Construction and Building Materials 25 (2011) 3345–3355

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Durability of concrete pipes subjected to combined steam and carbonation curing

Vahid Rostami, Yixin Shao ⇑, Andrew J. Boyd
Department of Civil Engineering and Applied Mechanics, McGill University, 817 Sherbrooke Street West, Montreal, Quebec, Canada H3A 2K6

a r t i c l e i n f o a b s t r a c t

Article history: A beneficial use of carbonation as an auxiliary curing regime for concrete pipes was studied in an attempt
Received 22 October 2010 to reduce steam curing time, improve durability performance and explore the possibility of using con-
Received in revised form 10 February 2011 crete pipe to sequester carbon dioxide. Durability performance of the carbonated concretes was charac-
Accepted 1 March 2011
terized by carbon uptake, strength gain, pH, calcium hydroxide content, permeability, sorptivity and
Available online 27 March 2011
sulfate and acid resistance. It was found that initial curing using steam is necessary to facilitate carbon-
ation. Although the contribution of early carbonation to strength gain is not noticeable after initial steam
Keywords:
curing, the process is unique in promoting enhanced durability performance of concrete. The early car-
Carbonation
Steam curing
bonation leads to a reduction in calcium hydroxide near the surface while maintaining a pH above the
Concrete pipe corrosion threshold value at the core. Carbonated concretes also exhibit improved resistance to sulfate
Durability attack, water absorption, and chloride ion penetration. A carbon uptake of 9% by cement mass makes con-
sulfate resistance crete pipe an ideal candidate for carbon dioxide capture and storage.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction proper compaction, also leads to improved transport properties,

Concrete pipes have been widely utilized for managing storm
water, sewage and drainage for more than a century. Deterioration
of concrete pipes constitutes a serious problem to society. It has
been reported that annual maintenance and repair costs of deteri-
orated concrete pipes in Canada in the early 1990s was approxi-
mately $2.5 billion dollars [1].
Concrete pipes are susceptible to chemical and biogenic attacks,
as well as freeze–thaw, corrosion, and abrasion damage. The in-
gress of water-borne sulfate ions into concrete pipes from the sur-
rounding environment can result in chemical degradation of the
calcium hydroxide and alumina phases, which leads to deteriora-
tive formation of ettringite and gypsum [2]. The source of acid at-
tack in concrete pipes is usually biogenic sulfuric acid which can
form in sewers and attack hydration products [3].
The primary methods for increasing durability of concrete pipes
against such threats include: using low alumina cement, using only
sulfate-free aggregates and additives, incorporating pozzolanic
materials, decreasing water-to-cement (w/c) ratio and ensuring
proper mixing and workmanship [4]. Low alumina cement and sul-
fate-free aggregates are two approaches to limiting the presence of
reactants in the pore solution, thus enhancing sulfate resistance.
Adding pozzolanic material such as fly ash helps sulfate resistance
of concrete pipes by lowering the concentration of calcium hydrox-
ide, which is vulnerable to sulfate attack, and improving transport
properties. A lower w/c ratio in concrete mix design, coupled with
⇑ Corresponding author. Tel.: +1 514 398 6674; fax: +1 514 398 7361.
E-mail address: yixin.shao@mcgill.ca (Y. Shao).
thus enhancing resistance to sulfate attack and freeze–thaw
damage.
In addition, the durability performance of concrete pipes can be
improved through controlled carbonation curing. The CO2 captured
at industrial point sources can be used as a carbonating agent.
Since carbonation is basically a lime consuming process, carbon-
ated concrete pipes so produced are expected to have enhanced
durability due to an overall reduction in reactive lime.
Studies to date on the effect of carbonation on concrete can be
categorized into two groups: carbonation of mature concrete and
carbonation of fresh concrete. The former is a phenomenon in
which well-developed hydration products react with atmospheric
CO2 through prolonged exposure, while in the latter case carbon-
ation is intentionally applied to freshly cast concrete wherein
CO2 reacts with the anhydrous phases to form strength contribut-
ing hydrates and carbonates.
Carbonation of mature concrete, also referred to as weathering
carbonation, and primarily resulting from long term exposure to
atmospheric CO2, has been extensively studied. It is generally ac-
cepted that carbonation of hardened concrete is detrimental to
long term performance of concrete because of the associated
reduction in pH of the pore solution, which can lead to the initia-
tion of reinforcement corrosion, and paste shrinkage leading to
cracking. The corresponding carbonation reactions are believed to
occur according to Eqs. (1) and (2). Prolonged carbonation is con-
sidered to be responsible for neutralizing the pore solution and
reducing its pH below the threshold value of about 10.5 in the
vicinity of reinforcement, leading to the loss of the protective film
on reinforcing steel and accelerated steel corrosion [5]. Long term

