TECHNICAL PAPER

A critical review of the

Water Sorptivity Index (WSI) Journal of the South African
Institution of Civil Engineering

parameter for potential

ISSN 1021-2019
Vol 63 No 2, June 2021, Pages 27–34, Paper 1123

durability assessment: AMY MOORE completed her BSc (Eng) and

MSc (Eng) degrees in the Department of

Can WSI be considered in

Civil Engineering at the University of Cape
Town. Her Master’s research focused on the
corrosion of reinforced concrete in marine

isolation of porosity?
environments. Since having completed her
studies, Amy has been pursuing a career in
education and publishing.

Contact details:
CoMSIRU
A J Moore, A T Bakera, M G Alexander Department of Civil Engineering
University of Cape Town
Private Bag X3, Rondebosch 7701, Cape Town, South Africa
T: +27 21 650 4012
The Durability Index (DI) approach used in South Africa is linked to transport-related properties
E: [email protected]
of the cover layer of concrete. This performance-based approach utilises three different
durability-related parameters, namely Water Sorptivity Index (WSI), Oxygen Permeability Index ALICE TITUS BAKERA is a PhD student in the
(OPI) and Chloride Conductivity Index (CCI). In the water sorptivity test, both the sorptivity Department of Civil Engineering at the
and the water-penetrable porosity of the concrete are measured. Until now, porosity has not University of Cape Town (UCT). Her current
research focuses on the biogenic acid attack
been considered directly as a separate durability-related parameter. This paper analyses these
of concrete subjected to the sewer
implications and suggests that porosity should be adopted as a further index. The paper environment. She is a member of the
concludes that sorptivity cannot be interpreted without also reflecting the porosity value since Concrete Materials and Structural Integrity
these two parameters are inter-related. Research Unit (CoMSIRU) at UCT and a
member of the Concrete Society of Southern Africa, Western Cape Branch.

Contact details:
CoMSIRU
INTRODUCTION AND It seems self-evident that measurement
Department of Civil Engineering
BACKGROUND of suitable parameters is essential to the University of Cape Town
After water, concrete is the most used achievement of durability, but the difficulty Private Bag X3, Rondebosch 7701, Cape Town, South Africa
material on earth, with over 25 billion has been the lack of appropriate durability- T: +27 21 650 4012
E: [email protected]
tons being produced worldwide annually related tests and parameters that can be
(World Business Council for Sustainable applied, especially on as-built structures.
PROF MARK ALEXANDER is an Emeritus
Development 2009). However, the premature DIs help to fulfil this requirement. Two Professor and a Senior Research Scholar at
deterioration of concrete structures is of other critical aspects in the selection of the University of Cape Town (UCT). He has a
increasing concern around the world. In suitable durability index parameters are the PhD from the University of the
Witwatersrand and is a Fellow of the
developing countries such as South Africa, purpose for which the durability param- University of Cape Town and of RILEM
the excessive cost of repairing and rehabili- eters are measured, and the ease and speed (International Union of Laboratories and
tating concrete infrastructure is detrimental of measurement so that they can be used Experts in Construction Materials, Systems
and Structures). He co-authored Aggregates in Concrete (2005),
to the needs of other sectors of the economy. as rapid indicators of quality in quality
Alkali-Aggregate Reaction and Structural Damage to Concrete (2011) and
Resources that could be directed towards control schemes. It is this combination of Durability of concrete – design and construction (2017) (CRC Press), and
urgently needed new construction are purpose and ease and speed of measure- was Editor of Marine concrete structures. Design, durability and
diverted to the cost of repairing existing ment that dictates whether a parameter performance (2016) (Woodhead Publishers). He is involved in the Concrete
Materials and Structural Integrity Research Unit (CoMSIRU) at UCT, which
infrastructure. These resources could also can be used as a durability indicator.
focuses on infrastructure performance and renewal research.
be better spent on education and developing
Contact details:
additional infrastructure. Therefore, it is Importance of durability
CoMSIRU
essential that ways be found to ensure that design and assessment Department of Civil Engineering
concrete structures are constructed in the Durability design is integral to structural University of Cape Town
most durable way possible. To respond to design. However, durability design needs to Private Bag X3, Rondebosch 7701, Cape Town, South Africa
T: +27 21 650 4012
this challenge, researchers are developing be emphasised during the design process E: [email protected]
new approaches and tools to address durabil- to ensure that concrete structures are
ity problems. A powerful new set of tools designed for longevity. Durability is defined
involves “durability indicators”, or “durability as the ability of a material or structure to
indexes (DIs)” which are linked with the pre- withstand the service conditions for which
ferred performance-based approach to dura- it was designed, without undue deteriora-
bility design and specification. These DIs are tion (Alexander & Beushausen 2019). As Keywords: concrete durability, durability index, water sorptivity,
discussed later (Alexander et al 2008). indicated, many concrete structures are porosity, absorption

