PIANC Report n° 162 - 2016

RECOMMENDATIONS FOR INCREASED

DURABILITY AND SERVICE LIFE OF NEW MARINE
CONCRETE INFRASTRUCTURE

The World Association for Waterborne Transport Infrastructure

 PIANC
The World Association for
Waterborne Transport Infrastructure

PIANC REPORT N° 162

MARITIME NAVIGATION COMMISSION

RECOMMENDATIONS FOR INCREASED

DURABILITY AND SERVICE LIFE OF NEW
MARINE CONCRETE INFRASTRUCTURE
2016
PIANC has Technical Commissions concerned with inland waterways and ports (InCom),
coastal and ocean waterways (including ports and harbours) (MarCom), environmental
aspects (EnviCom) and sport and pleasure navigation (RecCom).

This report has been produced by an international Working Group convened by the
Maritime Navigation Commission (MarCom). Members of the Working Group represent
several countries and are acknowledged experts in their profession.

The objective of this report is to provide information and recommendations on good

practice. Conformity is not obligatory and engineering judgement should be used in its
application, especially in special circumstances. This report should be seen as an expert
guidance and state-of-the-art on this particular subject. PIANC disclaims all responsibil-
ity in case this report should be presented as an official standard.

PIANC Secrétariat Général

Boulevard du Roi Albert II 20, B 3
B-1000 Bruxelles
Belgique

http://www.pianc.org

VAT BE 408-287-945

ISBN 978-2-87223-240-6

© All rights reserved

TABLE OF CONTENS

Contents
1. INTRODUCTION.................................................................................................................................... 4
1.1 GENERAL .............................................................................................................................................. 4
1.2 TERMS OF REFERENCE ..................................................................................................................... 5
1.3 TARGET READERSHIP ........................................................................................................................ 5
1.4 OBJECTIVE ........................................................................................................................................... 5
2. DURABILITY AND SERVICE LIFE ........................................................................................................ 7
2.1 DETERIORATION PROCESSES .......................................................................................................... 7
2.1.1 GENERAL.......................................................................................................................................... 7
2.1.2 CORROSION OF EMBEDDED STEEL ............................................................................................. 7
2.1.3 FREEZE-THAW RESISTANCE ....................................................................................................... 10
2.1.4 ALKALI-SILICA REACTION............................................................................................................. 10
2.1.5 CHEMICAL SEAWATER ATTACK .................................................................................................. 11
2.1.6 DELAYED ETTRINGITE FORMATION ........................................................................................... 11
2.1.7 EARLY-AGE CRACKING ................................................................................................................ 11
2.1.8 ABRASION ...................................................................................................................................... 12
2.2 CODES AND PRACTICE ..................................................................................................................... 12
2.3 QUALITY ASSURANCE AND ACHIEVED CONSTRUCTION QUALITY ............................................. 13
2.4 CONDITION ASSESSMENT AND PREVENTIVE MAINTENANCE .................................................... 14
2.5 LIFE CYCLE COSTING (LCC) ............................................................................................................. 14
2.6 LIFE CYCLE ASSESSMENT (LCA) ..................................................................................................... 14
3. DURABILITY DESIGN ......................................................................................................................... 16
3.1 PROBABILITY-BASED APPROACH ................................................................................................... 16
3.1.1 GENERAL........................................................................................................................................ 16
3.1.2 DURABILITY ANALYSIS ................................................................................................................. 17
3.1.3 CASE STUDY .................................................................................................................................. 18
3.2 PERFORMANCE-BASED APPROACH ............................................................................................... 21
3.2.1 3.2.1 GENERAL ............................................................................................................................. 21
3.2.2 CHOICE OF DESIGN LIFE.............................................................................................................. 22
3.2.3 ENVIRONMENTAL CONDITIONS AND MAIN RISKS OF DAMAGE .............................................. 22
3.2.4 SELECTION OF DURABILITY INDICATORS AND ASSOCIATED SPECIFICATIONS .................. 22
3.2.5 DESIGN OF THE CONCRETE MIX ................................................................................................ 23
3.2.6 TESTING BEFORE CONSTRUCTION AND FOR DIFFUSIVITY CONTROL ................................. 23
3.2.7 INITIAL DURABILITY POINT AND START OF STRUCTURE MONITORING ................................ 23
4. CHAPTER 4: ADDITIONAL STRATEGIES AND PROTECTIVE MEASURES .................................... 25
4.1 GENERAL ............................................................................................................................................ 25
4.2 STAINLESS STEEL REINFORCEMENT ............................................................................................. 25
4.3 NON-METALLIC REINFORCEMENT .................................................................................................. 26
4.4 CONCRETE SURFACE PROTECTION .............................................................................................. 27
4.5 CONCRETE HYDROPHOBATION ...................................................................................................... 29
4.6 PROTECTIVE SKIN SYSTEMS ........................................................................................................... 29
4.7 CATHODIC PREVENTION SYSTEMS ................................................................................................ 29
4.8 CORROSION INHIBITORS ................................................................................................................. 30
4.8.1 STRUCTURAL SHAPE.................................................................................................................... 30
4.8.2 PREFABRICATED STRUCTURAL ELEMENTS ............................................................................. 31
5. QUALITY ASSURANCE AND ACHIEVED CONSTRUCTION QUALITY ............................................. 33
5.1 CONCRETE QUALITY ASSURANCE ................................................................................................. 33
5.1.1 GENERAL........................................................................................................................................ 33
5.1.2 CONTROL OF CONCRETE QUALITY ............................................................................................ 33
5.1.3 CONTROL OF CONCRETE COVER .............................................................................................. 34
5.2 ACHIEVED CONSTRUCTION QUALITY ............................................................................................. 34
5.2.1 GENERAL........................................................................................................................................ 34
5.2.2 COMPLIANCE WITH DURABILITY SPECIFICATION .................................................................... 34
5.2.3 IN SITU QUALITY ............................................................................................................................ 34
5.2.4 POTENTIAL QUALITY .................................................................................................................... 35
5.3 EXAMPLE OF OFFSHORE QUALITY CONTROL IN THE 1970s ....................................................... 35
6. CONDITION ASSESSMENT, PREVENTIVE MAINTENANCE AND REPAIRS ................................... 38
6.1 GENERAL ............................................................................................................................................ 38
6.2 CONTROL OF CHLORIDE INGRESS ................................................................................................. 38
6.3 PROBABILITY OF CORROSION......................................................................................................... 39
6.4 PROTECTIVE MEASURES ................................................................................................................. 41
6.5 REPAIRS ............................................................................................................................................. 41
6.5.1 PATCH REPAIRS ............................................................................................................................ 41
6.5.2 CATHODIC PROTECTION.............................................................................................................. 41
7. REFERENCES..................................................................................................................................... 43
APPENDIX A: WORKING GROUP MEMBERS .................................................................................................... 50
APPENDIX B: TERMS OF REFERENCE.............................................................................................................. 52
APPENDIX C: CONCRETE COVER ..................................................................................................................... 54
APPENDIX D: DURACON MODEL ....................................................................................................................... 60
APPENDIX E: STADIUM® MODEL ....................................................................................................................... 70
APPENDIX F: Life-365TM MODEL ......................................................................................................................... 81
APPENDIX G: ADDITIONAL TEST METHODS FOR QUALITY CONTROL ....................................................... 85

2
EXECUTIVE SUMMARY

The durability of concrete structures in the marine environment is not only related to design and materials
but also to construction. Thus, much of the observed durability problems on marine concrete structures
can be ascribed to lack of proper quality assurance during concrete construction and poorly achieved
construction quality. Upon completion of new concrete structures, the achieved construction quality
typically shows a high scatter and variability, and during the operation of the structures, any weaknesses
and deficiencies will soon be revealed whatever durability specifications and materials have been
applied. To a certain extent, a probability approach to the durability design can accommodate the high
scatter and variability of quality. However, a numerical approach alone is insufficient for ensuring the
durability; greater control and improvements in durability also require the specification of performance-
based durability requirements which can be verified and controlled during concrete construction in order
to practically achieve quality assurance.

As a basis for the durability design and production of new major concrete infrastructure, all minimum
requirements in existing concrete codes and standards, as well as all established recommendations and
guidelines for good construction practice must be strictly followed. In recent years, however, a rapid
increase of international research and development has taken place and new experience gained on how
to obtain a more controlled and increased durability beyond what is possible when based only on existing
concrete codes and practice. Much of this research has been based on a probability approach to the
durability design, which makes it easier to accommodate the high scatter and variability of quality. Also,
much of the research has been based on the development of new performance-based durability
requirements, which provide a better basis both for durability specification and quality assurance during
concrete construction. As a result, a better durability design and quality assurance for new concrete
infrastructure can be achieved, and documentation of as-built construction quality and compliance with
the durability specification can be obtained.

For new major concrete infrastructure where probability-based durability design and performance-based
durability requirements have been applied in recent years, it has been very important for the owners to
receive documentation of achieved construction quality and compliance with the durability specification.
Moreover, it has been very important to receive this documentation before the structures were formally
handed over from the contractors, since this may have implications both for the obtained durability and
future operation of the structures.

In recent years, many owners of existing concrete infrastructure have experienced a significant and
rapidly increasing proportion of their limited construction budgets being spent on repairs and
maintenance of the structures. This development is not only unfortunate from a cost point of view; it
directly affects the sustainability of our society. Therefore, many owners are showing an increasing
interest to invest somewhat more at the outset of their new projects in order to obtain a better controlled
and enhanced durability of the structures beyond what is possible when based only on existing concrete
codes and practice ; even relatively small additional costs have proved to be an extremely good
investment.

In the current document, some additional recommendations and guidelines to existing concrete
standards for durability and service life are provided, the objective of which has been to obtain a better
controlled and enhanced durability of new marine concrete infrastructure beyond what is possible when
based only on existing concrete standards. This guidance is given with emphasis upon durability design
and quality assurance as well as condition assessment and preventive maintenance during the
operational life of the structures.

3
1. INTRODUCTION
1.1 GENERAL

Although a number of different deterioration processes can affect the durability of concrete structures in
the marine environment, extensive experience demonstrates that it is not the disintegration of the
concrete itself but rather an uncontrolled ingress of chlorides with subsequent corrosion of embedded
steel which represents the greatest challenge both to the durability design, execution and operation of
the structures.

Already in 1917, extensive field investigations of concrete structures in US waters demonstrated that
corrosion of embedded steel was a great problem affecting the durability of the structures [Wig and
Ferguson, 1917]. In 1924, Atwood and Johnson (1924) had assembled a list of approximately 3,000
references on durability of concrete in the marine environment. Later on, a large number of further
investigations were carried out in many countries, and a large number of papers, recommendations and
guidelines for durability of concrete structures in the marine environment produced. Still, however, a
poorly controlled durability of new major concrete infrastructure appears to be a big challenge for the
owners.

In recent years, existing concrete codes and standards, both for durability design and execution of
marine concrete structures, have been upgraded. Also, a number of new general recommendations and
guidelines for use of concrete in maritime engineering have been produced, the most recent and
comprehensive of which was published by CIRIA in 2010. Therefore, as a basis both for durability design
and production of new major concrete infrastructure, all minimum requirements in existing concrete
codes and standards as well as all general recommendations for good construction practice in existing
guidelines must be strictly followed.

However, compared to the rapid international research and development which have taken place in
recent years, existing concrete codes and practice both for durability design and execution of new
marine concrete structures have been slow to progress. Thus, it has taken more than 30 years for the
existing European Concrete Standards to reach the same strict durability specifications for concrete
structures in the marine environment as that specified for the first offshore concrete structures produced
for the oil and gas industry in the North Sea in the early 1970s [FIP, 1973]. Although the required service
life for the offshore concrete structures was much shorter than what was typically required for other
marine concrete structures, the international oil and gas industry was much more demanding in
obtaining more safe and regular operations for their installations. With more than 40 years of experience,
the durability of the offshore concrete structures in the North Sea has been much better than that of
other marine concrete structures produced during the same period. However, despite this good
performance, also the offshore concrete structures typically got a high scatter and variability of achieved
construction quality, and also some of these concrete structures got some very costly durability problems
which were the result of poorly achieved construction quality [FIP, 1996 ; Helland et al., 2010 ; Gjørv,
2014].

As soon as corrosion starts, the owners of marine concrete structures have a problem, which at the
outset only represents a cost problem but later on can develop into a more difficult safety and structural
problem. Internationally, deterioration of major concrete infrastructure has emerged as a severe and
demanding challenge both to the owners of the structures and to the professional society. Both public
and private owners are spending a significant and rapidly increasing proportion of their limited
construction budgets on repairs and maintenance of existing concrete infrastructure. A better controlled
and enhanced durability and service life of new major concrete infrastructure is not only important from
a cost point of view; it directly affects the sustainability of our society [Gjørv and Sakai, 2000].

A better controlled and enhanced durability of concrete structures in the marine environment primarily
depends upon preventing the initiation of corrosion of embedded steel, which is basically obtained by
ensuring the combination of high-quality, dense concrete and sufficiently quality concrete cover.
Additional protective measures can also be carried out to extend the service life of the structures, such
as sealers and coatings of the surface, corrosion resistant reinforcement and cathodic prevention
systems. In recent years, however, there has been a rapid international development of high-
performance concrete. By use of dense, high-performance concrete and sufficiently quality concrete
cover it should be possible to produce new marine concrete infrastructure with very high durability and

4
service life [Teng et al., 2014]. If the finely achieved concrete quality is dense enough to resist an
uncontrolled ingress of chlorides, current experience indicates that the concrete may also be good
enough to resist most of the other deterioration processes that threaten durability.

Also, some of the offshore concrete structures in the North Sea got some very costly durability problems
which were the result of high scatter and variability and poorly achieved construction quality. Both for
the offshore and other marine concrete structures, the durability specifications have typically been based
on prescriptive requirements for concrete composition and execution of concrete work, the results of
which are neither unique nor possible to verify for quality control during concrete construction. As a
result, much of the observed durability problems on concrete structures in the marine environment are
related to absence of proper quality assurance during concrete construction resulting in poorly achieved
construction quality.

In order to obtain a better controlled and enhanced durability of new major concrete infrastructure, rapid
international research and development has taken place in recent years. Much of this research has been
based on a probability approach to the durability design, which makes it easier to accommodate the
high scatter and variability of quality. Also, much of the research has been based on the development
of new performance-based durability requirements, which provide a better basis both for durability
specification and quality assurance during concrete construction. In Working Group 162 (WG), a review
of this new international knowledge and experience has been undertaken. As a result, some additional
recommendations and guidelines to existing concrete standards for durability and service life are
provided. This guidance is given with emphasis upon durability design and quality assurance as well as
condition assessment and preventive maintenance during the operational life of the structures. Again it
should be noted, however, that this guidance does not take the place of existing concrete standards for
the design and production of the structures.

1.2 TERMS OF REFERENCE

The current document is primarily aimed at providing guidance to owners and designers in order to
obtain a better controlled and enhanced durability and service life for new marine concrete infrastructure
beyond what is possible when based only on existing concrete codes and practice. The full text of the
‘Terms of Reference’ for the WG is shown in Appendix B.

1.3 TARGET READERSHIP

The target readership is primarily owners and designers of marine concrete infrastructure. Construction
engineers may also find some parts of the document of interest, although it was not the objective of this
WG to provide any guidance for concrete production or concrete construction which is amply covered
in other publications.

It was the aim of the WG that all recommendations and guidance would be complementary and
supplementary to relevant national standards and statutory regulations in force. In the event that
adopting one of the recommendations contained within this document would be a relaxation over the
relevant national standards and regulations, then it is the view of the WG that the national requirements
should take precedence.

1.4 OBJECTIVE

The objective of the WG has been to provide recommendations and guidance to owners and designers
of new marine concrete infrastructure worldwide in order to provide a safe, efficient and cost-effective
design and construction for such structures. However, the report should only be considered as guidance
in addition to existing concrete codes and practice for concrete durability and service life, with emphasis
upon:

 Durability design
 Quality assurance and achieved construction quality
 Condition assessment and preventive maintenance

5
For new concrete infrastructure, all minimum requirements in existing concrete codes and standards as
well as established recommendations and guidelines for good construction practice must be strictly
followed. However, existing concrete codes and practice both for durability design and execution of new
marine concrete structures have been very slow to progress compared to the rapid international
research and development on concrete durability in recent years. For new major concrete infrastructure,
therefore, more and more owners show an increasing interest to invest somewhat more at the outset of
their new projects in order to obtain a better controlled and enhanced durability of the structures beyond
what is possible when based only on existing concrete codes and practice. Even relatively small
additional costs have proved to be an extremely good investment compared to costly early repairs to
relatively new marine structures.

6
2. DURABILITY AND SERVICE LIFE
2.1 DETERIORATION PROCESSES

2.1.1 GENERAL

For concrete structures in the marine environment, many different deterioration processes can affect
the durability, a brief overview of these processes is given in the following.

2.1.2 CORROSION OF EMBEDDED STEEL

Chloride-Induced Corrosion

Normally, steel embedded in concrete is well protected against corrosion. This is because the concrete
is highly alkaline, and under such conditions, a strongly protective oxide layer forms on the steel surface
which prevents corrosion occurring. However, if the alkalinity of the concrete becomes reduced below a
certain critical level, the protective layer on the steel surface is broken and the corrosion process starts.
This is the case when the concrete has carbonated due to exposure to the atmosphere. Also, very small
amounts of chlorides in the concrete are capable of breaking the passivity of the steel, and the lower
the alkalinity of the concrete, the lower the amount of chlorides is required for breaking the passivity.

Although the chloride content in new concrete always is strictly limited by code specifications; chlorides
may later on penetrate the concrete cover from the external environment, and as soon as a sufficient
amount of chloride reaches the embedded steel, commonly known as the chloride threshold value,
corrosion starts. This chloride threshold which is statistical in nature is typically affected by the following:

 Composition of the cement and binder system

 Concrete quality and moisture content in the concrete
 Oxygen availability at the steel/concrete interface

When absolute values for the critical chloride content are given in current concrete codes and standards,
it should be noted that this is only based on empirical information on chloride contents giving a certain
risk for the start of corrosion. Such threshold values are therefore only based on previous experience
from older structures and generally assumed conditions for corrosion to occur. For new concrete
structures, based on new types of cement and binder system providing new levels of concrete quality,
however, it is difficult to estimate more accurate percentage values for the critical chloride content of
modern mixes.

It should also be noted that the chloride ions are not consumed in the corrosion process but just act as
a catalyst. Therefore, preventing further ingress of chlorides into the concrete once corrosion has started
does not stop further corrosion. As soon as the passivity of the steel is broken, however, the rate of
corrosion primarily depends on the oxygen availability at the steel surface and the ohmic resistance of
the concrete. Therefore, the rate of corrosion is very low for continuously submerged concrete due to
low oxygen availability [Gjørv et al., 1986a]. Thus, from the demolition of a concrete jetty in Oslo harbor,
very good overall condition of the embedded steel in the continuously submerged parts of the structure
was observed, even after more than 60 years of exposure [Gjørv and Kashino, 1986b]. Above water,
however, oxygen availability is not a limiting factor, and hence, corrosion can be observed after a short
period of time, where the rate of corrosion is primarily controlled by the ohmic resistance of the concrete
and the size of the anodic and cathodic areas in the very complex galvanic system developing along the
embedded rebar system.

During corrosion, expansive reaction products form, the result of which typically results in cracking and
spalling of the concrete cover (Figure 2.1). Also, the cross sectional area of the steel is gradually reduced
consequently reducing the load-bearing capacity of the structure (Figure 2.2). A type of corrosion known
as pitting corrosion in local parts of the steel represents a particularly difficult problem and challenge.

7
 Figure 2.1: Typical cracking and spalling in a corroding deck beam of a 29-year old concrete harbour structure
[Gjørv, 1968]

Figure 2.2: Reduced cross sectional area of the steel in the deck beam of a 25-year old concrete harbour
structure [Gjørv, 1968]

For prestressed concrete structures, both the high quality of the steel and the more highly stressed steel
make the tendons more sensitive to corrosion. For prestressed concrete, therefore, the chloride
threshold level is much lower. In addition, the tendons generally have a smaller cross-sectional area
which also makes them more susceptible to corrosion damage. Tendon fracture may be possible a long
time before any rust staining or corrosion-induced cracking becomes apparent at the surface, and as a
result, a sudden collapse may be possible. Although the members of the WG are not aware of any
collapse of prestressed marine concrete structures as a result of corrosion damage, there are instances
both in Europe and North America of highway bridges and multi-story car parks which have collapsed
as a result of chloride-induced corrosion, with no or little advanced warning.

8
Carbonation-Induced Corrosion

While both the alkalinity and the reserve basicity levels are very high in concrete based on pure portland
cements, other types of cement and binder systems may have much lower alkalinity and reserve
basicity. Therefore, concrete based on blast furnace slag cements, for example, having a much lower
alkalinity level, are more susceptible to carbonation compared to that of pure portland cements.
However, the rate of carbonation also very much depends on the density, porosity and permeability of
the concrete. Therefore, the more susceptible the binder system is to carbonation, the more important
is the density, porosity and permeability of the concrete. If the concrete is not dense enough, the
combination of chlorides and carbonation will typically accelerate the rate of chloride ingress. However,
for dense, high-performance concrete, e.g. water/binder ratio of ≤ 0.40, current experience
demonstrates that carbonation does not represent a significant practical durability problem, even for
concrete based on blast furnace slag cements with high slag contents.

High moisture contents in the concrete also contribute to reduced rates of carbonation. Carbonation
rates are typically the most rapid when the concrete has a relative humidity of about 65 %, and both
above and below this moisture level, the rate of carbonation is much lower. Thus, for concrete structures
in a moist, marine environment produced with dense, high-performance concrete, carbonation-induced
corrosion does not present a practical problem from a durability point of view.

Crack-Induced Corrosion

For structures with cracks in the concrete cover, the electrolytic conditions for start of corrosion may be
significantly affected. For cracked concrete, increased ingress and availability of chlorides, water and
oxygen will give an increased probability of local corrosion. Therefore, based on detailed procedures for
calculation of crack widths, existing concrete codes and recommendations specify upper limitations for
acceptable crack widths.

However, in spite of much field experience and a large number of investigations reported in the literature,
it appears that it is not possible to formulate a general relationship between crack width and probability
of corrosion for a given concrete structure in a given environment. Extensive research has revealed that
a number of factors affect the possibility for development of corrosion due to crack widths; both possible
mechanisms and the practical consequences are still the subject for much discussion. As shown in
Figure 2.3, the geometry of the cracks may be very complex, and different effects of the cracks running
parallel or perpendicular to the steel bars are also a factor. Cracks in different types of environment and
cracks due to different loading conditions such as static or dynamic loading, may therefore also affect
the corrosion process differently.

Figure 2.3: The geometry of cracks may be very complex [Beeby, 1977]

9
For most experimental investigations on the effect of cracks reported in the literature, it has not been
possible to establish a simple relationship between crack width and development of corrosion. Very
often, a certain effect of the cracks has been observed at an early stage of exposure, while later on, the
observed effect can be very small or almost negligible. For concrete structures in the marine
environment, it has typically been observed that the rate of corrosion in cracks has been reduced over
time by a clogging up of the cracks both by corrosion products and other densifying reaction products.

2.1.3 FREEZE-THAW RESISTANCE

Although the freeze-thaw resistance of concrete also has been the subject to extensive research for a
very long period of time, still this type of deteriorating process represents a great durability problem in
many countries. This is primarily due to the widespread use of de-icing salt on concrete pavements and
highway bridges. By comparing the field performance of concrete structures in various types of
environment, however, field experience clearly demonstrates that frost action in combination with de-
icing salt is far more aggressive to the concrete compared to that of freezing and thawing in a marine
environment [Petersson, 1995].

A general problem with both specification and assessment of the freeze-thaw resistance of concrete is
the lack of correlation between existing test methods and field performance. Also, different test methods
give different and conflicting test results. Another problem and challenge is the production of concrete
with a good and stable air void system during transportation, handling and placing of the fresh concrete.
This has been a problem for a long period of time, and extensive investigations of existing concrete
structures with intentionally entrained air have revealed that no air was present or what was there was
inadequate by current standards.

Also for the freeze-thaw resistance of concrete in the marine environment, extensive field experience
demonstrates that if only the applied concrete is dense enough, e.g. water/binder ratio of ≤ 0.40, it
appears that freeze-thaw deterioration does not represent so much of a durability problem as that of
concrete exposed to de-icing salt. Thus, for a large number of concrete harbour structures along the
Norwegian coastline, very good freeze-thaw resistance of dense, non-air entrained concrete in the tidal
zone even after exposure periods of up to about 50 years has been experienced [Gjørv, 1968 ; Gjørv,
2014]. Also after 25 years of severe exposure in the tidal zone on the East-Coast of Canada, even
concrete based on blast-furnace slag cement have shown a very good freeze/thaw resistance at a
water/binder ratio of 0.40 [Thomas et al., 2011].

2.1.4 ALKALI-SILICA REACTION

Alkali-silica reaction (ASR) occurs when certain types of concrete aggregate are not chemically stable
in the highly alkaline concrete [Gadea et al., 2010]. As a result, severe cracking and crumbling of the
concrete after some time may occur. During this reaction, an expanding gel is formed, the amount of
which depends on the alkalinity level of the concrete and the availability of moisture. Depending on type
of aggregate, deterioration may occur after 1-6 years, but occasionally for certain types of aggregate,
even after up to 20-30 years.

For concrete structures in the marine environment, the availability of moisture is generally very high.
Therefore, the most efficient way of avoiding problems due to ASR, is to use types of aggregate which
are not susceptible to ASR. Although rapid development of methods for testing the susceptibly of
aggregate to ASR have taken place in recent years, for certain types of aggregate, such testing can still
be inconclusive. In order to have more confidence in the aggregate, selection long-term records of
previous use of the particular type of aggregate should be obtained.

Most problems due to ASR are observed in concrete based on pure portland cements which typically
have a high level of alkalinity. Another approach of controlling ASR, therefore, is to use cements or
binder systems with lower alkalinity levels. However, ASR is not specific to the marine environment and
most recommendations and performance specifications for concrete production have special guidelines
for avoidance of problems due to ASR.

10
2.1.5 CHEMICAL SEAWATER ATTACK

Generally, concrete is susceptible to chemical attack by seawater; in particular, if the concrete is not
sufficiently dense. Thus, in areas where problems with placing of the concrete under water have
occurred during concrete construction resulting in a very porous concrete, it has later been observed
that the concrete matrix has been completely broken down even after a relatively short period of time
[Gjørv, 1968]; the weakened concrete could be characterised by up to 70 % loss of lime due to heavy
leaching and increased contents of magnesium oxide by up to 10 times that of the original content.

It is well documented that pure portland cements are much more vulnerable to chemical seawater attack
than most other types of binder system such as blast furnace slag cements or portland cements blended
with various pozzolanic materials. However, for chemical seawater attack as well as for most other
deterioration processes, the density and permeability of the concrete are the key for controlling the rate
of deterioration. The more vulnerable the binder system, the more important is the density and
permeability of the concrete to resist deterioration.