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.03.025
3346 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355

carbonation of hydrated concrete is also found to cause shrinkage Table 1

[6] and reduce the Ca/Si ratio in calcium silicate hydrates, mani- Curing scheme of concrete samples.

festing in decalcification of the phase [7]. However, it has been ob- Batch Initial curing Subsequent No. of Test age
served that long term carbonation of hardened concrete at curing samples
atmospheric pressure, from 1 week to 1 year, leads to a gain in 1 4S 4 h steam Sealed air 9 4 h/
compressive strength [8–12]. Carbonation of mature concrete can curing 28d/90d
also result in improved performance by producing a more refined 2 4S 4 h steam Water 3 90d
curing
pore structure. The prolonged atmospheric carbonation of hy- 3 2S + 2C 2 h steam + 2 h Sealed air 9 4 h/
drated concrete has been observed to decrease sorptivity due to carbonation curing 28d/90d
a reduction in surface capillary porosity through deposition of car- 4 2S + 2C 2 h steam + 2 h Water 3 90d
bonates [13]. Water permeability of hydrated mortars subjected to carbonation curing
5 2S + 2A 2 h steam + 2 h — 3 4h
long term exposure to CO2 can also be decreased due to reduced
sealed air
porosity [14]. 6 4C 4 h carbonation — 3 4h
7 2S 2 h steam — 3 2h
CaðOHÞ2 þ CO2 ! CaCO3 þ H2 O ð1Þ