Moore AJ, Bakera AT, Alexander MG. A critical review of the Water Sorptivity Index (WSI) parameter for potential durability assessment:
Can WSI be considered in isolation of porosity? J. S. Afr. Inst. Civ. Eng. 2021:63(2), Art. #1123, 8 pages. http://dx.doi.org/10.17159/2309-8775/2021/v63n2a4 27
showing inadequate durability, implying materials and mix proportions are impor- under “ideal” conditions, as well as actual
that the objectives of “durability design” tant, ultimately, it is equally important to concrete under site conditions. Since site
have not been particularly successful. To quantify in situ performance (Alexander & conditions are never “ideal”, allowance
help remedy the premature deterioration Beushausen 2019). must be made in design for exposure con-
problem, “potential durability” of the con- ditions and construction effects.
crete needs to be assessed and specified. Quantification of durability
Potential durability is taken here to mean From the above, it follows that the best Brief description of Durability
the ability of the concrete delivered to, or durability strategies for reinforced Index tests
mixed on, the job site to be durable in the concrete structures will be those that The South African DI tests consist of
structure for which it is intended, under seek to improve the quality of the cover three tests, namely, Oxygen Permeability
the given exposure conditions. concrete. This quality must be appropri- Index (OPI), Chloride Conductivity Index
Many of the deterioration processes ately measured, during construction, as (CCI) and Water Sorptivity Index (WSI)
that render reinforced concrete structures well as the actual cover depth achieved, tests. These tests have been discussed
vulnerable are concerned with the quality thus quantifying potential durability. in literature and are briefly described
and depth of the concrete cover to the rein- Potential durability, as defined above, is here (Alexander et al 2018, 2008, 1999;
forcing steel. Potential durability relates to readily quantified by means of durabil- Mackechnie & Alexander 2002). All the
the quality of the cover layer (which is a ity indexes. This concept can be taken tests are conducted on “standard” 70 mm
complex product of the concrete constitu- further by linking it to service-life models diameter, 30 mm thick specimens. The
ents and proportions) and site construction (Mackechnie & Alexander 2002) which can specimens are pre-conditioned by oven-
processes, including protection and curing then be used to predict deterioration of a drying and, for the CCI test, also by pre-
immediately post-construction. These structure. Durability should be quantified saturation with chloride solution.
all have a direct influence on the rate by considering indexes that represent the The OPI test measures the pressure
of ingress of deleterious substances into transport properties of concrete, while also decay of oxygen passed through a standard
the concrete. Cover quality, with a direct considering the aggressiveness of the expo- specimen in a falling head permeameter,
impact on potential durability, can be sure environment. This latter point is very the index being defined as the negative
accomplished by ensuring that appropri- important, since it was observed by Moore log of the coefficient of permeability.
ate materials are selected for the given (2014) that reinforced concrete structures Common OPI values range from 8.5 to 10.5
environment at the design phase, and that in the marine environment, of the same (on a log scale), a higher value indicating
construction processes such as compaction quality concrete, had higher rates of cor- a higher impermeability and thus a con-
and curing are correctly carried out. rosion in the splash zone, when compared crete of potentially higher quality. Using
with the tidal zone. this approach, carbonation depths may
Prescriptive versus performance be predicted for different environments
approach to durability (Salvoldi et al 2015).
design and specification DURABILITY INDEX APPROACH The CCI test measures the instantane-
Traditionally, durability design and The durability index approach has been ous electrical current due to the flux of
specification were achieved by applying developed in South Africa (SA) over recent chloride ions through a standard specimen,
prescriptive approaches, in which construc- decades (Alexander et al 2018, 2008, 1999; pre-saturated in 5 M sodium chloride
tion materials and other constraints (e.g. Mackechnie & Alexander 2002). The SA (NaCl) solution, in a two-cell conduction
water/binder ratio, concrete cover, and durability indexes are measurable durabil- rig in which both sides of the specimen are
grade of concrete) were adjusted to suit the ity parameters and serve as “indexes” of exposed to the NaCl solution. The chloride
environmental exposure class. However, the potential durability of the material or conductivity index (CCI) is related to the
this approach is now much criticised as it structure. They characterise the key mate- chloride diffusion properties of the concrete.
does not consider (a) the ability to measure rial properties that govern the durability Finally, the WSI test measures the rate
actual potential durability “resistance”, and issues of concern. Although being meas- of movement of a wetting front through
(b) the effects of the construction processes ured in the short-term (typically 28 days), concrete under capillary suction. A
which affect durability, mainly curing and research has shown that they can be used schematic presentation of the test setup
compaction. Further, modern concrete- as indicators of the likely durability per- is given in Figure 1 – the lower the WSI
making materials, specifically binders and formance of the structure in the long-term value, the better the potential durability
admixtures, can no longer be regarded as (Alexander & Beushausen 2019). of the concrete. This index is particularly
“simplistic” and need a more sophisticated As discussed, the indexes reflect the sensitive to the near-surface properties of
approach to characterise and use them potential durability of the concrete. This concrete and is, thus, a good indication of
properly. Therefore, performance-based can be affected if concrete is cast and the type and effectiveness of curing. As the
approaches, which must require the quality cured under laboratory conditions, in final step in the test, the water-penetrable
of the as-built structure to be assessed to which case the indexes generally (but not porosity is measured under vacuum. The
be truly performance-based, are crucial in always) represent concrete quality that is WSI parameter is in effect the rate of mass
achieving durable structures. The cover superior to both as-delivered quality, and change from water uptake by absorption,
zone of the as-built structure needs to as-built quality, with the latter reflecting divided (or “normalised”) by the water-
be tested to reflect the penetrability and the influence of construction practices. penetrable porosity of the specimen. This
thus potential durability of the concrete. Thus, it is possible to assess whether the has implications for the resulting sorptivity
While the composition of the constituent concrete can be expected to be durable values, which will be explored later.