2.1.6 DELAYED ETTRINGITE FORMATION

Ettringite is one of the minerals formed when portland cements react with water during hydration [BRE,
2001]. Ettringite occupies a larger volume than its original constituents, but as this normally forms prior
to the full setting of the concrete, the expansion is normally accommodated without any problems.
However, if the concrete is subjected to elevated temperatures during hydration, the formation of
ettringite can be delayed, hence the name ‘Delayed Ettringite Formation’ (DEF). If such a delayed
ettringite formation and expansion take place after the concrete has hardened, cracking may occur
provided sufficient moisture is available. Also this deteriorating process may take a long time; anything
from 2 to 20 years.

Availability of water is also important for this deterioration process. Since concrete structures in the
marine environment generally have high moisture contents, the most efficient way of avoiding problems
due to DEF is to limit the peak temperature in the concrete during concrete construction. Slightly different
guidance is provided in the literature, but even if sufficient moisture is available, there appears to be no
risk of DEF if the peak temperature of the concrete is kept below 60oC and very little risk if the
temperature is kept below 70oC. Higher peak temperatures can be tolerated for a short period of time,
but since the risk of DEF then increases, other types of binder systems than pure portland cements
should then be considered. Once the temperature exceeds 85oC, however, the risk of DEF is significant
irrespective of binder system.

Also for this deterioration process, DEF is not specific to the marine environment. Field experience
demonstrates that problems due to DEF are primarily a potential problem in the precast concrete
element industry, where elevated curing temperatures may be applied for accelerated production of
prefabricated concrete elements. However, If the peak temperature in the concrete during construction
exceeds 70oC, early-age cracking due to high thermal gradients may present an even more significant
problem.

2.1.7 EARLY-AGE CRACKING

Although early-age cracking is not a deterioration process in itself, such cracks can play an important
role in accelerating/exacerbating any of the above deteriorating processes [CIRIA, 2007]. Early-age
cracking may be due to either high thermal gradients or plastic shrinkage of the freshly placed concrete.

The role of cracking in the onset of a deterioration process in general and corrosion of embedded steel
in particular, is the subject of some debate. However, it is not practical to be able to produce concrete
structure entirely free of cracks, but the more frequent and the larger the cracks, the easier it will be for
seawater to penetrate and migrate into the concrete.

Early-age thermal cracking is the result of high temperature differentials both between the interior and
the exterior of the concrete and between the concrete and the adjacent surface against which it is cast.
Although a proper control of early-age cracking due to thermal gradients should be part of the structural
design, most recommendations for concrete production also include guidelines for avoiding early-age
cracking both due to the thermal gradients and plastic shrinkage. Since high-performance concrete with

11
low water/binder ratios are also more susceptible to plastic shrinkage, the potential cracking of such
concretes needs special attention in the mix design and trials.

2.1.8 ABRASION

In marine concrete structures, erosion of concrete by abrasion or cavitation caused by high velocity
water flow either from boat propellers or other sources may occasionally take place. Abrasive wear can
also be caused by heavy particles or ice scouring against the concrete surface. Although abrasion does
not appear to be a general problem for high-performance concrete with a high surface hardness, the
effect of abrasion such as heavy ice scouring or impacting waves carrying gravel may require some
special mix design and trials.

2.2 CODES AND PRACTICE

Although extensive field experience demonstrates that chloride-induced corrosion is one of the most
difficult processes to control and hence, represents a big challenge, all the above deterioration
processes represent a potential threat to the long-term durability. Therefore, as a basis for the durability
design and production of new major concrete infrastructure, all minimum requirements in existing
concrete codes and standards as well as adherence to established recommendations and guidelines
for good construction practice must be strictly followed. Both for the design, specification and execution
of concrete work, current concrete codes and standards give detailed regulations and requirements. In
order to demonstrate the level of detailing, the procedures for achieving correct concrete covers
according to the current European Concrete Standards are shown in Appendix C.

According to current concrete codes and practice, however, durability specifications are still primarily
based on prescriptive requirements to composition of the concrete mixture such as upper levels for
water/binder ratio and minimum levels for binder content. Although a low water/binder ratio also reflects
a high density and low permeability of the concrete providing both a high resistance to chloride ingress
and good durability, extensive investigations demonstrate that selecting a proper binder system may be
much more important for obtaining a high resistance to chloride ingress. For example, when the
water/binder ratio is reduced from 0.50 to 0.40 for a concrete based on pure portland cement, the
chloride diffusivity of the concrete, which reflects the resistance of the concrete to chloride ingress, is
reduced by a factor of two to three. However, by the incorporation of various types of supplementary
cementitious materials such as blast-furnace slag, fly ash or silica fume at the same water/binder ratio
reduced the chloride diffusivity by a factor of up to 20 [Thomas et al., 2011]. Also, while a reduced
water/binder ratio from 0.45 to 0.35 for a concrete based on pure portland cement may only reduce the
chloride diffusivity by a factor of two, a replacement of the portland cement by a proper blast-furnace
slag cement may reduce the chloride diffusivity by a factor of up to 50 [Bijen, 1998]. By also increasing
the fineness of the supplementary material and combining various supplementary materials, even for
the same water/binder ratio, very low chloride diffusivities can be achieved and hence very high
resistance to chloride ingress obtained [Teng et al., 2014]. If the concrete becomes dense enough to
resist the uncontrolled ingress of chlorides, current experience indicates that the concrete will also be
able to resist most of the other deterioration processes that threaten the durability.

For many years, when concrete was mostly based on pure portland cements and simple procedures for
concrete production, the concept of water/cement ratio was the fundamental basis both for
characterising and specifying concrete quality. Since a number of new cementitious materials and
reactive fillers are increasingly being applied in concrete production, the concrete properties are
increasingly being controlled by the various combinations of such materials. In addition, the concrete
properties are also more frequently being controlled by use of various types of processed concrete
aggregate, new concrete admixtures and sophisticated production equipment. As a result, the old and
very simple terms ‘water/cement ratio’ or ‘water/binder ratio’ for characterising and specifying concrete
quality have successively lost their meaning. As a consequence, there is need for performance-based
definitions and specifications for concrete quality; in particular, this is true for characterising and
specifying concrete durability [Bjegović et al., 2014].

According to most existing concrete codes and standards, the term ‘Service life’ is also typically defined
as the ‘The time until first major maintenance’, which is a vague concept for all parties involved. If the
‘First major maintenance’ is caused by corrosion of embedded steel, the owners of the structures already
have a serious maintenance problem. As soon as steel corrosion starts, extensive experience

12
demonstrates that it is both technically difficult and very costly to get the further corrosion process under
control (Chapter 6). More significantly the corrosion process may develop into a more difficult safety
problem (Figure 2.4). As a basis for the durability design of new marine concrete infrastructure, efforts
should therefore be made to obtain a best possible control of the chloride ingress during the initiation
period before any corrosion starts. It is in this early stage that it is both technically easier and much
cheaper to take precautions by selecting proper protective measures for the control of further
deterioration processes.

Figure 2.4: Deterioration of a concrete structure due to steel corrosion [Tuutti, 1982]

2.3 QUALITY ASSURANCE AND ACHIEVED CONSTRUCTION QUALITY

It is well established that many durability problems for concrete structures in the marine environment
can be ascribed to lack of proper quality assurance during concrete construction and poorly achieved
construction quality. Even before the concrete is placed in the formwork, the quality of the concrete may
show a high degree of scatter and variability. Depending upon a number of factors during concrete
construction, the achieved quality of the finally placed concrete may show an even higher scatter and
variability of quality (Chapter 5).

One of the most common quality problems in concrete construction, however, is the failure to meet the
specified cover thickness to embedded steel. Although, the specified concrete cover is normally carefully
checked prior to placing of the concrete, significant deviations can occur during concrete construction.
The loads imposed during concrete placement may cause movement of the reinforcement in the
formwork, or the chairs may have been insufficiently or have wrongly been placed.

In order to accommodate the potential high scatter and variability of quality, a probability approach to
the durability design as briefly outlined and discussed in Chapter 3 can be applied. However, since none
of the models for such design take into account the effect of cracks or other defects also typically
occurring during concrete construction, some additional strategies and protective measures can be
considered (Chapter 4).

With the probability approach to the durability design described in Chapter 3, however, performance-
based requirements both to concrete quality and concrete cover are established which later on provide
a basis for quality assurance and documentation of achieved construction quality (Chapter 5). A
performance-based durability design without any probability calculations can also be applied as that
described in Chapter 3.

Documentation of achieved construction quality and compliance with the durability specification should
be very important for the owner since such documentation may have implications both for the obtained

13
durability and expected service life of the structure. Experience from recent years has shown that where
such documentation has been required, it has typically clarified the responsibility of the contractor for
the quality of the construction process. As a result of this clear distinction, achieved construction quality
with a reduced scatter and variability has typically been observed [Gjørv, 2010, 2014, 2015].

2.4 CONDITION ASSESSMENT AND PREVENTIVE MAINTENANCE

Despite the best compliance to the achieved construction quality both from specification and
construction quality assurance, extensive experience demonstrates that all concrete structures in the
marine environment will have a certain rate of chloride ingress during the operational life of the
structures.

Therefore, upon the completion of a concrete structure, a service manual for regular monitoring and
control of the real chloride ingress taking place during the operation of the structure should be required
(Chapter 6). It is such a service manual that helps provide the ultimate basis for obtaining a controlled
and enhanced service life of the concrete structure in its environment.

2.5 LIFE CYCLE COSTING (LCC)

With a continuous pressure to confined construction budgets, most new projects are typically based
upon minimum durability requirements according to the existing concrete codes and standards. This
provides the lowest initial costs. Since any additional measure for a better controlled and enhanced
durability will cost somewhat more, this current practice has been detrimental to obtaining a more
durable concrete infrastructure. For new marine concrete infrastructure where high durability and service
life are of special importance, calculations of lowest costs over the whole service life of the structure
(Life Cycle Costing or LCC) should be adopted as the owner’s ultimate basis for the project.

Procedures for LCC (Life Cycle Costing) are well established in exiting documents and more information
about LCC is contained within the PIANC Report WG 31 [PIANC, 1998]. In essence, however, LCC
involves the summing up of all costs associated with the asset over its total service life; costs being
incurred in the future and discounted back to the net present value (NPV). This also takes into account
the cost of borrowing money and anticipated inflation. By carrying out this process for different
alternatives of design, construction and maintenance regimes, it should be possible to identify technical
solutions that provide the lowest costs over the whole service life of the asset. Additional costs for
shutting down of the operation of the asset may also be included in such calculations. LCC can then be
carried out at different stages of an asset’s life, including:

 The asset design stage in order to identify, amongst other things, what specific initial durability
provisions/maintenance scenario is likely to provide the optimum solution.
 The construction design stage in order to evaluate what is the most appropriate form and method
of construction.
 The service life of the asset to define a maintenance approach to be adopted throughout the
intended life of the asset.

LCC has been shown to be a very valuable tool for assessing the ‘cost-effectiveness’ of various technical
solutions in order to obtain an optimal durability design and construction of new marine concrete
infrastructure [Pruckner and Gjørv, 2004]. It may also be a valuable tool for assessment of various
technical solutions for condition assessment, maintenance and repair strategies during the operation of
the structures. Although calculations of costs against benefits are typically based on well established
procedures for LCC, such calculations can be carried out in specific ways by considering various types
of cost or benefit that may vary from one project approach to another.

2.6 LIFE CYCLE ASSESSMENT (LCA)

Internationally, an uncontrolled and premature deterioration of existing concrete infrastructure has

emerged as one of the most severe and demanding challenges facing the construction industry
[Horrigmoe, 2000]. Public agencies are spending significant and rapidly increasing proportions of their
construction budgets for repairs and maintenance of the structures. Repair projects will undoubtedly be
the subject to increasing economic constraints, so there will be a parallel increase in the consideration

14
of durability during the design and construction phases for new concrete infrastructure. Enhanced
durability and service life of new major concrete infrastructure are not only important from an economical
point of view; it directly affects the sustainability of our society [Gjørv and Sakai, 2000 ; Gjørv, 2014].

These and other factors have spurred the rapid development of Life Cycle Assessment (LCA) of new
concrete infrastructure. The framework and the methodology for quantifying the ecological effects and
impacts from design, production, and maintenance of the structures are available through current
standards such as ISO 14040 [ISO, 2006a] and ISO 14044 [ISO, 2006b]. As these standards show,
LCA includes assessment of consumption of materials and energy, generation of waste, and emission
of pollutants as well as the accompanying environmental and health risks. Therefore, LCA provides a
valuable tool both for quantifying and comparing the effects of various technical solutions for
improvements in the design, construction, and operation of new concrete infrastructure.

15
3. DURABILITY DESIGN
3.1 PROBABILITY-BASED APPROACH

3.1.1 GENERAL

In order to take the high scatter and variability of achieved construction quality into account, there has
been a rapid international development on probability-based durability design of new major concrete
infrastructure [Siemes et al., 1985 ; Siemes and Rostam, 1996 ; Stewart and Rosowsky, 1998 ; McGee,
1999 ; Gehlen and Schiessl, 1999 ; Siemes et al., 2000 ; DuraCrete, 2000a ; Gehlen, 2007].

As part of the design criteria for probability-based design, it should be noted that “A level of reliability for
not passing the limit state during the design service life” should be agreed upon between all parties
involved [ISO 2394, EN 1990, ISO 16204]. Also, the appropriate level of reliability shall reflect all possible
consequences by passing the ‘Limit state’. Even if ISO 16204 generally recommends 10 % as a limit
state level, also other levels can be selected; this should be agreed upon early in the procurement
process. Special conditions such as difficult inspections, critical parts of the structure etc. may also
influence the selection of an appropriate limit state level for a given structure.

Although different stages of a deteriorating process also can be chosen as a basis for the limit state
level, the onset of steel corrosion which is the most critical stage for deterioration of marine concrete
structures is typically being chosen as a basis for the durability design.

In 2000, the first international guidelines for probability-based durability design were introduced as a
result of the European research programme DuraCrete [DuraCrete, 2000b], and during the
following years, this design was applied to a number of new major concrete structures in many countries
[Gjørv, 2014]. Also in Norway such durability design was initially based on the DuraCrete guidelines, but
during the following years when practical experience using these guidelines for new marine concrete
infrastructure in Norwegian harbours was gained, they were further developed with more practical
implementation by the Norwegian Association for Harbor Engineers [NAHE, 2004]. Lessons learnt from
further practical experience with these guidelines to new commercial projects were incorporated into
subsequent revised editions and denoted the DURACON Model (Appendix D) [PIANC Norway/NAHE,
2009]. At the same time, this model was expanded to provide the basis for concrete quality control and
quality assurance, which upon completion of the concrete construction provides documentation of
achieved construction quality and compliance with the durability specification. This model also provides
a basis for control of the future chloride ingress as part of the regular condition assessment and
preventive maintenance during operation of the concrete structures (Figure 3.1).

Figure 3.1: The DURACON Model as a basis for durability design, quality assurance and operation of new major
concrete infrastructure in marine environments [PIANC Norway/NAHE, 2009].

In recent years, a number of new models and software for durability design have been introduced
internationally, a thorough review of which is given in the CIRIA Report [CIRIA, 2010]. Since the
commercial software based upon both the STADIUM® Model and the Life-365 Model have been widely
adopted in the USA in recent years, these two models are described in more detail in the Appendixes E

16
and F, respectively. It should be noted, however, that this information is only provided as ‘state-of-the-
art’ and does not mean any endorsement form PIANC. In order to demonstrate how such models can
be applied as a basis for the durability design of a new marine concrete structure, the DURACON Model
is briefly illustrated by the following.

3.1.2 DURABILITY ANALYSIS

In the DURACON Model, the overall durability requirement is based upon the specification of a given
‘Service period’ before the probability for the onset of steel corrosion exceeds an upper level of 10 %,
which is in accordance with current standards for reliability of structures. In order to calculate the
probability of corrosion, durability analyses are carried out which provide the basis for selecting proper
combinations of concrete quality and concrete cover which would meet the specified durability or
‘Service period’ for the given concrete structure in its given environment.

In principle, the probability of corrosion can be calculated by use of several mathematical methods and
available software. Based upon current experience for durability design of concrete structures in
chloride-containing environments, however, a simple combination of a modified Fick’s Second Law of
Diffusion and a Monte-Carlo Simulation was adopted as a basis for the calculation [Ferreira, 2004 ;
Ferreira et al., 2004]. Although such a combined calculation can also be carried out in different ways,
special software (DURACON) for this calculation was developed, for which proper information about the
following input parameters is needed:

 Environmental loading
- Chloride loading (CS)
- Age at chloride loading (t’)
- Temperature (T)
 Concrete quality
- Chloride diffusivity (D)
- Time dependence factor (α)
- Critical chloride content (CCR)
 Concrete cover (XC)

All the procedures and methods for determining and selecting the above input parameters are described
and discussed in more detail in Appendix D, where the link to a free downloading of the DURACON
Software is also given. It should be noted, however, that in the DURACON Model, the concrete quality
is characterised by the chloride diffusivity (D) according to the Rapid Chloride Migration (RCM) method
(AASHTO, 2003). Since the RCM method does not require any pre-curing of the concrete it can be
carried out very rapidly, independent of concrete age. Because the method provides a very strongly
accelerated test, however, this chloride diffusivity should only be considered as a simple relative index.
Although this index does not reflect the long-term capacity of the concrete for chloride binding, it reflects
both the porosity and permeability of the concrete as well as the ion mobility in the pore solution of the
concrete. Hence, it reflects both the concrete mixture’s resistance to chloride ingress as well as its
general durability properties. Using the 28-day chloride diffusivity (D28) as an input parameter to the
above durability design can be compared to using the 28-day compressive strength (f28) as an input
parameter to the structural design; both parameters are actually relatively simple indices that can be
used to establish that a concrete mixture is fit for purpose. However, it should be noted that the 28-day
chloride diffusivity (D28) is a much more sensitive parameter for detecting variations in concrete quality
compared to that of the 28-day compressive strength (f28).

For new major concrete infrastructure where high durability and service life are of special importance,
at least a 100-year ‘Service period’ should be specified before the probability of corrosion exceeds 10
%. For increased ‘Service periods’ of more than 100 years, however, the calculation of corrosion
probability gradually becomes less reliable. For ‘Service periods’ of up to 150 years, therefore, the
corrosion probability should be kept as low as possible and not exceeding 10 %, but in addition, some
further protective measures such as a partial replacement of the black steel with stainless steel or non-
corrosive reinforcement can be specified. For ‘Service periods’ of more than 150 years, however, any
calculation of corrosion probability is no longer considered valid. For such long ‘Service periods’, the
corrosion probability should still be kept as low as possible and not exceeding 10 % for a 150-year
‘Service period’, but in addition, one or more additional protective measures such as partial use of

17
stainless steel or non-corrosive reinforcement should be specified. If the risk for early-age chloride
exposure during concrete construction before the concrete has gained sufficient maturity and density is
also high, special precautions or protective measures may also need special attention (Chapter 4).

As a result of the above method of durability design, requirements for both concrete quality (D28) and
concrete cover (XC) are established parameters, and this provides the basis for the regular quality control
and quality assurance during concrete construction (Chapter 5). Upon completion of the concrete
construction, a new durability analysis is then required with the obtained average values and standard
deviations of both the 28-day chloride diffusivity and the concrete cover as new inputs. For this durability
analysis, all the other originally assumed input parameters which may have been somewhat difficult to
select during durability design, are now kept the same. Therefore, this documentation primarily reflects
the results obtained from the regular quality control of the 28-day chloride diffusivity (D28) and the
concrete cover (XC) during concrete construction, including the scatter and variability of quality
observed. Using this method, documentation of achieved construction quality and compliance with the
durability specification is obtained (Chapter 5).

During the operation of concrete structures, further calculations of corrosion probability should be carried
out as a basis for regular condition assessments and preventive maintenance of the structures. For this
monitoring method, the calculations are based on the apparent chloride diffusivities (Da) obtained from
the observed rates of chloride ingress in combination with site data on the concrete cover (XC) (Chapter
6). In this way, before this probability of corrosion becomes too high, appropriate additional protective
measure should be implemented (Chapter 4).

In order to further demonstrate how the above calculations of corrosion probability can be applied as a
basis for durability design of a new marine concrete structure, a case study is shown.

3.1.3 CASE STUDY

General

The overall durability specification for the new marine concrete structure was a 120-year ‘Service period’
before 10 % probability of corrosion would be reached. The given marine environment was characterised
by a quite severe chloride loading of 5.5 % by wt. of cement, typically observed on similar concrete
structures in similar environments, while the annual temperature was typically 200C. In order to select a
proper combination of concrete quality and concrete cover which would meet the above durability
specification of a 120-year ‘Service period’, two steps of durability analyses were carried out, the first of
which was to evaluate the effect of various concrete qualities, while the other was to evaluate the effect
of various concrete covers on the probability for the onset of corrosion.

Effect of Concrete Quality

In order to evaluate the effect of concrete quality, four different concrete qualities were compared for
which the 28-day chloride diffusivities (D28) were determined (RCM). Apart from type of binder system,
all of these concrete mixes were identical and fulfilled the minimum durability requirements according to
the current concrete standards for a 100-year service life, including a water/binder ratio ≤ 0.40 and a
binder content ≥ 330 kg/m3. The various binder systems included four different types of commercial
cements in combination with 10 % silica fume (CSF) by weight of cement; one high-performance
portland cement (Type 1), one fly ash cement with 20 % fly ash (Type 2) and two blast-furnace slag
cements with 34 and 70 % slag (Types 3 and 4), respectively (Table 3.1).

18
 Input parameter
Concrete quality CCR
D28
α (% by wt. of
(m2/s x 10-12)
binder)
Type 1 (CEM I 52.5 LA + 10 % CSF) N1)( 6.0; 0.64) N(0.40; 0.08)
Type 2 (CEM II/A – V 42.5 R + 10 % CSF) N(7.0; 1.09) N(0.60; 0.12)
N(0.4; 0.10)
Type 3 (CEM II/B – S 42.5 R NA + 10 % CSF) N(1.9; 0.08) N(0.5; 0.10)
Type 4 (CEM III/B 42.5 LH HS + 10 % CSF) N(1.8; 0.15)
1)
Normal distribution with average value and standard deviation

Table 3.1: Input parameters for analysing the effect of concrete quality

Based on the obtained 28-day chloride diffusivities (D28), a nominal concrete cover of 70 mm (XC) with
a standard deviation of 6 mm and estimated values both for the time dependence of the chloride
diffusivities (α) and the critical chloride content (CCR) as shown in Table 3.1, durability analyses were
carried out. The other input parameters for the calculations, estimated average values with standard
deviation both for chloride loading CS (5.5; 1.4 %), age at chloride loading t’ (28 days) and temperature
T (200C) were kept constant. For all input parameters, a normal distribution was assumed.

Although all four concrete mixes complied with the current standard requirements for a 100-year service
life, the ‘Service period’ before 10 % probability of corrosion was reached differed significantly
(Figure 3.2). These ranged from about 15 years for the portland cement type of concrete (Type 1) to
about 30 years for the fly ash cement concrete (Type 2) and to more than 120 years for the two slag
cement types of concrete (Types 3 and 4). Thus, only the two types of concrete based on slag cements
would meet the specified durability requirement for the given structure.

100
90 Type 1: Portland cement
Type 2: Fly ash cement (20%) Type 1
80
Probability of corrosion (%)

Type 3: Slag cement (34%)

70
Type 4: Slag cement (70%)
60
50
40
Type 2
30
20
Type 3 and 4
10
0
0 20 40 60 80 100 120
Service period (years)

Figure 3.2: Effect of cement type on the ‘Service period’.

It may be argued that the above durability analyses were only carried out on the basis of the obtained
28-day chloride diffusivities of the concrete, while the further reduction of chloride diffusivity for the
various types of concrete based on the various types of binder system would develop very differently.
Therefore, some additional durability analyses based on the chloride diffusivities obtained after longer

19
curing periods of up to 90 days were also carried out, but none significantly changed the relative basis
for comparing and selecting the concrete with the best durability properties with respect to the chloride
diffusivity.

Effect of Concrete Cover

In order to further evaluate the effect of increased concrete cover (XC) beyond the nominal cover of 70
mm, some additional durability analyses based on 90 and 120 mm concrete cover were also carried out
(Table 3.2). These analyses were only based on the above type concrete with portland cement (Type
1), while holding all the other input parameters constant. Figure 3.3 shows how an increased concrete
cover would also significantly affect the probability of corrosion. While a nominal cover of 70 mm for the
portland cement concrete would give a ‘Service period’ of about 15 years, 90 and 120 mm covers would
increase the ‘Service period’ up to about 30 and 70 years, respectively.

Average Standard
Input parameter Comments
value deviation
D28 6.0 0. 64 Chloride diffusivity (m2/s x 10-12)
α 0.40 0.08 Time dependence factor
Critical chloride content
CCR 0.40 0.10
(% by wt. of binder)
70 6
XC 90 6 Nominal concrete cover (mm)
120 6

Table 3.2: Input parameters for analysing the effect of concrete cover

100
90
Probability of corrosion (%)

80
70
70 mm
60
50
90 mm
40
30
120 mm
20
10
0
0 20 40 60 80 100 120
Service period (years)

Figure 3.3: Effect of concrete cover on the ‘Service period’ (Type 1 Concrete)

As can be seen from Figure 3.3, even a concrete cover of 120 mm for the Type 1 Concrete would not
be sufficient for meeting the overall durability requirement of a 120-year ‘Service period’ before 10 %
corrosion probability would be reached. It may be argued that increased cover thickness beyond 90 mm
would increase the risk for unacceptable crack widths. While this effect to some extent could be mitigated
by incorporating synthetic fibres in the concrete, increased cover would also have some secondary
effects such as increased total dead load. Therefore, an alternate solution would be to select one of the
other types of concrete with lower chloride diffusivities. Also, a replacement of the black steel with

20
stainless steel or non-corrosive reinforcement for the outer layer of the rebar system could also be a
viable alternative. An outer layer of less corrosive reinforcement would effectively increase the cover
thickness to the remaining black steel further into the section. In this way, the above durability analyses
can also be used as a design tool for quantifying how much of the traditional carbon steel needs to be
replaced by non-corrosive reinforcement in order to meet the required safety level against the onset of
corrosion. For important concrete infrastructure, current experience demonstrates that a partial use of
stainless steel in the most exposed and critical parts of the structure has proved to be a very simple and
robust technical solution, which in the long term has also proved to be an extremely good investment
[Gjørv, 2014].

Evaluation of Results

It should be noted that the above calculations of corrosion probability (durability analyses) are only
based on a very simplified diffusion model for a one-dimensional ingress of chlorides and a number of
very uncertain input parameters. In the field, also other transport mechanisms are controlling the rate of
chloride ingress. Therefore, the DURACON Model should not be used for any prediction of service life,
but only used as a simple engineering tool for quantifying the combined effects of various concrete
qualities and concrete covers, which provide a basis both for specifying and meeting a required durability
for the given concrete structure in the given environment. Nor should the specified ‘Service periods’ be
considered as the real-time until the onset of corrosion. The above ‘Service periods’ are only the result
of a simple assessment and judgement of the most important parameters related to the durability,
including the scatter and variability of quality involved. Thus, for the above concrete structure in this
example, both Concrete Types 3 and 4 based on blast-furnace slag cements in combination with a
nominal concrete cover of 70 mm would be the preferred selections.

For a number of new commercial projects where the above procedures for probability-based durability
design have been applied in recent years, it has been possible to reach optimal combinations of concrete
quality (chloride diffusivity) and concrete cover for the given concrete structures in the given
environments to obtain the required ‘Service periods’ [Gjørv, 2014]. As a result, concrete quality
assurance during concrete construction with documentation of achieved construction quality and
compliance with the durability specification has also been provided (Chapter 5).

3.2 PERFORMANCE-BASED APPROACH

3.2.1 3.2.1 GENERAL

A performance-based approach to the durability design can also be applied without any calculations of
corrosion probability. A performance-based approach also has a formula to describe and evaluate the
concrete for a given service period in a given environment, which constitutes a toolbox both for owners,
designers and engineers involved in the construction and management of new major concrete
infrastructure. In 2007, Japan Society of Civil Engineers (JSCE) published the guideline ‘Standard
Specifications for Concrete Structures – Design’, where all steps of a performance-based approach are
described and explained together with how it is implemented [JSCE, 2007]. Also, the CIRIA Report gives
detailed recommendations for a performance-based approach to the durability design [CIRIA, 2010].

The performance-based approach provides a basis for signing contracts which ensure a proper
durability, the basis for which represents a major change in the evaluation and specification of concrete
quality. The performance-based approach differs from the conventional prescriptive approach by setting
requirements for the durability properties rather than to the composition of the concrete mix and is well
suited for establishing the requirements for concrete durability for marine environments.

The performance-based approach is based on established knowledge and research results validated
experimentally. This approach has also led to a better understanding of the mechanisms of concrete
deterioration, particularly the process of chloride-induced corrosion. This approach also provides tools
for characterisation and monitoring the aging of the concrete which have led to development of a
methodological approach for the control of concrete durability. The requirements are determined in
terms of performance such as resistance of chlorides ingress rather than in terms of resources such as
minimum cement content. The approach also provides the tools for ensuring that the specified durability
requirements are achieved. However, the approach does not allow ignoring of established general
recommendations and good practice for the use of concrete in maritime engineering.

21
The performance-based approach provides greater freedom in the formulation of concrete mixes and
favours innovation and more flexibility in the selection for structures that require more high-performance
concrete qualities.
The performance-based approach requires:

 Define, prescribe, formulate and optimise the concrete for a given service life in a specific
environment.
 Evaluate the performance of a concrete.
 Act as an incentive for innovation and use of new materials by ensuring that the durability
requirements are met.

The performance-based approach is based on the concept of durability indicators associated with
performance-based criteria (thresholds) to be observed. Durability criteria are determined from tests of
physic-chemical characterisations. The performance-based criteria are normally calibrated from aging
models and experiments. Depending on the concrete environment (exposure classes, risk of potential
damage) and service life, the approach is based on selections of appropriate indicators of durability and
associated thresholds.

In general, the durability indicators provide a basis for evaluating the ability of a type of concrete to resist
a given degradation. If the objective is to prevent ingress of chlorides, appropriate durability indicators
will therefore be selected in order to give the concrete sufficient resistance to chloride ingress.

The performance-based approach is typically divided into the following six stages:

1. Choice of design life

2. Environmental conditions and main risks of damage
3. Selection of durability indicators and associated specifications
4. Design of the concrete mix
5. Testing before construction and for diffusivity control
6. Initial durability point and start of structure monitoring

3.2.2 CHOICE OF DESIGN LIFE

This is the period during which the structure is expected to remain in operation by being normally
maintained, but without the need for any major repairs. For important marine concrete infrastructure, the
project design life is mostly 100 years, but the owner may also require a longer service life.

3.2.3 ENVIRONMENTAL CONDITIONS AND MAIN RISKS OF DAMAGE

The exposure conditions of the structure must be analysed and the main degradation risks identified.
Then for each part of the structure, the exposure classes are defined.

In France, in accordance with EN 1992 and EN 206, the performance-based approach to marine
concrete structures is typically based on the following exposure classes (Appendix C):

 Buried parts of supports, foundations and retaining walls: XC2, XS1

 Parts constantly immersed of supports, foundations and retaining walls: XC1, XS2
 Parts in tidal zone or exposed to salt spray: XC4, XS3

In the case where the infrastructure is partly or fully exposed to chlorides, exposure class XD3 is also
specified.

3.2.4 SELECTION OF DURABILITY INDICATORS AND ASSOCIATED

SPECIFICATIONS

When the performance-based approach is applied, the specifications of concrete composition related to
exposure classes are substituted by performance specifications based on durability indicators.

22
For control of reinforcement corrosion, the following durability indicators may be chosen:

 Porosity accessible to water

 Chloride diffusion coefficients
 Apparent gas permeability
 Electric resistivity

In general, the number of indicators and the thresholds to be specified are related to the environmental
aggressiveness and the specified service life. The specification of stricter thresholds for the indicators
allows for a modification of the structural class, allowing a reduction in the concrete cover required. For
the same service life, this method is used to reduce the thickness of the concrete cover which will
improve the quality of the designed and specified concrete mix for the given environment.

3.2.5 DESIGN OF THE CONCRETE MIX

After the previous step, the concrete supplier completes the requirements to achieve the durability
indicators and any other additional requirements for concrete workability, transport time, setting time,
etc.

The testing of the durability indicators specified in Stage 3 is normally completed during a period of 90
days. However, both for the porosity accessible to water and the electrical resistivity which are used for
control, the testing is normally carried out at 28 days.

3.2.6 TESTING BEFORE CONSTRUCTION AND FOR DIFFUSIVITY CONTROL

Testing before construction in addition to standard mix design tests taking samples from the designed
concrete mix to determine the porosity of the concrete accessible to water and the resistivity at 28 days.

During construction, tests for concrete quality include compressive strength (strength class),
consistency of fresh concrete (class of slump or spread) and the regularity of the concrete composition.
The base compliance control is then completed by measurement of the durability indicators as specified
in the contract.

In addition, concrete cover measurements are carried out by use of covermeter on those parts of the
structure which are considered representative.

If significant variations are found (consistency of fresh concrete, compressive strength, etc.), the client
or his representatives may request additional measurements of durability indicators.

3.2.7 INITIAL DURABILITY POINT AND START OF STRUCTURE MONITORING

Initial Durability Point

At the end of the construction completion part of the project, all the results related to the performance-
based approach, obtained both before and during concrete construction are recorded as a base record.
Then, for each part of the structure, an initial durability point is identified related to the performance
specifications of the contract, the specified and measured values of indicators together with the specified
and achieved cover. These tests and information can include an assessment of the durability properties
of the concrete using non-destructive in situ measurements such as electrical resistivity, surface
permeability, etc.

Start of Structure Monitoring

The initial durability point is the handover between the contractor and the owner. This makes it possible
to initiate the durability monitoring of the structure. The parts of the structure which will be operationally
monitored are defined, specifying the nature and timing of the monitoring.

23
The monitoring of the structure is based upon the evaluation of lifetime reference measured from in situ
tests or cores. The lifetime references are the parameters allowing a monitoring of the progress of the
degradation process. In the case of reinforcement corrosion, the lifetime reference is the position of the
chlorides front in the cover (depth of concrete for which the chlorides concentration is above the critical
concentration for initiation of corrosion).

To refine and deepen the monitoring, evaluation of additional lifetime references are necessary:

 Position of the carbonation front.

 Evaluation of the electrochemical state of the concrete and the reinforcement (resistivity, potential,
corrosion current density).

In a conservative way, the lifetime in terms of reinforcement corrosion is the time taken by the chloride
front to reach the first reinforcement (initiation period).

Knowledge of durability indicators associated with lifetime references allow, through the use of predictive
models, an estimation of the potential service period (in the case of new structures) or residual service
period (in the case of existing structures). Such a procedure helps the owner evaluate the risk for
potential deterioration and provides the opportunity to apply cost-effective protective measures before
the onset of more costly remedial measures necessary.

24
4. CHAPTER 4: ADDITIONAL STRATEGIES AND PROTECTIVE
MEASURES
4.1 GENERAL

Whatever approach to the durability design taken, they all have some limitations, since none of them
takes into account the effects of cracks or other defects that also typically occur during the production
of concrete structures. Therefore, for all new marine concrete infrastructure where high durability and
service life are of special importance, additional strategies and protective measures such as those
outlined in the following should also be considered.

For concrete construction work in the marine environment, there may occasionally also be a risk of early-
age exposure to seawater during construction before the concrete has gained sufficient maturity and
density. If the risk for such early-age exposure is high, special precautions or protective measures
should also be considered.

Since all special protective measures may have implications both for the economy of the project and for
the future operation of the structure, such measures should always be discussed and agreed with the
owner of the structure before these measures are selected.

4.2 STAINLESS STEEL REINFORCEMENT

For a long period of time, stainless steel reinforcement has proved to be a very efficient way of enhancing
durability and service life of concrete structures in the marine environment. When reinforcement based
on stainless steel was applied in a concrete pier on the Yucatán Coast in Mexico in 1937, the additional
cost of this protective measure later proved to have been an extremely good investment for the owner
[Gjørv, 2014]. Traditionally, the costs of stainless steel have been so high that it has normally not been
considered viable for general concrete structures. During recent years, however, new experience has
shown that a selective use of stainless steel in the most exposed or critical parts of the structure can be
very beneficial for enhancing the durability of concrete structures in the marine environment. Also, in
order to counteract the effect of high scatter and variability of achieved construction quality, a partial use
of stainless steel in the most exposed or critical parts of the structure has provided a very simple and
robust technical solution.

For many years, it was believed that a galvanic coupling between reinforcing bars based on stainless
steel and carbon steel would present a potential corrosion problem. Both extensive experimental
investigations and practical experience have demonstrated, however, that a partial use of stainless steel
in combination with carbon steel does not increase the risk of corrosion [Bertolini et al., 2004]. Therefore,
a partial replacement of the carbon steel with stainless steel in the most exposed parts of the structure
provides a good technical solution, both from a protective and cost-effectiveness point of view.

There are many different qualities of stainless steel reinforcement on the market, but depending on both
the chemical composition and the microstructure of the steel, there are basically three groups:

● Ferritic steel
● Austenitic steel
● Austenitic-ferritic steel (duplex)

The corrosion resistance required for chloride containing environments mainly depends on the alloying
elements of the stainless steel such as chromium, nickel, molybdenum and nitrogen, but the
microstructure is also important. A classification of the various types of stainless steel is given both in
the European Standard EN 10088-1 [CEN, 1995] and the US Standard AISI.

In Table 4.1, approximate chemical composition and designation of some typical grades of stainless
steel used for reinforcing bars are shown. In this table, the notation ‘L’ indicates that the steel also has
low carbon content and is therefore weldable.

25
 Approximate chemical composition
Designation
(% by mass)
Grade Microstructure
Other
AISI EN 10088-1 Cr Ni Mo
elements

304L 304L 1.4307 Austenitic 17.5-19.5 8-10 - -

316L 316L 1.4404 Austenitic 16.5-18.5 10-13 2-2.5 -
22-05 318 1.4462 Duplex 21-23 4.5-6.5 2.5-3.5 N
23-04 - 1.4362 Duplex 22-24 3.5-5.5 0.1-0.6 N
21-01 - 1.4162 Duplex 21-22 1.4-1.7 0.1-0.8 Mn, N

Table 4.1: Approximate chemical composition and designation of some typical grades of stainless steel used for
reinforcing bars [Bertolini and Gastaldi, 2011]

4.3 NON-METALLIC REINFORCEMENT

In recent years, applications of fibre-reinforced polymer composites (FRP) as reinforcement in concrete
structures have been growing rapidly. Although most of the current experience and durability data on
FRP composite installations comes from the aerospace, marine and the corrosion-resistance industries,
FRP composites have also been used as a material for the construction industry since the mid-1950s
[ACMA MDA, 2006]. A major development of FRP for civil engineering has been the application of
externally bonded FRP for rehabilitation and retrofitting of existing concrete structures. During the late
1970s and early 1980s, however, a variety of new applications of composite reinforcing products were
demonstrated, and already in 1986, the world’s first highway bridge using composite tendons was built
in Germany. In recent years, non-metallic reinforcement based on FRP composites has had a wide
range of applications, and current experience demonstrates that such reinforcement systems have a
great potential for concrete structures in the marine environment [Newhook and Mufti, 1996 ; Newhook
et al., 2000 ; Tan, 2003 ; Serancio, 2004 ; Newhook, 2006] (Figure 4.1).

Figure 4.1: Concrete pier with FRP reinforcement at St. Lawrence Seaway, Quebec

Since FRP composites may show very high mechanical properties (Table 4.2), such products also
represent a viable alternative to conventional steel tendons for pre-stressed and post-tensioned
concrete. For a long time, suitable anchorage systems were a problem, but in recent years, new types
of anchorage systems for FRP tendons have been developed providing practical applications of pre-
stressed systems [Gaubinger et al., 2002].

26
 Armid fibre CFRP
Glass fibre Basalt Carbon fibre Steel strand
(Twaron wire
(E-glass) fibre (HT) (St 1570/1770)
HM)
Tensile strength
2600 2300 3200 3500/7000 2800 >1770
(MPa)
Young´s modulus
125 74 90 230/650 160 205
(GPa)
Ultimate strain
2.3 3.3 3.0 0.6/2.4 1.6 7
(%)
Density
1.45 2.54 2.6 1.8 1.5 7.85
(g/cm3)

Table 4.2: Mechanical properties of advanced composite fibres [Noisternig et al., 1998 ; ReforceTech, 2013]

For many years, glass fibre was the most commonly used type of reinforcing fibre, mostly in the form of
E-glass, but also in the form of alkali resistance glass (AR glass). Due to a rapid increase in production
capacity and new production methods, however, the costs of both carbon and aramid fibres have
become more attractive. In recent years, basalt fibres with improved properties compared to glass fibres
have also been introduced [ReforceTech, 2013]. These fibres do not have any durability problems in the
highly alkaline environment of concrete. Also, the cost of these fibres is equivalent to that of glass fibre
and approximately 1/10 of carbon fibre. Embedded in a proper type of a polymer-based matrix, all of the
above types of fibres are currently commercially available in various qualities and dimensions as
reinforcing bars for concrete reinforcement.

In principle, the fundamental design methodology for FRP-products is similar to that of conventional
steel-reinforced concrete. Cross sectional equilibrium, strain compatibility, and constitutive material
behaviour form the basis of all approaches to the structural design of reinforced concrete structures,
regardless of the reinforcing material. For a proper structural design, however, the non-ductile and
anisotropic natures of FRP reinforcing products need to be specifically addressed, which is properly
taken into account in current standards and recommendations for structural design [CSA, 2003 ; ACI,
2007 ; fib, 2007].

4.4 CONCRETE SURFACE PROTECTION

For those offshore concrete platforms in the North Sea which were given a thick protective epoxy coating
applied to the fresh concrete surface during concrete construction later proved to have been a very
efficient way of protecting the concrete against chloride ingress [FIP, 1996 ; Årstein et al., 1998 ; Gjørv,
2014]. This protective coating was continuously applied to the fresh concrete surface during slip forming
when the fresh concrete still had an under-pressure and suction ability, achieving a very good bonding
between the concrete substrate and the coating.

In recent years, a number of new surface protection products have been introduced either for retarding
or preventing chloride ingress into concrete. In principle, the effect of such products may either be to
make the concrete surface less permeable to chloride ingress or to reduce the moisture content in the
concrete, although many products combine these effects.

Different types of surface protection products may be grouped into the following four classes as
schematically shown in Figure 4.2, where (a) shows organic coatings that form a continuous film on the
concrete surface, (b) shows hydrophobic treatments that line the interior surface of the concrete pores,
(c) shows treatments that fill the capillary pores, and (d) shows a thick and dense surface layer. While
all surface treatments open to water vapour may have a longer service life with no or slower loss of
bond, the effectiveness of such surface treatments may generally be less than that of dense protective
systems. It therefore appears that the selection of the method will be dependent on the long term use
of the structures as there is a trade-off between good barrier properties and good long-term effect of the
selected surface treatment.

27
 Figure 4.2: Schematic representation of different types of concrete surface protection: (a) organic coating,
(b) pore-lining treatments, (c) pore-blocking treatments, (c) thick and dense surface layers [Bijen, 1989]

In recent years, surface hydrophobation of concrete structures has been widely adopted as a protective
measure for concrete structures in chloride-containing environments. Since the protective efficiency of
such a polymer-based surface treatment may be reduced over time due to weathering, most of these
types of treatments need to be reapplied from time to time in order to ensure proper long-term protection.
In order to reduce the effect of weathering and thus increase the duration of the protective effect, the
hydrophobic system should penetrate the concrete substrate as deep as possible.

Colloidal silica may also penetrate deeply into the concrete, the depths of which vary upon the
formulation of the colloidal silica and the porosity of the concrete substrate. Although technically a pore-
blocking technology, colloidal silica reacts with the free Portlandite in the concrete to produce more
Calcium Silicate Hydrate thus becoming part of the concrete structure itself rather than a coating or
foreign pore plugging material. Laboratory tests have shown a measureable increase in the total specific
gravity of the concrete treated with colloidal silica indicating that a densification of the concrete is also
taking place.

For concrete structures in the marine environment, high moisture contents in the concrete substrate
may reduce the ability of both the pore-blocking and the hydrophobic agents to properly penetrate the
concrete surface [Liu et al., 2005]. This is demonstrated in Figure 4.3, where field tests with a
hydrophobic surface treatments were carried out upon the completion of a new concrete harbour
structure. In order to test the efficiency of different surface treatments to protect the open concrete deck
against chloride ingress, different parts of the deck were treated with two different types of silane-based
hydrophobic agents, partly in one layer (0.25 mm) and partly in two layers (0.50 mm) of application,
respectively. The two silane-based agents were of Type A (Isooctyltrietoxy) and Type D (Isobutyltrietoxy)
with two different molecule sizes, respectively, and both agents were sprayed on in the form of a gel in
order to increase the suction time. After four years of severe exposure in the seawater splashing zone,
all types of the different surface treatments had effectively protected the concrete against chloride
ingress. As can be seen from Figure 4.3, however, it was difficult to detect any difference in protective
efficiency between the two different types of hydrophobic system and also between one layer and two
layers of application, respectively. Based on a number of concrete cores removed from the concrete
deck, no depths of ingress of any of the hydrophobic agents could be detected. It appears, therefore,
that all the applied hydrophobic surface treatments had mainly given a thin protective film on the
concrete surface. At the time of surface treatment, the moisture content in the concrete substrate
appeared to be very high, e.g. 95 % or more, which may often be the case for concrete structures in a
moist marine environment.

28
 Figure 4.3: Effect of hydrophobic surface treatment on the protective efficiency after four years of chloride
exposure [Liu, 2006]

It should be noted that if the surface hydrophobation is applied to a concrete structure at a later stage
of exposure when the chlorides have already reached a certain depth, experience shows that a further
chloride ingress still takes place for some time due to a redistribution of the already existing chloride
content in the surface layer of the concrete [Arntsen, 2001].

4.5 CONCRETE HYDROPHOBATION

For certain types of thin concrete elements such as precast formwork, the whole bulk of the concrete
can be made hydrophobic. By using silane-based hydrophobic agents as an admixture to the fresh
concrete mixture and allowing the concrete to properly dry out before exposure, investigations have
shown that this may be a viable alternative to the traditional treatment of the concrete surface [Årskog
et al., 2011].

4.6 PROTECTIVE SKIN SYSTEMS

As a special case of concrete surface protection, protective skin systems based on thin, fibre-reinforced
polymer sheets can be prefabricated and installed in the formwork before concrete construction; this
may be a viable protection for specially exposed parts of the structure. After demoulding of the formwork,
the protective skin systems are then left in place, properly anchored and attached to the concrete
substrate. Such protective polymer sheets of up to 6 x 15 m in size have been tailor-made and
successfully applied in order to protect specially exposed parts of major marine concrete structures in
Korea [Lee, 2015].

4.7 CATHODIC PREVENTION SYSTEMS

For concrete structures with heavy chloride-induced corrosion, experience shows that repairs based on
cathodic protection (CP) systems are one of the most effective ways of getting such corrosion under
control. For the design of an efficient CP system, however, the challenge is always to provide the
necessary protective current in a reliable, controllable and durable way [Bertolini et al., 2008]. For such
a protective measure to be effective, proper electrical continuity within the rebar system must be
obtained. According to the European Standard for cathodic protection of concrete structures [CEN,
2000], the electrical resistance between any two points in the rebar system shall not exceed 1 Ohm. To
provide such electrical continuity during a repair may be technically much more difficult and substantially
more costly compared to that of providing such continuity during the construction of the new structure.
Specifying a proper electrical continuity within the rebar system as part of the durability design may
therefore be claimed as an effective strategy for future repairs based upon installing a CP system later
during the operational life of the structure.

29
Another protective strategy of new concrete structures, however, would be to install a complete CP
system already from the beginning instead of installing it at a later stage of operation when the chlorides
have already reached the embedded steel and corrosion started. This approach for ‘cathodic prevention’
was suggested and introduced by Pedeferri in the late 1980s [Pedeferri, 1992]. Later on, cathodic
prevention systems have been successfully applied to a number of new concrete structures [Broomfield,
1997 ; Bertolini, 2000 ; COST, 2003].

In order to compare cathodic prevention systems with other protective measures, it may be difficult to
estimate the long-term performance and service life of the protective system. In addition to the initial
costs of installation, the costs for operation and maintenance of electrode systems, wires and sensitive
electronic equipment need to be taken into account as well as long-term monitoring (Chapter 6).

If provision for a future installation of a cathodic prevention system is part of the durability specification,
it should be noted that the cathodic prevention system must be installed at a stage before the first
chlorides have reached the embedded steel and initiated any corrosion. Hence, a very close control and
following-up of chloride ingress during the operation of the structure must be continued. Since all
concrete structures typically show a high scatter and variability of achieved construction quality, this
may represent a great challenge for the owner during the operation of the structure. Also, larger parts
of the concrete structure may later on not necessarily be easily accessible neither for control of chloride
ingress nor for any installation of a cathodic prevention system.

4.8 CORROSION INHIBITORS

Corrosion inhibitors both for prevention and delay of corrosion initiation have been on the market for a
very long period of time. A number of different inhibitors for addition to the fresh concrete exist, but from
one type of product to another, the protecting mechanisms may be very different [Büchler, 2005]. Of the
various types of product, calcium nitrite is probably the most extensively tested and widely applied so
far. Extensive investigations have shown that a proper addition of this inhibitor is capable of both
preventing corrosion and decreasing the corrosion rate [Hinatsu et al., 1990]. In order to be efficient,
however, a critical ratio in the range of 0.5 to 1.0 between nitrite and chloride has to be present [Andrade,
1986 ; Gaidis and Rosenberg, 1987]. Also, since corrosion activation will consume the substance over
time while the concentration of chlorides will increase, it may be difficult to predict and guarantee the
long term effect of such a protective measure [Hinatsu et al., 1990]. If the concentration of the calcium
nitrite over time becomes too low, accelerated corrosion may also take place [Nürnberger, 1988 ; Ngala
et al., 2002].

4.8.1 STRUCTURAL SHAPE

Already at an early stage of field experience with concrete structures in the marine environment, it was
observed that open concrete harbour structures with a flat slab type of deck typically showed a much
better durability than structures with a beam and slab type of deck [Gjørv, 1968]. For the structures with
a beam and slab type of deck, the deck beams started to corrode at an early stage, while the slab
sections in-between showed a much better performance.

For a long time it was assumed that the different durability behaviour of these different types of deck
construction was due to easier and better placing and compaction of the fresh concrete in deck slabs
compared to that of the deep and narrow beams. In Norwegian concrete harbour construction, the
practical consequence of this experience was seen in the early 1930s, when the first open concrete
harbour structure with a flat slab type of deck was introduced (Figure 4.4). From then on, a number of
new concrete harbour structures with a flat slab type of deck were constructed along the Norwegian
coastline proved to have much better durability and long-term performance than beam and slab deck
structures. However, the structural design based on flat slab decks was often more expensive, and the
slab and beam type of deck was gradually reintroduced. It was thought that if the beams were made
shallower and wider, it would be easier to place the fresh concrete in the wider section beams. However,
even the shallower and wider deck beams again showed early steel corrosion, whilst the flat slab deck
continued to perform much better in terms of corrosion.

30
 Figure 4.4: Open concrete harbour structures with a flat slab type of deck typically showed a much better
durability than structures with a beam and slab type of deck [Gjørv, 1968]

What was not appreciated in the earlier days of reinforced concrete marine construction was that the
more exposed deck beams of an open concrete harbour structure would always absorb and accumulate
more chlorides than the less exposed deck slab sections above the beams. As a consequence, the steel
in the more exposed deck beams soon became depassivated and developed anodic areas, while the
less exposed slab sections in-between the beams acted as catchments areas for oxygen and formed
cathodic areas. In this way, a complex system of galvanic cell activity along the embedded rebar system
in the concrete deck developed. As a result, accelerated corrosion in the deck beams took place, while
the slab sections in-between remained cathodically protected [Gjørv, 1968].

Therefore, the structural design of concrete structures in the marine environment should always consider
that certain parts of the structures will be more exposed to intermittent wetting and drying in the splash
zone than other parts of the structures, and as a result be more vulnerable to corrosion of the embedded
steel.

4.8.2 PREFABRICATED STRUCTURAL ELEMENTS

For many marine concrete structures, various types of prefabricated structural elements can be used.
Since the concrete element manufacturing and casting work can be carried out under more controlled
quality conditions, not only will this provide more consistent construction but will also be a good strategy
for protection against early-age chloride exposure during construction. Possible surface protection
systems to the most exposed parts of the structures may further be applied under more controlled and
optimum conditions before the prefabricated structural elements or parts of the structures are installed.

For marine concrete construction, prefabricated structural elements may vary from small prefabricated
formwork- or deck slab-elements to very large structural concrete elements. Some concrete structures
may also partly or completely be prefabricated in a dry dock and then moved into their final location. For
the Northumberland Strait Bridge Project on the east coast of Canada [Tromposch et al., 1998], both
the pier shafts and the cantilever girders of up to 190 m length were prefabricated on shore before they
were moved into position and installed using a heavy floating crane (Figure 4.5).

31
Figure 4.5: Installation of prefabricated cantilever girders for the Northumberland Strait Bridge on the east cost of
Canada [Malhotra, 1996]

32
5. QUALITY ASSURANCE AND ACHIEVED CONSTRUCTION
QUALITY
5.1 CONCRETE QUALITY ASSURANCE

5.1.1 GENERAL

Upon completion of new concrete structures, the achieved construction quality typically shows a high
scatter and variability, and in a severe environment, any weaknesses and deficiencies will soon be
revealed whatever durability specifications and materials have been applied. Even before the concrete
is placed in the formwork, the quality of the concrete may show a high scatter and variability [Gulikers,
2011]. Depending on a number of factors during placing and compaction of the concrete such as the
flowability and stability of the fresh mix, the actually achieved concrete quality may show an even higher
scatter and variability. For major concrete infrastructure in severe environments, the reinforcement
system may also be highly congested leading to additional screening and segregation of the fresh
concrete during placing. For air-entrained concrete, large variations in the obtained air void
characteristics can occur [Gjørv and Bathen, 1987]; this problem is exacerbated when fly ash cements
are used and the carbon content of the fly ash varies [Nagi et al., 2007], affecting both the chloride
diffusivity and other important properties of the concrete.

Probably the best known and well documented quality issue during concrete construction is the failure
to meet the specified requirements for concrete cover. Although the specified concrete cover is normally
carefully checked prior to placing of the concrete, significant deviations can occur during concrete
construction. The loads during placing of the concrete may occasionally be too high compared to the
stiffness of the rebar system, or the spacers may occasionally have been insufficiently or wrongly placed.
In recent years, therefore, improved codes and procedures for achieving the specified concrete cover
with more confidence have been introduced. Still, however, the variability of achieved concrete cover
appears to remain a big challenge.

In current concrete codes and standards, a large number of procedures and test methods for quality
control of both concrete production and concrete construction exist, a thorough review of which is given
in the CIRIA Report [CIRIA, 2010]. However, since current concrete codes and standards also mainly
specify some descriptive durability requirements which are not possible to verify and control during
concrete construction, it is not possible to carry out a proper quality assurance.

As a result of the probability-based durability design as described in Chapter 3, performance-based

requirements to concrete quality (28-day chloride diffusivity) (D28) and concrete cover (XC) are
established. As also described and discussed in Chapter 3, performance-based requirements for the
concrete quality can be specified as other separate durability indicators. As a result, documentation of
both compliance with the durability specification and achieved construction quality can be provided such
as outlined and described in the following. Such documentation is very important since it may have
implications both for the obtained durability and future operation of the structures.

During concrete construction, both the concrete quality and the concrete cover must be properly
controlled, which is achieved by providing ongoing verification and documentation. If the requirement to
concrete quality is based on the chloride diffusivity (RCM), the testing procedures for which are briefly
described and discussed in Appendix G. In this appendix, the methods for control of concrete cover are
also briefly discussed. When any unacceptable deviations in concrete quality and concrete cover occur,
immediate actions for correction should be taken and the anomalies and measures used recorded.

5.1.2 CONTROL OF CONCRETE QUALITY

As described and discussed in Chapter 3, the chloride diffusivity (RCM) of the given concrete is a very
important concrete quality parameter reflecting both the resistance of the concrete to chloride ingress
and the general durability properties of the concrete. Although the RCM method [AASHTO, 2003] is a
very rapid test method which provides data on the chloride diffusivity within a few days, this is not
sufficient for regular quality control during concrete construction. Therefore, a calibration curve relating
the general relationship between the chloride diffusivity (RCM) and the electrical resistivity of the given
concrete should be established before concrete construction (Appendix G). The regular control of the

33
chloride diffusivity during concrete construction is then based on measurements of the electrical
resistivity carried out, providing a quick, non-destructive test on the same concrete specimens being
used for the regular control of the 28-day compressive strength.

5.1.3 CONTROL OF CONCRETE COVER

For concrete structures in the marine environment, the applied concrete cover is relatively large (70 mm
plus), and the reinforcement is often highly congested making it difficult to measure cover thickness very
accurately with conventional cover meters. The use of stainless steel reinforcement may further
complicate such measurements as non-ferrous re-bar is more difficult to detect with conventional cover
meters, although cover meters based on pulse induction can then be used. Currently, also more
sophisticated systems for control of achieved concrete cover based on image scanning exist (Appendix
G). With such equipment available, there should be little excuse for allowing low cover in new concrete
structures, despite this still seems to continue to be a problem on even large construction sites.

5.2 ACHIEVED CONSTRUCTION QUALITY

5.2.1 GENERAL

Chapter 3 discussed how the obtained data from the quality control of both the 28-day chloride diffusivity
(RCM) (D28) and the concrete cover (XC) can be used as input parameters for new durability analyses
documenting the achieved construction quality. In this way, the achieved construction quality is
characterised and quantified in the form of probability of corrosion.

Since the specified chloride diffusivity is based on small and separately produced concrete specimens
cured in the laboratory for only 28 days, this may be quite different from that obtained on the construction
site during concrete construction. Therefore, some additional documentation of achieved chloride
diffusivity on the construction site during the construction period is required. Since neither the 28-day
chloride diffusivity from the laboratory nor the achieved chloride diffusivity on the construction site during
concrete construction reflects the potential chloride diffusivity of the concrete, additional documentation
about the long-term diffusivity is also required. For the owners of the structures, a best possible and
most complete documentation of the achieved construction quality is of great importance, since it may
have implications both for the actual durability and expected service life of the structures.

5.2.2 COMPLIANCE WITH DURABILITY SPECIFICATION

As a result of the durability design based on the DURACON Model, a certain ‘Service period’ with a
probability for corrosion of less than 10 % is specified. To show compliance with such a durability
specification, a new durability analysis must be carried out based upon the achieved average values
and standard deviations of both the 28-day chloride diffusivity (RCM) (D28) and the concrete cover (XC)
from the regular concrete quality control sampling. For this durability analysis, all previously assumed
input parameters from the original durability analyses are now kept the same. Hence, the documentation
of compliance with the durability specification primarily will reflect the results obtained from the regular
quality control sampling during concrete construction including the recording scatter and variability of
the quality.

5.2.3 IN SITU QUALITY

The achieved chloride diffusivity (RCM) on the construction site during concrete construction is primarily
based on the testing of a number of concrete cores removed from the structure under construction
(Figure 5.1). One or more representative dummy elements may also be produced on the construction
site, from which a number of additional cores can be removed and tested during the construction period.
Based on the achieved chloride diffusivity after one year on the construction site combined with the site
data on cover thicknesses as new input parameters, a fresh durability analysis is then carried out. Since
all the other previously assumed input parameters from the original durability design are kept constant,
the new analysis provides documentation of the achieved in situ quality during the construction period.

34
 Figure 5.1: Development of chloride diffusivity (RCM) on the construction site and in the laboratory during the
construction period [Gjørv, 2014]

5.2.4 POTENTIAL QUALITY

For most binder systems, the development of chloride diffusivity (RCM) tends to plateau after about one
year of water curing at 200C in the laboratory. Therefore, to estimate the potential construction quality
of the structure, the development of chloride diffusivity should also be determined on separately
produced, water cured specimens in the laboratory for a period of up to one year (Figure 5.1). This
chloride diffusivity in combination with the achieved site data on concrete cover as new input parameters
to the durability analysis provides a basis for documenting the potential construction quality of the
structure; all the other originally assumed input parameters to the durability design are kept the same.

5.3 EXAMPLE OF OFFSHORE QUALITY CONTROL IN THE 1970s

For all the offshore concrete structures produced for the oil and gas explorations in the North Sea, the
strictest durability specifications (FIP, 1973) and the strictest procedures for quality control of both
concrete production and concrete construction were applied; even stricter and more stringent than the
existing concrete codes and practice of the day. Also for some of these concrete structures, however,
very costly durability problems later on occurred which could be ascribed due to weaknesses and
deficiencies in achieved construction quality [FIP, 1996 ; Helland et al., 2010 ; Gjørv, 2014] (Figure 5.2).

(a)

35
 (b)

Figure 5.2: After 13 years of service, extensive and very costly repairs based on cathodic protection of the
Oseberg A Platform (1988) were carried out (Courtesy of Trond Østmoen)

Also, for the production of the offshore concrete structures in the North Sea, the durability specifications
were mainly based on descriptive requirements for water/cement ratio and cement content. As a result,
construction and monitoring test results were not possible to verify and therefore proper recorded quality
assurance was not carried out during concrete construction. While the Ekofisk Tank (1973) which was
the first offshore concrete structure in the North Sea, was produced with a water/cement ratio of 0.45,
all the other offshore concrete structures were produced with a water/cement ratio of 0.40 or less in
combination with nominal concrete covers for ordinary reinforcement and prestressing steel of 75 and
100 mm, respectively.

For all of the offshore concrete structures, the most rigorous methods for producing a highest possible
homogeneity of the concrete were also adopted [Gjørv, 2014]. As a measure of this consistency, it can
be seen from the regular concrete quality control of all platforms produced during the period 1972–1984
that the standard deviation from the 28-day testing of compressive strength typically varied from 2.3 to
3.9 MPa (Table 5.1). Such standard deviations are extremely low compared to that generally observed
even for new, large construction sites.

28-day compressive strength (MPa)

Platform (year)
Specified grade Obtained mean Standard deviation Obtained grade1)

452) 2.32) 41.62)

Ekofisk 1 (1972) 402)
57 3.5 51.9
Beryl A (1974) 45 55 3.0 50.7
Brent B (1975) 45 53 3.1 48.5
Brent D (1975) 50 54.2 2.5 50.6
Statfjord A (1975) 50 54.6 3.0 50.2
Statfjord B (1979) 55 62.5 3.9 56.9
Statfjord C (1982) 55 67.5 3.8 62.0
Gullfaks A (1984) 55 65.2 3.3 60.3
1)
Obtained grade = obtained mean – 1.45 x standard deviation
2)
150 x 300 mm cylinder

Table 5.1: Concrete quality control based on 100 mm cubes from platform construction during the period 1972-
1984 [Moksnes and Jakobsen, 1985]

36
From Table 5.1 it can be seen that the concrete for the Brent B Platform (1975) was produced with a
standard deviation of only 3.1. For this particular platform, extensive concrete coring was later on carried
out showing a very high scatter and variability of the actually achieved concrete quality [Gjørv, 2014].
All the Ø100 mm concrete cores were taken from two different elevations of the utility shaft above water
of + 14.4 and + 7.8 m, respectively, and one below water of - 11.5 m. All concrete cores were cut into a
number of slices for investigating various concrete properties. Some selected results showing the
chloride diffusivities (RCM) are given in Figure 5.3, demonstrating the very high scatter and variability
of the actually achieved concrete quality.
60
+14.4 m
50 + 7.8 m
-11.5 m
40
Percent

30

20

10

0
18-20 20-22 22-24 24-26 26-28 28-30 30-32 32-34
2 . -12
Chloride diffusivity (m /s 10 )

Figure 5.3: Achieved in-place concrete quality (chloride diffusivity) from three different elevations of the utility shaft
of the Brent B Platform (1975) [Gjørv, 2014]

37
6. CONDITION ASSESSMENT, PREVENTIVE MAINTENANCE
AND REPAIRS
6.1 GENERAL

For most operational concrete structures, maintenance and repairs are mostly reactive, and the need to
take appropriate repair or preventive measures is only identified at an advanced stage of deterioration.
For chloride-induced corrosion repairs at such a stage are then both technically difficult and
disproportionately costly compared to that of carrying out regular condition assessments and installing
preventive maintenance measures during operation of the structures. Therefore, for all new major
concrete infrastructure where high performance and long service life are of special importance, regular
condition assessments and preventive maintenance should be part of the operational maintenance
manual. This is not only important from a technical and economical point of view but also a good strategy
from a sustainability point of view [Gjørv and Årskog, 2012].

In recent years, a rapid international development of general systems for Life Cycle Management (LCM)
of important infrastructure facilities has taken place [Grigg, 1988 ; O`Connor and Hyman, 1989 ; Hudson
et al., 1987 ; RIMES, 1997 ; BRIME, 1997]. Depending on the number of facilities to be included,
established LCM systems for both network- and project level are commercially available. In many
countries, national authorities have also developed their own LCM systems. For many important
concrete structures, therefore, general condition assessment and preventive maintenance are already
part of the established LCM systems.

However, for all major concrete infrastructure in chloride-containing environments, special procedures
for control of chloride ingress during operation of the structures should be required. For new concrete
structures, the establishment of such procedures should be an important and integral part of the
durability design. Already at an early stage of the design, appropriate locations for the future control of
chloride ingress in critical parts of the structure should be selected and designed to be easily accessible
during the operation of the structure. If some of the critical parts of the structure are not readily accessible
for future control of chloride ingress, instrumentation based on embedded probes for chloride monitoring
can be applied. If protective measures such as cathodic prevention or provisions for such protective
measure are specified, necessary preparation for this must also be identified as the strategy at an early
stage of the durability design.

Although general condition assessment and preventive maintenance are already part of the established
LCM systems for structures, additional regular control of the chloride ingress during operation of marine
concrete structures is very important. The general basis for this is briefly outlined and discussed in the
following, and a brief outline of current repair experience is also included.

6.2 CONTROL OF CHLORIDE INGRESS

Even if the strongest requirements for the construction quality have both been specified and achieved
during concrete construction, extensive experience demonstrates that for all concrete structures in the
marine environment, a varying extent of chloride ingress will take place during the operation of the
structures. In the marine environment, an early-age chloride exposure may also take place before the
concrete has gained sufficient curing and maturity. If the risk for such early-age exposure during
concrete construction is high, this must also be taken into account in the durability design; under certain
exposure conditions, special protective measures may be required.

However, for the regular condition assessment and control of chloride ingress during the operation of
the structure, it is important to have a detailed monitoring plan for the concrete structure showing the
selected locations where the future control of chloride ingress is to be recorded. These locations must
be as representative as possible for the most exposed and critical parts of the structure. Since all
concrete structures typically show a high scatter and variability of achieved construction quality, it may
be difficult to know which parts of the structure will have the highest likely rate of future chloride ingress.
At an early stage of operation, therefore, more locations for chloride control should be selected. After
some time when it becomes clearer where the highest rates of chloride ingress are taking place, the
continuing monitor control of chloride ingress can then be concentrated in such ‘hot spot’ locations.

38
Normally, the rate of chloride ingress is faster at an early stage of operation. Therefore, the control of
chloride ingress should also be carried out more frequently at the early stages of operation. The first
control should preferably be carried out shortly after completion of the structure in order to provide a
proper reference level for the future chloride measurements. The next control sampling could then be
carried out after a service period of e.g. 10 years. Later, both the frequency and extent of control
measurements will depend on the observed rates of chloride ingress.

During the early stages of operation, the control of chloride ingress can be obtained in a more simple
way to obtain an approximate indication on how fast the chloride front is moving into the various parts
of the structure. From each location, measurements can be based on simple dust sampling from for
example by 5x16 mm drilling holes in 5 mm increments of depth. At a later stage when deeper chloride
ingress is observed, however, the measurements from each location should be based on one or more
concrete cores (Ø100 mm), from which very thin layers of dust samples are ground off and analysed for
a more detailed observation of the chloride ingress.

The chloride content can be measured by use of different procedures and test methods as described in
several standards and guidelines for condition assessment of concrete structures. Since field methods
for testing of chloride contents are generally less accurate, such measurements should only be used at
an early stage of condition assessment in order to obtain the rate of the chloride front migration. At a
later stage when it becomes necessary to obtain a more detailed understanding of the chloride front
ingress, the chloride analyses should be based on more accurate laboratory methods. Also, for the best
possible comparison of chloride data from one inspection period to another, the same procedure and
test method for control of the chloride ingress should be repeated.

If instrumentation based upon embedded probes is part of the given monitoring plan for future control
of chloride ingress, a design approach for this should also be made at an early stage for the durability
monitoring program. For example, instrumentation may be appropriate for monitoring of chloride ingress
in special or critical parts of the structure which later on may not be accessible for a manual control.

Although embedded probes for future monitoring of chloride ingress can provide valuable data on rates
of chloride ingress in certain locations of the structure, experience has shown that a few embedded
probes cannot replace a more complete control of the chloride ingress in large concrete structures.
Since concrete structures typically show a high scatter and variability of achieved construction quality,
it may be difficult to know where the highest rates of chloride ingress will take place. Also, since
embedded probes primarily provide information on the rate at which the chloride front moves into the
concrete, more detailed information about the chloride ingress is required in order to estimate or
calculate the probability of corrosion of reinforcement that will result.

6.3 PROBABILITY OF CORROSION

As soon as at least six incremental depths of chloride samples below the concrete surface can be
obtained, a regression analysis of the chloride data for the given exposure period can be produced with
a curve fitting to Ficks 2. Law of Diffusion. From this, a more complete chloride ingress curve as that
shown in Figure 6.1 can be drawn. This is referred to as the ‘Chloride penetration profile’, which then
provides the basis for determining some of the important durability parameters needed for calculating
the corrosion probability (Chapter 3).

39
 Figure 6.1: The result of a regression analysis of all observed data on chloride ingress for a given period of
exposure

Based on the above regression analysis, data on both the surface chloride concentration (CS) and the
apparent chloride diffusivity (Da) are obtained. A calculation of corrosion probability also needs
information about the time dependence of the chloride diffusivity (α) and the concrete cover (XC). Proper
information about the concrete cover may either be available from the concrete quality control carried
out during concrete construction or from separate measurements carried out later. At an early stage of
operation, however, reliable information about the time dependence of the chloride diffusivity (α) is not
available. For an early calculation of corrosion probability, therefore, the calculation must be based on
an estimated, empirical α-value for the given concrete in the given environment. However, as soon as
two or more values for the apparent chloride diffusivity (Da) become available from further control
measurements, a more accurate α-value can be determined as shown in the following [Poulsen and
Mejlbro, 2006]:
𝑡 𝛼
𝐷𝑎2 = 𝐷𝑎1 ∙ ( 1 ) (6.1)
𝑡2

where 𝐷𝑎1 and 𝐷𝑎2 are the apparent chloride diffusivities after exposure periods of t1 and t2 , respectively.
The time dependence of the chloride diffusivity (α) then becomes:

ln(𝐷𝑎2 /𝐷𝑎1 )
𝛼= (6.2)
ln(𝑡1 /𝑡2 )

As soon as three or more values for the apparent chloride diffusivity become available, even more
accurate α-values can be determined based on a regression analysis, and the more future values for
the apparent chloride diffusivity that become available, the more accurate the α-value will be. Hence,
the calculation of corrosion probability will gradually become more accurate and reliable during the
operation of the structure. Before the probability of corrosion becomes too high, appropriate protective
measures for control of the further chloride ingress can then be implemented (Chapter 4).

40
6.4 PROTECTIVE MEASURES

Depending on type of protective measure as discussed in Chapter 4, the observed rates of chloride
ingress can either be reduced or halted. If the chlorides have not penetrated very far into the concrete
cover, a surface treatment or coating may slow down the further rate of chloride ingress. If the chlorides
have penetrated significantly but the critical levels have still not reached the embedded steel, cathodic
prevention provides the principle protective measure that can arrest the further chloride ingress thus
avoiding the onset of corrosion. In this way, control of the chloride ingress in combination with
calculations of corrosion probability provides a basis for regular condition assessment and preventive
maintenance for the structure.

If the penetrating chlorides are allowed to reach the embedded steel and corrosion starts, it is just a
question of time before corrosion damage develops and costly repairs have to be carried out. As
described in the following, not only are the repairs costly but subsequent further maintenance of the
repairs and repair methods are necessary which also have significant cost implications.

6.5 REPAIRS

6.5.1 PATCH REPAIRS

For repairs carried out at an early stage of chloride-induced corrosion, patch repairs may be good
enough for controlling the further corrosion process; at least for a certain period. However, if the
deterioration has reached an advanced stage of corrosion, patch repairs are not able to bringing the
corrosion under control.

The poor and short term results produced for patch repairs for heavily chloride-induced corrosion was
first observed and reported on the San Mateo-Hayward Bridge (1929) in the Bay Area of San Francisco
in the early 1950s [Gewerts et al., 1958]. This bridge had been patch repaired due to chloride-induced
steel corrosion, first by local cleaning of the damaged areas and then patched by shotcreting, but after
a short period of time, steel corrosion was observed adjacent to the patched areas. During the extensive
field investigations of this bridge, a detailed potential mapping of the patch repaired structure by use of
half-cell copper-copper sulphate electrodes was also carried out; this was the first time potential
mapping of a concrete structure was carried out and reported in the literature. In this way, it was found
that the patched areas had typically formed cathodic areas, while anodic areas had formed adjacent to
the patched areas. As a result, increased rates of steel corrosion adjacent to the patched areas had
taken place. Subsequently, the poor results of patch repairs for heavily chloride-induced corrosion have
later been confirmed by a large number of field investigations reported in the literature.

6.5.2 CATHODIC PROTECTION

For heavily chloride-induced corrosion, Stratfull already in the early 1970s had demonstrated that
cathodic protection would be the most efficient way to get such corrosion under control [Stratfull, 1974].
A great deal of further experience has confirmed that cathodic protection is currently the most effective
way of getting such corrosion under control [Broomfield, 1997 ; Bertolini et al., 2004].

There is not much experience reported in the literature on the long-term performance of cathodic
protection systems. However, recent investigations of 105 installations of cathodic protection repairs in
the Netherlands were carried out, of which 50 installations had been operating more than ten years or
longer [Polder et al., 2012]. The service period without any major intervention varied between 10 and
20 years, but the need for some extent of intervention typically increased for the increased age of the
installation. In order to generalise the observations, a survival analysis was also carried out showing
that there was a 10 % probability for necessary intervention after seven years or less, and a 50 %
probability for necessary intervention after a service period of 15 years or less. As a basis for these
results, however, it was emphasized that a proper maintenance of the cathodic protection systems had
been carried out, which involved testing for depolarisation at least twice a year and visual inspection
once a year, as described by the European Standard for cathodic protection [CEN, 2000].

For all installations of a cathodic protective (CP) system, the maintenance is normally part of the
maintenance contract between the owner and the CP contractor. It should be noted that all CP systems
are based on very sensitive electronic equipment in combination with a number different types of

41
components typically being exposed to a very aggressive and corrosive environment; power units may
stop working, anode-copper connections may corrode, reference electrodes may fail and anode
materials or primary anodes may degrade [Broomfield, 1997 ; COST, 2003]. Therefore, the regular
control and maintenance of such systems must be carried out by qualified personnel, the following-up
and costs of which represent a further challenge for the owner to the operation of the structure.

42
7. REFERENCES
AASHTO (2003): “TP 64-03: Predicting Chloride Penetration of Hydraulic Cement Concrete by the Rapid
Migration Procedure”, American Association of State Highway and Transportation Officials, Washington,
DC.

ACI (2007): “Report on Fiber-Reinforced Polymer (FRP) Reinforcement for Concrete Structures”, ACI
Committee 440, Report 440R-7, Farmington Hills, ISBN 978-0-87031-259-5, 100 pp.

ACMA MDA (2006): http://www.mdacomposites.org.

Andrade, C. (1986): “Some Laboratory Experiments on the Inhibitor Effect of Sodium Nitrite on
Reinforcement Corrosion”, Cement and Concrete Aggregate, Vol. 8, No. 2, pp. 110-116.

Angst, U., Elsener, B., Larsen, C.K. and Vennesland, Ø. (2009): “Critical Chloride Content in Reinforced
Concrete – A Review, Cement and Concrete Research”, Vol. 39, pp. 1122-1138.

Angst, U. (2011): “Chloride Induced Reinforcement Corrosion in Concrete”, PhD Thesis 2011:113,
Department of Structural Engineering, Norwegian University of Science and Technology, NTNU,
Trondheim.

Arntsen, B. (2001): “In-Situ Experiences on Chloride Redistribution in Surface-Treated Concrete

Structures”, Proceedings Vol.1, Third International Conference on Concrete under Severe Conditions
– Environment and Loading, ed. by N. Banthia, K. Sakai and O.E. Gjørv, University of British Columbia,
Vancouver, ISBN 0-88865-782-X, pp. 95-103.

ASTM (2010): “C1202-10 Electrical Indication of Concrete’s Ability to Resist Chloride Ion Penetration”,
ASTM International, West Conshohocken.

Atkins, P.W. and De Paula, J. (2006): “Physical Chemistry”, 8. ed., Oxford University Press, ISBN 978-
0-19-870072-2, Oxford.

Atwood, W.G. and Johnson, A.A. (1924): “The Disintegration of Cement in Seawater, Transactions
ASCE”, V. 87, pp. 204-230.

Bamforth, P.B. (1999): “The Derivation of Input Data for Modeling Chloride Ingress from Eight-Year
Coastal Exposure Trials”, Magazine of Concrete Research, Vol. 51, No. 2, pp. 87-96.

Beeby, A.W. (1977): “Cracking and Corrosion”, UK Research Program on Concrete in the Oceans,
Report 2/11, London.

Bertolini, L. (2000): “Cathodic Prevention”, Proceedings, COST 521 Workshop, ed. By D. Sloan, P.A.M.
Basheer, The Queens´s University, Belfast.

Bertolini, L., Elsener, B., Pediferri, P. and Polder, R. (2004): “Corrosion of Steel in Concrete –
Prevention, Diagnosis, Repair”, Wiley-VCH, Weinheim, ISBN 3-527-30800-8, 392 pp.

Bertolini, L., Lollini, F., Polder, R.B. and Peelen, W.H.A. (2008): “FEM-Models of Chatodic Protection
Systems for Concrete Structures”, Proceedings, First International Symposium on Life-Cycle Civil
Engineering, ed. by F. Biondini and D.M. Frangopol, Taylor & Francis Group, London, ISBN 978-0-415-
46857- 2, pp. 119-124.

Bertolini, L. and Gastaldi, M. (2011): “Corrosion Resistance of Low-Nickel Duplex Stainless Steel
Rebars”, Materials and Corrosion, Vol. 62, pp. 120-129.

Bijen, J.M. (ed.) (1989): “Maintenance and Repair of Concrete Structures”, Heron, Vol. 34, No. 2, pp. 2-
82.

Bijen, J. (1998): “Blast Furnace Slag Cement for Durable Marine Structures”, VNC/BetonPrisma, DA´s-
Hertogenbosch, ISBN 90-71806-37-5, 62 pp.

43
Bjegović, D., Beushausen, H. and Serdar, M. (eds.) (2014): “Proceedings of RILEM International
Workshop on Performance-Based Specification and Control of Concrete Durability”, RILEM Publications
S.A.R.L., ISBN 978-2-35158-135-3, Bagneux, 657 pp.

BRE (2001): “Delayed Ettringite Formation”, Information Paper IP11/01, ISBN 1 86081 485 9.

BRIME (1997): “Bridge Management in Europe”, EC-DG-VII-RTD, The European Union – Program
Contract No. RO-97-SC.2220.

Broomfield, J.P. (1997): “Corrosion of Steel in Concrete”, E & FN Spon, London and New York, ISBN 0-
419-19630-7, 240 pp.

Browne, R. et al. (1980): “Marine Durability Survey of the Tongue Sand Tower”, Concrete in the Ocean
Program, CIRIA UEG Technical Report No. 5, Cement and Concrete Association, London.

Büchler, M. (2005): “Corrosion Inhibitors for Reinforced Concrete”, in: Corrosion in Reinforced Concrete
structures, ed. by H. Böhni, Woodhead Publishing, Cambridge, ISBN 1-85573-768-X, pp. 190-214.

CEB (1992): “Durable Concrete Structures – Design Guide”, Comité Euro- International du Béton – CEB,
Bulletin d´Information No. 183, Thomas Telford, London, ISBN 0-7277-1620-4, 112 pp.

CEN (1995): “EN 10088-1: Stainless steels – Part 1: List of stainless steels”, European Standard CEN,
Brussels.

CEN (2000): “EN 12696: Cathodic protection of steel in concrete”, European Standard CEN, Brussels.

CIRIA (2007): “Early-age thermal crack control in concrete”, CIRIA C660, Reprinted 2014, ISBN 978-
086107-660-2, London.

CIRIA (2010): “The use of concrete in maritime engineering – a good practice guide”, CIRIA C674, ISBN
978-086017-674-9, London, 364 pp.

Collepardi, M., Marcialis, A. and Turriziani, R. (1970): “Kinetics of Penetration of Chloride Ions in
Concrete”, l´Industria Italiana del Cemento, Vol. 4, pp. 157-164.

Collepardi, M., Marcialis, A. and Turriziani, R. (1972): “Penetration of Chloride Ions Into Cement Pastes
and Concretes”, Journal, American Ceramic Society, Vol. 55, No. 10, pp. 534-535.

Conciatori, D., Grégoire, E., Samson, É., Marchand, J., Chouinard, L. (2014): “Sensitivity of chloride
ingress modelling in concrete to input parameter variability”, Materials and structures, DOI
10.1617/s11527-014-0374-8.

Conciatori, D., Grégoire, E., Samson, É., Marchand, J., Chouinard, L. (2014): “Statistical analysis of
concrete transport properties, Materials and Structures”, Vol. 47, pp. 89-103.

COST (2003): “Action 521: Corrosion of Steel in Reinforced Concrete Structures”, Final Report, ed. by
R. Cigna,C. Andrade, U, Nürnberger, R. Polder, R. Weydert and E. Seitz, European Communities
EUR20599, Luxenburg, ISBN 92-894-4827-X, 238 pp.

CSA (2003): “S806-02 Fibre Reinforced Polymer Reinforcement for Concrete Structures”, Canadian
Standards Association, Rexdale.

DuraCrete (2000a): “The European Union – “Brite EuRam III Research Project: Probabilistic
Performance Based Durability Design of Concrete Structures”, Final Technical Report, Document BE95-
1347/R17, CUR, Gauda.

DuraCrete (2000b): “General Guidelines for Durability Design and Redesign”, The European Union –
Brite EuRam III Research Project: Probabilistic Performance Based Durability Design of Concrete
Structures,Document BE95-1347/R15, CUR, Gauda.

44
Enevoldsen, I and Sørensen, J.D. (1992): “Course on Statistical Analysis”, University of Ålborg, Ålborg.

Ferreira, M. (2004): “Probability Based Durability Design of Concrete Structures in Marine Environment”,
Doctor Dissertation, Department of Civil Engineering, University of Minho, Guimarães, 178 pp.

Ferreira, M., Årskog, V.N., Jalali, S. and Gjørv, O.E. (2004): “Software for Probability-Based Durability
Analysis of Concrete Structures”, Proceedings Vol. 1, Fourth International Conference on Concrete
Under Severe Conditions – Environment and Loading, ed. by B.H. Oh, K. Sakai, O.E. Gjørv and N.
Banthia, Seoul National University and Korea Concrete Institute, Seoul, ISBN 89-89499-02-X 93530,
pp. 1015-1024.

fib (2007): “FRP Reinforcement in RC Structures”, fib Bulletin No. 40, Lausanne, ISBN 978-2-88394-
080-2.

FIP (1973): “Recommendations for the Design and Construction of Concrete Sea Structures”,
Federation Internationale de la Precontrainte, London, ISBN 0-7210-0932-8.

FIP (1996): “Durability of Concrete Structures in the North Sea – State of the Art Report”, Federation
Internationale de la Precontrainte, London.

Gadea, J., Soriano, J., Martín, A., Campos, P. L., Rodríguez, A., Junco, C., Adán, I., Calderón, V. (2010):
“The Alkali–Aggregate Reaction for Various Aggregates Used in Concrete”.

Gaidis, J.M. and Rosenberg, A.M. (1987): “The Inhibition of Chloride-Induced Corrosion in Reinforced
Concrete by Use of Calcium Nitrite, Cement and Concrete Aggregate”, Vol. 9, No. 1, pp. 30-33.

Gaubinger, B., Bahr, G., Hampel, G. and Kollegger, J. (2002): “Innovative Anchorage System for GFRP-
Tendons”, Proceedings, First fib Congress, Session 7, Vol. 6, Japan Prestressed Concrete Engineering
Association, Tokyo, pp. 305-312.

Gehlen, C. and Schiessl, P. (1999): “Probability-Based Durability Design for the Western Scheldt
Tunnel”, Structural Concrete, P.1 (2), pp. 1-7.

Gehlen, C. (2007): “Durability Design According to the New Model Code for Service Life Design”,
Proceedings Vol. 1, Fifth International Conference on Concrete Under Severe Conditions – Environment
and Loading, ed. by F. Toutlemonde, K.Sakai, O.E. Gjørv and N. Banthia, Laboratoire Central des Ponts
et Chaussées, Paris, ISBN 2-7208-2495-X, pp. 35-50.

Gewertz, M.W., Tremper, B., Beaton, J.L. and Stratfull, R.F. (1958): “Causes and Repair of Deterioration
to a California Bridge due to Corrosion of Reinforcing Steel in a Marine Environment, Highway Research
Board Bulletin 182”, National Academy of Sciences – National Research Council Publication 546,
Washington DC, 41 pp.

Gjørv, O.E. (1968): “Durability of Reinforced Concrete Wharves in Norwegian Harbours”,

Ingeniørforlaget AS, Oslo, 208 pp.

Gjørv, O.E., Vennesland, Ø. and El-Busaidy, A.H.S. (1977): “Electrical Resistivity of Concrete in the
Oceans”, OTC Paper 2803, Annual Offshore Technology Conference, Houston, Texas, pp. 581-588.

Gjørv, O.E., Vennesland, Ø. and El-Busaidy A.H.S. (1986a): “Diffusion of Dissolved Oxygen through
Concrete, Materials Performance”, Vol. 25, No. 12, pp. 39-44.

Gjørv, O.E. and Kashino, N. (1986b): “Durability of a 60 Year Old Reinforced Concrete Pier in Oslo
Harbour, Materials Performance”, Vol. 25, No. 2, pp. 18-26.

Gjørv, O.E. and Bathen, E. (1987): “Quality Control of the Air-Void System in Hardened Concrete”,
Nordic Concrete Research, Vol. 6, pp. 95-110.

45
Gjørv, O.E. and Sakai, K. (eds.) (2000): “Concrete Technology for a Sustainable Development in the
21st Century”, E & FN Spon, London and New York, ISBN 0-419-25060-3, 386 pp.

Gjørv, O.E. (2003): “Durability of Concrete Structures and Performance-Based Quality Control”,
Proceedings, International Conference on Performance of Construction Materials in The New Millenium,
ed. by A.S. El-Dieb, M.M.R. Taha and S.L. Lissel, Shams University, Cairo, ISBN 977-237-191, 10 pp.

Gjørv, O.E. (2010): “Durability Design and Quality Assurance of Concrete Infrastructure”, Concrete
International, Vol. 32, No. 9, pp. 29-36.

Gjørv, O.E. and Årskog, V.N. (2012): “Preventive Maintenance and Sustainability of Concrete
Infrastructure”, Concrete International, Vol. 34, No. 5, pp. 36-40.

Gjørv, O.E. (2014): “Durability Design of Concrete Structures in Severe Environments”, Second edition,
CRC Press, Boca Raton, London and New York, ISBN 978-1- 4665-8729-8, 270 pp. Also published in
Chinese by Press of China Building Materials Industry, Beijing (2015), and in Portuguese by Oficina de
Textos, São Paulo (2015).

Gjørv, O. E. (2015): “Quality Control for Concrete Durability – A Case Study Provides Comparisons of
Work Performed Under Performance and Prescriptive Specifications”, Concrete International, Vol. 37,
No. 11, pp. 52-57.

Grigg, N.S. (1988): “Infrastructure Engineering and Management”, John Wiley and Sons, New York,
ISBN 978-0471849742, 380 pp.

Gulikers, J. (2011): “Practical Implications of Performance Specifications for Durability Design of

Reinforced Concrete Structures”, Proceedings, Workshop on Performance-Based Specifications for
Concrete, ed. by F. Dehn and H. Beushausen, MFPA Leipzig, ISBN 978-3-9814523-0-3, pp. 341-350.

Horrigmoe, G. (2000): “Future Needs in Concrete Repair Technology”, In: “Concrete Technology for a
Sustainable Development in the 21st Century”, ed. by O.E Gjørv and K. Sakai, E & FN Spon, London
and New York, ISBN 0-419-25060-3, pp. 332-340.

ISO (2006a): “ISO 14040 – Environmental management, Life cycle assessment, Principles and
framework”, International Organization for Standardization, Geneva, Switzerland, 20 pp.

ISO (2006b): “ISO 14044 – Environmental management, Life cycle assessment, Requirements and
guidelines”, International Organization for Standardization, Geneva, Switzerland, 46 pp.

JSCE (2007): “Standard Specifications for Concrete Structures – 2007 – Design”, JSCE Guidelines No.
15, Japan Society of Civil Engineetrs, Tokyo, 503 pp.
http://www.jsce.or.jp/committee/concrete/e/Standard_specification/JGC15_Design_1.0.pdf

Helland, S., Aarstein, R. and Maage, M. (2010): “In-Field Performance of North Sea HSC/HPC Offshore
Platforms with regard to Chloride Resistance, Structural Concrete”, Vol. 11, No. 1, pp. 15-24.

Hinatsu, J.T., Graydon, W.F. and Foulkes, F.R. (1990): “Voltametric Behaviour of Iron in Cement. Effect
of Sodium Chloride and Corrosion Inhibitor Additions”, Journal of Applied Electrochemistry, Vol. 20, No.
5, pp. 841-847.

Hudson, S.W., Carmichael, R.F., Moser, L.O. Hudson, W.R., Wilkes, W.J. (1987): “Bridge Management
Systems”, Transportation Research Board, National Research Council, NCHRP Report 300,
Washington, DC., 80 pp.

Kong, J.S., Ababneh, A. N., Frangopol, D. M. and Xi, Y. (2002): “Reliability Analysis of Chloride
Penetration in Saturated Concrete, Probabilistic Engineering Mechanics”, Vol. 17, No. 3, pp. 305-315.

Lee, S.W. (2015): Kookmin University, Seoul (private communication).

46
Liu, G., Stavem, P. and Gjørv, O.E. (2005): “Effect of Surface Hydrophobation for Protection of Early
Age Concrete Against Chloride Penetration”, Proceedings, International Conference on Water Repellent
Treatment of Building Materials, Hydrophobe IV, ed. by J. Silfwerbrand, Aedificatio Publishers, Zurich,
ISBN 3-931681-81-5, pp. 93-103.

Liu, G. (2006): “Control of Chloride Penetration into Concrete Structures at Early Age”, Ph.D. Thesis
2006:46, Norwegian University of Science and Technology – NTNU, Trondheim, ISBN 82-471-7838-9,
155 pp.

Malhotra, V.M. (ed.) (1996): “Proceedings, Third International Conference on Performance on Concrete
in Marine Environment”, ACI SP-163.

Moksnes. J. and Jakobsen, B. (1985): “High-Strength Concrete Development and Potentials for Platform
Design”, OTC Paper 5073, Annual Offshore Technology Conference, Houston, Texas, pp. 485-495.

Nagi, M.A., Okamoto, P.A., Kozikowski, R.L. and Hover, K. (2007): “Evaluating Air-Entraining
Admixtures for Highway Concrete”, NCHRP Report 578, Transportation Research Board, Washington,
DC, 49 pp.

Mangat, P.S. and Molloy, B.T. (1994): “Prediction of Long-Term Chloride Concentration in Concrete,
Materials and Structures”, Vol. 27, pp. 338-346.

McGee, R. (1999): “Modelling of Durability Performance of Tasmanian Bridges”, Proceedings, Eight

International Conference on the Application of Statisics and Probability, Sidney.

NAHE (2004): “Durable Concrete Structures – Part 1: Recommended Specifications for New Concrete
Harbor Structures, Structures, – Part 2: Practical Guidelines for Durability Design and Concrete Quality
Assurance”, Norwegian Association for Harbor Engineers (NAHE), TEKNA, Oslo (in Norwegian).

Newhook, J. and Mufti, A. (1996): “A Reinforcing Steel-Free Concrete Bridge Deck for the Salmon River
Bridge”, Concrete International, Vol. 18.

Newhook, J., Bakht, B., Tadros, G. and Mufti, A. (2000): “Design and Construction of a Concrete Marine
Structure Using Innovative Technologies”, In: “Proceedings of the Third International Conference on
Advanced Composite Materials in Bridges and Structures (CSCE)”, ed. by M. El-Badry, Ottawa.

Newhook, J. (2006): “Glass FRP Reinforcement in Rehabilitation of Concrete Marine Infrastructure”,

The Arabian Journal for Science and Engineering, Vol. 31, No. 1C, pp. 53-75.

Ngala, V.T., Page, C.L. and Page, M.M. (2002): “Corrosion Inhibitor Systems for Remedial Treatment
of Reinforced Concrete”, I. Calcium Nitrite, Corrosion Science, Vol. 44, No. 9, pp. 2073-2087.

Nilsson, L., Ngo, M.H. and Gjørv, O.E. (1998): “High-Performance Repair Materials for Concrete
Structures in the Port of Gothenburg”, In: “Proceedings, Vol. 2, Second International Conference on
Concrete Under Severe Conditions – Environment and Loading”, ed. by O.E. Gjørv, K. Sakai and N.
Banthia, E & FN Spon, London and New York, ISBN 0-419-23850-6, pp. 1193-1198.

Noisternig, F., Dotzler, F. and Jungwirth, D. (1998): “Development of CFK Prestressed Elements,
Seminarband Kreative Ingenieurleistungen”, Darmstadt- Wien (in German).

Nürnberger, U. (1988): “Special Measures for Corrosion Protection of Reinforced and Prestressed
Concrete”, Otto-Graf-Institute Series No. 79, Stuttgart (in German).

O’Connor D.S. and Hyman W.A. (1989): “Bridge Management Systems”, Report FHWA-DP-71-01R,
Demonstration Project 71, Demonstration Projects Division, Federal Highway Administration,
Washington DC.

Pedeferri, P. (1992): “Cathodic Protection of New Concrete Construction, Proceedings, International

Conference on Structural Improvement through Corrosion Protection of Reinforced Concrete, Institute
of Corrosion”, Document E7190, London.

47
Petersson, P.E. (1995): “The Salt Frost Durability of Concrete – Field Tests”, SP Report, Swedish
National Testing and Research Institute, Borås, 64 pp.

PIANC (1998): “Life Cycle Management of Port Structures – General principles”, Supplement to Bulletin
No. 99, Brussels.

PIANC Norway/NAHE (2009): “Durable Concrete Structures – Part 1: Recommended Specifications for
New Concrete Harbor Structures, – Part 2: Practical Guidelines for Durability Design and Concrete
Quality Assurance”, Norwegian Association for Harbor Engineers (NAHE), 3. ed., TEKNA, Oslo (in
Norwegian)

Polder, R., Andrade, C., Elsener, B., Vennesland, Ø., Gulikers, J., Weidert, R. and Raupach, M. (2000):
“RILEM TC 154-EMC: Electrochemical Techniques for Measuring Metallic Corrosion, Materials and
Structures”, Vol. 33, pp. 603-611.

Polder, R.B., Leegwater, G., Worm, D. and Courage, W. (2012): “Service Life and Life Cycle Cost
Modelling of Cathodic Protection Systems for Concrete Structures”, International Congress on Durability
of Concrete, Norwegian Concrete Association, Oslo, 12 pp.

Poulsen, E. and Mejlbro, L. (2006): “Diffusion of Chlorides in Concrete – Theory and Application”, Taylor
& Francis, London and New York, ISBN 0-419- 25300-9, 442 pp.

Pruckner, F. and Gjørv, O.E. (2004): “Life Cycle Design of Concrete Structures”, Proceedings Vol. 1,
Fourth International Conference on Concrete Under Severe Conditions - Environment and Loading, ed.
by B.H. Oh, K. Sakai, O.E. Gjørv and N. Banthia, Seoul National University and Korea Concrete Institute,
Seoul, ISBN 89-89499-02-X 93530, pp. 965-974.

ReforceTech (2013): “Basalt Fiber Reinforcement Technology”, Oslo, htt://www.reforcetech.com.

RIMES (1997) Road Infrastructure Maintenance Evaluation Study on Pavement and Structure
Management System, EC-DG-VII-RTD, Program Contract No. RO-97-SC 1085/1189.

Rosenblueth, E. (1981): “Two-point estimates in probabilities”, Applied Mathematical Modeling, Vol. 5,

pp. 329–335.

Samson E. and Marchand J. (2006): “Multiionic approaches to model chloride binding in cementitious
materials”, Rilem Proc. 51, pp. 101-122.

Samson E. and Marchand J. (2007a): “Modeling the effect of temperature on ionic transport in
cementitious materials”, Cement and Concrete Research, Vol. 37, pp. 455-468.

Samson E. and Marchand J. (2007b): “Modeling the transport of ions in unsaturated cement-based
materials”, Computers and Structures. Vol. 85, pp. 1740-1756.

Serancio, R. (ed.) (2004): “FRP in Civil Engineering, Proceedings of the Second International
Conference on FRP Composites in Civil Engineering”, Balkema Publishers, London.

Siemes, T., Vrouwenvelder, T. and Beukel, A. Van den (1985): “Durability of Buildings: A Reliability
Analysis”, HERON, V. 30, No. 3, pp. 2-48.

Siemes, A.J.M. and Rostam, S. (1996): “Durability Safety and Serviceability – A Performance Based
Design”, Proceedings, IABSE Colloquium on Basis of Design and Actions on Structures, Delft.

Siemes, T., Schiessl, P. and Rostam, S. (2000): “Future Developments of Service Life Design of
Concrete Structures on the Basis of DuraCrete”, In: “Service Life Prediction and Ageing Management
of Concrete Structures”, ed. by D. Naus, RILEM, pp. 167-176.

Stratfull, R.F. (1974): “Experimental Cathodic Protection of a Bridge”, Research Report No. 635117-4,
FHWA D-3-12, Department of Transportation, California, Sacramento, 54 pp.

48
Stewart, M.G. and Rosowsky, D.V. (1998): “Structural Safety and Serviceability of Concrete Bridges
Subject to Corrosion”, Journal of Infrastructure Systems, Vol.4, No. 4), pp. 146-155.

Takewaka, K. and Mastumoto, S. (1988): “Quality and Cover Thickness of Concrete Based on the
Estimation of Chloride Penetration in Marine Environments”, Proceedings, Second International
Conference on Concrete in Marine Environment, ACI SP 109, ed. by V.M. Malhotra, pp. 381-400.

Tan, K. (ed.) (2003): “Proceedings of the Sixth International Symposium of FRP Reinforcement for
Concrete Structures”, World Scientific, Singapore.

Tang, L., Nilsson, L.-O. and Basher, P.A.M. (2012): “Resistance of Concrete to Chloride Ingress”, Spon
Press, London and New York, ISBN 978-0-415-48614-9, 241 pp.

Tang, L. and Gulikers J. (2007): “On the Mathematics of Time-Dependent Apparent Chloride Diffusion
Coefficient in Concrete”, Cement and Concrete Research, Vol. 37, No. 4, pp. 589-595.

Teng, S., Divsholi, B. S., Lim, T. Y. D. and Gjørv, O. E. (2014): “Concrete with Very High Resistance to
Chloride Ingress”, Concrete International, Vol. 36. No. 5, pp. 30-36.

Tromposch, E.W., Dunaszegi, L., Gjørv, O.E. and Langley, W.S. (1998): “Northumerland Strait Bridge
Project – Strategy for Corrosion Protection”, In: “Proceedings Vol. 3, Second International Conference
on Concrete under Severe Conditons – Environment and Loading”, ed. by O.E. Gjørv, K. Sakai and N.
Banthia, E & FN Spon, London and New York, ISBN 0-419-23850-6, pp. 1714-1720.

Thomas M.D.A. and Bamforth, P.B. (1999): “Modelling Chloride Diffusion in Concrete – Effect of Fly Ash
and Slag”, Cement and Concrete Research, Vol. 29, pp. 487-495.

Thomas M.D.A., Shehata, M.H., Shashiprakash, S.G., Hopkins, D.S. and Cail, K. (1999): “Use of
Ternary Cementitious Systems Containing Silica Fume and Fly Ash in Concrete”, Cement and Concrete
Research, Vol. 29, pp. 1207-1214.

Thomas, M.D.A., Bremner, T. And Scott, A.C.N. (2011): “Actual and Modeled Performance in a Tidal
Zone”, Concrete International, Vol. 33, No. 11, pp. 23-28.

Tuutti, K. (1982): “Corrosion of Steel in Concrete”, Report Fo. 4, Swedish Cement and Concrete Institute,
Stockholm, ISSN 0346-6906, 489 pp.

Zhang, T. and Gjørv, O.E. (1996): “Diffusion Behavior of Chloride Ions in Concrete”, Cement and
Concrete Research, Vol. 26, pp. 907-917.

Wig, R.J. and Furguson, L.R. (1917): “What is the Trouble with Concrete in Sea Water?”, Engineering
News Record, V. 79, 532, 641, 689, 737 and 794.

Årskog, V., Borgund, K. and Gjørv, O.E. (2011): “Effect of Concrete Hydrophobation Against Chloride
Penetration”, Key Engineering Materials, 466, pp. 183-190.

Årstein, R., Rindarøy, O.E., Liodden, O. and Jenssen, B.W. (1998): “Effect of Coatings on Chloride
Penetration into Offshore Concrete Structures”, Proceedings Vol. 2, Second International Conference
on Concrete under Severe Conditions – Environment and Loading, ed. by O.E. Gjørv, K. Sakai and N.
Banthia, E & FN Spon, London and New York, ISBN 0-419-23850-6, pp. 921- 929.

49
APPENDIX A: WORKING GROUP MEMBERS
Odd E. Gjørv (Chairman)
Norwegian University of Science and Technology, NTNU
Faculty of Engineering Science and Technology
Department of Structural Engineering
NO-7491 Trondheim, Norway
Email: odd.gjorv@ntnu.no

Boy-Arne Buyle (Secretary)

UiT The Arctic University of Norway
NO-8505 Narvik, Norway
Email: boy-arne.buyle@uit.no

Pilar Alaejos
Concrete Department of Laboratorio Central
Research Center in Civil Engineering, Spanish Ministry of Public Works, Spain
Email: palaejos@cedex.es

Pascal Collet
TOTAL SA
DGEP/DEV/TEC/STR
Civil Engineering - Marine structures – Buildings, France
Email: pascal.collet@total.com

Tiffany Desbois
Cerema Ouest - Département Laboratoire Régional de Saint Brieuc, Ouvrages d'Art et Maritimes
22015 Saint-Brieuc CEDEX , France
Email: tiffany.desbois@cerema.fr

Steve Hold
E-mail: steveholdconsulting@gmail.com
Formerly at
ARUP, UK, Email: steve.hold@arup.com

Ema Kato
Port and Airport Research Institute
3-1-1 Nagase, Yokosuka, Kanagawa, Japan
Email: katoh-e@pari.go.jp

Stefan Kohn
bremenports GmbH & Co. KG
27568 Bremerhaven, Germany
Email: stefan.kohn@bremenports.de

Rodrigo García Orera

Sevilla Port Authority
41012 Sevilla, Spain
Email: rodrigo.garcia.orera@gmail.com

Thorsten Reschke
Federal Waterways Engineering and Research Institute (BAW)
Department of Structural Engineering, Section B3
76187 Karlsruhe, Germany
Email: thorsten.reschke@baw.de

50
Thomas Spencer
Moffatt & Nichol Blaylock
San Diego, CA 92108; USA
Email: tspencer@moffattnichol.com

Larissa Terdu
Shell Projects and Technology, Project & Engineering Services
Shell Global Solutions International BV
2280 AB Rijswijk, the Netherlands
Email: Larissa.Terdu@shell.com

Ad van der Toorn

University of Delft
Faculty of Civil Engineering and Geosciences
2628CN Delft, the Netherlands
Email: A.vanderToorn@tudelft.nl

Andrew Wareing
UK
Email: andrew.wareing@atkinsgobal.com

51
APPENDIX B: TERMS OF REFERENCE
B.1 HISTORICAL BACKGROUND – DEFINITION OF THE PROBLEM

In many countries, deterioration and repair of important marine concrete infrastructure has emerged as
a most severe and demanding challenge for the owners of structures. Although a number of deteriorating
processes may represent potential problems, extensive experience demonstrates that electrochemical
corrosion of the embedded steel poses the most critical and greatest threat to the durability and long-
term performance of the structures. Although current standards have been improved in recent years,
still an uncontrolled penetration of salt with subsequent steel corrosion can take place on relatively new
important marine concrete structures. As soon as the corrosion starts, the owner has a problem, which
in the beginning only represents a cost problem but later on also develops into a more difficult safety
problem.

The durability and service life of the structure is dependent upon preventing the initiation of
electrochemical corrosion. This is typically done by ensuring a quality concrete (low permeability, free
of chlorides, reactive aggregates, high temperatures during curing, etc.), concrete resistant to
environmental attack (freeze thaw, alkali-aggregate reaction, sulfate attack, carbonation, chlorides,
etc.), and concrete cover. Additional measures can be undertaken to extend the life of a structure, such
as sealers, coatings, corrosion resistant materials and cathodic protection systems. The durability of a
structure is dependent upon the system of cementitious materials, aggregates, water, admixtures, and
reinforcing.

Although all minimum durability requirements stated by existing standards must always be followed and
fulfilled for new concrete structures, more and more owners are willing to invest somewhat more in order
to obtain an increased and more controlled durability and service life beyond what is possible when only
based on current standards. New recommendations and guidelines for increased durability, service life
and service life modelling of new and important marine concrete infrastructure should be developed.

With the environmental constraints placed on the construction of new facilities in many countries, it may
be easier or more economical to extend the service life of an existing structure. Guidance should also
be developed for mitigation measures to extend the service life of existing structures.

B.2 OBJECTIVE OF THE WORKING GROUP

For all the structures where a partial use of stainless steel was applied, this protective measure proved
to be a much simpler and more robust technical solution. For all of these concrete structures, stainless
steel in the most exposed and vulnerable parts of the structures also proved to be economically
competitive, even on a short-term basis. On a long-term basis, additional high expenses both for
installation and operation of a cathodic prevention system are involved.

This Working Group would provide guidance to owners and designers of marine concrete infrastructure
worldwide, in order to provide a safe, efficient and cost-effective design and construction of these
structures.

The report of the WG is only to be considered as a guidance in addition to existing standards for concrete
durability and service life. It should also be considered as an additional document for improved quality
assurance during concrete construction as well as the regular condition assessment and preventive
maintenance during operation of the structures.

B.3 EARLIER REPORTS TO BE REVIEWED

 PIANC Norway and the Norwegian Association for Harbor Engineers (2009): “Durable Concrete
Structures – Recommendations and Guidelines for New Concrete Harbor Structures”.
 Any other standards, recommendations and reference documents from private or public
organisations.

52
Consistency shall be achieved with existing PIANC report on near topics, such as:

 PIANC WG 31 (1998): “Life Cycle Management of Port Structures – General principles”.

 PIANC WG 44 (2005): “Accelerated Low Water Corrosion”.
 PIANC WG 103 (2008): “Life Cycle Management of Port Structures – Recommended Practice
for Implementation”.

B.4 MATTERS TO BE INVESTIGATED

Review whether any companies or national or international public organisations have existing modern
engineering standards, recommendations or guidelines for increased durability and long-term
performance of marine concrete structures that are publically available or which the owners would be
willing to share with the Working Group for use.

Review available documents related to the service life, durability and degradation of concrete in various
environments that we operate in, as well as the methods of mitigating these degradation mechanisms
and of providing durable repair.

B.5 SUGGESTED FINAL PRODUCT

Recommendations for durability design, quality assurance as well as regular condition assessment,
repair and preventive maintenance of new marine concrete infrastructures.

B.6 DESIRABLE DISCIPLINES OF THE MEMBERS OF THE WORKING GROUP

In addition to the owners of marine concrete infrastructure and port authorities, the Working Group
members should represent all parties involved such as consulting engineers, contractors and public
authorities. Members with a research background shall also be welcomed.

B.7 RELEVANCE FOR COUNTRIES IN TRANSITION

The final product would help countries in transition through the increased durability and increased
service life of their new marine concrete infrastructure, taking advantage of the collective knowledge
gathered by this PIANC Working Group.

53
APPENDIX C: CONCRETE COVER
C.1 GENERAL

According to NF EN 1992-1-1, the three criteria for protection of the steel reinforcement against
corrosion are based on requirements to concrete quality, cover thickness and concrete cracking. The
concrete cover is defined as the distance between the concrete surface and the nearest reinforcement
(including links and stirrups, etc.). It is designed to ensure:

 Protection of the steel against corrosion

 Safe transmission of bond forces (adhesion steel/concrete)
 Adequate fire resistance

The Eurocode 2 recommendations propose to adjust and optimize the concrete cover regarding
exposure class, strength class, construction details, nature of reinforcement and design working life.
The design of the concrete cover includes the following nine stages:

Stage 1: Determining exposure classes

Stage 2: Determining structural classes

Stage 3: Modulation of structural classes

Stage 4: Taking the environment into account (Cmin,dur)

Stage 5: Taking specific protection into account (∆Cdur,st, ∆Cdur,add)

Stage 6: Taking adhesion of the concrete/reinforcement into account (Cmin,b)

Stage 7: Taking specific constraints into account (ki)

Stage 8: Taking execution allowance into account (∆Cdev)

Stage 9: Determining the nominal cover (Cnom)

The nominal cover shall be specified on the drawings. It is defined as a minimum cover plus an
allowance in the design for deviation. By default, this allowance is equal to 10 mm. It could be reduced
if the operational means of verification with respect to the cover during execution are justified.

The used value shall be the greater value for Cmin satisfying the requirements for both bond and
environmental conditions:

Cmin = max {Cmin,b;Cmin,dur+ΔCdur,γ-ΔCdur,st-ΔCdur,add;10mm)+ki

Where:

Cmin,b: Minimum cover due to bond requirement

Cmin,dur: Minimum cover due to environmental conditions

ΔCdur,γ: Additive safety element (recommended value: 0)

ΔCdur,st: Reduction of minimum cover for use of steel with a tested

corrosion resistance

ΔCdur,add: Reduction of minimum cover for use of additional protection

ki : Increase of minimum cover in case of concrete abrasion

54
C.2 STAGE 1: DETERMINING EXPOSURE CLASSES

C.2.1 TAKING INTO ACCOUNT THE PHYSIC-CHEMICAL AGGRESSIVENESS OF

ENVIRONMENT

Exposure classes reflect the actions of the environment (physical and chemical actions in addition to
the mechanical actions) that each part of the structure will be exposed to throughout the service period
of the structure. Each part of the structure may be subjected simultaneously to several aggressions and
attacks. It is therefore necessary to consider each part of the structure for determination of exposure
class (Table C.1).

Class designation Description of the environment

1 No risk of corrosion or attack
For concrete without reinforcement or embedded metal: all exposures except
X0 where there is freeze/thaw, abrasion or chemical attack

For concrete with reinforcement or embedded metal: very dry

2 Corrosion induced by carbonation
XC1 Dry or permanently wet
XC2 Wet, rarely dry
XC3 Moderate humidity
XC4 Cyclic wet and dry
3 Corrosion induced by chlorides
XD1 Moderate humidity
XD2 Wet, rarely dry
XD3 Cyclic wet and dry
4 Corrosion induced by chlorides from sea water
XS1 Exposed to airborne salt but not in direct contact with sea water
XS2 Permanently submerged
XS3 Tidal, splash and spray zones
5 Freeze/Thaw Attack
XF1 Moderate water saturation, without de-icing agent
XF2 Moderate water saturation, with de-icing agent
XF3 High water saturation, without de-icing agents
XF4 High water saturation with de-icing agents or sea water
6 Chemical attack
XA1 Slightly aggressive chemical environment according to EN 206-1, Table2
XA2 Moderately aggressive chemical environment according to EN 206-1, Table2
XA3 Highly aggressive chemical environment according to EN 206-1, Table2
Table C.1: Exposure classes related to environmental conditions in accordance with EN 206-1

The cover is designed to ensure durability regarding corrosion due to carbonation (XC) and chlorides
(XS, XD). The exposure classes XF related to freezing and XA related to chemical attacks cannot be
directly used for the design of cover. Nevertheless, it should be taken into account as shown in the
following.

The freezing may lead to damage of the cover (cracking, spalling), according to the freezing level (low,
moderate, high) and the exposure level of the structure to the chlorides other than marine chlorides
(Table C.2). It is therefore necessary to take it into account in the calculation of the cover using XC and
XD classes.

55
 Exposure classes
XF1 XF2 XF3 XF4
Exposure level to the Low XC4 Not applicable For concrete without air Not applicable
chlorides other than entraining agents XC4
marine chlorides With air entraining
agents XD1
Moderate Not applicable XD1, XD3 for Not applicable XD2, XD3 for
very exposed very exposed
elements elements
High Not applicable XD3 Not applicable XD3

Table C.2: Exposure classes in the case of structure exposed to the freezing and the chlorides other than marine
chlorides

C.2.2 TAKING INTO ACCOUNT THE MECHANICAL AGGRESSIVENESS OF

ENVIRONMENT

In addition to exposures classes XC, XS and XD, the Eurocode 0 introduces abrasion classes XM, which
may be relevant in maritime context. These exposure classes can increase the minimum cover to reflect
the risk of abrasion (sacrificial thickness), Table C.3.

Abrasion Conditions Coefficient

classes values
Moderate abrasion:
- Members of industrial site subjected to the traffic of vehicles with air tyres
XM1 K1 = 5 mm
- Friction of mooring lines or chains
- Sediment carried by the swell
Heavy abrasion:
XM2 - Members of industrial site subjected to the traffic of fork-lift with air or solid rubber tyres K2 = 10 mm
- ship hulls along a mooring quay

Extreme abrasion: Members of industrial sites frequented by fork-lifts with elastomer or

XM3 K3 = 15 mm
steel tyres or track vehicles

Table C.3: Abrasion classes

C.3 STAGE 2: STRUCTURAL CLASSES

The cover thickness has to consider the structure lifetime. The Eurocode 0 introduces the concept of
design (working) life which is the assumed period for which a structure or part of it is to be used for its
intended purpose with anticipated maintenance but without major repair being necessary. The Eurocode
2 tells the difference between six structural classes from S1 to S6. Buildings and usual civil engineering
structures come under the structural class S4. They are designed for a design life of 50 years. Bridges
and port and maritime infrastructures come under the structural class S6. They are designed for 100
years.

C.4 STAGE 3: MODULATION OF STRUCTURAL CLASSES

Structural class can be adjusted according to different considerations on the concrete properties, the
reinforcement characteristics and constructive provisions adopted. The general rule is: improve
durability properties allow to reduce structural class and thus reduce cover. Possible modulations
concern the strength class of the concrete, the binder nature and the cover compaction.

C.4.1 STRENGTH CLASSES

The strength class of the concrete plays the role of durability or compaction indicator. Table C.4 shows
the possible modulations of structural class based on the strength class of concrete, for each exposure

56
 class. Modulation -2 is justified by the significant reduction in the depth of carbonation and chloride
diffusion in high performance concretes and ultra-high performance, fibre reinforced concretes.

Structural Exposure class

class X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3
modulation
-1 ≥ C30/37 ≥ C35/45 ≥ C40/50 ≥ C45/55
-2 ≥ C50/60 ≥ C55/67 ≥ C60/75 ≥ C70/85

Table C.4: Modulation of the structural class according to the strength class

In the case of the application of the performance-based approach, modulation related to strength class
is replaced by modulation related to thresholds of durability indicators.

C.4.2 BINDER NATURE

The modulation of structural class depending of binder nature is justified by a significant reduction of the
carbonation depth for dense concretes (adequate strength) and concretes with appropriated chemical
nature (CEM I without fly ash), Table C.5.

Structural class Exposure class

modulation X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3
≥ C35/45 ≥ C40/50
-1 Not applicable
Concrete CEMI, without fly ash Concrete CEMI with fly ash

Table C.5: Modulation of the structural class according to the binder nature

C.4.3 COMPACT COVER

Modulation -1 of the structural class is also possible, subject to justify the execution for obtaining a good
compaction of the cover.

C.5 STAGE 4: TAKING ENVIRONMENT INTO ACCOUNT (Cmin,dur)

The Cmin,dur value is the minimum cover required to achieve the expected durability in terms of
environmental aggressiveness to which the concrete is exposed. The concrete cover Cmin,dur depends
on the structural class (after any modulations) and exposure classes as shown in Table 6 (case of
reinforcing steel) and Table C.7 (pre-stressing steel). These Tables are taken from the Eurocode 2.

Structural Exposure class

class X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3
S1 10 10 10 15 20 25 30
S2 10 10 15 20 25 30 35
S3 10 10 20 25 30 35 40
S4 10 15 25 30 35 40 45
S5 15 20 30 35 40 45 50
S6 20 25 35 40 45 50 55

Table C.6: Values of minimum cover requirements with regard to durability for reinforcement steel in accordance
with EN 10080

57
 Structural Exposure class
class X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3
S1 10 15 20 25 30 35 40
S2 10 15 25 30 35 40 45
S3 10 20 30 35 40 45 50
S4 10 25 35 40 45 50 55
S5 15 30 40 45 50 55 60
S6 20 35 45 50 55 60 65
Table C.7: Values of minimum cover requirements with regard to durability for pre-stressing steel

C.6 STAGE 5: TAKING SPECIFIC PROTECTION INTO ACCOUNT

(∆Cdur,st, ∆Cdur,add)

C.6.1 TAKING REINFORCEMENT TYPE INTO ACCOUNT (∆Cdur,st )

The recommended value, without further specification, is 0 mm. However, in the case of using
reinforcements whose corrosion resistance is proven (stainless steel or galvanised reinforcements for
example), the French national annex to Eurocode 2 provides the opportunity to reduce the cover by
introducing the ∆Cdur,st. Specifying the ∆Cdur,st value in a contract constitutes a dispensation to Eurocode
2.

C.6.2 TAKING ADDITIONAL PROTECTION INTO ACCOUNT (∆CDUR,ADD)

By default, ∆Cdur,add = 0 mm. However, in the case of use of adhesive coatings justified to prevent the
penetration of aggressive agents during the design life, the French national annex to Eurocode 2
provides the opportunity to reduce the cover by introducing the ∆C dur,add value. Specifying the ∆Cdur,add
value in a contract constitutes a dispensation to Eurocode 2.

C.7 STAGE 6: TAKING ADHESION OF THE CONCRETE/REINFORCEMENT

INTO ACCOUNT (CMIN,B)

The optimal transmission of forces between reinforcements and concrete requires an adhesion and a
minimal covering of reinforcements. This minimum cover is defined in Table C.8 for reinforcement of
concrete and in Table C.9 for post-tensioned circular and rectangular ducts for bonded tendons and pre-
tensioned tendons.

Cmin,b
Arrangement of bars
Dmax < 32 mm Dmax ≥ 32 mm
Separated Φ Φ + 5 mm
Bundled Φeq Φeq + 5 mm

Table C.8: Minimum cover Cmin,b – Reinforcement of concrete

With Dmax the maximum aggregate size; Φ the diameter of bar; Φeq the equivalent diameter of the
bundled bars.

Cmin,b
Bars in circulars ducts Min (Φ ; 8cm)
Bars in rectangular Max (a ; b/2)
ducts
Bars pre-tensioned Max (2Φ ; Dmax)

Table C.9: Minimum cover Cmin,b – Post-tensioned ducts and pre-tensioned tendons

With Dmax the maximum aggregate size; Φ the diameter of ducts and tendons; a, b the dimensions of
the rectangular ducts with b>a.

58
C.8 STAGE 7: TAKING SPECIFIC CONSTRAINTS INTO ACCOUNT (ki)

Some execution provisions affecting the cover characteristics are taken into account in the design. This
is the case of uneven facings and concrete cast against uneven surfaces.

In the case of uneven facings (such as ribbed finishes or exposed aggregate), the minimum cover C min
is increased by at least 5 mm (ki > 5 mm).

For concrete cast against uneven surfaces, the minimum cover should be increased to take into account
the unevenness importance. In France, the following values are used:

Concrete cast against prepared ground Cmin ≥ 30 mm

Concrete cast directly against soil Cmin ≥ 65 mm

In addition, when the facing is exposed to a risk of abrasion, the minimum cover Cmin is increased by a
coefficient ki, Table C.10, in accordance with the abrasion class (stage 1):

Abrasion classes Conditions Coefficients values

XM1 Moderate abrasion k1 = 5 mm
XM2 Heavy abrasion k2 = 10 mm
XM3 Extreme abrasion k3 = 15mm

Table C.10: Coefficient depending on the abrasion classes

C.9 STAGE 8: TAKING EXECUTION ALLOWANCE INTO ACCOUNT (∆Cdev)

The minimum cover is increased. By default, the value is 10 mm. In certain situations, the accepted
deviation and hence allowance may be reduced:

Fabrication subjected to a quality assurance system, in which

10 mm ≥ ∆Cdev ≥ 5 mm
the monitoring includes measurements of the concrete cover
It can be assured that a very accurate measurement device is
used for monitoring and non-conforming members are rejected 10 mm ≥ ∆Cdev ≥ 0 mm
(e.g. precast elements)

C.10 STAGE 9: DETERMINING THE NOMINAL COVER (Cnom)

The Eurocode 2 draws attention to the cracking problems which could lead a nominal cover C nom above
50 mm. In other words, in the case of particularly aggressive environments, such as the marine
environment, which may lead to an important cover, the cover shall be optimised by using the
opportunities offered to underestimate the structural class (Stage 3) and reduce the execution allowance
(Stage 8). In the case of a design resulting in a very important cover (>70 mm for example), the
opportunity to use specific reinforcements or additional protection may be considered (Stage 5).

59
APPENDIX D: DURACON MODEL
D.1 GENERAL

After the first international guidelines for probability-based durability design of concrete structures were
introduced by the European research program DuraCrete in 2000 [DuraCrete, 2000b], such design was
also applied to a number of new major concrete structures in Norwegian harbours. During the
following years, however, when practical experience with these guidelines for new projects was gained,
this design was further developed for more practical and appropriate implementation by the Norwegian
Association for Harbor Engineers (NAHE), but the basic principles remained essentially the same
[NAHE, 2004]. Lessons learnt from further practical experience with these guidelines to new commercial
projects were incorporated into subsequent revised editions and denoted the DURACON Model, the
third and last of which was also adopted by the Norwegian Chapter of PIANC [PIANC Norway/NAHE,
2009]. At the same time, the DURACON Model was expanded to provide the basis for concrete quality
control and quality assurance, which upon completion of the concrete construction provides
documentation of achieved construction quality and compliance with the durability specification. This
model also provides the basis for control of the future chloride ingress as part of the regular condition
assessment and preventive maintenance during operation of the concrete structures (Figure D.1).

Figure D.1: The DURACON Model as a basis for durability design, quality assurance and operation of new major
concrete infrastructure in marine environments [PIANC Norway/NAHE, 2009]

As a convenience for the reader of the current document, a more detailed description of the DURACON
Model is given in the following. It should be noted, however, that it is the creators of the model who are
responsible for the model.

In the DURACON Model, the overall durability requirement is based upon the specification of a given
‘Service period’ before the probability for the onset of steel corrosion exceeds a certain upper level, and
for this level, a probability of 10 % was adopted, which is in accordance with current standards for
reliability of structures (Chapter 3). In order to calculate the probability of corrosion, durability analyses
are carried out, which provide the basis for selecting proper combinations of concrete quality and
concrete cover which meet the required durability or ‘Service period’ for the given concrete structure in
the given environment. As already stated in Chapter 3, however, it should be noted that the DURACON
Model is not being applied for any prediction of service life, but only used as a simple engineering tool
for quantifying and specifying various levels of durability for the given concrete structure in the given
environment.

The durability analyses which are based on the combined calculations of chloride ingress and probability
of corrosion are briefly described in the following, the necessary input parameters for which are also
briefly outlined and discussed.

60
D.2 CALCULATION OF CHLORIDE INGRESS

Basically, rather complex transport mechanisms for ingress of chlorides into concrete exist [Zhang and
Gjørv, 1996 ; Poulsen and Mejlbro, 2006 ; Tang et al., 2012]. In a very simplified form, however, the rate
of chloride ingress can be estimated by use of Ficks 2. Law of Diffusion according to Collepardi et al.
(1970, 1972), in combination with a time dependent chloride diffusion coefficient according to Takewaka
and Mastumoto (1988) and Tang and Gulikers (2007), as shown in eq. D.1 and D.2:

  xC 
C x, t   C S 1  erf  
 2 D(t )  .t 
   (D.1)

In this equation, C(x,t) is the chloride concentration in depth XC after time t, CS is the chloride
concentration at the concrete surface, D is the concrete chloride diffusion coefficient and erf is a
mathematical function.

1 1 
D0  t '   t '   t 0 
D(t )  1         ke
1    t  t   t  (D.2)

In this equation, D0 is the diffusion coefficient after the reference time t0, and t’ is the age of concrete at
the time of chloride exposure. The parameter α represents the time dependence of the diffusion
coefficient, while ke is a parameter which takes the effect of temperature into account according to Kong
et al. (2002):

𝐸𝐴 1 1
𝑘𝑒 = 𝑒𝑥𝑝 [ ( − )] (D.3)
𝑅 293 273 + 𝑇

where exp is the exponential function, EA is the activation energy for the chloride diffusion, R is the Gas
constant and T is the temperature.

The criterion for steel corrosion then becomes:

C(x) = CCR (D.4)

where C(x) is the chloride concentration at the depth of the embedded steel, and CCR is the critical
chloride concentration in the concrete necessary for onset of corrosion.

D.3 CALCULATION OF PROBABILITY

For the structural design of concrete structures, the main objective is always to establish the combined
effects of external loads (S) and the resistance to withstand these loads (R) in such a way that the design
criterion becomes:

RS or R  S  0 (D.5)

When R < S, failure will occur, but since all factors affecting R and S show a high scatter and variability,
all established procedures for structural design have properly taken this into account.

In principle, the durability design takes the same approach as that of the structural design. In this case,
however, the loads (S) are the combined effects of both chloride loads and temperature conditions, while
the resistance to withstand these loads (R) being the resistance to chloride ingress, is the combined
effects of both concrete quality and concrete cover. Although neither S nor R is comparable to that of

61
structural design, the acceptance criterion for having a probability of ‘failure’ less than a given value is
the same.

In Figure D.2, the scatter and variability of both R and S are demonstrated in the form of the two
distribution curves along the y-axis. At an early stage, there is no overlapping between these two
distribution curves, but over time, a gradual overlapping from time t1 to t2 takes place. This increasing
overlapping will at any time reflect the probability of ‘failure’ or the probability for onset of steel corrosion,
and gradually, the upper acceptable level for the probability of ‘failure’ (tSLS) is reached and exceeded.

Figure D.2: The principles of a time dependent reliability analysis

In principle, the probability of failure can be written as:

p(failure) = pf = p(R – S< 0) < p0 (D.6)

where p0 is a measure for failure probability.

For the structural design, the safety of the structures is normally expressed in the form of
reliability, and in the current standards for reliability of structures, an upper level for probability
of failure of 10 % in the serviceability limit state is often specified. In a similar way as that for
structural design, an upper probability level of 10 % for onset of corrosion was also adopted
as a basis for the durability design.

Normally, the above failure function includes a number of variables, all of which have their own
statistical parameters. Therefore, the use of such a failure function requires numerical
calculations and the application of special software. Currently, there are several mathematical
methods available for evaluation of the failure function, such as:

 First-Order Reliability Method (FORM)

 Second-Order Reliability Method (SORM)
 Monte-Carlo Simulation (MCS)

D.4 CALCULATION OF CORROSION PROBABILITY

In principle, the calculation of corrosion probability can be carried out by use of any of the above
mathematical methods. In the following, however, the calculation of corrosion probability is based upon
the modified Fick’s Second Law in eq. D.1 in combination with a Monte-Carlo Simulation. Although such

62
a combined calculation can also be carried out in different ways, a special software DURACON for this
calculation was developed [Ferreira, 2004 ; Ferreira et al., 2004]; this software can be freely downloaded
from the following web page:

www.pianc.no/duracon.php

A Monte-Carlo Simulation can briefly be described as a statistical simulation method where sequences
of random numbers are applied to perform the simulation. When the ingress of chlorides is simulated by
use of eq. D.1, this requires that all the input parameters to this equation can be described by a
probability density function. Once these functions of the various durability parameters of the system are
known, the probability of failure is based on the evaluation of the limit state function for a large number
of simulations. The failure function is then calculated for each outcome. If the outcome is in the failure
region, then the contribution to the probability of failure is obtained.

For the DURACON Software, the concrete cover is chosen as the resistance variable (r), while the depth
of the critical chloride front is chosen as the load variable (s). The probability of failure is then estimated
by the following expression:

1 N
N j 1

p f    I g r j , s j   (D.7)

where N is the number of simulations, I [g(rj,sj)] is the indicator function and g(rj,sj) is the limit state
equation; s represents the environmental load and r represents the resistance of the concrete against
chloride ingress.

The standard error of the above calculation can be estimated by the following expression according to
Enevoldsen and Sørensen (1992):

p f 1  p f 
s (D.8)
N
from which it can be seen that the accuracy of the Monte-Carlo Simulation mainly depends on the
number of simulations.

Primarily, the above calculation of corrosion probability provides the basis for durability design of new
concrete structures. As a result, it is possible to specify a certain ‘Service period’ before 10 % probability
of corrosion is reached. For the given environmental exposure, requirements to both concrete quality
(chloride diffusivity, RCM) and concrete cover are then established. Upon completion of the concrete
construction work, new calculations of corrosion probability are then carried out (Chapter 5). In this case,
the probability calculations are used as a basis for documenting the achieved construction quality and
compliance with the durability specification. During operation of the structures, calculations of corrosion
probability are further used as a basis for condition assessment and preventive maintenance of the
structure as described in Chapter 6.

For all the above types of probability calculation, certain input parameters to the calculations are needed.
For the durability design, all the necessary input parameters are briefly described and discussed in the
following. For the documentation of achieved construction quality and compliance with the specified
durability, however, the input parameters are described and discussed in Chapter 5, while for the
condition assessment and preventive maintenance during operation of the structure, the input
parameters are described and discussed in Chapter 6.

63
D.5 INPUT PARAMETERS

D.5.1 GENERAL

In general, the durability design should always be an integral part of the structural design of the given
structure. Already at an early stage of the design, therefore, a certain ‘Service period’ should be specified
before 10 % probability of corrosion is reached. For the given environmental exposure, the durability
analysis then provides the basis for selecting a proper combination of concrete quality and concrete
cover. Before the final requirements to concrete quality and concrete cover are given, however, it may
be necessary to carry out several calculations for various combinations of possible concrete qualities
and concrete covers.

For all calculations of corrosion probability, proper information about the following input parameters is
needed:

 Environmental loading
- Chloride loading (CS)
- Age at chloride loading (t’)
- Temperature (T)
 Concrete quality
- Chloride diffusivity (D)
- Time dependence factor (α)
- Critical chloride content (CCR)
 Concrete cover (XC)

It should be noted that the above input parameters may have different distribution characteristics. If
nothing else is known for the distribution of the input parameters to the durability design, a normal
distribution with a coefficient of variation between 0.1 and 0.2 may be assumed. For the documentation
of achieved construction quality, the input parameters on both chloride diffusivity and concrete cover
are based on the obtained data from the regular quality control during concrete construction. Since these
data may occasionally have a high scatter and variability, other probability distributions for these input
parameters such as beta distribution may then be assumed. The same is true for the durability analyses
carried out as a basis for the condition assessment during operation of the structures. In this case, the
input parameter on chloride diffusivity is based on the observed data on the real chloride ingress taking
place, which may also have a high scatter and variability.

When a normal distribution of the parameters is assumed, the mean value μ and the standard deviation
σ are input parameters, while for a beta distribution, the parameters q and r must be calculated and
used as input for the variables:

(D.9)

(D.10)

In the following, some general guidelines for determination and selection of the above input parameters
to the durability design are given.

D.5.2 ENVIRONMENTAL LOADING

Chloride loading, CS

For all concrete structures in the marine environment, the chloride loading is normally defined as the
accumulated surface chloride concentration on the concrete surface (CS) after some time of exposure
as shown in Figure D.3. This chloride ingress curve is the result of a regression analysis of at least six

64
observed data on the chloride ingress after a given time of exposure and curve fitting to Ficks 2. Law.
The surface chloride concentration (CS) which is normally higher than the maximum observed chloride
concentration in the surface layer of the concrete (Cmax), is primarily the result of the local environmental
exposure, but both concrete quality, geometrical shape of the structure and height above water also
affect the accumulation of this surface chloride concentration. For all concrete structures, therefore, the
accumulated surface chloride concentration typically shows a very high scatter and variability. For the
durability analysis, however, it is important to estimate and select a proper value of the surface chloride
concentration (CS) which is as representative as possible for the most exposed and critical parts of the
structure. In some cases, it may be appropriate to select different chloride loads for different parts of the
structure and then carry out separate probability calculations for the various parts of the structure.