C—S—H þ 2CO2 ! SiO2 þ 2CaCO3 þ H2 O ð2Þ
Carbonation of fresh concrete has also been studied [15–17].
Accelerated setting of cement paste and rapid strength gain are
the primary advantages in employing carbonation curing at early
age. The main anhydrous cement phases which can be carbonated
are tricalcium silicate (C3S) and dicalcium silicate (C2S). These reac-
tions take place according to Eqs. (3) and (4), where calcium sili-
cate phases react with carbon dioxide in the presence of water,
and produce calcium silicate hydrates (CxSHy) and chemically sta-
ble calcium carbonate instead of calcium hydroxide [17]:
C3 S þ ð3  xÞCO2 þ yH2 O ! Cx SHy þ ð3  xÞCaCO3 ð3Þ
C2 S þ ð2  xÞCO2 þ yH2 O ! Cx SHy þ ð2  xÞCaCO3 ð4Þ
According to Reactions (3) and (4), the carbonation of calcium sili-
cate phases generates calcium silicate hydrates while producing no
calcium hydroxide.
From Eqs. (1)–(4) it can be seen that carbonation of concrete at
any age is a CO2 uptake process. Carbon dioxide sequestration and
material performance enhancement have been recently studied in
an attempt to integrate an early carbonation process into precast
concrete production to gain environmental, technical and eco-
nomic benefits [18].
This paper describes an investigation into the durability of con-
crete pipes subject to combined steam and carbonation curing. The
proposed carbonation treatment occurs within 24 h after the con-
crete is cast and is thus considered to be early age carbonation. In
this treatment, steam is utilized to accelerate the initial curing and
carbonation to improve the durability of the concrete. It was re-
ported that a 4-h steam curing period produced a hydration degree
of C3S of only 30% [19]. Therefore, the phases available for carbon-
ation, after the initial steam curing, can be both anhydrous calcium
silicates and hydration products. Since the treatment is applied at
early age, the subsequent hydration still plays a critical role in the
overall performance of the concrete. Intentional carbonation of
hardened concrete has also been investigated on concrete blocks
after steam curing to improve the block resistance to atmospheric
carbonation shrinkage [20]. This research examines the effect of
early carbonation on long term performance of concrete through
calcium hydroxide formation, pH value, sorptivity, permeability,
sulfate resistance and acid resistance.
2. Experimental procedure
2.1. Casting and curing
In the laboratory investigation, concrete pipes were simulated by slab samples
with dimensions of 40 mm thick, 76 mm height and 127 mm wide. The mixtures
were proportioned using ordinary Portland cement (CSA Type GU) at 286 kg/m3,
coarse aggregates 730 kg/m3, fine aggregates 1050 kg/m3 and water 100 kg/m3.
Both coarse and fine aggregates were crushed granite with a coarse to fine ratio
of 0.7. The coarse aggregates had a maximum aggregate size of 12 mm and the fine
aggregates had a fineness modulus of 4.1. The water to cement ratio was 0.35. No
admixtures were used. The concretes were dry mix concrete with zero slump. Slab
samples were formed by vibration casting following the procedure for commercial
production of 40-mm thick concrete pipes.
Table 1 shows the curing schemes that were applied to the concrete after vibra-
tion compact forming and upon removal of forms. Batches 1 and 2 were subject to
the same initial 4-h steam curing, which was followed by either air or water subse-
quent curing. They acted as reference specimens for Batches 3 and 4, in which 2-h
steam curing was followed by 2-h carbonation. Batch 5 was again a reference for
Batches 3 and 4 to compare the 2-h carbonation period with 2-h air curing. 4-h car-
bonation without steam was also performed immediately after demoulding (Batch
6). This was to examine whether steam could be totally replaced by carbonation at
early age. Batch 7 underwent 2-h steam curing representing the starting material
for the carbonation process. In steam curing, the temperature was raised at a rate
of 40 °C/h up to 63 °C. After steam at designated time, the temperature was dropped
by slow cooling in steam chamber. For combined steam and carbonation (Batches 3
and 4), the 2-h steamed samples were transferred to the carbonation chamber
immediately after steam for 2-h carbonation curing. The carbonation setup is
shown in Fig. 1, wherein CO2 of 99.5% purity, simulating the recovered CO2 from
industrial point sources, was injected into a sealed chamber and the interior pres-
sure was regulated at a constant of 1.5 bar so that the consumed CO2 could be
replenished. Using a digital balance, the mass gain due to CO2 uptake was recorded
throughout the carbonation process. Subsequent curing took place either in a sealed
bag or in water until 28 days and 90 days to investigate the effect of early carbon-
ation on later strength.
2.2. CO2 uptake
The degree of CO2 uptake was measured using two different methods, the mass
curve method and the mass gain method, and reported as the ratio of the bound CO2
to cement content. In the mass curve method, the mass the of entire chamber con-
taining the samples was recorded as a function of time and was corrected by a cal-
ibrated CO2 mass at 1.5 bar with nonreactive dummy samples of the same volume.
The generated mass curve represents reaction kinetics. The mass gain method cal-
culates CO2 uptake using Eq. (5) by measuring the mass of the samples before and
after carbonation. Evaporated water from the exothermic carbonation reaction was
collected by absorbent paper and added to the final mass. These two methods are
totally independent and can be used to verify consistency.
ðMassaftercarbonation þ Water lost Þ  Massbeforecarbonation
Mass gainð%Þ ¼ ð5Þ
Masscement
2.3. Compressive strength testing
To investigate the contribution of the carbonation treatment on strength
gain, compression tests were conducted on samples with a cross-section of
127 mm  40 mm and compressive strengths were calculated at 4-h, 28-day and
90-day ages.
2.4. pH measurement
In the carbonated samples, pH of the pore solution was measured in successive
layers 5 mm thick to examine the effect of carbonation treatment on pH value dis-
tribution with respect to depth from surface. Since calcium hydroxide is the major
