D7 Final

Deliverable D7
Decision on repair/replacement
Status (P)
BRIDGE MANAGEMENT IN EUROPE

(BRIME)
PL97-2220
Project
Coordinator:
Dr R J Woodward, Transport Research Laboratory (TRL)
Partners:
Bundesanstalt fuer Strassenwesen (BASt)

Centro de Estudios y Experimentacin de Obras Pblicas (CEDEX)
Laboratoire Central des Ponts et Chausses (LCPC)
Norwegian Public Roads Administration (NPRA)
Slovenian National Building and Civil Engineering Institute (ZAG)
Date: June 2000

PROJECT FUNDED BY THE EUROPEAN
COMMISSION UNDER THE TRANSPORT
RTD. PROGRAM OF THE
4th FRAMEWORK PROGRAM
Decision on repair/replacement
by R. Astudillo Pastor, J.M. Arrieta Torrealba,
C. Velando Cabaas and C. Lozano Bruna
Deliverable D7
P97-2220
CONTENTS
Page
Executive Summary
Abstract
1. Introduction
2. Literature review
2.1. Bridge Management Systems. Present and future. P.ThoftChristensen.
2.2. Managing the life-cycle safety of deteriorating bridges.
D. M Frangopol and G. Hearn.
2.3. Bridge Management Methodologies. P. C. Das.
2.4. Condition and repair cost estimate of the french road bridge assets.
C. Binet .
2.5. The actual state of the Bridge Management System in the state
national highway network of Spain. M. A. Yaez and A. J. Alonso.
2.6. Introduction to Bridge Management Systems. P. Clausen.
2.7. Bridge Management-the answer to the challenge. J. J. Ahlskog.
2.8. A Network Optimization System for Maintenance and
Improvement of Californias bridges. K. Golabi, P. D. Thompson.
2.9. DANBRO A Bridge Management System for many levels.
N. H. Andersen.
2.10. Emerging Methodologies for Bridge Management Systems.
K. C. Sinha, M. D. Bowman, Y. Jiang, S. Murthy, M. Saito, A. Tee.
2.11. New Developments in Bridge Management Methodology.
P. C. Das.
2.12. Bridge Management for New York City. B. S. Yanev.
2.13. Creating a Bridge Management System. J. Lauridsen, J. Bjerrum,
N. H. Andersen, B. Lassen.
2.14. Development of a Bridge Management System in Germany.
J. Naumann.
2.15. Bridge Maintenance Training: experience and future trends.
A. Huvstig.
2.16. The Pontis Bridge Management System. P. D. Thompson,
E. P. Small, M. Johnson, A. R. Marshall.
2.17. Whole-life costing. A. E. K. Jones, A. R. Cusens.
2.18. Concrete bridge assessment: an alternative approach.
C. R. Middleton.
2.19. The BRIDGIT Bridge Management System. H. Hawk, E. P. Small.
2.20. The Finnish Bridge Management M. K. Sderqvist, M. Veijola.
3. Review of existing decision system for bridge repair/replacement
3.1. Decision on maintenance and repair in various countries
3.2. Other management systems used abroad
3.3. Present state of bridge management in Spain
4. Review of commercial Bridge Management Systems
1
3
4
5
5
6
7
8
9
10
11
13
21
30
33
35
36
44
46
47
52
55
58
62
64
64
70
72
73
4.1. PONTIS
4.2. DANBRO
5. Theoretical models for repair/replacement
5.1. Frangopol
5.2. Branco&Brito
6. Decision system for repair/replacement
6.1. Introduction
6.2. Methodology
6.3. Identification of the factors
6.4. Evaluation of the factors
6.5. Comparison of the alternatives
7. Example
7.1. Approach
7.2. Conclusions
Bibliography
73
90
111
111
123
131
131
132
133
135
142
144
144
147
153
EXECUTIVE SUMMARY
SCOPE
Europe has a large capital investment in the road network including bridges,
which are the most vulnerable element. As bridges age, deterioration caused by
heavy traffic and an aggressive environment becomes increasingly significant
resulting in a higher frequency of repairs and possibly a reduced load carrying
capacity.
The purpose of this project is to develop a framework for the management of
bridges on the European road network that enables bridges to be maintained at
minimum overall cost i.e. taking all factors into account including condition of the
structure, load carrying capacity, rate of deterioration, effect on traffic, life of the
repair and the residual life of the structure.
SUMMARY
This report proposes a method for decision-making regarding possible actions in a
deteriorated bridge. The main objective of this work package is to prepare a
decision criteria that helps to choose the best repair decision considering safety,
durability, functionality and economy.
The method is based on a global cost analysis that considers all the costs involved
in designing, constructing, inspecting, maintaining, repairing, strengthening and
demolishing a bridge, as well as the road user costs associated with the service
life of the bridge.
The objective remains to develop a strategy that minimizes the global cost while
keeping the lifetime reliability of the structure above a minimum allowable value.
The method consists in the proposal of alternatives for the repair or replacement
of a deteriorated bridge with resistant or functional problems. The global cost of
each alternative will be evaluated through a set of factors. The selection of the
most suitable repair/replacement alternative will be based on the comparison of
these costs. This method allows the choice among alternatives that depend on
numerous factors that can be of a very different nature.
The possible alternatives must take into account the use of different types of
repairs, the different moments in which each one of the actions can be
implemented along the service life of the bridge and the possible replacement of
the structure, that will be considered as another alternative.
The actions considered in this method intend to restore the initial service level
(design) of the bridge, without considering an improvement of its initial
performances: dimensions, load carrying capacity, etc. Nevertheless, this method
could be also used when all the considered alternatives lead to the same level of
improvement in the bridge. In other cases it will be necessary to complete the
method taking into account the benefits that involve each alternative for the repair
or replacement of the bridge.
IMPLEMENTATION
The results of this study will be used to develop a framework for the management
of bridges. It should also stimulate further improvements of existing decision
procedures and development of new ones.
DECISION ON REPAIR/REPLACEMENT
ABSTRACT
An extensive literature review regarding the available documentation on bridge
management systems, in general, and, specifically, on the methods for selecting
the more convenient option for repair or replacement of a bridge has been made.
Also a brief study about decision procedures in the different countries has been
carried out analysing the replies of a questionnaire that was sent to the partners
and other documents.
A detailed study about two commercial bridge management systems has been
developed: PONTIS system and DANBRO system. Using the technical manuals
of these programs, each module and each component of the programs has been
analysed, emphasizing the decision system for bridge repair/replacement and the
methodology to evaluate user costs.
Two theoretical models have been selected among the models found in the
literature review because of its special interest: the model proposed by D. M.
Frangopol (University of Colorado) and the proposed by F.A.Branco and J. Brito
(University of Lisbon). A description of each one of this method can be found in
this Deliverable.
Finally a method for decision-making regarding possible actions in a deteriorated
bridge is proposed. The main objective of this work package is to prepare a
decision criteria that helps to choose the best repair decision considering safety,
durability, functionality and economy. The method is based on a global cost
analysis that considers all the costs involved in designing, constructing,
inspecting, maintaining, repairing, strengthening and demolishing a bridge, as
well as the road user costs associated with the service life of the bridge.
1 INTRODUCTION
The importance of bridge repair versus new bridge construction has become of great
concern to bridge authorities. This is due to the high structural deterioration rates that have
been observed in some structures or to the lack of functionality that sometimes has
occurred.
The disruption of each particular bridge has very high costs for society. It stops traffic
passing over it and forces thousands of persons to use alternative routes at extra cost and
time.
There is therefore a need for a set of rational criteria that ensure that bridges are maintained
in a safe and serviceable condition, with the required load carrying capacity. This must be
done throughout their design life at a minimum life-time cost whilst causing the least
possible disruption to traffic.
The analysis of the costs of alternative maintenance procedures highlights the need to
quantify such factors as the cost of traffic delays, the deterioration rate of bridges, the
effective life of repair systems and the time value of money. When the quantification of
items such as these has been achieved it is possible to put forward a programme of
maintenance optimised to achieve a set standard condition at minimum long-term cost.
2. LITERATURE REVIEW
The first stage in the elaboration of a decision system on repair/replacement of bridges was
an extensive literature review regarding the available documentation on bridge management
systems, in general, and, specifically, on the methods for selecting the more convenient
option for repair or replace a bridge.
A detailed search in Spanish and international databases and libraries has been done. In
addition documents received from other BRIME partners, technical information in
conference proceedings, scientific magazines and the proceedings of Bridge Congresses
over the last few years are being analysed.
After a final selection of the most interesting papers, publications and documents, a resume
has been prepared for every selected document. These resumes allow the reader to get a fast
but precise idea of the content of the documents.
The extension of resumes varies accordingly to the content of the analysed documents, i.e.,
the extension is larger in those works that directly deal with the subject of selecting the best
alternative among several possibilities for repair/replacement a bridge.
The selected and analysed information can be included in two groups:
Publications that regard, in a very general context, with the bridge management systems:
origin, need, fundament, characteristics, use, research and future, etc.
Documents that describe specific BMS developed in some countries or analysing
particular modules of these BMS mainly focused in the repair/replacement decision or
the maintenance strategies, the evaluation of user costs, etc.
2.1 BRIDGE MANAGEMENT SYSTEMS. PRESENT AND FUTURE.

P. Thoft-Christensen
Recent Advances in Bridge Engineering. J. R. Casas, F. W .Klaiber and A. R. Mar
(Eds.). CIMNE Barcelona 1996.
The article centres especially on the management systems for reinforced concrete bridges.
Three sections are developed after the introduction.
The introduction reasons with the need of maintenance in concrete structures and with the
possible damages that can affect their behaviour.
The principal characteristics that must observe the future bridge management systems are
briefly described in the first section, together with the parts or modules that should be
developed. The following aspects are specifically treated:
a) Adequate strategies for the inspection and maintenance of bridges; relation between
defects and inspection methods; definition of the fundamental parameters to be analyzed
and the correct methods to do it, etc.
b) Development of possible options for inspection, maintenance and repair. The optimum
decision should be based on the expected benefit and the total cost of inspection, repair
and maintenance, as well as the possible cost of an structural failure. In any case, the
reliability must be acceptable during the service life of the structure.
c) Application of Expert Systems. A division between the systems used by the inspector to
carry out the inspections and the systems that help the engineer in the analysis of the
bridge safety and in the selection of the best option for its maintenance or repair can be
outlined.
In the next section are briefly treated the most outstanding modules in the management
systems for reinforced concrete bridges: deterioration and its evolution in time, stochastic
models of inspection, stochastic models of repair, updating of the information and analysis
of the bridge reliability.
Finally, some commentaries on the results of the working group Assessment of
Performance and Optimal Strategies for Inspection and Maintenance of Concrete Structures
using Reliability Based Experts Systems, of the European Research Programme
BRITE/EURAM, are included.
2.2 MANAGING THE LIFE-CYCLE SAFETY OF DETERIORATING

BRIDGES.
D. M. Frangopol and G. Hearn
This article summarizes a research work carried out by the University of Colorado. A new
methodology for coding and recording conditions of bridge members that will make
possible meaningful automated evaluations of load capacity and safety is presented. Each
element is divided into several segments, with each segment receiving a single condition
rating in a field inspection. A segment of an element is a part of the element that is bounded
by physical landmarks that are present on the bridge or is a single physical unit of the
element. Each segment exists at a specific location on the bridge. Therefore, the load
demand in a segment is known. Also, condition ratings can be related to strength. With
information on both load demand and strength for each segment, a load rating can be
calculated for each segment. Segment based data includes information on the extent of
deterioration and the location of deterioration that is present on each element of the bridge.
Since load demands and element strength differ at different regions of an element, it is
necessary to know the location of deterioration, as well as the extent of the deterioration.
For example, load rating of a truss may be significantly reduced if deterioration is present in
a member that has high live loads and dead load demand, while deterioration in other
members may not affect the load rating of the truss.
To use segment based data for strength and load rating evaluations, it is first necessary to
relate element condition states to the remaining strength of the element.
Once condition state definitions are in terms of remaining strength, the remaining strength
of each segment of an element can be estimated based on the current condition data. Since
the location of each element is known, the live load demand from a standard truck loading
on each segment can be calculated. The dead load demand on each segment can also be
calculated. The current load rating for an element would then be lowest rating of all the
elements segments. The current bridge load rating would then be the lowest rating of all the
elements on the bridge.
Finally, some aspects related with the rational planning of the inspections and with the use
of the minimum total cost method during the bridge service life, in order to decide on the
actions to be taken in deteriorated structures, are included.
2.3 BRIDGE MANAGEMENT METHODOLOGIES.

P. C. Das
This article deals with the need to consider all the possible risks that can affect a structure in
a bridge management system. The bridge management systems available nowadays take
only into account the defects and deterioration when establishing the maintenance and
repair programmes. The need to develop new management systems that will cover other
type of risks, such as those derived of inadequate premises in the original project, is
considered in this work. The bridges subjected to some of these risks may not give signs of
damage or deterioration because they have not reached yet the pessimum situation.
To evaluate the risks, a series of concepts of structural reliability are introduced: probability
of failure, reliability index, safety coefficient, etc.
Finally, the actions that are being carried out in the Highways Agency of the United
Kingdom to develop or establish a bridge management system are described. A brief
description of the Programme PONTIS, together with the proposal of some improvements
or extensions, are given in this section.
2.4 CONDITION AND REPAIR COST ESTIMATE OF THE FRENCH

ROAD BRIDGE ASSETS.
C. Binet
The article describes the method used by the French Ministry of Transport to evaluate the
state of the bridges of its national road network.
The objective was to give an indicator called IQOA (Image de la Qualit des Ouvrages
dArt) representative of the overall condition of the bridge assets. After this first evaluation,
the project currently in progress has for objective to give an evaluation of the repair cost of
all the assets.
Numerous data on the French bridge patrimony are included in the first place: distribution
of bridges by materials, by structural typologies, by length, by age, etc.
The way to assess the condition of each bridge, the tools used and the results obtained in a
first evaluation made in 1994 of the whole patrimony are described immediately after.
Bridge conditions were rated according to the following system:
1 Bridges in apparently good condition.
2 Bridges showing defects of equipment or protection components, or minor structural
defects without need of repair.
2E As above but with urgent need of repair in order to prevent more advanced structural
deterioration.
2S As above but with urgent need of repair in order to guarantee the safety of the road users.
3 Bridges with structural damage.
3U As above but with urgent need of repair.
For each bridge, every component has been rated and an overall rating called IQOA has
been assigned.
In order to help all the actors in the classification and to ensure the homogeneity of the
rating on the whole territory, two following types of documents are produced:
-
inspection report guides made in such a way that people without specialised knowledge
of bridges should be able to note down the different defects observed; theses documents
are richly illustrated and include the complete list of all the defects that is possible to
find in a given type of bridge.
Catalogues of the main disorders explaining the defects and their possible causes
However these results must be taken with much care because the rating assigned to a
bridge reflects the condition of the worst component and not the average condition.
Finally, the results obtained of an experience carried out to know the repair cost of all the
patrimony are included. The method used consists of the final estimation of the repair cost
in a sample of bridges chosen at random for the later extrapolation of the results to all the
population.
2.5 THE ACTUAL STATE OF THE BRIDGE MANAGEMENT SYSTEM IN

THE STATE NATIONAL HIGHWAY NETWORK OF SPAIN.
M. A. Yaez, A. J. Alonso
This article presents the actions that are being carried out by the Directorate General of
Roads of Spain for the conservation, maintenance and repair of the national network
bridges, as well as the basis of a new project that has been started to establish a bridge
management system.
The type of inventory used in the national road network is described firstly and relevant
data included: percentage of bridges by materials, by provinces, etc.
The steps taken until now to classify and evaluate the bridges , in order to detect those that
require actions, are briefly described next.
Finally, a new programme under development for the establishment of a bridge management
system for the national network is presented. For the moment, a pilot project with 43
bridges of the network, that includes the results and conclusions that have been considered
of interest, has been carried out.
2.6 INTRODUCTION TO BRIDGE MANAGEMENT SYSTEMS.

P. Clausen
14 Th Congress New Delhi 1992 IABSE.
This article describes briefly the different modules that compose a bridge management
system and analyzes the principal characteristics of each of them and the possible
improvements and future developments.
The following modules are specifically analyzed:
-Databases.
-Inventory.
-Inspections.
-Maintenance.
-Planning of future repairs, rehabilitations or substitutions.
The use of cost models during the service life of the bridge, in order to compare the
different alternatives at all times, is proposed. A discount rate is used to pass from future to
present costs. To help to decide between the options of repair or substitution of the bridge,
the article includes a graphic in which, starting from the relation between the repair and
substitution costs, the extended service life (years) and the discount rate, the less costly
option is obtained.
According to the article, it is necessary to include in the costs of each alternative the users
cost, that can be determining in some occasions.
-
Classification and ordering of priorities among all the necessary actions in the network.
Study of routes for heavy transport.
10
Finally, the possible working lines to improve the bridge management systems are
enumerated, together with some reflections for the future users.
2.7 BRIDGE MANAGEMENT-THE ANSWER TO THE CHALLENGE.

J.J. Ahlskog
Bridge Evaluation, Repair and Rehabilitation. Edited by Andrzej Nowak.
The article begins with a series of reflections on the dilemma arising when it must be
decided if a bridge should be repaired, substituted, how it can be maintained, etc. This has
an incidence on the limitations entailed by the budgets allocated for these entries, that are
generally inferior to those necessary to keep all the patrimony under adequate conditions:
portage capacity, capacity, clearances, etc.
The future forecasts and the conclusions derived from this analysis are briefly analyzed
next: increase of the budget for bridge maintenance, development and improvement of the
repair techniques, national and regularized procedures to select the adequate option; repair,
substitution, etc.
According to the author, the use of these decision systems by the organizations and the
engineers in charge of the maintenance of the network bridges will make possible to select
the most adequate option in each case, independently of political questions, errors in the
engineers or responsible persons judgement, etc.
The elements that must comprise a bridge management system are also included: inventory,
databases, analysis, forecast of needs, assessment of options and planning.
In each case, the knowledge of the needs is complex, since these depend upon numerous
factors. The author proposes two definitions:
-
Needs are the actions of less cost that are necessary to reduce the differences between
the real conditions of the bridge and the optimum or desirable conditions, to that the net
benefit for the society is maximized.
Needs are the actions that maximize the net benefit for each bridge and, therefore,
represent the socially optimum choice among a wide series of alternatives.
By applying any of these definitions to a number of possible alternatives for the

improvement of a bridge, a single choice should result, the optimum for the bridge users.
11
According to the article, one method of applying bridge management techniques to bridge
improvement decisions is to determine the alternative improvement options for each
substandard bridge, estimate the improvement cost for each alternative and estimate the user
costs incurred by the public for each of the alternatives.
The incremental benefit/cost ratio is determined by taking each increment of benefit and
dividing it by each increment of cost. At some point there will be an increment of benefit
which equals the increment of cost. This is the optimal point for improvement.
If this process is repeated for all substandard bridges and the projects are listed in the order
which lists the highest incremental benefit/cost ratio projects first, the resulting list will be
the optimal list of projects in priority order.
The author includes some considerations and conclusions derived from the application of
this procedure, such as:
-
For normal values of discount rates, it is more convenient in most cases to repair or
improve a bridge than to replace it.
The long term tendency in the United States of America is to increase notably the
bridges maintenance and repair budget and to reduce the replacement budget.
Finally, it is analyzed the state of the legislation with respect to bridge management and the
future tendencies, including a small reference to the USA states that have a management
system and the extent of its application.
12
2.8 A NETWORK OPTIMIZATION SYSTEM FOR MAINTENANCE AND

IMPROVEMENT OF CALIFORNIAS BRIDGES.
K. Golabi, P. D. Thompson
This article describes briefly the Programme Pontis. The motives that led the Federal
Highway Administration FHWA to develop a bridge management system, called Pontis, for
the Department of Transport of California, with the intention to apply it afterwards to the
rest of the states, is presented in the introduction.
The author describes later the various tools that have been developed in several states in
order to facilitate the bridge management and reflects on the need to develop a system that
tackles the whole problem, formed by interrelated subsystems, flexible and based on
mathematical and economic concepts.
Some differences and special characteristics of the bridge management systems with respect
to other management systems, such as those for pavements, etc, are included in the next
section. A major distinguishing factor between the bridge management problem and most
other maintenance problem (pavements, pipelines, etc.) is the fact that improvement
implies two sets of distinct activities. The first set (calling maintenance, corrective and
rehabilitation) constitutes maintenance and corrective actions. Such actions, as correcting
scour conditions or replacing wearing surfaces, improve the condition of the bridge, which
may deteriorate again with time. The second set (calling improvements) essentially deals
with functional aspects, and once performed do not change with time. This set includes
actions such as deck widening or bridge raising to gain vertical clearance. Recommended
improvement actions are driven by state-specified level of service goals, or by the optimal
levels of those standards recommended by a consideration of budget availabilities, the
respective costs and benefits of various standards, and the backlog generated by any set of
level of service standards. The set of recommended actions dealing with maintenance, and
their budget requirements, need to be generated through a dynamic model, as there the
condition is time-depended and at least for some of the actions, there is a tradeoff between
their cost and the time until the next action is necessary. To be specific, if two actions
achieve a desiderable condition at this time, and one is less expensive then the other but has
a higher probability of deterioration in the future, the model must be able to address that
tradeoff and incorporate it in its calculations. Based in this understanding, a bridge
management system should have two distinct, but interrelated components to address
maintenance and improvement.
The following point deals with the objectives of Pontis. According to the author, these are
the following:
13
1. Provide a systematic procedure for setting repair and improvement priorities.

2. Provide a systematic procedure for helping decision makers set realistic level-of-service
goals for each of 12 predefined functional highway classifications.
3. Incorporate level-of-service goals in assessing bridge improvement needs and budget
requirements.
4. Provide a capability to consider the entire bridge network simultaneously in arriving at
recommendations, while retaining the flexibility to address any subset of bridges.
5. Provide priority orders and sequencing for bridges in need of improvement.
6. Coordinate maintenance planning decisions with future improvement decisions.
7. Be based on a methodology that is rigorous, general and flexible. The two latter
characteristics are essential for transferability of the system from one state to another
when state specific subsystems are modified.
8. Require a reasonable and realistic amount of data for model input, provide procedures
for data collection and checking the reliability of the collected data, assess the relative
importance of data items, establish reliability standards commensurate with relative
importance, and allow for incorporation of engineering judgement with objective data
whenever necessary.
9. Consider the differing inspection and repair needs of the major structural components
for bridges as well as the differing needs of the various types of bridges.
10. Address the probabilistic nature of bridge deterioration and the differences in the
deterioration patterns of various components. The model would be able to consider the
effect of design characteristics, geometry, environmental factors and maintenance
history on future deterioration of bridge components.
11. Allow for updating of predictive probabilities as the necessary data would become
available over time.
12. Consider the relative importance of the various bridges in terms of safety, risk exposure,
public convenience and regional economics.
13. Consider the immediate and future costs and benefits of the various courses of action
and their effect on future conditions. In particular, the model would weigh the benefits
of preventive maintenance versus costlier (but less frequent) corrective actions.
14
14. Allow sensitivity analysis of the recommended policies in terms of future conditions of
the bridge network and maintenance and replacement cost requirements. This would be
flexible to accommodate different state-specified maintenance and fiscal policy issues.
15. Provide least-cost estimates for correcting deficiencies if budget constraints are
removed. Compute costs over a planning horizon and prepare annual budget
requirements to address those needs.
16. Provide a basis for short-term and long-term maintenance and improvement budget
planning and resource allocation.
17. Consider the possible availability of Federal and special funds, the restriction on those
funds and the availability of matching funds.
18. Consider economies-of-scale involved in repairing/replacing either adjacent bridges or
several components of a single bridge.
19. Be easy to use by maintenance planners and engineers. Using the system would not
require knowledge of operations research, statistical methodologies or computer science.
20. Provide flexibility for bridge engineers to exercise their professional judgement in
making day-to-day decisions.
21. Provide a framework for reconciling divergent interests (e.g. increasing the ability of the
system to transport more loads versus preserving historical bridges).
22. Provide a basis for equitable funding to States bridge programs by the
government, and assist in allocating funds to different districts by each state.
Federal
23. Centralize the maintenance decision process and reduce gaps between the condition of
similar bridges within the network.
24. Provide rigorous procedures and an analytical framework for incorporation of expert
engineering judgement in the model.
Finally, the organization of the programme, its different components and the way in which
these are related to obtain its objectives, are described.
As a result of the interconnectedness of the system, the organization of the optimization
submodels has a determining effect on the data and organization of the other submodels in
the system. Two divisions within the problem which are most fundamental are:
15
Near-term vs. long-term programming. Long-term programming concerns itself with the
agencys vision of where it would like to be, eventually. It begins with a set of general
goals and, for an optimization problem, defines these goals in a manner which can be
succinctly stated and quantified (e.g. minimize social costs). The outputs of long-term
programming are maintenance and improvement policies in a general sense; funding
allocations as they apply to long-term strategies to ensure reliable funding levels; and
near-term objectives for the condition of the bridge infrastructure. Near-term
programming starts with these objectives and finds a way to get there from here. It is
concerned with the identification and resolution of specific identified or predicted bridge
needs; the listing of specific construction and maintenance projects; the development of
a program that can be shown to maximize movement of the infrastructure toward the
long-term goals (or minimize backsliding), within known budget constraints; and shortterm efforts to ease those constraints.
Maintenance and rehabilitation vs. improvement and replacement. Maintenance

management is the pursuit of a most efficient way to keep the existing bridges in
operation at their current level of service (except when the posted load has been reduced
because of deterioration of bridge components). It begs the question of what the
required level of services should be, or even whether the bridge should remain open;
instead, it assumes that operations must continue and that deterioration must be detected
and remedied before operations are affected, at minimal cost. Improvement
management, on the other hand, analyzes what the level-of-service standards should
be, identifies instances where the standards are not met, develops strategies to meet
them, and prioritizes and sequences such improvements.
The main components of the system are as follows:

1. INPUT DATA BASE
The input data base allows for storing all information needed to run the model as well as its
updating as more information becomes available over time. The data base retrieves and
stores the inventory, the physical characteristics of bridges, condition survey data and other
relevant information from the agencys data files. It will also store traffic information, load
characteristics, and cost information; as well as the main elements of each bridge, the
condition states defining the possible conditions of each element, and the set of feasible
corrective, maintenance and improvement actions associated with those conditions.
2. SELECTION OF FEASIBLE ACTIONS
The optimization procedures address the selection of feasible actions in a consistent,

integrated way. Algorithms determining feasible actions represent two basic relationships
involving maintenance and improvement activities.
16
They establish a correct mapping between the current condition of a bridge and the set of
activities that are technologically appropriate to correct the type, severity, and extent of
damage that is present, or to prevent damage that would otherwise occur in the future.
They embody criteria or references that help resolve overlaps among activities, and that
establish a clear hierarchy among the activities to be applied.
These algorithms are needed because there is not a simple, one-to-one relationship between
bridge damage and the activity available to correct or prevent that damage. Furthermore,
major activities may subsume the effects of one or more subsidiary activities.
3. COST SUBMODELS
The Cost Submodels consist of the following components:

1. Improvement Costs: Calculates the costs associated with improvement of each bridge
to achieve the level of service goals based on material quantities for each feasible
improvement and the States unit cost data. These costs are inputs to the Improvement
and Replacement Needs Submodel.
2. Replacement Costs: Calculates the replacement cost for each bridge when replacement
is a feasible option. These costs are input to the Improvement and Replacement Needs
Submodel and possibly to the Long-Term Maintenance Optimization Submodel.
3. User Cost Submodel: Calculates the user costs associated with the present and improved
condition of each bridge. The User Cost Submodel provides inputs to the Long-Term
and Near-Term Maintenance Optimization Submodels as well as to the Improvement
and Replacement Needs Submodel. Some of the measures (attributes) that are
considered in the development of the user costs are:
Travel time and fuel consumption due to detours
Travel time and fuel consumption due to traffic congestion
Travel time, fuel consumption, and vehicle depreciation due to deck roughness
Travel time due to sharp turns or other geometric problems
17
Accident costs
4. Maintenance, Rehabilitation & Corrective Action Costs: This submodel calculates the
costs of the feasible actions associated with preventive, rehabilitative and corrective
measures by computing the necessary material quantities and using the States unit cost
data. The model consists of a set of detailed equations that would attempt to estimate the
various costs as accurately as possible, associated with each element, condition, and
feasible action. Either bridge-specific or aggregate data may be used.
A desirable feature of the system is its ability to quantify the cost of deferred maintenance
and improvement. Generally, the costs associated with deferring actions that need to be
taken now, but because of budgetary restrictions need to be postponed, are quantified and
taken into consideration by the optimization models. However, the results can also be
directly provided as outputs. If the bridge is in need of improvement the deferred costs
consist mainly of user costs (including expected accident costs) until the deficiency is
removed. These costs are quantified for every bridge and provided for bridges which are not
chosen, as well as those chosen, for improvement. In its determination of the optimal
maintenance policy for each bridge, Pontis considers the possibility of future needs for
expensive corrective actions (including replacement) if appropriate corrective actions are
not taken at the present time. It would also consider the risks associated with different
conditions, and the conditions to which they are likely to deteriorate if proper maintenance
actions are not taken. These costs and risks are also quantified for those bridges whose
maintenance action has been deferred for budgetary reasons, and in particular for those
bridges where no tradeoffs between the various maintenance actions exist.
4. DETERIORATION PREDICTION SUBMODEL
This model estimates the deterioration rates associated with each major bridge element as a
function of its present condition, and the feasible maintenance and corrective actions
(including routine maintenance only). The deterioration of bridges is affected by many
factors. These factors can be classified as those related to the initial design of the bridge,
quality of construction methods and materials, and the external environment of the bridge.
To these one may add age and previous major maintenance and corrective actions.
Current models of deterioration tend to focus on the time-dependent change in some index
of bridge condition, and reflect very little of the cause and effect relationship that results in
damage and deterioration; therefore, they are not only subject to variability in their
predictions, but they also require further subjective assessments to relate levels of
deterioration to the activities that may be required. Realistically speaking, cause-and-effect
models of bridge damage mechanisms cannot be developed for all mechanisms on all bridge
components. The problem is too complex, and sufficient field data are not available.
Nevertheless, such theoretical or empirical relationships do exist (or can be derived from
published data) for certain components and for certain damage mechanisms; these are
18
employed in conjunction with expert opinion and statistical analyses of historical data to
develop a more informed judgment and calibration of the deterioration prediction model.
5. OPTIMIZATION OF MAINTENANCE, REHABILITATION, AND CORRECTIVE
ACTIONS
This optimization submodel has two parts, long-term and near-term. In the long-term, the
model combines the deterioration models with agency cost (and to a less important extent,
user cost) models to analyze the long-term performance of each individual structure and the
network as a whole. The products of the long-term maintenance optimization are expected
long-term conditions, expected maintenance budget needs, and optimal policies for
corrective and maintenance actions.
In the near-term, the focus is on finding optimal maintenance policies. The near-term is
much more specific about individual projects, and considers existing condition and
economies of scale. Although the processes, and therefore the models, are stochastic,
deterioration and environmental effects are more predictable in this model. Outputs in the
near-term are concerned with binding commitments which the agency must make:
scheduled maintenance actions and budgets. Both of these models emphasize the role of
agency expenditures in relieving deterioration, thereby extending the life of the structure
and reducing future agency costs. The models consist of smaller interrelated optimization
models and are capable of addressing a variety of policy and engineering issues, including
the lag time between planning and construction.
6. OPTIMIZATION OF IMPROVEMENTS AND REPLACEMENT
In this model level-of-service goals are considered explicitly. By definition, projects in this
category are the only ones which change the level of service offered by a structure. The two
submodels within this component of the system consider the possibility of allowing a bridge
to be posted or closed, and they consider the raising of posted limits or the improvement of
geometric or other functional attributes, including the replacement of a bridge with one
having better level-of-service characteristics.
The two outputs of this model are a list of improvement projects, and their priorization. In
the maintenance models, there is a long-term vs. near-term dichotomy, because the
perceived needs are always far greater than what can be done in the near future. The model
proceeds in the generation of improvement needs by first settling on a set of necessary
and desired level of service goals, which may vary according to various definitions of
sub-networks (e.g; by functional class). Next, it aggregates the bridges in the network into
units of analysis for the application of these goals.
19
The model is capable of functioning in two modes: (i) which bridges should be upgraded to
the desired level and which should be postponed, and (ii) what is optimal level-of-service
in the context of overall network benefits.
7. CAPITAL PROGRAM INTEGRATION
Since the above-mentioned models have as outputs two separate programs, one for
maintenance and one for improvements, what remains is to integrate the two into combined
long-term and short-term capital programs. The improvements optimization will have
identified bridges for upgrading and downgrading, and the maintenance optimization will
define a maintenance program for all of the remaining structures. Overall, the integration
process is a matter of coordinating the optimization results with each other and with prior
decisions, emergency work construction, and other over-rides. This is accomplished by a
separate model that considers the scheduling and Federal constraints on improvement and
major maintenance, as well as engineering dependencies of structural components on each
other.
20
2.9 DANBRO A BRIDGE MANAGEMENT SYSTEM FOR MANY

LEVELS.
N. H. Andersen
Danbro is a computer-supported BM-system that is in use in Denmark today. This paper
takes a look at the system from the user side with emphasis on the various levels at which
the system can be utilized to support bridge management. The concept of the system how
its components work together is explained. So is the organization which the system is
designed to support. The activities which are covered by the system are mentioned, and the
build-up of the EDP part is described what data the data bases contain and what computer
programs are developed to process the data. The setup of the EDP system for both central
and decentral use is explained together with the way information is exchanged between the
various locations where the system is used. Finally, the possibility of using the system in a
modified version in other countries is discussed.
The author defines, depending on the users and the ends to be achieved, four levels:
A. The executive level.
B. The planning level.
C. The administrative level.
D. The maintenance level.
A. THE EXECUTIVE LEVEL
This is the level at which the decision-makers decide on future policy for bridge
maintenance and thereby about future budgets. The decision-makers are primarily the
management group at the Road Directorate which is responsible for overall policy, but also
the politicians who approve the annual State Budget for road administration may be
considered as belonging to this group.
The information we can feed back to the executive with the help of the system includes:
1. The current condition of the bridge stock.
21
2. The need for funds, if we choose to use the optimal repair strategies on all
bridges.
3. The backlog we build up, if we do not exceed the given funds over the next few
years, together with the additional cost incurred.
4. The consequences for the road standard and traffic flow if the budgets are kept
unchanged or are suddenly changed.
B. THE PLANNING LEVEL
The chief of the Bridge Maintenance Division is in charge of the planning of bridge
maintenance. This has to be done according to the policy determined at the executive level.
At the planning level the following decisions are taken:
1. How much to use on routine maintenance out of the total budget.
2. An optimization of the repair and rehabilitation works, to suit the allocation set
aside for this type of work.
3. A decision on what new types of repair methods and material are to be
introduced and what older types are to be abandoned.
C. THE ADMINISTRATIVE LEVEL
This is the level at which data about the bridges are handled. Inspectors, office technicians
and budget supervisor collect and use these data for their work.
The data can include:
1. Administrative data on the bridges.
2. Technical data on construction of the bridges.
3. Inspection data on the condition of the bridges.
4. Repair strategies for bridges in need of remedial works.
5. Construction costs and repair costs of bridges and bridge components.
22
6. Test data about deterioration i.e. location, type, cause and development.
D. THE MAINTENANCE LEVEL
The practical routine maintenance is dealt with at this level. The local bridge engineer plans
and monitors the work.
He uses the system for the follow purposes:
1. Work orders for routine maintenance.
2. Work orders for periodic maintenance.
3. Control of the quantities for the above-mentioned works.
4. Control of budgets and costs for routine maintenance.
5. Monitoring the quality, costs and service life of materials used for maintenance.
After they have looked at the use they can make of the system they take a look at the
components of the system. These are the following:
-
A set of interrelated activities for bridge handling.
A set of codes and rules for the activities.
A set of data bases where data resulting from the activities are stored.
A set of EDP-tools to process the collected data.
In the following point The Maintenance Organisation are described how it is organized
the maintenance and bridge management in Denmark and the Organizations and human
resources in charge.
The author continues the article with a description of the programme activities, that are
summarized below:
-
Inventories
Principal inspections
23
Special inspection
Routine maintenance and inspection
Optimization and budgeting
Design of remedial works
Tendering
Execution of remedial works
Budget and cost control
Administration of special transports
Feed-back
A set of manuals describes how these activities are to be carried out and how data about
them are to be registered in the data base.
1. INVENTORIES
Inventories are carried out whenever a bridge has been completed or when remedial works
have been carried out.
At the inventories all documents such as drawings, calculations, soil reports, tender
documents and construction reports are collected, an index for all archive materials is set up
and the most important data are extracted. These data can be of administrative or technical
nature.
The administrative data which are of interest include: bridge name and number, owner,
designer, construction year, number and name or roads and their relation to the structure.
The technical data can be bridge type, construction materials, geometry including
clearances, and load-carrying capacity.
2. PRINCIPAL INSPECTIONS
Principal inspections are visual inspections which are carried out at set intervals to register
the condition of the structure. Important damage is registered and all components are rated
24
on a scale from 0 to 5. A rough estimate of the costs of remedial works to be carried out
over the next ten years are made and the times when they should be carried out are
registered. All registration is made on tape recorders.
3. SPECIAL INSPECTIONS
Special inspections are thorough investigations where the size and type of deficiencies are
established together with their causes. The present and future influence of the defects on the
stability of the structure are evaluated.
Two or three repair strategies are considered for a 25 year period and their economic
consequences evaluated. The consequences of a 5 year postponement are also evaluated.
The corresponding road user costs are taken into account, for each strategy.
4. ROUTINE MAINTENANCE AND INSPECTION
Many minor cleaning and maintenance works e.g. cleaning of manholes, maintenance of
bitumen joints, repair of pot holes and of concrete spallings have to be carried out at suitable
intervals. These types of jobs are carried out routinely together with superficial inspections,
where only major abnormal conditions are considered.
5. OPTIMIZATION
For bridges in need of remedial works an optimization is made. This involves setting up a
priority list that states which bridges are to be repaired when, and according to what repair
strategy. The optimization is based on the 5 year budgets together with repair strategies
from special inspections and the forecasts from principal inspections.
6. ADMINISTRATION OF SPECIAL TRANSPORTS
Special permission has to be given for a road transport if it is heavier, wider or higher than
the limits given in the traffic regulations. Permission is given by the police authorities, if all
the bridges on the road stretch to be used permit the transport to pass without problems.
The police use lists retrieved from the BM-system for the checking.
All bridges have been rated and given load bearing classes which are given in the list the
police use.
25
The heavy vehicles have to be classified so the load of the vehicle can be compared to the
bridge classes. The classification is done routinely at the Bridge Department in a couple of
hours after a request from a truck owner is received.
The EDP part of the system, how the programme operates, the updating of information
between different terminals and the process or working plan that is being carried out in
Denmark are treated in the last points.
The EDP part of the system is developed in the DOS version of a third generation language
called Dataflex. It runs on IBM-compatible personal computers. The system is built up of
the following modules:
1. THE BASIC MODULE
The basic module contains the administrative and structural data bases and the principal
inspection data bases.
Further, a set of programs is included which gives the user of the system access to screen
pictures composed of information from the various data bases. Once can choose a screen
picture that gives a chronological review, showing the most important events for a bridge
since construction. It is also possible to run through a road in this module, so that one can
obtain information on clearances and load-carrying capacities for the structures on a given
road stretch. This can be used for the administration of high and heavy transports.
Also programs for print-out of inspection reports and other reports are part of the module.
2. THE MAINTENANCE MODULE
The maintenance module contains data bases with information about components of the
bridges which have to be cleaned or maintained at regular intervals. Also a data base
containing possible remedial works on the components is included.
The programs incorporated in this module can automatically print out work orders at set
intervals for the local bridge engineer when he has decided upon which bridge components
are to be maintained, which maintenance works are to be carried out on these components,
the starting date of a maintenance job and the time interval between repetitions.
3. THE PRICE CATALOGUE
This module is set up to support the estimation of the costs of repairs in connection with
general inspection, special inspection, and design.
26
The data base in this module contains unit prices for various works, based on tender prices.
A built-in program secures that all prices in the module are adjusted to present-day prices
before they are used for estimation.
Unit prices for specified items classified by geographical area or by the size of the work can
be calculated and printed. Also combined prices for major works such as deck rehabilitation
can be obtained in this module.
Finally, bills of quantities can be printed for tender documents.
4. THE OPTIMIZATION MODULE
The data bases in this module contain the economic data for the various repair alternatives
set up at the special inspections. These data, together with the economic data from the
principal inspection, are processed by the optimization program that is part of this module.
The program operates at the network level but incorporates information from the project
level.
By means of an iterative process, the structures which it is most economic to repair, and the
repairs which will cost least to postpone when budgetary limits do not permit carrying out
all repairs at the optimum time, can be found.
The net present value method is used, and the optimum use of budgetary allocations for a 5
year period is found by iteration.
A program can calculate the consequences of having budgetary allocations lower than the
actual demand. The consequences are additional expenditure because of insufficient
resources to carry out repairs at the optimum time, and also accumulation of postponed
repairs.
5. THE BUDGET AND COST MODULE
The data bases in this module contain information about budgets for maintenance and
repairs. Furthermore, account figures on the running works are registered.
Changes in works and budgets are registered continuously over the financial year. Budgets
are typically changed after a tendering or after a change in the work decided on during its
execution. The module can give information at any time on the expected total expenditure
for the year.
27
6. THE EXPERIENCE MODULE

It includes data on damage as well as economic data.
Data on damage consist of information on the position, type, extent, and cause of the
damage, as well as the materials involved. Also the results of measurements made in the
field or on specimens in laboratories will be stored. All the data are collected in connection
with special investigations.
The economic data include registration of the expenditures on each structure over a number
of years.
The stored data will be used for analysis to obtain information about service life and service
life costs of bridge and material types.
Finally the writer explains how the interrelated activities interact with each other and how
the data for these activities are stored in the data base, and used in the computer program.
Superficial inspections and routine maintenance are carried out by roadmen according to
lists prepared by local bridge engineers with the aid of the maintenance module.
Inventories are carried out at the head office of the Bridge Department. The information
collected is stored in the basic module.
Proposals for remedial works and the cost of these are stored in the optimisation module.
The costs are based on the unit prices from the price catalogue. Information about the
damage and measurements are stored in the experience module.
Optimization is carried out once a year when funds are made available, or whenever
changes in the allocated funds are made during the year. Budgets for remedial works to be
carried out are stored in the budget and cost module, and the local engineers are informed
about funds made available for bridges to be repaired in their district.
Design of remedial works are carried out by consulting engineers according to the repair
method selected at the optimization, and instructions from the local engineers. The design is
approved by the inspection engineer before tendering.
Remedial works are carried out by contractors after tendering. The economic data from the
tendering are stored in the budget and cost module. The data from the two lowest bids are
stored in the price catalogue.
28
Long-term budgets are planned on the basis of experience and the forecast of the
optimization module.
Special transports are handled by the police authorities, with the help of information from
the basic module.
2.10 EMERGING METHODOLOGIES FOR BRIDGE MANAGEMENT

SYSTEMS
K. C. Sinha, M. D. Bowman, Y. Jiang, S. Murthy, M. Saito, A. Tee
This paper presents one particular application of emerging tools and techniques in bridge
management systems. These techniques were used in the development of a statewide bridge
management system in Indiana (IBMS). Theories of fuzzy sets were used in bridge
structural condition assessment and bridge traffic safety evaluation, while the analytic
hierarchy process was adapted for priority ranking. Techniques of dynamic and integer
programming were applied to develop an optimization module where bridge structural
conditions were dynamically updated by using Markov chain principles.
The author summarizes firstly the bridge management system of the Indiana Department of
Transportation.
The system has eight modules that are used to evaluate various bridge system factors, and
incorporate their influence on the overall bridge performance and integrity.
Fuzzy set mathematics was used in the condition rating assistance and bridge traffic safety
evaluation modules. A detailed statistical analysis was conducted to develop appropriate
information for the life-cycle cost analysis submodule. A modified analytic hierarchy
process (AHP) was developed incorporating utility curves in the bridge project ranking
submodule. The use of dynamic programming, in combination with integer linear
programming and Markov chain, provided an efficient tool to select bridge projects with the
system effectiveness or benefit being maximized subject to budget constraints over a given
program period. The remaining modules primarily involve input and output data of the
bridge management system. The application of these techniques made the input data
consistent and reliable, and the IBMS effective and flexible. The core of the system consists
of condition rating assessment, bridge traffic safety evaluation, ranking and optimization.
The application of the Fuzzy Sets in the IBMS is treated in the next point through the
following sections:
a) Fuzziness and randomness.
29
Information available to bridge inspectors as to the structural condition or traffic safety is

generally imprecise and often can be separated into objective and subjective components.
Fuzzy set theory (Zadeh 1965), a systematic technique to quantify imprecise information in
the subjective component, can be effectively used to improve the bridge condition and
traffic safety evaluation process.
b) Bridge condition assessment module.
Bridge condition rating is one of the key parameters that is used in determining the types of
repair necessary for a bridge. The usefulness of a bridge management system depends upon
the reliability and accuracy of the bridge inspection information. The current bridge
inspection practices, however, suffer from three inherent shortcomings.
1. The parameters in bridge inspections are not completely defined or cannot be precisely
measured;
2. Personal judgement bias and subjectiveness are often included but not systematically
accounted for in the evaluation process;
3. There is a lack of guidelines establishing the relationship between the extent of
deterioration and the assignment of values of condition rating.
The purpose of the condition assessment module was to filter the field inspection data of
any inconsistencies before entering the bridge management system. The fuzzy set theory
(Zadeh 1975) was utilized to minimize the bias of human judgement and to enhance the
overall accuracy of the bridge condition evaluation. One of the existing methods of
combining fuzzy knowledge, the fuzzy weighted average (Juang and Elton 1986), was used
in the module. The fuzzy weighted average has a simple mathematical form:
R=
1
( Wi ri ) [2.1]
Wi
where, R is a fuzzy number denoting a bridge components resultant rating, ri denotes the
fuzzy rating of the ith subcomponent and Wi denotes the fuzzy importance factor for the ith
subcomponent.
c) Bridge traffic safety evaluation module.
Bridge traffic safety information is often ignored in decisions regarding bridge improvement
projects. Similar to bridge condition evaluation, bridge traffic safety evaluation can also be
affected by many factors and subjective judgements may be necessary to assess it. Therefore
30
the bridge safety evaluation module (Murthy et al. 1989) was developed based on fuzzy set
theory and the concept of linguistic variables.
The module allows bridge inspectors to translate their subjective judgements on bridge
traffic safety related factors into a quantifiable value. In order to assist inspectors making
consistent judgements about bridge traffic safety, an interactive computer program was
developed. In the program, the bridge inspector is asked to input work ratings for the bridge
components associated with traffic safety and the program computes an appropriate index
based on fuzzy set principles.
The author continues the article with the description of the Project Selection Module.
To select projects primarily on the basis of costs, the life-cycle cost analysis sub-module is
used. However, when other factors, such as structural condition, traffic safety, community
impact and so on, are considered, the ranking and optimization sub-modules are employed.
The ranking sub-module can be used to compare bridge projects on the basis of multiple
objectives. Even though the ranking program may not give a system wide optimal solution,
the list of ranked bridges can be a guideline for the bridge manager to identify which bridge
projects should be placed in the following programming period. The optimization submodule, however, yields a system wide optimal solution for any given budget level and
program period.
1. RANKING SUB-MODULE
Decisions about bridge rehabilitation and replacement depend on a set of criteria, such as
bridge structural condition rating, traffic safety, and impact on the surrounding community.
This situation is a typical multi-attribute decision-making problem, requiring simultaneous
evaluation of several related factors. The technique of the analytic hierarchy process (AHP),
developed by Saaty (Saaty 1980), was adopted for the ranking module. The AHP method is
a useful tool to rank projects when subjective judgments are involved.
2. OPTIMIZATION SUB-MODULE
The optimization sub-module applies dynamic programming and integer linear
programming to select projects while the effectiveness or benefit of a bridge system is
maximized subject to the constraints of available budgets over a given program period
(Jiang and Sinha 1989). Markov chain transition probabilities of bridge structural conditions
are used in the model to predict or update bridge structural conditions at each stage of the
dynamic programming. The dynamic programming divides the federal and state budgets of
each year into several possible spending portions, and the integer linear programming
selects projects by maximizing yearly system effectiveness subject to different budget
expenditures. The dynamic programming chooses the optimal spending policy, which
31
maximizes the system effectiveness over a program period, by comparing the values of
effectiveness of these spendings resulted by integer linear programming.
2.11 NEW DEVELOPMENTS IN BRIDGE MANAGEMENT

METHODOLOGY
P. C. Das
Structural Engineering International. Volume 8, Number 4, November 1998.
This paper describes the principles and basic steps of a new bridge maintenance bidding and
bid prioritisation procedure being developed by the Highways Agency for implementing in
the near future. It has the following innovative features:
-
It combines the strategic needs of the network with the maintenance needs of individual
structures.
It takes into account whole life costing, risks and options of maintenance bids. The risk
of no choosing a particular maintenance option is considered in terms of resulting traffic
disruption or future premature rehabilitation.
It makes assessment an integral part of the bidding process.
All engineering decisions are left to the engineer and the network manager rather than
automatically provided by the computer.
It takes account of deterioration related maintenance as well as those arising from other
needs, such as design deficiency. This is because the maintenance needs are determined
by whether a structural element is sub-standard or not in terms of its structural adequacy
and not determined from its condition rating alone.
The author analyzes in the introduction some questions related with the present bridge
management, that are summarized below:
Currently available bridge management systems (BMSs) are broadly based on two
principles: firstly, the bridge maintenance needs are directly related to the condition states
of the structures and, secondly, the justification for any proposed work is that it will cost
more later if the work is not carried out now. However, contrary to the first assumption,
experience gained from the UK Highways Agencys bridge maintenance activities shows
that the major part of the work depends on the load carrying capacity (or structural
adequacy) of the structures rather than on their conditions alone.
32
Any global prediction of future deterioration rates is likely to be very approximate; network
level predictions can therefore be used only for overall, guidance. The maintenance
requirements of a particular bridge need to be based on its own specific assessment, which
would normally indicate a number of maintenance options. The network level strategy can
then provide the overall guidelines for choosing the best options.
Furthermore, in the face of many competing urgent needs for funds, the argument of future
cost increase is not a very strong one for securing funds for bridge maintenance.
On the other hand, possible traffic disruptions resulting from large-scale weight restrictions
imposed on unsafe bridges is likely to be a more compelling reason for justifying funds.
As such, it is unlikely in practice that project level actions recommended by an optimised
network strategy will be reflected by the actual maintenance activities determined for
individual bridges. Therefore, in recent years there has been a search for a better correlated
network and project level BMS methodology.
The assessment and prioritisation procedure developed by the Highways Agency is
analyzed in the next sections. The following aspects have been taken into account:
The annual list of maintenance bids is made up of the bids determined for individual
structures or structural elements by the maintaining engineers.
All elements of a structure will have a critical performance level (assessment level) that is
normally related to safety considerations. If the performance of the element falls below this
level (i.e., It is sub-standard) it has to be replaced or strengthened as soon as practicable.
Such work is considered as essential or rehabilitation work. If current performance is
above this level, and yet some work is deemed to be justifiable on grounds of economy,
such work is considered as preventive.
A strategic plan was prepared by the Highways Agency in 1997 to determine its bridge
maintenance needs for the future. The following describes the significant stages of this
exercise.
Maintenance items and their typical cost were based on examination of past records of work
done for the Highways Agencys bridges. The cost were grouped together into two items,
preventive and rehabilitation.
The next step was to divide the bridges into a number of distinct types such as concrete,
concrete/steel, etc., and to obtain, from the inventory database, the numbers of bridges of
these types with their years of construction. The different types were assumed to have
different rehabilitation rates, with and without preventive maintenance. These are now being
33
examined for future use with the help of maintenance records, and also using probabilistic
analysis, in much grater detail.
The next step was to multiply the numbers of bridges built in any year with the assumed
rates of rehabilitation to produce the numbers that will require rehabilitation in the future
years.
The next stage was to determine the appropriate maintenance strategy for each bridge type,
i.e., whether preventive maintenance was justifiable in terms of the future needs of the
bridge stock. For this, the estimated numbers of bridges requiring rehabilitation and
preventive maintenance in each future year were multiplied with the average unit costs of
rehabilitation and preventive work in order to produce the total maintenance cost profiles
for the future years.
Based on the above assumptions, the future maintenance cost profiles for the Agencys
bridges were calculated in terms of both essential (i.e., rehabilitation) and preventive work
that may arise from possible future condition deterioration of the present bridge stock. As
previously defined, rehabilitation will be needed when a structure becomes sub-standard,
i.e., inadequate in terms of the assessment loading, and preventive work is the work done on
structures which are not yet sub-standard.
The strategic plan is intended to provide an overview of maintenance needs and in itself will
not be sufficient for allocating funds, which requires the assessment of project level bids.
The project level bids are based on whole life assessments, where the current safety level
(minimum acceptable level) appropriate for the element concerned to determine if the work
is essential (i.e., rehabilitation) or preventative.
The future performance of the element is then projected using a number of alternative
maintenance strategies. For each strategy option, and for each maintenance action in the
strategy, the year of the action, various costs including planning, supervision and traffic
management costs and, in addition, the traffic delay costs if the work is not carried out, are
submitted as bids.
Once the maintenance bids from all agents are received, the network level prioritisation can
be carried out. This is done by prioritising first according to the nature of the work, i.e.,
contractually or otherwise committed, essential and preventative, etc.; and then by work
categories as determined by the authority (e.g., pier strengthening, parapet replacement,
protection against scour, etc.). following this, the cost profiles are discounted to present
value (PV) and the maintenance option with the lowest PV is selected for each structure or
element.
34
Finally, all the selected bid items are totalled up to check against the strategic plan
estimates. This is because the cheapest option for a scheme may not be the best option in
terms of the strategic considerations of the whole stock. For instance, if the cheapest options
for the whole stock totalled to, say, 20 bridges for rehabilitation in a particular year, and the
strategic plan indicates that 60 bridges are to be rehabilitated , the discrepancy needs to be
investigated. In such cases there is a possibility that an unacceptable backlog of
rehabilitation work may build up in the future. Hence, the project level bids may need to be
adjusted to include some options which may not be the cheapest project options in order to
bring some of the postponed rehabilitation work forward.
The final prioritised list should show, for instance, from the bottom of the list, decreasing
funding levels and increasing traffic delay costs for no providing the full funding. Those
involved in evaluating the options for different funding levels should then be able to draw a
line at the proposed level and get an indication of the extend of traffic disruption likely from
weight restrictions or other measures that would be necessary for maintaining safety if the
full bids is not funded. It should be noted that the traffic delay cost is only used here to
indicate the likely level of traffic disruption.
2.12 BRIDGE MANAGEMENT FOR NEW YORK CITY

