The document provides an overview of corrosion of steel in concrete. It discusses how steel is usually protected from corrosion by the alkaline environment of concrete but can corrode due to carbonation or chloride attack breaking down the protective layer. The corrosion process involves steel dissolving at the anode and oxygen being consumed at the cathode. This leads to a volume increase and the formation of rust, causing cracking and spalling of concrete. It also describes "black rust" that can occur in low-oxygen conditions without visible damage. The document aims to explain the basics of corrosion mechanisms to help understand investigation and repair techniques.
Report
Share
Report
Share
1 of 75
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
More Related Content
Introduction
1. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
1 Introduction
Reinforced concrete is a versatile, economical and successful construction material. It can be
moulded to a variety of shapes and finishes. Usually it is durable and string, performing well
throughout its service life. However, sometimes it does not perform adequately as a result of
poor design, poor construction, inadequate materials selection, a more severe environment
than anticipated or a combination of these factors.
The corrosion of reinforcing steel in concrete is a major problem facing civil engineers and
surveyors today as they maintain an ageing infrastructure. Potentially corrosion rehabilitation
is a very large market for those who develop the expertise to deal with the problem. It is also a
major headache for those who are responsible for dealing with structures suffering from
corrosion.
One American estimate is that $150 billion worth of corrosion damage on their interstate
highway bridges is due to deicing and sea salt induced corrosion. In a recent Transportation
Research Board Report on the costs of deicing (Transportation Research Board, 1991), the
annual cost of bridge deck repairs was estimated to be $50 to $200 million, with substructures
and other components requiring $100 million a year and a further $50 to 150 million a year on
multistory car parks.
In the United Kingdom, the Highways Agency’s estimate of salt induced corrosion damage is
a total of £616.5 million on motorway and trunk road bridges in England and Wales alone
(Wallbank, 1989). These bridges represent about 10% of the total bridge inventory in the
country. The eventual cost may therefore be ten times the Highways Agency’s estimate. The
statistics for Europe, the Asian Pacific countries and Australia are similar.
In the Middle East the severe conditions of a warm marine climate with saline ground wates
increase all corrosion problems. This is made worse by difficulty of curing concrete and has
led to very short lifetime for reinforced concrete structures (Rasheedezzafar et al., 1992).
In many countries with rapidly developing infrastructures, economies in construction have led
to poor quality concrete and low concrete cover to the steel resulting on carbonation
problems.
Corrosion became a fact of life as soon as man started digging ores out of the ground,
smelting and refining them to produce the metals that we use so widely in the manufacturing
and construction industries. When man has finished refining the steel and other metals that we
sets nature sets about reversing the process. The refined metal will react with the non-metallic
environment to form oxides, sulphates, sulphides, chlorides, etc. which no longer have the
required chemical and physical properties of the consumed metal. Billions of dollars are spent
every year in protecting, repairing and replacing corrosion damage. Occasionally lives are list
when steel pipes, pressure vessels or structural elements on bridges fail. But corrosion is a
slow process and it usually easy to detect before catastrophic rather than death or injury. The
corrosion of steel in concrete has previously caused only economic problems; however,
recently there have been cases of large pieces of motorist’s life in New York City. The failure
of post-tensioning tendons in a bridge caused a collapse with the loss of a tanker driver’s life
in Europe.
The economic loss and damage caused by the corrosion of steel in concrete males it arguably
the largest single infrastructure problem facing industrialized countries. Our bridges, public
- page 1-
2. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
utilities, chemical plants and buildings are ageing. Some can be replaced, others wold cause
great cost and inconvenience if they ware taken out of commission, power plants and other
structures we can build, it becomes crucial that the existing structures perform to their design
lives and limits and are maintained effectively.
One of the biggest causes of corrosion of steel in concrete is the use of deicing salts on our
highways. In the USA approximately 10 million tones of salt are applied per year to
highways. In the UK 1 to 2 million tones per year are applied (but to a proportionately far
smaller road network). In continental Europe the application rates are comparable, except
where it is too cold for salt to be effective, or the population density too low to make salting
worthwhile (in northern Scandinavia, for example). However, research in t he USA has shown
that the use of deicing salt is still more economic than alternative, more expensive, less
effective deicers (Transportation Research Board, 1991).
In the early years of this century corrosion of steel in concrete was attributed to stray current
corrosion from electric powered vehicles. It was only in the 1950s in the USA that was finally
accepted that there were corrosion problems on bridges far from power lines bit in areas
where sea salt or deicing salts were prevalent. During the 1960s attempts were made to
quantify the problem and in the 1970s the first cathodic protection systems were installed to
deal with the problem. In Europe and the Middle East sea water for mixing concrete and the
addition of calcium chloride set accelerators were acceptable until the 1960s and still used
until the 1970s in a mistaken belief that most of the chlorides would be bound up in the
cement and would not cause corrosion. This was found to be an expensive error over the next
20 years, particularly in the Middle East, where the low availability of potable water led to the
frequent use of sea water for concrete mixing.
In the UK the Concrete in the Oceans research programme in the 1970s did much to improve
our understanding of the fundamentals of corrosion of steel in concrete, particularly in marine
conditions (Wilkins and Lawrence, 1980). Government laboratories around the world have
worked on the problems since the 1960s when the problem started to manifest themselves.
The book frequently refers to the results of the Strategic Highway Research Program (SHRP).
The author was a member of the SHRP staff for the first three years and a consultant for the
last three years of SHRP. About $10 million worth of research was carried out over six years
(1987-1993) on corrosion problems on existing reinforced concrete bridges. A complete list of
SHRP reports on corrosion of steel in concrete is included in the appendix. SHRP was
probably the largest single effort to address the problem of corrosion of steel in concrete and
provide practical solution for engineers to use.
The text draws on corrosion problems and experience from around the world where possible,
giving balanced view of different approaches in different countries, particularly compring the
European and North American approaches. Inevitably it concentrates on the author’s primary
experience in the UK, North America and, to a lesser extent Europe, the Middle East and the
Pacific Rim countries.
The next two chapters following this introduction explore the basics of corrosion of steel in
concrete. The author has attempted to be thorough but also to start from the position of the
generalist, with a minimal memory of chemistry. A glossary is provided to remind the reader
of the basic terms used.
- page 2-
3. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
The first requirement when addressing a deterioration problem is to quantify it. Chapter 4
therefore discusses condition evaluation and the testing procedures and techniques we can use
to asses the causes and extent of the corrosion damage on a structure.
Chapter 5 and 6 are then concerned with repair and rehabilitation options, first the
conventional physical intervention of concrete repair patching, overlaying and coatings. The
electrochemical techniques of cathodic protection, chloride removal and realkalization are
dealt with in Chapter 6.
Chapter 7 discusses how to select our rehabilitation method. It summarizes the merits and
limitations of the different repair techniques covered in the previous chapters and the cost
differentials. The following chapter reviews models for determining deterioration rates and
service lives.
The penultimate chapter turns away the issues of dealing with the problem on existing
structures and reviews the options for new construction to ensure that the next generations of
structures do not show the problems associated with some present generation. Many materials
and methods of improving durability are on the market; which are cost effective?
The final chapter discusses future developments. As materials science, compute technology,
electronics and other disciplines advance, new methods for assessing corrosion are becoming
available. The author speculated about what will be on offer and perhaps more importantly
what we need to efficiently assess and repair corrosion damage on our fixed infrastructure in
the future.
1.1 REFERENCES
Rasheeduzzafar, Dakhil, F.H, Bader, M.A. and Khan, M.N. (1992) ‘performance of corrosion
resisting steel in chloride-bearing concrete’, ACI Materials Journal, 89 (5), 439-48
Transportation Research Board (1991) Highway Deicing: Comparing Salt and Calcium
Magnesium Acetate, Special Report 235, National Research Council, Washington, DC.
Wallbank, E.J. (1989) The Performance of Concrete in Bridges: A survey of 200 Highway
Bridges, HMSO, London
Wilkins, N.J.M. and Lawrence, P.F. (1980) Concrete in the Oceans: Fundamental
Mechanisms of Corrosion of Steel Reinforcements in Concrete Immersed in Sea Water,
Technical Report 6, CIRIA/UEG Cement and Concrete Association Slough, UK.
- page 3-
4. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
2 Corrosion of steel in concrete
This chapter discusses the basics of corrosion and how they apply to steel in concrete. The
glossary of term at the beginning of the books gives some definitions of the terms used here s
they apply to the corrosion of steel in concrete.
Why does steel corrode in concrete? A more sensible question is why steel does not corrode
in concrete. We know from experience that mild steel and high strength reinforcing steel bars
corrode (rust) when air and water are present. As concrete is porous and contains moisture
why does steel in concrete not usually corrode?
The answer is that concrete is alkaline. Alkalinity is the opposite of acidity. Metals corrode in
acids, whereas they are often protected from corrosion by alkalis.
When we say that concrete is alkaline we mean that it contains microscopic pores with high
concentrations of soluble calcium, sodium and potassium oxides. These oxides form
hydroxides, which are very alkaline, when water is added. This creates a very alkaline
condition (pH 12-13 –see the glossary for a definition of pH). The composition of the pore
water and the movement of ions and gases through the pores is very important when
analyzing the susceptibility of reinforced concrete structures to corrosion. This more fully
discussed in Chapter 3.
The alkaline condition leads to a ‘passive’ layer forming on the steel surface. A passive layer
is a dense, impenetrable film which, if fully established and maintained, prevents further
corrosion of the steel. The layer formed on steel in concrete is probably part metal
oxide/hydroxide and part mineral from the cement. A true passive layer is a very dense, thin
layer of oxide that leads to a very slow rate of oxidation (corrosion). There is some discussion
whether or not the layer on the steel is a true passive layer as it seems to be thick compared
with other passive layers and it consists of more than just metal oxides; but it behaves like a
passive layer and it is therefore generally referred to as such.
