfDB projects. In both
cases, the projects had underruns of less than 1 percent compared
with the PAR estimate. It was therefore not possible to draw any
conclusions from the limited sample size.
4.3.2.5 Summary Results Small projects are more susceptible
to cost overruns than larger projects.
Construction/Upgrading of roads is more susceptible to cost
overruns than underruns.
Rehabilitation of roads exhibits both characteristics
(overrun/underrun).
-
22
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.3.3 Median Rates4.3.3.1 Construction/Upgrading of Paved
RoadsSmall construction/upgrade of paved road projects have a
median unit rate (i.e. the unit rate in the middle of the dataset)
of about USD 228,000/lane km, but ranging from USD 61,000 below to
USD198,000 above. For large projects, the expected rate is USD
147,000 but ranging from USD 31,000 lower to USD 15,000 higher, as
indicated in Figure 4-24.
Table 4-5: Unit Rate Statistics for Construction/Upgrading of
Paved Road Projects (USD/lane km, rounded to 000)
Metric < 100 lane km 100 lane km
Quartile 3 425,400 162,000
Median 227,800 147,100
Quartile 1 166,300 115,900
Figure 4-24: Representative Dataset for Construction/Upgrading
of Paved Road Projects
Figure 4-25: Representative Dataset for Paved Road
Rehabilitation Projects
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
3,000,000
3,500,000
0 100 200 300 400 500
USD/lane km
lane km
Data
Median100km
0
100,000
200,000
300,000
400,000
500,000
600,000
0 500 1,000 1,500 2,000
USD/lane km
lane km
Data
Median100km
For both small and large projects, the inclu-sion of the 2010
AfDB dataset exerts an augmenting effect on unit rates. For small
projects, the 2008 AICD median unit rate was USD 201,000/lane km.
The 2008 AICD median unit rate for large projects was USD
145,000/lane km.
4.3.3.2 Rehabilitation of Paved RoadsFigure 4-25 shows the
dataset for paved rehabilitation projects after correcting for the
major outliers. Outliers includes two 2007 projects that exceeded
the PAR sig-nificantly, as well as the other very peaky West
African projects.
-
AfDB African Development Bank Group / Chief Economist
Complex
23
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
Accommodating the wide variance in unit rates for projects below
100 lane km, the summary statistics for paved road rehabili-tation
projects are shown in Table 4-6. For small rehabilitation of paved
road projects, the unit rate is expected to be about USD
180,000/lane km, but it can range from USD 70,000 lower to
USD110,000 higher. For large projects, the expected rate is USD
84,000, but ranging from USD 37,000 lower to USD 46,000 higher.
The median unit rates in the 2008 AICD study were USD
150,000/lane km for small projects and USD 79,000/lane km for large
projects. (The unit rate was not calculated in the 2007 AfDB
study.)
Table 4-6: Unit Rate Statistics for Paved Road Rehabilitation
Projects (USD/lane km, rounded to 000)
Metric < 100 lane km 100 lane km
Quartile 3 290,000 130,500
Median 180,300 84,400
Quartile 1 109,800 47,400
Table 4-7: Unit Rate Statistics for Periodic Maintenance of
Paved Road Projects (USD/lane km, rounded to 00)
Metric < 100 lane km 100 lane km
Quartile 3 N/A 72,200
Median N/A 64,600
Quartile 1 N/A 56,900
Figure 4-26: Representative Dataset for Periodic Maintenance of
Paved Road Projects
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
0 100 200 300 400
USD/lane km
lane km
Data
Median100km
4.3.3.3 Periodic Maintenance of Paved RoadsTwo 2010 AfDB outlier
projects (Figure 4-26) are excluded, leaving three fairly tightly
spaced unit rate data points for periodic maintenance of paved
roads.
