Materials and Structures (2019) 52:2

https://doi.org/10.1617/s11527-018-1301-1 (0123456789().,-volV)
(0123456789().,-volV)

50 YEARS OF MATERIALS AND STRUCTURES

Propagation of reinforcement corrosion: principles, testing

and modelling
Carmen Andrade

Received: 13 September 2018 / Accepted: 4 December 2018 / Published online: 26 December 2018
 The Author(s) 2018, corrected publication 2019

Abstract Reinforcement corrosion is the risk most The inability to ascertain the area affected when an
frequently cited to justify concrete durability research. electrical signal is applied to large-scale elements is
The number of studies specifically devoted to corro- described, along with the concomitant need to use a
sion propagation, once the object of most specialised guard ring to confine the current or deploy the potential
papers, has declined substantially in recent years, attenuation method. The reason that measurement with
whilst the number addressing initiation, particularly contactless inductive techniques is not yet possible
where induced by chlorides, has risen sharply. This (because the area affected cannot be determined) is
article briefly describes the characteristics of steel discussed. The method for integrating corrosion rate
corrosion in concrete that need to be stressed to dispel over time to find cumulative corrosion, Pcorr, is
certain misconceptions, such as the belief that the explained, together with its use to formulate the
corrosion zone is a pure anode. That is in fact seldom mathematical expressions for the propagation period.
the case and as the zone is also affected by microcells, The article concludes with three examples of how to
galvanic corrosion accounts for only a fraction of the use corrosion rate to assess cathodic protection, new
corrosion rate. The role of oxygen in initiating low-clinker cements or determine the chloride thresh-
corrosion, the scant amount required and why corro- old with an integral accelerated service life method.
sion can progress in its absence are also discussed.
Another feature addressed is the dependence of the Keywords Reinforcement corrosion  Oxygen
chloride threshold on medium pH and the buffering supply  Microcells  Corrosion rate  Propagation
capacity of the cement, since corrosion begins with period
acidification. Those general notions are followed by a
review of the techniques for measuring corrosion, in
1 Introduction
particular polarisation resistance, which has proved to
be imperative for establishing the processes involved.
Social appreciation for the built heritage is growing. The
conservation and maintenance required to prevent its
C. Andrade (&) deterioration entails high costs, however, given the
International Centre for Numerical Methods in characteristics of all the technologies and processes
Engineering, CIMNE- UPC-Spain, Barcelona, Spain
e-mail: candrade@cimne.upc.edu involved in rehabilitation and repair. Beyond such
economic considerations, deterioration diminishes func-
C. Andrade tionality and may even pose the risk of accidents with the
International Centre for Numerical Methods in loss of human life due to unexpected collapse. Buildings
Engineering, CIMNE- UPC-Spain, Madrid, Spain
2 Page 2 of 26 Materials and Structures (2019) 52:2

and infrastructure form an essential part of developed devoted exclusively to corrosion in 1981. A photo of
societies, and as such their usability to the highest standards the participants is reproduced in Fig. 1. One of
of safety is an indisputable priority. The material most scientists in the picture, Kyosti Tuutti, published a
widely used in the modern built heritage is concrete and much-cited and impactful thesis [3] in 1982, describ-
corrosion of its steel reinforcement heads the list of its ing a two-period (initiation and propagation) model.
possible causes of decay. A number of Rilem committees were subsequently
Concrete underwent ceaseless development in the created, but only one on reinforcement corrosion,
twentieth century for many reasons, one of which, directly RILEM TC-154-EMC: ‘‘Electrochemical Techniques
related to this study, was that most design engineers for Measuring Corrosion In Concrete’’ chaired by the
deemed it to be eternally maintenance-free, in the author, published Recommendations on the tech-
understanding that cement alkalinity affords chemical niques for measuring corrosion, which are still in
protection for the steel and the cover constitutes a physical effect [4–7]. Later committees went on to develop
barrier much more effective than paint. It was not until the service life models, in particular in connection with
nineteen seventies and early eighties when Rilem began to chloride diffusion [8], or determine the steel depassi-
explore reinforcement corrosion, a subject that aroused the vation limit or its governing parameters.
interest of very few researchers. As new construction was The many and very valuable scientific and technical
booming and time-mediated damage had not yet been papers that have been forthcoming in the interim attest to
detected, only a handful of cases of corrosion had been the economic implications of the subject and the
identified, despite one very distressing instance of collapse scientific challenges posed. The author apologises for
[1]. By way of example of the scant attention lent to listing but a few of those papers here [9–14], chosen
corrosion, when the author defended her Ph.D. thesis in because they deal specifically with active steel corrosion.
1973, she only found around 40 articles on the subject in This article summarises some of the principles
all the journals on concrete published at the time. associated with corrosion, its measurement and mod-
Pioneering the concern was Committee 60-CSC, Corro- elling in the propagation stage, as a tribute to Rilem in
sion of Steel in Concrete, chaired by Peter Schiessl, who its 75th anniversary and as a token of my gratitude for
published a summary of the state of the art at the time [2]. the inspiration I have drawn from its working
Engineer Hans Arup, who also alerted to the approach throughout my career. Rilem’s routine
problem early on, organised the first workshop practice of creating short-lived technical committees,

Fig. 1 Participants in the workshop entitled ‘Corrosion of steel O. Vennesland, (9) O. Gjorv, (10) C. Page, (11) R. Bakker, (12)
in concrete’ organised by Hans Arup at Copenhagen in 1981 In J. Dawson, (13) Byfors (14) P. Schiessl, (15) ??, (16) ??. (17) G.
the first row from right to left: (1) ??, (2) K. Tuutti, (3) ??, 4) J.A. John (18) C. Andrade, third row: (19) ??, 20) ??, (21) Wilkins,
Gonzalez, (5) ??, 6) C. Locke, (7) C. Hansson, second row: (8) (22) H. Arup, (23) ??, (24) ??, (25) ??, (26) ??, (27) ??
Materials and Structures (2019) 52:2 Page 3 of 26 2

consistently mandated to explore innovative and are often poorly understood, they are dealt with in the
topical concerns, expands the frontiers of knowledge, sections listed below.
invariably challenging presumably established princi-
1. Corrosion cell and acidification of the corrosion
ples. This review calls some of those principles into
zone
question and identifies the many gaps still extant in our
2. Microcells and macrocells (corrosión rate and
understanding of reinforcement corrosion.
galvanic current)
3. Effect of the oxygen
4. The role of potential
2 Basic principles and mechanisms
5. Composition of the rust formed
6. Effect of temperature
Concrete corrosion is an electrochemical process, for
it involves both chemical reactions and the circulation Although these effects feed back into one another,
of an electrical current, in which the reagents are they are addressed separately here for greater clarity.
electrical charges (electrons and ions). In essence,
1. Corrosion cell and acidification of the corrosion
corrosion is triggered when concrete pH at the
zone
interface with the steel drops to levels below around
8 [15], dissolving the passive layer that forms in the
As Evans showed [16] nearly 100 years ago with his
highly alkaline aqueous phase. Under those circum-
droplet experiment (Fig. 2), corrosion begins at the
stances the iron atoms in the steel convert to positively
atomic level with the formation of microcells. When a
charged ions (oxidation), generating an excess of
drop of salt water is left on steel the cell forming in the
electrons in the base metal:
centre of the drop is anodic because it contains less
Fe ! Feþ2 þ 2e ð1Þ oxygen than the border of the droplet, which becomes
cathodic. Evans proved that anodes and cathodes were
To maintain its electrical neutrality, the metal in formed (Fig. 1) by adding phenolphthalein (reduction
adjacent areas induces another reaction (reduction) generates hydroxyls, colouring the outer side of the
which in neutral and basic media uses dissolved drop pink) and sodium sulfocyanide (which when
oxygen: combined with iron yields a blue iron sulfocyanide
2O2 þ 4e ! 4OH ð2Þ complex). If left standing the colours mix in the drop
due to the formation of many nano- and microcells.
That oxidation–reduction (redox) mechanism sup- According to Pourbaix [15], the pH dips to around
ports the corrosion ‘cell’, an only apparently simple 3–5. Corrosion always entails local acidification in the
development, for neither oxidation nor reduction takes zones affected [17].
place in a single stage. Rather, both entail a series of
intermediate species, whilst electron migration from Feþ2 þ H2 O ! FeðOHÞ2 þHþ ð4Þ
the steel to the reagent depends on the so-called
‘tunnel effect’ to overcome the energy barrier
Consequently, the buffering capacity of portlandite
involved. Therefore, although the basic mechanisms
is of key importance not only in retarding carbonation
governing steel corrosion may be simplified, sight
but in chloride-induced depassivation in concrete.
should not be lost of their complexity, a feature they
When the first iron atoms are released into the solution
share with nearly all chemical reactions taking place in
by the depassivating action of the chlorides, they
successive or simultaneous stages. As a result, in the
hydrolyse the water, lowering the local pH. That effect
spontaneous reaction (corrosion-free potential), the
is neutralised by the buffering capacity of the
total anode and the cathode current densities are equal
hydroxyls, sourced primarily from the portlandite, as
and equal as well to the corrosion current density.
well as from the alkalis, which repair the passive layer,
ia ¼ ic ¼ icorr ð3Þ mitigating the decrease in pH. The critical chloride
content is reached when the amount of dissolved Fe??
As some of the most significant characteristics of
ions, generating protons by inducing water hydrolysis,
these basic processes and their governing parameters
cannot be neutralised. Consequently, the Cl-/OH-
2 Page 4 of 26 Materials and Structures (2019) 52:2