28 Volume 63  Number 2  June 2021  Journal of the South African Institution of Civil Engineering
 root of time t; t being the time (typically in
Concrete core specimen hours) after a specimen is first exposed to
70 mm diameter
30 mm thick water on its wetting face.
The denominator term (Msv – Mso) in
Wet paper towels Equation 1 is a proxy for porosity (Nilsson
Epoxy coating 2003) and thus Equation 1 normalises
the measured rate of mass gain “F” by a
Capillary rise porosity-related term, to give the sorptivity
S. Porosity n (in percentage) is defined in
Equation 3:

Msv – Mso
n= × 100(3)
Test surface Adρw

Figure 1 S chematic representation of the WSI test where Msv, Mso and d are defined above,
A is the cross-sectional area of the speci-
Table 1 Suggested range for durability classifications using index values (Alexander et al 1999) men, and ρw is the density of water. [Note
that the term (Msv – Mso) appears in
Durability class OPI (log scale) Sorptivity (mm/√h ) Conductivity (mS/cm)
the numerator in Equation 3, and in the
Excellent >10 <6 <0.75 denominator in Equation 1; it is in this
sense that the sorptivity can be thought of
Good 9.5–10 6–10 0.75–1.5
as the rate of mass gain “normalised” by
Poor 9.0–9.5 10–15 1.5–2.5 the porosity.]
In fact, water-penetrable porosity is
Very poor <9 >15 >2.5
the fundamental material parameter that
is measured in the WSI test, not sorptiv-
These tests are useful since they are SORPTIVITY VERSUS POROSITY ity. Porosity is a complex parameter that
rapid and can be conducted typically at In the WSI test, both the sorptivity and encompasses pore volume, pore intercon-
28 days after construction, allowing an the water-penetrable porosity of the nectivity, pore tortuosity, and pore path
early age evaluation of potential durability concrete are measured. The sorptivity length. In the test it is assumed that all
(Alexander et al 2008). Since the original value is effectively the rate of mass gain pore volumes that are penetrable by water
implementation of the tests, they have been (g/√h) “normalised” by the porosity of the are filled, giving a mass change that can
improved and updated (Mukadam et al concrete. The index is expressed as a rate be converted into a percentage volume
2016; Otieno & Alexander 2015). of penetration depth of a wetting front of the specimen. This, together with the
into the concrete, typically in mm/√h, measured rate of mass change due to water
Matrix of DI values and not purely as the rate of mass uptake. absorption, gives the sorptivity. This is
The durability indexes have been placed For deterioration mechanisms such as illustrated schematically in Figure 2, where
in a matrix of values which can be used to reinforcement corrosion, the rate at which the measured values are distinguished
evaluate the potential durability of concrete fluids approach embedded reinforcing bars from the inferred value of sorptivity.
structures at an early age (see Table 1). The is critical in evaluating the expected service Combining Equations 1 and 3 results in
table is qualitative and therefore has limita- life. Generally, concrete sorptivity is gov- Equation 4, which indicates that sorptivity
tions in its applications since the level of erned by a square root of time law, with the and porosity are related, showing that two
acceptable “performance” must be consid- relevant expressions given in Equations 1 specimens with the same measured rate of
ered in relation to the exposure conditions and 2: mass gain can have different sorptivity val-
and service environment of the structure. ues if their porosities differ, and vice versa.
Table 1 can be viewed as overly simplistic
because it does not directly consider the Fd F
S= (1) Sn = × 100(4)
latter two factors; for this reason more Msv – Mso Aρw
sophisticated specifications have been
developed by the South African National where S (mm/√h) is sorptivity; F (g/√h is For two different concretes (subscript 1 and
Roads Agency Limited (SANRAL) that take defined in Equation 2; Msv (g) is the satu- 2) with the same rate of mass change, the
these into account explicitly (but beyond rated mass of the specimen; Mso (g) is the F
term  × 100 is constant, indicating that
the scope of this paper) (SANRAL 2009). dry mass of the specimen; and d (mm) is Aρw
Durability studies must also address the the thickness of the specimen. the products of their sorptivity and poros-
relation of indexes to fundamental concrete ity values are equal (see Equation 5).
properties and microstructure. The work Mwt = F√t(2)
presented in this paper attempts such by S1n1 = S2n2(5)
considering the specific durability proper- where Mwt (grams) is the mass of the
ties of sorptivity and porosity, and how these sample, and F is the slope of the best fit It is the inter-relation between these two
are related in terms of potential durability. line from plotting Mwt against the square durability index parameters that is of