B. S. Yanev
Structural Engineering International. Volume 8, Number 3, August 1998.
Bridge Management will always remain a process dependent on the interaction between
structures, vehicles, users, designers, builders, economists, politicians or, to various degrees,
the general public. Despite the uncertainties built into the bridge management process, a
number of conclusions are emerging.
Bridge Management and the computer-based BMS developed over the last decade are
distinctly different. In one manner or another, bridge management is as old as bridges. The
recent interest in BMS is part of the general effort to draw attention to certain deficiencies
of bridge management, to clarify and improve the process of selecting optimal bridge
management strategies under defined constraints.
In many computer-based bridge management systems, the decision-making module is either
absent, under revision or limited to allowing for the intervention of a qualified individual.
This is likely to remain the case in the future. At best, BMS will be in a position to offer a
recommendation for certain types of action under certain constraints.
The most significant contribution of a bridge management system is its comprehensive and
accessible database. The amount of maintenance and reconstruction work that has been
35
performed on the bridge, a history of intervention measures that is currently missing in

databases, would be a valuable enhancement.
The quantification of bridge component conditions remains a challenging task for engineers,
with considerable room of improvement. Non-destructive testing techniques and on-line
bridge monitoring are rapidly improving the level of knowledge about how bridges age.
Bridge component conditions are increasingly associated with the events they depend on,
rather than merely with time. Efforts to obtain uniquely defined relationships between the
condition ratings of bridge components and bridges on the one hand, and measurable
quantities on the other, are underway. PONTIS, the bridge management system developed
by FHWA, has recently introduced a health index relating the condition state of each
bridge element to that of the whole bridge.
Bridge management tasks are more readily met for smaller bridge networks. In recognition
of this fact, FHWA has sponsored the development of different systems. Nonetheless, the
bridge network serving a large metropolis such as New York City cannot be fragmented,
since all important decisions have to address the global consequences. Special
considerations for unique structures or structural subnetworks have to be incorporated as
details of the system.
The overall average condition of a large bridge network is neither a particularly sensitive
nor, therefore, significant indicator. Simply averaging data from a large number of bridges
to determine deterioration or condition patterns can lead to false results, due to the
incorporation of undocumented and unknown factors. Deterioration rates of bridges and
bridge components, in particular, have been generated by a number of authors for various
bridge networks. Most results have shown a convex curve, asymptotically approaching a
certain rating higher than failure. There has been speculation about the reasons for such a
pattern. Observations indicate that convex deterioration patterns for bridges and bridge
components are due primarily to repair work. The most realistic pattern is also the most
pessimistic one. It is obtained by reviewing specific cases where the fastest know
deterioration has been observed.
2.13 CREATING A BRIDGE MANAGEMENT SYSTEM

J. Lauridsen, J. Bjerrum, N. H. Andersen, B. Lassen
Structural Engineering International. Volume 8, Number 3, August 1998.
Bridge Management involves a number of activities, such as the collection of inventory
data, inspections, assessment of damaged structures, administration of heavy transport,
allocation of funds for repair and maintenance, etc; all with the purpose of ensuring traffic
safety and maintaining the bridge stock in the desired condition at the lowest possible cost.
The purpose of a Bridge Management System is to assist in the management of these
activities.
36
In the course of managing bridges, decisions have to be made all the time. As help in
making the right decisions, guidelines should be established and followed, but normally it is
not possible to set up rigid and objective rules for these decisions. Therefore even if
computers are very useful for storing data and indispensable for processing the large
amounts of data involved, e.g.; in the optimisation process Bridge Management cannot be
performed by computer systems alone. Consequently, a complete Bridge Management
System is not just a computer program. It is:
-
a set of interrelated activities for handling bridges
a set of codes and guidelines for theses activities
an organization to manage and carry out these activities
a database holding data resulting from these activities
a set of computer tools for processing the data in the database.
In the development of a Bridge Management System, it must be kept in mind that a

comprehensive computer system is not the goal. The goal is to assist the decision-makers
and the administration in doing their job not to replace them.
A Bridge Management System should comprise guidelines and management tools to be
used at all levels of bridge operation and maintenance: the executive, the planning, the
administrative, and the maintenance levels.
One of the main objectives of a Bridge Management System is to help the administration
maintain the bridges at the lowest possible cost.
The principles described in this paper for a Bridge Management System have been applied
to the management of bridges on the Danish roads for more than a decade, and they have
been the basis for Bridge Management Systems implemented in several other countries as
well.
The article is structured in three sections:
1. Structuring a Bridge Management System.
2. Selecting Rehabilitation Strategies.
3. Developing Specific Bridge Management Systems.
37
2.13.1 Structuring a Bridge Management System

As a consequence of the realization that different bridge owners have different needs, the
structure of a Bridge Management System should be modular. Of course, there are interrelations between the individual modules, but in general each bridge administration should
be allowed to choose which modules to implement. And within the administration, each
individual user may be given access to selected modules, in order to distribute the
responsibility for data.
The principal modules are:
1. INVENTORY
2. PRINCIPAL INSPECTION
The principal inspection is a visual inspection of all visible parts of the bridge. Its purpose is
to maintain an overview of the general condition of the whole stock of bridges, and to reveal
significant damage at an early stage, so that rehabilitation works can be carried out in the
optimum way and at the optimum time.
For the inspection, the bridge is divided into a fixed number of standard components, one of
which is the bridge in general. For each standard component, the following information is
registered:
-
A condition rating, ranging from No damage to Ultimate damage/complete failure of

component
A short description of significant damage (less significant damage is not described)
Need for routine maintenance/cleaning (Yes/No)
Need for Special Inspection
Need for repair works (type of work, extent, cost estimate, timing, duration).
Part of the output from the principal inspections is a running 5-year cost estimate for
rehabilitation works.
3. ROUTINE INSPECTIONS AND MAINTENANCE
4. BUDGET AND COST CONTROLS
38
5. LONG-TERM BUDGETING
In addition to the 5-year budgets created by the Principal Inspection Module, there is a need
for long-term estimates for bridge rehabilitation. The Long-Term Budgeting Module is
intended for this purpose.
The general idea is that average repair intervals and corresponding average repair costs, as
well as average service lives and replacement costs, are registered for the standard
components of the bridges. The year of construction, type and principal dimension of all
standard components are registered for each bridge. Based on this data, the program
calculates total future budgets.
6. PRICE CATALOGUE
As a tool for collecting and updating unit prices of common rehabilitation works in a
systematic way, the Bridge Management System comprises a Price Catalogue Module.
7. HEAVY TRANSPORTS
An important part of any Bridge Management System is the administration of heavy
transports. Normally, the administration is based on assigning classes to the bridges and to
the vehicles. In addition to the normal class, the bridges are assigned special classes based
on the vehicle being alone on the bridge, travelling on a specified lane on the bridge, and
with reduced speed.
As a general rule, a vehicle may pass over the bridge if the class of the vehicle is not greater
than the class of the bridge.
2.13.2 Selecting Rehabilitation Strategies
In a case where there is no doubt about the cause and extent of damage, or the proper
rehabilitation method, the inspector determines the actions to be taken at the principal
inspection. This section describes how to determine the right rehabilitation strategy when
this is not the case.
1. SPECIAL INSPECTION
Normally, special inspections are initiated at the principal inspection. When the inspector is
not certain about the cause or the extent of damage, or the proper rehabilitation method, a
special inspection may be called for. Essentially, the purpose of the special inspection is to
determine what is wrong and what should be done? This is done step-by-step:
39
determine the cause and extent of damage
predict the development of damage if nothing is done
identify relevant rehabilitation strategies
determine the cost to society of each strategy
choose the strategy that is least costly the society
2. REHABILITATION STRATEGIES
When identifying and evaluating relevant rehabilitation strategies, it is necessary to have an
open mind and not focus only on the condition of the bridge itself. The purpose of the
bridge is to carry traffic in a reliable and safe way, not to conform to any preconceived
solution.
In this connection, it is worth recalling that a rehabilitation strategy is a series of actions
carried out in order to make sure that the bridge fulfils its purpose. The full range of
relevant strategies may normally be divided into four types:
-
make a thorough repair now, bringing the bridge back to good as new condition
make some superficial repairs now in order to postpone major repair
do nothing now, wait until the bridge, or the component in question, is no longer safe,
then replace it
do nothing at all; when the bridge is no longer safe, close it and accept the road-user
costs that follow.
Within each of these strategy types, the inspection/bridge engineer determines in detail
which kind of work is called for, the optimum time of execution, etc. At the end of this subtask, the range of relevant strategies has been reduced to a few significantly different ones,
each the economic optimum within its type.
3. ECONOMIC EVALUATION
An economic evaluation is carried out, comparing the relevant strategies and assessing, for
each strategy, the technical and economic consequences of postponing said strategy. These
consequences are to be used for the optimization discussed below.
40
The Net Present Value Method is used for the economic evaluation, giving the net present
value of executing, or postponing, each strategy, including direct and indirect costs to
society within the next 25 years. The strategy with the lowest net present value is the
economic optimum for the bridge, and will normally be the one proposed as the conclusion
of the inspection report.
4. DIRECT AND INDIRECT COSTS
The costs considered are not only the direct costs of the repair works. Indirect costs, such as
road-user costs and safety implications, are considered as well. The indirect costs may be
regarded as more or less fictitious, but in fact they are just as real as the direct costs. They
are simply paid by someone else, not the owner. This means that it is necessary to include
the indirect costs when determining which strategy is optimum for society but they should
not be included when deciding whether the available funds are sufficient.
The direct costs are the actual costs of the rehabilitation works, including design,
construction, supervision, and administration, i.e.; the money that the bridge administrator
actually has to pay to get the job done.
The estimation of the direct costs is based on a preliminary project and a database with unit
prices from similar works, the Price Catalogue discussed above, in combination with the
experience of the rehabilitation engineer.
Among the most significant indirect costs are those inflicted on the road users, caused by
the damage to the bridge. The most common circumstances causing road-user costs are:
-
restrictions on the permitted load on the bridge, imposing detours on heavy trucks
closure of the bridge for safety reasons or during rehabilitation, imposing detours on all
traffic
reduced traffic capacity during rehabilitation, resulting in waiting time and reduced
speed.
The traffic aspect is particularly important for bridges whose closure would involve a long
detour on which carry heavy traffic.
Road user costs are calculated in computer programs developed for the purpose based on
traffic counts (including the distribution of vehicle types), length of detours, lower speed on
detours or through working area, waiting time, and unit costs per km and per hour for each
vehicle type.
41
To the extent that they can be quantified in terms of money, other inconveniences for
society related to one or more of the investigated strategies may be included in the
economic analysis. The individual administration decides which of these to include, and
how to quantify them. Examples could be excessive wear on vehicles as a result of potholes
in the wearing layer, and increased risk of accidents because of works on the highway, or
inadequate safety barriers.
5. ADDITIONAL COSTS OF CHANGING STRATEGY
The normal situation is that the funds for bridge administration are not sufficient for
carrying out the optimum strategy on all bridges.
Therefore, the goal is to find the best solution with limited funds. This is handled by the
optimization module described below, with input from the special inspection. On some
bridges it will be necessary to select another strategy, or to postpone the execution of the
optimum strategy. Either way, the total cost for these bridges will increase, because a
strategy is chosen that is not the optimum.
In order to identify the bridges suitable for a change of strategy, the additional cost of
postponing or changing strategy is evaluated as a part of the special inspection. For each
strategy, all costs, direct and indirect, are reevaluated under the assumption that all activities
are postponed, e.g.; by five years. This may lead to higher repair costs because of prolonged
deterioration, and to additional indirect costs because of longer repair periods, or because
restrictions to traffic may have to be imposed when the bridge is not repaired earlier.
Thus, the input for the optimization module from the special inspection on a bridge is a set
of normally three strategies, each with a corresponding postponement strategy. Each of
these six strategies is described by the distribution of costs over the next 25 years.
6. OPTIMIZATION OF REHABILITATION WORKS
When the funds allocated for bridge maintenance are not sufficient for all the proposed
works, some sort of priority ranking must be made. The economic analyses from the special
inspections are taken as the basis for an optimization. This means that it is not necessarily
the bridges that are most severely damaged that are given highest priority. The bridges
selected for repair are those for which the economic consequences to society of postponing
the works are the worst. These consequences additional costs are calculated as part of the
special inspection as described above.
The funds for the coming five years, the discount rate, and the data from the economic
evaluations of the special inspections are input into the Optimization Module. The
optimization process is divided into three steps: automatic pre-processing, automatic
optimization and manual post-processing.
42
Pre-processing
The optimization program calculates the net present value of the optimum solution for each
strategy and that of the postponed solution. By interpolation it also finds the net present
value of the solutions in-between, postponed by 1, 2, 3, and 4 years.
This leaves six solutions for each strategy, being the optimum postponed by 0, 1, 2, 3, and 5
years, each solution having its net present value. Consequently, for each bridge there are 18
solutions (presuming there are three strategies). These 18 solutions are sorted by the net
present value, the cheapest first. This sorting may result in solutions for the three strategies
being mixed.
Automatic Optimization
The purpose of the optimization is to find the set of solutions one for each bridge for which
the following two criteria are met:
-
The total cost estimate for direct costs lies within the budget each of the first five years.
The economic consequences (the extra costs direct and indirect) are the lowest possible.
The result of the optimization is the optimum solution to the theoretical problem (provided
that the theoretically correct algorithm has been used), and consequently the economic
optimum for society, given the limited financial resources (presuming that the models used
for pricing the indirect effects are valid).
Manual Post-processing
It is important that the result of the automatic optimization is not immediately accepted as
the final priority-ranking. The result of the optimization must be studied carefully, and the
choice of strategies may be altered, taking into account factors that are not included in the
automatic process, because they cannot be expressed in terms of money. Such factors
include aesthetics or environmental aspects, political considerations, or co-ordination with
other works on the same road.
43
2.14 DEVELOPMENT OF A BRIDGE MANAGEMENT SYSTEM IN

GERMANY
J. Naumann
Operation and Maintenance of Large Infrastructure Projects, Vincentsen&Jensen
(eds).1998 Balkema, Rotterdam.
Because of the totally different political and economic developments in West and East
Germany after the Second World War, the stock and condition of the road networks in the
two halves of Germany had also developed in totally different ways until reunification in
1990.
In the east the most urgent priority was, in the short term, to take immediate action to make
the existing network able to cope with the sharp increase in the volume of traffic, and in the
medium term to satisfy the enormous requirement for the construction of new roads and the
upgrading of existing ones after decades of neglect.
Given these challenges, and the current budgetary restrictions, a high degree of priority has
to be given to the use of intelligent management systems for the optimum distribution of
resources, especially to preserve the stock of existing roads. A major component of this is
the development of a bridge management system.
The future BMS will have to satisfy two major requirements:
-
It will have to enable the Federal Government to obtain not only an overview of the
current condition of the structures at network level but also information of funding
requirements, and to realice strategies, long-term objectives and general conditions in
management practice.
It will have to provide the federal states and authorities with recommendations for the
implementation of improvements at project level which are compatible with the
strategies, long-term objectives, general conditions and budgetary constraints.
The concept for a future bridge management system in Germany provides for a division into
a total of seven modules, each of which is subdivided into a number of different subject
groups. Development work is being carried out jointly by the Federal Government and the
federal states. The current development phase is focusing on enlarging the stock of basic
data and improving the recording and evaluation of conditions.
44
In a highly simplified form, the overall model can be subdivided into seven principal subject
groups:
-
Data collection and evaluation
Examination of various strategies
Forecasting of requirements
Creation of management programmes
Execution of work
Success evaluation, taking stock
Module I contains the necessary basic data for bridges, tunnels and other civil engineering
structures, which are organized as part of a comprehensive road information base.
Module II concerns the provision of condition data derived from the results of bridge
inspections and the evaluation of the damage identified. In accordance with Standard DIN
1076 (2), bridge inspections are to be carried out every 3 years as general inspections and
every 6 years as major inspections. Bridges are inspected and damage information collected
and initially evaluated by specially trained bridge inspectors of the federal states. The result
of Module II is to provide a network-related condition rating and an initial calculation of
funding requirements.
Module III related to a more detailed project-related damage analysis, with experts being
consulted if necessary. To this end, damage development models are to be integrated at a
later date.
Module IV provides a detailed catalogue of management measures, some of which are
described in more detail in regulations. Using the catalogue of measures, suitable variants
can be selected and considered in each individual case.
Module V constitutes a further essential component of a BMS, although it is one that is still
in its infancy. It relates to the economic evaluation of management measures within the
framework of economic efficiency considerations. To this end, service life cost models and
cost-benefit calculations are to be integrated at a later date.
Module V is also an interface between project-related and network-related approaches.
Here, the results of the previous modules are combined and the maintenance measures are
allocated to an urgency rating within the framework of an overall analysis. This in turn
45
forms the basis for calculating, short, medium and long-term funding requirements, taking
into account the marginal political, economic and technical conditions. Whereas the more
operational tasks of Modules I to IV are principally allocated to the lower administrative
level, Module V is of primarily strategic importance and is allocated to the intermediate and
higher administrative levels of the federal states.
2.15 BRIDGE MAINTENANCE TRAINING: EXPERIENCE AND FUTURE

TRENDS
A. Huvstig
Operation and Maintenance of Large Infrastructure Projects, Vincentsen&Jensen
(eds).1998 Balkema, Rotterdam.
There are a great need for cost-effective strategies for maintenance and repair of bridges.
One important part is to describe the actual condition, for example bearing capacity or
degree of corrosion. Furthermore there is a need for understanding of the ageing
mechanisms, which are active, and at which speed the process of ageing is going on in the
actual environment. After this, you must connect these calculations and knowledge with a
calculation of the whole life costing for reparation at different years. If you only have a limit
amount of money, you also have to make a judgement of the lowest agreeable functional
characteristics, which can be permitted, for the bridge. The mentioned processes can be a
good ground for choice of strategy for maintenance and repair of bridges.
The most important part in these processes and systems are the people, who are working
with maintenance and repair of bridges. These people must have very much of experience
and knowledge from maintenance and repair of bridges. They also must have an
understanding in the theories behind deterioration of different materials in a bridge, and
creativity in order to make the right judgement in different situations.
Strengthening of structures is first a question of whether or not strengthening is necessary,
and if so, to choose the most suitable method for the strengthening. Already in this stage,
you ought to engage an experienced design consultant.
In some occasions, the choice has to be made between repairing or strengthening a structure
and a new bridge. This choice is dictated by numerous factors such as investment costs for
the various alternatives, residual service life without action, interest rates, anticipated
service life for a new bridge, financial prerequisites, etc.
For these decisions you need drafts of different design solutions in order to make
calculations of the whole life costing for the alternatives.
46
2.16 THE PONTIS BRIDGE MANAGEMENT SYSTEM

P. D. Thompson, E. P. Small, M. Johnson, A. R. Marshall
Structural Engineering International. Volume 8, Number 4, November 1998.
The Pontis bridge management system is the predominant bridge management system
employed in the United States. The system employs a network optimization model for
preservation, formulated as a Markov decision process with a linear program solution
procedure. On each bridge, a set of level-of-service standards determines functional needs,
whose benefits are calculated according to a user cost model. A multi-year program
simulation generates project alternatives by combining preservation and improvement needs
on each bridge. The program is optimized within budget constraints by means of an
incremental benefit/cost algorithm.
The Pontis analysis includes a network-level preservation model for maintenance, repair
and rehabilitation, a functional improvement model, and a program optimization model.
Each of these components is described separately below.
2.16.1 Preservation Model
The Pontis preservation model is formulated using a top-down analytical framework. The
optimization develops pure network-level policies first, and then uses these results to guide
project-level decisions. Since the policy analysis is strictly network-level, it does not
consider data on individual bridges. Instead, it analyzes the generic action alternatives that
are available to the agency in response to every possible condition state that can be observed
on every element present in the inventory.
Inspection Data
Elements are characterized by discrete condition states, which describe the type and severity
of element deterioration in visual terms. Inspectors record the percentage or quantity of each
element found to occur in each condition state.
Deterioration, Action Effectiveness and Action Cost Models
Pontis uses a Markovian deterioration model to predict the probability of transitions among
condition states each year. Since the number of condition states is limited to five for each
element, the transition probability matrix is very small. Markov models assume that the
condition states themselves incorporate all the information necessary to predict future
deterioration. Thus, condition predictions for any future year can be made simply by matrix
multiplication.
47
For each condition state, available actions and associated costs are defined. Up to three
actions, inclusive of the do nothing action, may be defined for each condition state of
each element. The do-nothing action represents the deterioration model. If actions 1 or 2 are
taken, there is an improvement in condition state, also represented by transition
probabilities, and a cost is incurred. Failure costs are defined for each element and failure
probabilities from the last condition state are determined.
In order to model some of the variation of deterioration and costs within a state, Pontis
allows the elements on each bridge to be classified in up to four categories of environments.
Each environment for an element can have its own deterioration and cost models.
Preservation Optimization Network Level
Network optimization in Pontis is performed at the level of generic elements. The
fundamentals of discounted dynamic programming are used to find the optimal, long-term
policy that minimizes expected life-cycle costs while keeping the element out of risk of
failure.
The objective function seeks to determine the optimal policy, where a policy is defined as a
rule for the selection of the appropriate action. Assuming stationary policies, the condition
state distributions of the element at sequential periods of observation form a Markov chain.
The optimality equation is the expected life-cycle cost, given the assumptions specified and
bounded action costs, shown as follows:
V(i) = C(i, a ) + Pij (a )V( j) [2.2]

j
where
i
= condition state observed today
= condition state predicted to occur one year in the future
V(i)
= long-term cost expected as a result of being in state i today
C (i,a) = initial cost of action a taken in state i
= discount factor for a cost incurred one year in the future
Pij(a) = transition probability of state j conditional on state i and action a
48
V(j)
= long-term cost expected as of next year if state j occurs.
Thus, the long-term cost is simply the sum of the initial action cost and the discounted sum
of the future long-term costs of each possible condition state which may result in the
following year. The optimality equation thus takes a recursive form through which the
network-level model seeks the set of actions which minimize the long-term costs for each
condition state. The combination of the minimum cost actions forms the optimal policy for a
given element in a given environment.
2.16.2 Improvement Model
In Pontis, functional improvements are considered separately from preservation. The model
relies upon the definition of minimum level-of-service standards, used to assess needs; and
design standards, used the assess the amount of work required to correct existing
deficiencies. The user may vary the level of service criteria by traffic volume class,
functional classification, bridge ownership, and national highway system designation.
Additionally, the user may specify feasibility criteria for improvement actions.
Improvement needs are determined by comparison of the level-of-service standards with
the physical characteristics of each bridge. Based on the design standards and a unit cost
matrix, the functional improvement costs are calculated.
Benefits of functional improvements in Pontis are assessed in terms of user cost savings.
The total user benefit of a project is therefore:
B = W(Ba + Bv +Bt) [2.3]
where
W
= weight given to user cost, a matter of agency policy
Ba
= savings in accident costs
Bv
= savings in vehicle operating costs
Bt
= savings in travel time costs
The method for estimating accident user cost savings in Pontis is derived from the North
Carolina Bridge Management System, using the following formula:
Ba = 365 V(R R ' )Ca [2.4]
49
where
V
= forecast average daily traffic on the bridge roadway
= estimate of the current accident risk per vehicle
= estimate of the accident risk per vehicle after improvement
Ca
= average cost per accident
Based on North Carolina research, an approximate way to estimate R based on bridge

attributes is given as follows:
(9 A)
R = 200 0.3048 W 6.5 1 + 0.5

[2.5]
7
where
W
= roadway width curb-to-curb (m)
= approach alignment rating (typically 2-9)
R is estimated in the same way for the roadway characteristics after the improvement.
For bridges having impaired vertical clearance or load rating. Pontis calculates the vehicle
operating cost associated with traffic on a detour route, and assumes that this entire cost is
saved if a functional improvement is undertaken. Only trucks are assumed to be affected.
The project benefit in terms of vehicle operating cost benefits is then:
B v = 365 VD C v D [2.6]
where
VD
= number of trucks detoured each day at the bridge
Cv
= average vehicle operating cost per km of detour
= detour distance for the bridge roadway in km
50
The number of trucks detoured VD is used in both the vehicle operating cost model and the
travel time cost model. It is calculated based on the distribution of trucks of different
heights and weights in the traffic stream.
Pontis calculated the travel time cost associated with traffic on a detour route, in a manner
similar to vehicle operating cost. Again, only trucks are affected. The project benefit in
terms of travel cost saving is:
Bv = 365 VD C t
D
[2.7]
S
where
VD
= number of trucks detoured each day at the bridge
Ct
= average travel time cost per hour of detour
= detour distance for the bridge road-way in km
= speed on the detour route (km/h)
2.16.3 Project-Level Programming

When each bridge in the inventory is inspected, the actual condition states of each element
on the bridge are observed. Pontis consults the network optimization results to find the
recommended maintenance, repair or rehabilitation (MR&R) action with its unit cost and
benefit for each observed condition state. The difference in long-term costs between the
optimal action and the do-nothing action is considered the net benefit of that action: thus a
large net benefit implies a strong recommendation. Summation of costs and benefits over all
elements for each bridge provides an estimate of the total MR&R needs, and associated
benefits of taking action, for each structure in the inventory.
The preservation needs may then be combined with the improvement needs. In cases where
a bridge has both MR&R needs and functional deficiencies, it may be considered for
replacement. Replacement may be triggered if the appropriate functional improvement
action is not feasible, if the combined MR&R and improvement cost exceeds a certain
fraction of the replacement cost, or if replacement provides a more attractive benefit/cost
ration.
The final decision as to whether to proceed with each project depends on funding
constraints and the relation-ships among competing project alternatives. Since there is
almost never enough funding to address all needs, an objective means is required to find the
51
program of projects which maximize the benefit achievable within a constrained budget.
Pontis uses an incremental benefit/cost method to do this.
Incremental benefit/cost analysis may be best understood within the context of zero-based
budgeting. The procedure starts with a zero budget, thus requiring every bridge to choose its
do-nothing alternative, with zero cost and benefit. As a small increment of funding is added
to the budget, it is allocated to the bridge that is able to maximize the marginal return, the
increase in benefit achievable for a given increase in cost.
2.17 WHOLE-LIFE COSTING (WLC)

A. E. K. Jones, A. R. Cussens
Structural Engineering, Volume 75, Number 7, April 1997.
2.17.1 Introduction
Whole-Life Costing (WLC) provides a method by which alternative solutions to a project
can be compared, in financial terms, over the total life of a structure. Whilst the basis of
WLC are simple , the assignment of values to the variables involved is more difficult, and
this is hindering wider application.
2.17.2 The Basis of WLC
The basis of WLC is that all costs associated with the solution of a project, over its total life,
can be added together to represent a total or 'whole-life' cost for that solution. Future cost
are normalized to a present value using :
Present value =
C
[2.8]
(1 + r ) t
where C is the cost at current prices, r is the discount rate and t is the time in years to when
the cost is incurred.
Once the whole-life cost has been calculated, it can then be used to compare different
solutions, involving different materials, or different levels of product quality. In addition,
straight comparisons do not allow for variations in the benefits to be derived from different
solutions. If these benefits can be costed, the whole-life cost can be deducted from the
benefit to give a net present value for the project.
To carry out WLC, the following information is required: first cost, test discount rate,
frequency and cost of maintenace, including disruption costs, and the proposed life of the
structure.
52
2.17.3 First Cost

In considering alternatives, much of each solution is identical and it will be appropiate to
evaluate only the WLC of the differences and the implications of these.
2.17.4 Test Discount rate
The test discount rate represents the fact that money not spent now can be invested (or at
least not borrowed), and would, therefore, be worth more at some time in the future. If it is
assumed that any costs will increase in line with inflation, the discount rate used should
include an allowance for inflation. Thus a discount rate net of inflation can be aproximated
to the expected growth (or cost) of money less inflation.
In the UK, the net discount rate used by the Department of Transport and recommended by
the Treasury is 8%. The use of such higt values for the discount rate is in contradiction with
engineering judgment and the design life required. For example, a bridge is designed with a
120 year design life, yet, with a discount rate of 8%, the effect of the cost of replacement at
60 years is approximately 1% of the first cost. There is an apparent discrepancy between the
design life and the economic life.
In addition, basing the discount rate for public structures on the growth of money in
industry may not be appropiate where the controlling authority is working to an annual
budget. For example, the use of 8% discount rate on a bridge owned by a local authority
implies that the money available for investment by that local authority increases by 8%
above inflation each year. Clearly, this is not the current situation.
In addition, the sensitivity of the solution to variations in the discount rate should be
investigated. In reality, because of the long-term nature of WLC for structures and the
variations in possible discount rates, the whole-life cost produced cannot be considered as
an absolute value and should be used for comparative purposes only.
2.17.5 Frequency and Cost of Maintenance
There are two methods which can potentially be used to estimate the frequency and cost of
maintenance. The first method is through deterioration models and the second is through
historical databases.
Deterioration models
The basic transport mechanism and deterioration process for concrete material is well
understood but predicting the effects of real environments, design detailing and execution
quality are not.
53
The advantage is that the method is even possible for new and innovative materials.
Historical databases
Performance prediction from historical databases incorporates allowances for the various
factors that affect performance in the field. However, any change in materials or
construction practices requires modification of the data. In addition, there is little
information available on historical performance.
The Department of Transport has set out basic lifetime maintenance costs for various
bridge types, and a supporting document specifies that the probability of major maintenance
being carried out on a concrete bridge is 0.02. Other authors argue that this value is too low.
Once a maintenance strategy is decided, the client needs to inform the engineer of the likely
disruption costs to his core business of such maintenance. However, when we consider the
public sector, the incorporation of these costs becomes more difficult. The Department of
Transport has developed a computer model, QUADRO, to calculate the delay costs incurred
when disrupting traffic.
Prediction of the cost and frequency of maintenance requires engineering judgment. Again,
the sensitivity in the selection of the most economic choice can be tested by varying the
various factors related to maintenance.
2.17.6 Life of Structure
The end of the life of the structure must be defined. There are several ways to do that: The
Code of Practice used (usually 50 or 120 years), the time to repay the investment, the time
of obsolescence or an estimation of the technical life.
The life of a bridge is typically assumed as 120 years but many reach functional
obsolescence owing to increased loads or widening long before then. It is really appropiate
to consider costs over the full 120 years?. Once the required life has been agreed (client and
engineer), the designer can then optimise the solution to meet this parameter.
Currently, assessment of the standard service life provides the main difficulty, and this
riterates the need for better predective models and historical data.
2.17.8 Conclusions
1.- WLC represents a rational way of comparing design and construction solutions over the
predicted life of a structure. Whilst there are some difficulties in the details, this should not
stop engineers using the process as a qualitative tool.
54
2.- Client and owner must take an active role in defining their requirements in terms of
return of investment, disruption costs and performance with time.
3.- The accuracy of the results from WLC can be improved by better quality and therefore
lower variability within the construction process.
2.18 CONCRETE BRIDGE ASSESSMENT: AN ALTERNATIVE