Fig. 2 a Evans droplet basic reactions (Note that water hydrolysis lowers the pH in the anodic zone and the formation of hydroxide ions
raise it in the cathodic zone) and b some initial pitting in a rebar

ratio governs the chloride corrosion threshold. In calcium carbonate forms with the calcium ion of
chloride attack, acidification is local because the portlandite, the pH declines and corrosion is driven
attack is localised and the rest of the pores retain their across the entire surface by many micro-anodes and
alkalinity, passivating the rest of the metal surface. micro-cathodes (Fig. 4). In that case acidity is gov-
Moreover, the polarisation and lower potential erned by carbonate/bicarbonate ions, which tend to
induced by cell formation beginning at the pit polarise maintain a neutral pH at values of around 7.
the surrounds to the pit potential (Fig. 3). That creates One important implication of acidification in the
a need for more chlorides to exceed the Cl- limit at the corroded areas is that the cathodic reaction may reduce
lower mixed potential induced by pit formation, not only the oxygen in the adjacent passive areas, but
ultimately protecting the surrounds. The attack there- also the protons in the corrosion zone itself. Corrosion
fore proceeds in depth instead of spreading laterally, may consequently proceed independently via the
which may nonetheless occur, depending on the interior microcells (self-catalysis) if the source of
degree of polarisation at the pit borders and pit growth. oxygen to the adjacent areas is interrupted (Fig. 3). In
In carbonation, hydroxil ions from portlandite other words, protons may be reduced in a cathodic
exert a buffering effect across the entire surface, reaction, a development that has been detected in the
attempting to offset the carbon dioxide-mediated form of H2 bubbles inside pits [18].
reduction of pH: as the hydroxil ions are consumed
H3 Oþ þ e ! 2H2 "ðgasÞ ð5Þ
Acidification and the repassivation capacity of the
passive layer are the factors that determine whether the

Fig. 4 Microcells generated by generalised carbonation-medi-

Fig. 3 Potential and pH values in a corrosion pit [15] ated depassivation
Materials and Structures (2019) 52:2 Page 5 of 26 2

attack is local (proceeding depth-wise) or laterally to corrosion current density, but unless the anodic area
widespread. In stainless steels with a very high is very small, they do not account for a majority
repassivation capacity, corrosion proceeds as very proportion in the total corrosion current. In other
small pits that bore quickly inward. Therefore, whilst words, the corrosion current measured is the sum of
the use of stainless steel ensures very high chloride the microcell and the galvanic currents. The inference
thresholds, once corrosion is triggered and the pits is that the galvanic current measurement alone cannot
become active, the attack is very dangerous because be directly likened, nor is it necessarily proportional,
the pit may grow depth-wise in a narrow area with a to the corrosion rate. This issue will be discussed more
scant loss of material. In contrast, in black steel with a fully in the section on measuring techniques.
lower repassivation capacity, the pits are shallower
Icorr ¼ Imicro þ Igalv ð6Þ
and spread sideways.
That corrosion is primarily supported by microcells
2. Microcells and macrocells (corrosión rate and
in the corrosion zone was confirmed long ago, as the
galvanic current)
earlier reference to Evans’s droplet experiment shows.
Atomic force microscopy (AFM) has recently corrob-
In microcell corrosion the anodes and cathodes are
orated those findings. In [19] a drop of saltwater lying
separated by just a few microns, whereas in macrocell
on a steel plate was observed under an atomic force
corrosion they may be spaced at several centimetres or
microscope (AFM) at different ages. Figure 5 shows,
even metres. Metres-wide macrocell separation can
in a 5 micron window, that the ferrite acting as the
only exist in concrete in marine media where electrical
anode dissolved whereas the cementite acting as the
charges can circulate across long distances thanks to
cathode remained intact (Fig. 5).
the low resistivity of seawater. Corrosion rate and
The behaviour clearly visible in Fig. 5 for the metal
galvanic current are different concepts. The reference
in a prestressing steel tendon in which the layers of
area for expressing the galvanic current, which is the
cementite and ferrite alternate, may vary in other steel
current circulating between the anodic and cathodic
microstructures. Generic corrosion is summarised in
zones, is the anodic area only, whereas corrosion rate
Fig. 6, which shows that at the nano- and micro- levels
refers to the sum of the anodic and cathodic areas.
some areas of the steel are anodic and others cathodic.
Corrosion, particularly in concrete, is invariably
The figure also describes the mechanism in which
supported mostly by microcells, with macrocells
water is hydrolysed by Fe??, generating protons that
contributing to the corrosion rate only marginally
acidify the anodic zones.
(see Figs. 3 and 4). If corrosion proceeds at the micron
On a somewhat larger (millimetre) scale, the trial
level, what role do passive zones much greater than the
conducted by Mansfeld [20] with a Kelvin probe,
corrosion area play in local corrosion? Whilst along
likewise with a drop of saltwater, is also illustrative.
with galvanic current, Ig, they indisputably contribute

Fig. 5 Three-dimensional AFM images of the sample surface after a 0 h 0 min; and b 2 h 15 min of exposure to a 0.05 M NaCl
solution [19]
2 Page 6 of 26 Materials and Structures (2019) 52:2

The mixed or corrosion-free potential measured on

a metal bar immersed in an electrolyte would lie in-
between the anodic and cathodic semi-reaction corro-
sion potentials. The exact value would depend on how
the process is conducted and whether the anode and
cathode are coplanar and positioned face-to-face. As
shown in Fig. 8, steel bars may be placed face-to-face
in separate chambers connected by a salt bridge
(Fig. 8, left) or in the same chamber separated by a
concrete disc acting as a diffusion cell (Fig. 8, right).
Electrical resistances are then introduced in the circuit
between anolyte and catholyte to simulate different
types of concrete and degrees of water saturation.
Figure 9 shows some of the results obtained from
Fig. 6 Representation of microcell activity
this experiment using galvanised steel – black steel
couples. In the figure Sc = cathodic bar surface and
SA = anodic (galvanised) bar surface [21].
Inasmuch as the salt bridge itself generates ohmic
resistance, in the graph on the left in Fig. 9 the
resistance drop prevents the potentials from concur-
ring. The mixed potential is any value from the
cathodic - 500 mV to the anodic - 1400 mV. In the
middle graph in Fig. 9 the mixed potential continues
to differ where the gap between anodic and cathodic
potential is wide in high resistance values, but as that
resistance lowers the mixed potential ultimately
concurs with the anodic potential. Given that the
anode is aeriated, it can support corrosion by itself
with its microcells. The graph on the right depicts the
Fig. 7 Mansfeld experiment with a Kelvin probe [20]: potential situation when the anode is non-aeriated. Here also, if
profile for a drop of saltwater resistance in the medium is very high the potentials
fail to concur. In contrast, when resistance declines
As Fig. 7 shows, the potential is positive only on the both the anode and the cathode are polarised, yielding
outer side that acts as a cathode, whereas the potential an intermediate mixed potential for the galvanic
on the inner side adopts very negative values. couple. The explanation is that due to non-aeriation,
corrosion in the anode (right in the figure) is

Fig. 8 Setup for studying galvanic currents in concrete with face-to-face alignment
Materials and Structures (2019) 52:2 Page 7 of 26 2

bar embedded in concrete bar in solution without O2

0
-100

Ecorr (mV)
-200
-300
-400
-500
-600
-700
0.1 1 10 100 1000 10000
Time (days)