The median unit rate for projects larger than 100 lane km is USD
65,000. The first quartile is about USD8,000/lane km lower, while
the upper quartile is about USD 8,000/lane km higher than the
median. Since only the 2008 AICD projects are used to calculate the
summary statistics, the median rate is the same as calculated in
that in the study.
-
24
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
4.3.3.4 Regraveling of Unpaved RoadsWith the exclusion of two
outlier projects, the DRC and Burkinab program, the regraveling
dataset is graphically presented in Figure 4-27.
Table 4-8: Unit Rate Statistics for Regraveling Projects
(USD/lane km, rounded to 00)
Metric < 100 lane km 100 lane km
Quartile 3 10,500 12,800
Median 9,600 11, 300
Quartile 1 8,100 9,600
Figure 4-27: Representative Dataset for Regraveling Projects
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
3,000,000
3,500,000
0 100 200 300 400 500
USD/lane km
lane km
Data
Median100km
This is the only type of road project reviewed that does not
demonstrate a clear pattern of economy of scale. The median unit
rate for small projects (less than 100 lane km) is about USD
10,000, ranging downward by USD 2,000 and upward by USD 1,000/lane
km. The median rate for large regraveling projects ( 100 lane km)
is USD 11,000, minus or plus USD 2,000/lane km.
In the 2008 AICD study, the median rate was calculated for all
(i.e. small and large projects) at USD7,800/lane km.
4.3.3.5 Summary ResultThe median unit rates clearly demon-strate
a pattern of economy of scale, i.e. smaller contracts have higher
unit costs compared to larger projects.
-
AfDB African Development Bank Group / Chief Economist
Complex
25
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
Conclusions
The analysis for this report was performed on a total of 172
projects (sourced from the 2008 AICD study, the 2007 AfDB study,
and newly selected projects obtained from AfDB during 2010), to
determine unit rates for different types of road infrastructure
investments. Of these 172 projects, unit rates could be calculated
for 161 projects in this study.
In response to the studys Terms of Reference, the following
conclusions are drawn:
5.1 Typical Road Unit CostsOne very important conclusion from
this review is that there is no such thing as a typical unit cost.
This is because (i) unit costs are calculated through a process of
standardizing projects that are broadly sim-ilar but which differ
in their design details and specific circumstances, and (ii)
the
Table 5-1: Summary of Unit Rate Statistics for Different Types
of Road Infrastructure Investment (USD/lane km, rounded to 00)
Type of Road Infrastructure Investment
Regraveling/ Periodic Maintenance of Unpaved
Roads
Periodic Maintenance of
Paved RoadsRehabilitation of
Paved Roads
Construction and Upgrading of Paved
Roads
< 100 lane km
Quartile 3 10,500 N/A 290,000 425,400
Median 9,600 N/A 180,300 227,800
Quartile 1 8,100 N/A 109,800 166,300
100 lane km
Quartile 3 12,800 72,200 130,500 162,000
Median 11,300 64,600 84,400 147,100
Quartile 1 9,600 56,900 47,400 115,900
Note: All values are given in 2006 USD.
size of the project invariably has an over-riding effect on the
unit rate (economy of scale). The first issue is largely overcome
by excluding major project and location-spe-cific factors (e.g.
bridges, taxes). The second issue is something that anyone
estimating or evaluating roads costs should be vigilant about.
Although it may be advantageous, from a statistical analysis
perspective, to recom-mend that road project costs should be cast
into a standard format (e.g. standard bills of quantities) that
will facilitate ex post comparison, this could add to the
administrative burden of national road agencies. Rather, it is
recommended that lenders through programs such as AICD classify
relevant data into standardized databases, as extracted by task
team leaders at PAR and subsequent stages.
Table 5-1 provides a summary of the unit cost findings.
5.2 Unit Cost Overruns/UnderrunsNearly a third of the projects
are from the 2008 AICD Study, where cost overrun was not an area
investigated. This implies that the conclusions on overruns are
based on a smaller dataset than the rest of the analysis.
