2nd International Symposium on Service Life Design for Infrastructure

4-6 October 2010, Delft, The Netherlands

THE CONSULTANT’S VIEW ON SERVICE LIFE DESIGN

Carola Edvardsen (1)

(1) COWI A/S, Parallelvej 2, DK-2800 Kongens Lyngby, Denmark

Abstract
There is a rapidly growing international demand for long-term well-performing concrete
structures without premature need for maintenance and repairs. Major structures like bridges
and tunnels are expected to have a long service life in the order of 100, 120 or even more
years to ensure that the investment is spent in a rational way.
Past decades have shown that the classical procedures for durability of reinforced concrete
structures have often failed to provide reliable long-term performance in aggressive
environments. Within Europe this awareness has led to the development of new service life
design approaches to provide necessary and valuable tools to satisfy present day design needs.
These new design tools allow the grossly over-simplistic deem-to-satisfy durability design
methods of traditional codes and standards to be replaced with rational scientifically sound
service life design methods using the same reliability based methods as used for decades for
structural design. These new service life design tools have been implemented in fib bulletin
No. 34 Model Code for Service Life Design and will be part of the new fib Model Code for
Service Life. Furthermore, it will be transferred into an ISO standard in the near future.
COWI spearheading the international developments for such rational service life design
methods during the last decades has adopted this new service life design approach for the
several larger infrastructures worldwide.
The paper presents one of today´s most upfront durability design methodology exemplified
by some case studies.

1. INTRODUCTION
During the past years, clients have asked for bridges, tunnels, prestigious engineering
structures to be designed to satisfy a specified service life, typically 100 and 120 years, and in
particular cases even 200 years. This substantially surpasses the assumed design life of most
traditionally used codes and standards.
Most commonly, the durability is generally “ensured” simply by adopting deem-to-satisfy
rules as given in the codes and standards such as Eurocode, AASHTO LRDF, BS, or DIN.
Experience shows that these rules based on a combination of experience, research and
intuition have many drawbacks and often they result in inadequate durability design. In short,
present codes and standards are often inadequate and not quantifiable.

249
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

The operational way of designing for durability is to define durability as a service life
requirement. In this way, the non-factual and rather subjective concept of “durability” is
transformed into a factual requirement for the “number of years” during which the structure
must perform satisfactorily without unforeseen high costs for maintenance. In this way, the
time factor is introduced as a design parameter.
In Europe this awareness has led to the development of more rational service life design
approaches to satisfy the above design needs. These approaches mainly developed within the
European research project ‘DuraCrete’ (Probabilistic Performance based Durability Design of
Concrete Structures, 1998), further developed in the subsequent ‘DARTS’ research project
(Durable and Reliable Tunnel Structures, 2004) have nowadays been implemented in the fib
Bulletin No. 34 Model Code for Service Life Design, 2006.

2. SERVICE LIFE DESIGN APPROACHES

2.1 Traditional durability design approaches

Most design codes and standards for concrete structures contain a prescriptive or deemed-
to-satisfy approach to the design and specification of reinforced concrete in various exposure
environments, including exposure to chlorides. This approach generally comprises the
specification of limiting values for maximum water/cement ratio, minimum cement content
and minimum cover for various cement types in a limited number of exposure environments.
Somewhat surprising, despite the obvious time-dependent nature of the risk of concrete
deterioration such as reinforcement corrosion, design life has not been included as a specific
consideration. One serious drawback of this deemed-to-satisfy approach is that the design life
is not quantifiable. Predicting the expected service life of concrete structures is not possible.
This philosophy makes the life-cycle cost analysis impossible and hinders the decision-
making process.
The development of the deemed-to-satisfy durability provisions in standards and codes has
generally not been based on fundamental principles but on experience, field observations and
limited research data with a significant element of compromise and engineering judgment
within code and standard committees.
The traditional approach for durability design has other serious limitations. For instance, a
method of testing the initial quality of the concrete in relation to the design life has not been
stated. In addition, the codes and standards often do not differentiate sufficiently with regard
to the actual exposure. For example, considering the concrete pier shown in Figure 1, each
concrete pier shown could be divided in three exposure zones. The first zone remains
submerged at all times. Another zone, which is located above the submerged zone, is
subjected to dry and wet cycles as water elevations change with possible chloride
accumulations due to evaporation effects (splash and tidal zone). Finally, there is a zone that
always stays above the water level and is subjected to atmospheric conditions only. These
three zones behave differently and require different approaches for detailing and design for
service life, a requirement that current codes and standards are not able to answer.
Traditional durability design guidance is usually specific to the area in which the guidance has
been developed and may not necessarily be applicable in other locations. A good example of
this is the Middle East, which when large-scale development began in the 1960’s and 1970’s,
had no local design codes or standards. Designers often used the design standards from their
own countries with no allowance for the often much more aggressive conditions, resulting
from the high temperatures, aggressive groundwater and high salinity sea water, and often