Fig. 9 Ecorr values for simulated macrocells versus the log of Fig. 10 Corrosion potential in a bar embedded in undersea
the resistance between anolyte and catholyte for an Fe/Zn concrete and in another submerged in a non-aeriated solution
galvanic couple [21]—[Legend: blank/solid squares: anodic
corrosion potential before/after connecting anode and cathode; 2.1 Passive conditions
blank/solid circles: cathodic corrosion potential before/after
connecting anode and cathode]. Ecorr values of simulated
macrocells in function of the inert resistances for the couple For oxygen to feed the cathodic reaction it must be
Fe/Zn dissolved in the pore solution. In the absence of
pollutants oxygen contributes to thickening the pas-
sive layer. The variation over time in the corrosion
potential characteristic of a bar embedded in an
less supported by microcells, raising the proportion of undersea structure with a 7 cm cover is graphed in
galvanic current in the corrosion current. Fig. 10. The potential evolves more positive up to
That is, if the anodic zone is as aeriated as the values of ? 80 mVAg/AgCl, denoting steel passivity
cathode (middle of Fig. 9), the mixed potential is the and the presence of a compact, late age passive layer
anode potential, whereas the potential adopts an due to an uninterrupted supply of oxygen. The
intermediate value when the corrosion zone has figure also shows the fluctuation in corrosion potential
restricted access to oxygen. In addition to these in a bar immersed in a 0.5 M NaCl solution bubbled
factors, the value of the mixed potential is affected with nitrogen throughout the test to maintain the
by the size of the cathodic relative to the anodic area. dissolved oxygen at 0.3 ppm or under. The potential
Another aspect that influences the contribution of the under these conditions is much more cathodic because
galvanic current to the corrosion rate is since in a given the low O2 concentration induces structural changes in
bar the anodic and cathodic zones are in the same the passive layer.
plane (coplanar), the galvanic current is transmitted The experiment with the non-aeriated solution
laterally, whereby the ‘transmission line model’ depicted in Fig. 10 confirms the near impossibility
should be applied in its calculation [22] and may of lowering the oxygen content of a solution below
result of smaller proportion than face-to-face 0.1 ppm by de-aerating with nitrogen [23], due to the
arrangement. oxygen impurities in this gas. The potential is
consequently around - 500 mVAg/AgCl and not lower.
3. Effect of the oxygen
If the oxygen level declines further, however, the
The role of oxygen in corrosion generation and potential would shift to more cathodic values
progress in concrete has been misconceived in some (Fig. 11), entering the generalised corrosion region
respects. Trying to control oxygen very seldom limits (Fig. 11) and, in the absence of depassivating ions,
concrete corrosion. Oxygen is essential for depassi- triggering passivation. Lower potential values would
vation but not for corrosion progression. The descrip- carry the current into the cathodic protection region,
tion of its role can be divided into two stages, as with the reduction of water as per the cathodic
discussed below: under passive conditions and in the reaction:
presence of active corrosion. 2H2 O þ 4e ! H2 " þ 2OH ð7Þ
2 Page 8 of 26 Materials and Structures (2019) 52:2

situations are seldom present in concrete, even when it

is underground or undersea.
The behaviour deducible from Fig. 10 likewise
stands as evidence that cover thickness has no impact
on oxygen availability around the reinforcement. That
is, tantamount to saying that although corrosion at
different depths logically begins at different times
because it takes chloride ions longer to penetrate
thicker covers, the corrosion rate is similar. Therefore,
oxygen concentration is either the same at all depths or
it does not control the process. Models that make
corrosion rate dependent upon oxygen permeability
are consequently erroneous.
Fig. 11 Possible trends of polarization curves of steel embeded
in concrete. If no breadown is found before around
? 650 mVAg/AgCl the steel is passive. If the curve bends before 2.2 The role of oxygen in depassivation
such potential with a sudden increase in the current, pits were and propagation of corrosion
generated. In the case of generalized corrosion, no decrease of
the current denoting passive film formation from around
- 650 mVAg/AgCl is noticed and the current continuously As said, sufficient oxygen is normally present in
increases concrete pore solutions to maintain passivity. That
notwithstanding, in alkaline media it plays an essential
role in the initiation of corrosion, but not in its
propagation. In the presence of a high enough chloride
To put it briefly, under passive conditions the value
content, oxygen at concentrations as small as a few
of the corrosion potential is governed by oxygen.
tenths of ppm suffices to initiate corrosion (localised at
When the oxygen concentration declines, the potential
the site where the oxygen is in contact with the metal
shifts to more cathodic values, hampering depassiva-
surface). At low oxygen concentration, corrosion is
tion. However, as Fig. 10 shows, oxygen has fairly
much localised, as shown in Fig. 12 and Table 1. The
ready access to the reinforcement not only in dry but
figure plots the fluctuations in potential and corrosion
also in saturated concrete, where the oxygen content in
rate in two bars immersed in saturated calcium
the pore solution suffices to maintain and raise steel
hydroxide solutions, one with 0.5 M NaCl and the
passivity. The reason is that water contains around
other with 2.5 M NaCl, bubbled with N2 gas through-
8 ppm of oxygen, which is more than enough to
out the test. (The bar in the 0.5 M solution is the same
heighten the passivation of concrete-embedded steel.
shown in Fig. 10) Despite an O2 content of \ 1 ppm,
Oxygen concentration in water depends on salt
corrosion was noticed, although even at the high
concentration and temperature and although oxygen
corrosion rates measured, a post-test visual inspection
solubility is lower in highly saline solutions, such

10
Icorr (µA/cm2)

0,5M NaCl
2,5M NaCl
0,1
0 200 400 600 800
Time (hours) Time (hours)

Fig. 12 Corrosion potential (left) and corrosion rate (right) of steels immersed in 0.5 M and 2.5 M NaCl solutions
Materials and Structures (2019) 52:2 Page 9 of 26 2

Table 1 Aspect of the bars tested in 0.5 M NaCl and 2.5 M NaCl solution deaerated bubbling nitrogen during 5 days and 30 days

0.5M NaCl solution 2.5M NaCl solution

Steel after immersion for 30 d Steel after immersion for 5 d

Table 2 Aspects of bars in 0.1 M NaCl and 0.5 M NaCl solutions submitted to potentiostatic testing

POTENTIAL VALUE 0.1 M 0.5 M

APPLIED

– 650mV

– 550mV

– 350mV

of the bars immersed in the 2.5 M solution revealed are indicative of active corrosion even when rust is
only sporadic corrosion staining and very small pits. A not visible to the naked eye until it reaches certain
bar tested for 5 d in 2.5 M NaCl (which exhibited a dimensions (hence the readier visibility in the month-
corrosion-free potential more anodic than that in the long than in the 5 day experiment).
0.5 M solution) is depicted in Table 1. When potential is not spontaneous but electro-
Rusting was more visible in the 30 day test at chemically induced (potentiostatic test), the results are
0.5 M NaCl. The corrosion products formed along a as given in Table 2. At potential values more cathodic
line in the bar surface which may have been than - 650 mV (Ag/AgCl electrode), no pits are
generated during manufacture (Table 1). Crevice detected. Pits form at - 550 mV, while generalised
corrosion was also observed at the edge of the tape corrosion appears at - 350 mV.
used to delimit the working area. In both the 0.5 M As commented, once the depassivation is produced
and the 2.5 M NaCl solutions only the areas with and sufficient rust formed, oxygen enhances the
surface imperfections such as seams or rolled laps, corrosion rate but it is not necessary for the corrosion
corroded. These imperfections seem to be the to progress, due to the protons and the Fe??? ions may
preferred sites for pit initiation. High Icorr values feed the cathodic reaction.
2 Page 10 of 26 Materials and Structures (2019) 52:2

300 corrosion may take place with scarce oxygen given

200 sufficiently high chloride content [23]. In concrete semi-
Potential (mV, SCE)

100
submerged in seawater, corrosion tends to begin in the
0
more aeriated tidal zone, which constitutes a sacrificial
-100
-200
anode for the rest of the structure, inasmuch as saltwater
-300
is a very good conductor. The differential aeriation
-400 between tidal and undersea zones gives rise to a galvanic
-500 macrocell. In the absence of such a difference in
-600 potential, corrosion may begin in the underwater zones.
-700 While not immune to corrosion, given the more cathodic
0 1 2 3 4
potential present, such underwater zones fail to depas-
% Total Cl (cement weight)
sivate until a higher chloride concentration is reached.
Fig. 13 Potentials applied to bars embedded in mortar (mean of 5. Composition of the rust formed
10 trials) vs total Cl- content needed to induce depassivation
(chloride threshold) [24, 25]
As the reaction products may vary widely and involve
4. The role of potential many intermediate species, establishing the basic
mechanism calls for very specific studies and the
The role of potential in depassivation is illustrated in application, for instance, of ring-disks [26] that are
Fig. 13, where breakdown potential is plotted against unusable in concrete. Although the description of the
chloride concentration threshold. Further to the time- mechanism is normally simplified when intended for
honoured fundamentals on breakdown potential in the practical purposes, sight should not be lost of the
presence of depassivating ions, for localised corrosion to complexity ensuing from its many stages and inter-
take place, the potential must be greater than the mediate species [15, 26–28] with charge imbalances
potential at which passivity breaks down, i.e., when due and solvated ions surrounded by different proportions
to the presence of oxygen, the potential exceeds the of water molecules. The species involved include:
breakdown potential threshold. Figure 13 [24], inspired [Fe(OH)?], [FeCl?], [Fe(OH)2], [Fe3O4], [c-Fe2O3],
by Pourbaix’s [15] findings, represents potentiostatic [c-FeOOH], or green rust gels with variable compo-
tests in mortars made with different types of cement sitions [29], etc. The most stable rusts are mixes of
[25].Different potentials are applied to bars embedded hydroxides and oxides (called oxy-hydroxides).
in specimens immersed in a NaCl solution and the The oxides involved in steel corrosion in concrete
current is recorded. When the amount of chlorides are the same as observed in its corrosion in the
arriving to the bar surface exceeds the threshold, the atmosphere, including nearly all the most common
current rises steeply, denoting breakdown of the passive iron oxides [27], such as magnetite, goethite and
layer. Potential, then, controls tolerance to one chloride lepidocrocite and, in the presence of chlorides, the
threshold or another. The more positive the spontaneous akaganeite. Given such diversity, all attempts to relate
potential, the more Fe??? found in the passive layer oxide composition under real exposure conditions to
(due to more oxygen present), the more vulnerable is corrosion mechanisms have proven futile, for compo-
passivity to breakdown and the lower is the chloride sition fluctuates constantly, in keeping with the
level required to initiate local corrosion. As the graph in balance between acidity (in turn dependent upon
Fig. 13 shows, when the potential applied is high, fewer chloride or bicarbonate ion concentration) and oxygen
chlorides are needed to induce corrosion, whereas when availability. When exposed to natural media, the
the potential shifts to negative values, the chloride composition and dynamics of the oxides present are
threshold is higher. The graph mimics the natural pattern neither predictable nor indicative of corrosion rate.
described earlier, in which a low enough potential Although composition is variable, certain differ-
prompts cathodic protection. ences exist between the oxides detected in the early
Briefly, potential is an indication of the level of stages of corrosion, depending on whether it is
oxygen present. Where it is not under - 650 mV, the carbonation- or chloride-induced (with the appearance
risk of corrosion subsists and local or generalised in the latter of akaganeite, an oxide that takes up
Materials and Structures (2019) 52:2 Page 11 of 26 2