Journal of the South African Institution of Civil Engineering  Volume 63  Number 2  June 2021 29
 data was that, sometimes, details of the
Measured Inferred concrete mixes were missing. Information
such as the compressive strength, w/c ratio
and binder type would have been useful in
Porosity making other inferences.
(percentage water–penetrable As shown in Equations 1 and 2, in the
volume) (%)
absence of data on the rate of mass change,
it is possible to obtain this parameter
from the sorptivity, knowing the porosity
(in effect the term (Msv – Mso)). As some
information sources that were analysed
did not have the rate of change of mass
Sorptivity
Uni-directional mass rate of recorded, the values were determined indi-
(rate of advance of wetting
change due to absorption (g/√h)
front) (mm/√h) rectly as indicated.
It was important to establish the rate
of change of mass F of the different data
points so that a comparison could be made
Figure 2 M
 easured versus inferred parameters in the WSI test between sorptivity and porosity for the
same rate of mass change. The rate of mass
interest here, with the question of whether in combination should be. This issue needs change is also useful in establishing con-
both parameters should be considered in to be addressed, so that durability tests are crete durability. It stands to reason that a
durability assessments, and if so, what rela- fully understood, and to improve predic- concrete with a higher rate of mass change
tive importance should be given to each. tions of concrete performance. should have a lower durability potential,
This paper analyses the implications of because a higher rate means that more
Porosity as a separate the above and suggests that porosity should water (and therefore potentially harmful
durability indicator be adopted as an index; importantly, the substances) may enter the concrete. This
Other sorption-type tests, e.g. ASTM paper recommends that sorptivity cannot increases the probability of corrosion or
C1585-13 (ASTM 2013), are based on be interpreted in the absence of porosity other deterioration occurring.
absorption only (amount of water absorbed since these two parameters are related and
over the surface area), which might be con- their interpretation regarding deteriora-
sidered of limited value. For example, for tion mechanisms needs to consider both RESULTS AND ANALYSIS
concrete with high porosity, high absorp- parameters.
tion may relate to low penetration depth Inter-relationship of
and potentially acceptable penetrability sorptivity and porosity
properties (i.e. low depth of ingress during OBJECTIVE OF THIS PAPER The sorptivity and porosity values from 601
a wetting cycle). However, such a measure The objective of this paper is, therefore, to data points recorded in previous research
does not account for the possibility that explore differing values of porosity for the were plotted on a scatter plot (Figure 3).
a lot of fluid might be absorbed into the same sorptivity, based on extensive previous Once the outliers were removed from
surface of the concrete, and, if contami- experimental results, and what implications the data, the correlation coefficient (R2)
nated (e.g. by chlorides), may result in high this might have for the potential durability between sorptivity and porosity was
contaminant concentrations at the surface, of concrete. However, at this stage, potential evaluated as 0.19. As the data set is large
resulting in low durability (e.g. from dif- durability has not been quantified by way of and consists of specimens of different mix
fusion of chlorides). Thus, a measure of any service-life model, since porosity is also designs and constituent materials, this is
porosity, representing the volume of acces- important in service-life predictions, imply- to be expected. However, a general trend
sible pores in the cover zone of concrete, is ing that a matrix of suitable values would is observed: higher sorptivity values tend
needed as well. have to be determined and provided. to have higher porosity values. This might
Until now, porosity has not been con- be somewhat surprising, considering the
sidered directly as a separate durability earlier argument on the expected inverse
parameter amid the suite of DI param- METHODOLOGY relationship between these two parameters.
eters. However, sorptivity and porosity For examining the sorptivity–porosity rela- However, it is also possible to have high
are inversely related (Equation 4), and tionship, data was utilised from our availa- sorptivity and corresponding high porosity
in laboratory measurements it has been ble laboratory database, in which sorptivity values if the pores are highly intercon-
noticed that there can be high or low and porosity values for various specimens nected and small in size. This is largely a
values of sorptivity associated with high had been evaluated over an extended consequence of taking diverse data and
or low values of porosity, and vice versa. period. A total of 601 data points, obtained combining them into a single data set. As
This brings into question whether poros- from different specimens from 2009 to will be shown, in general, sorptivity and
ity should be considered in conjunction 2015, tested as per Alexander et al (2018), porosity tend to be inversely related.
with sorptivity as a durability parameter, were collated and analysed. Each data point Box and whisker diagrams of all the
and therefore also the issue of what the consisted of sorptivity and porosity for a sorptivity and porosity values in the data
acceptable values of these two parameters given specimen. A limiting factor of this set were produced, in Figures 4 and 5

30 Volume 63  Number 2  June 2021  Journal of the South African Institution of Civil Engineering
 data represented concrete grades of 50,
18 40, 30 and 25 MPa. Following this, data
R2 = 0.19 points with similar rates of change of mass
16
(parameter F in Equation 2) were grouped
14 together, and the corresponding sorptivity
and porosity values were plotted against
Porosity (%)

12 each other. The categories of the groups

of rates of mass change were taken as 6–7,
10
7–8, 8–9, and 9–10 g/√h.
8 Figure 6 shows scatter plots of the
sorptivity and porosity values plotted for
6 different mass rates of change. The same
4
pattern can be observed across all four
2 4 6 8 10 12 14 16 scatter plots, with the sorptivity and poros-
Sorptivity (mm/√h) ity values being closely inversely related
for the same category of rate of change of
Figure 3 D
 ata sets (601) represented in a scatter plot: porosity versus sorptivity mass. The plots show that, for any given
absorption rate, the relationship between
sorptivity and porosity is essentially the
same regardless of concrete grade. This is
an important observation since it indicates
that the same sorptivity value is possible
for different grades of concrete, governed
largely by the different rates of change of
mass. The data also illustrates that for a
given concrete or closely related set of con-
cretes, low sorptivity values can result from
high porosity values and vice versa.
Further observation of the four plots
in Figure 6 shows that, as the rate of mass
0 2 4 6 8 10 12 14 16 18 change increases, the scatter bands move up
Sorptivity (mm/√h) and to the right in relation to the axes, imply-
ing that both sorptivity and porosity increase
Figure 4 B
 ox and whisker diagram showing distribution of sorptivity values with increase in rate of mass change, i.e.
complementing Equation 4. This is expected
for sorptivity, since it relates directly to rate
of mass change; however, the fact that poros-
ity also increases implies that the rate of mass
change increases faster than the increase in
pore volume. This is most likely linked to
greater pore interconnectivity at greater rates
of mass change. Both these aspects – greater
sorptivity and greater porosity at greater rate
of mass change – are undesirable for produc-
ing a durable concrete. This also intimates
that both parameters need to be considered
when assessing the potential durability of a
0 2 4 6 8 10 12 14 16 18 concrete mix.
Porosity (%) The scatter band plots in Figure 6
represent concrete strength grades over
Figure 5 B
 ox and whisker diagram showing distribution of porosity values a wide range of 25–50 MPa. While there
is some clustering of points for a given
respectively, showing a symmetrical distri- Analysis of data strength grade, the picture emerges that
bution of the data. The average sorptivity the two durability-related parameters are
value of the data set is 8.5 mm/√h, while Data categorised according to rate of mass not meaningfully linked to strength. This
the range of sorptivity is 3–14 mm/√h. The change, incorporating strength grade is also an important conclusion since many
porosity values are also found to be sym- Data points that had information pertain- engineers still relate durability to strength.
metrical, with the average porosity being ing to the stated grade of the concrete These results show that it is possible to
11.9%, and the range of porosity values (not measured grade), of which there were obtain a given sorptivity or porosity over a
being 5–17%. 151, were extracted and categorised. The wide range of concrete strengths.

Journal of the South African Institution of Civil Engineering  Volume 63  Number 2  June 2021 31
 6–7 (g/√h) 7–8 (g/√h)
10.5 15

10.0 14

9.5 13
Sorptivity (mm/√h)

Sorptivity (mm/√h)
9.0 12

8.5 11

8.0 10

7.5 9

7.0 8

6.5 7

6.0 6
9 10 11 12 13 14 15 16 8 9 10 11 12 13 14 15 16 17
Porosity (%) Porosity (%)
50 MPa 40 MPa 30 MPa 25 MPa 50 MPa 40 MPa 30 MPa 25 MPa

8–9 (g/√h) 9–10 (g/√h)

13 15

14
12
13
11
Sorptivity (mm/√h)

Sorptivity (mm/√h)

12
10 11

9 10

9
8
8
7
7

6 6
8 9 10 11 12 13 14 15 16 17 8 9 10 11 12 13 14 15 16 17
Porosity (%) Porosity (%)
50 MPa 40 MPa 30 MPa 25 MPa 40 MPa 30 MPa 25 MPa

Figure 6 S orptivity and porosity relationships for different categories of rate of change of mass, over a range of compressive strength grades

Relationship between sorptivity and rate Table 2 R

 ate of change of mass and associated porosity, for a sorptivity of 10 mm/√h
of change of mass for a constant porosity (data from Figure 6)
In general, an increase in the rate of water
Sorptivity of 10 mm/√h
absorption will result in a decrease in
potential durability. Regarding Figure 6, it is Rate of change of mass category (g/√h) 6–7 7–8 8–9 9–10
possible to obtain the same sorptivity over
Approximate porosity (%) 10.0 11.5 13.5 15.0
a range of values of rate of change of mass,
moderated again by the porosity. For exam-
ple, a sorptivity of 10 mm/√h is associated It is also important to consider the the pore available to expand. The amount
with a range of porosities, depending on the types of degradation that might occur of space and the pore spacing factor are
rate of change of mass, given in Table 2. because of different environmental exposure therefore critical in determining the amount
It can be argued that concrete with a conditions, as well as the function of the of damage. For other types of degradation,
sorptivity of 10 mm/√h will have a greater structure, as these play a role in whether such as corrosion of reinforcement, sorptiv-
potential durability at a porosity of 10% sorptivity or porosity is more important. For ity is more crucial. If the sorptivity is higher,
than at a porosity of 15%, particularly since, freeze-thaw degradation, the porosity and concrete exposed to a fluid will experience
as indicated above, the higher porosities critical pore size are probably more crucial deeper wetting in a shorter period, with the
may represent more interconnected pores. than the sorptivity, due to the mechanism fluid reaching the reinforcing bar and influ-
Thus, sorptivity cannot be considered in of freezing which involves the expansion encing corrosion. This might be critical in,
isolation of porosity since both can influ- of water in the pores. The larger the pore say, zones of wetting and drying such as the
ence potential durability. size, the larger the total amount of liquid in inter-tidal zone in marine structures.

32 Volume 63  Number 2  June 2021  Journal of the South African Institution of Civil Engineering
 Very poor (>15)
14 to 16 8.0 8.8 9.6 10.5 11.1 12.0 12.9 13.8 14.5 14.9
11 13
12 to 14 7.0 7.5 8.0 9.5 10.1 11.0 12.0 12.9 13.6 14.5
Poor (10–15)
Sorptivity (mm/√h)

9
10 to 12 6.0 6.8 7.4 8.0 8.6 9.2 9.8 10.4 11.4 12.2
7

8 to 10 5.2 5.8 6.1 6.6 7.0 7.6 8.4 8.7 9.5 10.0
5 Good (6–10)
6 to 8 3.9 4.4 4.9 5.5 6.0 6.4 6.8 7.4 7.9 8.3

3
4 to 6 2.7 3.1 3.5 3.8 4.3 4.5 5.0 5.5 6.0 6.5 Excellent (<6)

7 to 8 8 to 9 9 to 10 10 to 11 11 to 12 12 to 13 13 to 14 14 to 15 15 to 16 16 to 17
Porosity (%)

Values in white boxes on the contours represent rate of mass change in g/√h.

Figure 7 M
 atrix of porosity, sorptivity and associated rate of change of mass (in g/√h)

Semi-quantitative matrix of or within a set of contours, may have judgement was used to impose limits on
sorptivity and porosity values similar potential durability (given the same the values, such that both sorptivity and
The main objective of this work was to environmental conditions). For example, porosity values of greater than 15% were
determine whether sorptivity should be in this scheme a concrete with a sorptivity considered undesirable in a structural con-
considered in conjunction with porosity of 6 mm/√h and a porosity of about 12% crete. The desired durability classifications
when assessing the potential durability may be considered to have similar potential should also take into consideration the envi-
of concrete. The question that arises is: durability as a concrete with a sorptivity of ronmental exposure class of the structure.
what porosity values in combination with 8 mm/√h and a porosity of about 8%. A “poor” concrete might be quite acceptable
sorptivity values might be considered Figure 7 also gives a qualitative and in an environment where the structure is
“acceptable” or not? To assist with mak- semi-quantitative classification of durabil- not exposed to aggressive substances, such
ing this assessment, a semi-quantitative ity in the sorptivity categories on the right- as internal concrete in a normal building.
matrix of values of sorptivity and porosity hand side of the diagram, which derive Thus, engineers and specifiers should exer-
in combination was produced. The ranges from Table 1. Conventionally, these catego- cise caution in using Table 3, so as not to
of sorptivity and porosity were selected ries are extended leftwards by following over-specify a given concrete.
as 4–16 mm/√h and 7–15% respectively. the relevant horizontal category lines from
A gradation of 2 units was used for the their starting positions on the right-hand
sorptivity axis, and a gradation of 1 unit for side axis. This is the more conservative CONCLUSIONS
the porosity axis. The matrix is shown in approach. However, an alternative scheme
Figure 7. that bears further investigation is to extend Effect of strength grade on
Values of the average rate of mass the sorptivity categories leftwards by the relationship between
increase for possible combinations of sorp- following the relevant contour lines. For sorptivity and porosity
tivity and porosity were inserted on Figure 7, example, any concrete falling under the For any given absorption rate (i.e. rate of
associated with “contours” (see white boxes). mass rate of gain contour of, say, 7 g/√h mass change), the relationship between
A colour scale was used to show high (red) might be considered “excellent”, together sorptivity and porosity is essentially the
and low (green) rates of change of mass. It with other concretes within the contour same regardless of the nominal grade of
is reasonable to assume that the higher the limits with different porosities and sorptiv- the concrete (50, 40, 30 and 25 MPa). This
water absorption rate, the more vulnerable ity. It must be stressed that these potential
the concrete will be in terms of durability. durability classifications are tentative and Table 3 S uggested durability classification
From Figure 7, for a given sorptivity can only be taken in a general sense. They table for sorptivity and porosity values
value, there exists a wide range of rate of are presented here to stimulate debate on Sorptivity Porosity Durability
change of mass values. This shows again how these parameters might better be used (mm/√h) (%) classification
how concretes with the same sorptivity in durability design.
<6 <10 Excellent
value can have varying potential durability To make the matrix more “user-friendly”,
depending on the porosity. Concretes with the results were simplified into Table 3 that <10 Excellent to Good
6–10
higher porosity might be more penetrable gives the potential durability for a combina- >10, <12 Good to Poor
to greater depth over the long term than tion of different sorptivity and porosity
concretes with lower porosity. The con- values. These critical values for porosity <12 Good to Poor
10–15
tours attempt to show that concretes with were determined by using the contour lines >12, <15 Poor to Very poor
any combination of sorptivity and porosity of Figure 7 as a guide, but also considering
>15 – Very poor
values that fall on the same contour line, the data in Figure 3. Importantly, however,

Journal of the South African Institution of Civil Engineering  Volume 63  Number 2  June 2021 33
demonstrates that the same sorptivity value sorptivity and porosity values was devel- Alexander, M G & Beushausen, H 2019. Durability,
is possible for different grades of concrete, oped (Table 3). Judgement was used to service life prediction, and modelling for reinforced
and highlights the importance of consider- impose limits on very high porosity and/ concrete structures: Review and critique. Cement
ing a parameter such as porosity to fully or sorptivity values. Sorptivity and porosity and Concrete Research, 122: 17–29.
describe the durability potential. values greater than 15% were considered Alexander, M G, Mackechnie, J R & Ballim, Y 1999.
Also, both sorptivity and porosity undesirable. While this table is useful in a Guide to the use of durability indexes for achieving
increase with increase in rate of mass general sense, the environmental exposure durability in concrete structures. Resesearch
change. While this direct relationship conditions of the structure need to be Monograph No. 2. University of Cape Town &
between sorptivity and rate of change of considered when assessing the potential University of the Witwatersrand.
mass is to be expected, it is not intuitively durability. For example, a “poor” sorptiv- ASTM 2013. ASTM C1585-13 2013. Standard Test
expected for porosity, but implies that the ity classification may be acceptable for a Method for Measurement of Rate of Absorption
rate of change of mass increases faster concrete that forms part of the interior of Water by Hydraulic Cement Concretes. West
than the pore volume increases. This of a normal building, where deterioration Conshohocken, PA: ASTM International.
leads to the conclusion that higher rate is extremely unlikely. Further research is Mackechnie, J R & Alexander, M G 2002. Durability
of mass change is linked to greater pore needed to refine the table of “acceptable” predictions using early-age durability index testing.
interconnectivity (and not necessarily pore porosity and sorptivity values and explore Proceedings, 9th Durability and Building Materials
volume). An increase in both or either of the implications for concrete potential Conference, 17–21 April 2002, Brisbane, Australia,
these parameters, sorptivity and porosity, durability. Also, further work is needed p 11.
will result in a decrease in the potential to assess the effect of measured porosity Moore, A 2014. Effect of oxygen availability on the
durability of the concrete. For this reason, values and pore size distributions on the corrosion rate of reinforced concrete in marine
both sorptivity and porosity values should durability parameters considered. exposure zones: Inference from site and lab studies.
be considered when assessing the potential MSc dissertation. University of Cape Town.
durability of concrete. Mukadam, Z, Alexander, M G & Beushausen, H D
The results also show that a given ACKNOWLEDGEMENTS 2016. The effect of drying preconditioning on the
concrete strength could have varying com- The authors gratefully acknowledge the South African durability index tests. Cement and
binations of porosity and sorptivity. This financial support of the Concrete Materials Concrete Composites, 69: 1–8.
confirms the notion that strength is not and Structural Integrity Research Unit Nilsson, L O 2003. Durability concept: pore structure
indicative of durability, in contrast to what (CoMSIRU) at the University of Cape Town, and transport processes. In Newman, J & Choo, B S
is commonly thought by engineers. and the industrial sponsors of CoMSIRU, (Eds). Advanced Concrete Technology. Cambridge,
i.e. PPC Cement (Pty) Ltd, AfriSam, Lafarge, UK: Woodhead Publishing.
Porosity considered in SIKA, Haw & Inglis, and AVENG. Other Otieno, M & Alexander, M 2015. Chloride conductivity
conjunction with sorptivity sponsors of CoMSIRU are the Water testing of concrete. Past and recent developments.
As mentioned above, concretes with Research Commission of South Africa Journal of the South African Institution of Civil
the same sorptivity may have varying (WRC), the South African Department of Engineering, 57: 55–64.
porosity values (an increase in porosity Water and Sanitation, and Transnet. Salvoldi, B G, Beushausen, H & Alexander, M G 2015.
is associated with an increase in the rate Oxygen permeability of concrete and its relation to
of mass change). It can thus be argued carbonation. Construction and Building Materials,
that concrete with a given sorptivity rate REFERENCES 85: 30–37.
and “lower” porosity, will have a greater Alexander, M, Ballim, Y & Mackechnie, J 2018. Durability SANRAL (South African National Roads Agency
potential durability than a concrete with Index Testing Procedure Manual. Version. 4.5.1. Limited) 2009. Spreadsheet: Table 6000/1 : Concrete
a “higher” porosity, since higher porosities CoMSIRU. University of Cape Town & University of Durability Specification Targets (Civil Engineering
represent greater connectivity of pores. the Witwatersrand. http://www.comsiru.uct.ac.za. Structures only). Pretoria: SANRAL.
Thus, sorptivity cannot be considered in Alexander, M G, Ballim, A Y, Stanish, K & Ballim, Y World Business Council for Sustainable Development
isolation to porosity, as both parameters 2008. A framework for the use of durability indexes 2009. The Cement Sustainability Initiative (CSI).
influence the potential durability. in performance-based design and specifications Recycling of Concrete. http://www.wbcsdcement.org.
To attempt to quantify this, a table of for reinforced concrete structures. Materials and
suggested “acceptable” combinations of Structures, 41(5): 921–936.

34 Volume 63  Number 2  June 2021  Journal of the South African Institution of Civil Engineering