Corrosion engineers spend much of their time trying to find ways of stopping corrosion of
steel by applying protective coatings. Metals such as zinc or polymers such as acrylics or
epoxies are used to stop corrosive conditions getting to steel surface. The passive layer us the
corrosion engineer’s dream coating as it forms itself and will maintain and repair itself as long
as the passivating (alkaline) environment is there to regenerate it if it is damaged. If the
passivating environment can be maintained, it is far better than any artificial coatings such as
galvanizing or fusion bonded epoxy that can be consumed or damaged, allowing corrosion to
proceed in damaged areas.
However, the passivating environment is not always maintained. Two processes can break
down the passivating environment in concrete. One is carbonation and the other is chloride
attack. These will be discussed in Chapter 3. in the rest of this chapter we will discuss what
happens when depassivation has occurred.
- page 4-
5. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
2.1 THE CORROSION PROCESS
Once the passive layer breaks down then areas of rust will start appearing on the steel surface.
The chemical reactions are the same whether corrosion occurs by chloride attack or
carbonation. When steel in concrete corrodes it dissolves in the pore water and gives up
electrons:
The anodic reaction: Fe ->Fe2+ + 2e- (2.1)
The two electrons (2e-) created in the anodic reaction must be consumed elsewhere on the
steel surface to preserve electrical neutrality. In other words, it is not possible for large
amounts of electrical charge to build up at one place on the steel; another chemical reaction
must consume the electrons. This is a reaction that consumes water and oxygen:
The cathodc reaction : 2e- + H2O + 1/2O2 ->2OH- (2.2)
This is illustrated in Figure 2.1. You will notice that hydroxyl ions (2OH-) are generated in
the cathodic reaction. These ions increase the local alkalinity and will therefore strength the
passive layer, warding off the effects of carbonation and chloride ions at the cathode. Note
that water and oxygen are needed at the cathode for corrosion to occur.
The anodic and cathodic reaction (2.1 and 2.2) are only the first steps in the process of
creating rust. However, this pair of reactions is critical to the understanding of corrosion and
is widely quoted in any discussion on corrosion prevention for steel in concrete. The reactions
will be referred to often in this book.
Figure 2.1 The anodic and cathodic reactions.
If the iron were just to dissolve in the pore water (the ferrous ion Fe2+ in equation 2.1 is
soluble) we would not see cracking and spalling of the concrete. Several more stages must
occur for ‘rust” to form. This can be expressed in several ways; one is shown below where
ferrous hydroxide becomes ferric hydroxide and then hydrated ferric oxide or rust:
Fe2+ + 2OH- -> Fe(OH)2 (2.3)
Ferrous hydroxide
4Fe(OH)2 + O2 + 2H2O -> 4Fe(OH)3 (2.4)
Ferric hydroxide
- page 5-
6. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
2Fe(OH)3 ->Fe2O3.H2O + 2H2O (2.5)
Hydrated ferric oxide (rust)
The full corrosion process is illustrated in Figure 2.2. Unhydrated ferric oxide Fe2O3 has a
volume of about twice that of the steel it replaced when fully dense. When it becomes
hydrated it swells even more and becomes porous. This means that the volume increase at the
steel/concrete interface in two to ten times. This leads to the cracking and spalling that we
observe as the usual consequence of corrosion of steel in concrete and the red/brown brittke,
flaky rust in the bar and the rust stains seen at cracks in the concrete. The cracking and
spalling process is discussed more fully in Section 3.3 on corrosion damage.
Several factors in the explanation given in this section are important and will be used later to
explain how we measure and stop corrosion.
Figure 2.2. The corrosion reactions on steel.
- page 6-
7. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
The electrical current flow, and the generation and consumption of electrons in the anode and
cathode reactions are used in half cell potential measurements and cathodic protection. The
formation of protective, alkaline hydroxyl ions is used in cathodic protection, electrochemical
chloride removal and realkalization. The fact that the cathodic and anodic reaction must
balance each other for corrosion to proceed is used in epoxy coating protection of rebars. This
will be discussed later.
2.2 BLACK RUST
There is an alternative to the formation of ‘normal’ red rust described in reactions 2.3 to 2.5
above. If the anode and cathode are well separated (by several hundred millimeters) and the
anode is starved of oxygen (say be being underwater) the iron as Fe2+ will stay in solution. Tis
means that there will be no expansive forces to crack the concrete as described above and in
Section 3.3 so corrosion may not be detected.
This type of corrosion (known as ‘black’ or ‘green’ rust due to the color of the liquid seen on
the rebar when first exposed to air after breakout) is found under damaged waterproof
membranes and in some underwater saturated conditions. It is potentially as there is no
indication of corrosion by cracking and spalling of the concrete and the reinforcing steel may
be severely weakened before corrosion is detected. Rebars may be hollowed out in such
deoxygenated conditions, particularly under membranes or when water is permanently ponded
on the surface.
Figure 2.3 Rebars taken from under the end of a waterproof membrane. They have been
subjected to low oxygen conditions and therefore show local ‘wasting’ of bars with no
expansive oxide growth.
Examples of rebars attacked in this way are shown in Figure 2.3. These bars were taken from
underneath damaged waterproof membranes. Rust staining on the concrete surface may be
indicative of this type of attack, but obviously if water is getting under a membrane and
excluding oxygen it is unlikely that the iron in solution will get to the concrete surface where
it will then precipitate out to form rust stains.
- page 7-
8. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
2.3 PITS, STRAY CURRENT AND BACTERIAL CORROSION
2.3.1 Pit formation
Corrosion of steel in concrete generally starts with the formation of pits. These increase in
number, expand and join up leading to generalized corrosion usually seen on reinforcing bars
exposed to carbonation or chlorides. The formation of a pit is illustrated in Figure 2.4.
The chemistry of pitting is quite complex and is explained in most corrosion textbooks.
However, the principle is fairly simple, especially where chlorides are present. At some
suitable site on the steel surface (often thought to be a void in the cement paste or a sulphide
inclusion in the steel), the passive layer is more vulnerable to attack and an electrochemical
potential difference attracts chloride ions. Corrosion is initiated and acids are formed;
hydrogen sulphide from the sulphide (MnS) inclusion and HCl from the chloride ions if they
are present. Iron dissolves (equation 2.1), and the iron in solution reacts with water:
Fe2+ + H2O ->FeOH- + H+
MnS + 2H+ ->H2S + Mn2+
A pit forms, rust may from over the pit, concentrating the acid (H+), and excluding oxygen so
that the iron stays in solution preventing the formation of a protective oxide layer and
accelerating corrosion.
We will return to the subject of pitting corrosion later. It is related to the problems of coated
reinforcement and to the ‘black rust’ phenomenon discussed above.
2.3.2 Bacterial corrosion
Another complication comes from bacterial corrosion. Bacteria in the soil (thiobacilli) convert
sulphur and sulphides to sulphuric acid. This acid attack steel causing corrosion. There are
other species (ferrobacilli) that attack the sulphides in steel (FeS). This often associated with a
smell of hydrogen sulphide (rotten eggs) and smooth pitting, with a black corrosion product
when rebars are exposed having been in water saturated conditions. In anaerobic (oxygen
starved) conditions such bacteria can contribute to the pitting corrosion discussed above.
- page 8-
9. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 2.4. The classical corrosion model of pitting attack.
2.3.3 Stray current induced corrosion
As stated in the introduction, stray currents were originally blamed for all corrosion in
concrete in the USA, until the problem of chloride was identified in the 1950s. the main cause
of stray current induced corrosion was the direct current flowing through reinforcing steel due
to DC traction systems on trams (streetcars) and electric trains. When current has to jump
from one metal conductor to another through an ionic medium (e.g. from one reinforcement
cage to another via the concrete pore water), then one end becomes negative (a cathode, i.e
cathodically protected), while the other end becomes an anode and corrodes. The problem of
stray current corrosion within a cathodic protection system and an illustration of the
mechanism is given in Chapter 6 where the problems of current flow through isolated rebars
are discussed for electrochemical techniques.
- page 9-
10. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Today stray current corrosion in bridges, building and aboveground structures is a specialized
problem dealt with by engineers designing and maintaining light railway systems. Below
ground it is the province of the cathodic protection engineer, where interaction between a
cathodic protection system on pipeline, etc. can induce corrosion on adjacent buried
structures. Stray current induced corrosion does not occur on AC systems, and is most
unlikely on above-ground structures when the source is below ground.
2.3.4 Local versus general corrosion (macrocells versus microcells)
Corrosion s often local, with a few centimeters of corrosion and then up to a metre of clean
passive bar, particularly for chloride induced corrosion. This indicates the separation of the
anodic reaction (2.1) and the cathodic reaction (2.2), to form a ‘macrocell’. Chloride induced
corrosion gives rise to particularly well defined marcrocells. This is partly due to the
mechanism of chloride attact, with ppit formation and with small concentrated anodes being
‘fed’ by large cathodes. It is also because chloride attact is usually associated with high levels
of moisture giving low electrical resistance in the concrete and easy transport of ions so the
anodes and cathodes can separate easily.
In North America, this is used as a way of measuring corrosion by measuring ‘marcrocell
currents’. This is discussed later in Section 4.11.
2.4 ELECTROCHEMISTRY, CELLS AND HALF CELLS
The first two reactions we discussed in this chapter were the anodic and cathodic reactions for
steel in concrete. The terms ‘anode’ and ‘cathode’ come from electrochemistry wich is the
study if the chemistry if electrical cells. Figures 2.5 is a basic Daniell cell which is used at
high school to illustrate how chemical reactions produce electricity. The cell is composed o
two ‘half cells’, copper is copper sulphate and sinc in zinc sulphate. The total voltage of the
cell is determined by the metals used and by the nature and composition of the solutions.