Figure D.3: Definition of the surface chloride concentration (CS) based on a regression analysis of observed data
on chloride ingress

For a new concrete structure, it may not be easy to estimate and select a proper value for the chloride
loading as generally described above. If possible, therefore, data from previous field investigations of
similar types of concrete structures in similar types of environment should be applied. For the individual
concrete structure, experience shows that the surface chloride concentrations (CS) successively
accumulate over a certain number of years before they tend to level out to fairly stable values for the
given environmental exposure.

Although the selection of chloride loads for a new concrete structure should preferably be based on local
experience from similar concrete structures exposed to similar environments, general experience
available from the literature may also provide a basis for the selection of a proper chloride loading.
Based on general experience from marine concrete structures reported in the literature, some general
guidelines for estimation of chloride loading are given in Table D.1. Since the data on chloride
concentrations are often given in % by weight of concrete, a general conversion diagram to chloride
concentrations in % by weight of cement is also given (Figure D.4).

CS (% by wt. of cement)
Environmental loading
Mean value Standard deviation
High 5.5 1.3
Average 3.5 0.8
Moderate 1.5 0.5

Table D.1: General guidelines for estimation of chloride loading (CS) on concrete structures in the marine
environment

65
 Figure D.4: Conversion diagram for estimating chloride concentrations in % by weight of cement based on % by
weight of concrete with various cement contents [Ferreira, 2004]

Age at chloride loading, t’

Since the resistance of the given concrete to chloride ingress very much depends on the degree of
hydration of the given type of concrete, the age of the concrete at the time of chloride loading (t’) is also
a very important parameter for the assessment of chloride ingress. A proper selection of this input
parameter further depends on both local curing conditions and type of construction procedure.

Temperature, T

For a given concrete in a given environment, the rate of chloride ingress also very much depends on
the temperature as shown in eq. D.3. Based on local information on prevailing temperature conditions,
data on average annual temperatures may be used as a basis for the selection of this input parameter.

D.5.3 CONCRETE QUALITY

Chloride diffusivity, D

Although several methods for testing the chloride diffusivity (D) of concrete exist [Tang et al. 2012], the
Rapid Chloride Migration (RCM) method [AASHTO, 2003] was adopted for the DURACON Model; this
is the only method which can be carried out very rapidly independent of concrete age. Since the RCM
method provides a very strongly accelerated test, however, it was already stated in Chapter 3 that this
chloride diffusivity should only be considered as a simple relative index. Although this index does not
reflect the long-term capacity of the concrete for chloride binding, it reflects both the porosity and
permeability of the concrete as well as the ion mobility in the pore solution of the concrete. Hence, it
reflects both the concrete mixture’s resistance to chloride ingress as well as its general durability
properties. Further information about the testing of the RCM diffusivity is given in Appendix G.

For most durability analyses, the 28-day RCM diffusivity (D28) is normally used as input parameter to
the durability design in the same way as the 28-day compressive strength (f28) is used as input parameter
to the structural design. However, additional durability analyses based on chloride diffusivities obtained
after longer curing periods can also be carried out, and this may be appropriate if some of the possible
types of concrete are based on binder systems which hydrate very slowly such as those based on fly
ash. For the regular concrete quality control during concrete construction and documentation of
compliance with the specified durability, however, the testing is normally based on the 28-day chloride
diffusivity (Chapter 5).

As a basis for a general assessment of the resistance to chloride ingress of various types of concrete
based on the 28-day RCM diffusivity, some general values are shown in Table D.2.

66
 Chloride diffusivity, D28 m2/s x 10-12 Resistance against chloride ingress
>15 Low
10-15 Moderate
5-10 High
2.5-5 Very high
< 2.5 Extremely high
Table D.2: Resistance to chloride ingress of various types of concrete based on the 28-day diffusivity RCM
[Nilsson et al., 1998]

Time dependence factor, α

Since the chloride diffusivity is a time-dependent property of the concrete, this time dependence (α) is
also a very important quality parameter generally reflecting how the chloride diffusivity of a given
concrete in a given environment develops over time. In order to select a proper α-value, an empirical
value for the given type of concrete in the given type of environment is normally used as input parameter
to the durability analyses.

For new concrete structures, therefore, the same problem for selecting a proper α-value exists as that
already discussed for the selection of a proper chloride loading (CS). Again, current experience from
field investigations of similar concrete structures in similar environments may provide a basis for
selecting the α-value. Also, information from long-term field tests with similar types of concrete in similar
environments may be available from the literature. Based on such experience and information, some
general guidelines for selecting a proper α-value are given in Table D.3. This table shows some α-values
obtained for various types of concrete based on various binder systems exposed to the tidal and splash
zone of marine environments [Mangat and Molloy, 1994 ; Thomas and Bamforth, 1999 ; Thomas et al.,
1999 and Bamforth, 1999]. Although combinations of the various types of cement with supplementary
cementitious materials such as silica fume always will reduce the chloride diffusivity, current experience
indicates that Table D.3 may still be used as a general and rough basis for estimation of a proper α-
value.

α-value
Concrete based on various types of cement
Mean value Standard deviation
Portland cements 0.40 0.08
Blast furnace slag cements 0.50 0.10
Fly ash cements 0.60 0.12

Table D.3: General guidelines for estimation of α-values for tidal and splash zone exposure of concrete structures
in marine environments

Since the above empirical values for time dependence of the chloride diffusivity primarily are based on
experience from older structures, it may be difficult to estimate proper α-values for durability design of
new concrete structures based on new types of high-quality concrete and new types of cement and
binder system. Also, it may be argued that the chloride diffusivity of the concrete does not continue to
be reduced throughout the service life of the structure, so therefore, a certain truncation of the chloride
diffusivity by time should also be assumed. However, the DURACON Model is not being applied for any
prediction of service life, but primarily used in order to provide a simple basis for comparing and selecting
one of several technical solutions for the durability design. The assumption of a constant value for the
time dependence factor (α) by time may therefore be acceptable.

Critical chloride content, CCR

As already discussed in Chapter 2, even very small chloride concentrations in the pore solution of a
concrete are able to destroy the passivity of embedded steel and hence start corrosion. Since a number
of factors affect this depassivation, the critical chloride concentration in the pore solution may also vary
within wide limits [Angst et al., 2009 ; Angst, 2011]. Due to the very complex relationship between the

67
total chloride content in the concrete and the critical chloride concentration in the pore solution, it is not
possible to give any general values for critical chloride contents in concrete. When certain values for the
critical chloride content nevertheless are given in existing concrete standards and recommendations,
this is based only on empirical information on total chloride contents in the concrete which may give a
certain risk for start of corrosion. For traditional carbon steel, some empirical values are shown in Table
D.4. However, whether steel corrosion will develop or not also depends on other corrosion parameters
such as oxygen availability and electrical resistivity of the concrete. Thus, the risk for development of
corrosion may be very low both in submerged concrete due to very low oxygen availability and in very
dry concrete due to ohmic control of the corrosion process (Figure D.5).

Chloride content (%)

Risk of corrosion
by wt. of cement by wt. of concrete1)
> 2.0 > 0.36 Certain
1.0-2.0 0.18-0.36 Probable
0.4-1.0 0.07-0.18 Possible
< 0.4 < 0.07 Negligible
1) 3
Based on 440 kg/m of cement

Table D.4: Risk for development of corrosion of carbon steel depending on total chloride content
[Browne et al., 1980]

Figure D.5: Qualitative relationship between critical chloride content (CCR), environmental conditions and quality
of concrete [CEB, 1992]

Based on empirical information from a wide range of concrete qualities and moisture conditions, an
average value for critical chloride content of 0.4 % by weight of cement for reinforcement based on
carbon steel is often specified in current concrete standards. Since such an empirical value also is based
on previous experience with old, moderate concrete qualities and pure portland cements; new types of
high-quality concrete based on new types of binder system may show higher values of critical chloride
content. If nothing else is known, however, an average value of 0.4 % with a standard deviation of 0.1
% by weight of cement may still be selected as input parameter to the durability analysis.

For more corrosion-sensitive types of high-quality steel, an average value of 0.1 % with a standard
deviation of 0.03 % may be selected. For various grades of stainless steel reinforcement, however, the
critical chloride content may typically vary from 2.5 to 3.5 % by weight of cement, but grades with
threshold values of up to up to 5 to 8 % are also available [Bertolini et al., 2004].

68
D.5.4 CONCRETE COVER, XC

In current concrete standards such as shown in Appendix C, requirements to both minimum concrete
cover (XC,min) and tolerances are given for the given environment. Thus, the nominal concrete cover
(XC,N) is always specified with a certain value of tolerance (ΔXC), and different values for ΔXC may be
specified. For a tolerance of ± 10 mm, the minimum requirement to concrete cover then becomes:

XC,min = XC,N - 10 (D.11)

Although the specified concrete cover primarily gives the required concrete cover to the structural steel,
additional mounting steel for ensuring the position of the structural steel during concrete construction is
also often applied. Since the penetrating chlorides do not distinguish between structural and mounting
steel, however, the nominal concrete cover should preferably be specified for all embedded steel
including the mounting steel in order to avoid any cracking of the concrete cover due to premature
corrosion. As part of the structural design, great efforts to avoid any cracking of the concrete are always
made. Cracking of the concrete cover caused by corroding mounting steel may represent the same type
of weakness as that caused by any other types of cracking. Instead of using mounting steel which can
corrode, mounting systems based on non-corroding materials should preferably be applied.

If it is assumed that 5 % of the reinforcing steel has a concrete cover less than XC,min, the durability
analysis can be based on an average concrete cover of XC,N with a standard deviation of ΔXC/1.645.
Then, the effect of increased concrete cover beyond that required in current concrete standards can be
quantified. For the documentation of achieved construction quality as described in Chapter 5, however,
the durability analyses must always be based on the obtained values both for concrete cover and
standard deviation from the regular quality control during concrete construction.

69
APPENDIX E: STADIUM® MODEL

E.1 OVERVIEW
As a convenience for the reader of the current report, a brief outline of the STADIUN® Model is given in
the following. It should be noted, however, that it is the creators of the model who are responsible for
the model.

The software for the STADIUM® Model was developed to model the coupled transport of fluids and
chemical species in concrete exposed to aggressive environmental conditions. STADIUM® can be used
to simulate the ingress of external contaminants within concrete as well as the leaching of chemical
species initially present in the material’s pore solution to the external environment. The software
simultaneously tracks all modifications to the microstructure of concrete, e.g. formation of chloride-laden
hydrates, calcium hydroxide dissolution, C-S-H decalcification. The model also takes into account the
specific nature of cementitious materials by accounting for the high pH of the pore solution, the influence
of continuous hydration of the transport properties of concrete (the so-called aging effects) as well as
multi-ionic coupling. Depending on the environment, other alterations can be reproduced by the model,
such as sulfate exposure. Fully validated through extensive laboratory testing and field trials,
STADIUM® has been used in over 150 major engineering projects around the world. STADIUM® is the
model specified by the United-States Department of Defense to evaluate the service-life of concrete
waterfront structures. It is integrated in the Unified Facilities Guide Specifications (UFGS) document that
provides a complete methodology for the evaluation of concrete mixes for waterfront structures (see
section 2 of this report). Over the past five years, the UFGS has been used for the construction of
numerous major marine structures in the United-States and abroad.

The model was designed specifically for cementitious materials. It considers the electrodiffusion
coupling between ions and chemical activity gradients, key features since the pores of hydrated cement
pastes are filled with a high pH, high alkaline solution and that typical exposure cases involve highly
concentrated solutions (e.g. seawater, deicing salts). The model accounts for dissolution/precipitation
chemical reactions resulting from interactions between species and highly reactive hydrated paste. It
also allows considering the formation of solid solutions between AFm phases and Friedel’s salt, which
is the basic chloride binding mechanism in cementitious materials.

Contrary, to the first generation of chloride penetration models based on Fick’s second law,
STADIUM®’s mechanistic approach offers a more precise description of complex transport mechanisms
and provides more realistic long-term service-life predictions. The model was published in many peer-
reviewed journal and conference proceedings [Samson and Marchand, 2006 ; Samson and Marchand,
2007a and b].

E.2 DETAILS

The model is based on a Sequential Non Iterative Algorithm (SNIA) that separately solves the transport
equations and the chemical equilibrium relationships (Figure E.1). The transport equations (i.e. mass
conservation equations for the species, moisture flow, thermal conduction, and electrodiffusion coupling)
are discretized using the finite element method and solved simultaneously using a coupled algorithm.
The model accounts for both physical and chemical interactions. Physical chloride binding is modelled
by a Freundlich isotherm and directly incorporated in the transport equations. All chemical reactions,
including chloride binding due to formation of Friedel's salt, are treated by a separate chemical module.

70
 Figure E.1: General overview of STADIUM®

The transport of chemical species is described by the extended Nernst-Planck equation applied to
unsaturated materials. This equation accounts for the electrical coupling between ionic fluxes, chemical
activity effects, transport of species due to water content gradient and temperature effects:

𝜕𝑐𝑖𝑏 𝜕(𝑤𝑐𝑖 ) 𝐷𝑖 𝑧𝑖 𝐹
𝜌 + − div(𝐷𝑖 𝑤grad(𝑐𝑖 ) + 𝑤𝑐𝑖 𝑔𝑟𝑎𝑑(𝜓)
𝜕𝑡 𝜕𝑡 𝑅𝑇

𝐷𝑖 𝑐𝑖 ln(𝛾𝑖 𝑐𝑖 ) (E.1)
+𝐷𝑖 𝑤𝑐𝑖 grad(ln 𝛾𝑖 ) + 𝑤grad(𝑇) + 𝑐𝑖 𝐷𝑤 𝑔𝑟𝑎𝑑(𝐻)) = 0
𝑇

where 𝑐𝑖 is the concentration of species i in the pore solution, 𝑐𝑖𝑏 is the chloride bound to C-S-H due to
physical interaction, w is the volumetric water content, 𝜌 is the density of the material, 𝐷𝑖 is the diffusion
coefficient, 𝑧𝑖 is the valence number of species, F is the Faraday constant, 𝜓 is the electrodiffusion
potential, R is the ideal gas constant, T is the temperature, 𝛾𝑖 is the activity coefficient, H is the relative
humidity, and 𝐷𝑤 is the water diffusivity. The activity coefficients in the model are evaluated on the basis
of the Harvie, Moller and Weare (HMW) implementation of Pitzer's ion interaction model.

The physical binding term was estimated from binding experiments performed on hydrated C 3S pastes
exposed to different chloride concentrations. This term is zero for all ionic species except chloride, for
which 𝑐𝑖𝑏 is given by:
𝑏
𝑐Cl = 𝑘𝑐𝑖𝑛 (E.2)

where k and n are fitting parameters. As illustrated on Figure E.2, the binding experiments were
performed at two different pH conditions. A linear interpolation between these two values allows
estimating the physical binding at any pH.

71
 Figure E.2: Physical binding isotherms measured at different pH values

The electrodiffusion term in equation (E.1), involving the potential 𝜓 is mainly responsible for maintaining
the electroneutrality of the pore solution. Its role is to balance ionic mobilities so that there is no net
accumulation of charge at any location in the pore solution. In cementitious materials, where pore
solution concentrations are high (pH in the range 13-14), this term can have a significant influence on
the ingress rate of contaminants. To solve the diffusion potential 𝜓, the ionic transport equation is
coupled to Poisson's equation, which couples the electrodiffusion potential in the material to the ionic
profile distributions:
𝑁
𝐹
div(𝜏𝑠 𝑤grad(𝜓)) + 𝑤 ∑ 𝑧𝑖 𝑐𝑖 = 0 (E.3)
𝜀
𝑖=1

where 𝜀 is the permittivity of water, 𝜏𝑠 is the intrinsic tortuosity of the material (described later) and N is
the number of species in the pore solution. The intrinsic tortuosity is related to the diffusion coefficient
by the following relationship:
𝐷𝑖 = 𝜏𝑠 𝐷𝑖𝑜 (E.4)
where 𝐷𝑖𝑜 is the self-diffusion coefficient of species i. Self-diffusion coefficient values can be found in the
literature.

To account for moisture flow induced by capillary potential gradients in unsaturated materials, the
previous equations are coupled to:

𝜕𝑤 𝜕𝐻 𝜕𝑤 𝜕𝑇
+ − div(𝐷𝐻 grad(𝐻) + 𝐷𝑇 grad(𝑇)) = 0 (E.5)
𝜕𝐻 𝜕𝑡 𝜕𝑇 𝜕𝑡

where 𝐷𝑚𝐻 and 𝐷𝑚𝑇 are the moisture transport coefficients, given by:

𝑔
𝑘𝑠 𝑘𝑟𝑙 𝜌𝑙 𝑅 𝑇 𝜃𝑔 𝜏𝑠 𝜏𝑟 𝐷𝑣𝑜 𝑀𝑤 𝑝𝑣𝑠 1
𝐷𝐻 = + (E.6)
𝜇𝑀𝑤 𝐻 𝜌𝑙 𝑅 𝑇

𝑔
𝑘𝑠 𝑘𝑟𝑙 𝜌𝑙 𝑅 𝑇 𝑑𝑝𝑣𝑠 𝜃𝑔 𝜏𝑠 𝜏𝑟 𝐷𝑣𝑜 𝑀𝑤 𝑝𝑣𝑠 𝐻
𝐷𝑇 = (ln(𝐻) − 𝑠 )+ (E.7)
𝜇𝑀𝑤 𝑝𝑣 𝑑𝑇 𝜌𝑙 𝑅 𝑇2
where 𝑘𝑠 is the intrinsic permeability of the saturated material, 𝑘𝑟𝑙 is the relative permeability of the liquid
phase, 𝜌𝑙 is the liquid phase density, µ is the dynamic viscosity of the liquid phase, 𝑀𝑤 is the molar mass

72
of water, 𝜃𝑔 is the volumetric gas phase content, 𝐷𝑣𝑜 is the vapor self-diffusion coefficient, 𝑝𝑣𝑠 is the
𝑔
saturation vapor pressure, and 𝜏𝑟 is the relative tortuosity of the gaseous phase.

The terms 𝜕𝑤 ⁄𝜕𝐻 and 𝜕𝑤 ⁄𝜕𝑇 in equation (E.5) are associated with the moisture isotherm of the
material. In STADIUM®, the following relationship describes the moisture isotherm parameter:
𝜙
𝑤= (E.8)
𝛽𝜙(𝐻 𝛿 − 1) + 1
where 𝜙 is the porosity of the material, and 𝛽 and 𝛿 are material parameters that describe the
relationship between w and H.

Finally, the temperature distribution in the material is calculated from the classical heat condition
equation:
𝜕𝑇
𝜌𝐶𝑝 − div(𝜅grad(𝑇)) = 0 (E.9)
𝜕𝑡
where 𝐶𝑝 is the specific heat capacity of the material, and 𝜅 is the heat conductivity.

The key material parameter that determines the rate of ingress of chloride and other contaminants in a
concrete structural element is the diffusion coefficient 𝐷𝑖 (see equation E.1). This parameter is influenced
by many factors: temperature, saturation level of the material and age. It should be noted that the effect
of age, due to the hydration of cementitious materials, is often represented by the time-dependent aging
function:
𝑡𝑜 𝛼
𝐻(𝑡) = ( ) (E.10)
𝑡
where t is the age of the structure, 𝑡𝑜 is a reference value for the diffusion coefficient, generally taken at
28 days, and 𝛼 is a parameter that characterises the decreasing rate of the diffusion coefficient. This
function multiplies 𝐷𝑖 . This function however keeps decreasing with time, even after many years of
service-life, a fact not supported by data. To correct this, the following function was implemented in
STADIUM®:
𝑎
𝐻(𝑡) = (E.10)
1 + (𝑎 − 1)𝑒 −𝛼(𝑡−𝑡𝑜 )
where a represents the lower value reached by the function (e.g. 0.15 < a < 0.5 for fly ash mixtures).
Lab data have shown that after 2 years of curing, cementitious materials can be considered stable, and
their transport properties stop decreasing.

The system of nonlinear equations (E.1), (E.3), (E.5) and (E.9) is solved using the Newton-Raphson
method with all equations solved simultaneously. The spatial discretization of this coupled system is
based on the finite element approach using the standard Galerkin procedure. An Euler implicit scheme
is used to discretize the time-dependent part of the model.

The second module in STADIUM® consists in a chemical equilibrium code (see Figure E.1). Following
the transport step, the chemical equilibrium module verifies, at each node of the mesh, the equilibrium
between the concentrations and the solid phases of the hydrated cement paste, e.g. calcium hydroxide,
calcium silicate hydrates, ettringite, and monosulfates. The equilibrium of each phase is modelled
according to the classical law of mass action:
𝑁
𝜈 𝜈𝑚𝑖
𝐾𝑚 = ∏ 𝑐𝑖 𝑚𝑖 𝛾𝑖 (E.12)
𝑖=1

where 𝐾𝑚 is the equilibrium constant (or solubility constant) of the solid m, N is the number of dissolved
species, 𝑐𝑖 is the concentration of species i, 𝛾𝑖 is its chemical activity coefficient, and 𝜈𝑚𝑖 is the
stoechiometric coefficient of the ith ionic species in the mth mineral. If the solution is not in equilibrium
with the paste, solid phases are either dissolved or precipitated to restore equilibrium. The pore solution
is thus adjusted to enforce the equilibrium relationships (E.12) of the mineral phases. Solid phases can

73
also be formed when aggressive species penetrate into the porous network of the material, e.g.
ettringite, gypsum in the case of sulfate exposure.

The calculation of the chemical activity coefficients relies on the same Pitzer implementation that is
called in the transport module. Different databases can be used to provide the chemical equilibrium
constants. STADIUM® relies on the CEMDATA07 database.

After the pore solution concentrations are modified, the solid phases are also corrected according to:

𝑡 𝑡−1
𝑤𝑋𝑚 𝑀𝑚
𝑆𝑚 = 𝑆𝑚 − (E.13)
𝜌
𝑡
where 𝑆𝑚 is the amount of a given solid phase, t indicates the time step, 𝑀𝑚 is the molar mass of the
solid m, 𝑋𝑚 represent the amount of a given solid phase that has to dissolve to reach equilibrium, and
𝜌 is the density of the material.

Some solid phases do not act as pure mineral but instead form solid solutions. This is the case for the
formation of the chloride-AFm solid compound called Friedel's salts, which is the main reaction product
formed upon chloride penetration in cementitious materials. The equilibrium is calculated by assuming
that the AFm end-member acts as pure minerals and that Friedel's salts following the relationship:
(Cl)2 [AFm]
𝐾𝑆𝑆 = (E.14)
(SO4 ) [FS]
where 𝐾𝑆𝑆 is the equilibrium constant of the solid solution, (Cl) and (SO 4) are the activities of chloride
and sulfate in the pore solution, [AFm] represents the mole fraction of the AFm end-member, and [FS]
represents the mole fraction of the Friedel's salt end-member.

E.3 STADIUM® INPUT PARAMETERS

The previous sections emphasised the large number of parameters needed to perform a simulation. As
illustrated on Figure E.1, the parameters can be divided in three main categories:

 Geometry: dimensions of modelled structural elements. Since most cases can be simplified to 1D,
the characteristic length of structural elements must be provided (e.g. deck thickness). As
STADIUM® is based on the finite element method, information on the discretisation of the spatial
domain is also needed.

 Exposure conditions (environment): temperature, relative humidity, composition of solution in

contact with the structure. In the case of seawater, its composition is commonly expressed as the
salinity, which must be converted into individual species concentration. For example, a 32 ppt
salinity corresponds to (all values in mmol/L): 426.1 Na +, 9.0 K+, 25.6 SO42–, 9.3 Ca2+, 48.6 Mg2+,
3.8 HCO3–, and 495.8 Cl–. Depending on the exposure, all these parameters can be time-dependent.

 Material properties: they can be divided into two categories. Parameters are associated with the
transport equations (transport properties), while others are related to the chemical module. The
transport properties are summarized in Table E.1. Parameters needed by the chemical module are
provided in Table E.2. All other parameters listed in the previous section and not found in these
tables are either constants that can be found in textbooks or parameters for which a relationship
was found in the scientific literature and implemented in the model (e.g. relative permeability 𝑘𝑟𝑙 and
saturation vapor pressure 𝑝𝑣𝑠 ).

74
 Parameter
Volume of permeable voids
(porosity)
Diffusion coefficient
Intrinsic permeability
Moisture isotherm
Aging function
Thermal conductivity
Specific heat capacity
Symbols Test method Note
𝜙 ASTM C642 Also provides mixture density.
Modified version of ASTM C1202. Also provides
𝐷𝑖 Migration test
intrinsic tortuosity 𝜏𝑠 (see eq. E.4)
𝑘𝑠 ASTM C1792
The test is performed on a thin (10mm) sample.
𝛽, 𝛿 ASTM C1792
A model is then used to evaluate 𝛽 and 𝛿.
Tests performed at different curing duration can
a, 𝛼 Migration test
be used.
No standard test method for concrete. Since it
𝜅 – varies little from one mix to another, values from
the literature can be used.
No standard test method for concrete. Since it
𝐶𝑝 – varies little from one mix to another, values from
the literature can be used.

Table E.1: Transport module parameters

Parameter Symbols Test method Note

Initial solid phase content
Initial pore solution
The initial content in portlandite, C-S-H, AFm,
AFt, and other phases initially present in the
𝑆𝑚 – hydrated cement paste can be estimated on the
basis of the mixture proportions and cement
chemical analysis.
Although it is possible to physically extract
Pore solution solution from concrete samples to measure this
𝑐𝑖 parameter, it is expensive and inaccurate.
extraction
Models based on mixture proportions and
cement chemical analysis can be used.

Table E.2: Initial chemical composition of the hydrated cement paste

E.4 INTERFACING STADIUM® WITH EXTERNAL SOFTWARE

Parameters are provided to STADIUM® through regular text or .XML files. It is possible to launch
STADIUM® in interactive mode, where the model asks users for filenames (input and output files). It is
also possible to use STADIUM® in batch mode. In this case, filenames are provided directly on a
command line. This mode allows coupling STADIUM® with external software and graphical user
interfaces.

Over the years, the model has been wrapped in different graphical layers and embedded in external
programs. Here is a short list of key implementations of STADIUM®:

 STADIUM® v2.99: the main graphical user interface of STADIUM®. It allows setting up simulations
using predefined scenarios in chloride-laden environments and analysing output to determine time-
to-initiate corrosion of steel reinforcement. Material properties and exposure condition databases
are embedded in the interface.

 Kademuur Modellering Systeem (KMS): KMS is a waterfront structure management software

developed for the specific needs of the Port of Rotterdam. The web-based software calls
STADIUM® when inspection data triggers calculations of chloride ingress in specific quay wall
sections. The overall objective of the KMS is to improve maintenance of quay wall sections.

 Cementitious Barriers Partnership (CBP) Toolbox: the US Department of Energy funded CBP
project focuses on the development of tools dedicated to the prediction of service-life of concrete
structures used for nuclear waste storage. STADIUM® is used in this specific application to estimate
the durability of concrete in contact with highly corrosive grout materials filled with sulfate.