APPROACH
C. R. Middleton
Structural Engineering, Volume 75, Nos 23 & 24 December 1997.
In 1987 the Department of Transport launched a 15-year bridge rehabilitation programme
whose aim is to strengthen and safeguard the integrety of all motorway and trunk road
bridges in the UK ready for the introduction of a 40t load limit in 1999. In 1989 the local
authorities in the UK started their own complementary bridge assessment and strengthening
programme.
2.18.1 Progress with the Assessment Programme
Motorway and trunk road bridges (England)
The HA determined that 69% of the 10.000 bridges would require assessment. In March
1997, 94% of these had been assessed with 19% failed assessment. The 43% of these heve
been strengthened.
Local Authority bridges
With so many different local authorities it is difficult to obtain complete and up-to-date
statistics on progress with the assessment programme. In particular, with no single national
database of bridges, it has not been possible to quantify the primary causes of failure in the
bridge programme. It was stimated that 66.000 local authority bridges needed assessment
and only 19% of these had been assessed.
The causes of failure are varied and depend on the type of structure and also, to an extent,
on age and location. Most of the motorway and trunk road bridges most of the structures
were built post 1960 and ere predominantly concrete and the deficiencies are in shear and
flexural capacity, inadequate anchorage details, presstress corrosion and deterioration of
joints, piers ans cross heads. Local authorities have a large numbers of masonry arcg
bridges and also many older concrete bridges which have often been subject to significant
deterioration or were designed with inadecuate detailing, little or no top steel, and low
percentages of transverse steel.
55
2.18.2 The Analysis Credibility Gap

There is an underlying realisation that the analytical techniques developed for design are in
many cases unable to model the structural behaviour of existing bridges accurately. As a
result, assessment often significantly understimate the actual load capacity of bridges. This
brings into question the appropriateness of using elastic analysis for the determination of
ultimate strength for many types of bridge, in particular for short-span concrete slab bridges.
Clearly, there is a need to review and refine our existing methods and to develop improved
tecniques which can more realistically model the ultimate load capacity of bridges. These
must also be practical to apply and relatively quick to implement because of the large
numbers of bridges involved.
2.18.3 The scope for improved methods of assessment
Codes of Practice
The Highways Agency has lead the world in producing separate bridge standards for the
assessment of existing bridges, as distinct from design standarts for new bridges. By
considering the particular problems associated with assessing existing structures a number
of design-based rules heve been relaxed. In the case of concrete bridges these changes
include the use of worst-credible, rather than characteristic strength, parameters, modified
shear strengths and also provision for reinforcement details that are considered deficient by
current design standars but are often found in older concrete bridges.
In addition, the HA is currently funding research looking at ways of incorporating wholelife costing, deterioration profiles and reliability methods into assessment of bridges and for
determining optimal strategies for repair and maintenance.
Load testing
Load testing is currently not favoured as a primary method of assessment by the UK
Department of Transport because of the high cost involved, the possibility of causing
damage to the structure and the difficulty in interpreting any test results.
2.18.4 The assessment of short-span concrete bridge decks
The short-span bridge has received relatively little attention in the recent literature and
research programs. Although these may seem somewhat unexciting structures, they have
created significant problems in the current bridge assessment programme.
56
The fundamental philosophy adopted for assessment, as distinct from design, has been to
evaluate only the ultimate strength as the fundamental criterion for passing or failing a
structure. Serviceability criteria are not usually considered.
There are several methods of analysis:
a) Elastic analysis that is the conventional approach to assessing the load capacity
but should be conservative and hence safe.
A linear elastic analysis will not model accurately the distribution of streses in
the post-elastic range where non-linear effects dominate.
b) Non-linear finite element analysis (NLFE). Although NLFE methods have
developed to a sophisticated state, their pplicability is severely limited by their
high cost in computing time and the advanced level of expertise required to use
them. In addition, the technique is load-history dependent and very sensitive to
the choice of material parameters. Another disadvantage often cited is that finite
element programs usually generate a large amounth of output data and it is often
difficult to verify the rsults using some form of simple hand-calculation.
c) Plastic collapse analysis: the yieldline method. Yieldline analysis considers the
global collapse of a concrete slab rather than the failure of a single element, thus
utilising the full, distributed strength capacity of a structure. As a result, it is
usually less conservative then elastic methods.
Yieldline methods have been used extensively in research but have not been
widely adopted in engineering practice.
The author describes a new technique for performing yieldline analysis of concrete slabs
developed recently in Cambridge University implemented in a computer program COBRA.
The development of the COBRAS yieldline program provides a very powerful alternative
tool with which plastic collapse analyses of these bridges can be undertaken for a wide
selection of possible failure modes and assessment load cases. As it is an upper-bound
approach, care must be used in applying this technique.
It is likely that many of the concrete slab bridges which have failed their initial assessment
owing to inadecuate flexural capacity would in fact be found to pass the 40t assessment
requirements if yieldline analysis was used to reassess their ultimate capacity.
57
2.19 THE BRIDGIT MANAGEMENT SYSTEM

H. Hawk, E. P. Small
Structural Engineering International, Volume 8, Number 4, November 1998.
2.19.1 Introduction
Over 42% of the 600.000 highway bridges and culverts in the United States have been
considered structurally deficient or functionally obsolete. Effective bridge management
tools are required to ensure careful allocation of limited resources to bridges to provide the
most cost-effective treatment for preservation of the bridge infrastructure. The National
Cooperative Highway Research Project, sponsored by AASHTO and FHWA developed a
bridge management software tools for transportation agencies called BRIDGIT.
BRIDGIT is ideal for smaller departments of transportation and it can be a complement to
Pontis BMS, offering an independent analysis. BRIDGIT aids in the development of bridge
maintenance, rehabilitation and replacement programs based on life-cycle costing and
incremental benefit-cost analysis. Preservation considerations capitalize on Markovian
deterioration predictors. A level of service approach is employed for improvements, which
includes the consideration of user costs associated with traffic accidents and detouring.
2.19.2 System Structure and Components
The system is developed in a modular fashion structured around a FoxPro database and
compiled for use on a PC platform. The goal is to develop optimized analysis work-plans
for preservation and improvement over a defined analysis horizon. The method of
optimization is unique in that multiperiod analysis is performed as opposed to sequential
period analysis. Thus, the optimal action is considered in conjunction with the optimal
timing. actions may be delayed to later years where they would be more cost-effective,
which is particularly useful in unconstrained budget scenarios.
This multiperiod optimization is performed in two steps. First, life-cycle activity profiles
(LCAPs) are generated for each bridge in the inventory. The LCAPs estimate the present
and future costs of various MR&R (maintenance, repair and rehabilitation) measures, based
on element models, improvement scenarios and level-of-service standards. Second,
optimization is performed to prioritize needs and select the most cost-effective options for
given budgets over the planning horizon. Options may be considered simultaneously for
constrained and unconstrained scenarios.
58
2.19.3 Inspection Data Model

The MR&R data model is based on element-level definitions. Condition states and feasible
actions are defined. The user assigns A/M/U factors indicating whether the condition states
are acceptable, marginally acceptable or unacceptable.
Environment states are used in the system to account for differing deterioration rates
resulting from environmental conditions, element interactions, and loading conditions. Four
environments are defined: benign, low, moderate and severe. Each element and protective
system is assigned a corresponding environment states.
Repairs are triggered using the A/M/U designations. Actions and unit cost are defined for
each condition state of each element and protective system. Actions are assumed to restore
the element to condition state 1. While many different MR&R activities and unit costs may
be defined for each condition state, only one feasible action may be defined as the suggested
MR&R activity for each element in each condition state. This is considered along with the
do nothing and replacement alternatives in the development of LCAPs.
2.19.4 Deterioration Prediction
BRIDGIT utilizes Markovian deterioration models to predict the future condition states.
The model is applied at the element level with transition probability matrices pre-defined on
the basis of the condition states. Future condition state predictions are independent of the
element history. The current or initial condition state is assumed as the starting point for the
analysis, with a future condition state determined using the transition probability matrix.
Historical information can be used to update the deterioration transition probabilities over
time.
2.19.5 Effect of previous repairs and traffic
When repairs are undertaken, the deterioration characteristics are changed. BRIDGIT
accounts for the effects of previous repairs by artificially accelerating the deterioration
process. This artificial acceleration is performed by moving a percentage of the elements
from a higher condition state to the next lower condition state.
The effects of traffic, particularly truck traffic, affects the deterioration rate of deck
elements, joints, etc, and are accounted for through the use of an ADT (average daily traffic)
modifier.
Load carrying capacity LCC is used to determine user costs associated with detouring. The
required LCC of a structure is defined as that load capacity at which no trucks are detoured.
Prediction of future LCC depends only on the condition of the superstructure elements and
is determined as the lowest capacity of any singular primary superstructure element.
59
2.19.6 Agency Costs considerations

BRIDGIT employs a bottom-up optimization procedure based on various LCAPs. Agency
costs and user costs are considered in the development of these life-cycle activity profiles.
Agency costs include initial first costs of actions and future MR&R or improvement actions
over the life of the structure. The user costs are incurred due to functional deficiencies.
The LCAPs are determined for scheduled actions and repairs, replacement, improvements
and bridge replacement. If the cost of repairs exceeds 70% of the cost of replacement , then
replacement is chosen.
2.19.7 User Cost
Accident and detour costs are considered which may result from bridge load capacity
restrictions, inadequate clear deck width, number of lane deficiencies and/or vertical
clearance restrictions.
2.19.8 Life-Cycle activity profile
Expected agency and user costs over time are used to develop LCAPs. The costeffectiveness of different MR&R and improvement actions are considered by assessing the
present value of the life-cycle costs and benefits. Feasible activities are determined using
knowledge-based decision rules. The cost and benefits resulting from immediate and future
actions are considered.
Every bridge is assumed to require maintenance, repair, rehabilitation and, eventually,
replacement over its life span. With replacement, it is assumed that all functional
deficiencies are corrected. After replacement, load capacity deficiencies are corrected and it
is assumed that, over the life span of the structure, maintenance actions and repairs will
prevent load posting of the structure. Therefore, user costs associated with load capacity
deficiencies are not considered.
Routine maintenance costs are assumed to be incurred when bridge elements deteriorate to
condition state 2 or 3. Afterwards the maintenance costs increase linearly over time until
repair or rehabilitation corrects the deficiencies.
The BRIDGIT optimization model uses incremental benefit cost analysis to select optimal
MR&R and functional performance improvement actions for each bridge in the network
over a multi-year analysis. Unconstrained and constrained budget scenarios are considered.
60
2.20 THE FINNISH BRIDGE MANAGEMENT SYSTEM

Marja-Kaarina Sderqvist, Magnus Veijola
Structural Engineering International, Volume 8, Number 4, November 1998.
The Finnish National Road Administration started development of a BMS in 1986 for its
13.354 bridges and 2757 culverts (span length >2m). The network level system has been
running as a prototype since 1996, while the project level system was introduced at the end
of April 1998.
The goal is to provide a reliable support tool in decision related to fund allocation for
maintenance, rehabilitation and replacement minimizing costs and keeping bridges safe and
serviceable.
Bridge Management System
Damages and deterioration detected during inspections (and their location, extent, effect on
bearing capacity, repair urgency class, proposals for repair and costs) are registered in the
database together with structural, administrative and traffic data.
An improved multi-user client-server version was introduced in 1994.
Network level
a) Long-term Analysis
Is based on the idea that the bridge stock has an optimal condition distribution. There is an
optimum where the bridge stock can be kept at the same condition level year by year with
the smallest amount of funding, yet adhering to safety level and service requirements.
b) Sort-term analysis
Provides an economically optimal way to reach the long-term optimum condition during the
next few years. Short-term solutions are evaluated every year.
c) Network level
61
Offers the possibility of "What if?" experiments with respect to the safety and minimum
service level policy, repair measure cost, budget restrictions and other variables.
Project level System
Uses the results from the network level system to decide on the repair measures in
individual repair projects. It is an interactive computer program and includes a life cycle
cost analysis that gives the engineer the possibility to opt for repair measures that result in
the minimum optimal total cost over the entire life span of the bridge.
2.20.1 Mathematical Approach Network Level
The purpose of network level is to minimize the total yearly repair costs under given
restrictions by carrying out the right repairs at the right moment in the life span of specific
bridge components. The mathematical solution uses a set of probabilistic Markov chains
models to predict deterioration of various structural elements in the bridge stock. Together
with data on possible repair measures and their costs for every damage, condition
requirements and budgetary limitations, linear programming (LP) models can be formulated
and solved by computer giving a recommended long-term optimal solution for the condition
state distribution of the bridge stock. The LP models also give the distribution of repair
measures required each year to maintain the optimal cost-effective state.
2.20.2 Age Behaviour Modelling
Deterministic Deterioration Approach
The modelling of bridge deterioration acceleration and age behaviour is based on the
information of damage gathered during inspections. The present condition of the bridge
stock is calculated from the stored data of actual observed damage to the bridge structures.
Each damage is related to a specific part of the structure and each part can have several
damages. The severity is classified from 0 to 4.
Probabilistic Deterioration Approach
The optimisation algorithm works with populations of bridge structural parts that are similar
in construction and use: super/sub-structure, surfacing, furnishing as well as materials,
bridge type and construction techniques. Two environmental categories (salted / not salted
bridges) are considered. The system has a total of 50 categories.
The condition of a bridge structural part is evaluated with respect to one, two or three
damage groups: surface damage, structural damage, water leakage...
62
The deterioration model, repair measure models, repair measure cost and the
budget/condition constraints may be formulated as a linear programming model. Each
category of bridge items corresponds to one LP model. The objective is to minimize repair
costs due to given condition constraints or to minimize social costs.
Repair models
The Finnish bridge inspection manual recommends repair measures for each damage class
and type of structure. On the project level the inspector gives his judgement to a repair
measure recommendation regarding the observed damage on an individual bridge. Every
repair measure recommendation is saved in the database with its expected cost.
Bridge condition description at project level
-
Repair index: is used to arrange the bridges in an urgency order in the work programme,
an index describing the repair needs of individual bridges was taken in use on the
project level. The repair index is a function of bridge structural parts estimated
condition, damage class and repair urgency class.
Reconstruction index: is used to identify bridges with functional deficiencies:

narrowness, inadequate load carrying capacity, bridges at the end of their functional
life...
2.20.3 Reference Bridges

About 120 bridges have been selected for regular observation to improve the knowledge of
bridge ageing behaviour and durability. These reference group consists of bridges
throughout the country of different materials and types, ages and conditions.
2.20.4 Conclusions
The present condition of the bridge stock is well known. Condition prediction and scenarios
for different road maintenance policies can be calculated. Long-term condition targets and
the corresponding repair measures will be given for yearly bridge maintenance.
The detailed repair budgets for coming years are available. The annual repair programmes
are co-ordinated with the long-term condition target.
The benefits of the BMS include minimising the MR&R costs. Additionally, traffic safety
can be ensured through the choice of condition constraints. In time, increased knowledge
about the age of behaviour of materials and bridge elements will be obtained from the
detailed damage record of thousands of bridges.
63
3.
REVIEW
OF
EXISTING
REPAIR/REPLACEMENT
DECISION
SYSTEM
FOR
BRIDGE
3.1 DECISION ON MAINTENANCE AND REPAIR IN VARIOUS

COUNTRIES
In order to produce an outline framework for management of bridges on the road networks
across Europe, the first step consisted of a review of existing Bridge Management Systems
(BMS) used in Europe and abroad. This review was based on a questionnaire that was sent
to the partners of the BRIME project, but also to other European countries, as well as
countries outside Europe known to be well advanced in terms of Management of Bridges
(Canada, Japan, USA).
The responses were obtained from:
Germany - Bundesanstalt fr Strassenwesen.
Spain - CEDEX.
France - Road Directorate.
Great Britain - Highways Agency.
Norway- Directorate of Public Roads.
Slovenia - Road Directorate of the Republic of Slovenia.
Denmark - Danish Road Directorate.
Finland - Finnish National Road Administration.
USA - Federal Highway Administration.
New York City DOT.
California DOT.
This review shows a large variety of approaches among European countries. Most of the
countries use a computerised BMS (eight countries); two countries do not operate a BMS
(Germany and Spain), and one country declares operate a partial one (France).
64
The condition of the bridge (through bridge inspection results), information on maintenance
work and costs are recorded. Prediction models and taking account of indirect costs are not
very developed. Decision on maintenance and repair, as well as prioritization, is essentially
a matter of engineering judgement.
3.1.1 Decision on maintenance and repair
A series of questions on how each country decides the best option for the maintenance and
repair of a bridge was included in the questionnaire. Some of those questions are listed
below:
1. Does the BMS use a specific tool or criterion to decide whether to: repair, strengthen or
replace elements of a damaged bridge?
-
If YES, which kind of specific tool (program, decision chart,...) is your Organisation
using ?
If NO, what is the methodology used by your Organisation for this task ?
2. How do you decide when maintenance work is needed ?

3. How do you decide which is the best maintenance option to use ?
To know decision-making of different countries on the matter, a brief summary was made
by analyzing the answers to the questionnaire.
Most of the countries do not use a BMS tool for decision maintenance and repair. Except
Denmark, Finland and USA. Denmark uses a priorization program and Finland a repair
index.
In USA there are a variety of system used:
1. Pontis.
The optimal policies are developed on a network level based on the minimum expected lifecycle cots over an infinite planning horizon. Benefits are determined as agency cost saving
of performing an action versus postponement of the action by one year. The optimal policy
is then applied to a bridge and aggregated to determine the optimal preservation strategy for
each bridge.
65
2. Bridgit.
The bridge level actions are developed through minimization of expected life-cycle costs
over a 20 year planning horizon. The optimal sequence of actions and the optimal time to
take the action are considered. Actions may be triggered for improvements, to remove the
unacceptable condition states, for replacement, etc. Benefits are determined as user cost
savings.
3. State Specific Systems.
Five states have developed their own BMS: Alabama, Indiana, New York, North Carolina
and Pennsylvania. New York BMS has a specific tool but is not functioning and all
recommendations are subject to engineering approval.
For France, decision is only based on human shill with several level of control (technical
and financial). Likewise, engineering judgement are used in Germany and Spain. Great
Britain uses cost benefit analysis, future needs and engineering judgement.
In Norway, based on the damage description and the condition assessment, proposals for
repairing damage, shall be prepared from work descriptions/process codes. Cost estimates
are to be prepared for the proposed action, and an indication should be given which year
these activities shall/should be performed in so as to ensure that the specified standard is
maintained.
When the cost of necessary repair activities following from a major inspection or special
inspection exceeds 20% of a bridges replacement value, alternative strategies should be
investigated.
At least two different strategies will be investigated depending on what is relevant. In
addition to maintenance costs they should also include road user costs and any costs to
society if affected by the various strategies.
The following strategies may be considered:
1. Temporary action : Minor repair activity carried out during one period to postpone
major work or replacement of a bridge.
2. Major action : Extensive repair work during a brief period to significantly extend the
remaining service life of the bridge.
3. New element/bridge: No repair work undertaken; however, the existing element/bridge
is replaced at the end of its service life.
66
For each strategy different technical solutions may be considered.

When maintenance costs exceed 50% of replacement value, strategy 3 must be considered.
The present worth of the selected strategies shall be estimated and these will form the basis
for selecting an optimal strategy. Factors that normally do not enter into cost estimates shall
also be included before the final selection of a strategy.
-
Such factors may include :
Bridge age, remaining service life
Carrying capability
Bridge width/road curvature
Vertical clearance
Traffic safety
Mobility
Future usage
Aesthetics
Historic value
At present, in Germany, the Federal Ministry of Transport is aiming for a comprehensive

management system for structural maintenance. In this system, as the outcome of the overall
evaluation, and taking into consideration the intervention time, urgencies are assigned in
order to ascertain the sequence in which, and the structures on which maintenance measures
are required. However, the appropriate measures are as yet undetermined. Although the
overall evaluation provides an indication of the measures to be developed, the final
selection of the measures to be implemented is subject to further boundary conditions.
Taking into consideration the changes in condition to be expected in future, the overall
evaluation supplies important precepts for the decisions to be made as part of maintenance
management. In this case, decisions are to be understood as selections between several
alternatives with a view toward attaining one or more objectives. Goal-oriented selection
requires alternatives to be ordered by degree of preference. In this process, an investigation
67
is required as to the extent to which specified objectives can be attained by the available
alternatives.
Potential, mutually exclusive measures are determined during the development of
alternatives, (maintenance of variants, replacement, no measures). Furthermore, the effects
resulting from these measures are to be ascertained. This primarily includes the expenses to
be incurred by the owner and by users. A time-based evaluation of the effects of investment
must also be taken into account here. The complex configuration comprising
damages/measures/costs and the involved links are to be defined.
During the evaluation phase, the measures are rendered comparable with the help of
evaluation functions. The employed techniques of monetary evaluation include benefitscosts analyses (maximum-cost/benefit ratio) or a pure cost analysis (minimum overall cost)
if the zero-planning scenario i.e. non-implementation of measures cannot be defined.
Techniques without monetary evaluation include utility-value analysis and costeffectiveness analysis. A goal-oriented technique is to be developed for the evaluation step.
Most of the countries decide when maintenance work is needed on basis of inspections. In
France is a consequence of the prioritisation made by each Public Works District Directions
and human shill and experience. In Slovenian they decide on the basis of condition
assessment during the inspection and the rating of the bridge; the load restrictions; the
increased traffic volume and the importance of the bridge for the region. USA decide on
basis of safety needs and economic benefit. Denmark does it during routine inspection, and
Finland from reporting damages from the database.
Most of the countries decide which is the best maintenance option to use on basis of
engineering judgement. Great Britain precise that it depends on the solution available and
their cost, traffic management and traffic disruption; whole life costing is also used for
major schemes and will be used for all schemes in the future. Also, It has a computer
programme called QUADRO (QUeues And Delays at ROadworks) that provides a method
for assessing the total cost of major road maintenance works.
The total cost of maintenance works is made up of two elements:
The direct works cost of maintenance operations, for example, resurfacing or

reconstruction.
The cost imposed on road users while de the road works are being carried out, made up
of road user delays (value of time), vehicle operating costs and accident costs.
The QUADRO program requires a network, comprising the main route on which the works
are to take place and a representative diversion route, together with the traffic flows for both
routes in order to operate. It also requires details regarding the length of works site, its
location on the main route in relation to the diversion route and the duration, timing and
68
traffic management arrangements of each job. Works can be specified either individually (a
job) or a series of jobs (a profile).
For the main route without road works, the type of carriageway, length and geometry must
be known. QUADRO uses this information to determine the roads speed/flow curve. The
width of the road affects the length of any queues. Accident information must also be
chosen by the user. QUADRO also needs to know the overall proportions of each vehicle
category.
Time Costs
QUADRO calculates vehicle speeds and journey times on the without and with work
networks using the traffic flows and speed/flow curves. A monetary value is attached to
these times using different values for vehicles in work time and non-work time. Vehicle
speeds and journey times differ between the without and with works networks due to
queuing delays on the main route and the reduction in speed of all traffic on the diversion
route, cause by any diverting traffic. The effects of incidents in the with works case are
separately measured in terms of extra journey time caused by additional queuing.
Vehicle Operating Costs
Vehicle operating costs are calculated using the same relationships, based on average
vehicle sped. Theses costs tend to be lower in the with works case as sppeds through the site
will be slower. Also epeeds on the diversion route may be slower due to diverting traffic.
For diverting traffic this may be offset if the diversion route is longer. Fuel and non-fuel
costs are calculated by vehicle type using different parameters for each vehicle category.
Accident Costs
Accidents, as well as causing traffic delays, have their own associated costs. In QUADRO,
accidents are related to vehicle kilometres by an accident rate. Accidents rates and costs
vary with location and type of road. The costs associated with each accident assume that
there are particular number of fatal. serious and slight casualties per accident. An allowance
is made in the costs for damage only accidents.
Slovenia says that it depends on the importance of the bridge, load restrictions and
importance of the road and Finland uses repair guidelines. In the State of California uses
long term optimal strategies.
69
3.2 OTHER MANAGEMENT SYSTEMS USED ABROAD

Some countries, that have not taken part in the questionnaire, have developed or are
developing their own systems.
Japan
The first version of the Japanese management system was completed in 1995. This system is
intended to determine the most effective maintenance plan under a given set of basic
financial conditions.
The management system consists of two main modules, the condition module and the
planning module. The condition module is used to determine the condition of the entire
bridge based on the condition of its individual, main components. The planning module is
used to optimize planning. Specifications are made as to which bridges are to be repaired
and how urgent these repairs are under the given basic financial conditions. The system
contains cost tables which allow the comparative evaluation of various repair alternatives.
So far, it has been oriented towards planning on an annual basis.
Poland
Since 1989, Poland has developed a management system for maintenance planning, which
includes decision-making procedures for various organizational levels. The basic function
consists of the planning process with a 1-year horizon, taking into account data related to
inventory, construction and condition.
To optimize the allocation of resources, linear programming is used, taking into
consideration the replacement value of all the bridges in a region, the condition of the
bridges and additional statistic data 8number of bridges, area, etc.). These results are used to
determine the annual maintenance costs on the basis of cost tables.
As part of the optimization process, these resources are distributed among the individual
bridges. A number of parameters are taken into account here:
-
Cost carried by the operating company and users.
Comparison of service-life costs with the cost of a possible new construction.
Technical criteria.
Durability.
70
Influence of traffic.
Urgency, restrictions, etc.
Sweden
Essential constituents of the Swedish management system are inter-disciplinary strategies
for planning and control measures, as well as and operative planning and implementation of
measures.
Two models are available for operative planning and procurement. The first model is used
for routine maintenance, i.e. preventive maintenance and minor measures. The second
model, called SAFEBRO, is used for major maintenance measures. It consists of the
following components:
-
Inter-disciplinary management.
Bridge management activities.
Bridge databases.
Knowledge databases.
Processing modules.
Knowledge databases are used to ensure the quality of the information stored in the
databases. The database for technical solutions supports the planning procedure by
supplying suitable, technical alternatives. The database of prices per unit constitutes a price
list for these technical solutions.
Switzerland
Switzerland is developing the KUBA-MS system as a prototype of a BMS. The system is
intended to fulfils the following objectives:
-
Determination of economically ideal maintenance policies with and without budgetary

restrictions.
Determination of the effects resulting from deviations from this strategy.
Consideration of expenses incurred by operating companies and users.
71
Indication of optimal measures for any required planning horizon.
Determination of short-term and medium-term financial requirements.
Indication of the development of average structural conditions for various budgetary

restrictions.
3.3 PRESENT STATE OF BRIDGE MANAGEMENT IN SPAIN

Nowadays, an Integrated Bridge Management System applied to the main network does not
exist in Spain. Nevertheless, studies and works have been made or are under development
that could be considered as parts or phases of a future Management System.
Within the actions taken, there is the inventory of the structures of the Principal Network.
Some 15000 structures have been inventoried until now. Some activities related with
routine maintenance, main inspections and crossing viability of special transport have been
also carried out.
The fundamental task of the Bridge Management System under development in Spain is to
establish the rehabilitation strategy by ordering the repair priorities, direct function of
technical and economic criteria. The programs and plans of action are elaborated from this
classified information.
The future bridge management system is developed from:
1. The inventory of the bridges that conform the network, with a detailed definition of the
elementary geometric characteristics, typologies and materials of the different
constitutive elements and maintenance state.
2. The systematic execution of Principal Inspections.
3. The evaluation of the damages by assigning an state index or condition mark.
4. The establishment of repair priorities.
5. The execution of detail studies. Special Inspections.
6. The alternatives of maintenance and/or repair, with their cost.
7. The execution of the plans of action.
72
The study of the repair alternatives must be based on the setting out of, at least, two clearly
different solutions and in their later comparison. The costs entailed in the delay of the action
proposed up to five years, more or less, will be also examined; this will make possible to
define the opportunity of the action to be carried out.
The following items will be taken into account to estimate the costs of the repair works:
-
Cost of the repair to be made.
Cost of the users, both during the execution of the repair or its delay.
Cost of the routine maintenance. There are some repairs that, when they are not made at
once and are postponed in time, generate during that delay the need of a superficial and
preventive maintenance of a higher cost than the usual one.
4. REVIEW OF COMMERCIAL BRIDGE MANAGEMENT SYSTEMS