What is happening is that in each half cell the metal Is dissolving and ions are precipitating ,
i.e.
M <-> Mn+ + ne-
Copper is more resistant to this reaction than zinc so when we connect the two solutions by a
semie-permeable membrane (which allows charge to be exchanged through it but the ions
cannot pass through) and connect the two metals with a wire, the zinc goes into solution, and
the copper from the copper sulphate solution plates out (is deposited) on the copper electrode.
- page 10-
11. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 2.5 The Daniell cell. Cell voltage = 0.34 – (-0.76) = 1.10 V.
The voltage of any half cell can be recorded against a standard hydrogen electrode (half cell).
Table 2.1 gives the standard half cell potentials that are of interest to us as we evaluate
corrosion problems. Half cell potentials are a function of concentration as well as the metal
and the solution. A more concentrated solution is generally) more corrosive than a dilute one
so a current will flow in a cell made up of a single metal in two different concentrations of the
same solution. We can consider the corrosion of steel in concrete as a concentration cell.
We can measure the corrosion risk in that cell by introducing an external half cell. This is
most easily illustrated by a copper/copper sulphate half cell (see Section .7) moved along the
surface of the concrete which contains a rebar with anodic (corroding) areas and cathodic
(passive) areas on it.
Figure 2.6 Half cell measurement of corrosion potential.
As we see in Figure 2.6, by placing a half cell on the concrete surface and connecting it via a
voltmeter to the steel, we have a similar circuit to our Daniel cell (Figure 2.5). The electrical
- page 11-
12. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
potential difference will be a function of the iron in its pore water environment. If we move
the cell along the steel we will see different potentials because the iron is in different
environment. At the anode it can easily go into solution like the zinc in our Deniell cell. At
the cathode, the passive layer is still strong and being strengthened further by the cathodic
reaction, so the steel resists dissolution.
As a result we see higher potentials (voltages) on our voltmeter in anodic, corroding areas and
lower voltages in the cathodic, passive areas. The use and interpretation of half cell
measurements is discussed further in Section 4.7.
We must be careful in our use of electrochemical theory to explain what is going on in a
corrosion cell. Electrochemical theory generally applies to equilibrium conditions and well
defined solutions. Corrosion is not equilibrium, but a dynamic situation. The theory and
equation of electrochemistry are therefore an approximation and can lead to errors if the
model is stretched too far.
2.5 CONCLUSIONS
This chapter has discussed in detail the mechanism of whet happens at the steel surface. The
chemical reactions, formation of oxides, pitting, stray currests, bacterial corrosion, anodes,
cathodes and half cell have been reviewed. The next chapter will discuss the processes that
lead to the corrosion and the consequences of in terms of damage to structure, before moving
on to the measurement of the problem and how to deal with it.
- page 12-
13. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
3 Causes and mechanisms of corrosion and corrosion
damage in concrete
There are two main causes of corrosion of steel in concrete. This chapter will discuss how
chloride attack and carbonation lead to corrosion and how the corrosion proceeds once it has
started. The mechanism of corrosion damage is explained. There will also be discussion of the
variations that can be found when carrying out investigations in the field.
The main causes of corrosion of steel in concrete are chloride attack and carbonation. These
two mechanisms are unusual in that they do not attack the integrity of the concrete. Instead,
aggressive chemical species pass through the pores in the concrete and attack the steel. This is
unlike normal deterioration processes due to chemical attack on concrete. Other acids and
aggressive ions such as sulphate destroy the integrity of the concrete before the steel is
affected. Most forms of chemical attack are therefore concrete problems before they are
corrosion problems. Carbon dioxide and the chloride ion are very unusual in penetrating the
concrete without significantly damaging it. Accounts of (for instance) acid rain causing
corrosion of steel embedded in concrete have been shown to attack the steel and not the
concrete.
3.1 CARBONATION
Carbonation is the result of the interaction of carbon dioxide gas in the atmosphere with the
alkaline hydroxides in the concrete. Like many other gases carbon dioxide dissolved in water
to form an acid. Unlike most other acids the cacbonic acid does not attack the cement paste,
but just neutralizes the alkalis in the pore water, mainly forming calcium carbonate that lines
the pores:
CO2 + H2O -> H2CO3 (3.1)
Gas Water Carbonic acid
H2CO3 + Ca(OH)2 -> CaCO3 + 2H2O (3.2)
Carbonic acid Pore solution
There is a lot more calcium hydroxide in the concrete pores than can be dissolved in the pore
water. This helps maintain the pH at its usual level of around 12 or 13 as the carbonation
reaction occurs. However, eventually all the locally available calcium hydroxide reacts,
precipitating the calcium carbonate and allowing the pH to fall to a level where steel will
corrode. This is illustrated in Fugure 3.1. There is a discussion of the pH levels involved and
their measurement in Section 4.8.
Carbonation damage occurs most rapidly when there is little concrete cover over the
reinforcing steel. Carbonation can occur even when the concrete cover depth to the
reinforcing steel is high. This may be due to a very open pore structure where pores are well
connected together and allow rapid CO2 ingress. It may also happen when alkaline reserves in
content, high water cement ratio and poor curing of the concrete.
A carbonation front proceeds into the concrete roughly following the laws of diffusion. These
are most easily defined by the statement that the rate is inversely proportional to the thickness:
- page 13-
14. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
dx D0
= (3.3)
dt x
Where x is distance, t is time and D0 is the diffusion constant.
The diffusion constant D0 is determined by the concrete quality. At the carbonation front
there is a sharp drop in alkalinity from ph 11-13 down to less than pH 8(Figure 3.1). At that
level the passive layer, which we saw in the previous chapter was created by the alkalinity, is
no longer sustained so corrosion proceeds by the general corrosion mechanism as described.
Figure 3.1 Schematic of the carbonation front and its relationship to the corrosion threshold
and the colour change for phenolphthalein.
Many factors influence the ability of reinforced concrete to resist carbonation induced
corrosion. The carbonation rate, or rather the tome to carbonation induced corrosion, is a
function of cover thickness, s good cover is essential to resist carbonation. As the process is
one of neutralizing the alkalinity of the concrete, good reserves of alkali are needed, i.e. a
- page 14-
15. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
high cement content. The diffusion process is made easier if the concrete has an open pore
structure. On the macroscopic scale this means that there should be good compaction. On a
microscopic scale well cured concrete has small pores and lower connectivity of pores so the
CO2 has a harder job moving through the concrete. Microsilica and other additives can block
pores or reduce pores sizes.
Carbonation is common in old structure, badly built structures (particularly building) and
reconstituted stone elements containing reinforcement that often have low cement content and
are very porous. Carbonation is rare in modern highway bridges and other civil engineering
structures where water/cement rations are low, cement contents are high with good
compaction and curing, and there is enough cover to prevent the Carbonation front advancing
into the concrete to the depth of the steel within the lifetime of the structure. On those
structures exposed to sea water or deicing salts, the chlorides usually penetrate to the
reinforcement and cause corrosion long before carbonation becomes a problem. Wet/dry
cycling on the concrete surface will accelerate carbonation by allowing carbon dioxide gas in
during the dry cycle and then supplying the water to dissolve it in the wet cycle (equation
3.1). This gives problems in some countries in tropical or semi-tropical regions where the
cycling between wet and dry seasons seems to favour carbonation, e.g. Hong Kong and some
Pacific Rim countries.
When a repairer talks of repairing corrosion due to ‘low cover’ he usually means that the
concrete has carbonated around the steel leading to corrosion. As the cover is low it was a
quick process, perhaps within five year of construction. If the concrete were of the highest
quality carbonation may not have been possible and low cover might not have mattered.
Carbonation is easy to detect and measure. A pH indicator, usually phenolphthalein in a
solution of water and alcohol, will detect the change in pH across a freshly exposed concrete
face. Phenolphthalein changes from colourless of low pH (carbonated zone) to pink at high
pH (uncarbonated concrete). Measurements can be taken on concrete cores, fragments and
down drilled holes. Care must be taken to prevent dust or water from contaminating the
surface, is cheap and simple. Figure 3.1 shows a typical carbonation front in concrete with the
pH of the concrete before and after carbonation, the pH threshold of corrosion for steel and
the pH change of phenolphthalein taken from Parrott (1987). The measurement of the
carbonation front is covered in Section 4.6.
3.1.1 Carbonation transport through concrete
Carbon dioxide diffuses through the concrete and the rate of movement of the carbonation
front approximates to Fick’s law of diffusion. This states that the rate of movement is
inversely proportional to the distance from the surface as in equation 3.3 above. However, as
the carbonation process modifies the concrete pore structure as it proceeds, this is only an
approximation. Cracks, changes in concrete composition and moisture levels with depth will
also lead to deviation from the perfect diffusion equation. Integration of equation 3.3 gives a
square root law that can be used to estimate the movement of the carbonation front. The
calculation of diffusion rates is discussed in more detail in Chapter 8.