hydroxyl ion supplier in the pore solution, a pH drop in the carbonated zone of the
concrete is likely to occur. Meanwhile, the pH measurement could make it possible
to correlate between the carbonation induced consumption of the CH and the pH of
pore solution in different ages and at various depths. The pH measurements were
carried out on 5 mm thick slices which were also used in XRD phase analysis.
 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355
Fig. 1. Carbonation process set up.
A simple and convenient method was employed for measuring the pH of pore
solution of samples [21]. An absorptive paper of 10  10 mm size was placed on
the concrete surface and 100 ll of distilled water was added to it. After a 15 min
extraction, the hydroxyl ions came to equilibrium and an Extech PH110 pH meter,
which had a flat sensor at its tip, was used to measure the pH of the cut layers of the
samples. To ensure consistent results, the pH readings of three random spots on the
cut surface were averaged and reported.
2.5. Phase analysis
Following the pH test, the 5 mm thick slice was crushed, ground and passed
through a 38l sieve. The cement powder was analyzed with X-ray Diffraction
(XRD) to identify and quantify calcium hydroxide at different ages and depths of
the samples subjected to carbonation treatment. A Phillips PW1710 Power Diffrac-
tometer (Cu Ka radiation, scan interval 15–55° 2h, 0.02° and 0.5 s per step) was em-
ployed to perform the XRD analysis. Adding 20% of rutile (TiO2) by mass to the
powder sample, as a flushing agent, made it possible to quantify the CH phase using
Matrix Flushing Theory [22].
2.6. Chloride penetration
The effect of the early carbonation process on resistance to chloride ion penetra-
tion was investigated using the Rapid Chloride Permeability Test (RCPT). Disk sam-
ples of 100 mm in diameter and 50 mm thick were tested in accordance with ASTM
C 1202.
2.7. Sorptivity
To investigate the effect of early carbonation on the capillary suction behavior
of the concretes, sorptivity tests were conducted following ASTM C 1585 with sim-
ilar samples as those used in the chloride penetration testing.
2.8. Sulfate Resistance
Six mortar bars of size 25 mm  25 mm  285 mm with embedded pins at their
ends were cast in accordance with ASTM C 1012. The cement to sand ratio and
water to cement ratio were 1:2.75 and 0.35 respectively. Half of the bars were cured
for 4-h steam, to serve as a reference, while the other half experienced 2-h steam
followed by 2-h carbonation. After the initial 4-h curing, all of the bars were placed
in sealed bags until 24 h, when sulfate exposure was initiated. The samples were
submerged in a 5% sodium sulfate solution and the expansion of the bars was mea-
sured on a weekly basis.
2.9. Sulfuric acid resistance
Samples cured under both the steam and steam-carbonation procedures were
tested for resistance to sulfuric acid attack following general guidelines provided
by ASTM C 267. The specimens were submerged in sulfuric acid solution with a
pH value between 1 and 2, with their corresponding weight losses monitored
weekly.
3. Results and discussion
3.1. Carbon uptake
Fig. 2 presents a typical mass curve for a concrete during 2-h
carbonation after 2-h steam curing. The mass curve suggests that
3347
the reaction is rapid in the first 10 min, during which more than
40% of the total carbon uptake over 2 h takes place. The reaction
continued at a reduced rate and carbon uptake accumulated to
about 9% after a period of 2 h, relative to the dry cement mass used
in the concrete. Mass curve results were verified by the mass gain
method in Table 2, in which data from nine samples of Batch 3 are
presented. It is interesting to notice that, while the sample masses
stayed almost constant before and after steam curing, they were
quite different after 2-h carbonation. These discrepancies together
with the lost water collected, give an estimate of carbon dioxide
uptake. They are in a range of 8–9%, which is in the same order
of magnitude as the mass curve results.
Carbonation immediately after steam curing might be facili-
tated by the fact that the high temperature in steamed samples
vaporizes the free water, mostly from the surface, and creates void
spaces for carbon dioxide to penetrate. This phenomenon is illus-
trated in Fig. 3 in which the mass loss of a steamed sample exposed
to ambient condition is displayed. Mass loss of the steamed sample
is defined as the ratio of vaporized water from steamed concrete to
mixing water used in that sample. The curve was obtained at a
temperature of 24 °C and relative humidity of 30%. At the end of
the 2-h exposure, nearly 20% of the mixing water was evaporated,
making room for carbonation products to precipitate.
The achieved carbon uptake represents a promising result
regarding the capacity of concrete pipes to store CO2 in a beneficial
manner, since concrete pipe products have a large surface to vol-
ume ratio that gives them this particular carbon sequestration
potential.
3.2. Compressive Strength
Accelerated curing generated compressive strengths of con-
cretes are shown in Fig. 4. The 4-h steam curing batch served as
the reference. The 2-h steam plus 2-h carbonation procedure pro-
duced a strength which was 90% of the reference and very close to
that by 2-h steam plus 2-h air curing specimens. It is suggestive
that 2-h carbonation after 2-h steam curing did not generate high-
er strength than 2-h air curing. This is evident by the comparison
with the 2-h steamed concrete (2S). Subsequently carbonated
(2S + 2C) and air cured (2S + 2A) concretes gained about 10% and
8% additional strength, respectively, over 2-h steam (2S) curing
alone. The results from 4-h carbonation without steam are also
presented. The strength gain in this case was only about 3.6 MPa
with a carbon uptake of 2.1%. The low strength and low reactivity
of freshly-made concrete can be attributed to the saturated
surface due to high water to cement ratio and vibration compact
forming. It is clear that initial steam curing is beneficial to subse-
quent carbonation, though strength gain from carbonation is not
apparent.
3348 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355

Fig. 2. Typical mass curve of 2-h carbonation after 2-h steam curing.

Table 2 The 28-day and 90-day results in Fig. 5 indicate that subsequent
CO2 uptake of concrete due to 2 h carbonation after 2 h steam (batch 3). hydration in the sealed condition can take place in the carbonated
Mass Mass Mass (gr) Mass (gr) Water Uptake Uptake
samples. Although a slightly lower strength of the carbonated sam-
(gr) (gr) before after loss (%) (%) from ples is observed in comparison to the reference in air curing at
before after carbonation carbonation (%) from mass 28 days and 90 days, the carbonated concrete did reached the same
steam steam Eq. (5) curve strength as the reference after 90 days cured in water. This sug-
858.9 858.1 854.0 857.4 12.3 8.7 8.9 gests that the early age water loss of up to 12% due to the exother-
837.0 836.2 833.1 837.1 mic carbonation reaction slows down the subsequent hydration,
864.1 863.5 860.0 864.7
produces fewer hydration products and leads to a lower strength.
890.8 890.0 887.6 892.1 11.3 8.9 9.1
793.1 792.2 789.4 794.6 If sufficient moisture could be supplied during subsequent hydra-
851.5 850.8 848.0 852.4 tion, strength gain should not be hindered. Therefore, further
810.9 810.1 806.4 810.1 11.2 7.9 8.8 improvements in the subsequent hydration of concrete pipes, in
888.3 887.3 882.6 885.9 the form of either adding a wetting process after the carbonation
837.7 836.9 832.7 836.8
or utilizing internal curing is necessary.

Fig. 3. Mass loss of a steamed concrete due to evaporation.

 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355 3349

Fig. 4. Early strength of carbonated and steam hydrated concretes.

Fig. 5. Late strength of carbonated and steam hydrated concretes after subsequent curing.

3.3. Phase analysis by XRD
To understand the phase changes during the carbonation pro-
cess and avoid any influence by aggregates, ordinary Portland ce-
ment paste samples were first prepared by 2-h steam curing
followed by 2-h carbonation. Both steamed and steamed/carbon-
ated samples were analyzed using XRD. The results are presented
in Fig. 6. Comparing the XRD patterns before and after carbonation
indicates that the calcium hydroxide (CH) phase is significantly re-
duced while the calcium carbonate phase considerably increased
due to carbonation. Fig. 6 also shows that the calcium carbonate
produced is not solely due to the consumption of CH, since the
amount of non-hydrated phases (such as C2S and C3S) also exhibits
significant decreases.
Figs. 7 and 8 show the XRD patterns obtained from 5-mm layers
of the carbonated (A–D) and reference (E) concretes at ages of 4 h
and 28 days. It is obvious that the CH peaks are weak at the surface,
and gradually gain intensity toward the core. At the same time, the
calcium carbonate peaks are strong near the surface and slowly
disappear at the core. The phase analysis supports the observation
that carbonation occurred primarily near the surface of the sam-
ples, within a thickness of about 10 mm. It is worth noting that
the observed phase changes follow a similar trend at early (4-h)
and late (28 days) ages.
The calcium hydroxide distributions through the thickness of
the carbonated and hydrated samples at different ages are pre-
sented in Fig. 9. The effect of subsequent hydration after the car-
bonation process up to 28 days in sealed conditions is presented
3350 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355

Fig. 6. Phase analysis of cement paste before and after early carbonation.

Fig. 7. XRD patterns of concrete at 4 h: (A) 2S + 2C, 0–5 mm, (B) 2S + 2C, 5–10 mm, (C) 2S + 2C, 10–15 mm, (D) 2S + 2C, 15–20 mm, (E) 4S (Reference).

at different depths. The subsequent hydration up to 28 days didn’t
increase CH content at the surface. The CH remained constant,
suggesting there is no significant hydration reaction in the highly
carbonated zone. The lack of hydration reaction on the surface could
be partly due to the carbonation induced reduction of non-hydrated
phases. The reduced peaks of calcium silicate phases as presented in
Fig. 6 support this idea. Another cause of the termination of
hydration near the surface may be the coating of non-hydrated
cement particles with carbonation products in the heavily
carbonated zone.
Fig. 9 also reveals that the subsequent hydration reactions occur
to a greater extent at deeper layers, where less carbonation is
achieved. These reactions make significant contribution to the later
strength gain after the early carbonation.
3.4. pH value of the concrete
The pH values of concrete as a function of location through the
thickness at early and late ages are presented in Fig. 10. The 2-h
carbonation process reduced the pH mostly in the outer 10-mm
layer due to the conversion of CH to calcium carbonate; the pH
of concrete at a distance more than 10 mm from the surface was
almost equal to that of the reference samples; their pH values were
over 12.0. This result suggests that early carbonation treatment is
not likely detrimental with respect to carbonation induced corro-
sion in concrete pipes. The subsequent curing contributes to later
strength gain. It plays a critical role in this early carbonation curing
scheme. In the other words, early carbonation should always be
accompanied by proper subsequent hydration.
 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355 3351

Fig. 8. XRD patterns of concrete at 28 days: (A) 2S + 2C, 0–5 mm, (B) 2S + 2C, 5–10 mm, (C) 2S + 2C, 10–15 mm, (D) 2S + 2C, 15–20 mm, (E) 4S (Reference).

Fig. 9. Calcium hydroxide distribution in carbonated concretes at 4 h and 28 days.

Despite the significant decrease in CH content near the surface,
as observed from the phase analysis, the pH reduction at the car-
bonated surface was about one unit. It is well known that CH is
not the only contributor to the pH of pore solution; alkali ions play
equally important rule. The alkalis present in the concrete, namely
NaOH and KOH, can potentially be carbonated too. The signifi-
cantly reduced CH near the carbonated surface, and its relatively
high pH (over 11.0), may suggest that alkalis are not yet carbon-
ated and they are still the contributing phase to pH in carbonated
concrete.
The pH of the surface layer of the carbonated samples subjected
to subsequent hydration in sealed condition up to 28 days re-
mained same value compared to the pH immediately after carbon-
ation. It is in agreement with the phase analysis results which
verify that there is no hydration reaction in the heavily carbonated
surface layer after the early carbonation.
3.5. Chloride penetration and sorptivity
Fig. 11 compares the coulomb passing of carbonated and hy-
drated concrete at 2 days and 28 days. Two days were needed to
saturate both 4-h cured samples (4S and 2S + 2C) for RCP tests.
After 6 h of applying a 60 V electrical potential between the NaOH
and NaCl cells, the total coulomb passing in carbonated concrete is
only half of that in steam cured concrete. The total electrical charge
passed exhibits a linear increase with time, suggesting that the ini-
tial conductivity remained the same during the test and up to the
end of the electrical induction. Therefore, it could be concluded
3352 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355

Fig. 10. pH value of carbonated and hydrated concretes at 4 h and 28-day.

Fig. 11. Comparison of coulomb passing in carbonated and hydrated concretes at 2-day and 28-day.

that the change in temperature and ion migration during the test The effect of carbonation on the sorptivity of concrete is pre-
did not alter the conductivity of the samples. sented in Fig. 12. The initial and secondary sorptivity coefficients
The 28-day results of RCPT on steam cured reference and steam/ (S) were calculated using the following formula according to ASTM
carbonation cured samples are also presented in Fig. 11. Similar to C1585:
the RCPT results of 2-day tests, carbonated concrete still showed mt pffiffi
lower coulomb passing compared to steam cured specimens. It is I¼ ¼S t ð6Þ
a=d
also evident that both carbonated and hydrated concretes at
28 days are more resistant to chloride penetration than the early where I = the absorption (mm), mt = the change in mass of the sam-
age samples, suggesting the subsequent hydration after the early ple (gr), a = the exposed area of the sample (mm2), d = the density
pffiffi of
age curing could reduce ion penetration for both carbonated and the water (gr/mm3) and t = time (s). By plotting a=dmt
against t for
steamed concretes. steam reference and steam-carbonated concretes, the initial and
 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355 3353

Fig. 12. Absorption of carbonated and hydrated concretes at 28 days in sealed bag.

Table 3
suction in carbonated concrete could be evidence of shrinking cap-
Initial and secondary coefficients of sorptivity at 28 days. illary tubes and loss of connectivity in the pore space. This improve-
ment can possibly be attributed to increased volume of solid phases
Curing regime Initial sorptivity Secondary sorptivity
coefficient (mm/t1/2) coefficient (mm/t1/2)
within the pore network, mostly in the carbonated surface zone.
The reduced capillary suction after 24 h in both reference and
4 h steam + 28d sealed air 0.0126 0.0005
carbonated samples might suggest similar capillary porosity in
2 h steam + 2 h 0.0081 0.0006
carbonation + 28d the core of both samples. According to Fig. 12, the total water
sealed air absorption of steam reference samples on day 9 was higher than
steam/carbonated samples. This result explains reduced capillary
porosity due to the early carbonation.
secondary sorptivity coefficients were obtained. These results are
summarized in Table 3. The initial capillary suction of the carbon- 3.6. Sulfate and acid resistance
ated sample, up to 6 h, was noticeably reduced compared to the ref-
erence samples subjected to steam curing. However, the secondary Fig. 13 presents the expansion of the carbonated and hydrated
sorptivity coefficients for both reference and carbonated samples, reference mortar bars submerged in sodium sulfate solution con-
which represents capillary suction from 24 h to 9 days of testing ducted according to ASTM C 1012. Carbonated bars experienced
time, were of the same magnitude. The reduced initial capillary much less expansion compared to the reference specimens. The

Fig. 13. Expansion of mortar bars in 5% sodium sulfate solution.