Analysis of cost overruns indicates that for all the projects
where data on over/under-runs are available, the overruns amount to
48 percent and the underruns to -15 percent of the PAR values.
The main conclusions are that: (i) there appears to be a
correlation between the over/underrun and the size of the pro-ject
and (ii) the estimation error (i.e. PCR value minus PAR value) is
likely to be an underestimate (48 percent) rather than an
overestimate (-15 percent). These two conclusions are further
elaborated upon in the next paragraphs.
-
26
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
In terms of conclusion (i), the analysis shows that the smaller
the project, the larger the difference between the expected unit
rate (PAR value) and the contracted value. This may be because
larger contrac-tors are more sophisticated in their costing, and/or
that funding agencies expend more effort on the price estimates of
larger pro-jects. The implication is that unit rates for small
projects should be treated with some caution. However, care should
be taken not to spend more resources on refining designs,
feasibility studies, and other work underlying PARs than might be
expected to accrue in terms of the project benefits.
In terms of (ii), the estimation error varies according to the
type of project under-taken. For the rehabilitation of paved roads,
the difference occurs both above (overrun) and below (underrun) the
PAR value. In the case of the construction or upgrading of paved
roads, it appears that the pattern is for small projects to
overrun, rather than underrun. This suggests that PARs may have a
tendency to be overly optimistic.
5.3 Trend in Unit Costs The finding on an increase in unit cost
over time is inconclusive. This may be
purely because of data constraints, i.e. the limited sample size
for a specific year and standardization issues across projects in
the same class. The effect is that statistically extrapolated unit
cost curves (rather than rates) are compared.
Given these shortcomings, where com-parisons can be made, these
indicate that unit costs for large projects (>100 lane km) have
not increased during the last decade. It could even be inferred
that they have reduced, although this is counterintuitive, given
the field experience of AfDB task managers, which points to an
upward trend in unit rates.
5.4 General ConclusionsIt is important for lending agencies and
national road agencies to track road unit cost trends and issues so
that these can be reflected in their planning. Their ability to
influence the rates is probably fairly limited, other than in the
case of contractor market distortions (e.g. collu-sion). However, a
greater understanding of issues and trends will allow planners and
lending agencies to make more accurate projections.
Two general conclusions can be drawn, the first related to the
implications of the study findings and the second regarding the
analysis process:
1. The study findings point out that many of the traditional
causes of high road costs (e.g. geographic location, origin of
contractor) are not very useful to explain unit cost rates. The
principal explanatory factor appears to be project size and related
issues, e.g. that small contracts get less attention from lend-ers.
Teasing out deeper statistical rela-tionships will require a more
thorough and larger database.
2. In terms of the analysis of road unit costs, much greater
effort is required in two areas, namely: (i) the management of
documentation (enabling ex post analysis) and (ii) proactively
extract-ing relevant data, while issues are still recent and
well-understood. Although the data may only be analyzed
periodi-cally, the database itself should be kept up to date. The
AICD initiative, based on collaboration among the World Bank, AfDB,
and other lenders, is a useful platform to establish a perma-nent
database.
REFERENCES
Africon Limited. 2008. Africa Infrastructure Country Diagnostic
Study: Consulting Services for the Evidence on Unit Costs of
Infrastructure Projects in Sub-Saharan Africa. Washington, DC:
World Bank Group.
Africon Limited. 2007. Africa Infrastructure Country Diagnostic
Study: Study on Unit Costs of Infrastructure Projects in
Sub-Saharan Africa, Addendum 2: African Development Bank Study on
Roads Maintenance and Construction Costs in
Africa. Washington, DC: World Bank Group.
-
AfDB African Development Bank Group / Chief Economist
Complex
27
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
APPENDIX A: Detailed Description of Database Parts for AICD
Initiative
Africa Infrastructure Country Diagnostic Database (2008)
Africon Limited carried out a study on unit costs during 2008
under the auspices of the joint World BankAfrican Development Bank
initiative on Africa Infrastructure Country Diagnostic (AICD).
