See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250071710

Performance of concrete in XS1, XS2 and XS3 environments

Article  in  Magazine of Concrete Research · January 2008

DOI: 10.1680/macr.2008.60.4.261

CITATIONS READS

3 1,206

4 authors:

Gerry Starrs B.T. Linfoot

Heriot-Watt University Heriot-Watt University
56 PUBLICATIONS   899 CITATIONS    23 PUBLICATIONS   206 CITATIONS   

SEE PROFILE SEE PROFILE

T. M. Chrisp W. John McCarter

Heriot-Watt University Heriot-Watt University
62 PUBLICATIONS   1,005 CITATIONS    137 PUBLICATIONS   2,911 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

SENSOR-INFRA: Smart Engineered Cementitious Composites for Intelligent Infrastructure View project

All content following this page was uploaded by Gerry Starrs on 31 July 2018.

The user has requested enhancement of the downloaded file.

 Magazine of Concrete Research, 2008, 60, No. 4, May, 261–270
doi: 10.1680/macr.2008.60.4.261

Performance of concrete in XS1, XS2 and XS3

environments
W. J. McCarter, B. T. Linfoot, T. M. Chrisp and G. Starrs

Heriot-Watt University

An extensive chloride-profiling programme was undertaken over an eight-year period on nine concrete pier-stems
placed at a marine location to represent XS1, XS2 and XS3 environmental exposure conditions as defined in
European Standard EN 206-1. The pier-stems were 2.0 m high and octagonal in plan with each vertical face 0.66 m
wide and were constructed in groups of three (one each placed at the locations defined above): one group, which
was used as a benchmark, represented plain Portland cement concrete; the second group had a waterproofing agent
(Caltite) added at the time of mixing; and the third was treated with silane. Chloride concentration profiles were
taken at a number of locations on each pier-stem, which were subsequently used to evaluate the temporal change in
effective diffusion coefficient and surface chloride concentration for the different exposure conditions. In addition,
it allowed assessment of the relative performance of the plain concrete, Caltite concrete and silaned concrete pier-
stems with regard to chloride ingress.

BS 8500-1:20062 adopts the exposure environments

Introduction
defined in EN 206, although the XS2 exposure condi-
European Standard EN 206-11 uses a durability clas- tion is extended to include all concrete below mid-tide
sification system whereby a deterioration mechanism is level—that is, at a level were the concrete remains
considered with the ambient environment and, in all, 18 saturated and has no time to dry out.
exposure classes are presented. EN 206-1 (Annex J) The need to limit the amount of chloride entering
also permits the use of performance-based methods, reinforced-concrete structures has led to the use of a
which are based on experience with local practices in number of methods to alleviate the problem. This paper
local environments. Satisfactory guidelines for ensuring also examines the performance of two of these techni-
adequate concrete performance can only be developed ques
by monitoring real concrete structures under field ex-
(a) the use of a surface coating (silane)
posure conditions, ideally over an extended period of
(b) the addition of a waterproofing admixture (Caltite)
time. It is against this background that the work de-
to the concrete mix.
tailed below presents results from an eight-year field
monitoring programme on concrete exposed to a mar- Silane is a hydrophobic pore liner that permits the move-
ine environment. ment of water vapour but prevents the ingress of liquid
EN 206-1 defines three environments for corrosion water into the concrete. Caltite, on the other hand, is an
resulting from chlorides from sea water integral hydrophobic and pore-blocking ingredient
which is added to concrete at the time of mixing.3 It
provides a physical pore-blocking component when sub-
(a) XS1: exposed to airborne salt but not in direct
jected to hydrostatic pressure with the hydrophobic com-
contact with sea water
ponent, reducing water absorption by capillarity.
(b) XS2: permanently submerged
Although the effectiveness of silane in preventing or
(c) XS3: tidal, splash and spray zones.
slowing down chloride ingress is recognised, its per-
formance in structures in the marine environment is not
School of the Built Environment, Heriot-Watt University, Edinburgh,
clear, especially under long-term exposure; the long-
EH14 4AS, Scotland, UK
term performance of Caltite also warrants investigation.
(MACR-D-07-00079) Paper received 2 July 2007; last revised 3 Many studies on chloride ingress into concrete have
December 2007; accepted 2 January 2008 been undertaken using small-scale samples manufactured
261

www.concrete-research.com 1751-763X (Online) 0024-9831 (Print) # 2008 Thomas Telford Ltd

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
McCarter et al.

and tested under laboratory conditions or placed at out- the performance of concrete in a marine environment,
door exposure sites. In comparison, there are relatively in particular the tidal, splash and spray zone (see Fig.
few detailed studies on full-size concrete structures. To 1(b)). This programme entailed the construction of nine
this end, full-size bridge pier-stems were fabricated under ‘full-scale’ pier-stems that were placed on the east side
site conditions and tested for chloride ingress on a regular of the southern causeway leading to the road bridge.
basis extending over a period of eight years. Each pier-stem was 2.0 m high and fabricated using
cut-down sections of the steel formwork used in the
construction of the actual bridge piers (Fig. 2). The
Experimental details pier-stems were octagonal in cross-section with each
face being 660 mm; the reinforcement detailing was
Pier-stems identical to that used in the bridge piers and consisted
After the construction of the road bridge across the of a combination of 32 mm and 40 mm reinforcing bars
Dornoch Firth in north-east Scotland (see Fig. 1(a)), a in the form of a circular cage (Fig. 2). The cover to the
monitoring programme was undertaken to investigate reinforcement varied owing to the circular cage ar-
rangement and the octagonal shape of the pier-stems;
the minimum cover to steel at the centre of each face
Marine site was 65 mm.
location There was sufficient steel formwork to fabricate
moulds for three pier-stems. Since the formwork had to
be reused, this meant casting was staggered to allow
7 days’ curing with the formwork in place, stripping
and reassembling for the next cast. On removal of the
formwork, the pier-stems were covered with polythene.
Three pier-stems were constructed with concrete con-
taining Caltite waterproofing admixture; the remaining
(a) six used concrete to the exact mix specification as the
actual bridge piers (see Table 1), including the same

Zone under study

(b)
Fig. 1. (a) Location of marine exposure site; and (b)
Dornoch Firth road bridge and zone replicated at marine Fig. 2. Fabrication of pier-stems showing formwork and steel
exposure site reinforcement

Table 1. Mix details for pier-stems

Mix OPC*: 20 mm: 10 mm: Fine: Plasticiser: l/m3 Retarder: l/m3 Caltite: l/m3 w/c Slump: mm F28 : MPa
ref. kg/m3 kg/m3 kg/m3 kg/m3

Plain 460 700 350 700 1.80† 1.80‡ — 0.4 110 63

Caltite 460 653 327 770 5.52§ — 30 0.4} 105 48

* Ordinary Portland cement: CEM I:42.5N/ASTM Type I

† Sika FN
‡ Sika FR
§ Cementaid superplastet XL 1.3% w/w OPC
}
Includes 24 kg water from 83% by volume of Caltite
F28 : 28-day compressive strength (100 mm cubes)

262 Magazine of Concrete Research, 2008, 60, No. 4

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
 Performance of concrete in XS1, XS2 and XS3 environments
High-level
aggregate source (moranic gravel). Of these six, three 4 3 2
5 1 Seaward
received silane treatment and the remaining three acted Spray XS1 6 7 8
Mid-level
as references. The three pier-stems that were to have
Splash XS3
silane treatment were cast first in order to allow the 0·2 m Mean high-tide
minimum time to elapse between casting and applica- Low-level ⬇1·45 m XS3
tion (28 days); silane (Fosroc Nitocote SN511) was Riprap Mid-tide
⬇1·35 m
applied in accordance with current specifications.4,5 All XS2

pier-stems were transported to the exposure site and Mean low-tide

placed in position approximately 5 weeks after casting (a)

the first group of pier-stems.

Positioning of pier-stems
Figure 3(a) presents a schematic of the positioning of
the pier-stems at three different locations to provide
XS1, XS2 and XS3 exposure conditions. The pier-stem
positions have been designated high-level, mid-level
and low-level. One each of the plain concrete, Caltite
and silaned pier-stems was placed at all of the locations.
The top of each pier-stem was sealed by casting an in (b)
situ layer of a mortar containing Ronafix, a polymer-
based waterproofing admixture (www.ronacrete.co.uk).
The pier-stems were also bedded onto a similar mortar
on square, concrete levelling plinths (see Fig. 3(b)).
Regarding the latter, the three, low-level pier-stems had
an integral concrete plinth, so this was not required. By
taking these measures to prevent water and ionic ingress
through the tops and bases it was hoped that compari-
sons could then be made with the performance of the (c)
continuous, full-height bridge piers as shown in Fig.
1(b). Fig. 3(c) shows the pier-stems in position (note: in
this figure the low-level pier-stems are submerged);
Fig. 3(d) shows the alignment of the pier-stems to
7 8 9
simulate a full-height bridge pier. 4 5 6
With reference to Fig. 3(a), the low-level pier-stems
were completely submerged during the tidal cycle with
only the lower portion (
0.2 m) of the mid-level set 1 2 3
submerged at mean high tide. Mid-tide level was lo-
cated at
1.35 m above base of the low-level pier-
stems with the base of the high-level pier-stems (d)

2.0 m above mean high tide. The tidal range was

2.90 m, increasing to
3.90 m during spring tides. Fig. 3. (a) Schematic diagram showing placement of
Sea temperatures ranged, typically, from 5 to 148C be- pier-stems at marine site; (b) a mid-level pier-stem being
craned onto its levelling plinth with bedding mortar;
tween winter and summer months.6
(c) showing high- and mid-level pier-stems (in this
photograph the low-level pier-stems are submerged); and
Sampling and analysis (d) alignment of pier-stems (in this photograph, 1, 4 and 7
The main purpose of the experimental work was to are plain OPC; 2, 5, 8 are Caltite, and 3, 6 and 9 are
obtain a detailed picture of the performance of the con- silaned)
crete with respect to chloride ingress in a marine envir-
onment. With reference to Fig. 3(a), dust samples were files were taken over the full height of the mid-
obtained at the following locations level pier-stems: face 3 at 0.2, 0.4, 0.6, 0.8, 1.0,
1.2, 1.4, 1.6 and 1.8 m above base.
(a) low-level pier-stems: 1.0 m above base (XS2 envir-
onment according to BS 8500-1:20062 ) Samples were collected biannually (May and Novem-
(b) mid-level pier-stems: 0.2, l.0 and 1.8 m (XS3 en- ber) over an eight-year period from faces 1, 2 and 8
vironment) (see insert on Fig. 3(a)). Other faces were collected at
(c) high-level pier-stems: 0.5, 1.0 and 1.5 m (XS1 en- yearly intervals. At each of the locations, dust
vironment) was collected using a 30 mm drill bit with depth
(d) in addition, after 2.8 years’ exposure, chloride pro- increments 0–5, 5–15, 15–25, 25–35, 35–45, 45–55,
Magazine of Concrete Research, 2008, 60, No. 4 263

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
McCarter et al.

55–65 mm from the surface. The drill bit tool and drill 6·46% at 2·5 mm
6
hole were cleaned between depth increments to reduce Plain—1·3 years
the possibility of cross-contamination of samples from 5 Plain—7·3 years
Caltite—1·3 years
different depths. After drilling, the drill-hole was

Chloride: %wt cement

Caltite—7·3 years
4 Silane—1·3 years
grouted with mortar containing Ronafix. Typically, Silane—7·3 years
400–500 samples were collected every 6 months over
3
the test period. The dust samples were analysed for
total chloride content (i.e. acid soluble) in accordance 2
with BS 1881: Part 1247 and BS 6337: Part 4.8
1

0
Results and discussion 0 10 20 30 40 50 60
Depth: mm
The chloride ion concentration is based on the origi- (a)

nal cement content of the mix (460 kg/m3 ). The initial 6

chloride content of the concrete was measured by dril- Plain—1·3 years
Plain—7·3 years
ling cubes manufactured from the same batches of con- 5 Caltite—1·3 years
Caltite—7·3 years
crete as the pier-stems, which were found to be

Chloride: %wt cement

Silane—1·3 years
negligible (,0.01% by weight of cement). For concrete 4 Silane—7·3 years
containing steel reinforcement, chloride concentration
3
is limited to between 0.2% and 0.4%;1 as a conse-
quence, a value of 0.4% is used as a reference to study 2
the effectiveness of protection afforded by the concrete.
Unless otherwise stated, data are presented for the sea- 1
ward face (face 1) of each pier-stem.
0
0 10 20 30 40 50 60
Tidal zone: low tidal (XS2) and high tidal (XS3) Depth: mm
No part of the low-level pier-stems was permanently (b)

submerged; however, since concrete below the mid-tide

Fig. 4. Chloride concentration profiles at the beginning and
level does not have sufficient time to dry and remains end of test period for: (a) low-level pier-stems representing
in a fully saturated condition, it is classified as an XS2 XS2 exposure; and (b) mid-level pier-stems (at high water
exposure environment according to BS 8500-1:2006.2 mark) representing XS3 exposure
The sampling points on the low-level pier-stems were
located 1.0 m above the base of the pier-stem, which
categorises this region as XS2 exposure. The mid-level
pier-stem sampling point was, in this instance, 0.2 m tion is
1.80 m and would indicate that under this
above the base which coincides with high tide level and hydrostatic pressure, silane is ineffective under such
would represent XS3 exposure. conditions. Caltite is an integral hydrophobic pore-liner
Chloride profiles are presented in Fig. 4(a) for the and pore-blocking ingredient which is dispersed
three low-level pier-stems after 1.3 years and 7.3 throughout the entire volume of concrete. While the
years’ exposure respectively. The threshold value of hydrophobic action could, as with silane, be rendered
0.4% chloride is indicated on these figures. Over the ineffective under hydrostatic pressure, the pore-blocking
period presented, for XS2 exposure, chlorides have effect should stop the ingress of sea water under pres-
penetrated to a significant depth, with the threshold sure. Fig. 4(a) clearly indicates that the latter mechanism
value of 0.4% Cl being exceeded over the surface is not working as chloride levels are similar to the
50 mm for all pier-stems. The pier-stems are con- reference and silaned pier-stems.
structed of good-quality concrete—63 MPa for the un- Figure 4(b) presents the chloride profiles for the
treated and silaned pier-stems and 48 MPa in the case mid-level pier-stems with the sampling points just at
of the Caltite pier-stem—with a high cement content high tide level and subjected to XS3 exposure. The
and would, traditionally, be regarded as having a good main difference at this location is the lack of direct
resistance to the ingress of water and dissolved salts. contact with sea water under hydrostatic pressure.
It is apparent that both the Caltite and silane pier- Chloride concentrations are, for a given depth from the
stems have not prevented the ingress of chloride for surface of the pier-stem, approximately 50% less that
this particular exposure condition. the equivalent pier-stem in XS2 exposure presented in
Silane is a hydrophobic pore liner, repelling the in- Fig. 4(a). It is also noticeable that, at this exposure, the
gress of liquid water but allowing the movement of effectiveness of silane and Caltite in reducing chloride
water vapour. In the case of the profiles presented, the ingress is short term as, over the time-scale presented
maximum hydrostatic head at the sampling point loca- in Fig. 4(b), chloride concentrations at depths in excess
264 Magazine of Concrete Research, 2008, 60, No. 4

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
 Performance of concrete in XS1, XS2 and XS3 environments
1·0
of 10 mm attain similar values to the plain Portland
cement concrete. Fig. 4(b) also indicates that the Plain—1·3 years
0·8 Plain—7·3 years
threshold value of 0.4% chloride has advanced to a Caltite—1·3 years

Chloride: %wt cement

depth of 30–40 mm for all pier-stems after 7.3 years’ Caltite—7·3 years
Silane—1·3 years
0·6
exposure. Silane—7·3 years
After 3.7 years’ exposure, a series of six, 50 mm
0·4
diameter cores (approximately 55 mm long) were taken
from each of the high-level pier-stems (face 5). The
cores were returned to the laboratory and their porosity 0·2

determined by saturating and drying to constant weight

at 1058C; values are presented in Table 2 together with 0
0 10 20 30 40 50 60
the degree of saturation of the in situ concrete obtained Depth: mm
from the cores. Although the chloride concentration of
the North Sea was not measured at the Dornoch site, Fig. 5. Chloride concentration profiles for mid-level
sea water contains, typically, about 19.8 g/l of pier-stems at the beginning and end of test period for
chloride.9 If this concentration of sea water penetrates concrete in the splash zone (XS3 exposure)
the surface of the concrete, then the maximum attain-
able chloride concentration for the case where all avail-
able capillary pore space is filled at this concentration at reducing chloride concentrations within the surface
would be approximately 0.65% by weight of cement. It 10 mm.
can be seen from Fig. 4(a) that this value is easily
exceeded within the surface 5 mm, with values of 6% Airborne spray zone: XS1
attained. Thus, the ratio of the measured chloride con- Chloride profiles are presented in Fig. 6. In this
centration to that predicted from the chloride concen- position, any sea water contacting the surface will be as
tration in the sea water is approximately 10. Clearly, wind-blown spray. While the untreated concrete dis-
chloride binding and condensation10 effects must make plays a small concentration of chloride within the sur-
a significant contribution to the measured chloride le- face 10 mm, the Caltite concrete has almost eliminated
vels. the ingress of chloride. While silane reduces the chlor-
ide concentration in the short term, in the longer term
Splash zone: XS3 its effectiveness appears to be reduced.
Figure 5 presents chloride profiles at sampling points
located at 1.8 m above the base of the mid-level pier- Full-height profiles for mid-level pier-stems
stems (i.e. 1.6 m above high-water level). Although still From the discussion above, the location of the con-
classified as an XS3 environment, the main difference crete surface in relation to tide level has a major influ-
between this position and that discussed above, in rela- ence on the build-up of chlorides within the surface
tion to XS3 in Fig. 4(b), is the lack of contact with sea region. To investigate this aspect in more detail, parti-
water. cularly in the high-tidal and splash zones, after 2.8
Chloride concentrations for the untreated concrete years’ exposure, a series of chloride profiles was taken
presented in Fig. 5 (note change in vertical scale from over the full height of the mid-level pier-stems—face 3
Fig. 4) are, for a given depth from the surface of the in this instance. Figs 7(a)–(c) present the variation of
pier-stem, approximately 50–80% less that the equiva-
lent pier-stem in the XS3 position in Fig. 4(b), the 1·0
greatest reductions being most evident at increased
depths. Over the test period, the threshold value of 0·8
Plain—1·3yrs
0.4% chloride has advanced to a depth of 10–15 mm Plain—7·3yrs
Chloride: %wt cement

Caltite—1·3yrs
from the surface under this exposure. Unlike silane at Caltite—7·3yrs
0·6 Silane—1·3yrs
this location, Caltite would appear to remain effective Silane—7·3yrs

0·4

Table 2. Porosity and degree of saturation of concrete cores

0·2
from high-level pier-stems

Pier-stem Porosity: % Degree of 0

0 10 20 30 40 50 60
saturation: %
Depth: mm

Plain 15.5 80
Fig. 6. Chloride concentration profiles for high-level
Caltite 15.9 78
Silaned 14.3 74 pier-stems at the beginning and end of test period for
concrete exposed to airborne spray (XS1 exposure)
Magazine of Concrete Research, 2008, 60, No. 4 265

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
McCarter et al.
2·0
creasing linearly with distance above high water mark.
The silaned pier-stem, on the other hand, exhibits de-
2·5 mm
Height above base: m 1·5 10 mm creasing concentration up to approximately 1.0 m above
20 mm high water mark; it is also evident that, at this stage,
30 mm
40 mm silane is capable of maintaining lower chloride levels.
1·0 50 mm
Up to approximately 1.0 m above high water mark
represents a zone which is subjected to direct contact
0·5 with sea water during spring tides coupled with heavy
splashing from waves breaking on and around the pier-
Mean high tide
stems. The upper portion of the pier-stem (.1.0 m)
0 would be subjected to less intense wave action.
0 0·4 1 2 3
Chloride: %wt cement It is evident from Figs 7(a)–(c) that high chloride
(a) concentrations (1–2%) are attained at depths of up to
2·0 15–25 mm in a relatively short time (2.8 years).
2·5 mm
10 mm
20 mm Diffusion coefficient and surface chloride
1·5 30 mm concentration
Height above base: m

40 mm
50 mm
Although chloride ingress involves a complex inter-
1·0
action of mechanisms, it is commonly approximated to
a diffusion process. The most common approach is to
apply the error function solution to Fick’s diffusion
0·5 equation and derive an effective diffusion coefficient in
the form
  
x
0
0 1 2 3
C x, t ¼ C s 1  erf pffiffiffiffiffiffiffiffiffiffi (1)
2 Deff t
Chloride: %wt cement
(b) where Deff is the effective diffusion coefficient (m2 /s);
2·0 Cx,t is the chloride content at depth x (m) after expo-
2·5 mm
10 mm sure time t (s); Cs is the notional surface level of
20 mm chloride and erf is the error function. It is assumed in
1·5 30 mm
Height above base: m

40 mm equation (1) that the initial concentration of chloride in

50 mm
the concrete prior to diffusion is negligible. Both Cs
1·0 and Deff were calculated from the chloride against
depth profiles obtained from the pier-stems using
1
Mathcad’s Genfit general curve-fitting solution to
0·5 equation (1) with the first data point (i.e. 2.5 mm)
omitted from the analysis.
Figures 8(a)–(c) present the variation in chloride
0
0 1 2 3 concentration with depth for faces 1, 2 and 8 on pier-
Chloride: %wt cement stems 1, 2 and 3 respectively at the sampling height of
(c)
1.0 m. These are the seaward faces and the profiles
Fig. 7. Chloride concentration profiles for mid-level presented represent the situation after 2.8 years of XS2
pier-stems (face 3) taken over the full height of each exposure. These figures indicate the natural variations
pier-stem: (a) plain concrete; (b) Caltite concrete; and across the surface between different locations, particu-
(c) silaned concrete larly within the surface 5 mm (as would be anticipated);
however, at greater depths there is reasonable agree-
ment between the faces. Based on the profiles in Fig. 8,
chloride concentration at different depths within the and to highlight the range in results between adjacent
cover-zone with increasing height above the base of the faces, Table 3 presents Cs and Deff obtained from equa-
pier-stem. Mean high tide level and 0.4% chloride con- tion (1) above.
centration are indicated in these figures. Figures 9 and 10 present, respectively, the variation
The influence of distance above high water mark is of surface chloride concentration, Cs , and diffusion co-
most evident within the surface 0–25 mm, particularly efficient, Deff , over the duration of the testing pro-
for the ordinary Portland cement (OPC) and Caltite gramme. Figs 9(a)–(d) indicate the erratic nature of Cs
pier-stems. Considering the threshold value of 0.4% with the highest values occurring in the submerged
chloride, within the surface 0–5 mm, this value is ex- zone representing XS2 exposure—that is, below mid-
ceeded virtually over the full height of the OPC and tide level. From Fig. 9(a), Cs appears to be time invar-
Caltite pier-stems, with the chloride concentration de- iant with values fluctuating about a mean value of
266 Magazine of Concrete Research, 2008, 60, No. 4

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
 Performance of concrete in XS1, XS2 and XS3 environments
6
5.2% for the OPC pier-stem, 4.7% for the Caltite pier-
5 Face 1
stem and 5% for the silaned pier-stem. Cs values for
Face 2 concrete located at high water mark (Fig. 9(b)) repre-
Chloride: %wt cement

4 Face 8 senting XS3 exposure, are approximately 50% less than

the values in Fig. 9(a), and fluctuate about a mean
3 value of 3.4% for the plain concrete pier-stem, 2.0%
for the Caltite pier-stem and 2.9% for the silaned pier-
2
stem. Fig. 9(c) presents the variation of Cs for splash-
1
zone concrete (XS3); in this case, Cs values have not
yet stabilised and display a time dependency. Generally,
0 values increase with time, although these are lower than
0 10 20 30 40 50 60
Depth: mm
XS3 exposure for concrete in the tidal zone. Both the
(a) silane and Caltite would appear to be more effective in
6 reducing Cs at this location than the plain Portland
cement concrete. Cs values for concrete subjected to
5 Face 1 airborne spray (XS1) are displayed in Fig. 9(d); as is
Face 2
evident, values are generally very low and, in some
Chloride: %wt cement

Face 8
4 cases, no curve could be fitted to the data (hence values
at certain times are missing). Again, silane and Caltite
3
are effective in reducing Cs values in comparison to the
2
plain concrete.
Recommended values11 for the surface chloride con-
1 centration for use in design are presented in Table 4
(by weight of concrete). Converting these values to
0
0 10 20 30 40 50 60
weight of cement for the concretes used in the pier-
Depth: mm stems would give a typical value of Cs ¼ 2.20% and an
(b) upper value of 4.20%. Other studies,12–14 using consid-
6 erably smaller samples placed in the marine tidal zone,
have reported Cs values in the range 1.5–4.0% by
5 Face 1 weight of cement. Regarding the latter, both the
Face 2
Chloride: %wt cement

Face 8 strength and cement content of the concrete used work

4
were lower than that used in this study, although the
3
time-scale was similar (apart from Ref. 14 which was
conducted over 12 months). Measurements made on
2 marine concrete structures (8–40 years old)15 have also
reported Cs values between 1.8 and 3.9%. It is evident
1 that published results compare favourably with the re-
sults from this current study for concrete located at
0
0 10 20 30 40 50 60 high water mark (XS3).
Depth: mm The diffusion coefficients for pier-stems below mid-
(c) tide level (XS2) are presented in Fig. 10(a). In a
similar manner to Cs , Deff values at this location
Fig. 8. Variation in chloride concentration profiles across faces fluctuate about mean values of 2.8 3 1012 m2 /s for
1, 2 and 8 on pier-stems after 2.8 years’ exposure: (a) plain the plain concrete pier-stem, 3.6 3 1012 m2 /s for the
concrete; (b) Caltite concrete; and (c) silaned concrete Caltite pier-stem and 2.4 3 1012 m2 /s for the silaned

Table 3. Variation in surface chloride concentration and diffusion coefficient between adjacent faces on the pier-stem from
profiles in Fig. 8 (SD ¼ standard deviation). Data represent 2.8 years’ exposure in XS2 environment

Surface chloride, Cs : %wt cement Diffusion coefficient, Deff : 31012 m2 /s

Face OPC Caltite Silane OPC Caltite Silane

1 4.19 2.67 4.77 4.89 2.25 1.06

2 4.80 3.18 5.54 1.86 2.52 0.85
8 5.46 3.22 5.48 1.94 3.32 1.08
Mean 4.82 3.03 5.26 2.90 2.70 0.99
SD 0.63 0.31 0.43 1.73 0.56 0.13

Magazine of Concrete Research, 2008, 60, No. 4 267

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
McCarter et al.
8 8

7 7
Plain
6 6 Caltite
Silane
CS: %wt cement

CS: %wt cement

5 5

4 4

3 3

2 Plain
2
Caltite
1 Silane
1

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time: years Time: years
(a) (b)

8 8

7 7
Plain Plain
6 Caltite 6 Caltite
Silane Silane

CS: %wt cement

CS: %wt cement

5 5

4 4

3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time: years Time: years
(c) (d)

Fig. 9. Variation in surface chloride concentration, Cs , for: (a) concrete in the low-tidal zone (XS2); (b) concrete in the high-
tidal zone (XS3); (c) concrete in the splash zone (XS3); and (d) concrete subjected to airborne spray (XS1)

pier-stem. Fig. 10(b) displays the variation in Deff for tite and concrete treated with silane with regard to
that part of the pier-stem positioned at high-water chloride ingress when subjected to marine exposure
mark (XS3 exposure). Values, in this instance, fluctu- classes XS1, XS2 and XS3. To this end, a total of nine
ate about mean values of 1.5 3 1012 m2 /s for the full-scale pier-stems were fabricated, cast, cured and
plain concrete pier-stem, 2.3 3 1012 m2 /s for the Cal- placed under site conditions; the pier-stems underwent
tite pier-stem and 1.0 3 1012 m2 /s for the silaned a comprehensive chloride profiling programme and, as
pier-stem. Fig. 10(c) presents the diffusion coefficients a result, the following conclusions can be drawn for the
for concrete positioned in the splash zone at a height current study.
of 1.6 m above high tide level. At this location the
exposure environment is classified as XS3 and, (a) The XS2 (low-tidal) regime had the highest chlor-
although values fluctuate, there is a general decrease ide concentration at every depth with pier-stems in
over the initial five years’ exposure with values begin- this zone attaining critical chloride threshold con-
ning to stabilise by the end of the test period. Diffu- centration throughout the surface 50–60 mm after
sion coefficients for pier-stems in the airborne spray just 7.3 years’ exposure, despite the concrete mixes
zone (Fig. 10(d)) show a marked decrease over the having a high cement content.
seven-year period presented. In a number of instances, (b) Silane and Caltite did not significantly reduce
it was not possible to calculate a diffusion coefficient chloride ingress in cases where sea water comes in
due to a poorly defined chloride profile on which to direct contact with concrete exposed to XS2 (low-
curve-fit. tidal/submerged) and XS3 (high-tidal) environ-
ments. However, when compared to the plain OPC
pier-stems in the splash (XS3) and airborne zones
(XS1), both treatments performed better in redu-
Conclusions
cing chloride ingress.
The objective of this study was to investigate the (c) Diffusion coefficients displayed an erratic behav-
performance of plain concrete, concrete containing Cal- iour, although there was a general trend of decreas-
268 Magazine of Concrete Research, 2008, 60, No. 4

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
 Performance of concrete in XS1, XS2 and XS3 environments
10⫺10 10⫺10

Plain Plain
Caltite Caltite
Silane Silane
10⫺11 10⫺11
Deff: m2/s

Deff: m2/s
10⫺12 10⫺12

10⫺13 10⫺13
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time: years Time: years
(a) (b)

10⫺10 10⫺10

Plain Plain
Caltite Caltite
Silane Silane
10⫺11 10⫺11
Deff: m2/s

10⫺12 Deff: m2/s 10⫺12

10⫺13 10⫺13
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time: years Time: years
(c) (d)

Fig. 10. Variation in diffusion coefficient, Deff , for: (a) concrete below mid-tide level (XS2); (b) concrete at high tide level
(XS3); (c) concrete in the splash zone (XS3); and (d) concrete subjected to airborne spray (XS1)

Table 4. Surface chloride levels (% weight of concrete)11

(XS1), surface chloride levels displayed a gradual
increase over the test period. Cs values decreased
Type Typical Upper limit by almost an order of magnitude between the low-
tidal zone and the airborne zone: typical values for
Portland cement concrete 0.43% 0.80% all pier-stems being
(i) low-tidal XS2:
5%
(ii) high-tidal XS3:
2.5%
(iii) splash XS3:
1% (values had not stabilised
ing diffusion coefficient (Deff ) in moving from the within the test period)
low-tidal zone to airborne zone. Furthermore, for (iv) airborne XS1:
0.5%
concrete in direct contact with sea water—low-tidal
(XS2) and high-tidal (XS3)—diffusion coefficients
stabilised relatively quickly. In the splash (XS2)
and airborne (XS1) zones, diffusion coefficients Acknowledgements
displayed a gradual decrease over the test period. The work presented formed part of the concrete dur-
In summary, the diffusion coefficients for all the ability programme into monitoring and improving the
pier-stems fell within the following ranges performance of structural concrete in bridges underta-
(i) low-tidal XS2: 3–631012 m2 /s ken for the Scottish Executive Development Depart-
(ii) high-tidal XS3: 1.5–2.531012 m2 /s ment (SEDD). Funding from the SEDD is gratefully
(iii) splash XS3: 0.5–1.031012 m2 /s (values had acknowledged. The views expressed in this paper are
not stabilised within the test period) those of the authors and not necessarily those of the
(iv) airborne XS1: ,0.531012 m2 /s. SEDD.
(d) Surface chloride levels, Cs , displayed similar trends
to the diffusion coefficient in that values fluctuated
within a wide range. For XS2 and XS3 (high tidal) References
exposure, Cs values had stabilised by the end of 1. British Standards Institution. Concrete: Specification, Per-
the test period whereas for XS3 (splash zone), formance, Production and Conformity. BSI, London, 2000, BS
values were still increasing. In the airborne zone EN 206-1.

Magazine of Concrete Research, 2008, 60, No. 4 269

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
 McCarter et al.

2. British Standards Institution. Concrete: Complementary 10. Nagataki S., Otsuki N., Wee T.-H. and Nakashita K. Con-
British Standard to BS EN 206-1: Method of Specifying and densation of chloride ion in hardened cement matrix materials
Guidance for the Specifier. BSI, London, 2006, BS 8500-1. and on embedded steel bars. ACI Materials Journal, 1993, 90,
3. CementAid. How Everdure Caltite Works. See http://www. No. 4, 323–332.
cementaid.com/proven.htm (accessed 15 June 2007). 11. Bamforth P. B., Price W. F. and Emerson M. An International
4. Scottish Development Department. Technical Memorandum Review of Chloride Ingress into Structural Concrete. Transport
SB3/90: The Impregnation of Concrete Highway Structures. Research Laboratory, Crowthorne, Berkshire, 1997, Report No.
Scottish Development Department, Edinburgh, 1990, 10pp. 1303/96/9092, 162 pp.
5. Department of Transport Specification for Highway 12. Bamforth P. B. The derivation of input data for modelling
Works. BD27/86, Materials for the Repair of Concrete High- chloride ingress from eight-year UK coastal exposure trials.
way Structures, 6th edn. Vol. 3, Section 3, Department for Magazine of Concrete Research, 1999, 51, No. 2, 87–96.
Transport, London, 1986. 13. Thomas M. D. A. and Matthews J. D. Performance of pfa
6. Bundesamt für Seeschifffahrt und Hydrographie (Fede- concrete in a marine environment: 10-year results. Cement and
ral Maritime and Hydrographic Agency). See http://www. Concrete Composites, 2004, 26, No. 1, 5–20.
bsh.de/en/Marine%20data/Observations/Sea%20surface%20 14. Lindvall A. Chloride ingress data from field and laboratory
temperatures/anom.jsp#PercM (accessed 12 September 2007). exposure: influence of salinity and temperature. Cement and
7. British Standards Institution. Testing Concrete: Methods Concrete Composites, 2007, 2, No. 29, 88–93.
for Analysis of Hardened Concrete. BSI, London, 1988, BS 15. Polder R. B. and de Rooij M. R. Durability of marine concrete
1881-124. structures—field investigations and modelling. HERON, 2005.
8. British Standards Institution. General Methods of Chemi- 50, No. 3, 133–153. See http://heron.tudelft.nl/2005_3/Art1.pdf
cal Analysis: Part 4 (ISO 6227-1982). Method for Determina- (accessed 2 December 2007).
tion of Chloride Ions by Potentiometry. BSI, London, 1984, BS
6337.
9. Taywood Engineering. Duracrete: Modelling of Degradation.
Taywood Engineering, Middlesex, December 1998, Report Discussion contributions on this paper should reach the editor by
BE95-1347/R4-5 for Brite EuRam III. 1 November 2008

270 Magazine of Concrete Research, 2008, 60, No. 4

Delivered by ICEVirtualLibrary.com to:

IP: 137.195.49.43
Mon, 15 Feb 2010 19:57:23
View publication stats