Materials and Structures/Matériaux et Constructions, Vol.

33, December 2000, pp 603-611

RILEM TC 154-EMC: ELECTROCHEMICAL TECHNIQUES FOR MEASURING

RILEM TECHNICAL COMMITTEES

METALLIC CORROSION

Test methods for on site measurement of resistivity of concrete

Prepared by R. Polder, with contributions from C. Andrade, B. Elsener, Ø. Vennesland,
J. Gulikers, R. Weidert and M. Raupach
Recommendations
The text presented hereafter is a draft for general consideration. Comments should be sent to the TC Chairlady: Dr Carmen Andrade, Instituto de
Ciencias de la Construccion “Eduardo Torroja”, Serrano Galvache s/n - Aptdo 19.002, 28033 Madrid, Spain; e-mail: [email protected], by 30
June 2001.

TC MEMBERSHIP: Chairlady: C. Andrade, Spain; Secretary: B. Elsener, Switzerland; Members: C. Alonso, Spain; R. Cigna,
Italy; J. Galland, France; J. Gulikers, The Netherlands; U. Nürnberger, Germany; R. Polder, The Netherlands; V. Pollet, Belgium;
M. Salta, Portugal; Ø. Vennesland, Norway; R. Weidert, Germany; Corresponding members: C. Page, UK; C. Stevenson, South
Africa.

PREFACE tivity for loss of protection or the severity of corrosion.

Most non-destructive tests are electrochemical methods:
Reinforcement corrosion is the major threat to the potential mapping, polarisation resistance testing, concrete
durability of reinforced concrete structures. Although electrical resistance measurement. Embedding probes
many structures perform very well, various types of allows to monitor various factors involved in corrosion
structures in particular in the infrastructure, need signif- and environmental inf luences. To address these items,
icant maintenance due to the impact of aggressive envi- RILEM has set up TC 154-EMC on Electrochemical
ronments. Environmental actions are irreversible and Techniques for measuring Corrosion of Steel in Concrete.
aggressive substances build up over the years. Due to Through its existence, TC 154 has set out to prepare
adverse combinations of poor design, inadequate execu- RILEM Technical Recommendations on:
tion and aggressive environments, the world-wide – Half cell potential measurements
spending on maintenance and the impact on the perfor- – Test methods for on-site corrosion rate measurement
mance of structures are increasing. of concrete reinforcement by means of the polarisation
RILEM committees have studied reinforcement corro- resistance method
sion in concrete since 1960, which resulted in a successful – Test methods for on site measurement of resistivity of
State-of-the-Art report by TC 60-CSC (1988) and in a concrete
Draft Recommendation on Strategies for the Repair of – Embedded probes for corrosion.
Concrete Structures Damaged by Steel Corrosion by TC The present document intends to describe methods to
124 (1994). In this document, the importance of condition assess concrete resistivity on site for various purposes
assessment was pointed out as the first step in the repair related to corrosion and protection of steel reinforcement.
process and as a basis for further decisions. Decisions on
repair strategies should be based on knowledge of:
– the cause of damage or loss of protection 1. SCOPE
– the degree and amount of damage
– the expected progress of damage with time This RILEM Technical Recommendation intends to
– the effect of damage on structural behaviour and ser- give a general description of methods to assess concrete
viceability. resistivity on site for various purposes related to corro-
The most widely used assessment method is visual sion and protection of steel reinforcement. In addition, it
inspection, which at best allows to describe and quantify provides background, technical details and a guideline
the damage. Traditionally, intrusive testing is used to for the execution and interpretation of measurements of
reveal the causes of corrosion with a normally limited pos- concrete resistivity on site.
sibility to take samples. Non-destructive test methods When new structures are designed and built, resistiv-
have become valuable additions to the repertoire. They ity probes may be embedded. In addition to monitoring
should work rapidly and allow to cover the complete sur- chloride content, steel potential or corrosion rate, the
face of a concrete member. They should detect the main resistivity is measured periodically during the lifetime;
causes of corrosion, the condition of the steel, the sensi- together they may indicate the risk of corrosion of the