As part of this study, a database of infra-structure projects in
Africa was compiled, including roads, water, sanitation, and
electricity. The database was designed to consist of two parts,
namely: a) a general section common to all projects, capturing the
major exogenous variables; and b) an infrastructure-specific
section, adapted for each sector (e.g. roads, water, and
elec-tricity) and for different infrastructures and technologies
within each sector.
A short description of these parts is given in the next
sections.
a) General Contract Information The following general
information was
obtained for all contracts:
Project and contract information, including project/program
title, pro-ject/program unique number, contract title, and contract
unique number.
Task Team Leader information, including name and contact details
(email, telephone number).
Geographic information (country/ies, district/s and location/s)
and whether urban, rural, or deep rural. An attempt was made to
standardize the use of urban for cities, towns, and large
set-tlements, rural for small settlements, and deep rural for
locations totally
isolated from towns and villages. In the case of roads, the term
inter-urban was used for roads through rural areas but connecting
towns, and rural access for rural feeder roads.
Implementing agency/ies, includ-ing the main funding Development
Finance Institution (DFI), other DFIs and local funders and role
players.
Procurement, which may be National Competitive Bidding (NCB),
International Competitive Bidding (ICB) or direct appointment.
Bidders information, including the number of bidders (at Request
for Proposal/tender stage), the name of the winning bidder, and
nationality of the winning bidder.
Key dates, including the date the con-tract was signed, the
contracted com-pletion date, and the actual completion date (in as
far as these are available).
b) Specific Contract Information The following information for
each
infrastructure (e.g. roads) was obtained:
Infrastructure type, i.e. what tech-nology. Road projects were
differen-tiated according to whether they were urban, inter-urban
and rural access, whether with or without a shoulder, and whether
paved or unpaved.
Activity, including preservation, reha-bilitation, improvement,
and/or new construction.
Basis of costing, i.e. whether a pre-con-tract estimate,
contracted value, or the contract value as modified in the course
of the contract.
Input costs. This subsection divided the infrastructures into
their major components. This allowed some com-parison to be made
between technol-ogies, but importantly, it provided the basis on
which to isolate components of costs, which appeared to be
out-of-the-norm in the course of standardizing contract
information. The major com-ponents considered were:
Studies, design, land acquisition and environmental costs prior
to construction (note that for roads, environment is isolated as a
stan-dalone item to facilitate the use of the study data in
ROCKS);
Contractors establishment and mobilization (for roads,
demolition, dismantling, site clearance and other site preparation
are again isolated for use in ROCKS);
Mass earthworks;
Civils and structures. These were fur-ther subdivided into the
following:
Major bridges and structures; Minor bridges, culverts and
drainage; Accesses and junctions; Pavement courses; Shoulder
works; and Ancillary road works;
-
28
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
In the calculation of unit costs, the cost of major structures
and bridges was not included, as these were very contract-specific
and did not cast light on average road costs.
Mechanical, electrical, and control equipment;
Installation and commissioning (including dayworks);
Other (non-categorized costs);
Taxes;
Contingencies; and
Supervision (owners engineering costs).
Output descriptors, i.e. the number of units purchased with the
contract
amount, on the basis of which output unit costs are calculated.
For road projects, the key descriptors were the number of lanes and
road length (km).
Currency-related information, including the contract
currency/ies, the conversion date, the conversion factor/s (to USD)
and USD CPI factor (for conversion to 2006 USD).