75
E.5 USE OF STADIUM® FOR MIX SELECTION AND QUALIFICATION

In 2010, the US Department of Defense introduced its construction guide for new marine concrete
structures, the Unified Facilities Guide Specifications (UFGS-03 31 29). For more information on the
UFGS see http://www.wbdg.org/ccb/browse_org.php?o=70. The document describes the steps needed
to select concrete mixtures to achieve pre-defined service-life requirements for structural elements in
highly concentrated, chloride-laden environments. It replaces traditional prescriptive methods with
performance and design specifications. It is in line with performance specification language
progressively integrated in modern construction codes. For example, language for performance
specification was incorporated in Canadian standard CSA A23.1-04, and defined as a method for
specifying a construction product in which the final outcome is given in mandatory language, in a manner
that the performance requirements can be measured by accepted industry standards and methods.

The UFGS document describes the testing protocol for evaluating concrete mix durability parameters.
The guide also specifies the use of STADIUM® in order to determine the long-term durability of mixtures
and their service-life expectancy.

In the UFGS, the service-life is defined as the period of time before major repairs are required. In
chloride-laden environment, it is associated with corrosion initiation of steel reinforcement. The following
paragraphs summarize the steps detailed in the UFGS document to find concrete mixtures that can
achieve a specified service-life.

The process of mixture selection starts with the preparation of laboratory concrete. The objective is to
prepare samples to evaluate the following parameters:

 Volume of permeable voids (porosity), measured in accordance with ASTM C642 test procedure:
Standard Test Method for Density, Absorption, and Voids in Hardened Concrete

 Diffusion coefficients, measured on the basis of the migration test, a modified version of the ASTM
C1202 procedure: Standard Test Method for Electrical Indication of Concrete's Ability to Resist
Chloride Ion Penetration

 Water permeability, measured on the basis of ASTM procedure ASTM C1792: Standard Test
Method for Measurement of Mass Loss versus Time for One-Dimensional Drying of Saturated
Concretes

A minimum of three trial batches must be prepared to establish tolerance limits of diffusivity. Concrete
from each batch is tested with the listed procedures after 28 days of curing in moist conditions (limewater
immersion, 100 %-RH moist room). A statistical analysis of the results allows estimating the expected
effective chloride diffusion coefficient, which is the effective diffusion coefficient not expected to be
surpassed for more than 10 % of batches, at the 90 % confidence level. This tolerance limit calculates
the upper limit on the expected effective diffusion coefficient based on the observed mean and variance
of laboratory testing. Additional laboratory testing will generally decrease this maximum expected value
because more data leads to higher confidence in the measurements. A second test series is performed
after 90 days of curing. The 28-day and 90-day data are used to calculate the aging function that
simulates the improvement of diffusive properties over time due to cement hydration. The methodology
assumes that after 90 days, no additional improvement of properties should be considered when
modelling service-life expectancy.

Once data from laboratory tests become available, simulations are performed with STADIUM® to
estimate the time-to-initiate corrosion and assess the capacity of the mixtures to reach the predefined
service-life target. Beside data generated from the laboratory tests, the simulations must account for
structural element geometry, local exposure conditions and steel type. A series of five simulations is
performed. The first simulation is based on the maximum expected chloride diffusion coefficient
calculated on the basis of laboratory data. In addition to this, simulations with ±25 % and ±50 % of the
effective diffusion coefficient are performed. The results from the five simulations allow assessing
whether a given mixture reaches the service-life expectancy. The simulations results are also used to
determine the highest acceptable chloride diffusivity value that allows reaching the target service-life
requirement. At the end of simulations, possible mixtures candidates are selected among those which
cleared the service-life requirement. The simulation process is illustrated on Figure E.3.

76
 Figure E.3: Chloride ingress calculations for a concrete mix with 20 % Type C fly ash used in a concrete deck
exposed to chloride (Rebar depth: 4 cm, Rebar type: stainless steel)

During concrete production, the UFGS methodology plans for regular concrete testing to assess the
quality of on-site production. During production, concrete samples collected from production sites must
be collected and placed in cylinders for 28-day moist curing. At this point, samples must be tested for
volume of permeable voids (ASTM C642) and diffusion coefficients (modified ASTM C1202) to make
sure that the effective chloride diffusion coefficient is below the acceptable maximum limit calculated
from numerical modelling.

E.6 COUPLING STADIUM® WITH PROBABILISTIC ANALYSES

Analysis and design of concrete structures rely on the concept of calculating the external loads applied
upon the system, which is noted the solicitation (S), and the resistance (R) of the system to these loads.
Failure occurs when the solicitation is equal to or greater than the resistance of the system. These
principles apply equally to structural design and durability analysis.

Analysing the durability of a system using a probabilistic process allows for the calculation of the
evolution of risk associated with a degradation mechanism. Both the solicitation and the resistance of a
system are susceptible to intrinsic, environmental, and material variability. Probabilistic analysis
considers the probability density associated with both the solicitation and resistance. The risk is
calculated as the probability that the solicitation of the system will equal or surpass the resistance. This
probability is known as the probability of failure. Risk assessment also allows for multiple optimization
calculations, including life-cycle cost analysis and the determination of risk mitigation alternatives over
the intended service life.

E.6.1 CALCULATION OF CORROSION PROBABILITY

The calculation of the corrosion probability requires the use of mathematical calculations for the failure
function. It is solved in two parts. First, the solicitation and resistance probability densities must be
calculated for the entire service life of the structure. Then, the probability of failure is calculated by using
one of several methodologies available.

Figure E.4 shows a representative analysis for the calculation of the probability of corrosion over a given
service life. Over time, the chloride content (solicitation, red) distribution will increase at the face of the
reinforcement, eventually overlapping the chloride threshold (resistance, blue) distribution necessary for
the initiation of corrosion. This overlapping is seen in the increase of the risk of corrosion initiation. When
the solicitation curve is entirely below the resistance curve, the risk is null (0 %). As the solicitation
increases, the risk grows. It is a common misconception that the risk is represented by the overlapping

77
area of the curves. As a matter of fact, perfectly overlapping curves with equal means represent exactly
50 % risk. When the solicitation curve is entirely above the resistance curve, the risk is 100 %.

Figure E.4: Probability of failure associated with the initiation of corrosion of a concrete element with consideration
of environmental and material variability in the solicitation, and chloride initiation threshold variability in the
resistance of the system – inlaid graphs shows the probability density of the chloride content and chloride
threshold at specific moments during the service life

E.6.2 DETERMINATION OF PROBABILITY DENSITY OF SOLICITATION

(CHLORIDE CONTENT)

The solicitation probability distribution is calculated with the distribution of input variables of the model.
For the case of the STADIUM® model, these input variables include the environmental conditions (salt
loading, humidity exposure, etc.) and material properties (volume of permeable voids, ionic diffusion
coefficients, water permeability, cement chemistry, etc.). A systematic analysis of material properties for
repeated tests on varying concrete mixes has established guidelines for the distribution types and
variability that can be used for this type of analysis when modelling with STADIUM® [Conciatori et al.,
2014].

For simple mathematical models, algebraic methods can be used to determine the solicitation probability
distribution for the service life of a structure. The propagation of uncertainty in an iterative, nonlinear
finite element models (such as STADIUM®) through algebraic manipulation is not feasible.

For more complex models, numerical integrations methods or Monte-Carlo simulations (MCS) can be
used. Monte-Carlo modelling uses a random number generator to generate values for the input
variables to the model. The inputs are generated following valid input distribution shapes, such as those
available in the literature for the material properties, and known environmental variations for the
environmental conditions. A service life simulation with the generated input values and the response of
the model is stores. This methodology of generating inputs and storing the outputs is run multiple times
until the response shape of the simulation results distribution is stable. The number of simulations

78
necessary to obtain a stable model response shape varies approximately with the square of the number
of probabilistic input variables. As an indication, Monte-Carlo modelling used to generate the solicitation
distribution probability of a specific exposure case, while considering only the material properties as
probabilistic, necessitates approximately 1,000 individual simulations [Conciatori et al., 2014]. For
advanced service life models such as STADIUM®, Monte-Carlo modelling is possible; however, it is
very computationally expensive.

An alternative to MCS is using point estimators to calculate the statistical moments of the model output,
based on the variability of the input parameters. When using appropriate point estimators, such as the
Rosenblueth point estimators [Rosenblueth, 1981], the statistical moments of the model response can
be calculated. The first three statistical moments of a distribution are the mean, variance, and skewness,
which respectively give the location, spread, and asymmetry shape of a distribution. The statistical
moments contain enough information to approximate the complete solicitation probability distribution.

The methodology was validated by comparing the Rosenblueth point estimator methodology to a Monte-
Carlo simulation. This comparison was performed by analysing the 95 % confidence interval of the
solicitation probability distribution of two exposure cases of concrete samples, for which laboratory
experimental validation was available.

These concrete samples were exposed to either constant saturation in a NaCl solution, or to 4 days of
wetting followed by 3 days of drying. The samples were exposed to these conditions for three and six
months. Chloride content profiles were evaluated from the exposed faces after these exposure periods.
Figure E.5 shows the confidence interval (C.I.) of the Monte-Carlo simulation, the Rosenblueth point
estimator method, and the experimental chloride profiles taken from the cores for the wetting/drying
exposure case. Similar results are available for the constant exposure case. It was shown that the
solicitation distribution obtained by the Rosenblueth point estimator method exhibit less than 1 %
difference with those generated by Monte-Carlo simulations, at a much lower computational cost (16
instead of 1,000 service life simulations) [Conciatori et al., 2014].

Figure E.5 (taken from Conciatori et al., 2014)1: 95 % confidence interval of chloride content prediction with
Rosenblueth points estimators, with Monte Carlo verification, and experimental chloride profile points of laboratory
specimens for wetting-drying case after (left) 3 months, and (right) 6 months – the range indicates the thickness of
the milling from which the powder sample was taken [Conciatori et al., 2014].

79
E.6.3 DETERMINATION OF PROBABILITY DENSITY OF RESISTANCE (CHLORIDE
CONTENT THRESHOLD FOR THE INITIATION OF CORROSION AT THE FACE
THE REBAR)

In the probabilistic analysis of corrosion initiation, the resistance distribution function is the chloride
content necessary to initiate corrosion at rebar depth. Two sources of variability must be considered:
the variability of the concrete cover over the reinforcement, and the variability in the chloride content
threshold. The location of the rebar can be determined by fabrication tolerance or through testing the
concrete elements. The determination of the chloride threshold is dependent on the type of rebar, and
specific concrete material chemistry considerations. A specific analysis of most available literature data
has shown that modelling the chloride threshold of black steel for corrosion initiation using a normal
distribution with a mean of 0.5 % and a standard deviation of 0.075 % (chloride content by weight of
binder) gives the most representative analysis.

E.6.4 DETERMINATION OF PROBABILITY OF FAILURE

Once the solicitation and resistance probability density function have been determined, the probability
of failure over the service life can be calculated using different methodologies:

 Monte-Carlo simulations (MCS)

 First-order reliability method (FORM)
 Second-order reliability method (SORM)

Once again, Monte-Carlo simulations can be used to determine the probability of failure. From the
previously recorded solicitations values, resistance values are randomly generated and the probability
of failure is calculated from the proportion of failed cases over the service life. However, the
computational time drawback must be considered, but to a higher degree since the addition of the
resistance variability leads to more simulations necessary to acquire a sable output model.

Second order reliability methods (SORM) are necessary to calculate failure probability when the failure
function is discontinuous or has local minima. This occurs when multiple solicitation or resistance factors
affect the same failure modes. In these cases, the FORM algorithm can remain stuck in a local
optimisation problem and not correctly determine the failure probability. In the case of corrosion initiation,
the failure function is continuous and only has one global minimum.

For probabilistic simulations using the STADIUM® model, the reliability problem is solved using first-
order reliability method (FORM). It is solved using the improved Hasofer & Lind, Rackwitz & Fiessler
algorithm (iHLRF). This methodology relies on the statistical moments and distributions of the solicitation
and resistance distribution functions, which has been determined as shown earlier. The application of
this algorithm has shown good agreement when compared to MCS. Figure E.6 shows the results.

Figure E.6: Probability of failure (corrosion initiation) for two different rebar covers over a 20-year simulation, red
dashed line shows the FORM results with 16 simulations using Rosenblueth point estimators, and the blue circles
shows the MCS results after approx. 2,000 simulations

80
APPENDIX F: Life-365TM MODEL
In the following a brief outline of the Life-365TM Model is given with the permission of the owners. It
should be noted, however, that it is the creators of the model who are responsible for the model.

The software for the Life-365TM Model is designed to estimate the service life and life-cycle costs of
alternative concrete mix designs. It follows a methodology, created by the Life-365 Consortium I and II
groups of companies, that gives research-based estimates of the effects of concrete design, chloride
exposure, environmental temperature, concrete mixes and barriers, and steel types on this service life
and life-cycle cost.

In the following, a quick overview of how Life-365TM is structured and used is given; for further
information, reference is made to the Life-365TM User Manual.

When you first start Life-365TM, you will see splash screen in the main window, a navigation panel on
the left side, which you can use to start new projects, open existing projects, and get help with steps in
the analysis; and a series of tabs at the bottom which allow you to set your default values (for example,
using U.S. or S.I. units of measure) and access on-line help.

To start a new project, select ‘Open new project’ from the left-side navigation panel.

81
When the new project is created, a series of tabs will appear in the main window. To conduct the project
analysis, access each tab, starting from the left side and then working toward the right. In the first tab,
shown above, you set the project name and date, the type of structure (slab or column) and its
dimensions, and the number of alternative concrete mixes to be analysed.

In the second tab, you set the environmental conditions to which your structure will be exposed, in
particular the maximum chloride surface concentration (one of the factors that determines how long it
takes the chloride concentrations on the reinforcing steel to reach critical levels) and monthly variations
in temperature (which affect the diffusivity of the concrete). Life-365 comes with a database of
concentration and temperature values for regions in North America; for other areas, you can input your
own values.

82
The third tab is used to specify the concrete mix for each of your project alternatives. The top of the
screen has a list of concrete mixes (each corresponds to one of the projects); select an alternative from
this list to then select the values for w/cm ratio, SCMs (silica fume, fly ash, and slag), steel, inhibitor,
and barrier used. Once you specify each of the concrete mixes, press the Recalculate service lives
button at the top. This will calculate the initiation period (time for the chlorides to initiate corrosion of the
reinforcing), the propagation period (time between first corrosion and time to repair), and service life (the
sum of initiation and propagation periods). It will also create a series of graphs at the bottom that help
interpret the service life estimates.

Once the concrete service lives have been calculated, the next tab, above, is used to input the costs of
concrete, steel, inhibitors, and barriers, and the amount, cost, and timing of repairs. Because the costs
of concrete itself vary widely, it is strongly recommended that you input your own costs in the upper-left,
Set Concrete Costs table. Once all of these costs have been reviewed and if necessary modified, the
lower panel shows, for each alternative, a timeline of the costs (amount, year of occurrence).

83
Finally, the life-cycle costs tab above allows you to see and compare the life-cycle costs of each project
alternative; it does this through summary bar charts of life-cycle cost; stacked bar charts of component
life-cycle costs (construction, barrier, and repair costs), annual and cumulative present-value and
current-dollar costs. These calculations follow the ASTM standard approaches for estimating life-cycle
costs. You can also conduct sensitivity analysis of some of the factors that affect life-cycle cost.

Life-365TM offers a wide range of ways to print or export results; for example, you can:

 printout a concrete mix report and a life-cycle cost report

 export the project data (by selecting Export project data from the navigation panel)
 copy table data to the clipboard and then paste in Excel or into Word tables; and copy timeline graph
data to clipboard and then paste in Excel or Word

See the User Manual for a more complete description of how to use Life-365TM.

84
APPENDIX G: ADDITIONAL TEST METHODS FOR QUALITY
CONTROL

G.1 GENERAL

As already pointed out in Chapter 5, a large number of test methods and procedures for quality control
of both concrete production and concrete construction are given in current concrete codes and
standards, a thorough review of which is given in the CIRIA Report [CIRIA, 2010]. For a concrete quality
control and quality assurance as that described in Chapter 5, however, an ongoing control of both the
chloride diffusivity (RCM) and the concrete cover must be carried out, the test methods and procedures
for which are briefly described and discussed in the following. If any additional protective measures such
as cathodic prevention or provisions for such a protective measure are also part of the durability
specifications (Chapter 4), an ongoing quality control and quality assurance of the electrical continuity
within the rebar system must be carried as also briefly outlined and discussed in the following.

G.2 CHLORIDE DIFFUSIVITY

G.2.1 GENERAL

Although the duration of the Rapid Chloride Migration (RCM) method (AASHTO, 2003) only takes a few
days, this is not good enough for the regular quality control and quality assurance during concrete
construction. For any porous materials, however, the Nernst–Einstein equation expresses the following
general relationship between the diffusivity and the electrical resistivity of the material (Atkins and De
Paula, 2006):

R T ti
Di   (G.1)
Z 2  F 2  i  ci  

where:

Di = diffusivity for ion i

R = gas constant
T = absolute temperature
Z = ionic valence
F = Faraday constant
ti = transfer number of ion i
i = activity coefficient for ion i
ci = concentration of ion i in the pore water
 = electrical resistivity

Since most of the factors in eq. G.1 are physical constants, the above relationship can for a given
concrete with given temperature and moisture conditions be simplified to:

1
Dk (G.2)


where D is the chloride diffusivity, k is a constant and  is the electrical resistivity of the concrete. Since
the electrical resistivity of the concrete can be tested in a very rapid and simple way compared to that
of the chloride diffusivity, it is primarily a regular quality control of the electrical resistivity of the concrete
which provides the basis for an indirect control of the chloride diffusivity (D28) during concrete
construction [Gjørv, 2003]. Therefore, the above relationship between chloride diffusivity and electrical
resistivity for the given concrete must be established before concrete construction starts. This is done
by producing a certain number of concrete specimens, from which parallel testing of chloride diffusivity
(RCM) and electrical resistivity are carried out at different periods of water curing at 20 0 C, a typical
example of which is shown in Figure G.1.

85
 Figure G.1: A typical calibration curve for an indirect control of the 28-day chloride diffusivity (RCM) based on
electrical resistivity measurements

After the above relationship between the chloride diffusivity and the electrical resistivity of the given
concrete has been established, this relationship is later on used as a calibration curve for an indirect
control of the 28-day chloride diffusivity based on regular testing of the electrical resistivity during
concrete construction. Since the testing of the electrical resistivity is a non-destructive type of testing,
these measurements are carried out as a quick test on the same concrete specimens as that being used
for the regular quality control of the 28-day compressive strength. Of all these control data, any individual
value should never exceed 30 % by that of the specified 28-day chloride diffusivity.

In order to establish the above calibration curve, the measurements of chloride diffusivity are carried out
on three 50 mm thick specimens cut from Ø100 x 200 mm concrete cylinders after approximately 7, 14,
28 and 60 days of water curing. After cutting, the specimens are water saturated according to an
established water saturation procedure before further testing. In parallel, the corresponding
measurements of the electrical resistivity are carried out on three concrete specimens of the same type
as that being used for the regular quality control of compressive strength; these measurements are
carried out on water cured concrete specimens after curing periods as shown above.

It is not the purpose here to describe all the details for the measurements of chloride diffusivity based
on the RCM method; all procedures are described in more detail elsewhere [AASHTO, 2003]. Such
measurements also require special testing equipment and qualified experience which are only available
in professional testing laboratories. However, for a better evaluation and application of the obtained
results, a brief outline of the test method is given in the following.

G.2.2 TEST SPECIMENS

The testing is normally based on 3 x 50 mm cut slices of a concrete cylinder with diameter Ø100 mm.
The 50 mm thick slices are either cut from the separately cast concrete cylinders or from the concrete
cores either removed from the structure under construction or the corresponding dummy elements
produced on the construction site.

G.2.3 TESTING PROCEDURE

Immediately before testing, the test specimens are pre-conditioned according to a standardised water
saturation procedure. Then, the specimens are mounted into rubber sleeves and placed in a container
with a 10 % NaCl solution as shown in Figure G.2, while the inside of the sleeves are filled with a 0.3 N
NaOH solution. A set of concrete specimens under testing with the RCM method is shown in Figure G.3.

86
 Figure G.2: Experimental set up for the RCM testing of chloride diffusivity (D0), where (a): rubber sleeve, (b):
anolyte, (c): anode, (d): concrete specimen, (e): catholyte, (f): cathode, (g): plastic support and (h): plastic box
[AASHTO, 2003]

Figure G.3: Concrete specimens under testing with the RCM method [AASHTO, 2003]

By use of the separate electrodes placed on each side of the specimens, an electrical voltage gradient
is applied and the electrical current through the specimens observed. Depending on the level of the
observed current which reflects the resistance of the concrete to chloride ingress, the applied potential
is adjusted in order to obtain a proper duration of the testing. By an applied DC potential which may vary
from 10 to 60 V, the chloride ions are forced into the concrete specimens during a relatively short period
of time. For a normal dense concrete, a testing duration of 24 hours may be typical, while for a denser
concrete, a longer time is needed.

Immediately after termination of the accelerated exposure to the chloride solution, the test specimens
are split into two halves, from which the average depth of chloride ingress is observed on the freshly
split concrete surface by use of a colorimetric technique (Figure G.4). Based on the observed depth of
chloride ingress and the applied testing conditions, the chloride diffusivity is calculated according to an
established procedure. As a result, the chloride diffusivity (D) of the concrete is calculated with an
average value and standard deviation from the testing of three different test specimens. Depending on
the resistance to chloride ingress of the given concrete, the whole test period including preparation of
the concrete specimens may take a few days.

87
 Figure G.4: Observed depth of chloride ingress on the split concrete surface after spraying of the fresh surface
with a standard AgNO3 solution [AASHTO, 2003]

G.2.4 EVALUATION OF RESULTS

When the resistance of the concrete to chloride ingress is determined and characterized by such a
strongly accelerated test method as that described above, it should be clear that the results obtained
are quite different from the real ingress of chlorides taking place under more normal conditions in the
field. As already discussed in Chapter 3 and Appendix D, however, it should be noted that the above
Rapid Chloride Migration (RCM) diffusivity is only a simple, relative index reflecting the resistance of the
concrete to chloride ingress and the general durability properties of the concrete; the testing conditions
for this method are very much similar to that of the Rapid Chloride Permeability (RCP) method [ASTM,
2010 ; Teng et al., 2014].

Since the RCM method does not require any pre-curing of the concrete so the testing can be carried out
very rapidly, this independence of pre-curing is very important for a performance-based concrete quality
control and quality assurance during concrete construction.

For a more complete evaluation and comparison of the resistance to chloride ingress of various types
of concrete, however, it should be noted that not only the 28-day chloride diffusivity should be evaluated.
It is the more complete development of chloride diffusivity from an early age and up which more properly
reflects the durability properties of the given concrete and hence, the different resistance to chloride
ingress of various types of concrete.

G.3 ELECTRICAL RESISTIVITY

G.3.1 GENERAl

Also for the testing of electrical resistivity of the concrete, several experimental techniques and test
methods exist, all of which give different results [Gjørv et al., 1977 ; Polder et al., 2000]. Basically,
however, there are two different types of test method which are typically being used for quality control
of concrete, one of which is the two-electrode method, while the other is the four-electrode method or
the so-called Wenner method. Since the two-electrode method is the most well-defined and accurate
test method, however, this method should preferably be applied for the regular concrete quality control.
The two-electrode method better reflects the bulk properties of the concrete, and this method is also
less dependent of the operator during testing.

G.3.2 TEST METHOD

Schematically, the two-electrode method is shown in Figure G.5, where the applied current flows through
the whole bulk of the concrete specimen which is placed in-between two solid steel plates. The electrical
resistivity of the concrete is then observed by use of a LCR Meter at a frequency of 1 kHz. The resistivity
of the concrete (r) is calculated by the following equation:

88
 (G.3)

where R is the resistance, A is the surface area of the specimen and t is the height of the concrete
specimen.

Figure G.5: The two-electrode method for electrical resistivity measurements of concrete

G.3.3 EVALUATION OF RESULTS

Since both the moisture and temperature conditions of the concrete also are very important factors
affecting the observed electrical resistivity, all measurements of the electrical resistivity must be carried
out under defined and controlled conditions in the laboratory. Immediately before testing of the water
cured concrete specimens, all free water on the surface of the specimens must be carefully wiped off.
Also, it is very important to ensure a good electrical connection between the electrodes and the concrete
surface, which is ensured by use of wet burlaps between the steel plates and the concrete surface. In
order to avoid any current drain during measurements, the specimens must further be placed on a dry
electrically insulated base plate, and any touching of the concrete specimen during testing by hand
avoided.

G.4 CONCRETE COVER

For concrete structures in the marine environment, the specified concrete cover is normally very thick,
and the reinforcement is often highly congested, making it difficult to measure the cover thickness
accurately based on conventional cover meters. The use of stainless steel reinforcement may further
complicate such measurements, although cover meters based on pulse induction can then be used.
Also more sophisticated scanning systems for control of achieved concrete cover exist (Figure G.6).

Figure G.6: Control of achieved concrete cover based on image scanning

89
Often, a more pragmatic approach to the control of cover thickness is applied, where manual readings
of the cover thickness on protruding bars in construction joints during concrete construction are specified
(Figure G.7). If the amount of such control measurements is sufficiently high to produce reliable
statistical data, such a simple approach may be considered sufficiently accurate for the regular control
and quality assurance during concrete construction. As long as an ongoing control and documentation
of the achieved concrete cover are required, experience demonstrates that the increased focus and
attention on achieved concrete cover also in itself is very important for an increased quality of
workmanship [Gjørv, 2010, 2014, 2015].

Figure G.7: Control of achieved concrete cover on protruding bars in construction joints

G.5 ELECTRICAL CONTINUITY

G.5.1 GENERAL

If cathodic prevention or provisions for such a protective measure are specified, requirements to the
electrical continuity within the reinforcement system are given in the European standard for cathodic
protection of concrete structures EN 12696 [CEN, 2000]. For heavily reinforced concrete structures,
there may already be a sufficient electrical continuity without any additional measures, but normally,
both welding and special electrical connections between various parts of the rebar system are needed.
For each step of the concrete construction work, however, the specified electrical continuity of the
reinforcement must be properly controlled and secured. Such control should preferably be carried out
and assured by qualified people with experience from cathodic protection systems.

G.5.2 TESTING PROCEDURE

In principle, the electrical continuity is controlled by measurements of the ohmic resistance between
various parts of the reinforcement system; between any two points within the system, the resistance
shall not exceed 1 Ohm. In order to avoid the uncertainty of measurements when a traditional multimeter
is applied, however, the measurements should preferably be carried out by use of a relatively high
current (1 A) between the various parts of the reinforcement system. Commercial equipment for such
measurements is available (Figure G.8), where both the ohmic resistance and the rest voltage between
the measuring parts 0.1 seconds after interruption of current are observed.

90
Figure G.8: Equipment for control of the electrical continuity within the reinforcement system

91
 PIANC Secrétariat Général
Boulevard du Roi Albert II 20, B 3
B-1000 Bruxelles
Belgique

http://www.pianc.org
VAT BE 408-287-945

ISBN 978-2- 87223-240-6

EAN 9782872232406

9782872232406