A detailed study about two commercial bridge management systems has been developed:
PONTIS system and y DANBRO system. Both of them have a modular structure. Using the
technical manuals of these programs we have analysed each module and each component of
the programs, emphasizing the decision system for bridge repair/replacement and the
methodology to evaluate user costs.
4.1 PONTIS
4.1.1 Introduction
Bridge management consist of a series of activities involving information gathering,
interpretation, prediction, cost accounting, decision making, budgeting and planning.
The Pontis, consists of a set of interconnected models that address these functions. It is a
flexible and interactive tool which allows user input in every stage of the process and uses
mathematical models to help in generating and evaluating alternatives.
Budgeting for expected expenditures to address current and future rehabilitation,
maintenance and corrective measures, as well as improvements and possible replacements
and additions, are intimately related to the desire of decision makers to meet several
objectives. In addition to cost and budgetary issues, the main objectives that planners and
administrators are concerned with are as follows:
1. Meeting and maintaining the highest standards of safety for the traveling public
2. Improving riding comfort and convenience of the public
73
3. Preserving the considerable investment in structures

4. Providing efficient routes for emergency services
5. Minimizing disruptions and delays, and costs to users
6. Providing economical routes for transport of industrial goods and agricultural products
7. Correcting deficiencies within reasonable time
8. Equitable allocation of resources to the various geographical areas and bridge activities
9. Avoidance of costly repairs though appropriate preventive maintenance
10. Efficient utilization of engineering and maintenance personnel
11. Efficient utilization of funding sources
12. Minimization of total expected cost over the long run
The basic approach to development of Pontis is built on several simple ideas:
1. Separate maintenance, repair and rehabilitation (MR&R) decisions from improvement
decision: MR&R activities retard or repair the effects of deterioration but they do not
directly change the level of service of the bridge, while improvement activities usually
changes it.
2. Divide each bridge into a reasonable number of elements, the sum of which would
describe all bridges in the network.
3. Specify the set of possible condition states that each part of each element can be in.
4. For each element define a homogenous unit in order to represent the parts of a given
element in percent of the total number of units for it.
5. For each condition state define an appropriate set of feasible actions.
6. Define environments in such a way that interactions among elements can be addressed.
7. For each bridge specify the percentage of each element in each condition state.
74
8. Find optimal MR&R policies for each unit, and then bring them together to find optimal
MR&R actions for each bridge.
9. Use a separate optimization procedure to find the optimal set of bridges that should be
chosen for each MR&R budget, and their priority orders.
10. Use functional deficiencies, or instances of failure to meet level-of-service standards, in
order to find candidates for improvement actions.
11. Use reduction of user costs as a basis of determining the benefits of carrying out
improvements for each candidate bridge.
12. Use an optimization procedure to find the optimal set of bridges that should be improved
for each improvement budget, and their priority orders.
13. Bring all the actions specified for MR&R and improvement for a bridge together,
calculate the total benefit of recommended actions on the bridge, and find its priority
order.
14. Integrate all actions and budget requirements to specify the current work plan.
15. Simulate traffic growth and deterioration of components to estimate budget needs in the
future, and for every budget scenario find the future backlog and network conditions.
4.1.2 Organization of Pontis
Figure 4.1 shows schematically the major components of Pontis and the manner in which
they connect to one another in order to meet the objectives concerned. As a result of the
interconnectedness of the system, the optimization models have a determining effect on the
data and organization of the other submodels in the system.
In Pontis each Bridge is divided into its constituent elements, and for each element define
the possible conditions that each unit of element can be in. The definitions of elements and
their units, condition states, and the set of feasible actions for each condition constitute the
Condition State and Feasible Action Model. To capture the environmental effects, this
model places each element of each bridge in one of the following environmental categories:
benign, low, moderate and severe.
The Prediction models estimate deterioration rates for each element and quantify the
uncertainties inherent in such predictions. Prediction models have the capability to learn
from actual experience and automatically update the predictive functions as more data
become available over time.
75
The MR&R optimization model derives its data needs from the Predictive, MR&R Cost,
and the Condition State and Feasible Action models. The outputs are optimal MR&R
policies for each element and each condition, long-term conditions of the network, an
MR&R work plan for each bridge, and the set of bridges that should be scheduled for
MR&R work when budgetary constrains exist.
The Improvement Optimization model weights the benefits of improvement against its cost
and prioritizes the bridges in need of improvement.
The Integrated Project Programming Model schedules the projects to conform to budget
constraints. This model not only combines the results of MR&R and Improvement
Optimization models but is a powerful tool for predicting future network conditions, needs
and backlogs as a function of budget allocations, traffic growth and changes in the level of
service goals and standards.
As seen from figure 4.2, the modular structure of Pontis software necessitates a central role
for the database in the software. The database allows for storing all information needed to
run the models as well as its updating as more information becomes available over time. It
provides the information necessary for running each model, and finally, it stores the results
and outputs of the various models.
76
Fig. 4.1 Pontis Major System Components
77
Fig. 4.2
4.1.3 The Models

4.1.3.1. Database Management System
The Pontis software is structured in a way that implements each of its models as a separate
module communicating with a central database. The principles that govern the database
features of the software are:
Flexibility
Transportability
78
Speed
Capacity
What-if Capacity
One of the important benefits of the modular structures is that Pontis is expandable.
Figure 4.3 shows graphically the internal structure of the database. Most of data in Pontis
are structured into a two-dimensional table having one row per bridge. The columns, or
fields, in the database include all primary record items in the National Bridge Inventory
(NBI), plus additional fields needed for Pontis models. Pontis permits the maintenance of
multiple databases in separated directories. For each database, Pontis maintains separate
indexes, selection files and report files.
Internally, Pontis keeps the data in a highly compressed format, which reduces the hard disk
space impact of having multiple databases, and increases the speed of database access. The
database is physically divided into multiples files, called segments, which also helps to
increase the speed of routine database access. One of these segments is the Program
Segment, that is devoted to project scheduling information. Pontis allows multiple Program
Segment in the same database.
Another segment of the database is devoted to bridge elements. Each bridge may have up to
40 elements. Each element on each bridge is classified in one of the four environmental
categories to represent the effects of climate, operating policy, and other environmental
factors. The element number and environment coded for each element of each bridge are
used to related the bridge-level data to several separate network-level files.
Along with each database, Pontis maintains a set of utility files to speed certain routine
database operations associated with modelling, editing and reporting. These supporting files
are:
Formula files: to create several different models and a set of analytical reports.
Index files: provide a way for Pontis to pre-sort the bridges to make productions of
sorted reports faster.
Manual selection files: offer a similar capability for selecting subsets bridges.
Report files: Pontis has standardized reporting and editing modules which take
advantage of all the above features and which combine them in ad hoc ways.
79
Main Bridge Data Segments
Program Segments
Segment A
Segment B
Bridge 1
Bridge 2
Bridge 3
Bridge 4
Bridge 5
Up to 50,000 bridges
Element 1
Element 2
Element 3
Element 4
Element 5
Element 6
Element 7
Element 8
Up to 40 elements per bridge

Supporting Files
-Index Files
-Manual Selection Files
-Formula Files
-Report Files
Fig.4.3. Internal Structure of Bridge Database
80
4.1.3.2. Condition State and Feasible Action Model

In Pontis each Bridge is divided into its constituent elements, and for each element define
the possible conditions that each one can be in. Bridge elements make a transition among
condition states as the result of deterioration and maintenance. The definitions of elements
and their units, condition states, and the set of feasible actions for each condition constitute
this model.
The number of the condition sates is typically four for each element, and for each condition
state there are one to three feasible actions.
When a bridge is inspected, each element is rated by dividing it among 3 to 5 condition
states. Several important points should be noted about condition states:
The continuous process of deterioration is divided into a small number of discrete steps.
Parts of a given element can be distributed among any or all of possible states.
The parts are represented in percent of the total number of units for this element.
The units of the element migrate over time to worse condition states, so that the percent in
state 1 decrease and the percent in the last state increases. MR&R actions reverse this
process.
The behaviour of each bridge element over time is governed by its environmental and
random effects of traffic and age. To capture the environmental effects, each element of
each bridge is placed in one of the following environmental categories: benign, low,
moderate and severe.
4.1.3.3. Cost models
In Pontis, there are cost models that provide inputs to other models:
Improvement costs: calculates the costs associated with improvement of each bridge to
achieve the level of service goals based on material quantities feasible improvement and the
States unit cost data. These costs are inputs to the Improvement and Replacement Needs
model.
Replacement costs: calculates the replacement cost for each bridge when replacement is a
feasible option. These costs are input to the Improvement and Replacement Needs model
and possibly to the Long-Term Maintenance model.
81
MR&R Costs: This model calculates the cost of the feasible actions associated with
preventive, rehabilitative and corrective measures by computing the necessary material
quantities and using the States unit cost data. It consists of a set of detailed equations that
estimates the various cost associated with each element, condition, and feasible action.
User Cost Model: provides inputs to the improvement optimization model which compares
the savings in user costs due to replacement or improvement with the cost of the investment.
The user cost corresponding to a particular year is computed as the sum of the following
three components:
At = Accident Costs
Ot = Vehicle Operating Costs
Tt = Travel Time Costs
Ut = At + Ot + Tt [4.1]
Some of the measures that are considered in the development of the user costs are:
-
Travel time and fuel consumption due to detours
Travel time and fuel consumption due to traffic congestion
Travel time, fuel consumption and vehicle depreciation due to deck roughness
Travel time due to sharp turns or other geometric problems
Accident costs
The result of this models can also be directly provided as outputs. Pontis considers the
possibility of future needs for expensive corrective actions if appropriate corrective actions
are not taken at the present time.
4.1.3.4. Prediction model
This model estimates deterioration rates for each element and quantifies the uncertainties
inherent in such predictions. This model can learn from actual experience and update its
predictions as more data become available.
82
The prediction model consists of two separated models. The first model, called the Prior
Model, quantifies the likelihood that a unit of a particular element would make a transition
from one condition state to an inferior one within a period. The probabilities are derived
from a series of questions asked of experienced engineers. A computer program assist
engineers in providing the elicited information. The transition probabilities generated by the
elicitation process is the primary data source during the first two years of Pontis usage.
The second model, called the Posterior Model, is used to update the prior transition
probabilities as data become available. After each inspection, this model analyses the
collected data, weights them against the prior transition probabilities, and generates a new
set of transition probabilities that consider the prior information as well as the new data. In
the next cycle, the last posterior transition probabilities would act as prior probabilities for
the next round of updating.
4.1.3.5. MR&R Optimization Model
The objective of the MR&R optimization model is to find for each element in each
environment, the policy which minimizes the long-term maintenance funding requirements
while keeping the elements out of risk of failure.
An important concept in the MR&R optimization is the steady-state condition. This
concept means that, for any given state, the amount of elements passing into it equals the
amount of elements passing out of it every year.
The MR&R model, in reality, consist of two interrelated models. The first model determines
the optimal action for each condition of each element, while the second model calculates the
steady-state network conditions if optimal actions are followed. The models are
probabilistic as they consider the uncertainties associated with deterioration rates for each
conditions. These two models recommend network-wide MR&R policies.
A third component of the MR&R optimization model applies the optimal network policies
to individual bridges to specify what action should be taken for each bridge. It then
calculates the associated cost and prioritizes the bridges in need of work. The mathematical
structure of the model allows to achieve the optimal selection of bridges , and priorization
of those bridges, by calculating a benefit/cost ratio, rank them according to the ratio and
choose the bridges that fall within any cut-off budget level.
The MR&R model receives its data inputs from several other models. The first two
submodels of MR&R optimization model require that, for each element, the set of condition
states and for each condition the set of actions that can be taken be known. This information
is derived from the Condition State and Feasible Action Model. The unit cost for each
feasible MR&R action is derived from MR&R Cost tables. Another important input is the
set of probabilities governing the transition of each element from one condition state to
83
another when an action is taken. This information is provided to the MR&R optimization
model by the Prediction Model, which calculates these transition probabilities. The third
component of the MR&R optimization combines the results of the optimal-action model to
calculate the cost of MR&R for each bridge and requires the quantities and distribution of
the current condition of each element.
4.1.3.6. Improvement Optimization Model
The objective of the improvement optimization model is to maximize the benefit gained, in
terms of user cost savings, from any given level of investment. Replacement is also an
improvement action, which is considered in a manner that integrates improvement
considerations with MR&R considerations.
Improvement decisions usually changes the level of service, but once an action is taken , the
physical characteristics remain the same and no new action to be considered until future
traffic growth makes an action necessary again. The optimization model for improvement is
a static model, and since all parameters are assumed known, it is deterministic.
The mathematical structure of this model is maximization of benefits subject to budget
constraints. The natural way to formulate this problem is a linear integer program. This
formulation allows the incorporation of a variety of constraints in the model. This
mathematical structure and the solution method are similar to the third component of the
MR&R optimization model. We are interested in finding the set of bridges that provide the
highest benefits within a budget limit leading to the optimality of a ranking procedure
between saving in user costs if the action is taken now versus later and are provide by the
Users Cost model.
The results of the improvement model are applicable for any budget constraints and hence
can be used with or without such constraints. By itself the model can recommend the set of
bridges that should be improved for any given budget, and can give the priority rank for
each improvement. The original candidate bridges for improvement are chosen based on
whether or not they meet current levels of service goals for their respective traffic levels.
4.1.3.7. Integration and Program Planning Model
The Integrated Project Programming model not only combines the results of MR&R and
Improvement Optimization models but is a powerful tool for predicting future network
conditions, needs and backlogs as a function of budget allocations, traffic growth and
changes in the level of service goals and standards. It is, by itself, not an optimization model
but a device for bringing the results of the optimization models together, and for simulating
future events according to the criteria set by the other models.
84
In the first period, the algorithm proceeds as follows:

1. Generate a sorted list of the selected bridges having recommended actions. Projects
already in the pipeline go first, followed by the other projects sorted by total benefit/cost
ratio.
2. Scan down the list, programming each project in the first year until the budget is
exhausted.
3. Continue scanning, programming each project in the second year until its budget is
exhausted.
At the beginning of each subsequent period, the algorithm must begin by calculating a new
MR&R benefit/cost ratio to account for intervening deterioration. It does this as follows:
1. For each element on each bridge, it predicts the next periods condition state distribution
by applying the transition probabilities.
2. It determines the recommended actions from the MR&R optimization results, then sums
up the benefits and the costs over the states and elements of the bridge.
3. If the improvement on the bridge have not yet been programmed, it adds the
improvement costs and improvement benefits to the MR&R cost and benefits to give a
new overall benefit/cost ratio for the bridge. Replacement projects are also handle in this
way and projected traffic growth is also handle at this step.
The Integration and Program Planning Model results are presented in two report modules:
The first describes the MR&R and improvement schedules and backlogs, showing how
they grow or shrink over time.
The second shows the detailed first-year project list, with the recommended actions and
the list of the additional actions which, though not scheduled today, are most likely to
need action soon.
In addition, two other reports are available to print any ad hoc reports about the program
and schedule, including lists of scheduled actions by element on each bridge.
85
4.1.4 Maintenance, Repair and Rehabilitation Optimization Model

4.1.4.1 The model and its components
The objective of the MR&R optimization model is to find for each element in each
environment, the policy which minimizes the long-term maintenance funding requirements
while keeping the elements out of risk of failure.
The fundamental difference between MR&R and improvement models is that MR&R
activities are geared towards keeping a bridge in the best possible condition but at its
current level of service, retarding or repairing the effects of deterioration, while
improvement activities usually change the level of service of the bridge.
In the Pontis MR&R model framework, three important things happen every year:
1. Bridge elements deteriorate, making transition from one condition state to another worse
one.
2. Actions taken on specific bridge elements, incurring a cost.
3. The action taken causes an improvement in condition.
Thus, in the network, any given state has elements passing into it due to deterioration or
actions taken in the previous year, and will have elements passing out of it due to
deterioration or actions taken in the following year. Because of this, an important concept in
the MR&R optimization is the steady-state condition. This concept means that, for any
given state, the amount of elements passing into it equals the amount of elements passing
out of it every year. So, the optimal long-term MR&R policy is the one which minimizes the
annual expenditure while satisfy the requirement of a steady-state.
The MR&R model consists of two interrelated models. The first model determines the
optimal action for each condition of each element, while the second model calculates the
steady-state network conditions if optimal actions are followed. These two models are
dynamic optimization and probabilistic models.
A third component of the MR&R optimization model applies the optimal network policies
to individual bridges to specify what action should be taken for each bridge. It then
calculates the associated cost and prioritizes the bridges in need of work.
4.1.4.2. Optimization procedure
MR&R optimization is formulated as a Markov decision model. The planning horizon is
infinite and the future cost are discounted. For each condition of the element it seeks the
86
action that would minimize total discounted costs over long-run, given that the element may
deteriorate, and the optimal policy continues to be followed. There are various methods for
solving this model , but the most efficient is to convert this model into a linear programming
model for which efficient solution methodologies exist.
While the first model is efficient in finding optimal MR&R policies, the second model
would give the steady-state probabilities that the process would be in a given condition state
and that we would take a given action. This information is used to give the steady-state
distribution of conditions of the elements and the proportion of time each action is followed.
This second model allows the incorporation of minimum performance standards and
budget constraints at the element level in the decision-making process if desired at a later
time, so, it provides flexibilities in planning that cannot be obtained through the first model
alone. Hence, we use the first model for specifying actions, and the second model to derive
long-term conditions.
Mathematical formulation
Let Xn denote the state of the process at time n. If a stationary policy is employed, then the
sequences of states [ Xn, n = 0,1,2 ] forms a Markov chain.
The Optimality Equation is the expected life-cycle cost, given the assumptions specified and
bounded action costs, shown as follows:
V(i) = C(i, a ) + Pij (a ) V( j) [4.2]

j
where
i
= condition state observed today.
= condition state predicted to occur one year in the future.
V(i)
= long-term cost expected as a result of being in the state i today.
C(i,a) = initial cost of an action a taken in the state i.
= discount factor for a cost incurred one year in the future.
Pij (a) = transition probability of state j conditional on state i and action a.

V(j)
= long-term cost expected as of next year if state j occurs.
87
Thus, the long-term cost is the sum of initial cost and the discounted sum of the future longterm costs of each possible condition state which may result in the following year. The
optimality equation takes a recursive form through which the network-level model seeks the
set of actions which minimize the long-term costs for each condition state. The combination
of minimum cost actions forms the optimal policy for a given element in a given
environmental.
Data Requirements
The MR&R model receives its inputs from several other models:
-
The set of condition states for each element and the set of action that can be taken for
each condition, from the Condition State and Feasible Action Model.
The unit cost for each feasible MR&R action from the MR&R Cost Tables (MR&R cost
submodel).
The set of probabilities governing the transition of each element from one condition
state to another when an action is taken from the Prediction Model.
Markovian decision process

From the data of the Condition State and Feasible Action Model and the Prediction Model,
the transition probabilities matrix is obtained for each element for an specific period of
years.
Markov models assume that condition states themselves incorporate all information
necessary to predict future deterioration. So, the probability of deterioration depends only
on the current condition data and external factors such as the environmental and action
taken, not on any previous conditions or actions. Thus, condition predictions for any future
year can be made simply by matrix multiplication. For instance the four-year transition
probability matrix is the product of multiplying the two-year probability matrix by itself.
Once the matrix for the period of years selected, the percentage of each element (or unit) in
each state and the unit cost of each possible action are known, the value of the optimization
equation is calculated for each of them.
V(i) = C(i, a ) + Pij (a ) V( j)

j
The action that makes minimum the value of the optimization equation will be selected, for
each state of each element, obtaining in this way the long term optimal policy.
88
After this optimal policy is determined, the dual form of this model is solved using
simultaneous equations to find the steady-state condition state distribution which would be
achieve if optimal policy were follow for a sufficiently long period of time.
4.1.4.3. Priorization procedure
The third component of the MR&R model applies the optimal network policies to individual
bridges to specify what action should be taken for each bridge. It then calculates the
associated cost and prioritizes the bridges in need of work. For each budget, it can specify
the optimal set of bridges that should be selected for MR&R work and their priorities by
simply calculating a benefit/cost ratio, rank them according to the ratio, and choose the
bridges that fall within any cut-off budget level.
Mathematically, the criterion of the third component is maximization of benefits subject to
budget constraints. Benefits are defined as the cost saving resulting from performing all
MR&R work on a bridge in the current year versus postponing them for one period and then
following the optimal action for the condition that the bridge would be in at that time. To
calculate this benefits it needs the predicted condition of each bridge after a n-year cycle, as
well as the cost of carrying out the optimal policy for those conditions. This information is
obtained internally from the Prediction Model, the Cost Tables and the output of the first
MR&R optimization submodel. The cost is the cost to carry out all specified MR&R work
on the bridge.
It is important to notice that , in practice, there is not likely to be enough funding to restore
the whole network to long-term optimal condition right away, so the backlog projects
compete with the more routine projects for the limit funding available. Because of this, they
can be made a part of the same priorization process.
4.1.4.4. MR&R Costs submodel
In the element-level segment of the bridge database, each element on each bridge is
measured, usually in units of linear feet or a count of the occurrences of the element.
MR&R costs are expressed in dollars per unit. Pontis contains a file of MR&R cost factors,
showing for each action possible on each element in each environmental and condition state,
the cost of that action per unit.
In addition to the normal action costs, each element also has failure costs. These represent
the cost to replace the element if the element were allowed to fail without intervention. The
MR&R optimization uses the failure cost along with the probability of failure to assess the
risk of allowing an element to deteriorate. Elements with high failure costs tend to be
recommended for more preventive maintenance and are kept, on average, in better
condition.
89
4.1.4.5. MR&R Results

The outputs of the MR&R optimization model are:
-
Optimal MR&R policies for each element and each condition.
Long-term conditions of the network
MR&R work plan for each bridge
The set of bridges that should be scheduled for MR&R work when budgetary constraints
exit.
Both the MR&R model and the Improvement model generate unconstrained needs and
provide the information necessary to prioritize them. The Pontis Integrated Programming
Model schedules the projects to conform to budget constraints. It can recognize eligibility
requirements and funding constraints for specific funding programs and separately program
eligible projects. It can also simulate the possibility of future year projects and prioritize
them according to their expected benefit/cost ratio to generate rough future schedules.
4.2 DANBRO
4.2.1 Introduction
Bridge management consist of a series of activities involving information gathering,
interpretation, prediction, cost accounting, decision making, budgeting and planning.
DANBRO is a computer-supported BM-system that is use in Denmark today.
The main objectives of this Bridge Management System are:
1. Ensuring traffic safety
2. Optimizing utilization of allocated funds
3. Minimizing maintenance costs
4. Preserving the road network capacity
5. Forecasting budget needs
90
6. Organizing preventive maintenance.