-
AfDB African Development Bank Group / Chief Economist
Complex
29
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
APPENDIX B: Statistical Terms
Note on Statistical Terms
This report employs a number of statis-tical terms to describe
the methodology and findings. These all have to do with the
distribution of unit costs, i.e. how unit rates distribute around a
value that indi-cates the norm for the particular type of project
(construction, rehabilitation, etc.). In the analysis, there are
two major steps to interpret the data statistically:
Best-fit curve. The first step is to determine a formula that
eliminates the noise of the scattered unit cost data, so that the
data can be shown as a curve. The best-fit curve is the
mathematical
formula that reduces most noise. The formula used in the
analysis is a nega-tive exponential curve (y = a.x-b). The best-fit
curve allows conclusions to be drawn about the typical unit rates
across all project sizes.
Statistics of distribution. The best-fit curve allows projects
to be divided into large (> 100 lane km) and small projects. For
each group, the central value is determined with reference to the
median. The central value shows a point around which the data
cluster. The median is the value in the middle
of the data range, i.e. there are as many values smaller than
the median as val-ues larger than it. The median is gener-ally
accepted to be a truer reflection of the central value than the
average, as it gives a lower importance to outlier val-ues (very
high or very low values in the data set). The data clustering or
noise around the median is measured by the data variance. The more
scattered the data, the higher the variance and the lower therefore
the confidence that can be placed in the median.
-
30
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
APPENDIX C: Consolidated Data
Source Class Project Title Country/iesContractor Nationality
Dates Contract Price DimensionsOriginal Currency
Exchange Rate
USD CPI
Factor USD/lkmPAR Contract Completion PCR PAR PCR Lanes
Length
Periodic Maintenance/Regravel (Unpaved roads)
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.049 2 0.30 FCFA 1.00 1.00 -
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.035 2 1.00 FCFA 1.00 1.00
17,485.40
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.076 2 4.20 FCFA 1.00 1.00
8,988.80
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea Guinea 10/22/2002 3/1/2003 - 0.067 2 4.20 FG 1.00 1.00
8,004.86
AICD 2008 Regravel Nakivubo Channel Rehabilitation Project
Uganda Uganda 2/2/2004 10/13/2004 - 0.976 2 4.25 USH 1.00 1.00
-
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.045 2 5.10 FCFA 1.00 1.00
4,422.17
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.094 2 6.50 FCFA 1.00 1.00
7,192.47
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 3/21/2005 6/7/2006 - 0.101 2 6.66 FCFA
1.00 1.00 7,603.35
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.099 2 7.80 FCFA 1.00 1.00
6,353.96
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.130 2 8.30 FCFA 1.00 1.00
7,812.75
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea Guinea 11/5/2002 3/24/2003 - 0.161 2 8.40 FG 1.00 1.00
9,590.55
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.155 2 9.10 FCFA 1.00 1.00
8,506.85
AICD 2008 Regravel Second Transport Sector Program