Empirically, a number of equations have been used to link carbonation rates, concrete quality
and environment. Table 3.1 summarizes some of those equations and shows the factors that
have been included. Generally there is a t dependence. As discussed above the other factor are
exposure, water/cement ratio, strength and CaO content (functions of cement type and its
alkali content). For example if we consider the basic equation:
- page 15-
16. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
d = At 0.5
Table 3.1 A selection of carbonation depth equations
Equation Parameters
d = At n
d = carbonation depth
t = time in years
A = diffusion coefficient
n = exponent (approximately ½)
d = ABCt 0 .5
A = 1.0 for external exposure
B = 0.07 to 1.0 depending on surface finish
C = R (wc − 0.25 ) / (0.3(1.15 + 3wc ))
1/ 2
for
water cement ration (wc) > = 0.6
C = 0.37R(4.6wc – 1.7) for wc < 0.6
R = coefficient of neutralization, a function of
mix design and additives
d = A( Bwc − c)t 0.5 A is a function of curing
B and C are a function of fly ash used
d = 0.43(wc − 0.4 )(12(t − 1)) + 0.1 28 day cured
0 .5
d = 0.53(wc − 0.3)(12t ) + 0.2 Uncured
0 .5
(
d = 2.6(wc − 0.3) + 0.16 t 0.5
2
) Sheltered
d = ((wc − 0.3) )
+ 0.07 t 0.5 Unsheltered
2
d = 10.3e −0.123 f 28 at 3 years Unsheltered
fx = strength at day X
d = 3.4e −0.34 f 28 at 3 years Sheltered
d = 680( f 28 + 25) − 0.6 at 2 years
−1.5
d = A + B f 28 + c (CaO − 46) CaO is alkali content expressed as CaO
2 0.5
(
d = 0.508 f 35.5 − 0.047 (365t )
0
) 0.5
d =⎜
(
⎛ 0.846 10 wc (10 f 7 )0.5 )⎞(12t )
⎟ 0.5
− 0.95
⎜ − 0.193 − 0.076 wc ⎟
⎝ ⎠
d = A(T − t i )t 0.75 (C1 C 2 ) ti = induction time
0.5
T = temperature in °K
C1 = CO2 concentration
C2 = CO2 bound by concrete
Source: Parrott (1978)
And carbonation depths are 16mm in 16year in typical poor concrete and 4mm in 20years in a
good concrete, we would expect the diffusion coefficient A to range form 0.25 to 1.0 mm
yr1/2. In bridges it is common to see negligible carbonation after 20 years or more. On
buildings it can range from negligible to 60mm or more. This is a function of the higher
cover, lower water cement ratios, better specification and construction required to produce the
high strength concrete required in a bridge. However, high strength is not always synonymous
with a low diffusion constant. Permeability measurement (Section 4.12) should always be
made on structures in harsh environments rather than just relying on the ocnrete strength as a
measure of its durability.
- page 16-
17. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
3.2 CHLORIDE ATTACT
3.2.1 Sources of chlorides
Chlorides can come from several sources. They can be cast into the concrete or they can
diffuse in from the outside. Chlorides cast into concrete can be due to:
• deliberate addition of chloride set accelerators (calcium chloride, CaCl2, was widely
used until the mid-1970s);
• use of sea water in the mix;
• contaminated aggregates (usually sea dredged aggregates which were unwashed or
inadequately washed).
Chlorides can diffuse into concrete as a result of:
• sea salt spray and direct sea water wetting;
• deicing salts;
• use to chemicals (structure used for salt storage, brine tanks, aquaria, etc.).
Much of our discussion will centre in the diffusion of chlorides into concrete as that is the
major problem in most parts of the word either due to marine salt spray or use of deicing salts.
However, the cast-in chlorides must not be overlooked, especially when they are part of the
problem. A low level of chloride cast in can lead to rapid onset of corrosion if further
chlorides become available from the environment. This often happens in marine conditions
where seawater contaminates the original concrete mix and then diffuses into hardened
concrete.
Figure 3.2 Chloride profiles of marines bridge substructure (Yaquina Bay bridge soffit)
3.2.2 Chloride transport through concrete
Like carbonation, the rate of chloride ingress is often approximated to the laws of diffusion.
There are further complications here. The initial mechanism appears to be suction, especially
when the surface is dry. Salt water is rapidly absorbed by dry concrete. There is then some
- page 17-
18. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
capillary movement of the salt-laden water through the pores followed by ‘true’ diffusion.
There are other opposing mechanisms that slow the chlorides down. These include chemical
reaction to form chloroaluminates ans adsorption on to the pore surfaces.
The other problem with trying to predict the chloride penetration rate is defining the initial
concentration, as chloride diffusion produces a concentration gradient not a ‘front’. In other
words we can use the square root relationship for the carbonation front as the concrete either
is or is not carbonated, but we cannot use it so easily for chlorides as there is no chloride
‘front”, but a concentration profile in the concrete. A typical chloride profile is shown in
Figure 3.2. This particular profile is a very convincing fit to a diffusion curve but show no
error bars. Many profiles show far more scatter. The calculation of chloride diffusion rates is
discussed more fully in Chapter 8.
3.2.3 Chloride attack mechanism
In Chapter 2 we discussed the corrosion of steel in concrete and the effectiveness of the
alkalinity in the concrete pores producing a passive layer of protective oxide on the steel
surface which stops corrosion. In the previous section we observed that alkalinity in the
concrete pores is neutralized by carbonation. The depassivation mechanism for chloride attack
is somewhat different. The chloride ion attacks the passive layer but, unlike carbonation, there
is no overall drop in pH. Chloride act as catalysts to corrosion when there is sufficient
concentration at the rebar surface to break down the passive layer. They are not consumed in
the process but help to break down the passive layer of oxide on the steel and allow the
corrosion process to proceed quickly. This is illustrated in Figure 3.3. This makes chloride
attack difficult to remedy as chloride are hard to eliminate.
Figure 3.3 The breakdown of the passive layer and ‘recycling’ chlorides.
Obviously a few chloride ions in the pore water will not break down the passive layer,
especially if it is effectively re-establishing itself when damaged, as discussed in Chapter 2.
There is a ‘chloride threshold’ for corrosion given in terms of the chloride/hydroxyl ration. It
has been measured in laboratory tests with calcium hydroxide solutions. When the chloride
concentration exceeds 0.6 of the hydroxyl concentration, corrosion is observed (Hausmann
1967). This approximates to a concentration of 0.4% chloride by weight of cement of
chlorides are cast into concrete and 0.2% they diffuse in.
- page 18-
19. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
In the USA a commonly quoted threshold is 1lb chloride per cubic yard of concrete. Although
these figures are based on experimental evidence, the actual values are a function of practical
observations of real structures.
All these thresholds are approximations because:
1. Concrete pH varies with the type of cement and the concrete mix. A tiny pH change
represents a massive change in hydroxyl ion (OH-) concentration and therefore
(theoretically) the threshold moves radically with pH.
2. Chlorides can be bound chemically (by aluminates in the concrete and physically (by
absorption on the pore walls). This removes them (temporarily or permanently) from
the corrosion reaction. Sulphate resisting cements have low aluminate (C3A) content
which leads to more rapid diffusion and lower chloride thresholds.
3. in very dry concrete corrosion may not occur even at vary high Cl- concentration as
the water is missing from the corrosion reaction, as discussed in Chapter 2.
4. In sealed or polymer impregnated concrete, corrosion may not occur even at a very
high Cl- concentration if no oxygen or moisture is present to fuel the corrosion
reaction.
5. Corrosion can be suppressed when there is total water saturation due to oxygen
starvation, but if some oxygen gets in, then the pitting corrosion described in Chapter
2 can occur.
Therefore corrosion can be observed at a threshold level of 0.2% chloride be weight of
cement if the concrete quality is poor and there are water and oxygen available. In different
circumstances no corrosion may be seen at 1.0% chloride or more if oxygen and water are
excluded. If the concrete is very dry or totally saturated (as in (3) or (5) above) then a change
in conditions may lead to rapid corrosion.
3.2.4 Macrocell formation
As stated in the previous chapter, corrosion proceeds by the formation of anodes and cathodes
(Figure 2.1 and 2.2). In the case of chloride attack they are often well separated with areas of
rusting separated by areas of ‘clean’ steel (illustrated in Figure 2.3). This is known as the
macrocell phenomenon. Chloride induced corrosion is particularly prone to macrocell
formation as a high level of water is usually present to carry the chloride into the concrete and
because chlorides in concrete are hygroscopic (i.e. they absorb and retain moisture). The
presence of water in the pores increases the electrical conductivity of the concrete. The higher
conductivity allows the separation of anode and cathode as the ions can move through the
water filled (or water lined) pores.
The separation of corroded areas does not necessarily represent the distribution of chlorides
along the rebar. The anodic and cathodic reactions are separated with large cathodic area
supporting small, concentrated anodic areas.
For carbonation to occur the concrete is fenerally drier (otherwise the CO2 does not penetrate
far). Corrosion is therefore on a ‘microcell’ level with apparently continuous corrosion
observed along the reinforcing.
Stray currents from cathodic protection systems on other objects (such as a nearby pipeline or
DC electric railway lines) can also cause corrosion by accelerating the anodic reaction and
suppressing the cathodic reaction. In anaerobic conditions, bacterial attack has also been seen.
These causes all give rise to the same corrosion mechanism discussed above, but may need
specialist investigation and treatment.
- page 19-
20. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
3.3 CORROSION DAMAGE
In most industries corrosion is a concern because of wastage of metal leading to structural
damage such as a collapse, perforation of containers and pipes, etc. Most problems with
corrosion of steel in concrete are not due to loss of steel but the growth of the oxide. This
leads to cracking and spalling of the concrete cover.
Structural collapses of reinforced concrete structures due to corrosion are rare. The author
knows of two multistory parking structures in North America which collapsed due to deicing
salt induced corrosion. A post-tensioned concrete bridge collapsed in Wales due to deicing
salt attack on the strands (Woodward and Willicams, 1988), and so did one in Belgium.
Concrete damage would usually have to be well advanced before a reinforced concrete
structure is at risk.
Particular problems arise when the corrosion product is the black rust described in Chapter 2
and in prestressed, post-tensioned structures where corrosion is difficult to detect as the
tendons are enclosed in ducts. Tendon failure can be catastrophic as tendons are loaded to
50% or more of their ultimate tensile strength and modest section loss leads to failure under
load. The particular problems of assessing prestressed post-tensioned concrete structures are
addressed in Section 4.15.
The most common problem caused by corrosion is spalling of concrete cover. A man was
killed in New York City by a slab of concrete which spalled off a bridge substructure due to
deicing salts, and a vehicle was badly damaged in Michigan in a similar incident. Special
metal canopies have been built around the lower floors of high rise building where corrosion
has led to risk of failing concrete. This enables the investigators to collect the fallen concrete
at regular intervals and weigh it. In that way they can determine whether the corrosion rate is
stable, increasing or decreasing.
The important factors in corrosion of steel in concrete compared with most other corrosion
problems are the volume of oxide and where it is formed. A dense oxide formed at high
temperatures (such as in a power station boiler) usually has twice the volume of the steel
consumed. In most aqueous environments the excess volume of oxide is transported away and
deposits on open surfaces within the structure. For steel in concrete two factors predominate.
The main problem is that the pore water is static and there is no transport mechanism to move
the oxide away from the steel surface. The second problem is that the oxide is not dense. It is
very porous and takes up a very large volume, up to ten times that of the steel consumed. This
is illustrated in Figure 3.4
- page 20-
21. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 3.4 Rust growth forcing steel and concrete apart
The thermodynamics of corrosion, coupled with the low tensile strength of concrete, means
that the formation of oxide breaks up the concrete. It has been suggested that less than 100µm
of steel section loss are needed to start cracking and spalling the concrete. The actual amount
needed will depend upon the geometry in terms of cover, proximity to corners, rebar spacing,
bar diameter and the rate of corrosion. There is further discussion of the rate and amount of
corrosion leading to damage in section 4.11 on corrosion rate measurement.
Corners tend to crack first on corroding reinforced concrete structures. This is because the
oxygen, water, chlorides and carbon dioxide have two faces as path ways to the steel.
Delaminations occur as corrosion proceeds on neighbouring rebars and the horizontal
cracks/join up as shown in Figure 3.5.
Figure 3.5 Corrosion induced cracking and sapalling.
3.4 VERTICAL CRACKS, HORIZONTAL CRACKS AND CORROSION
The importance of vertical cracks in accelerating corrosion by allowing access of corrosion
agents to the steel surface has been widely discussed. If reinforcing steel is doing its job in
areas of tension in structure, small cracks will occur as the tensile load exceeds the tensile
- page 21-
22. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
strength of the steel. Most of these are small cracks (less than 0.5mm) intersecting the
reinforcing steel at right angles. They do not lead to corrosion of the steel as any local ingress
of chlorides, moisture and carbonation is limited and contained by the local alkalinity.
Obviously there is a limit to this ‘self-healing’ ability. If large cracks stay open (greater than
0.5mm), then corrosion can be accelerated. Such cracks may be due to plastic shrinkage,
thermal expansion or other reasons. The relationship between cracks in concrete and
reinforcement corrosion is fully discussed in Concrete Society (1995) Technical Report No.
44.
Corrosion causes horizontal cracking long the plane of the rebar and the corner cracking
around the end rebar. This leads to the loss of concrete cover as shown in Figure 3.5. This is
the main consequence of reinforcement corrosion with its subsequent risk of failing concrete
and unacceptable appearance.
3.5 THE SYNERGISTIC RELATIONSHIP BETWEEN CHLORIDE AND
CARBONATION ATTACK, CHLORIDE BINDING AND RELEASE
We have already discussed the fact that chlorides can be bound be the concrete. One of the
constituents of cement paste is C3A, a complex inorganic aluminium salt. This reacts with
chloride to form chloroaluminates. This removes the chloride from availability in the pore
water to cause corrosion. This binding process is strongest for chlorides cast into concrete,
and is why it was considered acceptable to use sea water to make concrete for many years.
The extent of binding and its effectiveness is not well understood. However, it is lnow that a
reduction in pH as caused by carbonation will break down the chloroaluminates. This leads to
a ‘wave’ of chloride moving in front of the carbonation front. Consequently structures with
chlorides in them that carbonate are more susceptible to corrosion than those with only one
source of problem.
3.6 REFERENCES
Concrete Society (1995) The Relevance of Cracking in Concrete to Corrosion of
Reninforcement, Technical Report No 44, The Concrete Society, Slough, UH.
Hausmann, D.A (1967) ‘Steel corrosion in concrete: how does it occur? ’, Materials
Protection, 6, 19-23
Parrott, L.J. (1987) Review of Carbonation in Reinforced Concrete, a review carried out by
the Cement and Concrete Association under a BRE contract.
Wooward, R.J and Williams, F.W (1988) ‘Collapse of Ynes-y-Gwas bridge, West
Glamorgan’, Proceding of the Institution of Civil Engineers Part I, 84
- page 22-
23. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4 Condition evaluation
We have considered the main mechanisms of corrosion in Chapter 2. We have seen that the
chemical process in the same regardless of whether the cause is carbonation or chloride
attack, as described in Chapter 3. But if we are to perform an effective repair we must fully
understand the cause and extent of damage or we risk wasting resources with an inadequate or
unnecessarily expensive repair. This chapter explains how to evaluate the condition of
corroding reinforced concrete structures.
A full evaluation is normally two-stage process. The preliminary survey should characterize
the nature of the problem and give guidance in planning a detailed survey. The detailed
survey will confirm the cause and quantify the extend of the problem. The Concrete Society
Technical Report 26 (Concrete Society, 1984) and the American Concrete Institute
Committee 222 Reports (American Concrete Institute, 1990) give excellent review of how to
conduct the surveys.
It is important to remember that corrosion is not the only deterioration mechanism in
reinforced concrete. Alkali-sillica reactivity (ASR), freeze-thaw, plastic shrinkage, thermal
movement, settlement and other movement can all lead to cracking and spalling of concrete.
Some structures may be prone to unusual chemical attack of the steel or the concrete. For
example:
• Storage vessels can contain liquids that will attack aggregates, cement paste or the
steel
• Carbonates in water can attack concrete pipelines and underground structures.
However, we will concentrate in corrosion of atmospherically exposed reinforced concrete
structures and elements.
A condition evaluation as described here is not a structural survey. A structural engineer must
be consulted if there are concerns about the capacity of the structure either because of
corrosion damage or for any other reasons. Any excessive deflection of structural elements,
misalignment, impact damage, excessive cracking, loss of concrete or loss of steel section will
require a structural evaluation before repairing corrosion damage.
4.1 PRELIMINARY SURVEY
This normally involves a visual inspection, probing of cracks and spalls to see their extent,
reinforcement cover measurement, possibly a few carbonation measurements, half cell
measurements and the taking of samples (sometime taking broken pieces of concrete rather
than coring) for laboratory testing. Particular attention must be paid to safety and structural
integrity from concrete spalling or steel section loss. Other causes of concrete cracking (e.g.
ASR, freeze-thaw, thermal movement, structural movement, impact damage) must not be
overlooked at this stage.
4.2 DETAILED SURVEY
- page 23-
24. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
The purpose of a detailed survey is to define the extent and severity of deterioration as
accurately as possible. We will need to know much damage has been done and whet has
caused the damage. Quantities for repair tenders will probably be based on the results of this
survey, so a full survey of all affected elements may be required. Alternatively a full visual
and hammer (delamination) survey may be required, with detailed measurements of half cell
potentials, chloride contents carbonation depth, cover, etc. at a few representative locations. It
is usual to produce pro formas for noting down all deterioration, as well as reading and
samples taken so that everything can be tied together in the analysis. The weather conditions
are also recorded as these can affect some readings. As example of a report drawing of a
condition evaluation is given in Figure 4.1.
- page 24-
25. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.1 Typical visual survey of a RC framed building suffering sea salt spray induced
corrosion on top cast-in chlorides.
- page 25-
26. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.3 AVAILABLE TECHNIQUES
The following sections explain the available techniques, their advantages and limitations and
the resources needed to employ them. If all the techniques available were used to survey a
corroding reinforced concrete structure thoroughly then huge resources could be used before
anything was done to stop the problem. However, expensive repairs can be useless if the
problem is not properly diagnosed. So what techniques should be used and what are their
limitations and capabilities?
Table 4.1 list most of the techniques that are or can be used for condition surveying and have
relevance to corrosion or the damage it causes. A minimum requirement would usually be a
visual survey, a delamination survey, carbonation and chloride measurements, and cover
measurements. A petrographic analysis of the concrete is also usually required. These will be
done in one or two representative or more severely damaged areas. These measurements will
tell the engineer the cause of corrosion, the extent to which chloride or carbonation have
depassivated the steel and the extent of existing damage.
Table 1 Methods for condition surveying
Method Detect Use Approximate speed
Visual Surface defects General 1 m2s-1
Hammer/chain Delaminations General 0.1 m2s-1
Cover meter Rebar depth and size General 1 reading in 5min
Phenolphthalein Carbonated depth General 1 reading in 5 min
Chloride content Chloride corrosion General Core in 10 min or
drillings in 2 min +
lab or special site
analysis
Half cell Corrosion risk General/specialist 1 reading in 5s
Linear polarization Corrosion rate General/specialist 1 reading in 5-30 min
depending on
equipement used
Resistivity Concrete General/specialist 1 reading in 20 s
resistivity/corrosion
risk
Permeability Diffusion rate General/ specialist 1 reading in 5 min or
core + lab
Impact/ultrasonics Defects in concrete Specialist 1 reading in 2 min
Petrography Concrete conditions, General Core+lab
etc.
Radar/radiography Defect, steel location, Specialist <1 m² per second for
condition vedicle system or
1m² in 20s for hand
system + interpreting
Half cell surveys are increasingly popular as a rapid way of showing how corrosion is
spreading ahead of the damage it causes. The other techniques are used for specific
requirements as described below.