250
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

poor construction quality. Much premature deterioration of concrete structure occurred,

mainly due to chloride induced reinforcement corrosion.

Figure 1: Importance of differentiation in exposure classes

First, the latest generation of Eurocodes and European Standards such as EN 2 and EN 206
do reflect the time-dependant deterioration behaviour and take the service life into
consideration on some way.

2.2 European approach for design for service life - the fib bulletin 34 approach
Based on experience showing that the traditional procedures for durability of reinforced
concrete structures have often failed to provide reliable long-term performance, European
research focus has therefore shifted towards studying the mechanisms, which govern
deterioration of concrete and corrosion of reinforcement and their interrelations. The type of
cements used and the quality of concrete, particularly the influence of the denseness
represented by the diffusivity of the concrete, have attracted more focus.
Between 1996 and 1999, with financial support of the European Commission, a series of
studies were undertaken to develop scientifically verified methods to design and evaluate
concrete structures for durability or service life. The project is referred to as the DuraCrete
Project as an abbreviation for ‘Probabilistic Performance-based Durability Design of Concrete
Structures’. The project was led by COWI and included 12 partners, all from Europe.
Seven major tasks were undertaken under the DuraCrete project ending up with a new
design tool (computer modelling) for service life of reinforced concrete structures. It is a
probabilistically and performance-based service life design approach which considers the
probabilistic nature of the environmental aggressiveness, the degradation processes, and the
material properties involved. This ‘full probabilistic’ approach is basically based on the same
design methodology as generally used for structural designs and, among others, represented
by the LRFD methodology (Load and Resistance Factor Design). Similar to structural design
codes for load, this means that safety requirements and limit states must be defined for the
design service life.
The approach can be used for the design of new structures and in the verification of the
service life of existing structures (re-design). It addresses mainly chloride and carbonation
induced-reinforcement corrosion. These two types are often the decisive deterioration
processes of concrete structures. Chloride- and carbonation-induced corrosion is addressed
through deterioration and transport models capable of predicting the time that it takes to start

251
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

the corrosion. Therefore, the approach is performance-based as the time factor of these effects
is taken into consideration.
Other sources of deterioration in concrete such as sulfate attack, alkali-silica reaction
(ASR) and freeze and thaw attacks are addressed by another approach, the ‘Avoidance of
Deterioration’ approach. In this case, deterioration is prevented up-front by using appropriate
quality concrete, i.e. non-reactive aggregates, sulfate resistant cements, low alkali cements
and concrete with artificial air entrainment. The use of stainless steel reinforcement belongs to
the ‘Avoidance of deterioration’ approach as well and may be an alternative to the
probabilistic design approach in case of reinforcement corrosion.
This approach for durability design has been adopted by national authorities (e.g. the
Dutch Ministry for Transport) and individual clients all over the world. It has been
implemented in the fib Bulletin No. 34 Model Code for Service Life Design. The flow chart in
Figure 2 illustrates the flow of decisions and the design activities needed in a rational service
life design process with a chosen level of reliability. Currently, fib is working on a full
revision of the CEB/FIB Model Code 1990, where the fib Bulletin 34 approach for durability
design will be fully integrated. The fib Bulletin 34 identifies four levels of sophistication in
the performance-based design of concrete structures:

• A full probabilistic design, also called ‘DuraCrete’ approach

• A partial factor (semi-probabilistic) with factors calibrated with level 1 above
• A deem-to-satisfy design corresponding to methods in current codes and standards
but requirements calibrated with level 1 above to the extent possible
• Avoidance of deterioration.