Fig. 14 Clustering of Fe ion tetrahedra into octahedra and solvated particles

chlorides in its structure). Chlorides generally give rise reinforcement had corroded completely and disap-
to more soluble oxides that diffuse farther than those peared entirely in extensive areas [31].
involved in carbonation. As corrosion progresses the rust no longer migrates
Since the corrosion zone acidifies, the oxide ions but remains at the bar/concrete interface where,
form a suspension that progressively gels due to depending on the steel microstructure or the presence
coagulation of the individual constituents (Figs. 14, of moisture in the concrete, it spreads by layers or
15). The figure shows that the OH--solvated tetrahe- cycles (Fig. 15). This rust detaches readily when
dron housing the Fe ion evolves into an octahedron accessed, but may afford misleading information, for
that begins to coagulate with others, forming larger the residual sound diameter may appear to be larger
particles that ultimately constitute a network of than the actual uncorroded diameter (Fig. 15g).
crystalline oxides. Another fact worth mentioning is that section losses
These grains can migrate across the pores and of 20% or higher may render the steel more brittle, for
spread, precipitating and clustering until they appear the hydrogen generated during corrosion may diffuse
on the surface. When the resulting gels [29] find toward the steel.
alkaline cathodic zones they coagulate further and If the pressure exerted by the lack of space to
precipitate, ceasing to spread as they convert into accommodate the rust forming suffices to crack the
laminar rust or solid particles. They diffuse neither concrete, cracking begins at the reinforcement surface
continuously nor at a continuous rate. Rather, diffu- (Fig. 16) and proceeds toward the closest concrete
sion adapts to the steady increase of concrete resistiv- surface, where it emerges and spreads practically
ity over time, rendering corrosion rate resistivity- linearly [32] with corrosion penetration, Pcorr between
dependent [30]. 10 and 50 microns. Corrosion-induced cracking has
been a popular line of research in recent years, with
2.3 Cover cracking effect any number of papers on analytical and numerical
calculation methods. Many fewer experimental stud-
If the reaction proceeds, the rust on the concrete ies have been conducted, however [33, 34].
migrates across its pores. Although the rust exerts The general consensus is that as an attack of under
pressure from the outset, due to the porosity of the 10 lm to 50 lm suffices to crack the cover, cracking
steel/concrete interface the concrete is unaffected by may be a visual symptom of the presence of corrosion.
the full thrust of the pressure. When the pores at the The earlier corrosion is detected the better, whilst
interface become saturated, the full radial pressure is attempts to prevent it by adding fibres to the concrete
transferred to the concrete, ultimately prompting or wrapping with carbon based textiles, may be
cracking at the surface of the bar that extends to the counterproductive.
surface of the concrete (Fig. 15). The more porous the
concrete, then, the longer it takes for rust to exert 2.4 Alternative cathodic reactions
effective pressure that could crack it. Very porous or
moist concrete may not crack at all. The cantilevered Another consideration to be borne in mind with
roof over Maracaná Stadium, for instance, contained respect to the oxides formed during corrosion is that
chlorides due to the haste with which it was built. Sixty the ferric oxides forming with Fe?? oxidation may in
years later, no cracks were detected even though the
2 Page 12 of 26 Materials and Structures (2019) 52:2

Fig. 15 Rust in different

physical conditons in
concrete

(a) Initial ocalized corrosion (b) Rust in wet conditions (c) The same rust after drying
showing that the rust is
falling down and is gelous

(d) Oxides diffusing out of the rebar in (e) Oxides diffusing out of the rebar in
concrete containing chlorides carbonated concrete

(f) Old rust sjhowing parallel-to-the-bar (g) Rebar so corroded that has a thick oxide
layering layer in the surface

turn be reduced to ferrous ions, binding electrons and when corrosion is said to be self-catalytic and explains
contributing to the cathodic reaction. that it may proceed with no or scant access to oxygen.
Feþ3 þ e $ Feþ2 ð8Þ 6. Effects of temperature
Whether corrosive action continues therefore Temperature is known to accelerate chemical reac-
depends not only on oxygen availability in the still tions by reducing activation energy. The Arrhenius
passive zones. It may also be supported by the equation is normally used to calculate the effect of a
oxidation–reduction of the oxides forming in the change in temperature on the rate constant and
corrosion zones [35] and by the proton reduction therefore on the rate of reaction [36]. However, this
generated in water hydrolysis. That is what is meant
Materials and Structures (2019) 52:2 Page 13 of 26 2

Fig. 16 Left: Strain gauges on concrete in a prismatic specimen. Right: crack induced corrosion at initial stages

equation can be applied (providing the result is

considered an Apparent Activation Energy), if the
concrete is saturated, but not if it can exchange
moisture freely with the environment, because then the
changes are due not only to the activation energy
barrier. Thus, higher temperature entails evaporation,
rendering Arrhenius inapplicable. In other words,
temperature cycles involve variations in humidity Fig. 17 Section and diameter loss due to generalized (left) and
localized (right) corrosion
inside concrete that preclude using Arrhenius
equation. penetrate the concrete from outside, the outcome is
Fluctuations in temperature also change the solu- normally an asymmetrical loss of diameter (Fig. 17,
bility of the components of the aqueous phase, which left). The residual diameter, Ut, is therefore equal to
may not necessarily rise with temperature. Portlandite the initial diameter, U0, less the corrosion penetration,
and sulfates, for instance, are less soluble at higher Pcorr. When corrosion is localised (Fig. 17, right), the
temperatures. Furthermore, temperatures of over depth of the deepest pit (Ppit) determines the residual
60 C may decompose sulfoaluminate phases, releas- diameter.
ing sulfates into the aqueous phase. Another signifi- Loss can be measured gravimetrically, i.e., (a) by
cant effect of rising temperature is the decline in gas weighing the sample before and after testing it in
solubility in liquids. Oxygen is consequently expelled specific environments, to find the cumulative loss,
at higher temperatures, which translates into less Pcorr, expressed in g/cm2 of area exposed to corrosion,
rather than more corrosion. For all the foregoing, the or (b) by using electrochemical techniques able to
Arrhenius equation is not recommended to be applied measure the metal weight loss through Faraday’s law.
in corrosion in concrete.
By way of summary, all these developments make 3.1 Electrochemical techniques
corrosion much more than a simplistic process
involving anodes and cathodes in which iron oxidises The oldest of such techniques, the corrosion potential
and oxygen is reduced. A rigorous interpretation must measurement, is qualitative, indicating only the risk of
take all the intermediate steps and the many stages corrosion, as specified in (4). A second technique,
governing them into consideration. which measures concrete resistivity, which is an
indication of the degree of material saturation (5),
was recently proposed by the author to characterise
3 Corrosion rate and measuring techniques overall concrete durability, as discussed in a subse-
quent section.
Corrosion entails a loss of metal. Since in the case of The technique that has been widely shown to
reinforcement steel the aggressive substances quantify corrosion most accurately consists in
2 Page 14 of 26 Materials and Structures (2019) 52:2

measuring polarisation resistance, Rp, or linear polar- The ratio between Rp and corrosion rate, Icorr, is
isation (6). Over the years, this technique, first applied known as the ‘Stern formula’.
to concrete in the early nineteen seventies [37, 38] to
study a series of corrosion inhibitors, has proven [39] B
Icorr ¼ ð10Þ
to be the most appropriate, yet to be excelled by any Rp
other, for quantifying and monitoring material loss
Between 1970 and 1978 [37], further to an
both in concrete and any other electrolyte/metal
approach developed by the author, constant B in that
system. Inasmuch as corrosion is an electrochemical
expression when applied to concrete, was quantified
event, such techniques are indisputably the most
by calibration with simultaneous gravimetric losses
suitable for quantification purposes.
(Fig. 18) and application of Faraday’s law to convert
Linear polarisation, Rp [6, 37], routinely used in
lA/cm2 to corrosion penetration in mm/yr.
all metal/electrolyte systems, consists in applying no
The constant B was found to be 52 mV in the
more than 20–30 mV above the corrosion-free
passive state and 26 mV in the active or corroded
potential (DE) of the embedded steel to induce an
state, although 26 mV can be used in all cases
electrical current (DI). The duration of the increase
inasmuch as an error of a factor of 2 in passive steel
in potential should be at least 30–60 s to obtain the
is negligible, in light of the gravimetric losses. Three
correct Rp value, for at shorter times the value is
important issues, have to be taken into account
affected by the transitory electrical capacity induced
regarding this technique (6).
by the double layer at the steel/electrolyte interface.
Measurements lasting only a few seconds, such as in 1 lA/cm2 \ [ 0:0116 mm/year ð11Þ
some devices, yield erroneous values due to the non-
dissipation of that effect. The duration of testing to
ensure optimal results is one of the features 1. As the area exposed to corrosion must be taken
addressed in the earliest studies of this technique into account, Rp is expressed as Xcm2 and hence
[37, 39]. corrosion rate in lA/cm2. Rp,ap is the value not
  normalised to the area of the polarised metal.
DE
Rp ¼ ð9Þ 2. In concrete the upper case ‘I’ rather than the lower
DI
case ‘i’ (=uniform corrosion current density) is used

Gravimetric loss ( mg / cm2 )

100 1000
OPC 300kg/m3-w/c 0.5
FAC 300kg/m3-w/c 0.5
OPC 400kg/m3-w/c 0.38
Electrochemical Weight Loss (g)

OPC 400kg/m3-w/c 0.6 NaCl 0.5M

100 14

15
HCl
7
10 8
12
FeCl
6 5 2

10
3 13
17 16 20 FeCl2 0.2M
18
19 1 2 +
4
Ca(OH)2
10
sat
11

1 1
1 10 100 1 10 100 1000
2
Gravimetric Weight Loss (g) Electrochemistry loss ( mg / cm )

Fig. 18 Examples of comparison of gravimetric weight loss to electrochemical loss in steel embedded in concrete: (left) submerged in
sea water for several years [40], (right) of steel submerged in solutions at different pH’s values Fig. 19 [41]
Materials and Structures (2019) 52:2 Page 15 of 26 2

1000 solutions with different pH used to simulate the values

Averaged corrosion rate (uA/cm2.day)

inside and outside a pit induced by chlorides (x) given

100 in Fig. 19. In this figure, it is appreciable that only
Goñi et all
CCR 1990 solutions with a pH of over 12.5 exhibit Icorr values of
10 FeCl2 (N2)
less than 0.1 lA/cm2. That does not mean, however,
FeCl2 (O2) that corrosion begins at pH values of \ 12.5, for the
1
HCl (N2)
pH attendant upon carbonation-mediated corrosion
has been shown to be around 8 (15). Depassivation as a
0.1
HCl (O2) result of a general decline in pH would, then, depend
0.01 FeCl2 + on the composition of the pore solution and call for
Ca(OH)2 values of around 8.
literature
0.001 values
0 2 4 6 8 10 12 14
pH 3.2 Electrochemical impedance
Fig. 19 Icorr values (log scale) vs pH (aeriated and non-aeriated
solutions) Electrochemical impedance spectroscopy (EIS) deliv-
ers results very similar to those found with linear
as an indication that corrosion may be localised (not polarisation. As noted many years ago [39], the
visible in concrete). The use of icorr is only correct parameters are the same in the two, but expressed
when corrosion is generalised and uniform. differently: as direct current in terms of time (linear
3. In the absence of coloured rust (=where the steel polarisation) or as frequency (EIS). Since EIS is more
remains passive), Icorr is less than 0.1–0.2 lA/ work-intensive, its use is entirely unnecessary in most
cm2 and it is greater than those values, when cases. Its application is only justified in specific studies
coloured rust is detected. Values of over 0.1 lA/ where, for instance, the object is to determine
cm2 may be temporarily recorded soon after the capacity-related properties in the double layer of the
steel is immersed in the alkaline solution due to concrete or the corrosion mechanisms involved. It is
passive layer formation. Once the layer becomes often used to establish concrete resistivity, which can
uniform the Icorr values decline to under 0.1 lA/ be found quickly with a single pulse using only one
cm2, whether the steel is immersed in a solution frequency. It is not discussed in detail in this review
or embedded in concrete. When the steel is paper in light of such specific applicability.
actively corroding, Icorr values lower than
0.1 lA/cm2 may nonetheless be found when 3.3 Contactless measurement
the concrete is dry.
As accessing the reinforcement entails breaking the
The aforementioned technique is applicable to all concrete cover, corrosion should ideally be measured
systems, as shown by the mean corrosion rates for

STEEL IN DISTILLED WATER.

POLARISATION OF THE STEEL
-0,160
-0,170
-0,180 Iap = 500 µA
ELECTROLYTE
E (V)

-0,190
Ie Re -0,200

A B -0,210 I ap = 50 µA

-0,220
Iap RM
IM -0,230
0 5 10 15
METAL t (s)

Fig. 20 Left: Arrangement of electrodes in the Inductive method, Middle: simplified analogue model, Right: polarization detected
when applied external current
2 Page 16 of 26 Materials and Structures (2019) 52:2

3.4 Measurements in small- and large-scale

elements

Devices able to measure Rp are based on potentiostat-

galvanostats, which can apply a fixed potential or
current. The technique can be readily deployed in
small specimens and all commercial devices include
the option. The problem in large members in real
Critical length structures is that as the counter-electrode is much
smaller than the reinforcement area, the potential
Fig. 21 Scatter with distance of the steel bar of the current scatters across large distances, as shown in Fig. 21.
applied through the small counter The problem of finding the critical length can be
countered in one of two ways.
on the surface, with no physical contact with the bar.
That approach, which had never before been used to • In the guard ring method for current confinement
measure corrosion (even though the corrosive power [44] the ring must be modulated [adapting the outer
of stray current had long been known), was shown by ring current to confine it to the circle pre-
the author to be viable in small-scale specimens in a determined by the two inner Vss electrodes
paper published in 2001 [42] and was subsequently (Fig. 22)]. The current is not correctly confined
patented in Spain (Patent No. ES 2 237 241 B). It was without those electrodes (6). Only one commercial
based on a simplification of the analogue electrical device correctly confines the current with this
model (Fig. 20) [43]. method; other market devices without such inner
Contactless measurements in large specimens need electrodes deliver erroneous corrosion rate values,
to account for the reinforcement length (critical under- or over-confining the central current.
length, see later) reached by the electrical pulse, as • In potential attenuation method [45], electrodes
is made with normal corrosion measurements. No aligned with the counter-electrode (Fig. 23) mea-
publications were found by the authors on how to sure the distance reached by the potential applied
know this critical length in the contactless method. and use the total area polarised (6). Only one
Then, the application to large structures of this new commercial device (the one correctly confining the
non-contact method with the same level of accuracy of current) features this method, which should only be
the traditional one should be questioned, until the implemented when the structure is very moist and
scientific basis and convincing results are published. satisfactory confinement is not possible.

Fig. 22 Modulated
confinement of the current:
uses a Guard ring which is
controlled with the two
(Vss) electrodes in between
guard ring and central
counter-electrode [44]
Materials and Structures (2019) 52:2 Page 17 of 26 2

Fig. 23 Attenuation of the potential method: the distance reached by the current is measured through the side electrodes at different
distances [45]

Fig. 24 Values of the Icorr-Ro

relation Icorr-resistivity 100
obtained in real structures
and trend established
Corrosion rate Icorr (uA/cm2)

10

high
1

moderate
low
0.1
negigeable
corrosion
0.01
0.01 0.1 1 10 100 1000 10000
Resisvity (Kohm·cm)

All that need be done to verify whether a device i.e., is proportional to the resistance of the medium.
delivers correct values is to measure samples contain- Empirical verification of that fact [30] led to the so-
ing passive steel. Devices that fail to confine the called corrosion rate-resistivity (Icorr-q) diagram
current satisfactorily cannot deliver values below (Fig. 24), from which the following expression was
0.1 lA/cm2, the correct rate for passive steel. This is derived:
relevant, for instance, to determine whether corrosion h  i 26
inhibitors effectively reduce the corrosion rate [44]. Icorr lA cm2 ¼ ð12Þ
q½kX  cm
3.5 Resistance control. Relationship between Rp
and Rohm The relationship exhibits some scatter, depending
on the temperature at which resistivity is measured and
One issue with significant implications that has not the porosities and resistivities of the concrete pore
been addressed sufficiently in the literature to date is solution, although the general pattern is consistently
that in concrete Icorr is subject to resistance control, observed. With that empirical relationship, Icorr can be
2 Page 18 of 26 Materials and Structures (2019) 52:2

related both to resistivity and hence to the effective

diffusion coefficient [46], from which the apparent
diffusion coefficient can be deduced. Thus,
k
De ¼ ¼k r ð13Þ
q

26  105 26;000
Then q¼ ¼ ð14Þ
De Icorr
which can be rearranged as:
26;000 Fig. 25 Quantification of the propagation period from several
Icorr ¼  De ¼ De  108 and corrosion rates
26  105
De ¼ Icorr  108
classified into four levels as shown in Fig. 25 and the
This reasoning can be carried one step further, by difference between instantaneous corrosion rate, Icorr,
again comparing that equation [12] to the basic and cumulative corrosion penetration, Pcorr (6), can be
equation for Rp [10], whereby Rp can be equated to expressed.
resistivity: Icorr values of over 100 lA/cm2, equivalent to a
26 26 diameter loss of over 1 mm/year, are extremely high,
Icorr ¼ ¼ ! Rp ¼ q ð15Þ for the very highest values measured in concrete [48]
Rp q
are on the order of 100–200 lA/cm2. Those should
In a concrete specimen with actively corroding steel consequently be the maxima used to artificially
and a resistivity of 20 kXcm, for instance, the bar accelerate corrosion with an external current [33].
corrosion rate can be found with Eq. (15): Tests with Icorr [ 200 lA/cm2 create unrealistic con-
26 ditions that yield types of rust differing from those
Icorr ¼ ¼ 0:254 lA/cm2 ð16Þ actually found due to the high acidity generated by
20
such high currents.
As the relationship only holds in the presence of Integrating Icorr over time has also provided
active reinforcement corrosion, it is not applicable insight into the effect of climate, leading to the
when the steel is still passive, for in such circum- establishment of patterns and the definition of the
stances deducing the Rp from the q value would yield yearly representative corrosion rate, Icorr,REP. Fig-
wholly erroneous results. ure 26 reproduces an example of instantaneous
corrosion rate in carbonated concrete over several
years [50] and Fig. 25b the respective cumulative
4 Propagation period model corrosion penetration values, Pcorr. The graphs in
Fig. 25c, d show the same parameters for concrete
Once the corrosion rate is quantified, the effect of any with chlorides and the Fig. 25e, f are of concretes
variable that alters it [47], such as inhibitors or the submerged in seawater.
presence of moisture in carbonated or chloride-bear- The inference drawn from these behaviours is that
ing concrete, can be studied, the condition of a real corrosion can evolve over time presenting different
structure can be assessed, and the propagation period trends. Either the corrosion evolves linearly (Fig. 25
[48] quantified by integrating the value over time. line A) or accelerately (line C) or it may attenuate (line
Pcorr ½mm B). The difference can be impacted by cracking. In
tp ½years ¼ h i ð17Þ carbonated concrete, once formed, the cracked region
mm
Vcorr year
on the steel surface remains dry longer than the
uncracked zones (Fig. 24a, b). When cracks form in
On those grounds and bearing in mind the structural
underwater structures, the opposite occurs, for the
consequences of corrosion [49] the Icorr values can be
ingress of seawater across the cracks accelerates
Materials and Structures (2019) 52:2 Page 19 of 26 2

Fig. 26 a Evolution of the instantaneous corrosion rate of a d accumulated corrosion of the specimen in c, e evolution of the
carbonated specimen, b Accumulated corrosion penetration instantaneous corrosion rate of specimen immersed in sea water,
(mm) of the carbonated specimen, c evolution of the instanta- f accumulated corrosion of the specimen in sea water
neous corrosion rate of a specimen mixed with 3% CaCl2 and

corrosion (Fig. 24e, f). In summary, the propagation 4.1 Diameter loss and structural implications
period can be expressed as a linear relationship
between time and Pcorr when Icorr is constant, or as Although the relationship between corrosion and
two bilinear periods that may also be expressed structural decay is not addressed in this paper,
mathematically as exponential or potential, (Fig. 27). diameter loss in reinforcement is acknowledged as
A Representative yearly corrosion rate, Icorr, REP the basic parameter for quantifying the loss of bearing
can be obtained from the slope of the evolution of Pcorr capacity (Fig. 28) [43]. In future, this link should be
with time. In the linear trend the value will be unique used for developing better models of corroded struc-
when in the bilear trend, there would be two tures in order to optimize the calculation of the
representative corrosion rate. residual safety.
2 Page 20 of 26 Materials and Structures (2019) 52:2

1. Verification of the efficacy of cathodic protection

C 2. Accelerated service life testing: diffusion coeffi-
cient, chloride threshold and corrosion rate in
Pcorr (mm)

A concrete; inhibitor efficacy

3. Corrosion resistance of new binders.
B

5.1 Verification of the efficacy of cathodic

tp tip2 protection

tim The uncertainties associated with the verification of

Initiation Propagation
cathodic protection efficacy are well known, for none
Fig. 27 Bilinear trend of the propagation period [51] of the methods tested to date ensures that the steel has
repassivated or corrosion has been reduced to negli-
gible levels. The methods most commonly used are:
• ‘instant-off’ potential measurement
Structural performance

• potential change after 4 h of depolarisation

• 24 h depolarisation potential.
Present performance
Efficacy can be accurately measured, however, if in
addition to any of these methods, Rp is measured on
Minimum acceptable site. Icorr should be measured by suitably confining the
current (modulated confinemenet). The condition can
RESIDUAL LIFE
be measured when the steel is disconnected from the
Initiation of corrosion propagation power source. Depolarising after 4–24 h, its Icorr value
indicates whether or not the steel is passive. Figure 29
Time illustrates the case of a bridge. In the left of the figure it
is represented the decay in potential registered when
Fig. 28 Structural decay curve in Manual Contecvet [43]
depolarizing after 6 months with cathodic protection.
The decay is smaller than the required (100 mV) while
5 Examples of use of corrosion rate measurements
in the right of the figure is represented the Icorr before
and after the 6 months. The values of Icorr have
Three examples are presented for the practical appli-
however shown a clear decrease in the corrosion rate.
cation of the principles described before.

10
100

2 h decay 24 h decay
Icorr (uA/cm2)

80
1
E (mV)

60

40 0,1

20
0,01
0 P4 P9 P14 P25 P32 P37
P4 P9 P14 P25 P32 P37 21st July 2010
15 and 16th December 2010

Fig. 29 Left: Depolarizing decay after 6 months of application of cathodic protection. Right: the values of corrosion rate before and
after 6 months of application of the PC
Materials and Structures (2019) 52:2 Page 21 of 26 2

Fig. 30 Set up for the accelerated Integral method for chloride ingress testing

5.2 Accelerated service life testing: diffusion • The other specimen is split open to:
coefficient, chloride threshold and corrosion
• Extract a sample of the concrete in contact with
rate in concrete; inhibitor efficacy
the chloride solution from which to find the
chloride surface concentration prevailing dur-
Accelerated testing to determine the service life of a
ing the test
concrete mix is one of the challenges facing specialists
• Observe the corrosion induced in the steel by
today. The procedure normally involves testing the
the chloride front visualised by spraying the
mix for chloride penetration, either under accelerated
material with silver chloride
or natural diffusion, to find the diffusion coefficient
• Extract a sample of the concrete/steel interface
(Dap). That, in conjunction with the respective age
to determine the (threshold) chloride content
factor value, the surface concentration and an assumed
that induced such corrosion.
chloride threshold, can be used to predict steel
depassivation time. • The time to depassivation is then used to calculate
Present tests [52] deliver only the value of, the apparent diffusion coefficient (Fig. 31) through
coefficient Dap and of surface concentration: all the the application of the equation of Table 3. Fig-
others are assumed. An accelerated ‘integral’ method ure 31 also shows the chloride threshold
has been developed [53] to find the values of the measured.
service life prediction parameters determining:
Figure 32 shows the registration of the corrosion
• surface concentration (likewise derivable from rate Icorr with time from the initiation of the test. In the
other tests) figure several concentrations of the same inhibitor are
• the chloride threshold compared. It is evident that A0 with no inhibitor
• corrosion rate after depassivation. presents the earliest depassivation while the rest
present later depassivation times as the inhibitor
The method, now standardised in Spain [54],
concentration increases. The findings were reasonable
consists in embedding one bar each in two cubic
and while the long-term results regarding inhibition
concrete specimens and conducting an accelerated
power are unknown, comparisons that help classify the
chloride migration test by applying 12 V between two
starting materials in the concrete may be drawn from
electrodes until the steel depassivates, as shown in
the short-term data. Until now it was found that the
Fig. 30). The subsequent steps are as follows.
values obtained with this accelerated test are consis-
The drop in exterior voltage is interrupted in one of
tent with those found with the natural diffusion test
the specimens, allowing corrosion to proceed for at
[55].
least 15 days to measure the Icorr values that appear
This accelerated method delivers more experimen-
spontaneously.
tal parameters than the diffusion coefficient test alone.
2 Page 22 of 26 Materials and Structures (2019) 52:2

Dap inhibitor A
1.00E-07 0.3
inhibitor A

Chloride threshold (% dry mas

0.25
8.00E-08
Dap (cm2/s) integral test
0.2

of concrete)
6.00E-08
0.15
4.00E-08
0.1

2.00E-08
0.05

0.00E+00 0
A0 A1.5 A2 A 2.5
Dns inhibitor A

Fig. 31 Values of (left) Apparent Diffusion coefficient and chloride threshold (right) for different concentration of inhibitor. A0 is the
reference without inhibitor