- page 26-
27. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.4 VISUAL INSPECTION
The visual inspection is the first step in any investigation. It may star as a casual ‘look over’
that spots a problem and end up as a rigorous logging of every defect seen on the concrete
suface (Concrete Society, 1984).
4.4.1 Property to be measured
The aim of the visual survey is to give a first indication of what is wrong and how extensive
the damage is. If concrete is spalling off then that can be used as a measure of extent of
damage. Weighing the concrete tat spalls off over set periods can be used as a direct measure
of the deterioration rate.
4.4.2 Equipment and use
The main equipment is obviously the human eye and brain, aided with a notebook, pro forma
or hand-held computer and a camera. Binoculars may be necessary, but close inspection is
better if access can be arranged. A systematic visual survey will be planned in advance. Many
companies that carry out condition surveys will have standardized systems for indicating the
nature and extent of defects. These are used in conjunction with customized pro formas for
each element or face of the structure. It is normal to record date, time and weather conditions
when doing the survey, also noting visual observations such as water or salt run down and
damp areas. Examples are given in Figures 4.1 and 4.2.
4.4.3 Interpretation
Interpretation is usually based on the knowledge and experience of the engineer or technician
conducting the survey. The Strategic Highway Research Program (SHRP) has produced an
expert system, HWYCON (Kaetzel et al., 1994). This guides the less experienced engineer or
technician through the different types of defects seen on concrete highway pavements and
structures, including alkali-silica reaction, freeze-thaw damage and corrosion.
4.4.4 Limitations
The main limitation is the skill of the operative. Some defects can be mistaken for others.
When corrosion is suspected, it must be understood that rust staining can come from iron
bearing aggregates rather than from corroding reinforcement. Different types of cracking in
the SHRP manual on ASR (Stark, 1991) and in HWYCON (Kaetzel et al., 1994). Visul
surveys must always be followed up by testing to confirm the source and cause of
deterioration.
4.5 DELAMINATION
As corrosion proceeds, the corrosion product formed takes up a larger volume than the steel
consumed. This builds up tensile stresses around the rebars. A layer of corroding rebars will
often cause a planar fracture at rebar depth, before the concrete spalls, as shown earlier in
- page 27-
28. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figures 3.4 and 3.5. This can be detected at the surface by various means from hitting the
surface with a hammer and listening for a hollow sound to sophisticated techniques using
radar, infrared, sonic and ultrasonic equipment.
4.5.1 Property to be measured
The aim is to measure the amount of cracking between the rebars before it becomes apparent
at the surface. It should be noted that this can be a very dynamic situation. Figure 3.4 shows
the way that cracks propagate between corroding bars. The horizontal cracks are detected be a
hollow sound when the surface is hit with a hammer or a chain dragged across the surface?. It
can be detected from the effect on its physical properties associated either with the presence
of a layer or air in the concrete or the phase change from concrete to are to concrete, when
subjected to radiation or ultrasound.
4.5.2 Equipment and use
The hammer survey or chain drag (‘on decks) is usually quicker, cheaper and more accurate
than the other alternatives such as radar, ultrasonic or infrared thermography. However, these
techniques do have their uses, for instance in large-scale surveys of bridge decks (rada and
infrared thermography), or=f waterproof membranes or other concrete defects (ultrasonics
and radar). They are discussed later in this section.
The delamination survey with a hammer is often conducted at the same time as the visual
survey. Hollow sounding areas are marked directly in to the surface of the structure with a
suitable permanent or temporary marker and recorded on the visual survey pro forma.
Correctly tuned infrared cameras can be used to detect the temperature different between solid
and delaminated concrete. This is best done when the concrete is warming up or cooling down
as the delaminated concrete heats and cools faster. This means that the technique’s sensitivity
depends upon the weather conditions and the orientation of the face being surveyed. Infrared
thermography tends to work best on bridge decks in the early morning or late evening of clear
days. The best systems incorporate a visible light camera for joint recording of visual and IR
image so that the location of defects is recorded simultaneously.
Radar measures changes in the dielectric constants associated with the concrete/air phase
change. However, the radar also senses the dielectric changes at the steel-concrete interface,
the presence of water and, to a small extent, chlorides. This makes interpretation of radar
images a difficult process. In North America the main use of radar and infrared has been for
bridge deck surveys with vehicle mounted systems. In Europe and the UK hand-held systems
have been used for surveys of building and other structures. The reader is recommended to
review the literature for further information (Cady and Gannon, 1992; Bungey, 1993; Titman,
1993). There is further discussion of this topic in Section 4.12.3.
Radar is used in North America for surveying bridge decks using truck mounted rapid data
acquisition. Radar is not accurate in defining the size and location of individual delaminations
but can be used for generalized condition surveys or comparative ranking of damaged decks
(Alongi et al., 1993). Radar and infrared thermography have been used in combination in
North America. This increases the accuracy of the measurement of delaminations and other
defects but at the expense of doubling equipment and interpretation costs. The problem of
- page 28-
29. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
needing the right weather conditions for the infrared thermograms has led to a decline in the
popularity of this approach.
4.5.3 Interpretation
The interpretation of radar and infrared is a specialist process usually carried out by the
companies who have the equipment and are hired to conduct such surveys.
With a hammer or chain drag survey the experience of the operative is vital. A skilled
technician who is experienced in carrying our delamination surveys will often produce better
and more consistent results than the more qualified but less experienced engineer.
- page 29-
30. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
- page 30-
31. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.2 Visual delamination survey on a cross beam on a motarway bridge in the UK
suffering from deicing salt ingress (developed elevation of beam viewed in the direction of
increasing bent numbers)
4.5.4 Limitations
The trapping of water within cracks, deep delaminations (where bars are deep within the
structure) and heavy traffic noise can complicate the accurate measurement of delaminations
for hammer techniques. Water and deep delaminations also cause problems for radar and
infrared thermography.
It is common during concrete repairs for the amount of delamination to be far more extensive
than delamination surveys indicate. This is partly due to the inaccuracy of the techniques
available but also because of the time between survey and repair. Once corrosion has started,
delaminations can initiate and grow rapidly. An underestimate of 40% or more is not unusual
and should be borne in mind when budgeting for repairs.
Radar is reasonable accurate in predicting the amount of damage on a bridge deck, but not the
precise location of the damage. The problems with infrared thermography are getting the right
weather conditions to carry out a useful survey.
4.6 COVER
Cover measurement is carried out on new structure to check that adequate cover has been
provided to the steel according to the specifications. It is also carried our when corrosion is
observed because low cover will increase the corrosion rate both by allowing the agents of
corrosion (chloride and carbonation) more rapid access to the steel, and also allowing more
rapid access of the ‘fuels’ for corrosion, moisture and oxygen. A cover survey will help
explain why the structure is corroding and show which areas are most susceptible to corrosion
due to low cover.
4.6.1 Property to be measured
A cover survey requires the location of the rebars to be measured in there dimensions, i.e.
their position with regard to each other and the plane of the surface (X,Y) and depth from the
surface (Z). If construction drawings are not available then it may be necessary to measure the
rebar diameter as well as its location.
4.6.2 Equipment and use
Magnetic cover meters are now available with logging capabilities and digital outputs. A
spacer can be used to estimate rebar diameter. Other alternatives such as radiography can be
used to survey bridges or other structures but this is rarely cost effective (Cady and Gannon,
1992); Bungey, 1993). Cover meters are surprisingly difficult to use. They are slow, and deep
cover and closely spaced bars affect the readings. A typical device is shown in Figure 4.3. An
alternating magnetic field is used to detect the presence of magnetic materials such as steel
rebars.
- page 31-
32. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.6.3 Interpretation
One if the few standards for cover meters is BS1881, Part 204. This refers to the measurement
on a single rebar and is more concerned with the meter than taking a cover survey. Alldred
(1993) discusses cover meter accuracy when several rebars are close together and suggests
that different types of head are more accurate in different conditions. The smaller heads are
better for resolving congested rebars.
4.6.4 Limitations
The mains problem with cover measurements is the congestion of rebars giving misleading
information (Alldredd, 1993). Iron bearing aggregates can lead to misleading readings as they
will influence the magnetic field. Different steels also have different magnetic properties (at
the extreme end, austenitic stainless steels are non-magnetic). Most cover meters have
calibrations for different reinforcing steel types. The devices are slow and are not very
accurate in the field, as anyone who has tried to use one to locate steel and excavate the steel
will know. You often miss the steel that is indicated when excavating to find it to make
electrical connections or visually examine it. It is always advisable to check cover meters by
excavating and exposing one or more rebars.
4.7 HALF CELL POTENTIAL MEASUREMENTS
The electrochemistry of corrosion, cells and half cells was discussed in Section 2.4. Th half
cell is a simple device. It is a piece of metal in a solution of its own ions (ss copper in copper
sulphate, silver in silver chloride, etc.). If we connect it to another metal in a solution of its
own ions (such as iron in ferrous hydroxide, Fe(OH)2, see equation 1.1 and 1.3) there will be
a potential difference between the two ‘half cells’. We have made a battery (or an electrical
single cell to be precise). It will generate a volatage because of the different position of the
two metals in the electrochemical series (Table 2.1) and due to the difference in the solutions
(Figure 2.6). This is a galvanic cell in that the corrosion and current flow between different
metals is known as galvanic action. A second type of cell is a concentration cell that will
generate a voltage depending upon differences in the concentration of the solution (strictly the
activity), around otherwise similar electrodes.