The different levels can be combined within the same structure, but for different parts with
different degrees of exposure.
The direct service life calculation following the full probabilistic design will mainly be
used for larger infrastructures with a required particularly long service life such as bridges,
tunnels, airports, marine structures etc., whereas the deemed-to-satisfy approach is meant for
everyday buildings and normal structures.
It shall be noted that fib Bulletin 34 also has its limitation as it is mainly concerned with
the service life of concrete structures as related to concrete durability. Other factors affecting
the service life of structures such as expansion joints, bearings, coatings etc. are not addressed
to the same level. Further, load effects are not considered. For instance, fatigue caused by
dynamic loading leading to time-depending material degradation and corrosion fatigue
developed by simultaneous action of corrosion and environmental factors are not addressed.
In the following practical examples for the full probabilistic design and the avoidance of
deterioration approach are presented for exemplification.

3. EXAMPLES FOR FIB BULLETIN 34 DESIGN APPROACHES FOR

DURABILITY

3.1 Full probabilistic approach

3.1.1 General
The Busan-Geoje Fixed Link project comprises a 8.2 km motorway link from Busan,
Korea’s southernmost and second largest city, to the island of Geoje (Figure 3). The
connection includes a 4 km immersed tunnel - the deepest in the world at a water depth of 50

252
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

metres - and two cable-stayed bridges, each 2 km long. The project is scheduled for
completion at the end of 2010. COWI is the leading consultant for both the bridges and the
tunnel; DAEWOO E&C is the leading contractor.
The bridges and the tunnel should be designed for a service life of 100 years. A project
specific design basis was developed as an initial part of the design, and details were agreed
with the client and owner. Chloride-induced reinforcement corrosion was identified as the
governing deterioration mechanism in the design for service life, as all other potential
deterioration mechanism (sulfate attack, alkali-silica reaction, frost) were solved through the
‘Avoidance of deterioration’ approach.

Figure 2: The fib Bulletin 34 approach for durability design showing four different levels of
sophistication for service life design

253
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

Figure 3: Busan-Geoje Fixed link with bridges and immersed tunnel designed for 100 years’
service life using the full probabilistic service life approach following the DuraCrete
approach
The following operational basis for the service life design was adopted:
• The design life is 100 years
• The initiation phase of the deterioration by chloride ingress represents the design life.
During the initiation phase chloride penetrates the concrete and reaches the
reinforcement level. The initiation phase ends when the chloride content reaches the
critical level at the reinforcement (threshold level). Thus, the chloride initiation is
defined as the nominal end of service life, see Figure 4.
• A 90% probability of not having any corrosion initiated before the 100 years has
passed, corresponding to 10% probability of premature corrosion initiation. Or, in
other words, adopt the corresponding reliability index (β = 1.3) as design basis. The
fib bulletin 34 indicates a value for β = 1.5-1.8 (6.7% - 3.6% probability of corrosion
initiation).
• Design the minimum concrete cover using the DuraCrete tool. This will provide the
minimum concrete cover and the maximum chloride diffusion coefficient of the
concrete complying with the above-given design criteria.

The surface chloride concentration being generated by the local marine environment is the
force driving the chlorides into the concrete and towards the reinforcement. This value was
determined based on extensive data from similar types of exposure.
The determination of the design quality of the concrete with respect to its chloride
penetrability is the basic protective parameter against premature corrosion initiation and shall
be determined at the initial stage together with the testing of the design compressive and
tensile strength needed for structural design.

Figure 4: Service life of concrete structures. A two phased time-related modelling of

deterioration - initiation and propagation phase
3.1.2 Different design steps and input parameters
Regarding the general concept of the full probabilistic approach and the reliability- and
performance-based principles reference is made to e.g. DuraCrete report, Rostam, Gehlen &
Schiessl, Siemes. One important issue of the approach is that the probabilistic nature of the
environmental aggressivity, the degradation processes and the materials’ properties involved
will be taken into account by stochastic variables, i.e. all input parameters will be defined by
mean values and distribution functions. In the following, the different design steps (1-4) and
the important input parameters (mean values) are listed for the bridge design:

254
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

1. Identification and quantification of the environmental exposure of the different

structural members and their location:
It is assumed that the main potential deterioration risk for the bridges is chloride induced
reinforcement corrosion due to marine exposure. With regard to chloride-induced corrosion
the following different exposure classes have been investigated:
• Splash zone (piers, pylons, caissons)
• Submerged zone (caissons)
• Atmospheric environment (piers, pylons, abutments, bridge decks).