3354 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355
Fig. 14. Weight loss of concretes in sulfuric acid solution (1 < pH < 2).
accelerated sulfate attack performed on the samples simulated a
sodium sulfate attack on concrete pipes buried in a sulfate rich
environment with a sufficient supply of water. It is agreed that
the sulfate initially reacts with calcium hydroxide to form gypsum.
The gypsum can subsequently react with C3A, usually via forma-
tion of mono-sulphoaluminates, leading to a disruptive formation
of ettringite [2]. Therefore, the reduced calcium hydroxide content
in the concrete samples due to carbonation makes contribution to
lower expansion and better sulfate resistance.
The acid resistance of the concretes cured under steam and
steam-carbonation regimes is compared in Fig. 14. Despite the
slightly higher weight gain of the carbonated samples in first week,
both the carbonated and the steamed concretes showed similar
weight loss trends. This result suggests that carbonated concretes
are not more susceptible to sulfuric acid attack than hydration
products.
4. Conclusions
The use of carbonation as an auxiliary curing regime for con-
crete pipes was studied in an attempt to reduce steam curing time,
offer improved durability performance and explore the possibility
of using concrete pipes to sequester carbon dioxide emitted from
cement production. It was found that early carbonation introduced
after an initial 2-h steam curing enabled a 9% carbon dioxide up-
take (by cement mass) and reduced calcium hydroxide content at
the concrete surface. This chemical modification could significantly
improve sulfate resistance, while maintaining the pH of the pore
solution at the core above the corrosion threshold value. Carbon-
ated concretes are not more susceptible to strong acid attack.
The early carbonation treatment of concrete resulted in de-
creased chloride ion migration, possibly as a result of the elimina-
tion of hydroxyl ions and precipitation of calcium carbonate within
the surface layer. Thus, early carbonation could lower the conduc-
tivity of the pore solution. Although this result was obtained
through RCPT, which is an electrically induced chloride migration
test, the reduced ion migration through the carbonated concrete
might be indicative of improved resistance to ion diffusivity in
the carbonated surface zone, serving as a barrier to ion movement.
The resistance to initial sorptivity of carbonated concrete could
be noticeably enhanced. This improvement suggests a microstruc-
tural modification in the carbonated surface zone in terms of a
reduction in capillary water suction through the unsaturated con-
crete. This decrease in capillary suction and total capillary porosity
of the carbonated concrete could be considered as another
enhancement in the durability performance due to the carbonation
treatment.
An early carbonation process applied after steam curing did not
make noticeable contribution to the rapid strength gain, nor did
immediate carbonation after compact forming without steam
curing. Steam seems to be necessary to facilitate carbonation. Sub-
sequent hydration with additional moisture is needed for carbon-
ated concrete to be comparable to hydration reference.
The proposed carbonation curing procedure for precast concrete
pipes has an added-value. It offers an economically feasible carbon
sequestration technology. Carbonation of concrete is a CO2 uptake
process. Precast concrete pipes can thus be used for carbon storage.
In the United States of America, annual production of concrete
pipes is approximately 22 million tonnes [23]. If the cement con-
tent in pipe is assumed to be 12.5% by total mass and carbon up-
take is 10% by cement mass, the US concrete pipe industry could
consume 275,000 tonnes of CO2 per year. If the carbon uptake
can be increased to 20%, the carbon storage capacity by concrete
pipes would be doubled. Early carbonation curing technology has
shown environmental, economical and technical benefits.
References
[1] NRC. Construction innovation. Urban infrastructure; 1996. <http://www.nrc-
cnrc.gc.ca/eng/ibp/irc/ci/volume-2-n1-25.html>.
[2] Neville A. The confused world of sulfate attack on concrete. Cem Concr Res
2004;34:1275–96.
[3] Attal A. Biological mechanisms of H2S formation in sewer pipes. Water Sci
Technol 1992;26:907.
[4] Skalny J, Marchand J, Odler I. Sulphate attack on concrete. 1st ed. London: Spon
Press; 2002.
[5] Mehta PK, Monteiro JP. Concrete microstructure, properties and material. 3rd
ed. New York: McGraw Hill; 2006.
[6] Powers TC. A hypothesis on carbonation shrinkage. J PCA Res Develop Lab
1962;4:40–50.
[7] Castellote M et al. Chemical changes and phase analysis of OPC pastes
carbonated at different CO2 concentrations. Mater Struct 2009;42:515–25.
[8] Chang CF, Chen JW. Strength and elastic modulus of carbonated concrete. ACI
Mater J 2005;102:315–21.
[9] El-Turki A, Ball R, Allen GC. The effect of carbonation on the mechanical and
microstructural properties of lime and cement. In: Proceedings of the 11th
international conference on non-conventional materials and technologies
(NOCMAT) 2009. Bath, UK.
[10] Chang JJ et al. Mechanical properties of carbonated concrete. J Chin Inst Eng
2003;26:513–22.
 V. Rostami et al. / Construction and Building Materials 25 (2011) 3345–3355
[11] Jerga J. Physico-mechanical properties of carbonated concrete. Constr Build
Mater 2004;18:645–52.
[12] Matsusato H et al. Studies on the carbonation of hydrated cement and its effect
on microstructure and strength. In: Ninth international congress on the
chemistry of cement; 1992. India.
[13] Dias WPS. Reduction of concrete sorptivity with age through carbonation. Cem
Concr Res 2000;30:1255–61.
[14] Song HW, Kwon SJ. Permeability characteristics of carbonated concrete
considering capillary pore structure. Cem Concr Res 2007;37:909–15.
[15] Goodbrake CJ, Young JF, Berger RL. Reaction of beta-dicalcium silicate and
tricalcium silicate with carbon dioxide and water vapor. J Am Ceram Soc
1979;62:168–71.
[16] Goodbrake CJ, Young JF, Berger RL. Reaction of hydraulic calcium silicates with
carbon dioxide and water. J Am Ceram Soc 1979;62:488–91.
[17] Young JF, Berger RL, Breese J. Accelerated curing of compacted calcium silicate
mortars on exposure to CO2. J Am Ceram Soc 1974;57:394–7.
3355
[18] Shao Y, Mirza MS. CO2 sequestration using calcium-silicate concrete. Can J Civ
Eng 2006;33:776–84.
[19] Alunno-Rossetti V, Chiocchio G, Collepardi M. Low pressure steam hydration of
tricalcium silicates. Cem Concr Res 1974;4:279–88.
[20] Shideler JJ. Carbonation shrinkage of concrete masonry units. J PCA Res
Develop Lab 1963;5:36–51.
[21] Heng M, Murata K. Aging of concrete buildings and determining the ph value
on the surface of concrete by using a handy semi-conductive pH-meter. Anal
Sci 2004;20:1087–90.
[22] Chung FH. Quantitative interpretation of X-ray diffraction patterns of
mixtures. I. Matrix-flushing method for quantitative multi-component
analysis. J Appl Crystallogr 1974;7:519–25.
[23] Rinker, <http://www.rinker.com/downloads/TargetStatement.pdf>.