Senegal Senegal 11/23/2001 - 0.300 2 10.00 FCFA 1.00 1.00
14,985.06
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.051 2 10.60 FCFA 1.00 1.00
2,421.70
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea Guinea 11/5/2002 3/26/2003 - 0.196 2 11.50 FG 1.00 1.00
8,532.42
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.247 2 13.00 FCFA 1.00 1.00
9,504.16
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 4/11/2005 9/21/2005 - 0.150 2 15.00
FCFA 1.00 1.00 4,983.96
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.280 2 16.00 FCFA 1.00 1.00
8,750.07
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea Guinea 10/22/2002 3/24/2003 - 0.228 2 16.30 FG 1.00 1.00
6,994.72
AfDB 2007 Pobe-Ketou-Illara (regraveling)
0 Benin Benin 9/1/2004 8/1/2006 4.000 4.312 2 16.50 UA 1.47 0.99
-
-
AfDB African Development Bank Group / Chief Economist
Complex
31
A f r i c a n D e v e l o p m e n t B a n k
Study on Road Infrastructure Costs:Analysis of Unit Costs and
Cost Overruns of Road Infrastructure Projects in Africa
Source Class Project Title Country/iesContractor Nationality
Dates Contract Price DimensionsOriginal Currency
Exchange Rate
USD CPI
Factor USD/lkmPAR Contract Completion PCR PAR PCR Lanes
Length
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 3/21/2005 7/6/2005 - 0.838 2 18.30
FCFA 1.00 1.00 22,894.44
AICD 2008 Regravel Second Transport Sector Program
Senegal Senegal 11/23/2001 - 0.436 2 19.70 FCFA 1.00 1.00
11,053.69
AICD 2008 Regravel Second Transport Sector Program
Senegal Senegal 11/23/2001 - 0.407 2 20.00 FCFA 1.00 1.00
10,181.31
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 3/21/2005 7/26/2005 - 0.179 2 20.50
FCFA 1.00 1.00 4,369.03
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 3/21/2005 7/6/2005 - 1.055 2 20.80
FCFA 1.00 1.00 25,358.84
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 4/11/2005 9/22/2005 - 0.298 2 23.00
FCFA 1.00 1.00 6,481.25
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea Guinea 10/28/2002 3/24/2003 - 0.277 2 25.10 FG 1.00 1.00
5,520.62
AICD 2008 Regravel Second Transport Sector Program
Senegal Senegal 11/23/2001 - 0.570 2 26.40 FCFA 1.00 1.00
10,798.40
AICD 2008 Regravel & related reinstatement
Village Communities Support Program (Phase 1)
Guinea China 10/24/2002 2/12/2004 - 0.435 2 26.60 FG 1.00 1.00
8,178.04
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.341 2 30.50 FCFA 1.00 1.00
5,583.90
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso NA 12/1/2006 - 0.489 2 35.00 FCFA 1.00 1.00
6,980.63
AICD 2008 Regravel Second Transport Sector Program
Senegal Senegal 11/23/2001 - 0.778 2 39.30 FCFA 1.00 1.00
9,899.65
AICD 2008 Regravel & related reinstatement
Transport Sector Project
Burkina Faso Burkina Faso 3/21/2005 7/21/2005 - 0.701 2 50.25
FCFA 1.00 1.00 6,972.42
AICD 2008 Regravel Burkina Faso Burkina Faso Burkina Faso
7/31/2001 - 1.501 2 75.00 FCFA 1.00 1.00 10,009.68
AICD 2008 Regravel Emergency Recovery and Rehabilitation
Project
Ethiopia Ethiopia 9/11/2003 11/29/2006 - 2.704 2 120.00 ETB 1.00
1.00 11,264.74
AICD 2008 Regravel Emergency Living Conditions Improvement
Support
DRC DRC 1/18/2006 2/27/2006 - 13.665 2 168.00 USD 1.00 1.00
40,668.29
AfDB 2007 Wacha-Maji (upgrade to gravel)
0 Ethiopia China 8/11/2003 5/1/2007 23.310 51.102 2 173.00 UA
1.53 0.97 -
AICD 2008 Regravel Emergency Living Conditions Improvement
Support