2. Determination of the design quality of the concrete with respect to its design
penetrability for the aggressive substances and their concentrations, as identified
from the environmental exposure:
The following input parameters are decisive: The design surface chloride concentration
(Cl-s) as expected for the quality of concrete and foreseen when exposed to the different
environments:
• Splash zone: Cl-s = 4 % by weight of binder
• Submerged zone: Cl-s = 3 % by weight of binder
• Atmospheric environment: Cl-s = 2 % by weight of binder
The background chloride concentration foreseen in the concrete mix: Cl-0 =0.1% by
weight of binder.
The chloride diffusivity (DCl-), typically the decisive design transport parameter
measured by standard NT Build 492. The test method determines the chloride migration
coefficient, which in the scope of this paper is comparable with the chloride diffusion
coefficient. The diffusion coefficient is a functional requirement within the concrete
specification with which the contractor has to comply: DCl- between 2-7x10-12 m2/s represents
the possible spectrum of different types of cement and binder combinations based on
European experience.
The critical chloride concentration (Cl-cr) is decisive for the chloride concentration at
the level of reinforcement, as it triggers corrosion of the reinforcement:

• Splash zone: Cl-cr = 0.6 by weight of binder

• Submerged zone: Cl-cr = 1.8 by weight of binder
• Atmospheric environment: Cl-cr = 0.8 by weight of binder

The ageing factor (α) represents the ability of the concrete to develop an increased
denseness. It is represented by a decreasing diffusion coefficient with increasing age: It
depends on the type of binder and the environmental conditions:
• Splash zone: α = 0.4, 0.5 and 0.6
• Submerged zone: α = 0.3, 0.4 and 0.5
• Atmospheric environment: α = 0.4 and 0.5

3. Adoption of design life and acceptance criteria:

See above: Pf = 10% corresponding to a reliability index β = 1.3.

4. Determine the nominal concrete cover using the DuraCrete methodology.

255
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

3.1.3 Service life calculations - DuraCrete results

In the following, the service life verification for the Busan-Geoje Fixed Link is
exemplified for the bridge and the different exposure zones.
Tables 1-3 show the spectrum of the obtainable reliabilities β depending on the maximum
chloride diffusion coefficient and the age factor α for the concrete structures for the selected
final nominal concrete cover of 75 mm.
Table 1: Splash zone (piers, pylons and caissons (external). Interrelation chloride diffusion
coefficient, DCl-; age factor, α; reliability, β
Nominal cover Max. DCl- at 28 days of maturity α = 0.40 α = 0.50 α = 0.60
2
(mm) (m /s) β
2 x10-12 2.2 3.0 3.3
3 x10-12 1.5 2.3 2.7
4 x10-12 1.1 1.9 2.3
75
5 x10-12 - 1.6 1.9
6 x10-12 - 1.3 1.7
7 x10-12 - - 1.5
- Not determined (β << 1.3)
Table 2: Submerged zone (piers, pylons and caissons (external).
Interrelation chloride diffusion coefficient, DCl-; age factor, α; reliability, β
Nominal cover Max. DCl at 28 days of maturity α = 0.30 α = 0.40 α = 0.50
2
(mm) (m /s) β
2 x10-12 3.2
3 x10-12 2.5
4 x10-12 2.0 >>2.0 >>2.8
75
5 x10-12 1.7
6 x10-12 1.4
7 x10-12 1.2 2.0 2.8

Table 3: Atmospheric zone (piers, pylons, abutments and bridge decks).