1359-5997/00 © RILEM 603

Materials and Structures/Matériaux et Constructions, Vol. 33, December 2000

embedded steel. However, this is beyond the present – to assess the (range of) value(s) of the concrete resistiv-
scope. For structures without embedded probes, the ity of a particular structure, in order to estimate the risk
resistivity can be measured non-destructively using elec- of corrosion in case passivation will be (or has been) lost
trodes placed on the concrete surface. Together with – to locate the most permeable parts of a structure, in order
other information the corrosion risk can be determined. to define further investigations or protective measures
This Recommendation describes into detail methods to – to locate spots with the most severe exposure to water
determine concrete resistivity on site. and dissolved aggressive species
The resistivity of cement-based materials (concrete) – to help design systems for cathodic protection and
is a function of porosity (inherent, time dependent other electrochemical treatments
materials property), the chemical composition (conduc- – quality control of concrete in the production phase.
tivity) of the solution in the pores and the number and
distribution of pores filled with solution as a result of the
interaction with the environment. Based on these three 3. DEFINITIONS
aspects, the electrical resistivity of a given concrete struc-
ture or element may give information related to corro- Concrete resistivity is the ratio between applied volt-
sion of embedded steel in the initiation as well as in the age and resulting current in a unit cell that is a specific
propagation period. geometr y independent mater ial property, which
describes the electrical resistance, The dimension of
resistivity is resistance multiplied by length, its unit is
2. SIGNIFICANCE AND USE usually Ω m (ohm meter).
Conductivity is the inverse of resistivity.
The electrical resistivity of concrete is a materials
property that may be useful for monitoring and inspec-
tion of concrete structures with regard to reinforcement 4. THEORETICAL BACKGROUND
corrosion in combination with other non-destructive
techniques, e.g. potential mapping and corrosion rate. 4.1 General
The resistivity of a given structure exposed to chlo-
ride load gives information about the risk of early corro- The electrical resistivity of concrete may vary over a
sion damage, because generally a low concrete resistivity wide range, from 101 to 106 Ω m, mainly influenced by
is correlated to rapid chloride penetration and to a high the moisture content (environment) and the composi-
corrosion rate. In addition resistivity mapping may show tion (material) of the concrete, as was shown by Gjørv &
the most porous spots, where chloride penetration is Vennesland [20] and Tuutti [42]. In concrete, electrical
likely to be fastest and future corrosion rates will be current is carried by ions dissolved in the pore liquid.
highest; preventive measures may be taken accordingly. More pore water (wet concrete) as well as more and
After damage occurs, resistivity is relevant for possible larger pores with a higher degree of connectivity and a
maintenance actions as well: electrochemical repair lower tortuosity (high w/c) cause a lower resistivity. For a
methods are inf luenced by concrete resistivity and its constant relative humidity and in stationary conditions,
variation over the structure. resistivity is increased by a lower water to cement ratio
Resistivity does not show whether steel in concrete is (w/c), longer curing times (hydration) or by the addition
in an active state of corrosion or not. That information of reactive minerals such as blast furnace slag, f ly ash
must be obtained in another way: from chloride analysis, and/or silica fume. The resistivity of concrete increases
carbonation depth measurement, potential mapping, when the concrete is drying out and when the concrete
polarisation resistance measurement and visual inspec- carbonates (in particular in Portland cement concrete).
tion of the steel. If the steel is actively corroding, resistiv- Carbonation reduces the amount of ions available for
ity measurements may give additional information: it carrying the current and may densify the concrete. For
may show where in the structure corrosion may be non-carbonated concrete, the effect of the penetration of
strongest. In general terms and within one structure, rel- chloride ions on the resistivity is relatively small.
ative corrosion rates can be predicted. The choice Because the current is transported only by the ions of
between local or more general repair measures can be the pore liquid in the cement paste, concrete is not a
related to the variation of the corrosion rate, as deduced homogeneous conductor. Aggregate particles are essen-
from resistivity measurements. tially isolating bodies. The coarse aggregate may have a
Resistivity measurements (from the concrete surface) similar size as the concrete cover to the steel or the spac-
can be performed on all parts of concrete structures that ing of measuring electrodes. Consequently, on the scale
are exposed to air. It cannot be measured on buried or of centimetres, the current f low is non-homogeneous.
submerged parts. Resistivity measurements can be used With measuring electrodes far apart, the current f low
at any time during the service life of a concrete structure will be more homogeneous. Local disturbances of
and under any circumstances, provided the temperature homogeneous current flow may be also due to a differ-
is higher than 0°C. ent resistivity of the surface concrete layer and to the
Resistivity measurements may be useful for the fol- presence of steel bars.
lowing purposes:

604
TC 154-EMC

4.2 Temperature dependence given structure (after passivation has been lost), it is
likely that areas with low resistivity will have a higher
Temperature changes have important effects on con- corrosion rate than areas with high resistivity.
crete resistivity. A higher temperature causes the resistiv-
ity to decrease and vice versa (for a constant relative
humidity). This is caused by changes in the ion mobility 4.4 Concrete resistivity and chloride penetration
in the pore solution and by changes in the ion-solid
interaction with the cement paste. As a first approach an From theoretical and experimental work there appears
Arrhenius equation can be written to describe the effect to be a relationship between resistivity of and chloride dif-
of temperature on conductivity: fusion in a particular concrete composition [2, 33, 35]. For
example, concrete with a high percentage of blast furnace
σ(T ) = σ (T ) ∗ exp (b [T -1 – T -1])
i 0 0 i (1) slag has a high resistivity and a low chloride diffusion coef-
with ficient. Complete quantitative evaluation may require
σ the conductivity [1/Ω m] knowing the pore water conductivity [5] and information
T0 the reference temperature [K] on chloride binding (and other ion-solid interactions).
Ti the actual temperature [K] In practical terms:
b an empirical factor [K]. – within a particular existing structure, more permeable
For steady state conditions, b was found to be in the areas will have a comparatively lower resistivity and
range of 1500 to 4500, [7, 8, 11, 16], and: stronger chloride penetration
– to increase with decreasing relative humidity for a – for a new structure to be exposed to chlorides, a mini-
given cement paste, mortar or concrete mum resistivity can be specified based on the empirical
– to decrease with w/c ratio of the mix for a given rela- relationship between resistivity and chloride penetration
tive humidity. and when the concrete is produced, the resistivity may
The temperature dependence of the conductivity of be measured of each day’s production as part of the qual-
bulk pore solution differs significantly from that of cement ity control system.
paste or mortar with the same ion concentration in the
pores [7]. This is due to strong ion-solid interactions. The
humidity dependence of the temperature exponent in 4.5 Concrete resistivity and carbonation
cement paste or concrete can be explained by the fact that
at lower RH the pore solution becomes more concen- In OPC concrete, carbonation induces a significant
trated and is present in more narrow pores, so the ratio of increase of resistivity. When exposed to sufficiently wet
pore wall surface area to liquid increases and consequently conditions, the corrosion rate of steel in carbonated con-
the degree of interaction between ions and solid increases. crete may be comparatively high, following the general cor-
These interactions may be different for cements with dif- relation between inverse resistivity and corrosion rate [21].
ferent chemical compositions (slag, fly ash cement).
From the foregoing, it will be clear that accurate tem-
perature correction of resistivity data is very complex. The 4.6 Concrete resistivity and electrochemical
concrete composition and the moisture content both influ- maintenance methods
ence the resistivity itself and its temperature dependence.
For simplicity, it may be assumed that in the range of 0°C Concrete resistivity is an important factor in the effec-
to 40°C, doubling of resistivity takes place for a 20°C tiveness and durability of cathodic protection (CP) sys-
decrease, or that a change of 3% to 5% per degree occurs. tems applied to chloride contaminated structures [23, 30,
34]. Similarly resistivity is important for electrochemical
chloride removal and realkalisation [31]. For a large varia-
4.3 Concrete resistivity and corrosion rate tion in resistivity within a structure it is more difficult to
obtain uniform protection (CP) or a well-predicted result
From the electrochemical nature of the corrosion (chloride removal, realkalisation). In practice, parts of
process, a relationship may be expected between the structures with different resistivities (either due to differ-
resistivity of concrete and the corrosion rate of rein- ent concrete composition or different exposure condi-
forcement after depassivation. In the corrosion cell cir- tions) should be placed in separate electrical zones in order
cuit, the (hydroxide) ion transport between anode and to achieve sufficient control of current flow. Theoretical
cathode is one of the rate controlling factors [21, 37]. An work in this area was reported by Hunkeler and
alternative view is that a low resistivity increases the Holtzhauer [24]. Repair mortars used in connection with
anodic (dissolution) process [21]. Using a simplified electrochemical techniques should have a similar resistivity
approach based on the work of Bazant [6], the corrosion to that of the old concrete [9, 36]. Differences in resistiv-
rate of depassivated steel in concrete should be inversely ity in the material or different resistance between anode
proportional to the resistivity of the concrete. This was and reinforcement can be accepted to a certain degree; the
confirmed in a general sense [1, 3, 21]. Further work has resulting difference in current density may have a positive
shown indeed that this relationship may be dependent effect: where conditions are more corrosive, a stronger
on concrete composition [7, 19]. In any case, within a current will be flowing.

605
Materials and Structures/Matériaux et Constructions, Vol. 33, December 2000

5. METHODS FOR ON SITE MEASUREMENT

OF CONCRETE RESISTIVITY
5.1 Basic aspects of on site measurement of
resistivity
All methods for on site measurement of concrete resis- Fig. 1 – Setup of
tivity involve at least two electrodes, of which one may be four-electrode
a reinforcing bar. A voltage is superimposed between the measurement of
concrete resis-
electrodes and the resulting current is measured (or vice tivity.
versa). The ratio gives a resistance (in Ω). The resistivity is
obtained by multiplying the measured resistance by a geo-
metrical conversion factor, called the cell constant (in m).
For a given test geometry the cell constant can be
obtained in various ways, either from theoretical consid- applicability of this formula has been shown by Millard
erations or from calibration using either standard concrete [26] and Elsener [12]. Elsener studied concrete blocks
samples or electrolytes of known resistivity. For compara- with 200 mm thickness using electrodes with a spacing
tive purposes such as monitoring changes in time, using varying from 20 to 100 mm. Electrode spacing from 20
the resistance is sufficient. to 80 mm gave the same resistivity with a maximum
The assumption of converting the measured resis- error of 20%. A good correlation was found between
tance to a resistivity value is valid only for a homogeneous calibrated data measured with cast in electrodes and 4-
material. Instead, concrete usually contains steel rebars point resistivity obtained from the surface over a wide
and may show different resistivity distributions as a func- range of values. From laboratory tests on various speci-
tion of the depth to the surface. The surface layer may be mens [43, 45] it was shown that the “true” resistivity
carbonated or strongly dried out, resulting in a higher was obtained within an error of 25%.
resistivity than the concrete bulk. On the other hand, Because rebars conduct current much better than con-
shortly after rain, a surface layer may have a lower resis- crete, they will disturb homogeneous current f low. In
tivity than the bulk. When using a four-probe method, extreme cases (four probes on top of one rebar), an artifi-
these effects may be minimised by placing the measuring cially low resistivity is measured. Measuring over bars at 10
electrodes at a larger distance. However, this is less or 20 mm depth, errors can be made of a factor 2 to 6 [45].
favourable with regard to the inf luence of rebars. The However, even if only one of the four electrodes is near a
effect of a carbonated surface layer is not very large, pro- bar, current f low will be far from ideal and erroneous
vided its depth is less than the electrode spacing, in prac- results may be produced. The measured result may be arti-
tice if it does not extend beyond the reinforcement [26]. ficially low or high, depending on which electrode is near a
Possibly more strict boundary conditions apply [22]. rebar. To minimise this effect, none of the measuring elec-
trodes should be placed above or near rebars. With the usual
rebar spacing this causes the need to place the measuring
5.2 Four-point method electrodes quite close. This may be conflicting with the
objective to have the electrodes far apart in order to avoid
On site, the resistivity can be measured using a probe effects of aggregate particles. Measuring errors resulting
according to Wenner, consisting of four equally spaced from the contact resistance between the electrodes and the
point electrodes that are pressed onto the concrete sur- concrete is discussed by Ewins [17] and several other
face (4-point method), illustrated in Fig. 1. The two sources of error by Millard [26]. A practical compromise
outer point electrodes induce the measuring current appears to be an electrode spacing of 30 to 50 mm.
(usually AC with a frequency between 50 and 1000 Hz,
normally sinusoidal) and the two inner electrodes mea-
sure the resulting potential drop in the electric field. The 5.3 Methods involving the rebar network as
resistance is the ratio of the voltage and the current. This one electrode
method has long been known and used for determining
soil resistivities [44], and was studied for the application A possible method is to place one metal electrode on
to concrete structures by Stratfull [40] and Naish and co- the concrete surface and to measure the resistance
workers [28]. between this electrode and the reinforcement. This
The resistance R calculated from the four-point mea- requires a connection to the reinforcement cage and full
surement can be converted to resistivity r using a cell steel continuity. The method is illustrated in Fig. 2 and
constant based on theoretical considerations by: is actually a 2-electrode type of measurement.
ρ=2∗π∗a∗R (2) The conversion of the measured resistance between
disc electrode and steel bar, R(disc-bar), into resistivity ρ
with a the electrode spacing. This formula applies in according to:
principle only for homogeneous semi-infinite volumes
of concrete and infinitely small electrode points. The ρ = k ∗ R(disc-bar) (3)

606
TC 154-EMC

6.1.1 Measuring instruments

Instruments for measuring concrete resistivity on site
may be:
– earth resistance meters (for 4-electrode methods)
– regular resistance meters (for 2-electrode methods)
– parts of other instruments, like ohmic drop compensa-
tion circuits in potentiostats or frequency analysers.
The resistance meters shall apply sinusoidal current
of frequencies between 50 Hz and 1 kHz, or other signal
forms if their suitability is demonstrated. The electrolyti-
cal resistance of concrete does not seem to depend sig-
Fig. 2– Setup of one electrode (disc) measurement of concrete nificantly on frequency in this range.
resistivity. One manufacturer uses a square waveform with 13
Hz and fast sampling at a particular point in the cycle,
based on advanced circuit analysis [17].
is more complex than for the four-point measurements, DC instruments (like simple multimeters) shall not
because the cell constant k depends on disc size, con- be used, because the direct current will induce electrode
crete cover, rebar spacing and rebar diameter. For disc polarisation which may cause serious errors.
sizes smaller than the distance to a large electrode (the
rebar system), Feliu et al. [18] have shown that the resis- 6.1.2 Probes (electrodes)
tivity is given by: According to the measuring method, a probe consists
ρ = 2 ∗ a ∗ R(disc-bar) (4) of one or more electrodes. For one-electrode measure-
ments, the instrument shall have a good electrical con-
with a the diameter of the disc (in m). tact to the reinforcement (less than 1 Ω). For four-elec-
In practice, this may require that the disc is not trode probes, the electrodes shall be firmly fixed to keep
placed immediately over the rebars, but at some distance, their spacing constant.
for instance in the middle of the rebar mesh. This is par-
ticularly necessary if the cover depth is low or the disc 6.1.3 Contact between electrodes and concrete
diameter is more than 10 to 20 mm. Each electrode shall make contact with the concrete
For large disc electrodes the magnitude of the cell con- surface via a conducting electrolyte, usually in the form
stant is different. Using a disc of 200 mm diameter, calibra- of a wetted sponge or a wetted wooden plug. If the con-
tion with the four-point method was carried out [13] on a crete is dry, it will strongly absorb contact liquid and fre-
reinforced concrete wall with quite homogeneous cover quent wetting of the plugs is necessary. The presence of
and rebar network. For resistivities in the range from 200 a water film on the surface should be avoided. Extensive
to 1500 Ω m, the cell constant was about 1.5 m. pre-wetting the surface is not recommended because it
may change the bulk concrete resistivity.
5.4 Other arrangements 6.1.4 Cables and connections
All resistivity measurements are sensitive to poor
Concrete resistance can be measured also using two conduction of cables and connections. All cables and
electrodes placed on the concrete surface. Compared to connections shall be checked carefully before making
the one-electrode method, this will avoid the need to measurements. A resistance of less than 1 Ω is usually
make electrical contact to the reinforcement. The pres- taken as indicating good electrical connection.
ence of rebars will disturb this measurement as described
for the four-point method above. Because the amount of
current running (for a fixed voltage) depends on the size 6.2 Calibration and standardisation
of the contact between electrode and concrete [26], the
method is more sensitive to variations in the execution 6.2.1 Procedure
than other methods described above. For a given elec- In the laboratory, four-electrode probes, disc probes
trode diameter and precise operation, this method may and instruments must be calibrated by placing them with
give good resistance results. Conversion to resistivity is the electrode tips just in contact with liquids of known
not recommended. conductivity. The container shall measure at least
200 × 100 mm2 surface area and at least 100 mm deep for
electrode distance 30 mm and 200 mm deep for elec-
6. PRACTICAL EXECUTION trode distance 50 mm. Similarly, disc-electrode arrange-
ments can be calibrated using a container filled with
6.1 Apparatus electrolyte solution.
True calibration is very difficult however to be per-
The apparatus consists of a measuring instrument, formed in the field. A reasonable indication of good
one or more probes and cables to connect them. operation of the equipment can be obtained as follows.

607
Materials and Structures/Matériaux et Constructions, Vol. 33, December 2000

Disc method (one electrode): the electrode is placed cedure may be as follows. Cubes are cast as described
on a sheet of metal via the wetted sponge such as will be above. Resistivity is measured using steel plates as
used on the concrete. The resistance is measured described above. Slabs are cast with reinforcing bars at
between the electrode and the metal sheet. The value two or more cover depths. The measuring electrode is
shall be less than 10 Ω. If a higher resistance is mea- placed on the concrete surface and resistance is mea-
sured, cables, connections and the wetting of the sponge sured. The “true resistivity” is calculated from the mea-
shall be checked until < 10 Ω is measured. surement on the cubes with the two plates. The disc-
Four-electrode method: the probe is placed with all electrode device is calibrated by equating A/L to 2 ∗ a
four electrodes touching a metal sheet via their wooden (Equation (4)).
plugs. The resistivity is measured in the normal way, and If measurements have to be carried out on concrete
the reading shall be less than 1 Ω or 1 Ω m (depending with particular high resistivity, it may be recommended
on the instrument readout). to cast calibration specimens with high resistivity, for
example by using cement with a high percentage of blast
6.2.2 Standard reference measurement furnace slag (70%) fly ash (30%) or silica fume (10%).
Reference measurements shall be carried out on lab- It is emphasised that calibration measurements shall
oratory specimens of sufficient size. A possible proce- always be carried out on wet concrete (surface dry).
dure is as follows.
Concrete cubes are cast with a rib length not less
than 150 mm and preferably 200 mm, with aggregate 6.3 Precision and bias
not larger than 32 mm. The current distribution should
not be limited by the borders of the cube in order to Considerable scatter is present in most sets of resistiv-
have similar current distribution in the laboratory and ity measurements, even if they concern four-point data
when measuring on real structures. Cement type may be from laboratory specimens cast from the same mix and
ordinary portland cement or different (see below). The exposed to the same environment. In any set of measure-
cubes shall be stored in a fog room for at least four ments on the same concrete in the same conditions,
weeks. For calibration measurements, cubes are taken coefficients of variation of 10% are good and 20% must
from the fog room, their surface is dried with a cotton be considered normal. In the field, a coefficient of varia-
cloth and they are exposed to dry laboratory air for 10 tion of 30% is normal.
minutes to one hour, in order to have the surface loose
its film of water. For calibration, the resistivity is mea- 6.4 Measurement procedure
sured in two steps:
– with metal plates pressed to two opposing (mould) sur- 6.4.1 Concrete conditions and surface preparation
faces via wetted cloth (see Fig. 3) Concrete shall be clean and in particular free from oil
– with the four-point probe diagonally over the four ver- and other types of contamination.
tical mould surfaces. Before measurements are taken, the concrete surface
The “true resistivity” is calculated from the measure- may be wetted slightly to improve the contact, for
ment with the two plates using a geometrically calcu- instance by quickly moving a soaked sponge over the
lated cell constant (= area/length) by area to be measured. Strong wetting will inf luence the
resistivity, which shall be avoided if the resistivity under
ρconcrete = Rmeasured ∗ A / L (5) the prevailing conditions must be obtained.
with ρconcrete is the resistivity (in Ω m), Rmeasured the Sponges and wooden plugs shall be kept moist,
resistance between the plates (in Ω), A the area of the which may require frequent wetting as the concrete may
cube faces (in m2) and L the length of the cube (in m), absorb the wetting solution (from wooden plugs in par-
and the four-electrode measurement is calibrated by ticular), thus preventing sufficient contact.
equating A/L to 2 ∗ π ∗ a (Equation (1)). Because temperature has a significant effect on con-
For calibrating disc-electrode measurements the pro- crete resistivity, it is recommended to avoid measuring in
very hot and very cold weather conditions. In all cases,
the surface temperature of the concrete shall be mea-
sured and reported.

6.4.2 Procedure
For one-electrode measurements a simple procedure
(with relatively poor precision) is as follows:
– locate the rebars (using a magnetic cover meter)
Fig. 3 – Resistivity deter- – expose rebars on at least two spots, check electrical conti-
mination of a concrete nuity and make a firm connection to the reinforcement
core or cube. – wet the surface if considered necessary
– measure the resistance with the disc electrode at 5 to
10 closely spaced spots, record the values and take the
median as the reading for that position

608
TC 154-EMC

6.5 Complementary tests: additional samples

for laboratory testing
It may be useful to take core samples from the struc-
ture and expose them in the laboratory in standardised
saturation conditions to determine the resistivity.
Furthermore, core samples can be analysed to determine
the cement type and further compositional details.
Core samples taken from the structure are exposed in
a fog room or saturated under vacuum and the resistivity
is measured by using metal plates (see Fig. 3) to deter-
mine its value in saturated conditions. Obtaining satura-
tion may require a long time, in particular for very dense
Fig. 4 – Taking resistivity (four electrode) at various spots in the
same area to minimise influence of rebars. concretes. However, imperfect saturation does not make
a large difference, as long as the wider pores are filled.
For cores the resistivity is calculated by:
– calculate the resistivity using the estimated cell con-
ρconcrete = Rmeasured ∗ B / L (6)
stant using Equation (4).
For four-electrode measurements the procedure is as with ρconcrete is the resistivity (in Ω m), Rmeasured the
follows: resistance between the plates (in Ω), B the area of the
– locate the rebars using a cover meter and mark the core face (in m2) and L the length of the core (in m).
rebar mesh In some cases, it has been useful to expose cores in a
– wet the surface if considered necessary climate room with other than saturated conditions in
– measure with the four electrodes as far from the rebars order to obtain resistivity values representative of more
as possible, usual diagonally inside the rebar mesh (see sheltered conditions. This can be achieved by storing the
Fig. 4), and take five readings from the same position, cores in chambers with controlled humidity and temper-
moving the electrodes at least a few mm between each ature, for instance 20°C and 80% Relative Humidity.
reading Over some time (preferably months or alternatively until
– record all five readings and take the median as the constant weight is achieved) resistivity is measured as
reading for that position described in previous paragraphs.
– calculate the resistivity using the theoretical cell con- Complementary tests can be carried out on cores,
stant from Equation (2). supporting the interpretation of resistivity values. For
instance cement type and further compositional infor-
6.4.3 Selection of measurement locations mation can be determined using polarising and fluores-
Measurements shall be carried out on areas which are cence microscopy (PFM).
representative with regard to the following aspects:
– concrete composition
– exposure (wetting by rain and splash or sheltered, ori- 7. INTERPRETATION AND USE OF RESULTS
entation to prevailing winds)
– importance of structural element. 7.1 General
Within each area, sufficient measurements must be
taken to get a representative set of data. This may include To take full advantage of resistivity data measured on
the average resistivity but also the variation. For resistivity a structure, they should be compared to reference data of
mapping with the four-electrode technique, a grid spacing similar concrete types. In Table 1, some reference data
of 1 m is usually suitable [26]. are given, depending on cement type and exposure,

Table 1 – Global reference values at 20°C for the electrical resistivity of dense-aggregate concrete
of existing structures (age > 10 years); conditions between [ ] are the comparable laboratory climates
Concrete resistivity ρconcrete (Ω m)
Environment Ordinary Portland cement Blast furnace slag (> 65% slag, CEM III/B) or fly ash (> 25%)
concrete (CEM I) cement or silica fume (>5%) concrete
Very wet, submerged, splash zone, [fog room] 50 - 200 300 - 1000
Outside, exposed 100 - 400 500 - 2000
Outside, sheltered, coated, hydrophobised 200 - 500 1000 - 4000
[20°C/80%RH], not carbonated
ditto, carbonated 1000 and higher 2000 - 6000 and higher
indoor climate (carbonated), [20°C/50%RH] 3000 and higher 4000 - 10.000 and higher

609
Materials and Structures/Matériaux et Constructions, Vol. 33, December 2000

derived from various laboratory studies. The variations suggested interpretation of resistivity values with regard
given within each group (e.g. CEM I, very wet: 50 - 200 to corrosion risk for OPC concrete. As before, it must
Ω m), are caused by variation of water-to-cement ratio be emphasised that the values refer to 20°C.
and chemical cement composition [10, 29]. In view of
the discussion of temperature effects on resistivity, it is
important to realise that these reference values have been 8. DATA PRESENTATION REPORT
obtained during exposure at 20°C.
The report must contain:
– date of testing
7.2 Meaning of absolute values – description of the structure and individual measure-
ment locations
The absolute value of resistivity measured on a struc- – the weather conditions (temperature, humidity) at the
ture can be interpreted as follows. If the cement type is time of testing and preferably over a few days before testing
known, the observed resistivity is compared to the refer- – the measuring method used and calibration data
ence value for that cement type and for the appropriate – the measuring grid used
exposure conditions. If for example, a wet structure made – the results of the measurements, preferably with maps
with OPC has a mean measured resistivity of 50 Ω m, it or location sketches.
means that the water-to-cement ratio must be quite high.
The risk of corrosion is relatively high if factors for depassi-
vation are present (e.g. exposure to de-icing salts). REFERENCES
As discussed before, resistivity data must be corrected
for temperature effects. In general, a decrease in temper- [1] Alonso, M. C., Andrade, C. and Gonzalez, J. A., ‘Relation
between resistivity and corrosion rate of reinforcement in car-
ature of about 20°C doubles the measured resistivity, if bonated mortar made with several cement types’, Cement and
all other factors remain constant (see 4.2). If concrete Concrete Research 8 (1988) 687-698.
cores are taken and exposed in a laboratory, temperature [2] Andrade, C., Sanjuán, M. A., Recuero, A. and Río, O.,
correction can be made more accurately. ‘Calculation of chloride diffusivity in concrete from migration
experiments, in non steady-state conditions’, Cement and Concrete
Research 24 (7) (1994) 1214-1228.
[3] Andrade, C., Alonso, M. C., Gonzalez, J. A. and Feliu, S.,
7.3 Mapping for moisture content, ‘Similarity between atmospheric/underground corrosion and
homogeneity, etc reinforced concrete corrosion’, in ‘Corrosion of Reinforcement
in Concrete’, C. L. Page, K. W. J. Treadaway, P. F. Bamforth,
If the concrete composition is relatively homoge- eds., Elsevier (1990) 39-48.
[4] Andrade, C., Sanjuan, M. A. and Alonso, M. C., ‘Measurement
neous, mapping resistivity may show wet and dry areas. of chloride diffusion coefficient from migration tests’, paper 319,
If an OPC structure has resistivity values between 100 NACE Corrosion ‘93 (1993)
and 500 Ω m, the extreme values can be interpreted as [5] Arup, H., Sørensen, B., Frederiksen, J. and Thaulow, N., ‘The
indicating relatively wet and relatively dry areas. rapid chloride permeation test - an assessment’, paper 334,
If on the other hand, the exposure (so the moisture NACE Corrosion ‘93 (1993)
[6] Bažant, Z., ‘Physical model for steel corrosion in concrete sea
content) is homogeneously wet, variations in resistivity structures - Theory & application’, J. Structural Division, Am. Soc.
(say from 50 to 200 Ω m) can be interpreted as caused by Civ. Eng. 105 (ST6) (1979) 1137-1154 and 1155-1166.
local variations in water-to-cement ratio. Areas with [7] Bertolini, L. and Polder, R. B., ‘Concrete resistivity and rein-
50 Ω m will be more susceptible to penetration of chlo- forcement corrosion rate as a function of temperature and
ride from the environment than areas with 200 Ω m. humidity of the environment’, TNO report 97-BT-R0574,
March 1997, 85 pp.
[8] Bürchler, D., Elsener B. and Böhni, H., ‘Electrical resistivity and
dielectric properties of hardened cement paste and mortar.
7.4 Relationship to corrosion rate measurements Electrically based Microstructural Characterization’, ed. R.A.
Gerhardt, S.R. Taylor and E.J. Garboczi, Mat. Res. Soc. Symp.
The relationship between concrete resistivity and Proc. Vol. 411 (1996) 407.
[9] CEN, ‘Cathodic protection of steel in concrete – Part 1:
corrosion rate is still subjected to study. However, as a Atmospherically exposed concrete’, European Standard EN
general rule some reference can be given. Table 2 shows 12696-1, CEN/TC 219 (2000).
[10] COST 509, ‘Corrosion and protection of metals in contact with
concrete’, Final report, Eds. R.N. Cox, R. Cigna, O.
Vennesland, T. Valente, European Commission, Directorate
Table 2 - Risk of corrosion of reinforcement associated General Science, Research and Development, Brussels, EUR
with concrete resistivity [1, 10] for 20°C and OPC concrete 17608 EN, ISBN 92-828-0252-3, (1997) 148 pp.
[11] Elkey, W. and Sellevold, E. J., ‘Electrical resistivity of concrete’,
Concrete resistivity ρconcrete (Ω m) Risk of corrosion
Norwegian Public Roads Aministration Publication No. 80 (1995).
< 100 high [12] Elsener, B., ‘Electrochemical methods for evaluating reinforced
concrete structures, SIA Documentation D020, Non-destructive
100 - 500 moderate test of reinforced concrete structures, Swiss Association for
500 - 1000 low Engineers and Architects, Zürich (1988) p. 27-37 (only available
in German).
> 1000 negligible [13] Elsener, B., Flückiger, D., Wojtas, H. and Böhni, H., ‘Methods

610
TC 154-EMC

for the evaluation of corrosion of steel in concrete’, Swiss Federal Proc. of the Int. Conf. on Blended Cements in Construction,
Department of Traffic and Energy (EVED), ASB Research for Sheffield, 1991, ed. R.N. Swamy, Elsevier, 401-415.
the Maintenance of Bridges, VSS Report 521 (1996) 140 p (only [30] Polder, R. B. and Nuiten, P. C., ‘Design and installation of a
available in German). multi-element cathodic protection system’, Proceedings RILEM
[14] Elsener, B., Klinghoffer, O., Frolund, T., Rislund, E., Schiegg, International Conference on Rehabilitation of Concrete
Y. and Böhni, H., ‘Assessment of Reinforcement Corrosion by Structures, eds. D. W. S. Ho & F. Collins, Melbourne (1992)
Means of Galvanostatic Pulse Technique’, Proc. Int. Conf. 257-266.
Repair of Concrete Structures, Svolvær, Norway 28.-30. May [31] Polder, R. B., Hondel, A. J. van den, ‘Electrochemical realkali-
1997, ed. A. Blankvoll, Norwegian Public Roads sation and chloride removal of concrete; state-of-the-art, labora-
Administration, 391-400. tory and field experience’, Proc. RILEM International
[15] Elsener, B., ‘Corrosion Rate of Steel in Concrete - From Conference Rehabilitation of Concrete Structures, Melbourne
Laboratory to Reinforced Concrete Structures’, in ‘Corrosion of eds. D. Ho & F. Collins (1992) 135-148.
reinforcement in concrete - monitoring, prevention and rehabili- [32] Polder, R. B., Valente, M., Cigna, R. and Valente, T.,
tation’, Papers from Eurocorr’97, Mietz, J., Elsener, B., Polder, ‘Laboratory investigations of concrete resistivity and corrosion
R., Eds. The European Federation of Corrosion Publication rate of reinforcement in atmospheric conditions’, Proceedings
number 25, The Institute of Materials, London, ISBN 1-86125- RILEM International Conference Rehabilitation of Concrete
083-5, 1998, 92-103. Structures, Melbourne, eds. D. Ho & F. Collins (1992) 475-486.
[16] Elsener, B., ‘Corrosion of Steel in Concrete’, in “Corrosion and [33] Polder, R. B., Bamforth, P. B., Basheer, M., Chapman-
Environmental Degradation”, Vol. 2, Materials Science and Andrews, J., Cigna, R., Jafar, M. I., Mazzoni, A., Nolan, E. and
Technology, John Wiley (2000). Wojtas, H., ‘Reinforcement Corrosion and Concrete
[17] Ewins, A. J., ‘Resistivity measurements in concrete’, British Resistivity- state-of-the-art, laboratory and field results’, Proc.
Journal of NDT 32 (3) (1990) 120-126. Int. Conf. on Corrosion and Corrosion Protection of Steel in
[18] Feliu, S., Andrade, C., Gonzalez, J. A. and Alonso, C., ‘A new Concrete, ed. R. N. Swamy, Sheffield Academic Press, 1994,
method for in situ measurement of electrical resistivity of rein- 571-580.
forced concrete’, Mater. Struct. 29 (1996) 362-365. [34] Polder, R. B. and Nuiten, P. C., A multi-element approach for
[19] Fiore, S., Polder, R. B. and Cigna, R., ‘Evaluation of the con- cathodic protection of reinforced concrete, Materials Performance
crete corrosivity by means of resistivity measurements’, Proc. 33 (6) (1994) 11-14.
Fourth Int. Symp. on Corrosion of Reinforcement in Concrete [35] Polder, R. B., ‘Chloride diffusion and resistivity testing of five
Construction, eds. C. L. Page, P. B. Bamforth, J. W. Figg, concrete mixes for marine environment’, Proc. RILEM.
Society of Chemical Industry, Cambridge, UK, 1-4 July 1996, International Workshop on Chloride Penetration into Concrete,
273-282. St-Remy-les-Chevreuses, October 15-18, 1995, Eds. L.-O.
[20] Gjørv, O. E., Vennesland, Ø., El-Busaidy, A. H. S., ‘Electrical Nilsson, J.-P. Ollivier, RILEM, 1997.
resistivity of concrete in the oceans’, 9th Annual Offshore [36] Polder, R. B., ‘Cathodic Protection of Reinforced Concrete
Technology Conference, paper 2803, Houston (1977). Structures in The Netherlands - Experience and Developments’,
[21] Glass, G. K., Page, C. L. and Short, N. R., 1991, ‘Factors affect- in ‘Corrosion of reinforcement in concrete - monitoring, pre-
ing the corrosion rate of steel in carbonated mortars’, Corrosion vention and rehabilitation’, Papers from Eurocorr’97, Mietz, J.,
Science 32 (12) 1283-1294. Elsener, B., Polder, R., Eds. The European Federation of
[22] Gowers, K. R. and Millard, S. G., ‘Measurement of concrete Corrosion Publication number 25, The Institute of Materials,
resistivity for assessment of corrosion severity of steel using London, ISBN 1-86125-083-5, 1998, 172-184.
Wenner technique’, ACI Materials Journal (1999) 536-541. [37] Raupach, M., ‘Chloride-induced macroelement corrosion of
[23] Hunkeler, F., ‘The essentials for reinforced concrete monitor- steel in reinforced concrete’, German Committee for Reinforced
ing, particular emphasis on CP onset and future responses’, Concrete, 433, Beuth Editions, Berlin-Cologne, (1992) (only
Conference on Structural Improvement through Corrosion available in German).
Protection of Reinforced Concrete, Institute of Corrosion, 2-3 [38] Schiessl, P. ed., ‘Corrosion of steel in concrete’, report of
June 1992, London. RILEM Technical Committee 60-CSC, Chapman and Hall,
[24] Hunkeler, F. and Holtzhauer, K., 1993, ‘Water content, poros- London (1988).
ity and electrical resistance of mortars and concrete’, Swiss [39] Schiessl, P. and Raupach, M., ‘Influence of concrete composi-
Association for Engineers and Architects (1993) (only available in tion and micro-climate on the critical chloride content in con-
German). crete’, in “Corrosion of Reinforcement in Concrete”, eds. C. L.
[25] Millard, S., Ghassemi, M., Bungey, J. and Jafar, M., ‘Assessing the Page, K. W. J. Treadaway, P. F. Bamforth, Elsevier (1990) 49.
electrical resistivity of concrete structures for corrosion durability [40] Stratfull, R. F., Materials Protection 29 (1968).
study’, in: Corrosion of Reinforcement in Concrete, eds. C. Page, [41] Tritthart, J. and Geymayer, H., ‘Changes of electrical resistance
K. Treadaway and P. Bamforth, Elsevier (1990) 303-313. in drying concrete’, Zement und Beton (Cement and concrete) 1 (1)
[26] Millard, S. G., ‘Reinforced concrete resistivity measurement (1985) 23-28 (only available in German).
techniques’, Proc. Institution Civil Engineers, part 2, March [42] Tuutti, K., ‘Corrosion of steel in concrete’, CBI Stockholm, 1982.
1991, 71-88. [43] Valente, M., Polder, R. B., Cigna, R. and Valente, T.,
[27] Monfore, G. E., ‘The electrical resistivity of concrete’, Journal of ‘Experimental investigation of concrete resistivity and corrosion
the PCA Research and Development Laboratories (1968) 35-48. rate of reinforcement in atmospheric conditions’, TNO report
[28] Naish, C. C., Harker, A. and Carney, R. F. A, ‘Concrete BI-91-173 (1991).
inspection: Interpretation of potential and resistivity measure- [44] Wenner, F., ‘A method for measuring earth resistivity’, Bulletin
ments’, in “Corrosion of Reinforcement in Concrete”, eds. C. of the Bureau of Standards 12 (1915) 469-478
L. Page, K. W. J. Treadaway, P. F. Bamforth (1990) 314-332. [45] Weidert, R. and Gehlen, C., ‘Electrolytic resistivity of cover
[29] Polder, R. B., Ketelaars, M., ‘Electrical resistance of blast fur- conrete: Relevance, measurement and interpretation’, Eighth
nace slag cement and ordinary Portland cement concretes’, in Conf. on Durability of Materials and Components (1999).

611