Pilot Implementation of HDM-4 in Bangladesh

Overview

This article describes the on-going full-scale pilot implementation of the HDM-4 system in Bangladesh. The project is funded by the UK Department for International Development (DFID) as part of the Institutional Development Component (IDC) managed by WSP International. This is part of the second Road Rehabilitation and Maintenance Programme (RRMP) in Bangladesh which is co-funded by the World Bank, the Asian Development Bank, DFID and other bilateral donors. The IDC project, has amongst its objectives, the implementation of a sustainable road maintenance management system (RMMS) with HDM-4 as one of the key management tools within the overall RMMS for Bangladesh.

The Roads and Highways Department (RHD) within the Ministry of Communications, is responsible for the management of approximately 20 850 km comprising three categories of road classes; National, Regional and Feeder type 'A' roads.

Organisational structure

The network activities of RHD have been divided into seven zones (regions) headed by Additional Chief Engineers. The zones are further sub-divided into divisions and sub-divisions. Different categories of engineering personnel from senior level to junior level are based within each zone and are responsible for design, planning and management of road maintenance. Road works are carried out by both contractors and the RHD labour force.

Table 1:The RHD road network in Bangladesh

Within the RHD headquarters, management of the road network is done by the Network Management wing which comprises 5 circles (departmental sections) responsible for; Planning and programming, Procurement, Economics, Maintenance, and the HDM system. Each circle is headed by a Superintending Engineer.

Road Network Asset Database

Records of the entire RHD road network are held within a central database managed by the HDM circle within RHD. There are separate controlled databases for bridges, road links, pavement condition attributes and traffic data (see Figure 1). The RHD network is divided into road links defined as sections of road with homogeneous traffic characteristics, typically between two major junctions. Each link is further subdivided into segments on the basis of following parameters;

Several road inventory attributes are stored for each link in addition to the zone, circle, division, and sub-division through which the road link passes. Road links are demarcated by nodes (i.e. intersections, boundaries, population centres, etc.). The bridges database holds records of all bridge structures under the jurisdiction of the RHD. Attribute data held include the road link, geographic location, bridge type, condition, etc. The traffic database contains records of the traffic volumes counted at fixed locations within the network. The traffic data is assigned to each link in the road network through a cross reference table. The above databases were initially developed using Microsoft Access and FoxPro, but have since been ported to an SQL Server RDBMS in order to provide multi-user access. A geographic information system (GIS) provides facilities for displaying management information in map format. Data from all databases can be extracted and displayed in map format using ArcView GIS software.

Pavement Condition Data

Data on pavement condition are collected at the zone level by Executive Engineers within each division. Local field survey teams are supervised at sub-division level to carry out visual inspections of all road links and record these against the distance chainage. The data collected include the extent and severities of cracking, potholes, rutting, edge break, and depressions as well as the condition of the shoulders, side drains, and any embankments. Training on data collection was provided by the IDC consultants, WSP International, who continue to monitor and co-ordinate the annual data collection.

In addition to the visual condition data, roughness surveys are conducted using vehicle mounted bump integrators (VMBI), calibrated using the TRL MERLIN device. Data on pavement strength are collected periodically by local consultants on contract to RHD. Benkelman beam deflection surveys and DCP tests are carried out to determine the pavement structure and pavement strength measured in terms of the pavement structural number (SNP).

The Economics circle within RHD is responsible for collecting and updating vehicle operating cost (VOC) data. These are based on the vehicle fleet calibration carried out by Mott MacDonald consultants as part of the IDC project. The HDM-VOC model was used in previous years by the IDC consultants to determine the calibration data for the vehicle fleet in Bangladesh.

Integration with HDM

The HDM circle within RHD is responsible for producing annual maintenance programmes for the road network. This is done in conjunction with the Economics circle. In the past, annual road maintenance programmes as well as the 5-year road development programme, were prepared manually by consolidating lists of candidate road projects put forward by the zones.

One of the key objectives of the IDC technical assistance funded by DFID, is to impart the expertise for a more rational analysis of road maintenance and development needs through the application of economic criteria embodied in the HDM model in order to prioritise and appraise candidate road maintenance and development projects. Over the past three years, the HDM-III model has been successfully adopted by the HDM circle who are now able to use the model for the appraisal of individual road projects. However, the preparation of annual work programmes requires the analysis of several road sections selected from the network using defined criteria.

The implementation of the HDM-4 system is designed to provide a tool for both project appraisal and for the prioritisation of a multi-year periodic road maintenance programme and a road development programme. In addition, HDM-4 will be used to conduct strategic analyses to determine long term funding requirements in order to achieve road network performance targets set by the government.

The first step prior to the HDM-4 analyses involves the dynamic segmentation of the road links in the database into homogeneous road sections based on the measured road condition. The segmentation criteria apply user defined tolerance levels to demarcate homogeneous road sections defined in terms of pavement structure, condition, traffic levels and geometric characteristics. A road sections table is then created within the RMMS database in preparation for the economic analysis using HDM-4.

The HDM circle has established a number of criteria for screening candidate road projects including;

• road sections whose pavement condition exceed defined intervention levels (e.g. wide cracking > 10% or roughness > 6 IRI on National roads, etc.)
• road sections with inadequate traffic flow capacity (e.g. volume/capacity ratio > 0.8 or average speeds less than 40 km/h for more than 8 hours of the day on National roads, etc.)
• unpaved roads with average annual daily traffic (AADT) exceeding specified thresholds (e.g. AADT > 400 on Regional roads, etc.)
• road sections with high occurrence of accidents (from the TRL road accident database - MAAP)
• road sections affected by flood damage (based on data from inspections carried out immediately after the monsoon season)
• other criteria defined by RHD (e.g. road sections with high volumes of non-motorised traffic, or road sections which require immediate routine maintenance).

The physical attributes of the selected list of road sections are then exported to a data exchange file format defined for HDM-4. This permits all road section data required by HDM-4 to be imported directly from the RMMS SQL server. Data transformation rules have been implemented for converting the pavement condition data held in the RMMS database to the form used by HDM-4. For example, pothole data is recorded in the RMMS database in terms of the % area of the pavement surface. This is converted to the equivalent number of standard pothole units (10 litres by volume) required in HDM-4. Similarly, other data required by HDM-4, such as pavement deterioration calibration factors, are inserted as pre-defined default values according to the type of pavement, road class, and other defined factors. Other data required for the HDM-4 analyses are stored within the HDM-4 internal database. These include data on vehicle fleet characteristics, road maintenance and improvement standards, unit costs and economic analysis parameters (e.g. discount rate, analysis period, etc.)

Figure 1

The main outputs expected from HDM-4 within the RMMS are the multi-year work programmes prioritised in descending order of the economic Net Present Value divided by the economic cost of the project (i.e. incremental NPV/C ratio) (see figure 2). This provides an efficient ranking index in line with the efficiency frontier concept. Other outputs include project level economic indicators (NPV, IRR, FYB, etc.) and, at the network level, the long term budget requirements, and long term average network performance indicators.

Figure 2

Future enhancements

It is planned that additional features will be implemented to enhance the overall functionality of the RMMS. These will include a project monitoring module for keeping track of road works in progress. Records of monthly achievements are kept by the RHD Procurement circle in order to provide data for monitoring project status.

A budget allocation system is also planned. This will allow senior RHD engineers to select projects from annual work programme in order to assemble the budget allocation for each zone and for each road class.

A comprehensive training programme for RHD has been implemented by the RHD consultants. Over the past 4 years, a number of RHD staff have received training to Masters degree level in a range of disciplines including highway management, transport economics, information technology, GIS, etc. Special training on HDM has been provided to several RHD staff and this is planned to continue.

In order to ensure the long term sustainability of the RMMS, the SQL database as well as the GIS were developed and are maintained by local IT consultants within Bangladesh. A high level of interaction has been maintained between the IDC consultants, local IT consultants and the HDM-4 team from the University of Birmingham. It is anticipated that the RMMS implementation will be completed by July 2000.

Further details on the above pilot implementation of HDM-4 and the RMMS may be obtained from the HDM-4 Technical Secretariat. Contact Dr Henry Kerali and Mr David Wightman. E-mail: ISOHDM@bham.ac.uk

Acknowledgement: The project described in this Newsletter is funded by DFID and the Bangladesh Government. The project is managed by WSP International as part of the institutional development of the Roads and Highways Department in Bangladesh. The project would not succeed without the assistance provided by these organisations.