Interrelation chloride diffusion coefficient, DCl-; age factor, α; reliability, β
Max. DCl-at 28 days of
α = 0.40 α = 0.50 α = 0.40 α = 0.50
maturity
nominal cover = 40 mm nominal cover = 50 mm
(m2/s)
β
-12
2 x10 1.3 2.1 2.1 2.5
3 x10-12 0.9 1.6 1.5 1.9
4 x10-12 - 1.2 1.2 1.4
- Not determined (β << 1.3)

256
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

3.1.4 Final selection of the durability related design parameters

The maximum chloride diffusion coefficients DCl- and external concrete covers (nominal
covers) shown in Table 4 have been selected as design basis. Since the composition of the
actual concrete mix(es) and thus the properties of the selected mix(es), particularly the age
factor, have been unknown at the design stage, the age factor was chosen at the low end of
achievable values. Thus, the design parameters recommended below are considered to be on
the safe side.
Table 4: Selected durability design parameters
Exposure zone Level Max. DCl-(m2/s) Cover (mm)
Atmospheric 50
above level -3.5 3.5x10-12
Splash 75
Submerged below level -3.5 6.5x10-12 75

3.1.5 Concrete durability testing

The durability strategy has been implemented by the elaboration of project specific
concrete specifications. The chloride diffusion coefficient is one of the functional
requirements with which the contractor has to comply. During pre-testing and production
testing the contractor has to verify that the actual diffusion coefficient to be achieved for the
actual concrete fulfils the requirement stated in the concrete specifications on the specific
location, with the available concrete constituents (type of cement, cement replacements, etc.)
and with the competence of the available workforce and with respect to local traditions.
A comprehensive pre-testing of different concrete mixes has been undertaken by
DAEWOO E&C Institute of Construction Technology (DICT) to find concrete mixes, which
comply with the design requirement for the chloride diffusion coefficient, see Figures 5 and 6.
As shown in the figures several concrete mixes have the potential to be as durable as required
as the chloride migration coefficients are lower than the required limits of 3.5 x10-12 m2/s
(splash zone) and 6.5 x10-12 m2/s (submerged zone).

Figure 5: Chloride migration Figure 6: Chloride migration coefficient determined

coefficient determined during pre- during pre-testing for the Busan-Geoje Fixed Link.
testing for the Busan-Geoje Fixed Different potential concrete mixes with slag cement
Link. Different potential concrete (Class I = mix design for Cut & Cover, Vent. B/D,
mixes with Portland Cement (A5 = 5% Class II = mix design for Ballast, Class III = mix
air content and A10 = 10% air content) design for immersed Tunnel)

257
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

3.1.6 Other projects

A number of other projects have been designed and assessed along the outlined durability
design approach. Reference is made to e.g. the following recent and current COWI projects:
• Bahrain-Qatar-Causeway, Qatar, 2009. Durability design of the bridges for service
life of 120 years (Figure 7)
• Cityring, Denmark, 2009. Tender durability design for service life of 100 years
• Salahal and Seeb Airport Project, Oman, 2008. Durability design for service life of
50 years
• Lusail Development Project, Qatar, 2007. Durability design for service life of 50
years
• Chong Ming Bridge, Shanghai, China, 2005. Durability design for service life of 100
years
• SuTong Bridge, China, 2003. Durability design for service life of 100 years

The 40 km long Qatar-Bahrain Causeway was one of the first bridge projects in the Middle
East where the client required a durability modelling following a recognized methodology.
One specific task was to document the anticipated range of input parameters or confidence
levels of the various assumptions. The durability modelling was based on the fib Bulletin 34
approach. The use of the approach required specific knowledge of the surface chloride
concentration, the chloride diffusion and related ageing coefficient and how these change with
time and temperature and how the extreme aggressive environment in the Middle East ‘loads’
the structure.

Figure 7: Qatar-Bahrain Causeway at night (rendering by Thomas Lavigne and Christophe

Cheron Architects)

3.2 Avoidance of deterioration approach

3.2.1 Stainless steel reinforcement
The use of stainless steel reinforcement (SSR) in zones exposed to high chloride
concentrations is considered a highly reliable solution following the ‘Avoidance of
deterioration’ approach. This can ensure a very long problem-free service life in the part of
the structure exposed to these high chloride concentrations provided the concrete itself is
made sufficiently resistant to avoid other types of deterioration such as alkali-aggregate
reactions, sulfate attack, or salt scaling.