The basic features of DANBRO system are:
1. This system can be used for maintenance work at several levels: executive level,
planning level, administrative level, maintenance level.
2. The components of the system are:
A set of interrelated activities for bridge handling
A set of codes and rules for the activities.
A set of databases.
A set of software to process the collected data.
3. The System is built up of several modules, so, it is possible to adapt it to conditions in

other countries by omitting modules or incorporating new modules.
4. The activities covered by the system are:
Inventories
Special inspection
Tendering
Budget and cost control
91
Feed-back
4.2.2 The Levels

Depending on the user of the program and the end to be achieved, this Bridge Management
System should be utilized at four different levels.
4.2.2.1. The executive level
This is the level at which the decision-makers decide on future policy for bridge
maintenance and thereby about future budgets.
The information we can feed back to the executive with the help of the system includes:
1. The current condition of the bridge stock
2. The need for funds, if we choose to use the optimal repair strategies on all bridges
3. The backlog we built up, if we do not exceed the given funds over the next few years,
together with the additional cost incurred.
4. The consequences for the road standards and traffic flow if the budgets are kept
unchanged or are suddenly changed.
4.2.2.2. The Planning Level
The system enables the planners to decide the amounts to be allocated for preventive
maintenance, repairs and replacements
At the planning level the following decisions are taken:
1. How much to use on routine maintenance out of the total budget.
2. An optimization of repair and rehabilitation works, to suit the allocation set aside for
this type of work.
3. A decision on what new types of repair methods and material are to be introduced and
what older types are to be abandoned.
92
4.2.2.3. The Administrative Level

The system should be able to manage tasks such as scheduling principal and special
inspections, permits for special transports, tendering and control of maintenance works and
budget control.
This Administrative level is the level at which data about the bridges are handled.
Inspectors, office technicians and budget supervisor collect and use these data for their
work.
The data can include:
1. Administrative data on the bridge
2. Technical data on construction of the bridges
3. Inspection data on the condition of the bridges
4. Repair strategies for bridges in need of remedial works
5. Construction costs and repair costs of bridges and bridge components
6. Test data about deterioration: location, type, cause and development.
4.2.2.4. The Preventive Maintenance Level
The system should be used as a tool for writing out work orders, checking quantities for
maintenance works and monitoring the costs of routine maintenance.
The local bridge engineers use the system for the following purposes:
1. Work orders for routine maintenance
2. Work orders for periodic maintenance
3. Control of the quantities for the above mentioned works
4. Control of budgets and costs for routine maintenance
5. Monitoring the quality, costs and service life of materials used for maintenance
93
4.2.3 Organization of DANBRO System

Figure 4.4 shows graphically the main principles of the system:
DANBRO SYSTEM
Administrative/
Condition report
technical data
Bridge database
Inspection data
Maintenance plans
Priority-ranking/optimization
Calculation of
economic
consequences
Economic data
Fig. 4.4. Main Principles
94
Input
Three types of data are entered into the system:
Firstly, the administrative and technical data on the bridges and structures, including
location, size, year of completion, constructions principles and materials.
Secondly, the information on the present maintenance condition of the bridges, partly
from routine registration in connection with the usual maintenance, and partly from
registration of the condition and maintenance needs of the bridges. These needs are
observed at a general inspection, which is carried out every four or five years, and where
each bridge component is given a detailed inspection.
Thirdly, the economic information including budgets and allocations as well as various
repair solutions and their prices.
The bridge data base and the calculation module

All data are registered in the central bridge data base. Another central part of the system is
the calculation module, which can be used to make priority-ranking and optimization
calculations.
Output
Condition report: it is possible to make a damage report for each bridge and to obtain a
description of the changes in the overall condition of the bridges or to find out where the
major problems are located
Maintenance plans: given a certain allocation for the coming year, it is possible to
describe which bridges should be repaired, how the repair should to be carried out and
the estimated costs.
Calculation of economic consequences: the future conditions of the bridges can be

describe as a function of various allocation levels, and the politicians are able to see
what the additional costs will be later on, if they prefer a short-sighted retrenchment
solution, which will result in accelerating damage.
The structure of this Bridge Management System is modular. There are interrelations
between the individual modules, but, in general, each bridge administration should be
allowed to choose which modules to implement.
95
As seen from Figure 4.5, the interrelated activities interact with each other and the data for
these activities are stored in the data base and is used in the computer program.
Inventories are carried out at the head office of the Bridge Department. The information
collected is stored in the basic module.
The bridge inspectors from Bridge Department carry out principal inspections of all bridges
at regular interval(between two and six years).The data from the inspections are stored in
the basic module. Estimates of costs are obtained with the help of the price catalogue.
Special inspections are carried out by experienced materials engineers whenever the need is
recognized during principal inspections or if an unusual event has happened, which may
have a change in a structural part of the bridge.
Proposals for remedial works and costs of these are stored in the optimization module. The
costs are based on the unit prices from the price catalogue. Information about the damage
and measurements are stored in the experience module.
Optimization is carried out once a year when funds are made available, or whenever
changes in the allocated funds are made during the year. Budgets for remedial works to be
carried out are stored in the budget and cost module.
Design of remedial works are carried out by consulting engineers according to the repair
method selected at the optimization, and instructions from the local engineers. The design is
approved by the inspection engineer before tendering.
Remedial works are carried out by contractors after tendering. The economic data from the
tendering are stored in the budget and cost module. The data from the two lowest bids are
stored in the price catalogue.
Long-term budgets are planned on the basis of experience and the forecast of the
optimization module.
Special transports are handle by the police authorities, with the help of information from the
basic module.
96
Inventory
Basic
Principal
Inspection
Tender
prices
Experience
Price Catalogue
Optimization
Optimization
Budget
Policy
Planning of
Maintenance
Budget and
Costs
Maintenance
Special Investigations
Clearance Bearing
Capacities
No Repair
Repair Design
Police authorities
Maintenance
Works
Construction Works
Fig.4.5
4.2.4 Components of the DANBRO System
The components of the system are:
A set of interrelated activities for bridge handling: design, construction, inspection,

maintenance, repair and budgetary work, including the storing of data resulting from the
activities.
A set of codes and rules for the activities, such as road legislation, design, construction
and repair codes and standards and the manuals which have been specially written for
the system.
A set of databases to store information resulting from the activities.
97
A set of software to process the collected data.
4.2.4.1. Interrelated activities

The activities covered by the system are:
Inventories
Special inspection
Tendering
Budget and cost controls
Feed-back
The structure of this Bridge Management System is modular. The most important activities
and the modules that register and use the data supplied by the said activities are described
below.
Inventory
All documentation regarding the design and construction of the bridges is stored on paper
and microfilm in the archives. In order to facilitate daily management, selected data for
which easy access is required, is registered in inventory module.
The inventory contains
98
Administrative data: road register, bridge identifications, etc.
Technical data: bridge types, dimensions, materials, etc.
Passage data: data on roads, water-ways, etc. including clearances and load carrying
capacity classes for the bridges.
Archive references: information on the contents of the archives.
Chronological overview: a list of important events in the life of each bridge, such as
construction, inspections, rehabilitation works, including the most important data from
each event.
The inventory module contains report programs to print out selected data on a single bridge
or on a selection of bridges, and to provide various statistics on the bridge stock. The data
base may be connected to an electronic bridge map which enables the user to find data for a
specific bridge by pointing it out on the screen, or to show all bridges on the map with
special properties.
The data are normally only changed in connection with major repairs or changes in the
administration.
Principal inspection
The principal inspection is a visual inspection of all visible parts of the bridge. For the
inspection the bridge is divided into a fixed number of standards components, one of which
is the bridge in general. For each standard component, the following information is
registered:
-
a condition rating(scale from 0 to 5), ranging from No damage to Ultimate

damage/complete failure of component.
a short description of significant damage.
need for routine maintenance/cleaning
need for Special Inspection
need for repair works
99
An essential part of the principal inspection is to determine the year of the next inspection
for the individual bridge. The interval may vary between one and six years depending on the
condition of the bridge.
Part of the output from principal inspection module is used to make a estimate of the costs
of remedial works to be carried out over the next ten years.
Special inspection
Special inspection are initiated at the principal inspection. When the inspector is not certain
about the cause or the extent of damage, or the proper rehabilitation method, a special
inspection may be called for. Special inspections are carried out by engineers with
experience in deterioration mechanisms, inspection methods and rehabilitation design. They
comprises destructive and non-destructive test carried out in situ, as well as laboratory test
on collected samples.
The aim of the inspection is to identify the problems of a bridge and the factors which have
caused the problems. This is done step by step:
-
determine the cause and extent of damage
predict the development of damage if nothing is done
identify relevant rehabilitation strategies
determine the cost to society of each strategy
choose the strategy that is least costly to society
Two or three repair strategies are considered for a 25 year period and their economic
consequences evaluated. The consequences of a 5 year postponement are also evaluated.
The corresponding road user costs are taken into account, for each strategy.
Frequent routine inspections are carried out in order to monitor the safety and the day-today serviceability of the bridges and to plan the routine maintenance. The Routine
Maintenance Module is based on a list of standard bridge components. For each component,
standard works are linked with corresponding unit prices and specifications of materials to
be used. For each bridge, the principal dimension of each standard component are
registered, together with the standard works applicable to the individual bridge. At the
routine inspection (one a year) the need for each work is registered based on the actual
100
condition of each component and the experience of the inspector, possible in connection
with a general maintenance policy that may have economic implications.
Based in this registrations, work orders for each individual bridge are automatically
generated and printed by the maintenance module, to be used as documentation for the
ordered and executed works.
Optimization
For bridges in need of remedial works an optimization is made. This involves setting up a
priority list that states which bridges are to be repair when, and according to what repair
strategy. The optimization is based on the 5 years budgets together with repair strategies
from special inspections and the forecasts from principal inspection.
Budget and cost controls
The Bridge Management System may comprise a module for managing the flow of money
used for bridge maintenance. This module controls the development of the budgets for the
individual rehabilitation works, as well as the total for all works, thus keeping track of the
budgets throughout the fiscal year.
Long-term budgeting
In addition to the 5 year budgets created by the Principal Inspection Module, there is a need
for long-term estimates for the bridge rehabilitation. The Long-Term Budgeting Module is
intended for this purpose.
The general idea is that average repair intervals and corresponding average repair cost, as
well as average service lives and replacement costs, are registered for the standard
components of the bridges. The year of construction, type and principal dimension of all
standard components are registered for each bridge. Based on this data, the program
calculates total future budgets.
Special permission has to be given for road transport if it is heavier, wider or higher than the
limits given in the traffic regulations. Permission is given by the police authorities. The
police uses lists retrieved from the BMS for the checking.
A special transport is rated according to a set of standard vehicles. Subsequently a route for
the transport can be selected from the information on geometry and bearing capacity stored
in the data base of the DANBRO system.
101
4.2.4.2. Codes and Rules

These are the basic documents that initiate the activities.
Firstly, the Road Legislation is the basis for all activities on the roads, as it states who is
responsible for the construction and maintenance of the roads.
Secondly, there are design codes and standards for bridge construction and repairs together
with standards and guidelines for inspections and special investigations.
Finally, The manuals which have been specifically written for the system give guidelines for
activities not covered by the other documents mentioned.
4.2.4.3. The modular structure of the system
As a consequence of the realization that different bridge owners have different needs, the
structure of a Bridge Management System should be modular. The system is built up of the
following modules:
The Basic Module
The Basic Module contains the administrative and structural data bases and principal
inspection data bases.
A set of programs is included which gives the user of the system access to the screen
pictures composed of information from various data bases. It is possible to choose a screen
picture that gives a chronological review, showing the most important events for a bridge
since construction, and it is also possible to run through a road in this module to obtain
information on clearances and load-carrying capacities for the structures on a given road
stretch. This can be used for the administration of high and heavy transports.
Also programs for print-out of reports are part of the module.
The Maintenance Module
The Maintenance Module contains data bases with information about components of the
bridges which have to be cleaned or maintained at regular intervals. Also a data base
containing possible remedial works on the components is included.
The programs incorporated in this module can automatically print out the work orders at set
102
the starting data of a maintenance job and the time interval between repetitions.
The Price Catalogue
This module is set up to support the estimation of the costs of repairs in connection with
general inspections, special inspections and design.
The data base in this module contains unit prices for various works, based on tender prices.
A built-in program secures that all prices in the module are adjusted to present-day prices
before they are used for estimation.
Unit prices for specified items classified by geographical area or by size of the work can be
calculated and printed. Also combined prices for major works such as deck rehabilitation
can be obtained in this module.
Finally, bills of quantities can be printed for tender documents.
The Optimization Module
The data bases in this module contain the economic data for the various repair alternatives
set up at special inspections. These data, together with the economic data from the principal
inspections, are processed by the optimization program. The program operates at the
network level but incorporates information from the project level.
By means of an iterative process, the program can find the structures which it is most
economic to repair, and the repairs which will cost least to postpone when budgetary limits
do not permit carrying out all repairs at the optimum time.
The net present value method is used and the optimum use of budgetary allocations for a 5
year period is found by iteration.
The program can calculate the consequences of having budgetary allocations lower than the
actual demand. The consequences are additional expenditure because of insufficient
resources to carry out repairs at the optimum time, and also accumulation of postponed
repairs.
103
The Budget and Cost Module

The data bases in this module contains information about budgets for maintenance and
repairs. Changes in works and budgets are registered continuously over the financial year.
Budgets are typically changed after tendering or after a change in a work decided on during
its execution. The module can give information at any time on the expected total
expenditure for a year.
The Experience Module
This module includes damage and economic data. Damage data consist information of the
position, type and cause of the damage, as well as the materials involved. Also the results of
measurements made in the field or on specimens in laboratories will be stored. All the data
are collected in connection with the special investigations.
The economic data includes registration of the expenditures on each structure over a
number of years.
The stored data can be used for analysis to obtain information about service life and service
life costs of bridge and material types.
4.2.5
Optimization of repair and rehabilitation works
4.2.5.1. Selecting Rehabilitation Strategies

In the case where there is no doubt about the cause and extent of damage, or the proper
rehabilitation method, the inspector determines the actions to be taken at the Principal
Inspection. In other case the inspector must call for a Special Inspection.
Special inspections are carried out by engineers with experience in deterioration
mechanisms, inspection methods and rehabilitation design. They comprises destructive and
non-destructive test carried out in situ, as well as laboratory test on collected samples.
Based on the results of these tests, the state of damage is assessed as well as its probable
future development, and various rehabilitation strategies are evaluated determining the cost
to the society of each one and choosing the best strategy.
Rehabilitation strategies
A rehabilitation strategy is a series of actions carried out in order to make sure that the
bridge fulfils its purpose. The full range of relevant strategies may normally be divided into
four types:
-
make a thorough repair now, bringing the bridge back to the No deteriorate condition.
104
make some superficial repairs now in order to postpone major repair.
do nothing now, wait until the bridge, or component, is no longer safe and then replace
it.
do nothing at all; when the bridge is no longer safe, close it and accept the consequent
road-user costs.
Within each of these strategies types, the inspector engineer determines which kind of work
is called for, the optimum time of execution, etc. At the end, the range of relevant strategies
has been reduced to a few significantly different ones, each the economic optimum within
its type.
Economic evaluation
The Net Present Value Method is used for economic evaluation, giving the net present
value of executing or postponing each strategy, including direct and indirect costs to society
within the newt 25 years. The strategy with the lowest net present value is the economic
optimum for the bridge, and will normally be the one proposed as the conclusion of the
inspection report.
The cost considered are the direct costs of repair works and the indirect costs such as roaduser costs and safety implications.
1. Direct costs
The direct cost are the actual costs of the rehabilitation works, including design,
construction, supervision and administration. The estimation of direct costs is based on a
preliminary project and a database with unit prices from similar works (The Price
Catalogue) in combination with the data from the Experience Module.
2. Indirect costs. Road-user costs
It is necessary to include indirect costs when determining which strategy is the optimum for
society, but they should not be include when deciding whether the available funds are
sufficient.
The most significant indirect costs are those inflicted on the road users, cause by the
damage to the bridge. The most common circumstances causing road-user cost are:
-
restrictions on the permitted load on the bridge, imposing detours on heavy trucks.
105
closure of the bridge for safety reasons or during rehabilitation, imposing detours on all
traffic
reduced traffic capacity during rehabilitation, resulting in waiting time and reduced
speed.
The traffic aspect is particularly important for bridges whose closure would involve along
detour or which carry heavy traffic.
Road user cost are calculated based on:
-
Traffic counts
Distribution of vehicle types
Length of detours
Lower speed on detours or through working area
Waiting time
Unit cost per km
Unit cost per hour for each vehicle type
The following formula is specifically applied:

RUC = ADL t (CH , L t W , L + C KM , L d L ) + ADH t (C H , H t W , H + CKM , H d H ) [4.3]
where:
RUC
= Road user cost
ADL
= Average Diary Light Traffic Flow
ADH
= Average Diary Heavy Traffic Flow
CH,L y CH,H
= Unit cost per hour.
CKM,L y CKM,H = Unit cost per kilometre for light and heavy vehicles.
106
tW,L y tW,H
= Additional Waiting time, in hours, for light and heavy vehicles.
dL y dH
= Length of detours in Km for light and heavy vehicles.
3. Additional costs of changing strategy

The normal situation is that the funds for bridge administration are not sufficient for
carrying out the optimum strategy on all bridges. The goal is to find the best solution with
limited funds. This is handle by the Optimization Module with input from the Special
Inspection.
In order to identify the bridges suitable for a change of strategy, the additional cost of
postponing or changing strategy is evaluated as a part of the Special Inspection. For each
strategy, all costs, direct and indirect, are re-evaluated under the assumption that all
activities are postponed by five years.
Thus the input for the Optimization Module from the Special Inspection on a bridge is a set
of normally three strategies, each with a corresponding postponement strategy. Each of
these six strategies is described by the distribution of costs over the next 25 years.
4.2.5.2. Optimization Module
When the funds allocated for bridge maintenance are not sufficient for all the proposed
works, some sort of priority ranking must be made.
The economic analyses from the Special Inspection are taken as the basis for an
optimization. The bridges selected for repair are those for which the economic
consequences to society of postponing the works are worst. These consequences are
calculated as a part of the Special Inspection.
The data requirements for the Optimization Module are:
-
Funds for the coming five years, from the Budget and Cost Module
The discount rate
Economic evaluation of the Special Inspection
The optimization process is divided into three steps:
107
1. Pre-processing
The optimization program calculates the net present value of the optimum solution for each
strategy and that of the postponed solution. By interpolation it also finds the net present
value of the solution in-between, postponed by 1, 2, 3 and 4 years.
The leaves six solutions for each strategy, being the optimum postponed by 0, 1, 2, 3, 4 and
5 years, and each solution having its net present value. So, for each bridge, presuming three
strategies, there should be 18 possible solutions. The 18 solutions are sorted by the net
present value.
2. Automatic optimization
The purpose of the optimization is to find the set of solutions, one for each bridge, for which
the following two criteria are met:
-
The total cost estimate for direct costs lies within the budget each of the first five years.
The economic consequences are the lowest possible.
The problem may be solved using an iterative process, starting with the optimum solution
for each bridge and postponing one strategy at a time until the budgets are met.
3. Post-processing
It is important that the results of the automatic optimization is not immediately accepted as
the final priority-ranking. The result of the optimization must be studied carefully, and the
choice of strategies may be altered taking into account factors that are not included in the
automatic process, because they cannot be expressed in terms of money. Such factors
include aesthetics or environmental aspects, political considerations or co-ordination with
other works on the same road.
4.2.6 The Maintenance Organization
4.2.6.1. The set up of the Maintenance Organization
All activities concerning bridge maintenance are responsibility of the Bridge Maintenance
Division. The head of the Bridge Maintenance Division runs the maintenance organization.
He has a staff to assist him with the planning and the budgetary work carried out at the
planning level. He has three inspectors who take care of ordinary structures and five bridge
maintenance experts who take care of the special structures. Each of these is responsibility
for their part of the daily work at the administrative level.
108
The set up of the organization is shown in Figure 4.6.

Head of Bridge Maintenance Division
Staff
Ordinary
Special
Structures
Structures
Bridge Inspector
Bridge Maintenance Experts
-------------------------------------------------------------------------------------------------------------
Counties
Local Maintenance
Engineers
Fig. 4.6
4.2.6.2. Objectives
The main objectives of the maintenance are:
-
To ensure the daily functioning of the structures.
To prevent that small damages have a non-controlled influence on the service life of the
work or on any of its elements.
109
4.2.6.3. Activities
The activities are grouped in three levels:
-
Level 1: Superficial maintenance: this includes superficial maintenance, emergency

measurements, clearing of obstacles, cleaning, renewal of replaceable parts, reports on
serious damages, etc.
Level 2: Preventive maintenance: this includes painting, exploitation works, small

repairs, etc.
Level 3: maintenance in accordance with Specific Instructions.
4.2.6.4. Routine Maintenance Module

Data requirements
The maintenance module contains data bases with information about components of the
bridges which have to be cleaned or maintained at regular intervals. For each component
standard works are linked with corresponding unit prices and specifications for the materials
to be used. At the routine inspection, once a year, the need for each work is registered based
on the actual condition of each component and the experience of the inspector, possibly in
connection with a general maintenance policy.
Maintenance Module
The programs incorporated in this module can automatically print out the work orders at set
the starting data of a maintenance job and the time interval between repetitions.
The module may comprise facilities for tendering out the routine maintenance works. Based
on the needs for routine maintenance registered at the extended routine inspections, the
program prints out bills of quantities for the tendering.
110
5. THEORETICAL MODELS FOR REPAIR/REPLACEMENT

Two theoretical models have been selected among the models found in the literature review
because of its special interest: the model proposed by D. M. Frangopol (University of
Colorado) and the proposed by F.A.Branco and J.de Brito (University of Lisboa).
5.1 FRANGOPOL
Frangopol proposes a general methodology for determining the optimum inspection/repair
program for new and existing bridges based on minimizing the expected life-cycle costs
while maintaining an acceptable level of reliability. The method determines the optimum
inspection technique, number of lifetime inspections, number of lifetime repairs, and the
timing of these inspections/repairs.
5.1.1 Problem approach
The expected total life-cycle cost CET includes the initial cost CT and the costs of preventive
maintenance CPM, inspection CINS, repair CREP and failure CF. Accordingly, CET can be
expressed as
C ET = CT + C PM + C INS + C REP + C F [5.1]
The objective remains to develop a strategy that minimizes CET while keeping the lifetime
reliability of the structure above a minimum allowable value.
To implement an optimum lifetime strategy, the following problem must be solved:
Minimize CET subject to Pf ,life Pf*,life [5.2]
Where
Pf*,life = maximum acceptable lifetime failure probability (also called lifetime target
failure probability). Alternatively, considering the reliability index

= 1 (1 Pf ) [5.3]
where is the standard normal distribution function, the optimum lifetime strategy is
defined as the solution of the following mathematical problem
Minimize CET subject to life *life [5.4]
111
Where life and *life are the lifetime reliability index and the lifetime target reliability index,
respectively.
The best lifetime strategy is found by solving the optimization problem [5.2].
The solution takes into account the quality of various inspection techniques, all repair
possibilities based on an event tree, the effects of aging and corrosion deterioration, the
damage intensity, the effects of repair on structural reliability, and the time value of money.
The cost analysis includes all components of the overall life-cycle cost according to [5.1].
5.1.2 Preventive Maintenance Cost
Preventive or routine maintenance includes replacing small parts, patching concrete,
repairing cracks, changing lubricants, and cleaning and painting exposed parts. The
structure is kept in working condition by delaying and mitigating the aging effects of wear,
fatigue, and related phenomena.
The cost of routine maintenance is difficult to predict. Traditionally, an engineering cost
associated with the routine maintenance expenditure is used for estimating budgets and
planning. Such estimates are obtained by summing the products of inputs quantities and
their units rates. For example, an organisation might use the average cost per mile for bridge
repair multiplied by the number of miles of bridges as part of an estimate of repair costs.
These average cost rates are derived from observed costs and quantities from a large number
of bridges. They rarely account for factors such as weather, bridge age, and bridge
condition.
The routine maintenance work is proportional to the size and the age of the bridge. It may
become more attractive at some point to replace a bridge rather than spend a large sum of
money to maintain it. Because the maintenance cost increases with time, an estimate of the
routine cost must consider the effect of time. For a given bridge, the cost of routine
maintenance at any time t, Cmain,t, may be assumed a linear function defines as:
C main , t = C main t [5.5]
where Cmain = cost of preventive maintenance at year one; and t = age of the bridge in years.
Assuming a service of 75 years and routine maintenance scheduled once every two years,
the preventive work stars at t = 2 years and continues until t = 74 years. Consequently,
preventive maintenance work will be performed 37 times during the life of the structure.
Therefore, the lifetime routine maintenance cost is:
C PM = C main , 2 + C main , 4 + C main ,6 + .... + C main , 74 [5.6]
112
If the future maintenance costs are converted to their present values, then the lifetime
preventive maintenance cost becomes:
C PM = C main , 2
1
1
1
1
+ C main , 4
+ C main ,6
+ .... + C main , 74
[5.7]
2
4
6
(1 + r )
(1 + r )
(1 + r )
(1 + r ) 74
where r = net discount rate of money.

Numerous factors such as type of bridge, average daily truck traffic, and bridge
environment influence the level of bridge maintenance expenditure. Non-linear cost
functions may be necessary to forecast routine maintenance expenditures based on these
factors.
5.1.3 Special Inspections Cost
When performing special inspection, the ability to detect damage is dependent on the
quality of the inspection technique being used . A higher quality inspection method will
provide a more dependable assessment of damage. No repair will be made unless the
damage is detected.
The detectability function d() can be defined as the probability of detecting damage given
the damage intensity (0 1) as follows:
d() = P(damage detection/) [5.8]
The probability of a defect being detected independent on the damage intensity and the
inspection technique being used:
d() = 0 for 0 min
0.5
for min < < max [5.9]
d() =
ins
d () = 1 for max
where is the distribution function of the standard normal variable, 0.5 = damage intensity
at which there is a 50-50 chance of detection; ins = standard deviation of the detection
ability of the inspection; min damage intensity below which detection is impossible; and
max damage intensity above which detection is absolutely certain.
113
In general, the cost of inspection is dependent on the quality of the NDE (non-destructive
evaluation) method. A higher quality inspection is usually more expensive. Assuming that
the cost for the ideal inspection [i.e., d() = 1 >0] is ins, the cost associated with a real
inspection method, Cins, can be estimated based on the quality of detectability as follows:
Cins = ins (1 min ) 20 [5.10]
where min > 0 is the minimum detectable damage intensity. ins usually is assumed to be a
fraction (i.e., 0.07) of the initial cost CT.
For a strategy involving m lifetime inspections, the total expected inspection cost is
m
C INS = Cins
i =1
1
[5.11]
(1 + r ) Ti
where Cins = inspection cost based on the inspection method used (see Eq.[5.10]); Ti is the
time of inspection and r = net discount rate.
5.1.4 Cost of a repair activity
In most inspection and repair works, it is assumed that the repair cost is constant during the
life of the structure. This is not generally true. Since the repair is part of the life-cycle cost,
it is reasonable to assume that its cost will be a fraction of the replacement construction
cost. In this study, the repair cost is considered to be a function of the replacement cost and
the effect of the repair activity.
Assuming that the after repair damage reduces from br to ar (i.e. ar <br ), where br and
ar are the damage intensities before and after repair, respectively, the corresponding mean
moment capacity increases from Mr,b to Mr,a (i.e., Mr,a > Mr,b), where Mr,b and Mr,a are the
mean moments capacities before and after repair, respectively. The effect of a repair
activity, erep, is defined as the amount by which this activity improves the condition of a
structural component. Because most structural components are evaluated based on their
moment resistances, the effect of a repair activity can be quantified as:
e rep =
M r,a M r,b
Mr0
[5.12]
where Mr0 is the original mean moment capacity of the beam, and 0 < erep < 1. The repair
cost can be expressed in terms of the repair effects as follows:
114
C rep
M M r,b
= repe rep
[5.13]
= rep r , a
M
r0
where = a model parameter; and rep = replacement cost. In this study it is assumed that
rep is equal to the initial cost, rep = CT, and = 0,5.
5.1.5 Repair and Failure Cost
The lifetime (also called life-cycle) cost, target lifetime reliability, inspection interval, and
quality of repair must all be considered when optimizing the inspection/repair strategy of
structural systems. There is a trade-off between a higher reliability and minimum expected
total cost. The goal of an optimal inspection and repair strategy is to minimize the lifetime
cost of a given structure while ensuring that the structure maintains an acceptable reliability
level throughout its expected service life.
An event tree is used to investigate all possible repair events associated with the
inspections. For each case, the structural cross sectional dimensions, corrosion rate v,
number of inspections, loads, allowable reliability level, and median detectability of the
inspection method 0.5 are given. The expected total costs associated with different
inspection/repair strategies are obtained.
The assumptions used to compute the optimal lifetime solution are as follows:
1. The initial design is given and the associated reliability index t = 0 under
nondeteriorating condition is computed.
2. The reliability index is assumed to be a nonincreasing function with time t if no repair
is performed.
3. The deterioration mechanism considered is associated with general corrosion.
4. The loading, material properties, and time-dependent limit state function that describes
the moment capacity of a reinforced concrete T-girder subjected to corrosion are those
described in Lin (1995) and Frangopol et al. (1997).
5. The time value of money is considered using a constant interest rate over time. The net
discount, r, is used to convert the future cost to present cost.
6. If damage is found then a repair action will follow. If the damage is not found then the
repair action will be postponed until the next inspection.
115
Event Tree Analysis

The event tree model provides a systematic means of structuring and evaluating the repair
possibilities related to an uncertain inspection/repair environment. It clearly and precisely
defines the total environment. In this study, the event tree is used as a model to represent all
possible events associated with repair or no repair actions.
To construct an event tree, it is recognized that a decision to either repair or not repair needs
to be made after every inspection. Repair decisions made after every new inspection are
influenced by decisions made in the past. For example, the decision whether or not to repair
after the second inspection will be influenced by whether or not the structure was repaired
after the first inspection. As the number of inspections, m, increases, the number of
branches, 2m, in the event tree increases much faster.
Consider a bridge with three inspections during its entire lifetime using an inspection
method with 0.5 ( Fig. 5.1). The values 0 and 1 represent no repair and repair action,
respectively, and Ti is the time of inspection. Let bij represent the event corresponding to
the occurrence of branch j at time Ti. The inspection interval is ti and the ith inspection
occurs at time
i
Ti = t j [5.14]
j =1
Assume that the bridge is placed in service at t = T0. At time t = T1 - b = T1-, where b
represents a small time interval (e.g; 1 day) before the first inspection occurs, the
probability of failure of this structure is
Pf ,T1 = P[g M (T1 ) 0] [5.15]
where g M (T1 ) 0 defines the failure of the concrete bridge girder due to moment for the
reference time interval (T0, T1-).
116
Fig 5.1
The probabilities associated with the events in Fig 5.1, the probabilities of failure before
each inspection and at the end of lifetime, the lifetime failure probability, and the expected
total failure and repair costs are computed as follows:
1. At t = T1 + a = T1+ where a represents a small time interval (e.g., 1 day) after the first
inspection has been performed, there are two possible events depending on the result of
inspection. In Fig 5.1 the event b11 indicates that the structure is repaired and event b12
indicates no repair. P(b11 ) and P(b12 ) are, respectively, the probabilities that events b11
and b12 occur at time t = T1+ . According to the repair policy the probability of event b11
is:
T 0.5
[5.16]
P(b11 ) = 1
117
The repair effort e rep,1,1 associated with the event b11 can be estimated from [5.12].
Clearly, the probability of event b12 is:
P(b12 ) = 1 P(b11 ) [5.17]
The repair effort e rep,1, 2 associated with the event b11 is nil since no repair effort is
required.
Prior to the second inspection at time t = T2 b = T2 the probability of failure is
calculated for both branches. Given b11 , the failure probability is:
Pf1, T2 = P g1M (T2 ) 0 [5.18]

where g1M ( t ) defines failure given repair at time T1 . The damage intensity is 1T2 .
The superscripts 1 and 0 indicate whether a repair has or has not been performed
after an earlier inspection, respectively. In the same manner given b12 , the structure
was not repaired at T1 so that the probability of failure is:
Pf0, T2 = P g 0M (T2 ) 0 [5.19]

and the damage intensity is 0T2 .
2. After the second inspection given the occurrence of event b11 , there are two possibilities
represented by the complementary events, b12 and b 22 (see Fig. 5.2) that indicate repair
and no repair, respectively. The probabilities of these events are:
1T 0.5
[5.20]
P(b12 ) = 1 P(b 22 ) = 2
In the same manner, there are also two complementary events, b32 and b 42 . The
probabilities that events b32 and b 42 occur are:
0T2 0.5
[5.21]
P(b ) = 1 P(b ) =
3
2
4
2
118
Fig 5.2
The repair effects associated with this events are:

e rep, 2,1 , e rep, 2, 2 = 0, e rep, 2,3 and e rep, 2, 4 = 0.
The four probabilities of failure before the third inspection at t = T3 associated with
damage intensities ijT3 are:
Pfij, T3 = P g ijM (T3 ) 0 [5.22]

This process of possible events is the same for all the inspection. Each branch in the
event tree represents a specific sequence of events bij . Assuming that these events are
mutually statistically independent, then the probabilities of occurrence of paths B1, B2,
B3,B8, respectively, are:
P(B1 ) = P(b13 ) P(b12 ) P(b11 )
P(B2 ) = P(b32 ) P(b12 ) P(b11 )
......
[5.23]
P(B8 ) = P(b83 ) P(b 42 ) P(b12 )
119
The probabilities of failure at the end of lifetime t = T associated with the eight paths
shown in Fig. 5.2 are as follows:
Pfijk, T = P g ijk
M (T ) 0 [5.24]
For each branch, there are four probabilities of failure as follows: before first inspection,
before second inspection, before third inspection, and at the end of lifetime. The lifetime
probabilities of failure for branches 1-8 are defined as:
Pf , life,1 = max(Pf , T1 , Pf1, T2 , Pf11, T3 , Pf111
,T )
.....
[5.25]
Pf , life,8 = max(Pf , T1 , Pf0, T2 , Pf11, T3 , Pf111

,T )
Finally, the lifetime probability of failure of the structure is:
8
Pf , life = Pf , life, i P(Bi ) [5.26]

i =1
If the failure cost is Cf, then the expected failure cost is:
C F = Cf Pf , life [5.27]
The cost of repair associated with branches 1-8 are (see Eqs.[5.7] and [5.13])
C rep,1 =
repe0rep.5 ,1,1
(1 + r ) T1
repe 0rep.5 , 2,1

(1 + r ) T2
repe0rep.5 ,3,1
(1 + r ) T3
....
C rep,8 =
[5.28]
e
0.5
rep rep ,1, 2
T1
(1 + r )
0.5
rep rep , 2 , 4
T2
(1 + r )
0.5
rep rep , 3,8
T3
(1 + r )
Then the expected total repair cost CREP is

2m
C REP = C rep, i P(Bi ) [5.29]

i =1
120
5.1.6 Optimum Strategy

The method to select the optimal strategy consists, as described in the paragraph 1. Problem
Approach, in minimizing the expected total life-cycle cost.
The posible strategies to be considered must be defined in the first place, taking into
account the state of the structure and the possible repairs. The expected total life-cycle cost
will be calculated for each of those strategies through the following expression:
C ET = CT + C PM + C INS + C REP + C F [5.1]
where:
CT = Initial cost
CPM = Preventive maintenance cost. The lifetime routine maintenance cost is:
C PM = C main , T1
1
1
1
+ C main , T2
+ .... + C main , Tn
[5.30]
T1
T2
(1 + r )
(1 + r )
(1 + r ) Tn
where r = net discount rate of money and Ti is the time of preventive maintenance works.
CINS = Total expected inspection cost. For a strategy involving m lifetime inspections, the
total expected inspection cost is
m
C INS = Cins
i =1
1
[5.11]
(1 + r ) Ti
where Cins = inspection cost based on the inspection method used (see Eq.[5.10]) and Ti is
the time of inspection.
CREP = The expected total repair cost:
2m
C REP = C rep, i P(Bi ) [5.29]

i =1
where Crep,i is the cost of repair associated with branch i (see Eq.[5.28]) and P(Bi) the
probability of occurrence of path Bi (see Eq.[5.23]).
121
CF = The expected failure cost:

C F = Cf Pf , life [5.27]
where Cf is the failure cost and Pf,life is the lifetime probability of failure of the structure (see
Eq. [5.26]).
We choose the strategy that minimizes CET and keeping the lifetime reliability of the
structure above a minimum allowable value. This is the optimum lifetime strategy (see
Eq.[5.2]).
5.1.7 Conclusions
In this paper a general methodology was introduced for optimum design of bridge
inspection/repair programs based on reliability and cost. The proposed methodology
minimizes expected life-cycle costs and maintains an expected level of lifetime performance
for a deteriorating bridge. The result identifies the optimum inspection technique, number of
lifetime inspections, number of lifetime repairs, and the timing of these inspections/repairs.
An event tree accounts for all possible repair outcomes following the inspections and the
optimum strategy is based on the likelihood of following various paths on the event tree.
We wish to emphasize some objections of this methodology:
1) One of the assumptions used to compute the optimal lifetime solution is that if damage
is found then a repair action will follow. If the damage is not found then the repair
action will be postponed until the next inspection.
2) The focal points of damage models are oriented towards structural problems (see
definition of erep). The european road network exhibits a different environment. The
medium age of structures is lower than that of USA structures. Faults and occurrences of
damage on european bridges are thus not as much of a structural nature. Instead, the
problems are related more to traffic safety and durability.
3) The objective of this method is to determine the optimum inspection/repair program for
a bridge along its service life and there are many problems when we need to know the
best repair alternative at a certain time. This fact leads to a more complex selection
process which needs a lot of initial data that generally are unknown and have to be
estimated.
122
5.2 BRANCO&BRITO
The importance of bridge maintenance and rehabilitation versus new bridge construction
has become of great concern to bridge authorities. This is because of the rapid structural
deterioration rates that have been observed in some structures or sometimes the lack of
functionality that has occurred with the increasing traffic volumes. The disruption of each
particular bridge has high costs for society: it stops traffic going over it and forces
thousands of users to use alternative routes at extra cost and time.
Nevertheless, bridges are still frequently built according to the criterion of the minimum
initial cost to achieve the proposed level of functionality. This often results in bridges that
are difficult and expensive to inspect and maintain, or that are repaired frequently because
the design was strictly structural with inadequate consideration given to durability. Due to
lack of long-term planning, many bridges become functionally obsolete many years before
the end of their structural life. This leads to high functional failure costs and /or extensive
upgrading works.
In this paper methodology developed to support decision-making on bridge repair strategies
is presented. The decision criteria of this methodology are part of a global management
system which includes a periodic inspection strategy and the selection of repair works
which are performed with a knowledge-based interactive system. The repair decision
module is based on a cost/value economic analysis which compares repair costs and their
subsequent benefits for the expected remaining service life of the structure for each repair
alternative, helping authorities to reach rational decisions.
5.2.1 Global Cost Analysis
For decision-making in bridge management it is necessary to be able to quantify the global
costs of building, using and replacing each bridge and to predict the benefits during the life
cycle. To perform this analysis, a global cost function C was developed:
C = C0 + C I + C M + C R + C F B [5.31]
In which C0 are the initial costs, CI the inspection costs, CM the maintenance costs, CR the
repair costs, CF the failure costs and B the benefits.
123
C0
C0D
C0C
C0T
CI
CIL
CILD
CILT
CIE
CM
CR
CRSA
CRSR
CF
CFSF
CFFR
CFFL
CFFH
CFFF
CFFFD
CFFFV
CFFFL
CFEI
B
BD
BV
BL
initial costs
design costs
construction costs
testing costs
inspection costs
labour costs
displacement costs
testing costs
equipment costs
maintenance costs
repair costs
structural assessment costs
structural repair costs
failure costs
structural failure costs
bridge replacement costs
loss of lives and equipment costs
architectural,
cultural
and
historical costs
functional failure costs
traffic delayed costs
traffic flow rerouted costs
heavy traffic rerouted costs
environmental impact/social costs
benefits
traffic delayed benefits
traffic low rerouted benefits
heavy traffic rerouted benefits
Table 5.1
The benefits from a bridge must also take into account the road of which it is part. To
consider this, the notion of area of influence of each bridge was introduced in the analysis to
quantify the percentage of the benefits of the road that are attributed to each bridge. This
coefficient is equal to the percentage of the initial cost of the bridge in the cost of all the
bridges in the road. In some of the costs listed in Table 5.1 an extra amount is included to
allow for the costs of the area of influence of the bridge.
124
1) Initial costs
Initial costs are those involved in the design and construction of the bridge. They include
preliminary studies, structural and traffic design, construction of the bridge (including its
approaches and the area of influence) and eventual load testing before use.
C 0 = C0 D + C0 C + C0 T [5.32]
Design costs C0D include all the expenses of the preliminary studies (traffic, environmental
impact, geological/drilling, hydrological, economic feasibility, industrial impact, regional
authorities consulted and so on) and the various stages of structural design itself.
Construction costs C0C include all the labour, materials, equipment, building site
management and quality control costs involved in the actual construction of the bridge, its
approaches and area of influence. Some miscellaneous costs, such as demolition of the
existing bridge, right-of-way, utilities, creek diversion and detours, should also be included.
Testing costs C0T are associated with eventual global test of the bridge before it is open to
traffic and with a reference to future maintenance activities.
If the cost analysis is performed after the construction of a bridge, the initial costs are
known. If a new bridge is being analysed, the construction costs, can easily be predicted.
Usually, average costs per square metre of the deck are used, costs that should take account
of the expected structural type, the type and location of the road, and the length of the area
of influence. The design and testing costs are typically predicted as a percentage of
construction costs, based on current engineering fees and local experience in testing. In a
cost analysis, all these costs must be divided over time, and to do this, experience of design
and construction period is important.
2) Inspection costs
Within a bridge management system, inspection costs are those involved in regular
inspection of the bridge and its area of influence within the maintenance framework. They
do not include structural assessment when a main structural deficiency is suspected or the
benefits obtained in terms of increase in bridge safety reliability.
The inspection costs can be divided into:
C I = C IL + C IE [5.33]
125
The labour costs CIL include all the fees of the personnel that perform the inspection and of
those who feed the data into the computer database. The equipment costs CIE include the
amortization of the more expensive inspection equipment (including expendables) being
used at the site and also take into account the time for its transport from one bridge to the
next.
Within the economic analysis, several options can be used to predict the bridge inspection
costs: an automatic computation based on the bridge dimensions and location, current costs
of the authorities (e.g. man-hour cost and daily cost of equipment) and a prefixed calendar
of inspections; use of regression techniques with data from previous years for similar
bridges; and a year to year imposed rate prediction. (To predict the maintenance costs of the
adjoining road, the options are similar.)
3) Maintenance costs
Maintenance costs are those involved in keeping the bridge and its area of influence at their
design level of service and exclude any main structural work. They are often uniformly
distributed over the life of a bridge, as they include only the small repairs that are suggested
as a result of periodic inspections.
Within the cost analysis, there are several options for the estimation of these costs: an
automatic computation in which the yearly maintenance costs of the bridge are a percentage
of its construction costs (this can vary with its age); regression techniques using data from
previous years for the same or similar bridges, and an automatic computation based on the
global current maintenance costs of the authorities and on the bridge dimensions. Fixed
average annual rates for bridge maintenance, with values varying from 1.0% to 2.0% of the
cost construction, have been suggested as the easiest means of predicting maintenance costs.
4) Repair costs
Repair costs are those for the main structural work (repair, strengthening and deck
widening) and include all the costs of the structural assessment usually associated with the
repair decision. For the cost analysis, it is considered that there are no structural repair
works in the area on influence of the bridge.
The bridge repair costs can then be divided into
C R = C RSA + C RSR [5.34]
The structural assessment costs CRSA include all the fees of the personnel that perform the
inspection, the cost of amortization and expendables of the equipment used and the fees
126
involved in the preliminary structural design of the options of repair or enhancement of the
bridge that are proposed.
The structural repair cost CRSR include all the labour, materials, equipment, administration
and quality control involved in the construction of the option that is approved. Cost related
to impairment to traffic are included in the functional failure costs of the year of the repair.
The repair costs usually occur in short periods of the service life of the bridge, frequently
close to the end of that period. Within a management system, if a repair decision has to be
made, the economic analysis must consider computed repair costs based on a repair
solution. To define this solution, a previous structural assessment is needed, with costs that
can be related to the construction costs of the bridge, its deck area and the type of inspection
necessary.
For global economic analysis, average total repair costs can be predicted as distributed over
the service life of the bridge. In this case, they can be estimated by simplified methods, e.g.
regression techniques using data from previous years at the same or similar bridges; an
automatic computation in which the yearly repair costs of the bridge are a percentage of its
construction costs (increasing with its age, after an initial period after construction without
costs); and a estimation of repair works, in discrete years, considering average repair costs
for each technique. Typical average annual values of up to 5% of the structures initial cost
are considered for roughly predict the repair costs.
5) Failure costs
Failure costs are associated with the partial or total impairment of a bridge to fully comply
with its design function and can be divided into,
C F = C FSF + C FFF + C FEI [5.35]
in which CFSF are the structural failure cost, CFFF the functional failure costs and CFEI the
environment/social costs.
Structural failure costs CFSF include all the costs resulting from a structural collapse of a
bridge (or a situation in which such a collapse is imminent and the bridge has to be closed to
traffic). The cost associated with the structural failure can then be obtained from the
probability of failure Pf and the cost of collapse CFF (even though collapse does nor occur
under normal circumstances, these costs can still be considered in an economic analysis as
insurance costs).
C FSF = Pf CFF [5.36]
127
In the economic analysis, the estimation of the probability of failure considers, in a

simplified way, a linear variation in time during the service life of the bridge. A probability
failure path based on degradation mechanisms and the associated reliability index could also
be implemented . Such an approach has a disadvantage in that it involves the need to
consider in the mathematical modelling a large number of parameters which affect the
partial factors for design and assessment and, hence, the acceptable reliability level. Among
other parameters, the following would have to be included: size and importance of the
structure, degrees of redundancy and ductility, design life, type and modes of failure,
inspection and maintenance periodicity and scope and data acquired from in-situ
inspections. The complexity of such an analysis and the difficulty in obtaining reliable data
currently limit its use to very important and onerous projects.
The cost of collapse can be divided into bridge replacement costs CFFR, loss of lives and
equipment CFFL, and architectural, cultural and historical costs CFFH.
C FF = C FFR + C FFL + C FEH [5.37]
The bridge replacement costs CFFR include all the extra expense involved in having to
replace a bridge that theoretically still had a few years of service. To estimate this the
situation of the immediate collapse and replacement of the bridge has to be compared with
the situation of replacement in a few years time. The replacement costs are essentially
those for constructing a new bridge and for the traffic impairment during that period.
The loss of lives and equipment costs CFFL comprise the value of the lives of the people
using the bridge (or what society is prepared to pay to save them) and the value of their
vehicles supposing there is a sudden collapse of the bridge. These costs can be estimated
from current traffic values and normal insurance values for vehicles and people.
The architectural, cultural and historical costs CFFH are a way of over-evaluating bridges
which are especially important from these point of view. An increase in the collapse cost
CFF and thus an increase in the global failure costs, may weaken the replacement option
versus the repair or strengthening options. These costs are difficult to estimate and, in the
economical analysis, are correlated with empirical coefficients that overevaluate the bridge
replacement costs.
Functional failure costs CFFF correspond to the value attributed to the facts that not all the
predicted traffic can use the bridge or that its design speed has to be reduced for certain
periods during the day. The costs of detours of traffic unable to use the bridges must also be
taken into account. For the analysis of each bridge, it is assumed that the other bridges in the
same road have no direct effects on these costs. They can be divided into,
C FFF = C FFFD + C FFFV + C FFFL [5.38]
128
The costs due to traffic delayed CFFFD are those caused by the slowing down of the traffic
crossing the bridge, especially during rush hours. They are estimated from consideration of
the average delay time and hourly value of the average user.
The cost due to detours of traffic in terms of volume CFFFV are those that arise when traffic is
rerouted from one bridge, because the saturation of the bridge in terms of traffic flow, to a
nearby bridge. They are estimated from consideration of the costs associated with additional
travel time, vehicle running expenditure and the increase in the traffic accident rate. The
costs associated with traffic disruption during repavement, widening and lengthening
operations are also included in CFFFV and are frequently larger than the repair costs
themselves.
The costs due to detours of heavy traffic in terms of load CFFFL are those that arise when a
certain margin of exceptionally heavy traffic has to be rerouted from one bridge, because of
its insufficient structural capacity, to a nearby bridge. They are estimated from
consideration of the additional total running costs (including personnel) and traffic accident
costs.
Each of these costs is divided into costs in terms of time wasted by drivers, fuel costs,
vehicle maintenance costs and costs of the increase in traffic accidents. For the economic
analysis, these costs are calculated using several data, e.g. traffic surveys (yearly and daily),
the service design level of the road, future traffic estimates, existing alternatives to each
bridge, its traffic and structural capacity, energy and vehicle maintenance average costs and
estimated repair calendars.
Environmental impact and social costs CFEI correspond to the value attributed to the
consequences of constructing the road and the bridges along it in a particular environment
with its respective population. These costs are usually quite difficult to quantify and are
essentially related to the road in a global way. They can be considered in a simplified way
by taking account, for each bridge, a percentage of the cost of impact-minimizing measures
related to the area of influence of the bridge.
In order to estimate functional costs, it is necessary to predict future traffic. This can be
done in terms of annual volume, by a regression technique or other statistical techniques.
The daily distribution of traffic flow at the bridge (definition of rush hours, etc.) in terms of
vehicles and weights must also be considered, based on measurements or on typical
distributions.
6) Benefits
Benefits are the value of enhancing a bridge in order to provide a better service (or for it to
be of wider use) than the bridge provided at a certain standard situation (usually the design
129
stage). They are necessarily associated with functional failure costs (a benefit is equivalent
to a negative functional failure costs) and are therefore measured in the same manner.
B = BD + BV + BL [5.39]
In the economic analysis, the example situation is that before the road and its bridges were
built. Therefore, all the services it provides are considered to be benefits. However, these
benefits are divided in terms of initial costs between each bridge in the road according to its
relative importance, contrary to the functional failure costs that are attributed totally to the
bridge under analysis. This notion can also be used to compare two solutions of repair,
strengthening, deck widening or replacing of a certain bridge (in such a situation, one of the
options is considered the standard situation with a benefit nil and the other is evaluated
against it).
7) Sensitivity analysis
The economic analysis considers a certain number of parameters the accuracy of which
cannot always be guaranteed (future values of discount rates, inflation rates, initial costs,
inspection costs, maintenance costs, repair costs, probability of structural collapse, global
traffic volume, rerouted traffic volume, etc.). It is therefore useful to know the ones to
which the final results are more sensitive in order to try to estimate them more carefully.
5.2.2 Decision-Making
Every repair decision is made according to the costs effectiveness index (CEI) of each
option that indicates how well the proposed repair compares to the no-action option. The
greater the coefficient for a particular option, the better investment that option is. In the
calculation of CEI, the repair costs CR, the failure costs CF and the benefits B are
considered. For each option the CEI may be quantified by:
CEI =
(C R + C F B) Re pair
(C R + C F B) No action
[5.40]
The CEI coefficient may be used at different levels of action, namely:

-
Level 1. To compare different solutions for the repair of the same anomaly.
Level 2. To obtain the priority of action among the repair of different anomalies within a
bridge. The maximum CEI of each anomaly is considered for comparison between
different anomalies.
130
Level 3. To obtain the priority of action among the bridges of a network. The
accumulated maximum CEIs of each group of repairs within each bridge are considered
to compare different bridges.
6 DECISION SYSTEM FOR REPAIR/REPLACEMENT

6.1 INTRODUCTION
The importance of bridge repair versus new bridge construction has become of great
concern to bridge authorities. This is due to the high structural deterioration rates that have
been observed in some structures or to the lack of functionality that sometimes has
occurred.
The disruption of each particular bridge has very high costs for society. It stops traffic
passing over it and forces thousands of persons to use alternative routes at extra cost and
time.
A method for decision-making regarding possible actions in a deteriorated bridge is
proposed in this study. The decision criteria helps to choose the best repair decision
considering safety, durability, functionality and economy.
This method is based on a global cost analysis that considers all the costs involved in
designing, constructing, inspecting, maintaining, repairing, strengthening and demolishing a
bridge, as well as the road user costs associated with the service life of the bridge.
For decision-making in bridge management it is necessary to be able to quantify the global
costs of building, using and replacing each bridge and to predict the road user costs during
the life cycle. To perform this analysis, a global cost function C was developed:
C = C C + C I + C M + C R + C F + C U + C O VS [6.1]
In which CC are the construction costs, CI the inspection costs, CM the maintenance costs,
CR the repair costs, CF the failure costs, CU the road user costs, CO other costs and VS the
salvage value of the bridge.
The objective remains to develop a strategy that minimizes C while keeping the lifetime
reliability of the structure above a minimum allowable value. To implement an optimum
lifetime strategy, the following problem must be solved:
Minimize C subject to Pf ,life Pf*,life [6.2]
131
where Pf*,life = maximum acceptable lifetime failure probability (also called lifetime target
failure probability).
The actions considered in this method intend to restore the initial service level (design) of
the bridge, without considering an improvement of its initial performances: dimensions,
load carrying capacity, etc. Nevertheless, this method could be also used when all the
considered alternatives lead to the same level of improvement in the bridge.
6.2 METHODOLOGY
A method that consists in the proposal of alternatives for the repair or replacement of a
deteriorated bridge or a bridge with problems for its correct functionality is proposed in this
work.
The global cost of each alternative will be evaluated through a series of factors. The
selection of the most suitable repair/replacement alternative will be based on the
comparison of these costs. This method permits the choice among alternatives that depend
on numerous factors that can be of a very different nature.
In this method the factors are considered independent or, at least, semi-independent,
although in fact it is not always like that.
The possible alternatives must take into account the use of different types of repairs, the
different moments in which each one of the actions can be implemented along the service
life of the bridge and the possible replacement of the structure, that will be considered as
another alternative.
The method proposed is this work is structured in the following phases:
1)
Identification of the factors.
2)
Evaluation of the factors.
3)
Comparison of alternatives and selection.
Any cost involved in the analysis period must be taken into account in the evaluation of the
global cost for every alternative updating its value to a specific instant T0 common to all the
alternatives (For example T0 could be the moment when the study is done).
If in the year Ti (taking as a reference T0 ) a cost Ci is produced, this cost in the instant T0
will be:
132
Ci ,T0 = Ci
1
[6.3]
(1 + r ) Ti
where r = net discount rate of money.

In this way all the other costs along the analysis period will be updated to the reference
instant T0, resulting a total updated cost:
n
C = Ci
i =1
1
[6.4]
(1 + r ) Ti
This cost will be used for the comparison of alternatives.

Regarding the updating procedure, all the costs along a year will be considered as produced
at the end of the said year.
6.3 IDENTIFICATION OF THE FACTORS

The identification of the factors to be taken into account for the comparison of the
repair/replacement alternatives is of great importance, since those aspects not considered in
this phase will be ignored in the rest of the study.
In fact, the identification phase implies a certain prevaluation in which, in a global and
approximate way, roughly values are considered, this making possible to discard effects
already estimated from the beginning as insignificant.
On the other hand, the identification of the factors establishes the degree of detail of the
study to be developed, a general study with few factors highly aggregated or a detail study
with many factors highly disintegrated.
The degree of detail that establishes the identification of factors conditions their evaluation
and the comparison of the alternatives later on. It is difficult to evaluate the highly
aggregated factors since each of them comprises many variables of different nature that are
hard to analyze as a whole. In the opposite case, the highly disintegrated factors are easier to
evaluate, although the selection process is more complicated and more entry data are
necessary.
A list of the factors to observe in a generic bridge is given below; this must be taken as an
indication on the structure and types of factors to be considered and it should be adapted to
each specific bridge.
133
This list contains the factors that can have more relevance in the choice of the best
alternative for repair or replacement. Nevertheless, in some occasions it is advisable to add
or to eliminate factors to the list or either to divide some of them in detail.
134
CI
inspection costs
CM
maintenance costs
CR
repair costs
CRA
structural assessment costs
CRR
structural repair costs
CF
failure costs
CU
road user costs

CUD
traffic delayed costs
CUR
traffic rerouted costs
CURT
time costs
CURO
vehicle operating costs
CURA
accident costs
VS
salvage value
CO
other costs
135
6.4 EVALUATION OF THE FACTORS

It will be necessary to establish a common value unit, that will serve as a reference, to
evaluate the factors. The economic cost has been selected as common unit for this work.
The value of most factors tends to have an objective base and almost always gives a good
adjustment to quantitative evaluations. In spite of the objectivity of the setting out, there are
sometimes difficulties to estimate the value due to several reasons: lack of data, their
accuracy, etc. For example, increase in the accidents rate when the width of the lanes is
reduced in a 15%.
In other occasions, the value of specific factors is more subjective, depending on social and
economic factors, among others, that make difficult to justify its evaluation. For example,
the value of a life, destruction of the artistic or cultural patrimony, social impact produced
by the breakdown of a bridge, etc.
In the case that an specific factor in an alternative implies a benefit, this will be introduced
in the cost evaluation of the said alternative as a negative cost. For example, if in a given
alternative there is a reduction in the time consumed by the users.
In any case, when a study of alternatives is brought up, only those factors whose value
presents differences in some of them will be considered, disregarding those factors whose
value is the same for all of them, since this will not affect the comparison of the alternatives.
For example, the cost of the construction of the original bridge is not considered in decision
making because it is the same for all the alternatives.
The evaluation of all alternatives must be done for the same analysis period, even if not all
them correspond to the same service life. To account for the differences in the service life,
two methods are available:
-
To assume that shorter service life alternatives will be replaced as many times as
necessary to obtain equal service life.
The second method involves shortening the analysis period to the expected service life
of the shorter alternative and attributing a salvage value to the longer service life
alternatives that will be taken into account in the cost evaluation.
The guidelines to estimate the values that can take the factors included in the above list are
given next, taking into account that the best estimation always comes from the real data that
have been obtained.
136
6.4.1 Inspection costs

Within a bridge management system, inspection costs are those involved in regular
inspection of the bridge within the maintenance framework. They do not include structural
assessment when a main structural deficiency is suspected or the benefits obtained in terms
of increase in bridge safety reliability.
The inspection costs can be divided into labour costs and equipment costs.
The labour costs include all the fees of the personnel that perform the inspection and of
those who feed the data into the computer database. The equipment costs include the
amortization of the more expensive inspection equipment (including expendables) being
used at the site and also take into account the time for its transport from one bridge to the
next.
Within the economic analysis, several options can be used to predict the bridge inspection
costs:
An automatic computation based on the bridge dimensions and location, current costs of
the authorities (e.g. man-hour cost and daily cost of equipment) and a prefixed calendar
of inspections.
Use of regression techniques with data from previous years for similar bridges.
A year to year imposed rate prediction.
6.4.2 Maintenance costs

Maintenance costs are those involved in keeping the bridge at their design level of service
and exclude any main structural work. They are often uniformly distributed over the life of a
bridge, and they include only the small repairs that are suggested as a result of periodic
inspections.
The maintenance work is proportional to the size and the age of the bridge. It may become
more attractive at some point to replace a bridge rather than spend a large sum of money to
maintain it. Because the maintenance cost increases with time, an estimate of the routine
cost must consider the effect of time.
Within the cost analysis, there are several options for the estimation of these costs:
An automatic computation in which the yearly maintenance costs of the bridge are a
percentage of its construction costs (this can vary with its age).
137
Regression techniques using data from previous years for the same or similar bridges.
An automatic computation based on the global current maintenance costs of the
authorities and on the bridge dimensions.
Fixed average annual rates for bridge maintenance, with values varying from 1.0% to 2.0%
of the cost construction, have been suggested as the easiest means of predicting maintenance
costs.
6.4.3 Repair costs
Repair costs are those for the main structural work and include all the costs of the structural
assessment usually associated with the repair decision. For the cost analysis, it is considered
that there are no structural repair works in the area of influence of the bridge.
In the case that the replacement of a bridge is raised in any of the alternatives, the cost of
this action will be included in this section of Repair Costs.
The bridge repair costs can then be divided into
C R = C RA + C RR [6.5]
The structural assessment costs CRA include all the fees of the personnel that perform the
inspection, the cost of amortization and expendables of the equipment used and the fees
involved in the preliminary structural design of the options of repair of the bridge that are
proposed.
If the replacement of the bridge is considered as alternative, all the costs derived from the
execution of the project for the new bridge will be included in this section, together with the
demolition project of the existing bridge.
The structural repair cost CRR include all the labour, materials, equipment, administration
and quality control involved in the construction of the option that is approved.
If the replacement of the bridge is considered as alternative, all the costs derived from the
construction, supervision and administration of the new bridge and the demolition of the
existing bridge will be included in this section.
For global economic analysis, repair costs can be estimated using data of other repairs of the
same type in similar bridges, taking into account the extension and the placing of the
damages, the conditions of accessibility, deck area, repair method, etc.
138
6.4.4 Failure costs

Failure costs CF include all the costs resulting from a structural collapse of a bridge or a
situation in which such a collapse is imminent and the bridge has to be closed to traffic. The
cost associated with the structural failure can then be obtained from the probability of
failure Pf and the cost of collapse CFF. Even though collapse does nor occur under normal
circumstances, these costs can still be considered in an economic analysis as insurance
costs.
C F = Pf C FF [6.6]
In the economic analysis, the estimation of the probability of failure considers, in a
simplified way, a linear variation in time during the service life of the bridge. A probability
failure path based on degradation mechanisms and the associated reliability index could also
be implemented . Such an approach has a disadvantage in that it involves the need to
consider in the mathematical modelling a large number of parameters which affect the
partial factors for design and assessment and, hence, the acceptable reliability level. Among
other parameters, the following would have to be included: size and importance of the
structure, degrees of redundancy and ductility, design life, type and modes of failure,
inspection and maintenance periodicity and scope and data acquired from in-situ
inspections. The complexity of such an analysis and the difficulty in obtaining reliable data
currently limit its use to very important and onerous projects.
The cost of collapse can be divided into bridge replacement costs, loss of lives and
equipment, and architectural, cultural and historical costs.
The bridge replacement costs include all the extra expense involved in having to replace a
bridge that theoretically still have a few years of service. To estimate this the situation of the
immediate collapse and replacement of the bridge has to be compared with the situation of
replacement in a few years time. The replacement costs are essentially those for
constructing a new bridge and for the traffic impairment during that period.
The loss of lives and equipment costs comprise the value of the lives of the people using the
bridge (or what society is prepared to pay to save them) and the value of their vehicles
supposing there is a sudden collapse of the bridge. These costs can be estimated from
current traffic values and normal insurance values for vehicles and people.
The architectural, cultural and historical costs are a way of over-evaluating bridges which
are especially important from these point of view.
Anyway, in usual cases the failure costs can be left out for the comparison of alternatives
because of the limitation of the probability of failure value and the little differences in the
cost of collapse among all the alternatives.
139
6.4.5 Road user costs

Road user costs CU correspond to the value attributed to the facts that not all the predicted
traffic can use the bridge or that its design speed has to be reduced for certain periods
during the day. The costs of detours of traffic unable to use the bridges must also be taken
into account. For the analysis of each bridge, it is assumed that the other bridges in the same
road have no direct effects on these costs. They can be divided into,
C U = C UD + C UR [6.7]
In order to evaluate road user costs, it is necessary to predict future traffic. This can be done
in terms of annual volume, by a regression technique or other statistical techniques. The
daily distribution of traffic flow at the bridge (definition of rush hours, etc.) in terms of
vehicles and weights must also be considered, based on measurements or on typical
distributions.
The costs due to traffic delayed CUD are those caused by the slowing down of the traffic
crossing the bridge, especially during rush hours. They are estimated from consideration of
the average delay time and hourly value of the average user.
C UD = ADTL C H ,L t L + ADTH C H ,H t H [6.8]
where:
CUD: costs due to traffic delayed.
ADTL: average daily light traffic flow.
ADTH: average daily heavy traffic flow.
tL: additional waiting time, in hours, for light vehicles.
tH: additional waiting time, in hours, for heavy vehicles.
CH,L: unit cost per hour for light vehicles.
CH,H: unit cost per hour for heavy vehicles.
The costs due to detours of traffic CUR are those that arise when traffic is rerouted from one
bridge, because of the saturation of the bridge in terms of traffic flow or because of its
insufficient structural capacity, to a nearby bridge. They are estimated from consideration of
140
the costs associated with additional travel time CURT, additional vehicle running expenditure
CURO and the increase in the traffic accident rate CURA.
C UR = C URT + C URO + C URA [6.9]
The costs associated with additional travel time CURT due to detours of traffic can be
calculated by the following formula:
C URT = ADTL C H ,L t L + ADTH C H ,H t H [6.10]
where:
CURT: costs due to additional travel time due to detours of traffic.
tL:
additional travel time, in hours, for light vehicles.
tH:
additional travel time, in hours, for heavy vehicles.
CH,L:
unit cost per hour for light vehicles.
CH,H: unit cost per hour for heavy vehicles.

The costs associated with additional vehicle running expenditure CURO due to detours of
traffic can be calculated by the following formula:
C URO = ADTL C km ,L d L + ADTH C km ,H d H [6.11]
where:
CURO: costs due to additional vehicle running expenditure due to detours of traffic.
141
dL:
additional length of detour in km for light vehicles.
dH:
additional length of detour in km for heavy vehicles.
Ckm,L: unit cost per km for light vehicles.

Ckm,H: unit cost per km for heavy vehicles.
The additional accident costs CURA may be calculated as:
C URA = t ADT ri c i [6.12]
i
ADT: average daily traffic flow.

t:
time when the increment of the accident rate occurs.
ri :
increment of the accident rate for type i accidents.
ci :
the cost of type i accident.
i:
type of accident. These are classified in three groups: fatal accidents, bodily injured
accidents and material damage accidents.
Programmes have been developed in several countries to evaluate the road user costs under
different circumstances. For example, In the United Kingdom a computer programme
called QUADRO (QUeues And Delays at ROadworks) provides a method for assessing the
cost imposed on road users while the road works are being carried out, made up of road user
delays (value of time), vehicle operating costs and accident costs, etc.
6.4.6 Salvage value
The salvage value of a bridge is its value at the end of the analysis period. The estimation of
this salvage value, complex in many cases, must be made when the analysis period is lower
than its service life in a given alternative. As a starting point, the salvage value of a bridge is
considered null at the end of its service life and it is equal to the cost of construction when
the bridge is put into operation. Interpolations between this values can be made for
intermediate instants.
142
6.4.7 Other costs

The section, other costs (Co), intends to include other aspects of different nature that can
originate additional costs in some alternatives and whose influence can be important in
some cases. Some orientative examples are given below:
Reduction of gauges or dimension of deck platforms (lanes, shoulders, etc).
Influence of the proposed alternative in other traffic (pedestrian, bicycles, etc).
Absence of alternative routes for light and/or heavy traffic.
If the bridge is used for public transport (buses, coaches, school transport, etc), absence
of other alternative public transport to cover that route (for example, railway).
Influence on other ways of transport (railway, high speed, etc) that can provoke traffic
disruption, limitation of working hours, need of night working hours, etc.
Economic influence of the works in nearby localities: for example, the interruption of a
road can affect shops, industries, etc of the inhabitants of those localities crossed by that
line.
Environmental impact of the works in the locality where they take place: noise, dust,
contaminants, etc.
Loss of the historic, patrimonial, aesthetic, religious and traditional values of the bridge
at all levels: national, provincial, local, etc.
Appearance of additional functioning expenses: personnel, boards, beacons, other
signalling, etc.
Convenience of a given alternative from the point of view of the use of available
equipment, stocked materials, similar actions in nearby places, etc.
6.5 COMPARISON OF THE ALTERNATIVES

As it has been said before, the selection of alternatives is based on the minimization of the
generalized total cost along the analysis period.
In the proposed model every repair decision is made according to repair index (RI) of each
option that indicates how much the proposed repair costs compared with the no-action
option or with respect to any other alternative taken as a reference. The smaller the
143
coefficient for a particular option, the better investment that option is. In the calculation of
RI, the inspection costs CI, the maintenance costs CM, the repair costs CR, the failure costs
CF, the road user costs CU, other costs CO and the salvage value VS are considered. For each
option the RI may be quantified by:
RI =
(C I + C M + C R + C F + C U + C O VS ) Re pair or replacement
[6.13]
(C I + C M + C F + C U + C O VS ) No action or reference alternative
The economic analysis considers a certain number of parameters the accuracy of which
cannot always be guaranteed: values of discount rates, inspection costs, maintenance costs,
probability of structural collapse, evolution of traffic, etc. It is therefore useful to know the
sensitivity of final results to each parameter in order to try to estimate more carefully those
which denote more influence.
The RI coefficient may be used at different levels of actions, but its principal goal is to
compare and select the best alternative of repair/replacement for a bridge.
On the other hand, the method allows to calculate the global cost of all the alternatives. This
makes possible to order the alternatives per cost and to evaluate the difference of cost that
implies to delay in time some actions by only considering this possibility as one more
alternative.
This method can be also used to give data useful to compare different actions among several
bridges of the network, considering that the assessment of the costs should be homogeneous
for all the bridges.
144
7 EXAMPLE
7.1 APPROACH
This example consists in the proposal of six alternatives for the repair or replacement of a
deteriorated bridge. The general characteristics of the bridge are:
Construction and opened to traffic: 1950.
Assumed service life: 100 years.
Year of the analysis: 2000.
At present (2000) the bridge has problems with its structural carrying capacity, therefore the
heavy traffic is rerouted.
Discount rate: 1,5
The following six alternatives are considered:
ALTERNATIVE 1
The bridge will be replaced in 2016. Between 2001 and 2015 the heavy traffic will be
rerouted. In 2016, year of the construction of the new bridge, all the traffic will be rerouted.
The assumed service life of the new bridge is 100 years.
ALTERNATIVE 2
The bridge will be repaired (Repair 1) in 2001. During the repair (three months) the heavy
traffic and half of the light traffic will be rerouted.
ALTERNATIVE 3
The bridge will be repaired in a lighter way than Repair 1 (Repair 2) in 2001 and in 2026.
During the two repairs (two months each one) the heavy traffic and half of the light traffic
will be rerouted.
ALTERNATIVE 4
Is the same as the alternative 2 (Repair 1) but delayed five years. The bridge will be
repaired in 2006.
145
ALTERNATIVE 5
The bridge will be repaired (Repair 3) in 2001. The repair cost of this alternative is lesser
than the repair cost of the alternative 2 (Repair 1) but the repair works last six months.
During the repair the heavy traffic and half of the light traffic will be rerouted.
ALTERNATIVE 6
The bridge will be replaced in 2001. In 2001, year of the construction of the new bridge, all
the traffic will be rerouted. The assumed service life of the new bridge is 100 years.
REPLACEMENT
Alternative
Minimum performance level
Year
2016
2050
ALTERNATIVE 1
146
R E P A IR 1 A lte rn a tiv e 2
R E P A IR 3 A lte rn a tiv e 5
M in im u m
p e rfo rm a n c e le v e l
2001
2050
A L T E R N A T IV E S 2 A N D 5
R E P A IR 2
R E P A IR 2
M in im u m p e rfo rm a n c e le v e l
Y ear
2001
2026
2050
A L T E R N A T IV E 3
147
R E P A IR
M in im u m
1 A lte rn a tiv e 4
p e rfo rm a n c e le v e l
2 0 0 6
2 0 5 0
A L T E R N A T IV E
148
R E P L A C E M E N T A lte rn a tiv e 6
M in im u m p e rf o rm a n c e le v e l
Y ear
2001
2050
A L T E R N A T IV E 6
7.2 CONCLUSIONS
Table 7.1 presents the cost evaluation of the Alternative 1 as example. The penultimate and the
last colum are respectivaly the total discounted cost and the budget for each year.
149
TABLE 7.1
Disc.Rate
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
1.5
ADTL
1800
1813
1825
1838
1851
1864
1877
1890
1903
1917
1930
1944
1957
1971
1985
1999
2013
2027
2041
2055
2069
2084
2099
2113
2128
2143
2158
2173
2188
2204
2219
2235
2250
2266
2282
2298
2314
2330
2346
2363
2379
2396
2413
2430
2447
2464
2481
2498
2516
2533
ALTERNATI
VE 1
ADTH
Inspection C. Mainten. C.
450
7128
23760
453
7128
23760
456
7128
23760
460
7128
23760
463
7128
23760
466
7128
23760
469
7128
23760
473
7128
23760
476
7128
23760
479
7128
23760
483
7128
23760
486
7128
23760
489
7128
23760
493
7128
23760
496
7128
23760
500
7128
11880
503
7128
11880
507
7128
11880
510
7128
11880
514
7128
11880
517
7128
11880
521
7128
11880
525
7128
11880
528
7128
11880
532
7128
11880
536
7128
11880
539
7128
11880
543
7128
11880
547
7128
11880
551
7128
11880
555
7128
11880
559
7128
11880
563
7128
11880
566
7128
11880
570
7128
11880
574
7128
11880
578
7128
11880
583
7128
11880
587
7128
11880
591
7128
11880
595
7128
11880
599
7128
11880
603
7128
11880
607
7128
11880
612
7128
11880
616
7128
11880
620
7128
11880
625
7128
11880
629
7128
11880
633
7128
11880
Heavy U.C.
Time
126144
127027
127916
128812
129713
130621
131536
132456
133384
134317
135257
136204
137158
138118
139085
2376000
140058
Repair C.
Km
394200
396959
399738
402536
405354
408192
411049
413926
416824
419741
422680
425638
428618
431618
434640
437682
Light U.C.
Time
560233
Km
Salvage V.
437682
-1995840
Total C.
551232
554874
558542
562236
565955
569701
573472
577271
581095
584947
588825
592731
596664
600624
604612
3970663
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
-1976832
Tot. Disc. C.
543086
538595
534144
529730
525354
521015
516714
512450
508221
504029
499873
495752
491667
487616
483600
3129006
14758
14539
14325
14113
13904
13699
13496
13297
13100
12907
12716
12528
12343
12161
11981
11804
11629
11458
11288
11121
10957
10795
10636
10478
10324
10171
10021
9873
9727
9583
9441
9302
9164
150 -939004
10269486
Budget
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
30888
2395008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
19008
Figure 7.1 shows the different costs of each alternative: inspection costs, maintenance costs,
repair costs, etc. Figure 7.2 presents the comparison among the alternatives for each cost.
Table 7.2 presents the main results of the economic analysis. The main conclusion from this
analysis is that the best option is the second followed by the fifth. As you can see the RI of
the first option is very high that shows that this option is inadvisable.
RI
Ranking
Alternative 1
4.692
Alternative 2
1.000
Alternative 3
1.063
Alternative 4
2.347
Alternative 5
1.001
Alternative 6
1.793
Table 7.2
151
Alternative 2
12000000
2500000
10000000
2000000
8000000
Euros
Euros
Alternative 1
6000000
4000000
1500000
1000000
500000
2000000
0
0
CI
CM
CR
CU
VS
CI
Alternative 3
CM
CR
CU
VS
CU
VS
CU
VS
Alternative 4
6000000
5000000
Euros
Euros
3000000
2000000
1000000
0
4000000
3000000
2000000
1000000
CI CM CR CU VS
0
CI
2500000
1500000
Euros
Euros
2000000
1000000
500000
0
CM
CR
CR
Alternative 6
Alternative 5
CI
CM
CU
VS
4500000
4000000
3500000
3000000
2500000
2000000
1500000
1000000
500000
0
CI
CM
CR
CL: Inspection costs
CR: Repair costs
VS: Salvage costs
CM: Maintenance costs
CU: User costs
C: Total costs
Figure 7.1
152
Maintenance Cos ts
300000
1200000
250000
1000000
200000
800000
Euros
Euros
Ins pection Cos ts
150000
100000
600000
400000
200000
50000
0
A1
A2
A3
A4
A5
A1
A6
A2
2500000
1000000
2000000
800000
1500000
600000
1000000
500000
A5
A6
400000
200000
0
A1
A2
A3
A4
A5
A6
A1
User Costs
A2
A3
A4
A5
A6
T o ta l C o s ts
10000000
12000000
8000000
10000000
6000000
8000000
Euros
Euros
A4
S a lva g e Va lu e
Euros
Euros
Repair Costs
A3
4000000
2000000
6000000
4000000
2000000
A1
A2
A3
A4
A5
A6
A1
A2
A3
A4
A5
A6
Figure 7.2
153
Bibliography
[1] P.Thoft-Christensen, Bridge Management Systems. present and future. Recent Advances
in Bridge Engineering, J. R. Casas, F. W. Klaiber and A. R. Mar (Eds.). CIMNE Barcelona
1996.
[2] D. M. Frangopol and G. Hearn, Managing the life-cycle safety of deteriorating bridges.
Recent Advances in Bridge Engineering, J. R. Casas, F. W. Klaiber and A. R. Mar (Eds.).
CIMNE Barcelona 1996.
[3] P. C. Das, Bridge Management Methodologies. Recent Advances in Bridge Engineering,
J. R. Casas, F. W. Klaiber and A. R. Mar (Eds.). CIMNE Barcelona 1996.
[4] C. Binet, Condition and repair cost estimate of the french road bridge assets. Recent
Advances in Bridge Engineering, J. R. Casas, F. W. Klaiber and A. R. Mar (Eds.). CIMNE
Barcelona 1996.
[5] M. A. Yaez and A. J. Alonso, The actual state of the Bridge Management System in the
state national highway network of Spain. Recent Advances in Bridge Engineering, J. R.
Casas, F. W. Klaiber and A. R. Mar (Eds.). CIMNE Barcelona 1996.
[6] P. Clausen, Introduction to Bridge Management Systems. 14 Th Congress New Delhi
1992 IABSE.
[7] J. J. Ahlskog , Bridge Management-the answer to the challenge. Bridge Evaluation,
Repair and Rehabilitation, Andrzej Nowak. Kluwer Acadamic Publishers 1990.
[8] K. Golabi, P. D. Thompson, A Network Optimization System for Maintenance and
Improvement of Californias bridges. Bridge Evaluation, Repair and Rehabilitation, Andrzej
Nowak. Kluwer Acadamic Publishers 1990.
[9] N. H. Andersen, DANBRO A Bridge Management System for many levels. Bridge
Evaluation, Repair and Rehabilitation, Andrzej Nowak. Kluwer Acadamic Publishers 1990.
[10] K. C. Sinha, M. D. Bowman, Y. Jiang, S. Murthy, M. Saito, A. Tee, Emerging
Methodologies for Bridge Management Systems. Bridge Evaluation, Repair and
Rehabilitation, Andrzej Nowak. Kluwer Acadamic Publishers 1990.
[11] P. C. Das, New Developments in Bridge Management Methodology. Structural
Engineering International, Volume 8, Number 4, November 1998.
154
[12] B. S. Yanev, Bridge Management for New York City. Structural Engineering
International, Volume 8, Number 3, August 1998.
[13] J. Lauridsen, J. Bjerrum, N. H. Andersen, B. Lassen, Creating a Bridge Management
System. Structural Engineering International, Volume 8, Number 3, August 1998.
[14] J. Naumann, Development of a Bridge Management System in Germany. Operation and
Maintenance of Large Infrastructure Projects, Vincentsen&Jensen (eds). 1998 Balkema,
Rotterdam.
[15] A. Huvstig, Bridge Maintenance Training: experience and future trends. Operation and
Maintenance of Large Infrastructure Projects, Vincentsen&Jensen (eds). 1998 Balkema,
Rotterdam.
[16] P. D. Thompson, E. P. Small, M. Johnson, A. R. Marshall, The Pontis Bridge
Management System. Structural Engineering International, Volume 8, Number 4, November
1998.
[17] A. E. K. Jones, A. R. Cusens, Whole-life costing. Structural Engineering, Volume 75,
Number 7, April 1997.
[18] C. R. Middleton, Concrete bridge assessment: an alternative approach. Structural
Engineering, Volume 75, Nos 23&24, December 1997.
[19] H. Hawk, E. P. Small, The BRIDGIT Bridge Management System. Structural
[20] M. K. Sderqvist, M. Veijola, The Finnish Bridge Management System. Structural
[21] Pontis Technical Manual.
[22] Pontis Users Manual.
[23] Road Directorate, Denmark Ministry of Transpot, Bridge Management System.
[24] Road Directorate, Denmark Ministry of Transport, Bridge Maintenance on the Danish
main road network. 1995.
[25] COWI, Bridge Management System, January 1999.
155
[26] C. A. Collier, C. R. Glagola, Engineering economic and cost analysis. Third edition.
Addison Wesley Longman, Inc. 1998.
[27] Proceedings of the International Symposium on Advances in operation and
maintenance of large infrastructure projects. Copenhagen, Denmark 10-13 may 1998.
Operation and maintenance of large infrastructure projects. Edited by L. Vincentsen and J.
S. Jensen. A.A. Balkema, Rotterdam 1998.
[28] P. Haardt (Federal Highway Research Institute), Concept of a Management System for
the Maintenance of Bridges and Engineering Structures. Bergisch Gladbach 1998.
[29] The Department of the Environment, Trasport and the Regions, Getting the best roads
for our money An introduction to QUADRO. March 1998.
[30] Proceedings of the US-Canada-Europe Workshop on Bridge Engineering. Dubendorf
and Zurich, 11 15 July 1997. Recent advances in bridge engineering. Advanced
rehabilitation, durable materials, nondestructive evaluation and management. U. Meier and
R. Betti (Eds.) 1997.
[31] Proceedings of the NATO Advanced Research Workshop on Bridge Evaluation, Repair
and Rehabilitation. Baltimore, Maryland, U.S.A. April 30 May 2, 1990. Bridge
evaluation, repair and rehabilitation. Edited by A. S. Nowak. NATO ASI Series. Vol. 187.
1990.
[32] Proceedings of the US-Europe Worshop on Bridge Engineering. Barcelona, Spain 1517 July 1996. Recent advances in bridge engineering. Evaluation, management and repair.
J. R. Casas, F. W. Klaiber and A. R. Mar (Eds.). CIMNE Barcelona 1996.
[33] Asociacin Tcnica Espaola de Pretensado. Recomendaciones para la conservacin
de puentes pretensados H. P. 7-92. Edita Colegio de Ingenieros de Caminos, Canales y
Puertos, Madrid.
[34] D. M. Frangopol, K. Y. Lin, A. C. Estes, Life-cycle cost design of deteriorating
structures. Journal of Structural Engineering. October 1997.
[35] D. M. Frangopol, A.C. Estes, Lifetime Bridge Maintenance Strategies Based on System
Reliability. Estructural Engineering International Number 3 1997.
[36] D. M. Frangopol, A.C. Estes, Optimum design of bridge inspection/repair programs
based on reliability and cost. Symposium on The Management of Highway Structures
organized by the Institution of Civil Engineers and The Highways Agency in London. May
1998.
156
[37] A.C. Estes, D. M. Frangopol, Optimum reliability-based inspection/repair strategy

based on minimum expected cost. Structural Safety and Reliability. N. Shiraishi, M.
Shinozuka, and Y. K. Wen, eds., Balkema, Rotterdam 1998.
[38] Papers based on presentations made at The International Symposium The Safety of
Bridges in London July 1996. Safety of Bridges. Edited by P. C. Das. London 1997.
[39] F. A. Branco, J. Brito, Bridge management from design to maintenance. J. R. Casas, F.
W. Klaiber and A. R. Mar (Eds.). CIMNE Barcelona 1996.
[40] F. A. Branco, J. Brito, Decision criteria for concrete bridge repair. Structural
Engineering International Number 2 1995.
[41] J. Brito, F. A. Branco, M. Ibaez, A Knowledge-Based System for Concrete Bridge
Inspection. Concrete International, ACI, Detroit, MI, Feb 1994.
[42] J. Brito, F. A. Branco, A Decision System for Bridge Management. 14Th Congress New
Delhi IABSE. March 1992.
[43] EC BRITE/EURAM. Assessment of Performance and Optimal Strategies for
Inspection and Maintenance of Concrete Structures Using Reliability Based Expert Systems.
Report P3091. Aalborg, Denmark, 1993.
[44] Asociacin Tcnica Espaola de Pretensado. Reparacin y refuerzo de estructuras de
hormign. Gua FIP de buena prctica. Edita Colegio de Ingenieros de Caminos, Canales y
Puertos, Madrid 1995.
157

D7 Final

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

D7 Final

Загружено:

Авторское право:

Доступные форматы

Deliverable D7

BRIDGE MANAGEMENT IN EUROPE

Dr R J Woodward, Transport Research Laboratory (TRL)

Bundesanstalt fuer Strassenwesen (BASt)

Date: June 2000

2.1 BRIDGE MANAGEMENT SYSTEMS. PRESENT AND FUTURE.

2.2 MANAGING THE LIFE-CYCLE SAFETY OF DETERIORATING

2.3 BRIDGE MANAGEMENT METHODOLOGIES.

2.4 CONDITION AND REPAIR COST ESTIMATE OF THE FRENCH

2.5 THE ACTUAL STATE OF THE BRIDGE MANAGEMENT SYSTEM IN

2.6 INTRODUCTION TO BRIDGE MANAGEMENT SYSTEMS.

Study of routes for heavy transport.

2.7 BRIDGE MANAGEMENT-THE ANSWER TO THE CHALLENGE.

By applying any of these definitions to a number of possible alternatives for the

2.8 A NETWORK OPTIMIZATION SYSTEM FOR MAINTENANCE AND

1. Provide a systematic procedure for setting repair and improvement priorities.

Maintenance and rehabilitation vs. improvement and replacement. Maintenance

The main components of the system are as follows:

The optimization procedures address the selection of feasible actions in a consistent,

The Cost Submodels consist of the following components:

Travel time and fuel consumption due to detours

Travel time and fuel consumption due to traffic congestion

Travel time due to sharp turns or other geometric problems

2.9 DANBRO A BRIDGE MANAGEMENT SYSTEM FOR MANY

A set of interrelated activities for bridge handling.

A set of codes and rules for the activities.

A set of EDP-tools to process the collected data.

Routine maintenance and inspection

Optimization and budgeting

Design of remedial works

Execution of remedial works

Budget and cost control

Administration of special transports

6. THE EXPERIENCE MODULE

2.10 EMERGING METHODOLOGIES FOR BRIDGE MANAGEMENT

Information available to bridge inspectors as to the structural condition or traffic safety is

2.11 NEW DEVELOPMENTS IN BRIDGE MANAGEMENT

It makes assessment an integral part of the bidding process.

2.12 BRIDGE MANAGEMENT FOR NEW YORK CITY

performed on the bridge, a history of intervention measures that is currently missing in

2.13 CREATING A BRIDGE MANAGEMENT SYSTEM

a set of interrelated activities for handling bridges

a set of codes and guidelines for theses activities

an organization to manage and carry out these activities

a database holding data resulting from these activities

a set of computer tools for processing the data in the database.

In the development of a Bridge Management System, it must be kept in mind that a

2.13.1 Structuring a Bridge Management System

A condition rating, ranging from No damage to Ultimate damage/complete failure of

A short description of significant damage (less significant damage is not described)

Need for routine maintenance/cleaning (Yes/No)

Need for Special Inspection

determine the cause and extent of damage

predict the development of damage if nothing is done

identify relevant rehabilitation strategies

determine the cost to society of each strategy

choose the strategy that is least costly the society

make some superficial repairs now in order to postpone major repair

2.14 DEVELOPMENT OF A BRIDGE MANAGEMENT SYSTEM IN

Data collection and evaluation

Examination of various strategies

Creation of management programmes

Success evaluation, taking stock

2.15 BRIDGE MAINTENANCE TRAINING: EXPERIENCE AND FUTURE