DRC DRC 12/20/2005 2/27/2006 - 16.347 2 208.00 USD 1.00 1.00
39,296.73
AICD 2008 Regravel Emergency Living Conditions Improvement
Support
DRC DRC 1/18/2006 2/27/2006 - 18.072 2 224.00 USD 1.00 1.00
40,339.22
AfDB 2010 Second Road Program
Periodic Maintenance on Earth Roads - Lot 1
Burkina Faso Burkina Faso 11/14/2001 4/28/2004 6/1/2008 9/1/2008
5.011 4.978 2 276.00 UA 1.48 1.07 14,347.32
AfDB 2010 Second Road Program
Periodic Maintenance on Earth Roads - Lot 2
Burkina Faso Burkina Faso 11/14/2001 5/6/2004 6/1/2008 9/1/2008
3.825 3.799 2 377.80 UA 1.48 1.07 7,953.41
-
32
A f r i c a n D e v e l o p m e n t B a n k
African Development Bank Group / Chief Economist Complex . May
2014 AfDBStudy on Road Infrastructure Costs:Analysis of Unit Costs
and Cost Overruns of Road Infrastructure Projects in Africa
Source Class Project Title Country/iesContractor Nationality
Dates Contract Price DimensionsOriginal Currency
Exchange Rate
USD CPI
Factor USD/lkmPAR Contract Completion PCR PAR PCR Lanes
Length
AfDB 2010 Rural Feeder Roads Maintenance Program
Rural Feeder Roads Maintenance Program
Uganda ? 10/30/1991 2/1/1996 6/1/2001 11/1/2005 8.860 8.360 2
4,003.00 UA 1.45 1.30 -
Periodic Maintenance (Paved roads)
AfDB 2010 Road Maintenance and Construction (ROMAC II)
Project
Resealing of two paved roads (Benga-Nkhotakota-Dwangwa)
Malawi UK 2/15/1990 11/1/1993 10/1/1998 10/1/2005 6.362 3.747 2
109.00 UA 1.40 1.38 33,243.18
AfDB 2010 Road Project III Resurfacing of main roads (132
km)
Kingdom of Morocco
Morocco 12/15/1994 11/1/1995 8/31/2001 9/1/2002 4.869 3.516 2
126.00 UA 1.52 1.31 27,863.11
AICD 2008 Paved: Inter-urban Periodic Maintenance
Second Transport Sector Program
Senegal Senegal 7/29/2004 11/16/2006 - 9.986 2 89.00 - 1.00 1.00
56,098.72
AICD 2008 Paved: Inter-urban Light Reconstruction
Emergency Multisector Rehabilitation amd Reconstruction Project
(EMRRP)
DRC NA 9/27/2006 - 13.526 2 93.85 FCFA 1.00 1.00 72,064.04
AICD 2008 Paved: Inter-urban Periodic Maintenance
Transport Corridors Improvement Project
Mali France 10/4/2005 3/29/2006 - 21.725 2 150.00 USD 1.00 1.00
72,418.14
AICD 2008 Paved: Inter-urban Periodic Maintenance
0 Burkina Faso Other 1/19/1999 - 19.801 2 173.00 FCFA 1.00 1.00
57,229.57
Rehabilitation (Paved roads)
AICD 2008 Paved: Inter-urban Rehab/Reconstruct
Roads and Bridges Management and Maintenance Program
Mozambique Italy 11/11/2004 10/5/2006 - 1.841 2 1.95 MT 1.00
1.00 472,098.28
AICD 2008 Paved: Urban Rehab/Reconstruct
Road Maintenance and Rehabilitation Project
Malawi Malawi 4/28/2004 - 1.192 2 2.60 MKW 1.00 1.00
229,284.64
AICD 2008 Paved: Urban Rehab/Reconstruct
Urban Development and Decentralization Project
Mali Senegal 12/29/2000 3/7/2005 - 1.753 4 1.48 FCFA 1.00 1.00
296,722.27
AICD 2008 Paved: Urban Rehab/Reconstruct
Road Maintenance and Rehabilitation Project
Malawi Malawi; Zimbabwe
10/5/2005 - 1.048 2 3.50 MKW 1.00 1.00 149,762.87
AICD 2008 Paved: Urban Rehab/Reconstruct
Road Maintenance and Rehabilitation Project
Malawi Malawi 4/11/2004 - 1.715 2 3.64 MKW 1.00 1.00
235,535.15
AICD 2008 Paved: Improvement/Upgrade
Nakivubo Channel Rehabilitation Project
Uganda Uganda 3/2/2004 10/13/2004 - 0.743 2 3.70 USH 1.00 1.00
100,338.29
AICD 2008 Paved: Inter-urban Rehab/Reconstruct
Roads Sector Invetment Program
Zambia South Africa 12/16/2003 4/12/2004 - 3.251 2 4.10 ZMK 1.00
1.00 396,479.45
AICD 2008 Paved: