hdm 4_revised_draft_final_report_v03

Government of the Federal Republic of Nigeria

Road Sector Development Team

Configuration and Calibration of HDM-4 to

Nigerian Conditions

Draft Final Report

Infrastructure Management and Engineering Services Limited Birmingham United Kingdom

Integrated Engineering Associates Limited

Kaduna Nigeria

February 2014

TABLE OF CONTENTS

i

TABLE OF CONTENTS

TABLE OF CONTENTS ................................................................................................................. i ACCRONYMS AND ABBREVIATIONS ........................................................................................iii LIST OF TABLES ......................................................................................................................... iv LIST OF FIGURES ...................................................................................................................... vi EXECUTIVE SUMMARY .............................................................................................................vii

1 INTRODUCTION ........................................................................................................................... 1

1.1 Background ...................................................................................................................... 1 1.2 Project Aim and Objectives .............................................................................................. 1 1.3 Project Scope ................................................................................................................... 2 1.4 The Study Team ............................................................................................................... 3 1.5 Purpose and Structure of the Report ................................................................................ 3

2 OVERALL STUDY METHODOLOGY ........................................................................................... 5

2.1 Introduction ....................................................................................................................... 5 2.2 Configuration .................................................................................................................... 5

2.3 Calibration ........................................................................................................................ 5 2.4 Approach and Methodology .............................................................................................. 6

2.5 Stakeholder and Consultations ......................................................................................... 8 2.6 Main Assumptions ............................................................................................................ 9

3 FIELD SURVEYS, DATA COLLECTION AND PROCESSING .................................................. 11

3.1 Vehicle Fleet Characteristics and Unit Costs .................................................................. 11 3.2 Road Network Characteristics ........................................................................................ 24 3.3 Traffic Characteristics ..................................................................................................... 27

3.4 Road Works Data ........................................................................................................... 34 3.5 Recommendation ........................................................................................................... 34

4 ROAD DETERIORATION MODEL CALIBRATION .................................................................... 36

4.1 Introduction ..................................................................................................................... 36 4.2 Climate Zones ................................................................................................................ 36 4.3 Parameter Sensitivity ...................................................................................................... 40 4.4 Roughness – Age – Environment Factor ........................................................................ 41

4.5 Cracking Initiation and Progression ................................................................................ 43 4.6 Ravelling Initiation and Progression ............................................................................... 45 4.7 Rutting ............................................................................................................................ 45 4.8 Potholing ........................................................................................................................ 48 4.9 Edge Break ..................................................................................................................... 48

4.10 Roughness ..................................................................................................................... 49 4.11 Gravel Loss on Unsealed Roads .................................................................................... 50 4.12 Rigid Concrete Pavements ............................................................................................. 51

5 ROAD WORKS EFFECTS MODEL CALIBRATION .................................................................. 52

5.1 Introduction ..................................................................................................................... 52 5.2 Effects of Road Works .................................................................................................... 52

6 ROAD USER EFFECTS MODEL CALIBRATION ...................................................................... 54

6.1 Introduction ..................................................................................................................... 54 6.2 Speed Prediction Model ................................................................................................. 54

TABLE OF CONTENTS

ii

6.3 Side Friction ................................................................................................................... 56 6.4 Fuel Consumption .......................................................................................................... 60

6.5 Spare Parts Consumption .............................................................................................. 61 6.6 Tyre Wear ....................................................................................................................... 62 6.7 Vehicle Exhaust Emissions ............................................................................................ 64

7 HDM-4 CONFIGURATION .......................................................................................................... 65

7.1 Climate Zones ................................................................................................................ 65 7.2 Traffic Flow Pattern ........................................................................................................ 65

7.3 Speed Flow Type ............................................................................................................ 66 7.4 Road Network Aggregate Data ....................................................................................... 68

8 CONCLUSION ............................................................................................................................ 75

9 REFERENCES ............................................................................................................................ 76

APPENDIX A: TERMS OF REFERENCE .................................................................................. 78 APPENDIX B: HDM-4 ANALYTICAL FRAMEWORK ................................................................. 87 APPENDIX C: MINUTES OF NEGOTIATION MEETING .......................................................... 92 APPENDIX D: OBSERVED, CALIBRATED AND DEFAULT DETERIORATION PLOTS ......... 96

ACCRONYMS AND ABBREVIATIONS

iii

ACCRONYMS AND ABBREVIATIONS

AADT Annual Average Daily Traffic ADT Average Daily Traffic AfDB African Development Bank CDB Construction Defect Indicator for the Road Base CDS Construction Defect Indicator COMP Relative compaction of the whole pavement FCT Federal Capital Territory FERMA Federal Road Maintenance Agency FGN Federal Government of Nigeria FMW Federal Ministry of Works FRSC Federal Road Safety Corps GDP Growth Domestic Product HDM-4 Highway Development and Management Tools IMES Infrastructure Management and Engineering Services Limited IQL Information Quality Level IRI International Roughness Index LTTP Long Term Pavement Performance Sites MT Motorised Traffic NBOS National Bureau of Statistics NWT Non-Work Time PCSE Passenger Car Space Equivalence RD Road Deterioration RSDT Road Sector Development Team RUE Road User Effects SEE Social-Economic Effects ToR Terms of Reference TTC Travel Time Costs VDF Vehicle Damage Factor VEDSIR Desired Speed of Travel VOC Vehicle Operating Costs WB The World Bank WE Works Effects

LIST OF TABLES

iv

LIST OF TABLES

Table 1.1: Project Scope ...................................................................................................... 2

Table 1.2: New vs Initial Schedule of Deliverables ............................................................... 2

Table 1.3: Consultant’s Team .............................................................................................. 3

Table 1.4: Counterpart Staff ................................................................................................. 3

Table 3.1: Vehicle Categories ............................................................................................ 12

Table 3.2: Basic Vehicle Fleet Characteristics.................................................................... 15

Table 3.3: Economic Costs of Vehicle Characteristics (in US Dollars) ................................ 16

Table 3.4: New and Replacement Vehicle Financial Prices ................................................ 19

Table 3.5: Existing Pavement Types .................................................................................. 25

Table 3.6: Pavement Age ................................................................................................... 26

Table 3.7: Number of Road Section ................................................................................... 26

Table 3.8: Minimum Sample Size Requirements ................................................................ 28

Table 3.9: Observed Free Speeds on Paved Roads .......................................................... 30

Table 3.10: Representative Traffic Flows on Federal Road Network .................................. 31

Table 3.11: Summary of Results of Independent Axle Load Survey ................................... 32

Table 4.1: Climate Zones ................................................................................................... 36

Table 4.2: Climate Zones Parameters ................................................................................ 39

Table 4.3: HDM-4 Sensitivity Classes ................................................................................ 40

Table 4.4: Sensitivity of Road Deterioration Models ........................................................... 40

Table 4.5: Number of Road Sections ................................................................................. 40

Table 4.6: Environmental Coefficient by Climate Zones .................................................... 42

Table 4.7: Road Construction and Drainage Effects Factor ............................................... 42

Table 4.8: Effective Environmental Coefficient ................................................................... 43

Table 4.9: Roughness-age-environment Calibration Factor (Kgm) .................................... 43

Table 4.10: Estimated Observed Time to Cracking Initiation .............................................. 44

Table 4.11: Summary of Cracking Initiation and Progression Factors ................................ 44

Table 4.12: Average Rut Depths by Climate Zone, Pavement Type and Age Group .......... 46

Table 4.13: Summary of Rutting Progression Calibration Factors ...................................... 47

Table 4.14: Average Edge Break by Climate Zone, Pavement Type and Age Group ......... 49

Table 4.15: Average Edge Break by Climate Zone, Pavement Type and Age Group ......... 49

Table 4.16: Typical Observed Gravel Loss ......................................................................... 50

Table 4.17: Typical Observed Gravel Loss ......................................................................... 50

Table 4.18: Summary of Observed and HDM-4 Predicted Gravel loss and Calibration Factor

51

LIST OF TABLES

v

Table 5.1: Summary of class, type of works (activities) and effects adapted for Nigeria ..... 53

Table 6.1: Estimated “β" Values for each Representative Vehicle ...................................... 55

Table 6.2: Vehicle Driving Power ....................................................................................... 56

Table 6.3: Rating Friction due to MT .................................................................................. 57

Table 6.4: Estimated “" Values for each Representative Vehicle ...................................... 60

Table 6.5: Parts Consumption Data and Model Calibration Factors ................................... 61

Table 6.6: Calibration Coefficient for Tyre Wear Model ...................................................... 63

Table 7.1: Traffic Flow Pattern ........................................................................................... 66

Table 7.2: Capacity and speed-flow model parameters ...................................................... 68

Table 7.3: Default Traffic Volume on Bituminous Roads .................................................... 70

Table 7.4: Default Traffic Volumes on Unsealed Roads ..................................................... 70

Table 7.5: Default Geometry Characteristics ...................................................................... 71

Table 7.6: Default Construction Defect Indicators .............................................................. 71

Table 7.7: Default SNP Values ........................................................................................... 72

Table 7.8: Default Pavement Layer Thicknesses ............................................................... 72

Table 7.9: Default Riding Quality Data ............................................................................... 72

Table 7.10: Paved Surface Condition Default Values ......................................................... 73

Table 7.11: Unsealed Surface Condition Default Values .................................................... 73

Table 7.12: Default Texture/Skid Resistance Values .......................................................... 74

LIST OF FIGURES

vi

LIST OF FIGURES

Figure 2.1: Study Methodology ............................................................................................ 7

Figure 2.2: The impact of the accuracy of data on road deterioration predictions .............. 9

Figure 2.3: Concept of Information Quality Level ................................................................ 10

Figure 3.1: Illustration of Moving Car Observer Study ........................................................ 29

Figure 3.2: Schematic Representation of ADT on Federal Road Network .......................... 33

Figure 3.3: Schematic Representation of Heavy Vehicle ADT on Federal Road Network ... 34

Figure 4.1: Map of Annual Rainfall in Nigeria including Proposed Climate Zones ............... 37

Figure 4.2: Map of Annual Minimum Temperature ............................................................. 37

Figure 4.3: Map of Annual Maximum Temperature ............................................................ 38

Figure 4.4: Plot of Environmental Coefficient with Mean Annual Rainfall ............................ 42

Figure 6.1: High level of friction .......................................................................................... 58

Figure 6.2: Intermediate level of friction, high level of road side activities ........................... 58

Figure 6.3: Intermediate level of friction due to presence of animals .................................. 59

Figure 6.4: Low level of friction ........................................................................................... 59

Figure 6.5: Standard Tyre Typology ................................................................................... 62

Figure 7.1: Illustration of Speed Flow Model ...................................................................... 67

EXECUTIVE SUMMARY

vii

EXECUTIVE SUMMARY

This is the draft final report of the study titled “Configuration and Calibration of HDM-4 to

Conditions in Nigeria”. The purpose of the study is to improve decision-making on

expenditures in the road sector in Nigeria by enabling effective and sustainable utilisation of

the latest HDM-4 knowledge. The study basis for an effective implementation of decision-

support methods and computerised tools for use by the ‘Federal Ministry of Works (FMW),

Road Sector Development Team (RSDT), Federal Road Maintenance Agency (FERMA) and

other related agencies to achieve sustainable operation of Nigerian road management system.

The adaptation of the model for Nigeria was based data from field studies carried out as well

as data from statutory agencies. Data collected and analysed for aspects of the assignment

were on climate, vehicle operating cost, traffic characteristics and on the various road

pavement types dominant on the network. Due to the nature of data available cross-sectional

method was used for the calibration of the HDM-4 model to simulate the local condition of

Nigeria. The default values have all been updated to be consistent with. The relevant

pavement deterioration factors have all been updated based what pertains in Nigeria. The

information from the road agencies on construction practices and local specification used in

the selection of the road pavement layer materials for bituminous (surface dressed and

asphaltic concrete) and unsealed roads in order to reflect local quality control regime with

respect to completed road works.

The road user effect with respect to speed flows at various periods of the day has been

established. The fuel consumption patterns of the representative vehicles a fleet under various

road conditions on road types in Nigeria were ascertained. The rate of tyre wear, fuel

consumption and vehicle maintenance with respect to spare parts usage under the respective

road condition were confirmed through field survey carried by interviewing drivers in Nigeria.

This Draft Final Report is the third project report of the series of deliverables and it covers the

period from the project start date of 15 November 2012. The project activity schedule was

revised to meet the agreed time scales of the Project. The Consultant worked closely with and

was guided by our direct client RSDT and the designated officials of the FMW and the FERMA.

This ensured that the opinions and views of all those involved in the road transport sector were

incorporated or considered in the adaptation and calibration of HDM-4 to conditions in Nigeria.

The study team applied modern experience and techniques in transport investment

appraisal/evaluation, and that encapsulated in road asset development and management

technology, and have provided a system whose results will be internationally acceptable.

The calibrated HDM-4 model can now be used to carry out specific project, programming and

strategy analysis under the operating environment and climatic condition of Nigeria.

Programming of works to be executed over a given period of time say 5 to 10 year horizon

based on current road condition for the various classes of road in under the jurisdiction of the

respective Road Agencies in Nigeria can now objectively be assessed. Strategy Analysis can

now be carried for the entire road network and use as the basis to focus budget in a

constrained budget scenario. The total budget required to fix the perennial maintenance

backlog can now be addressed and used as a basis to engage the development partners in

soliciting of loans and grants. The government outfit with ministerial oversight on the road

agencies can now use the HDM-4 analysis to direct attention in a given area of the road

network through policy.

EXECUTIVE SUMMARY

viii

In order to effectively use the calibrated HDM-4 model, base data must be updated and

sustained and a system put in place to enable annual data collection thereafter. There is the

need for staff training of the road agencies’ staff in the proper use of the HDM-4 model to

enable its use for feasibility studies, programming of works and strategic planning of the road

network. The workspace has also been customised in accordance with Nigerian local condition

and the relevant Look-up tables all reviewed.

INTRODUCTION

1

1 INTRODUCTION

1.1 Background

Following the implementation of several projects aimed at improving decision-making on expenditures in the road sector, there is currently a widespread recognition in Nigeria of the importance of road development and maintenance and the value placed on the issue both by users and the wider community. There is also an increasing understanding of the serious consequences of failure to invest adequately and effectively in maintaining the national road network. The Federal Government of Nigeria (FGN), through the Federal Ministry of Works (FMW) with financial assistance from the World Bank, procured the services of Infrastructure Management and Engineering Services Limited (IMES) to carry out Configuration and Calibration of the Highway Development and Management Model (HDM-4) to conditions in Nigeria.

Since the HDM-4 model simulates future changes to the road system from current conditions, the reliability of the results is dependent upon two primary considerations:

1. How well the data provided to the model represent the reality of current conditions and

influencing factors, in the terms understood by the model; and,

2. How well the predictions of the model fit the real behaviour and the interactions between

various factors for the variety of conditions to which it is applied

Application of the model thus involves two important steps:

(i) Data input: a correct interpretation of the data input requirements, and achieving a

quality of input data that is appropriate to the desired reliability of the results. This

includes configuration of HDM-4 and this will focus on default inputs such as vehicle

fleet, speed-flow types, traffic flow pattern, climate zones, accident rates, and the

relationships between detailed and aggregate data.

(ii) Calibration of outputs: adjusting the model parameters to enhance how well the forecast

and outputs represent the changes and influences over time and under various

interventions. Calibration of the HDM-4 model focuses on the components that

determine the physical quantities, costs and benefits predicted for the road deterioration

(RD), works effects (WE), road user effects (RUE) and Socio-Economic Effects (SEE)

analysis

The configuration and calibration has been done for the latest version of HDM-4 (Version 2.08) to suit observations, norms and practices in Nigeria.

1.2 Project Aim and Objectives

The study is aimed at improving decision-making on expenditures in the road sector by enabling effective and sustainable utilisation of the latest highway development and management knowledge. The study has developed the basis for an effective implementation of decision-support methods and computerised tools for use by the FMW, Road Sector Development Team (RSDT) and other related agencies with the aim of achieving sustainable operation.

INTRODUCTION

2

There are three levels of calibration for the HDM-4, which involves low, moderate and major levels of effort and resources, see Section 2. For this project, the target level of the configuration and calibration is Level 2 Calibration, and the calibration carried out concentrated on the most sensitive parameters embedded in the HDM-4 system.

1.3 Project Scope

The project scope is structured under seven phases described in Table 1.1. The progress of each of the phases is also provided in Table 1.1.

Table 1.1: Project Scope

Phase Description Status

1 Inception, reviews and consultations including: review of existing information and stakeholder engagement, and formulation of homogeneous groups of road sections;

Completed

2 Field surveys as necessary including: pavement performance related distresses, and RUE data

Completed

3 HDM-4 configuration including: representative vehicles, climate zones, road network aggregate data, traffic flow patter, speed flow, and accident rates/classes;

Completed

4 HDM-4 road deterioration model calibration Completed

5 HDM-4 works effects model calibration Completed

6 HDM-4 road user effects calibration including speed prediction model parameters, fuel consumption, spare parts consumption and tyre wear.

Completed

7 HDM-4 Customisation including: economic parameters, lookup tables and customised HDM-4 workspace.

Ongoing

The project deliverables include an inception report, interim report, draft final report, and final report and a customised HDM-4 workspace. The reporting schedule for the deliverables including status is provided in Table 1.2. The final report will be delivered in hard copy and softcopy on CD.

Table 1.2: New vs Initial Schedule of Deliverables

Reports Initial Timing -Months from Start of Project

New Proposed Dates Status

Inception Report 1 17th December 2012 Submitted

Interim Report 4 23rd October 2013 Submitted

Draft Final Report 7 17th December 2013 This report

Final Report 9 28th January 2014 Programmed

INTRODUCTION

3

1.4 The Study Team

1.4.1 Consultant’s Team

The consultant’s team and the key roles of each member are presented in Table 1.3.

Table 1.3: Consultant’s Team

Name Role Main Tasks

Jennaro B. Odoki Team Leader / Highway Engineer

Study approach and methodology, support on road deterioration, works effects and road user effects model calibration

Akli Ourad Transport Economist

Road user effects models and economic analysis, and provide support on road deterioration model calibration and HDM-4 configuration

Michael Anyala Road Maintenance Management Specialist

Road deterioration and works effects models, data collection and support on road user effects model calibration, and HDM-4 configuration

Isa Yunusa Chedi Environmental Management Specialist

Vehicle data and exhaust emissions models

Stephen Kinyera Otto Traffic / Road Safety Expert

Road safety and traffic flow

Nobert Omony Support Economist Data collection, traffic and economic analysis parameters

Aliyu Amiru Ubaidullah Office Administrator General clerical duties, documentation, etc.

1.4.2 Counterpart Staff

The counterpart staff assigned to this project is listed in Table 1.4.

Table 1.4: Counterpart Staff

Staff Designation

1. Chike Ngwuocha Project Manager, Road Sector Development Team

2. Victor I. Ajah Federal Roads Maintenance Agency

3. David Yiltong Federal Ministry of Works

4. Ebere Izunobi Federal Ministry of Works

1.5 Purpose and Structure of the Report

The purpose of the Draft Final Report presents the results obtained by the Study Team. After the introductory chapter, the report is structured as follows. The overall approach and the methodology adopted for this study is presented in Chapter 2. This covers the full spectrum

INTRODUCTION

4

of the work and identifies the specific methods used for each component and phase of the study.

Chapter 3 describes the activities carried out for field surveys and data collection on vehicle fleet characteristics and unit costs, road network and functional classes, pavements and traffic characteristics. Chapter 4 discusses road deterioration model calibration. Chapter 5 describes the road works effects model calibration. Chapter 6 describes the road user effects model calibration. Phase 3 of the study. Chapter 7 presents the configuration of HDM-4 to the norms, practices and conditions e Nigeria. Chapter 8 concludes the report.

OVERALL STUDY METHODOLOGY

2 OVERALL STUDY METHODOLOGY

2.1 Introduction

This chapter describes the technical approach and methodology used for this assignment. It shows not only our appreciation of the conditions and requirements of the commission but how the study team delivered the outputs stipulated in the Terms of Reference, given in Appendix A.

The study team applied modern experience and techniques in road management system development, and that encapsulated in road asset development and management technology, in order to provide a product that meets the aim and objectives of the study.

In order to appreciate the scope of work and the objectives of the study, a summary of the HDM-4 analytical framework is provided in Appendix B.

2.2 Configuration

The primary objective of configuration is to make the analysis from the model relevant and compatible to the environment in Nigeria by restructuring default configuration data in line with local conditions, standards and practices.

HDM-4 configuration involved a number of activities that included the following:

Provision of information on the climatic conditions prevailing in Nigeria, different road

types and functional classes, and the pavement types that constitute the road network.

Definition of the general characteristics of traffic flow on the different road types in the

network; the traffic bands, traffic composition by representative vehicle types and traffic

growth rates pertaining to each road type/class. Types of accidents predominant on

each road type and accident rates have to be determined.

Definition of road surface condition in aggregate form (e.g. good, fair, poor) based on

measures of surface distresses (e.g. cracking, ravelling, rutting, potholes, edge break,

roughness, thickness of gravel) to conform to local standards and practices.

General assessment of quality of road construction in Nigeria using strict adherence to

technical specifications and design standards as a measure of full compliance in order

to reflect local quality control regime.

Estimation of pavements strength of the various road types and classes expressed in

terms of structural number and deflection.

2.3 Calibration

Calibration of HDM-4 is intended to improve the accuracy of predicted pavement performance and vehicle resource consumption. The pavement deterioration models incorporated in HDM-4 were developed from results of large field experiments conducted in several countries. Consequently, the default equations in HDM-4 if used without calibration, would predict pavement performance that may not accurately match that observed on specific road sections. A fundamental assumption made prior to using HDM-4 is that the pavement performance models will be calibrated to reflect the observed rates of pavement deterioration on the roads where the models are applied. The extent of HDM-4 calibration may be defined as follows:


6

1. Level 1:

Application: Determines the values of required input parameters based on a desk study of available data and engineering experience of pavement performance, adopts many default values and calibrates the most sensitive parameters with best estimates

2. Level 2:

Verification: Requires measurement of additional inputs and moderate field surveys to calibrate key predictive relationships to local conditions

3. Level 3:

Adaptation: Experimental data collection required to monitor the long-term performance of pavements within the study area, which data should be used to enhance the existing predictive relationship or to develop new and locally specific relationships for substitution in the source code for the model

The Nigerian road agencies have been maintaining a road database for its road network for some years and stored in their respective databases. For the purpose of calibration work, the data has been retrieved from the road database and processed to suitable forms to use as input in the HDM-4 model.

2.4 Approach and Methodology

At the inception meetings with the client’s representative in October and November 2012, we discussed our methodology, timelines for the project and reporting mechanisms amongst other issues. The minutes of the meeting are attached in Appendix C.

After examination of the Terms of Reference and HDM-4 analytical framework, desk studies and inception discussions with the client, the study was divided into seven phases as follows:

Phase 1 – Inception, reviews and consultations;

Phase 2 – Field surveys;

Phase 3 – HDM-4 configuration;

Phase 4 – HDM-4 road deterioration model calibration;

Phase 5 - HDM-4 works effects model calibration;

Phase 6 – HDM-4 road user effects calibration and;

Phase 7 – HDM-4 customisation.

The study phases, main tasks and outputs are illustrated in Figure 2.1. Note that the activities in phases 4, 5 and 6 were carried out in parallel. The details of how each of these phases was completed are given in the subsequent chapters.


7

Figure 2.1: Study Methodology


8

2.5 Stakeholder and Consultations

To the extent that transport infrastructure development has a transformational effect on the

lives of all the people served, every Nigerian is a stakeholder in transport infrastructure, to

differing extents and levels. Stakeholders include both public and private sector actors.

In carrying out this assignment the Study Team consulted extensively with the stakeholders

in order to produce the desired outcome. To date the following organisations have been

consulted:

Road Sector Development Team (RSDT) – Traffic data

Federal Ministry of Works (FMW) – for road works classification, works effects and unit

costs

Federal Road Maintenance Agency (FERMA) – for road network data for road

deterioration modelling and traffic data

Federal Road Safety Corps (FRSC) – for accident and speed-flow data

National Infrastructure Agency Facilitation (NIAF) and TRL – for road condition surveys

and pavement strength data

Public-Private-Partnership Department of FMW

National Bureau of Statistics – for data on economic parameters related to road

transport, exchange rates, import and export

Vehicle Fleet Operators in Abuja – for RUC data in Abuja

Vehicle Fleet Operators in Kaduna – for RUC data in Kaduna

Vehicle Fleet Operators in Kano – for RUC data in Kano

Vehicle Fleet Operators in Lagos – for RUC data in Lagos

Vehicle Fleet Operators in Port Harcourt – for RUC data in Port Harcourt

The stakeholders and key representatives listed in Table 2.1 have been consulted in the

course of executing the project.

Table 2.1: List of key stakeholder consulted

No. Stakeholder Consulted Designation

1. Ishaq Mohammed Unit Manager, Road Sector Development Team

2. Chike Ngwuocha Project Manager, Road Sector Development Team

3. Dr Ibitoye Road Sector Development Team

4. Dr Emeka Federal Roads Maintenance Agency

5. Victor I. Ajah Federal Roads Maintenance Agency

6. David Yiltong Federal Ministry of Works

7. Ebere Izunobi Federal Ministry of Works

8. Ikene Deputy Director of Planning, Federal Ministry of Works

9. George Federal Ministry of Works

10. Dr Greg Morosiuk National Infrastructure Agency Facilitation, TRL

11. Tunde Ekunsumi Public-Private-Partnership Department

12. Dr Terry Mene Project Manager and Advisor (Safe road corridors), Federal

Road Safety Corps

13. National Bureau of Statistics


9

2.6 Main Assumptions

The main assumptions made in this study include the following:

Background information on available data and reports of relevant studies that have been completed could be used for configuration and calibration of HDM-4; Samples of additional data were collected to fill in the gaps identified in existing data.

Homogeneous groupings of road sections were used to select representative calibration road sections based on key parameters that are likely to have significant impacts on the deterioration of the Nigerian road network. The key parameters drawn from our experience in undertaking similar tasks typically included climate zones, pavement types, pavement ages and traffic levels.

The data accuracy required for HDM-4 model calibration is dictated by the level of sensitivity of the model to each parameter. For a less sensitive model there is no need to quantify the input data to a very high degree of accuracy. Conversely, for a very sensitive model it is important to quantify the data as accurately as is practical given the available resources.

Figure 2.2 illustrates the impact of the accuracy of input data on road deterioration predictions and the timing of future maintenance interventions (Bennett and Paterson, 2000). HDM-4 uses incremental-recursive models and the existing condition (denoted by point C1 or C2) is the start point for the modelling. The pavement will deteriorate and reach that condition, defined by a given set of criteria for maintenance intervention, in a certain period of time depending on the existing condition. The difference in the start point will have as great, if not greater impact, on when the treatments are triggered as will the calibrated deterioration factor.

Figure 2.2 also illustrates a second point: That HDM-4 model predictions are based on the mean deterioration rate and therefore will have a certain time interval within which a particular treatment will be triggered by a given set of intervention criteria. Typical values that define the slower and faster rates of deterioration into a band vary across the different distresses modelled. The further into the future one predicts the deterioration, the greater the spread in the trigger interval. Consequently, this will impact on the analysis results as costs incurred in the future are discounted to the base year value.

Source: Bennett and Paterson (2000) Figure 2.2: The impact of the accuracy of data on road deterioration predictions

Time

Observ

ed C

on

ditio

n

Confidence Interval

Intervention Threshold

C1

C2

Condition 1 Trigger Interval

Good

Poor

Condition 2 Trigger Interval


10

In the situations of lack or missing data, the concept of Information Quality Level (IQL) was used to configure HDM-4 aggregate data for Nigeria. The concept of Information Quality Levels (IQL) as depicted in Figure 2.3 allows data to be structured in ways that suit the needs of different levels of decision making and the variety of effort and sophistication of methods for collecting and processing data. In the IQL concept, very detailed information at a low level (low-level data) can be condensed or aggregated into progressively fewer items at successively higher levels of IQL (high-level data) as shown in Figure 2.3.

Figure 2.3: Concept of Information Quality Level

Performance

Structure Condition

Ride Distress Friction

IQL-5

IQL-4

IQL-3

IQL-2

IQL-1

System Performance Monitoring

Planning and Performance Evaluation

Programme Analysis or Detailed Planning

Project Level or Detailed Programme

Project Detail or Research

HIGH LEVEL DATA

LOW LEVEL DATA

FIELD SURVEYS, DATA COLLECTION AND PROCESSING

3 FIELD SURVEYS, DATA COLLECTION AND PROCESSING

3.1 Vehicle Fleet Characteristics and Unit Costs

Following the desk studies and reviews carried out in Phase 1, the consultant determined the availability of suitable data for calibration of RUE models. The following samples of additional data were collected to fill in the gaps identified in existing data:

Vehicle fleet resources including fuel consumption, spare parts consumption and tyre

wear collected through questionnaires and structured interviews with vehicle fleet

owners;

Vehicle fleet economic and financial costs were also be collected during the field

surveys;

Accident data on different road types and travel time values from available sources.

The data collected are held in an EXCEL database and were analysed to determine calibration factors for the different HDM-4 RUE models.

3.1.1 Representative Vehicles

As it is not possible to model the operating costs of each individual vehicle in the fleet, representative vehicles of the national fleet are used instead. These are vehicles whose characteristics can be considered to be representative of all vehicles within a certain class. The number of representative vehicles is usually influenced by factors such as the composition of traffic, functional differences between different types of vehicles, the objectives of the study, and the availability and quality of data.

There are 14 vehicle categories adopted to represent the Nigeria Fleet. These have been taken from the FHWA classification currently used by the FMW.

Class Vehicle Category

01 Motorcycles

02 Small Car

03 Medium Car

04 Large Car

05 Four Wheel Drive

06 Small Bus

07 Medium Bus

08 Big Bus

09 Light Delivery

10 Medium Delivery

11 Truck Rigid 2 Axles

12 Truck Rigid 3 or 4 Axles

13 Truck Rigid more than 4 Axles (with trailer)


12

14 Articulated Truck (divided into 4 sub-categories)

These representative vehicles are elaborated in Table 3.1.

Table 3.1: Vehicle Categories

No. Vehicle

Category Typical Engine Capacity (cc)

Typical Illustration

1 Motorcycle 125

2 Small Car 998

3 Medium Car 1400

4 Large Car 1998

5 Four Wheel

Drive 4200

6 Bus (Small) 2500

7 Bus (Medium) 4200

8 Bus(Big / Coach)

8867

9

Light Delivery Vehicle

(Utilities & Pickups)

2500

10 Medium Delivery Vehicles

2982


13

No. Vehicle

Category Typical Engine Capacity (cc)

Typical Illustration

11 Trucks (Rigid 2

- axle) 7545

12 Trucks (Rigid

3/4 - axle) 8867

13 Trucks (Rigid

& Trailer) 8867

14 (i)

Trucks (Horse & S-Trailer, 3&4 axles)

11750

14 (ii)

Trucks (Horse & S-Trailer, 5&6 axles)

11750

14

(iii)

Trucks (Horse

& S-

Trailer,7axles)

11750

14

(iv)

Trucks (Horse

& 2 Trailers) 11750

Sufficient data was collected for the customization of vehicle fleet and calibration of RUE

models. Vehicle fleet characteristics and data on unit costs of vehicle resource consumption


14

were collected through specific surveys in Abuja, Kaduna, Kano, Lagos and Port Harcourt

areas in February and March 2013. The data was collected through interviews with:

Haulage companies

Transport operators

Taxi companies

Car hire companies

Private vehicles

Other companies with important fleets (e.g. NGOs)

A full VOC survey was conducted in Abuja, Kaduna, Kano, Lagos and Port Harcourt areas. A minimum of 50 surveys were performed for each vehicle category defined in Table 3.1. The surveys were evenly spread around the five cities mentioned above. The team has processed and analysed the data. The summarised VOC data determined as a result of the survey are provided in Tables 3.2 and 3.3.

15

Table 3.2: Basic Vehicle Fleet Characteristics

Veh

icle

Nam

e

Pas

sen

ge

r C

ar

Sp

ace

Eq

uiv

ale

nce

Nu

mb

er

of

Wh

eels

Nu

mb

er

of

Ax

les

Tyre

Typ

e

Base N

um

ber

of

Recap

s

Re-t

read

Co

sts

as a

Perc

en

tag

e o

f N

ew

Tyre

Co

sts

Avera

ge N

um

ber

of

Kilo

metr

es D

riven

per

Yea

r

Wo

rkin

g H

ou

rs p

er

Yea

r

Avera

ge S

erv

ice L

ife

(Years

)

Perc

en

tag

e o

f V

eh

icle

Use o

n P

rivate

Tri

ps

Avera

ge N

um

ber

of

Pas

sen

ge

rs

Perc

en

tag

e o

f

Pas

sen

ge

r T

rip

s t

hat

are

Wo

rk R

ela

ted

Avera

ge O

pe

rati

ng

Weig

ht

(To

nn

es)

Nu

mb

er

of

Eq

uiv

ale

nt

Sta

nd

ard

Axle

Lo

ad

s

(a) Motorcycle 0.5 2 2 Bias-ply 1.3 15 10,875 1,691 10 75 1 25 0.2 0

(b) Car Small 1.0 4 2 Radial-ply 1.3 15 32,155 1,666 12 75 1.5 25 1.0 0

(c) Car Medium 1.0 4 2 Radial-ply 1.3 15 29,703 1,467 13 75 1.5 25 1.5 0

(d) Car Large 1.0 4 2 Radial-ply 1.3 15 41,459 1,479 11 75 2 25 1.8 0

(e) Four Wheel Drive 1.0 4 2 Bias-ply 1.3 15 43,298 1,076 12 0 1 0 2.0 0.02

(f) Bus Small 1.2 4 2 Radial-ply 1.3 15 120,188 2,329 11 0 10 75 2.7 0.01

(g) Bus Medium 1.5 6 2 Bias-ply 1.3 15 58,465 1,524 13 0 27 75 5.7 0.70

(h) Bus Large/Coach 1.6 10 3 Bias-ply 1.3 15 132,734 2,858 11 0 60 75 18 2.00

(i) Light Delivery Vehicle 1.0 4 2 Radial-ply 1.3 15 37,591 1,368 13 0 0 0 2.6 0.01

(j) Medium Delivery Vehicle 1.0 4 2 Bias-ply 1.3 15 42,510 1,662 9 0 0 0 4.0 0.01

(k) Truck Rigid 2-axle 1.3 4 2 Bias-ply 1.3 15 105,511 1,966 13 0 0 0 11.6 7.00

(l) Truck Rigid 3/4 Axle 1.6 10 3 Bias-ply 1.3 15 106,433 2,453 16 0 0 0 22.6 12.00

(m) Truck Multi-axle Truck & Trailer 1.8 18 5 Bias-ply 1.3 15 111,492 2,244 16 0 0 0 55.2 14.00

(n) Truck Horse & S-Trailer 3/4 Axles 1.6 10 3 Bias-ply 1.3 15 111,492 2,244 16 0 0 0 34.8 14.00

(o) Truck Horse & Semi-Trailer 5/6 Axles 1.8 18 5 Bias-ply 1.3 15 116,552 2,035 16 0 0 0 34.8 14.00

(p) Truck Horse and semi-Trailer 7 Axles 1.8 18 7 Bias-ply 1.3 15 111,492 2,244 16 0 0 0 34.8 14.00

(q) Truck Horse & 2 Trailers 1.8 18 7 Bias-ply 1.3 15 111,492 2,244 16 0 0 0 67.4 14.00


16

Table 3.3: Economic Costs of Vehicle Characteristics (in US Dollars)

Ve

hic

le N

ame

Ave

rage

Eco

no

mic

C

ost

s o

f N

ew V

ehic

le

Ave

rage

Co

sts

of

Sin

gle

Tyre

Ave

rage

Co

sts

of

Fue

l

(pe

r lit

re)

Ave

rage

Co

sts

of

Lub

rica

nts

(p

er

Litr

e)

Ave

rage

Co

sts

of

Mai

nte

nan

ce L

abo

ur

(pe

r H

ou

r)

Tota

l Cre

w W

age

s (p

er

Ho

ur)

Ave

rage

An

nu

al

Ove

rhe

ads

Co

sts

An

nu

al In

tere

st

Ave

rage

Co

sts

of

Pas

sen

ger

Wo

rkin

g Ti

me

(p

er

Ho

ur)

Ave

rage

Co

sts

of

Pas

sen

ger

No

n-

Wo

rkin

g Ti

me

(p

er

Ho

ur)

A

vera

ge C

ost

s o

f C

argo

D

ela

y (p

er

Ho

ur)

(a) Motorcycle 1342 24 0.77 1.95 1.0 0 220 11 0.91 0.18 0.00

(b) Car Small 15880 56 0.77 1.95 1.5 0 1300 11 1.10 0.22 0.00

(c) Car Medium 22998 73 0.77 1.95 1.5 0 1300 11 2.11 0.42 0.00

(d) Car Large 29295 119 0.77 1.95 1.5 0 1300 11 2.58 0.52 0.00

(e) Four Wheel Drive 33950 139 0.88 1.95 2.29 0 2400 11 2.26 0.45 0.00

(f) Bus Small 33128 139 0.88 1.95 2.29 1.69 2760 11 2.90 0.58 0.00

(g) Bus Medium 50651 134 0.88 1.95 3.87 3.23 4040 11 0.65 0.13 0.00

(h) Bus Large/Coach 175794 123 0.88 1.95 4.27 4.39 5740 11 2.86 0.57 0.00

(i) Light Delivery Vehicle 20534 141 0.88 1.95 2.58 1.39 1920 11 0.00 0.00 0.90

(j) Medium Delivery Vehicle 31759 453 0.88 1.95 2.9 2.26 2400 11 0.00 0.00 1.47

(k) Truck Rigid 2-axle 41068 324 0.88 1.95 3.55 2.74 2400 11 0.00 0.00 1.78

(l) Truck Rigid 3/4 Axle 89994 626 0.88 1.95 3.89 3.67 2660 11 0.00 0.00 2.39

(m) Truck Multi-axle Truck & Trailer 110983 503 0.88 1.95 4.75 4.54 4600 11 0.00 0.00 2.95

(n) Truck Horse & S-Trailer 3/4 Axles 132538 707 0.88 1.95 4.75 4.54 4600 11 0.00 0.00 2.95

(o) Truck Horse & Semi-Trailer 5/6 Axles 129262 522 0.88 1.95 5.23 4.9 6100 11 0.00 0.00 3.19

(p) Truck Horse and semi-Trailer 7 Axles 127624 637 0.88 1.95 5.23 4.9 6100 11 0.00 0.00 3.19

(q) Truck Horse & 2 Trailers 149022 561 0.88 1.95 5.23 4.9 6100 11 0.00 0.00 3.19


17

3.1.2 Vehicle Utilisation and Service Life

There are three vehicle utilisation parameters that have been calibrated for the Nigeria vehicle fleet.

a) Annual Utilization - Number of Kilometres Driven

b) Annual Hourly Utilization - the Number Hours per Year

c) Percentage Private Use

Vehicle annual utilization is expressed both in kilometre coverage and in hours driven per year. The kilometre coverage depends on the mobility of vehicles while the total hours include both driving time and non-driving activities: loading and unloading, refuelling, layovers, etc.

3.1.2.1 Annual Utilization - Number of Kilometres Driven

Annual kilometreage is the number of kilometres driven per year. This parameter is used in calculating the parts consumption and the interest costs. The annual kilometreage is obtained from information detailing the ages of vehicles (vehicle age spectrum) and the distances that they have travelled. The utilization of a vehicle generally varies with age. In several studies older vehicles have been found to have lower utilization than newer ones. It was therefore important that any data collected not be biased in favour of vehicles of a given age.

The consultant conducted a VOC survey in February/March 2013 and the information on annual utilization was collected directly from drivers and transport operators the information was processed and analysed to determine the average annual utilization for each vehicle category defined in Sub-section 3.1.1.

The average annual Utilization - Number of Kilometres Driven per vehicle type is provided in Table 3.2.

3.1.2.2 Annual Hourly Utilization - the Number Hours per Year

There are three definitions for the hourly utilization:

HAV - the number of hours the vehicle is available per year

This is the number of hours per year (8760), less the time allowed for crew rest, time

lost loading, unloading, refuelling, finding cargo, repairs, etc.

HRD - the numbers of hours driven

This is the hours that the vehicle is operated. It can be calculated from the annual

number of kilometres driven divided by the average annual speed.

HWK - the number of hours worked

This is similar to the hours driven (HRD), except it includes the time spent loading,

unloading and refuelling.

The utilization model in HDM-4 is based on the number of hours-worked approach. Using a standard working week, a vehicle is typically available for approximately 1800 hours per year. However, since there are substantial periods of time when the vehicle is not in use, for example due to loading/unloading, the driving time would often be less than 50 per cent of this value. Trucks and buses usually have the highest utilizations; utilities the lowest.


18

The hours driven were calculated based on data collected through the February/March 2013 VOC survey. From this, the annual average number of hours worked for each vehicle was derived. The average utilization per vehicle type is provided in Table 3.2. It was important to ensure that the value adopted for hours driven be consistent with the annual utilization in kilometres driven and the average speed. If not, the predicted costs could be distorted.

3.1.2.3 Percentage Private Use

The percentage private use was established through the VOC survey data of users. The resulting values were ultimately used in HDM-4 to calculate travel time cost. The average utilization per vehicle type is provided in Table 3.2 for vehicle in Nigeria.

3.1.2.4 Vehicle Service Life

A vehicle, or any physical property, has three measures of its life, namely the:

1. Service life defined as the period over which the vehicle is operated

2. Physical life defined as the period which the vehicle exists (even if it is not being used)

3. Economic life defined as the period which the vehicle is economically profitable to

operate

The service life is used in HDM-4 to calculate the depreciation costs of vehicles that can have a significant impact on the RUC. In HDM-4 the user needs to define the expected service life in years for a vehicle operating on a smooth pavement. This value is then used to determine the effect of roughness on service life when using the Optimal Life technique. This expected service life is the distance at which it becomes appropriate to scrap the vehicle.

There are a number of different techniques available for calculating the service life. For Level 2 calibration the ages were obtained from the VOC survey conducted as part of this study. The life of vehicle depends on road characteristics depreciate faster with less service on bad roads. Data in respect of vehicle life was collected from operators based in Abuja, Kaduna, Kano, Lagos and Port Harcourt. The data was analysed to determine the average service life of each representative vehicle type defined in Table 3.1. The average service life per vehicle type is provided in Table 3.2.

3.1.3 Parts Consumption and Maintenance Labour

Vehicle maintenance and repair costs are usually the largest single component of VOC. In HDM-4, this cost component is modelled as a function of vehicle age (expressed in terms of cumulative number of kilometres), riding quality (i.e., road roughness) and speed change cycles.

Maintenance labour costs relate to the labour component of fitting spare parts and repairing vehicles. Maintenance labour cost is affected by the road conditions since vehicles require more spare parts and frequent repairs on poor roads resulting in higher labour costs. If the road is improved to a better condition, the corresponding labour cost will be reduced. Maintenance labour costs differ for different vehicles. The type and composition of mechanics involved on the maintenance of bigger vehicles is different from smaller vehicles and accordingly, payments for mechanics differ based on their skills, extent of experiences and training.

Data on spare parts requirements and maintenance labour costs were collected from different transport operators in Abuja, Kaduna, Kano, Lagos and Port Harcourt. Calibration of parts


19

consumption model is presented in Chapter 5. The parts consumption and maintenance labour costs per vehicle type are provided in Table 3.2.

3.1.4 Vehicle Prices

Vehicle price is an important component of the Vehicle Operating Costs (VOC) as the parts consumption is expressed as a percentage of the vehicle price. The VOC survey carried out in February/March 2013 provides data to determine the current average price per vehicle category.

The financial prices of different kinds of vehicles were collected from dealers in Abuja, Kaduna, Kano, Lagos and Port Harcourt. Financial prices are the actual prices that transport operators pay for the acquisition of the vehicles. The financial prices normally include taxes and duties. However, financial prices do not reflect the real costs to the national economy where resources are actually used in providing transport services. Items such as taxes and duties are transfer payments from the private sector to the public sector and thus, they do not represent the actual consumption of resources.

Essentially for economic appraisal purposes, the costs and benefits that are associated with the proposed improvement works are expressed in resource or economic terms rather than in market prices. This normally is done to avoid distortion that results from market imperfections. In view of the above, taxes and duties are subtracted from financial prices in order to arrive at economic prices. Data on taxes and duties will be collected from the National Bureau of Statistics (NBOS) and other sources. New and Replacement financial prices are provided in Table 3.4 and economic prices are provided in Table 3.3.

Table 3.4: New and Replacement Vehicle Financial Prices

Vehicle Category Average Price (in Naira)

01 Motorcycles 173,093

02 Small Car 690.651

03 Medium Car 946,373

04 Large Car 1,950,717

05 Four Wheel Drive 3.941,860

06 Small Bus 2,846,379

07 Medium Bus 3,503,404

08 Big Bus 41,510,305

09 Light Delivery 533,544

10 Medium Delivery 1,437,181

11 Truck Rigid 2 Axles 1,476,744

12 Truck Rigid 3 Axles 9,953,771

13 Truck Rigid + 4 Axles 14,887,500

14 Multi-axle Truck 18,000,000


20

3.1.5 Standard Conversion Factor

To convert financial costs into economic costs a standard conversion factor (SCF) was derived from the following expression:

SCF = [border price value of all imports plus border price value of all exports] divided by

[(value of all imports plus all taxes on imports) plus (value of all exports minus all taxes on

exports)]

An average SCF value of 0.85 was estimated using data on Nigeria exports and imports from 2003 to 2010 obtained from National Bureau of Statistics (NBOS). The annual SCF values ranged from 0.82 to 0.89.

3.1.6 Interest Costs

The interest cost is the annual charge on the capital required to purchase the vehicle. The annual cost of capital is derived from the interest charges made by suppliers of vehicles or financiers. When dealing with financial analysis, the nominal interest rate prevailing in the market is considered, while for economic analysis interest is defined as the opportunity cost of using the money needed for purchasing vehicles in other alternative sector of the economy. The money needed for purchasing vehicles could have been used elsewhere in the economy of the country. The opportunity costs of potential usage of this money in other sectors of the economy should be calculated in economic terms. The average lending rate at the Central Bank of Nigeria has been determined from the merchant banks and provided in Table 3.3.

3.1.7 Crew Costs

Crew costs relate to payments made to crew members who travel with a vehicle to carry out different duties. The size of the crew members depends on the vehicle type and the use of the vehicle. Larger public transport vehicles usually have more crew members in addition to the drivers in order to work as conductors and assistants to the drivers.

Crew time costs are costs per crew-hour of vehicle operation. They are applicable for commercial vehicles only i.e. vans, minibuses, large buses and trucks. Pertinent data were collected in major commercial cities in Nigeria mainly in Abuja, Kaduna, Kano, Lagos and Port Harcourt. Accordingly, both financial and economic costs have been estimated from the data collected. The average crew costs are provided in Table 3.3.

3.1.8 Overhead Costs

Overhead costs include such items as garaging and insurance costs with the latter functioning as a surrogate for accident costs. For commercial operators they may also include costs of administration and support staff and office premises. Overhead costs also include the annual license fees of drivers.

Information on overhead costs was obtained from different transports operators in Nigeria based in Abuja, Kaduna, Kano, Lagos and Port Harcourt. Exact data of overhead costs is difficult to obtain in Nigeria as is common in most developing countries. The average overhead costs are provided in Table 3.3.


21

3.1.9 Fuel and Lubricants Prices

This is an important component of vehicle operating costs. Fuel consumption is affected by vehicle speed, which in turn is affected by the characteristics of road surfaces. Rise and fall as well as degrees of curvature of roads and also gross vehicle weight are among the contributing variables for fuel consumption. The two major types of fuel consumed on Nigeria roads are petrol and diesel. Data on current pump prices of petrol and diesel per litre have been collected from a number of Filling Stations and these are the financial costs.

Lubricants are consumed during the operation of vehicles. Lubricant consumption is the amount of lubricants used by a vehicle. This includes engine oil and grease measured in litres; the most important of them all is engine oil. Financial costs of lubricants were collected from fuel stations in Abuja, Kaduna, Kano, Lagos and Port Harcourt.

The fuel and lubricants prices determined by the survey are provided in Table 3.3.

3.1.10 Tyre Prices

Tyre consumption refers to the volume of tyre used per unit of distance travelled and thus, tyres are consumed continuously as vehicles travel. As the vehicles travel, materials are removed from the tyres and eventually the tyres wear out because of abrasive wear. Road characteristics and vehicle weight are the two major variables that affect the consumption of tyres. Road roughness directly affects tyre consumption because of the abrasion on the tyre surfaces by the road. Thus, the improvement in road surfaces will reduce tyre consumption; but on the other hand, the sidewalls of tyres will be damaged as a consequence of vehicle overloading.

Financial prices of tyres were collected from major commercial centres in Abuja, Kaduna, Kano, Lagos and Port Harcourt. Economic costs of tyres were derived from those of financial costs by removing taxes, duties or subsidies (if any). The tyre prices per vehicle are provided in Table 3.3.

3.1.11 Accident Costs

The HDM-4 system allows users to define a series of look-up tables for accident rates. These are basically broad, macro descriptions of the expected accident rates defined according to a particular set of road and traffic attributes. For each road type users are required to specify the accident rate for each severity (that is, fatal, injury or damage only), in terms of the numbers of accidents per 100 million vehicle-kilometres. When a road is improved a new set of accident rates can be specified based on data observed for roads with similar traffic flow and geometric characteristics. Thus, it is possible to analyse the change in total numbers of accidents and the costs resulting from the improvement.

The study team investigated accident rates on different types of road sections in Nigeria using data provided by Federal Road Safety Corps (FRSC). In particular, the team investigated how accident rates for each severity vary with parameters like road type, traffic level and flow-pattern, the presence of non-motorized transport, road geometry, and road surface characteristics. Detailed analysis of the data received from FRSC revealed that the data was not exploitable as the road accident locations were not related to road sections

Although it is not easy to attribute monetary values to the losses arising from accidents, estimates of accident costs are an essential aid to decision-making in the road safety aspects and investment choices. Costs of road accidents arise from the following areas, TRRL (1988), Overseas Road Note 5:


22

Damage to vehicles and other property

Costs of hospital treatment, police work, administration, etc.

Loss of life and injury

The first two areas of losses involve material resources and are normally readily defined, even though their values may be uncertain. They can be translated into economic terms without great difficulty. Costs relating to the loss of life and injury are subjective, involving the need to value human life and ‘pain, grief and suffering’. The valuing of human life is a difficult and often contentious process. Several methods/approaches of valuing human life exist including the gross output, net output, life insurance, court award, value of risk-charge, and implicit public sector valuation.

The study team investigated different accident cost methodologies and selected those that are relevant to the objectives being pursued by the country taking into consideration data availability and quality in Nigeria.

It is recommended that average accident costs should be as follows:

Damage only – US$ 1,000

Serious Injury – US$ 37,320

Fatal - US$ - 186,600

In 2012, GDP per capita (average income) was estimated to be US$ 1,555. Then using the 120 multiplier from Miller (2000) this equals US$ 186,600 for 2012. Allowing for a real increase of 2% in per-capita income over the previous year, economic value of preventing one road death (VSL) equals $59,600 for 2012.

The valuation of prevention of serious injury - involving police costs, hospital costs, and loss of earnings, across a wide spectrum of people, both adults and children - is of necessity simplified, and is based on a research report for TANROADS (SweRoad, 2004). Serious injury cost is estimated as 20% of the VSL, or US$ 37,320 in 2012. Note that these values should be adjusted to account for real increases in personal incomes in future years.

3.1.12 Travel Time

3.1.12.1 Passenger Travel Time

Savings in time when journeys are related to work clearly have a value; if less time is spent travelling more time in the working day can be used for economically productive purposes.

Another way of looking at this is the employer pays the employee an hour’s wages for no return. The employer would be willing to pay equal to an hour’s wages to reduce travel time by one hour. It can be argued that due to overheads and social charges the employer would be prepared to pay even more, but the common practice in developing countries is to equate the value of work time to the earnings rate of the traveller. In developed countries, where there are often large social costs, this gross wage is increased by the employer’s on-costs.

The use of wage rates is complicated by the fact that official statistics on wages probably underestimate the earnings of travellers. Wage statistics do not usually cover the earnings of the highest paid workers, and the wages of those travelling during working time may be higher than the average. Often, there are also regional variations in wages that make it impractical to adopt a national average.


23

Those in the informal sector or travelling in leisure time are not considered to be productive in the same way as those travelling in work time. Ultimately, the value of non-working time should reflect Government policy. If the policy is to maximize GDP, ignoring leisure time preferences and increasing the welfare of passengers, then a zero value should be placed on non-work time. It must be recognized that assigning a zero value to the time for those in the informal sector will serve to bias the results in favour of those who contribute to the cash economy.

There is evidence that the leisure time savings are valued, particularly since these travellers still prefer their trips to be faster than slower and are often willing to pay more for this to happen. How much a person is prepared to pay for a quicker trip is based upon their income and wealth. It is therefore common practice to assume a value of personal time related to the individual’s income. Various percentages have been assumed in different studies, usually in the range 20-50 per cent, but 20-25 per cent seems to be the most common.

Many who travel in personal time do not earn any income and so using this approach would have no value for their time. In affluent societies this would not be true, but it is argued by some that in some countries a zero value of time is appropriate.

In these instances the mean income is used to calculate the value of time. The alternative approach is to calculate the value of time based only on those working and then to apply the value to all travellers. This will yield a higher value of time than using the mean income of travellers.

To summarize, there are three sets of passenger time values to be considered:

Employed, travelling in work time,

Employed, not travelling in work time; and,

Unemployed or in non-paid activities.

There is evidence that travel time values are higher for traffic travelling under congested as opposed to free-flow conditions. When establishing the value of time, particularly for truck and bus operators, it is important to include extra income that may be obtained above the base salary, for example:

Daily allowances - to cover food and rest,

Carrying passengers (trucks) or extra, non-reported passengers (buses); and,

Backhaul of goods by trucks - where the operator instead of the owner keeps the

income.

It is also common to differentiate between modes of travel. This, for example, sees different passenger time costs for passengers in private transport and those in buses or other public transport.

Table 3.3 gives the method that was used to calculate passenger time and crew costs, using basic data from the VOC survey along with some assumed additional values. The notes at the bottom of the table detail how the individual values will be calculated.

It is widely accepted that non-working travel time cost represent around 20% of the passenger working travel time cost. This estimate was used in this study and the results obtained are provided in Table 3.3.


24

3.1.12.2 Cargo Delay Cost

Cargo delay cost is the number of vehicle-hours spent in transit. The delay of vehicles normally results in late arrival of cargoes at the destinations reducing the expected benefits to the shippers. Any decrease in duration of trips has effect on saving vehicle time. Road improvements result in the increased use of a vehicle within the same length of time in terms of distance travelled. Cargo delay cost mostly applies to trucks since these vehicles strive to increase their services by completing more trips within a given time period. It should also be considered that different kinds of cargo have different values and durability before usage. This reality will affect the delay cost that should be assigned to the vehicles ferrying the goods.

During surveys, shippers of freight were asked what their benefits would be, in monetary terms, if their cargoes arrive on time or even earlier. It was noted that they may not be able to express in quantitative terms except the subjective assessment of the benefits. It was also difficult to express in monetary values how much shippers would benefit from marginal early arrival of their goods at the prevailing slow status of economic activity in the country, except for perishable commodities.

It is widely accepted that cargo time cost represent around 1.5 – 2.0 times the passenger working travel time cost. This estimate has been used in this study and results are provided in Table 3.3.

3.2 Road Network Characteristics

3.2.1 General

The Nigerian road network from the colonial days to the present day, have been classified into three namely Trunk A, B and C (Source: State of Infrastructure Report on Nigerian Highways by MCS Consulting, et al, December 2011):

Trunk A. These roads form the skeleton of the national road grid. They cut across regional boundaries in the country and even extend to the international borders of neighbouring West African countries. These categories of roads are under Federal Government’s ownership. They are designed, constructed, maintained and financed by the Federal government through the Federal Ministry of Works. The Federal Road Maintenance Agency (FERMA) is in charge of carrying out maintenance of this class of roads.

Trunk B. These roads are local feeder roads constructed and maintained by the Works Department of Local Government Authorities in Nigeria. This class of roads are primarily not concrete asphalted and are affected by seasonal weather changes. The roads link villages and communities in the remote parts of each local government region. These roads are the second category of main roads in Nigeria. They link the major cities within States with the State capitals.

Trunk C. These roads are designed, developed, financed and maintained by the State governments through their Ministries of Works, Transport or Infrastructure. The primary objectives of Trunk B roads are to enhance the socio-economic development of the various States in the country.

The total national road network is approximately 200,000 km made up of 33,000km, 50,000km and 117,000km for Federal, State and Local Government respectively depicted by the chart below. Only about 65,000 km of the 200,000 km are paved mostly in bituminous layers others are earth roads. Out of this, the Federal Government owns about 35,000km representing 54%


25

of the entire bituminous road network in Nigeria. The balance is shared between the 36 States and the 774 Local Government Areas.

Most of the freight and passenger movement in the country are conveyed by road. Even though Federal roads constitute only 17% of the total national stock, they carry more than 80% of the National vehicular traffic, thus underscoring their crucial importance to the economy of the country. The Federal roads have been subjected to severe pressure as a result of increased vehicular traffic as well as freight especially given the near absence of rail, marine and other forms of transport to convey heavy goods.

Advancing towards a self-governing Federal Highway Authority, RSDT has a mandate to initially manage the implementation of RSDMP on behalf of the Federal Ministry of works over a 10-year period with the assistance of a USD330million credit from the International Development Association - IDA (i.e. World Bank - WB), and a USD162million loan from the African Development Bank (AfDB).

3.2.2 Pavement Types

The classified road network is made of different pavement types. Road sections with different pavement types must be analysed separately, as their performance is often different. Most importantly, in HDM-4 each pavement type has its own set of performance modelling relationships. Based on the classified road network constitution, the following HDM-4 pavement types have been adopted for this study on the basis of the type of roads encountered within the classified road network in Nigeria are bituminous and unsealed.

3.2.2.1 Bituminous Pavements

Asphalt Mix on Granular Base, referred to in the modelling as “AMGB”. This

pavement type includes roads with an asphalt concrete surfacing on granular base.

Asphalt Mix on Asphalt Base, referred to in the modelling as “AMAB”. This type of

pavement includes road with an asphalt concrete surfacing on asphalt base.

Surface Treatment on Granular Base, referred to in the modelling as “STGB”. This

pavement type includes roads with a surface dressing surfacing on a granular base.

3.2.2.2 Unsealed Pavements

Gravel, referred to the modelling as “Gravel”.

Earth, referred to the matrix identity as “Earth”.

Data samples for use in the calibration of Road Deterioration models were required for the ten pavement types presented in Table 3.5. The extent of data that were available for each pavement type is discussed in section 3.2.3 and Chapter 4.

Table 3.5: Existing Pavement Types

No. Description Notation Remarks

1 Asphalt Mix on Granular Base AMGB

2 Asphalt Mix on Asphalt Base AMAB

3 Asphalt Mix on Asphalt Pavement AMAP Bituminous road that has been overlaid with asphalt

4 Asphalt Mix on Stabilised Base

AMSB


26

No. Description Notation Remarks

5 Surface Treatment on Granular Base STGB

6 Surface Treatment on Stabilised Base

STSB

7 Surface Treatment on Asphalt Base STAB

8 Surface Treatment on Asphalt Pavement

STAP Bituminous roads that have been surface treated.

9 Rigid Concrete Pavement JPCP, JRCP, CRCP

10 Gravel Gravel

3.2.3 Pavement Ages

Five pavement age bands have been defined as shown in Table 3.6. This is useful for cross-sectional analysis of road deterioration where time series data does not exist, as it is in this study.

Table 3.6: Pavement Age

No. Description Pavement Age since construction, reconstruction or major

periodic maintenance

1 New <2 Years

2 Recent 3 to 5 Years

3 Middle 6 to 10 Years

4 Advanced 11 to 15 Years

5 Old >15 Years

3.2.4 Number of Road Sections

Pavement defects data for road calibration modelling were elicited from a total of 338 road

sections. Table 3.7 summarises the number of road sections by climate zone and pavement

type. The actual defects elicited from the road sections are discussed in Chapter 4.

Table 3.7: Number of Road Section

Pavement Type Climate Zone

1 Climate Zone

2 Climate Zone

3 Climate Zone

4

Asphalt Mix on Granular Base (AMGB)

24 85 92 68

Surface Treatment on Granular Base (STGB)

5 8 7 9

Asphalt Mix on Asphalt Pavement (AMAP)

- - 4 -

Asphalt Mix on Stabilised Base (AMSB)

- - - 36

Note to Table 3.7:


27

(1) Data for climate zones and pavement types denoted as “-“ was not available.

3.2.5 Pavement Defects

Data collected from the representative sections have been divided into two groups. The first group of data, mainly of cross-sectional nature, are being used for an initial calibration of the HDM-4 models. For this study the following data have been collected (or derived) at predefined intervals along the representative bituminous road sections:

Rut depth, measured using a 2-metre straight edge or similar equipment, in mm

All cracking, expressed as a percentage of the total carriageway area

Wide cracking (>3 mm), expressed as a percentage of the total carriageway area

Ravelling, expressed as a percentage of the total carriageway area

Potholes, in number per km

Sand patch test to determine texture depth, in mm

Deflection measurement using the Falling Weight Deflectometer (FWD)

Coring to ascertain the exact thickness of the asphaltic layers

DCP testing to establish the granular layer thicknesses and strengths, and the type

and strength of the subgrade

Trial pits to confirm DCP data

Laboratory tests on the recovered cores to establish their Marshal and elastic

properties together with the properties of the recovered bitumen.

Ride quality (roughness) expressed in terms of the International Roughness Index (IRI

in m/km).

For unsealed road sections

Ride quality (roughness) expressed in IRI m/km

Material thickness

For all sampled road sections, data on classified traffic counts have been collected or derived from secondary sources.

The second group of data is time-series based data and will be used to progressively improve the calibration of HDM-4 to Nigerian conditions. Data collected over a minimum of four years will be required for this purpose.

3.3 Traffic Characteristics

3.3.1 General

The two main sources of data on traffic characteristics used for this study were:

Traffic flow, speed and flow-pattern field surveys using non-traffic intrusive reliable

methods and SDR radar equipment;

Traffic volume, axle loading and speed-flow relationships from available sources,

especially recent study reports;


28

The following sub-sections describe the data collection methods and processing.

3.3.2 Traffic Flow, Speed and Pattern

For this study, vehicle speeds were observed on selected straight, flat, smooth, non-congested road segments of different road types (2-lane, 4-lane, 6-lane roads) for both paved and unpaved roads on the Federal road network. The survey was carried out from September to November 2013.

The free speed of travel for vehicle types were determined on a number of selected road sections on the Federal Road Network as part of the traffic flow, speed and pattern survey carried out in September to November 2013. The observed mean speeds are shown in Table 3.9.

The objective of the field study was to collect data for use in deriving the traffic flow pattern and configuring the speed-flow relationship embedded in HDM-4 to reflect local conditions, as presented in Chapter 7.

The approach used is the moving car observer method, which provides estimates of hourly volume; average travel time and space mean speed as a vehicle makes round trips through the sections being investigated. These estimates are obtained by measuring travel time, opposing vehicles, overtaking vehicles and passed vehicles.

The minimum number of runs on each study section was determined from Table 3.8. The values in the table have a confidence level of 95% and are based on desired permitted error and the average range in running speed. Robertson (1994) proposed the following error ranges for estimation of mean travel speeds for different types of studies:

Transportation planning and highway needs studies

± 3.0 mph (4.83 km/h) to ± 5.0 mph (8.05 km/h)

Traffic operations, trend analysis and economic evaluations

± 2.0 mph (3.22 km/h) to ± 4.0 mph (6.44 km/h)

Before and after studies ± 1.0 mph (1.61 km/h) to ± 3.0 mph (4.83 km/h)

The data collected in this study was used to inform economic evaluations as well as

transportation planning and highway needs, therefore, a maximum error of 6.44 km/h was recommended.

Table 3.8: Minimum Sample Size Requirements

Average range in running speed

Minimum number of runs for a permitted error of:

2.0km/h 3.5km/h 5.0km/h 6.5km/h 8.0km/h

5.0 4 3 2 2 2

10.0 8 4 3 3 2

15.0 14 7 5 3 3

20.0 21 9 6 5 4

25.0 28 13 8 6 5

30.0 38 16 10 7 6

The average range in running speed (C) in Table 3.8 is calculated using the following equation


29

𝐶 = ∑ 𝐴

𝑁−1 (3.1)

Where: C = average range in running speeds in km/h, ∑A = Sum of absolute differences

between each pair of sequential runs (km/h) N = Number of completed test runs. The average range in running speed (C) can be derived following two initial runs but a minimum of four runs is recommended.

Study Approach This approach assumes that traffic flows are in a relatively steady state condition. It is based on a survey vehicle that travels in both directions on the road. Considering Figure 3.1, the survey vehicle is driven at an average speed within the speed limit of the particular road while another observer holds a video camera and at the same time records the starting time and ending time at points A and B along the study road section. The video camera captured the opposing traffic volume, number of vehicles passed by the study car and the number of vehicles overtaken by the test car. Where a video camera was not available, counts were manually recorded on predesigned forms.

Figure 3.1: Illustration of Moving Car Observer Study

The study was undertaken on roads with two-way flows where opposing vehicles were visible at all times. The driver was able to turn around instantaneously at end of the pair run. Both volume and speed measurements are obtained simultaneously through this method. Classified counts (for each vehicle type i) should be made of vehicles met on the opposite stream (Xi), vehicles overtaking the observation vehicle (Yig), vehicles overtaken by the observation vehicle (Yin) and the journey times of the observation vehicle while travelling against and with the traffic stream.

Flows were calculated for each vehicle category for each of the three road sections for the inbound and outbound traffic streams using equation 3.2.

Study Vehicle

A

B

LOutbound

(Tw, Xi)

Inbound (Ta, Yig, Yin)


30

𝑞𝑖 = 𝑋𝑖+𝑌𝑖

𝑇𝑎+𝑇𝑤 (3.2)

where qi = flow of vehicle category i; Xi = average number of category i vehicles met by the observation vehicle when travelling against the stream; Yi = average number of category i vehicles overtaking the observation vehicle (denoted by Yig) less number overtaken (denoted by Yin) by the observation vehicle; Ta = the journey time of the observation vehicle over a section while travelling against the traffic stream; and Tw = the journey time of the observation vehicle over a section while travelling with the traffic stream.

The average speed each category of vehicle (i) for the outbound and inbound stream for each section is calculated using equations 3.3 and 3.4.

𝑣𝑖 = 𝐿

𝑇𝑖 (3.3)

where vi = average speed for category i vehicles and L = section length.

𝑇𝑖 = 𝑇𝑤 −𝑌𝑖

𝑞𝑖 (3.4)

The following resources and equipment were used for each run: Survey forms, a stop watch, a video camera (manual counts are also possible), ensure batteries are fully charged a survey vehicle with a driver, 2 observers, a length measure of the survey section. The road segment of length "L" selected was located between two junctions and a minimum length of 1 was used. At least 4 runs to and 4 runs fro were done at around 7am, 10am, 1pm, 4pm, 6pm and 9pm.

Table 3.9: Observed Free Speeds on Paved Roads

Vehicle Category

Free Speeds (in Km/hr)

Range Mean

01 Motorcycles 104 - 136 122

02 Small Car 119 - 143

136 03 Medium Car

04 Large Car

05 Four Wheel Drive 115 -148 132

06 Small Bus 89 -124 107

07 Medium Bus 80 -120

100 08 Big Bus

09 Light Delivery 80 - 122

101 10 Medium Delivery

11 Truck Rigid 2 Axles 76 - 113 95


31

Vehicle Category

Free Speeds (in Km/hr)

Range Mean

12 Truck Rigid 3 Axles 70 -98

84 13 Truck Rigid + 4 Axles

14 Multi-axle Truck

3.3.3 Traffic Counts

Existing traffic data on the Federal Road Network was obtained from the RSDT and FMW. This consisted of estimates of Average Daily Traffic (ADT) on numerous points on the network. For the purpose of this Study the estimate of ADT obtained from the short term counts was assumed to represent the AADT. This assumption is unlikely to adversely affect the results of the Study.

From the information obtained, representative traffic flows on each of the major links in the Federal Road Network were determined. The analysis of these provided traffic flows on the basis of weighted average flows on the numerous segments in each of the main links. The weighted average figures on the defined links are used to illustrate the major link flows and relative importance of these. Figure 3.2 presents these weighted average link flows.

Table 3.10: Representative Traffic Flows on Federal Road Network

Major Federal Road Link ADT Heavy Vehicles /Day

Lagos - Shagamu 40,000 5,000

Shagamu - Benin City 22,000 3,100

Shagamu - Ibadan 8,900 2,800

Benin - Warri - Port Harcourt 5,000 350

Port Harcourt - Aba 18,000 2,200

Aba - Enugu 12,000 2,000

Aba - Nlagu 8,900 2,000

Nlagu - Calabar 4,500 1,000

Enugu - Nkalagu 6,000 1,209

Nkalagu - Mfom 4,000 500

Benin City - Onitsha 14,500 2,500

Onitsha - Enugu 18,000 1,500

Benin City - Lokoja 7,300 1,000

Ibadan - Ilorin 10,000 2,500

Ilorin - Jebba 5,000 2,200

Mokwa-Bida 4,500 1,500

Bida - Abuja 2,200 550


32

Major Federal Road Link ADT Heavy Vehicles /Day

Lokoja - Abuja 9,000 900

Abuja - Kaduna 8,000 800

Abuja - Akwanga 5,700 250

Enugu - Makurdi 6,000 920

Makurdi - Akwanga 6,200 400

Akwanga - Jos 4,000 220

Jos - Bauchi 7,000 380

Bauchi - Yola 4,200 370

Kaduna - Zaria 11,000 920

Zaria - Sokoto 5,100 420

Sokoto - Illela 3,000 100

Zaria - Kano 10,000 700

Kano - Katsina 5,600 630

Kano - Potisku 4,000 300

Potisku - Maiduguri 5,000 920

Maiduguri - Ngala 3,000 1,000

Source: SSI Engineers and Environmental Consultants (Pty) Ltd., 2008

These figures have been obtained as a weighted average and have been “rounded” to representative values.

Schematic representation of the above ADT and heavy vehicles per day is provided to give visual illustration of the traffic and heavy vehicle distribution throughout the country. These are illustrated in Figure 3.2 and 3.3 respectively.

An Independent axle load survey was conducted, in which 1,350 heavy vehicles were weighed in 16 surveys conducted at 8 points on the Federal Road Network. The results of this survey formed the basis for all the proposals in this report and are summarised in Table 3.11.

Table 3.11: Summary of Results of Independent Axle Load Survey

Heavy Vehicle Class

No. counted

No. weighed

No. of Overloaded

vehicles

% vehicles Overloaded

No. of Overloaded

axles

Average % Overloaded per axle

2 Axle 1479 316 97 30.7 114 36.1

3 Axle 656 162 100 61.7 192 57.4

4 Axle 1696 691 407 58.9 1112 51.1

5 Axle 648 166 89 53.6 287 46.8

6 & more Axle

111 15 7 46.7 29 35.4

Totals 4590 1350 700 51.9 1734 48.8


33

If ECOWAS limits are used instead of the official legislated limits, the number and extent of the overloading changes slightly, bringing the overloading down to approximately 684 of the sampled vehicles, but with more of the axles of those vehicles overloaded, at 1782. Table 3.14 shows that 3 axle and 4 axle vehicles are more likely to be overloaded and recorded a higher average overload per axle. This information was used to determine the equivalent axle load factors, ESAL, for each representative vehicle type given in Table 3.1.


Figure 3.2: Schematic Representation of ADT on Federal Road Network

Mfum

Enugu

Benin City

Abuja

Kano

NIGER

CAMEROON

CHAD

BENIN

Kaduna

Katsina

Illela

Sokoto

Mokwa

Maiduguri

Yola

Lagos

Calabar

Port Harcourt

Akwanga

BIGHT OF BENIN

Shagamu

TRAFFIC BAND WIDTH SCHEMATIC

Zaria

Ilorin

Ngala

BauchiJos

MakurdiLokoja

40,000

Thickness = VPD

20,000

10,000

5,000

2,000

LINK TRAFFIC KEY

Onitsha

Potisku


34


Figure 3.3: Schematic Representation of Heavy Vehicle ADT on Federal Road Network

The discontinuities in the heavy vehicle flows highlighted by the arrows in Figure 3.3 represent areas in which the Federal Road Network is probably not the preferred route used by the heavy vehicles on the particular link in question.

3.4 Road Works Data

Road works data have been collected from FMW and the Federal Road Maintenance Agency (FERMA) for bituminous and unsealed surface classes. The information collected included different road works activities under routine and periodic maintenance, improvement and development works, and special works. Data have also been collected on maintenance history for various treatment types, and the condition of the road pavement before and after treatment. This information was used to calibrate HDM-4 works effects models for different types of treatments such as new construction or reconstruction, overlay, and reseal. The financial unit costs of road works items were also collected and expressed in economic terms.

3.5 Recommendation

In our comments on the ToR, we proposed that RSDT should set up Long-Term Pavement Performance sites at the representative sections which we would identify under this commission (refer to minutes of the inception included in Appendix C of this report). Such sites should be clearly identified and permanent warning signs of conventional nature should be placed at both ends of the section to inform the public that these are research sites that will be subjected to frequent inspections and there may be delays in carrying out maintenance works on these sections. If RSDT decide to extend the scope of this work to include the setting of LTTP sites, then the second group of data will be collected for use in establishing sets of time-series data from the selected representative road sections. These sets of data will subsequently be used to refine the calibration and achieve full adaptation of HDM-4 to Nigerian conditions. Regular monitoring of the selected calibration sections should be carried

Lokoja

1000

Mfum

Enugu

Benin City

Abuja

Kano

NIGER

CAMEROON

CHAD

BENIN

Kaduna

Katsina

Illela

Sokoto

Jebba

Maiduguri

Yola

Lagos

Calabar

Port Harcourt

Akwanga

BIGHT OF BENIN

Shagamu

HEAVY VEHICLE TRAFFIC BAND WIDTH SCHEMATIC

Zaria

Ilorin

Ngala

BauchiJos

Makurdi

5,000

Thickness = HVPD

2,000

1,000

500

200

LINK HEAVY VEHICLE

TRAFFIC KEY

Potiskum

4,000

5,00

0

2800

2200

3,100 1000

900

570

900

800

900

420

100

630

300920

1000

25001500

350

2200 1000

920

400

250

220

380

3701500


35

out; initially this may be conducted biannually, but subsequently dropping to annually once the basic patterns of behaviour have been established.

ROAD DETERIORATION MODEL CALIBRATION

36

4 ROAD DETERIORATION MODEL CALIBRATION

4.1 Introduction

The HDM-4 road deterioration models have a number of adjustment factors (or parameters) and it is important to acknowledge the general level of sensitivity of the model to each parameter so that appropriate emphasis can be given to important parameters and less emphasis to second or third order effects. The influences of individual parameters differ according to the nature of the parameter, the result being considered, and the values assigned to other parameters for a given analysis. The sensitivity of results to variations in a parameter therefore varies somewhat under different circumstances. For a detailed treatment of this issue refer to Volume 5 of HDM-4 documentation series (Bennett and Paterson, 2000).

4.2 Climate Zones

Climate has a significant influence on road deterioration. In HDM-4, climate zones are defined using temperature and moisture classifications. The rainfall map for Nigeria given in Figure 4.1 and temperature maps in Figures 4.2 and 4.3 were used to define climate zones for the adaptation and calibration of HDM-4 to Nigerian conditions.

The maps suggest that the temperature ranges in Nigeria largely fall within the Tropical temperature classification with temperature ranging between 18oC and 35oC. Delineation of the country into climate zones was therefore largely influenced by the large variation in annual rainfall. On this basis, four (4) climate zones described in Table 4.4 and illustrated in Figure 4.1 are proposed.

Table 4.1: Climate Zones

Climate Zone Description

Zone 1 areas receiving average annual rainfall in the range of 400 – 800 mm per year);


Zone 3 areas receiving average annual rainfall in the range of 1200 – 1800 mm per year; and

Zone 4 areas receiving average annual rainfall in the range of >1800 mm per year


37

Figure 4.1: Map of Annual Rainfall in Nigeria including Proposed Climate Zones

Figure 4.2: Map of Annual Minimum Temperature


38

Figure 4.3: Map of Annual Maximum Temperature

Climate parameters specified for each zone for use in the HDM-4 model are provided in Table 4.2.


39

Table 4.2: Climate Zones Parameters

Clim

ate

Zon

e

Description

Mo

istu

re C

lassific

ation

Tem

pera

ture

Cla

ssific

atio

n

Nu

mb

er

of D

ays w

he

n

Tem

pera

ture

Exce

ed

s 3

2oC

Ave

rag

e T

em

pera

ture

Ra

ng

e

(oC

)

Fre

eze In

de

x

Mo

istu

re In

de

x

Me

an

Mo

nth

ly P

recip

itatio

n

Me

an

Tem

pera

ture

Du

ratio

n o

f D

ry S

easo

n

Pe

rce

nta

ge o

f D

rivin

g o

n

Sn

ow

Co

vere

d R

oad

s

Pe

rce

nta

ge o

f D

rivin

g D

one

on W

ate

r C

overe

d R

oad

s


Semi-arid Tropical 136 3 0 -40 57.5 26 8 0 5


Sub-humid

Tropical 94 4 0 0 72 26 9 0 10

Zone 3 areas receiving average annual rainfall in the range of 1200 – 1800 mm per year; and

Sub-humid

Tropical 179 9 0 0 95 23 6 0 15

Zone 4 areas receiving average annual rainfall in the range of >1800 mm per year

Humid Tropical 104 1 0 60 246 26 3 0 20


40

4.3 Parameter Sensitivity

HDM-4 deterioration models are categorised according to the sensitivity classes summarised in Table 4.3. This study aimed to calibrate model coefficients with high to moderate impacts (Sensitivity classes I and II) as well as models for defects such as potholes and edge break that have low impact but are prevalent on roads in Nigeria. The sensitivity of road deterioration models are summarised in Table 4.4.

Table 4.3: HDM-4 Sensitivity Classes

Impact Sensitivity Impact Elasticity

High S-I >0.5

Moderate S-II 0.2 – 0.5

Low S-III 0.05 – 0.2

Negligible S-IV <0.05

Table 4.4: Sensitivity of Road Deterioration Models

Description Sensitivity to Outputs

Roughness-age-environment factor High

Cracking initiation High

Cracking progression High

Rut depth Low

Potholing Low

Edge break Low

The general roads roughness progression factor has low priority, despite its moderate sensitivity, because its range is small based on many inter-country validation studies conducted. The following sections describe the calibration of these HDM-4 deterioration models to Nigeria roads conditions using the data that were available. Table 4.5 provides a summary of the number of road sections used to inform the calibration for each pavement defect. The observed defects are shown in the plots in Appendix D.

Table 4.5: Number of Road Sections

Climate Zone Pavement Type Rut Depth Edge Break Cracking

1 AMGB 24 24 12

STGB 5 5 4

2 AMGB 85 85 51

STGB 8 8 8

3

AMAP 4 4 4

AMGB 92 92 28

STGB 7 7 7

4

AMGB 68 68 51

AMSB 36 36 36

STGB 9 9 9


41

4.4 Roughness – Age – Environment Factor

The roughness-age-environment factor determines the amount of roughness progression occurring annually on a non-structural time-dependent basis (due to environmental effects). The factor adjusts the environment coefficient, m, which has a base value of 0.023 in the model, representing 2.3 percent annual change in roughness that is independent of traffic as illustrated in Equation 4.2.

∆𝑅𝑡𝑒 = 𝐾𝑔𝑚0.023𝑅𝑡 (4.1)

Where Kgm is the roughness age-environment calibration factor, Rte is the change in the roughness component due to environment in the 1st-year of analysis time increment and Rt the roughness at the beginning of the year.

Sufficient data were not available to allow a detailed calibration to Level 2 hence Level 1 calibration was performed in accordance with guidance provided by Bennett and Paterson (2000). This involved establishing Kgm values for each climate zone and ranges of road construction material standards and drainage maintenance standards practiced in Nigeria in accordance with the following steps set out in Volume 5 of the HDM-4 documentation:

Step 1 Identify environment applicable to roads in Nigeria and select appropriate values of environmental coefficient m;

Step 2 Adjust the environmental coefficient (m) by multiplying with a factor Km which represents the standard of road construction and drainage maintenance in Nigeria to give an effective m-value, meff; and

Step 3 Calculate Kgm from meff as follows:

𝐾𝑔𝑚 = 𝑚𝑒𝑓𝑓

0.023 (4.2)

Step 1: Environment Coefficient m

Figure 4.4 adapted from Bennett and Paterson (2000) shows the variation of the environmental coefficient in semi-arid, sub-humid and humid tropical climate zones which are typical of conditions in Nigeria. The environmental coefficient (derived from Figure 4.4) corresponding to the climate zones in Nigeria (discussed in section 4.2) are summarized in Table 4.6.


42

Figure 4.4: Plot of Environmental Coefficient with Mean Annual Rainfall

Table 4.6: Environmental Coefficient by Climate Zones

Nigerian Climate Environmental

Coefficient (m) Climate Zone Moisture

Classification

Typical Annual

Rainfall (mm)

Zone 1 Semi – arid 690 0.011

Zone 2 Sub-humid 864 0.017

Zone 3 Sub-humid 1140 0.022

Zone 4 Humid 2952 0.025

Step 2: Effective Environment Coefficient meff

The standard of road pavement construction and drainage in Nigeria ranges from the use of high standard materials to materials of variable quality. To account for the effect of this variability, the environmental coefficients m given in Table 4.7 were adjusted by multiplying with factors provided in Table 4.8 for high, normal, and variable road construction and drainage standards to give a climate zone and construction standard matrix of effective environmental coefficient summarized in Table 4.9.

Table 4.7: Road Construction and Drainage Effects Factor

Standard Description Factor

Km

High Standard: Materials and drainage of high engineering standards

Motorways

Raised formation

Free draining materials

Special drainage facilities

0.6

Normal Standard: Material quality to normal engineering standards

Drainage and formation adequate for local moisture conditions

Drainage moderately maintained

1.0

Variable Standard: Variable material quality in pavement

Including moisture susceptible material

Inadequate drainage or poorly maintained

Formation near water table

1.3


43

Table 4.8: Effective Environmental Coefficient

Climate Zones Construction and Drainage Standards

High Normal Variable

Zone 1 0.007 0.011 0.014

Zone 2 0.010 0.017 0.022

Zone 3 0.013 0.022 0.029

Zone 4 0.015 0.025 0.033

Step 3: Roughness-age-environment Calibration Factor

The roughness-age-environment calibration factor determined for each pavement and construction and drainage standard using Equation 4.2 are given in Table 4.9.

Table 4.9: Roughness-age-environment Calibration Factor (Kgm)

Climate Zones Construction and Drainage Standards

High Normal Variable

Zone 1 0.287 0.478 0.622

Zone 2 0.443 0.739 0.961

Zone 3 0.574 0.957 1.243

Zone 4 0.652 1.087 1.413

4.5 Cracking Initiation and Progression

HDM-4 models the time to initiation of cracking and the progression of the cracking using two separate prediction models. Cracking initiation is predicted in terms of the surfacing age when first visible crack appears on the road surface. Cracking is deemed to have started at the age of the surface. HDM-4 effectively initiates cracking when 0.5% of the carriageway surface area is cracked. During the progression phase, cracking gradually spreads to cover, eventually, the entire pavement area if not treatment is applied.

While HDM-4 models both all and wide cracking, only data for all cracking was available. Consequently wide cracking is not considered in this report. Table 4.10 gives the estimated average observed time to cracking initiation by pavement time and climate zone.

Ideally the calibration of cracking initiation and progression should cover the complete spectrum of pavement types and climate zones that exist on the Nigerian road network. Data necessary to facilitate this detailed level of calibration was not available for use in this study, instead an average observed time to cracking initiation on paved National roads of 5 years estimated from limited data was used to derive calibration factors by pavement type and climate zones.


44

Table 4.10: Estimated Observed Time to Cracking Initiation

Climate Zone

Pavement Type

AMAB1 AMAP2 AMGB AMSB3 STGB STSB4

Zone 1 6.5 6.5 7.5 6.1 6 6

Zone 2 6.5 6.5 6.5 6.1 6.1 6.1

Zone 3 6.5 6.5 6 6.1 5.8 5.8

Zone 4 6.5 6.5 5.8 6.1 5 5

Notes to Table 4.10:

1. Data for Asphalt Mix on Asphalt Base (AMAB) pavement type was not available, time to cracking initiation for this pavement type was assumed the same as the observed cracking initiation time for Asphalt Mix on Asphalt Pavement (AMAP).

2. Due to lack of data, time to cracking initiation on AMAP for Zones 1, 2 and 3 were taken to be the same as the cracking initiation time determined for Zone 3.

3. Due to lack of data, time to cracking initiation on AMAP for Zones 1, 2 and 4 were taken to be the same as the cracking initiation time determined for Zone 4.

4. Data for estimating the average time to cracking initiation for Surface Treatment on Stabilised Base (STSB) was not available, hence values for Surface Treatment on Granular Base (STGB) were used.

The default HDM-4 cracking initiation calibration factor (Kcia) of 1 was adjusted to achieve estimated time to cracking initiation in Table 4.10, results are summarized in Table 4.11. The calibration factor for cracking progression model (Kcpa) were generally calculated as the inverse of the cracking initiation factor according to Bennett and Paterson (2000), these factors are summarized in Table 4.11.

Table 4.11: Summary of Cracking Initiation and Progression Factors

Pavement Type

Climate Zone 1

Climate Zone 2

Climate Zone 3

Climate Zone 4

Kcia Kcpa Kcia Kcpa Kcia Kcpa Kcia Kcpa

Asphalt Mix on Asphalt Base (AMAB)

1.18 0.85 1.22 0.82 1.23 0.81 1.50 0.67

Asphalt Mix on Asphalt Pavement (AMAP)

1.11 0.90 1.14 0.88 1.1 0.91 1.21 0.83

Asphalt Mix on Granular Base (AMGB)

1.36 0.74 1.13 0.88 1.15 0.87 1.56 0.64

Asphalt Mix on Stabilised Base (AMSB)

0.31 3.23 0.30 3.33 0.38 2.63 0.08 1.00


45

Pavement Type

Climate Zone 1

Climate Zone 2

Climate Zone 3

Climate Zone 4

Kcia Kcpa Kcia Kcpa Kcia Kcpa Kcia Kcpa

Surface Treatment on Granular Base (STGB)

0.7 1.43 0.72 1.39 0.7 1.43 0.60 1.00

Surface Treatment on Stabilised Base (STSB)

0.75 1.33 0.80 1.25 0.8 1.25 0.67 1.49


1. Kcia = Calibration factor for Cracking Initiation model

2. Kcpa = Calibration factor for Cracking Progression model

3. Climate zones are defined in sub-section 4.2 of this report.

4.6 Ravelling Initiation and Progression

The adjustment factor for ravelling initiation and progression has low impact on most applications, since observed ravelling data was not available, it is reasonable to retain the default value of 1 for it.

4.7 Rutting

There are four components of rutting modelled in HDM-4, namely:

Initial densification;

Structural deformation;

Plastic deformation; and

Surface wear due to studded tyres.

Initial densification is a function of degree of relative compaction of the base, subbase and selected subgrade layers. The effect is predominant during the first year of the pavement life. Observed data was not available so the initial densification calibration factor Krid was set to the default value of 1.

Plastic deformation occurs in thick bituminous roads (asphalt) under high temperature and heavy traffic loading and stationary or slow moving loads as experienced at road intersections and on heavily traffic roads. The default calibration factor (Krpd) for plastic deformation was used due to lack of data for the calibration. This factor can be adjusted on particular roads that exhibit this defect, e.g. at the project level.

The structural wear due to studded tyres component of rutting is not applicable in Nigeria and thus not considered.

Progression of rutting was therefore modelled using the structural deformation model. Average observed rutting trend was derived from cross-sectional analysis of rutting data and pavement age by climate zone. Table 4.12 sets out the average observed rutting data used for the calibration.


46

Table 4.12: Average Rut Depths by Climate Zone, Pavement Type and Age Group

Climate Zone

Pavement Type

Sections (Number)

Rut Depth (mm) by Pavement Age Group (Years)

New (<2 Years)

Recent (3-5 Years)

Middle (6-10 Years)

Advanced (11-15 Years)

Old (>15 Years)

1 AMGB 24 0.9 3.6 8.2 21.5 26.6

STGB 5 4.3 2.5 7.7 - 22.25

2 AMGB 85 2.5 3.3 7.7 14.5 26.2

STGB 8 - 4.1 - - 27.4

3

AMAP 4 - 4.6 8.8 - 28.2

AMGB 92 1.1 4.3 7.2 13.4 23

STGB 7 - 4.8 7.8 - 24.5

4

AMGB 68 1 3.5 7.4 11.9 25.9

AMSB 36 4.3 8.8 5.4 12.5 24.1

STGB 9 - 3.5 7.5 11.9 23.9


1. ‘–‘ denotes age groups for which average observed rutting data were not available;

The following assumptions were used to estimate observed rut depth trends on representative calibration sections with missing data:

Climate Zone 1: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB), Asphalt Mix on Asphalt Pavement (AMAP) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed rut depth trends on these sections were assumed on average to be similar with observations within Climate Zone 1 for Asphalt Mix on Granular Base (AMGB). Furthermore, representative rut depth data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 1 for Surface Treatment on Granular Base (STGB);

Climate Zone 2: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB), Asphalt Mix on Asphalt Pavement (AMAP) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed rut depth trends on these sections were assumed on average to be similar with observations within Climate Zone 2 for Asphalt Mix on Granular Base (AMGB). Representative rut depth data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 2 for Surface Treatment on Granular Base (STGB);

Climate Zone 3: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed rut depth trends on these sections were assumed on average to be similar with observations within Climate Zone 3 for Asphalt Mix on Granular Base (AMGB). Representative rut depth data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 3 for Surface Treatment on Granular Base (STGB); and


47

Climate Zone 4: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB) and Asphalt Mix on Asphalt Mix on Asphalt Pavement (AMAP) were not available, observed rut depth trends on these sections were assumed on average to be similar with observations within Climate Zone 4 for Asphalt Mix on Granular Base (AMGB). Representative rut depth data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 4 for Surface Treatment on Granular Base (STGB).

Calibration was performed by climate zone since structural rutting is affected by cracking and subsequently rainfall. Table 4.13 gives the derived rutting progression calibration factors which were determined as the ratio of the slope of the observed rut depth trend to the slope of the trend predicted by HDM-4 using a default calibration factor of 1.

Table 4.13: Summary of Rutting Progression Calibration Factors

Climate Zone

Pavement Type Slope (Observed) Slope (Predicted) Calibration Factor

(Krst)

1

AMAB 1.278 0.234 5.46

AMAP 1.278 0.23 5.56

AMGB 1.278 0.228 5.61

AMSB 1.278 0.257 4.97

STGB 1.012 0.271 3.73

STSB 1.012 0.271 3.73

2

AMAB 1.149 0.238 4.83

AMAP 1.149 0.234 4.91

AMGB 1.149 0.239 4.81

AMSB 1.149 0.282 4.07

STGB 1.363 0.277 4.92

STSB 1.363 0.281 4.85

3

AMAB 1.323 0.256 5.17

AMAP 1.323 0.251 5.27

AMGB 1.085 0.269 4.03

AMSB 1.085 0.33 3.29

STGB 1.215 0.312 3.89

STSB 1.215 0.317 3.83

4

AMAB 1.323 0.298 4.44

AMAP 1.323 0.322 4.11

AMGB 1.01 0.41 2.46

AMSB 1.137 0.727 1.56

STGB 1.065 0.544 1.96

STSB 1.065 0.577 1.85


48

4.8 Potholing

According to the HDM-4 Calibration Manual, the calibration factor for pothole progression has a low impact on maintenance alternatives except patching, it is therefore reasonable in most cases to adopt the default value of 1 in most cases.

If patching alternatives are expected to dominate and the prediction of potholes using the default factor 1 is significantly different from observations, then the calibration factor should be adjusted. Sufficient observed data was not available to allow such assessments and calibration to be performed.

4.9 Edge Break

Edge break refers to the loss of surface, and possibly base materials from the edge of the pavement. It is prevalent on roads in Nigeria particularly where unpaved shoulders exist and traffic activities are high. Time series records of age break on Nigerian roads were not available, to this end, information elicited from various states for selected representative road sections were used to inform the calibration. The data were banded by climate zone, pavement types and age groups and is summarised in Table 4.14 following cross-sectional analysis.

The following assumptions were used to estimate observed edge break trends on representative calibration sections with missing data:

Climate Zone 1: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB), Asphalt Mix on Asphalt Pavement (AMAP) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed edge break trends on these sections were assumed on average to be similar with observations within Climate Zone 1 for Asphalt Mix on Granular Base (AMGB). Furthermore, representative edge break data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 1 for Surface Treatment on Granular Base (STGB);

Climate Zone 2: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB), Asphalt Mix on Asphalt Pavement (AMAP) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed edge break trends on these sections were assumed on average to be similar with observations within Climate Zone 2 for Asphalt Mix on Granular Base (AMGB). Representative edge break data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 2 for Surface Treatment on Granular Base (STGB);

Climate Zone 3: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB) and Asphalt Mix on Stabilised Base (AMSB) were not available, observed edge break trends on these sections were assumed on average to be similar with observations within Climate Zone 3 for Asphalt Mix on Granular Base (AMGB). Representative edge break data for Surface Treatment on Stabilised Base (STSB) were assumed to be similar to that observed within Climate Zone 3 for Surface Treatment on Granular Base (STGB); and

Climate Zone 4: representative data for the pavement types Asphalt Mix on Asphalt Base (AMAB) and Asphalt Mix on Asphalt Mix on Asphalt Pavement (AMAP) were not available, observed edge break trends on these sections were assumed on average to be similar with observations within Climate Zone 4 for Asphalt Mix on Granular Base (AMGB). Representative edge break for Surface Treatment on Stabilised Base (STSB)


49

were assumed to be similar to that observed within Climate Zone 4 for Surface Treatment on Granular Base (STGB).

Table 4.15 gives the derived edge break calibration factors for each pavement type by climate zone.

Table 4.14: Average Edge Break by Climate Zone, Pavement Type and Age Group

Climate Zone

Pavement Type

Sections (Number

)

Edge Break (m/km) by Pavement Age Group

New (<2 Years)

Recent ( 3 - 5 Years)

Middle (6 - 10 Years)

Advanced (11 - 15 Years)

Old (>15 Years)

1 AMGB 24 0 0 2.8 6.1 36.9

STGB 5 0 0 2.9 - 34

2 AMGB 85 0 0 2.8 13 33.3

STGB 8 - 0 - - 40.4

3

AMAP 4 - 0 4.2 - 46.1

AMGB 92 0 1.3 2.5 11 29.7

STGB 7 - 0 0 - 35.4

4

AMGB 68 0 0 1.9 8.9 33.8

AMSB 36 0 4.1 0.9 8.8 32.2

STGB 9 - 0 2.3 8.4 27.3

Table 4.15: Average Edge Break by Climate Zone, Pavement Type and Age Group

Pavement Type

Edge Break Calibration Factor (Keb) by Climate Zone

Zone 1 Zone 2 Zone 3 Zone 4

Asphalt Mix on Asphalt Base (AMAB) 1.296 1.216 1.053 0.654

Asphalt Mix on Asphalt Pavement (AMAP) 1.365 1.286 1.109 0.64

Asphalt Mix on Granular Base (AMGB) 0.587 0.51 0.41 0.275

Asphalt Mix on Stabilised Base (AMSB) 1.001 0.82 0.72 0.485

Surface Treatment on Granular Base (STGB) 0.302 0.297 0.26 0.211

Surface Treatment on Stabilised Base (STSB) 0.451 0.443 0.377 0.233

4.10 Roughness

Calibration of roughness progression on paved roads was not performed because of its dependence on the various types of calibrated models described in previous sub-sections. Further adjustment would require time series data that were not readily available for use in this study.


50

4.11 Gravel Loss on Unsealed Roads

Gravel loss is defined as the change in gravel thickness over a period of time. It is used to estimate when the thickness of the gravel wearing course has decreased to a level where regravelling is necessary. Annual quantity of gravel loss is predicted by HDM-4 as a function of monthly rainfall, traffic volume, road geometry and characteristics of the gravel.

The calibration of the gravel loss deterioration model requires data on the properties of typical unsealed surfacing and subgrade material typical of Nigeria together with recorded data on gravel loss which are normally obtained over a number of years following field observations. Table 4.16 summarises average material properties (provided by the Ministry of Works) for three climate zones (see section 4.2). Data for climate zone 1 was not available.

Table 4.16: Typical Observed Gravel Loss

Pavement Layer Climate

Zone

Maximum Particle

Size

Plasticity Index

Percentage Passing Sieve Size

2 0.425 0.075

Surfacing

Zone 2 10.1 13.8 81.2 44.5 30

Zone 3 12.6 12.2 68 49.3 33

Zone 4 17 25 61 41 27

Subgrade

Zone 2 6.1 8.2 88 58.1 30.4

Zone 3 11.5 23 77 59 51

Zone 4 3.1 5.4 86 57 15.7

Regravelling interval in Nigeria typically varies between 3 to 8 years. Based on this interval, typical scenarios of gravel loss rates given in Table 4.17 were derived on the basis that the surfacing thickness of newly regravelled unsealed roads is 150mm.

Table 4.17: Typical Observed Gravel Loss

Scenarios Regravelling Interval (Years) Typical Gravel Loss

(mm/Year)

A 8 18.75

B 7 21.4

C 6 25

D 5 30

E 4 37.5

F 3 50

The HDM-4 gravel loss model was then calibrated for each climate zone by adjusting the calibration factor for gravel loss (Kgl) so that the predicted rate matched the observed rate on each site. The calibration was performed for the following three traffic levels:

Low - 100 AADT;

Medium - 300 AADT;

High - 800 AADT;


51

Table 4.18 gives the default HDM-4 predicted gravel loss rates (before calibration) as well as the calibration factors for adjusting the model to achieve the gravel loss rates reported in Table 4.17.

Table 4.18: Summary of Observed and HDM-4 Predicted Gravel loss and Calibration Factor

Climate Zone

Traffic Level

Default HDM-4 Predicted

Gravel Loss (mm/Year)

Gravel Loss Calibration Factor (Kgl) by Gravel Loss Scenario (See Table 5.13)

A B C D E F

Zone 2

Low 19.05 0.98 1.13 1.32 1.58 1.99 2.65

Medium 29.91 0.61 0.71 0.84 1 1.26 1.69

High 55.05 0.32 0.37 0.44 0.53 0.67 0.91

Zone 3

Low 19.03 0.99 1.13 1.32 1.58 1.99 2.65

Medium 31.79 0.57 0.66 0.78 0.94 1.18 1.58

High 60.73 0.28 0.33 0.39 0.48 0.6 0.81

Zone 4

Low 13.48 1.39 1.59 1.85 2.23 2.78 3.71

Medium 19.91 0.94 1.08 1.26 1.51 1.9 2.55

High 29.94 0.61 0.7 0.83 1\ 1.25 1.68

The results presented in Table 4.18 does not include calibration factors for climate zone 1 due to lack of data on typical properties of materials that are available or used in this zone. Since climate zones 1 and 2 are geographically adjacent to each other, calibration factors derived for zone 2 could in the meantime be used for analysis of unsealed roads located in climate zone 1. It is however recommended that future improvements to the calibration should consider deriving specific calibration factors for unsealed roads in zone 1 on the basis of the properties of surface and subgrade material properties in zone 1.

4.12 Rigid Concrete Pavements

Rigid concrete constitute a very small percentage of the Federal road network. Reliable data for rigid concrete pavements were not available to enable a meaningful calibration of the RD models in HDM-4. The calibration of HDM-4 RD models for rigid concrete models was therefore not performed. The Consultant recommends that data on the performance of rigid concrete pavements should be collected in the future to enable HDM-4 RD model calibration.

ROAD WORKS EFFECTS MODEL CALIBRATION

52

5 ROAD WORKS EFFECTS MODEL CALIBRATION

5.1 Introduction

Road works are divided under two main categories under HDM-4 namely: preservation and development. Preservation of the existing pavements involves performing maintenance works to offset the deterioration of roads as well as lowering road user cost by providing a smooth running and keeping the road open on a continuous basis. Development works are aimed at expanding the capacity of the network through the provision of stronger pavement and the improvement of the geometric characteristics in order to minimize the total cost of road transportation and to mitigate environmental impacts. Within the above two broad categories, road works are considered in classes. The works classes consider road works in terms of their frequency of application and the budget head used to fund them. The work classes adapted in line with planning, programming and project practices in Nigeria are summarised in Table 5.1.

5.2 Effects of Road Works

Definition of the effects of road works activities given in Table 5.1 is an important requirement in life cycle modelling of road pavement deterioration. These effects vary by pavement type and construction practice. The effect of overlay on roughness is of high sensitivity because it dictates future deterioration rates and thus maintenance activities.

Meaningful calibration of the HDM-4 work effects models requires data recorded immediately before and after road works on representative road sections. Such data were not available for use in this study, instead, effects of each work activity were defined where defined using expert knowledge elicited through structured interviews with the Ministry of Works Engineers. Default HDM-4 work effects model parameters is suggested for some work activities. A review of the proposed works effects (Table 5.1) is recommended when suitable data is available.

ROAD WORKS EFFECTS MODEL CALIBRATION

53

Table 5.1: Summary of class, type of works (activities) and effects adapted for Nigeria

Work Category

Work Class Work Type Work Activity/Operation Effects

Preservation

Routine Maintenance

Spot Regravelling Derived in HDM-4 using defaults

Grading Derived in HDM-4 using defaults

Drainage Maintenance

Ditch Cleaning

Very Good

Re-excavation of Drainage Ditches

Cleaning and Repair Culverts

Desilting Culverts

Repair of Cracks on Drainage Structures

Erosion and Scour Repairs

Periodic Maintenance

Regravelling

Placing of adequate subbase gravel on an existing gravel road to strengthen the pavement.

Roughness = 4 IRI

Resealing

Placing of a fresh seal coat on an existing bituminous surfaced to seal cracks and improve resistance.

Derived in HDM-4 using defaults

Overlay

Placing of asphaltic concrete on an existing bituminous surfaced or asphaltic concrete road to strengthen the pavement

Roughness = 2.5 IRI, Rut Depth = 0mm

Partial Reconstruction

Scarifying of existing bituminous surfaced road, strengthening the base layer with addition of adequate thickness of base material and applying surface treatment.

Roughness = 2.5 IRI, Rut Depth = 0mm

Development

Improvement

Reconstruction

Full Pavement construction and drainage structures may involve widening and re-alignment

Roughness = 2 IRI, Rut Depth = 0mm

Major Rehabilitation

Mainly Partial Pavement reconstruction and drainage structures may not involve widening and re-alignment.

Roughness = 2 IRI, Rut Depth = 0mm

Upgrading

Gravel to Bituminous Surface Treated (BST)

Roughness = 2 IRI, Rut Depth = 0mm BST to Asphaltic Concrete

New Construction

New Section

Dualization Roughness = 2 IRI, Rut Depth = 0mm

Missing Links

By-pass Construction

ROAD USER EFFECTS MODEL CALIBRATION

54

6 ROAD USER EFFECTS MODEL CALIBRATION

6.1 Introduction

Road User Costs (RUC) comprises of Vehicle Operating Costs (VOC), travel time, vehicle emissions, accident and energy use along with developmental effects.

6.2 Speed Prediction Model

The two most sensitive factors in the speed prediction model include the free speed and the driving power. For most of the other factors level 1 or HDM-4 default values have been adopted.

6.2.1 Free Speed

The HDM-4 speed prediction model is mechanistic, being based on physical and kinematic principles, as well as behavioural constraints. Consequently, the basic physical model is highly transferable and Level 2 Calibration should give first priority to speed and capacity. The HDM-4 speed model predicts that speeds are the probabilistic minimum of five constraining speeds based on:

Power

Braking

Curvature

Roughness

Desired speed This assumes that the physical performance of the vehicle has been properly calibrated by identifying valid representative values of other important vehicle characteristics, namely:

Vehicle mass

Used driving power

Braking power

The key behavioural constraint parameters are as follows;

VDESIR - the desired speed of travel This can be expected to differ considerably between countries, and even regions within the same country. VDESIR represents the maximum speed of travel adopted by the driver of a vehicle when no other physical constraints, such as gradient, curvature, roughness or congestion, govern the travel speed. The value of VDESIR is influenced by factors such as speed limits and enforcement, road safety, cultural and behavioural attitudes. For this study, vehicle speeds were observed on selected straight, flat, smooth, non-congested road segments of different road types (2-lane, 4-lane, 6-lane roads) for both paved and unpaved roads as described in Chapter 3.

β - the ‘draw down


55

Indicates how far from the constraining speeds the predicted speed will be. β is the Weibull Shape Parameter. It is functionally related to the dispersion of the underlying distribution of the constraining speeds. The value of β were determined using the data collected on vehicle free speeds presented in Chapter 3.

The desired speed of travel for vehicle types were determined on a number of selected road sections on the Federal Road Network as part of the traffic flow, speed and pattern survey carried out in September to November 2013. The observed mean speeds are shown in Tables 3.12 and 3.13. The HDM-4 model was run on a road section with average representative condition reflecting observations during the speed estimates. Adjustments were then made to the VDESIR values

by applying the ratio of the observed and predicted speed plus modifications of ‘' following each model run until the predicted speed was the same as the observed speed for each

representative vehicle type. The final values are given in Table 6.1.

Table 6.1: Estimated “β" Values for each Representative Vehicle

Representative Vehicle β Value

(a) Motorcycle 0.182

(b) Car Small 0.201

(c) Car Medium 0.201

(d) Car Large 0.201

(e) Four Wheel Drive 0.217

(f) Bus Small 0.217

(g) Bus Medium 0.289

(h) Bus Large/Coach 0.321

(i) Light Delivery Vehicle 0.201

(j) Medium Delivery Vehicle 0.201

(k) Truck Rigid 2-axle 0.318

(l) Truck Rigid 3/4 Axle 0.164

(m) Truck Multi-axle Truck & Trailer 0.164

(n) Truck Horse & S-Trailer 3/4 Axles 0.164

(o) Truck Horse & Semi-Trailer 5/6 Axles 0.164

(p) Truck Horse and semi-Trailer 7 Axles 0.164

(q) Truck Horse & 2 Trailers 0.164

6.2.2 Vehicle Driving Power

The driving power only has a significant effect on speeds when the gradient is positive and higher than 4% for light vehicles and 2-3 % for heavy vehicles. In Nigeria, this case particularly applies to the plateau regions and the central hilly areas and where we encounter some hills. For a level 2 calibration the used power can be estimated from the attributes using the following equations:

HPDRIVE = 0.70HPRATED (Diesel Vehicles) (6.1)

HPDRIVE = 2.0HPRATED0.7 (Petrol Vehicles) (6.2)


56

Where:

HPDRIVE is the used driving power (kW)

HPRATED is the Society of Automotive Engineers (SAE) maximum rated engine power (kW)

Table 6.2 gives the estimated vehicle driving power for each vehicle.

Table 6.2: Vehicle Driving Power

Representative Vehicle Fuel Type

Observed HPRATED (kW)

HPDRIVE (kW)

(a) Motorcycle Petrol 8.8 6.2

(b) Car Small Petrol 88.0 45.9

(c) Car Medium Petrol 98.0 49.5

(d) Car Large Petrol 133.0 61.3

(e) Four Wheel Drive Diesel 138.0 96.6

(f) Bus Small Diesel 100.0 70.0

(g) Bus Medium Diesel 110.0 77.0

(h) Bus Large/Coach Diesel 210.0 147.0

(i) Light Delivery Vehicle Diesel 113.0 79.1

(j) Medium Delivery Vehicle Diesel 129.0 90.3

(k) Truck Rigid 2-axle Diesel 240.0 168.0

(l) Truck Rigid 3/4 Axle Diesel 300.0 210.0

(m) Truck Multi-axle Truck & Trailer Diesel 320.0 224.0

(n) Truck Horse & S-Trailer 3/4 Axles Diesel 300.0 210.0

(o) Truck Horse & Semi-Trailer 5/6 Axles

Diesel 340.0 238.0

(p) Truck Horse and semi-Trailer 7 Axles

Diesel 375.0 262.5

(q) Truck Horse & 2 Trailers Diesel 460.0 322.0

6.3 Side Friction

For modeling traffic flows and effects on vehicle operating costs (VOC), HDM-4 considers three types of friction:

Friction to motorized transport arising from roadside activities, XFRI (e.g. different

types of land use and encroachment to the road right of way). XFRI value also ranges from 0.4 (high friction level) to 1.0 (no friction)

Friction to motorized transport due to the presence of non-motorised transport XNMT

(e.g. pedestrians, bicycles, animal carts). XNMT value ranges from 0.4 (high friction level) to 1.0 (no friction)

Friction to non-motorised transport arising from motorized transport using the road, XMT. Values of XMT also ranges from 0.4 to 1.0.


57

The Consultant has made observations on a number of road sections of different types and analysed the information collected to provide a guide to determine values of XNMT, XMT and XFRI on different roads in Nigeria. Defaults values provided in HDM-4 would need to be

adjusted to reflect side specific observations as well as analysis objectives.The following three parameters, called speed reduction factors, are used to model the three types of friction described above. These are defined in Table 6.3.

Table 6.3: Rating Friction due to MT

Grade Description

1 No friction

2 Moderate friction

3 High friction

For Grade 1 rating of friction the values of XNMT, XFRI and XMT will be 1.0. For Grade 3 rating, the procedure for determining the values of XNMT and XFRI will be as follows:

(i) Identify a sample of the road sections most severely affected by friction (ii) Determine the average free speed of motorized traffic (SF) on the road

sections, or use the average design speeds of the sampled roads instead (iii) Observe the average operating speed of motorized traffic (SO) on the road

sections (iv) Calculate the ratio of SO to SF. The ratio denoted by FR should then be

compared with the minimum default in HDM-4 which is (0.4 x 0.4 = 0.16). If FR is less than 0.16 reset the value of FR to be 0.16. This means that if SF is 100 km/h the average operating speeds of motorized traffic will be reduced to 16 km/h due to friction arising from both roadside activities and the presence of non-motorised transport.

The values of XNMT and XFRI for Grade 3 rating of friction will be given by the following equation: X = (FR)0.5 (6.3) Where X = XNMT or XFRI value for Grade 3 For any intermediate level of friction (e.g. Grade 2 rating of friction) the values of XNMT and XFRI will be obtained from the following equation: Y = 0.5(1 + X) (6.4) Where Y = XNMT or XFRI value for Grade 2 X = XNMT or XFRI value for Grade 3 For modeling NMT, the values of XMT should be as follows: For Grade 1 XMT = 1.0; for Grade 2 XMT = 0.7; and for Grade 3 XMT = 0.4 Intermediate values can be obtained by interpolation when data is required at a more detailed level. The photographs in Figures 6.1 to 6.4 show different levels and types of side friction in Nigeria.


58

Figure 6.1: High level of friction

Figure 6.2: Intermediate level of friction, high level of road side activities


59

Figure 6.3: Intermediate level of friction due to presence of animals

Figure 6.4: Low level of friction


60

6.4 Fuel Consumption

The HDM-4 fuel consumption model predicts fuel consumption as proportion of the power as follows:

𝐼𝐹𝐶 = 𝑀𝐴𝑋(∝, ξPtot) (6.5)

Where:

IFC is the instantaneous fuel consumption (ml/s);

α is the idle fuel consumption (ml/s);

is the power-to-fuel efficiency factor (ml/kW.s); and

Ptot is the total power requirements (kW).

Data required for the calibration of the fuel consumption model was elicited from road side interviews of drivers who normally travel on several national roads throughout the country. Data collected for each vehicle included: vehicle type, vehicle category, age of vehicle, type of fuel used, trip origin, trip destination, approximate trip distance, average travel time, and amount of fuel used.

The collected data was used to derive the following data types: average fuel consumption (litres per 1000 vehicle kilometres), average road gradient, and total power requirement (kW).

The calibration parameter was estimated for each vehicle type by running HDM-4 for a typical

road section and adjusting .until the predicted fuel consumption closely matches the observed values for each vehicle type. The results are shown in Table 6.4.

Table 6.4: Estimated “" Values for each Representative Vehicle

Vehicle Type

Default Fuel Consumption (l/100 veh-

km)

Observed Fuel

Consumption (l/100 veh-

km)

Default Calibrate

d

(a) Motorcycle 33.86 41.049 0.067 0.084

(b) Car Small 117.81 108.283 0.067 0.062

(c) Car Medium 132.04 127.466 0.067 0.065

(d) Car Large 154.45 129.239 0.067 0.056

(e) Four Wheel Drive 158.36 124.456 0.057 0.045

(f) Bus Small 175.04 127.860 0.067 0.049

(g) Bus Medium 212.21 142.315 0.057 0.038

(h) Bus Large/Coach 378.24 229.167 0.057 0.034

(i) Light Delivery Vehicle 156.76 187.619 0.067 0.080

(j) Medium Delivery Vehicle 217.87 142.315 0.067 0.044

(k) Truck Rigid 2-axle 322.05 223.494 0.057 0.040

(l) Truck Rigid 3/4 Axle 486.3 332.000 0.057 0.038

(m) Truck Multi-axle Truck & Trailer 760.6 524.367 0.055 0.038

(n) Truck Horse & S-Trailer 3/4 Axles 550.48 403.133 0.056 0.041

(o) Truck Horse & Semi-Trailer 5/6 Axles 697.21 480.665 0.055 0.038


61

Vehicle Type

Default Fuel Consumption (l/100 veh-

km)

Observed Fuel

Consumption (l/100 veh-

km)

Default Calibrate

d

(p) Truck Horse and semi-Trailer 7 Axles 722.48 498.087 0.055 0.038

(q) Truck Horse & 2 Trailers 927.39 639.354 0.055 0.038

6.5 Spare Parts Consumption

Calibration of parts consumption model was achieved by adjusting the rotational parts consumption calibration factor so that the observed parts consumption for each vehicle type was the same as the predicted values for an average road condition. Predicted default vehicle parts consumption presented in Table 6.5 were mostly higher than values elicited from surveys. The calibration was thus performed on the basis that a large number of vehicles used in Nigeria are second hand imports. Furthermore, the road network may on average be described as being in fair to poor condition with medium vehicle utilisation in general.

The generally high ages of the vehicles implies higher parts consumption but this is counteracted by the following factors:

Lower standard of vehicle maintenance and service which results in lower parts consumption;

Medium vehicle utilisation hence less parts consumption;

Availability of cheaper locally manufactured or improvised vehicle parts;

The existence of cannibalism practices i.e. removing spare parts from on vehicle and fitting them on another vehicle.

Table 6.5: Parts Consumption Data and Model Calibration Factors

Vehicle Type

Parts Consumption per 1000 veh-km Rotational

Calibration Factor Observed

Default Prediction

(a) Motorcycle 0.081 0.079 1.023

(b) Car Small 0.105 0.215 0.489

(c) Car Medium 0.113 0.220 0.514

(d) Car Large 0.139 0.209 0.666

(e) Four Wheel Drive 0.115 0.146 0.789

(f) Bus Small 0.140 0.226 0.619

(g) Bus Medium 0.112 0.112 1.001

(h) Bus Large/Coach 0.143 0.103 1.392

(i) Light Delivery Vehicle 0.158 0.239 0.662


62

Vehicle Type

Parts Consumption per 1000 veh-km Rotational

Calibration Factor Observed

Default Prediction

(j) Medium Delivery Vehicle 0.174 0.214 0.815

(k) Truck Rigid 2-axle 0.105 0.150 0.699

(l) Truck Rigid 3/4 Axle 0.174 0.303 0.574

(m) Truck Multi-axle Truck & Trailer 0.139 0.333 0.418

(n) Truck Horse & S-Trailer 3/4 Axles 0.157 0.303 0.518


0.131 0.333 0.394


0.148 0.333 0.445

(q) Truck Horse & 2 Trailers 0.139 0.333 0.418

6.6 Tyre Wear

The tyre type and the number of wheels are used in HDM-4 for establishing the rolling resistance. Bias ply tyres have greater rolling resistance than radial tyres, and the resistance increases with an increasing wheel diameter and number of wheels. Tyre sizes have a standard typology. The two most common types are shown in Figure 6.5 (HDM-4 Documentation, Volume 5) along with a description of what each term means. The top typology is common with truck tyres and is based on the nominal section width being expressed in inches. The second is used for all vehicles and has the nominal section width given in mm along with the aspect ratio. The wheel diameter can be estimated from the tyre typology using the following equation:

(6.6)

Figure 6.5: Standard Tyre Typology


63

Where: DIAM: is the wheel diameter (mm) xxx: is the nominal width of the tyre (mm) yy: is the aspect ratio zz: is the rim size (inches)

The numbers of wheels and tyre type have been established through the VOC survey. The full results are presented in Table 6.6. Since the vehicle and trailers may have different tyre types, the tyre consumption has be calculated separately for the vehicle and trailer and then combined to give the total tyre consumption. We determined tyre consumption using the HDM-4 standard methodology provided in Volume 7 of the HDM-4 documentation. In calibrating the HDM-4 Tyre wear model, we determined the tyre wear coefficient and the wearable rubber volume using data on tyre types and frequency of tyre replacements by vehicle fleet operators and the general public in Nigeria. Tyres with thinner grooves are considered to last longer. From previous experience in calibrating this model, we have often found that legal limits in terms of depth of the grooves on the tyres are not normally fully followed. We investigated if this situation is prevalent in Nigeria and found that is the case too. We, therefore, use it to inform the calibration of the model. It should also be noted that the quality of the rubber used in manufacturing the car tyres has an effect on the rate of wear and the volume of wearable rubber.

Table 6.6: Calibration Coefficient for Tyre Wear Model

Representative Vehicle Constant Term, dm3

Wear Coefficient dm3/J-m

Wearable Rubber Volume, dm3

(a) Motorcycle 0.00575 0.0005 0.35

(b) Car Small 0.02354 0.00204 1.4

(c) Car Medium 0.02354 0.00204 1.4

(d) Car Large 0.02354 0.00204 1.4

(e) Four Wheel Drive 0.02160 0.00187 1.6

(f) Bus Small 0.02160 0.00187 1.6

(g) Bus Medium 0.02397 0.00207 6

(h) Bus Large/Coach 0.02779 0.00241 8

(i) Light Delivery Vehicle 0.02160 0.00187 1.6

(j) Medium Delivery Vehicle 0.02160 0.00187 1.6

(k) Truck Rigid 2-axle 0.02160 0.00187 1.6

(l) Truck Rigid 3/4 Axle 0.03176 0.00275 8

(m) Truck Multi-axle Truck & Trailer 0.03589 0.00311 8

(n) Truck Horse & S-Trailer 3/4 Axles 0.03176 0.00275 8


0.03589 0.00311 8


0.03589 0.00311 8

(q) Truck Horse & 2 Trailers 0.03589 0.00311 8


64

6.7 Vehicle Exhaust Emissions

HDM-4 can be applied to predict vehicle emissions considered to be the most damaging to the natural environment and human health. These includes the following types of vehicle exhaust emissions: Hydrocarbon (HC), Carbon monoxide (CO), Nitrous Oxide (NO), Sulphur Dioxide (SO2), Carbon Dioxide (CO2), Particulates (Par) and Lead (Pb).

The emission models included in HDM-4 predict vehicle exhaust emissions as a function of vehicle speed, fuel consumption and vehicle service life. Data suited for calibrating the emission models to conditions in Nigeria were not available. However, since the fuel consumption, speed and service life models were calibrated, the effects will be reflected in the emissions model outputs. Default HMD-4 emission model parameters is recommended.

HDM-4 CONFIGURATION

65

7 HDM-4 CONFIGURATION

7.1 Climate Zones

Four climate zones were used to categorise the Nigerian road network. Detailed description including HDM-4 climate input data for each zone are given in Section 4.2.

7.2 Traffic Flow Pattern

The data that describes the hourly traffic flows and the volume-capacity ratios required for modelling congestion effects on vehicle speeds and vehicle operating costs will be derived from traffic data obtained from DGER. The key parameters are as follows:

Qp: The traffic flow, in PCSE per hour, during each traffic flow period p.

VCRp: The volume-capacity ratio for each traffic flow period p.

Hourly flow-frequency distribution data are specified for each road use category. This reflects the fact that the predominant use of different roads requires different shapes of flow-frequency distribution curves. The curves are defined in terms of the number of hours per year that the traffic volume is at a certain percentage of the AADT. Each specified flow-frequency distribution can then be assigned to a group of road sections

To determine traffic flow pattern encountered within the Nigeria road network, the traffic database has been scanned and data from annual automatic counts and temporary stations have been evaluated.

The hourly traffic flow for each flow-frequency period is expressed as a proportion of AADT, and is given by:

(7.1)

Where:

HVp is hourly traffic flow in period p, as a proportion of AADT

PCNADTp is percentage of AADT in period p

HRYRp is number of hours per year in period p

The traffic flow during each flow period is calculated as follows:

(7.2)

Where:

HDM-4 CONFIGURATION

66

Qp: hourly traffic flow in period p (PCSE per hour)

AADTk: annual average daily traffic of vehicle type k

PCSEk: passenger car space equivalent of vehicle type k

The volume-capacity ratio during each traffic flow period is expressed as follows:

Table 7.1 summarises traffic flow pattern defined for Nigeria for four road use types.

Table 7.1: Traffic Flow Pattern

Road Use Flow Period Hours Per Year

(HRYRp) % AADT (PCNADTp)

Commuter

A 87.6 3.10

B 350.4 11.33

C 613.2 17.00

D 2978.4 58.26

E 4730.4 10.31

Inter-Urban

A 87.6 2.17

B 350.4 7.59

C 613.2 11.00

D 2978.4 40.24

E 4730.4 39.00

Seasonal

A 87.6 4.25

B 350.4 13.24

C 613.2 16.60

D 2978.4 40.32

E 4730.4 25.59

Free Flow A 8760 100

7.3 Speed Flow Type

The average speed of each vehicle type is required for calculating vehicle operating costs, travel time, energy use and emissions. The speeds of MT vehicles are influenced by a number of factors, which include:

Vehicle characteristics

Road severity characteristics, for example, road alignment, pavement condition, etc

The presence of non-motorised transport (NMT)

HDM-4 CONFIGURATION

67

Roadside friction, for example, bus stops, roadside stalls, access points to roadside development, etc.

Total MT traffic volume

The speed-flow model adopted for each motorised transport (MT) is the three-zone model proposed by Hoban et al. (1994). This model is illustrated in Figure 7.1 below.

Figure 7.1: Illustration of Speed Flow Model

The following notation applies to Figure 7.1:

Qo - the flow level below which traffic interactions are negligible in PCSE/h

Qnom - nominal capacity of the road (PCSE/h)

Qult - the ultimate capacity of the road for stable flow (PCSE/h)

Sult - speed at the ultimate capacity, also referred to as jam speed (km/h)

Snom - speed at the nominal capacity (km/h)

S1 to S3 - free flow speeds of different vehicle types (km/h)

PCSE - passenger car space equivalents

The model predicts that below a certain volume there are no traffic interactions and all vehicles travel at their free speeds. Once traffic interactions commence the speeds of the individual vehicles decrease until the nominal capacity where all vehicles will be travelling at the same speed, which is estimated as 85% of the free speed of the slowest vehicle type. The speeds can then further decrease towards the ultimate capacity beyond which unstable flow will arise.

HDM-4 CONFIGURATION

68

Table 7.2: Capacity and speed-flow model parameters

Speed Flow Type

Width (m) XQ1 XQ2 Qult

(PCSE/hour/lane) Sult

(km/hr)

Single-Lane Road

< 4 0 420 600 10

Intermediate Road

4 – 4.5 45 630 900 20

Two-Lane Narrow

5.5 - 7 130 1170 1300 23

Two-Lane Standard

7 - 8 140 1255 1375 25

Two-Lane Wide 8 - 12 320 1440 1600 30

Four-Lane Road >12 800 1900 2000 40

Six-Lane Road >21 1040 2400 2600 40

7.4 Road Network Aggregate Data

Configuration of aggregate data involves the definition of aggregate information for the following:

Traffic levels: e.g., low, medium, high;

Geometry class: in terms of parameters reflecting horizontal and vertical alignment;

Pavement characteristics: structure and strength parameters defined according to

pavement surface class;

Road condition: ride quality, surface distress and surface texture; and

Pavement history: construction quality, pavement age, etc

The aggregate data represent a set of default values that may be used in case data is lacking. It also sets the classes and categories used for the segmentation of the road network into homogeneous representatives or physical sections. The proposed aggregate classes include the following:

Road Class - This will cover all the road classes defined for the Nigeria road network

Traffic Volume - Five traffic classes have been proposed as follows: o Very Low o Low o Medium o Heavy o Very Heavy

Geometry Class - Seven geometry classes have been proposed for the Nigeria road network to represent the different types of terrain encountered in the country:

o Straight and level o Mostly straight and gently undulating o Bendy and generally level o Bendy and gently undulating

HDM-4 CONFIGURATION

69

o Bendy and severely undulating o Winding and gently undulating o Winding and severely undulating

Construction Quality - The following five default construction quality classes have been proposed for this configuration.

o Good o Fair-brittle o Fair-soft o Poor-brittle o Poor -soft

Structural Adequacy - Three Classes have been proposed for the Nigeria Road Network

o Warning o Acceptable o Strong

Roughness - Three Roughness Classes have been proposed as follows: o Poor o Fair o Good

Surface Condition - Three Surface Condition Classes have been proposed as follows: o Good o Fair o Poor

Surface Texture/Friction - Three Surface Texture/Friction Classes have been defined as follows:

o Coarse o Medium o Fine

Section Aggregate Tables Traffic volume

For each road section, traffic level is specified in terms of annual average daily traffic (AADT) flow. At aggregate data level, traffic volume is defined in bands or levels. Detailed data values are associated with these in terms of the mean AADT.

The following defines a traffic band: description and road surface class. The road surface class to which the traffic band applies that is bituminous, concrete or unsealed. Traffic Band

Five traffic bands were defined for each road surface class. Tables 7.3 and 7.4 show the detailed traffic levels assumed related to the five road classes, with the following default levels:

Very Low

Low

Medium

Heavy

Very Heavy

HDM-4 CONFIGURATION

70

The representative ADT weighted average value for each traffic band was calculated from existing data. The default values indicated are mid-point measures and must be replaced with calculated averages wherever possible.

Table 7.3: Default Traffic Volume on Bituminous Roads

Traffic Class

Traffic Band & Representative Value

Default AADT (in vehicles/day) by Road Class

A B C D

Very Heavy

Band >5001 >5001 >5001 >5001

Rep. Value 7500 7500 5500 5500

Heavy Band 2001 - 5000 2001 - 5000 2001 - 000 2001 - 5000

Rep. Value 3500 3500 3500 3500

Medium Band 1001 - 2000 1001 - 2000 1001 - 2000 1001 - 2000

Rep. Value 1500 1500 1500 1500

Low Band 501 - 1000 501 - 1000 501 - 1000 501 - 1000

Rep. Value 750 750 750 750

Very Low

Band <500 <500 <500 <500

Rep. Value 250 250 250 250

Table 7.4: Default Traffic Volumes on Unsealed Roads

Traffic Class

Traffic Band & Representative Value

Default AADT (in vehicles/day) by Road Class

Trunk A Trunk B Trunk C Trunk D

Very Heavy

Band >1001 >1001 >1001 >1001

Rep. Value 1250 1250 1250 1250

Heavy Band 501 - 1000 501 - 1000 501 - 1000 501 - 1000

Rep. Value 750 750 750 750

Medium Band 251 - 500 251 - 500 251 - 500 251 - 500

Rep. Value 375 375 375 375

Low Band 101 - 250 101 - 250 101 - 250 101 - 250

Rep. Value 175 175 175 175

Very Low

Band <250 <250 <250 <250

Rep. Value 100 100 100 100

Geometry Aggregate Table

At the aggregate level, road geometry in Nigeria has been defined in terms of various parameters reflecting horizontal and vertical curvature. These represent geometry classes and apply to a group of roads.

The following detailed data (see Table 7.5) defines a geometry class:

Description

Average rise plus fall (m/km)

Number of rises and falls per kilometre (no/km)

Average horizontal curvature (deg per km)

Superelevation (at bends) - represented as a percentage (%). If not available as in the

case of Nigeria, values of superelevation can be derived from the average horizontal curvature, C, as follows: bituminous roads (e = 0.012*C), unsealed roads (e =

0.017*C).

HDM-4 CONFIGURATION

71

The following data have also been defined together with a road geometry class:

Speed limit (km/h)

Speed limit enforcement factor (default = 1.1)

The number of geometry classes defined is at the discretion of the user. Table 4.7 resents the detailed data assumptions relating to the default geometry classes for the Nigeria road network. The values given can be amended if necessary.

Table 7.5: Default Geometry Characteristics No. Geometry

Class Rise +

Fall (m/km)

Number of rises and

falls per km

Horizontal curvature (deg/km)

Superelevation

(%)

Speed limit

(km/h)

1 Straight and level 1 1 3 2 120

2 Mostly straight and gently undulating

10 2 15 2.5 120

3 Bendy and generally level 3 2 50 2.5 110

4 Bendy and gently undulating 15 2 75 3 80

5 Bendy and severely undulating 25 3 150 5 70

6 Winding and gently undulating 20 3 300 5 60

7 Winding and severely undulating

40 4 500 7 50

Construction Quality Aggregate Table

The construction quality for bituminous pavements is described at the aggregate data level by values such as good, fair, poor, etc. The actual data details to be specified relate to construction defect indicators. In the case of the Nigeria road network, five construction quality classes have been used, and for each class the following information has been provided:

Description

Construction defect indicator for bituminous surfacing, CDS

Construction defect indicator for road base, CDB

Relative compaction, (%)

The default values for construction quality relating aggregate to detailed data are shown in Table 7.6.

Table 7.6: Default Construction Defect Indicators

Construction Quality

Construction defect indicator for

Bituminous Surfacing (CDS)

Construction defect indicator for Road Base

(CBS)

Relative Compaction (%)

Good 1.0 0 98

Fair-brittle 0.75 0.8 91

Fair-soft 1.25 0.8 91

Poor-brittle 0.50 1.5 85

Poor-soft 1.50 1.5 85

Structural Adequacy Aggregate Table

The parameters that are used in HDM-4 to describe pavement characteristics vary according to road surface classes. The strength of bituminous pavements is defined by their structural adequacy to carry traffic loading. At aggregate data level, structural adequacy is defined in terms of qualitative descriptors/measures such as good, fair, poor, etc. The detailed data

HDM-4 CONFIGURATION

72

values relating to these are in terms of the Adjusted Structural Number of the Pavement (SNP). The detailed data have been assigned by traffic level. The default data assumptions for structural adequacy of bituminous pavements in terms of three qualitative descriptors Warning, Acceptable, and Strong, for different traffic bands (that is, low, medium, and high), defined for Nigeria are shown in Table 7.7.

Table 7.7: Default SNP Values

Structural Adequacy Traffic Bands

High Medium Low

Warning 2.5 2.0 1.5

Acceptable 3.5 2.5 2.0

Strong 4.0 3.5 2.5

Bituminous Layer Aggregate Table

The SNP values defined as measures for structural adequacy is also used in HDM-4 to represent pavement layer thickness. For each SNP range, and for each pavement type (plus the option of “all pavement types”), the following have been defined for Nigeria and their default values indicated in Table 7.8.

most recent (new) surfacing thickness (mm)

previous/old surfacing thickness (mm)

road base thickness if base type is stabilised base SB (mm)

Table 7.8: Default Pavement Layer Thicknesses Structural Adequacy Surface Thickness (mm) Stabilised Base Thickness (mm)

New Old

SNP < 2.5 10 15 150

2.5 < SNP < 4.0 15 20 200

4.0 > SNP 25 25 300

Ride quality is an indication of the roughness of the road. It is an important parameter for indicating road condition and maintenance needs, and for predicting vehicle operating costs. At the aggregate level, ride quality is defined in terms of qualitative measures such as good, fair, poor, etc. The detailed data values related to these are in terms of roughness IRI (m/km), and are assigned by road class. The default data assumptions for Nigeria for ride quality of bituminous pavements, and unsealed roads in terms of three qualitative measures (good, fair, and poor), for different road classes. The weighted average roughness values are calculated for each cell from existing data using roughness ranges indicated in Table 7.9.

Table 7.9: Default Riding Quality Data

Road Class

Roughness ranges and Rep. Values

Paved Road Roughness (m/km)

Unsealed Road Roughness (m/km)

Good Fair Poor Good Fair Poor

A Band < 3.0 3.1 – 4.0 > 4.1 < 5 5.1 - 7 >7.1

Rep. Value 2.5 3.5 4.5 4 6 9

B Band < 4.0 4.1 – 5.0 > 5.1 < 7 7.1 – 10 > 10.1

Rep. Value 3.5 4.5 5.5 6 9 12

HDM-4 CONFIGURATION

73

Road Class

Roughness ranges and Rep. Values

Paved Road Roughness (m/km)

Unsealed Road Roughness (m/km)

Good Fair Poor Good Fair Poor

C Band < 4 4.1 – 5.5 > 5.6 < 9 9.1 - 14 > 14.1

Rep. Value 4 5 6.5 8 12 16

F Band < 6 6.1 – 8.5 > 8.6 < 12 12.1 - 16 > 16.1

Rep. Value 5.5 7 8.5 10 14 16

Surface Condition Aggregate Table

Surface condition is modelled by a number of distress modes. At the aggregate level surface condition can be defined by a qualitative measure (for example, new, good, fair, poor) that represents several distress modes. These distress modes differ depending on whether the surface class is bituminous or unsealed. For each of the two paved road surface classes available in Nigeria, and for each qualitative measure of surface condition used, a default value for each of the distress modes has been defined for the Nigeria road network The weighted average distress values calculated for each cell from existing data using distress ranges indicated in Table 7.10.

Table 7.10: Paved Surface Condition Default Values

Surface Condition

Distress Band & Rep. Value

Levels of Distresses by mode

Total area Cracking

(%)

Ravelling (%)

Potholing (no/km)

Edge break

(m2/km)

Mean rut depth (mm)

New Band 0 - 0 0 - 0 0 - 0 0 - 0 0 - 0

Rep. Value 0 0 0 0 0

Good Band 0 - 5 0 - 5 0 - 0 0 – 5 0 – 8

Rep. Value 1.5 2 0 0 3.5

Fair Band 5 - 10 5 - 10 0 - 5 0 - 10 8 - 15

Rep. Value 6.5 7 3 6 9

Poor Band > 10 > 10 > 5 > 10 > 15

Rep. Value 25 30 12 25 17

For unsealed roads, surface condition and structural adequacy are both related to the traffic level and are represented by the thickness of the gravel surfacing. For each qualitative measure of surface condition and for each of the pre-defined traffic bands, a set of default values of gravel thickness for unsealed roads are given in Table 7.11.

Table 7.11: Unsealed Surface Condition Default Values

Pavement Structural Adequacy

Distress Band & Rep. Value

Surfacing material thickness (mm) by Traffic Band

High Medium Low

Good Band 200 - 250 150 - 200 150 - 180

Rep. Value 200 180 150

Fair Band 150 - 200 100 - 180 80 - 150

Rep. Value 190 160 135

Poor Band < 150 < 100 < 80

Rep. Value 80 50 30

HDM-4 CONFIGURATION

74

Texture/Friction Aggregate Table

At the aggregate data level, surface texture can be defined by a qualitative measure (for example, coarse medium, fine or slippery, etc) that gives an indication of the texture depth and skid resistance of the surface. In the current version of HDM-4, these parameters are modelled only for bituminous pavements. For each qualitative measure of surface texture, and for each of the surface types (AM or ST), mean values have been assigned for the sand patch texture depth and skid resistance at 50 km/h (SCRIM). The default values used for Nigeria at the detailed data level are given in Table 7.12.

Table 7.12: Default Texture/Skid Resistance Values

Texture Class ST Pavements Asphalt Pavements

Texture Depth (mm)

Skid Resistance (SFC50)

Texture Depth (mm)

Skid Resistance (SFC50)

Very Coarse 2.5 0.6 1.5 0.5

Coarse 1.5 0.5 1.0 0.4

Medium 0.5 0.4 0.3 0.35

Fine 0.3 0.3 0.2 0.3

CONCLUSION

75

8 CONCLUSION

The Draft Final Report has been prepared addressing all the objectives in the Terms of Reference. The Draft Final Report has presented the HDM-4 configuration and calibration work and the results obtained. Phase 1 activities have been completed. Phase 2 activities have taken longer than initially expected because of the difficulties and unforeseen situations regarding data acquisition. Nevertheless, this has now been completed. Phase 3 activities which relate to HDM-4 Configuration have been completed. The work activities under phases 4, 5, 6 and 7 have been completed.

Comments received at the Inception and after the submission of the Interim have all been addressed. .

The calibrated HDM-4 model can now be used to carry out specific project, programming and strategy analysis under the operating environment and climatic condition of Nigeria. Programming of works to be executed over a given period of time say 5 to 10 year horizon based on current road condition for the various classes of road in under the jurisdiction of the respective Road Agencies in Nigeria can now objectively be assessed. Strategy Analysis can now be carried for the entire road network and use as the basis to focus budget in a constrained budget scenario. The total budget required to fix the perennial maintenance backlog can now be addressed and used as a basis to engage the development partners in soliciting of loans and grants. The government outfit with ministerial oversight on the road agencies can now use the HDM-4 analysis to direct attention in a given area of the road network through policy.

In order to effectively use the calibrated HDM-4 model, base data must be updated and sustained and a system put in place to enable annual data collection thereafter. There is the need for staff training of the road agencies’ staff in the proper use of the HDM-4 model to enable its use for feasibility studies, programming of works and strategic planning of the road network. The workspace has also been customised in accordance with Nigerian local condition and the relevant Look-up tables all reviewed.

REFERENCES

76

9 REFERENCES

1. Bennett, C.R. and Paterson, W.D.O. (2000). A Guide to Calibration and Adaptation – Volume 5. International Study of Highway Development and Management Series, World Road Association (PIARC), PARIS. ISBN: 2-84060-063-3

2. Bennett, C.R. and Greenwood, I. (2004). Modelling Road User and Environmental Effects in HDM-4 – Volume 7. International Study of Highway Development and Management Series, World Road Association (PIARC), PARIS. ISBN: 2-84060-103-6

3. MCS Consulting, December 2011. State of Infrastructure Report on Nigerian Highways

4. Morosiuk G., Riley M.J. and Odoki J.B., (2004). Modelling Road Deterioration and Works effects - Volume 6 of The Highway Development and Management Series. International Study of Highway Development and Management (ISOHDM), World Road Association PIARC, Paris. ISBN: 2-84060-102-8

5. Nigerian Metrological Department, 2013, Abuja

6. Odoki, J.B. & Kerali, H.G.R. (2000). Analytical Framework and Model Descriptions – Volume 4. International Study of Highway Development and Management Series, World Road Association (PIARC), PARIS. ISBN: 2-84060-062-5

7. SSI Engineers and Environmental Consultants (Pty) Ltd., 2008. Axle Load Study

8. TANROADS - Vehicle Operating Costs Study, Final Report - July 2004 (DHV)

9. Ted R. Miller, Variations between Countries in Values of Statistical Life. Journal of Transport Economics and Policy, Vol 34, Part 2, May 2000

10. TRRL Limited, (1988). A guide to road project appraisal, Road Note 5. Transport and Road Research Laboratory, Crowthorne, Berkshire, UK

APPENDICES

77

APPENDIX A: TERMS OF REFERENCE

78



79

CONSULTING SERVICES FOR CONFIGURATION AND CALIBRATION OF HDM-4 TO

NIGERIA CONDITIONS

GENERAL

The Federal Government of Nigeria (FGN) through the Federal Ministry of Works (FMW) with financial assistance from the World Bank intends to procure the services of a consultant to carry out Configuration and Calibration of the Highway Development and Management Model (HDM-4) to Nigeria conditions aimed at improving decision-making on expenditures in the road sector by enabling effective and sustainable utilisation of the latest HDM-4 knowledge. The proposed configuration and calibration exercise will develop the basis for an effective implementation of decision-support methods and computerised tools for use by the FMW, RSDT and other related agencies with the aim of achieving sustainable operation. The configuration and calibration will be done for the latest version of HDM-4 (Version 2.08).

HDM-4 simulates future changes to the road system from current existing conditions. The reliability of the results is dependent upon two primary considerations:

How well the data provided to the model represent the reality of current existing conditions and influencing factors, in the terms understood by the model; and

How well the predictions of the model fit the real behaviour and the interactions between various factors for the variety of conditions to which it is applied.

The former relates to the correct interpretation of data input requirements and achieving a quality of input data that is appropriate to the desired reliability of the results. The latter refers to calibration of outputs, and it concerns adjusting the model parameters to enhance how well the forecast and outputs represent the changes and influences over time and under various interventions in Nigeria.

The prediction of road deterioration constitutes the backbone for life-cycle analysis and economic assessment in the HDM-4. Thus, the importance of carrying out a proper calibration of the road deterioration models to Nigeria conditions can scarcely be overstressed.

It is required that the HDM-4 works activities be adapted to model the performance and effects of the different road works types in Nigeria.

OBJECTIVES

The HDM-4 system requires several inputs that require configuration and comprises several models that are designed to be calibrated to local conditions in order to ensure accurate prediction of pavement performance and road user costs. As part of the HDM-4 implementation project in Nigeria it is necessary to configure the model and calibrate the various models in HDM-4:

Road Deterioration and Works Effects (RDWE)

Road User Effects (RUE)Traffic Characteristics

Configuration of the model is needed in the following areas:


80

Vehicle Fleet Characteristics

Road Works Unit Costs

Asset Value Calculation Parameters

Maintenance and Improvement Standards

Accident Rates

Others (Climate zones, speed flow type, traffic flow patterns, road network aggregate data, economic parameters, etc.)

Each of these will be configured and calibrated by the appointed consultant jointly with the aid of staff of RSDT. This section describes the proposed tasks required for calibration and adaptation of HDM-4 to Nigeria conditions.

There are three levels of calibration for the HDM-4, which involves low, moderate and major levels of

effort and resources, as follows:

Level 1 – Basic Application (Configuration)

Determines the value of required basic input parameters, adopts many default values, and calibrates the most sensitive parameters with best estimates, desk studies or minimal field surveys.

Level 2 – Calibration

Requires measurement of additional input parameters and moderate field surveys to calibrate key predictive relationships to local conditions. This may entail slight modification of the model source code.

Level 3 – Adaptation

Undertakes major field surveys and controlled experiments to enhance the existing predictive relationships or to develop new and locally specific relationships for substitution in the source code of the model.

In terms of effort, these levels can be viewed as weeks, months and years. A Level 1 calibration can be conducted in weeks. For a Level 2 calibration there is an increase in the amount of effort required so it will take months. Level 3 calibrations require a long-term commitment to basic data collection so their duration will be for a year or more.

For this project, the target level of the configuration and calibration study is Level 2 calibration, and the calibration will concentrate on the most important parameters as measured by impact sensitivity. The ultimate objective is to adapt the HDM-4 works activities to model the performance and effects of the different road works types in Nigeria and also for budgetary planning process.

Impact elasticity is the ratio of the percentage change in a specific result to the percentage change of the input parameter, holding all other parameters constant at a mean value. For example, if a 10 per cent increase in traffic loading causes a 2.9 per cent increase in roughness developed after 15 years, the impact elasticity term of traffic loading for that roughness result is 0.29. If there were a 2.9 per cent decrease, the value would be -0.29.

HDM-4 has four classes of model sensitivity, which have been established as a function of the impact elasticity. The higher the elasticity, the more sensitive the model predictions are. These classes are listed in Table A2.1.


81

Table A2.1 HDM-4 sensitivity classes

Impact Sensitivity class Impact elasticity

High S-I > 0.50

Moderate S-II 0.20 - 0.50

Low S-III 0.05 - 0.20

Negligible S-IV < 0.05

Source: HDM-4 Series Volume 5 “A Guide to Calibration and Adaptation”.

HDM-4 Volume 5 recommends that efforts should be based on the results of these sensitivity analyses. Those data items or model coefficients with moderate to high impacts (S-I and SII) should receive the most attention. The low to negligible impact (S-III and S-IV) items should receive attention only if time or resources permit. One usually assumes the default HDM-4 values for S-III and S-IV items since these will generally give adequate results.

HDM-4 Volume 5 discusses the configuration, calibration and adaptation of HDM-4, thus, shall be used by the consultants as a guide for their work. The consultants shall review the HDM-4 calibration efforts done in other countries that are documented on the http://www.lpcb.org/lpcb/ website at:

http://www.lpcb.org/lpcb/index.php?option=com_docman&task=cat_view&gid=22&Itemid=32

SCOPE OF THE CONSULTANCY SERVICES

Study

The Level 2 calibration will be based on analysis of records held in the FMW, FERMA and RSDT. However, it will be necessary to carry out field surveys as well as additional monitoring of pavement performance and collection of vehicle fleet and traffic data in order to provide complete data sets necessary to achieve this level of calibration.

The Consultant during the calibration of the HDM-4 to Nigeria conditions needs to take information from the existing records, historical and present survey of road inventory, pavement condition, traffic, maintenance and improvement standards unit costs of road works, vehicle characteristics, and vehicle operating costs (VOC) such that a Level 2 calibration can be achieved. This will require a careful review and understanding of:

Past working methods and materials used for the construction and maintenance of the road network.

Deterioration profiles for each road surface class and pavement structure type observed from survey data collected.

Current methods of road works, standards and works effects applicable to periodic maintenance; the material properties, expected life and unit costs of works activities.

Traffic information and speed-flow relationships

http://www.lpcb.org/lpcb/

http://www.lpcb.org/lpcb/




82

Road user effects including vehicle characteristics and utilisation, vehicle resource consumption and unit costs, and road safety.

Environmental impacts

Subsequently, the Consultant would be required to conduct studies to calibrate the different components of HDM-4 as follows:

Pavements and materials - to calibrate road deterioration and works effects models.

Road user effects – to calibrate the VOC models, fuel consumption, travel time valuation and accident costing and other road user components.

Traffic – traffic characteristics, vehicle speeds, and accident rates.

The studies and final reports are expected to be completed in Nine (9) months.

The calibration procedure will involve the following aspects:

Desk study – to identify data and information sources, to collate information and carry out basic analysis

Sampling – to select pavement sections from different climatic zones, vehicle fleet operators and workshops from different regions, and road sections of different traffic flow characteristics, which are representative of Nigeria. Statistical analysis may be used in the selection process.

Field surveys and measurements – to collect data on pavement ages, traffic loading, road condition, pavement structural number, pavement type, and drainage environment type; to collect data on vehicle resources e.g. fuel consumption.

Evaluation – to determine the calibration factors and model coefficients, by comparing the initiation and progression of the different distress modes from the collected data and data available from existing records in Nigeria against the prediction of HDM-4 (with default parameters).

After the configuration and calibration work is done, the consultants shall setup an HDM-4 workspace containing all the configuration and calibration data collected for the study. This HDM-4 workspace and its accompanying documentation could then be distributed to persons interested on running HDM-4 in Nigeria. This HDM-4 workspace shall contain at least:

Nigerian Road network broken down by road classes for network strategic analysis;

Sample cement concrete road sections

Sample asphalt concrete road sections

Sample surface treatment road sections

Sample gravel road sections

Sample earth road sections

Vehicle fleet characteristics and calibration parameters

Traffic growth sets


83

Default unit costs of road works in financial and economic terms

Asset Value Calculation Parameters

Basic maintenance standards

Basic improvement standards

Sample evaluations for projects module

Sample evaluations for programmes module

Sample evaluations for strategies module

HDM-4 configuration parameters (traffic flow patterns, speed flow types, accident classes, climatic zones currencies, aggregate data, road deterioration calibration parameters)

Road Deterioration and Works Effects

Road deterioration is broadly a function of the original design, material types, construction quality, traffic volume, axle load characteristics, road geometry, environmental conditions, age of pavement, and the maintenance policy pursued. HDM-4 includes relationships for modelling Road Deterioration (RD) and Road Works Effects (WE). The RD relationships are used for the purpose of predicting annual road condition and the WE relationships are used for evaluating road works strategies and estimating road agency resource needs for road preservation and development. Thus, the relationships link standards and costs for road construction and maintenance to road user costs through road user cost models.

The calibration of RD and WE models will ensure that the true deterioration of the pavements are appreciated and that any modification to the HDM-4 source codes can be carried out, to predict the effect of both past works and new works on the future condition of the road infrastructure.

To calibrate the Road Deterioration and Works Effects models (RDWE), will require complete data sets on the different pavement types from different climate zones. For road deterioration, this will include both cross-sectional and time-series data. Pavement data should cover all the surface classes in use in Nigeria (bituminous, unsealed and concrete), as well as the different pavement types within each class (e.g. for bituminous pavements: asphalt mix on granular base, surface treatment on granular base, asphalt mix on stabilised base, etc.).

In terms of road works effects, data will be required on the immediate effects of maintenance treatments, unit costs, and unit rates of energy use, for each of the various works activities applicable in Nigeria. As some of these data will vary from one region to another (e.g. unit costs), the calibration and adaptation exercise should reflect this.

The calibration of RDWE should be divided into the following groups of tasks:

Adaptation of data – mapping of pavement types, strength measures, condition measures, and roadworks activities to those used by HDM-4.

Calibration of road deterioration models – for bituminous, rigid concrete pavements and unsealed roads.

Calibration of road works effects models – for bituminous, rigid concrete pavements and unsealed roads.

Definition of unit costs of road works.


84

The calibration should be focused on the roughness deterioration, but should not be limited to roughness, because roughness is the key road condition indicator for an economic evaluation.

The Consultant shall recommend any additional monitoring of pavement performance and collection of data that may be required to refine the calibration of HDM-4 models for Nigeria.

Traffic

Traffic impacts all the other models in HDM-4 and in particular the Road User Effects (RUE) and Road Deterioration and Works Effects (RDWE) models. The results of economic analyses are quite sensitive to traffic data, and most benefits that justify road improvements arise from savings in road user costs. To perform economic analyses in HDM-4, traffic characteristics of roads therefore need to be described and represented at an appropriate level of detail. Traffic volume is a Class I parameter for the RDWE model and annual loading a Class II parameter. Speed and vehicle damage parameters are Class II and III parameters for the RUE model

The calibration of the HDM-4 traffic models will involve collection of data which reflect the different traffic situations (e.g. urban, rural, inter-urban and seasonal), and different road types (e.g. four-lane, two-lane, single-lane). Both motorised transport (MT) and nonmotorised transport (NMT) from different regions of Nigeria will be considered in the study. Data will also be collected on the accident rates occurring on the various classes of roads and junctions on the Federal road network.

The tasks for the calibration of traffic relationships have been grouped under the following:

Prediction of total annual traffic volumes, traffic composition, and growth rate

Vehicle mass and axle loading to determine vehicle damage (to pavement) factor

Vehicle free speeds prediction for the different road surface classes

Speed-Flow relationships (capacity restraint model) for different road types

Traffic flow pattern (hourly distribution of traffic volume) for different categories of road use

Accident rates - categorised in severity levels: fatal, injury, and damage only for different road types.

The Road User Effects Model

Road user effects comprises vehicle operating costs and travel time costs for both motorised and non-motorised transport, road safety, and environmental effects such as pollution caused by exhaust emissions and traffic noise.

The calibration of the RUE models will ensure that the predicted magnitude of each VOC component and the relativity between the different components conform to that observed in Nigeria. This will require the collection of vehicle fleet data (including non-motorised transport) from different regions of Nigeria. The RUE model predicts vehicle operating resources as functions of the characteristics of each type of vehicle and the geometry, surface type and current condition of the road, under both free flow and congested traffic conditions. The operating costs are obtained by multiplying the predicted quantities for the various resource components by the unit costs or prices, which are specified by the user in economic terms.

The tasks for the calibration of RUE relationships have been grouped as given as follows:

Representative vehicles - A detailed definition of representative vehicles, both MT and NMT is required.


85

Vehicle resources - the following components of vehicle resources are considered, and each has its separate model: fuel and lubricating oil consumption, tyre wear, vehicle utilization, parts consumption, maintenance labour hours, capital costs (depreciation and interest), crew hours, and overheads.

Unit costs of vehicle resources in financial and economic terms

Travel time - considered in terms of passenger-hours during working and non-working time, and cargo holding hours.

Accident costs – for the different severity levels: fatal, injury, and damage only.

Vehicle basic characteristics and key calibration parameters (for example: annual utilization, hours driven, service life, number of passengers, power, aerodynamic coefficients, fuel consumption, etc.)

The impact elasticity of RUE parameters is set out in Table 4.2, Chapter 4, Volume 5 of HDM-4 documentation series (Bennett and Paterson, 2000). The most sensitive Class I RUE parameter is the new vehicle price, which affects the depreciation, interest and parts consumption models. The parts consumption model itself is also very sensitive. Priority will, therefore, be given to the calibration of the depreciation and parts consumption models. Most of the Class II parameters affect the depreciation and speed models.

A procedure will be recommended for the regular updating of the vehicle fleet characteristics. This will include the identification of the minimum annual data set required to achieve this purpose.

Other Parameters

There are other general parameters apart from the models that need to be configured to local conditions for the correct application of HDM-4. These include climate zones, road network aggregate data, look-up table for missing data, unit costs of road works, traffic growth sets and economic parameters.

Climate Zones

The climate in which a road is situated has a significant impact on the rate at which the road deteriorates and on vehicle movement. Important climatic factors are related to temperature and precipitation. The principal climatic data that is used to model the deterioration of the different categories of roads considered in HDM-4 is described in section 4.1, Part C1 Volume 4 of HDM-4 documentation series (Odoki and Kerali, 2000).

HDM-4 requires the climate parameters to be specified for each distinct climate “zone” within a country. It is currently envisaged that there would be a minimum of six zones: North-East (NE), North-West (NW), North-Central (NC), South-West (SW), South-East (SE) and SouthSouth (SS).

Road Network Aggregate Data

HDM-4 requires the calibration of eight Road Network Aggregate Parameter Tables. These contain default data values for use in appraisal when more detailed survey data is not available. These parameters are:

AADT

Geometry

Compaction Quality

Roughness


86

Pavement Strength

Bituminous Layers

Surface Distresses

Surface Texture

Data is divided into a number of classes such as low, medium or high traffic volumes. For road geometry five classes are used to summarise average geometry from a generally level and straight road to a winding severely undulating road. The number of classes can be varied and should be defined by the consultants, who will also characterize numerically each class.

Look-up Table

This consists basically of a list of representative road sections with all the data requirements completed and used as defaults. The missing data for a given road section will assume the default values of a representative section onto which it is mapped.

The consultants will also create a matrix of representative road classes and will estimate the breakdown of the Nigerian road network length considering these road classes. This matrix of road classes could then be used for a network strategic analysis with HDM-4.

Economic Parameters

The economic parameters that need to be defined are:

Discount rate

Analysis period

Appropriate ‘base case’ alternatives

Salvage value of road

The choice of discount rate and analysis period will be agreed with the RSDT and the relevant funding institutions. A discount rate of 8 - 12 per cent and an analysis period of 20 years have generally been used to date in road appraisals. A method to estimate the salvage value of road works at the end of the analysis period will also be defined. This is most important with new road construction.

APPENDIX B: HDM-4 ANALYTICAL FRAMEWORK

87



88

HDM-4 Analytical Framework

The basic unit of analysis in HDM-4 is the homogeneous road section. Several investment options can be assigned to a road section for analysis. The vehicle types that use the road must also be defined together with the traffic volume specified in terms of the annual average daily traffic (AADT).

The analytical framework of HDM-4 is based on the concept of pavement life cycle analysis, which is typically 15 to 40 years depending on the pavement type. This is applied to predict road deterioration (RD), road works effects (WE), road user effects (RUE), and socio-economic and environmental effects (SEE) (Odoki and Kerali, 2000). The underlying operation of HDM-4 is common for the project, programme or strategy applications. In each case, HDM-4 predicts the life cycle pavement performance and the resulting user costs under specified maintenance and/or road improvement scenarios. The agency and user costs (i.e. RAC and RUC, respectively) are determined by first predicting physical quantities of resource consumption and then multiplying these by the corresponding unit costs.

Two or more options comprising different road maintenance and/or improvement works should be specified for each candidate road section with one option designated as the base case (usually representing minimal routine maintenance). The benefits derived from implementation of other options are calculated over a specified analysis period by comparing the predicted economic cost streams in each year against that for the respective year of the base case option. The discounted total economic cost difference is defined as the net present value (NPV). The average life cycle riding quality measured in terms of the international roughness index (IRI) is also calculated for each option.

The overall logic sequence for economic analysis and optimisation is illustrated in Figure B1. This figure shows the following (Odoki and Kerali, 2000):

The outer analysis loop - enables economic comparisons to be made for each pair of investment options, using the effects and costs calculated over the analysis period for each option, and it allows for variations in generated and diverted traffic levels depending on the investment option considered.

Effects, costs and asset values - how annual effects and costs to the road agency and to the road users, and asset values are calculated for individual road section options.

Optimisation procedures and budget scenario analysis - these are performed after economic benefits of all the section options have been determined.

Multiple criteria analysis - provides a means of comparing investment options using criteria that cannot easily be assigned an economic cost. Note that this capability has not been used in the present study.


89

Figure B1: Analytical Framework Input data requirements


90

The main data sets required as inputs for HDM-4 analyses are categorised as follows (Kerali et al., 2000):

Road network data comprising: inventory, geometry, pavement type, pavement strength, and road condition defined by different distress modes;

Vehicle fleet data including vehicle physical and loading characteristics, utilisation and service life, performance characteristics such as driving power and braking power, and unit costs of vehicle resources;

Traffic data including details of composition, volumes and growth rates, speed-flow types and hourly traffic flow pattern on each road section;

Road works data comprising historical records of works performed on different road sections, a range of road maintenance activities practised in the country and their associated unit costs.

Economic analysis parameters including time values, discount rate and base year.

Predicting road deterioration

Road pavements deteriorate as a consequence of several factors, most notably: traffic volume and loading, pavement design, material types, construction quality, environmental weathering, effect of inadequate drainage systems, and works on utilities. The HDM RD models are deterministic models which were developed using a structured-empirical approach (Paterson, 1987). This combines the advantage of both theoretical and experimental bases of mechanistic models with the behaviour observed in empirical studies. The type of model used for predictive purposes are incremental recursive and this gives the annual change in road condition from an initial state as a function of the independent variables.

Road deterioration is modelled in terms of cracking, ravelling, potholes, edge-break, rutting, roughness, friction and drainage. Roughness draws together the impacts of all other pavement distresses and maintenance. It is the dominant criterion of pavement performance in relation to both economics and quality of service as it gives most concern to road users. For each pavement type and each distress type there is a generic model which describes how the pavement deteriorates. To take account of the different behaviour of a particular pavement type constructed with different materials, the coefficients of the generic model depend on the different combinations of the materials. After maintenance treatments the generic pavement type can change.

Choice and effects of maintenance actions

Standards refer to the levels of conditions and response that a road administration aims to achieve in relation to functional characteristics of the road network system. The choice of an appropriate standard is based on the road surface class, the characteristics of traffic on the road section, and the general operational practice in the study area based upon engineering, economic and environmental considerations. In HDM-4, a standard is defined by a set of works activities with definite intervention criteria to determine when to carry them out. In general terms, intervention levels define the minimum level of service that is allowed. Road agency resource needs for road maintenance are expressed in terms of the physical quantities and the monetary costs of works to be undertaken. The annual costs to road agency incurred in the implementation of road works are calculated in economic and/or financial terms depending on the type of analysis being performed. The cost of each works activity is considered under the corresponding user-specified budget category (capital, revenue or special).


91

When a works activity is performed, the immediate effects on road characteristics and road use need to be specified in terms of the following: pavement strength, pavement condition, pavement history, road use patterns, and asset value. The long-term effects of a works operation are considered through the relevant models, for example: rate of road deterioration, changes in road user costs, changes in energy use and environmental impacts. Thus, both the immediate and long-term effects are combined to determine the benefits of carrying out different sets of roadworks activities at different times over the analysis period.

Predicting road user effects

The impacts of the road condition, road design standards, and traffic levels on road users are measured in terms of road user costs, and other social and environmental effects. Road user costs comprise vehicle operation costs (fuel consumption, tyre wear, oil, spare parts, depreciation, interest, crew hours and overheads), costs of travel time - for both passengers and cargo holding, and costs to the economy of road accidents (i.e., loss of life, injury to road users, damage to vehicles and other roadside objects). The social and environmental effects modelled in HDM-4 comprise vehicle emissions and energy consumption (Odoki and Kerali, 2000).

Motorised vehicle speeds and operating resources are determined as functions of the characteristics of each type of vehicle and the geometry, surface type and current condition of the road, under both free flow and congested traffic conditions. The operating costs are obtained by multiplying the various resource quantities by the unit costs or prices. Thus, the annual road user costs are calculated for each vehicle type, for each traffic flow period and for each road section alternative.

Optimisation

The NPV computed for the different section alternatives are used by the optimisation process to select the best alternative for each road section subject to the budget constraints not being exceeded. The optimisation problem therefore becomes one of searching for the combination of road investment alternatives that optimises the objective function (e.g. maximisation of economic benefits) under a budget constraint. The set of investment alternatives to be optimised is user-defined in such a way that they result in a different selection of treatments and it is not the set of all possible options for the network.

APPENDIX C: MINUTES OF NEGOTIATION MEETING

92



93


94


95

APPENDIX 1: OBSERVED, CALIBRATED AND DEFAULT DETERIORATION PLOTS

96

APPENDIX D: OBSERVED, CALIBRATED AND DEFAULT DETERIORATION PLOTS


97


98


99


100


101


102


103


104


105


106


107


108


109


110


111

hdm 4_revised_draft_final_report_v03

Documents