By using a standard half cell that is in constant state, and moving it along the concrete surface,
we change our full cell by difference in conditions of the steel surface below the moving half
cell. If the steel is passive the potential measured is small (zero to -200mV against a
copper/copper sulphate half cell, or even a positive reading); if the passive layer is failing and
increasing amounts of steel are dissolving (or if small areas are corroding but the potential is
being averaged out with passive area), the potential moves towards -350mV. At move
negative than -350mV the steel is usually corroding actively. By convention we connect the
positive terminal of the voltmeter to the steel and the negative terminal to the half cell. This
gives a negative reading.
Very negative potentials can be found in saturated conditions where there is no oxygen to
form a passive layer; but with no oxygen there can be no corrosion. This shows the weakness
of potential measurements. They measure the thermodynamics of the corrosion, not the rate of
corrosion. Corrosion potentials can be misleading. Their interpretation is based on empirical
observation, not rigorously accurate scientific theory. The problem is that the potential is not
- page 32-
33. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
purely a function of the corrosion condition but also of other factor, and that the corrosion
conditions is not the corrosion rate.
The half cell potential measurement gives an indication of the corrosion risk of the steel. The
measurement is linked by empirical comparisons to the probability of corrosion.
The half cell potential is a function of the amount of iron dissolving (corroding). This is a
function of the extent to which the steel is depassivated, i.e. the extent of carbonation around
the steel or the presence of sufficient chloride to break down the passive layer, and the
presence of oxygen to sustain the passive layer. Without oxygen, iron will dissolve but will
remain stable in solution as there is no compensating cathodic reaction so the potential may be
very negative. However, the corrosion rate will be low. However, there may be a very large
(negative) potential against a standard half cell. A detailed description of half cell and its use
is given in Vassie (1991).
4.7.1 Equipment and use
The equipment used is a standard half cell. Silver/silver chloride (Ag/AgCl) and
mercure/mercure oxide (Hg/HgO) are recommended Copper/copper sulphate (CSE) cells are
also used but are not recommended because of the maintenance needs, the risk of
contamination of the cell, the difficulty of use in all orientations and the leakage of copper
sulphate.
It is important to record the equipment used. Different half cells have different ‘offsets’. The
silver/silver chloride half cell gives potentials that are a function of the chloride concentration
in them. This is usually about 130 mV more positive than a copper/saturated copper sulphate
electrode. This can be compensated for internally in the logging equipment if used or during
reporting if the ASTM criteria are being used (see below).
Standard electrode potentials are given against the ‘hydrogen scale’. That is against a standard
cell consisting of one atmosphere (strictly unit fugacity) of hydrogen gas in a solution
containing one mole (strictly unit activity) of hydrogen ions. The cell itself has platinum
electrode with hydrogen bubbling over it and is in a 1M solution of hydrochloric aci.
It is more usual to measure potentials in the laboratory against a saturated calomel electrode
(mercury in saturated mercuric chloride). This cell is recommended for calibrating field helf
cells. It is possible to adjust the chloride content of a silver/silver chloride cell so that it
behaves like a calomel cell. A calomel cell is not usually used in the field because it contains
mercury.
A high impedance digital voltmeter is used to collect the data in the simplest configuration.
Other options are to use a logging voltmeter (or logger attached to a voltmeter), an array of
cells with automatic logging or a half cell linked to a wheel for rapid data collection
(Broomfield et al., 1990). Examples are shown in Figure 4.4 (a) – (d).
Loggers can be linked to individual half cells or build into the wheel or array systems. They
will store readings and position. They will download to computers and some have built in
printers to output the data on site.
- page 33-
34. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.4 (a) A half cell and ‘Canin’ logging voltmeter. Courtesy of Proceq SA and
Hammond Concrete Services.
- page 34-
35. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.4 (b) Half cell potential measurement with the ‘Great Dane’ logging voltmeter
display. Courtesy of Germann Instruments.
The measurement procedure with a half cell is as follows:
1. Decode on the area of measurement, usually a whole element such as a bridge deck or
cross-head beam, or representative areas of an element or the structure, several square
meters in size.
2. Use a cover meter to locate the steel and determine rebar spacing
3. Make an electrical connection to the steel either by exposing it or using already
exposed steel.
4. Check that the steel is electrically continuous with a DC resistivity meter between two
points that are well separated and on well separated rebars
5. Mark out a grid. This typically be 0.2 to 0.5 m² but may be smaller, larger or
rectangular depending on the steel spacing, the geometry of the element being
surveyed and other factors determined by the experience of the investigator. The grid
may coincide with the rebar spacing on small surveys, but not usually on larger scans.
6. Check and calibrate the half cell and voltmeter
7. If necessary wet the whole area to ensure good electrical contact. Alternatively wet the
immediate area of the measurement. Tap water, soap solution and even saline
solutions have been recommended for wetting. The author prefers tap water (for a
reading to be made charged ions must flow from the steel to the half cell; the concrete
must therefore be damp enough for an ionic path; direct contact to the steel must not
occur, the current must flow as ions, not electrons).
8. take and record the readings. For manual measurements (without automated logging
equipment) it is good practice to take two immediately adjacent readings to check that
they are within a few millivolts of each other.
9. Examine for anomalies, check most negative reading areas for signs or causes of
corrosion.
Data are normally recorded on a plan reflecting the survey grid. The interpretation and
presentation are discussed below. The potential map should be drawn up while still on site in
order to check that the date are sensible and that apparent ‘corrosion host spots’ are
investigated as part of the survey.
- page 35-
36. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.4(c) The potential voltmeter for half cell potential measurements Courtesy CNS
Electronics
4.7.2 Interpretation and the ASTM criteria
ASTM C867 presents one way of interpreting half cell potentials in the field. ASTM quotes
the values against a copper/copper sulphate half cell. However, as noted above, the
copper/copper sulphate cell is not recommended and cells should be calibrated against a
calomel cell. Therefore the criteria are given in Table 4.2 against a saturated calomel
electrode (SCE), the standard hydrogen electrode (SHE) and a routinely used silver/silver
chloride electrode. The negative sign is by convention and will depend upon how the leads are
connected to the half cell and the rebar the millivolmeter.
This interpretation was devised empirically from salt induced corrosion of cast-in-place
bridge deck in the USA. Also offsets are seen for different types of concrete (precast or
containing cement replacement materials). Figure 4.5 shows the ranges of potentials seen on
different structures with different concretes and different conditions.
- page 36-
37. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
The major problems with this interpretation occur when there is little oxygen present,
especially where the concrete is saturated with water and the potentials can go very negative
without corrosion occurring. The wet bases of columns or walls often show more negative
potentials regardless of corrosion activity. Very negative potentials have measured below the
water line in marine environments; however, the lack of oxygen will often slow the corrosion
rate to negligible levels. Other problems arise in the presence of carbonation.
Figure 4.4 (d) The ‘Canin ‘multiple wheeled half cel array. Courtesy of Proceq SA and
Hammond Concrete Services.
Table 4.2. ASTM criteria for corrosion of steel in concrete for different standard half cells
Copper/copper Silver/silver Standard Calomel Corrosion
sulphatee chloride / 4M hydrogen condition
KCl electrode
>-200mV >-106mV >+116mV >-126mV Low (10% risk
of corrosion)
-200 to -350mV -106 to -256mV +116mV to - -126mV to - Intermediate
34mV 276mV corrosion risk
< -350mV < -256mV < -34mV <-276mV High (<90% risk
of corrosion)
< -500 mV < -406mV < -184mV < -426mV Severe corrosion
- page 37-
38. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.5. Potential rnges (rebar potential vs. a Cu/CuSO4 half cell). From Baker (1986).
In carbonated concrete the anodes and cathodes are so close together that a ‘mixed potential’
an average of the anode and cathode) is measured. Also carbonated concrete wets and dries
quickly as the pores are partly blocked by the calcium carbonate deposits. This means that the
resistivity of the concrete will affect the measurement. If the reading is taken with no wetting
a very positive reading may be found. If the reading is taken after wetting the measurement
may be drift more negative for many hours.
A third problem arises due to the existence of the carbonation front. This is a severe change in
the chemical environment from pH 12 to pH 8, i.e. a factor of 104 difference in the
concentration of the hydroxyl and hydrogen ions. There are similar changes in the calcium
and other metal ion concentrations in solution as they precipitate out on carbonation. This can
lead to a ‘junction potential’ superimposed on the corrosion potential giving rise to
misleading results.
For carbonated concrete the best method is first to do a potential survey with minimum
wetting (if stable potentials can be established). Then wet the surface thoroughly and leave it
for at least two hours or until potentials are stable. Resurvey once potentials have stabilized
and then look for the most anodic (negative potential) areas. These are most probably active if
a potential difference of 150mV or more exists over a space of 1.0m or less. A physical
investigation is essential to see if there is reasonable correlation between corrosion and anodic
areas.
- page 38-
39. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Given the thermodynamic nature of the measurement and our current understanding of
potential measurements, that is probably the best that we can do presently on interpresentation
of half cell potentials. A fuller description of interpretation methods for chloride
contamination is given in Vassie (1991). For all its limitations, the half cell is a very
powerful diagnostic tool for corrosion investigation. Its main problem is that some people rely
on too simplistic interpretation. Half cell potentials are not corrosion rate measurements.
Corrosion rate measurement is discussed below.
Stray electrical currents can also influence the readings. These were discussed in Section 2.3.
the effect of stray current on half cell potentials can be used as a diagnostic tool where stray
current corrosion is suspected in the presence of DC fields. If a half cell is mounted on or in
the concrete linked to a logging voltmeter, then any fluctuations in potentials may be linked to
t he operation of nearby DC equipment, especially if the equipment can be deliberately turned
on and off and the potentials fluctuate accordingly.
There has been a tendency to correlate half cell potentials with corrosion rates. The half cell
potential is a mixed potential representing anodic and cathodic areas on the rebar. It is not the
driving potential in the corrosion cell. Any correlation between potential and corrosion is
fortuitous and is often due to holding other variables constant in laboratory tests.