258
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

From a practical point of view, this technology is particularly interesting because it “only”
solves the corrosion problem. All other techniques and technologies within design,
production, and execution of reinforced concrete structures remain practically unchanged,
which is a very attractive fact to the traditionally very conservative construction industry. Of
particular importance is the often overlooked fact that SSR can be coupled with normal black
steel reinforcement (carbon steel) without causing galvanic corrosion. The reason is that the
two types of steels reach nearly the same electro-chemical potentials when cast into concrete.
This leads to the possibility to use SSR only in those parts of the structure where this is
considered necessary, and then reinforce the remaining parts with ordinary black steel
reinforcement. Such highly exposed parts of bridges could be e.g. edge beams, parapets and
crash barriers exposed to de-icing salts, splash zones of bridges in marine environment, lower
parts of bridge columns and abutments exposed to salty groundwater etc. Another benefit is
the added value which follows from the possibility of accepting the use of locally available
materials, even with chloride contamination, and also accepting the qualifications of the local
workforce as it is.
The Stonecutter Bridge in Hong Kong is one example where stainless steel (grade 1.4301)
has been used successfully. The pylons are heavily reinforced with a multi-layer of ø 50 mm
bars. To achieve a service life of 100 years SSR is used for the outer layer of reinforcement,
the remaining reinforcement is ordinary black steel. In total approximately 3000 tons of
reinforcement were supplied to the project.

Figure 8: Stonecuttter Bridge, Hong Kong. Figure 9: The outer layer of reinforcement of
Design: COWI A/S in Joint Venture with the Stonecuttter Bridge is stainless steel. The
Arup remaining is ordinary black steel.

Similarly, the Shenzhen Corridor Bridge (between Hong Kong and mainland China), the
Sheikh Zayed Bridge (Abu Dhabi) and the Sitra Bridge (Bahrain) have adopted the same
approach. About 5400 tons of SST have been used at the Sitra Bridge. Also for piles the
selective use of SSR has been attracted focus at some recent projects located in the Middle
East. The coupling of SSR and carbon steel can easily be done.

259
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

Figure 10: Sheikh Zayed Bridge, Abu Dhabi, selective use of stainless steel reinforcement

Figure 11: Mix of stainless steel and Figure 12: Danish Motorway. Selective use of
ordinary black steel at piles, AL Khail stainless steel reinforcement for edge beams
Road Interchange, Dubai

In addition, an additional benefit of SSR is the fact that SSR is a poorer cathode than
carbon steel. Therefore, SSR can be beneficial in connection with repairs where ordinary
carbon steel has corroded to such an extent that local replacement or added reinforcement is
needed as part of the repair. A current example of such replacement is the repair of corrosion-
damaged bridge edge beams on Danish motor- and highway bridges using SSR. As a
consequence of this deterioration, the Danish Road Directorate now requires the use of SST
for all edge beams of new bridges.
3.2.2 Steel-fibre reinforced concrete
The use of steel-fibre reinforced concrete is another design option following the
‘Avoidance of Deterioration’ approach.
Fibre reinforced concrete (FRC) is a concrete material modified, typically by adding steel
or synthetic fibre reinforcement. The fibre reinforcement reduces the inherent brittleness of
concrete and results in an improved and potentially high performance material suitable for
civil infrastructure applications. FRC can be utilised together with or without conventional
reinforcing bars (rebars).
For many applications, FRC is a very suitable solution for structural members. Besides
advantages in terms of construction and cost, FRC provides structural benefits and superior
durability properties compared to steel bar reinforced, conventional concrete (RC). The