Table 3 Equations to calculate the Dap from the integral test types. The smaller proportion of portlandite attendant
[53] upon the reduction in clinker content translates into
2
x RTL x2   ½cm2 =s less effective buffering of the decline in pH induced by
Dns ¼ 2tmig zFðD/1:6Þ ½cm2 =s Dns;dif
2teq;dif 
t0
ti
corrosion. As noted earlier, the buffering capacity of
the so-called ‘alkaline reserve’ in cement plays a
pivotal role in long-term concrete durability.
Steel passivation capacity or corrosion resistance in
any type of cement, including new binders with much
less clinker, has traditionally been studied on the
grounds of electrochemical techniques. The method-
ology was exemplified in [55].
A thorough study of corrosion resistance calls for
specific passivation tests and plotting polarisation
curves for different amounts of chlorides, a discussion
of which lies outside the object of this article. Be it
said, however, the comparison with a type I reference
cement can be made by testing the chloride and
carbonation resistance.
• Chloride resistance can be determined with the
‘integral’ method described above. Mortars or
Fig. 32 Values of Icorr with time recorded during the integral
test which show that the A0, without inhibitor, depassivates first concretes can be prepared with the cements to be
and the increasing concentrations of inhibitor give progressive tested, comparing with a cement type I the Dap’s,
later depassivation times the chloride thresholds, the surface concentrations
and the corrosion rates.
More specifically, it constitutes an accelerated method • Carbonation resistance calls for two types of
for determining critical chloride content. specimens.

5.3 Corrosion resistance in new binders • Cubic non-reinforced specimens are used to
observe carbonation penetration in natural or
The intense efforts presently being made to lower the accelerated (B 3% CO2) tests.
clinker content in cement as a way of reducing its • The second test type comprises reinforced
carbon footprint entail the use of additions of different specimens with a thin (around 0.5 cm) concrete
Materials and Structures (2019) 52:2 Page 23 of 26 2

Fig. 33 Scheme of testing new binders

cover to shorten the time required for carbon- corrosion rate and in fact account in general for a very
ation to reach the steel/concrete interface. CO2 small percentage of that rate. Their measurement may
concentrations of up to 20% may be applied for therefore detect only part of the process.
accelerated carbonation, as the aim is full Oxygen is needed to trigger corrosion in concrete,
carbonation rather than rate comparison. After but as tenths of ppm suffice for localised corrosion
the specimen is fully carbonated, it is exposed where chloride content is high, the process may occur
to high humidity (submerged conditions) to even underwater. Less aeriated zones cannot therefore
induce corrosion. The Icorr values observed for be said to be immune to corrosion. Oxygen is not
the new binders are then compared to those required for corrosion to proceed (due to water
found for a type I reference cement. hydrolysis and the reduction of the oxides formed),
although its presence intensifies the process, which is
Experiments of this nature are summarised in [55]
enhanced by the sum of all the cathodic processes.
and Fig. 33. Reports of resistance to depassivation
Concrete was one of the first systems in which the
must specify whether chloride or carbonation testing
instantaneous corrosion rate, also known as ‘polarisa-
was involved. Concrete resistivity and cement buffer-
tion resistance’ and ‘linear polarisation’ (alluding to
ing capacity are both instrumental in chloride-medi-
the fact that a small alteration in potential is used to
ated depassivation. In contrast, buffering capacity is
induce a linear increase in current), was measured. The
the predominant mechanism for resisting the effects of
technique has been widely used ever since it was
carbonation.
described in the ASTM G59-97 (2014) standard for all
metal/electrolyte systems. Attempts to discredit it
have been unable to disprove its good correlation with
6 Final comments
weight losses simultaneously measured by gravimetry
and converted using Faraday’s law. At this writing its
Steel corrosion takes place in concrete in much the
direct current version (more readily interpreted than as
same way as in the atmosphere or the soil, although
electrochemical impedance) continues to be the easi-
cement generates a unique alkaline medium. Corro-
est and most convenient technique for quantifying
sion is triggered by local carbonation or chloride
measuring corrosion in any metal/electrolyte system.
attack-mediated acidification and proceeds primarily
The application of Rp to real structures is influenced
via microcells. Corrosion zones in concrete, even
by the quasi-infinite area of the reinforcement spray-
where involving a single pit, are not pure anodes
ing longer than the auxiliary electrode border: inas-
connected to a cathode. Galvanic currents between
much as the actual area of the metal polarised by the
corrosion and passive zones are not equivalent to
signal must be known, the area of the auxiliary
2 Page 24 of 26 Materials and Structures (2019) 52:2

electrode is not the correct reference. That circum- Acknowledgements The author acknowledges the
stance necessitates specific devices in which measure- collaboration of all those contributed to the work presented
here. The original research reported in this review paper was
ment is based on modulated current confinement or in supported by the following grants: public Agencies belonging to
measuring the attenuation of potential with the the several Ministries of Research of Spain, the European
distance from the auxiliary electrode border. Although Commission (BRITE-EURAM and Marie-Curie -Nanocem
a method for contactless corrosion measurement was grant) and the private firm GEOCISA (corrosion-rate meter).
Also would like to mention Prof. José Calleja who had the vision
published in 2001, to date no market device has proven to start the study of reinforcement corrosion from the end of
able to take accurate measurements in large-scale 1960’s in the Institute of Construction Sciences of Spain.
structures (despite claims to the contrary), for the area
affected has not yet been determined nor have studies Compliance with ethical standards
on its calibration been forthcoming. Conflict of interest The author declares that she has no con-
Periodic measurement of corrosion rate has pro- flict of interest.
vided the grounds for calculating cumulative corro-
sion and proposing corrosion propagation period Open Access This article is distributed under the terms of the
models. In outdoor structures, seasonal cycles can be Creative Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits use,
applied to show that corrosion may be constant if duplication, adaptation, distribution and reproduction in any
expressed as a yearly mean (Icorr,REP). In ‘resistance medium or format, as long as you give appropriate credit to the
control’ of corrosion, also shown to exist in concrete, original author(s) and the source, provide a link to the Creative
the rate is proportional to resistivity. This provides Commons license and indicate if changes were made.
grounds for certain analogies, such as deducing
corrosion rate from resistivity (under conditions of References
active corrosion) or relating it to the effective diffusion
1. Isecke B (1983) Failure analysis of the collapse of Berlin
coefficient. Congress Hall. In: Crane AP (ed) Corrosion of reinforce-
An incorrect diagnosis of the causes of corrosion ment in concrete, chapter 5. Elseiver, Amsterdam, p 79
and how it proceeds is known to be one of the major 2. Schiessl P (1989) Corrosion of Steel in Concrete. Report of
reasons for the limited durability of concrete repairs. RILEM TC 60-CSC RILEM, Chapman & Hall, New York
3. Tuutti K (1982) Corrosion of steel in concrete. Swedish
Corrosion rate measurement is essential to studying Cement and Concrete Institute (CBI), Stockholm. pp 4–82
and diagnosing progression of this process in a given 4. Elsener B, Andrade C, Gulikers J, Polder R, Raupach M
structure. Activity in each zone of the structure can (2003) RILEM TC 154-EMC: electrochemical techniques
only be reliably determined by measuring corrosion for measuring metallic corrosion half-cell potential mea-
surements—potential mapping on reinforced concrete
rate. Engineers engaging in the repair of deteriorated structures. Mater Struct 36:461–471
structures must acquire more specialised knowledge to 5. Polder R, Andrade C, Elsener B, Vennesland O, Gulikers J,
ensure effective intervention. Structural behaviour is Weidert R, Raupach M (2000) Test methods for on-site
in itself complex, but the basic processes triggering measurement of resistivity of concrete’’, RILEM TC
154-EMC: electrochemical techniques for measuring
corrosion are much more complex than the mere metallic corrosion. Mater Struct 33:603–611
presence of a corrosion cell with an anode and 6. Andrade C, Alonso C, Gulikers J, Polder R, Cigna R,
cathode. Another challenge to be confronted is to Vennesland Ø, Salta M, Raharinaivo A, Elsener B (2004)
understand the structural behaviour of concrete with RILEM TC 154-EMC: electrochemical Techniques for
Measuring Metallic Corrosion. Recommendations Test
microcracks around the reinforcement, for with the methods for on-site corrosion rate measurement of steel
concomitant impact on its bond with the steel it ceases reinforcement in concrete by means of the polarization
to behave as a composite material. resistance method. Mater Struct 37(273):623–643
Intense basic research on reinforcement corrosion 7. Vennesland Ø, Raupach M, Andrade C (2007) Recom-
mendation of Rilem TC 154-EMC: ‘‘electrochemical
will continue to be needed in the years to come to fill techniques for measuring corrosion in concrete- measure-
the existing gaps in the knowledge of its effect on ments with embedded probes’’. Mater Struct 40:745–758
structural behaviour and, using the findings, to develop 8. Castellote M, Andrade C (2006) Round-Robin test on
calculation tools to accurately assess the residual methods for determining chloride transport in concrete.
RILEM recommendation of TC-178-’’ testing and modeling
strength and hence the safety of the structures chloride penetration in concrete’’. Mater Struct 39:990–995
concerned. 9. Page CL, Treadaway KWJ (1982) Aspects of the electro-
chemistry of steel in concrete. Narute 297(5862):109–115
Materials and Structures (2019) 52:2 Page 25 of 26 2

10. Gouda VK (1966) Anodic polarization measurements of 29. Thierry D, Persson D, Leygraf C, Boucherit N, Hugot-le
corrosion and corrosion inhibition of steel in concrete. Br Goff A (1991) Raman spectroscopy and XPS investigations
Corros J 2:13 of anodic corrosion films formed on Fe-Mo alloys in alka-
11. Kaesche H (1959) Testing corrosion danger of steel rein- line solutions. Corros Sci 32(3):273–284
forcement due to admixtures in concrete. Zement-Kalk- 30. Andrade C, Fullea J, Alonso C (1999) The use of the graph
Gips 7:289 corrosion rate-resistivity in the measurement of the corro-
12. Pedeferri P (1989) Corrosione e protezione delle strutture sion current. In: Rilem Proceedings no. 18: Measurement
metalliche e in cemento armato negli ambienti naturali-, and Interpretation of the on-Site Corrosion Rate.
Clup-Milano. Piazza Leonardo da Vinci, 32, Milano pp 157–165
13. Cigna R (1966) Studio sula corrosione dei ferri affogarti in 31. Andrade C, Pazini E (2012) Opciones teóricas de actuación
malte cementicie effettuato mediante curve i polarizza- en la marquesina del estadio de Maracana- PATORREB-
cione. L’Industria Italiana del cementpo, p 740 2012- 48 Congreso de Patologia y Rehabilitacion de edifi-
14. Whiting D (1981) Rapid determination of the chloride cios-Santiago de Compostela
permeability of concrete-, Report no. FHWA-RD-81-119, 32. Alonso C, Andrade C, Rodriguez J, Dı́ez JM (1998) Factors
NTIS DB no. 82140724, Federal Highway Administration, controlling cracking of concrete affected by reinforcement
Washington, DC, p 174 corrosion. Mater Struct 31:435–441
15. Pourbaix M (1973) Lectures in electrochemical corrosion. 33. Andrade C, Alonso C, Molina FJ (1993) Cover cracking as a
Plenum Press, New York function of rebar corrosion: Part I - Experimental test. Mater
16. Evans UR (1948) Metallic corrosion, passivity and protec- Struct 26:453–464
tion. Longmans Green and Co, New York 34. Torres-Acosta AA, Sagüés AA (2004) Concrete Cracking
17. Galvele J (1983) Pitting Corrosion-Treatise on Materials by Localized steel corrosion. Geometric effects. ACI Mater
Science and Technology 23:1–57 J 101:501–507
18. Pickering HW, Frankenthal RP (1972) On the mechanism of 35. Andrade C, Keddam M, Novoa XR, Perez MC, Rangel CM,
localized corrosion of iron and stainless steel: I. Elec- Takenouti H (2001) Electrochemical behavious of steel
trochem Stud J Electrochem Soc 119(10):1297–1304 rebars in concrete: influence of environmental factors and
19. Sanchez J, Fullea J, Anmdrade C, Gaitero JJ, Porro A (2008) cement chemistry. Electrochim Acta 46:3905–3912
A study of the early corrosion of a high strength steel in a 36. Andrade C, Garcı́a S, Toro L, Alonso C, Castellote M
diluted sodium chloride solution. Corros Sci 50:1820–1824 (2006) Reinforcement corrosion in chloride environment of
20. Mansfeld F (1997) Potential distribution in the Evans drop different concentrations. Concrete Durability and Service
experiment. Corros Sci 39(2):409–413 Life Planning (ConcreteLife’06)- Ein Bokek, Israel
21. Andrade C, Rodriguez-Maribona I (1992) Macrocell versus 37. Andrade C, Gónzalez JA (1978) Quantitative measurements
microcell corrosion of reinforcement placed in parallel. of corrosion rate of reinforcing steels embedded in concrete
Corrosion- NACE, paper no 194 using polarization resistance measurements-. Werkst Kor-
22. Andrade C, Rodriguez-Maribona I, Feliu S, Gonzalez JA, Feliu ros 29:515
S Jr (1992) The effect of macrocells between active and passive 38. Stern M, Geary AJ (1957) Electrochemical polarization: I.
areas of steel reinforcements. Corros Sci 33(2):237–249 A. theoretical analysis of the shape of polarization curves.
23. Andrade C, Fullea J, Toro L, Martinez I, Rebolledo N Electrochem Soc 104(1):56–63
(2012) Reinforcement corrosion in chloride media in 39. González JA, Algaba S, Andrade C (1980) Corrosion of
absence of oxygen. RILEN event TC -226 CNM- Long term reinforcing bars in carbonated concrete. Br Corros J
oerformance of cementitous barriers and reinforced con- 3:135–139
crete in Nuclear Power Plant and Radioactive Waste Storage 40. Andrade C, Rebolledo N (2009) Corrosion of reinforced
and Disposal- NUCPERF - Cadarache- France concrete made with different binders and exposed for
24. Alonso C, Castellote M, Andrade C (2002) Chloride 20 years in natural sea water. In: 2nd international RILRM
threshold dependence of pitting potential of reinforcements. workshop on concrete durability and service life planning-
Electr Acta 47:3469–3481 ConcreteLife’09- Haifa, Israel
25. Izquierdo D, Alonso C, Andrade C, Castellote M (2004) 41. Garces P, Andrade C, Saez A, Alonso MC (2005) Corrosion
Potentiostatic determination of chloride threshold values for of reinforcing steel in neutral and acid solutions simulating
rebar depassivation. Experimental and statistical study. the electrolytic environments in the micropores of concrete
Electrochim Acta 49:2731–2739 in the propagation period. Corros Sci 47:289–306
26. Joiret S, Keddam M, Nóvoa XR, Pérez MC, Rangel C 42. Andrade C, Martinez I, Alonso C, Fullea J (2001) New
(2002) Use of EIS, ring-disk electrode, EQCM and Raman advanced electrochemical techniques for on site measure-
spectroscopy to study the film of oxides formed on iron in ments of reinforcement corrosion. Mater Constr
1 M NaOH. Cem Concr Compos 24(1):7–15 51(263–264):97–107
27. Stratmann M, Bohnenkamp K (1983) Engell, H-J- Elec- 43. Andrade C, Sanchez J, Martinez I, Rebolledo N (2011)
trochemical study of phase-transitions in rust layers. Corros Analogue circuit of the inductive polarization resistance.
Sci 23(9):969–985 Electrochim Acta 56:1874–1880
28. Albani A, Gassa LM, Zerbino JO, Vilche JR, Arvia AJ 44. Feliú S, González JA, Feliú S Jr, Andrade C (1990) Con-
(1990) Comparative study of the passivity and the break- finement of the electrical signal or in situ measurement of
down of passivity of polycrystalline iron in different alka- polarization resistance in reinforced concrete. ACI Mater J
line solutions. Elecfrochimica Acta 35(9):1437–1444 87:457
2 Page 26 of 26 Materials and Structures (2019) 52:2

45. Feliú S, González JA, Andrade C (1996) Multiple-electrode parameters. In: International RILEM conference on mate-
method for estimating the polarization resistance in large rial science- MATSCI, Aachen-Germany. Springer, Berlin.
structures. J Appl Electrochem 26:305–309 vol II, pp 43–51
46. Andrade C (1993) ‘‘Calculation of chloride diffusion coef- 51. Andrade C, Rebolledo N (2017) Generic modelling of
ficients in concrete from ionic migration measurements’’. propagation of reinforced concrete damage. In: Proceed-
Cement Concr Res 23:724–742 ingsfib Symposium -Maastricht-The Netherland
47. Andrade C, Sanchez J, Fullea J, Rebolledo N, Tavares F 52. CEN/TS 12390-11: 2013: Testing hardened concrete.
(2012) On-site corrosion rate measurements:3D simulation Determination of the chloride resistance of concrete. Uni-
and representative values. Mater Corros 63(12):1154–1164 directional diffusion
48. Andrade C (2017) Reliability analysis of corrosion onset: 53. Andrade C, Rebolledo N (2014) Accelerated evaluation of
initiation limit state. J Struct Integr Maint 2(4):200–208 chloride corrosion by means of the integral test. Structural
49. CONTECVET -A validated user’s manual for assessing the 186:09
residual life of concrete structures, DG Enterprise, CEC, 54. UNE 83992-2-Concrete durability. Test methods. Chloride
(2001). (The manual can be downloaded from the web site penetration tests on concrete. Part 2: Integral accelerated
of http://www.ietcc.csic.es/index.php/es/publicaciones-2/ method. Structural 186, p 09 (2014)
manual-contecvet 55. Andrade C, Bujak R (2013) Effects of some mineral addi-
50. Andrade C, Castillo A (2010) Water content of concrete in tions to Portland cement on reinforcement corrosion. Cem
natural atmospheres and its impact in the corrosion Concr Res 53:59–67