There has been some discussion above and in the literature of the ‘junction potentials’ created
by the change in chemical concentrations within the concrete (Bennett and Mitchell, 1992).
This effect was severe in a concrete slab subjected to chloride removal, but than may be due
to the treatment (discussed in Section 6.10), rather than being a real problem under normal
conditions. However, the junction potential may explain the erratic changes in potentials seen
in carbonated structures. This is because potentials exist across the carbonation front due to
the pH change, and there are concentration changes as carbonated concrete wets and dries
quickly because of the lining of the pores by calcium carbonate.
A histogram or cumulative frequency plot will show what proportion of measurements exceed
the criteria to show the extent of high risk (see ASTM C876, 1991). Where the ASTM criteria
do not apply they will show the distribution of readings so that high risk areas can be
identified.
The best way of interpreting half cell potential data is to expose areas of rebar which show the
most negative potentials, intermediate and least negative potentials to correlate corrosion
condition with readings. If there are severe potential gradients across the surface then
corrosion is likely to be localized with pitting present. Care must be taken in areas of
moisture, e.g. puddles that remain on the surface of where moisture comes up from the
ground. These may show very negative potentials due to chloride accumulation or to oxygen
starvation. Corrosion rte measurements may be required to determine whether the potential is
an artfact or due to high corrosion rates.
4.7.3 Half cell potential mapping
A further understanding of the corrosion conditions is given by drawing a potential map of the
area surveyed. This is a printout of the readings where lines are drawn separating the levels of
potential, i.e. joining the points of equipotential with “contour lines”. This shows the high
corrosion risk areas and the low corrosion risk areas. A rapid change in potential is seen as a
steeper gradient.. this indicates a greater risk of corrosion. Figure 4.6 shows a typical half cell
potential plot for a brigde beam, which correlates with the visual survey in Figure 4.2. While
- page 39-
40. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
these isopotential contour plots are not as quantitative as simply following the ASTM criteria
such mapping of anodic areas is valid over a wider range of structures and conditions. The
distribution of points between the different contour levels is given.
A line potential measurements can be plotted on a distance vs. potential plot. This will show
which points exceed the ASTM criteria and where the steepest potential gradients are and the
most anodic areas are with the most negative values. A set of contour plot, three dimensional
plot and line scan is show in figure 4.7 (from Broomfield et al., 1990).
- page 40-
41. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
- page 41-
42. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.7.4 Cell to cell potentials
If it is not feasible to make direct connections to the reinforcing steel it is possible to obtain
comparative potential data by measuring the potentials between two half cells, with one kept
in a fixed position and the other moved across the surface. This is the linking of two full cells,
one kept constant (the fixed cell and the steel directly below it), while the other half cell
moves, changing the steel to concrete half cell and thus the second full cell. We will not know
the absolute value to the steel to concrete half cell, but we will measure how it changes from
point to point. Interpretation of data from fixed versus moving half cell surveys is more
difficult than interpreting absolute values, but again, the contour plot is probably the most
useful method of interpretation.
- page 42-
43. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.8 Carbonation depth measurement
Carbonation depth is easily measured by exposing fresh concrete and spraying it with
phenolphthalein indicator solution. The carbonation depth must then be related to the cover
(the average and its variation) so that the extend to which carbonation has reached the rebar
can be estimated, and the future carbonation rate estimated.
There was some discussion in Section 3.1 about whether the carbonation front is truly as well
defined as the indicator shows it to be; but for most practical considerations it is a very
accurate and reliable technique. Some dark coloured fine aggregates can cause problems by
making the colour transition difficult to see. Very poorly consolidated concrete and concrete
underground exposed to dissolved carbonates in the water may not show clearly defined
carbonation fronts due to the non-uniform progress of the carbonation front.
The cutting up of cores to expose a fresh surface should be done carefully to prevent dust
from carbonated areas contaminating the uncarbonated surface and vice versa.
4.8.1 Equipment and use
Carbonation is easily measured by exposing fresh concrete and spraying on phenolphthalein
indicator as shown in Figure 4.8. This can be done either by breaking away a fresh surface
(e.g. between the cluster of drill holes used for chloride drilling as described in Section 4.9),
or by coring and splitting or cutting the core in laboratory. The phenolphthalein solution will
remain clear where concrete is carbonated and turn pink where concrete is still alkaline.
The best indicator solution for maximum contrast of the pink colouration is a solution of
phenolphthalein in alcohol and water, usually 1g indicator in 100 mm of alcohol/water (50:50
mix) or more alcohol to water (Building Research Establishment, 1981; Parrott, 1987). If the
concrete is very dry then a light misting with water before applying the phenolphthalein will
also help show the colour. Care must be taken that dust from drilling, coring or cutting does
not get on the treated surface. Other indicator such as thymolphthalein, Alizarin yellow and
universal indicator have been used, along with pH meters. However, phenolphthalein is the
mst reliable, convenient and widely used indicator (Parrott, 1987).
4.8.2 Interpretation
Carbonation depth sampling can allow the average and standard deviation of the carbonation
depth to be calculated. If this compared with the average reinforcement cover then the amount
of depassivated steel can be estimated. If the carbonation rate can be determined from
historical data and laboratory testing then the progression of depassivation with time can be
calculated.
- page 43-
44. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
Figure 4.8 Carbonation test on a reinforced concrete mullion. In this case the carbonation had
only penetrated about 5mm on an element with a 25 mm cover. Phenolphthalein solution
sprayed on to the freshly exposed concrete shows pink in the alkaline areas (the darker central
area around the rebar), and remains clear in the carbonated areas near the surface.
4.8.3 Limitations
Phenolphthalein changes colour at pH 9. The passive breaks down at pH 10-11. If the
carbonation front is 5 to 10 mm wide, the steel can be depassivated 5mm away from the
colour change of the indicator as shown in Figure 3.1. This should be considered when using
phenolphthalein measured carbonation depths to determine rate and extend of depassivation.
Some aggregates can confuse phenolphthalein readings. Some concrete mixes are dark in
colour and seeing the colour change can be difficult. Care must be taken that no
contamination of the surface occurs from dust and phenolphthalein sprayed surface must be
freshly exposed or it may be carbonated before testing.
It is also possible for the phenolphthalein to bleach at very high pH, e.g. after chloride
removal (or possibly realkalization). If the sample is left for a few hours it will turn pink.
There can also be problems on below ground structures where carbonation by ground water
does not always produce clear carbonation front induced by atmospheric CO2 ingress.
The number of test areas is usually limited by the amount of concrete that those responsible
for the structure will allow to be broken off or holes drilled and cores taken. It is often
necessary to do such sampling away from public gaze so it is rarely possible to carry out
carbonation tests systematically. They are usually done at accessible locations where damage
is not too obvious and repair is easy.
- page 44-
45. Corrosion of Steel in Concrete : Understanding, Investigation and Repair
4.9 Chloride determination
Chlorides are usually measured by dissolving powder samples in acid. The samples are taken
from drillings or from crushed cores. It is preferable to collect a series of drillings at different
depths so that a chloride profile can be produced. Alternatively a core can be cut into slices
and the slices crushed. Like the carbonation tests the chloride profile must be related to cover
so that the extend to which rebars are exposed to high chlorides can be determined.
Chloride profiles can also be used to determine the diffusion coefficient and thus predict the
ongoing rate of ingress.
The carbonation thresholds were discussed in Section 3.2.3. It is important to recognize that
these are approximate. It is also important to realize that the chloride level at the rebar
determines the present extend of corrosion, but the profile (Figure 3.2) determines the future
rate, as that is what drives more chlorides from the concrete surface into the steel surface.
Chloride contents can be measured by several methods. In the laboratory, powdered samples
are usually digested in acid and then titrated to find the concentration in the conventional wet
chemical method.
In the field there are two well known methods of measuring chlorides: Quantab strips and
specific ion electrodes. The former are of modest accuracy and can be made inaccurate by
certain aggregate types. The latter can be highly accurate. However, the equipment is
expensive and requires training and a good methodology to use effectively. Any field
technique should be checked against laboratory analysis of duplicate samples.
The results of the above methods are referred to as the ‘total’ or ‘acid soluble’ chloride
contents. There are also methods for measuring the ‘free chlorides’ or water soluble rather
than the acid soluble chlorides. This refers to the fact that it is the chloride dissolved in the
pore water that contributes to the corrosion process. Any chlorides chemically bound up in the
cement paste (chloroaluminates or C3A), or bound up in the aggregates are ‘background’
chlorides that should not contribute to the corrosion threshold (Section 3.5).
Unfortunately the water soluble techniques produce results that are difficult to reproduce so
they are rarely used in the UK or Europe. However, the ‘Soxhlet extraction technique’, a
method of refluxing concrete chips in boiling water to extract the chloride is a standard
technique used in North America. For example, AASHTO T260 Sampling and Testing for
Total Chloride Ion in Concrete and Concrete Raw Materials includes a procedure for water
soluble chloride ion analysis, covering sample preparation and analysis.
Chlorides can be cast into concrete or can be transported in from the environment. The
chloride ion attacks the passive layer even through there is no significant, generalized, drop in
pH (puts in the steel can become very acidic, see Section 2.3). Chlorides act as catalysts to
corrosion. They are not consumed in the process but help to break down the passive layer of
oxide on the steel and allow the corrosion process to proceed quickly.
Chloride testing will show:
1. Whether chlorides are present in a high enough concentration to cause corrosion.
Typically concrete with more than 0.4% chloride by weight of cement is at risk of
corrosion.
2. 2. Whether chloride were cast in or diffused in later. The spatial distribution of
chloride with depth and about the structure will show a profile with depth if chlorides
- page 45-