260
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

capacity of the fibres to reduce crack width and deflection is often more important than the
increase in tensile strength of the concrete. Furthermore, the addition of fibres enables
significant load carrying capacities after cracking and allows for stress redistribution, which
makes the otherwise quasi-brittle behaviour of the concrete more ductile and the whole
structure more robust.
For bored tunnels, the use of steel-fibres instead of the conventional steel bar
reinforcement of the segments is already a common picture nowadays.
Several research investigations have shown that the durability of steel fibre reinforced
concrete (SFRC) under chloride exposure is superior to the one of steel bar reinforced
concrete (RC). Among others, it has been demonstrated that the chloride threshold of fibres in
SFRC is 5-10 times higher than the one of rebars in RC. Under practical conditions, corrosion
in SFRC is generally limited to fibres protruding from the surface or to the surface layer that
may be affected by leaching or carbonation; rust stains on the surface may occur, but chloride-
induced corrosion within a SFRC member is highly unlikely. Furthermore, even in the highly
unlikely event of corrosion of internal fibres, spalling and cracking due to the formation of
voluminous corrosion products (a common durability issue for RC) cannot take place for
SFRC, because the individual cross-sections of the fibres are limited.
The superior durability properties of steel fibres have motivated the designer to build a 16.5
km long sewage tunnel in Abu Dhabi as SFRC bored tunnel lining. The tunnel is up to 50 m
deep and has an internal diameter of 5 m. The project is located in an environment with
extreme high concentrations of chlorides and sulfates in the soil and groundwater. Corrosion
of steel fibres can be safely excluded for the entire cross section of SFRC lining, even if
exposed to the worst chloride concentration of 9%. Hence, the durability design of the sewage
tunnel is determined by the risk of sulfate attack of the concrete itself and the corrosion risk of
additional splitting rebar reinforcement. This carbon steel reinforcement might be necessary
along the longitudinal joints due to the combination of high hoop forces in this deep tunnel
and the ovalisation due to the relatively soft surrounding rock. It should be noted that all
internal concrete surfaces are protected by an HDPE membrane to prevent microbiologically
induced corrosion due to the aerobic bacterial activity (sulfuric acid).
In conclusion, this means that corrosion of fibres is foremost an aesthetic problem, but it
neither interferes with the durability nor with the structural performance of an uncracked
SFRC member. The potential of adopting stainless steel fibre reinforcement will definitely
gain more momentum in the future, not only for tunnels, but also for bridges or structural
parts of bridges.
For bridges the use of SFRC looks promising in case of SFRC reinforced bridge deck
overlays and edge beams. In other Nordic countries such as Sweden and Norway there is a
long tradition for using steel fibres for overlays on bridges. SFRC overlays can potentially
replace the traditionally used bridge insulation completely, see Figure 17. SFRC overlays are
expected to be more durable and tough with respect to both reinforcement corrosion and wear
from traffic. Thus, SFRC overlays are expected to be more cost effective than the traditional
bridge insulation which needs replacement after 20-30 years depending on the amount of
traffic. Similar enhanced durability is expected for edge beams of bridges or other bridge
barrier systems which are often prone to heavy reinforcement corrosion if the traditional steel
reinforcement can be partly or fully replaced by steel fibre reinforcement.
An ongoing Scandinavian research project financed by the Road Authorities from
Denmark, Sweden and Norway aims to document the long-term performance of SFRC used
as bridge overlays in the Nordic countries. The purpose of the research project, which

261
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

includes the application of this layer on several new bridges in Denmark and Sweden, is to
document the service life of bridges without traditional insulation with focus on the concrete
durability and resistance against chloride ingress. The first bridges built in Sweden for about
20 years with this steel-fibre overlay show very promising results.
Another possible area where steel fibres may be used are bridge columns and pre-stressed
elements where the conventional steel reinforcement is replaced by steel fibres. In Denmark
the first pedestrian bridge (designed for tractor loads) will be built as steel-fibre reinforced
pre-stressed concrete.

Figure 14: Steel-fibre

Figure 13: Heating Tunnel Copenhagen with steel-fibre reinforced concrete of tunnel
reinforced concrete for the segmental lining segment

Figure 16: SFRC tunnels and access shaft, STEP

Figure 15: Surface alignment of the project, Abu Dhabi
new STEP sewers in Abu Dhabi

4. WHOLE-LIFE COSTING TECHNIQUE FOR DESIGN EVALUATION

It has been acknowledged that the use of an initial costs basis, i.e. the lowest possible
tender, for selecting a design and construction process is not the most cost-effective in the
longer run. Some clients are thus requiring whole-life cost comparison during the design. The
use of whole-life-cycle costing to evaluate the relationship between initial and in-service
costs, and hence to determine the most cost-beneficial solution, has underlined the need to

262
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

select materials and systems that will enhance durability and avoid costly maintenance and
repair costs. Such a whole-life-cycle costing requires reliable prediction of deterioration rates
and should include disruption costs to take account of any maintenance and repair work that
requires closure of the structure.
Calculations show that whole-life costing evaluations are greatly influenced by the
discount rate that is used. Because of the uncertainties about the future value of money it is
very difficult to select an exact value for discount rates, so it is very difficult to calculate the
absolute value for the whole-life cost of a structure. However, within the decision-making
process, using the best available estimate of discount rate, whole life-cycle costing will give a
good indication of the relative merits of alternative design options.

Figure 17: Steel-fibre reinforced concrete overlay instead of Figure 18: Steel-fibre reinforced
traditional bridge insulation concrete overlay

Figure 19: Comparison of CO2 emission for different types of cement and steel reinforcement

5. SERVICE LIFE AND SUSTAINABILITY

Material selection and construction methods for civil structures are today mainly based on
technical performance criteria such as durability and service life. A change in this might come
in the coming years, because the society will demand that public authorities make sustainable
choices. Owners and contractors will be able to meet this demand if they deliberately consider
carefully a range of technical and environmental factors in different combinations.

263
 2nd International Symposium on Service Life Design for Infrastructure
4-6 October 2010, Delft, The Netherlands

One main factor is the use of cement. When building a structure the production of cement
is usually the main source of CO2 emissions. CO2 emissions can be reduced significantly by
replacing a proportion of the cement with fly ash, see Figure 19. Another approach is to
choose an environmentally friendly type of cement, one which causes minimum CO2
emission to produce.
For the first time a large project has been tendered in Denmark, the Cityring, where the
client sets out strict environmental requirements. The Cityring is a 17 km long new subway
line in the centre of Copenhagen. Among the requirements to the contractor is a minimum 30
per cent reduction of CO2 emissions compared to the existing Metro line.

6. CONCLUSION
Today, valuable tools are available to perform performance-based service life designs for
concrete structures including optimal life-cycle costing. The merits of a good concrete quality
and alternative durability enhancing measures can be quantified. Furthermore, the
consequences for the service life of the actually achieved concrete qualities as measured after
completion of the structures can be used to update the service life.
Several international large infrastructure projects have proven that the probabilistically
based durability approaches are viable and of benefit to the owner, the society and the
construction engineering profession as a whole. In addition, design options such as the use of
stainless steel and steel-fibre reinforced concrete with the aim to avoid the deteriorations are
getting more focus.
However, some educational work is still necessary. First of all, a durability-related quality
needs to be enforced by the owner. The owner must have help to clearly formulate
requirements identifying the service life he wants. As the owner is very often not an expert in
these matters, he has to be advised or educated in order to open up for these new design
approaches and to overcome the still existing traditional way of durability thinking.

REFERENCES
[1] DuraCrete – Final Technical Report, ‘Probabilistic Performance Based Durability Design of
Concrete Structures’, Document BE95-1347/R17, European Brite-EuRam Programme, Published
by CUR, The Netherlands, 2000.
[2] fib Bulletin No. 34 ‘Model Code for Service Life Design’, 2006.
[3] Rostam, S., ‘Performance-Based Design of Structures for the Future’, in ‘Structures for the Future
- The Search for Quality’, Proceedings of IABSE Symposium, Rio de Janeiro, Brazil, 1999.
[4] Gehlen, Ch. & Schiessl, P., ‘Probability-based durability design for the Western Scheldt Tunnel’,
Structural Concrete, Journal of the fib (1999) 1-7.
[5] Siemes, A.J.M. & Edvardsen, C., ‘DuraCrete service life design for concrete structures - A basis
for durability of other building materials and components’ in Proceedings of the 8th Conference on
Durability of Building Materials and Components (8DBMC), Vancouver, BC, Canada, 1999.
[6] Nordtest method NT Build 492 ‘Concrete, mortar and cement-based repair materials: Chloride
migration coefficient from non-steady migration experiments’, 1999
[7] Nordtest method NT Build 492, ‘Concrete, mortar and cement-based repair materials: Chloride
migration coefficient from non-steady migration experiments’, 1999
[8] Nordic Research project of using steel-fibre overlayers instead of traditional bridge isolation.
2007-2009.
[9] Strobach, C.P., Kurth, H., Vojtech, P. and Grunert, J.P., ‘Precast concrete: Steel-fibre-reinforced
prestressed precast beams made of self-compacting concrete’ Concrete Engineering International,
10(3) (2006) 37-40.

264